ARABIDAD This is the concluding part of the monographic study Ground Beetles (Carabidae) of Fennoscandia. The book begins with some basic remarks on the prerequisites (material and methods) of zoogeographical research. The author has stated his outlook on these questions more precisely. The somewhat strongly subjective interpretation, in the words of the author, “has the purpose of provoking opposition and open discussion.” The three main sections of the analytic part contain a detailed treat- ment of three special cases, each of which describes one of the three decisive area-limiting factors: the existence requirements of the animal; the dynamic characteristics of the animal; and the influence of time. The comments in the synthetic part are to a large extent based on the results obtained in the analytic part. By a combination of conclusions drawn from concrete material of fossil records and somewhat abstract deduc- tions from the present-day distribu- tion pattern of the species the author has attempted to answer the question of the biological significance of quaternary glaciations. Lindroth’s original English summary is reproduced at the end. BI 5 Hakan A i \ CE Raute KIN MAAN EI RENHINN RO IN ER ta aha CRSA Oe Nn mye as f ar DR Hey I. IN i) JS. ea ca ny DENN Rn) n N AN yon | He di we oy REN. ne hu Ground Beetles (Carabidae) of Fennoscandia Smithsonian Institution Libraries WASHINGTON, D.C. 20560 202-357-2240 The Smithsonian Institution Libraries has selected your library to receive gratis copies of volumes published in our Translations Publishing Program. Under this program, signi- ficant non-English-language works that related to a field of Smithsonian research are translated and published with Special Foreign Currency funds according to the provisions of Public Law 480. We hope that you will find this publication a useful addition to your collection. We would appreciate receiving a copy of all reviews of this work. Additional copies of this work may be purchased from the National Technical Information Service (N.T.LS.), Springfield, Virginia, 22161. Please cite the TT (Technical Translation) number printed on the Copyright page when ordering from N.T.I.S. Let us know if your copy of this book is incomplete or damaged. Please direct all inquiries and address corrections to: Translations Publishing Program Smithsonian Institution Libraries Natural History Building 22 Washington, D.C. 20560 202/357-1850 August 1992 Ground Beetles (Carabidae) of Fennoscandia A Zoogeographic Study PART II General Analysis With a Discussion on Biogeographic Principles CARL H. LINDROTH Scientific Editor JOACHIM ADIS Scientific Coordinator TERRY L. ERWIN Smithsonian Institution Libraries and The National Science Foundation Washington, D.C. 1992 TT 79-52039/03 Die Fennoskandischen Carabidae: Eine Tiergeographische Studie III. Allgemeiner Teil. Zugleich eine biogeographische Prinzipdiskussion Bröderna Lagerström Boktryckare Stockholm, 1949 © 1992 Amerind Publishing Co. Pvt. Ltd., New Delhi Translated from the German Translator: Dr. J.S. Bhatti General Editor: Dr. V.S. Kothekar Library of Congress Cataloging-in-Publication Data Lindroth, Carl Hildebrand, 1905- Ground beetles (Carabidae) of Fennoscandia. Translation of: Die fennoskandischen Carabidae. Bibliography: p. Contents: pt. 1. Specific knowledge regarding the species—pt. 2. maps— pt. 3. General analysis, with a discussion on biogeographic principles. Supt. of Docs. no.: SI 1.2.:B39 1. Carabidae—Geographical distribution—Collected works. 2. Insects—Scandinavia—Geographical distribution—Collected works. 3. Insects—Finland—Geographical distribution—Collected works. I. Adis, Joachim, 1950— . II. Erwin, Terry L., 1940- . III. Title. QL596.C2L5313 1989 595.762 87-600154 Translated and published for the Smithsonian Institution Libraries, pursuant to an agreement with the National Science Foundation, Washington, D.C., by Amerind Publishing Co. Pvt. Ltd., 66 Janpath, New Delhi 110 001 Printed at Pauls Press, New. Delhi, India Dedicated to Rolf Krogerus Helsinki Published with the support of Längmanska Kulturfonden To the Reader We recommend to follow the instructions given in the “Foreword to this Translation” in Part I of the book. Manaus/Plon, Joachim Adis July 1988 Scientific Editor CONTENTS PREFACE xi INTRODUCTION TO PART III 1 BASIC REMARKS ON MODERN INSECT TAXONOMY 5) ON THE CONSIDERATION OF LITERATURE 15 ON THE TASK OF THE MUSEUMS 17 ON SYNECOLOGY AND “SYNGEOGRAPHY” 22 ANALYTIC PART THE REALIZED EXPERIMENTS 32 THE “LIMESTONE SPECIES”: An Example of the 97 Influence of Existence Factors THE FAUNA OF THE ISLANDS: An Example of the Significance of the Dynamic Properties of Animals 1077, WING DIMORPHISM: An Example of the Zoogeographical Influence of Time 295 Vill SYNTHETIC PART THE DEFINITION OF AREA 368 THE RELIABILITY OF DISTRIBUTION MAPS 369 THE RELATIONSHIPS OF THE FENNOSCANDIAN FAUNA 379 SPECIES DISTRIBUTION AMONG DIFFERENT PLANT REGIONS 384 EXISTENCE FACTORS 397 Climate 397 Temperature 398 Precipitation and Humidity 429 Other Climatic Factors 437 On Lococlimate and Microclimate 439 Indirect Evidence of Climatic Factors 442 Soil 445 Size of Soil Particles 446 Chemical Properties of the Soil 456 Food and Feeding Habits 469 Competitors and Enemies 488 Stenotopy and Eurytopy 495 Types of Development 499 DYNAMIC FACTORS 503 Flight Capacity and Wind Dispersal 503 Water Dispersal 524 Transport by Animals 529 Transport by Man 529 Other Dynamic Factors 531 Barriers against Dispersal The Sea The Mountains Final Remarks on Area Limits FAUNAL HISTORY Faunal Changes in Recent Times Fossil Records Relicts THE POSTGLACIAL IMMIGRATION I. The Southern Immigrants II. The Eastern Immigrants II. The Problem of “Wurm Hibernation” Concluding Remarks on the History of the Fennoscandian Fauna ENGLISH SUMMARY SUPPLEMENT TO PARTS I AND II BIBLIOGRAPHY INDEX TO PART III BULL, u i's eg PREFACE 7 Biogeographical research can be pursued at three different places: in nature, in the laboratory and at the desk. This series indicates an order of precedence. One cannot escape the fact that the “natural method” not only provides the primary material, knowledge of the area (cf. p. 477; suppl. scient. edit.) of the species concerned, but also direct observations in the field give us first-hand information on the require- ments of life for the species and thus offer a clarification of the distributional pattern. One may go to the desk fresh from the field to tackle the problems, bypassing the laboratory. That has been the way with much biogeographical research, especially in the past. This is quite apart from the fact that in many cases the material was first collated at the desk. Once I too felt in this way until I gradually realized that the decision as to which factor determines the distribution of a particular species in a given case, cannot be arrived at with fair certainty without experiments. The belief that one may dispense with all experiments is actually a kind of arrogance. Apart from exceptional cases, anyone who can single out the decisive factor right there in nature is truly endowed with extraordinary powers. It is only by experimentation that hypotheses made in the field (or at the desk) can be confirmed or rejected. The strikingly low reputation that biogeographical research has generally 8 earned in the eyes of other biologists is undoubtedly due to just this regret- table lack of precise methodology and the consequent lack of unambiguous, unprejudiced results. But it is possible to pursue this science more objectively. This means the laboratory must no longer be bypassed. This concept, which I owe above all to the influence of my friend Rolf Krogerus, has brought me immense relief. The responsibility has been to a great extent lifted from the shoulders of the poor scientist and passed on to the insect, which “answers” the questions through its behavior in various apparatuses. On this is based the relative objectivity of the results achieved, but t (Pagination of the original German version to which page citations in the text of this trans- lation are referring to; suppl. scient. edit.) Xi not just because they can usually be expressed numerically. Modern biology generally shows altogether too strong an inclination to illustrate everything numerically, without always considering how these figures were obtained. The series of steps (first: study of nature; second: experiments; third: con- clusions) recommended here in the work of a zoogeographer suffers from the weakness of all democratically functioning systems due to opposition on the part of workers. Above all, it is cruel not to be permitted to think until one has all the necessary premises to hand. On the contrary, inclination, which is more intuitive than rational, always tempts me to form enthusiastic, bold conclusions as soon as the primary material becomes available, to look for answers when the questions are yet to be properly framed for lack of suffi- cient premises. Such attempts to take the desk into the field have often proved unsuccessful. It means bringing preconceived ideas to the solution of the prob- lem. How often the subsequent experiments which are “dutifully” executed to confirm the established hypothesis are unexpected and unwelcome! And what self-discipline it takes to junk the beautiful structure and find a new approach to the problem instead of fiddling the data to make them fit! The present book is thus not the result of a calm, purposeful work. It is on the contrary a heterogeneous mixture of scarcely interrelated questions which, uninvited, clamored for an answer. And the answer was often a compromise between the rival claims of a “heurella” (I’ve found it!) and a “nitschevo” (never mind!). However, it is no misfortune when traces of this erratic method of work peep out between the lines. Science is after all not an industry, it is the grand adventure. Djursholm, September 1948 Carl H. Lindroth 10t 11 INTRODUCTION TO PART III The first two parts of this work, published in 1945, present the material for the general zoogeographical aspects which are dealt with in this concluding part. This treatment is necessarily both lengthy and to a certain extent one- sided. If the Nordic carabid fauna, in all 362 species, would be considered and discussed under all those aspects of interest for zoogeography, then the present work would cross all bounds. “For the sake of completeness” some sections would have to be included that would be irrelevant to the main issue. But as already pointed out in the Preface to Part I, this is the postglacial—and possibly the glacial, history of the Fennoscandian fauna. As examples of such aspects ignored here or only touched on, however interesting and productive, we may mention: synecological and generally quan- titative research, qualitative distribution into ecological groups (p. 35); studies of races and analysis of variations in the material (p. 20); and division of the Fennoscandian territory into zoogeographical regions (p. 43). Even then the tasks undertaken involved a disquieting tendency to enlarge and refine during the course of the investigation, to the extent that I sometimes despaired on seeing only fragments which defied incorporation in the main body of the text. So I had to restrict the scope of this work. At the same time, degeneration into a general superficial study had to be avoided. I therefore selected definite problems, and hence definite species, so as to be able to carry out more detailed Studies, in the hope that from the results obtained and the methods developed a broader concept of the conditions for life and the history of the entire Fennoscandian carabid fauna could be developed. In the general arrangement of the text this finds expression in the division of the main part of the book into two, one dealing with analysis and the other with synthesis. The former contains more original research and more conclusions based on experiments. It turned out to be practical to begin it with the details of experiments. Of course the “synthetic” part also contains references to experiments, but the arrangement of the data at the beginning of the book seemed a good idea because it is more convenient to refer to them there than at the end amid t (Pagination of the original German version to which page citations in the text of this trans- lation are referring to; suppl. scient. edit.). 12 2% the supplement to the first two parts, list of references, index and English summary. Moreover, the “analytic” part is also an example of extensive research. It deals with too many species to delve thoroughly into the physiology of even one of them. But it was not the intention to strive for precise values concerning the response of one or other species of animals. The physiological results were only the secondary objective, or more correctly “preliminary objective.” They were designed to provide for a more viable comparison of the ecological requirements, of the dynamics, etc. of an entire series of species. I wanted to find out whether the behavior of these species in diverse ways in nature corresponds with various responses to definite and as far as possibile isolable factors. The book begins with “some basic remarks.” This is a series of small articles apparently bearing little relation to the main theme. Actually they involve the basic prerequisites (material and methods) for zoogeographical research. I wanted to state my outlook on these questions more precisely, and the somewhat strongly subjective interpretation has the purpose of provoking Opposition and open discussion. The three main sections of the analytic part contain a detailed treatment of three special cases, each of which describes one of three decisive area-limiting complexes of factors: the existence requirements of the animal (exemplified by the “limestone species”), the dynamic (“dispersal-ecological”) characteristics of the animal (exemplified by the fauna of the islands), and the influences of time (exemplified by species exhibiting wing dimorphism). The comments in the synthetic part are to a large extent based on the results obtained in the analytic part. The area-limiting factors are here re- viewed in the usual way, with the present-day conditions considered first. But it is clear that these are inexplicable without a knowledge or at least an es- timation of the prehistoric conditions. An attempt must therefore be made to acquaint ourselves with the Fennoscandian fauna of earlier eras. It is best to go backward and to begin with the study of ascertainable changes in the fauna of recent decades. By a combination of conclusions, drawn on the one hand from the concrete material of woefuily incomplete fossil records, and on the other hand by more or less abstract deductions from the present- day distribution pattern of the species, it is possible to enter cautiously into the earlier parts of the postglacial period and to approach the central ques- tion that this work sets out to answer: the biological significance of quaternary glaciations. The objective of this investigation may be considered achieved only to the extent that I have been able to answer this question convinc- ingly. teef. page 13 of this volume and page 15 of Part I; suppl. scient. edit.). 13 At the end there are four supplementary sections: 1. English summary. Here, only the most important results of Part III are briefly discussed. 2. Supplement to the first two parts. This includes any new records that contribute to the significant improvement of the earlier distribution maps, and confirmations of older or doubtful records. New information on ecology, “biology”, dynamics, and fossil records has also been incorporated, and in some cases taxonomic remarks (Amara communis, Badister bipustulatus). Only one species new to the area could be added: Lionychus quadrillum. 3. List of references. In addition to the literature used in Part III, such further references concerning the first two parts of the book are included as have appeared recently or were omitted earlier. 4. Index to Part III. This includes only the Fennoscandian Carabidae, including the place where each species was recorded. Finally there are two questions on terminology: 1. The word postglacial is used here (as by most biologists) in the widest sense, meaning the entire period since the maximum of the last glaciation. Geologists specializing in the quaternary period, on the other hand. divide this period into the late glacial and the postglacial (s. str.) periods. 2. Ekman (1922, p. 308) is credited with recognition of two well-defined groups of area-limiting factors affecting an animal or plant species. He calls these existence ecology (“Existenzokologie”) and dispersal ecology (“Aus- breitungsökologie”)!. The first of these includes the requirements of life for the animal (or the plant) in its environment and the second concerns its capacity to seek areas suitable for habitation. Corresponding with this we generally speak respectively of existence ecological (“absolute,” Reinig, 1938, p. 46) and dispersal ecological (“interim,” Reinig, l.c.) area limits. However, these terms are very unwieldy, and the second has the draw- back of not being internationally understood and being difficult to translate. The German translation of the good Swedish word “spridnings-ekologi,” which clearly expresses the dynamic character of the term, has often been inappropri- ately rendered “Verbreitungs-Okologie” (= distribution ecology) (for instance, Heinze, 1932-35), although “distribution” indicates a static condition! In Part I of this book, instead of dispersal ecology, I have throughout used the term dynamics (“Dynamik”). Here I will use the term dynamic limits (“dy- namische Grenzen,” lignes dynamiques) for dispersal ecological area limits. The existence ecological area limits will be called existence limits (“Existenz- grenzen,” lignes d’existence); they could also be called static limits (“statische Grenzen,” lignes statiques). It seems superfluous to drag on the stem “‘ecol- ogy” with these terms. tcf. pp- 43 and 203: suppl. scient. edit.). In the present work the term “ecology” is always understood to denote existence ecology, and the word is generally used in this restricted sense in the literature. 13 Basic Remarks On Modern Insect Taxonomy Contemporary insect taxonomy has three main features: 1. The splitting of genera. 2. The renaming of species. 3. The splitting of species into categories of lower rank (subspecies, “race,” variety, form, aberration, etc.). All three features are especially strong in the taxonomy of Lepidoptera. However, they are basically to be considered as modern characteristics related to the entire taxonomic-biological nomenclature, both zoological and botani- cal. Each of these problems will be considered separately. 1. The process of splitting the old, large genera, having simple names, easy to pronounce and remember, into numerous, less comprehensive genera, has been carried to extremes in the family Noctuidae. Momentarily even the experienced specialist in this field may not know the genus to which one or other species, otherwise familiar to him, belongs. The result may be heard in any conversation between lepidopterologists: they consistently use only the species name for their favorite insects. Any one wishing to describe a new species of Noctuidae looks for a name that is new not only in the genus but also in the family, and if possible in the entire order Lepidoptera. The splitting of genera has actually led to the end of the binomial nomenclature. Instances of such splitting of old genera, disastrous from the practical viewpoint, are also found in other insect orders. Clear evidence is to be found in the new, otherwise exemplary, revision of the Nordic species of Ci- cadina by Ossiannilsson (1946-47), and in the otherwise perfect monograph on Coleoptera, unfortunately not completed, concerning the Palaearctic Ceu- thorrhynchinae by Hans Wagner (1938, p. 172). But coleopterologists have at least generally been more prudent. In the systematics of Carabidae specialists like G. Muller and Schauberger have even merged some of the generic entities, earlier considered separate, into the genus Agonum and Harpalus. Jeannel (1941-42) has, on the contrary, taken the oppo- site course, and has split ail the larger carabid genera. For instance, Bembidion has been split into 17 genera! 15 16 If the procedure adopted by Jeannel were to serve as a model for carabidologists—which I certainly do not anticipate—it would signify a great loss, the loss of familiar generic names which we have used for many generations as tags for these animals. It is of course correct that the higher taxonomic categories (genus, family, etc.) should, wherever possible, reflect the true relationship (i.e. common origin) of the animals. But this is only one of its functions, and it is best not to insist on it too strongly. It seems it is no longer possible to take bold, categorical decisions as to what is “primitive” and what is “derived” in the organization of an animal. If such considerations alone were to serve as the guiding principles for monographers, then very soon every author would have his own “system.” The other aspect of taxonomy is purely practical. The special purpose of the generic name is to give the nonspecialist an indication as to where the species in question belongs. In a zoogeographical, ecological or physiological study, etc. it would be very undesirable if every time a species was introduced into the discussion it was necessary to give its systematic position just because its name was known only to taxonomic specialists. The primary task of the systematist is to give the animals easily understandable and, as far as possible, stable names that can be utilized by the main body of biologists the world over. The idea is not to pro- vide themselves with counters for playing the adventurous game of phylogeny. In terms of nomenclature this branch of biological research has been adequately endowed with the erection of subgenera or species-groups, whose names need not be used when the animal is mentioned in non-phylogenetic contexts. From the viewpoint of a practical nomenclature the genera are better too large than too small. It is no misfortune if the isolated species-groups which might merit the rank of a genus (among the Carabidae for instance, Europhilus or Ophonus) are maintained as subgenera (also compare Handlirsch, 1913a, p. 71). With regard to the generic names which were changed as a result of application of the rules of priority, see below. 2. The entire realm of zoology and botany of the last century is marked by an extensive change of species names, a trend that has increased rather than decreased during recent decades. Thus we find that in the Fennoscandian carabid fauna, since Grill (1896) 35 species (11%) have been given new names, and this of course excludes species which were simply misidentified or were later split. Going still farther back to Thomson (1885), the number increases to 77 species (28%). Even since the publication of the Nordic Catalogue (1939) six names have been changed. If all the “new” names used by Csiki (1927-33) and Jeannel (1941-42) were accepted the number of name changes would further increase. The same is substantially true of the generic names. The chronic instability is evidently undesirable. Is it necessary? Some of the name changes are unavoidable. Wrong identifications must be corrected (for example, Bembidion scandicum as against B. macropterum). Be- sides, more careful modern studies have shown that some of the earlier species 17 7 must be split (for example: Notiophilus “palustris” and Harpalus “luteicornis” of early authors). But this is not true of the majority of cases. On the other hand, there are cases where, probably by the study of a specimen named as the type by an earlier author, it becomes clear that the name he gave has subsequently been used for a different insect. Or, more often, rediscovered works reveal a name which is ingeniously ascribed to a species now known generally by a more recent name. Most of the mischief is due to the “law of priority” (Regles Internat., 1905; Handlirsch, 1913b). A clear instance in the systematics of carabids is the case of Harpalus (Ophonus) “brevicollis” and “rufibarbis,” which I have discussed earlier (1943, p- 25). Each of these names has been used at various times either for seladon Schaub. or for schaubergerianus Puel. Hence it seems inappropriate to retain any of them. Another kind of example is that of Nebria gyllenhali Schonherr (1806), whose name Jeannel (1937) sought to change to rufescens Strom (1768). Yet the former name has been in use almost as the only name for more than a century in Europe, despite the fact that rufescens was undoubtedly proposed as a modification of forma rufino, and was named after it (see Lindroth, 1939b, p- 59). Why does Jeannel wish to follow the law of priority most rigidly in this case, whereas in the case of Abax ater Villers (1789) he (1941-42, p. 776) refuses to replace it by parallelepipedus Piller & Mitterpacher (1783), among other reasons, because “l’espece est bien connue sous son nom d’ater Vill.”? No wonder the consequences of the rigid application of the law of pri- ority have aroused serious misgivings and clear opposition from the most level-headed entomologists. Heikertinger, particularly, champions the oppo- site view. In a long series of articles* (see list of references at the end of this book), he adduces a wealth of evidence in favor of the principle of continuity as against the principle of priority. The principle of continuity is emphasized in the following declaration (and also Heikertinger, 1935, pp. 147-148; 1939a, p. 221; 1942a, p. 26; 1942b, p. 1): “The valid name of a genus or species is that found in use in the scientific literature, whether or not it was given first. If there are two names in use for a genus or species, then the revisor should choose the name whose retention will produce the fewest nomenclatural changes in the existing literature.” These views represent a true anarchy against the principle of priority. Must they prevail in order to prevent the continuation or even accentuation of the present intolerable situation? Or are more moderate solutions possible, which may satisfy the “priority people,” at the same time ensuring nomenclatural stability? Heikertinger (i.e. 1939b, p. 561) suggested such a compromise. *Heikertinger’s article of 1942b includes a complete list of his papers on nomenclatural problems. 18 One idea is the “limitation proposition,” i.e. species and genus names that have been the only ones in use for a definite number of years (Horn, 1938, p. 2, suggests 20 years) should not be changed. A second suggestion is “nomina conservanda,” i.e. a list of names of animals should be prepared that must not be changed in future under any circumstances. In this way, one has attempted to save from destruction at least some of the most commonly used scientific names of animals, and of course among insects primarily those related to applied entomology. Escherich (1940) has suggested the preparation of a long list of nomina conservanda, which should primarily include the economically most important insects. But Heikertinger (for instance, 1941, p. 230) has rightly pointed out that systematics cannot be helped with such lists. On which principles are certain names to be selected for the list? How could agreement on these be achieved among specialists, in Commissions and at Congresses? How much time and how much paper would be required for selecting the “deserving” names from among hundreds of thousands! Finally, if a nomina conservanda list would be prepared on such a large scale and one (really?) assumes that it represented a definite advantage for taxonomic research, then the names would no longer be “exceptions” to the rules of priority. The rules could as well be suspended and all such names as are currently valid, declared “conserved.” If after the above reasoning I still have definite reservations about the un- conditional acceptance of Heikertinger’s “principle of continuity” it is mainly on account of the possible abuse of the expression “the most frequently used names.” It cannot be expected that every zoological systematist, not even every distinguished specialist, would strive to undertake a really objective investiga- tion in doubtful cases as to which of the names of a genus or a species are “cur- rently” used most frequently. I am afraid the decision would generally favor one’s own geographical or linguistic region, for instance, the Americans would tend to conserve their own names. | indeed concur with Horn (1938, p. 2) when he says: “It would be no misfortune if temporarily here and there two or three names were considered valid at the same time (i.e. in different countries) for the same insect ferm.” It would probably not be a misfortune, but a great in- convenience which would lead to many misunderstandings. It is all the more so since the word “temporarily” used by Horn seems to be rather too optimistic. If the simultaneous use of two or more names for the same species of animal (or genus) poses only a possible inconvenience, then the gloomy word “misfortune” is appropriate for the converse situation, i.e. where the same name is used for more than one species of animals. This would not only make the work of the researcher more laborious but for long periods it would be impossible for him to decide which species was involved at one or other place in the literature, and detailed studies on ecology, physiology, etc. might become nearly worthless. 19 20 9 It therefore seems to me that one of the most important tasks of the Rules of Nomenclature is to prevent the transfer of a name from one species of animals (genus, subspecies, etc.) to another. Where no synonymic names exist it is bet- ter to give an entirely new name if necessary. A proposal to this effect was made at the Zoological Congress in Monaco (1913), but was frustrated by “compromises.” Secondly, I would like to concur with the “limitation proposition” men- tioned above. Perhaps it would be appropriate to declare definite “limiting years,” somewhat as follows: “Names of genera and species appearing before 1850 and not used in scientific literature after 1900 (or used only as synonyms) must not supersede a name used later.” The period would be better shortened than protracted. The above proposals might be considered as a compromise between the principles of continuity and priority. Nevertheless, I would approve the first without the least reservation, if only I could share Heikertinger’s confidence that enough systematists could be won over. However, gentlemen with a pen- chant for systematics are usually constitutionally conservative and less open to purely practical proposals. I believe they would generally submit only to the decisions or recommendations of entomological or zoological congresses and similar bodies. And many of those who were undecided would prefer to follow the law of priority as “safer.” There is another reason to annul the laws of priority in its rigid form. Changes in the names of species, “name-hunting,” is usually the result of dig- ging out older, long-forgotten descriptions. Now these descriptions are almost without exception so trivial that the species in question could be identified simply and solely by a study of the so-called “Type” unless the secondary char- acteristics of the animal were involved (geographical distribution, biological evidence, etc.). These “type-specimens” of the earlier entomologists—which could often be selected only as the most probable representatives-—are now preserved largely in public museums and are not loaned out in some cases (for example, by the British Museum). If the specialist cannot visit that museum he has to depend on second-hand descriptions and comparisons by someone else, photographs, etc. Preparation of genitalia and other structures, which are often essential, sometimes is not permitted by the museum directors. Worse, in the meantime many of these types have been destroyed during the Second World War, especially in Germany, and the political situation scarcely augurs a lasting peace. “Types” are even less enduring than books. To base the nomenclature largely on the former is highly impracticable. The scientific description of an animal, which cannot be interpreted without examination of the type, would be better not undertaken. The principle of continuity or limitation would do away with this exaggerated cult of types (Handlirsch, 1913b, pp. 88-89) and would leave the museums for more important matters. At any rate the rigid laws of priority in their present form must be set | 22 10 aside, so that the zoological and botanical literature can free itself from the ban of this “end in itself.” 3. Detailed studies on the variability of species—both in animals and plants—have shown that they are less homogeneous than originally believed. In many cases splitting into several taxa of a lower systematic rank has resulted, representing minor requirements upon characters which separate species. In other cases the demarcations between two or more species recognized earlier proved to be so diffuse that it seemed appropriate to merge them—if necessary into “Rassenkreise”* or “Formenkreise”** (Rensch, 1929). Even where the species constantly manifested its individuality vis-a-vis other members of the same genus, fairly large and more or less stable differ- ences were always found between individuals, groups of individuals (popula- tions), or larger entities. This led to the proposal of a fairly large number of names within the species concept. These categories have been named very diversely. But the one immedi- ately below the rank ofa species is universally known as subspecies (“Unterart,” race). The general requirement is that every subspecies be separated geograph- ically from others within the same species (for example, Rensch, 1929, 11). In keeping with this, new subspecies are frequently described almost exclusively on the basis of their more or less isolated geographical occurrence. The unsatisfactory result of this criterion is evident from the case of Carabus problematicus in Fennoscandia (Strand, 1935). There are three main types of the species found here (Fig. 1) which, if “pure,” positively deserve equal rank. One of these (strandi) lives geographically quite isolated in the far North and hence fulfills the strictest requirements for subspecies rank. The other two (wockei and scandinavicus) occur in many regionst of central Scan- dinavia so close together that at some places (particularly in southwestern Norway) the ranges of the two overlap. There they evidently hybridize, since truly intermediate forms are not exceptional in these zones of intermixing. The original status of wockei as a montane form and that of scandinavicus as one inhabiting the plains is not strictly maintained here and a geographical line of demarcation between the two forms can no longer be drawn. But there is no doubt that both wockei and strandi survived the last glaciation within the region, whereas scandinavicus represents a race that has immigrated during the postglacial period. A different nomenclatural consideration of wockei and strandi would therefore be absurd. In the usual systematic categorization these three are to be treated as subspecies. Assuming that to qualify as a subspecies it is essential that a group of individuals be isolated from other members of the species, it must be realized that isolation can be due to other factors. These include altered ecology, food *(Polytypic species: suppl. gen. edit.). **(A collective category of allopatric subspecies or species; suppl. gen. edit.). t (cf. p. 822: suppl. scient. edit.). 11 21 Fig. 1. Carabus problematicus. The Fennoscandian races. Plain circles—scandinavicus Born.; Spotted circles—strandi Born.; Black circles—wockei Born.; Crossed circles—“relicrus Hellén.” Black and white circles—intermediates between wockei and scandinavicus forms (in south- western Norway). 23 12 biology, sensory physiology, etc., thus geographical isolation must not be con- sidered as the essential prerequisite for a particular systematic category (cf. Huxley, 1940, p. 27 ff.). It would also probably be incorrect to want to deny subspecies rank to the two forms of Calathus mollis (see Supplement), just because they coexist on Bornholm, and certainly in northern Germany too (according to data in Gersdorf, 1937, pp. 81-82). The situation in the case of the chrysomelid Galerucella nymphaeae L. is still more questionable, since the two forms nymphaeae s. str. and sagit- tariae Gyll., rightly considered as subspecies by Palmen (1945), largely occupy common areas. But they seem to be constantly associated with different food plants, and attention could be drawn to the isolating factor by the designation “subspecies trophica.” The generally advanced criterion for a subspecies, as the capacity to breed with other subspecies of the same species, although to a “limited” extent, is a quite ambiguous definition and cannot be used in practice. It is in any case clear that the correct delimitation of a subspecies is far more difficult than the description of a new species. The characteristics of a species, according to its definition, have a higher degree of constancy (among insects, for instance, often manifested by distinctly marked characters of the genitalia) whereas species that can be divided into subspecies are almost with- out exception characterized by a strong general variability of minute characters. If it sometimes appears rather adventurous to describe a new species on the basis of a single specimen, a little consideration shows that such a procedure is quite unscientific when followed for a new subspecies. For the subspecies can hardly ever be defined by clearly marked characters, but only by the amplitude of variation around certain mean values. And to establish that, particularly extensive material is necessary. But such discretion in characterizing new subspecies is not always exer- cised, at least by the lepidopterists. Bryk (1942, p. 5) may be mentioned as a typical example: “In describing new species I proceeded from the irrefutable postulate that every form of a species already divided into subspecies must belong either to a known subspecies or to a subspecies to be newly recognized and described... In having before me one specimen [emphasis mine] which deviates from the known species the question arises, even if it might be an aberration, whether the specimen belongs to a new subspecies or might be referred to one already known. If yes, the specimen undoubtedly represents an individual of a new subspecies.”—The more meager the material, the more numerous the subspecies! It would be a kindness if the specialists would try to confine the characteri- zation of new subspecies exclusively to very clear cases. The lower categories are of course poorly, often subjectively, separated from subspecies. There is a whole series of terms of higher and lower or- 24 25 13 der, which sometimes seem to be merged (possibly including the subspecies category) under the term “var.” (varietas, variety, variant). An example of an exhaustive and extremely consistent subdivision of the Species is provided by the monograph of the large genus Carabus by Breuning (1932-36). He follows Semenov (1910; see also Handlirsch, 1913a, pp. 68 ff.) and uses, in descending order, the terms subspecies, natio, morph and aberration (= individual variants, which usually do not deserve a name). This imposing work offers a much-needed remedy for the confusion of names in this genus, where the most insignificant individual color variations have been solemnly named and described by ambitious “researchers.” But Breuning’s system is rigid and completely unnatural. This is especially evident from the detailed studies carried out on the mor- phology and physiology of Carabus nemoralis by Krumbiegel (1932). As is very characteristic of a taxonomist of the “old school,” Breuning (l.c., p. 665) cites this work without utilizing it in systematics. Krumbiegel has most clearly shown that C. nemoralis exhibits gradual morphological and physiological changes (at least some of which are heritable), from the northeast (eastern Germany) to the southwest (southern France), which coincide with the change from a noc- turnal to a diurnal life. Krumbiegel provided a good example of a cline (“Klin”) even before the term was proposed. The enormous advantage of the concept of cline, formulated by Huxley (1939), is that one must precisely determine and trace out the gradation of every single variable character within the area of the species. According to Huxley (for example, 1945, p. 226) the cline method in no way replaces the division into subspecies. On the other hand there is no doubt that its consistent application to a strikingly variable species would demonstrate that some (I should like to say most) of the subspecies described so far are untenable. It is deplorable that Petersen (1947), the first-ever entomologist (and the only zoologist) to analyze variations by the cline method in the Fennoscandian Region, in his studies on certain widely distributed butterflies not only largely retains the older subspecies but also describes new ones, some of which are very poorly delineated. Certain carabids of the Fennoscandian Region may also be considered Suitable subjects for future research by the cline method. One might men- tion Carabus glabratus, C. violaceus, and Cychrus caraboides, in which, among others, the following characters may be considered: body size, comparative length of pronotum: elytra, length index of pronotum: width, curvature and surface (microsculpture) of elytra; in violaceus, also the color; in Cychrus the shape of the posterior angles of the pronotum and the proportions of the legs (especially the hind legs). In the welter of confusion in the Central European violaceus-purpurascens complex some degree of order can be brought about only in this way. The lowest systematic entities, wherever possible, should be denoted only 26 14 by code words (“Kennworte”) (such as f. nigrino, f. coerulescens, f. impunc- tata, f. rufipes, etc.) without the names of authors. Otherwise they are best left unnamed (as in Lepidoptera, suggested among others by Kiriakoff, 1948). Naming of aberrations is useful only if the nature of the characters as a mod- ification or a mutation has been determined, in this case the expressions “f. mod.” or “f. mut.” may be used. The wing dimorphism dealt with in this book shows that even a strong morphological as well as a physiological modification can be due to the influence of a single gene. With the exception of the cases mentioned above, members of the Fennoscandian fauna of Carabidae are unusually homogeneous, since they show very little regional variation within the region. Therefore little attention has been paid to this aspect in the present work, nor is there a special chapter devoted to this problem. The reason for this uniformity in the fauna is undoubtedly the infancy of the fauna, as well as the extensive enlargements of its area during the later post-glacial period (< 16,000 years), which have been of the nature of enormous and incessant intermixing of populations. Nevertheless the stocks which remained isolated along the Norwegian coast during the Wurm glacia- tion, of which numerous examples are given below, show a tendency to form subspecies only in exceptional cases (Carabus problematicus). Thus normally the formation of new systematic entities, even of a lesser rank than species, requires longer periods of time than is generally assumed (Lindroth, 1941, p. 437 ff.). Among aberrations in Fennoscandian Carabidae the following might be mentioned: all gradations of nigrino and rufino forms, the former of these predominating in Calathus melanocephalus, the latter being especially marked in Amara alpina, A. torrida, Bembidion prasinum, Nebria gyllenhali, Patrobus septentrionis, Pelophila. In my opinion both of these represent modifications. On the other hand the black- or red-legged character of some species is probably genetically determined, for instance in Anisodactylus binotatus, Ca- lathus fuscipes, Chalaenius nigricornis, Harpalus aeneus, H. fuliginosus, Nebria gyllenhali (balbii), Pterostichus cupreus. Of these, the red-legged Nebria gyllen- hali balbii alone is regionally restricted. —The color is especially variable in most of the metallic species, chiefly Harpalus aeneus. Distinct pairs of subspecies are found only in Badister bipustulatus, Ca- lathus mollis, Carabus cancellatus, and Patrobus septentrionis. In all these cases there is an eastern and a western subspecies, which partly coexist only in the case of Badister and Calathus. Some very closely interrelated pairs of species, such as Amara cursitans and municipalis, Badister dorsiger and sodalis, might possibly represent subspecies. The status of the “subspecies silvicola” of Amara quenseli is doubtful. 27, On the Consideration of Literature The volumes of the Zoological Record of recent decades placed side by side convey an idea of the relative output each year in zoology. The inhibiting effect of the two world wars is all too evident. Still, during the 20th century there has been a sharp increase in the zoological literature. This growth has been especially marked in entomology. The unavoidable consequence is that every careful researcher must sac- rifice much time to the study of the literature in his field. This burden has become so great that the question arises whether this time could not be better spent. Some authors go all out for a “complete” list of references in their par- ticular field almost as a sport. It is all the more annoying when at a Swedish inaugural dissertation the opponent puts it to the poor respondent that he has “forgotten” a certain reference, regardless of whether this would have had a vital bearing on the theme. As long as the zoologist or botanist, as a specialized systematist, works within the confines of a well-demarcated group of animals or plants, the requirement of knowledge of all the literature, if possible, in that particular field is justified. The specialist must not only be familiar with the animals but must also be aware of the views published by other authors on the systematic position of every species, genus, etc. Otherwise many statements contradicting each other will frequently remain uncontested, which would result in confu- sion. A good example among carabidologists is provided by Jeannel, author of the exemplary monograph on Trechinae (1926-28). When he later turned to other carabids, in order finally to revise the entire family in Faune de France (1941-42), for unknown reasons he largely ignored the literature. This natu- rally resulted in completely unnecessary mistakes and misunderstandings. In particular, his data on the total area of each species are frequently misleading because they are incomplete. On the other hand, the reverential consideration of the literature can be overdone, even by a specialist. Especially among phytogeographers, for every species treated a detailed report is often provided on the history of its discov- ery in the region, including all the literature (for instance, Degelius, 1935). Although it reveals an interesting part of the history this method seems inad- visable to me. It tends to enlarge the scope of a biogeographical monograph beyond reasonable limits, without adding to the material content. The situation is somewhat different in the fields of physiology, ecology, zoogeography, etc. Here a knowledge of the animals studied is not an end in itself, but the material under study is to be used as a basis for conclusions with general validity or at least a validity beyond the scope of the material. Conversely, the results obtained by other researchers on other animals can 29 16 naturally. be of great significance in such a study. The question is: To what extent is one obliged to take such literature into consideration? The zoogeographer is in an especially difficult position when he requires information from the auxiliary sciences like general geography (even cultural geography), geology, climatology, also often soil chemistry, the theory of heat in physics, even archaeology, etc. Besides, he must always keep an ~ eye on the experience gained by botanists. In these fields he is usually a complete novice who must rely on the literature or on statements by other experts. My personal opinion concerning study of the literature by a zoogeographer is as follows: He must thoroughly acquaint himself with the primary material. He should put together as much data as possible on the taxonomy, distribution, ecology, etc. of the animals concerned, and spare no pains to completely gather this information from the literature, the museums, or private individuals. Such data form the basis of the first two parts of this book. Next follows the collation of the primary material, the endeavor to answer all the questions that arise during this compilation, and those characterizing the complicated problem of every species: How has the area taken its present shape? A judgment can be arrived at by two methods. Either one can iook for many, perhaps similar, cases and problems in the literature and check the experience of other researchers against the case at hand. Or, one can treat it as an isolated problem using observations in nature, experiments and wholesome “farmer’s sense,” without prejudging the issue. Although in general perhaps a combination of the two methods seems desirable I decidedly prefer the latter. This choice is not one of convenience, and it must not be judged as an expression of haughtiness (“Hybris”). But as far as I can decide, a thorough knowledge of the case in hand is more important than of the method used by other researchers. Finaliy, I did not have enough time to go through the relevant literature on zoogeography, phytogeography, ecology, physiology, etc., and I am convinced that such a study would have deprived the working task of much of its liveliness. After all, the method adopted is therefore an expression of personal preference. This is my defense against the reproach, certainly justified in some cases, that I have not sufficiently taken the literature into consideration. I have uti- lized such works as I have often come across accidentally that might contribute to a solution of the problems posed. A complete historical review of a partic- ular branch of research has never been achieved. On the contrary, I would like to assert that it is often useful to re- searchers concerned with the general problems of ecology, physiology, etc., to go more deeply into the elementary characteristics of the objects of their investigation. By this I mean that such investigations should be preceded by a thorough knowledge of the taxonomy, biology (in the broader sense) and 30 19 distribution of the species of animals concerned; but this is not always the case. On the Task of the Museums The public and private collections of natural history specimens had their be- ginning as showcases containing objects of curiosity. As descriptions of the animals began to appear in print from the 18th century and they were classified into a system, they ceased to share the status of philately, yet the collections largely remained at that level to this day. In the meantime, the requirements of entomological researchers, especially in the fields of zoogeography and ecology, have also greatly added to the tasks of public museums and hence created more difficulties for them. To get an idea of the extent to which the larger museums can satisfy such modern requirements, and of what improvements seem to be especially desir- able, Ihave made a compilation of the existing conditions mainly through en- quiries addressed to the curators of leading museums of the Nordic countries. The curators of museums who answered all of my enquiries in detail, for which I sincerely thank them here, were: Zoological Institute of the University of Lund! (ML). Lecturer Kj. Ander. Zoological Museum of the University of Oslo! (ML). Dr. L.R. Natvig. Zoological Museum of the University of Helsinkit (“Helsingfors”, MH). Dr. R. Frey. Zoological Museum of the University of Abot (MA). Professor K.J. Valle. Zoological Museum of the University of Copenhagen (“Kobenhavn”; MC). Dr. S.L. Tuxen. Through my own experience I was sufficiently acquainted with the en- tomological collections of the Natural History Riksmuseum, Stockholm (RM; Director Professor O. Lundblad) and the Natural History Museum, Goteborg* (MG; Director Fil. lic. H. Lohmander). For the sake of clarity the result of the enquiries is presented here in the form of 10 questions, which were directed to the museums concerned, and in each case some remarks have been added. It was assumed that each museum was concerned with the fauna of its own country, which is actually the case at present in all the museums mentioned. The following summary takes into account only the collections of Coleoptera, but in principle might be relevant to other orders of insects. 1. Has all the indigenous material been brought together into a single tcf. pages 19, 21, 22 of Part I: suppl. scient. edit.). *In MG, for lack of staff, so far only Carabidae. Heteromera, and Longicornia have been provisionally put together. 31 18 collection? (see also question 6!). Or are there also special collections which are maintained separately for reasons of reverence, as voucher material of certain publications, as special geographical collections, etc.? The answers revealed that all the museums mentioned have tried to avoid, so far as possible, any distribution of the indigenous material among sev- eral collections. In some cases donations of material were even refused be- cause of the condition attached that the collection must be maintained sepa- rately. However, in the museums mentioned, with the exception only of MO and MA, besides the main indigenous collection the following special collections of Coleoptera are present: a. For reasons of reverence: coll. Boheman in RM: coll. Roth in ML; coll. .B. Ericson in MG. b. As voucher material of scientific publications (partly also acquired pri- vate collection): coll. Zetterstedt (“Insecta Lapponica”) and coll. Thomson in ML; coll. Lindroth (only carabids) in MG; coll. Mannerheim (world collection) in MH; coll. Schigdte in MG. c. Special geographical collections are maintained only in RM from the nature reserves of Gotska Sandon as well as Sarek and Abisko in Lapland. d. In addition, all museums connected with universities have special ref- erence collections which are not taken into account below. Horn (Horn and Kahle, 1935-37, p. 497) has listed reasons why the sep- arate arrangement of a collection may be justified. “Reasons of reverence” is not one of them. It does not seem to be justified to maintain separately the above-mentioned collections of Boheman, I.B. Ericson, and Roth, since their owners published comparatively little on the indigenous Coleoptera, and the collections can scarcely serve as voucher material. The present curator of ML is of the same opinion. On the other hand, the four collections mentioned in par. “b” above, brought together by leading entomologists of their time, are historically valu- able. Even the unlabeled insects among them, which would be put away as a result of merging with the main collection, are valuable as voucher specimens for the opinions of the respective authors. The Zetterstedt collection may also be considered a special geographical collection. Such collections are otherwise present only in RM, where the cura- tor also seems to be inclined to discontinue their separate maintenance. But in my view special geographical collections, for instance of islands or regions that have recently undergone big changes as a result, say, of damming operations, are very important for entomologists working with zoogeographical problems. So are special ecological collections, for instance from sampling areas where an exact inventory of the fauna has been realized. I would prefer to see an increase rather than a decline in this practice. 2. Has a certain maximum number of specimens been laid down for every 32 19 species in the main collection? What is generally done with new material that does not earn a place in the existing box for the species? The answers show that at present a limit on the number of specimens is imposed only in RM and MA, in the former case according to provinces (see question 3), and in Abo (for the time being) restricted to 20 (—40) specimens of each species. 3. Are specimens of every species arranged geographically? If so, is there a limit on the number of specimens from each province? All the answers expressed the desirability of having a geographical ar- rangement of the material. But in MG, MO, and MC it had to be dispensed with for lack of staff. A consistent limit on the number of specimens from each province is followed only in RM, where it is “half a transverse row” (= 3-4 pins, each row often with several insects). 4. Is each province represented by as many localities as possible? Or if the Space assigned to the province has been filled up, are any specimens coming in later considered duplicates? All but one of the museums try (probably within limits due to lack of space), to acquire representative specimens from as many localities as possi- ble. The exception is RM where a full “half-transverse row” in the box means no more specimens from the province are to be incorporated in the main collection, even if all the specimens aiready present originate from a single locality. This procedure is in no way consistent with the present-day require- ments of a geographical collection. 5. What is to be done with large series from the same locality? In MO the maximum number of specimens from every locality is fixed at six. Other restrictions are evident from the answers to questions nos. 2 through 4. Only MG, MH and MC try to maintain a large undivided series. 6. Is there a special collection of duplicates, that can be utilized for ex- change or donation? There is in RM, MO, and MA. In MG, ML, MH, and MC all the material is maintained in the main collection (at least for the time being), or some of it is temporarily kept in supplementary boxes. An answer especially congenial to me was given by S.L. Tuxen (MC): “The term of “duplicate” simply does not exist!” 7. Must all the material be uniformly prepared? Or is there a perfectly prepared special “systematic” main collection, while the “geographical” col- lection is more heterogeneous? Only in RM does the curator insist on perfectly mounted material (on rectangles) for the large (geographical) main collection, established about 15 years ago. Some of the “better” older insects are remounted for this purpose. All the other curators of museums entirely discount a uniformly prepared main collection. 33 34 20 None of the museums has separated the “systematic” from the “geograph- ical” collection, but Kj. Ander (ML) considers such a distinction desirable. 8. Must specimens be “perfect”? What is done with defective specimens? All but one of the curators replied that “perfection” counts only in the sense that highly defective specimens are replaced by better ones (from the same locality or at least from the same province). The exception is RM where, with the exception of very rare species, a defective specimen is not included in the main collection if as little as an antenna or tarsal segment is missing. It can be said that such an attitude to a geographical collection is quite unscientific. An identifiable fragment from a distant locality is far more inter- esting than any “show-piece” from an insignificant region. The most far-sighted definition of the term “defective,” even from the viewpoint of the systematic entomologist, was given by Ganglbauer (Heikertinger, 1914, p. 137): “So long as a specimen possesses one complete antenna and three different legs un- damaged it is not defective.” 9. Has the attempt been made to achieve uniformity in the contents and design of the locality label? How is the indication of the collector’s diaries given, if they exist? In all the museums at present the collector’s original locality labels have been retained (if they exist), which is undoubtedly correct. If these labels are much abbreviated or badly written, occasionally a more detailed printed label is attached, at least in ML. It would be welcome if this procedure was adopted more commonly for indistinctly labeled, especially older material. To date the exemplary practice of affixing a register number (“Journal- Number”) to every insect is followed as a rule only in Finland. General compliance would be highly desirable. In this case, it is furthermore important that the museum acquires a complete copy of the daily diary of the collector. This has been the case with MH since about 1850. 10. What is the procedure for maintenance of genitalia and other prepa- rations, larvae and pupae preserved in liquid, pieces of food, etc. relating to a Species or a specimen of the pinned collection? I was informed by ML and MA that genitalia preparations, wherever pos- sible, are attached to the pinned insects on a strip. If film or similar material is used, preparations in Canada balsam, etc. can also be affixed in this way. Sep- arately maintained preparations are indicated by definite numbered labels in all museums. It is recommended that for such labels colored paper indicating the particular collection be used. The above exposition is a survey of the Ihe at present followed with regard to the indigenous collections of our museums. In various remarks my own views on these questions may have found expression to some extent. However, I will here briefly recapitulate them more precisely. 35 21 All accurately and reliably labeled material is valuable. There are no “du- plicates.” Identical stamps do exist but not identical animals. It is therefore insufficient to strive for representative specimens from as many localities as possible; the museums must maintain large series of animals collected at the same time. The present work clearly shows, in the treatment of carabids with wing dimorphism, how important the presence of larger series can be. Cer- tainly in the future, when according to all indications, pure systematics will also get involved in anaiysis of variation, the need for quantitative completeness of material will be more and more urgent. The objection that the museums have neither space nor time (i.e. staff) for this can be met in the first place by the argument that less important pursuits should be restricted! As soon as more reasonable conditions with regard to entomological nomenclature are achieved, the decline in the type- cult on the part of museum officials will save many valuable working hours. Also the mistaken pretension of some museums of being in possession of a Palaearctic or even world collection, deprives them of much energy. If every museum accepts as its main task the maintenance and compilation of the indigenous fauna it will be possible to treat this collection according to modern concepts. This does not mean that special collections of another kind (for instance, material collected from foreign countries or world collections of smaller systematic groups, which have been acquired incidentally by donation, exchange, profitable purchase, etc. or as a result of foreign expeditions by local researchers), must be neglected. From the purely practical viewpoint the indigenous collection must be arranged in such a way that supplementary boxes can everywhere be fitted in and all the boxes replaced without any problem. It is impossible, even af- ter the establishment of large geographical collections, to estimate how much space a species will require after 50 or 100 years. It is impracticable to put all the specimens that cannot be accommodated in the proper boxes in a special supplementary cabinet. It may be said that such an enormous collection with hundreds of specimens of almost every species must be difficult to survey. It is, for the usual systematic purpose, i.e. identification. I am inclined to favor the solution given above, that besides the large main geographical collection an indigenous main systematic collection be maintained. In this, normally two pairs of every species (several specimens of variable species) would be pre- served, all mounted perfectly and “flawlessly” and provided with the necessary genitalia or other preparations and “authoritative” determination labels*. This collection should also contain all the uniques, which would be indicated in the geographical collection where appropriate by a small label. On the whole I am no opponent of special collections of the kind indicated *In general the museums should strive to supply a determination label with every animal bearing the name of the determinator. 36 22 above (Point 1). But it is also appropriate that if a species is not represented in the geographical collection from the province concerned, an indication on a small colored label be provided referring to the special collection. If such indications are consistently entered in the geographical collection, along with information on genitalia preparations (when separately preserved), any larval collection, pieces of food, etc., this will serve as a kind of central card index, making it unnecessary to organize a special card index. On Synecology and “Syngeography” The word autecology means the branch of biological research in which the responses of the individual species of animals and plants are studied in rela- tion to the environmental conditions. On the other hand synecology signifies not the species but the community of organisms, the biocoenosis (for instance, Krogerus, 1932, p. 190; Thienemann, 1939). However, the non-quantitative de- termination of the dominant or characteristic species in a particular biotope, or of the more or less regular association of animal and (or) plant species, also falls within the realm of synecology (Bodenheimer, 1938, p. 134). Thus the fol- lowing section on the “limestone species” is to some extent to be considered as synecology. In practice, however, the material of synecology can also be used for quan- titative collection. For this study such collections were not undertaken, so I suppose I should explain why. In my opinion the methods at present favored in zoological synecology (at any rate so far as relates to terrestrial animals) are unsatisfactory. The present situation may be considered under the following heads: 1. A quantitative determination of the carabid fauna of different biotopes or selected sampling areas, without taking into account the other animal or plant inhabitants, would be rather futile. In this way only a small part of the biological group of “consumers” (Thienemann, 1939) would be taken into account. This part cannot be expected to represent a restricted functional unit and consequently a stable large component of the fauna. Even where larger systematic entities, such as the entire fauna of Coleoptera (Brundin, 1934; Renkonen, 1938) are selected as the subjects of quantitative study, from the viewpoint of food biology it means the treatment of only a fraction of the consumer group of the biocoenosis in question. In the case of Coleoptera, besides, it involves a mixture of various types. On the other hand, anyone who wants to take into consideration the entire (macroscopic) animal world must restrict himself to a very small number of biotopes (for example, Krogerus, 1932; Franz, 1943a). A quantitative study that takes into consideration only the carabids would actually elucidate nothing more of the “community” of the recorded biotope 37 38 23 than—using a human example—the professional distribution within the family with respect to the business conditions of a city. This is also true of the work of Renkonen (1944), where only carabids and staphylinids are statistically treated. On the other hand, this can naturally be used to take up other questions, especially the problem of competition (cf. p. 554)t. 2. Even if we abandon our overall objective to describe the entire biocoeno- sis of a small sampling surface, and thus restrict ourselves to the members of a single group of animals (for instance, insects), to determine their qualitatively and quantitatively changing stock in different biotopes—then too it is essen- tial to give a detailed description of every sampling area including its primary inhabitants, the autotrophic plants. The persistent dispute as to whether sam- pling areas should be considered and named according to the nature of their flora or exclusively to the composition of their fauna* is inconsequential here. The most important point is that the consumers cannot be judged without the producers, insofar as one actually wishes to causally explain with reasons the existence of one or other species of animals. However, in synecological investigations the causality is all too often ig- nored. Most people are satisfied with purely statistical descriptions of the animals or plants of units which are called biocoenoses, etc. and at best make comparisons, mainly quantitative, with other biocoenoses, ignoring the rela- tionships among the members in the biocoenosis under study. With few exceptions synecology has therefore become a purely descriptive science. Its units, the biocoenoses, are compared with taxonomic units, the species (cf. Taylor, 1935). It is implied that a detailed description possible in the latter, is as important in the former, in order to form the basis for future, more causally-based research. Synecology today should therefore be at the same Stage as autecology was at about 100 years ago. 3. But in my opinion it is not appropriate, or even possible, to divide syne- cological research into a descriptive and a later “causal” period. The syneco- logical unit, the biocoenosis, is a far more abstract concept than “the species.” It simply represents the mean value of a large series of individuals (of the sam- pling areas or sampling volumes), and “hybrids” are frequently more common than typical cases. Each of its units possesses a stronger individuality than the individual animals within the species. The relationships between the animal and plant inhabitants of a sampling area must therefore be studied in nature itself and not only quantitatively. Even when describing them, the formula- tion of the problem must be clear, namely: To what extent are members of the biocoenosis mutually dependent? tcf. footnote page 10; suppl. scient. edit.). *For literature on the subject see Hesse, 1924, p. 143; Krogerus, 1932, p. 221; Brundin, 1934, p. 42 ff.; Franz, 1939, p. 376 ff.; 1943a, pp. 402, 483; Agrell, 1941, p. 62; Kuhnelt, 1943; Tuxen, 1944, p. 171; Backlund, 1943, p. 175; Gisin, 1947. 39 24 There is a precondition: One must be acquainted with the autecology of the organisms. And that takes us to the main problem: Ir is misleading to wish to establish complex units before their components are sufficiently known. Before one is familiar with the autecology of the species, synecology is impossible. Otherwise we will deal with statistics, using numbers which were achieved in an obscure way, thus are incomprehensible. One can describe H,SO, and utilize it in practice, but its true nature cannot be understood without a knowledge of its components, its basic substances, and their characteristics. According to Thienemann (1939, 1941), the biocoenosis is “an association of organisms within the same space, in which the individual components of the community have determined, vital relationships with one another.” Now these relationships must be determined before one is justified in considering the biocoenosis as a functional unit (see Bodenheimer, 1938, pp. 134 ff.), as an “organism of a higher rank” (Friederichs, 1930, pp. 232 ff.). 4. The prerequisite for expressing the composition of a biocoenosis sta- tistically, i.e. the abundance of its members (of the species or at least the food-biological groups) shows constancy, is that it be “saturated.” This indi- cates that there exists a balance not only among the producers, consumers, and reducers (bacteria and fungi*), as well as the consumers of different levels (herbivore, predator, parasite, etc.), but also “laterally,” for instance among various species of predators. A judgment on this subject hits the most difficult complex of questions in autecology, the problem of competition between ecologically related species, which is touched on below (p. 554). It is not possible at this time to decide on the significance of competition. It should be taken only as the expression of a subjective understanding when I assume that the terrestrial communities of animals are generally not saturated, at least not constantly. The basis of such an assertion may be found in the often highly vari- able number of individuals of a species in different years in the same place, especially among insects. Above all, if the competition factor is strongly in- fluential, one should expect a definite succession of zoocoenoses, especially in newly emerged or much altered biotopes. Here, the stronger competitors among the species that at first have arrived accidentally should gradually be- come prominent, even when the vegetation and other environmental factors remain unchanged (or else after they stabilize). In such a recent region as Fennoscandia, where numerous species of animals are still in the process of Spreading out, such a succession of biocoenoses ought to be the rule. To me, however, it appears improbable that it would have remained so completely un- noticed. In the phytocoenoses it is easy to see, for instance on newly formed *In the soil, fungi are to a large extent nitrogen accumulators. Animals that feed on mycelia (especially acarids; Forsslund, 1943, pp. 167, 175), and thus set nitrogen free may to some extent be considered as reducers. Concerning the significance of other soil animals in the turnover of nutrients, see Franz (1943b). 40 25 islands or in burned-over fields. It may seem bold to oppose on such feeble grounds the assumption appar- ently shared by most zoosynecologists, as to the “saturation” even of terrestrial animal communities. But it seems to me the onus of proof is on them. It would be incomprehensible to simple “farmer’s sense” if the animal members of a biocoenosis, with all their fluctuations from year to year and in the course of a single year, were actually capable of using completely all the food afforded by the plants (directly or indirectly) at all times. By far the larger part of the plant material is not utilized as nutrition by animals but is left for the “reducers” (bacteria, fungi, etc.). I doubt there could be any proof that the animals might be able to take over even a certain percentage of these plant materials, so that the normal metabolism of the producers, the green plants, was not deranged. Moreover, part of the nutriment utilized by the animals is passed on to the reducers. In the limnetic and marine soil zoocoenoses, where the animals are more or less sedentary, the competition factor in my view plays a far bigger part. For that matter water is a far more stable medium than air, so far as temper- ature, light, movement, etc. are concerned. The abiotic environmental factors, especially the climate, but also enemies and diseases (which represent some- thing different than “competitors’’), may exercise an appreciably stronger influ- ence on the constantly changing terrestrial zoocoenoses than the competition factor (see Bodenheimer, 1938, p. 135, and the example given by Elton, 1930, p- 17 ff.). Their influence on various members of the “community” is again a task of autecology. An extreme example is provided by the three more or less constantly oc- curring species of animals (two diptera, one snail) of the hot springs (“the ab- solutely hot springs”) of Iceland (Tuxen, 1944). All of them are phytophagous, but competition among them for food could not be established. It is very im- probable, since the Cyanophyceae on which they live occur in great numbers. The sole factor responsible for the occurrence of these three species in the hot springs appears to be their resistance to high temperatures, the accurate determination of which is the task of autecology. It is therefore difficult to have understanding what advantages or which higher forms of truth are re- vealed by considering the species in question as a “community,” which is so much emphasized by Tuxen (pp. 59 ff., 100 ff.). To some extent the above observations are also applicable to plant sociol- ogy, which can otherwise function (almost) independently of animal sociology. My reservation refers to the stronger, and at least easier detectable competition among individuals and species within the plant community. It is distinctly a competition “for space” (for example, P. Palmgren, 1930, pp. 15-16; Krogerus, 1932, p. 10). In the case of certain trees, such as spruce (Picea abies) against pine (Pinus montana), this competition is a common knowledge. There is no doubt that the “biocoenotic” factor has a far bigger role in the phytocoenosis 41 42 26 than in zoocoenosis. For this reason, statistical data on the number of indi- viduals (density) or the “degree of plant cover” (“Deckungsgrad”) within a sampling area may reflect more of a regular reality. From another viewpoint plant sociology is even worse off than animal sociology, owing to the sampling area-method. A botanist surveys the object he wishes to analyze fairly completely, i.e. the vegetation, even before he de- marcates his sampling areas. His choice is not random: it is determined by his conception of what is “typical.” The botanist might consider this an advan- tage, but it really means that with the best of intentions he cannot undertake a study of his object without preconceived ideas. He chooses his sampling areas, and the result of his study, the description of the phytocoenosis, is more or less tainted by his preconceived opinions about its composition. The animal sociologist, who works with terrestrial soil communities, is far more fortunate. In choosing his sampling area he takes into consideration the general (biotic and abiotic) characteristics, including the plant cover, and cannot be misled into a preconceived selection based on the characteristics of the actual object of his study, the faunal composition. The animal sociologist does not need to be that strong in morality! Like animal sociology, plant sociology is predominantly a descriptive sci- ence, which is pursued statistically. In particular, it has an extensive termi- nology of its own, which often gives the impression of being an end in itself. To what an appalling extent nature is forced into a straitjacket becomes clear from the comprehensive account by Du Rietz (1932) (cf. P. Palmgren, 1930, ‘p. 10). Naturally, in plant sociology the question of the causes determining phy- tocoenosis obliges us to indicate the characteristics of the individual species of plant (especially of the dominant species) (for instance, Du Rietz, loc. cit., p. 474). Why then is it not clearly stated that: The synecology is incomprehen- sible without sufficient knowledge of the autecology of every plant? Probably because the latter cannot be determined satisfactorily without experiments. At any rate the Uppsala school, which is mainly engaged in plant sociological research, is not just indifferent, it is almost hostile toward experimental work. They seem to posit a sixth sense, enabling the trained researchers in the field to pinpoint the area-limiting factor in each case. Actually the most difficult problem, decisive for an understanding of the biocoenosis, is the factor of competition, which can be successfully grasped only by experiments. An in- stance is the “Taraxacum” cultures by Sukatschew (1928). Tansley (1946, p. 27) states: “The ‘ideal’ method of study might be to investigate each species separately, till we know in detail its life history, the methods by which and the rate at which it could spread, its behavior under different conditions of climate and soil; and only when we had obtained this knowledge proceed to study the species as it existed in communities with other 43 27 species.”? As a plant sociologist Tansley afterwards understandably considers this method as “quite impracticable”, with “such complete knowledge” being unattainable. The method is not necessary indeed! Actually in certain cases exact knowledge of a single characteristic (such as the temperature requirement, limestone requirement, parasitism, etc.) of the organism concerned can suffice to explain its choice of biotope. This in- formation is then more important than which biocoenoses the species is a member of. Furthermore, if the method indicated as “ideal” by Tansley is out of reach, it shares this characteristic with all other “ideals”, since this very concept implies unattainability. But this does not mean we should not aim toward it. Finally, comprehensive studies on the modes of dispersal of plants are urgently needed. It is a great pity that among the many projects suggested by Sernander just this one is hardly pursued any more in Sweden. If today one wishes to get information on the methods of dispersal of a very common plant of our own flora one still often depends entirely on foreign literature. At the beginning of the period of plant sociology in Swedish botany the “for” and “against” were expressed in the following significant phrase by Kylin (1923, p. 233) (translated from Swedish): “The different species organize them- selves independent of one another, according to the ecological conditions.” Th. Fries (1926, p. 5) finds this formulation agreeable, but wishes to make a small correction: He wants “independent of’ to be replaced by dependent on! In my view, the truth lies somewhere between these two concepts. It is senseless to deny the influence of biotic factors among various organisms, but they are not so dominant or so regularly defined as to represent the bio- coenoses systematic entities comparable with species, genera, etc. The defini- tion of biocoenosis by Thienemann, quoted above, according to which “the in- dividual components of the community show definite interrelationships which are essential for life,” might represent a beautiful thought from Plato’s world of ideas. I prefer the view stated by Uvarov (1931, p. 161): “.... the theory of stable equilibrium is based on the assumption that the numbers of an organism depend mainly on the numbers of their enemies and on the quantity of food, i.e. on factors which in their turn are dependent on other organisms. No one will deny the controlling value of these factors, but the evidence... should go far towards proving that the key to the problem of balance in nature is to be looked for in the influence of climatic factors on living organisms.”* Species with more or less similar ecology can be combined in groups (as, for instance, by Larsson, 1939, pp. 433 ff., in the case of Danish carabids). The zoogeographers may also find it appropriate to treat species in groups on the t (Original quotation in English; suppl. scient. edit.). *(Original quotation in English: suppl. gen. edit.). 45 28 basis of more or less similar total or partial distribution. This is syngeography in its most modest form. This branch of study reaches its full development when we pass from groups of species to surfaces. Hence the zoogeographical and phytogeograph- ical regions are the true counterparts of the biocoenoses of ecology and like them, can be quantitatively (statistically) treated. The most recent and the best division of Scandinavia into zoogeographical regions and subregions is that by Ekman (1922, pp. 547 ff., Fig. 142; Fig. 2 in the present work). It is predominantly based on the vertebrate fauna. The carabids are a group so rich in species, comprising so many zoogeo- graphical elements, that they could well form the basis for an independent regional division of Scandinavia or, better, the whole of Fennoscandia. Cer- tain changes in Ekman’s map would then be called for. First, the Norwegian “western country” would have to be elevated to the rank of a separate region: its carabid fauna and the beetle fauna in general, possesses quite sufficient pos- itive indications (as also the terrestrial mollusk fauna, according to Okland, 1925, p. 149). But the division of the “high boreal region” adopted by Ekman can scarcely be upheld. In particular the “northern Baltic coast” would be too weakly characterized. The birch-tree region (but not the more extensive subarctic region in Ekman’s sense) ought to be downgraded to a subregion. Within the “south Scandinavian region” doubts arise only in the case of the Norwegian southeastern coast. Otherwise the entomology only confirms Ek- man’s subregions. If Ekman’s map is compared with one of the more recent phytogeograph- ical maps of the same region (for example, Du Rietz, 1925, 1935; Hard, 1939), irrespective of the changes suggested above, the common features are striking. In particular, the southern boundary of the “high boreal region” and that of the “north Swedish coniferous forest region” are almost identical. Considering Coleoptera, it could be clearly identified in Varmland (the valley of the Klaralv River) (Palm and Lindroth, 1936, p. 40). There are considerable differences only in southern Sweden, where the botanists divide the south Swedish moun- tain region with a more or less strict North-South line into a “Subatlantic” and a “Central Baltic” region (Hard, 1924, p. 226). This is based for instance on the common occurrence of Erica tetralix or the distribution of Narthecium _ ossifragum, but these regions are not sufficiently corroborated by the fauna (see map, Fig. 61). On the whole, however, the correspondence between the phytogeograph- ical and zoogeographical regions in Scandinavia is so great that the question arises: Is it really necessary or appropriate to undertake separate division according to the zoological and botanical specimens? Would it not be possi- ble to have a synthesis or could one branch of research not utilize the regions recognized by another? These questions recall the corresponding and likewise different views concerning the foundation of biocoenoses. . Fig. 2. The zoogeographical division of Scandinavia. (After Ekman, 1922, 44 576). p. 46 30 “The regional zoogeography should divide the earth into regions that ex- press the actual distribution of animals and the phylogenetic correlation of the faunas as truly as possible” (Ekman, 1940, p. 18). On the other hand, Reinig (1937) exclusively follows the historical method, which is believed to “reveal the centers of evolution,” “centers whose characteristic is more lasting and more easily comprehensible than the very hypothetical boundary lines of organized zoogeography” (l.c., p. 72). By “centers of evolution” Reinig under- stands glacial refuges. Study of the glacial history offers by far the most important basis for judg- ing the present-day fauna of the Northern Hemisphere. However, one fully agrees with Ekman when he rejects this viewpoint as the sole guiding princi- ple, since it has to be pursued with altogether too many purely hypothetical preconditions. Regional zoogeography is therefore primarily a descriptive science, and one might expect that on account of its relative objectivity it would lead to fairly unequivocal conclusions. Quite the contrary. Let us compare the two maps of regions by Holdhaus (1929) and Semenov (1935), based on ento- mogeographic facts. The former covers the entire globe and the latter the Palearctic Region. Even apart from the fact that Semenov’s map divides the subregions into provinces there are considerable differences, especially in Cen- tral and Eastern Asia. If other animal groups are considered, still greater diver- gences result. It is scarcely an exaggeration to state that every zoogeographer dealing with these questions in detail has his own system. An example of set- ting up a very unnecessary faunistic province is “the northern temperate, east Atlantic faunal region” (Iceland, The Faeroes, The British Isles, ? northwest- ern France, ? western Norway) named by Braendegaard (1932, p. 33). What then is the purpose of these regions, including all of their subdivisions? Is it arrangement (“Ordnung”) just for the sake of arrangement? The regional boundaries are often so ill-defined and the results so diverse, depending on the group of animals considered, that one is inclined to take this irreverent view. And any future incorporation of the entire terrestrial fauna into one regional system would certainly be so full of compromises that the transitional zones would occupy a larger total surface than the regions and their subdivisions. Where the regional zoogeographical boundaries are sharp, for example, the northern boundary of the Mediterranean Region and the forest-steppe boundary in Asia, they coincide almost without exception with the phytogeo- graphical boundaries. The two most distinct zoogeographical boundaries in Scandinavia are the timber line in the fjeldst and the southern boundary of the northern coniferous forest region. So I do not consider it a misfortune not to have a special zoogeographical division of Fennoscandia into regions. t(= barren plateau of the Scandinavian upland; suppl. scient. edit.). Sil 47 Regional zoogeography can under no circumstances be studied without plants, especially the forest patterns, and, besides, these are the most concrete expo- nents of the climate of the landscape. My answer to these questions is: At least within limited geographical re- gions it may be practicable for the zoogeographer to accept the phytogeographical regions and work with them as units. For my purpose the map of the forest — regions of Fennoscandia (Fig. 61) sufficed. I was chiefly concerned with an evaluation of the climatic requirements of the individual species according to their distribution in different “regions,” including the high altitude belts of the fjelds. The most important task of regional zoogeography is not delineation and description of regions along with divisions of a lower rank but comparison be- tween the faunal stocks of the larger or smaller regions. This should stimulate causal research, especially a historical reflection. I do not know whether such faunal stocks have a homogeneity comparable with the large phytogeographi- cal regions of the earth, but I think they may well have. It need not be especially emphasized that the above discussion relates dnly to the regional geography of terrestrial fauna. The marine, and to a greater or lesser extent the limnetic fauna is not directly dependent on the plant world. - The content of this section has a distinctly negative character, and betrays the author’s limited understanding of every form of collective treatment of the species of animals and plants. This is due to his humble acknowledgment of the species as, in any case, a relative physiological entity, whose responses must be studied, before it can be combined with other entities into larger categories. This is because “every grouping of the material all too easily conceals the basic fact that every species poses its own problem.” 4gt 49 Analytic Part The Realized Experiments The experiments represent an attempt to isolate the external factors which operate on the animals in nature and to judge their influence. However, isola- tion signifies something unnatural. For instance, we can never claim to know the accurate temperature preferendum of any particular species, because there is no such thing. It is among other things dependent on humidity. it would of course be possible to determine the temperature preferendum for a partic- ular stage of the animal at constant humidity of the air, but this too would not determine any “natural” complex of factors. In nature it is namely the most essential point how the animal in its biotope reacts to daily and yearly changes, amongst other things to exactly those of temperature and humidity. And these facts can be determined only on the spot, by means of extensive microclimatic measurements, which are not available at present. —Besides, the same species may respond differently at different times (even in the adult stage), many examples of which are given below. These and other considerations directed me towards always obtaining the comparative values in all experiments. This means that every experiment was carried out, if possible, simultaneously (or at least under similar external con- ditions, as far as possible) with two or more species which, if possible, were collected at the same time and kept in captivity under identical conditions. It is best to select two systematically closely related species, differing in their distribution or ecology. If the experiments show a correspondingly different response of these species one may then be justified in seeking the cause of their different behavior in nature within the established complex of factors. So too when unrelated species that are more or less identical in the aspects mentioned show similar responses in the experiments. —Nevertheless, we are still far from this and from any determination of the exact decisive factor, which indeed is unattainable so far as my experience goes. A further reason for my relatively modest expectations from these ex- periments is that they were conducted under primitive conditions with simple t (Pagination of the original German version, to which page citations in the text of this translation are referring to; suppl. scient. edit.). 50 33 instruments in the open air, in my apartment house, or in the laboratory at my school (only the experiments using a refrigerator were carried out in Statens Vaxtskyddsanstalt, Stockholm). For instance, the possibility of controlling the humidity of the air in the experimental room was extremely limited. Under such conditions it would be wrong to expect “accurate values.” The weakest point of all my experiments is undoubtedly the fact that only the responses of adults were studied. Only in two cases was adequate larval material available. In one case, Prerostichus anthracinus, the larvae were reared for genetic studies (Lindroth, 1946), and I did not venture to jeopardize them by other experiments. In the other case, Oodes gracilis (Lindroth, 1943a), the “response points” were actually determined (l.c., Diagram 6) and a tempera- ture gradient apparatus ( “Temperaturorgel”) experiment was also run (l.c., p. 136). This did not succeed because the larvae in the cold part of the apparatus immediately became torpid. The greatest differences between the larva and the imago are not to be ex- pected in the preferenda values, since the parents, at least during the breeding period, must of course seek the most suitable biotopes for their young ones, i.e. at least at this time they possess the same or very similar preferenda. But the resistance values, for instance the range of the activity temperature largely determined by the “lower response point” (see p. 104) would no doubt usually show greater sensitivity for the larva, which was revealed by experiments with the larvae of Oodes gracilis (see above). The species used in such experiments, with few exceptions (Amara equestris, Calathus and Cymindis), were imago hibernators. They experience the critical periods, i.e. spring and autumn (at any rate with regard to temperature) in the adult stage, and the responses of the adults may therefore actually be decisive. As far as the humidity of the air and soil is concerned, the summer is probably more fateful and hence also affects the larvae. The results of the experiments with such factors must therefore be treated with greatest caution. It may be justly objected that my experiments were all conducted with too little material. But one must consider the many very diverse questions asked by these experiments and therefore greater concentration, otherwise very useful, became impossible. Superficiality was the price for tackling so large a task. Also, some of the species used are so uncommon that more material could not be efficiently gathered. Finally, it should be mentioned that it is best to use about equal numbers of males and females in experiments. In many cases (which it is unnecessary to justify at this point) I kept the sexes separate and found no constant differ- ences. But it is conceivable that there could be differences. The most important sources of error intrinsic to different designs of ex- periment are assessed below with the description of each apparatus. Here only some series of experiments will be described, those of prime significance for any judgment of all other aspects. 51 34 1. A fundamental question is whether the different behavior of the indi- viduals of a species of animals in the experiments (especially the preferenda experiments) is due to individually different (hereditary or environmentally de- termined) characteristics or is only fortuitous. It would be especialiy important to decide whether that difference is between markedly stenotopic (especially stenothermic) and distinctly eurytopic (especially eurythermic) species, which is manifested in the experiments by a greater dispersion of the eurytopic species in the apparatus (particularly in the temperature gradient apparatus). Is this difference to be understood as meaning that a eurytopic species is more het- erogeneous, i.e. consists of a larger number of physiologically different “bio- types,” or that in each individual there is a different sensitivity toward the factor studied? Pterostichus nigrita turned out to be a suitable experimental animal. Com- pared with P. anthracinus, it shows not only a lower temperature preferendum (12.4°C as against 20.2°C; Experiments 27, 26a, p. 72) but also a considerably greater dispersion in the gradient apparatus (Diagram 1). As a geographically and ecologically ubiquist, the species also shows a pronounced eurytopy. Each of the 15 specimens of P. nigrita was differently marked with zinc white and their sequence (but not the exact temperature preferendum of every animal) in the temperature gradient apparatus was noted in 10 successive experiments. The result (Diagram 2) shows that the relative placement of every individual was as good as random. The “coldest” animal (d), in all 10 experiments, of course stayed in the colder half, but settled down at the low- est temperature only three times. The “warmest” animal (1) on six occasions remained at the highest temperature, but once even entered the colder half. The medium preferendum as well as the dispersion* figure (mean deviation of all insects from the medium preferendum) would not significantly change even if both these “extreme” animals were excludedt. The following statement therefore seems justified: The eurythermic character of Pterostichus nigrita is not (or is only slightly) due to physiological heterogeneity of the populations, but to the insensitivity of the individuals. Although further experiments with other species would be desirable I am inclined to assume that the same principle holds well for all eurythermic (and generally eurytopic) species. It does not follow that such species do not form physiological races, or that populations in distant parts of the total area of a widely distributed species would nec- essarily be physiologically alike. But the eurythermic character of a species that is geographically and ecologically ubiquitous must be primarily due to its less specific requirements for life. Conversely the stenotopic species have sharply specific requirements. Or they are stenotopic only at the periphery of *] was prevented from investigating the two “extreme” specimens of Pterostichus nigrita more closely in comparative temperature experiments because one of the insects perished. (Contradictory to the ecological term “dispersal”. cf. p. 203; suppl. scient. edit.) 51 35 their area, close to the minimum of a particular factor, and may then occur elsewhere as more or less eurytopic species. The following case may be cited as an example of individual behavior in the temperature gradient apparatus (Experiment No. 17b, p. 72); 17 specimens of Harpalus punctatulus were tested after 4 months in captivity. Five specimens were collected positively at the cold end (about 12.5°C), where they became torpid, whereas the remaining individuals did not settle down at temperatures below 17.8°C. The five “cold” individuals which were thereafter subjected to a new temperature gradient apparatus experiment (Experiment 23b, p. 72), again stayed (despite the use of H,SO, at the cold end) at the lower temperatures (11.9 to 15.6°C) than the other 12 specimens. —Individually different response in the temperature gradient apparatus in the case of Calathus erratus and Harpalus pubescens has also been observed by Agrell (1947). It is usually impossible to decide whether such individual differences in response are due to a fortuitous physiological condition (for instance, the age of the insect) or whether they are genetically determined. 2. The second major question therefore concerns the constancy of the responses of a particular species of animal. How much do the responses of the individual observed during the experiment depend on environmental factors and how much on the internal physiological condition (such as age)? I have already touched on this question (Lindroth, 1943a, pp. 136-137) with regard to Oodes gracilis and O. helopioides. In the temperature gradi- ent apparatus (Diagram 3), the former showed a stable preferendum, which always stayed at +20°C, irrespective of the environmental factors, whereas O. helopioides exhibited a variable preferendum. This “unmistakably varied according to the initial temperature on the day of the experiment” (or even- tually increased during the summer, independent of small variations in the day temperature). Similar results were obtained by Herter (1923, p. 284) in experiments with Formica rufa L., and by Bodenheimer and Schenkin (1928, | | I I een Diagram 1. Distribution of insects in the temperature gradient apparatus (“Temperaturorgel”). A—Pterostichus anthracinus (Experiment 26a, p. 72; 17 specimens ); B—P. nigrita (Experiment 27, p. 72; 19 specimens). 54 55 37 pp. 3, 10) with storage pests. In Calathus erratus and Harpalus pubescens, Agrell (1947) found a distinct decline in the preferendum after four days’ exposure of the test animals to a low temperature (+10°C). He got the same result af- ter exposure to a higher temperature (+30°C). In nature the latter response, with the exception of markedly cold-seeking species, has an insignificant role, since high enough temperatures rarely occur and usually can be avoided by the animals through active movements. A prolonged period of thirst (Heerdt, 1946, p. 28) or hunger (Agrell, 1947) also causes a decline in the preferendum. The specimens of Oodes available for observations was very low in number. However, experiments with other species have shown that this division of species into those with a “stable” and those with a more or less “variable” temperature preferendum (which evidently show no sharply distinct types) is justified. Pterostichus anthracinus provided highly suitable test material, of which both, freshly collected animals (Up! Djursholm) and animals of the next gen- eration, bred from these in the laboratory, were tested in the temperature gradient apparatus. All animals therefore belonged to the same population, but the parents, which had hibernated as adults, had been subjected to all the changes of weather in nature. However, the animals of the following generation had passed their lives at constant room temperature (about 20°C) from the egg on (May through August). The temperature gradient apparatus experiments (Diagram 4) showed a medium preferendum on the part of the parents of 20.2°C, and on the part of the new generation of 20.5°C. The dispersion of field animals was of course somewhat greater. The insignificant difference is a *fe I De ee Ts Tl 20.2 u, 1 RMU OW m © cr ll | | Il II PS | | 20.8” a ofan. | kl let le aA ee | | | | 14.9° II b ?%/e Mele Il | 15,3 ee vane Pe etal anal | Ps ies 54 Diagram 3. Oodes gracilis (I) and O. helopioides (II). Distribution in the temperature gradient apparatus on three different dates (a, b, c). After Lindroth, 1943a. 38 within the limits of experimental error or chance, all the more as the parents numbered only 17 specimens. It can therefore be stated that Pterostichus an- thracinus is distinguished by an unusually stable temperature preferendum. The highly variable environmental factors have exercised no demonstrable influ- ence in this case, nor the fact that in one case hibernated individuals more than one-half year old were taken and in the other less than two-month-old adults. Harpalus serripes (Diagram 5) also shows a remarkably stable temperature preferendum. Freshly collected insects showed (during early May, 1945) a me- dian preferendum of 27.89°C in the temperature gradient apparatus and a dis- persion (mean deviation) of only 1.95°. During late August, after four months of captivity indoors, the values obtained from the same animals were almost 56 exactly the same: 27.93°C and 2.13°C. Only in the following year, by which time the 12 survivors of the original 23 individuals in every respect looked weak and infirm with age*, did their responses decline. The medium preferendum then was 25.9°C and, significantly, the dispersion had increased to 4.8°. l Q Ü t 1 | ! t i Ü | ! | l ! | ! I I Mean 110% 15 20 25 20.17° | N Se B ee TE TEE TUTTO | 20.53" 55 Diagram 4. Pterostichus anthracinus. Temperature gradient apparatus. A— Field specimens (P generation); B—Fı generation bred from them. Experiment 20a DCH 2: IR el ie Oy Pato ata fee ie Car eon ean in (eae 4 ! Mean 20° 25 30° A | | MT BENENNEN 27.89" B I Il ee 27.93° el | | ot elas | 25.90° 55 Diagram 5. Harpalus serripes. Temperature gradient apparatus. A—Freshly collected field material; B—After 4 months in captivity; C—After 14 months in captivity. Experiment 23, p. 72. * However, the last specimen of H. serripes lived until January, 1948 thus became more than three years old. 39 Brachynus crepitans (Diagram 9; see also p. 60) behaved similarly, with its medium temperature preferendum of 25.87°C after 42 months in captivity. In the case of freshly collected material the value was only slightly higher at 26.65°C. The dispersion in the former case was likewise only slightly greater, 3.2° as against 2.5°. Low stability was shown by Harpalus punctatulus (Diagram 6), which was collected simultaneously with H. serripes and was similarly treated. During spring the medium temperature preferendum was 27.08°C, with a dispersion of only 1.65°. After four months of captivity, however, it was 21.06°C and the dispersion had risen to 4.74°. It is probably nat primarily the decrease of the temperature preferendum but the increased moisture requirement (possibly due to abnormal conditions in captivity; cf. below), which automatically must cause this effect. In nature a corresponding variation with the temperature is not demonstrable in this species. On June 25, freshly collected specimens showed a medium preferendum of 28.83°C and a dispersion of 1.53°. Normally the labilityt of the preferendum, when significant, might be considered as a physiological adaptation to the environmental temperature, which is expressed by increasing values during summer. In addition to Oodes helopioides and Harpalus punctatulus, which were studied, Harpalus melleti and H. rupicola appear to behave similarly. At the beginning of May the species showed (Diagram 7) a medium preferendum respectively of 23.04° and 18.07°C (dispersion of 3.07° and 2.06°, respectively). But in late June (fresh material) it was respectively 26.98° and 23.81°C (dispersion 1.47° and 2.33°, respectively). Especially in eurytopic and widely distributed species a labile temperature preferendum might be normal. The question calls for a thorough investigation based on more extensive material, and the explanation should be sought in purely physiological terms. From the results so far obtained it appears at any rate that populations that respond differently in the experiments must not Ee ity in Me ost ON aie Wminta nctay einem: eae Mean 115” 20 25 30 A | ee ||| 27.08" 218 1 | ata coe laa Late cee 21.06° Cc III EE MMT EE 28-83° Diagram 6. Harpalus punctatulus. Temperature gradient apparatus. A—Freshly collected field material (early May); B—After 4 months in captivity; C—Freshly collected field material (last June). Experiment 17, p. 71. t(= ecological term: suppl. scient. edit.). 5 © 59 57 40 be just declared as constitutionally different, because if we are dealing with “Jabile” species they may be in different stages of adaptation. It might also be possible to determine this adaptability of some animals with other preferenda experiments (but apparently to a lesser extent with re- sistance experiments). The behavior of Harpalus punctatulus in the “humidity gradient apparatus” (“Feuchtorgel”) might be cited here as an example (Dia- gram 8). Twenty-one specimens were collected in late April, 1945. After three weeks in Captivity they were tested in the “humidity gradient apparatus” (Ex- periment 72a, p. 79). The attraction toward the dry end of the humidity gra- dient apparatus was as strong as in H. serripes (Diagram 22, p. 134). The following year, freshly col'ected material (75 specimens) were tested in the same way in late June (Experiment 72b). The animals distributed themselves fairly uniformly in the apparatus, whereas H. serripes (150 specimens) always showed (Diagram 21) a clear attraction to the dry boxes. If the boxes are num- bered 1 to 7 from the dry to the moist part of the apparatus we can calculate the “mean box” in each test. This rose from 2.8 to 4.2 in A. punctatulus but from 2.8 to 3.4 in H. serripes. The corresponding experiments in the “universal gradient apparatus” (“Universalorgel”) (Experiment 112b; Diagram 29) were even more clear. The mean box place (10 boxes) increased in the same specimens, which was tested in June (after 2 months’ captivity) and in August (after 33 months), in A. punctatulus from 3.65 to 6.5, but in H. serripes (Experiment 116b, Diagram 33) from 3.85 to 4.3. Assuming that environmental factors (conditions in the culture container) could have influenced E. punctatulus, because the animals were, for instance, | | | I l 1 I | l I I I | | l l l | Mean IE! ul 1 ea NEN] | 23.04° OR SIE LIEDER NL I. 288° bel alba II 18.07° ID 8 ll alle klasse 23.81° Diagram 7. Harpalus melleti (1) and A. rupicola (II). Temperature gradi- ent apparatus, using freshly collected field material. a—Early May; b—Late June. Experiments 15, 21, p. 71. 41 % 50 Moist Dry ie) \ 40— \ Oe= “~=O— — —Oc 6.0 Box No 7, 6 5 4 3 2 1 Diagram 8. Harpalus punctatulus. Distribution according to varying humi- dity in the substratum gradient apparatus (“Substratorgel”). a—May, after three weeks in captivity; b—June, after one week in captivity; c—July, after two weeks in captivity and two days’ confinement in moisture-saturated air. “Mean box” is noted on right. Experiment 72, p. 79, Diagram 16. kept too dry and were thirsty, I kept them (50 specimens) for two days with moisture-saturated air and again tested them in the humidity gradient apparatus (Experiment 72c). The result was surprising. The distribution was exactly the reverse from that obtained in spring, i.e. a strong clustering in the moist end. Evidently the water balance of the animals had greatly altered during their stay in damp air. The lability of the moisture preferendum of Harpalus punctatulus is evident from these experiments. It can be altered by a change in the environmental factors and besides, an increase in the humidity requirement evidently takes place during summer. Thereby (as shown on p. 56 and on p. 67) the tempera- ture preferendum is also affected (lowered), but conversely the environmental temperature has certainly a considerable influence on the water balance of the animal. Specific studies should be able to reveal which factor is primary. On the other hand, according to Bodenheimer (1931, p. 741), in Calathus fuscipes the temperature preferendum is independent of changes in the humidity of the air. From the practical experimental viewpoint the above examples show that 60 61 42 in some species absolute response values cannot be achieved, and that the species to be compared must, so far as possible, be represented by equal material (of the same age, similarly treated, studied at the same time of year, etc.). 3. In nature many factors simultaneously influence individual animals and the species. It is usually extremely difficult to decide which of the factors (throughout the area or in some part of the area) is the most important. An attempt to approach the problem more closely by studying the influence of one or more factors experimentally at the same time is represented by the “universal gradient apparatus” described below. But a special case may be discussed here, whereby the question is ap- proached as it were through the back door, that is the case of Brachynus crepitans and Agonum dorsale. These two species are such obligatory companions that Brachynus actually never occurs without the Agonum species (on the other hand, Agonum can live without Brachynus, as for instance in Ska). The distribution maps, especially for eastern Sweden, are therefore almost identical. This relationship led to the belief that Brachynus, whose development is still unknown, was associated with Agonum as a parasite. As mentioned elsewhere (p. 548), this assumption is erroneous. Neither of the two species is associated with the other through food biology. Brachynus crepitans and Agonum dorsale thus represent a rare instance of two species that show exactly the same ecology and almost the same distribution on account of their identical life requirements. It is therefore natural to test both the species together in all the usual preferenda and resistance experiments, in order to determine the factors to which they respond most similarly. It may then be concluded that these factors are the decisive ones. The following experiments were conducted: a. Temperature preferendum (Diagram 9; Experiments 1 a-d, 6 a-e). In all concurrent experiments Brachynus thus shows a higher temperature preferendum than Agonum dorsale. Because of this the dispersion in the lat- ter species is somewhat less. The mean temperature preferendum for all the experiments is*: Brachynus (110 specimens) 25.64°C , dispersion 2.67°C. Agonum dorsale (70 specimens) 22.52°C, dispersion 2.15°C. b. The response points of temperature. The lower limit of activity (Experiments 124, 128, 129), i.e. the temper- ature at which the animals became motionless after supercooling, was deter- mined concurrently on a total of 42 specimens of each species. The following figures were obtained: *The total mean values were calculated without taking into consideration the number of individuals used in each experiment. 43 nn ale site alia ae lh 20° 25 30 Sr a **/s Hy I TAM ante tat ome ea) Hl I b**/e enon ed Uk EO ACTH NC il inl un | Ic */s IA ESS ath lis staged | d ®ı Il el al lulellss | ent: bel | I) a */s | ish ocan kT MII ht | be ELITE alg Il eth, U taal N ae alles det | | Diagram 9. Brachynus crepitans (1) and Agonum dorsale (II). Distribution in the temperature gradient apparatus on different dates. Data of each ex- periment noted on left. cf. Diagram 11, Experiments 6 and 1, pp. 69-70. Brachynus: 4.1, 4.4, 5.9, 6.4, 6.4, 6.8, 6.9, 7.0, 7.6, 7.6, 7.7, 7.7, 7.7, 7.7, 7.7, 7.8, 7.9, 8.0, 8.1, 8.4, 8.5, 8.5, 8.6, 8.6, 8.7, 8.7, 8.8, 8.8, 8.8, 8.9, 8.9, 8.9, 9.1, 9.5, 9.6, 9.8, 10.5, 10.5, 10.7, 10.8, 11.0, 12.6. Asonum; dorsale: 6.5, 6.6, 6.8, 7.4, 7.3, 7:5, 7.6, 1.7, 7.7, 1.1,.7.8, 1.8.8.0, 8.0, 8.1, 8.1, 8.1, 8.2, 8.2, 8.3, 8.3, 8.4, 8.6, 8.6, 8.6, 8.7, 8.7, 8.7, 8.7, 8.8, 8.8, 89489,9.0, 9.1, 9.5, 9:5,.9.6,.9.8, 9.8, 10.0, 12.2. The mean values, 8.35°C (Brachynus) and 8.45°C (Agonum dorsale) respectively, are so close that they may be considered identical. However, the upper limit of activity (Experiment 137) was very different. The temperature at which the first sign of paralysis appeared, and that at which total paralysis resulted, were measured individually on 20 specimens each. Since the figures are very close together, enumeration is unnecessary. First sign Total paralysis of paralysis Brachynus 46.2-49.6°C 48.8-54.2°C Agonum dorsale 42.4-47.1°C 42.8-48.8°C The mean values for Brachynus are 48.7°C and 51.0°C and for Agonum dorsale 45.5°C and 47.1°C. Thus in both cases there is a difference of 3-4°C. 62 44 c. Humidity preferendum (Diagram 10; Experiments 82, 83, p. 80). This was determined in the circular universal gradient apparatus, since the usual substratum gradient apparatus is unsuitable for these non-digging insects. In recording three observations every second hour, 120 observations were obtained in each case. The calculated mean box place for Brachynus is 3.1, for Agonum 3.0, thus showing an extremely close relationship. d. Drought resistance (Experiment 141). The length of time for which the animals were able to live without water was determined for 20 specimens each of the two species. The mean maximum duration of life in Brachynus was found to be 137 hr 45 min, and in Agonum dorsale 123 hr 48 min—a difference of more than half a day. In these fairly closely correlated experiments with Brachynus crepitans and Agonum dorsale, which have the same ecological requirements in nature, the two species exhibited such close similarity with regard to temperature for the lower limit of activity and their moisture preferendum that these values may be considered identical. Hence the conclusion that among the factors measured (and among all those measurable?) these two are decisive for the species con- cerned. Evidently this conclusion must not be applied too freely to other animals. For instance, in the hygrophilous species of Oodes the preferenda appear to give a more correct indication of the temperature factors decisive for life than of the “lower response point” (p. 457; also see Lindroth, 1943a, p. 137). However, it seems at least justified to suppose that the factors decisive for Brachynus and Agonum dorsale can also be especially important for other more Moist Dry fi 50 40 30 20 10 rfl Tt +o EEE ew el Box No. 10 9 8 2 6 5 4 3 2 1 Diagram 10. Brachynus crepitans (black) and Agonum dorsale (white). Distri- bution during a common experiment in the “humidity gradient apparatus.” Experiments 82, 83, p. 80. 45 63 or less xerophilous carabids. Accordingly they have been considered important in the treatment of the “limestone species,” among others. Finally, it should be mentioned that I have not analyzed the data obtained from the experiments by the latest method of statistical analysis (as in Bonnier and Tedin, 1940). The mean values have only been calculated arithmetically and, besides, the dispersion index has also been given, i.e. the mean deviation of all the experimental animals from the calculated mean value. As an example how the mean values turn out by one or the other method, I have also analyzed the above temperature gradient experiments with Brachynus and Agonum dorsale with the statistics of variation (after Herter, 1924, pp. 234-236). The observations were divided into frequency classes of 2°C each (Diagram 11); these calculations are presented in Table 1. The mean values differ only slightly from the arithmetic mean. Unfortunately I have had absolutely no schooling in mathematics. Never- theless I would venture to cast doubt on the validity of applying purely mathematical methods to an analysis of the temperature gradient apparatus experiments. The distribution of insects in the gradient apparatus (and hence the values obtained by interpolation) does not appear to me purely accidental. It can be easily observed (as discussed on p. 69) how the animals show a choking effect as soon as they come from their more or less broad zone of preferendum towards the warm end of the apparatus, whereupon they hastily turn back. On one side of the preferendum zone there is, so to speak, a barrier, which has no counterpart on the other side (i.e. towards the cold end). The deviations from the mean value (whether obtained arithmetically or statistically) are therefore not determined fortuitously. 35% 30 25 20 15 10 5 be 16° 18 20 22 24 26 28 30 32 34° Diagram 11. Brachynus crepitans (continuous lines) and Agonum dorsale (broken lines). Frequency curves of temperature preferenda according to Diagram 9. a—Arithmetic mean value; b—Statistical mean value, and calcu- lated amplitudes of variations of latter (horizontal lines). 65 46 Table 1. Statistically calculated figures obtained from experiments on Brachynus crepi- tans and Agonum dorsale (Diagrams 9, 11) with the temperature gradient apparatus M m M+3m o n +°C De a2 1 (G; 2€ Brachynus crepitans 25.836 + 0.345 24.802- + 3.620 110 26.870 Agonum dorsale 22.286 + 0.341 21.263- + 2.854 70 23.309 Note: M—Mean value; m— Mean error thereof; o—Standard deviation; n—Number of observa- tions. Probably because of this the calculated mean preferenda are consistently too low. But it is more important that different species behave differently toward the “heat barriers.” In the case of some species, such as Brachynus, the response starts after contact with the highest temperature, such that they usually proceed toward the cold end of the apparatus without resting. In the case of others, such as Agonum dorsale, the same stimulus usually causes only lesser movements. This difference is clearly evident when the curves of the two species are superimposed (Diagram 11). I even believe that the strange growth of the “27° frequency class” of the Agonum curve resulted from the aggregation of individuals that were repulsed by the heat barrier. Another manifestation, which is evident especially in most of the experi- ments with the universal gradient apparatus (Diagram 25 ff., p. 141 ff.), is that the more or less distinct preferendum zone (hence the “main maximum” of the diagram) is demarcated by a distinct minima on one or both sides, after which the curve rises steeply. The probable explanation is that the preferendum zone in the immediate vicinity of the insects exercises especially strong “absorption” on them—further evidence that the animals are not distributed accidentally outside their maximum of frequency. Also the striking association of various species—for instance, Cymindis humeralis, in addition to Brachynus and Agonum dorsale—which often cluster together in the gradient apparatus, must interfere with the fortuitous distri- bution. It is always risky to treat biological manifestations purely mathematical as numbers. But these are only symbols whose true content is all too easily forgotten during such studies. After these remarks on the principles we will now proceed to the report on the experiments conducted. The information given below with each experiment may to some extent appear trivial and inconsequential. But I am of the view that the primary material presented should always be so complete that it can be objectively tested by other researchers. In this way it is possible to discover 66 47 sources of error overlooked by the author, or eventually to utilize the material for quite different purposes. It is advisable to describe these experiments neither in temporal sequence according to my diary nor in the order in which they are discussed in the text of this book, but to arrange them alphabetically by species for each group of experiments. With each experiment an indication is given of the place where it is dis- cussed in the text. Of course, the text also contains pointers in the opposite direction, to the transcript of experiments. The number of “specimens” cited for the experiments indicates the cases observed (even if the same individual was utilized more than once). With few exceptions only those experiments are described whose results are utilized in this book. The results obtained are given only in the running text, and not in the transcript of experiments. A. PREFERENDA EXPERIMENTS As the name indicates, in these experiments the animal is given freedom of choice among various factors or among various gradations of the same factor. At the outset attention must be drawn to the fact that the preferendum must not be automatically equated with the optimum. For instance, it is quite possible that the preferred temperature is not the one at which the life func- tions of the animal, such as reproduction and development, are optimal. On the other hand a large difference between these two “points” is not to be ex- pected, since it would be inappropriate (nonadaptive)—with the exception of the theoretically conceivable case where the animal never came across its pre- ferendum in nature. If a distinct difference between the preferendum and the optimum is to be experimentally detected, it should be established whether this has not resulted from an abnormal situation in the experiment or under the influence of some factor that was overlooked, before the results are attributed to nature. The noteworthy exceptions, such as the strong attractive—usually disturbing—effect of light on nocturnal flying insects, are to be considered as “unnatural situations” (Mast, 1911, pp. 227, 237 ff.). I. Temperature Gradient Apparatus (“Temperaturorgel”) This device, which has become the most important aid for experimental ecol- ogy, was developed by Herter (1924, pp. 225 ff.). It consists of an elongated glass box with a metal floor (copper in my device). The cover has any de- sired number of holes through each of which a thermometer is inserted so 67 68 67 48 that the bulb touches the floor. I have used this simple original form of the temperature gradient apparatus (Fig. 3). Many investigators have devised modifications of this apparatus in one form or another, including Herter himself (1934; 1939, p. 744). He fitted the bulb of the thermometer into the metal base so that only the temperature of the floor is measured and not that of the air (or of both combined). I used the original simple type because I was interested in comparative, not absolute values. It seemed to me indifferent whether I measured the temper- ature of the floor or of the air. It is a matter for discussion whether animals always respond more to the temperature of the soil than to that of the air (compare, for instance, Thomsen and Thomsen, 1937, p. 346). If the under- neath of the bulb is spherical, as in my experiments, so that it touches the floor only at a point, the readings actually show the temperature of the air next to the floor (not of the floor itself). A particle of sawdust interpolated between the thermometer and the floor, slightly separating the two, caused not the slightest change in the temperature. The metal floor was sticking out of the box and bent down at both ends into glass beakers, of which one was full of water and warmed by an adjustable spirit lamp and the other was filled with running cold water. On the hottest summer days ice had to be used for cooling. The difference of temperature over the entire apparatus was 20-30°C. The duration of exposure was about 2 hours (only in Experiment 28 was it considerably less). The location of each insect was not recorded mechanically after a stipulated period but only after it had come to rest. Hence it often happened that the individuals were not all recorded at the same time (but in any case they would be within half an hour). The main source of error in my temperature gradient apparatus experiments was undoubtedly the humidity of the air (cf. also Bodenheimer, 1931, and Palmén and Suomalainen, 1945, pp. 38 ff.). It is impossible to maintain the humidity uniform throughout the apparatus without specialized Fig. 3. Temperature gradient apparatus. Floor measurement 7 x 70 cm. 69 49 appliances. At the warm end the relative humidity goes down. If one places a moist swab of cotton at the warm end and a water absorbent (for instance, H,SO,) at the cold end the preferendum of the animals markedly increases. Their choice of temperature is therefore not a function of their sensitivity to temperature alone. The largest variations are observed in pronouncedly hygrophilous species. In these species, it is sometimes impossible to prevent clustering at that cold end without making the above arrangement for uniform humidity. However, even if it is possible to maintain the relative humidity through- out the apparatus at 100%, the use of H,SO, cannot prevent the condensa- tion of very fine droplets of water on the wall at the cold end (at least when the temperature there drops below +10°C). This can attract the animals to drink. Besides, a moist floor offers mechanical advantages, greatly increasing the adhesive capacity of the tarsi. At least I could not otherwise explain the great difference between the two experiments with Badister unipustulatus (Dia- gram 12; Experiment 5, p. 70). In both cases the air was moisture-saturated throughout the apparatus whereas its floor was first covered with dry sawdust and afterwards with wet sawdust. These serious sources of error with regard to the humidity enjoin extreme caution in judging that the absolute temperature preferendum of a particular species has been determined. This is less consequential for my modest goal of comparative values. But care must be taken to handle the species compared in an identical way in the gradient apparatus; the best of all is to carry out a common experiment with both species. The responses of the animals when they find themselves outside their “preferred zone” in the gradient apparatus differ according to which side of it they are. The warm part causes an immediate strong escape response, and the animal usually hurries back to the cold end without resting. On the other hand, a long stay at the cold end leads to fairly complete torpid, and if the animal is able to reach its preferred zone at all it does so very slowly. The result is a clustering at the cold end and thus a decline in the average preferendum, | | ! 1 | | | j | | | I ) ! \ | Mean 10° 15 20 25% A We tele | Kal Tee ıl kl al 15.51 Seal le] Ir AU Nitta ER I Sees Diagram 12. Badister unipustulatus. Temperature gradient apparatus. A— with dry; B—with wet sawdust on floor. Moisture-saturated air in both cases. Experiment 5, p. 70. 70 50 especially in more or less hygrophilous species (cf. p. 64 above). If the group of torpid specimens at the cold end of the apparatus was separated by a distinct gap from the main group of animals in the preferred zone, and looked “unnatural,” I changed the approach. I aimed a small lamp at the animals until they became active and sought a new place within the preferred zone. I am of course aware that this eventually raised even the “true” preferendum value. However, the treatment was always identical for both (or all) the species compared, except in the cases recorded below in the transcript of experiments. In more or less xerophilous species the preferred zone is clear and well-defined without any interference (even without wet cotton and H,SO,). Uniform illumination of the temperature gradient apparatus is necessary in the case of species that shun the light, or the animals will usually cluster in the dark corners of the apparatus. All experiments were therefore carried out in darkness or in uniform dim artificial light. For further literature concerning the temperature gradient apparatus expe- riments, reference may be made to Herter (1939) and Heerdt (1946). Concerning the species of Harpalus, see Diagram 19 (p. 129). Experiment 1. Agonum dorsale. Diagram 9, 11, p. 60. a) May 2, 1945. Gtl. Horsne, April 29, 1945. 20 specimens. Room temper- ature 17.5°C. Thermometer readings: 8.0, 11.5, 13.9, 16.6, 19.2, 22.3, 26.9°C. b) August 24, 1946. Ogl Mogata, August 17-18, 1946. 20 specimens. Room temperature 22.0-22.2°C. Thermometer readings: 12.8, 17.0, 20.2, 23.1, 26.6, 31.6, 3913. c) July 5, 1947. Old Halltorp area. June 12-23, 1947. 15 specimens. (to- gether with Brachynus). Room temperature about 28°C. Moist cotton at warm end. Lamp at cold end. Thermometer readings: 16.8, 19.5, 22.6, 25.1, 27.8, 31.2, S802E- d) Same as (c). Room temperature about 27°C. Thermometer readings: 15°45 19:05 22.5) 25:2, 28:5, 32:694 20°C Experiment 2. Agonum lugens. July 13, 1947. Upl Sigtuna, July 13, 1947. 20 specimens. Moist cotton at warm end. Room temperature 22°C. Thermometer readings: 10.8, 14.4, 17.2, 20.1, 23.3, 27.9, 35.5°C. Diagram 48, p. 457. Experiment 3. Agonum viduum. June 25, 1946. Upl Angby, Rockstasjon Lake, May 26, 1946. 30 specimens. Moist cotton at warm end. Room temper- ature 22-23°C. Thermometer readings: 9.1, 13.0, 16.5, 19.4, 22.6, 26.8, 33.5°C. Diagram 48, p. 457. Experiment 4. Badister dilatatus. June 20, 1946. Öld Halltorp, June 11-12, 1946. 20 specimens (together with B. unipustulatus). Moist cotton at warm end, very thin layer of wet sawdust on floor. Room temperature 20.4-20.5°C. Thermometer readings: 11.0, 14.7, 17.8, 20.3, 23.6, 28.5, 35.0°C. Diagram 48, posi. Experiment 5. Badister unipustulatus. a) June 20, 1946. Old Halltorp. June 11-12, 1946. 20 specimens. Moist 71 31 cotton at warm end, few dry sawdust particles on floor. Room temperature about 20.5°C. Thermometer readings: 12.0, 15.4, 18.5, 21.0, 24.8, 30.4, 39.0°C. Diagram 12, p. 68. b) Same as Experiment 4. Same 20 specimens. Diagrams 12, 48, p. 457. Experiment 6. Brachynus crepitans. Diagrams 9, 11, p. 60. a) August 25, 1945. Upl Lovon, April 2, 1945 (i.e. after 43 months in captivity). 30 specimens. Room temperature 22-23°C. Thermometer readings: 10:2215.2,,193, 226,726.95 31.4, 38:6°C. b) August 26, 1945. Upl Lovon, August 23, 1945. 30 specimens. Room temperature 22.6-22.7°C. Thermometer readings: 12.2, 16.2, 19.6, 22.6, 26.5, 31.238.923. c) August 24, 1946. Upl Lovon, August 23, 1946. 20 specimens. Room temperature 22°C. Thermometer readings: 13.9, 17.2, 20.2, 22.7, 26.1, 31.1, S7.92E: d) Same as 1c. 15 specimens (together with Agonum dorsale). e) Same as 1d. 15 specimens. Experiment 7. Bradycellus collaris, macropterous form and brachypterous form from same population. Jtl Revsund, collected on August 31, 1947. Lamp at cold end. Diagram 46, p. 359. a) September 5, 1947. Macropterous form. 30 specimens. Room temper- ature about 23°C. Thermometer readings: 10.1, 14.3, 17.4, 20.5, 24.2, 28.9, Spas b) Same as (a). Brachypterous form. 30 specimens. Thermometer readings: 10.3914.5317.5,205724:0728:8, 36.5°C. c) September 6, 1947. Macropterous form [same animals as in (a)]. 30 specimens. Room temperature 24°C. Thermometer readings: 12.0, 15.6, 18.5, 21.4, 24.8, 29.8, 37.5°C. d) Same as (c). Brachypterous form [same animals as in (b)]. 30 specimens. Thermometer readings: 11.5, 15.4, 18.7, 21.6, 25.5, 31.0, 39.5°C. Experiment 8. Cymindis angularis. June 25, 1946. Öld Greby, June 11-15, 1946. 14 specimens. Room temperature 23.0-23.2°C. Thermometer readings: 14.2, 17.9, 21.2, 24.6, 27.2, 32.0, 39.9°C. Moist cotton at warm end. Diagram 24, p. 138. Experiment 9. Cymindis humeralis. June 22, 1946. Old Greby, June 11-15, 1946. 20 specimens. Room temperature 22.8-23.1°C. Thermometer readings: 13.8, 16.9, 20.1, 22.6, 26.1, 30.6, 37.4°C. Diagram 24, p. 138. Experiment 10. Cymindis macularis. June 22, 1946. Old Stora-Rör, June 13, 1946. 20 specimens. Room temperature 22.9-23.6°C. Thermometer readings: 13.9, 17.4, 20.7, 23.3, 27.1, 31.9, 39.6°. Diagram 24, p. 138. Experiment 11. Harpalus aeneus. June 23, 1946. Öld Greby, June 11-15, 1946. 20 specimens. Room temperature 24°C. Thermometer readings: 13.9, 16.7, 20.4, 23:5, 27.5, 32.8°C (not read). Experiment 12. Harpalus anxius. June 21, 1946. Öld Stora-Ror, June 13, 1946. 20 specimens. Room temperature 21.5°C. Thermometer readings: 12.8, 52 15.9, 18.8, 21.6, 25.1, 30.2, 38.6°C. Experiment 13. Harpalus azureus. June 21, 1946. Öld Greby, June 11-15, 1946. 20 specimens (all brachypterous). Room temperature 21.5°C. Ther- mometer readings: 12.8, 16.0, 18.9, 21.6, 25.1, 29.8, 37.7°C. Experiment 14. Harpalus hirtipes. June 24, 1946. Öld Stora-Rör, June 13, 1946. 15 specimens. Room temperature 22.2-22.6°C. Thermometer readings: 13.6, 16.8, 20.1, 22.9, 26.9, 31.9, 40.2°C. Experiment 15. Harpalus melleti. a) May 5, 1945. Gtl Visby, April 28-30, 1945. 20 specimens. Room tem- perature 18°C. Thermometer readings: 9.0, 13.5, 16.3, 19.3, 23.3, 27.5, 36.1°C. Diagram 7, p. 57. b) June 25, 1946. Gtl Visby, June 17, 1946. 20 specimens. Room tempe- rature 23°C. Thermometer readings: 13.1, 16.9, 20.2, 23.2, 26.9, 31.8, 38.0°C. Experiment 16. Harpalus neglectus. June 24, 1946, Old Stora-Rör, June 13, 1946. 15 specimens. Room temperature 22.5°C. Thermometer readings: 14.0, 17.0, 20:2, 23.0, 27.0) 32.0r40°7.@ Experiment 17. Harpalus punctatulus. a) May 3, 1945. Gtl Visby, April 28, 1945. 19 specimens (in addition one specimen had lost one antenna and showed aberrant behavior). Room temperature 17.3°C. Thermometer readings: 10.5, 14.2, 16.9, 19.1, 22.3, 26.7, 32.1°C. Diagram 6, p. 56. b) August 26, 1945. Material same as in a) (i.e. after 4 months in captivity). 17 specimens (preference of 5 specimens laying torpid at the cold end was later determined in repeat Experiment 23b together with A. serripes, where it was again found to be lower than for the remaining individuals). Room temperature 22.6-22.7°C. Thermometer readings: 14.5, 17.8, 21.0, 23.9, 27.4, 31.5, 38.5°C. Diagram 6, p. 56. c) June 25, 1946. Gtl Visby, June 17, 1946. 20 specimens. Room temper- ature 23.0-23.2°C. Thermometer readings: 14.0, 17.4, 20.6, 23.6, 27.6, 32.5, 38.8°C. Experiment 18. Harpalus puncticeps. June 25, 1946. Old Halltorp, June 15, 1946. 13 specimens (together with H. rufitarsis). Room temperature 23.0723.3°C. Thermometer readines: 11.8, 15.9, 19.4, 222, 25.7, 299 sonic: Experiment 19. Harpalus rubripes. June 22, 1946. Öld Greby, June 11-15, 1946. 20 specimens. Room temperature 23.1-23.2°C. Thermometer readings: 13.9, 17.0, 20:2,,229, 229, 304. 37€ Experiment 20. Harpalus rufitarsis. a) Same as Experiment 18. Old Stora-Ror, June 13, 1946. 7 specimens. Diagram 19, p. 129. b) June 25, 1946. Same animals as above. Room temperature: 23.1-23.4°C. Thermometer readings: 12.0, 16.3, 20.0, 23.0, 27.2, 32.8, 40.3°C. Experiment 21. Harpalus rupicola. a) May 4, 1945. Gtl Visby, April 28-30, 1945. 20 specimens. Room temper- ‚ature: 17.19@.) Thermometer readings»8.8,12.4,15.0,18.0922.9127373138@: WZ; Diagram 7, p. 57. b) June 22, 1946. Gtl Visby, June 17, 1946. 20 specimens. Room tem- perature 22.5-22.9°C. Thermometer readings: 11.6, 15.1, 18.5, 21.3, 24.5, 28.7, 34.2°C. Experiment 22. Harpalus seladon. May 3, 1945. Gtl Visby, April 30, 1945. 20 specimens. Room temperature 17.9°C. Thermometer readings: 9.1, 12.6, 159 17.9, 21.3) 26.0, 32/4°C: Experiment 23. Harpalus serripes. a) May 2, 1945. Gtl Visby, April 29, 1945. 20 specimens. Room tempe- rature 17.9°C. Thermometer readings: 8.7, 11.9, 14.7, 17.6, 21.4, 26.3, 32.5°C. Diagram 5, p. 55. b) August 26, 1945. Same animals as above (i.e. after 4 months in cap- tivity; together with 5 specimens of H. punctatulus, see Experiment 17b). 20 specimens (H,SO, at cold end, which certainly had no influence on this ex- ‘tremely xerophilous animal). Room temperature 22.1-22.6°C. Thermometer readings: 12.4, 16.6, 20.1, 23.1, 27.3, 32.3, 40.2°C. c) June 28, 1946. Same animals as above (i.e. after 14 months in captivity). 12 specimens. Room temperature about 21.5°C. Thermometer readings: 13.0, 16.0, 19.1, 21.8, 25.4, 30.5, 36.5°C. Diagram 5, p. 55. Experiment 24. Harpalus smaragdinus. June 23, 1946. Old Stora-Rör, June 13, 1946. 20 specimens. Room temperature 24.1-24.4°C. Thermometer rea- dines15.0,18.1,21.7, 24.3, 28.1,:32.8, 395°C. Experiment 25. Harpalus tardus. June 22, 1946. Oland and Gotland, June 12-17, 1946. 20 specimens. Room temperature 22.7-22.8°C. Thermometer neaGines: mile, 15.3, 18.8,21.7, 25.3, 293, 36.9°C. Experiment 26. Pterostichus anthracinus. a) April 10, 1945. Upl Djursholm. Ekebysjon lake, April 10, 1945. 17 specimens (6 macropterous specimens, 11 brachypterous specimens). Room temperature 18°C. Thermometer readings: 10.1, 13.0, 15.2, 17.6, 20.1, 23.4, 28.9°C. Diagrams 1, 4, pp. 51, 55. b) September 1, 1945. F, generation obtained by crossing animals in Ex- periment 26a (emerged on July 4-17, 1945; reared in room from eggs). 15 macropterous, 15 brachypterous specimens, the latter marked with zinc white. Room temperature 21.2-22.0°C. Thermometer readings: 12.8, 16.8, 19.7, 22.4, 26.2, 31.1, 37.1°C. Diagrams 4, 43, pp. 55, 356. c) Same as b). September 2, 1945. 10 specimens each. Room temperature 21.2-21.5°C. Thermometer readings: 13.0, 17.1, 20.5, 23.3, 27.7, 33.6, 42.1°C. Diagrams 4, 43, pp. 55, 356. d) Same as b). September 16, 1945. 13 specimens each. Moist cotton at warm end. Room temperature about 19°C. Thermometer readings: 11.5, 14.7, 17.2, 19.8, 23.4, 28.6, 31.5°C. Diagram 43, p. 356. Experiment 27. Pterostichus nigrita. April 23, 1945. Upl Danderyd, Nora, April 22, 1945. 19 specimens. Room temperature 16.5°C. Thermometer read- 1n95:9.0, 10:6°.12.9, 15.7, 18.4, 22.1, 27.42C. Diagram: 1, p. 31. 74 54 Experiment 28. Pterostichus nigrita. May 9-10, 1945. Upl Danderyd, Nora, April 22, 1945. 15 specimens (same animals as in Experiment 27), all in- dividually marked with zinc white. Room temperature 16.8-19.1°C. In 10 experiments (3 on May 9, 7 on May 10), individual sequence of animals was recorded (but not temperature) as soon as they came to rest. Exposure be- tween + hr and more than 2 hrs. Diagram 2, p. 51. II. Substratum Gradient Apparatus (“Substratorgel’’) This name here designates the apparatus (Fig. 4) developed by Krogerus (1939, p- 1230) which is here briefly described. The design, which he later modified, is a zinc box (80 x 8 x 8 cm) which contains a series of small zinc boxes of 7.5° cm (hence about 400 cm?), standing next to each other. The upper two- thirds of the juxtaposed sides of the boxes are perforated with big holes, so that the insects can move freely not only along the surface of the box, usu- ally half-filled with substratum, but also in the substratum itself through the boxes. A fine-mesh metal wire gauze was used as cover. Time of exposure 1-2 days. The experiments were conducted partly as serial experiments, i.e. with a smooth gradation of one factor from one end of the gradient apparatus down to the other, and partly as alternating experiments, in which only two or three (quantitative or qualitative) different kinds of substratet were used, mostly in regularly alternating boxes. The main source of error in these experiments is the tendency of the animals to gather more densely in the boxes at the two ends, partly because further movement is impeded and partly because darker hiding places are eas- ily found there. Hence one must try to provide uniform illumination in the apparatus. This difficulty does not appear in “alternating experiments,” since each end of the apparatus has a box containing each of the two substrata. In “serial experiments” it can be compensated at least partly by eliminating the end boxes from the experiment and if necessary by filling these with a substra- tum having a strong negative influence (i.e. dry or wet sawdust respectively for hygrophilous or xerophilous species). A circular gradient apparatus would perhaps have been better, but here again the method used can be judged as sufficient for the desired comparative values (of two or more species). In order to study, at least in one case, the distribution of animals in the substratum gradient apparatus without the influence of experimental error, thin this case the apparatus is also called substrate (or substratum) choice apparatus; suppl. scient. edit.]. 75 55 I made a control experiment with Amara ingenua (Diagram 13, Experiment 107, p. 85) for which all the boxes were filled with the same kind of soil. The experiment disturbing clustering of individuals at the two ends of the apparatus was evident. It would now be possible to correct the figures obtained in different serial experiments (in the case of alternating experiments no control experiment is necessary) from the values given in Diagram 13. I faced the same problem with the “universal gradient apparatus” (p. 85) and found that in this case an empirical correction curve must be plotted for every species. However, apart from the huge amount of work required for this, every factor in an experiment clearly influencing the animals, might not indeed eliminate the above mentioned source of error, but might reduce it to such an extent that a correction of this kind would be too difficult. I have tried to avoid these difficulties—as far as possible—in realizing 0 Box No haa DMs Od EN] 8 9,10 Diagram 13. Amara ingenua. Distribution of 200 specimens in the sub- Stratum gradient apparatus without influence of any measurable factors. Continuous line—Empirical curve; broken line—Symmetrically leveled curve. Experiment 107, p. 85. 76 56 alternating experiments (so that the preceding serial experiments served the purpose to determine only the desired quantities of the factors to be exam- ined), or ignoring the end boxes in the experiment (with few exceptions). The exact procedure is given with the description of each experiment below. In studying some of the smallest species (Bembidion, Dyschirius) I had to use other arrangements since the animals were able to crawl into the interstices between the boxes of the substratum gradient apparatus. These experiments are described below in detail (Experiments 92 ff., 95 ff.). In one case (Experiments 82, 83) the circular universal gradient apparatus (p- 93) was used for a simple substratum experiment (humidity). a. Serial experiments with CaCO, Limestone-free, washed, fairly coarse siliceous sand was mixed with CaCO, (precipitated chalk) in the following series: Box 02 3 4 5 En 8a ug Sand, cm? 200 199.61 199.22 198.44 196.88 193.75 187.5 175 150 100 Limestone. cm — 03 078 156 313 625 1259 5 50 100 The two mixtures in boxes 2 and 3 were used only in Experiment 29. Exposure 2 days. For the species of Harpalus, see Tables 2-3 (pp. 121, 122). Experiment 29. Brachynus crepitans.* a) April 10, 1945. Upl Lovon, April 2, 1945, 50 specimens. To each box 25 cm? of spring water was added. b) Same as a). April 12, 1945. c) Same as a). April 14, 1945. Experiment 30. Brachynus crepitans.* a) Material and method of experiment as in Experiment 29, but to each box CO,-saturated water (15 cm?) was added. April 16 through 18, 1945. 50 specimens. b) Same as a). April 23, 1945. Experiment 31. Harpalus melleti. May 4, 1945. Gtl Visby, April 28-30, 1945. 24 specimens. Only 8 experimental boxes (Nos. 2 and 3 above excluded), the end boxes had dry sawdust. In each box 20 cm? distilled water was added. Experiment 32. Harpalus melleti. May 19, 1945. Material and method of experiment as in Experiment 31, but with addition of 10 cm’ CO,-saturated water in every box. Experiment 33. Harpalus punctatulus. May 10, 1945. Gtl Visby, April 28, 1945. 24 specimens. As in Experiment 31, but end boxes had wet humus. * Limestone was found to have no influence. WY 3 Experiment 34. Harpalus punctatulus. May 14, 1945. Material and method of experiment as in Experiment 33, but with 10 cm? CO,-saturated water added in every box. Experiment 35. Harpalus rupicola. May 2, 1945. Gtl Visby, April 28-30, 1945. 24 specimens. As in Experiment 31. Experiment 36. Harpalus rupicola. May 24, 1945. Material and method of experiment as in Experiment 35, but with 10 cm? CO,-saturated water added to each box. Experiment 37. Harpalus seladon. May 12, 1945. Gtl Visby, April 30, 1945. 24 specimens. As in Experiment 33. Experiment 38. Harpalus seladon. May 21, 1945. Material and method of experiment as in Experiment 37, but with 10 cm? CO,-saturated water added to each box. Experiment 39. Harpalus serripes. a) May 6, 1945. Gtl Visby, April 29, 1945. 23 specimens. As in Experiment Sil b) As in a), but with wet humus instead of sawdust in the end boxes. May 8, 1945. Experiment 40. Harpalus serripes. May 16, 1945. 21 specimens. Material and method of experiment as in Experiment 39b, but with 10 cm? CO,-saturated water added to each box. b. Alternating experiments at different pH levels Three substrata were used and their pH determined electrometrically. 1. Soil with humus from the locality of Harpalus seladon. Gtl Visby (April 30, 1945), pH 7.5. 2. Soil with humus from the locality of Pterostichus anthracinus, Upl Djur- sholm, Ekebysjon lake (April 10, 1945), pH 4.8. 3. Mixture of 1 and 2, hence with pH of about 6. Only 7 boxes were used and, with the exception of Experiment 43b, ar- ranged as follows: Sawdust; pH 6; pH 7.5; pH 6; pH 4.8; pH 6; sawdust. The moisture of the “pH boxes” was regulated with distilled water. Expo- sure 2 days. Concerning the species of Harpalus, see Table 4 (p. 124). Experiment 41. Harpalus melleti. May 14, 1945. Gtl Visby, April 28-30, 1945. 25 specimens. End boxes with wet sawdust. Experiment 42. Harpalus punctatulus. a) May 6, 1945. Gtl Visby, April 28, 1945. 25 specimens. End boxes with dry sawdust. b) Same as a), but with wet sawdust in the end boxes. May 8, 1945. Experiment 43. Harpalus rupicola. 58 a) May 12, 1945. Gtl Visby, April 28-30, 1945. 25 specimens. End boxes with wet sawdust. b) Same as a), but with a different sequence of pH boxes: 7.5, 6.0, 6.0, 6.0, 4.8. May 16, 1945. 23 specimens. Experiment 44. Harpalus seladon. May 4, 1945. Gtl Visby, April 30, 1945. _ 25 specimens. End boxes with dry sawdust. Experiment 45. Harpalus serripes. May 10, 1945. Gtl Visby, April 29, 1945. 23 specimens. End boxes with wet sawdust. Experiment 46. Pterostichus anthracinus. August 26 through September 9, 1945. F, generation obtained by crossing the animals in Experiment 26a ( i.e. from one and the same population) (emergence of adults on July 4-17, 1945; maintained from the egg onward in the room with the substratum pH 4.8). 12 macropterous and 14 brachypterous specimens tested simultaneously, the latter marked with zinc white; and boxes with dry sawdust. The experiment was repeated 5 times, altogether 39 + 39 specimens. Exposure 2 days in each case. Table 28, p. 357. c. Alternating experiments with limestone gravel and siliceous gravel The limestone was weathered gray gravel from Gtl Visby and the silicate was limestone-free diluvial gravel from Upl Djursholm. Particles larger than 4 mm or smaller than 3/4 mm were removed by double filtration and by washing. Medium-size (1-2 mm) particles were somewhat more plentiful in the sam- ples of siliceous gravel. The following difference is noteworthy: The limestone gravel consisted of flat particles, the siliceous gravel of more or less spherical particles. For weight by volume see p. 127. Only 8 boxes of the substratum gradient apparatus were used. They were almost half-filled (180 cm?) with the limestone and siliceous gravel alternately. End boxes contained wet sawdust or humus. Moistening (about 10 cm? per box) with spring water. Exposure 1 day. See Table 5 (p. 126). Experiment 47. Cymindis humeralis. June 21-25, 1946. Oid Greby, June 11-15, 1946. Repeated 4 times, total 100 specimens. Experiment 48. Harpalus anxius. June 26 through 29, 1946. Old Stora-Rör, June 13, 1946. Repeated twice, total 100 specimens. Experiment 49. Harpalus azureus. June 24-29, 1946. Old Greby, June 11-15, 1946. Repeated 4 times, total 100 specimens. Experiment 50. Harpalus melleti. July 20 through August 20, 1945. Gtl Visby, April 28-30, 1945. Repeated 5 times, total 100 specimens. Experiment 51. Harpalus punctatulus. July 12-21, August 18-19, 1945. Gtl Visby, April 28, 1945. Repeated 5 times, total 100 specimens. Experiment 52. Harpalus rubripes. June 21-24, 1946. Old Greby, June 11-15, 1946. Repeated 3 times, total 100 specimens. 78 59 Experiment 53. Harpalus rufitarsis. June 26 through July 2, 1946. Old Stora-Ror, June 13, 1946. Repeated 5 times, total 34 specimens. Experiment 54. Harpalus rupicola. a) July 20-26, August 18-20, 1945. Gtl Visby, April 28-30, 1945. Repeated 5 times, total 50 specimens. b) June 24-26, 1946. Gtl Visby, June 19, 1946. Repeated twice, total 50 specimens. Experiment 55. Harpalus serripes. July 12-21, August 18-20, 1945. Gtl Visby, April 29, 1945. Repeated 6 times, total 100 specimens. Experiment 56. Panagaeus bipustulatus. June 26 through July 2, 1946. Old Greby, June 11-15, 1946. Repeated 5 times, total 39 specimens. d. Alternating experiments with limestone gravel and schist gravel Limestone gravel as described above. The schist gravel consisted of limestone-free argillaceous slate from Dlr Osmundberget, which was triturated and subjected to filtering and washing, as in the case of limestone. The particles were even flatter than those of limestone. For weight by volume, see p. 128. Moistening with distilled water. Otherwise exactly as in the preceding series of Experiments 47-56. See Table 6 (p. 127). Experiment 57. Harpalus anxius. July 3-9, 1946. Old Stora-Rör, June 13, 1946. Repeated 5 times, total 100 specimens. Experiment 58. Harpalus azureus. July 2-16, 1946. Old Greby, June 11-15, 1946. Repeated 7 times, total 100 specimens. Experiment 59. Harpalus melleti. July 3-16, 1946. Gtl Visby, June 19, 1946. Repeated 6 times, total 100 specimens. i Experiment 60. Harpalus rufitarsis. July 2-14, 1946. Old Stora-Ror, June 13, 1946. Repeated 5 times, total 35 specimens. Experiment 61. Harpalus rupicola. July 2-15, 1946. Gtl Visby, June 18, 1946. Repeated 6 times, total 100 specimens. Experiment 62. Panagaeus bipustulatus. July 3-9, 1946. Öld Greby, June 11-15, 1946. Repeated 5 times, total 34 specimens. e. Serial experiments with variable humidity Only 7 boxes were used (6 in Experiment 81). In each of these, a different quantity of water was added to 200 cc of dry sawdust (of aspen poplar) in the following series: Exposure one day (during spring 2 days) unless otherwise stated (the heat of summer caused a too rapid drying of the sawdust). 79 60 The experiments with Harpalus (and Cymindis) were carried out simulta- neously using several species, without this being noted in each case. It should be mentioned that delicate freshly emerged beetles should not be used in these experiments, since they have a demonstrably higher moisture requirement. Concerning the species of Harpalus, see Diagrams 21-22 (p. 133). Experiment 63. Cymindis angularis. July 3-23, 1946. Old Greby, June 11-15, 1946. Repeated 4 times, total 40 specimens’. Experiment 64. Cymindis humeralis. July 18-23, 1946, Old Greby, June 11-15, 1946. Repeated 3 times, total 60 specimens . Experiment 65. Cymindis macularis. July 3-23, 1946. Old Stora-Ror, June 13, 1946. Repeated 4 times, total 60 specimens.* Experiment 66. Harpalus aeneus. June 25-27, 1946. Old Greby, June 11-15, 1946. Repeated twice, total 75 specimens. Experiment 67. Harpalus anxius. June 28 through July 2, 1946. Old Stora- Ror, June 13, 1946. Repeated 3 times, total 75 specimens. Experiment 68. Harpalus azureus. June 23 through July 1, 1946. Old Greby, June 11-15, 1946. Repeated 3 times, total 75 specimens. Experiment 69. Harpalus hirtipes. June 25 through July 2, 1946. Old Stora- Ror, June 13, 1946. Repeated 4 times, total 50 specimens. Experiment 70. Harpalus melleti. a) May 19-23, 1945. Gtl Visby, April 28-30, 1945. Exposure 2 days, re- peated twice, total 40 specimens. b) Same as a), but with only one day of exposure. June 23 through July 1, 1946. Gtl Visby, June 17-19, 1946. Repeated 3 times, total 75 specimens. Experiment 71. Harpalus neglectus. June 25 through July 3, 1946. Old Stora- _ Ror, June 13, 1946. Repeated 5 times, total 50 specimens. 80 Experiment 72. Harpalus punctatulus. a) May 19-23, 1945. Gtl Visby, April 28, 1945. Exposure 2 days, repeated twice, total 41 specimens. Diagrams 8, 22, pp. 57, 134. b) Same as a), but with only one day of exposure. June 23 through July 1, 1946. Gtl Visby, June 17, 1946. Repeated 3 times, total 75 specimens. Diagrams 8, 2A pps) 37.133: c) Same as b), but before the experiment the animals were kept for 2 days in moisture-saturated air, July 2-3, 1946. 50 specimens. Diagram 8, p. 57. Experiment 73. Harpalus puncticeps. June 23 through July 3, 1946. Old Halltorp, June 15, 1946. Repeated 5 times, total 50 specimens. Experiment 74. Harpalus rubripes. June 28 through July 2, 1946. Old Greby, June 11-15, 1946. Repeated 3 times, total 75 specimens. Experiment 75. Harpalus rufitarsis. July 18-24, 1946. Old Stora-Rör, June 13, 1946. Repeated 4 times, total 28 specimens. *The three species of Cymindis did not behave distinctly different. 61 Experiment 76. Harpalus rupicola. a) May 19-23, 1945. Gtl Visby, April 28-30, 1945. Exposure 2 days, re- peated twice, total 42 specimens. b) Same as a), but with only one day of exposure. June 23 through July 1, 1946. Gtl Visby, June 17, 1946. Repeated 3 times, total 75 specimens. Experiment 77. Harpalus seladon. a) May 19-23, 1945. Gtl Visby, April 30, 1945. Exposure 2 days, repeated twice, total 39 specimens. ...b) Same as a), but with only one day of exposure. August 24-27, 1946. Old Greby, June 11-16, 1946. Repeated twice, total 100 specimens. Experiment 78. Harpalus serripes. a) May 19-23, 1945. Gtl Visby, April 29, 1945. Exposure 2 days, repeated twice, total 42 specimens. b) Same as a), but with only one day of exposure. June 25 through July 2, 1946. Old Greby, June 11 through 15, 1946. Repeated 6 times, total 150 specimens. Experiment 79. Harpalus smaragdinus. June 25-28, 1946. Old Stora-Rör, June 13, 1946. Repeated 3 times, total 75 specimens. Experiment 80. Harpalus tardus. June 25-28, 1946. Old and Gtl., June 14-18, 1946. Repeated 3 times, total 75 specimens. Experiment 81. Pterostichus anthracinus. September 16 through October 1, 1945. F, generation obtained by crossing the animals in Experiment 26a (i.e. from one and the same population) (adults emerged from July 4-17, 1945; maintained from the egg onwards in the room in wet humus soil). 13 macropterous and 13 brachypterous specimens simultaneously tested, the for- mer marked with zinc white. 6 boxes (those with 3 1/8 cm? water excluded; see p- 78). Exposure one day, repeated 8 times, altogether 100 + 100 specimens. Diagram 44, p. 357. * * * Experiment 82. Agonum dorsale. July 8, 1947. Old Halltorp area, June 12-23, 1947. The experiment was carried out in the circular universal gradi- ent apparatus (p. 93). About 1/2 cm thick layer of sawdust in 20 boxes of 10 moisture gradations (each duplicated as in Experiment 108 ff.). A lamp in the center of the gradient apparatus uniformly illuminated all the boxes. A glass plate was used as cover, through which observations of the beetles were made every 2 hours without touching the gradient apparatus. 40 specimens (including Brachynus). 3 replicates, hence total 120 specimens. Diagram 10, p- 61. Experiment 83. Brachynus crepitans. July 8, 1947. Old Halltorp area, June 12-23, 1947. As in Experiment 82. 40 specimens (including Agonum dorsale). Diagram 10, p. 61. 81 62 f. Serial and alternating experiments with salts and ammonia The purpose of the serial experiments was to determine a possible pre- ferendum to the concentration of substances in question for each species, with which alternating experiments could then be carried out. Coarse siliceous sand was used as neutral substratum. The experiments with Amara were carried out in the usual substratum gradient apparatus (p. 73). In serial experiments all 10 boxes were used with 2 days of exposure. In alternating experiments only six of these were used with one day of exposure (except in Experiment 84b). A different arrangement was adopted only in the ammonia experiment (Experiment 88). See also p. 525. In the case of the species of Bembidion the substratum was placed in a series of 6 small cuvettes (about 100 cm’), which were placed in a row at the bottom of a large glass container with the broad sides touching each other. The container was filled with water almost to the edge of the cuvettes, and strips of paper placed in the water enabled beetles that fell in to climb out again. Exposure 1 day. See p. 523. Experiment 84. Amara ingenua. a) Serial experiment with Ca(NO,),. September 16-26, 1945. Upl Solna, Bergshamra, September 15, 1945. Each box with 160 cm? of sand and 25 cm? of liquid in the following sequence: Box 1 2 3 4 5) 6 Y 8 9 10 Salt Distilled 1/4% 12% 1% 2% 3% 4% 5% 73% 10% solution water Repeated 4 times, total 100 specimens. b) Alternating experiment with Ca(NO,),. October 23 through November 2, 1945. The same animals as above. 6 boxes, alternately with distilled water and 5% Ca(NO,), solution. Exposure 2 days. Repeated 3 times, total 66 specimens. Experiment 85. Amara ingenua. Serial experiment with Thomas phosphate (commercial quality). November 10-14, 1945. Upl Solna, Bergshamra, Septem- ber 15, 1945. Each box with 160 cm? of sand and 25 cm? of liquid in the following sequence: Box 1 2 3 4 5) 6 7 8 9 10 Salt Distilled 0.08% 0.15% 03% 0.63% 1.25% 25% 5% 750% 10% solution* water Repeated twice, total 44 specimens. Experiment 86. Amara ingenua. a) Serial experiment with superphosphate (7-Ca(H,PO,), + 5-CaSO,). January 27 through February 8, 1946. Upl Solna, Bergshamra, September 15, “Thomas phosphate is merely suspended in water, but I tried to obtain as uniform a distri- bution of the particles as possible by shaking. 82 63 1945. Each box with 160 cm? of sand and 25 cm? of liquid in the following sequence: Box 1 22 3 4 5 6 7 8 9 10 Salt Distilled 0.25% 05% 1% 2% 3% 4% 5% 75% 10% solution water Repeated 4 times, total 72 specimens. b) Alternating experiment with superphosphate. February 14-18, 1946. The same animals as above. Six boxes, alternately with distilled water and 2% superphosphate solution. Exposure one day. Repeated 4 times, total 72 specimens. Experiment 87. Amara ingenua. a) Serial experiment with KCI. December 22-31, 1945. Upl Solna, Bergshamra, September 15, 1945. Each box with 160 cm? of sand and 25 cm? of liquid in the following sequence: Box 1 2 3 4 5 6 y 8 9 10 Salt Distilled 1/4% 1/2% 1% 2% 3% 4% 5% 73% 10% solution water Repeated 4 times, total 83 specimens. b) Alternating experiment with KCI. January 6-15, 1946. The same animals as above. Six boxes alternately with distilled water and 1% KCI solution. One day of exposure. Repeated 5 times, total 109 specimens. Experiment 88. Amara ingenua. a) Experiment with ammonia (commercial quality) February 23, 1946. Upl Solna, Bergshamra, September 15, 1945. Six boxes, each with 160 cm? of sand and 25 cm? of liquid in the following sequence: Box 1 2 3 4 5 6 Solution Distilled Distilled Distilled 1% 2% 3% ammonia. water water water 17 specimens. Exposure one day. b) Same as a), but with lower concentration of ammonia. March 3-4, 1946. Box 1 2. 3 4 5 6 Solution Distilled Distilled Distilled 1/8% 1/4% 1/2% ammonia. water water water 15 specimens. Exposure one day. Experiment 89. Amara praetermissa. Alternating experiment with Ca(NO,),. August 24-28, 1946. Upl Djursholm, August 24, 1946. Exactly as in Experiment 84b. Repeated twice, total 100 specimens. 83 64 Experiment 90. Amara praetermissa. Alternating experiment with super- phosphate (7-Ca(H,PO,), + 5-CaSO,). September 1-5, 1946. Up] Djursholm, August 24, 1946. Exactly as in Experiment 86b, but exposure 2 days. Repeated twice, total 100 specimens. Experiment 91. Amara praetermissa. Alternating experiment with KCl. August 28 through September 1, 1946. Upl Djursholm, August 24, 1946. Exactly as in Experiment 87b, but exposure 2 days. Repeated twice, total 100 specimens. Experiment 92. Bembidion aeneum. a) Alternating experiment with NaCl (together with B. minimum). August 24 through September 4, 1946. Boh Samstad, August 10, 1946. 6 cuvettes (see p. 81), each with 100 cm? of coarse siliceous sand and 15 cm? of liquid (alternating spring water and 1% NaCl solution). Exposure one day. Only 5-7 specimens, hence repeated 11 times, total 50 specimens. b) Same as a). June 25 through July 12, 1947. Old Mockelmossen, June 18, 1947. Only 5 specimens, so repeated 19 times, total 50 specimens. Experiment 93. Bembidion minimum. a) Serial experiment with NaCl. August 20-22, 1946. Boh Samstad, August 10, 1946. 6 cuvettes (see p. 81), each with 100 cm? of coarse siliceous sand and 10 cm? of liquid in the following sequence: Cuvette 1 2 3 4 5 6 Salt Spring 1/8% 1/4% 1/2% 1% 2% NaCl. solution water Exposure one day, repeated twice, total 29 specimens. b) Same as 92a. Alternating experiment with NaCl (together with 2. .aeneum). August 24-31, 1946. Boh Samstad, August 10, 1946. Repeated 8 times, total 50 specimens. g. Serial experiments with sand of various particle size The sand obtained from “Statens Vaginstitut,” Stockholm, was divisible into the following six categories: _ 1) 2.0 to 1.0 mm particle size. Gst Gavle, airfield. Isolated limestone par- ticles (detected with HCl). 2) 1.0 to 0.5 mm = 1. 3) 0.5 to 0.25 mm. His Sandarne. Limestone-free. 4) 0.25 to 0.125 mm. Vrm Karlstad, airfield. Limestone-free. 5) 0.125 to 0.075 mm. Mdp Sundsvall, airfield. Limestone-free. 6) < 0.075 mm = 5. Except for Dyschirius (Experiments 95, 96), the usual substratum gradient apparatus was used in all experiments. Six boxes were placed in the container, each with 175 cm? of sand of categories 1-6 in simple sequence. See Diagram 51 (p. 506). 84 65 For Dyschirius a deep dish was used, divided by vertical strips of paper into 6 equal sectors, each with a different type of sand. In this case a circular gradient apparatus was used. See Diagram 52 (p. 507). Since completely dry sand could not be used, I tried to judge its water- holding capacity, i.e. the “relative” moisture content, as follows: 20 cm? of each category of sand was placed in a reagent tube, as tightly packed as possible. Water was carefully added until the air was completely excluded and the water formed about a one mm thick layer on the surface. The quantities of water required for this purpose were: Category of sand 1 2 3 4 5 6 Water, cm? 9 9.5 10 9.5 10.5 9.7 According to these proportions different quantities of water were added to the sand, beginning with 5 cm? (10 cm? in the case of Dyschirius) per 100 cm? of type 3. Doubtless it would have been more accurate to determine the moisture preferendum of every species first for each category of sand and then carry out a comparative study of the six samples obtained in this way. Exposure one day in every case. Experiment 94. Cymindis macularis. July 24-30, 1946. Old Stora-Rör, June 13, 1945. Repeated 5 times, total 50 specimens. Irregular distribution of the individuals. Experiment 95. Dyschirius obscurus. August 18-26. 1947. Nl Tvarminne, around August 1, 1947. Deep dish with 6 different sectors for sand (see above). Repeated 8 times, total 100 specimens. Experiment 96. Dyschirius thoracicus. June 20-28, 1947. Old Hornsjön lake, June 15, 1947. Exactly as in Experiment 95. Repeated 5 times, total 150 spe- cimens (in two experiments, 63 specimens, the moisture content of the sand was checked only with the fingers). Experiment 97. Harpalus anxius. July 29-30, 1946. Old Stora-Ror, June 13, 1946. Repeated twice, total 100 specimens. Experiment 98. Harpalus hirtipes. July 24-29, 1946. Old Stora-Rör, June 13, 1946. Repeated 5 times, total 75 specimens. Experiment 99. Harpalus neglectus. July 24-30, 1946. Old Stora-Ror, June 13, 1946. Repeated 6 times, total 50 specimens. Experiment 100. Harpalus rufitarsis. July 24-31, 1946. Old Stora-Rör, June 13, 1946. Repeated 6 times, total 35 specimens. Experiment 101. Harpalus serripes. July 24-27, 1946. Öld Greby, June 11-15, 1946. Repeated 3 times, total 100 specimens. Experiment 102. Harpalus smaragdinus. July 28-31, 1946. Öld Stora-Rör, June 13, 1946. Repeated 4 times, total 100 specimens. 85 66 Experiment 103. Harpalus tardus. July 28-31, 1946. Öld and Gtl, June 14-18, 1946. Repeated 3 times, total 75 specimens. h. Experiments on the relationship between two species The suspected “passive” species was confined within wide-meshed net or wire gauze walls, and the suspected “active” species was left free to settle anywhere, or close to the walls, through which of course the smell of the beetles passed. Experiment 104. Brachynus crepitans versus Agonum dorsale. May 26 through June 1, 1945. Upl Lövon, April 2, 1945 (Agonum, Gtl Horsne, April 29, 1945). 5 boxes of the substratum gradient apparatus (half-filled with sand) were used, the middle one was closed with a thin cloth after releasing 10 specimens of Agonum into it. The open boxes contained 24 specimens of Brachynus. Exposure for 3 days. Each experiment was repeated 3 times, total 72 specimens of Brachynus; p. 549. Experiment 105. Dyschirius obscurus versus Bledius arenarius Payk. August 14-18, 1947. NI Tvarminne, around August 1, 1947 (both species). Deep dish with humid sand. Two small wire gauze cages (devised by Palmen), 2.5 x 2.5 x 2.5 cm, were placed in the sand exactly opposite one another, close to the edge of the dish (Fig. 5). One contained 6-8 specimens of Bledius and the other was empty. The 4 sectors (see figure) were made only the following day during observation, and the distribution of Dyschirius in these was recorded. Repeated 4 times, total 50 specimens; p. 546. Experiment 106. Dyschirius thoracicus versus Bledius arenarius. June 18-20, 1947, Old Hornsjön lake, June 15, 1947 (both species). Exactly as Experiment 105. Repeated twice, the second time with 5 specimens of Stenus spp. in the cage that was empty the first time, total 65 specimens; p. 546. i. Distribution of insects in the substratum gradient apparatus without the influence of determinable factors (control experiment) Experiment 107. Amara ingenua. September 29 through October 18, 1945. Upl Solna, Bergshamra, September 15, 1945. All 10 boxes were used, each with 160 cm? of coarse sand and 25 cm? of distilled water. Exposure for 2 days. Repeated 9 times, total 200 specimens. Diagram 13, p. 74. III. The Universal Gradient Apparatus (“Universalorgel”) I have given this pretentious name to an apparatus developed by alterations of Krogerus’ substratum gradient apparatus (p. 73). In addition to this “linear” universal gradient apparatus, I have also developed a “circular” one, which is described below in detail, although it proved less suitable for my purpose. 86 67 Fig. 5. Circular substratum gradient apparatus (plate) used for studying relationship between Dyschirius and Bledius. Experiments 105, 106, p. 84. A-C are sectors partitioned when recording observations. Striated squares are small cages of wire gauze. Dotted area—sand. The principle of the universal gradient apparatus involves study of the simultaneous effects of 2 or more factors (3 in the present case) in various combinations. It is therefore a substratum gradient apparatus which also func- tions as a temperature gradient apparatus with the help of the bent copper sheets soldered to the ends. The boxes of the universal gradient apparatus do not have perforated walls as in the substratum gradient apparatus, but are provided with a 5 mm high horizontal slit (Fig. 6), only 5 mm above the bottom, through which animals of that particular size can pass, but which fairly effectively isolates the air contained in the adjacent boxes. In my experiments, the factors temperature, humidity (of the bottom sub- stratum, to a lesser extent that of the air), and /ight were investigated*. Each of these is here treated separately, and the more important experimental errors are pointed out. Temperature. As in the case of the temperature-gradient apparatus, the copper plate at one end was immersed in water heated by an adjustable spirit lamp, whereas at the other end there was continuous cooling with cold running water. It differs from the usual temperature gradient apparatus in the fact that the position and temperature preferendum of individual animals could not be determined, since these were distributed over 10 temperature categories (= boxes), whose mean temperature must be calculated. For this purpose two sets of temperature readings were taken for all the boxes at the same temperature of hot and cold running water 50°C and 11.8°C *The chemical characteristics (for instance, the NaCl content) of the water could easily be considered as a fourth factor. 87 88 68 Fig. 6. Box of the (circular) “universal gradient apparatus.” respectively). In one case the bulbs of all the thermometers rested on the bottom against the warmer wall and in the other against the cooler wall of each box. The following figures were recorded: Box 1 2 3 4 5 Temperature 16.4-17.5 18.6-19.8 205-2 lez 21.9-22.6 PPL I 232 Mean 17.0 19 20.9 22.3 U 23.0 Box 6 7 8 9 10 Temperature 23.5-24.0 24.4-24.8 26.0-26.9 27.5-28.4 31.8337 Mean 23.8 24.6 26.5 28.0 320] The fall in temperature in the universal gradient apparatus thus occurs by steps, with high levels at the warm end of the apparatus along the edge of the two boxes (Diagram 14). These falls may exercise an interfering, choking effect on the beetles and serve as barriers. The diagram also shows that the temperature plot of the universal gradient apparatus is less uniform than that of the temperature gradient apparatus, because the fall in temperature in the middle parts of the apparatus is smoother. Direct comparison of the preferenda obtained in the two gradient apparatuses is therefore not possible. With a steeper fall in temperature in the universal gradient apparatus, the temperature of the adjacent boxes at both ends of the apparatus differed still more. A simultaneous reading at the warm and cold walls in the three warmest and two coldest boxes provided the following data: Box 1 2 De 8 9 10 Nemperature, 143153 17.0180... 273 28:8 Sill 34:0 55.4393] Mean 14.8 17:5 28.1 32.6 313 69 35. 30° DEE 20° Universal gradient 9 apparatus, box number 1 2 3 4 5 6 7 8 9 io Temperature gradient 1 2 3 4 5 apparatus, thermometer number Diagram 14. Comparison between the fall in temperature in the univer- sal gradient apparatus (circles) and in the temperature gradient apparatus (crosses). Open circles—Extreme temperatures; Black circles—Calculated mean value for each box in the universal gradient apparatus. In all the experiments with beetles the thermometers in the 5 “warm” boxes touched the warm wall and in the 5 “cold” boxes touched the cold wall. The average temperature of every box was estimated from readings so taken. All measured and estimated temperatures, as in the case of the temper- ature gradient apparatus (p. 66), are of the layer of air next to the bottom, but not of the bottom itself. Several control experiments, both with dry saw- dust and with different grades of moist sawdust, showed that the temperature reading was not dependent upon whether the thermometer directly touched the bottom or not. Humidity. Every box in all the experiments contained 25 cm? of sawdust (of aspen poplar), which was uniformly spread over the bottom in an almost 5 mm thick layer. If the humidity factor was studied simultaneously, differ- ent quantities of water were added to the sawdust samples, resulting in the following smoothly ascending series: Box 1 2 3 4 5 6 7 8 9 10 H:0, cm? 0 0.5 1 2 3 4 5 6 8 10 89 70 Before carrying out the experiment, water was distributed in each sample as uniformly as possible by careful mixing. One sample was never used in more than two successive experiments. The differences between these samples lay in the moisture content of the substrata, rather than in that of the air, which was also taken into consideration when testing species with burrowing habits (only the genus Harpalus). How- ever, in order to determine the humidity of the air in the boxes the following measurements were taken with a calibrated hair-hygrometer: a) Humidity series as above, without temperature differences (room tem- perature 22.7°C, relative humidity of the air 69%): Box 1 2 3 4 a 10 Water content of sawdust 0 0.5 1 2 N 10 (H20, cm?) A TAN % relative humidity 69 84 99 100 of air b) Simultaneous humidity and temperature series (“warm-dry”) (room temperature 25.0°C, relative humidity of the air, 64%): Box 1 2) 3 4 5 6 U 8 9 10 Temperature (mean) 41.977344 773127728:07726.0772437,723:07 20.8 32152 H>0, cm? 082.05 1 2 3 4 5 6 8 10 % relative humidity 70 92 100 100 100 100 100 100 100 100 of air c) Inverse humidity and temperature series (“warm-moist”) (room tem- perature 24.7°C, relative humidity of air, 64%): Box 1 2, 3 4 5 6 V 8 9 10 Temperature (mean) 39557 34.0, 30/6 26:06. 2553235912297 18T Ra HO, cm? 10 8 6 5 4 3 2 1 0.5 0 % relative humidity 100 100 100 100 100 100 96 90 75 64 of air d) Temperature data as in b) and c) (38.7-15.2°C) but throughout with dry sawdust (room temperature 22.7°C, relative humidity of the air 69%). The relative humidity of the air in the boxes varied only from 72% at the warm end to 68% at the cold end. 90 zul These data show: first, that the relative humidity of the air reaches 100% even with low moisture in the sawdust, and it therefore has a very small influ- ence on the distribution of beetles in the apparatus; second, that the volumes of air in different boxes are apparently well isolated from one another, since adjacent boxes can show for a long period such different percentages of relative humidity of the air as 64, 75, 90 or 70, 92, 100%. The exchange of air through the narrow slit at the bottom of the boxes is evidently very little, which must also be conducive to the maintenance of constant temperature during the ex- periment. In experiments where the humidity factor had no importance, dry or almost dry sawdust (2 cm? H,O per 25 cm?) was used. Light. Corresponding to each box, four small squares were cut out in thick black opaque paper; the total exposed surface area of the boxes decreased in the following sequence: 1 (without paper, all 55 cm’ open), 1/2 (273 cm?), 1/4 (132 cm?), ..., 1/256, O (dark). Cf. Fig. 7. The paper, together with a tightly fitting glass lid, formed the cover of the apparatus. Quantitative measurements of the light entering each box were not taken. They would have been useful only if the same (artificial) source of light was always used, which was not the case. All the “light experiments” were carried out in the same room with indirect daylight, never with a completely overcast sky, from June through early August. Hence the results are in no way precise. However, since light was studied only as a “regulating” factor in relation to temperature and humidity, the procedure might be considered suitable. All experiments without the light factor were carried out in the dark, using black paper all over. The experiments with the universal gradient apparatus are also subject to errors other than those mentioned above, some of which affected the experiments on the temperature and substratum gradient apparatus as well (see pp. 66 and 73). In addition, the following points may be emphasized: 1) The short duration of exposure (always 2 hours). This was necessary when very dry sawdust was used, since this has a harmful effect on the less xerophilous species, and such animals soon become restless. 2) The readings in each experiment were recorded as follows: After re- moving the lid the boxes were taken as quickly as possible and placed in a porcelain dish whereupon the animals were counted. A certain amount of disturbance, which often drove the faster, “more nervous” species from their resting position into another box, could not be avoided. When a large num- ber of beetles are involved, and one need not take care (as I had to do) that some individuals are not crushed each time, the boxes should be isolated from one another quickly, without removing them, by inserting thin metal plates in between. 3) At high temperatures, on account of the increased loss of water from 91 V2 the animal, dry sawdust has a stronger negative influence, and therefore the temperature preferendum is higher with moist sawdust. Furthermore, two experiments with dry sawdust must not be carried out in quick succession, because the animals get thirsty and behave abnormally. 4) Because of the slight temperature differences among the boxes in the middle (see Diagram 14), the beetles are distributed more sparsely there and exhibit less distinct maxima than in the pure temperature gradient apparatus. Box No 1 2 3 4 5 6 7 8 9 10 Diagram 15. Distribution of 100 specimens each of given species in the universal gradient apparatus without influence of any measurable factor. a—Harpalus punctatulus; b—H. seladon; c—H. melleti; d—H. serripes. Con- tinuous line—Empirical numbers; broken line—Symmetrically leveled num- bers. Experiment 119, p. 103. 93 94 173 5) As in the case of the linear gradient apparatus (see p. 73), most species show a tendency to cluster at the ends. Since the warm end has a repelling effect, this means that in an experiment where the temperature factor is also involved, over-representation results in the cold part of the gradient apparatus. Besides, (at every temperature), the moist part of the apparatus to a lower extent, slows down the movements of the animals. As in the case of the substratum gradient apparatus (Diagram 13), some species were investigated, with respect to these errors, i.e. their distribution in the universal gradient apparatus was studied (Diagram 15), without the influ- ence of measurable factors. The result is that even species that are taxonom- ically and ecologically closely related may behave differently in this respect. Evidently correction of the results of the remaining experiments can be un- dertaken on the basis of these diagrams. An example of how this can be done is provided by the experiment with Harpalus punctatulus concerning the “hu- midity” factor (Diagram 16). Otherwise I have not made any correction of the experimental values of the four species studied (or any others) (cf. p. 74). The main reason is that the experiments with the universal gradient apparatus were primarily intended to be a study of the variations shown by the same species under the influence of various combinations of factors, whereat the correction would be an unimportant constant. On the other hand, I have tried to avoid this disproportionate distribution in the apparatus by devising a circular universal gradient apparatus (Fig. 7). This contains 20 boxes, double the linear apparatus. Unfortunately it was found that taking readings for every experiment (point No. 2 above) with so many boxes resulted in excessive disturbance of the active carabids and I had to give up this apparatus after several trials. However, the apparatus can be useful with slow- moving species and those that do not hide in the ground (or are not studied with respect to the substratum), since the results can be recorded through the glass lid. Moreover, in the circular gradient apparatus the “unnatural” clustering in the cold part, especially if it was also moist, can be eliminated only in certain cases (Diagram 17); in other cases it was unavoidable (Diagram 18). For my purpose, therefore, the apparatus was no better than the linear arrangement. In conclusion one might ask, in view of the numerous demonstrable ex- perimental errors and inaccuracies in the results with the universal gradient apparatus, whether it was useful to spend so much time on these experiments. I think it was. In part, the results are not quite in the nature of fortuitous hap- penings, as might have been expected, which was shown by the often remark- able identical distribution of the animals in repeated experiments (Diagrams 25-35, pp. 141 ff.). Then again, as in the case of all the similar experiments in- cluded in this book, I am convinced that from the comparative values obtained with simultaneous studies on two or more species, certainly no physiological, (66 ‘d “-qZ11 Juswumodxq) suaunoads [7 ‘Oy isndny g isnsny—o ‘(66 -d “.gzıı wotliodxq) suswpads oz ‘zı Aine ‘oT Ainr—aq ‘(66 ‘d “azıı qwoutliadxq) suswmads Op ‘97 aung—e “BCT WEISEIT UI 9AIND O} SUIPIOIOR UONI91109 Jaye pouleiqo AInJ— au USXOIg ‘aamno Jesundwg—oun snonunuoy -g wweiserg JO 'syyuow jusıayyıp ul snyeledde jusıpei1s jesısarun oy) Ul Aıpııuny 0) ZumpIone uvonnginsiqg "snynivisund snypdıoyy 9] wesseiq ¢ OL 6 8 L 9 S v € @ | ONxog 0 =O =o C6 96 94 75 but satisfactory ecological and zoogeographical, answers can be derived, since they often show only a ranking order among a series of species. Finally, the experiments with the universai gradient apparatus actually highlighted certain points on the interaction of two factors (temperature and humidity), about which I was unable to obtain an explanation in any other way. In the universal gradient apparatus only species of the genus Harpalus were studied. The entire experimental material used (with the exception of some specimens of H. seladon) originated from Oland and Gotland. It comprised old hibernated adults. It should therefore be as suitable for comparison as any. The results are all given coherently (p. 139 ff.); an indication with every experiment would therefore be superfluous. The mean temperature calculated for each box in each experiment is evi- dent from Diagrams 25-35 (p. 141 ff.). With each of the 11 species treated the following 13 different experiments (always duplicated) were carried out: Fig. 7. Circular universal gradient apparatus. Above: Black paper used in experiments with light factor. 76 ‘Wy 6c ‘dd ‘snjesedde juaipess Jesıaarun 107 sweigerp Jo uoneueldxa osje 998 “(Ajaanoodsai EgT “d ‘07T Juowedxg ‘66 'd “DEI Juswuadxg) sodugnı pp pue 9 ‘(Ajaanoadsai Eo -d ‘pozI uswnadxg ‘COI 'd ‘az Tt uawnadxg) smuipspsnwus snjodivzj—q pue e ‘snjesedde Jusıpeı$ sejnos9 Jo qed pjoo ul syoasut Jo Suusisnjo psonpay ‚snyeredde juatpess jesiaatun (p ‘g) Jejnosto ul pue (9 ‘e) seoury ut (Ayıprtuny ‘ainjesadua}) Suswmadxa sojoey-z aieodng “11 welselg ISIOW Aig ISIOW) hig “08 66 We), ‘IJ 6¢1 ‘dd ‘snyesedde juatpess jesıaamun 107 sweigeip JO uoneueldx> osje 399 “(Afaamaadsaı go ‘d ‘qozt Juswuadxgq ‘66 'd ‘pzt] Juawadxg) smyniwjound p—p pure > ‘(Ajoanoodsai gor “d ‘e0ZI quauodxg ‘86 ‘d ‘ar {] wouedxq) sayjaus snypdıry —q pue e ‘saseo je ul snyeiedde Jo pua pjoo ie syoasul Jo SulsaysniD -snjesedde quoipess jessoatun (p ‘q) 1ejns11 ul pue (9 ‘e) eaun ul (Aıpruny ‘ainjesodtuay) syuatutadxa Joyoej-Z aeodnq 'gI weiseig] ‚Or sgL Ov GE O€ ST 07 sl ie) fe} ® e 0 0 OL OL 07 07 %0€E e %0€ ® 9 ® Aiq Isıo N JSIOW Aig 97 78 a-c. Experiments with one factor (temperature, humidity, light), each taken separately. d-i. Two-factor experiments. d. Temperature + humidity (“warm-dry”). e. Temperature — humidity (“warm-moist”). f. Temperature + light (“warm-dark”). g. Temperature — light (“warm-light”). h. Humidity + light (“dry-dark”). i. Humidity — light (“dry-light”). j-m. Three-factor experiments. j. Temperature + humidity + light (“warm-dry-dark”). k. Temperature + humidity — light (“warm-dry-light”). l. Temperature — humidity + light (“warm-moist-dark”). m. Temperature — humidity — light (“warm-moist-light”). Unless otherwise mentioned the /inear universal gradient apparatus was used. The exposure always lasted 2 hours, at the most 10 minutes more or less. Experiment 108. Harpalus aeneus. July 15-24, 1946. Old Greby, June 11-15, 1946. a'. Temperature. July 18 through July 20, 2 experiments (with dry sawdust), total 60 specimens (together with rubripes). a°. July 20. 2 experiments (with slightly moist sawdust), total 60 specimens (together with rubripes). b. Humidity. July 23. 2 experiments, total 60 specimens (together with rubripes). c. Light. July 15, July 17. 2 experiments, total 60 specimens (together with anxlus). d. Temperature + humidity. July 17, July 18. 2 experiments, total 60 spe- cimens (together with rubripes). e. Temperature — humidity. July 15. 2 experiments, total 59 specimens (together with anxius). f. Temperature + light. July 15, July 17. 2 experiments, total 60 specimens (together with anxius). g. Temperature — light. July 20, July 21. 2 experiments, total 60 specimens (together with rubripes). h. Humidity + light. July 20, July 23. 2 experiments, total 60 specimens (together with rubripes). i. Humidity — light. July 24. 2 experiments, total 60 specimens (together with rubripes). j. Temperature + humidity + light. July 17, July 18. 2 experiments, total 60 specimens (together with rubripes). k. Temperature + humidity — light. July 21, July 23. 2 experiments, total 59 specimens (together with rubripes). 79 l. Temperature — humidity + light. July 15. 2 experiments, total 60 spe- cimens (together with anxius). m. Temperature — humidity — light. July 23. 2 experiments, total 60 spec- imens (together with rubripes). Experiment 109. Harpalys anxius. July 14-25, 1946. Old Stora-Rör, June 13, 1946. a’. Temperature. July 24, July 25. 2 experiments (with dry sawdust), total 59 specimens (together with smaragdinus and tardus). a°. July 22, July 24. 2 experiments (with slightly moist sawdust), total 60 specimens (together with smaragdinus and tardus). b. Humidity. July 24, July 25. 2 experiments, total 60 specimens (together with smaragdinus and tardus). c. Light. July 15, July 17. 2 experiments, total 60 specimens (together with aeneus). d. Temperature + humidity. July 14, July 19. 2 experiments, total 60 spe- cimens (together with smaragdinus and tardus). e. Temperature — humidity. July 15. 2 experiments, total 59 specimens (together with aeneus). f. Temperature + light. July 15, July 17. 2 experiments, total 60 specimens (together with aeneus). g. Temperature — light. July 14. 2 experiments, total 60 specimens (to- gether with smaragdinus and tardus). h. Humidity + light. July 19, July 22. 2 experiments, total 60 specimens (together with smaragdinus and tardus). i. Humidity — light. July 21, July 24. 2 experiments, total 60 specimens (together with smaragdinus and tardus). j. Temperature + humidity + light. July 14, July 19. 2 experiments, total 60 specimens (together with smaragdinus and tardus). k. Temperature + humidity — light. July 19, July 22. 2 experiments, total 60 specimens (together with smaragdinus and tardus). l. Temperature — humidity + light. July 15. 2 experiments, total 60 spe- cimens (together with aeneus). m. Temperature — humidity — light. July 21, July 22. 2 experiments, total 60 specimens (together with smaragdinus and tardus). Experiment 110. Harpalus azureus. July 3-10, 1946. Öld Greby, June 11-15, 1946. a!. Temperature. July 3, July 7. 2 experiments, total 60 specimens (July 7, together with rupicola). a’. July 25. 2 experiments (with slightly moist sawdust), total 60 specimens (together with melleti and serripes). b. Humidity. July 8, July 10. 2 experiments, total 60 specimens (July 8, together with rupicola; July 10, together with punctatulus). 98 80 c. Light. July 8. 2 experiments, total 60 specimens (once together with rupicola). d. Temperature + humidity. July 3, July 7. 2 experiments, total 60 speci- mens (July 7, together with rupicola). e. Temperature — humidity. July 5. 2 experiments, total 59 specimens (once together with rupicola). f. Temperature + light. July 3, July 6. 2 experiments, total 60 specimens (July 6, together with rupicola). g. Temperature — light. July 5, July 6. 2 experiments, total 60 specimens (July 5, together with rupicola). h. Humidity + light. July 9. 2 experiments, total 60 specimens (once to- gether with rupicola). i. Humidity — light. July 9, July 10. 3 experiments, total 90 specimens (twice together with rupicola). j. Temperature + humidity + light. July 3, July 7. 2 experiments, total 60 specimens (July 7, together with rupicola). k. Temperature + humidity — light. July 7. 2 experiments, total 60 speci- mens (once together with rupicola). l. Temperature — humidity + light. July 5, July 6. 2 experiments, total 60 specimens (July 5, together with rupicola). m. Temperature — humidity — light. July 5, July 6. 2 experiments, total 60 specimens (July 5, together with rupicola). Experiment 111. Harpalus melleti. June 29 through August 9, 1945 (with the exception of 111a*). Gtl Visby, April 28-30, 1945 (111a°, June 17, 1946). a'. Temperature. July 2-6, 1945. 3 experiments, total 114 specimens (once together with punctatulus, twice with rupicola, 3 times with serripes). a’. July 25, 1946. Gtl Visby, June 17, 1946. 2 experiments (with slightly moist sawdust), total 60 specimens (together with azureus and serripes). b. Humidity. June 30, August 8, 1945. 2 experiments, total 65 specimens _ (once together with punctatulus, once with rupicola, twice together with ser- ripes ). c. Light. June 29, August 8. 2 experiments, total 64 specimens (together with serripes, once also with rupicola). d. Temperature + humidity. July 2, July 31. 2 experiments, total 66 spe- cimens (together with serripes, once also with punctatulus and rupicola). e. Temperature — humidity. July 5, August 2, 2 experiments, total 62 specimens (together with serripes, once with rupicola, once with seladon). f. Temperature + light. July 3, July 31. 2 experiments, total 64 specimens (together with serripes, once with rupicola). g. Temperature — light. July 5, July 31. 2 experiments, total 62 specimens (together with serripes, once each with rupicola and seladon). h. Humidity + light. Juiy 1, August 9. 2 experiments, total 65 specimens (together with serripes, once with rupicola). 99 81 i. Humidity — light. July 1, August 9. 2 experiments, total 64 specimens (together with serripes, once with rupicola). j. Temperature + humidity + light. July 4, August 2. 2 experiments, total 63 specimens (together with serripes, once with punctatulus and rupicola, once with seladon). k. Temperature + humidity — light. July 4, August 2. 2 experiments, total 63 specimens (together with serripes, once with rupicola). l. Temperature — humidity + light. July 5, August 2. 2 experiments, total 62 specimens (together with serripes, once with rupicola). m. Temperature — humidity — light. July 3, August 1. 2 experiments, total 63 specimens (together with serripes, once each with rupicola and seladon). Experiment 112. Harpalus punctatulus. June 20 through August 10, 1945 (with the exception of 112a* and b’). Gtl Visby, April 28, 1945. (112a°; b’, June 17, 1946). a’. Temperature. June 20 through July 2, 1945. 3 experiments, total 119 specimens (twice together with seladon, once with melleti, rupicola, serripes). a’. July 25, 1946. Gtl Visby, June 17, 1946. 2 experiments (with slightly moist sawdust), total 60 specimens (together with rupicola and seladon). b!. Humidity. June 26, 1945. 1 experiment, 40 specimens (together with seladon). b?. August 8, August 10, 1945. 3 experiments, total 71 specimens (once together with seladon, once with melleti and serripes). b>. July 10, July 12, 1946. Gtl Visby, June 17, 1946. 2 experiments, total 70 specimens (once together with azureus). c. Light. June 21, August 8, 1945. 2 experiments, total 65 specimens (to- gether with seladon). d. Temperature + humidity. June 20, July 2. 2 experiments, total 80 spe- cimens (once together with seladon, once with melleti, rupicola, serripes). e. Temperature — humidity. June 27, June 30. 2 experiments, total 80 specimens (together with seladon). f. Temperature + light. June 6, June 29. 2 experiments, total 80 specimens (together with seladon). g. Temperature — light. June 27, July 5. 2 experiments, total 80 specimens (together with seladon). h. Humidity + light. June 27, August 10. 2 experiments, total 64 specimens (together with seladon). i. Humidity — light. June 28, August 11. 2 experiments, total 64 specimens (together with seladon). j. Temperature + humidity + light. June 28, July 4. 2 experiments, total 80 specimens (once together with seladon, once with melleti, rupicola, serripes). k. Temperature + humidity — light. June 28, August 1. 2 experiments, total 66 specimens (together with seladon). 100 82 |. Temperature — humidity + light. June 27, August 1. 2 experiments, total 66 specimens (together with seladon). m. Temperature — humidity — light. June 28, June 30. 2 experiments, total 80 specimens (together with seladon). Experiment 113. Harpalus rubripes. July 16-24, 1946. Öld Greby, June 11-15, 1946. al. Temperature. July 18, July 20. 2 experiments, total 60 specimens (to- gether with aeneus). a’. July 20. 2 experiments (with slightly moist sawdust), total 60 specimens (together with aeneus). b. Humidity. July 23. 2 experiments, total 60 specimens (together with aeneus). c. Light. July 16, July 18. 2 experiments, total 60 specimens (together with smaragdinus and tardus). d. Temperature + humidity. July 17, July 18. 2 experiments, total 60 spe- cimens (together with aeneus). e. Temperature — humidity. July 16. 2 experiments, total 60 specimens (together with smaragdinus and tardus). f. Temperature + light. July 16, July 18. 2 experiments, total 60 specimens (together with smaragdinus and tardus). g. Temperature — light. July 20, July 21. 2 experiments, total 60 specimens (together with aeneus). h. Humidity + light. July 20, July 23. 2 experiments, total 60 specimens (together with aeneus). i. Humidity — light. July 24. 2 experiments, total 60 specimens (together with aeneus). j. Temperature + humidity + light. July 17, July 18. 2 experiments, total 60 specimens (together with aeneus). k. Temperature + humidity -- light. July 21, July 23. 2 experiments, total 60 specimens (together with aeneus). |. Temperature — humidity + light. July 16. 2 experiments, total 61 spe- cimens (together with smaragdinus and tardus). m. Temperature — humidity — light. July 23. 2 experiments, total 60 spe- cimens (together with aeneus). Experiment 114. Harpalus rupicola. June 29 through July 5, 1945 and July 5-25, 1946. Gtl Visby, April 28-30, 1945 and June 18, 1946, respectively. a'. Temperature. July 2, July 5, 1945. 2 experiments, total 39 specimens (together with melleri and serripes, once also with punctatulus). a’. July 25, 1946. 2 experiments (with slightly moist sawdust), total 59 specimens (together with punctatulus and seladon). a’. July 7, 1946. 1 experiment (with dry sawdust), 29 specimens (together with azureus). 83 b!. Humidity. June 30, 1945. 1 experiment, 20 specimens (together with melleti and serripes). b?. July 8, 1946. 1 experiment, 30 specimens (together with azureus). c!. Light. June 29, 1945. 1 experiment, 20 specimens (together with melleti and serripes). c?. July 8, 1946. 1 experiment, 30 specimens (together with azureus). d'. Temperature + humidity. July 2, 1945. 1 experiment, 20 specimens (together with melleti, punctatulus, serripes). d’. July 7, 1946. 1 experiment, 30 specimens (together with azureus). e!. Temperature — humidity. July 3, July 5, 1945. 2 experiments, total 40 specimens (once with melleti and serripes). e*. July 5, 1946. 1 experiment, 29 specimens (together with azureus). f'. Temperature + light. July 3, 1945. 1 experiment, 20 specimens (together with melleti and serripes). f’. July 6, 1946. 1 experiment, 30 specimens (together with azureus). g'. Temperature — light. July 5, 1945. 1 experiment, 20 specimens (together with melleti and serripes). g°. July 5, 1946. 1 experiment, 29 specimens (together with azureus). h'. Humidity + light. July 1, 1945. 1 experiment, 20 specimens (together with melleti and serripes). h?. July 9, 1946. 1 experiment, 30 specimens (together with azureus). i’. Humidity — light. July 1, 1945. 1 experiment, 20 specimens (together with melleti and serripes). . i’. July 10, 1946. 1 experiment, 30 specimens (together with azureus). j’. Temperature + humidity + light. July 4, 1945. 1 experiment, 20 speci- mens (together with melleti, punctatulus and serripes). j’. July 7, 1946. 1 experiment, 30 specimens (together with azureus). k'. Temperature + humidity — light. July 4, 1945. 1 experiment, 20 spe- cimens (together with melleti and serripes). k’. July 7, 1946. 1 experiment, 30 specimens (together with azureus). l!. Temperature — humidity + light. July 5, 1945. 1 experiment, 20 speci- mens (together with melleti and serripes). I. July 5, 1946. 1 experiment, 30 specimens (together with azureus). m'. Temperature — humidity — light. July 3, 1945. 1 experiment, 20 spe- cimens (together with melleti and serripes). m?. July 5, 1946. 1 experiment, 30 specimens (together with azureus). Experiment 115. Harpalus seladon. June 20 through July 5, 1945; Gtl Visby, April 30, 1945. July 31 through August 11, 1945; Upl. Varmdon, Vasterskagga, July 1945. July 25, 1946; Old Greby, June 11-15, 1946. a'. Temperature. June 20, June 28, June 29, August 1, 1945. 4 experiments, total 129 specimens (three times with punctatulus, once with serripes). a’. July 25. 1946. 2 experiments (with slightly moist sawdust), total 59 Specimens (together with punctatulus and rupicola). 101 84 b. Humidity. June 26, August 8, 1945. 2 experiments, total 59 specimens (together with punctatulus). c. Light. June 21, August 8, 1945. 2 experiments, total 66 specimens (to- gether with punctatulus). d. Temperature + humidity. June 20, August 1, 1945. 2 experiments, total 72 specimens (once together with punctatulus, once with serripes). e. Temperature — humidity. June 27, June 30, August 2, 1945. 3 experiments, total 92 specimens (twice with punctatulus, once with melleti and serripes). f. Temperature + light. June 29, August 3, 1945. 2 experiments, total 62 specimens (together with punctatulus). g. Temperature — light. June 27, July 5, July 31, 1945. 3 experiments, total 92 specimens (twice with punctatulus, once along with melleti and serripes). h. Humidity + light. June 27, August 10, 1945. 2 experiments, total 65 Specimens (together with punctatulus). i. Humidity — light. June 28, August 11, 1945. 2 experiments, total 62 specimens (together with punctatulus). j. Temperature + humidity + light. June 28, August 2, 1945. 2 experiments, total 67 specimens (once along with punctatulus, once with melleti and serripes). k. Temperature + humidity — light. June 28, August 1, 1945. 2 experiments, total 65 specimens (together with punctatulus). 1. Temperature — humidity + light. June 27, August 1, 1945. 2 experiments, total 66 specimens (together with punctatulus). m. Temperature — humidity — light. June 28, June 30, 1945. 2 experiments, total 54 specimens (together with punctatulus). Experiment 116. Harpalus serripes. June 29, through August 9, 1945 (except 116a°). Gtl Visby, April 29, 1945 (except 116a°). a'. Temperature. July 2, 3, 5, August 1, 1945. 4 experiments, total 83 specimens (three times together with melleti, twice with rupicola, once each with punctatulus and seladon). a?. July 25, 1946. Old Greby, June 11-15, 1946. 2 experiments (with slightly moist sawdust), total 60 specimens (together with azureus and mellett). b. Humidity. June 30, August 8, 1945. 2 experiments, total 42 specimens (together with melleti, once each with punctatulus and rupicola). c. Light. June 29, August 8. 2 experiments, total 42 specimens (together with melleti, once with rupicola). d. Temperature + humidity. July 2, 31, August 1. 3 experiments, total 63 specimens (twice along with melleti, once each with punctatulus, rupicola, seladon). e. Temperature — humidity. July 5, August 2. 2 experiments, total 43 specimens (together with melleti, once each with rupicola and seladon). f. Temperature + light. July 5, 31. 2 experiments, total 43 specimens (to- gether with melleti, once each with rupicola and seladon). 102 85 g. Temperature — light. July 5, 31. 2 experiments, total 43 specimens (together with melleti, once each with rupicola and seladon). h. Humidity + light. July 1, August 9. 2 experiments, total 42 specimens (together with melleti, once with rupicola). i. Humidity — light. July 1, August 9. 2 experiments, total 42 specimens (along with melleti, once with rupicola). j. Temperature + humidity + light. July 4, August 2. 2 experiments, total 43 specimens (along with melleti, once each with rupicola and seladon). k. Temperature + humidity — light. July 4, August 2. 2 experiments, total 42 specimens (together with melleti, once with rupicola). 1. Temperature — humidity + light. July 5, August 2. 2 experiments, total 43 specimens (together with melleti, once with rupicola). m. Temperature — humidity — light. July 3, August 1. 2 experiments, total 43 specimens (together with melleti, once each with rupicola and seladon). Experiment 117. Harpalus smaragdinus. July 14-25, 1946. Old Stora-Rör, June 13, 1946. al. Temperature. July 24, 25. 2 experiments, total 61 specimens (together with anxius and tardus). a’. July 22, 24. 2 experiments (with slightly moist sawdust), total 60 spe- cimens (together with anxius and tardus). b. Humidity. July 24, 25. 2 experiments, total 60 specimens (together with anxius and tardus). c. Light. July 16, 18. 2 experiments, total 60 specimens (together with rubripes and tardus). d. Temperature + humidity. July 14, 19. 2 experiments, total 60 specimens (together with anxius and tardus). e. Temperature — humidity. July 16. 2 experiments, total 60 specimens (together with rubripes and tardus). f. Temperature + light. July 16, 18. 2 experiments, total 60 specimens (together with rubripes and tardus). g. Temperature — light. July 14. 2 experiments, total 60 specimens (to- gether with anxius and tardus). h. Humidity + light. July 19, 22. 2 experiments, total 60 specimens (to- gether with anxius and tardus). i. Humidity — light. July 21, 24. 2 experiments, total 60 specimens (to- gether with anxius and tardus). j. Temperature + humidity + light. July 14, 19. 2 experiments, total 60 specimens (together with anxius and tardus). k. Temperature + humidity — light. July 19, 22. 2 experiments, total 60 specimens (together with anxius and tardus). l. Temperature — humidity + light. July 16. 2 experiments, total 60 spec- imens (together with rubripes and tardus). 103 86 m. Temperature — humidity — light. July 21, 22. 2 experiments, total 60 specimens (together with anxius and tardus). Experiment 118. Harpalus tardus. July 14-25, 1946. Old and Gtl June 14-18, 1946. a'. Temperature. July 24, 25. 2 experiments, total 48 specimens (together with anxius and smaragdinus). a’. July 22, 24. 2 experiments (with slightly moist sawdust), total 49 spe- cimens (together with anxius and smaragdinus). b. Humidity. July 24, 25. 2 experiments, total 48 specimens (together with anxius and smaragdinus). c. Light. July 16, 18. 2 experiments, total 54 specimens (together with rubripes and smaragdinus). d. Temperature + humidity. July 14, 19. 2 experiments, total 54 specimens (together with anxius and smaragdinus). e. Temperature — humidity. July 16. 2 experiments, total 54 specimens (together with rubripes and smaragdinus). f. Temperature + light. July 16, 18. 2 experiments, total 54 specimens (together with rubripes and smaragdinus). g. Temperature — light. July 14. 2 experiments, total 54 specimens (to- gether with anxius and smaragdinus). h. Humidity + light. July 19, 22. 2 experiments, total 51 specimens (to- gether with anxius and smaragdinus). i. Humidity — light. July 21, 24. 2 experiments, total 50 specimens (to- gether with anxius and smaragdinus). j. Temperature -, humidity + light. July 14, 19.2 ea total 53 specimens (together with anxius and smaragdinus). k. Temperature + humidity — light. July 19, 22; 2 experiments, total 51 specimens (together with anxius and smaragdinus). 1. Temperature — humidity + light. July 16. 2 experiments, total 54 spe- cimens (together with rubripes and smaragdinus). m. Temperature — humidity — light. July 21, 22. 2 experiments, total 50 Specimens (together with anxius and smaragdinus). Distribution in the linear universal gradient apparatus without the influence of measurable factors (control experiment) Experiment 119. Harpalus melleti, punctatulus, seladon and serripes. August 11-13, 1945. Gtl Visby, April 28-30, 1945; except seladon, which was collected in July 1945 near Upl Varmdon, Vasterskagga. Exposure 2 hours, 2 species tested every time with 20 specimens of each. All species tested 5 times, thus total 100 specimens per species. Dry sawdust in the boxes. Diagram 15, p. 91. 104 87 Control experiments in the circular universal gradient apparatus Only experiments covered in the text are mentioned here (Diagrams 17, 18, p- 95). Experiment 120. July 10-11, 1946. a. Harpalus melleti. Temperature — humidity. Gtl Visby, June 17, 1946. 2 experiments, July 10-11, total 80 specimens. b. Harpalus punctatulus. Temperature + humidity. Gtl Visby, June 17, 1946. 2 experiments, July 11, total 80 specimens. c. Harpalus rubripes. Temperature — humidity. Öld Greby, June 11-15, 1946. 2 experiments, July 10-11, total 80 specimens. d. Harpalus smaragdinus. Temperature — humidity. Öld Stora-Rör, June 13, 1946. 2 experiments, July 10, total 80 specimens. IV. Food Preference Only one species, Amara ingenua, was experimentally studied in this connec- tion. One animal was placed in each glass dish with moist filter paper, where it was simultaneously provided with different seeds and fruits, in one case also animal food. The signs of feeding were recorded every morning. Experiment 121. September 5-9, 1946. Upl Solna, Bergshamra, August 30, 1946. 8 specimens isolated from one another in glass dishes and provided with seeds and fruits of 13 different species of plants. Table 33, p. 539. Experiment 122. September 18-19, 1946. The same 8 specimens. Each one of which was provided with 3 fruits of Polygonum aviculare and in addition 3 young pods each of Capsella and Erysimum cheiranthoides; p. 539. Experiment 123. September 21-22, 1946. 4 of the above specimens were each provided with 3 fruits of Polygonum aviculare and in addition a crushed insect of their own species; p. 540. B. EXPERIMENTS ON RESISTANCE I. Low Temperatures The animals were cooled in a cooling chamber until complete cessation of all movement, and then (by disconnecting the electricity) very slowly warmed up (about 1°C in 10 minutes). All animals were initially placed on their backs and the temperature was noted when each individual reverted to its normal position. In several cases the temperature at which the first signs of life were noticed was also recorded. Observation was carried out through the front window of the refrigerator. The thermometer was placed in the glass dish (the bottom was covered with filter paper) containing the animals. These experiments were carried out following Krogerus’ model (1932, pp. 145-146; 1937; p. 306). However, I have not taken into account his third 105 88 response point (“full activity of the experimental animals”) because I found it impossible to determine this exactly. Besides, “Point 1,” initial movement, is rather vague, especially in the case of small species, since this often involves the slightest quivering of the palps or tarsi. Due to uneven refraction through the glass panel the slightest movement of the observer’s head can give the impression of such a micromovement. It also happened that an animal that had been lying apparently lifeless suddenly turned over and immediately showed full activity. The “point of turnover” is undoubtedly the most important, for it is an expression of the lower thermic level of activity of the animal. The two most important sources of error in these experiments are: 1. The animals frequently show different individual responses (more with respect to “the first movement” than to the “point of turnover”). The cool- ing occasionally causes death; in other cases there is a distinct shock effect (paralysis of the extremities). 2. The velocity of and the amount of cooling, which, on the one hand, depend on the starting temperature and, on the other hand, the reduction necessary to reach the condition of immobility, the cooling time may cause different intensities of paralysis in the animals. It is clear that in these experiments, it is as impossible to obtain absolute values for each species as with the temperature gradient apparatus. The “point of turnover” after a brief cooling (as in these experiments) probably lies below the actual temperature of the lower limit of activity at the time of waking up of the animal in spring. It is therefore once again advisable to seek only comparative values, i.e., the experiments must always proceed simultaneously (or at any rate as far as possible under identical conditions) with 2 or more species for the purpose of comparison. That the values obtained with extensive material at least do not seem to coincide too much has been shown by their identity in the case of Brachynus versus Agonum dorsale (p. 61). All the experiments were carried out in the “Statens Vaxtskyddanstalt” (State Institute of Plant Protection), Stockholm. Harpalus, see Diagram 20 (p. 130); Cymindis, see p. 139. Experiment 124. June 27, 1946. Room temperature 21.1°C. a. Agonum dorsale. Old Greby, June 11-15, 1946. 2 specimens; p. 61. b. Amara lucida. Old Greby, June 11-15, 1946. 3 specimens (+ 3.5, 5.4, 6.0 and 7.6, 7.6, 7.8°C respectively) *. c. Brachynus crepitans. Upl Lovon, May 1946. 2 specimens; p. 61. d. Calathus melanocephalus. Old Greby, June 11-15, 1946. 3 specimens (—1.3, +1.0, 1.0 and 4.9, 5.6, 5.7°C respectively)”. e. Calathus mollis. Old Greby, June 11-15, 1946. 3 specimens (+ 1.2, 3.9, 6:Ovand 6.2, 7.1,72:72@ respectively). *Not treated in the text. 89 Experiment 125. June 28, 1946. Room temperature 20.4-20.8°C. a. Cymindis angularis. Öld Greby, June 11-15, 1946. 3 specimens. b. Cymindis humeralis. Öld Greby, June 11-15, 1946. 2 specimens. c. Cymindis macularis. Öld Stora-Rör, June 13, 1946. 3 specimens. d. Harpalus aeneus. Öld Greby, June 11-15, 1946. 3 specimens. e. Harpalus azureus. Öld Greby, June 11-15, 1946. 3 specimens. f. Harpalus melleti. Gtl Visby, June 19, 1946. 3 specimens. g. Harpalus punctatulus. Gtl Visby, June 18, 1946. 3 specimens. h. Harpalus puncticeps. Öld Halltorp, June 16, 1946. 2 specimens. i. Harpalus rupicola. Gtl Visby, June 18, 1946. 3 specimens. j. Harpalus seladon. Old Greby, June 11-15, 1946. 3 specimens. Experiment 126. June 29, 1946. Room temperature 20.1-20.3°C. a. Harpalus hirtipes. Öld Stora-Rör, June 13, 1946. 3 specimens. b. Harpalus neglectus. Öld Stora-Rör, June 13, 1946. 3 specimens. c. Harpalus rubripes. Öld Greby, June 11-15, 1946. 1 specimen. d. Harpalus rufitarsis. Öld Stora-Rör, June 13, 1946. 3 specimens. e. Harpalus tardus. Old and Gtl, June 11-18, 1946. 3 specimens. f. Harpalus vernalis. Öld Stora-Rör, June 13, 1946. 3 specimens (+0.9, 3.6, 3.8 and 6.9, 7.0, 7.6°C respectively)*. Experiment 127. July 1, 1946. Room temperature about 21°C. a. Cymindis humeralis. Old Greby, June 11-15, 1946. 1 specimen. , b. Cymindis macularis. Old Stora-Rör, June 3, 1946. 1 specimen: c. Harpalus aeneus. Old Greby, June 11-15, 1946. 2 specimens. d. Harpalus anxius. Old Stora-Rör, June 13, 1946. 3 specimens. e. Harpalus rubripes. Old Greby, June 11-15, 1946. 3 specimens. f. Harpalus serripes. Old Greby, June 11-15, 1946. 3 specimens. g. Harpalus smaragdinus. Old Stora-Ror, June 13, 1946. 3 specimens. Experiment 128. June 26, 1947. Old Greby, June 12-22, 1947. Room tem- perature about 23°C. a. Agonum dorsale. 10 0‘, 10 2 (recorded separately); p. 61. b. Brachynus crepitans. 10 9‘, 10 2 (recorded separately); p. 61. c. Calathus mollis, macropterous form. 5 9, 5 2 (recorded separately); p. 358. d. Calathus mollis, brachypterous form. 5 0‘, 5 ? (recorded separately); P..398. e. Harpalus azureus. 5 9, 5 ? (recorded separately); p. 191. Experiment 129. June 28, 1947. Old Greby, June 12-22, 1947. Room tem- perature about 23°C. a. Agonum dorsale. 10 9°, 10 (recorded separately); p. 61. b. Brachynus crepitans. 10 7, 10 2 (recorded separately); p. 61. c. Harpalus azureus. 5 9‘, 5 $ (recorded separately); p. 191. *Not treated in the text. 107 90 Experiment 130. September 19, 1947. Room temperature 23°C. a. Bradycellus collaris, macropterous form. 20 specimens. Jtl Revsund, August 31, 1947. p. 359. b. Bradycellus collaris, brachypterous form. 20 specimens. Jtl Revsund, August 31, 1947. p. 359. N c. Lionychus quadrillum. 3 specimens. Nke Orebro, August 20, 1947. “Point of turnover”: — 1.8, + 2.3, 3.2°C. II. High Temperatures The animals were placed in a tin box whose bottom was covered with filter paper and the top with a glass lid. A thermometer was placed at the bottom. The box was slowly warmed up (about 1°C in 2 minutes) in a water bath, and for each individual the temperatures of the first sign of paralysis and the last body movement (Points 4 and 5, Krogerus, 1932, p. 146; 1937, p. 306) were recorded. The individual variations in the above-mentioned temperatures were gen- erally not large. The chief external sources of error might be the initial tem- perature and the humidity of the air. However, all the experiments were car- ried out in the same room in July at a temperature of 23.8-26.2°C. In the relative humidity of the air greater variation than 60 to 70% could not have occurred. Concerning the species of Harpalus, see Diagram 20 (p. 130); Cymindis, p. 139. Experiment 131. July 24, 1946. Room temperature about 26°C. The figures obtained are not treated in the text. a. Amara equestris. Old Greby, June 11-15, 1946. 2 specimens (43.8, 43.8 and 45.7, 46.1°C respectively). b. Calathus mollis. Old Greby, June 11-15, 1946. 3 specimens (36.3, 38.8, 40.3 and 42.7, 43.6, 43.7°C respectively). c. Panagaeus bipustulatus. Old Greby, June 11-15, 1946. 3 specimens (46.3, 46.3, 46.9 and 47.6, 47.6, 48.6°C, respectively). Experiment 132. July 27, 1946. Room temperature 25.5-25.9°C. a. Harpalus azureus. Old Greby, June 11-15, 1946. 3 specimens. b. Harpalus melleti. Gt! Visby, June 19, 1946. 3 specimens. c. Harpalus punctatulus. Gtl Visby, June 19, 1946. 3 specimens. d. Harpalus rupicola. Gtl Visby, June 19, 1946. 3 specimens. e. Harpalus seladon. Gtl Visby, June 19, 1946. 3 specimens. f. Harpalus serripes. Öld Greby, June 11-15, 1946. 3 specimens. Experiment 133. July 28, 1946. Room temperature 25.6-26.2°C. a. Harpalus aeneus. Old Greby, June 11-15, 1946. 3 specimens. b. Harpalus anxius. Öld Stora-Rör, June 13, 1946. 3 specimens. 108 91 c. Harpalus puncticeps. Öld Halltorp, June 16, 1946. 3 specimens. d. Harpalus vernalis. Old Greby, June 11-15, 1946. (48.1, 48.7, 48.8 and 49.1, 49.2, 49.3°C respectively)*. Experiment 134. July 28, 1946. Room temperature 23.8°C. a. Cymindis angularis. Old Greby, June 11-15, 1946. 3 specimens. b. Cymindis humeralis. Old Greby, June 11-15, 1946. 3 specimens. Experiment 135. July 29, 1946. Room temperature 23.9°C. a. Harpalus hirtipes. Old Stora-Rör, June 13, 1946. 3 specimens. b. Harpalus rubripes. Old Greby, June 11-15, 1946. 3 specimens. Experiment 136. July 30, 1946. Room temperature 24.6-24.9°C. a. Amara lucida. Old Greby, June Ur -15, 1946. 3 specimens (46.4, 46.7, 47.6 and 47.6, 48.3, 48.6°C respectively)”. b. Cymindis macularis. Öld Stora-Rör, June 13, 1946. 3 specimens. c. Harpalus aeneus. Old Greby, June 11-15, 1946. 3 specimens. d. Harpalus azureus. Old Greby, June 11-15, 1946. 3 specimens. e. Harpalus hirtipes. Old Stora-Rör, June 13, 1946. 3 specimens. f. Harpalus neglectus. Old Stora-Rör, June 13, 1946. 3 specimens. g. Harpalus rufitarsis. Öld Stora-Rör, June 13, 1946. 3 specimens. h. Harpalus smaragdinus. Öld Stora-Rör, June 13, 1946. 3 specimens. i. Harpalus tardus. Old and Gtl, June 11-19, 1946. 3 specimens. ı Experiment 137. July 10-11, 1947. Room temperature 26°C. Old Greby, June 12-23, 1947; p. 61. a. Agonum dorsale. 10 o (2 experiments). b. Agonum dorsale. 10 2 ( 2 experiments). c. Brachynus crepitans. 10 o (2 experiments). d. Brachynus crepitans. 10 2 (2 experiments). III. Drought The animals were placed, each species separately (occasionally form or sex as well), in a small completely dry glass dish with filter paper at the bottom, and the maximum life-span of each individual without water and food was recorded. Observations were recorded every 4 to 9 hours. An important source of error is undoubtedly the humidity of the air. However, since the experiments were carried out in the room during the period from June 26 through August 22, the values obtained might be comparable. Experiment 142 (September 23 through October 1) was an exception; but the purpose here was to exclusively compare the 2 forms of this species. Experiment 138. From July 25, 1946 onward. Diagram 23, p. 135 (with the exception of Amara, Cymindis, and Harpalus vernalis). *Not treated in the text. 92 a. Amara lucida. Öld Greby, June 11-15, 1946. 3 specimens (58, 91, and 106 hours respectively)*. b. Cymindis angularis. Öld Greby, June 11-15, 1946. 3 specimens; p. 139. c. Cymindis humeralis. Öld Greby, June 11-15, 1946. 3 specimens; p. 139. d. Harpalus aeneus. Old Greby, June 11-15, 1946. 3 specimens. e. Harpalus anxius. Old Stora-Ror, June 13, 1946. 3 specimens. f. Harpalus azureus. Old Greby, June 11-15, 1946. 3 specimens. g. Harpalus melleti. Gtl Visby, June 19, 1946. 3 specimens. h. Harpalus punctatulus. Gtl Visby, June 19, 1946. 3 specimens. i. Harpalus puncticeps. Old Halltorp, June 16, 1946. 3 specimens. j. Harpalus rubripes. Old Greby, June 11-15, 1946, 3 specimens. k. Harpalus rupicola. Gtl Visby, June 19, 1946. 3 specimens. l. Harpalus seladon. Gtl Visby, June 19, 1946. 3 specimens. m. Harpalus serripes. Öld Greby, June 11-15, 1946. 3 specimens. n. Harpalus vernalis. Old Greby, June 11-15, 1946. 3 specimens (45, 58 and 125 hours respectively)”. Experiment 139. From August 1, 1946 onward. Diagram 23, p. 135 (without Cymindis). a. Cymindis macularis. Old Stora-Rör, June 13, 1946. 3 specimens; p. 139. b. Harpalus hirtipes. Old Stora-Rör, June 13, 1946. 3 specimens. c. Harpalus neglectus. Old Stora-Rör, June 13, 1946. 3 specimens. d. Harpalus rufitarsis. Old Stora-Rör, June 13, 1946. 3 specimens. e. Harpalus smaragdinus. Old Stora-Rör, June 13, 1946. 3, specimens. f. Harpalus tardus. Old and Gtl, June 11-19, 1946. 3 specimens. Experiment 140. From June 26, 1947 onward. Calathus mollis. Öld Greby, June 12-22, 1947, p. 359. a. Macropterous form. 4 0’, 4 $ (recorded separately). b. Brachypterous form. 4 0, 4 (recorded separately). Experiment 141. From July 10, 1947 onward. Old Greby, June 12-22, 1947; p- 61. a. Agonum dorsale.5 9' + 5 9,5 9° +5 (2 experiments). b. Brachynus crepitans. 5 3 + 5 9,5 0' + 5 ? (2 experiments). Experiment 142. From September 23, 1947 onward. Bradycellus collaris. Jtl Revsund, August 31, 1947, p. 359. a. Macropterous form, 10 specimens. b. Brachypterous form, 10 specimens. IV. Water The few experiments on exposure to water were carried out in small vertical test tubes which were filled up to two-thirds with spring water. In each tube *Not treated in the text. 109 110 93 there was aiways une animal. Each time two series of experiments were under- taken: in one tne animals were allowed to swim undisturbed on the surface, in the other the test tube was inverted three times a day in order to simulate to some extent the natural conditions during transport in water. The temperature of the air in the room was high (20-25°C), and therefore the values obtained on the duration of life are undoubtedly lower than in similar situations in nature. Cymindis, see p. 248. Experiment 143. Cymindis angularis. From June 25, 1947 onward. Old Greby, June 12-23, 1947. a. 3 specimens. Left undisturbed. b. 3 specimens. Regularly shaken. Experiment 144. Cymindis humeralis. From June 25, 1947 onward. Old Greby, June 12-23, 1947. a. 3 specimens. Left undisturbed. b. 3 specimens. Regularly shaken. Experiment 145. Cymindis macularis. From June 25, 1947 onward. Old Stora-Ror, June 14, 1947. a. 3 specimens. Left undisturbed. b. 3 specimens. Regularly shaken. Experiment 146. Broscus cephalotes. From August 16, 1947 onward. NI Tvarminne, August 12, 1947, p. 599. a. 2 specimens. Left undisturbed. b. 2 specimens. Regularly shaken. C. EXPERIMENTS ON THE DIRECTION OF FLIGHT A special apparatus was constructed whose appearance and dimensions are shown in Fig. 8 (see also Lindroth, 1948d). The lower part shown in the Figure was covered with a tight-fitting 45 cm high bell jar covered with cellophane. The principle involved is that beetles flying out from a glass dish in the center of the apparatus would collide with the transparent walls of the bell and fall into the corresponding funnels, whose spouts were directed into a glass jar. It was intended to study the influence of the position of the sun on the flight of the insects. Therefore all the experiments were carried out around 6 p.m. (5-7 p.m.) and one of the sectors was directed towards the sun, i.e. west- ward. Due to the distribution of the animals in different sectors the direction of flight was recorded by constructing “flight paths” (“Flugrosen”) (Fig. 21, p. 255). See also Table 19 (p. 257). Experiment 147. Acupalpus consputus. Old Halltorp, June 13-15, 1947 (about 60 specimens). All experiments carried out near Old Greby, June 19-23, 5:30-7:00 p.m., total 200 observations. Experiment 148. Acupalpus dorsalis. 94 Fig. 8. Lower part of the flight direction apparatus. Experiment 147 onward. a. Öld Möckelmossen, June 18, 1947 (8 specimens). Experiments near Öld Greby, June 19-23, 1947. 5:30-7:00 p.m., total 14 observations. b. NI Tvärminne, August 10-11, 1947 (about 70 specimens). Experiment in Djursholm, August 14, 5:15-6:00 p.m. (sunny, followed by overcast sky), total 77 observations. Experiment 149. Badister dilatatus. Öld Halltorp, June 13-15, 1947 (about 30 specimens). Experiments carried out in Old Greby, June 19-23, 5:30-7:00 p.m., total 34 observations. Experiment 150. Badister peltatus. Old Halltorp, June 13-15, 1947 (12 spec- imens). Experiments carried out in Öld Greby, June 19-23, 5:30-7:00 p.m., total 15 observations. y Experiment 151. Badister unipustulatus. Old Halltorp, June 13-15, 1947 (15 specimens). Experiments conducted in Old Greby, June 19-23, 5:30-7:00 p-m., total 27 observations. aa 152. Oodes gracilis. Upl Djursholm, Ösby Lake, May 31, 1947 (about 50 specimens). The experiments were carried out in Djursholm, June 8, Gtl Visby, June 10, Old Greby, June 12-18, 6:00-6:30 p.m., total 100 ob- servations. Fig. 21. D. EXPERIMENTS WITHOUT INSECTS All these experiments are related to the studies on “limestone species” (pp. 112 ff.). They were carried out with the purpose of comparing the character- istics of the limestone rock and the basement complex (besides the loose de- posits Originating from these). Since these experiments are discussed in detail in the running text only brief notes are given here. Unless otherwise mentioned 111 95 these experiments were carried out in Djursholm. Experiment 153. Daily temperature conditions on underside of pieces of lime- stone and granite. Diagram 36, p. 179. Studies carried out on the roof of the Institute in Djursholm from July 1, 1945, 5 a.m., through July 2, 1945, 6 a.m. Detailed account on p. 178. Experiment 154. Daily temperature conditions in small volumes of limestone gravel and siliceous gravel. Diagram 37, p. 180. Simultaneously realized with Experiment 153. Description of gravel as on p. 77. Temperature recorded in 4 samples, each in a 400 cm? zinc box, to which 100, 50, 25, and 0 cm? of water was added respectively. Description on p. 178. Experiment 155. Temperature conditions in limestone and siliceous gravels over longer period of time. September 4 through October 29, 1945. Same boxes and categories of gravel as in Experiment 154. But a minimum thermometer was used, inserted horizontally through the walls of the boxes. The boxes were placed in the open but protected from rain by an overhanging roof. The minimum temperature was read every day and usually also at 11 a.m., 4 p.m., and 9 p.m.; p. 178. 3 Experiment 156. Temperature conditions in limestone and siliceous gravels with heating. The same methods as in Experiment 155. a. February 10, 1946. The boxes remained in the open for several months. On February 10, with an outside temperature of —10.7°C, they were abruptly brought into a room at a temperature of +17.6°C and the temperature of the gravel was recorded regularly (every minute during the first hour-and-a- quarter). The gravel was apparently completely dried and hence considerable condensation of water must have taken place. Diagram 38, p. 182. b. March 3, 1947. Repetition of a). Outside temperature —9.3°C, room temperature +18.2°C. The boxes were cooled only for one day prior to the experiment and the gravel was apparently not quite dry. The curves did not intersect. p. 181. : c. March 7, 1947. Outside temperature —11.0°C, room temperature +17.1-17.4°C. Repeated as in b) with the same result. p. 181. Experiment 157. Temperature conditions in limestone and siliceous gravels during strong cooling. March 14-15, 1946. Opposite to Experiment 156. The boxes were kept inside the room from February 10 through March 14. Starting temperature +18.1°C. Outside temperature —6.1 to —16.1°C (at the end of the experiment at 1:35 a.m.). Since the air temperature fell constantly during the experiment (minimum night temperature: —19.0°C), the curves did not meet. Diagram 39, p. 182. Experiment 158. Evaporation of water from limestone gravel and siliceous gravel. September 8 through October 3, 1946. The same boxes and categories of gravel as in Experiments 154-157. To the volumes of limestone and siliceous gravel (400 cm? each) 100 cm? of distilled water was added, respectively. The boxes were placed in a dry, shady location in the room (central heating) and 96 were weighed every day. The decrease in weight was equated wıth the evapo- ration of water. Diagram 41, p. 188. Experiment 159. Daily temperature conditions in soil on limestone and on granite. Dir Rattvik, from 1700 hours on June 4 through 7 p.m. on June 5, 1946. The observation area situated near Sjurberg is described in detail (p. 181) and illustrated with drawings and a photograph (Figs. 14-16). The vegetation on both sides of the fault fissure (orthoceratite limestone as against granite) was similar, meadow type; Primula farinosa and Carex ornithopoda were found on both sides of the fissure; the moss Thuidium abietinum (det. Frisendahl) also requires limestone. —The day before was quite rainy. The soil was therefore unusually moist, which certainly limited the extreme values of temperature. Diagram 40 (p. 185). 112 113 The “Limestone Species” An Example of the Influence of Existence Factors The great importance of limestone (CaCO,) for plants has been known for a long time, and there is a very extensive literature (among others: Thurmann, 1849; Kraus, 1911; Warming and Graebner, 1918, p. 118 ff.; Frodin, 1919; Salis- bury, 1920; Tamm, 1921; Mevius, 1921, 1924; Th. Fries, 1925; Arrhenius, 1926; Pesola, 1928; Brenner, 1930; Lundegärdh, 1930, p. 335 ff.; Eklund, 1931, 1933). Many animals, including insects, are often commonly named “limestone species” or at least “limestone favored” species, depending on their more or less pronounced regular occurrence on calcareous soil. In Part I of this work, such contentions or assumptions on my own part or others’ are mentioned with the relevant species. In Fennoscandia the occurrence of the exposed limestone rock is ex- tremely limited*, and the more or less limestone-rich loose deposits (especially moraine and marl) are very unevenly distributed (maps in Figs. 9 and 10). The species associated with limestonet, if such species occur at all, must therefore occupy a corresponding area. Actually to date it is almost always on the ba- sis of such comparisons and considerations that certain insects are designated “limestone species” in Fennoscandia (and usually elsewhere). On this basis, there are primarily two geographical groups in our area: 1. Species which in Scandinavia occur in Oland and Gotland: a) restricted to these islands; and b) with their northern limit in one of these islands. Ex- amples: a) Harpalus azureus, Harpalus punctatulus, Harpalus rupicola**; b) Cy- mindis humeralis, Harpalus melancholicus, Harpalus melleti, Harpalus neglectus, Harpalus picipennis, Harpalus serripes, Harpalus servus, Harpalus vernalis, Ma- soreus wetterhalli, Pterostichus punctulatus. (cf. p. 455 and species mentioned on following pages). *Siliceous rocks with a greater or lesser limestone content which (as in Varmland) exercise the same influence as limestone, although to a lesser extent (for example, Tamm, 1921, p. 112; Saxén, 1928), are not referred to in the map (Fig. 9). t(= “calciphilous” species; cf. p. 816; suppl. scient. edit.). **Besides these three species of Harpalus, the only other Scandinavian carabid which has been found only in Oland and (or) Gotland, is Calosoma investigator (one specimen). 4 Crecaceous Kambrosilur a Cambro- Silurean N Primitive N limestone Archacic 114 Fig. 9. Compact limestone rock as superstratum in Sweden. Cf. footnote on p. 113. From appendix to “Kalkutredningen” (Sveriges Geol. Unders., 1931; unpublished). Fig. 10. Loose deposits containing limestone in Sweden. After Lundqvist 115 (1925, p. 117; 1943). Actual limits in nature are not so distinct. 116 100 2. Species that show a relict-like restricted distribution in or on the small isolated Cambrian-Silurian areas of central Sweden. Examples: Harpalus anxius, Harpalus rufitarsis, Microlestes minutulus, Panagaeus bipustulatus; possibly also Acupalpus consputus, Leistus rufomarginatus, Microlestes maurus. In Norway the sirangely isolated occurrence of Abax ater represents a similar case. On the basis of the distribution pattern alone it is of course not possible to prove that a particular species is associated with limestone, still less to establish which factors operate in this regard. For this reason I decided to study the problem experimentally. Formulation of the problem was simple: it involves the following considerations: 1. The species in question were to be examined as thoroughly as possible for their ecology, comparing them with related species with a different mode of life, and also with geographically and ecologicaily more or less pronounced cosmopolitan species. It was necessary to find any characteristic requirements for life common to the so-called limestone species. 2. The limestone rock (especially the Cambro-Silurian) and the limestone- rich loose deposits had to be studied for their chief chemical and physical characteristics, in comparison with siliceous rock (basement complex), which is dominant in Fennoscandia, and siliceous soil. 3. It should be possible to determine the influence of limestone on the animals by comparing the results of 1 and 2. If it was possible to correlate the characteristics of “limestone species” and calcareous soil, a causal relationship could be established. Characteristics of the “Limestone Species” For several reasons (lack of time, material, etc.) it was impossibie to investigate experimentally the limestone requirement of all species. With this limitation it was found advantageous to work with species as closely related as possible, of which it might be assumed that they would not show excessive physiolog- ical differences. On the other hand these must include both the well-defined “limestone species” of different types and the ecologically and geographically more or less distinct ubiquists. These considerations revealed that the genus Harpalus would be by far the best for a solution of the problem in hand, the more so because all the three carabid species restricted to Oland and Gotland belong to this genus (subgenus Ophonus). In the case of some rare species it was impossible to gather sufficient live material. It was possible to get live material of 15 species, which can be divided into the following 6 groups on the basis of their distribution in Scandinavia (for further details see the maps in Part II): 1. Species restricted to Oland and Gotland: Harpalus azureus 117 101 Harpalus punctatulus Harpalus rupicola 2. Species whose northern limit in Scandinavia is Old or Gtl: Harpalus hirtipes“ Harpalus melleti Harpalus neglectus Harpalus Puncticeps“ Harpalus serripes* 3. Species with relict-like occurrence in the central Swedish limestone areas: Harpalus anxius Harpalus rufitarsis 4. Markedly southern species, but continuously distributed up to northern Upl: Harpalus rubripes Harpalus smaragdinus 5. Species with their northern limit in southern Norrland (Mdp): Harpalus seladon Harpalus tardus 6. Geographical ubiquist (missing only in the fjelds): Harpalus aeneus. The presumed “limestone species” are the 10 species of the first three groups (marked “K” in the reports on experiments in the following pages). The remaining 5 species will be taken into consideration for purposes of com- parison. There was also sufficient material of Harpalus vernalis (northern limit in Gtl), but this small, sluggish species proved unsuitable for my experiments with the gradient apparatus and is therefore not treated below. All the species mentioned are animals of open terrain and more or less markedly xerophilous**. The few species of Harpalus that deviate ecologically to the extent that they thrive in shady and somewhat moist places (latus, luteicornis, quadripunctatus, winkleri), are not considered here, being of little interest. On the other hand some experiments were carried out with other “lime- stone species” (along with other species for comparison), including Amara lucida, species of Cymindis, and Panagaeus bipustulatus, some of which are described beiow. For an idea of the typical biotopes of the species treated here see the two photographs. *Only a single more or less accidental locality is known more northerly on the Swedish mainland. **For Harpalus aeneus, see Palmen and Suomalainen (1945). 102 117 Fig. 11. Öld Räpplinge, southern margin of Greby Alvar, seen from the north. Locality of Harpalus azureus, H. serripes and other “limestone species.” (Photo: June 15, 1946). The first of these (Fig. 11) is from Old Greby (Rapplinge parish) and 118 shows the southern edge of the Greby Alvart. The stony soil on top of the rocky outcrop of the mountain is about 2 dm thick; pH 7.9 (electrometrically measured). The vegetation is very low (partly because of sheep grazing), but is dense and rich in species. The following plants are prominent (regardless of their abundance): Achillea millefolium; Bellis; Calamintha acınos; Dianthus deltoides; Erodium; Festuca ovina and F. rubra; Fragaria viridis; Galium verum; Herniaria; Heiracium pilosella; Plantago lanceolata; Poa alpina; Potentilla ar- gentea; Sagina procumbens; Sedum acre and S. album; Thymus serpyllum; Tri- folium procumbens and T. repens. Solitary shrubs of Juniperus, Prunus spinosa, Crataegus, and Rosa. Of the species used in experiments the following species lived here: Harpalus aeneus, H. azureus, H. puncticeps (1 specimen), H. rubripes, H. serripes, H. smaragdinus (2 specimens); Amara lucida, Cymindis humeralis (together with C. angularis), Panagaeus bipustulatus. Other possible “limestone species” are Agonum dorsale, Brachynus, Calathus mollis, Harpalus vernalis, Ne- 119 bria salina. The curculionid Lepyrus capucinus Schall. was common. The second photograph (Fig. 12) shows the “locus classicus” (“typical locality”) of Harpalus rupicola (Part I, p. 509) near Gil Visby, a scree slope tA type of rendzina; suppl. gen. edit.). 118 103 Fig. 12. Gtl Visby. Railroad track embankment south of town, view from the south. Locality of Harpalus melleti, H. punctatulus, H. rupicola, and other “limestone species.” (Photo: June 9, 1947). formed by falling weathered gravel, used for a railroad track embankment. The gravel is extremely coarse, mixed with larger stones, and has very high porosity between the particles; pH 7.4 (electrometrically determined). The vegetation is rather sparse but tall. Dominant flora: Arrhenatherum elatius,; Centaurea scabiosa, Dactylus glomerata, Daucus, Hieracium sp., Medicago sativa; patchy distribution: Anthemis tinctoria, Artemisia campestris and vulgaris, Cichorium, Potentilla reptans, Ranunculus polyanthemus, Torilis. Shrubs of Sorbus suecica. In addition to Harpalus rupicola, also numerous here were H. melleti and H. punctatulus; H. aeneus and A. rubripes appeared only sporadically. Agonum dorsale, Amara curta, Brachynus, Dromius linearis and larvae of Licinus also occurred, among others. The mineral components of the soil in the two biotopes described above originated from limestone. The admixture of humus is small. In summer the surface layers dry up completely. 120 104 The Food Phytophagous species, dependent on one kind of limestone-bound species of plant, are thereby secondarily associated with limestone. Predatory animals that feed exclusively on phytophagous animals mav be considered as the third level of species associated with limestone. It is therefore very important to determine the diet of the suspected “limestone species.” It has been known for a long time that many species of Harpalus (and species of Amara) are strikingly polyphagous, consuming both vegetable and animal focd. As shown elsewhere (p. 531), in this respect they are not excep- tional among the carabids, as has been assumed so far, although members of the genus Harpalus (and of Amara) appear to prefer plant food more than other genera. However, they are not fastidious in this respect. The 15 species of Harpalus mentioned above (besides A. vernalis) used in my experiments, as well as the species of Cymindis (angularis, humeralis, macularis), Panagaeus bipustulatus, Brachynus, and many others, were fed exclusively on bread for many months and were found to thrive (Harpalus serripes for more than 3 years!). They did not attack living insects (not even small tender ones such as the collembolans) but greedily devoured crushed conspecific individuals and other species of Harpalus. At any rate this was established for FH. aeneus, H. anxius, H. punctatulus, H. rubripes, H. smaragdinus and H. tardus. H. serripes also consumed a dead Tenebrio and A. hirtipes, H. neglectus and H. punctatulus ate crushed flies. The last named species also accepted a dead Amara aenea. The earlier observations, given in Part I, showed pronounced and sponta- neous polyphagy of H. aeneus, H. calceatus, H. distinguendus, H. pubescens and H. puncticeps. Members of the subgenus Ophonus have been observed quite frequently on the umbels of Umbelliferae attacking unripe fruits; there was also spontaneous feeding on seeds of another kind by H. griseus. The above observations apply to aduits. Little is known about the nou- rishment of the larvae; in general they are considered polyphagous carnivores. But the larva of H. pubescens is also believed to consume vegetable food. At any rate it has been established that species of Harpalus are not de- pendent on a particular plant or a particular prey for their nourishment. Their somewhat restricted distribution, due to which it was assumed in certain cases that they require limestone may in no way be due to food habits. Preferenda Experiments with Limestone The first main experimental task naturally was to determine how the species in question react to limestone (CaCO,). The behavior of a series of species toward siliceous gravel was first studied with increasing admixtures of CaCO, (precipitated chalk) in the substratum 121 122 124 105 gradient apparatus (devised by Krogerus) (Experiment 31 ff., p. 75; Table 2)*. None of the 5 species studied revealed attraction to the chalk-rich end of the gradient apparatus. Since it might possibly be presumed that the animals are able to consume chalk only in dissolved form, the experiment was changed so that in place of pure water, water saturated with CO, was added to the soil, in which a smaller or larger amount of chalk was mixed (Experiment 32 ff., p. 76; Table 3). The same 5 species were used. The result thus obtained showed no striking difference from that obtained earlier: CaCO, on its own does not attract the species in question. But it may seem strange that the animals do not distribute themselves uniformly in the gradient apparatus and instead show a pronounced tendency to congregate in the half that is poor in chalk. The reason probably lies in the extremely small particle size of chalk, as a result of which the porosity of the substratum is reduced, and the animals have difficulty burrowing into the substratum. The experiments would therefore have required repetition after the corre- sponding correction, possibly supplemented with alternating instead of serial experiments, and with much more material. However, this proved unneces- sary aS a result of the experiments with limestone gravel and siliceous gravel discussed later. It has often been emphasized by botanists that most limestone plants do not require limestone as nutrient (e.g. Lundegardh, 1930, p. 339), but are on the other hand dependent on factors secondarily caused by limestone or at least usually characteristic of limestone-rich soil. These factors have been summed up by Brenner (1930, p. 85) under the same “complex of limestone factors.” “The most important limestone factor for most plants is undoubtedly the relatively neutral and stable reaction”! (l.c.; also cf. Arrhennius, 1926, and Lundegärdh, 1930, pp. 341 ff.). Even though high pH of the soil is not always caused by limestone, this is evidently the most important factor for the neutralization of acids and for geographical distribution of different pH values in soil that is otherwise of the same xind, coinciding to a large extent with the distribution of limestone (both as outcropping rock and in the form of loose deposits) (Fig. 13). This pH map therefore shows values on the high side (less acidity) in areas with limestone mountains (cf. map in Fig. 9, p. 114) and in those which, starting from the former, are located in the direction of the movement of inland ice, where as a result loose limestone was transported chiefly as moraine. Examples are the areas between Jtl Storsjon lake and the Bothnian sea, and the northern parts of the “Swedish highlands” (Vgl, Sma, Ögl). The relatively high pH *For further details concerning the experiments, reference may be made to the transcript of experiments (p. 66 ff.). t(=“Response”; suppl. scient. edit.). 106 121 Table 2. Distribution of 5 species of Harpalus in the substratum gradient apparatus on siliceous gravel with different admixtures of CaCO; (chalk) Percentage values refer to volumes. “K”—suspected “limestone species.” Experiment 31 ff., p. 75 Silicate Silicate ee 1.56 % 3: 13% 6) 6.25 WA 125 % 257 9%, 50% CaCOs punctatulus K* == OO serripes K a 4 seladon (6) | O | (6) | I u nn DU Ss I rupicola K 3 | o | o | I melleti K ie) (©) 16 122 Table 3. As in Table 2, but soil moistened with CO;-saturated water, Experiment 32 ff., p. 76 nee) ||(Silicate Silicate |--0.78 % 1.56 % 313% 6.25 % 12.5 % 25% ea % CaCOs CaCOs punctatulus K serripes K seladon rupicola K melleti K values of the last-named are somewhat surprising, since this region is poor in flora, primarily due to the shortage of nutrient-rich loam and not because of the pH. Fig. 13. pH of peat soil. Average values of pH of cultivated bogs. From sup- plement of “Kalkutredningen” (Sveriges Geol. Unders., 1931; unpublished). Area with higher pH in Vbt is actually somewhat larger. 125 108 At any rate it was found absolutely essential to study the behavior of our “]imestone species” against different pH values, the more so because Krogerus (1939, p. 1222 ff.) had already shown that the moor fauna comprises various insects, including Coleoptera (also carabids), that are “stenoionic,” i.e. intol- erant of large variations in soil reaction. The experiments (Experiment 41-ff., p. 76; Table 4) were carried out with the same 5 species studied with respect to CaCO,, also in the substratum gradient apparatus. The arrangement of the boxes is shown in Table 4; the animals were given the choice of 3 pH values (4.8 and 7.5, and a mixture of both, hence approximately 6.0). The result was expressed as the mean number of individuals per box. It was found that none of the 5 species was attracted to the alkaline sample (pH 7.5). This was all the more surprising in that this soil was taken from a locality of H. seladon in Gotland. Therefore at least for this species there was no question of any unfavorable influence of other chemical factors. The pH of the common locality of H. melleti, H. punctatulus and H. rupicola in Gti Visby (Fig. 12) was 7.4 and that of A. serripes in Old Greby (Fig. 11) was 7.9. The preference for the acidic samples by all species might be due to the somewhat smaller size of the particles and higher water-storage capacity. However, these differences are not large, and the experiment shows that the (chemicalt) reaction of the soil, within given limits, has no demonstrable etfect on the species tested. 124 Table 4. Distribution of 5 species of Harpalus in the substratum gradient apparatus with soil of 3 different pH values. Experiment 41 ff., p. 76 Arrangement of boxes pH > 6.0 pH 7.5 pH > 6.0 pH 4.8 pH > 6.0 Mean number of individuals pH pH pH 75 >6.0 4.8 punctatulus K 33 specimens 4.5 2.2 Ses) serripes K_ 22 specimens 3 4.3 6 seiadon 24 specimens 2 4.3 6 rupicola K 46 specimens 2 4 9 melleti K 25 specimens 1 3 12 t(Suppl. Translator). 109 The small dependence on reaction and characteristics of the ground water is now easily understandable in the case of the more or less xerophilous species. The beetles meet their requirement of water orally (by drinking or feeding), the impermeability of the cuticle of carabids being evident from the virtual absence of cuticular transpiration (Eder, 1940, p. 208). The main food of the species of Harpalus, plant seeds, are deficient in water. The losses due to transpiration must be compensated chiefly by drinking (to a lesser extent by consuming other xerophilous animals). But the surface layers of their habitats usually contain no free water, so they depend on rain and especially on nighttime dew, which, before evaporating in the morning, cannot be noticeably influenced by the high pH of the soil. Concerning the undoubted significance of pH for some hygrophilous carabids, see below (p. 527). The experiments described so far were intended to study the association of “limestone species” with the chemical characteristics of CaCO,. The result was negative. For this reason I conducted experiments (still in the substra- tum gradient apparatus) to find out whether the species in question are able to distinguish limestone gravel from other kinds of gravel. To exclude incon- sequential differences as far as possible, the particle size of the gravel used was always limited to about 3/4 to 4 mm by sifting. Since these experiments must be considered highly significant they were carried out with a consider- able number of species (10) and a larger number of specimens (mostly 100 of each species). The 2 kinds of substrata were placed alternately, with an equal number of boxes. Table 5. Distribution of some species of Harpalus, Cymindis humeralis and Panagaeus bipustulatus in the substratum gradient apparatus on limestone (“Kalk”) gravel and siliceous gravel. Experiment 47 ff., p. 77 Limestone _Siliceous : gravel gravel H. punctatulus K 40 60 = 100 specimens H. serripes K 40 60 = 100 specimens H. rubripes 41 59 = 100 specimens Cym. humeralis K 25 75 = 100 specimens H. azureus K 76 24 = 100 specimens H. anus K 80 20 = 100 specimens H. rupicola K 81 19 = 100 specimens H. melleti K 83 17 = 100 specimens H. rufitarsis K 21 (62%) 13(38%) = 34 specimens Pan. bipustulatus K 32 (82%) 7 (18%) = 39 specimens 126 127 128 110 First, the distribution on limestone gravel and siliceous gravel was studied (Experiment 47 ff., p. 77; Table 5). On the basis of the experiment, the species studied can be divided into two distinct groups. The first 4 species in the Table (Harpalus punctatulus, H. rubripes, H. serripes, Cymindis humeralis) showed no attraction to limestone. On the contrary, at least 59% of the individuals preferred the boxes with siliceous gravel. In contrast, the remaining species (Harpalus azureus, H. anxius, H. melleti, H. rufitarsis, H. rupicola, Panagaeus bipustulatus) exhibited a clear preference for limestone; 62-83% of the individuals were found in limestone. Harpalus punctatulus and H. serripes thus seem to be excluded from the species suspected of being “limestone species,” which is in fact not surprising. The former has been found during recent years at many places in southern Finland, including the limestone-deficient parts of the southeast (see map in Eskola, Hackman, etc., 1929). Even though it may be an accidental, climatically determined feature these occurrences prove that the species is able to survive for several years on more or less limestone-free soil. H. serripes occurs not only on Alvar soil but also on marine and other deposits of sand (Ska, Gtl, Old Stora-Rör; Central Europe), which are highly deficient in limestone even in Oland (Sterner, 1938, pp. 17, 19-20). The distinct inclination for limestone gravel shown by the 6 species men- tioned above may give rise to some queries. Is it the chemical or other charac- teristics of limestone that are decisive in this respect? There are such “other” differences between limestone and siliceous stone. For instance, the former is lighter (weight by volume of gravel, including air spaces, when closely packed is 1.4 as against 1.7 for siliceous gravel)* and its particles, produced by the weathering of a kind of stratified rock, are flattened (hence also lighter for a given diameter). There are thus distinct mechanical differences between the two kinds of gravel. Such characteristics of the soil can be important for noc- turnal soil animals, which must burrow into the ground every morning. It was therefore important to test the “remaining” 6 species on a kind of gravel possessing mechanical characteristics, identical as far as possible to those of limestone gravel, but limestone-free. I obtained such material in eroded soil from the clay schist of Dir Boda, Osmundsberg. The “gravel ex- periment” was repeated with the 6 species mentioned, with schist substituted for siliceous gravel (Experiment 57 ff., p. 78; Table 6). The result is unequivocal, if surprising. None of the “limestone species” preferred limestone to schist! Only the two species for which the least (and too scanty) material was available showed the same values. The remaining 4 species preferred clay schist, apparently because it had more favorable me- *On the other hand the weight by volume of each kind of gravel without air spaces is greater than in the case of limestone, 2.34 as against 2.20. The two kinds of rock used in Experiment 153 had the following values: limestone 2.63, granite 2.65. 111 Table 6. Distribution of some species of Harpalus and Panagaeus bipustulatus in the substratum gradient apparatus on limestone gravel (“Kalk”) and CaCO;-free clay schist (“Schiefer”) gravel. Experiment 57 ff., p. 78 Limestone Clay gravel schist H. azureus K 40 60 = 100 specimens H. anxius K 39 61 = 100 specimens H. rupicola K 30 70 = 100 specimens H. melleti K 34 66 = 100 specimens H. rufitarsis K 17 (49%) 18(51%) = 35 specimens Pan. bipustulatus K 16 (47%) 18(53%) = 34 specimens chanical characteristics (weight by volume: 1.27 against 1.40; with even more flat particles. Weight by volume without air spaces: 2.24). As far as I can see, the preferenda experiments so far described, which were carried out with sufficient material at least in the case of “gravel experiments,” showed that limestone attracts the presumed “limestone species” not because of its chemical, but because of other characteristics, either mechanical or not manifested in these experiments. However, these assumptions lead to the con- clusion that such preferenda experiments are on the whole reliable and a preferendum is also an optimum or at least comes very close to it. Would it not be possible, on the other hand, that the adult beetles exclusively used for these experiments would be indifferent to limestone, whereas the larva would show a certain requirement of limestone? I do not think so. If the beetle shows no positive limestone preferendum (whether or not it requires limestone or not), then there is the obvious danger that the female may deposit her eggs at a place where a larva with a limestone requirement could not live. It is true in this as in all other cases (discussed at greater length on p. 66) that the greater the difference between the preferendum and the optimum, the less viable is the species. And for this purpose a comparison should be made in the first place between the preferendum of the most activet stage and the optimum of the most inactivett stage. Thermal and Hygric Characteristics of “Limestone Species” Since the experiments described above did not indicate any dependence of the “limestone species” on the chemical characteristics of CaCO, I proceeded to test other physiological characters of these species. In particular, their re- sponses to temperature and humidity were studied. t( = “Mobile”: suppl. Translator). tt(= “Immobile”; suppl. Translator). 131 112 The temperature preferenda were studied (Experiment 11 ff., p. 71) in the usual temperature gradient apparatus (description on p. 66). The temperature of the place chosen by each individual was computed by interpolation and depicted graphically (Diagram 19). It is easiest to draw a comparison between different species by calculating the median preferendum of the total individuals of each species. An idea of the stenothermy (the more or less firm dependence on a definite temperature range) permits calculation of the dispersion, i.e. the mean deviation of individuals from this median preferendum. The critical temperatures (“response points”) were determined (Ex- periment 125 ff., p. 105 ff.) according to the method of Krogerus (1932, pp. 143-146). For each species 3 specimens were tested, and the mean was calculated from the most resistant animals (in each “point”). This prevents displacement of numbers by a diseased or otherwise debilitated individual. A further difference from Krogerus is that I have excluded his “Point 3,” since to me it appears impossible to determine exactly the lowest temperature at which the test animals remain active (cf. p. 104). In Diagram 20 the species are arranged in alphabetical order, since they do not show a ranking order for the various reaction (= “response”)! points. The response of the animals to humidity was as well investigated, partly by preferenda experiments and partly by determining their resistance. I obtained the humidity preferenda in the substratum gradient apparatus by serial experiments with 7 categories of completely pure sawdust of different moisture contents (Experiment 79). It must be noted that the distribution of the beetles was studied with regard to the moisture content of the substratum, not of the air, because the apparatus was not closed. At least in the four samples with the highest moisture content (4 to 7) the relative humidity of the air between the substratum particles was certainly identical (100%). In the case of the species of Harpalus that live subterraneous for most of the day, the humidity of the substratum not only plays a role in the water balance of the animal, but also determines the possibility of burrowing itself, which was especially evident from experiments described elsewhere (p. 505) with sand of different particle sizes (Experiments 94 ff.). This twofold response of the animals undoubtedly reflects an experimental error. However, since it involves ecologicaliy very closely related species, and only comparative values were intended, I consider the conclusions justified. The same goes for the reservation on the inevitable larger clustering of animals at the end of the linear gradient apparatus.—For each species, the average place in the series of boxes was calculated. The diagrams (Diagram 21) illustrate the series from the least to the most “xerophilous” species. In the initial experiments, with comparatively little material, which were 134 carried out during spring (May) (Diagram 22), serripes and punctatulus t (Suppl. scient. edit.). 136 113 showed equally strong xerophilous characteristics. But the repetition with more material (June through July), however, showed that serripes has a much stronger inclination for dry conditions, which undoubtedly correspond better to the actual situation. V. Peritunen, who later took over the material of both species, tested it in a circular gradient apparatus at different levels of relative humidity of the air and found a much higher preferendum in the case of punctatulus. However, in this species the humidity preferendum (and so the temperature preferendum) is extremely labilet (see above, p. 57 ff.), so that their place in the “ranking list” (Table 7) remains uncertain. The time the beetle survived in a small dish without water was determined as the relative resistance to desiccation (Diagram 23). Three specimens of each species were used. The first indications of paralysis were recorded. As in the case of the “critical” temperatures (Diagram 20), the mean endurance was cal- culated only from the 2 most resistant individuals. In the diagram (Diagram 23) the species are arranged according to resistance to desiccation. It is necessary to give a summary of the experiments on the response of the species of Harpalus to temperature and humidity as described above. It was intended to establish which of these animals are especially thermophilous (heat requiring) or xerophilous (dryness requiring). The aim of the experiments was to provide only comparative values and it is therefore advisable to consider this summary as a ranking order of the species, on the one hand with regard to their thermophily and on the other to their xerophily. However, first of all one must be clear that not all the values obtained are equally significant. In the temperature studies special importance must be attached to the mean preferendum, which shows (excluding experimental errors) the temperature that the species voluntarily seeks out in nature, if possibilities for them are existing. Compared with this, the “mean deviation” from the mean preferendum—the dispersion in the gradient apparatus—is less important. It gives an idea of the relative stenothermy of the species, of the fineness of its temperature sensitivity, and surely also about its persistence in a biotope of a particular temperature type, but not necessarily of a high temperature. For there also exist cold-requiring stenothermous species (Krogerus, 1939, p. 1223). The determination of “critical temperatures” has its greatest importance in that it provides an understanding of the temperature margin within which the species in question remains active, hence about the duration of the annual life period. At any rate, for our climatic conditions it is quite clear that the “lower points” are decisive, i.e. point b (Diagram 20). In this connection, I consider the “upper points” (c, d) rather unimportant. The species studied (with the exception of A. aeneus) are nocturnal animals or at least animals of the twilight, which remain buried during the hot day and are not affected by the highest temperatures of the upper layer of soil. As t (Ecological term; suppl. scient. edit.). -uopvjas ‘FI 107 sainsy MO] u paynssı Alqegoıd sey aiep Alles SyL, “I LS d 998 ‘snynpjound “Fy Jo epuasayaid Jo Ayyiqeiea SuILI99UOJ, 7 17, dp IT qowredxg 'snyeredde juaipess oinjesoduay oy) ur smppdıvg Jo salWads ST Jo vonnginsiq] 6] WeiseIq 621 O€ GC 07 oGb | Me =) I i Ina al uoppJas Ka Ee ll Ill N pjosdna | | Ho mem IA | | snpan) tI er eM tele ft 1 | | STP ODADUS eel ram AM | snguan He NV stelle tail ==) | N sadıyay IE ee ee 1 y smuxrup a R EN ee en | IF] ee | fg y smysajbau ei es = TH th 8/9 N“ Sadtssas elle 1 | P/ ge sadıugna 1 nal A oz y sdaayound Pd i ott EN az y snynjojound bt niin Vil 2 y snaansp Ate allslet “eet Joe XN sısanpfna ua O€ GC 07 cs -uedxa saı9adsS yo Aeq ‚sısäjered jeay [E10 L—P :(s33] pury Jo Allewadsa ‘sis -Ajesed Jensed) uonsneyx> Jo uSIs 18.11I—0 ‘(pa “juaros 'Iddns : .1910U.1m1 Jo yuIOd,, =) JOSıı PAIySLI sey PISUI—q JUSLU9AOLU ISILJ—e “LOI-SOL “dd “yy Zei “JJ szı quotuledxg “(pajeed) jou osimayjo ‘synudiaa “FY osje) snypduvzyz Jo sanads 9] Jo sjulod asuodsai ainqesodtuay gz weisey 0EI OS o GV Or ‚Ol G O+ eS SSS SSS ESS of OF y sipusga OL —————— ee ———— — 118 snpam „Lg a4 een, — So 0708 snuipbvavus ‚cd = rt = A .1'05 Y sodıusas .g$ neh --4___--- =) -—-FReE oP ab noppJas 99 —— Fe er 7 ZZ ———— ser y porns ,0°8 — en F-FF gos y sisappfna del Sets = See 4/08 sadiaqna sg I Eu F3+ 94/08 y sJaoıysund ao) ——— Pr 7 a — „p‘oS y snynyoj2und Way _——— Te 20708 NY 5nj92160u er | u ZZ —— .g'Lr N nallaW oe Fee Oe a ‚115 M sadyary £9 SSS SS Sy Pr VE GG ‚LES MY Sn24nSD oc, ——— bee cols 4 smauv re he Z Zend fr ——: 915 snauao „us ———— 000 =3-- 1 ne Sp „or | suawıaads suawıaads Kl G OF 3o7 Z 10} p juiod 10} Z 10} q julod 104 | ainjyesadwa} aunjyesadwa} | ued ues Dry Moist 30. „208 seladon 5.73 __ oD : : OZ mu am ia | | 30. 20_ rupicola K 5.57 : 10_ vu i _ om i 40_ $ : 20_ aeneus 539 oy, Zz pad a m EN 40 30 es Do 10_ MM Se] ei h j = 20. puncticeps K 4.6 - 10_ 0 Es S| a a 30_ 20. rubripes 4 81 oA a 0% zu BEE BE 30. 20_ melletiK 4es 10_ Mm Em 3 4 5 6 7 Box number 1 2 133 Moist D 20. ae smaragdinus 4.31 ; of Bi mu BEE Pa | 20_ punctatulus K 4.24 a 10 0 3 20. azureus K. 3.77 10. ‘Beaune OG 20_ anxius K 3.63 $ ir ee a = EB ie 30_ by 20_ tardus 452 er 10. & E ME =: BB EHE ga neglectus Ke 2 20_ ane | 0% NEE m 4 30_ 3 20_ Serripes K. 3.39 2 a a if g uo ow 30. t 20_ hirtipes K 2.2 i BD BE um un Box number 1 2 3 4 5 6 7 Diagram 21. Distribution of 15 species of Harpalus in the substratum gradi- ent apparatus according to varying moisture. 7 boxes, number from 1 (dry) to 7 (moist). Calculated “mean box” for each species after name of species. All experiments carried out simultaneously, June 23 through July 24. Ex- periment 66 ff., p. 79. 134 137 118 Moist Dry %, 50 — +++++®++ ++ nelleti © — punctotulus N —:-9:— rupicola / \ sec ee Be cee seladon 40— : mr Tome serri pes Box number, 7 6 5 4 3 2 1 Diagram 22. Distribution of 5 species of Harpalus in the substratum gradient apparatus according to varying moisture. All species were tested simultane- ously, May 19-23. Experiments 70 ff., pp. 79-80. Cf. Diagram 8 (p. 58) and Diagram 21. far as I am aware, the highest soil temperatures measured in Sweden are from a Norrlandish pine forest, where temperature of 59°C was found in the upper mossy layer (Forsslund, 1943, p. 60). Such high values were not even obtained by Kraus (1911, p. 109) in the thermally very favored undulating limestone area in the Maintalt. In the humidity experiments the preferenda should likewise be given prior- ity, partly because they were carried out with more material and partly because the animals in the resistance experiments showed fairly large individual vari- ations. After long thought I decided to do justice to the very variable significance te = valley of the River Main; suppl. translator). “SATE POAIISIO SEM JBLUIUL USYM SUI] ISET—Q :(SIUSLUIAOLL JO UONEPIEIOI PUNSIP JO sıskjeied jenied) uonsneyxa Jo uSıs 111,7 —e ‘g01 'd ‘HEI-BEI sjuowsdxg “(PaIean J0U asIMIOYIO ‘sybUlaA “FY osje) snpodinzy Jo saioods 97 ul uonNenIsap 0} DOULISISAY “EZ WeISPIG SET a Sears = ——— iz 209 MY vjondna FS „9 M Yaa > O—=ES ie 05 M sıypu4aa a Zn » 198 uopvgas SS n 90) 4 snaanzn Sia a oS me ee Fee 25 33651 MY sdosysund +>4-=+F>F= a ig! M snivun IR re == r= vezlgl y snnipysund Fa = 3 ee oe = 02T snauaD == SSS — ae tree sagiaqns Se 2 SS — ue Lle y sijrapbau =] FF FF — — — u SLO N sısanpifna OT SS SS gg — m ——————— uw EE snp4n} TH nu O6€¢ sninpbv4pus ru FEST —— >, gg9r M sodyyay en ra er 2 == sınoyS$'z6t M Sadissas 0589 008 OS 00% ose 00€ 0527 00z ost 001 09 Oa esueulseds Z JO ueayy 120 137 Table 7. Thermophilous and xerophilous rank of 15 species of Harpalus based on experiments. Each of 10 suspected “limestone species” is in bold type in Table where it has a higher position. a: mean preferendum and “point b” (Diagram 20, p. 130) calculated twice. b: mean preferendum calculated twice. preferendum Mean deviation from preferendum Lower ‚temperature limit “point b” Mean preferendum Resistance to desiccation hirtipes K serripes K neglectus K tardus 11 azureus K 2 puncticeps K 3punctatulusK 4 rufitarsis K 5 melleti K 6 rupicola K 7 vubripes 8 Tardus azureus K 9 hirtipes K 9 rubripes 10 serripes K 7 10 rufitarsis K II aeneus tt melleti K 12 seladon 12 puncticeps K 13 neglectus K 13 aeneus 14 anxius K 14 rupicola K 15 smaragdinus 15 seladon OPW 4 smaragdinus anxius K punctatulus K Ques ON Awm BW DN - co NI ODOAMUM ORWDN & Se tlt wm Bh WwW 138 of the experiments as follows: The “zhermophilous rank” was calculated from the series of species with respect to cold resistance (Diagram 20, point b), mean preferendum, and mean deviation therefrom (Diagram 19), so that the two first-mentioned values were counted twice. The heat resistance (Diagram 20, points c, d) was not taken into consideration. The “xerophilous rank” was determined in the series with regard to the preferendum (Diagram 21) and resistance (Diagram 23, point b), and the former was counted twice. The result is evident from the two preceding Tables (Table 7a, b). Each of the 10 suspected “limestone species” (groups 1-3, p. 117) is printed in bold face in the Table where it is ranked higher. It is noticeable that the first three positions in both Tables are taken by “limestone species”, none of them rank lower than the 6th position in either Table. 121 We speak of “xerothermic” species, i.e. animals which require both warmth and dryness. Evidently it would be possible to prepare a common “xerothermic table” on the pattern of the above Tables. However, data produced as a result of an admixture of diverse characters and viewpoints can hardly be satisfac- tory. It is more wise to treat each parameter separately. Accordingly the con- clusions based on the experimental results can be expressed as follows: Some of the species of (Harpalus) presumed to be associated with limestonet on the basis of their ecological or geographical distribution are markedly heat requiring (thermophilous), whereas others are markedly dryness requiring (xerophilous ). 139 Cymindis humeralis is also a distinctly thermophilous species. Its preferred temperature is much higher than in angularis and macularis (Diagram 24), and shows a very small dispersion. The mean lower and mean upper limit of ac- tivity (as calculated in Diagram 20) respectively is: humeralis +2.1° and 48.1°; angularis +5.4° and 45.4°; macularis +2.4° and 47.3°C (Experiments 125, 127, 134, 136, pp. 105 ff.). Resistance to desiccation (as in Diagram 23): humeralis 985; angularis 883; macularis 86 hours (Experiments 138, 139, p. 108). Experiments with the “Universal Gradient Apparatus” If, as in the preceding section, the effect of two or more factors on one species is studied one of course wishes to compare the degree of their influence on the animal. For instance, if a species exhibits distinct thermophily as well as xerophily the question arises whether the two responses have equal magni- tude or whether one is stronger. It is therefore necessary to experimentally test the simultaneous influence of two, three or more factors in various combi- Mean 1 1 | 1 I { N | { ! N 1 | 1 i 1 1 15 20 25° humeralis 1 MEU AIMEE I 25-37" angularis! I | 3; I 1 | | || | 120721 macularts || | | 7 a bok li Gl ie | ial | | | 19.48” Diagram 24. Distribution of 3 species of Cymindis in the temperature gradi- ent apparatus. Experiments 8-10, p. 70. Values for C. angularis are certainly too high, because moist cotton was used at the warm end. t(= “calciphilous species; cf. p. 816; Suppl. Scient. Edit.) 140 122 nations. For this purpose I modified Krogerus’ apparatus described elsewhere (p. 85), which then received the rather pretentious name “universal gradient apparatus.” In the study of those “limestone species” that are rather insensitive to chemical factors, it was found advisable to restrict the experiments to tem- perature, humidity and light factors. First, each of these factors was studied separately in the apparatus, and then in different combinations of two or three factors. Each species was therefore subjected to 13* different experiments in the universal gradient apparatus, and each experiment was repeated at least twice (each time with 20 or more specimens). Sufficient living material was not available simultaneously for all 15 species of Harpalus treated in the “lime- stone experiments.” The experiments in the universal gradient apparatus were restricted to the following 11 species: H. aeneus H. rupicola H. anxius H. seladon H. azureus H. serripes H. melleti H. smaragdinus H. punctatulus H. tardus H. rubripes Even this resulted in 154 (11 x 14) different experiments, which can be represented only by as many different diagrams. I was undecided for a long time whether publication of all these diagrams might not be considered a waste of space, since only a small number of them are significant in the present context. On the other hand singularity of behavior of one species under the influence of a given combination of factors can be meaningful only when compared with a sufficient number of “normal” species. Finally, it was possible that useful conclusions might be drawn by other researchers from the basic data I did not utilize. I therefore decided to publish all the experiments carried out with the help of the universal gradient apparatus in the form of diagrams (Diagrams 25-35). The 2-4 identical experiments (with the same species) were combined in one diagram and depicted by different types of circles (open, filled, crossed, etc.). Like this, one can always look up the readings for a particular experi- ment and to some extent draw inferences as to whether the species concerned behaved consistently or not vis-a-vis the particular constellation of factors. A “mean distribution curve” was derived by joining all the values simultaneously obtained (in the same experiments) and subsequent interpolation. The response to temperature was studied by transforming the substratum gradient apparatus into a temperature gradient apparatus (details on p. 86). * Actually 14 experiments each, since the temperature factor was studied partly with dry and partly with uniformly slightly moist sawdust. 174 123 This differed from the usual apparatus only in that it was not possible to read off the actual location of each individual. The beetles were classified according to the number of boxes in 10 temperature categories (Krogerus, 1937, p. 299), whose mean temperature was calculated. The humidity can be equated with the humidity of the substratum. As in the experiments described above (p. 78) on humidity preferenda, sawdust of varying moisture content was spread on the bottom of the boxes in a layer about 5 mm thick. The humidity of the air was subject to much less alteration (p. 88). The light was regulated through squares of different sizes cut out of the black, opaque paper lid (four per box). All the experiments with light were carried out in the laboratory in broad, but indirect daylight (from June 20 through August 1), never under an overcast sky. Since in itself the light factor was of no significant interest for these experiments, being used chiefly to regulate the other two factors, its somewhat variable intensity due to cloud or time of day may be of no great importance. —All experiments in which the light factor was not taken into consideration were carried out in dark- ness. Both, humidity and light were always regularly graded from one end of the gradient apparatus to the other. An exhaustive discussion on the conclusions that can be drawn from the experiments with the universal gradient apparatus would not be relevant here. I shall therefore restrict myself to a few points which may be important for the “limestone problem.” Since we are dealing with more or less markedly thermophilous, xerophilous and heliphobous animals, in the following account I use the term similarly directed factors for two-factor and three-factor experiments, as long as the same end of the apparatus is warm, dry and dark or any two of these. Otherwise I use the term differently directed factors. Admittedly, these terms do not quite fit certain species, for instance, melleti, rupicola and seladon, which are not markedly xerophilous. It is clear that the results of the experiments depicted graphically in the diagrams represent only a very general expression of the responses of the species concerned. The experimental material (mostly 40-80 readings) is too small for the plots to be considered authentic in detail. This is especially true of the plots with two or more peaks. On the other hand some species show an astonishing consistency in the position of the maximum, over several (repeated or related) experiments, examples of which are given below. The intention was not to establish the exact distribution of the animals in one or Other experiment, but to find the relative changes produced by a new or converse factor. It is advisable to begin with a comparison of two especially pronounced types, on the one hand H. azureus, which in the above correlated temperature Moist Dry 80%, Light Dark 9° 50% 40_ 30_ 20_ OZ 15° 30% 20_ 10= a? OW oo 141 Diagram 25 a-c, h-i. Harpalus aeneus. Universal gradient apparatus. Experi- ment 108 (p. 96). a!—Temperature (dry); a’—Temperature (slightly moist); b—Humidity; c—Light; h—Humidity + light; i—Humidity — light (cf. p. 176). 30% Dry 10_ Moist 40° Dark 30. 20- 10- Light 142 Diagram 25 d-g. Harpalus aeneus. Universal gradient apparatus. Experiment 108 (p. 96). d—Temperature + humidity; e—Temperature — humidity; f—Temperature + light; g—Temperature — light (cf. p. 176). 30% 20_ 10- 30- 20- 155 143 \ loist Diagram 25 j-m. Harpalus aeneus. Universal gradient apparatus. Experi- ment 108 (p. 96). j—Temperature + humidity + light; k—Temperature + humidity—light; I—Temperature — humidity + light; m—Temperature — humidity—light (cf. p. 176). Dry Dark Dark 20_ 40_ 20- 144 Diagram 26 a-c, h-i. Harpalus anxius. Universal gradient apparatus. Experiment 109 (p. 97). See Diagram 25 and p. 175. 40% 30- 20_ 10_ 20_ 10- = 10- 145 Moist Dry Diagram 26 d-g. Harpalus anxaus. Universal gradient apparatus. Experiment 109 (p. 97). See Diagram 25 and p. 175. Dark 40° oO 50% 40_ 30— 20— Moist Light 146 o Diagram 26 j-m. Harpalus anxius. Universal gradient apparatus. Experiment 109 (p. 97). See Diagram 25 and p. 175. Moist Dry Light Dark 40% 40% 6 Cc ® 20% 202 0 on 30% 20_ 30% 20_ 147 Diagram 27 a-, h-i. Harpalus azureus. Universal gradient apparatus. Experiment 110 (p. 97). See Diagram 25 and p. 174. 30%, Dry 20_ Moist 40° Dark 20— Light 0_ 148 Diagram 27 d-g. Harpalus azureus. Universal gradient apparatus. Experiment 110 (p. 97). See Diagram 25 and p. 174. 30- 20_ 30— 149 Dry Light Diagram 27 j-m. Harpalus azureus. Universal gradient apparatus. Experiment 110 (p. 98). See Diagram 25 and p. 174. Dry Dark Moist Dark 10 Moist Dry Light Dark O0 40%, Light Dark 40% Dark Light 40° ® © Ske % a 6 © fo) o © 150 Diagram 28 a-c, h-i. Harpalus melleti. Universal gradient apparatus. Experiment 111 (p. 98). See Diagram 25 and p. 176. 30% 20 10_ ‘Dry Moist 20. 40° 30- 20= Dark Light 0_ ® fe) ® ® 151 Diagram 28 d-g. Harpalus melleri. Universal gradient apparatus. Experiment 111 (p. 98). See Diagram 25 and p. 176. 20— 10— 30% 20— 10_ 20% 10- 102 152 Diagram 28 j-m. Harpalus melleti. Universal gradient apparatus. Experiment 111 (p. 98). See Diagram 25 and p. 176. Dry Dark oo Dry Light Moist Light 20% 10_ 108 fe) Diagram 29 a-c, h-i. Harpalus punctatulus. Universal gradient apparatus. Experiment 112 (p. 98). See Diagram 25 and p. 175. 154 Diagram 29 d-g. Harpalus punctatulus. Universal gradient apparatus. Experiment 112 (p. 99). See Diagram 25 and p. 175. Dry Dark 20_ 10_ 0_ ® 40% 30- 20_ Dry 10_ Light 0_ ® Moist 20% Dark 10_ © Oz Moist 20% Light 10_ 0_ © 155 Diagram 29 j-m. Harpalus punctatulus. Universal gradient apparatus. Experiment 112 (p. 99). See Diagram 25 and p. 175. 40% 20 0 e Moist Dry Dark Light 40% 20— 0= 156 Diagram 30 a-t, h-i. Harpalus rubripes. Universal gradient apparatus. Experiment 113 (p. 99). See Diagram 25 and p. 176. Dry 30_ 20_ fe) 10_ Moist 40° Dark Light 10 157 Diagram 30 d-g. Harpalus rubripes. Universal gradient apparatus. Experiment 113 (p. 99). See Diagram 25 and p. 176. Moist Dry Light Dark 20 9 C ® 10_ = 3 0_ Dry 30% Light () 208 © 10_ ro) e 0- fe) 15° 20 30% Dry Moist Light Dark 20_ ® fo) 102 ie) @ O Oo 0_ 30% fo) 205 fe) © ® 10- |Dry Moist Dark N Light 0_ fo) fo) 158 Diagram 30 j-m. Harpalus rubripes. Universal gradient apparatus. Experiment 113 (p. 99). See Diagram 25 and p. 176. Moist Dry Light Dark 40% 40% Moist Dry Moist > Day Light Dark Dark Light 30% 10_ 30 20- 10_ 159 Diagram 31 a-<, h-i. Harpalus rupicola. Universal gradient apparatus. Experiment 114 (p. 100). See Diagram 25 and p. 176. 30% 502 40 10- Moist Dry Dark 160 d e) e 35 f g ® Diagram 31 d-g. Harpalus rupicola. Universal gradient apparatus. Experiment 114 (p. 100). See Diagram 25 and p. 176. Dry Moist Light 10_ 10- 20_ 1O= 10— 161 Diagram 31 j-m. Harpalus rupicoia. Universal gradient apparatus. Experiment 114 (p. 100). See Diagram 25 and p. 176. Moist Dark 40_ 30- 20- 162 Diagram 32 a-c, h-i. Harpalus seladon. Universal gradient apparatus. Experiment 115 (p. 100). See Diagram 25 and p. 176. Dry 20— Dark 10_ 40— 30_ 20— 2 ® Dark Light 163 Diagram 32 d-g. Harpalus seladon. Universal gradient apparatus. Experiment 115 (p. 100). See Diagram 25 and p. 176. 30% 164 Diagram 32 j-m. Harpalus seladon. Universal gradient apparatus. Experiment 115 (p. 101). See Diagram 25 and p. 176. 60% @ 60% Moist Dry 165 Diagram 33 a-c, h-i. Harpalus serripes. Universal gradient apparatus. Experiment 116 (p. 101). See Diagram 25 and p. 174. 40_ Moist Dry 30_ 20. 10 202 Moist 10) 166 Diagram 33 d-g. Harpalus serripes. Universal gradient apparatus. Experiment 116 (p. 101). See Diagram 25 and p. 174. 40° 10_ 102 20_ 10_ 50— 40_ 30— 20— 167 Diagram 33 j-m. Harpalus serripes. Universal gradient apparatus. Experiment 116 (p. 101). See Diagram 25 and p. 174. 40% 60% 20- 40. 02 20_ 02 60% 40_ 20_ 0- 168 Diagram 34 a-c, h-i. Harpalus smaragdinus. Universal gradient apparatus. Experiment 117 (p. 101). See Diagram 25 and p. 176. 30% 20_ Dry 20_ Dark 20_ 169 Diagram 34 d-g. Harpalus smaragdinus. Universal gradient apparatus. Experiment 117 (p. 102). See Diagram 25 and p. 176. 30%, Moist Dry Light Dark 20_ > fe) e) 10. e o e S 0_ eo 30% Moist Dry Dark Light 20 e 10— D [o) ® fo) e ¥ 119% 20 bd 40° 30% Dry Moist Ligh Dark a ight [6) 108 oO o © 022216 0) fe) ® 40% 30 20- fe) @ 10_ dry ® ° Moist Dark = Light 0- fe) fo) © 170 Diagram 34 j-m. Harpalus smaragdinus. Universal gradient apparatus. Experiment 117 (p. 102). See Diagram 25 and p. 176. 30_ 200 Sick Moist Diya © 60% 40- 20_ Moist Dr Y 0 Light Dark ® Light © 171 Diagram 35 a-<, h-i. Harpalus tardus. Universal gradient apparatus. Experiment 118 (p. 102). See Diagram 25 and p. 176. 30% 20. 10_ 20. 10— 10— 20— 10— O— Light Dark 172 ® Le) ® Oo ° © @ ® [e) o Q ® Oo ° ie) @ ® ® [6) @ ® Diagram 35 d-g. Harpalus tardus. Universal gradient apparatus. Experiment 118 (p. 102). See Diagram 25 and p. 176. Dry C) Moist [e) 40° Dark ox) Light (0) @ 20% 10_ 10— 173 Diagram 35 j-m. Harpalus tardus. Universal gradient apparatus. Experiment 118 (p. 102). See Diagram 25 and p. 176. Moist Dark Moist Light 175 157 experiments was the most thermophilous of the 15 species tested (but only 8th in xerophilous ranking: Table 7); and on the other hand H. serripes, which was the most xerophilous species after H. hirtipes (only 10th in thermophilous ranking). This difference is already evident in the one-factor experiments (Dia- grams 27 a-c, 33 a-c) in the universal gradient apparatus. The temperature preferendum in azureus is far more pronounced and lies (especially with dry sawdust (a!)) about 3° higher. On the other hand serripes shows a stronger attraction to the dry end of the apparatus (b). Besides, it is much more helio- phobous (c). In the two-factor experiments (with the temperature factor) the differ- ence becomes pronounced. If the temperature and humidity are similarly di- rected (d) the beetles behave almost identically with a very sharp maximum at 30-31°C. If the same factors are differently directed (e), serripes at once moves to the cold end, whereas azureus is remarkably undecided. In one of the two experiments most of the animals congregated in the colder (drier) half, but in the other experiment, in the warmer (moister) half of the apparatus, and the mean plot has two marked peaks. —Likewise the light factor (f, g) produced much stronger variation of the temperature preferendum in serripes than in azureus. The three-factor experiments (j-m) are still more informative. A. azureus, in all combinations, retains a distinct maximum at about 32°C, but serripes does so only if the humidity factor is similarly directed with the temperature (j, k). If not, it moves mostly to the cold end, even if the light factor, against which it otherwise reacts very strongly, is differently directed. Certainly the results of the experiments with azureus-serripes also reveal incongruities that cannot be explained by insufficient observation material. Examples are the secondary maximum of serripes at 20°C with all factors simi- larly directed (33j), or the question, why a different direction of the humidity factor alone (e) causes a steeper decline in the temperature preferendum than both humidity and light together (m). However, the general tendency in both these species is quite clear. The universal gradient apparatus experiments greatly strengthen the earlier con- clusions that for azureus the temperature factor is decisive and for serripes the humidity factor. A counterpart of the constantly high temperature preferendum of azureus was not found among the remaining species tested in the universal gradient apparatus. H. punctatulus, which comes after azureus (and puncticeps) in the ranking list given in Table 7, shows very irregular responses (Diagram 29), which is certainly related with the peculiar labile humidity preferendum of this species (pp. 57 ff.). H. anxius belongs to the serripes type. But in the latter species the tem- perature factor evidently has a much stronger effect. The experiments with 176 19707 158 differently directed temperature and humidity have especially to be compared (Diagrams 26e, 33e). In the earlier ranking order of thermophily (Table 7), anxius certainly has too low a place, chiefly because of its low “lower response point” (“point b”). In contrast with the species discussed so far, which show a sharply de- limited preferendum zone in most of the temperature gradient apparatus experiments, H. aeneus (Diagram 25) may be singled out. It reacts strongly and consistently to the light factor (c, f, g, etc.), and to some extent also to variations in humidity. But the temperature and different combinations of fac- tors with it, especially in the two-factor experiments (d-g), elicit very diffused responses. It is highly characteristic that a strongly pronounced maximum ap- pears only in the three-factor experiment in which all three factors are similarly directed (Diagram 25j). It is characteristic of an eurytopic and at the same time widely distributed species that its temperature preferendum is not only low but is also less fixed and consequently can easily be displaced by other factors. —An exactly similar characteristic is shown by H. seladon (Diagram 32). In the simi- larly widely distributed but somewhat less eurytopic (xerophilous) H. tardus (Diagram 35) a deviation is discernible only to the extent that the inclination for dryness is somewhat more established (d, e, 1). A median position is occupied by H. rubripes (Diagram 30) and H. smaragdinus (Diagram 34), which have almost identical Fennoscandian distribution (Swedish northern limit in Northern Upl) and are also very closely related ecologically (occurring on more or less dry sandy and gravely soil). Their behavior in the universal gradient apparatus is largely identical. Their temperature preferendum is much more fixed than in aeneus, seladon, and tardus (cf. especially the two-factor experiments; they are “southern” species). But the humidity factor, much as in tardus, shows a more distinct influence throughout (they are markedly more xerophilous). The two remaining species, H. melleti and H. rupicola, are distinct in several ways. The earlier one-factor experiments showed (see ranking list in Table 7, p. 137) that they are neither distinctly thermophilous nor xerophilous. Yet both species occur in the extreme south, rupicola being actually confined to Old and Gtl. The universal gradient apparatus experiments (Diagrams 28, 31) confirm that their temperature preferendum is not only comparatively low but is also easily displaced by other factors, for instance light (g). A. rupicola is found to be at least as little xerophilous as seladon (Diagram 31e, k-m), whereas melleti has somewhat higher requirements in this respect. Those factors, which are responsible for the restricted southern distribu- tion of melleti and rupicola, apparently were not revealed by the universal gradient apparatus experiments. However, an indication is perhaps provided by the earlier finding of the strikingly high “lower response point” with regard to temperature in both these species (Diagram 20; Table 7), i.e. as in azureus and punctatulus, the short duration of their continuous annual activity period. 178 159 At present it is not known whether a particular thermal sensitivity of the lar- vae is decisive, but it seems less probable, since this stage is passed through in both cases in midsummer. The conclusions drawn from the experiments with the universal gradient apparatus may be summarized as follows: 1. If one of the factors studied exercises an especially strong effect it is possible to determine it. For example: temperature (azureus), humidity (ser- ripes). 2. Species that are widely distributed and at the same time eurytopic (aeneus, seladon, tardus), have weakly fixed temperature and humidity pre- ferenda, so that the distribution of the animals in the apparatus is determined mainly by the light factor. A factor that certainly has no more than a subordi- nate role in the choice of biotope in the case of more or less subterranean soil animals and which probably has no role at all in the geographical distribution (with regard to area limits). According to responses, such geographically and ecologically ubiquists can be designated “euryphysic” (“euryphys”) species in contrast to the more or less “stenophysic” (“stenophys’”) species. 3. The examples of melleti and rupicola nevertheless show that restricted distribution (in our case markedly southern) is not always related to the re- quirements for life demonstrated by the preferenda experiments. Probably, dis- tribution is yet influenced by temperature. A conformity exists in the results obtained with Agonum dorsale and Brachynus (p. 61), according to which “the lower response point,” i.e. the resistance to low temperatures, may have a greater role than the position and magnitude of the thermal preferendum. Characteristics of Limestone Rock and Limestone Gravel On the basis of the experiments carried out with the presumed “limestone species” the chemical characteristics of CaCO, appear to have no influence on the animals concerned, but these animals are characterized by pronounced requirements of heat or dryness (or both). In the following experimental study of limestone rock and limestone gravel it was especially important to determine the thermal and hygric characteristics thereof. The temperature experiments were first carried out on a small scale as laboratory studies. A series of day experiments with readings every hour were run on the roof of the school in Djursholm from July 1, 1945 (5 a.m.) through July 2 (6 a.m.), with rock and gravel (Experiments 153, 154, p. 110). The two squarish pieces of rock were about 15 cm’ in area and about 3 cm thick. They comprised Gotland limestone (weight by volume 2.629) and fine- grained Uppland granite (weight by volume 2.646). These rocks had a very similar gray color (Thurmann, 1849, pp. 109-111, already stated that color 181 160 has an important role in the absorption of heat by rock). Thermometers were placed underneath the rocks on aluminum foil, and this face was isolated with thick saddler felt. The gravel consisted of the types described above (p. 77): partly limestone and partly limestone-free siliceous stone. During the experiment it was placed in small cubic zinc boxes of 400 cm’, with the bulb of the thermometer in the center. Four boxes with different humidity contents were used for each of the two kinds of gravel, but only 3 plots of each are given here. The four pairs of plots (Diagrams 36, 37) are largely similar. In particular it is evident that siliceous stone, whether rock or gravel, is strongly warmed during full sunshine. The greatest dıfference with limestone in this respect lies in the gravel being moderately moist. This can perhaps be explained by a greater loss of heat due to stronger evaporation of water in the case of limestone, which is discussed below (p. 188). During the night the minima differ less. But it is obvious that those of limestone always lie a little higher (up to 0.6°C in moderately moist gravel). In the gravel with the highest moisture content the difference from siliceous stone is only 0.1°C. It was evidently very important to test the general validity of the findings by further experiments, that is the higher temperature of the limestone by night, which, if borne out, would be of great biological significance. In the Fall (September 4 through October 29, 1945) the lowest night temperatures were measured with minimum thermometers in the same boxes placed in the open with the gravel as dry as possibie. Out of 54 readings higher values were shown by limestone in 31 cases, the values were identical in 20, and a difference of more than 0.1°C was recorded only in one case. The amounts of gravel used (400 cc) were very small. I also wanted to study the temperatures of gravel under sudden heating or sudden cooling, and therefore recorded a series of readings at close in- tervals on a cold day after transferring the boxes of gravel, which had been outside for weeks, into the room. Later, the converse experiment was car- ried out (Experiments 156, 157). The temperature plots (Experiments 38, 39) show the same trend as in the 25-hour series (Diagram 37): With cooling, the temperature of siliceous stone always declines faster. With warming the same phenomenon was observed during the first half-hour (temperature rise of about 11°C); thereafter the temperature of the limestone always ran ahead (by up to 0.3°C). Identical temperatures were achieved after 33 hours only. Probably condensation heat was involved, which is more influential in the case of limestone due to the greater surface area (cf. evaporation plots, Diagram 41, p. 188). However, the precondition for this evidently was the complete drying of the limestone. For when the experiment was later repeated with the gravel cooled for only one day, both temperature plots showed exactly the reverse course to Experiment 157 (Diagram 39), i.e. the siliceous stone plot 161 “Me Jo Sinelsdıus L—oul] wiyy, ‘ONUEIQ—aUN usyoig ‘ouoisow1q—ouy snonuNuogy ‘OTT 'd ‘Est Juawnadxg ‘ayuess pue Juorsatum Jo sa991d yey YIeauıapun sjojd sınye1adtus) INoY aAy-K1UIMT, “9¢ wesseig 641 I0L 9,01 007 o0E ew ee ee ....- sınoy9 9 v € [4 EV ER 210500 6 Sl Al “Ol Gh are El cl lol ö 8 (Ga) G sınoy 162 hoursın, BO 7 8. ©) to 11 12 13 140 15 \e 17. 18 19.20.21 29) 23 pa 2 30,20 oWsonare . va 40° 7 ae u EI ~ 40° 30° i = 30° I u a é si 20°— 8 - 20° 10cm 1 0 eee BY 40° - 40° 30° 20° 30° 20° WL fl pa Te ER VERDERBEN ee UCR UNEASE NE TOURS I UE Ee LE DERTTERSEETE ER hours5 6 7. 8 9 io 11 4243 1415 16) 17. 18, 19,20 ‚21 22 723724771727 3 4 576%hours 180 Diagram 37. Twenty-five-hour temperature plots in limestone gravel (con- tinuous line) and siliceous gravel (broken line). Temperature of air and weather as in Diagram 36. a—Dry gravel; b—400 cm? gravel with 25 cm? water; c—With 100 cm? water. Experiment 154, p. 110. 184 163 constantly ran ahead. In all the temperature measurements here recorded for limestone and siliceous stone very small amounts were involved, moreover in unnatural loca- tions. It was therefore deemed necessary to test the results in nature. Presup- position was that the temperature be measured simultaneously on limestone and siliceous stone under as far as possible similar situations (with respect to humidity, exposure to the sun and wind, etc.). Such requirements could be completely fulfilled only at a point where limestone and siliceous stone met edge to edge. Since such a situation hardly occurs in the case of loose deposits (moraine, fluvial rubble, etc.), it was necessary to find a suitable fault fissure. After consultations with several geologists Dr. P. Thorslund suggested a suit- able area which he had studied some years before (Thorsiund, 1936), namely, the railroad section near Sjurberg, north Rattvik in Dir. The location was visited on June 4-5, 1946 and measurements were taken. All expectations were confirmed: The fault line is very sharp. The orthoceratite limestone and granite are not separated by any loose deposits, only the rock is brecciated in the immediate vicinity of the fault and is traversed by cracks (Fig. 16). The exposure on the two sides is exactly the same. The horizontal rock is overlain by a sandy limestone-containing moraine, which rises steeply from < 1 cm (at the edge of the profile) to a height of about 2 m. The slope is toward the east (12°N), thus receives the sun for almost the whole forenoon which is not impeded by any trees or bushes (Fig. 14). Here (Experiment 159) 5 thermometers were stuck (Fig. 15) in the ground (5-6 cm deep) in pairs on each side of and equidistant from the fissure, and the temperature was measured every hour from June 4, 5 p.m. through June 5, 7 p.m. The 4 thermometers along the edge (K,, K,, G,, G,) touched solid rock, the thermometers K, and G, were positioned over about 3 dm thick moraine and the 4 uppermost (K,, K,, G,, G,) over about 1 m thick moraine. There were no perceptible differences in the moisture content of the ground and vegetation. The data obtained revealed striking differences. Those obtained with the thermometers K, and G, lying at the bottom of the fissure had to be omitted. Especiaily the plot of G,, which was based on the strongly fissured and brec- ciated granite (see photo, Fig. 14), was greatly irregular. The 6 thermometers (K,_;, G,_,) sticking into thick moraine showed very small differences: Ku Gain Ko GG, Maxima 18.75 186 182 17.85 Minima 104 103 10.45 10.35 164 183 Fig. 14. Dir Rättvik, Sjurberg. Place where temperature was measured on either side of the fault fissure. Ten thermometers are indicated by small pieces of paper. Experiment 159. Fissure UA SaR pee aaa ie a a Ks K4 G4 G5 [0] Sp (>) oO : 7 63 Hz o G2 Gi = E33 = 183 Fig. 15. Sketch showing the arrangement of thermometers in Fig. 14. Kı to Ks on limestone, G; to Gs on granite. A granite block is seen at the mouth of the fissure. It is evident that the identical moraine on both sides of the fissure has neutralized any probably existing thermal differences between limestone rock and granite rock. +6° +4° 2° ene -8° S10? 182 Diagram 38 (on the left). Course of temperature rise in limestone gravel (continuous line) and siliceous gravel (broken line) from — 10.7° to +17.6°C after sudden transfer. Experiment 156, p. 111. Curves met only after 7 hours (at 107). Diagram 39 (on the right). Course of temperature fall in limestone gravel (continuous line) and siliceous gravel (broken line) from +18.1° to — 6.1°C (gradually, down to —16.1°C) after sudden transfer. Experiment 157, p. 111. Because of the continuously falling in external temperature, curves never met during the period of observation (9 hours). NI2-v Ee ee cena toringic BESsjpnenoz enalkite a fault fiss E53 weatherea “=~ Aeposit hmestone Ei 183 Fig. 16. Railroad section at Rattvik, Sjurberg. After Thorslund (1936, p. 7). Fault fissure (“förkastning”) indicated in Figs. 14 and 15 to the right. On the other hand the plots of the two thermometers K, and G, (Dia- gram 40) which touched the bedrock and were farthest from the fissure, con- sistently showed large differences. The maximum of the granite thermometer (18.8°C), reached 2 hours earlier, was 2.6°C above that of the limestone ther- mometer (16.2°C). But during the night, the minimum temperature of the latter was 1.4°C higher (10.6° as against 12.0°C). We thus have, on a larger scale, con- firmation of the earlier conclusion obtained experimentally with small quan- tities of rock and gravel: limestone shows lower maxima and higher minima, it provides an oceanic microclimate with regard to temperature (!). 186 There is evidently a distinct thermal difference between the limestone (at any rate of Cambro-Silurian origin) and the basement complex (at any rate granite). Constants with respect to this are particularly found in the technical literature. The figures for specific heat are little different; in the case of granite and gneiss 0.20, in the case of limestone rock and marble 0.21 (“Hutte,” 1931, p. 488; Suenson, 1942, p. 34). The thermal conductivity is different: In the case of granite and gneiss it is 2.7-3.5 kcal/degree m hr. (“Hutte,” 1931, p. 494; Suenson, 1942, p. 35, gives 1.8 as the lower limit); in Sweden 3.0 is given as the mean value (“Anvisn. till byggnadsstadgan”, 1946, p. 32); in the case of limestone rock, 0.6-0.8 according to “Hutte,” 0.5-2.2 according to Suenson. The limestone (at any rate of Cambro-Silurian origin) is a much poorer heat conductor than the basement complex. We are also interested in the fact that marble is very close to the basement complex, since its thermal conductivity is given as 1.7-3.0 (“Hutte,” I.c., Suenson, l.c.). The different day plots of temperature on outcrop granite and limestone are thus undoubtedly due to the different thermal conductivity of the two kinds of rock. In the laboratory experiments with small quantities of material (Experiment 154; p. 178 above), dry and moderately moist gravel of siliceous stone and of limestone showed still greater thermal differences with regard to the daily maxima and minima. The pronounced differences in the thermal conductivity of /oose deposits, which must also be present in nature, are un- doubtedly due to the greater porosity and consequently the greater air content of the limestone gravel. Nevertheless, in nature this relationship can perhaps 167 “We JO JInelsdwa ] — au uly], ‘aywuess uUGO— UI] Udyog ‘(daop wo 9-5) Juoysawmm IMEIHOYLIO UO—aUI] snonuNuo) ‘(9I-FI 'sF14 79) aanssıy IIneJ Jo sapis yIOG UO ‘S-F aung ‘Oj aımyessdıus} Inoy-97 "Zisganls ‘YANneY JIG ‘Or weiserq sınoy 61 LL Gif EL LL 6 fh g € L ez Lz 61 Li sınoy ach oGl „02 90 UNS 0} |IOS JO sınsodxJ — - UNS 0) |JEM 4901 JO a1nsodx3 ee —— — — PUIAA -- = — uns 187 168 never be verified as clearly as by the Rattvik plots on solid rock, since loose deposits (such as moraine) of pure limestone and pure siliceous stone hardly border on one another. However, it is justified to conclude that the greatest thermal differences between limestone and the basement complex occur where on one side thin, loose limestone deposits (especially purely residual soil) rest on limestone rock, and on the other side where siliceous sione deposits rest on the basement complex. The data obtained near Rattvik, where the higher minimum temperature of 1.4°C on limestone has the main role biologically, must be considered as minimal differences for the following three reasons: 1. The adjoining solid volumes of limestone and granite are small (see profile by Thorslund, Fig. 16). 2. The loose deposits on both sides of the fissure are completely identical (limestone-containing moraine). 3. The weather before the readings were taken was very rainy all day. The ground was abnormally moist and as a result, extreme temperatures undoubt- edly were more uniform (cf. the curves obtained in the laboratory experiments above, Fig. 37). How little these things have been noticed until now is clear from the latest summary in the field of microclimatology, the work of Geiger, 1942. On p. 30 of this work the thermal conductivity of “cliff rock” is given as 0.011 cal/degree cm sec. without giving any amplitude! The literature, however, is full of differ- ent figures for the heat conductivity of various kinds of rock (cf. Stiny, 1929, pp. 436-437). The numerous readings of ground temperature by Kraus (1911) already show that the limestone content of loose deposits exercises a favorable thermal effect. But it may be added that Kraus strangely draws no general con- clusions out of it. Otherwise literature frequently contains the statement that limestone is “warmer” (for instance, Thurmann, 1849, pp. 108-109; Warming- Graebner, 1918, p. 123; Hesse, 1924, p. 427; Brenner, 1930, p. 85), without substantiating the idea with data. The above expositions are based on differences between limestone and siliceous stone with identical humidity. However, the natural conditions show, as has often been noticed by field botanists, that loose deposits of limestone or deposits with a big admixture thereof are drier than siliceous soil in the same situation. “Limestone plants” are therefore mostly evident as more or less markedly “xerophilous” (Thurmann, 1849, p. 268). In order to test this difference more precisely, I performed a simple ex- periment (Experiment 158, p. 111; Diagram 41). Equal quantities (400 cm?) of air-dried limestone gravel and siliceous gravel of the same particle size (types and quantities of gravel in boxes as in Experiment 154, p. 110) were mixed with 100 cm* water and kept in the room. Evaporation of the water, which was determined by recording the weight daily, was much faster in the case of limestone: 74.5% after 10 days. Within the same period only 50.5% water 169 188 was lost by the siliceous stone. The reason is undoubtedly the flatter shape of 189 188 the limestone particles (larger surface!) and the herewith connected greater porosity and aeration of the gravel mass. Humidity then is of essential importance for the thermal conductivity of rock. Suenson (1942, p. 35) states: “In water-saturated condition granite, gneiss, marble, and limestone have approximately the same conductivity” (tra- nslated from the Danish). The faster evaporation and moreover the greater subterranean drainage in limestone, both reduce the thermal conductivity and by this increase the thermal difference when compared with silicates. Corre- sponding with this, the temperature plots in Experiment 154 (Diagram 37) with wet gravel were the least pronounced. Hygrophilous animals are not favored on limestone. Among such species only Agonum krynicki, Bembidion clarki and B. lunulatum have their northern Fennoscandian limits in areas with Cambro- Silurian limestone. Effect of Limestone on Beetles The experiments described and evaluated in the two preceding sections have shown that the “limestone species” are markedly xerophilous or (and) ther- mophilous animals and that limestone rock and loose deposits thereof in nature are comparatively dry, and particularly that they are thermally cha- racterized by high minima of the ground temperature. This latter characteristic is particularly significant for heat-requiring soil animals. It is for instance clear that all the species assumed here to be influ- enced by limestone live in Fennoscandia at the extreme northern boundary of their total area. Then it would not be too audacious to say that for them there is a definite problem of survival: how to attain a suitable hibernating 10 8 6 0% %0 1 5 10 15 20 Days Diagram 41. Evaporation of water from equal quantities of limestone gravel (continuous line) and siliceous gravel (broken line) during 3 weeks. Experi- ment 158, p. 111. Each sample (400 cm?) received 100 cm” water and was kept indoors. 191 192 193 170 stage during the short summer deficient in heat. In other words, the length of the continuous period of life for these species may be especially decisive. We can now estimate the length of these periods (in days or weeks), thanks to the experiments on cold resistance (Experiments 125 ff., Diagram 20). But how this period is utilized in relation to the speed of development of the species in question can be revealed only by comparative breeding experiments. These are so far lacking. I must therefore content myself with an exposition of the first of these two aspects, by way of an example. The meteorological measurements at the Visby station in Gotland, recorded over many years, make a suitable starting point. The thermometer for measuring the temperature of the air during this period (it was transferred to the harbor in 1937) was hung in a box 7.1 m above the ground on the wall of the posting house at “Donnersplan” close to the harbor. This house is situated on loose deposits below the steep limestone slope and is only a few meters above sea level. The effect of the firm limestone rock and of the soil on the temperature was therefore very low, and at any rate not of a “microclimatic” character. The mean of the daily minimum temperature from April 10 through De- cember 1 was calculated for the 20-year period (1917-1936) (Diagram 42). On the basis of this plot, the average continuous annual period of activity of each of the 16 tested species of Harpalus (of which only neglectus, puncticeps, and rufitarsis are missing in Gtl) was roughly calculated. The lower temperature limit of activity (Diagram 20, p. 130, “point b”) determined by the cold resis- tance experiments (aeneus 5.2°C, anxius 3.4°C, etc.) was taken as the starting point. The daily minima of the ground temperature (at a depth of about 5 cm) were calculated (at least for limestone the values are not too favorable!) according to the Rattvik plots (Diagram 40), both on limestone and on granite. One may now undertake a hypothetical analysis of the situation: /f Gotland were made of granite instead of limestone, how much lower would the minima of the ground temperature be? Or, more correctly, how much shorter would the annual period of activity on granite be? Following the Rattvik plot we proceed from the hypothetical prerequisite that the minimum ground temperature (about 5 cm deep) on granite is 2.4°C and on limestone 3.8°C higher than the minimum temperature of the air. This gives us the above estimates (Table 8). The average increase in the period of activity of all 16 species on limestone is 23.4 days. These values suggest that an average increase in the minimum ground temperature by only 1.4°C has great biological significance. I have also tried to depict cartographically (Fig. 17) the favorable effect of limestone ground on the basis of the Rattvik plots, an attempt of a purely hy- pothetical character. I used the map of the air temperature showing the mean of the minimum temperature for May + September (Fig. 69, p. 471), adding 3.8° in the case of areas with Cambro-Silurian limestone, or 2.4°C in other 191 171 cases. These temperatures therefore hold good only for thin soil (5-6 cm) on bedrock*. Primitive limestone might well be equated with the basement com- plex, since the thermal conductivity of the two is almost the same (p. 186); the Mesozoic mountain materials (especially chalk) in southernmost Sweden, particularly in Skäne, are never exposed to the same extent. The map shows the strikingly favorable temperature of the soil in Öland and Gotland, which is at best also attained in Skäne, and furthermore in the isolated Cambro-Silurian areas of Central Sweden, areas where Harpalus ru- Table 8. Harpalus. Average yearly periods of continuous activity on granite and lime- stone respectively, in a macroclimate of the Visby type. According to data in Diagram 20 (“point b”) and diagram 40 oe ; Number N Number of days ale jan 25 days 16/5 — 30/11 28 20/5 — 13/10 19 11/, 24/49 24 16), ——ı2/10 20 27/4 —183/jı 26 1/5 — 16!) % 20 18/5 ia), 10/5 —s6/10 14/, — 20/10 Increased duration on limestone aeneus anxius K asureus K' hirtipes K melleti K neglectus K punctatulus K puncticeps K rubripes rufttarsis K rupicola K seladon serripes K smaragdinus tardus vernalis K 191, ub ay 10/5 25/19 5/5 — 7/11 28/4 _19/,, N 11/, oy 10 18/5 — 21/0 l«Point b” of azureus (8.2°, Diagram 20) may actually lie still higher. In experiments with more material (20 specimens, Experiments 128, 129) the mean value of 10.8° was obtained. However, the difference on limestone and on granite here again would be 19 days (128 as against 109 days). *The measurements recorded near Rättvik (see p. 181) indicate that the temperature in deeper loose deposits, even when composed chiefly of limestone, is more unfavorable than in thin soil resting on limestone rock, especially at night. ‘(gr WeiseIq) siojd ylaney pue (0EI ‘d ‘oz weiseiq) „q qulod,, 0) SurpIoase spel 919m suoge]na[e) ‘(QUIT SNONUNUO9) JUOJsaWI] UO puke (JUN UdYOI1q) aque uO ‘soul [eJUOZIOY YIM poyseUI ose Syewmpolsıu OU) UT Aanoe In} MOYS pInom snainzy snypdıory yaıym SulNp spotted jenuue jeonayiodéyy “30jd poouejeq pue jeoidwy ‘9¢61-LI61 poled Jo ueow ‘oinjesodwis) wnwmiunu Aled “AQSIA 19 ‘Zp WeIseIq 061 .oF .0F 7 l AN 4 - Z 7 € Nn : \) 4 7 N : V auo)sawı] Jo sAep GOL : S ee 2227 ee eee ejiuei6 uo Shep 9 HL > ; :ENIUNZO 5 : 8 : 6 „ol = = LL = zı - ei VL vr y / vl 1 N t | N ! ! N N | ı 1 { 1 1 { ! t | 1 ! ! 0E 02 OL UL Oc Ob tL Oy Ob —f 02 Ol tL org Of oc Ol t 02,0, 0¢ Ol, 1 19qW8AON 1390190 Jaquiajdas isniny Aine sunf Ae Judy 192 173 fitarsis, H. anxius, Panagaeus bipustulatus, Leistus rufomarginatus, Microlestes maurus, and M. minutulus have their relict-like northernmost occurrence. It must seem daring (too daring?) to publish these two structures, the one in Diagram 42 and Table 8 and the other in the map just considered, on the basis of the Rattvik plots alone. However, I should emphasize again (cf. p. 186) that the temperature differences found near Rattvik must be con- sidered as abnormally small. Besides, it is not known where in Sweden (but perhaps in southeastern Norway?) a location may be found that has a similarly favorable rock condition and could serve as an object of study. The geologists 2044 OY > N: FELL d CL, isk ’ Oy SNK) % > [/ &, Poe’. HIT Fig. 17. May + September. Hypothetical mean of the minimum temperature of the soil at a depth of 5-6 cm with bedrock immediately below. Cf. map, Fig. 69 (p. 471). 194 174 are not aware of any cases that can be compared with the Sjurberg fissure*. If generalization may be made, for which the biologist feels a strong urge, they must be based on very limited material and consequently be purely hy- pothetical. The intention was only to show that the thermal characteristics of the southern Swedish Cambro-Silurian limestone actually have an important biological role. The conclusion seems to be justified that these factors are decisive, especially at the periphery of the area of a soil-bound species, i.e. they may determine the northern limit in Fennoscandia. It is noteworthy that the fairly rich occurrence of primitive limestone, especially in central Sweden (Sdm, Upl, Vst, Dir, etc.), for instance of marble, has no apparent influence on the soil fauna (however, it has on the flora). The south-facing slopes of the Kolmärden mountains at the boundary between Sdm and Ogl, where marble occurs, show a strikingly poor fauna. The widely distributed but dispersed occurrence of limestone in Finland seems to exercise just as little direct influence on the insect fauna. This seems to be related to the fact that the occurrence of Cambro-Silurian deposits near Äland is only submarine (Metzger, 1927), whereas on the Finnish mainland only primitive limestone rock is found (Eskola, Hackman, etc., 1929). Even areas that are characterized (Fig. 10, p. 115) by a striking admixture of limestone in loose deposits (moraine, diluvial gravel, etc.), but lack Cambro- Silurian outcroppings, for instance, in northern and western Upland (cf. map, Fig. 13, p. 123), are not characterized by any “limestone species” or “heat- requiring species” of the soil fauna. Both these facts provide further weighty evidence against the assumption of a chemical effect of limestone on the beetles studied here. Strangely, Holdhaus (1911a, pp. 742-743; 1911b, p. 342), in his study on the effect of rocks on the fauna of the Central and South European mountains, concludes that in the case of rock-bound (petrophilous) species “the chemical factors ...are more important [than the physical factors],” in spite of the fact that, with the exception of terrestrial snails, he could find no species of animals associated with a particular type of rock (1911a, p. 732). He of course implies a very indirect dependence (especially in the case of predators): “carnivorous animal—phytophagous animal—plant—soil (water)—rock” (p. 743), although he believes that the specific nutrient requirements of the animal is decisive. I cannot agree with him in this regard. Moreover, I am convinced that if his concept were correct, the distribution of animals on chemically different kinds of rock and soil would be far more defined and consistent. And, the carabids would not have turned out to be such distinctly polyphagous animals (see also p: 33). The concept developed here, that limestone has no role as a chemical *The Cambro-Silurian of the fjeld areas is markedly transformed, and even the limestone there is physically very different from that of southern Sweden. 195 196 Gs) factor for terrestrial insects—or only a subordinate role—would have received substantial support from breeding experiments, by rearing the carabids and other “limestone species” from the egg, on a completely limestone-free sub- stratum. I have repeatedly carried out such experiments, particularly with species of Ophonus (azureus, melleti, punctatulus, rupicola), but so far I have not managed to make these species oviposit, either on limestone-free or on chalky substratum. I hope the efforts of some more patient researchers will be rewarded with success. Conclusions The observations in nature and the experiments carried out on the effect of limestone (CaCO,), at any rate on the carabids studied, justify the following conclusions: 1. The chemical characteristics of limestone are of no perceptible direct or indirect (e.g. through pH) importance. 2. The physical characteristics of Cambro-Silurian limestone are decisive. Among these, the effect of three groups of factors is noticeable: a. Thermally, limestone is primarily characterized by high minima of tem- perature. For the animals investigated, this is the most important characteristic of limestone. Especially in Oland and Gotland, these and other thermal factors are intensified by the extensive horizontal surface of the limestone tableland, which is largely exposed to the sun, and by the small thickness of the overlying layer of weathered material. Furthermore, it must not be forgotten that flat limestone rocks are rapidly warmed through by the sun. Particularly in spring, this can be an important factor for insects buried underneath. Thus limestone favors thermophilous insects. b. Hygrically, limestone is characterized by its dryness. This is the effect of several factors: Limestone rock has a tendency to form fissures and easily undergoes weathering, as a result of which downwards directed drainage is facilitated. Evaporation of water from the residual soil is rapid, since the particles are flat (with a large surface for evaporation) and irregular in shape, which results in high porosity with good aeration. —Thus xerophilous animals are favored on limestone. c. Mechanically, low weight and flatness of the particles, as well as the porosity of the soil arising from these two factors, are characteristic of limestone—thus burying animals (with a nocturnal mode of life) are favored. This summary must lead to the conclusion that at least among the insects studied here (family Carabidae), none of the species can be called “limestone- dependent.” In central Europe, there are certainly some other kinds of stratified rocks, which largely correspond with Cambro-Silurian limestone in their ther- mal, hygric, and mechanical characteristics. Holdhaus (1911a, p. 732), the only entomologist who seriously studied the dependence of animals on rock, also 176 claims that in the Alpine area, with the exception of the “calciphilous” terres- trial snails, there is no other species of animal occurring only on limestone. However, it is very probable that at the periphery of its area and also along its northern Fennoscandian limit, one or other species may live only on limestone for thermal reasons. According to our present knowledge, absolute dependence on limestone is found in insects only if a species is directly or indirectly de- pendent on a limestone-bound plant, which means secondary dependence. Evidently we cannot apply our conclusions about insects to autotrophic plants without alterations. On the other hand there are certain responses common to all living organisms and not at least those against thermal factors. Hence it may not be presumptuous to ask the botanists too, especially the phytogeographers, to consider these entomological resuits. At present these researchers seem to consider the chemical effect of limestone as the main, if not the only, decisive area-limiting factor for so-called limestone plants. In fact some botanists apparently regard the pH as the only decisive factor (Olsen, 1921, pp. 144, 145). It would be more prudent to regard it as an indi- cator (Lundegardh, 1930, p. 350; Naumann, 1932, p. 10). Particularly in cases where a species of plant prefers soil with different pH-values in different parts of its area (for example, Pesola, 1928, pp. 208, 243), it would be interesting to find out whether thermal factors operate as well. In this connection the well-known phenomenon might be recalled that certain species of plants are completely dependent on limestone only at the 197 periphery of their total area, chiefly along the northern limit* (Andersson and Birger, 1912, p. 44 ff.; Salisbury, 1920, p. 208) or at higher altitudes (Adamovic, 1909, p. 70). As true counterparts of our species of Harpalus, which require the most heat, we should not forget to investigate the “pure” species of plants oc- curring only in Oland and Gotland (enumerated on p. 298) with regard to their dependence on limestone and requirement of heat, particularly in the more southern areas. It would undoubtedly be productive if the phytogeographers made a new examination of Thurmann’s old concept (1849) on the effect of the physical characteristics of soil on plants. Study of these questions, however, demands the close cooperation of the field botanist and the laboratory botanist (preferably combined in one person). At present such a situation unfortunately does not exist within the domain of Swedish botany. On the other hand the strong dependence of many plants on primitive limestone (for example, Eklund, 1931; cf. also Hope-Simson, 1938) also shows the undoubted effect of chemical factors, which, as far as is known, has no counterpart among insects. *If a dependence on limestone along the northern limit can be explained by the higher night- lime minima, it is conceivable that the lower day maxima of temperature on limestone (Diagrams 36, 37, 40) favor a northern species that is limestone-dependent only at its southern limit. 198 199 The Fauna of the Islands An Example of the Significance of the Dynamic Properties of Animals For the zoogeographer working on ecology there is hardly a more attractive task than to investigate the fauna of an island. The boundaries are drawn by nature and one need never doubt what should or should not be taken into con- sideration. The frequent pronounced poverty of fauna resulting from isolation and the limited number of biotopes permits the collection of sufficient material in a reasonable time. But the first question is: How did the animals get there? We cannot apply the fundamental question of the origin of new species and what caused it in our islands. There is as little justification for classifying them as continental or oceanic. All of them lie on the continental shelf and even though a few, like Gotska Sandon, have as far as is known never been connected with the mainland, they basically do not differ, at least biologically, from the numerous islands in the Baltic Sea basin, which had such a connection only during the interglacial or preglacial periods. On account of repeated Quaternary glaciations, biologically speaking, the Fennoscandian islands are very “young,” for none of them completely escaped it. However, the paucity of very primitive ancient forms of animals on these islands has a certain advantage. We can estimate the maximum age of each member of our insular fauna without too many uncertainties; we can judge with fair certainty any changes in their area from the very beginning, possible land connections, climate, etc. In an area that underwent glaciation during the Quaternary period the islands show the contemporary faunistic activity almost as Clearly as atolls emerging out of the sea. And thanks to the postglacial rising of the land, new islands are continually being formed even now. The principal theme of this chapter is: How does an island become popu- lated? And if it is possible to answer this question, at least for some cases or in a broad outline, then we will probably also be justified in trying to answer a general question: How does a species of animai generally extend its area? The choice of islands whose carabid fauna is analyzed here in detail was on the one hand simple, on the other hand quite difficult. It was simple because 201 178 only islands whose fauna have been sufficiently explored could be considered and it was difficult because as many different types of island as possible had to be represented. In a few cases a compromise was unavoidable. After some thought the following island regions were selected (see map in Fig. 18): 1. Hailuoto (Karlo) in Ob, in the northern part of the Gulf of Bothnia. 2. Aland (Al), main island. 3. The southwestern Finnish Skargard between Aland and the Finnish mainland, eastward up to Ab Nagu. 4. Hogland (Ka) in the middle of the Gulf of Finland. 5. The remaining “outer islands” in the Gulf of Finland: Tytarsaari, La- vansaari, and Seiskari. 6. Valamo (Kl), in Ladoga. 7. Värmdö-Ingärö and Djuro (Upl), in the Skärgärd of Stockholm. 8. Osel and Dagö (situated outside the region). 9. Gotska Sandon, North of Gotland. 10. Faron, immediately northeast of Gotland. 11. Gotland (excluding Karlsoarna). 12. Oland. 13. Bornholm (situated outside the region). 14. Ven (Ska), in Oresund. 15. Islands in the Skargard of Goteborg (Vgl, Boh): Branno, Styrso, Donso, Ockero, and Hono. 16. Orust, in the middle of the province of Bohuslan (Boh). 17. Hvaler (1), on the Swedish border. 18. Hitra and adjoining islands (9, 26), west of Trondheim. 19. Donna, Alstenoy, Heroy, and Lokta (31). 20. Lofoten and Vesteralen (the whole of Province 34) in northwestern Norway. 21. Five islands in Troms and Finnmark. These are not considered in the following tables because their inclusion would be unnatural. It was unfortunate that the Solovetsk islands in the White Sea have been studied too little to be considered. But fortunately the islands selected form a diverse collection. The following are the most important variations: a. Size. The largest island, Gotland, is 2960 km’, and the smallest, Ven, is 7.5 km? (some islands in the groups of islands considered are even smaller). b. Isolation. Faron is farthest from the mainland (125 km) but it is only 0.5 km from Gotland. The greatest absolute isolation is that of Gotska Sandon, with 38 km of distance from any other land. The outer Finnish islands and Bornholm are also quite isolated. c. Variety of biotopes. The greatest uniformity is shown by the outer islands in the Gulf of Finland, which are made of moraine or quicksand, and by Gotska Sandon. They almost totally lack fresh water and are therefore uninhabitable & Q = 2 200 Fig. 18. Islands studied and regions on mainland compared with them (a, b). 1—Hailuoto; 2—Aland; 3—Skärgärd east of Aland; 4—Hogland: 5—Remain- ing “outer islands” in the Gulf of Finland; 6—Valamo (in Ladoga); 7—Varm- dö-Ingarö and Djurö; 8—Osel and Dagö; 9—Gotska Sandön; 10—Färön; 11— Gotland; 12—Oland; 13—Bornholm; 14—Ven (in Oresund); 15—Skar- gard of Goteborg; 16—Orust; 17—Hvaler; 18—Hitra, Smola, Dolmoy and Fröya; 19—Donna, Alstenöy, Heröy, and Lokta; 20—Lofoten and Vester- älen. I-V—-5 islands of Northern Norway: I—Tromso; H—-Hillesöy; II—Nordfu ,- löy; IV—Kvalöy; V—Mageröy. 202 180 for many hygrophilous species. Moreover Ven, the islands in the Skärgärd of Goteborg and those in northernmost Norway are markedly uniform. The most vivid variety of biotopes is shown by the biggest islands, especially Äland and Gotland. d. Age, not geological but biological. In this sense some of the West Nor- wegian islands are the oldest: probably they were partly outside the ice cover during the last glaciation, and were never flooded during the postglacial period. In the Baltic Sea area only Hogland and Bornholm appeared as islands im- mediately after the ice melted. The youngest, also geologically speaking, are Hailuoto and two of the Finnish outer islands. e. Postglacial land connection. Such a land connection did not exist for any Baltic Sea island north of about 59° N latitude. This assumption is much disputed for the islands lying south of this. Such a land connection has been generally assumed only for Bornholm and Ven. —The fauna of the West Norwegian islands may have partially survived in situ during the entire period after an inter-glacial land connection. f. Ice corridor during winter. Biologically speaking such a connection can replace a land connection to a limited extent. In the Bothnian Sea such a firm land corridor is an annual phenomenon, but in the remaining Baltic Sea region as well as on the Swedish west coast such a corridor is formed only in especially severe winters. It is never formed along the Norwegian Atlantic coast. g. Influence of man. Unfortunately none of the islands considered is com- pletely uninhabited. However the influence of human culture on the Finnish outer islands, on Gotska Sandon, and on some tiny islands in northernmost Norway is minimal. The characteristics of each island or group of islands, in this and other respects, are described below along with an analysis of the fauna. A comparison between two or more different insular faunas should not be undertaken just to determine the common and peculiar species (Table 9). A more rewarding insight into the faunal character of a region is obtained by dividing the species into different ecological, dynamic and other groups (Tables 10, 12-14). But even the results so obtained must be used with caution. For instance, it could easily be imagined that the absence or presence of a species (or an entire group of species) capable of flight depends on its ecology, and that this is a decisive factor, whereas the flight capacity is of minor importance. This is true to a certain extent. Hygrophilous species, particularly nearly all of the ripicolous fresh-water forms, have well-developed wings. So on an island, which lacks such bodies of water, a decrease in the percentage of species capable of flight could be expected. —The xerophilous component in the fauna decreases going north, the flightless species become more common toward the west, and the same is true of the species that hibernate as larvae. Hence a comparison of the dynamical, ecological and other groups from the 203 181 islands alone would be subject to too many sources of error. For a more reliable criterion for deciding what is “normal” and what is peculiar in one or other insular fauna, regions of comparison, shown in the map (Fig. 18), were selected on the mainland. They were generally demarcated as circles with a radius of 100 km each, the center at the coastal point closest to the island. In the case of long island (Oland, Gotland) or widely separated islands (outer islands in the Gulf of Finland), the region of comparison was correspondingly increased in length. Irregularly formed regions of comparison are Estonia, Kurland, and Sjaelland, where the extact localities of some species were not known. For the same reason a desirable region of comparison could not be found in North Germany for Bornholm. I did not let the regions 19a and 20a in northwestern Norway overlap the Swedish fjelds, since they are unsuitable for comparison, and for that reason (in 20a) the diameter of the regions was correspondingly increased. Most of the Baltic Sea islands, as well as Ven in Oresund, were assigned wo regions each of comparison. It is necessary to emphasize that the regions of comparison were not se- Jected as the presumed areas of origin of the insular fauna concerned. They were intended only to show how the fauna has evolved under as similar con- ditions as possible but with better possibilities of immigration, and to serve as a standard against which the peculiar character of the insular fauna stands out more sharply. Some islands situated far outside the region in the North Sea and in the northern Atlantic were also considered for the purpose of comparison (Table 25). They are discussed at the end of this section (pp. 325 ff.). Before we pass on to a study of the individual insular regions it is advisable to give some more tabulated survey to facilitate comparison. It is imperative to form a judgment on the capability of dispersal (the dynamics) of the insular beetles. The easiest way to do this is to divide the species into “flight groups” (m, b, d, etc.) as in Table 10. In contrast with the areas of comparison most of the islands are con- stantly deficient in macropterous species. This is surprising, since the nume- rous species appearing more or less accidentally on the islands mostly belong to this group. Only the outer islands in the Gulf of Finland and Gotska Sandon show a clear preponderance of such species. The brachypterous species con- stantly occur in about one-half the number of islands with plus or minus values. However, the most characteristic feature is the large component of dimorphic species in the insular faunas. Only Hailuoto Island shows slightly minus values. Nevertheless, in the case of this island, the region of comparison (1a) has the largest number of dimorphic species of all, resulting in that the number of the island itself is relatively but not absolutely low. t (Contradictory to the ecological term “dispersion”; cf. p. 53; suppl. scient. edit.). 204 205 182 This large element of dimorphic species, which on different islands is be- tween 15.6% (Gotska Sandon) and 26.4% (Orust), is functionally quite hetero- geneous. Such a species may occur in one region exclusively in macropterous form, in another in brachypterous form, or it may even be dimorphic in the region in question. So a reliable division of the insular carabids was not possi- ble without a study of the representative material of the dimorphic forms from all the islands. Such material was not available from Osel and Dagö, which had to be omitted in the following tables. An account of all the specimens of dimorphic insular carabids examined is given in Table 11. We are now in a position to provide a reliab ¢ tabulation (Table 12) of the dynamic groups of insular carabids. The regions of comparison are taken into consideration but it must be emphasized that the data on these are only estimates, since complete material from all the regions of comparison could not be studied. However, the resulting errors cannot be large, since both forms of a dimorphic species are distributed on the mainland far more regularly than on the islands. The most important figures in Table 12 are the percent deviations from the regions of comparison especially the column “b + d,” that is, the sum of the species found to be brachypterous and dimorphic in the area concerned. In the case of the latter the macropterous form is generally far more rare than the brachypterous form, and if the material studied is too small, often the former may not be proved at all. In this connection Bornholm offers a clear example. The number of species found there only in brachypterous form (column “b”) is much larger than on Öland and Gotland, since insufficient material was available. On the other hand the sum of “b” and “d,” which undoubtedly corresponds better with reality, is somewhat smaller on Bornholm. The study of the dimorphic forms in the following section will show that the brachypterous form can immigrate through the macropterous form rarely if ever (either through a macropterous female that has already copulated with a brachypterous male, or by recurrent mutation). It is therefore valid to con- sider a dimorphic species that occurs in the region concerned exclusively or partly in brachypterous form as functionally belonging to the dynamic group of brachypterous species. This group shows the highest numbers (both absolute and relative) on the islands along the Norwegian west coast and the lowest on Gotska Sandön and the outer islands in the Gulf of Finland. The brachypterous group (b + d) in Table 12 is of particular interest from the viewpoint of the history of immigration. They include species whose dis- persal was especially difficult and for which air passage is virtually impossible (see also p. 590). Besides being passively dispersed by man or by animals they were able to colonize most of our islands by passive transport with water and ice, since, with few exceptions, there has been no land connection during the postglacial period. Obviously such accidental transport can only lead to successful permanent 234 183 colonization of a particular island where possibilities exist for breeding. This is dependent not only on the presence of a suitable biotope but also on the essential prerequisite that either an impregnated female reaches the island, or two or more individuals arrive more or less simultaneously so that pairing can take place in the new region. The latter possibility might be applicable especially in the case of dimorphic species, when a brachypterous individual that has arrived by one or other method is able to pair with a wind-transported specimen of the macropterous form. The pronounced occurrence of dimorphic species in the insular faunas (Table 10) can perhaps be explained this way. At any rate, transportation of imagines must be much more advantageous than that of larvae (or other immature stages). In view of their weak chitiniza- tion, the latter are more susceptible in every way, especially to changes in the humidity of the air and substratum. Compared with imagines their concealed mode of life seldom exposes them for passive dispersal. The above considerations suggest that the colonization of an island by any species takes place mainly in the season when it is adult. In this respect our carabids in no way exhibit identical behavior. Most of them are indeed adult hibernators (with spring breeding), but a few are quite irregular and seem to be able to hibernate at almost every stage. Finally there is a fairly large group (increasing toward the west) of 72 species (20% of the Fennoscandian fauna) that hibernate only or normally as larvae (cf. p. 568). It seems to be correct to take passive transport by water as the most important mode of colonization of islands by flightless species. To answer the question whether this transportation occurs mostly in the winter half-year (with drifting ice), especially in early spring, or in summer, we need to know whether the brachypterous and truly dimorphic species of one or other island (and the “regions of comparison”) are mainly adult or larval hibernators. If the latter, winter transport would seem to be of lesser importance. Table 13 was prepared to provide a firmer basis for such considerations. It classifies the flightless and (in the region concerned) dimorphic species into more or less pronounced larval hibernators [O + L + (L), see Table 9] and exclusively or predominantly imago hibernators [I + (I)]. The relationship between the sum of both groups of species (fourth and eighth columns) is expressed as an index number (ninth column). More important is the deviation from the index number of the regions of comparison (tenth column). Finally, it is naturally of special interest to undertake an ecological group- ing of the insular carabids (Table 14). It is not only an expression of the biotope characteristics of the island itself, but may occasionally also provide a clue to the origin and the immigration route not of the entire fauna, but of one or other species. Keeping in view the frequent occurrence of accidental, wind-driven migrants on the islands, which do not belong to the native fauna of the island concerned (see f. ex. Gotska Sandon, p. 282), which on the other hand cannot consistently be omitted, it seemed to me best to consider in this 184 case only the brachypterous species and the “true” dimorphic species (b + d, Table 12). The reduced reliability of the data resulting from the smaller number of species is perhaps compensated by the elimination of all accidental migrants. The ecological groups and symbols (x, n, h, etc.) are those used in Table 9. In the last two columns, the percentage deviations from the regions of comparison have been especially noted in the case of xerophilous and forest species. Details of the six tables given below are discussed more fully in the fol- lowing separate treatment of each insular region. 206-221 Table 9. List of carabids found on Fennoscandian islands studied Presence of species on an island (or island group) is indicated by + and absence by —. In the first 3 columns species are grouped as follows: A. Dynamic groups m—Constantly macropterous species; flight observed; (m)—Constantly macropterous species; flight not observed: b—Constantly brachypterous and flightless species; d—Dimorphic species; macropterous form observed in flight; (d)—Dimorphic species; flight not observed. Dimorphic species found in Fennoscandia only in one form, macropterous or brachypterous, are listed with this group (=A.) B. Hibernation groups I—Constantly adult hibernator; (I)—Predominantly adult hibernator; I’—Adult hibernator in the north, probably with 2-year development; L—Constantly larval hibernator; (L)—Predominantly larval hibernator; L?—Larval hibernator in the north, probably with 2-year development; O—Species with irregular hibernation, i.e. both stages (larva and adult) C. Ecological groups x—Strongly xerophilous; (x)—Weakly xerophilous; n—Mesophilous (including ecologically ubiquitous species); (h)—Weakly hygrophilous; h—Strongly hygrophilous; a—Arboricolous; w— Pronounced forest species; (w)—Predominantly forest species; k—Exclusively synanthropous species = 35 I 3 4 5 aC S E = & ZEN LE 5 fa) = gs &% Ss esog = 3 = Sd a 18 Os m = 2 S Z $ > ij are ee ee pe | —~—{+]4+)4+] +] +1 +|I+| +|-1+1]- 1) - | - a a ee | | | ee IR a ee | | ee ee +/|- | - | +-.]|+|+!|+1|+1|- | - 1|I-1|- | - | - us td we wer ane u Zu — Lu zus = + ei cue - | +1|1-|- | - | +|+I|+|+1|I+1-1- | - | - am pees es u aan ‚aa ne = en Lil Ze au + Bites en Bun = — _ + — == — — ne a ae 186 Agonum dorsale SUNN SEAS AA SARA LAA RA RA RAR AR ARR A RA Rp ericeti fuliginosum gracile gracilipes Krynicki livens lugens mannerheimi marginatum micans moestum mülleri obscurum piceum quadripunctatum ruficorne sexpunctatum thoreyt versutum viduum mara aenea apricaria aulica bifrons brunnea communis consularis converiuscula crenata cursitans curta equestris erratica eurynota famelica fanuliaris fulva fusca mfpfina ingenua interstitialis Dynamic group (m) BEER EZEREREFE EEE EEE EEE — ws = ° = 3 is _ o = x (1) (I) Ecological +] Be Ms Seo ee ~~ -*~ i—~ Stes) Za — = — — n? Lal Hailuoto N Aland ° aie || [ote latest IF FF FH Hr HH HH Hl 44441 I++++ 1441 | =a | fe} w I 25 ee a8 A) I+++ | lar | | ++ I + I +++ tt HH Hl HH ti) ttt este | | | 1+! I ++1Iı Hogland + Er +++ I ++ tei ++i ti 5 8 & © (-4 5 be © ~ =) io} to") = este steel Kerala “a uc) =| 8 = “a — | JL ++++ ++ +4 187 "939 ne | ee een fat cae ale ae ey 3) gyal She tah) ale 5) Se ae Se ay ayy sep a St 2 Masta obese eee ee aa FL] L+H ttt] P ttt tl ptt ti tsti Dee eee eee fo a Se cores (ales te eld ele ee Ea ESUPSaenenene sae crane seer ar nena ar ane 12 moon fot a aa ee Ee eS hae | eee | + ++++++ = v0. +++ 1+++ +L tHE] + HH HH HH HH HH HH HH HH HH I HH | Its tt tt tt HH HH HH HH HHH Fee tees IH FI +++ HH HH I FH HH HL TFT III FF IH | reed tree II FI II FI I III tel bree FF HH HH HH EI FI ttt I IH | 188 Amare littorea lucida lunicollis majuscula municipalis nitida ovata plebeja praetermissa quensel simtulata Spreta tibialis : torrida Anisodactylus binotatus Asaphidion flavipes A, pallipes | Badister bipustulatus! dilatatus peltatus sodalis striatulus V. HANSEN unipustulatus embidion aeneum Andreae polonicum argenteolum articulatum assimile azurescens biguttatum bipunctatum Clarki dauricum dentellum difficile Doris femoratum fumigatum gilvipes grapei guttula hasti SS OSS wae Baers BR snes dose Do ete 1The subspecies lacertosus Sturm has not been omitted here (comp. supplement). Dynamic group (m) B55 558 8 z Boa 8 8.95 8 8 Bos E iS) ‘> 3 = baal oO x 2 ee ee | wv wv ee oe = O I Ecological group Gee —_ YY nz Perser Efe = 3 saps S — h = Hailuoto PEI b + I t+tttts4 | I+++ | sel 3er ised | | Fer se ee | [+l + I + I +++ | I+++1 Hogland + [steel ER Sele] eae td Lae sl £ Gs) = © ia 5 — 34 =} °5 wna & —_— I++ I +1 | Lord On Valamo 189 "919 U2107077 ee ee ee ee ee ee oe poesia acess zer eb eee le lee oe le ele eek Gee i eee ee alla ete el | dele el else ll el eal le ieal tek = Pee eee ee emo EEESEEEEZEEEEEESTEIERETEIENEIEEZEEEHENEEN ee ae De es ee ee a HI | = A ri + = a PL TS ee ele as |e ee ee a ee Se is fen oe ee ae oO umoyuog} | ++] | |+t+++ttl+++t+ttee+ ltl ter eel I HIHI HH IH 2 FHtttHtHHHHH +l ttl HHH HT tt THF HH ei ti II | 13 ree: FHEEEEEHEHE HEHEHE ISL HH Feet HIT et HH TH HH tit — Er cles ete) lect ee let tee | ot et Ea eee el eel uopueg 248109 Fee let ete eet ee ete cet ol een Spey | Bu a a u VFL PEEL tte ee Be Be ttt ttt "339 |. | ae tele tet ee ze rer 190 Bembidion humerale illigeri lampros litorale lunatum lunulatum minimum nitidulum normannum DFJ. obliquum obtusum octomaculatum pallidipenne prasınum properans quadrimaculatum quinquestriatum rupestre saxatile Schippeh semipunctatum Stephensi transparens unicolor ustulatum varıum velox virens Blethisa multipunctata Brachynus crepitans Bradycellus collarıs Panes I IT, ı 5 5 h , h DT LS m m D B. harpalinus B. similis B. verbasci Broscus cephalotes Calathus ambiguus (& erratus CG: fuscipes Ge melanocephalus CG: micro pterus @ mollis tenellum Er., WGN. Dynamic group = ° 'S 3 I im o > ae) Ce a ne en — ee AHA A Ecological group sr er |) 8 Sse er se Beer er ser Br er — >) = Hailuoto seele ser D o III +11 1 +++ Äland I++ 1441 IH I + I +++++| | t++4+4+4| to} 3 Ks} ond a ae ali | ars locale gle ae Bee eae || I++++ 41 Hogland + I+++ +1 | | [eal Sete ate Fee 5 bs} E © fad 5 =I 34 = °o5 Ces) eos — ase sale all al | | ae ae tein I) | geal seal Ov Valamo J++ + 191 "979 ua}ojor7] hs aishe steels bes healseeeledisecearetisie eqs siy se lores Bee Sector suusa 2 ee eae ee ee ole ele pote ee | ak ie el ee ee pose ls eels es ee lea ele dele sl ee || el | bale ea eae | = Pitt) lth I FI I HI ++ ++ 2 Pitti II FI I FI I tb tte I ++ Ba Flee youl lag sal sal ee III + I I I tttttest | -- ob Blase! Peel eofeesleactoeliemte heel ele see le te ot [ote maleate starstyle ee I tet laste et | LHHHHHHHEH | LEH LH HHH tt Lt tt HH HH HH HH [+ttl Ltt a +l eeeee tt ttt tt Pee tte ee ae [3 peo, Ft] 1 Li ltt iti teeee et lp tttt ttt) bd +4+44444 pa Pitt LPH Hee Peel tt ett rte el Pe b+ tttstes uopueg eon | | LLL LT EE EEE PEEP HEHEHE II III FF FI FI FI HH tt tr HH © pu | a Ba Be a a a rte ib FF a a Be a bette be a Be Bee er Botello ose) oslo tle ee | 192 Calosoma auropunctatum @& inquisitor (Ge investigator (& reticulatum (Co sycophanta Carabus arvensis cancellatus C clathratus M C convexus G coriaceus @ glabratus G granulatus C hortensis CG: intricatus C. nemoralis C nitens Gy problematicus C. violaceus Chlaenius nigricornis €. nitidulus SCHRK. @* quadrisulcatus Gz sulcicollis C. tristis Ge vestitus Cicindela campestris & hybrida G: maritima C: silvatica Clivina fossor Cychrus caraboides Cymindis angularis G humeralis C. macularis (Oy vaporarıorum Demetrias imperialis ID); monostigma Dichirotrichus pubescens Dolichus halensıs Dromius agilis D. angustus D. fenestratus D. linearis coOCB BS SRBBR BEES Ce ler ler fen er ep len ten ep fer {24 (ep er! — 2 — BE Ele) El ee) = je — (=) ° = Par) Gs) = ha oO Be) — = -—_ HOOOOR HHH eH HP POR HH OS e ses (L) Ecological group f=] pie eRe tl IA — = Hailuoto 2 3 4 5 ne = Ido <= | lee + SH et, un + kes Aa Fu el ee | m + Ss yess ae | Eı el — —> + oo Ge beam | © m | ae = El NE ae ll ar - I - | =] + Alten eel at Se oe at ee alate sah al haan +. + ee lee Fu Bl ae) | se len || is le islands Fre ee 193 "29 u9]0j07] | © 2 eng a] = in Pigdıens ™ s31oqs}0n +++tt] tt) +HHH i++) HH I t+4+4+141 uopueg BYS}]OD | | | = | = ° oa pe eae pue [9sQ | i ee a ee ae ee a ee ee PEELE EIFP PHI FIT ad det I FH et ee ea eae I III IT IT FF TI HI TH II TI a Va Du a Be a Ba Be a a a Ba a BB a Be a titi ti Ua I Pas u u Bes Be Ver a ea PPP ee Be a a Be a ae EB BE +t Ua Ver Ta a ee a Be ee Ba ae a u u a Sa Ba Be a a Be a Be a BE ee Fre les Peet eee ee a ae +++ il I HI FHHHH tHe Fee HH HH HH HH HH I TH HH I++t+t+ttt+ I ++ 14144 II III I tee FH FH I FI FI I FH HH HT HH 4tt4+ III + I III ++ I FH IE III FI IF I FI FI FI I Ir + rettete | sesearsesearaesr lsese ee el oe sears ie ares eee | PT ee ee se ie cP secs Tse ee: 194 Dromius longiceps D. marginellus ID), melanocephalus ID) nigriventris ID), quadraticollis JD) quadrimaculatus D. quadrinotatus D. sigma Dyschirius acneus angustatus globosus impunctipennis intermedius Lüdersi obscurus politus salinus septentrionum SEI : thoracicus slaphrus cupreus lapponicus riparıus uliginosus Harpalus aeneus anrıius azureus calceatus distinguendus Frölichi nf m © fuliginosus griseus hirtipes latus luteicornis melancholicus melleti neglectus picipennis pubescens punctatulus puncticeps puncticollis MEET TET ATESet Tats Dynamic group = ° > 3 = kee vo = x [OR ET nn TE ee ne a oe I ee ee ee ce oe en Eu} CS oS Ecological group _ Hailuoto | ar Pe ee ae | ar [arar | N Aland ore Se |e Se a eee Se eae sea ae tse ae | w ac) iS} 8 < o A Go IH I ++t+H+ I + I + | er ee ee I ee ass el | | sy om. on be Ho} a ep) Hogland + = ss) E oO [4 5 mi 25 3 °o5 oe” fey re BEE Se] Feel tl ee | 933 U9J0J07T 195 ‘39 vH Pr P= pmo in PABBILAS = s31043J09 eg Peete = uopueg BYS}OH os odeq ” pue [2sQ "939 N UNPUNEA et ee ee ee | er i a i Bl lee a ate ate +++ | er lla te | ++i) ++ 14441 1441 ++++1 4 }++++14+41 IH I + I ++i +++ I ttt East 4 eae ale elle Fade [|| ee ee eerrereleeeeeee ee een ee ee Pee Eee! +l ++t+tttt++e+ HH HH 4 ++++++ ) +4444) +4+4+4+4+4+4+4+ ++++Ht1 I | I t+t+t44e0 1 +41 +i tl tt) I I + I HH IH IH Da Pu Bu yer Bu Be a BE ee u +++) u Be Bu Bu se u Bu BB a ES +++ I te I I II HI I II IH I + 196 Harpalus quadripunctatus rubripes rufilarsıs rufus rupicola seladon serripes servus smaragdinus tardus vernalis winkleri Lebia chlorocephala It crux-minor 1D cyanocephala Leistus ferrugineus L. rufescens L. ruformarginatus Licinus depressus Loricera pilicornis Masoreus wetterhalli Metabletus foveatus Bee: M. truncatellus Microlestes maurus M. minutulus Miscodera arctica Nebria brevicollis N. gyllenhali N. livida N. salina Notiophilus aquaticus N. biguttatus N. germinyi N. palustris N. pusillus Odacantha melanura Olisthopus rotundatus Omophron limbatun Oodes helopioides Panagaeus bispustulatus 12. crux-major signaticornis DFT. (m) 3 nn SBOE SE Se ee ESB Du 7 —_~— wa wa g ° > S fa fee o = E — — SS — wa Ecological group ii Hailuoto [414 144 | Aira © +++ + tei ttl ttt+ttt 1 ++i +1 +) +++4++4++4 | a I =} = og Al (7 I ++ I I+H++H+ | oe Kara |++++ | lead ar | ° ar Skärg Hogland + lose Beer t | less} bi FI + I I ++ IH | ' i=) 3 E © fa ing outer ™ islands oO Valamo Fee] lar 197 "939 SE eneı Serra ar eae ee eee ee ee el = ole ruugg Be Peel tt a alse al sl sleet ak eb oh al Sh Pa] etal Se] fo REO Sa eee Fr SEI eee Poe eee Sse | +I LPP titi ttt petted I FI FF I HI IT ttt eet mo FH II IF I I EI FI I FI I FH II FI I FI TI FI FH I FH HI Pe tess ede Fe ER re etre ease poe eee Pied Bey IH I I II I I FI III TI FI TI FI HF HIFI FI FF HH I IH + 2 eee HH HH tt H HH HH HH HH HH HH HH HH HH HH HH HI HH HH HH + = s31oq3109 So | ++ FF FF HH FF HF HH HH HH HH Hr HH HH HI H HH HH HH HH +++ = ro | ut a BB a na a BB nn 2 Ba na a a | BB ea Be a a a a a a a a a +t uopueg DB PHIL LLL Pee PE Ba Be ee a ee a a ttttaeei tt = ersıon wel el lee tes IESE: te lee dee lea lrte te lle eee ole |e al odeq > pue jasQ PILL pte Pe Leet HH tee TI IT t+ teeet 1 tet Il ae ehe are rar Ss la are ee in 198 Patrobus assımilıs b P. atrorufus b IB septentrionis m P. _sepientrionis australis | (m) Pelophila borealis m Pristonychus terricola b Pterostichus adstrictus m P. angustatus (m) ER anthracınus d IP, aterrimus i m JP coerulescens m 12. cupreus m i diitgens (d) IP, gracilis m 12, lepidus (a) JP, minor (d) IP. niger m IP, nigrita m B: oblongopunctatus (m) 12, punctulatus (m) 1% strenuus d I vernahs (d) 1% vulgarıs (d) Sphodrus leucophthalmus m Stenolophus mixtus m SS skrimshiranus STEPH. (m) I: teutonus (m) Stomis pumicatus b Synuchus nivalis (d) Tachyta nana (m) Trechus discus m cn fulvus b 2: obiusus 5 10, quadristriatus m T. rivularis (a) I, rubens - 7m secalis 5 Trichocellus cognatus = an, placidus (m) Zabrus tenebrioides Total species Hibernation group — — bw OH FORE u = pn u lL een I aoe cn ce oe as En En mn an [| fe — YS —— — ma See Ln tan il ae il een Bl GD Gen en on Lo eae ihe HHH Hee Ecological group = Hailuoto J ++++ | lSsteal: sel IH t+ i + w Alands Skargard | +++ + + + + + + + el ee licks leche: leche Hogland = ae bear | | +++] +++ [Peter leche ete al ing outer ™ islands I FI +++ +++e4 IH | Fat it | ++ | +++ | | | +++ ar 199 3 U9]0J07] tet le er Sle er elles Be Fee! 64 le le oe el et ls lee oe BER ER FE Fe 46 2 99 eI sae | De) ISNIQ = te ole ela ae lee es eee | et ELF Bee EEE 49 (Pte te eta ese EB ee eae ee eee ee ee | 87 Bessere un PABSIZAS ™ $910q9}05) oo uopueg 2?21sI09 104 Fa FELL ELF FELL EFF | KELLER LT EL LEI HT IF I II FI I FFFFFHH FH HEHEHE HI HH THF HH HH | 221 | FHP III | PEE EHH EEE HH HH ET HH TI FH HH HH 195 tea el lest te ker er Ei le lee le le lee le ee | 134 ° O el leis | ela) ce let) tl | lel Wel tel tet el elt et 90 oo ned pue |2sQ Bere ae) Stee el eine Ele tS | ee else er ltt fe | 131 | Fe |e 104 | |++++ 1414 +++ 200 Table 10. Comparison of dynamic groups of carabid species of islands and of the mainland. Cf. explanation with Table 9 Last three columns give percentage deviation of insular faunas from those of mainland regions compared. 1. Hailuoto Region of comparison 2. Äland Region of comparison a 3. Alands Skärgärd Region of comparison 4. Hogland Region of comparison a 5. Remaining outer islands Region of comparison a 9 b 6. Valamo Region of comparison 7. Värmdön Region of comparison 8. Ösel and Dagö Region of comparison a 9 b 9. Gotska Sandön Region of comparison a 10. Färön Region of comparison a 92 b 11. Gotland Region of comparison a 9 b 12. Öland Region of comparison a 93 b 13. Bornholm Region of comparison m Species % 31=58,5 61=51,3 89 = 56,7 94 = 53,4 65=52,4 94 = 53,4 46 = 52,9 82 = 55,8 124 = 53,0 48 = 62,3 112=52,3 124 = 53,0 30 = 50,8 100=53,2 55 = 52,8 99 = 51,9 69 = 52,6 124 = 53.0 113 =52,6 61 = 67,7 99 = 54,1 113 =52,6 71 = 53,0 99 = 54,1 113 = 52.6 106 = 54,4 109= 52,9 113 = 52,6 118=53,4 111=52,1 113 = 52,6 122=54,0 138 = 52,5 Species% 4= 75 18= 15,1 20=12,7 28=14,8 29=16,5 12=9,7 29=16,5 12=13,8 19=12,9 43 = 18,4 9=11,7 38 = 17,8 43=18,4 8= 13,6 35 = 18,6 12= 11,6 29=15,2 13 = 10,0 43= 18,4 33= 17,7 9=10,0 24=13,1 38=17,7 17 =12,7 24 = 13,1 38 = 17,7 30=15,4 30=14,6 38=17,7 33 =14,9 32=15,0 38=17,7 39=17,2 48 = 18,2 | d Species % 4= 76 75:9 10= 6,4 Io= 5,3 8= 4,5 10= 81 8= 4,5 6= 69 7= 4,8 eS NA, 6= 78 9= 42 Il= 4,7 = 3,4 =) AR = 5,8 I1= 5,8 ıı= 84 II= 47 II= 5,! = 6,7 I1= 6,0 ILL 5,1 IZ2= 9,0 I1= 6,0 11= 5,1 I2= 6 12= 5,8 II= 51 17= 77 12= 5,6 II= 5,1 14= 62 I2= 4,6 (d) Species% 7=13,2 19=15,9 19 =12,1 22=11,6 23=13,1 20 = 16,1 23=13,1 12= 13,8 18=12,2 25=10,7 10= 13,0 26 =12,1 25=10,7 11 = 18,6 24=12,8 15=14,4 22=11,7 16 =12,2 25 =10,7 20= 9,3 8= 89 21=11,5 20= 9,3 18=13,4 21=11,5 20= 9,3 23=11,8 27 =13,1 20= 9,3 25=11,3 28=13,2 20= 9,3 23= 10,2 30=11,4 b Species % 7 =13,2 14=11,8 19 =12,1 26= 13,8 22=12,5 17 =13,7 22=12,5 Il =12,6 21=14,3 31=13,2 4= 5,2 29=13,€ 31=13,2 = 13,6 21=11,2 16= 15,4 27=14,4 22 =16,8 31 =13,2 33=15,3 6= 6,7 28=15,3 33=15,3 16= 11,9 28=15,3 33=15,3 24 = 12,3 28 = 13,6 33 =15,3 28 = 12,7 30=14,1 33=15,3 28=12,4 35 = 13,3 Percentage deviation from region of comparison + + + nn 0,4 — ,0 Sully 0,2 |+ 1,25| — 1,05 78 + 6,6 + 1,2 3,35| + 45 |— 1,15 325|+ 4,95| — 8,2 74|+50|+ 24 37|+ 2,7 |+ 10 8,25 + 5,7 + 2,55 8,95| — 0,35| — 8,6 3,05) + 6,45| — 3,4 0,9 |}-+ 1,25) — 2,15 0,4 |+ 24 |— 2,0 0,5 |+ 0,4 Tai 0,9 } 14. Ven Region of comparison a b 15. Göteborgs Skärgärd Region of comparison 16. Orust Region of comparison 17. Hvaler Region of comparison 18. Hitra & c. Region of comparison 19. Dönna & c. Region of comparison 20. Lofoten, Vesterälen Region of comparison m Species % 58= 55,8 140= 52,4 128 = 50,8 46 = 52,0 115 =53,2 45=51,8 115 =52,5 56= 52,4 104=52,8 19= 38,7 55=50,4 23 = 50,0 40= 49,4 29 = 45,3 34 = 45,3 (m) Species% 13=12,5 50=18,7 52= 20,6 10= 11,5 34=15,7 6= 69 36= 16,5 15=14,0 30=15,3 4= 82 17=15,6 4= 87 15=18,5 I1=17,2 19= 25,4 d Species% 10= 9,6 Il= 4,1 Il= 4,4 IL = 12,6 Il= 5,1 10=11,5 II= 5,0 = 75 10= 5,! = 82 5= 4,6 4= 87 = 49 6= 9,4 3= 4,0 Species % 9= 86 36= 13,5 32=12,7 9=10,4 28= 13,0 13=14,9 29 = 13,2 15= 14,0 31=15,7 17 = 34,7 18=16,5 9=19,6 13=16,1 11 =171 10= 13,3 — 45 —10,3 — 1,7 —I19,I 201 Percentage deviation from region of comparison — 2,95|+ 7:45 +71 + 8,6 + 3,4 + 09 517 + 43 — 45 — 2,6 +17 — 17 +18,2 1.135 + 38 202 224-227 Table 11. Insular specimens of dimorphic species examined m—-Macropterous specimens; b—Brachypterous specimens; + —Recorded on the island but material not available Alands x By © Skargard o Hogland # Remaining © iS) oO 3 aD) Agonum fuliginosum = 4b 2b 2b Ib Ib I9b — + A. moestum — = im — = =! =e oe Tha A. obscurum — Om 2b 3b — en a ı8b Bart, aie Amara infjima a — — — we — a en KM Bembidion aeneum — — — Er N wi, a ie er D assimile = 2 by || 2,m 3b _ — _ — Im 4m B Clarki Se ri FE = rs DE m zZ as B gilvipes = 2b ii |) Dam lim = — au 2 B Lrapet FF = Im — — 2 = aa en B. guitula eo 5m 6b um Sb 2m — = Im 2b Nam 7m B lampros Ib | 2m 6b Alo) || Sm Im Ib Ib| Im 6b — 2b B obtusum gs == — — = _ = = 4m DB: properans m 7b Iob — —_ — = — ıb B. schippelt Ib = = Ib == — = aa = IB. transparens 29 | ipa BOY win 7b || Boo Im — = — 3m B. ustulatım ET 2b — — we — 2b =a Ber Bradycellus collarts a Ib — — Im — ı3b| 2m1b tas Bs harpalinis | fae a — — nae — ue au 2m Calathus erralus FR 1om 4b/ am ıb | 2m 3b | ım4b == Iom 9b IIb 7b Gs melanocephalus Ib 4b 7b Ib] Im Ib 8b | 2m 3b 9b & mollis Sg = — — = ar as Iom = Carabus elathratus Ib 3b Ib Su Su eu une iss Ib Cymindis macularis Ib 3b 2b. Sm — = = ou 2b & vaporariorum Bar — Ib Im an an u NN we Dromius linearts ms 2b ıob — = -— 4b _ 2b /D). nigriventris En 3b 29 b — nite — a 6m 2b Ib D. sigma Ib 4b 8b — Mult a zn Par 2% Harpalus azureus — — — = | Su sa a a 3m Fp neglectus — — | — = a aes N Au EN Jets picipennis — = — = ay as an ans Bi Masoreus welterhalli — — — = faith = u ıb| Im 3b] Metabletus tr uncatellus = 6b 9b 3b = Ib 5b 5b| Im 6b Microlestes maurus — — oe — == = = un Ib Notiophilus aquaticus 2b Irb 8b = Ib = 2m 5b 1b 4b bl = Gotland 4b Im 14b Im Iob Byane 1 19) 13b Im 28b 5m Iob Im 2Ib 5m 13b 2m. 6b 34 b 5b Iob 25 b 15 b Im 3m me (5b | ı2b| 7b 44 nm IO4A b 3b IIb 16b 13b Im 4m 3m - N Oland 2b 5m 9b] 8b 6b 34m 7m 3b 2m 4b 3m 3b sm 3b 28m 4b 8mIob 14 b 3b 3m 6b 4m Im39b 9b 25m 47b 3b 13b 1b Iob 3b 86 b 25m 2b 2miz 3h 20b 6b 13b IIm 2b ies) Bornholm Ib Im 15 un Ba Q 0 2g 3 zb 6b 16 17 2 s é | g Im Im Ib 6b Im =. 3b 9b + + Imob + 4b IIb 3b = Im 4m Im 3b 4b Ib = 4b 2b | Im 3b 2b 6b 203 = Hailuoto Alands a Skargard Hogland # Remaining Notiophilus biguttatus N. Lerminyt N. palustris Olisthopus rotundatus Pterostichus anthra- cinus diligens lepidus minor strenuus vernalis vulgaris Synuchus nivalis Trechus rivularis We ge oY oo oS _ ies) Bornholm Goteborgs Skirgard “" = OV Orust = N Hvaler 2m2b 205 Mageröy 206 228 Table 12. Comparison between dynamic groups in insular and mainland faunas after separating dimorphic species (see Table 11) d—Dimorphic on island or in mainland region; d?—Species for which no material was available (ignored in calculating percentages). Osel and Dagö could not be taken into consideration. Num- bers for mainland regions (= regions of comparison) are largely based only on calculations. Last two columns give percentage deviation of insular faunas from those of respective mainland regions Percentage devia- tion from region of comparison ee m + (m) d Species% | Species% 1. Hailuoto | 36=679| I= 1,9 16=302| — +60 | +0,5 Region of comparison 82=68,4| 10= 8,3| 28=23,3 2. Äland 109=69,4| 8= 5,1] 40=25,5| — | +43 | +0,7 Region of comparisona | 132=69,5| 18= 9,5 | 40 = 21,0 i b | 124=70,8] 15= 8,6] 36=20,6 3. Alands Skärgärd | 81=65,3| 6= 49| 37=208| — | +92 | +5,5 Region of comparison | 124=70,8| 15= 8,6| 36 = 20,6 4. Hogland 65=74,7| 4= 46| 18=20,7| — +05 | —4,3 Region of comparisona | 101=68,7| 15=10,2| 31 =2r,1 i b 168 =71,5| 21= 8,9| 46= 19,6 5. Remaining outer islands | 63=82,9| 4= 5.3} 9=1 1,8 I --8,3 | —i1,4 Region of comparisona | 153=71,5| 17= 7,9| 44=20,6 PR b | 168=71,5| 21= 8,9| 46= 19,6 6. Valamo 39=67,2| 0 19= 32,8 I +157 | +7,1 Region of comparison 139=74,3| 16= 8,6| 32= 17,1 7. Värmdön 67=64,4| 8= 7,7| 29=279| — +67 | +44 Region of comparison |130=68,8| 19=10,0 | 40=21,2 9. Gotska Sandön 76=84,5| 3= 3.3) ıı=1ı22| — —88 |—15,6 Region of comparison a | 125 =67,9] 20=10,9 | 39=21,2 e b 151=69,9| 20= 9,3] 45 =20,8 10. Färön 97=73,5| 3= 2,3| 32=24,2| 2 +32 | —4,6 Region of comparison a | 125=67,9| 20=10,9 | 39=21,2 | a2 b | 151 =69,9] 20= 9,3] 45 =20,8 11. Gotland 136 = 69,8] 20=10,2| 39=20,0| — —1,0 | — 0,7 Region of comparison a | 142 =68,3] 22=10,6] 44 =21,1 ü a b | 151=69,9| 20= 9,3 | 45=20,8 12. Oland 153=69.2| 21= 9,5| 47=21,3| — +02 | -0,1 | Region of comparison a | 147=68.4| 22= 10,2 | 46=21,4 | b 151=69,9| 20= 9,3] 45=20,8 207 Percentage devia- m + (m) d b d: tion from region ie of comparison Species% | Species% | Species% |Species Vi 13. Bornholm 164=72,9| 10= 4,4| 51 =22,7 I +3.4 | —23,0 Region of comparison | 188=70,9| 26= 9,8| 51=19,3 14. Ven 79=76,7| 5= 49| 19=18,4 I —0,755 | — 5.4 Region of comparison a | 191=71,0| 28=10,4| 50=18,6 i DJ 182 71.061,22 18,7| 50-197 15. Göteborgs Skärgärd 56=65,1| ı= 1,2| 29=33,7| 1 +14.4|+ 5,1 Region of comparison | 153=70,2] 23=10,5 | 42= 19,5 16. Orust 56=65.1]| 2= 23| 23=326| 1 +13,6 | +5,0 Region of comparison |155=70,1| 24=10,9 | 42= 19,0 17. Hvaler 73=68,9| 4= 38| 29=27,3| I +55 |+9,1 Region of comparison | 136=69.0| 18= 9.2] 43=21,8 18. Hitra & c. 24= 49,0] 0 25=51,0| — | +235 |+19,8 Region of comparison 75 =68,8] 4= 3.7| 30=27,5 19. Donna & c 27=58,7| © 19=413| — || +129 | +92 Region of comparison 55=67,9| 3= 3.7| 23= 28,4 20. Lofoten, Vesterälen 39=609| ı= 1,6| 24=375| — | +145 |+ 11.1 Region of comparison 54=720| 3= 4,0] 18=24,0 Note: Following dimorphic species are denoted “b” (brachypterous), although no material was available: Agonum fuliginosum (from Hitra), Bembidion transparens (from Lofoten), Cyminidis vaporariorum (from Bornholm, Hitra), Dromius nigriventris (from Orust), D. sigma (from Born- holm), Metabletus truncatellus (from Bornholm), Notiophilus aquaticus (from Hitra), Prerostichus diligens (from Orust), P. vulgaris (from Hvaler). Bembidion lampros from Hvaler is denoted “d” (dimorphic). 208 230 Table 13. Comparison between flightless and dimorphic species (b and d, Table 12) of islands and mainland (= regions of comparison according to hibernation type (O, L, I, etc., Table 9) Ösel and Dagö omitted. b—Constantly brachypterous species; b(d)—Dimorphic species, but brachypterous in region of study; d—Constantly dimorphic species, including region of study. Last column: Index of deviation from fauna of mainland district compared. A. Larval hibernators B. Imago hibernators en eviation O+L-+(L) I + (1) Index from region A:B of Jorofa [reat Petal a fro] *® | oman 1. Hailuoto 3 4 0 7 4 5 I 10 0,70 — 0,03 Region of comparison 8 5 3 16 6 9 7 | 22 0,73 2. Aland II 5 4 20 8 | 16 5 | 29 0,69 —0,14 Region of comparisona | 14 4 7 25. 100 | Kroll rar 0,81 “ b| ı 6 Rll 2s 9 8% ToW 27, 0,85 3. Älands Skärgärd 10 7 2 19 6 14 2/22 0,86 +o,01 Region of comparison 2 6 5 23 9 8 | 10 | 27 0,85 4. Hogland 7 2 I 10 4 5 3 | 12 0,83 +0,14 Region of comparisona | jo 4 5 19 || 10 Gel Tor 1836 0,73 ni eG) | Sa Re 1a |) 28 || eee 5. Remaining outer islands | „ a a 5 3 4 - 8 0,63 —0,03 Region of nenn 13 ; 6 24 || 15 Tol oat 136 0,67 : 12 6 7 25 15 9 14 | 38 0,66 & Valen N 6 5 o II 2 6 0 8 1,38 +0,75 Region of comparison Värmdön Io 4 5 19 12 7 II 30 0,63 % Sum 5 10 4 5 19 6 10 2| 18 1,06 +0,28 Region of comparison x T4 3 8 25 II Ic II 32 0,78 9. Gotska Sandön \ N > ı 9 R ‘i 2 5 1,80 +1,12 Region of comparison a 8 » bel 4 2 25 12 7 13 | 32 0,7 vs b II 5 6 22 17 7 14 38 0,58 10. Färön 9 8 2 19 7 8 1 | 16 1,19 +0,51 Region of comparisona | 14 4 GN) 250 12 7 | 13 | 32 0,78 DD b| ıı 5 6 22 17 7 742038 0,58 11. Gotland 13 5 6 24 I II 13 | 35 0,69 0,018 Region of comparisona | 14 7 m 280 12 om 25011036 0,78 : ag Dalai 5 6 22 || ı7 za 1a 1038 0,58 12. Oland 14 8 5 27 | ı3 | 12 | ı5 | 40 0,68 +0,01 Region of comparisona | 4 7 7 28 | 13 9) | 15.137 0,76 i by oo 5 6 22 17 7 14 38 0,58 13. Bornholm 13 9 3 25 14 14 7 35 0,71 +0,01 Region of comparison tae 6 9 31 rere | 2 laa 0,70 209 A. Larval hibernators B. N ee ne & eviation 2 a is als (L) i from region 3 of CSC NS salle 14. Ven 4 0 9 4 7 5| 16 0,56 —0,17 Region of comparison a i ZA NETS 32,.10.172.1..10) 1.17. 44 0,73 id 15 7 7.110820) | #165 11,0. s15, | 40 0,73 15. Göteborgs Skärgärd 4 7 ra) II 5 13 I 19 0,58 —0,10 Region of comparison 14 5 7 26 | 13 9 16 38 0,68 16. Orust 7 6 I 14 6 9 I 16 0,88 +0,20 Region of comparison | ,4 4 8 26 || 13 9 16 38 0,68 17. Hvaler 8 8 0 16 7 6 4 17 0,94 +0,15 Region of comparison | 15 4 7 26 | 14 8 | ıı 35 0,79 18. Hitra & c. 12 4 0 16 4 5 0 9 1,78 +0,24 Region of comparison | 13 4 3 20 4 8 I 13 1,54 19. Dönna & c. 7] 3 o 10 2 7 0 9 1,11 —0,25 Region of comparison | 10 3 2 15 3 7 I II 1,36 20. Lofoten, Vesterälen 8 4 I 13 2 7 I 10 1,30 —0,33 Region of comparison 8 3 2 13 3 4 I 8 1,63 Note: Bembidion dauricum and Pristonychus terricola omitted. Hibernation type of former is not known. Pristonychus is certainly introduced all over. 10. 11: 13, . Hailuoto 2=11,8 | o 8=47,1 | 4=23,5 1=59 Region of comparison ZZ 38,0 2213,20 005023915 =21,0 2=5,3 _ Aland 6=125 | 6=12,5 | 19=39,6 | 8=16,7 1=2,1 Region of comparisona | 7=!2:5 | 7=12,5 |22=39,3 | 9=16,1 3=5,4 2 b | 6=12.2 5=10,2 | 21=42,9 9 =18,4 3=6,1 . Alands Skargard 5=122 | 7=17,1 |15=36,5 | 7=171 1=2,4 Region of comparison 6=12.3 | 5=10,2 |21=42,8 | 9=18,4 3=6,1 . Hogland ı= 4,6] 2= 9,1 | 9=409 | 5=22,7 1=46 Region of comparisona | 3= 6,5 | 6=13,0 | 20=43,5 | 9= 19,6 2=4,3 22 b| 8=12,3 |11=16,9 |22=33,9 | 10=15,4 3=4,6 . Other outer islands 2=15.4 | I= 7,7 | 4=30,75| 4=30,75| 2=15,4 | 0 Region of comparisona | 6=100 | 8=13,3 |24=40,0 | 10=91,7 | 12=20,0]} 5=8,3 f b | 8=12,3 |11=16,9 |22=33.9 | 10=15,4 | 14=21,5 3=4,6 . Valamo (0) .. PR 3 . Värmdön 4= 10,8 5=13,5 17 = 46,0 4= 10,8 7= 18.9 1=2,7 7 210 232 Table 14. Ecological grouping of flightless and dimorphic species (b and d, Table 12) of islands and the respective mainland (= regions of comparison). See explanation with Table 9. Osel and Dagé omitted. Last two columns give percentage deviation from fauna of mainland district compared. Percentage devia- (w) tion from region of comparison Ww Species % | Species % | Species % || Species % | Species % 3=158 |11=57,9 | 1= 53 | 4=210 | 1=5,3 Region of comparison = 61 | 5= 10,2 |19=38,8 | 9= 18,4 | 13=26,5 || 2=4,1 Region of comparison =12,3 | 8=14,0 |22=38,6 | 8=14,0 | 12=21,1 || 3=5,3 Gotska Sandön 3=21,4 | 4=28,6 | 7=50,0 | 0 o — an Region of comparisona | 6= 10,5 | 8=14,0 |23=40,4 | 9=15,8 | 11=19,3 3=5,3 “ b 10=16,7 | 9=15,0 |21=35,0 | 8=13,3 |12=20,0 || 3=5,0 Färön 6=17,1 | 7=20,0 |15=429 | 5=143 | 2= 57 | 1=29 Region of comparisona | 6105 | 8=14,0 |23=40,4 | 9=15,8 |11=193 | 3=5,3 » p [10=16,7 | 9=15,0 |21=35,0 | 8=13.3 |12=20,0 || 3=5,0 Gotland 9=15,3 | 8=13,5 |24=40,6 | 9=15.3 | 9=15,3 1=1,7 Region of comparisona | 11 =17,2 | 10=15,6 |23=35,9 | 9=14,1 | 11=17,2 || 2=3,1 A b | 10=16,7 | 9=15,0 |21=35,0 | 8=13,3 |12=20,0 || 3= 5,0 - Oland 13=19.4 |10=149 | 25=37,3 | 9=13,5 |10=149 | 2=3,0 | 9=13,5 |t 23)— 12 Region of comparisona | 11=16,9 | 10=15,4 | 24=36,9 | 9=13,9 | 11=16,9 || 2=3,1 | 9=13,9 2 bl 10—16.,7 | 9150 2135.01 3-33 112- 20,0) 3 ol 81333 Bornholm 13=21.3 | 9=148 |23=37,7 | 8=131 | 8=13,1 | 3=49 | 9=148 |T 2814 24 Region of comparison | 13= 17.3 |12= 16.0 27 = 36,0 |10= 134 | 13=17,3 || 4=5,3 | 9= 12,0 14. Ven Region of comparison a 9 15. Göteborgs Skärgärd Region of comparison 16. Orust Region of comparison 17. Hvaler Region of comparison 18. Hitra & c. Region of comparison 19. Dönna & c. Region of comparison 20. Lofoten Region of comparison 211 Percentage devia- “ila lon cane ek ano a vcs ERE Species % | Species % | Species % | Species % | Species % || Species % | Species % x+(x) w+(w) 3=12.0 | 4=16,0 | 11=44,0 | 5=20,0 | 2= 80 o 3=12,0 |— 42|— 6,5 13—17,1°112—15,8°|27= 35,5 | 11 = 14,5 | 131751 5=6,6 9=11,8 REIS — 15479 20— 37,2 TU — 15.79 (Tl SET A] Q= 12,9 3=10,0 | 6=20,0 | 10=33,3 | 6=20,0 | 5=16,7 © (0) — 1,2|—17,2 10=15,6 |IO=15,6 | 24=37,6 | 1o=15,6 | 10=15,6 ZT 9=IA4,I 0 5=16,7 |14=46,6 | 5=16,7 | 6=20,0 1-33 5=36,7 13,0|+ 1,2 9=14,1 |10=15,6 |24=37,5 | 10=15,6 | 11=17,2 3= 407, 9=14,1 6=18,2 | 4=12,1 |14=425 | 4=12,1 | 5=15,1 I =3,0 4=12,1 |+ 4,0|— 7,7 G—TOmm O05 Sul 22-38 Gul 117—19,3, | 9—15.3 1 417,00 79-1558 I= 4,0 "ı= 40 |15=60,0 4=16,0 | 4=16,0 2=8,0 5=20,0 |—-10,2|+ 0,7 DO KATZ ITS — 525 1 55,2 5) A a oan 210 7212 0 = 5,3 |12=63,1 | 315,8 | 3=158 | © 5=26,3 | 6,3 |— 8,3 fo) 3=11,6 |16=61,5 | 4=15,4 | 3=11,5 DEIN 7=26,9 2= 83 | 3=12,5 |12=50,0 | 3=12,5 | 4=16,7 ı=42 | 5=20,8 |+ 6,5 |— 84 fe) 3=14,3 | F1=52,4 | 3=14,3 | 4=19,0 1=4,8 6 = 28,6 Note: Pristonychus terricola is not taken into consideration. 53 species, that of the mainland region of comparison, 119 species. In view of the far greater surface of the latter, the inevitable impoverishment of the fauna is rather less than expected. This is largely attributable to the striking variety of biotopes in Hailuoto. The missing species, according to present knowledge, belong to no particular dynamic group (Tables 10, 12) or group based on a developmental stage (Table 13). With regard to ecology (Table 14) the most striking feature is under-representation of the xerophilous species, but especially of forest species. Nevertheless, it may be concluded that Hailu- oto represents a remarkable example of how an island can be colonized in a relatively short period (< 2000 years) by a fauna completely “normal” in com- position. This cannot be expiained solely by the overall development of biotopes on the island: the possibilities of immigration must have been exceptionally favorable. It is not easy to decide whether the main reason was the short dis- iance from the mainland or the formerly busy shipping. The assumption of an anthropochorous dispersal is contradicted by the fact that only two species 237 212 found in Hailuoto (Dromius longiceps and Dyschirius lüdersi) seem to be absent from the mainland (= region of comparison), in spite of the active shipping contact which apparently occurred earlier between distant localities in the southern part. It is difficult to attribute the formation of a fauna, which is “normal” in all respects, to anthropochorous dispersal as the most important factor. The rapid and uniform colonization of Hailuoto by the fauna is in my view due to its short distance from the mainland, and to the fact that a big river, the Ule, empties in its immediate vicinity. During spate in spring this river brings enormous quantities of various alluvial material into the sea. Especially at this time of year, when drift ice is also brought down by the river, all kinds of objects lying on the ice or frozen into it, such as tree trunks, twigs, reeds, soil, etc. can be transported far out to sea. Successful landing of such transported objects at Hailuoto, lying only 27 km from the mouth of the river, might take place every year. —Comparison with the islands of the Province Oa (Replot, Bjorko, etc.) situated in Kvarken, which are about the same age (Valovirta, 1937, p. 35 ff.) and are similarly isolated but lack such favorable location close to a river mouth, would undoubtedly be highly informative. Unfortunately their coleopteran fauna is imperfectly known. 2. Aland We will consider here only the main island (Swedish “Fasta Aland”) and Eckero which is separated from it by the narrow Marsund. The Skargard far- ther east has faunistic peculiarities and is best treated separately. Fasta Aland is an extremely fragmented island with an area of 937 km? (including Eckero), which has developed gradually by the joining of small islands rising out of the sea (see map in Hausen, 1910a) and is continuously growing in size. The highest point, Orrdalsklint in Saltvik, is at an altitude of 132 m above sea level. The present rate of land emergence varies around 5 mm per year (Witting, 1943, p. 28; Hausen, 1946, p. 82). The oldest known culture (“Ganggrift” period) corresponds to a location on the shoreline at 34 to 36 m above sea level, with a maximum age of 5000 years (Hausen, 1910a, p. 56; Munthe, 1940, pp. 189, 201, Plate XV). The highest points in Aland may have risen from the sea as naked rocky skerries as late as the Ancylus period, perhaps around 7000 BC (Munthe, l.c., Plate XI; Sauramo, 1942, p. 240). The nearest distance to the mainland, Vaddo in Upl, is 38 km (from Eckero). The nearest point on the Finnish mainland (Ab Taivassalo) is 76 km away, and Estonia is a good 200 km away. The population is about 20,000 persons, a remarkable density, which gives the island the striking appearance of a cultivated landscape. There are active trade connections with both Sweden and the Finnish mainland. In earlier times these were predominantly through Stockholm, but since the penne separation from Sweden (in 1809) mainly through Abo. 238 218 Äland has an exceedingly high variety of biotopes (cf. the summary in Swedish by A. Palmgren, 1943-44, p. 20 ff.). Particularly striking are the luxu- riant meadows of leaves (= “Laubwiesen’’) and groves (A. Palmgren, 1915 to 1917); otherwise coniferous forest is dominant. The numerous lakes belong to diverse types (Cedercreutz, 1937), likewise the seashores. In Eckero even quicksand is found (Krogerus, 1929). The mountainous terrain consists almost without exception of granite (Hausen, 1946, p. 31). The large number of species of the flora (vascular plants) in Aland is very striking (650 “large species”; A. Palmgren, 1925, p. 44), which even surpasses that of the corresponding region on the South Finnish mainland (l.c., p. 54). —Unfortunately there is no comprehensive account of the beetle fauna of Aland but it is possible to extract an almost complete list of species from the Catalogus (1939). The two contributions by Hak. Lindberg (1924, 1925) are only reports of collecting. Most records are unpublished and are only represented by specimens in museums and private collections. The carabid fauna, according to our present knowledge, comprising 157 species is strikingly rich, though not to the same extent as the flora. In the num- ber of species it is of course surpassed not only by the region of comparison in Uppland (with 189 species) but also by that of the Finnish mainland (with 176 species). The composition of the fauna on the basis of dynamic groups is fairly normal; the preponderance of dimorphic species (Table 10, “d + [d]”) and functionally brachypterous species (Table 12, “b + d’) is small, i.e. only two and one species respectively. More striking is the deficit of species that hibernate in the larval stage (Table 13, last column). Ecologically (Table 14) there is a small preponderance of xerophilous species (x + [x]) and a larger deficit of forest species (w + [w]). The origin as well as the time, direction and mode of immigration of the flora and fauna of Aland present a series of captivating problems on which much has already been written. The recent origin of the island and its location almost midway between two areas of mainland which during the postglacial period obtained their fauna largely independent of one another from different directions, render the questioning extremely concrete. A. Palmgren, in a series of contributions, dealt thoroughly with the im- migration of the flora to Aland, and in 1927 summarized his results. After a detailed comparison with the flora of neighboring regions (Upl; Ab, NI; Es- tonia) he concluded that “a predominant part of the flora of Aland has been received from Sweden. The immigration from Finland has seemingly been very slight. Immigration from the eastern Baltic was stronger, though scarcely sig- nificant” (1927, p. 79). He bases this idea mainly on the gradual reduction in the number of species within the landscape limits of Aland from west to east. The reason is “the increasing distance to the east from a center of dispersal in 239 214 the west (in Sweden)!...” (1921, p. 71). On this Palmgren bases his postulate on “the distance as a phytogeographical factor.” A detailed and effective criticism of this point of view was provided by Ek- lund (1931, pp. 71-92). In the Skärgärd of Äboland (Korpo and Houtskär) he found a similar reduction, especially in the flora of leaf meadows (=“Laub- wiesen”), which there diminishes from north to south and is cleariy due to edaphic factors. He concludes that the reduction in the number of species in the landscape of Aland toward the east is due to the decrease in limestone in the same direction. It is moreover really astonishing that Palmgren bases a “theory of distance” exclusively on statistical grounds without considering the dispersal ecology of the plants. It would have been natural to attempt a division of the species into anemochorous, hydrochorous, etc., whence his “theory” might have received causal confirmation. But even in his comprehensive contribution “The routes of immigration of the flora to the islands of Alandt” (1927), in which each of over 50 species is treated separately in detail, no data at all are provided on the nature and mode of dispersal of the diaspores. On the other hand Eklund (1931, pp. 88 ff.) devoted special attention to the dynamics of plants and in particular dealt with the possibilities of hydro- chorous passive dispersal (also experimentally; 1927a, 1927b, 1929). He sum- marized his conclusions on the routes of immigration (1931, pp. 105-106) as follows: 1. “A large main immigration of the flora from Sweden to Aland and in general to the Finnish rocky skerries may not have taken place.” 2. “...an immigration from the south (i.e. from the eastern Baltics) seems far more Significant than earlier assumed.” Palmén (1944, pp. 224 ff.) accepted this in- terpretation with regard to insects, and went somewhat too far in ascribing to Eklund a greater role of the ‘Baltic’ immigration route than he actually proposed (see quote). Let us examine to what extent the carabid fauna of Aland can contribute to a solution of these questions. It includes a number of species, although small, for which the immigration route seems more or less clear, solely with regard to their distribution on the mainland (see maps in Part II). Swedish origin may therefore be assumed for the following 12 species: Agonum lugens Harpalus winkleri Badister bipustulatus s. str. (see Supplement) Licinus depressus Bembidion illigeri Metabletus foveatus Calathus fuscipes Nebria brevicollis Dormius linearis Odacantha melanura D. nigriventris Olisthopus rotundatus. t(Original German quotation translated into English; suppl. scient. edit.). 240 215 With the exception of Licinus, Olisthopus and possibly the form of Badis- ter, aS far as is known they are absent from the northern coast of Estonia and from Dago (altogether 3 or 4 of them are found in Estonia; Harpalus winkleri uncertain). On the Finnish mainland Agonum and Nebria are completely ab- sent and possibly the form of Badister; Bembidion and Dromius linearis appear to occur there as more or less accidental immigrants. The rest were found only as rarities in the southwestern corner (certainly immigrated from Aland) (Metabletus and Calathus also in the extreme southeast) with the exception of Harpalus, which is unknown in Ab. The immigration route of Olisthopus is Clarified (p. 373) by the map showing wing-dimorphic forms (Fig. 32; p. 373). An eastern origin, from the Finnish mainland (F) or from the Baltics (B), can be assumed for the following 11 Aland species: Acupalpus exiguus F B. varium B? A. flavicollis B Cymindis macularis F Amara majuscula F? Dyschirius impunctipennis B? Bembidion andreae polonicum B D. salinus B B. humerale F Microlestes minutulus F? B. minimum B? A distinction between a southeastern (“Baltic”) and a purely eastern im- migration from the Finnish mainland is especially difficult in cases where southwestern Finland also apparently received its stock of the species in ques- tion from the Baltics (but possibly at earlier times), as in the case of Bembidion minimum and B. varium. However, in the present context it is only of minor importance since the intention was to present a group for comparison with the above “Swedish” group. —Amara majuscula might also occur in Esto- nia (although confused with apricaria). Dyschirius impunctipennis may have immigrated from Gotska Sandon or Faron, as assumed by Krogerus for the ecologically corresponding species Bledius tibialis Heer. All 11 species are ei- ther missing from the parts of the Swedish mainland (Sdm, Upl) lying opposite (e.g. Bembidion andreae, Cymindis, Dyschirius impunctipennis, D. salinus, Mi- crolestes) or are found there very rarely, more or less sporadically (e.g. Acupal- pus flavicollis, Bembidion humerale, B. minimum, B. varium), or have evidently immigrated very late (e.g. Acupalpus exiguus, Amara majuscula; cf. pp. 622 ff.). Do the members of these two groups of immigrants in the fauna of Aland show any basic differences? A division based on ecology (Tables 9, 14) reveals the following: “Swedish group” “Eastern group” xerophilous [x, (x)] 5 species 2 species mesophilous (n) 4 species 1 species hygrophilous [h, (h)] 3 species 8 species 216 In the “eastern” group, therefore, the hygrophilous species predominate. Since the same group includes 5 species (the species of Bembidion and Dyschi- rius, with the exception of B. humerale) which always lead a littoral existence (along the sea) it is natural to consider hydrochorous transport from the Baltics (cf. Eklund, 1931, pp. 93 ff.). However, the 5 species mentioned are all capable of steady flight, and so could use anemohydrochorous transport (in Palemen’s sense, 1944). The idea is strongly supported by the fact that 7 of the 11 species of the “eastern” group (all, with the exception of Bembidion andreae, cymindis, and species of Dyschirius) were found in his drift material (1944, pp. 37-39). This brings us to a more important difference between the two groups treated here, which is evident from the following division according to the dynamics of the species (Tables 9, 10): “Swedish group” “Eastern group” macropterous [m, (m)] 6 species 10 species dimorphic [d, (d)] 3 species 1 species brachypterous (b) 3 species 0 species Table 11 shows that 3 ofthe 4 dimorphic Äland species are found exclusively in the brachypterous form and the fourth (Olistkopus) predominantly so. A comparison with the entire fauna of Äland is best carried out according to the division in Table 12: m + (m) d b od percentage deviation The entire fauna of 109 species 8 species 40 species Aland = 69% = 5% =26% “Swedish group” 6 species 1 species 5 species = 50% = 8% = 42% +19 “Eastern group” 10 species 0 1 species = 91% = 9% —21 The “Swedish group” in the fauna of Aland thus includes a large part for which anemochorous dispersal can scarcely be considered. For a closer acquaintance with this western group, and following Palm- gren’s “distance principle,” I also analyzed the dynamics of the species that occur on Aland exclusively in Eckerö (the western part). These are given below: *Acupalpus exiguus A. dolens Agonum assimile A. quadripunctatum 242 244 217 Amara ingenua Carabus granulatus *A. majuscula C. nemoralis A. municipalis Chlaenius nigricornis A. ovata *Cymindis macularis Badister dilatatus Dromius marginellus *Bembidion andreae polonicum Dyschirius aeneus B. gilvipes *D. impunctipennis *B. humerale Notiophilus pusillus Bradycellus collaris Pterostichus gracilis It is amazing to find that six of these species (asterisk) belong to the “east- ern” group treated above. This agrees with Eklund’s idea (1931, pp. 88, 96), according to which at least the hydrochorously dispersed plant disseminules have “far greater prospects of even reaching western Aland from the eastern Baltics than from Sweden.” Although there is hardly any purely hydrochorous dispersal of the beetles in question, their distribution shows that Palmgren’s “distance theory” is not applicable even to organisms independent of lime- stone, i.e.: The distribution of carabids within the confines of Aland does not Provide any adequate clues to their immigration route. Therefore I proceeded to examine and consider the entire coleopteran fauna, the species found in Finland (borders prior to the Second World War) exclusively in the landscape Aland (Al, also including the Skärgärd of Aland). The material was obtained from the Catalogus (1939), from further references in “Notulae Ent.” and “Ann. Ent. Fenn.”, and from Mr. Hellén and Mr. An- ton Jansson (Table 15). The records from the eastern Baltics were taken from the relevant literature. Of the 42 species in Table 15, which could not have immigrated to Aland from the mainland of Finland, 30 are found in east-central Sweden and 28 in Upl. As far as is known, 22 of the species in question are native to the eastern Baltic. Thus immigration from parts of Sweden next to Aland is inconceivable or at any rate most improbable for at least 12 species, 3 of which (Bagous, Diphyl- lus, Limnobaris) are nevertheless found northward up to Ogl. —Bledius tibialis, unknown in the eastern Baltics, may have come from Gotska Sandon, as pro- posed by Krogerus (1929, p. 72), and possibly Drilus from Gotland (p. 248). As far as is known, therefore, no fewer than 20 of the 42 species are absent from the eastern Baltics and another 6 from Estonia; in Osel and Dago altogether only 11 species are found. Nevertheless, a small component of the eastern Baltic element in the group is quite natural, considering the poor exploration of these regions. Had their members arrived in Aland from the Baltics, it would have been very strange if they had not simultaneously colonized the mainland of southwestern Finland. With the exception of the 12 species mentioned they may be considered as some of the most definite 246 248 218 representatives of the Swedish element in the coleopteran fauna of Äland. We ncw proceed to an opposite group: those species of Aland province (also including the Skargard of Aland), that are not found in east-central Sweden (Table 16). The list of 83 species is quite impressive, especially when compared with the 42 species in Table 15. The two lists are not directly comparable: in the first case we are dealing with the species of Aland entirely absent from Fin- land, in the second only those absent from east-central Sweden. But the great role played by eastern or southeastern immigration of the coleopteran fauna of Aland is clearly evident. Eleven species (“p” in list) are represented in Palmen’s drift material. Eleven species (Airaphilus, Apion pisi, Bledius tibialis, Brachygluta, Cassida, Ceuthorrhynchus schönherri, Diphyllus, Drilus, Limnobaris, Stenus lindbergi, Tro- gophloeus nitidus) are common to both lists; they may have arrived from the south (or the southeast). Moreover, it is striking that out of the second list (Table 16) only 3 species (Chromoderus, Heterocerus obsoletus, Triarthron) are missing from southwestern Finland (Ab, Nl, St, Ta), and also (with the ex- ception of Chromoderus) as far as is known, from the Baltics, where many of the remaining species are also unknown. The most important immigration route of the Coleoptera to Aland might have been from the east, from the Finnish mainland. To return to the point where we felt obliged to extend the discussion to the entire coleopteran fauna, we have the question of the modes of dispersal of the beetles that immigrated to Aland. It is seen from an examination of the “non-carabids” in the two lists above that they are fully winged almost without exception*, i.e. they have in all probability immigrated in the anemochorous (or anemohydrochorous) mode. Exceptions are: in the “Swedish” group (Table 15, excluding the common species of both lists) Chrysomela haemoptera, Galeruca laticollis, and Lycoperdian succincta**; in the “eastern” group (Table 16 excluding the common species) Eonius and Scleropterus. From among the “common” species only Drilus (female) must have immigrated from the south (or from the southeast). The last species (as larva and female) lives as a parasite in the shells of Helix, and therefore must have extraordinary opportunities to be transported hydrochorously with the host over long distances (possibly as impregnated *] have not examined all the species in Tables 15 and 16 regarding the formation of the hind wings but only those which on the basis of their mode of life (soil insects!) could be presumed flightless. The following “suspected” species turned out to be macropterous: Anthicus umbrinus, Apion pisi, Brachygluta helferi, Chromoderus fasciatus, Oxypoda brachyptera, Reichenbachia impressa, Stenus scabriculus, Tachyporus tersus. ** West (1940-41, p. 391) erroneously considered Lycoperdina succincta as capable of flight. Two specimens from Denmark (Asserbo Overdrev) and from Ögl (Mogata) show completely rudimentary wings. 242-243 Table 15. Coleoptera (except Carabidae) of Al province (Äland) not found elsewhere in Finland b—Brachypterous; d—Dimorphic Upl Sdm| cu Ösel | Est- Ost- | Vst Nke Dagö| land | balıikum Agabus chalconotus PANZ. + Agathidium varians BECK 4 4- — a = En Airaphilus perangustus HAR. LINDB| — _ — _ -— Aphodius contaminatus HBST. — + Apion pist FBR. — Bagous brevitarsis V. HANSEN — Bledius tibialis HEER Brachygluta helferi SCHM.-GOEB. Cassida stiymatica SUFFR. Ceuthorrhynchus pollinarius FORST. @ schönherri BRIS. b Chrysomela haemoptera L' Coccidula scutellata HBST Coeliodes ruber MRSH. Cryptophagus affinis STURM @ populi PAYK. Cteniopus flavus SCOP. Dasytes coeruleus DE G. Dermestes laniartus ILL. Diphyllus lunatus F BR Q6 Drilus concolor AHR. b Galeruca laticollis SAHLB.” Graphoderes cinereus L. Graptodytes bilineatus STURM Gymnetron pascuorum GYLL. Hedobia imperialis L. Hypocyptus suecicus PALM Laccobius striatulus FBR. Limnobaris Teitteri MUNST. Liodes curta FAIRM. ] issodema cursor GYLL. b Lycoperdina succincta L. Micropeplus fulvus ER. Octotemnus mandibularis GYLL. Oedemera croceicollis GYLL. Olibrus bicolor FBR. Reichenbachia impressa PANZ. Silis ruficollis FBR. (@) Sitona humeralis STEPH.® Stenus lindbergi RENK. Tetratoma jungorum FBR. Trogophloeus nitidus BAUDI | | | b+) ++ + | + | + | | | +++ | | | [+14 [++++++44+ 14 }+++++4+4+4+141 IH I +++++4 | |+1 +41 | +1 I+I++1 I++++++ +1 }++)4+++4+4+4+14+4+4441 p++ i t+4+4+14+i ++ t+] tL +1 +44 | | | + | | | | bl ++ tt tt HH HH HH HH HH HH | | + | | + | | + | IH I +++++4+++ + | II + I ++ 1 +414 | ‘Wings of Chrysomela haemoptera are comparatively less developed than those of C. hyperici (see p. 245): particularly the apex is shorter. According to Rüschkamp (1927, p. 29) they are occasionally rudimentary, and I believe he is right in considering the species flightless. 2Galeruca laticollis has reflexed wings with strong veins. But they are not broader than the elytra, and the apex is rudimentary (cf. G. oelandica, p. 304). 3Sitona humeralis is dimorphic, according to Jackson (1928, p. 172). One individual from Aland and 2 from Ska were found to be macropterous. Later Palmen (in lit.) found one individual (macropterous) in drift material near N! Tvarminne. 244 245 220 female). Thus direct dispersal in this way to Äland from Ösel-Dagö or Gotland cannot be ruled out. The nature of the currents in the Baltic Sea in autumn is highly favorable for such transport (Fig. 19b). Eonius may also be suitable for hydrochorous transport. It is a stenotopic quicksand species found inland (Krogerus, 1932, p. 248) only as an Ancylus relict, otherwise living exclusively at the seashore, and therefore regularly get- ting into the sea at high tide. The insect must have arrived on the outer islands in the Gulf of Finland (Hogland, Lavansaari, Seiskari), possibly also in Färön, by hydrochorous means. Table 16. Coleoptera (excluding Carabidae) of Al province (Aland), not found in east- central Sweden (Upl, Sdm, Vst, Nke) *—Absent throughout Sweden; P—Represented in drift material reported by Palmen (1944); b—Brachypterous; d—Dimorphic P Chrysomela hyperici Forst.! P Coclambus marklini Gyll Acmaeops septentrionis Th. Airaphilus perangustus Har. Lindb. * * Aleochara lygaea Kr. Anthicus umbrinus Laf. Aphodius piceus GylL Colon viennense Hbst. Corticaria lambiana Sharp Cryptocephalus aurcolus Suffr. A. plagiatus L. * C. coerulescens Sahlb. P Apion hookeri Kirby P Cryptophagus corticinus Th. A. pisi For. C. instabilis Bruce. Atheta aquatica Th. P A. castanoptera Mnh. Diphyllus lunatus For. Dorytomus salicis Walt. * A. dubiosa G. Bck b(2) Drilus concolor Ahr. A. dvinensis Popp. Dryops similaris Boll A. fungicola Th. b Eonius bimaculatus IL * A. livida Muls. rey. Ernobius pini Sturm. Atomaria bicolor Er. Gnathoncus punctator Rtt A. södermani Sjob. Haemonia pubipennis Reut * Attagenus schaefferi Hbst. Heterhelus scutellaris Heer. Bagous limosus GylL Heterocerus flexuosus Steph. Berosus spinosus Stev. H. intermedius Kies. Bidessus hamulatus GylL H. obsoletus Curt. Bledius diota Schiß. * Hololepta plana Sulz * B. rastellus Schif. P Hylastes attenuatus Er. B. tibialis Heer. * H. brunneus Er. B. tricornis Hbst. P Hylobius transversovittatus Gze. * * Brachygluta helferi Schm. Goeb. Bryoporus crassicornis Mäkl Calodera protensa Mnh. Cassida stigmatica Suffr. Ceuthorrhynchus schonherri Bris. Chromoderus fasciatus Müll Laccobius decorus GylL Limnobaris reitteri Munst Liodes badia Sturm. * L. flavescens Schm Mecinus collaris Germ. Meligethes morosus Er. 221 Monochamus galloprovincialis OL Rhantus notaticollis Aubé (d) Mycetoporus aequalis Th? Rhynchaenus fagi L. P Nanophyes circumscriptus Aubé. *b Scleropterus serratus Germ. Noterus clavicornis Deg. * Seymnus testaceus Motsch. P Ochthebius marinus Payk Sitona cylindricollis Fähr." Oxypoda brachyptera Steph. * Stenus lindbergi Renk Oxytclus sculpturatus Gr. S. scabriculus J. Sahlb. * Pachyta quadrimaculata L. Tachyporus tersus Er. Philonthus addendus Sharp. Triarthron maerkeli Schm. P Phytobius velaris GylL Trichomicra sahlbergiana Bernh. P Psylliodes isatidis Heik. Trogophloeus foveolatus Sahlb. T. nitidus Baudi. 1 In one specimen of Chrysomela hyperici examined from Skä the wings, well developed and with strong veins, are 1.5 times as long as the elytra. Ruschkamp (1927, p. 26), who had not examined C. hyperici, curiously believes that all species of Chrysomela being flightless genus is represented by 5 species, including C. hyperici, in drift material reported by Palmen (1944). 2 Mycetoporus aequalis is dimorphic (Hellén, 1925, pp. 33, 40; Strand, 1946, p. 177); one individual from Al turned out to be macropterous. 3Palmén studied 17 Finnish specimens of Scleropterus for me, 15 from Aland. In all, the hind wings were totally reduced. 4 According to Jackson (1928) Sitona cylindricollis is always macropterous. Eonius is, ecologically and possibly dynamically, a perfect counterpart of Cymindis macularis among the carabids; they are the only Coleoptera of Aland of which it must be assumed that they arrived only by the hydrochorous mode from the east (or southeast). It is significant that both are found in Aland only on Eckero, in the extreme west (for Eonius, see Hak. Lindberg, 1931, ps 153): To assess the capacity of Cymindis macularis to tolerate hydrochorous transport, at least to get an idea, I carried out simple experiments on exposure to water (Experiments 143-145, p. 109) with 6 individuals each of this species and simultaneously of C. angularis and C. humeralis. Two serial experiments were attempted each with 3 x 3 specimens. In one the insects were left swim- ming undisturbed on the surface of the water, in the other the container was regularly shaken. The room temperature was high (20-25°C). The following results were obtained with exposure to water. A. Undisturbed insects No. of surviv- ing specimens after 67h 71h 115h 123h 147h 157h 195h 243h 315h 326h Cymindis BENS 3 2 2 2 2 2 1919: angularis C. humeralis 3 3 3 3 3 2 1 1 a. Ei’macularis 13° 2 2 2 2 1 1 u a oe 249 250 222 B. Insects in shaken containers No. of surviving specimens after 13h 28h 43h 52h 60h Cymindis angularis 2 1 1 1 Dar C. humeralis 2 1 1 an C. macularis 3 2 — — — These experiments show (even taking into consideration the abnormally high temperature) that Cymindis macularis has little resistance to water, especially if splashed, and even less resistance than the 2 related species whose distribution provides no ground for presuming that they are dispersed by the hydrochorous mode. It is probable that C. macularis can tolerate long-distance hydrochorous transport (cf. p. 600) only with the help of floating objects like fascicles of Elymus and Psamma. The situation is quite different for Lycoperdina succincta, Chrysomela haemoptera and Galeruca laticollis. None of these is a seashore species; the first two are xerophilous animals living in open, gravelly or sandy terrain; Lycoperdina is associated with puffballs, Chrysomela with species of Plantago (West, 1940-41, p. 494); Galeruca lives on green meadow soil, among others on Thalictrum flavum (West, l.c., p. 506). All three are to be considered as species with extremely poor dynamics. They naturally come next to the 6 flightless carabids discussed above (p. 241). —Scleropterus, which has quite rudimentary wings, is enigmatic. In Fennoscandia it is found only in southwestern Finland (Ab, Nl) and in Aland. Its biology (host plants) and the modes of dispersal are unknown. The coleopteran fauna of Aland on the one hand consists of an eastern (Finnish-Baltic) element, which with few exceptions is capable of flight and has certainly immigrated by the anemochorous mode (including anemohydrochorous), and on the other hand of a western (Swedish) element, of which at least 9 species must have immigrated by other modes (hydrochorous, zoochorous, or anthro- pochorous). Evidently now an important task is to examine the validity of the rule that mainly aerial immigration has taken place from the east, so that it can be applied to other organisms. It is most natural to study the phanerogamous flora. From a survey of the contributions by A. Palmgren (particularly 1927 and 1943-44) and Eklund (1927a, b; 1931) it might be possible to draw up lists of the “western” (Swedish) and of the “eastern” (Finnish-Baltic) species of the flora of Aland. Even though the allocation of many of these species in the group concerned was based on more or less strongly motivated conjectures on the part of these authors, we may be justified in considering each group as a whole as fairly representative. The most important mode of dispersal of each species was then assessed, partly from the works of Eklund, Sernander Ü g 4 SS N T N | i \ \ Dic : —— re ar: Pats rn rrents in the Northern Baltic Sea Basin: a—in June; b—in October. Adapted from Atlas öfver Finland 1910 (Simplified). Fig. 19. Surface sea cu 247 224 (1901), Heinze (1932-35), Lagerberg (1937-39), and Romell (1938), and partly according to my own judgment. For nomenclature, Stockholmstraktens Vaxter (1937) is followed. 253 A survey of the plants listed above (Tables 17, 18) reveals the following: The 129 “western” species in Aland include 27 anemochorous species, i.e. 2176 The 63 “eastern” species in Äland include 9 anemochorous species, i.e. 14%. The importance of anemochorous immigration to Äland is also evident for other plant species. Gustaf Haglund told me that all the Taraxacum species occurring in Aland are found in central Sweden. But T. rubicundum Dahist., for instance also found there, is in no way synanthropic, and is unknown elsewhere in Finland and the entire eastern Baltics. —According to a communication by Herman Persson, among the mosses occurring in Äland, Ulota drummondi and Douinia ovata are unknown elsewhere in Finland and in the eastern Baltics but are found along the east coast of Sweden. 250-251 Table 17. Phanerogamous plants (“large species”) of Äland province that can be con- sidered (at least partly) as immigrants from Sweden ; *—Dispersal (mainly) anemochorous Adoxa moschatellina C. extensa Agrimonia odorata C. hirta Aira praecox C. lepidocarpa Androsace septentrionalis C. livida Anthyllis vulneraria C. loliacea Aquilegia vulgaris C. montana Arctium nemorosum C. nemorosa Asperula odorata C. ornithopoda A. tinctoria C. paradoxa Athamanta libanotis C. polygama Barbaraea stricta C. remota Brachypodium silvaticum C. riparia Bromus benekenii C. vaginata * Calamagrostis arundinacea C. vulpina Calamintha acinos Ceratophyllum demersum Callitriche autumnalis * Chimaphila umbeliata * Campanula latifolia Cochlearia danica Cardamine flexuosa * Coeloglossum viride Carex aquatilis Corydalis intermedia C. arenaria C. laxa * C. caespitosa Crepis paludosa C. caryophyllea * C. praemorsa C. digitata Cypripedium calceolus * 225 * Epipactis latifolia * FE. palustris 251 * Eriophorum gracile * Orchis incarnata (coll.) * O. sambucina * Orchis traunsteineri E. latifolium Erysimum hieraciifolium Fragaria viridis Fritillaria meleagris Gagea lutea G. minima Galium trifidum Gentiana campestris G. uliginosa Geranium columbinum G. dissectum G. lucidum Glechoma hederacea Herminium monorchis Hippophaé rhamnoides Hydrocharis morsus-ranae Hypochaeris maculata Jasione montana Lactuca muralis Phloeum boehmeri Polygonum viviparum Potamogeton crispus P. poiygonifolius P. praelongus Potentilla tabernaemontani Ranunculus circinatus R. lingua Rhynchospora alba Rosa tomentosa Rubus pruinosus Rumex hydrolapathum Sagina maritima Salix livida Salsola kali Samolus valerandi Saxifraga granulata Scheuchzeria palustris Scirpus maritimus Lathraea squamaria S. rufus Lemna trisulca Sedum annuum Litorella uniflora S. rupestre Malaxis paludosa Medicago lupulina Melica uniflora Mentha litoralis Mercurialis perennis Microstylis monophylla Monotropa hypopitys Myricphyllum verticillatum Myosotis palustris Najas marina Neottia nidus-avis Nuphar luteum N. pumilum Odontites litoralis Oenanthe fistulosa Ophrys muscifera S. sexangulare Selinum carvifolia Solidago virgaurea Sorbus fennica S. suecica Sparganium affine S. ramosum microcarpum Spergula vernalis Stellaria nemorum Taxus baccata Torilis anthriscus Ulmus glabra Veronica anagallis-aquatica V. beccabungae Vicia lathyroides Viola rupestris V. stagnina 254 It may be concluded that the indication obtained from Coleoptera, that anemochorous dispersal has taken place to Aland mainly from east to west, is not confirmed by a study of plants. 226 252 Table 18. Phanerogamous plants (“large species”) of Aland province that can be con- sidered (at least partly) as immigrants from Finland or Baltics *—Dispersal (mainly) anemochorous * Acer platanoides Agrimonia eupatoria Ajuga pyramidalis Alliaria officinalis Allium ursinum Alopecurus ventricosus Arbis hirsuta Arrhenatherum elatius Artemisia campestris Atriplex hastata Avena pratensis Brachypodium pinnatum Cakile maritima Calystegia sepium Cardamine hirsuta Carex brunnescens C. distans C. glareosa * Cephalanthera longifolia * Cirsium heterophyllum Clinopodium vulgare Crambe maritima Crataegus curvisepala C. monogyna Daphne mezereum Draba incana Geranium pratense G. sanguineum * Gymnadenia conopea Honckenya peploides Hypericum hirsutum Isatis tinctoria Juncus balticus Knautia arvensis Lepidium latifolium Melampyrum cristatum Melandrium viscosum Myosotis collina Orchis mascula Origanum vulgare Polygala amarella Polygonatum multiflorum Polygonum dumetorum Picea excelsa Plantago lanceclata Ranunculus cassubicus R. ficaria Rhamnus catharctica Salicornia europaea Salix rosmarinifolia Saxifraga tridactylites Scleranthus annuus Scutellaria hastifolia Silene venosa Stachys silvatica Stellaria holostea Suaeda maritima Succisa pratensis Thymus serpyllum Tilia cordata Trifolium montanum Veronica spicata Viola uliginosa Actually the documented evidence of a western anemochorous element in the case of plants, apart from the short distance across the sea, also seems to be supported by the wind conditions prevailing around Aland (Fig. 20). Easterly or southeasterly winds do not prevail at any time of the year. The dominance of southwesterly winds in autumn, the most important period for dispersal of most of the species, is striking. Even in the higher atmosphere in these areas, especially during the summer half-year, westerly winds prevail (Ostmann, 1933, p. 30). 227] OR RB BB RA Fig. 20. Wind rosest in Aland (1886-1905). I—Salskar (N of Aland); II—Bogskär (S of Aland). a—Spring; b—summer; c—Autumn; d—Winter. From atlas ofver Finland, 1910. On the other hand I do not believe that the major component of an eastern anemochorous element in the beetle fauna can be explained simply by the fact that the animals fly mainly in spring—which actually happens in the case of carabids (see p. 580)—for even at that time of year southerly and southwesterly winds are commoner than easterly or southeasterly ones. —The habit of carabids of flying (p. 584) against the wind may be ignored as well, since they are such weak fliers that they are carried away even by very light winds (> 2m/sec). However, the physiological difference between animals and plants is very important. The diaspores of plants can drift for weeks and months in the Baltic Sea water without losing their capacity to germinate (Eklund, 1927a, b; 1929). This amply explains the substantial admixture of those plants in the flora of Aland, which have dispersed hydrochorously from the eastern Baltics (Eklund, 1931). On the other hand, in the drift material of insects along the southern coast of Finland no species was found that must have crossed the Gulf of Finland by purely hydrochorous means (Palmen, 1944). Such cases may occur (p. 603), but only as rare exceptions. T(“Windrosen” = wind directions; suppl. scient. edit.) 255 256 228 Two problems have here to be dealt with more specifically: 1. In the colonization of Äland, why has anemochorous immigration com- ing from the east played a greater role in the case of beetles than in the case of plants? 2. How have the flightless “Swedish” beetles reached Aland? 1. The first question may concern not only Aland but the Baltic Sea region in general. Later we will deal with the considerable eastern element in the fauna of Gotland and neighboring islands, which has arrived by the anemo- chorous mode. Then it is also appropriate to point out the oddity that 4—5 carabid species have reached Sweden from Finland by crossing Kvarken (the narrowest part of the Gulf of Bothnia) through the air (p. 593), whereas there is no evidence to show that immigration of any species has taken place from the opposite direction. Thinking about this, it occurred to me that: many beetles capable of flight, including some carabids, fly in the evening or afternoon when the air has warmed up. Could it be that these animals, especially those flying in the afternoon, fly toward the sun, or at least show a preference for this direction? Then the dispersal of the species would be mainly toward the west. It appeared possible to study this problem experimentally. For this pur- pose I devised a “flight direction apparatus” (Fig. 8, p. 109), in which certain species that are comparatively active fliers were studied (Experiment 147 ff., p. 110). The experiments were all carried out between 5 p.m. and 7 p.m. and one of the eight sectors of the apparatus was directed exactly toward the sun (west). Only in the case of 3 species (Oodes gracilis, Acupalpus consputus, A. dorsalis) was the observation material large enough (> 90 cases) for the “flight roses” to be drawn (Fig. 21). The result of these experiments is so unambiguous that there is no room for experimental error or coincidence. A strong inclination to fly toward the sun is shown in particular by Oodes. That this is not so consistently evident in both the species of Acupalpus may be due to the fact that they take off in the air more or less spirally, in contrast with Oodes, which usually takes off in the final direction of flight. I believe that by using a larger apparatus the species of Acupalpus would also show “better” results. It is advisable to express the results numerically in such a way that the animals that entered the three “western sectors” are compared with those of the three “eastern sectors.” In this way species of Badister are also taken into consideration, although the material available was insufficient (Table 19). We find that in the species observed, with the exception of Badister peltatus (for which too little material was available), the inclination to fly toward the sun is very pronounced. Being more or less typical evening fliers, they must therefore disperse more toward the west than toward the east, provided there are no strong winds in the opposite direction. This seems to be an important factor which has influenced the considerable immigration of flying Coleoptera 229 ß CG 255 Fig. 21. “Flight roses”. Percentage distribution of individuals observed in the flight direction apparatus (Fig. 8, p. 109), between 5 p.m. and 7 p.m. in bright sunshine. a—Oodes gracilis (100 specimens); b—Acupalpus consputus (200 specimens); c—A. dorsalis (91 specimens). Experiment 147ff., p. 110. to Aland from the east and has also influenced the fauna of the other Baltic 257 islands (see below). The general consequences of this observed phenomenon will be reported in a later section (p. 592 ff.). 2. It remains to be considered why the transport of flightless species to Aland has been more from the west than from the east (p. 249). Two possible explanations may be given. 258 230 a. The species in question could have arrived anthropochorously, predo- minantly from Sweden, since in earlier times Aland had active trade connec- tions with that country. b. The dispersal could have taken place mainly hydrochorously, and the shortest water crossing west of Aland would have facilitated transport from Sweden. Theoretically speaking, zoochorous dispersal (Supposedly only by birds) is also conceivable. But as discussed elsewhere (p. 605), hardly any signifi- cance should be attached to this mode of dispersal, particularly with respect to carabids, so that it can be ignored here. a. The anthropochorous dispersal of carabids is also discussed elsewhere (p. 606) in detail, where I suggest that it has generally been overestimated. In the case of Aland in particular, the following facts may be stated: The only two functionally brachypterous species occurring in Aland that can be cer- tainly designated as synanthropous, namely, Bembidion ustulatum and Carabus nemoralis, are both extremely rare in Aland. Altogether 3 specimens of the former species have been found at 3 different localities, and Carabus has been found only once in Eckero in the west. All these observations are recent, at any rate after 1930. The climatic conditions do not provide any explanation for this sparse occurrence in Aland, since the species in question extend much farther north on the mainland on both sides of the Bothnian Sea. The two remaining brachypterous carabids of the Stockholm region (and of the en- tire “region of comparison” 2a) which are markedly synanthropous, namely, Pristonychus terricola and Stomis pumicatus*, have not reached the mainland of Aland, but the former was found in 1944 in Ab Korpo, an island that has trade connections chiefly with Abo, where the species has long been recorded. It is thus striking that the immigration of synanthropous species to Aland Table 19. Distribution of 6 carabid species in the flight direction apparatus. Experi- ments 147-152 (p. 110) Number of observed Western sectors Eastern sectors Ratio cases Number % Number % West: East Oodes gracilis 100 63 63 Wy 17 3.7 Acupalpus consputus 200 107 5335 Dy 255 2.1 A. dorsalis 91 39 43 23 25 1:7 Badister dilatatus 34 19 56. 7 21 20, B. unipustulatus 27 16 59 3 11 3,5 B. peltatus 15 3 20 11 73 0.3 *Stomis is markedly synanthropous only in the northern part of its area. 259 231 that are flightless and thus dependent chiefly on anthropochorous dispersal has taken place in an extremely fortuitous and ineffective manner. It is all the more improbable to explain the presence of the entire functionally brachypterous element in the fauna of Äland by such transport (chiefly from Sweden). The completely normal composition of the element is evident from the fact that of the constantly brachypterous non-synanthropous carabids of the Swedish area of comparison in Upl (2a, map in Fig. 18), only the following 4 species are missing from the mainland of Aland: Agonum ericeti, Carabus coriaceus, C. glabratus, C. problematicus. b. We therefore have to assume immigration by a predominantly hydro- chorous mode of transport, chiefly from Sweden* for the functionally brachypter- ous carabids of Äland. This applies chiefly to the tokens 6 species of carabids and 3 species of chrysomelids: Calathus fuscipes Olisthopus rotundatus Dromius linearis Chrysomela haemoptera D. nigriventris Galeruca laticollis Licinus depressus Lycoperdina succincta Metabletus foveatus Keeping this possibility in view, when we study the surface currents of the surrounding sea (Fig. 19) we nevertheless find that in the strait which sepa- rates Aland from Sweden (Upl) they follow a highly unfavorable course. Along the Swedish (Upplandish) coast the direction of the current throughout the year is markedly southward. In spring this is true for the entire strait reaching up to Aland. Only in autumn does a northward current also appear immedi- ately off the west coast of Aland. Animals which happen to get in distress at sea along the part of the coast of Uppland that lies nearest to Aland must inevitably be carried south, and they must have already passed the latitude of Aland before they are able to cross over, even when they are gradually pushed eastward by waves caused by a strong westerly or southwesterly wind. Trans- portation to Aland is possible, at least theoretically, only if the starting point of the hydrochorous transport is placed farther north (at least up to north- ern Upl). However, the Swedish northern limit is situated farther south than Aland for no fewer than 5 of the 9 species susceptible of such transport, i.e., Dromius linearis, Licinus depressus, Chrysomela haemoptera, Galeruca laticollis, Lycoperdina succincta. The objection that these limits were possibly situated farther north during the postglacial warm period is irrelevant in the present context, since colonization of Aland was impossible during most of this period. Even during the Old Stone Age, Eckero comprised at the most two bare rocky skerries above the sea level at that time (Hausen, 1910a, map). *For the possibility of anemochorous transport, of flightless insects too, see p. 590. 260 232 There is, however, an exceptional case of hydrochorous transport, where the currents play a smaller role, and the winds have a correspondingly larger role: the drifting of ice in early spring (for its occurrence in the sea W of Aland, see maps by Erik Palmen in Hausen, 1946, pp. 107-108). Given sus- tained westerly to southwesterly winds (see wind roses, Fig. 20), ice-floes emanating from the coast of Uppland, crossing the relatively weak currents (Hausen, l.c., p. 97), can be carried to western Aland. The ice-floes may in- clude ones that were earlier squeezed together and pushed onto the beaches of the Swedish coast by easterly winds, where they might easily be covered with reeds, litter, etc. Such ice-floes, strewn with remains of plants, are fre- quently observed during spring passages in the Skargard of Stockholm. It may also be noted that at this time the hibernating insects are still more or less stiff and motionless, with low respiration, so they can better endure the ad- versity of a passage in spite of being drenched from time to time, and make no attempt to escape. Shore beetles can even endure freezing in solid ice for months, if reeds or other hiding places are present (Palmen, 1945a; see also p- 600). A detailed argument was provided above (p. 205) as to why passive trans- port of insects is more likely to lead to permanent colonization if it takes place in the adult stage. If the preceding exposition is correct, according to which the flightless carabids reached Äland from Sweden chiefly by ice trans- port, it must be expected that those flightless species that hibernate as adults are favored for the colonization of Aland. Table 13 (p. 230; last column) actually shows that out of all the islands of the Baltic Sea area considered, Aland shows the fewest larval hibernators. This is strong evidence for the above hypothe- SiS. In conclusion the following statement may be made on the immigration of the beetle fauna of Aland: The small number of flightless forms is characteristic for the Fasta-Äland fauna, especially as compared with that of the Skargard situated east of it. This element has colonized Aland chiefly from Sweden, and by means of hydro- chorous transport. The main immigration route of the flying species, however, is from the east and southeast (Finland and eastern Baltics). Some of this immigration has also reached the central Swedish mainland (p. 719). 3. The southwest-Finnish Skargard In the tables this region is indicated as the “Skargard of Aland.” This is Strictly incorrect, because most of the islands are located in the Province of Ab, eastward as far as Nagu. It is a very scattered archipelago, consisting of tens of thousands of islands and bare rocky skerries, situated at a much lower level than the main island of Aland. Even the larger islands are almost all only 20-30 m above sea level (Hausen, 1910b, pp. 2, 6). Only some points reach a few meters (< 50 m above sea level) higher (Bergroth, 1894, p. 8; Eklund, 261 233 1931, p. 11; Hausen, 1946, p. 17). Hence the entire area of rocky skerries is very young, and the first bare rocks emerged from the sea only at the beginning of the Littorina period. The population is fairly dense and the larger islands are more or less cultivated. The trade connections, especially with Abo, are active. The rich flora, which is largely similar to that of Fasta-Äland, is dealt with in the works cited above (p. 249). With regard to Coleoptera, only a few of the larger islands or island groups have been explored more thoroughly, primarily Foglo, Kokar, and Korpo, and in addition Sottunga, Kumlinge, Houtskar and Nagu. The composition of the carabid fauna naturally also shows a close relationship with Fasta-Aland. It is of course poorer, with only 124 species (as against 157), which include the following 10 species not found on Aland: Agnoum moestum Dromius quadrinotatus Amara crenata Harpalus luteicornis Bembidion grapei H. punctatulus Carabus convexus Panagaeus bipustulatus Cymindis vaporariorum Pristonychus terricola Of these, Agonum, Amara and Panagaeus are to be considered easily as accidental immigrants from the south, and perhaps also Bembidion grapei and Harpalus punctatulus, although at least the former may have originated from the Finnish mainland. Pristonychus is synanthropous and anthropochorous. The two functionally brachypterous species, Carabus convexus and Cymindis vaporariorum, are of special interest. On the whole, the brachypterous element of this Skargard region deserves special attention. Compared with the “region of comparison,” it is /arger than that of any of the Baltic Sea islands here considered (Table 12). This is not to be understood as having immigrants of anthropochorous transport (cf. Aland above). The reason must be that the Skargard region in question, due to its special features, is suitable for hydrochorous colonization. This is all the more striking in view of the young age of these islands, which is evident from a comparison with the outer islands of the same age in the Gulf of Finland, and with Gotska Sandon (see below). The usual freezing of the entire Skargard sea (Fig. 22) seems to be espe- cially favorable for transport by ice in early spring. However, if this represents the most important mode of colonization, one would expect a deficit of lar- val hibernators, as in the case of the mainland of Aland. But this is not the case (Table 13). It is therefore not rash to assume, in addition, a strictly hydro- chorous transport during the summer half-year from the Finnish mainland. The surface currents of the sea are especially in favor of it in autumn (Fig. 19, p- 247). 262 263 234 The tens of thousands of islands and rocky skerries of the Skärgärd region of southwestern Finland have functioned as a fine-mesh sieve, representing a big obstacle for the dispersal of hydrochorously transported insects from Fin- land to Fasta-Aland. This is hardly less true for anemochorously dispersed animals. Let us visualize concretely what happens if a living, hydrochorously transported flightless animal or an anemochorously carried animal capable of flight lands on an islet of the Skargard in vital condition. If favorable con- ditions for survival exist both kinds will stay and, under the most propitious circumstances (impregnated females or several individuals arriving simultane- ously), may realize a permanent colony. But in the vast majority of islands, at any rate on small rocky skerries the requirements for life for these species are lacking. What happens then? The winged animal flies on after forlorn search, toward an uncertain (but possibly lucky) fate. But the flightless animal stays on. It does not voluntarily jump into the sea in order to swim to new areas. Only an accident can take it there, just as originally it was an accident that it came into the water. On the contrary, the winged animal at one time, at the original starting point, voluntarily opened its wings, and can always repeat this action in voluntary flight. For such an animal the thousands of islands are welcome places of rest, at least after an anemohydrochorous adventure, and before starting for new trips. We may thus conclude that a dense, widely scattered Skargard offers an excellent springboard for the flying forms for further dispersal; for the flightless forms it functions as an obstacle, as compared to the open (but moderately wide) sea. This provides an important additional clarifica- tion of the fact that Fasta-Aland has obtained flying forms chiefly from the east and flightless forms largely from the west. However, these points are evidently not applicable to plants. However, it cannot. be denied that, especially for the colonization of the smaller outer rocky skerries, anthropochorous transport in hay-boats has played an important role, also with respect to insects, as has been vividly il- lustrated by Eklund (1931, p. 69). Such a role is nevertheless quite modest in my opinion. 4. Hogland Hogland differs from the remaining so-called outer islands of the Gulf of Finland in several significant ways. Primarily it is much older, older even than Aland. The highest point is 158 m above sea level, which shows that Hogland was never submerged completely during the postglacial period (Sauramo, 1942, maps on pp. 227-228). The distance from the Finnish mainland is 43 km, that from the nearest wooded islets about 20 km; Estonia lies 55 km away. Hogland actually represents a continuous basement complex rock forma- tion. It is 11 km long and up to 3 km broad and almost everywhere rises steeply from the sea. The fegion is about 20 km’, one-half of which is covered with 262 264 Fig. 22. Maximum extent of solid ice in the Skärgärd of southwestern Finland in mild winter (hatched and crossed [“blue ice”]) compared with normal winter (broken line). After Nordman (1943). forest (chiefly Abies and Pinus) (Valikangas, 1936, p. 513). In the lowlands along the coast there is also an admixture of deciduous trees, but there is no purely deciduous forest (Nordling, 1904, p. 122). In the interior there are a few small lakes (altogether covering about 19 ha); there are no swamps or bogs. The seashore is uniformly rocky or stony; only at one place is there a short sandy strip. The population of the two villages is about 900 people. They originally came from Virolahti (Vederlaks) on the Finnish mainland (Province of Ka) (Ramsay, 1896, p. 57), and there have always been stronger connections with Finland than with Estonia. There is not much cultivation. Hogland is to be considered a region strikingly poor in biotopes, and the flora is correspondingly poor (M. Brenner, 1871; Saelan, 1900). However, the carabid fauna, with 87 species, is unexpectedly rich, especially in comparison with the remaining outer islands, and some species possibly remain to be discovered. The xerophilous element is more poorly represented than on any of the other Baltic Sea islands (Table 14). Instead there is a preponderance of forest species. With respect to dynamics, as on almost all the islands, the fauna shows a predominance of dimorphic species (Table 10). But so far as is known, in seven out of 18 cases these occur only in the macropterous form 266 236 (Table 11), so that there is a distinct deficit of the functionally brachypterous group (Table 12). It is much less than in the case of the remaining outer islands to be considered below. For a more precise comparison it is advisable to place the functionally brachypterous (or dimorphic) species of all the outer islands side-by-side. The few records from Peninsaari Island were not considered, nor the dimorphic species Carabus clathratus, the condition of whose wings could not be deter- mined (Table 20). It is at once evident that of these 27 species only one, Prerostichus diligens, is common to all four islands. Among the 89 functionally macropterous cara- bids of the outer islands (cf. list on p. 270), on the other hand, the following 9 species (a good 10%) are common to ail four islands: Agonum marginatum B. saxatile A. sexpunctatum Harpalus aeneus Bembidion obliquum H. pubescens B. quadrimaculatum Pterostichus nigrita. B. rupestre Of these only the two species of Harpalus are to some extent favored by cultivation, and may thus have been introduced anthropochorously. The obvious conclusion is: The capacity to fly allows a rapid and appa- rently “consequent” colonization of habitable regions. But if sufficient time is available even the soil-bound, flightless faunal element reaches its destination, even though it may be unfavorably located (within reasonable limits). Hence of the above 27 flightless species of the outer islands only four are missing from Hogland, all of which are species of open terrain, and a fifth (Bembidion lampros) was found there only in macropterous form. None of the 22 functionally brachypterous species of Hogland is associated with human culture (anthropobiont), and only 2 species (Patrobus atrorufus and Prerostichus vulgaris) are to some extent considered cultural beneficia- ries (anthropophilous). Among species of the same ecological type apparently missing on Hogland, although widely distributed on both sides of the Gulf of Finland and more or less common, mention may be made of the follow- ing: Amara apricaria, A. eurynota, A. ingenua, Bembidion ustulatum, Carabus cancellatus, C. nemoralis, Pterostichus coerulescens. Of these the species of Carabus and Bembidion are functionally brachypterous. In my view these facts indicate that anthropochorous dispersal had at the most a subordinate role in the colonization of Hogland. As in the case of Aland and the entire southwestern Skargard, we should thus ascribe the richly represented flightless faunal element of Hogland chiefly to a hydrochorous transport. Since, in clear contrast to the insular regions mentioned, it shows a considerable preponderance of larval hibernators 267 (Table 13), it must be assumed that this transport took place chiefly in summer. 268 237 I am unable to decide whether this is due to unfavorable ice-drift conditions (cf. Tytarsaari, p. 272). It of course needs to be pointed out that the southern shore of the Gulf of Finland, where the main current of the surface water touching Hogland originates in spring (Atlas öfver Finl., 1910, map 9), is ice- free earlier than Hogland (Ibid., map 6b). There is very small possibility of the ice from this coast landing at Hogland. If we study the surface currents of the sea in the Gulf of Finland (Fig. 19, p- 247), which vary considerably in different years (Witting, 1911, p. 46), we find that especially in early summer they are not unfavorable to a fairly straight journey from deep in the gulf to Hogland. The turbulence occurring around this island might even bring about a concentration of transported materials. The map given by Palmén (1944, p. 66) shows the currents in late summer, when Hogland comes in contact chiefly with the eastward current, which signi- fies a longer transport route from the north coast of Estonia. I therefore find that Palmén’s discussion (l.c., p. 81), according to which more or less direct hydrochorous transport across the Gulf of Finland (Estonia—southwestern Finland) is impossible, does not apply to the conditions prevailing in spring and early summer. It is indubitable that the flightless beetles have also colo- nized Finland by this route (p. 603). It may not be possible to decide whether Hogland received its hydro- chorously immigrant species from Finland or from Estonia. Its functionally brachypterous species are widely distributed in both countries, with the excep- tion of Bembidion schuppeli and Carabus problematicus, which require further discussion. The species of Bembidion occurs here in brachypterous form, whereas only the macropterous form was found in the drift material from the Tvarminne area (NI). The south Finnish-Estonian subarea of the species has a markedly relict character, early postglacial stock, which established itself in the regions first colonized. The question arises whether Hogland was part of these regions. Still more striking is the record of Carabus problematicus, which is otherwise known in Finland exclusively from Petsamo, and is missing from northern and western Russia. Its occurrence in the Baltics is based on dubious records. The animal from Hogland was described as a different “morph” (Hellén, 1934, p. 41). But to this, much importance cannot be attached since it was based on a single specimen (Krogerus later found fragments of a second specimen). At any rate this undoubtedly represents a very old inhabitant of the island. In Scandinavia C. problematicus is certainly a Wurm hibernator, even in the highest northern part. It is to be assumed that even in Finland it closely followed the melting ice edge as well. It must be borne in mind that at the time of the “Baltic Ice Sea,” Hogland remained above water, whereas the entire southern coastland of Finland was covered with water up to the edge of the ice. A few islands occurred at the Isthmus of Karelia and NW of Ladoga (Sauramo, 1942, map on p. 227). 265 238 Table 20. Functionally brachypterous (or dimorphic) species of the outer islands in the Gulf of Finland m—AMacropterous form only; d—Both forms of dimorphic species Hogland Tytärsaari Lavansaari Seiskari Agonum fuliginosum + = + a Bembidion lampros (m) _ + (m) B. schüppeli + = er = B. unicolor + = = ke Calathus erratus +d + _ +d C. melanocephalus + (m) = sil C. micropterus +P eo = it Carabus granulatus + — + + C. hortensis + = a A C. nitens = + & BE C. problematicus + = a by Cychrus caraboides + = a2 pas Cymindis angularis _ + = & Dyschirius globosus + Sr — + Leistus ferrugineus + = = a L. rufescens + = = is Metabletus truncatellus + = a at Notiophilus aquaticus = os + ue N. palustris +d ~ Es we Patrobus atrorufus + = Lu u Pterostichus diligens + +f + + P. lepidus = a ut + P. minor +d (m) gr zy P. strenuus + + a Si P. vernalis +d +d (m) ir P. vulgaris ar +d + en Trechus secalis + = us ae Poa alpina and Viscaria alpina in the flora of Hogland (M. Brenner, 1871, p. 13) can be considered as counterparts of these beetles. Miscodera arctica likewise represents a northern and certainly an early immigrant. It is numerous only in the fjelds, but has been recorded in southern Finland by a few specimens, in Estonia actually by only one specimen. —In its history of immigration, Nebria gyllenhali corresponds with the three species considered. However, since it is widely distributed along the north coast of Estonia a later immigration from there is conceivable. It of course also occurs on Aland. —Both species are capable of flight. 269 239 5. The Remaining Outer Islands Only Tytärsaari, Lavansaari, and Seiskari are treated here because the other, smaller islands received only cursory study (Rosberg, Grotenfelt, Hilden, and Wecksell, 1923). Tytarsaari, about 17 km SSE of Hogland, is the highest and consequently oldest of these islands. It is almost circular, 8.2 km? in surface. The highest point, about 50 m above sea level (Rosberg, 1898; Valikangas, 1936, p. 513) is in the north where, as in the west, it has mountainous terrain (basement complex). This may have already emerged from the sea at the end of the Yoldia period, existing only as bare rock for thousands of years (Sauramo, 1942, map on p. 231). The distance from Estonia is 43 km and from the Finnish mainland about 65 km. In the south and the east the island is quite sandy, in the east there is even quicksand along the seashore. There are also gravelly and stony shores and at one place a small moist shore-meadow. The only bodies of fresh water are small lagoons and a brook near the village. The island is largely wooded, chiefly with Pinus, in the south predominantly with Abies and here and there an admixture of Betula, etc. (M. Brenner, 1871; Olsoni, 1927). —The population, more than 500, lives in one village in the south. The people came from Finland but the contacts have been mainly with Estonia, from where there was once considerable import of potatoes, hay, etc. Valikangas (1930, p- 76) believes that this explains the surprising occurrence of Eliomys quercinus on the island. Since there is no harbor, all freight must be transhipped in small rowboats. Cultivation is slight and consists chiefly of dry meadows and potato fields. The native flora is not rich. The carabid fauna includes more species than those of the other two islands, which is primarily due to better exploration. Lavansaari is situated 43 km E of Hogland and about 33 km from Tytarsaari. It is likewise nearly circular but has a more irregular shape; the former islet Suisaari has now merged with it. The region is almost 16 km’. There is no outcropping mountainous terrain, but the entire island consists of sandy moraine, which at some places on the shore merges into quicksand. The highest point is about 20 m above sea level (Krogerus, 1932, p. 61). Thus, the island may have finally stabilized only after the maximum rise in sea level (“Transgression”) of the Littorina period, about 4,700 years ago (Sauramo, 1942, p. 242). The distance from the coast of Ingermanland is 25 km and from the Finnish mainland a good 50 km. In the south there is a lake more than one km long, from which a rivulet runs into the sea. Furthermore, the island is poor in biotopes. Deciduous trees have a smaller role than in Tytarsaari. On the other hand, the littoral flora is richer (M. Brenner, 1871, p. 10). —The population, about 1,340 (1921), is restricted to three small villages in the northwest. Cultivation (gardens, potato fields) is very poor. In recent decades the trade connection has been chiefly with Kotka and Viborg, but earlier with Leningrad and Narva too. 270 270-271 240 The coleopteran fauna has not been fully explored. Seiskari is situated 25 km E of Lavansaari, a good 20 km from the coast of Ingermanland and almost 40 km from the Finnish mainland (Ik). It is 4 km long and about 4 km’ in surface. The highest point is about 15 m above sea level. Seiskari is thus the youngest of the three islands. The eastern part of the island is gravelly-stony, wooded (mostly Pinus), and the shores are mainly stony and slimy (Krogerus, 1932, p. 60). The western shore consists of highly mobile quicksand. There is no running fresh water, only a few bogs and ponds. The vegetation is generally scanty and the flora is poor in species. The population, about 850 (in 1921), is restricted to three villages on the west coast. Cultivation is extremely poor. In earlier times there was active contact with Leningrad and Narva but now chiefly with Finland. The coleopteran fauna has been insufficiently explored. The entire carabid fauna of the three outer islands considered here com- prises 77 species (Table 9). However, since they have not been treated sep- arately in Tables 9-14 it is advisable at this stage, for purposes of compari- son, to give an account of the functionally macropterous species of the three islands (Table 21). The functionally brachypterous species were listed above (under Hogland, Table 20), where the partially macropterous species, Bem- bidion lampros, Calathus melanocephalus, Pterostichus minor and P. vernalis, excluded from the present tabulation, were also considered. Table 21. Functionally macropterous (excluding dimorphic) species of three outer islands in the Gulf of Finland. Cf. Table 20 (p. 265) Tytarsaari Lavansaari Seiskari Also on Hogland Acupalpus dorsalis Ta _ + Agonum dolens

Diagram 46. Bradycellus collaris, macropterous form (continuous lines); brachypterous form (broken lines). Frequency curves for temperature pre- ferenda. Experiment 7, p. 70. Vertical lines indicate arithmetic and statistical mean values (in macropter- ous form coinciding), horizontal lines represent calculated amplitudes of variation of the latter. macropterous form 82.6 hours; brachypterous form 75.6 hours. d. Longevity under optimal conditions. Fourteen specimens each of the two forms were put in a glass jar with branches and “forna” of Calluna. The contents were kept moderately moist at room temperature. The animals were fed on breadcrumbs. During three months (September 23 through December 23, 1947) altogether 5 specimens died, of which 3 were macropterous and 2 brachypterous. The culture was then destroyed by mistake. These experiments apparently used far too little material. Hence they can- not support definitive conclusions on the selection value of the macropterous or the brachypterous form of dimorphic species on account of higher “general vitality.” For, as discussed below (p. 365), even positive selective features, which in the long run cause significant differences in the composition of a population, may show such slight effectiveness in one generation that they can be experimentally established with certainty only with very large material (> 1,000 readings). Nevertheless it is striking that of the 9 series of experiments carried out, only 2 (Bradycellus collaris, b and d) suggested a somewhat lower resistance capacity of the macropterous form. The other experiments indicated the op- posite. However, the difference in all cases is so insignificant that it lies well within the limits of statistically possible variation, and we may assume that—apart from flight capacity—there are no demonstrable physiological dif- ferences between the two forms of the species investigated (and probably among carabids in general). In any case we are not justified, as explained 317 below, nor is it necessary, to posit a general higher selection value of the brachypterous form, as Darlington does (1943). I will mention two other phenomena which in my view contradict the idea that the brachypterous form of the dimorphic carabids generally has to show higher vitality than the macropterous form. Such a selective advantage for the brachypterous form would be partic- ularly significant at the periphery of the area, as in cases where the species in question has attained its existence limits. In an area of distribution sta- bilized in this way it could be expected that an outer zone of predominantly brachypterous forms would develop (but not ofa “purely” brachypterous stock, since most individuals are heterozygous). However, no such situation can be demonstrated in the Fennoscandian fauna, and | doubt if the only reason is the short duration of the postglacial period. The apparent exceptions in the form of more or less “purely” brachypterous subareas in northern Finland and, particularly, in western Norway, are due to very different causes (the his- tory of immigration), which are discussed below in detail. In northern Sweden such a phenomenon is demonstrable only where the stock represents a direct offshoot of the Finnish or of the Norwegian area. An interesting special case is that of Notiophilus biguttatus. It is almost universally distributed in Fennoscandia, and in southern Scandinavia certainly represents one of the earliest immigrants, which has had long enough time in the area to achieve an equilibrium between the two forms by selection. Yet in southern Sweden, unlike most other dimorphic species, the macropter- ous forms are even in a majority (see p. 408). The reason seems clear: since the species is an eurytopic forest inhabitant the danger of elimination of macropterous forms that stray away in flight is small, i.e. selection does not 361 work against the flight capacity. A possibly existing advantage of vitality in the brachypterous form would have caused such a selection—which did not happen! 2. Selective importance of flight capacity. In its simplest form this factor was cited by Darwin (1859, pp. 135-136) when he set out to explain the richness of flightless Coleoptera on Madeira. He believed the flying forms were gradually swept out to sea by the wind. Darlington (1943) is certainly correct when he argues that the situation is not so simple. Exposure to the wind, namely, is not decisive; in the southern Appalachians as well as in the Greater Antilles totally wooded mountains are also characterized by the same high number of flightless carabids. Darlington tried to explain the significance of flight capacity on a purely ecological basis (p. 58): “The principal function of flight among geophile Cara- bidae is found to be to maintain sparse, unstable populations in large, unstable areas!.” He mentions a few examples of such “unstable” biotopes (also in his t (Original quotation in English; area = region; suppl. scient. edit.) 362 318 work of 1936, pp. 156 ff.). For the Fennoscandian Region the following eco- logical groups have to be taken into consideration: First, the ripicolous species. “Shores in general may be the most unstable of all biotopest” (Lindroth, 1943a, p. 126). In our region there is no con- stantly brachypterous carabid that is strictly ripicolous in this sense, with the characteristic exception of the two species, A&pus marinus and Trechus fulvus confined to the seashore, and Elaphrus angusticollis in eastern Fennoscandia. Nevertheless, Agonum moestum, Bembidion assimile, B. schüppeli and B. trans- parens are dimorphic, as also Carabus clathratus; also dimorphic are the less strictly ripicolous species Bembidion clarki, B. gilvipes, B. guttula, Dromius sigma and Pierostichus anthracinus. Second, the species of open terrain, in the large high boreal forest re- gions, where only small, often variable patches are available. It is quite cha- racteristic that in the most undisturbed forests of Fennoscandia, such as those of eastern Karelia, such dimorphic species as Amara infima, Bradycellus col- laris, and Pterostichus strenuus are found predominantly in their macropterous form. Third, the species of cultivated soils. The farther north, that is the more re- cent the cultivation and the more predominant the macropterous carabids on cultivated soils. North of latitude 64°N there are absolutely no brachypterous species associated with cultivation or even favored by cultivation. The excep- tions are just two solitary records of Pterostichus vulgaris (possibly resulting from passive dispersal), in addition to Bembidion lampros, which in the north is slightly favored by cultivation. “In small, stable areas (= regions)?! where populations are dense and sta- ble, flight presumably loses this function” (Darlington, l.c.; cf. quoted above). Such biotopes occur in Fennoscandia mostly on the seashore and in high re- gions of the fjelds, both of which have at least a relative stability. Much more pronounced ones are found outside our regions on old islands and (from the biological viewpoint) old montane regions. It is strange, however, that Dar- lington on the one hand ascribes a great positive role to the flight capacity for those species that are able to protect themselves in this way against alter- ations in the biotope, on the other hand he does not concede the correspond- ing negative role to the same characteristic for “stable” species. He repeatedly emphasizes that for these species the flight capacity is “useless, but not nec- essarily harmful.” I find this illogical. The flight of all or some individuals of a given population signifies rapid dispersal or thinning out over a larger region*. If thereby new habitable land is colonized the result is positive (from t (Original German quotation translated into English; suppl. scient. edit.). tT (suppl. scient. edit.). * Concerning the flight capacity of carabids and its ecological significance, see a later section (pp. 573 ff.). 363 319 the viewpoint of preservation of the species), especially if the original habitat has changed in a disadvantageous direction. But if the emigrants get an un- suitable region they are destroyed without producing progeny and this must happen the ofiener, the smaller and more isolated are the regions habitable by these populations. Concrete examples are provided by the drift material of insects, especially at the south coast of Finland (see also Bembidion gilvipes and B. lampros in Part I). Palmen (1944) assumed with reason that they originated predominantly from Estonia. The occasional occurrence of a larger number of dimorphic species* in the drift material exclusively in macropterous form must signify a considerable impoverishment thereof in the original populations in Estonia, without any conceivable corresponding influx of immigrants to the regions of population. Selection has taken place in favor of the brachypterous form. The same is true of the frequent “stragglers” of dimorphic species of the plains, also observed by Darlington (for instance, 1943, pp. 59-60) on the “Presidential Range” (New Hampshire) in the regio alpina, where they must die with no possibility of reproducing. Finally the maps on the distribution of the two forms of a number of cara- bids, discussed in greater detail below, show that selection actually takes place, which at least in part depends on the flight capacity of the macropterous forms. Attention may also be drawn to an experiment with Drosophila which, however, was somewhat too simple (l’-Heritier, Neefs and Teissier, 1937). A mixed population of “wild” and “vestigial” types was exposed to a moderate wind in the open over a long period of time and was continually studied statis- tically. The experiment showed the not unexpected result, that the component of vg in the population gradually increased. Darlington’s reluctance to accept Darwin’s old concept**, even with reser- vations, is evidently due to the fact that he thought more of passive wind trans- port (“blown out to sea”: Darwin, l.c.) than on the active flight of insects. He claims wind exposure for those localities where the brachypterous forms are in a majority, to support Darwin’s hypothesis. However, the voluntary flight of the animal is the essential precondition, and a selection in favor of brachypterism may take place equally well in pro- tected (for example, wooded) places, provided the habitable areas (for example, a certain region of altitude in the mountains) are sufficiently small and isolated. *The dimorphic carabids of the Finnish drift material (Palmén, 1944) are: Agonum fuligi- nosum, A. moestum, A. obscurum, Bembidion assimile, B. gilvipes, B. grapei, B. guttula, B. lampros, B. schüppeli, B. transparens, Bradycellus collaris, Calathus erratus, C. melanocephalus, Metable- tus truncatellus, Notiophilus aquaticus, N. biguttatus, N. palustris, Pterostichus diligens, P. minor, P. strenuus, P. vernalis, Trechus rivularis. **Darlington (1943, p. 58) states: “Facts here given contradict the old Darwinian idea that insects on mountains and islands are flightless, with atrophied wings, because flying forms are blown or straggle away from exposed environments.” 364 365 320 Probably anabatic mountain winds (anabatic valley winds; Geiger, 1942, p. 236) in the afternoon have an important role in transporting “stragglers” to higher, unfamiliar areas. A good example is provided by the frequent mass occurrence of woodland insects in snowfields of Fennoscandian mountains. The only situation where selection in an isolated population of dimorphs would not affect the macropterous form would be where this form does not fly. Nevertheless, the above list (Table 27, p. 377) shows that 25 of the 67 European dimorphic species have been observed flying, and this feature could apply to the macropterous form of all species. I regard as confirmation of the above exposition the fact that among the dimorphic species widely distributed in our region the macropterous form has been maintained the longest—also in regions colonized early in the post- glacial period—in such species that are markedly eurytopic (examples: Bem- bidion lampros, Bradycellus collaris, Calathus erratus, Notiophilus aquaticus) or live in extensive and continuous biotopes (forest species: for instance, Notiophilus biguttatus, see p. 360; eurytopic hygrophilous species: Bembidion guttula, Pterostichus minor, P. vernalis), apparently because in such cases the macropterous migrants do not perish so often. On the other hand, in the case of stenotopic species with limited possibilities of colonization, selection sooner favors brachypterism. Examples are: shore species (on the seashore and larger lakes), such as Bembidion assimile and B. schuppeli; and xerophilous species, such as Amara infima, Bembidion nigricorne, Cymindis macularis, Dromius lin- earis, Pterostichus lepidus. But there are many exceptions to this rule, due not to present but to past conditions of the environment. These are considered below. Stability, area restriction, and isolation of the biotopes favor brachypterous species and forms. Variability, extension, and moderate splitting up of biotopes favor macropterous species and forms. From this it follows that in periods of alternating stability and variability—above all with respect to the climate—species with wing dimorphism are favored. This is applicable for the Quarternary period with its alternating glacial and interglacial epochs. Hence Darlington and I came to the same final conclusion; our conception differs only with regard to the way selection operated. Since the development of wing dimorphic species in the Fennoscandian region now apparently opposes any increase in the brachypterous form, it would be interesting to study the purely theoretical calculations, that have been made on the effect of selection in closed! populations. Perhaps in this way an idea of the length of time required can be achieved. Such calculations were undertaken by Haldane (1932, p. 184), in part cited in Huxley, 1942, p. 56 etc. A pertinent case in this connection is selection in favor of a dominant gene located in an autosome. Other factors (recurrent mutation, chance) are here ignored. The assumed selective value is very low, 0.001. That is, the ratio tin the sense of “isolated”; suppl. scient. edit.). 366 32 between the phenotypic dominants and phenotypic recessive increases from 1 to 1.001 per generation. In the carabids (with very few exceptions) one generation can be equated with one year. Increase of the favored dominant takes place at the following rate: Population increase from 0.001 to 1% in 6,920 generations. Population increase from 1 to 50% in 4,819 generations. Population increase from 50 to 99% in 11,664 generations. Population increase from 99 to 99.999% in 309,780 generations. If the selective value of the dominant is higher, the process will be that much quicker in the beginning. But the ominous last one percent is not com- pletely eliminated even with total selection of the recessive. The latter can: affect only the homozygotes, since only these express the recessive gene pheno- typically. Even in the extreme case where these homozygotes constitute only 0.001% of the individuals of a population, not less than 0.2% of the dominants are heterozygotes (Haldane, 1932, p. 185). Applying these considerations to mankind one came to the conclusion that sterilization of the recessive homozygotes with one or other hereditary defect is successful only where the gene in question occurs fairly frequently, as perhaps in hereditary imbecility. It is evident from these calculations that even a selection that operates lightly brings about a fairly rapid increase in the favored mutants up to 99%, but the last one percent can never completely disappear by selection alone if this mutant is dominant. This is the problem of Achilles and the tor- toise: an unending series of gradually decreasing numbers. —However, if a population is small and remains isolated over a long period of time the in- ferior form is finally eliminated by chance. If we imagine a population nor- mally with 100 individuals on an average in which the gene for the recessive character with low selection value is present in only 1%, thus on an average present in one individual at any time (heterozygotic), then it might happen in the course of generations that as a result of this or that mishap—i.e. by chance—this heterozygote one day fails to develop or find an opportunity to reproduce. In the case in question, where the inferior factor represents a recessive gene, it is naturally impossible to decide whether this gene has completely disappeared or not in a given population, be it experimentally or by observa- tion in nature. In mammals and other larger animals it is perhaps practicable to verify the existence of the recessive gene by active crossing of all indi- viduals of the population (with the recessive homozygotes). In insects this is simply not possible since we cannot track down all the individuals of even smallest populations in nature. Remainders of “the macropterous gene” are even certain in many carabid species, that we today consider (rightly, from the phenotypic point of view) “constantly short-winged.” However, similar cases where the rare accidental pairing of two heterozygotes has resulted in solitary 367 322 macropterous individuals (for example, in Dromius sigma, Notiophilus germinyi, N. reitteri, of which only one macropterous individual is known in each case), have provided occasion for use of the dubious term “atavism” (“regressive form,” Kolbe, 1920, p. 393), All brachypterous and apterous species of Coleoptera have undoubtedly arisen secondarily (cf., on the contrary, Horn, 1907). The study of species with wing dimorphism reveals the evolutionary path taken: mutation, selection and chance*. Species whose wings are particularly heavily reduced, have perhaps been affected by more than one mutation. Because of this, and perhaps due to the markedly favorable effects of the wing-reducing mutation on the general yitality of the insect, certain species can achieve a condition of genetically “pure” brachypterism more rapidly than others. However, the main precondi- tion might have been temporary isolation in small populations. Distribution of Dimorphic Forms The geographical distribution of the two forms of a wing-dimorphic species of insect has been little studied. J. Sahlberg (1868) and Hak. Lindberg (1929) give a very general statement that brachypterous Hemiptera in the north and in the outermost parts of the Skargard of southern Finland (Lindberg) are more frequent. With regard to Gerris asper L., Ekblom (1941, p. 61) denies this. Darlington (1938) attempted to map the three North American species of Patrobus. I have published (Lindroth, 1939) preliminary dimorphic maps of Calathus mollis, Bembidion transparens, and B. aeneum. In the zoogeographical context, as in the present contribution, the chief concern is to map the dimorphic species and to explain the results. For species whose macropterous form is extremely rare a map would be superfluous. For other species sufficient material was not available or the distribution of the two forms did not appear to support any conclusions. Nevertheless, of the 50 dimorphic carabids occurring in Fennoscandia exactly one-half were found to be of zoogeographical interest in one way or another, and for this reason 25 dimorphic maps are provided in the present account. Before I discuss the individual maps a few remarks on their method of notation might be in order. Each locality (or complex of localities located close together) is indicated by a circle whose size** is proportional to the number *] cannot consider Reinig’s (1938) “elimination” as a factor comparable with the three men- tioned above. It is a result of chance, to a certain extent reminding one of Sewall Wright’s “drift effect” (for instance, 1931, p. 147, Fig. 18; 1932; see also Huxley, 1942, p. 58 and elsewhere). More- over, the term chosen is unfortunate since what is involved is not actual elimination (= destruction) but a sifting out and uneven distribution of the genotypes throughout the total area of a species. **The relative size is easily underestimated. The largest circle of Bembidion grapei (38 Alta), for instance, corresponds to 80 individuals (Fig. 50). 369 323 of individuals examined*. It must be borne in mind that an aggregation of large circles does not necessarily signify a “focal point” of the area, but rather a well-explored region. However, the larger the circle, the more reliable in general** is the picture it provides of the actual distribution of macropterous and brachypterous forms. Distribution of macropterous individuals is shown by blank circles or blank sectors, of brachypterous individuals (intermediate specimens are also included) by black circles or sectors. The reliability of a dimorphic map is completely dependent on the amount and versatility of the material studied. However, it is almost always impossible to obtain voucher specimens of a species from all the Fennoscandian localities for study. In some cases this is to be lamented, in most other cases the gaps were more or less filled with nearby localities. To avoid unnecessary confusion, localities that have not been taken into account have generally been left out of the distribution maps; in a few cases they have been marked with crosses. Comparison with the general distribution maps in Part II of this work allows at least a judgement on the completeness of the dimorphic forms. The arrows in some of the maps mark the more or less established postglacial migration route. Occasionally areas populated predominantly or exclusively by brachypterous forms are demarcated with special lines. For these and other indications see key with each map. It is advisable to begin with Calathus mollis (Fig. 28, cf. also Fig. 25). The distribution in Sweden distinctly illustrates the principle that is more or less clearly valid for most of dimorphic species (originally probably for all): Ar the periphery of the area of distribution (in this case in the north, particularly inland) there are predominantly or exclusively macropterous specimens; a prepon- derance of brachypterous specimens is found in regions colonized early. According to the modern “invasion terminology” it may be said that the macropterous forms represent the “parachutists,” the first to come in by air, whereas the brachypterous form is the soil-bound infantry, which has to march on its own. Since the macropterous form represents homozygotes, no brachypterous spec- imens can result from their pairing (unless a wing-reducing mutation recurs); on the other hand the brachypterous stock more or less comprises heterozy- gotes, from which macropterous specimens can be produced. In extensive material from western Denmark and Norway (respectively 95 and 19 specimens), not a single brachypterous specimen was found. Apparently an isolated, genetically pure macropterous stock with a western distribution predominates, for instance, in West Germany and in the British Isles. It can also be separated by insignificant morphological characters (including those of the parameres of the male) from the Swedish form (“f. typica”; see Sup- *Only the map of Bradycellus collaris (Fig. 48) is not quantitative. **Exceptions are mass catches of wind-driven individuals, which are exclusively macropte- rous. For instance, see northern Oland in the map of Bembidion lampros (Fig. 40). 324 368 Fig. 28. Calathus mollis. Distribution of wing-dimorphic forms. Blank circles and sectors: Macropte- rous specimens. Black circles and sectors: Brachypterous specimens. The area of circles is proportional to the number of individuals examined. Two circles outside the map indicate Braunschweig (one specimen) and Berlin (53 specimens). 370 371 325 plement). Both forms occur together only on Bornholm, but there only the eastern one (“subspecies erythroderus’’) is dimorphic. The species undoubtedly reached Norway by flying across the sea from Jutland (or from the old “Doggerland”), but Sweden was not reached via Denmark. In Denmark (excluding Bornholm) just one macropterous specimen of the eastern form was found near Copenhagen. Immigration to Sweden must have taken place directly from the south. Another noteworthy fact which now emerges is the strong preponderance of the brachypterous form not only in the southern Skane but also on Oland and Gotland (to a lesser extent on Smaland along the Kalmarsund strait), whereas in the intervening regions, on Smaland and Blekinge, only the macropterous form occurs almost everywhere. Study of more material from the last-mentioned province would be highiy desirable, but even if the percentage of the brachypterous form was found to be fairly high, we could not speak of a coherent, homogenous brachypterous stock in southeastern Sweden. Three hypotheses might explain the situation: 1. The environmental factors on Oland and Gotland operate in favor of a stronger selection of the brachypterous form (or even the emergence of recurrent mutations). 2. The desired homogeneous brachypterous stock (Ska—Ble—Old—Gtl) actually existed earlier during a climatically more favorable period. With the worsening of climatic conditions, the species as a whole was pushed back toward the southeast coast and was restricted to two subareas perhaps due to a gap in Ble. Subsequent recolonization in the present comparatively favorable climate took place chiefly by the macropterous form. 3. Oland and Gotland obtained their mollis stock by separate immigration and nor through Skane. The map shows that I consider the last of these the only acceptable hy- pothesis. Hypothesis 1 should also have to explain why selection (or mutation frequency) on Skane has been stronger on the south than on the other coasts. Moreover, some of the following maps of the other species show that the operation of selection on Oland and Gotland generally does not favor the brachypterous form in general (for example, Harpalus azureus, Bembidion as- simile, B. obtusum). The problem of the origin of recurrent wing mutations is considered later (p. 403). There are more serious objections to hypothesis 2. Such a “pulsating” postglacial immigration, even with periodic retreats, is certainly true of most species. Such a hypothesis cannot be ruled out. However, the following fact strongly supports hypothesis 3: Calathus mollis reached Skane and (partly) Bornholm not via Denmark but directly from the German Baltic Sea coast (even in brachypterous form). It is difficult to see how this could have come about other than by direct land connection. Under such standard conditions in the southern Baltic Sea, if not a firm land connection, then a far more favor- able situation must have existed between Oland—Gotland and the continent, 372 326 which made possible a direct immigration in that direction. This question has already been considered above in detail in the section on “The Fauna of the Islands” (pp. 306 ff.). It was clear that Calathus mollis is only one of numerous species on Oland—Gotland which on faunistic grounds indicate the mentioned land connection. To the north a counterpart of Gotland in Gotska Sandon, which has remained continuously isolated since its postglacial origin. C. mollis certainly reached there by flight; the 10 specimens studied are macropterous. Harpalus picipennis and H. neglectus (Figs. 29, 30) provide interesting com- parisons with Calathus mollis and are to some extent its counterparts. The former species was found in Skane exclusively in macropterous form (13 spe- cimens), whereas both forms occur on Oland. On the other hand neglectus is dimorphic on the Swedish mainland (Ska, Hll, Ble) with a preponderance of the brachypterous form, whereas on Oland of 26 specimens studied only 2 were brachypterous. The assumption that picipennis reached Oland via Skane is improbable. On the contrary, a separate immigration to this island must have taken place, as assumed for Calathus mollis. In some cases the morphic map provides very limited indication of the history of the immigration of the species, valid only for part of the area. Prero- stichus anthracinus may be mentioned as an example (Fig. 31). Its northern limit in Sweden, formed by the River Dalalven to the east, is apparently stable and relatively old, since it has been attained by both forms. But in the west (in Dlr and Vrm) only macropterous individuals have been recorded which apparently represent a later immigration. Four brachypterous specimens were Fig. 29. Harpalus picipennis. Fig. 30. Harpalus neglectus. Distribution of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). 373 375 377 327 found in Sv north of the mouth of the River Swir (outside the map), which is the only eastern Fennoscandian locality. Evidently this population did not arise from an accidental migration. In the case of Olisthopus rotundatus (Fig. 32) both forms coexist almost throughout its area, with the noteworthy exception of the mainland of east- ern Fennoscandia, where so far only macropterous individuals (altogether 18) have been collected, except for a single specimen from Ik Metsapirtti (coll. LBG!)*. But on the main island of Aland the brachypterous form is already predominant (24 of 27 specimens studied), which conclusively supports the Swedish origin of this stock. For further remarks see the section on the Fauna of the Islands (pp. 239 ff.). In Amara infima (Fig. 33) the macropterous form is rare. It has been recorded only in eastern Karelia as the only form occurring there (however only 3 specimens). Moreover, the brachypterous form has also been found in the northernmost isolated localities. These are thus no accidental migrants, which is of considerable interest in view of the discovery of this species only recently in Finland. The discontinuous distribution of the species probably reflects the concealed, almost subterranean mode of life of the animal. Harpalus azureus (Fig. 34) is the only example of a species that seems to occur on Oland exclusively in brachypterous form (86 specimens studied), but exists on Gotland in both forms. The distribution of the species there is not completely irregular, but the macropterous form predominates, especially in the north, including Faron, and also in the eastern part of the island. In the entire northern one-third, only one brachypterous specimen (Farosund, JNS!) was found, as compared to 29 macropterous specimens. Apparently this part was colonized later. The problem is dealt with in more detail in the section on the Fauna of the Islands (p. 310). The situation with regard to Calathus erratus (Fig. 35) is unusually clear. This species exemplifies the principles formulated and discussed for C. mollis on a large scale. The hind wings of the macropterous form of C. erratus are weak and somewhat variable in size. Perhaps all the macropterous specimens (with the reflexed apical part of the wing) are capable of flight. However, they might be genetically uniform. That the markedly macropterous specimens have flight capacity is proven by the detection of this form only (64 specimens) in Finnish marine drift material (Palmén, 1944, p. 146). In all three countries, north of latitude 62-63° N C. erratus occurs exclusively in macropterous form, whereas in the south, especially in Sweden, the brachypterous form predominates. It is quite clear from the map that the latter form has been able to advance in Norway along the Gudbrands valley *One old brachypterous specimen (MH), bearing the label “Viborg. coll. Mannerheim,” is better ignored. Fig. 31. Pterostichus anthracinus. Distribution of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). Outside the map area, 4 brachypterous specimens were found at the mouth of River Swir in eastern Karelia. almost up to latitude 62° N, but the watershed toward the Atlantic seaboard formed an effective barrier which could be crossed only by the macropterous forms. To the inner Sogn (Province 19), too, the species apparently came from the east by air. But in southwestern Norway (Provinces 4-6) exclusively the brachypterous form was found (26 specimens), which is evidence of an old colonization. In wooded areas, especially in central Finland and eastern Karelia, the macropterous form of this species associated with open terrain is conspicuously favored. Noteworthy is the exclusive occurrence of the macropterous form in wes- tern Norway between latitudes 60° and 69° N. Evidently external factors (cli- 329 373 Fig. 32. Olisthopus rotundatus. Distribution of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p- 368). Crosses indicate localities where no material was available. matic or others) are not alone responsible for the origin or advantage of the brachypterous form in western Scandinavia. It is advisable to keep this fact in mind in the following treatment of other distribution types, which are nearly contrary to Calathus erratus. The distribution of Pterostichus vernalis (Fig. 36) in the southern half of Scandinavia corresponds with that of Calathus erratus. The peripheral zone (northern and western), which is exclusively populated by the macropterous form, is less extensive. Unfortunately no reference specimens were available from 4 localities in western Norway north of Bergen, or from one locality in the upper Gudbrands valley, but with fair certainty they belong to the purely 378 macropterous zone. In contrast with Calathus erratus, Pterostichus vernalis is Fig. 33. Amara infima. Distribution of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). represented in southwestern Norway (Provinces 5-6) almost exclusively by the macropterous form, which appears to have immigrated there relatively late along the south coast. Interesting conditions, differing from the aforementioned, are found in the Bothnian coastland, where the south Swedish stock extends only up to Mdp and is represented at this northern limit only by the macropterous form. But on the Finnish side along the entire coast the area is continuous, with its Outermost point represented by one locality in Sweden (Nbt), which here contains brachypterous individuals. This is strange, because a broad apparently “purely” macropterous zone seems to straddle central Finland. The possibility of brachypterism having arisen in the Bothnian coastland by late recurrent mutation cannot be denied. Or instead, the occurrence there is of relict nature. 331 375 Fig. 34. Harpalus (Ophonus) azureus. Distribution of wing-dimorphic forms on Gotland. For explanation see p. 367 and Fig. 28 (p. 368). The following species, Prerostichus lepidus (Fig. 37) and Carabus clathra- tus (Fig. 38), show a high degree of correspondence. Both occur in the south almost exclusively in brachypterous form: In Norway this is the only form recorded, in Sweden south of latitude 60° N only 4 and 2 brachypterous in- dividuals were found respectively (among 197 and 85 brachypterous speci- 376 Fig. 35. Calathus erratus. Distribution of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p- 368). In southern Norway the main watershed is indicated. Macropterous drift material from southern Finland (Palmen, 1944) has not been consi- dered. 378 333 Fig. 36. Pterostichus vernalis. Distribution of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). Crosses indicate localities where no material was available. mens respectively); in Finland and east Karelia south of latitude 63°N several macropterous individuals of the Carabus species were found (5 among 27 brachypterous specimens). On the other hand only one macropterous indi- vidual of P. lepidus was found (among 96 specimens studied). No conclusions concerning immigration can be drawn from these parts of the area. But in the north, especialiy on the Swedish side, the macropterous form occurs more frequently. In the case of P. /epidus, on both sides of this “macropterous zone” the area is continuous, toward the southwest through Scandinavia and toward 379 Fig. 37. Pterostichus lepidus. Distribution of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p- 368). 380 Fig. 38. Carabus clathratus. Distribution of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p- 368). Arrows indicate possible routes of postglacial migration. 382 383 336 the southeast through Finland and east Karelia. We saw earlier that a high representation of the macropterous form indicates a recent colonization. We are consequently justified in seeing the zone of P. /epidus as the “scar,” where the western and eastern stocks met and merged. It is characteristic that it lies on the Swedish side, which is not clearly evident from the usual distribution map. Carabus clathratus actually differs only in one point, which is important: In Sweden south of the “scar” there is no deteciable connection with the south Scandinavian stock. The old record in Hls is of course enigmatic, but may be connected with the northern subarea, because in northern Upl, in Dir and Gst and in some regions that have been very well explored, this splendid species could hardly escape the collectors. I suggest we can understand the riddle better from the records on three small islands in Kvarken (the narrowest part of the Gulf of Bothnia). Probably the insect was successful in colonizing Sweden too by this route, in predominantly or (genetically) purely brachypterous form, so evidently by hydrochorous transport. The “scar” would then represent not the boundary between an eastern and a southern stock but between two eastern stocks. In all probability, independent origin must also be ascribed to the isolated fairly numerous southwest Norwegian stock. This would mean that Carabus clathratus reached Scandinavia by 4 different routes. The two species, Prerostichus lepidus and Carabus clathratus, show a further correspondence in a strip of macropterous specimens across central Finland (at about the height of Ule-trask), which is most evident in the case of Prerostichus species. We find the same manifestation in the text below, especially in the case of Pterostichus minor. It is therefore quite possible that in this region too we have the last trace of a boundary between two originally separate immigrant . stocks. Pterostichus vulgaris (Fig. 39) represents a dimorphic species, whose macropterous form is extremely rare and is apparently continuously declining. From south Sweden (south of about latitude 62°20’ N), among hundreds of specimens so far collected, only 3 were macropterous (Ska Bastad; Boh Fjällbacka; Nke Hasselfors). As was to be expected, an exception is provided by the border zones, which could be checked only with Swedish material. There the macropterous form occurs somewhat more numerous partly along the boundary in the Province of Vbt and in northern Ang (8 macropterous, 3 brachypterous specimens) and partly in the localities that are highest and farthest from the coast in the parish of Lima in Dir (10 macropterous, 23 brachypterous specimens). The last-mentioned material (Axel Olsson) was collected at irregular intervals over a period of almost 30 years. I tried to find out whether during this period a shift took place in Lima in favor of the brachypterous form. But I was unable to establish this, since the macropterous individuals were distributed fairly uniformly over the years. It is particularly interesting that the macropterous form occurs at the 382 337 Fig. 39. Pterostichus vulgaris. Distribution of wing-dimorphic forms in Sweden (and Äland) north of lati- tude 60°N. For explanation see p. 367 and Fig. 28 (p. 368). northern limit in the coastal region but is unknown in the interior parts of Äng and Jtl. The former region was apparently colonized later, even though some inland localities (Jtl Ulriksfors, Ang Tasjo) are at least equally far from the area of departure of the immigrating animal. This is a distinct example that river valleys are important migration routes in high boreal forest regions, where the intervening distances, which are partly forest, partly swamp, and only slightly cultivated, pose great difficulties especially to soil-bound animals of the open terrain, if they attempt to emigrate from one riparian region to another. The dimorphic map of Bembidion lampros (Fig. 40), the most frequent and most eurytopic of our Bembidion species, has been worked out only for Sweden and north Finland. It provides no indications as to the immigration routes, since the brachypterous form, at least in Sweden (from north Finland too little material is available) has everywhere reached the outermost limits of the area. At first glance the picture appears to be completely uninteresting. But, this is not so. Actually such a map holds a very special interest, although 385 387 338 in a field that has not been touched on so far. What can be the significance of the fact that the brachypterous form has everywhere caught up with the macropterous one, in the north as well as in the mountains? Evidently, that the species in question under the prevailing conditions (climate, cultivation, etc.), has reached its existence limit (existence-ecological limit). If not, if the species would still be in the state of expansion, then a border of leading “parachutists” should be expected. —The reader will perhaps ask whether in zoo- and phytogeography generally such existence limits are not taken into consideration, since isotherms, pH lines, occurrence of limestone, etc. are repeatedly used to explain the area of distribution. But the difference is that in most cases this is on the assumption only that the species concerned has achieved its existence limit, whereas in the case of Bembidion lampros we are justified in asserting that it has. It must therefore be of great general interest to map as precisely as possible the geographical limits of all such “stable” dimorphic species (provided that the macropterous form occurs more than as solitary individuals, which are inconsequential for dispersal biology), so we can determine the typical course of true existence limits, for instance, the northern limit within the Fennoscandian region. “Per analogia,” we could then also distinguish between the existence and dynamic limits of species lacking wing dimorphism (details see below, p. 616). Among the species considered earlier, the Swedish northern limit of Prero- stichus vulgaris (Fig. 39) in particular, has the same stable type as that of Bem- bidion lampros. This is also true of B. gilvipes (Fig. 41), a species that occurs in all three countries in its brachypterous form up to the northern limit. Mapping was therefore limited to regions north of about latitude 60° N. In Bembidion guttula (Fig. 42) we come across a new type in a way, where the geographical distribution of the macropterous and brachypterous form is not what might be expected. In the south, at least until latitude 62° N, the two forms are found intermixed; in one region the macropterous form and in an- other, somewhat more often, the brachypterous form predominates, but there is no extensive area where brachypterous specimens are found exclusively. The exception is found, strangely enough, in the far north, in Finland and in the Russian parts north of latitude 64°30’ N where solely the brachypterous form was found (altogether 11 localities, with 14 specimens). This popula- tion advances in Nbt also to the Swedish region, and characteristically shows macropterous specimens only at its outermost limit toward the west, much like Pterostichus vulgaris. The markedly mixed stock of south Scandinavia has evidently advanced to Vbt; the comparatively narrow gap between Vbt Vin- deIn and Nbt Alvsbyn may therefore be a real one. No corresponding gap has been demonstrated in Finland. What is the explanation for this northeastern, almost purely brachypterous stock? We might be able to assume that north Sweden was colonized earlier from the east than from the south. But the tip of this eastern advancing stock me Wit | | 4D | gist SS 384 Fig. 40. Bembidion lampros. Distribution map of wing-dimorphic forms in Sweden, in Norway north of latitude 63° N, and in Finland north of 64° N. For explanation see p. 367 and Fig. 28 (p. 368). The numerous macropterous specimens from north Oland originate from drift material. 340 in Nbt, indicated by the occurrence of macropterous specimens cannot of course be blocked from advancing to the south by existence limits. It must be assumed that this stock has already advanced farther in its development toward brachypterism. It thus does not represent the northernmost outpost of the stock native to south Finland, which, especially in the southwest, reveals a strong macropterous component. But for this species it cannot be decided where in Finland the boundary between these two eastern stocks is to be drawn. Pterostichus minor (Fig. 43) corresponds in important features. Only the presence of an actual gap in the distribution in north Sweden is doubtful (the Province of Ang has not been sufficiently explored). Instead this map 389 shows in a highly instructive way the difference between a group consisting predominantly of macropterous specimens immigrating late into central Nor- way (Trondheim region), and the north Finnish stock, which reveals its latent macropterism for the first time on the Swedish side. That this immigration was not very late is indicated by a fossil record of the species (probably from the early Littorina period) near Nbt Alvsbyn (p. 671). The two arrows show the possible migration routes. Here too the insects immigrated to Finland by two separate routes, but once again it is difficult to draw the boundary be- tween them in central Finland. This is easier for Bembidion transparens and Pterostichus strenuus, which are considered below in detail. + 8S Ä 385 Fig. 41. Bembidion gilvipes. Distribution map of wing-dimorphic forms (excluding south Sweden). For explanatıon see p. 367 and Fig. 28 (p. 368). 341 Fig. 42. Bembidion guttula. 386 Distribution map of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). 392 342 In the case of Bembidion nigricorne (Fig. 44), which has very limited dis- tribution, the conditions correspond in principle to the extent that in the northern subarea, which forms a peculiar belt across central Finland around latitude 65° N, only brachypterous individuals were found, but in south Fin- land and east Karelia, the macropterous form was also found. Both specimens found far from each other in central Sweden were brachypterous as well, but the only two localities on the west coast are represented by 3 macropterous individuals. In the north we apparently have the remains of a once continuous area which in the case of this stenotopic species of the open sandy Calluna heaths suffered greatly from the changes in the landscape and least of all secondarily in the climate. Bembidion transparens (Fig. 45) shows total separation between a northern and southern stock in eastern Fennoscandia with a broad intervening gap. The latter has spread to central Sweden, where the Malar lake region represents a secondary center. From here the species has dispersed in all directions, with a continuous decline of the brachypterous form; at the outer periphery (in the province of Gst, Vrm, Vgl, HIl, Ögl) the species is purely macropterous. Solitary records on Ska, Bornholm and Gotland are undoubtedly due to chance migration from the east. —The total area reveals still more clearly the eastern origin of the species. In central Europe the species has been found only along the German Baltic Sea coast, westward to Rugen*. Especially interesting is the highly broken up area in north Fennoscandia. Here the brachypterous form predominates to such an extent that macropte- rous specimens were found only at the periphery, partly in the west near 35 Tromsdal (2 macropterous and 14 brachypterous specimens), and partly in the south at the northern end of the Gulf of Bothnia (2 macropterous and 7 brachypterous specimens). In the northeast, however, all 62 specimens studi- ed were found to be brachypterous. That the north Fennoscandian stock as a whole does not represent an offshoot of the more southern stock is evident already from the usual distribution map (in Part II); but the dimorphic map also shows that the northern range must be considered the older of the two. There is also an ecological difference between the two stocks. In the south it is a stenotopic lakeside species and occurs elsewhere, for example, along the seashore, only accidentally and solely in the macropterous form. On account of the instability of lakeside biotopes, selection there in general does not favor brachypterism. It is significant that in the Fennoscandian southern area of Bembidion transparens a marked preponderance of brachypterous species is found only around the large central Swedish lakes Malaren and Hjalmaren, *] was able to examine the wings of only a few specimens from the localities outside Fennoscandia: Pomerania, Köslin, 1 macropterous specimen, Wollin, 2 macropterous, 1 brachypte- rous specimen; eastern Prussia, Rauschen near Königsberg, 1 macropterous specimen; Poland, Pinsk, 1 brachypterous specimen; Siberia, Tobol, 3 macropterous specimens. 343 388 Fig. 43. Pterostichus minor. Distribution map of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). Arrows indicate presumed postglacial immigration routes. ehr N id Te TEN Learn x | Ze 390 Fig. 44. Bembidion nigricorne. Distribution map of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). 393 345 where the water level hardly varies (in part thanks to man-made regulation devices). In the north the species is partly and apparently primarily an inhabitant of the seashore, which in any case was established in Lp Petsamo (Häk, Lindberg, 1933, p. 118) and in the southern part of the Kola Peninsula (Poppius, 1905, p. 91). With this mode of life, selection favors brachypterism (see p. 361), which has resulted in almost complete elimination of the macropterous form. In the inland of the north, dispersal of the species has thereby been greatly impeded. Especially watersheds have provided hurdles which to cross “is al- most as difficult for a flightless riparian animal as it is for a fish”! (Lindroth, 1939, p. 262). The watersheds are thereby marked on the map. —In the cited work, where I briefly dealt with the species and supplied a map, I was very uncertain about the connection between the small subareas of the north. Later on, the species was found in considerable numbers near Ks Salla, close to the watershed between the Gulf of Bothnia and the White Sea. It seems thus very probable that the species reached the northern end of the Gulf of Bothnia from this region (and not from Norway) all the more as among the species considered above (and below) we repeatedly find examples of immigration by the same route, which was not clear to me at that time. The Norwegian subarea, where the brachypterous form is still more promi- nent despite the most difficult possibilities for immigration, may actually be as isolated as the map suggests, and must be very old. It would be especially interesting if the somewhat doubtful record from 34 Lofoten (see Part I) could be confirmed. The most important question is whether selection during the postglacial period can explain the preponderance of the brachypterous form in the north. Unfortunately we cannot attach too much importance to the fact that the northern Fennoscandian area appears completely isolated on the map even toward the east, offering no clue to the postglacial origin of the northern stock as a whole, —because northern Russia has been very poorly explored. On the other hand we can establish, that selection cannot operate more in favor of brachypterism in the north than in the south, as soon as the animal lives on the shores of fresh-water. But, if it is assumed that the northern stock originates entirely from seashore populations the situation is understandable. We must then go far back into the period when the species was pushed to the seashore—to the “Ice Age.” The problem is summed up below (p. 412). The central Swedish stock of B. transparens poses another problem of general relevance. The species is missing in all the older collections from this region. The earliest record dates to 1910 (Upl Uppsala), and presently this species is a characteristic animal of the eutropic Phragmites shores of the region! But if it actually immigrated so late, how has selection operated t (Original German quotation translated into English; suppl. scient. edit.). 391 Fig. 45. Bembidion transparens. Distribution map of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). Crosses: No material available. Arrows indicate presumed postglacial immigration routes. In north Finland main watersheds are shown. 394 395 347 so rapidly in favor of brachypterism and how did the species arrive at all in its brachypterous form from the east, where no postglacial land connection spanned the Baltic Sea? Is it not more probable that Sweden was originally colonized by a pure (homozygous) macropterous stock, so that we have here a case of recurrent mutation, which originated on Swedish soil? Is this not supported by the isolated occurrence of 2 brachypterous specimens (among 2 macropterous ones) on Lake Hornborgasjön (Vgl)? The problem cannot be solved without extensive experiments over many years. However, as a further example which indicates the same possibility, reference may be made to Bembidion assimile (Fig. 46). This species, primarily a seashore inhabitant which occurs in the west and on the large central Swedish lakes (excluding Vattern) predominantly in the brachypterous form, is mostly macropterous in the southeast. This is particularly the case on Gotland, where among 63 specimens only one brachypterous specimen was found. Perhaps here, too, a recurrent mutation is in hand? For it seems improbable that selection along the seashores of Gotland would have operated differently from that along the west Scandinavian shores, on condition that the species arrived on Gotland right from the start in both forms. Hypothetically one might be inclined to examine these difficulties in the case of Bembidion transparens and B. assimile (i.e. the isolated and rare occur- rence of brachypterous specimens) on the assumption that brachypterism here is due to a recessive mutant. But the distribution in central Sweden of the two wing-dimorphic forms of B. transparens (and still more of B. aeneum, Fig. 49, in which the rudimentariness of wings is similar) can be explained only under the presupposition that the macropterous forms are homozygotes. However, there is a third possible explanation for the isolated and al- most sporadic occurrence of brachypterous individuals in an otherwise purely macropterous population: that the brachypterous specimens are the progeny of an immugrant macropterous female that had previously paired with a brachypte- rous male. The question as to the extent to which fertilized females of carabids can actually fly is specially dealt with below (p. 595). Here it may be pointed Out that the apparently purely macropterous populations at the periphery of the area of Calathus mollis or C. erratus clearly prove that such an event at least is not common. These species are pronouncedly xerophilous. —On the other hand it is perhaps no chance that the two species involved (Bembidion transparens, B. assimile), which might support the above assumption, are both ripicolous. Although flight after pairing is certainly unusual, one might ima- gine that species with this mode of life might be compelled to undertake an abnormal late flight (after pairing) if a small body of water at which edge they were living dried up (cf. Bembidion doris, p. 582). In the case of Pterostichus strenuus (Fig. 47) three immigration groups are evident. In south Sweden there is a large preponderance of the brachypterous form, but farther north as well as in Finland the macropterous form gradually 397 Fig. 46. Bembidion assimile. Distribution of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). increases in number. We have here two stocks which have immigrated during the postglacial period from the south. Their northern limit in Scandinavia is marked on the map by a broken line. It is not so clear in Finland, because here we have the same remarkable phenomenon as in the case of P. minor and Bembidion guttula (cf. also Bembidion transparens and B. nigricorne, as well as Pterostichus lepidus and Carabus clathratus): a marked, so to speak sudden increase in the number of brachypterous specimens at the northern end of the Gulf of Bothnia. The “brachypterous zone” in the case of P. strenuus is all the more evident because immediately to the south, in the inland regions of central Finland and east Karelia, there is marked preponderance of the macropterous form (between latitudes 62° and 65° N, 42 macropterous as against 31 brachypterous specimens), found nowhere else in Fennoscandia. 349 396 Fig. 47. Pterostichus strenuus. Distribution of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). Presumed limits between three different stocks are shown with broken line. 399 350 This is another example of a separate stock that immigrated early from areas of the White Sea, which has also reached the Swedish province of Nbt. Cha- racteristically, much as in the case of Bembidion guttula and B. transparens, only in this last area colonized the recessive macropterous gene manifests. The postglacial immigration route from the east across Kuusamo and Salla has been of great importance. Still clearer are the conditions in Norway where, with the exception of a small region in the southeast, only the brachypterous form has been found (al- together 159 specimens examined). It is certainly impossible to claim that the macropterous gene has been completely eliminated here. But since it cannot be presumed that selection at least in the inland Norway is presently operating in a different way than in the corresponding parts of Sweden or Finland, it must be concluded that the Norwegian stock of P. strenuus, at least in the west and north, did not immigrate postglacially from Sweden and that it, on the other hand, is very old in order that the elimination of the macropterous form has progressed to such a degree, whereby the dispersal capacity of the animal was greatly reduced as well. This brings us to the most important result of the present study of dimor- phism, the history of the west Scandinavian faunal element. This problem will now be taken up with the help of further examples. The map of Bradycellus collaris (Fig. 48) is not quantitative. Each locality is marked only to indicate whether brachypterous (black), macropterous (blank), or both forms were found. The reason is partly that from certain localities (in connection with crossing experiments) so much material was available to me (for example, 268 specimens from Upl Djursholm) that the usual depiction by circles was not practicable and partly that the macropterous specimens are mostly so rare that their distribution can be expressed clearly only by the method used here. In eastern Fennoscandia both forms occur side-by-side almost everywhere. The macropterous form was found even on the southern side of the Kola Peninsula. Hence nothing contradicts a uniform postglacial immigration to this part of the region. In Scandinavia the macropterous specimens decrease in number toward the north and markedly toward the west. In Norway the conditions remind very much those of Prerostichus strenuus, with the important restriction that the macropterous form was not so thoroughly eliminated by selection. Still, only 4 macropterous specimens among 114 brachypterous specimens were found there, excluding the southeast (as in the case of Prerostichus strenuus). It is moreover interesting that this predominantly brachypterous western stock has also spread to the Swedish fjeld regions, where so far only brachypterous specimens have been found from northern Dir to Tol. The map of Bembidion aeneum (Fig. 49) is schematically fairly clear. The usual distribution map (in Part II) already shows the strong geographical iso- 351 23 y cy RN ie oe US f ies Ui Ki N SE N SG \ Uo, A) : Ä | ” ¢ LY r o meee 2%) ‘ne % 398 Fig. 48. Bradycellus collaris. Distribution map (non-quantitative) of wing-dimorphic forms. Blank circles— Only macropterous specimens; Black circles—Only brachypterous specimens; Divided circles—Both forms. 52 ’ i 1 ae jae ee ne ee 400 Fig. 49. Bembidion aeneum. Distribution map of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). Cross indicates Up! Uppsala, from which no material was available. Arrows indicate presumed postgiacial migration routes. 401 353 lation of the western Norwegian subarea, which must have its own history. The dimorphic map clarifies the recent postglacial history throughout the region. In the south, in spite of its wide distribution, the species occurs as an original inhabitant of the seashore. Here, from southwestern Skä to the Oslo region, both forms live side-by-side in nearly equal numbers (specimens studi- ed: 78 macropterous, 70 brachypterous). Only in Skäne, near Lund on the bank of the River Höje-a and near Klippan on the Ronne-a, were brachypte- rous specimens found at some distance (respectively 8 and 24 km) from the sea, in addition to one intermediate (flightless) specimen on the south shore of Lake Bullaren in Boh (15 km from the sea). All other Scandinavian inland records, as well as those along the coast in the southeast (Ble, Sma, Öld), represent macropterous specimens. They are the result of a late immigration from the southwest and the sifting out of flying animals from the mixed popu- lations of the seashores. This route, obliquely across central Sweden through Vgl to the Mälar region, has been used by many other southern immigrants with flight capacity, especially the species associated with loam (p. 717). It may appear strange, that unlike B. assimile—the dimorphic map (Fig. 46) is otherwise similar—B. aeneum strangely failed to establish “relict stocks” of the brachypterous form on the shores of the large central Swedish lakes. A possible explanation is that during the Yoldia period the southern aeneum stock had not yet extended so far north. In the strangely isolated area of western Norway only brachypterous spe- cimens (61 of them were examined) were found. Naturally we cannot decide whether the macropterous gene has completely disappeared there, but in the context of dispersal biology the species is evidently to be considered brachypterous here. Since it is moreover a stenotopic seashore inhabitant it is not surprising that it has remained static. We have here the oldest Scandi- navian stock of Bembidion aeneum, which cannot have immigrated during the postglacial period. However, the strangest record of Bembidion aeneum is from Ks Paanajarvi lake, where 2 specimens were discovered in different years. Both are brachypte- rous; there is no question of accidental occurrence. In the preliminary treat- ment of this species (Lindroth, 1939, pp. 262-263), I assumed immigration from the east. When later on the great importance of this immigration route via Salla and Kuusamo for a whole series of species became clear (See Preros- tichus strenuus, and summarizing discussion on p. 724), this assumption was greatly strengthened. Probably the species still lives on the shores of the White sea, which have unfortunately been very inadequately explored, especially in the south. Before looking at the dimorphic map of Bembidion grapei (Fig. 50) it is advisable to study the usual distribution map (in Part II) more closely. What conclusions could have been drawn concerning the immigration of this species on the basis of this map alone? It is a picture of a widely distributed species 354 402 Fig. 50. Bembidion grapei. Distribution map of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). Limits of presumed Würm refuges in Norway (after Nord- hagen, 1933, 1935) are indicated. 403 355 of the high boreal coniferous forest area, somewhat of the type of Pterostichus adstrictus or of Pelophila borealis. The apparently continuous distribution over the whole of Finland until the southeast of Karelia suggests a firm connec- tion eastward through Russia to Siberia, thus a component of the “Siberian woodland fauna” immigrated during the postglacial period. Only the more or less complete absence from the Kola Peninsula may appear strange, as also, particularly, the isolated localities of the south Norwegian records. But the dimorphic map gives a completely different, so to speak reverse picture. In the whole of eastern Fennoscandia, with the exception of the far north (north of latitude 67° N) only the macropterous form (53 specimens) has been found, and the same in Sweden south of latitude 64° N (40 specimens). However, toward the north and especially the west, that is in Norway, the brachypterous form gradually increases in number. Symptomatic was extensive material (80 specimens) from 38 Alta, which comprised 62 brachypterous and 18 macropterous specimens. Is it now possible, if only from a purely theoretical viewpoint, to regard Bembidion grapei as a postglacial immigrant of Fennoscandia, in view of this distribution of short-winged and long-winged individuals of the species? Like the cat around the stew pot, we have been going round this question so far. Now we must seriously take it in hand. The concentration of the brachypterous form of B. grapei in north and west Fennoscandia could give rise to any of the following attempted explanations (cf. a similar formulation of the question for Calathus mollis, p. 370), provided that the character is genetically determined. 1. The species immigrated to Fennoscandia exclusively (or predominantly) in macropterous form during the postglacial period, and the brachypterous form in the west and north arose thereafter (or enriched): a. by recurrent mutations; b. by a particular influence of the selection; —or by a combination of both factors. 2. The regions with predominantly or exclusively brachypterous popula- tions were colonized earliest. These are Ice Age refuges, from which a post- glacial emigration took place into the regions once covered with ice (hence predominantly to the east). la. Recurrent mutations have been experimentally established, especially in Drosophila; of the known wing-reducing mutation “vestigial” at least 29 alleles are known (thus phenotypicaliy not or hardly separable), almost all in the same locus (chromosome 2). Earlier (p. 393) we hypotheticaliy set up the possibility of recurrent mutations in Bembidion transparens and B. assimile. However, who wishes to explain the occurrence of brachypterous B. grapei only in certain areas of Fennoscandia by mutations “in situ”, must ascribe the definitive role to environmental factors in west Scandinavia, not only for the species in question but for all the other species here considered that attained 404 356 maximum brachypterism in these regions. One would have to maintain that the west Norwegian climate (since other environmental factors do not come into question) would have caused a similar (Or corresponding) mutation in an entire series of species. But this totally contradicts our present experience, achieved experimentally and theoretically, on the nature of mutations. —More important is the finding that by mutation alone, however high its frequency owned is, suppression of the “normal type” cannot result. The more or less complete preponderance of the brachypterous form in a subarea of the species can be explained only by selection (whether this has taken place within the limits of the area or outside it before the immigration). 1b. Some particular effect of selection in west Scandinavia deserves serious consideration. One might thus surmise that B. grapei immigrated during the postglacial period (for instance, from the east along the northern margin of the shrinking ice, as soon as the coastal areas became ice-free) in macropterous form (possibly, but improbably, with a slight admixture of the brachypterous form). After the origin of the brachypterous gene by mutation, selection in the west should have operated strongly in favor of brachypterism. One might remember in this case the exposed location of the largely unforested coastal areas of Norway and the high frequency of atmospheric minima with the at- tendant, often strong winds. In the case of B. grapei we have first of all to notice that the species was found within the region not above the timber line (see Part I) and does not live at all in very exposed places, but mostly in association with forest and undergrowth. And if we transfer this observation on species like Bradycellus collaris and particularly Notiophilus aquaticus (see below), we come across the unexplained phenomenon that the brachypterous form on the Swedish side as well in the fjelds as in the valleys, shows an equally strong preponderance. And the climatic conditions here are totally different. Large continuous sub- areas of a species, inhabited predominantly or (apparently) exclusively by the brachypterous form, cannot have arisen solely by selection operating at present. For none of the 3 species mentioned here (even the little Prerostichus strenuus), ecologically considered, has such a stenotopy that flight capacity could have dangerous consequences (for Bembidion aeneum, which is associated with the seashore in west Norway, it might). And a positive selection value must be ascribed to the brachypterous form (more correctly: denied to the macropte- rous form), if the short postglacial period has sufficed to allow the origin of a mutation (see calculations on p. 365) and an increase up to preponderance (in the case of Prerostichus strenuus, Bembidion aeneum, Notiophilus aquaticus apparently up to 100% strength). We must also refer to Calathus erratus and Pterostichus vernalis, in which the Macropterous form is predominant in west Norway. Together they show that the climate in these parts does not favor “a priori” the brachypterous form of a dimorphic species. 405 408 337 2. The preceding account has probably shown that ihe geographical distri- bution of the two wing-dimorphic forms of Bembidion grapei (besides various other species) cannot be explained by the currently prevailing conditions of climate, etc. We must go back to the last glaciation (Würm). Accordingly the map shows the Würm refuges presumed particularly by Nordhagen (1933, 1935) (but also cf. map in Fig. 111, p. 775). Supposing that B. grapei “win- tered” in some (or all) of the ice-free coastal regions, the map is readily un- derstandable without any hypothetical construction. The rule first formulated for Calathus mollis is applicable: At the periphery there are macropterous specimens and a preponderance of the brachypterous form is to be found in the regions colonized early. Finally the question remains whether the entire Fennoscandian area of B. grapei has arisen by emigration from the Norwegian coastal refuges. Inas- much as to the east, south of the White Sea, there is probably a continuous distribution across Russia (see Part I), the answer is not forthcoming. Now Poppius (1911, p. 36) published an early postglacial subfossil record (“Dryas zone”) of this species from Finland Ik. Unfortunately it was not possible to examine the elytron that he studied (probably in MH), and the determination is certainly not definite. However, the find would not be surprising, since it appears most probable, that from the start south Finland received its grapei stock from the east or south during the postglacial period. Otherwise the dis- persal of the glacial stock proceeding south would have been considerably more effective in Finland than in Scandinavia, where, especially in south Nor- way, the degree of colonization is perceptibly low. This last condition will be explained if future study of more numerous material from Norway south of latitude 62° N actually confirms the total absence of the macropterous form here. The last two dimorphic maps show how even in the case of more or less universally distributed species, whose geographical distribution in itself provides no clues, the mapping of wing-dimorphic forms may be “revealing.” The species in question both belong to the genus Notiophilus. Since the publication of Part I many new records of Notiophilus bigutta- tus have come in, and a couple of records noted there have turned out to be uncertain (see Supplement to this part). It seemed best to give a revised “usual” distribution map (Fig. 51) to provide an instructive comparison with the dimorphic map (Fig. 52). Notiophilus biguttatus is almost universally distributed in Fennoscandia. As a pronounced forest species, which occurs very sporadically and possibly accidentally in the Regio alpina, it is missing from the high mountains and the tundra. More noteworthy is the gap in north Finland, which also extends into the wooded regions of the Kola Peninsula. Such a map gives no idea of the immigration routes. The dimorphic map does. The distribution of the two forms becomes clear from a rough compari- 358 son between the latitudinal zones of the three countries (excluding the Baltic Sea islands): Index = Macropterous _ Brachypterous (macropterous/ specimens brachypterous) Finland South of latitude 62° N 69 21 2.55 Between latitude 62° N and 64° N 32 12 2.66 Between latitude 64° N and 66° N 12 4 3.0 Between latitude 66° N and 68° N 4 1 4.0 Sweden South of latitude 58° N 66 22 3.0 Between latitude 58° N and 60° N 52 26 2.0 Between latitude 60° N and 62° N 41 59 0.7 Between latitude 62° N and 64° N 14 32 0.44 Between latitude 64° N and 66° N 15 31 0.48 North of latitude 66° N 15 5 3.0 Norway South of latitude 60° N 26 46 0.57 Between latitude 60° N and 62° N 3 26 0.12 North of latitude 62°N 4 105 0.04 In Finland the distribution is easily understandable. To the north the brachypterous form gradually declines in number. The map very much gives the impression of a stock that immigrated during the postglacial period from the south. The only exceptions are the two localities in Petsamo, which have connection with the Norwegian area. In Sweden the conditions south of latitude 60° N and in the far north (beginning from latitude 66° N) largely correspond with those in Finland. But in the rest of Sweden the brachypterous form preponderates, with increasing number to the west. These are the parts connected with Norway by wooded regions or, in the north, by numerous wooded passes. 359 406 Fig. 51. Notiophilus biguttatus. Completed and corrected distribution map (cf. Part II of this book). 360 mk 407 Fig. 52. Notiophilus biguttatus. Distribution map of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). Minimum limit of postglacial stock westward and northward is indicated by a broken line. Arrows indicate migration routes of the Wurm “overwintering” element. 409 411 361 Finally, in Norway the macropterous specimens play a significant role only in the extreme south, chiefly in the coastal regions. In all other parts they occur sporadically or are apparently totally missing in large regions. This colossal west Scandinavian stock, in which the macropterous indi- viduals have been nearly eliminated, has undoubtedly arisen through Wurm hibernation. This species is especially interesting because the above-mentioned old stock has met with the postglacial immigrated stock on a broad front. Study of more material—more than was available to me—would certainly establish such origins of the individual populations in this mixed zone. —I had assumed that the macropterous specimens, wherever they were still found, must have had a southern, postglacial origin, in other words that the interglacial stock had become “purely” brachypterous by selection (and finally by accident; see p- 365). A later study of material of the same species from Iceland, where the species was certainly a Wurm hibernator as well (Lindroth, 1931), showed, however, that there, too, the macropterous form had not been completely eliminated (four macropterous among 148 brachypterous specimens). So it is impossible to establish the outermost limits of the postglacial immigration in Scandinavia by means of the isolated, sporadically occurring macropterous specimens in Norway. Such a conclusion would be justified only if they oc- curred regularly. The broken line on the map thus is the presumed minimum limit of the postglacial stock to the west. However, the regions of Scandinavia where the brachypterous form shows a distinct preponderance (established by a study of sufficient material!) must have been colonized to a greater or lesser extent starting from Ice Age refuges. The arrows on the map show such emigrations to the Swedish region. - Notiophilus aquaticus (Fig. 53) is the most widely distributed of all Fennoscandian carabids. It is the only species that lives uninterruptedly from southernmost Skane to the high alpine zone of the fjelds and the tundra of the Kola Peninsula. Ecologically too it is very eurytopic, but it cannot be considered ubiquitous because it avoids wet places and densely wooded regions. Therefore one cannot imagine a distribution map that is less appropri- ate to speculation about immigration routes. The accumulation of points, especially in certain fjeld regions, is greatly explained by more thorough ex- ploration and cannot be considered as centers.” Much more instructive is the dimorphic map, which largely corresponds with that of N. biguttatus. We find the macropterous form distributed fairly uniformly over most of Finland, over Sweden (with the exception of the fjeld regions), and finally in southeastern Norway. In general, however, it occurs in a pronounced minority and is best represented in Swedish Norrland. The most important difference with N. biguttatus is that only brachypterous specimens have been found on the west coast between 6 Stavanger and 40 Varanger, as well as in a broad adjoining region which also touches Sweden from Hjd 412 362 onward, and in the Kola Peninsula (with the exception of the southwest). This is the “interglacial stock,” from which the macropterous form seems to have been completely eliminated. In N. biguttatus we found that the interglacial stock had not completely lost the macropterous gene, that this even manifested itself phenotypically in 4-5% of the individuals as a result of crossing between heterozygotes. In N. aquaticus the selection has proceeded farther, so that among 400 specimens from the “purely” brachypterous region in west and north Fennoscandia, no macropterous specimen was found. The reason may be on the one hand, that the “initial stock” of N. aquaticus was brachypterous to a greater extent than in N. biguttatus, which is indicated by the condition of the postglacial stocks of the two species (perhaps by the higher probability of emigration of the brachypterous form of N. aquaticus to the refuges at the beginning of the Wurm period, given the pronounced resistance of this species to adverse climate). On the other hand it may be, that selection in the case of N. aquaticus favored the elimination of macropterous specimens, since this species lives in rather exposed places, whereas N. biguttatus lives mainly in localities protected by woodland or bushes. Evidently we can never prove that the brachypterous interglacial stock of N. aquaticus is completely homozygously brachypterous. But since macropte- rous specimens in this region occur at most completely isolated, it is at any rate justified in taking the broken line on the map (Fig. 53) as the maximum limit attained by the stock that immigrated in the postglacial period. This can be fixed with greater certainty than the corresponding minimum limit of N. biguttatus (Fig. 52). But it is not possible to follow the sporadic advances of the interglacial stock to the east as in the case of N. biguttatus, on account of the preponderance of brachypterous specimens even in the postglacial stock of N. aquaticus. What remains to be done is to examine more closely the Ice Age condi- tions, which for many species caused selection in favor of brachypterism, and furthermore to answer the question why dimorphic species generally seem to be favored in an area glaciated during the Quaternary period (cf. p. 364). The Fennoscandian Wurm refuges, which are treated in greater detail else- where (pp. 752 ff.), are now generally believed to have been ice-free coastal regions, situated—in any case predominantly—in western and northern Nor- way. The peaks in the mountains (“nunataks”), completely surrounded by ice, biologically had at the most a very subordinate role. The coastal refuges were comparatively small, surrounded on all sides by obstacles (from the viewpoint of terrestrial biology), namely ice and sea. They were like isolated islands or mountain peaks, encircled by uninhabitable areas, and the fauna, like that of the above-mentioned areas, was subject to a selection that favored flightless individuals, forms and species. As soon as an insect in these places took off into the air there was the danger of its arriving, actively or passively, on the ice or in the sea, resulting in death by freezing or drowning. The latter eventuality 363 410 Fig. 53. Notiophilus aquaticus. Distribution map of wing-dimorphic forms. For explanation see p. 367 and Fig. 28 (p. 368). The broken line indicates the maximum limit ofthe southern postglacial stock toward the west and north. 413 364 was probably equally frequent, since it must not be forgotten that the winds around the high pressure regions above inland ice chiefly move centrifugally, which, with the large differences in altitude in western Norway, resulted in fohn winds (fall winds). These may have favored the fauna and flora of the refuges climatically, but especially temporary higher temperatures have mis- lead flying insects to swarm flights, as a result of which they got in danger of being carried across the sea. For this reason I believe that selection in the Ice Age refuges could have caused an elimination of flying insects more than in most of the islands or in the mountains. It was stated earlier (p. 338) that the Fennoscandian carabid fauna show a strikingly high percentage of dimorphic species. A closer examination will reveal that dimorphism has been of direct advantage in regions that were repeatedly glaciated during the Quaternary period (see p. 364; cf. also Dar- lington, 1943, p. 44). During periods of stronger alterations (chiefly of climatic nature) the winged species and winged forms of dimorphic species were clearly favored. Such periods were primarily phases of increase or of melting away of the various glaciations. At the end of each interglacial period the upper existence limits of each animal species were gradually pushed downward and to the south, the biotopes were strongly altered by the fall in temperature, by edaphic alterations (partly catastrophic in nature, such as the summer-like glacial rivers) and impoverishment of species, and finally became uninhabit- able for the species concerned. The faster these alterations occurred, the more difficult it became for other than flying individuals to find a suitable new area: selection operated in favor of macropterism. Within the limits of Fennoscandia only such populations escaped destruc- tion which managed to reach regions by a combination of dispersal capacity and chance, that remained ice-free during the maximum of the ensuing glacia- , tion. One might ask if under such circumstances the refuges of the Atlantic coast were reached more frequently only by the winged form of dimorphic species—whether therefore the dimorphism of the species would not have been lost as a result of this escape, and whether brachypterism thereupon must have arisen anew by recurrent mutation within the confines of each refuge. —It is correct that the brachypterous specimens normally must have reached the refuges on foot (“per pedes”) since they cannot arise from the homozygous macropterous form without mutation. But during this emigration (which, being chiefly directed downwards, would be supported by passive modes of dispersal, such as running water) the possibility existed of flying individuals having had a direct positive role in the dispersal of the brachypterous form, by reproducing with stray brachypterous specimens in more or less chance unions due to searching for the other sex. From every such pairing a progeny of 50 to 100% brachypterous specimens would have arisen (depending on whether the short-winged or both parents were heterozygous or homozygous). To the extent that fertilized females fly (a question dealt with elsewhere, on pp. 395 414 365 and 595), it is possible that a female that has paired with a brachypterous male spreads brachypterism so to speak by air. In western Scandinavia, with its large altitude differences and consequently with different zones of climate and vegetation closely following one another, the route between the interglacial “place of residence” of the species and the glacial refuge need not have covered many miles. Conversely, we find that even alpine plants (such as Papaver, Nordhagen, 1931, 1933) now live in close proximity to the presumed refuges. During the glacial periods, lasting thousands of years, conditions were probably to some extent stable again and selection, as we discussed above, operated in the reverse direction—in favor of brachypterism—to the extent that the hibernating stocks of species, such as Bembidion aeneum (p. 399), Pterostichus strenuus (p. 395), Notiophilus aquaticus (p. 409) apparently (or actually?) became “purely” brachypterous. The alterations at the end of each glaciation may not have been so di- sastrous as during the growth phase, but the improvement in the climate and the melting of ice sometimes proceeded (for example at the end of the Wurm period) very rapidly, so that especially the fauna of open terrain would be threatened to a great extent by forest (See also p. 710). Besides, there were slower but more comprehensive alterations in the distribution of land and water. Hence during this period there were enormous alterations in the biotope which again favored flying animals. This was true, for instance, of the Yoldia period in central Sweden and Finland, and of the Littorina period (and later) in the southern Baltic Sea area, when the problem of overcoming the aquatic barriers arose. During the more stable middle eras of the interglacial epoch the brachypte- rous individuals were favored by selection, but to a very variable extent, depen- ding on the ecological stenotopy of the species and most strongly (see p. 364) in the case where flying individuals could not easily locate habitable surfaces, and least of all in more or less ecologically ubiquitous species, such as Bembidion lampros, Notiophilus aquaticus and N. biguttatus. The importance of wing dimorphism for the animals of a repeatedly glaciated area can be summarized as follows: The presence of winged indi- viduals at the beginning and end of a glaciation and of wingless individuals during its maximum, reduces the danger of extinction. By this, the occurrence of dimorphic forms was generally favored in parts of the earth that were glaciated during the Quaternary period. A particular problem is whether the present geographical distribution of the two forms of one or other dimorphic carabid species allows drawing any conclusions as to the locality of the Wurm refuges. As is well known, such conclusions were drawn by Nordhagen (1933, 1935), generally on phytogeo- graphical grounds, and, following his idea, the presumed refuges are shown on the map of Bembidion grapei (Fig. 50). 415 416 366 Most of the dimorphic species that must be considered as west Scandina- vian “winterers”, in the west, however, have a continuous, widely distributed brachypterous stock, which provides no further clues in this connection. The following exceptions are noteworthy: Bembidion grapei, and possibly also Calathus erratus indicate a refugium in the Stavanger region. Bembidion grapei and B. aeneum indicate a refugium in Möre (Province 9). Bembidion grapei and B. transparens indicate a refugium in the Lofoten region (33-35). | Bembidion transparens and possibly also B. grapei indicate a refugium in Varanger (40-41) or Petsamo. From an Ice Age refuge must also have originated those dimorphic (and other) species which come from the region around the White Sea by separate immigration, such as Bembidion aeneum, B. guttula, B. transparens, Pterostichus minor, and P. strenuus, and possibly also Bembidion nigricorne, Carabus clathra- tus, and Pterostichus lepidus. For the high brachypterism of these stocks must also be due to selection within an isolated region, and the separation from this stock, which (with the exception of Bembidion aeneum) colonized the more southerly parts of Finland, would be incomprehensible if they had a common glacial origin at the southeastern edge of the large mass of inland ice. The question of the glacial overwintering of the fauna is further discussed below in a special section (p. 735 ff.). Table 29. Distribution of 9 “Würm winterers” among dimorphic carabids within the vegetation zones of the fjelds Regio alpina Regio betulina Regio coniferina Higher Middle Lower Bembidion aeneum - — = = ae B. grapei = — +" + + B. guttula — _ — ae fu B. transparens _ _ = on a Bradyceilus collaris _ _ (+) an ai Notiophilus aquaticus + + + st ap N. biguttatus _ - (+) a N Pterostichus minor _ = _ = 2 P. strenuus — _ = = ye * Outside the region (Siberia, Alaska, Greenland). 416 367 The most important result shown by this analysis of the dimorphic carabids of Fennoscandia is that at least 7 species must have survived the last glaciation within the confines of the region, in addition to 2 species (Bembidion guttula, Pterostichus minor) probably along the eastern edge, in the vicinity of the White Sea. An ecological study of these species may provide an understanding of the climatic and other conditions in the refuges during the Wurm period. It will be interesting at this stage to give the present distribution pattern of these 9 species in the usual vegetation zones (“regions”) of the Fennoscandian fjelds (Table 29). This is an excerpt from the larger Table 30 (p. 440). We conclude that apparently only Notiophilus aquaticus can servive under high arctic conditions. In the Regio alpina only 4 of the 9 species have been recorded. The occurrence of two of them is more or less accidental and a third does not occur within the Fennoscandian region. Four species have not even been recorded even once in the Regio betulina. —These are facts which cannot be reconciled with the traditional concept of Ice Age conditions. Synthetic Part The Definition of Area 417 The concept of “area” (Arealt; region of distribution) of an animal or plant species is fictitious. What is involved is not a terrain occupied, but a sum of points—individuals— which in the case of animals possess a more or less pronounced totality of mobility. Furthermore, this totality of points cannot be added up all at the same time. The fiction is therefore predicated on space, and also by time. The zoogeographical term “area”, which is nevertheless very useful, is best defined by its limits. These do not delineate more or less homogeneous areas, such as the forest limit of the fjelds; these limits are homologous with the treeline. An area limit is therefore a line drawn to join up the outermost (northernmost, highest, etc.) localities of an animal or plant species. It would of course be preferable if instead of “localities” we could fall back on popula- tions, definite occurrences, or the like. But as far as insects are concerned, in practice it is possible to decide only in exceptional cases whether the animal found at the sampling region has appeared accidentally or is actually native to that place. The area limits are therefore easily drawn too far out. The more demanding a species is, the more stenotopic it is, the more dis- junct is its area (see examples on p. 563), since most terrains are uninhabitable for it. A hypothetical example of how one can visualize the actual and mo- mentary distribution of a stenotopic insect species within a limited part of its periphery, is provided by the accompanying map (Fig. 54). It may be noted that to the north the species in question selects only the most favorably situated (sun exposed) places, and that individuals that have strayed—including those found in biotopes alien to the species—easily misrepresent the true distribu- tion. What then are the “area” and “area limits” in the present case? —To me it appears that the answer will vary, in keeping with the task set by the researchers. For the pure ecologist the area of the species can include only the 419 actually, constantly inhabited terrains (marked black in the map). A continu- t(Suppl. scient. edit.). 420 369 ous area limit is of little actuality for ihe ecologist. On the other hand, for the zoogeographer, whose chief interest is historical, and who wishes for instance to interpret the postglacial processes of immigration, the above broader (how- ever less exact) use of the term is more practicable, not least since he mostly works with much more extensive regions. He is satisfied with the following definition: The area limit is a line, up to which the species concerned constantly occurs in suitable terrains. However, because of enemies, competition with other animals, chance, etc. no species occupies all the terrains suitable for it. On the model map (Fig. 54) the suggested zoogeographical area limit is marked with a crossed line. The area limits—in the above sense of the word—are not only highly unnatural lines, they are sometimes completely hypothetical. Highly isolated occurrences, be it in advanced outposts or relicts, should be left outside the continuous area limit. Wherever two separate immigrating stocks come close, it is often impossible to decide how to allocate solitary records of the in- tervening region (for instance Andersson and Birger, 1912, p. 335, Anemone nemorosa; Ekman, 1922, p. 201, Sterna paradisaea). Such considerations led me to decide not to draw any area limits in the maps given in Part II. How- ever, in the present part the terms “area” and “area limit” are always used in the broad, zoogeographical sense of the word. I believe even critics of these “constructs” do not mind speaking of the northern limit of an animal species. Finally, it may be mentioned that it seems not unjustified in working with the “area of a species” even where the species has been divided into subspecies, ecotypes, etc. One can even speak of areas of a genus, etc. The Reliability of Distribution Maps Occasionally, even being awake, I have a terrible dream. The real distribution image of one of my carabid species is presented to me! —However, (unfor- tunately or fortunately?) we never achieve such an image. There is no avoiding the most important question: How reliable actually are maps published in a book like this one? The answer is best obtained by considering the most important sources of error. These are: 1. The region in question has been inadequately or in any case erratically explored. Actually, not even a small terrain is being completely explored, not only because the researcher does not come across everything, but also because the composition of the insect population is subject to continuous alteration. That faunistic or floristic research can never end is all the more true of large regions such as Fennoscandia. Goethe’s words, cited by Horion (1941), apply here: “Such work is actually never complete; one has to pronounce it com- plete.” Now, in judging the stage of the map at which this “declaration” may be made, biogeographers apparently hold extremely divergent opinions. I have N er ae x Ders u ~ S40 at WY Ay K N wife NECN 67 N TEN aoe SSE S| ) +++ tX = 418 Fig. 54. Northern peripheral region of the hypothetical, momentary “area” of an insect species bound to shores and banks of a particular type. Black areas—Populations; Black circle—Single individual; Stippled—Suitable biotopes. The presumed isotherm is “macroclimatic,” ie. obtained by usual meteorological measurements. Crossed line—Zoogeographical “area limit”. 422 371 (1939, pp. 242-243) already formulated my criteria for the distribution map to be ripe (i.e. for publication): “...the map must be at least ‘credible.’ The evidence that this stage has been attained can be tested when any new record can be naturally fitted into the earlier picture without ‘destroying’ the integrity of the map. If a really surprising locality of a carabid species in Sweden [in the present context: Fennoscandia] has been added only two or three times in recent years, I con- sider this proof that my carabid maps have almost reached the stage of ‘credi- bility’.” Today, nine years later, almost every year other “surprising localities” are added, which cannot all be attributed to late immigration. Nevertheless, I have ventured “to declare the maps ready,” i.e. they have been published, not as final documents but as a basis for zoogeographical discussion. If the general or so to speak the average reliability of the maps is suf- ficient for this purpose, the often erratic exploration of the region must not be overlooked. Some gaps in the distribution can readily be considered artifi- cial. To these, for each species, attention was drawn in Part I. Some of them have subsequently been filled in with additional data provided in the Supple- ment. Assessment of the detailed situation will be facilitated by a map on the frequency of coleopterological exploration of the region (Fig. 55). In 1938, I published such a map for Sweden. As already mentioned, the decision as to a distribution map being ripe for publication varies greatly with different researchers. As an example of a “premature map” in the Fennoscandian region, mention may be made of the work of Somme (1937) on the zoogeography of Odonata. One is not justified in basing zoogeographical conclusions on distribution maps that are so incomplete. The situation is different when the author himself is aware of the imperfection of his maps and draws no biogeographical conclusions from them, but regards the maps as an impetus for further research. P. Palmgrén (1939, 1943) did this in the case of spiders of Finland and also provided a map on the frequency of exploration of the region (1939, p. 81). But personally I doubt if such maps are deserving publication, having been prepared only for “pedagogic” purposes. All too easily greater significance than is intended can be attributed to them by other researchers, especially foreigners, who have not been apprised of the correct situation and may thus misuse them. A map that is not only the clearest and simplest method of depicting the known distribution of an animal or plant species, it is also a wonderful puzzle of nature, which one wishes to solve at once. It is not very nice to present an “illusory puzzle.” Let us pass over those maps, especially of the Southern Hemisphere, that comprise solitary, widely separated points which have prompted assumptions about land connections in all directions across the oceans. And if one is not sufficiently acquainted with the distribution of the species, one can prepare maps of genera and families and draw smart conclusions. 2. The map does not reflect the present and unique state. The Fennoscandian SO Qe N fe Lf SWR N £ 1% 4 7 5 Nea / ¥ 2 / 3 fi ii a £ Ce, SOF at; NR Sp L 2: a A, G a Bi SUN U Eh FE) 7 ES a, ZH = 4 ER 494 X & U VY OR $a, Fig. 55. Relative exploration of Coleoptera in the Fennoscandian region. 421 Biack— Well explored; Squared—Moderately explored; Hatched—Poorly ex- plored; White—Not explored at all. Assessment, which is subjective, is based on records of Carabidae up to 1947. 423 425 313 carabid maps are the result of nearly 150 years of collecting. During this period the fauna has demonstrably altered in both negative and primarily positive directions, a problem to which a separate section is devoted below (pp. 621 ff.). Some especially clear examples of very recent immigrations are cartographically depicted. Theoretically it would always be more correct to prepare a separate map of the animal species concerned for each epoch (for instance, every decade). In practice that is not feasible, since the number of entomologists is not large enough to bring together material for a nearly complete map. Moreover, there is no date on the labels of any of the oider collections of Coleoptera or, (alas!) even of some of the collections still being built up. Finally, in Fennoscandia the faunistic changes of the last century were predominantly of positive nature: New species immigrated, some older ones expanded their area (especially northward and westward). Hence with remarkably few exceptions the maps provided here reflect the present areas of the species, of course in the most important features with regard to the area limits, if not in detail, since there have been vast changes in the biotopes, especially in the south. In our region the possibility of immigration of transgrading species from all possible directions is not as great as, for instance, in Germany. Hence Horion’s repeated exhortations (for example, 1941, p. 5) to note the yearly fluctuations in the faunal stock need be accepted only to a limited extent. 3. The maps do not record abundance and frequency. Yet these determine the focal point of the area, and when it lies in a relatively poorly explored region it may even show up as a patch of sparser points compared to better- explored peripheral parts where the animal is not as common. It would be especially interesting to find out whether the species gradually decreases in numbers toward the limit of its area or whether its abundance remains un- diminished right up to this boundary, which would help to answer the question whether there is an existence limit or a dynamic limit. Preparation of a map of the desired type is impracticable mostly because of the paucity of material. Frequency or abundance can hardly ever be determined on the basis of the collections, since very few entomologists make the relevant notes in the field. Miscodera arctica (Fig. 56) has been chosen as a simple example of what such a “quantitative” map might look like. If 1, 2, 3, or more specimens were collected on the same day, this has been indicated by four types of circle of different sizes. In spite of the numerous records the pronounced rarity of the species becomes evident in the south and along the coast, and it shows up much more clearly as an animal of the north and the fjelds. Similarly we can illustrate the southward tapering off of a high boreai species, such as Pterostichus adstrictus, or the decline of Amara similata or A. aenea toward the north. A few words on the best mode of marking a distribution map. I have selected the simplest method: each locality is represented by a point. This is also the most correct method. As soon as illustration by indication, by 424 Fig. 56. Miscodera arctica. “Quantitative” map. Largest circles indicate 4 or more specimens collected in one day, smallest only one specimen (or an unknown number). 426 375 hatching, etc. of large regions is used, precision suffers. Examples are the publications of Sainte-Claire Deville (1930a), Holdhaus and Lindroth (1939), etc. The two methods are combined by Sainte-Claire Deville (1930b), Somme (1937) and Hulten in his botanical studies (see below). The “surface method” seems to me justified in cases where a very large region, for instance a whole continent, is mapped. The exact siting of the points has a smaller role and a point-map would give a very unnatural impression, since without exception exploration of large areas has been more or less erratic. Unavoidably, the “surface method” glosses over actual gaps in the distribution, and such maps may easily prompt erroneous conclusions. For even the smallest lacunae of the area, as far as their existence can be actually determined, may be most significant for the history of immigration, many examples of which are provided in the present work. This is also true of plants. The method used by Hulten on a grand scale in his work which is being published on the Fennoscandian distribution of vascular plants, namely, the representation of the area largely by hatched surfaces, is of course understandable. Who would want to map Calluna vulgaris by the “point method”? But some important details of the distribution have been lost. In the case of insects, which never extend over large terrains as a continuous “stock,” and for which the observation material is still in no case insurveyable, the point method of mapping a region the size of Fennoscandia must be considered in principle as the only reliable method. But it seems to me important to emphasize that a// known localities must be marked on the map (as long as they are far enough apart on the particular scale). The procedure adopted by Somme (1937, p. 134: “Where the localities are very close, and not every single record is indicated”) is decidedly to be rejected, being an attempt to mask the uneven exploration of the region. An aberration of the point method, used mostly for economy in print- ing, is the representation of several species on the same map, using different signs. Striking examples are found in Breuning (1932-36), Heberdey (1939), and Malaise (1945, p. 51), where 8 to 20 (in Breuning up to 60!) species or subspecies are represented together on a single map. This method is under- standable where it is intended to illustrate the total distribution of a genus or a group of species. In all other cases it seems to me completely misguided. If I were given the choice of publishing 10 maps of the distribution of 20 species I would prefer to leave out half of them rather than to put 2 species on a common map (with the exception of cases where the two areas are mutually exclusive). The superiority of the picture provided by a map over a list of localities lies in the fact that just one glance gives an understanding of the relative continuity of the area and of the connection between its parts. And, this picture harbors the explanatory attempts. The “collective” maps lack these advantages. An original approach was adopted by Borchert (1938), who published the 427 376 European distribution of all the German beetles in the form of “model maps” of 93 species. Each of the remaining species was referred to one of these “models” in the text, with additions and deletions. Distribution data given in the text in this form are hard to follow, partly because of the abbrevia- tions; otherwise the data seem to have been very carefully gathered from the literature. This brings us back to the main basic question: Are the distribution maps of the Fennoscandian beetles true to nature to the extent that they could serve as the basis for determining the area-limiting factors and the postglacial history of immigration of these insects? Compared with other distribution maps of in- sects they are good, but it would be apposite to pose the question: Is it possible for the insect collector—or insect collectors over an entire century—to come across such a large portion of the insects occurring in nature in a region, a province, or an entire country, etc. that the species composition can be judged adequate for a fairly true-to-nature mapping? More simply stated: How many of the actually existing species does the collector find? Eigin Suenson (1934) made an interesting calculation on the number of species of carabids that he collected in Denmark on various excursions. Sum- marizing the results of his collecting trips, which represent 112 days of collect- ing (i.e. fairly corresponding to an intensively utilized summer), he recorded a total 191 species, which is 65% of the Danish fauna. My experience in the summer of 1936 was similar, when I undertook a trip to supplement the carabidological survey of the most poorly studied parts of Sweden (map in Fig. 57). By coincidence, the result, 211 species in 73 days of collecting, constitutes exactly the same component of the Swedish fauna, 65%, as Suenson obtained for Denmark. It may be mentioned that the trip was not undertaken to collect as many species as possible but to complete the distribution map of some particular species, and the number could have been easily increased by short visit to Öland-Gotland or to the fjelds. Within smaller areas comparisons can be still more advantageous. Suenson (l.c.), in only 9 days of collecting on Bornholm, found 81 carabids (= 36%)* of its fauna. On Gotland in the spring of 1940, in a total of barely 15 days I collected 134 species, constituting no less than 69% of the carabid fauna of this island. The best results of collecting I have ever had for one day were in the environs of the city of Mariestad (Vgl) on June 10, 1936. In a region of at the most 5 km’, along the shore of Lake Vanern, 79 carabid species were collected, i.e. 24% of the entire Swedish fauna and 37% of that of the province of Vgi (214 species). From the above figures it is clear that an experienced collector who is acquainted with the ecology of every species and restricts himself to a moder- ately large group of animals can obtain a major part of the fauna in a short *Perhaps still more, since it seems that he did not take along some of the commonest species. 428 429 430 3 period of time. The conditions with regard to carabids are especially favorable, since they have a relatively exposed mode of life and a relatively long adult lifespan, so that, with few exceptions, they can be collected during the long period from May through September. As rightly noted by Suenson (l.c., p. 476) it is the “rare,” often biogeo- graphically interesting species (relicts, transgrading species, etc.), that escape attention. But experience shows that the word “rare” can often be replaced by “stenotopie.” Once the correct locality of the species has been discovered the “rarities” often reveal an astonishing abundance. An example is Trechus rivularis. Distribution types which are distinctly divided in subareas are fre- quently explicable just because of their restriction to special biotopes. This is especially evident in the shore beetles, which depend on a definite grain size of shore material (for example, Bembidion), in burrowing sand beetles (Harpalus, etc.), peatbog insects, such as Agonum ericeti, etc. (see p. 563). Of course it may be gladly conceded that some species at the periphery of their area or in their relict areas are really rare. Miscodera, and more so Pterostichus adstrictus, can very easily escape attention in south Sweden, and probably have remained undiscovered in some regions. Records of a trans- grading, only accidentally occurring species (i.e. Agonum gracilipes, Harpaius calceatus and H. griseus) are determined purely by chance; however, in such cases the completeness of the map, even the exact position of each locality is of minor importance. At the outset some facts were mentioned in support of the view that the maps presented in this work are to a great extent true to nature. In support of this we may Cite: 1. The frequency of remarkably regular lay of the northern limit of a num- ber of southern species, for instance Agonum obscurum, Broscus, Pterostichus vulgaris. Likewise the Swedish eastern limit and the Finnish western limit respectively of Carabus concellatus and Cicindela hybrida. Even inconspicu- ous carabids like Bembidion illigeri and Trechus secalis, which have not been collected by every entomologist, show sharply defined area limits. 2. The extremely rare discovery in recent decades of any new carabid species in Sweden. Altogether 20 species (6.8% of the fauna) have been collected for the first time in Sweden during this century. Of these only 3 have been collected during the last 10 years (1938-1947). In the context of changes in the fauna (pp. 621 ff.) these “new arrivais” are studied in more detail. This study shows that 6 (possibly 10) of the above-mentioned 22 species must have actually immigrated to Sweden very recently. Of the remaining species, 5 were found exclusively and 6 others at least predominantly at localities, such as the Abisko region, that were not visited by entomologists before 1900. Three others are markedly stenotopic species (Agonum munsteri, Perileptus, Tachys bisulcatus), whose discovery almost presupposes a knowledge of their concealed mode of life. Only 4 species are left, Carabus intricatus, Dromius longiceps, Dyschirius 428 378 Fig. 57. Author’s collecting localities in 1936 (May 28 through August 15). Blank circles indicate superficially explored localities; Black circles more care- fully investigated ones. septentrionum and Harpalus rupicola (if the last of these has not recently immi- grated), that escaped the notice of earlier entomologists. However, this actually forms a very small component of a fauna of 326 species. In expressing our ad- miration for the earlier collectors, like Gyllenhal, Zetterstedt, Boheman, or Thomson, we must recognize this as evidence for the remarkable completeness 431 319 of the results of their entomological collecting activity. Conversely we are also justified in considering carabids such as Agonum bogemanni or Harpalus nigritarsis, which have not been found in Sweden in this century, as vanishing or even extinct species. What has been said for Sweden applies to Finland almost to the same ex- tent. On the other hand, Norway has been irrefutably less explored. It is thus to be expected that future noteworthy changes in the picture of the Fennoscan- dian distribution map of one or other species will take place mainly in that country. The Relationships of the Fennoscandian Fauna Only two carabids, Bembidion scandicum and Bradycellus ponderosus, are known exclusively from Fennoscandia, and are therefore (provisionally) to be consi- dered as endemic to the area. In all other cases it is clear that only a precise knowledge of the total area of the species concerned can provide sufficient clues to its history, for example on the dynamic center. Distribution maps of the total area, at least of the most interesting species, would thus be highly desirable. I have simply not ventured to undertake the preparation of such maps. When the sources of error are already so large in a relatively small region like Fennoscandia, and the need to verify all the records published and unpublished imperative, the preparation of such maps of the total area must be considered unattainable, even for a limited number of species. I have been impelled to exercise the utmost caution with regard to the locality data in the literature, following the sound critical approach laid down by Horion (1941, Preface). Attention has already been drawn in Part I of this work (and in the Supple- ment at the end of this part) to the maps already to be found in the literature on the total area of one or other carabid species. The maps by Sainte-Claire Deville, and the maps of Bembidion by Netolitzky, seem to be the most reliable. In Part I, with every species a characteristic of the distribution type was given. This was done partly with “code words” (circumpolar, palearctic, etc.; see Part I, p. 13). But it must be emphasized that these terms are valid so to speak “proceeding from Fennoscandia” and comprise—especially the term “palearctic”—species with very different total areas. A more natural division into immigration groups will be considered below (p. 703). Nevertheless it might be useful to summarize the Fennoscandian fauna according to the above “group designations.” This gives us the following results: Circumpolar species (in Europe, Asia, and North America): 26 Palearctic species (in Asia, at least to longitude 60° E): 225 432 433 434 435 380 West Palearctic species (in Asia, as far as West Turkestan): 46 Euro-Caucasian species (possibly also in North Africa): 11 Euro-Mediterranean species (in North Africa and (or) the Near East): 10 Purely European species (also missing from Caucasus): 42 Amphi-Atlantic species (in Europe, the Mediterranean region and North America): 2 A better general view is provided by a cartographic representation of the number of species that Fennoscandia has in common with various subareas (Figs. 58, and 59). Since these figures vary widely in accordance with the size of the selected region for comparison, and evidently depend on the more or less thorough exploration thereof, they offer no absolute value, but give a definite impression of the extent of the faunistic relationship. The first thing that strikes one is the vast, in most cases, probably con- tinuous distribution of our species to the east, with about one-half of the Fennoscandian fauna having been found even in central Siberia. The high numbers in North Asia are all the more striking, considering that in East Fennoscandia there are only 298 species compared with 326 in Sweden (Nor- way has 244 species, and litile Denmark has 308 species). The distribution of the species within the confines of Fennoscandia (Fig. 60) also shows that south Sweden has incomparably the richest fauna, suggesting a relationship “directed” predominantly toward the south. Especially instructive is an analysis of the above stereotyped 7 patterns of distribution types according to the dynamic characteristics of their compo- nents (Fig. 47). The “Euro-Caucasian” and “Euro-Mediterranean” species are appropriately combined and the two “amphi-Atlantic” species are ignored. As usual, a distinction has been made between macropterous, brachypterous and dimorphic species: all the carabids showing wing-dimorphism in some part of their area have been placed in the last group. Diagram 47 unmistakably shows that the species with especially restricted distribution—the “purely European” species—are largely flightless. The per- centage of macropterous species increases gradually and steadily with the size of the area and attains its highest value (85%) among the circumpolar species. It is also interesting to note the intermediate position of dimorphic species, which form the largest component among species having moderately wide dis- tribution (groups b and c). A word of caution against possible misinterpretation of the two maps (Figs. 58, 59) is that these tend to give the impression of Fennoscandia as a center, from which the species emigrated in different directions for different distances, depending on their dynamic characteristics. Nothing could be more incorrect. In the total area of individual species, Fennoscandia usually occupies a peripheral position; for a smaller portion of the entire population of a species the region may have been important at the most as a secondary center (during 432 436 381 OW OVP? KAT OR Fig. 58. Approximate distribution of 362 Fennoscandian Carabidae in the rest of Europe. the Wurm Ice Age). The method of collective mapping of species with more or less similar distribution according to the “theory of equiformal progressive areas” (Hulten, 1937) also poses the danger of overestimating the role of the selected starting region as a distribution center. Hultén, who studied the distribution of higher plants, failed to avert this danger. In line with his choice of the starting region, he clearly overestimated northeastern Asia as the original center (Holdhaus and Lindroth, 1939, p. 273). bidae. 383 Zaun 7 BE: er ; Ly r CM Cys Ata #5 Pr Y > J! © = BEN AN > 4 & » A eS 7 N 4 % Sr © 8 Ns KI GL ~ 4 Y) A Set Fig. 60. Approximate number of carabid species in different parts of Fennoscandia. 434 384 90— 2 80- ip 70- “2 60- a 50- 4 40- ti 30- 20 10- ap d e Number of 42 21 46 225 26 species a 435 Diagram 47. Percentage composition of macropterous (blank), dimorphic (hatched), and brachypterous (black) species of Fennoscandian carabid fauna within different distribution groups. a—Purely European species; b—Euro-Caucasian and Euro-Mediterranean species; c—West Palearctic species; d—Palearctic species; e—Circumpolar species. 436 Species Distribution among Different Plant Regions For the reasons given above (pp. 43 ff.) no distribution of the Fennoscan- dian carabid fauna by region was undertaken. On the other hand, it might be valuable to determine the distribution of each species according to the vari- ous Fennoscandian vegetation zones, which are especially sharply delimited to the fjeldt areas, where they appear as high-altitude zones. This is the simplest way of estimating the climatic requirement (chiefly thermic) of a beetle. These results can be very useful in judging the climatic and biogeographical influence of the “Ice Age.” The division utilized here represents a simplification of that undertaken by the Swedish botanists. Seven regions are distinguished (cf. map in Fig. 61): t (Barrier plateau of the Scandinavian upland; suppl. scient. edit.). ei T SD 5 : 5 Ga = g Bet ZB aXe N ER N AY oe er = 3 = | bah s DL, i N ao (= J \ a ey We BEE SF, ms: 9 ae I8 SS oma ; UKE Pa) ~~. afd 2 h St Sed b d i N og o 6 R 27% by AT . > ° 5" DAN NL L 3 Q\ 3 & « @ =) VE zum ER Sg ee CAR 437 Fig. 61. Simple division of Fennoscandia into forest regions. Slant-hatched—Regio fagina; vertical-hatched—Regio quercina; White—Regio coniferina (and Regio betulina); Stippled—Regio alpina. Northern coniferous limit is drawn only in east Fennoscandia. Isolated north- ernmost coniferous forest stands are shown in black. Arrows indicate wooded passes through the Scandinavian watershed; those in Regio betulina are in- dicated by blank circles, those in Regio coniferina by black circles. (Ekman, 1922, p. 363; Enquist, 1933)*. After Kihlman (1890); Atlas öfver Finland, 1910; Du Rietz (1925, 1935); Enquist (1933); Hjelmqvist (1940). *The pass in Pite-lappmark, north of the Arctic Circle, is referred to as wooded only by Enquist (1933). It is missing from the map by Hard (1939). Fig. 62. Vegetation regions of Sweden (according to Du Rietz, 1935a; Härd, 1924). a— Absolute limit of Corylus; b—Absolute limit of Fraxinus; c—Absolute limit of Quercus; d—Limit of abundant Corylus; e—Southern limit of abundant Betula nana; f—Eastern limit of abundant Erica tetralix; g—Absolute eastern limit of Narthecium; h—Southern and western limits of spontaneous Picea. 439 387 1. Regio alpina superior (Higher alpine regiont). No continuous vegetation cover of vascular plants. Lichens and mosses dominate (Du Rietz, 1930, p. 358; 1942, p. 184). 2. Regio alpina media (Middle alpine region!). In Fennoscandia character- ized chiefly by heaths with Carex rigida—Juncus trifidus, in the north also with Cassiope tetragona (Du Rietz, l.c., p. 357 and p. 183). 3. Regio alpina inferior (Lower alpine regiont). Dwarf shrub heaths rich in species dominate (in the south with Calluna). Sometimes with Salix bushes (determined by edaphic conditions), for which reason an “osier region” was proposed (for example, Ekman, 1922, p. 549; 1944, p. 20). Luxuriant meadows on good soil, for instance with Trollius, Geranium silvaticum (Brundin, 1934, p- 85 ff.). 4. Regio betulina. The upper limit (or the northern limit) is formed by the timber line, but not by the treeline. 5. Regio coniferina. The upper (or northern) limit is the coniferous forest line formed by Pinus or Picea. Solitary coniferous trees or small groups of them may also occur in the Regio betulina. 6. Regio quercina. The northern (eastern in West Norway) limit is formed by the absolute northern or the high-altitude limit of Quercus. 7. Regio fagina. The region with abundant occurrence of stand-forming Fagus silvatica (according to Hjelmqvist, 1940). The division of the Regio quercina of south Sweden, generally adopted by botanists, into a western and an eastern part (p. 45) is ignored here because it is hardly climate-based. The position of these and other botanical lines of division, corresponding chiefly with the southern limit of the Regio coniferina, is illustrated in the map (Fig. 62). Let us discuss the most appropriate definition of the southern limit of the Regio coniferina, the “limes norrlandicus,” which has been recently studied by M. Fries (1948). According to him this limit follows a course that reflects the geographical as well as the climatic and biological viewpoints. But the choice of the absolute northern limit of Quercus is actually rather arbitary. In the following Table, which lists all the Fennoscandian carabids, occur- rence of the species outside the area is, as far as possible, also taken into consideration. Data exclusively based on such occurrence are given in square brackets. The tundra species are allocated only approximately, according to their advance toward the north. Occurrence in the northern peripheral area of the taiga is noted for the Regio betulina. Records in the north Norwegian coastal area, northerly to 70°20’N and easterly to Porsanger, are allocated to the Regio coniferina. t (suppl. scient. edit.). 388 440- Table 30. Distribution of Fennoscandian Carabidae within the different vegetation 448 zones. Cf. map in Fig. 61. Round brackets indicate marginal cases or + occasional occurrences (apparent cases of wind transport—for instance, in snowdrifts at high localities—are ignored). Square brackets indicate records outside of the region Regio alpina Regio Regio Regio Regio Higher Middle Lower betulina coniferina quercina fagina Abax ater (+) Au Acupalpus consputus Ar sh A. dorsalis Ar Ar + A. dubius oh: A. exiguus (+) Ar SF A. flavicollis Sr a= Ar A. meridianus (+) air tr Aépus marinus (+) 35 [+] Agonum aldanicum + [+] [?] A. archangelicum [+] == A. assimile SF ar ae A. bogemanni ar ar A. consimile Sr ate (+) A. dolens ei ap ar A. dorsale er ar A. ericeti =F an ar A. fuliginosum (+) Ar ate Sr ats A. gracile (+) (+) ae af aa A. gracilipes (+) (+) Tr A. impressum Ar ai = A. krynicki + Ar A. livens ap te ar A. longiventre [+] ala [+] A. lugens ate ap A. mannerheimi (+) A. marginatum ap ai A. micans == + =F A, moestum + a A. mülleri a5 ar an A. munsteri SF oh [+] A. obscurum Ar AP Ar A. piceum = ar “Ir 4. quadripunctatum Sr rn cia A. ruficorne Ga) 7 al A. sexpunctatum ai er ap 389 Regio alpina Regio Regio Regio Regio Higher Middle Lower betulina coniferina quercina fagina Agonum thoreyt A am tr A, versutum att ar =e A, viduum ar at al Amara aenea Ar al Sn A. alpina np: Ar aim ae (+) A. apricaria (+) IF =F =a ai A: aulica ictal ar ar ie A. bifrons a ir nm A, brunnea (+) + An + + + A. communis oF ar ae A. consularis + =i AR A. converiuscula ar as A, crenata al! A. cursitans Er) Tr ah A. curta air a ate A. equestris are +s air A, erratica [+] Sr an = A. eurynota Gi) ar zum 7 A, famelica ar = ate A. familiaris ae ae al A, fulva at ais ar A. Fusca a A. infima m ar an A. ingenua ae a ef A. interstitialis = ar (3) A. littorea Ct) AE Ar Sr A. lucida ai ah A. lunicollis a an. an al A. majuscula Ar + tr A. montivaga ei MIR ar A, municipalis (5) 3 ne ab Zu A. nigricornis ap aL Ar A. nitida Ar ar + A. ovata Ar is ai A. peregrina Gr) [?] [+] A, plebeja at ae ie A. pretermissa aR TUR ote te on A. quenseli Ar er ai + + ain A, similata ar ae oth A, Spreta Gap) at + A. tibialis air = ng A. torrida Ce) m ae Anisodactylus binotatus at Tr AR A. nemorivagus (Cam) Hr GR A. poeciloides ay + ny Asaphidion flavipes 390 Asaphidion pallipes Badister bipustulatus B. B. B. B. dilatatus peltatus sodalis unipustulatus Bembidion aeneum Dwr Do iu du Du bu In Do Du Du ui bo Du Du Du in bu Du ws andreae polonicum argenteolum articulatum assimile azurescens biguttatum bipunctatum chaudoiri Clarki crenulatum dauricum dentellum difficile doris fellmanni femoratum fumigatum gilvipes grapei grapeioides -guttula harpaloides hasti hirmocoelum humerale hy perboraeorum illigeri lampros lapponicum litorale lunatum lunulatum minimum monticola nigricorne nitidulum obliquum obtusum octomaculatum Lower an [+] Regio betulina + DZ (+) (+) Regio coniferina + + Gis) + Gi) + Cy) + Gi) Ci) (+) + + vw — DZ SE VS Peete ne dee Pe a er 8 Regio quercina = aU) — en an en +S++++++++++++t+H+ ¢+ + + + ++ FH+t+t+t+t+ ++ + Regio fagina ++ H4++H4++4t+4+4+444 = ir — =) +3 = +++ + + Ee +t+++t++++4++ ++ + Bembidion pallidipenne Do Be Da in Bu by By du u oy OW in Du yo du iu iu Du In du du in Do prasinum properans punctulatum pygmaeum quadrimaculatum quinquestriatum repandum ruficolle rupestre saxatile scandicum Schüppeli semipunctatum stebket stephensi striatum tibiale tinctum trans parens unicolor ustulatum varıum velox virens Blethisa multipunctata Brachynus crepitans Bradycellus collaris bo by by by be csikit harpalinus ponderosus similis verbasci Broscus cephalotes Calathus ambiguus BEGET SN aaa erratus fuscipes melanocephalus micropterus mollis piceus “alosoma auropunctatum denticolle inquisitor investigator reticulatum Higher Regio alpina Middle Lower (+) (+) Regio betulina + + +++ Et t+H+ HH E+ terttt+e ++4+4+4 Regio coniferina +2 4- ++++ Regio quercina + = nd -~ ++tttt¢+ + + ++++H4++H+++ FEF +4 444444444 391 Regio fagina ++ ++ ++++++ — — “-_ +EHH+etet+et+et +4444 £4444 = u — 392 Calosoma sycophanta Carabus arvensis (Gs auratus Ge cancellatus & clathratus @ convexus (Gs coriaceus Gs glabratus Ce granulatus C; hortensis (Gi intricatus C. . Menetriesi (0% monlis Ge nemoralis Ge nitens C. problematicus Gi violaceus Chiaenius costulatus C. nigricornis Ge quadrisulcatus & sulcicollis C. tristis G: vestitus Cicindela campestris Gi hybrida C. maritima (©; silvatica Clivina collaris G@ fossor Cychrus caraboides Cymindis angularis (Ge humeralis (Cy macularıs GC vaporariorum Demetrias imperialis 1D). monostigma Diachila arctica D. polita Dichirotrichus pubescens iD rufithorax Dolichus halensis Dromius agilis angustus fenestratus linearis longiceps Sboy Regio alpina Higher Middle Lower - “- + - + HH + + + = - + + + Regio betulina +++ Regio coniferina ++++ 4 ++ +++ ++++ +4 + ++ + $34 Regio quercina ~~ (+) + Gi) +++ H+ +++ + +4+4+4+4+4+4+ -o- wa -~ C+t++ + +F+++++++ De +++++ ++ A pn Regio fagina m Sie [+] ae + er aad = u FH HHHHH Er H HH HH HH He ttt $44444 — — ++++++++ Dromius marginellus SSsyss Sess I momo SOySysyyoysyy Rt BE SS is Be Be By be Be melanocephalus nigriventris quadraticollis quadrimaculatus quadrinotatus sigma yschirius aeneus angustatus chalceus globosus helleni impunctipemnis intermedius lüdersi neresheimeri nitidus obscurus politus „rufipes” salinus septentrionum thoracicus aphrus angusticollis cupreus lapponicus riparius uliginosus arpalus aeneus anxius azureus calceatus distinguendus frölichi fuliginosus griseus hirtipes latus luteicornis melancholicus melleti neglectus nigritarsıs picipennis pubescens punctatulus Higher Regio alpina Middle Lower ar I] +++ Regio beiulina +++ Regio coniferina +f +4+4+++ 4444 — in ~S (+) 393 Regio quercina Regio fagina + + Dr - + u ls + +++++++ ++ F+tt++tt++++Hette + + + +44444+4 — m nd +++ ++++++3++4 394 Regio alpina Regio Regio Regio Regio Higher Middle Lower betulina coniferina quereina fagina Harpalus puncticeps (+) + als puncticollis + + + Jal, quadripunctatus + + =F + H. rubripes + =; + ie rufitarsis =F Sr H. rufus (+) (+) Jal rupicola + [+] H. seladon + + + H. serripes + + H. servus (+) + ele smaragdinus + ae + EI tardus + + + H. vernalis + + H. winkleri + + + ir + Lebia chlorocephala + + + IL, crux-minor + + + It cyanocephaia (+) + + Leistus ferrugineus + + + + IE rufescens + + + + IE: rufomarginatus (+) + Licinus depressus zig == Lionychus quadrillum (+) [+ Loricera pilicornis (+) SF + + + Masoreus Wetterhalli + + Metabletus foveatus (+) + oF M. truncatellus + + + Microlestes maurus (+) + ae M. minutulus Sr Ar I Miscodera arctica + Ar ar ats Ar (+) Nebria brevicollis + + N. gylienhali Ge > an m ae N. livida + + + N. nivalis Ar + =F (+) N. salına (+) + + Notiophilus aquaticus + + + + + + N. biguttatus (+) + + + + N. germinyi +a + + + + N. palustris + + + N. pusillus (+) + + N. reitteri (+) + + N. rufipes + Odacantha melanura (+) + + Olisthopus rotundatus + + + Omophron limbatum 4- + Oodes gracilis + [+] ©); helopioides + + + Panagaeus bipustulatus J crux-major Patrobus assimilis IZ atrorufus iP: septentrionis Pe sept. australis Pelophila borealis Perileptus areolatus Pogonus luridipennis Pristonychus terricola Pterostichus adstrictus aethiops angustatus anthracinus aterrimus coerulescens cupreus diligens fastidiosus gracilis lepidus madidus middendorf fi minor niger nigrita oblongopunctatus punctulatus strenuus vernalis ; vulgaris Sphodrus leucophthalmus Stenolophus mixtus iS teutonus Stomis pumicatus Synuchus nivalis Tachys bistriatus TR, bisulcatus Tachyta nana Trachypachys Zetterstedti Trechus discus fulvus micros obtusus quadristriatus rivularis u No ID Nu Ju Su Nu Nu Su VU VU Saar Regio alpina Middle Lower Regio betulina + +++ Regio coniferina (+) 395 Regio Regio quercina fagina = + + ~ + + + let [+] CG) 7 [eel ar LJ a + (Gis) +2 aN wa — — ++ +++ tt Hr HtHtHtHH Ftt+ +4444 — +++ +++ tt +++ Ft+t+4He+e+ EHH+ +4+4+4+4+4++ — — ps] — — ++++++ EH Regio alpina Regio Regio Regio Regio Higher Middle Lower betulina coniferina quercina fagina Trechus rubens ate or ir IR; secalis ar Ar + Trichocellus cognatus Sr Ar SF a= ar 1% mannerheimi + [+] [+] IF placidus ate + + Zabrus tenebrioides + 448 Summation gives the following figures for each of the seven regions: Higher regio alpina: 4+ (1) hence 5 species (max.) Middle regio alpina: 12 + (3) + [1] hence 16 species (max.) Lower regio alpina: 43 + (20) + [9] hence 72 species (max.) Regio betulina 67 + (17) + [8] + 6 hence 99 species (max.) Regio coniferina 206 + (58) + [6] + 6 hence 276 species (max.) Regio quercina 273 + (34) + [8] + 1 hence 316 species (max.) Regio fagina 272 + (5) + [38] + 1 hence 316 species (max.) This summary highlights the extreme paucity of the carabid fauna in high alpine Fennoscandia. The only species exclusively native to the Regio alpina is Nebria nivalis. Earlier I gave a survey of all the Coleoptera known at that time from the Swedish Regio alpina (Lindroth, 1935b, pp. 25 ff.). 449 450 Existence Factors Climate Generally—and probably correctly—climatic factors have been considered decisive in determining the area limits of an animal or plant species. This is certainly valid as long as we study the main features of the distribution of animals and plants from the Equator toward the poles, or when we take into consideration larger regions, such as the African continent, which have remained more or less unaltered through long geological epochs. On the other hand it might be more than dubious to consider the fauna of a newly originated island as determined mainly by the climate. An intermediate position is occupied by regions which, like Fennoscandia, were colonized only in a geologically late period, chiefly in the postglacial period. According to the definition of the term “climate” it is important to de- termine not only its characteristic as the average condition of the atmosphere during any given number of years but also the spatial validity of the measure- ments taken. As a rule it is true that the higher the recording apparatus is placed above ground level, the greater the area (at the same height) for which the representative values are obtained. Thermic and most other meteorological instruments are usually mounted at a height of 1.5-2 from the ground. If the site is not extreme in any way the results may hold good as representative for square miles (100 km?). This is the macroclimate. On the other hand if one wants the extremes in a region, in order to com- pare them climatically with the help of meteorological measurements, one is in the domain of lococlimate* (“ecoclimate,” Uvarov, 1931, p. 128; meso- climate, Geiger, 1942, p. 3). Suitable examples are provided by slopes with different exposure to the sun, especially in the fjeld regions (examples: Ham- berg, 1908, pp. 8-9, Frödin, 1915; Krogerus, 1937). The figures obtained may be considered representative for hundreds of square meters (100 m?). Finally, the microclimate operates within a much smaller unit area. These are the climatic conditions of square meters (m?) and their fractions. Hence *This expression was orally suggested to me by Lohmander. 451 398 microclimate can never be studied with the meteorological instruments set up in the usual way. The measurements have to be taken just above the ground and in the soil itself. Temperature Every attempt to correlate thermal conditions with animal or plant areas (for instance, isotherms and area limits) must at present be based on macrocli- matic observations. To date sufficient measurements are available only in this field and—what is more important—lococlimatic and microclimatic condi- tions cannot be cartographically depicted for large geographical regions, such as Fennoscandia. It is always useful to keep this difficulty in mind. For the thermic microclimate at the most represents a reflection of the actual con- ditions over and in the soil, which is not only highly generalized but is even distorted (for example, Geiger, 1942, Fig. 34, p. 75). Easily the commonest procedure is the correlation of the area limit—in our case primarily the northern limit—of an animal or plant species with an isotherm of the mean temperature for one month. Examples in the literature are so abundant that it seems superfluous to cite any. At any rate, in our climate, which is marked by very pronounced seasons, the use of isotherms for the whole year or even for one season (3 months or more) is to be avoided. But even the isotherms for a single month have a highly variable “bio- logical importance.” For instance, the correspondence found by Heydemann (1931) between the January isotherms and the northeastern limit of Selidosema plumaria Schiff. (ericetaria Vill.), and that between the southern limit of Colias palaeno L. and the January isotherm of —1° or —2°C (Hesse, 1924, p. 387; Friederichs, 1930, p. 146), can only be coincidental and hence inconsequen- tial, as shown by Warnecke (1931). For as soon as there is a snow cover, with its strong thermal isolation effect (Geiger, 1942, p. 159), the temperature of the air in winter, as such, has virtually no effect on soil organisms (the ge- ometrid mentioned above hibernates as a half-grown caterpillar). On the other hand it is quite possible that the January temperature is a definite expression of the area-limiting factor for trees sensitive to cold, such as lex (Holmboe, 1913). In spring and autumn the daily variations are so large, with the tem- peratures often plunging below + 0°C, that data on the mean temperatures also have little “biological importance.” This situation is different in sum- mer, when only in exceptional cases can the daily minima, and the maxima in our climate, do any direct harm to a very few stenothermic cold-loving organisms. The quantum of the “heat-sum” may be decisive at this time of year (see below for details), and this is fairly well expressed by the average monthly temperature. In choosing which of the three summer months to use, June is excluded because it shows large temperature anomalies in different 453 399 years (Ängström, 1938, Plate IV) and a considerable daily temperature range (Hamberg, 1914, Plate VI) with frequent night frosts (l.c., p. 39). Compared to July, August has the advantage of a smaller daily range, but normally has some- what more marked minima (Hamberg, 1934, Plate XXVIII), in the north even fairly regular night frosts. Moreover, July is the warmest month of the year throughout Fennoscandia, whether the calculation is based on mean tempera- ture, maxima, minima, or constant figures, and is hence in a way an exponent of the entire summer climate (Wahlgren, 1913, p. 162). A. Summer Temperature For the above reasons July must be considered the month whose average temperatures correspond closest to the biologically effective thermal compo- nent of the climate. I have therefore worked out an isotherm map based on the mean July temperature for Fennoscandia (Fig. 63). The data for Sweden are a simplification of those provided by Angstrom (1946, Plate III), for Nor- way they are taken from the Norske Meteorological Institute’s Yearbook, and for Finland from Mänadsöversikt (1946) and correspondence with J. Keranen. However, I must note that the interpolation method used by Angstrom (l.c.), which gives an extraordinary degree of detail on the map, does not seem to be reliable. As I understand it, the isotherms in regions with scattered sta- tions simply follow the altitude contours, which are subject to speculation especially in the cutoff (“Kupiert’”) parts of Norrland. It is characteristic that on the lower courses of the Pite and Angerman rivers, and on the depres- sion lakes in Halsingland small “warm regions” with > 16°C can be indicated just because meteorological stations are located in or near these regions. Why should not similar high July temperatures occur in the other river valleys of Norrland, such as the regions of Over—Kalix, Harads, and Hallnas, and on the rivers Kalix, Lule and Ume? The numerous botanically indicated “southern mountains,” chiefly in Lappland, should also be shown on such a detailed map, but they cannot be depicted by any “interpolation method.” A broader generalization of the isotherms, especially in north Sweden, would have made a more reliable impression. If the new July isotherm map is compared with that by Ekman (1922, p. 311), significant differences are evident. Specifically, the center of heat along the big central Swedish lakes is much more pronounced. Likewise, southeast Norway is evidently more favored, and the “cold gaps” between the 15°C isotherms in central Norrland and in Osterbotten on the Finnish side have either, disappeared or shrunk to a narrow strip in northern Angermanland. The alterations may be partly due to actual climatic improvement (see p. 643), although this is less obvious in July than in the winter and spring months. But they are partly due to the denser network of stations. We now come to the difficult task of selecting the species of carabids whose areas—on the basis of their distribution and ecology —may be infiu- x ee 2 % . = \ & 5 DR Ns N H y moet ; > Nt 5 > 16°C; 400 1.5-2 m above ground level, 1901-1930. Fig. 63. Mean temperature for July, 452 Manadsoversikt av vaderleken > 17°C; hatched After Norsk Meteor. Arbok; Ängström (1946); and Rubinstein (1927). Black broken line = 15° isotherm. The standard 13° and 10°C isotherms are 2 i Finl. (1946); indicated only in Norway and in the high North, respectively. 454 401 enced by the July temperature (summer temperature). For this it will be best to take definite, thermally characteristic regions of Fennoscandia. a. Plus-Districts | 1. Southeastern Norway, especially the regions around the Oslo fjord. This is the only part of Norway where the July temperature reaches 17°C. A more or less isolated northward advanced occurrence in this region is shown by the following species: Abax ater Cicindela hybrida Amara montivaga* Lebia cyanocephala Carabus convexus Licinus depressus. 2. The interior valleys of east-central Norway, chiefly the Gudbrands valley. In these valleys the 15°C isotherm swings conspicuously to north. The north- ernmost occurrence in Norway is particularly shown by the following species: Amara fulva Cymindis angularis Anisodactylus binotatus Dyschirius politus Badister bipustulatus Harpalus puncticollis Bembidion gilvipes H. smaragdinus. B. properans 3. The innermost part of Sognefjord. This is the most isolated and the mildest region in the whole of Fennoscandia. Correspondingly no fewer than 8 species occur here in complete isolation: Agonum sexpunctatum Harpalus rubripes Amara curta (also near geiranger) H. tardus Anisodactylus binotatus Lebia crux-minor Carabus cancellatus Metabletus truncatellus. These three Norwegian “warm regions” are of course also characterized by other climatic features, primarily by relatively low precipitation and more hours of sunshine, as described below (p. 497) in detail. Yet we might be fully justified in considering the summer heat—here represented by the mean July temperature—as the most important factor for the above-mentioned species. This will be evident from the comparison below with species that are favored by the oceanic west Norwegian climate. Especially interesting among the above 8 “Sogn species” are Carabus can- cellatus** and Metabletus truncatellus, both of which are flightless. Under the *However, the present distribution of Amara montivaga is strongly influenced by the history of immigration (p. 631). **If the record of a solitary specimen of Carabus cancellatus in Sogn is to be attributed to transport by man, the same cannot be said of Metabletus truncatellus, which is widely distributed in the above-mentioned region. Metabletus is dimorphic, but the macropterous form is extremely rare; 7 specimens studied from the inner Sogn were all brachypterous. 455 456 402 present climatic conditions they could hardly have immigrated to the Sogn region, which is completely cut off to the north, east and south by the Regio alpina of the fjelds (see also p. 688). In the small “17°C regions” along the Swedish west coast no especially heat-loving carabids have been discovered. At present it is not possible to decide whether the isolated records of Calosoma reticulatum and Pogonus luridipennis, the first was possibly more or less accidental, were made at places where meteorologically a favorable “lococlimate” has still not been estab- lished. It would be better to consider Anthicus gracilis Panz (Jansson, 1927, p. 222), which in addition to the warm region of North Bohuslän also occurs in southeast Norway. The matter is still clearer in the case of Sphingonotus coerulans cyanopterus Charp., which is common in the region of Strömstad and was also found along the Swedish west coast only in the “warm region” near Göteborg (Hansson, 1902, p. 33); also in southeast Norway (see also Kvifte, 1941, p. 42) but not in Denmark. The species has therefore been con- sidered a “xerothermic relict” (Wahlgren, 1917, p. 100; Ander, 1942, pp. 16 ff.). Ander (1942, 1947) also applies this explanation to the occurrence of Scolia hirta Schrk. near Saro in northern Halland and in southeastern Norway. The July isotherm map renders this assumption plausible. On the border between Sweden and Norway close to the coast insolation conditions are, in addition, extremely favorable (p. 497). 4. The central Swedish lake district, especially around lakes Malaren and Hjalmaren. The character of this part of Sweden as the biggest summertime warm region of the whole of Scandinavia is becoming evident with the new July isotherm map (Fig. 63) (cf. the older map by Ekman, 1922, p. 311). The fauna of this region is actually extremely rich and includes a strong markedly southern element, which I have already considered (Lindroth, 1943a, p. 139) very briefly in connection with Coleoptera. Among the carabids, in the first place, the following are the “central Swedish heat-requiringt species” (x = xerophilous): Agonum dorsale x Harpalus anxius x A. lugens Licinus depressus X Badister sodalis Odacantha melanura B. unipustulatus Oodes gracilis Brachynus crepitans x Panagaeus bipustulatus x Demetrias imperialis Secondly mention may be made of: Harpalus rufitarsis and Microlestes minutulus, which have been found only in the Vattern region, and Demetrias monostigma, Leistus ruformarginatus and Microlestes maurus, found only in the Vanern region. ticf. p. 305 ff.; suppl. scient. edit.). 457 458 403 On the other hand the more or less isolated occurrence, in the same regions, of species like Amara montivaga, Bembidion transparens and Dichi- rotrichus rufithorax is chiefly due to their history of immigration. All of them are demonstrably late immigrants to Sweden (see p. 630). Ecologically the 11 species listed above can be readily divided into two groups: xerophiles, marked with an “x” and hygrophiles, i.e. the remaining 6 species. I paid special attention to this latter group in the contribution cited above (1943a). Apart from other Coleoptera it includes numerous representatives of carabids. Zoogeographically it is easy to arrange them in a continuous series and also to classify them into four main types: a) There is a more or less uninterrupted connection of the central Swedish area along the east coast (where the July map shows some high values) to- ward the south. These are evidently less typical heat-requiring species and are not considered above. Examples are: Agonum livens, Badister dilatatus, Oodes helopiodes, Pterostichus gracilis. b) The direct connection of the area is interrupted southward, but the species concerned are found again on Oland and Gotland and in the southern- most mainland provinces of Sweden: Agonum lugens, Badister sodalis, B. uni- pustulatus, and Odacantha melanura. c) The species are also missing from Oland and Gotland, so that the gap in distribution becomes very wide. Examples of these are found only outside the family Carabidae: Oedemera croceicollis Gyll., Psammobius bipunctatus Fbr., Reichenbachia impressa Panz., Silis ruficollis Fbr. d) The species are completely missing from Scandinavia outside the cen- tral Swedish lake district: Demetrias imperialis (for a possibly accidental oc- currence on Faron, see p. 287), Oodes gracilis, Euconnus rutilipennis Mull., Stenus solutus Er. (the record from Skane in the Catalogue, 1939, remains unconfirmed). Here we need not venture an opinion as to whether a relict phenomenon is involved here (cf. p. 687). However, these distribution types at any rate show that the central Swedish lake district must be climatically favored in some way. The following account of other climatic factors will show that such a pronounced advantage derives only from summer temperature con- ditions. The view expressed earlier (Lindroth, 1943a, p. 139) might therefore be justified, namely that “the locally (microclimatic) elevated midsummer tem- perature of shallow, eutrophic lakes accounts for the favorable nature of the central Swedish lake district.” However, even the 5 xerophiles (“x”) of the above list show a similar oc- currence in central Sweden, which is more or less isolated (especially in the case of Harpalus anxius, least in the case of Panagaeus bipustulatus). We are therefore justified in assuming that the favorable thermal conditions in cen- tral Sweden affect chiefly, but not exclusively, the limnetic biotopes, i.e. it is 404 ‘Repel ‘UIOIPUIT 0} Zuıpıone sawads sapoQ '(Z) sapıoıdojay ‘EC pue (1) syrav48 sspoo— ‘(OL 'I) S-r sounsedxg (zZ) smuviwpp 'g pue (1) smvpmsndium .sıpog—g ‘(ox ’d) €-7 stuswnadxy 7) wnnpia “y pue (1) suadny unuosy—e "DAneI21 payngunsıp Ajapim 10m e YIM paseduoe yoea ‘satoods „Surmmba1-Jeau,, USIPaMS JENUS 991 JO epua1sja1ld oinjelodwiay ‘gp weiseiq [ 4 ( | t U i] 0 l v 4 t 1 | 6 [| 4 i] | 4 4 ( 4 | peek Lit | Wd EM UM EW tte R) „102 TT MI A b 08'S! | | | | = WE A Ele Il 2 q 098! al eu | ll iz St st i 2 | LEW mh EM N 2 e zz) Kae Sle al = All dee il ; 00€ GZ 0z Gt OL ueay\ i] [1 l l 4 4 4 4 | 1 4 4 8 | i] L) U ’ | ( 4 6 4 | LSt 459 405 a macroclimatic situation affecting the entire living world, and this concept is therefore fully confirmed by the new July map by Angstrom (1946). It is naturally important to verify experimentally whether the assumed “heat-requiring species” of central Sweden are actually thermophilous. A. few such experiments were performed, chiefly in the temperature gradient apparatus, where the “comparison species” were always simultaneously studied. Diagram 48 shows the behavior of three pairs of strongly hygrophilous, ecologically closely-related species in the temperature gradient apparatus. In each pair—geographically speaking—the “heat-requiring species” shows a preference 2.8-5.6°C higher than that of the more widely distributed comparison species.* The lower response point was studied only in the species of Oodes and Demetrias imperialis (Lindroth, 1943a, pp. 137, 143). Both the adult and the larva of gracilis showed considerably higher values than the adult of helopiodes. But in Demetrias these figures were much lower. The preferenda in this case might reflect a zoogeographically more important characteristic than the response point. Of the xerophiles, Harpalus anxius em H. rufitarsis weret tested from among the presumed “limestone species” (pp. 130 ff.). In the latter species | the temperature preferendum is extremely high, in anxius it is lower but still higher than in Airtipes (and in the widely distributed species aeneus, tardus and seladon). Also with regard to the lower response point to temperature, rufitarsis Shows far higher values than anxius, in which, strangely, it is ex- tremely. low. Brachynus and Agonum dorsale, which have been thoroughly compared (pp. 59 ff.), have low preferenda (in comparison with the species of Harpalus) but a much higher low response point (even higher than in Harpalus azureus). The experiments thus give the clear impression that the central Swedish “heat-requiring species” do actually have a thermal requirement. As a rule this is evident from the high temperature preferendum. Hence the temperatures in summer seem to be most significant, and our idea that the July isotherms provide a true picture of the central Swedish warm region of the fauna is strengthened. 5. The inner parts of south Finland. The July isotherms here, especially those of 16° and 17°C, swing strikingly northward. Let us find out whether this can explain the well-known fact that many animals in Finland have their northern limit much beyond that in Scandinavia. Among the carabids the following more or less pronounced cases may be mentioned: * The low values for the species of Badister are probably to be explained by their pronounced hygrophily (see p. 68). t(CE. p. 215 ff.; suppl. scient. edit.). 460 406 Acupalpus dorsalis B. unicolor A. flavicollis B. varium Agonum livens Broscus cephalotes A. obscurum Carabus cancellatus Amara aenea C. granulatus A. equestris Cicindela hybrida A. montivaga Dromius marginellus A. similata Dyschirius lüdersi Anisodactylus binotatus Harpalus anxius Asaphidion flavipes H. smaragdinus Badister bipustulatus s. ]. H. tardus B. dilatatus Lebia chlorocephala B. peltatus Microlestes maurus B. sodalis M. minutulus Bembidion andreae polonicum Nebria livida B. articulatum Oodes helopioides B. biguttatum Pterostichus angustatus B. humerale .P. cupreus B. nigricorne P. gracilis. B. ruficolle In the case of some of these species, for example, Bembidion nigricorne, B. ruficolle, Dyschirius lüdersi, it is natural to look for the basis of their wider distribution in Finland in the history of immigration. This is especially clear in species (for example, Panagaeus crux-major and Pterostichus vernalis, as also Amara nitida and Pterostichus niger) that have already spread into the Swedish region (in Nbt). Even in an extreme case, such as that of Bembidion biguttatum, which is known in Scandinavia with certainty only from Skane, one would not suppose that the species has actually reached its existence limit there. Otherwise one would be misled into believing that the East Baltic stock is physiologically different, with a lesser heat requirement. These questions are discussed elsewhere (p. 732). But it is out of the question that the marked incidence of a more northerly northern area limit in Finland is determined by existence ecology. Ekman (1922, pp. 357 ff.) alluded to this problem. He gives a list of no fewer than 23 mammals and birds that have spread east of the Baltic Sea much farther north than in Scandinavia. He implies that factors relating to the immigration history have had the main role, but does not examine the important ques- tion whether the climatically favorable conditions in Finland may not have contributed to the slanting Fennoscandian northern limit. It is not improb- able that, for instance, the northern limit of the thrush-nightingale (Aedon luscinia)* is dependent on the thermal conditions in early summer. In recent * Now known as Luscinia luscinia—General Editor. 461 407 decades, characterized by increased temperatures (see pp. 643 ff.) it has consid- erably expanded its Swedish area northward. The northern limit of the polecat (Putorius putorius*, according to Kalela (1940a), is dependent on thermal con- ditions as well, at least in Finland. That the high northern limit of some carabids in Finland must actually be thermally determined, so that many species have reached their existence limits on both sides of the Baltic Sea, is evident from the fact that even some of the species whose more or less isolated occurrence in the central Swedish lake district characterizes them as pronouncedly “heat-requiring,” belong to the group with the “slanting northern limit.” Of the 16 species mentioned above (p. 455) one-half attain a more northerly position in Finland**: Badister sodalis Microlestes maurus B. unipustulatus M. minutulus Demetrias monostigma Odacantha melanura Harpalus anxius Panagaeus bipustulatus. The heat requirements of 2 of these species were experimentally studied (see above). One may also adopt the consideration of Ekman (1922, p. 368), that the pronounced east Scandinavian species ceases to occur in coastal southern Nor- way: If this limit were determined by dispersal factors “it would be rather strange that all the species should have remained static just at the western boundary of the country” (original in Swedish). In the present case it is strik- ing that among the carabids considered here, several species have almost the same northern limit, which forms a slanting line through central Finland at about the latitude of Kuopio. Clear exponents of this type are: Agonum ob- scurum, Amara aenea, A. montivaga, Asaphidion flavipes, Badister bipustulatus s. 1., Broscus, Harpalus smaragdinus. This conjunction of the northern limit of a larger number of species would be incomprehensible if it were assumed that it was determined by dynamics (dispersal ecology, history of immigra- tion). Yet the July isotherms correspond completely. It is noteworthy that the same 7 species—with the exception of Amara montivaga, which is in process of dispersal west of the Baltic Sea (p. 632)—show a largely common northern limit in Scandinavia too. A more southerly group in Finland, with the northern limit in the east around latitude 62° N, comprises Badister dilatatus, Bembidion articulatum, Microlestes minutulus, Pterostichus angustatus, etc. In Scandinavia the northern limit of these species also runs much farther south than that of the earlier group. The assumption is therefore fully permissible, that animals that have ad- vanced considerably farther north in Finland than in Scandinavia have in some *Now known as Mustela putorius—General Editor. ** Agonum dorsale and Brachynus are excluded because their occurrence in Finland appears to be more or less accidental. 462 408 cases actually found their existence limit on both sides of the Baltic Sea. It is difficult to detect the deciding factors outside the climate. Moreover, among the climatic factors I am not aware of any in the interior of Finland that might prove as favorable a condition as the summer temperature, as expressed in the lay of the July isotherms. These mean temperatures evidently do not represent the decisive factor—it would be more correct to work with the heat-sums of the summer (also see p. 479). But they give a general idea of the temperature con- ditions, which cannot be far wrong for the summer (cf. p. 451). Moreover, the inner parts of Finland compares favorably with Scandinavia with regard to the frequency of the temperature maxima in summer (Enquist, 1929, p. 20, Fig. 8). 6. The surroundings of the northern end of the Gulf of Bothnia. In Ek- man’s July isotherm map (1922, p. 311) the 15°C isotherm shows an iso- lated warm region here. In the new map (Fig. 63), based on the figures of the period 1901-1930, this isolation—at least partly on account of climatic improvement—is accentuated toward the east, and in Sweden in the plains there is an insignificant gap with < 15°C only in northern Angermanland. Now if the isotherm of 15.5°C were drawn—the necessary meteorological ob- servations are not available with me—the characteristic of this region as a heat center in midsummer, even though poorly marked, would be evident. Species like Agonum versutum, Amara tibialis, Pterostichus vernalis and Synuchus ni- valis here have their northernmost occurrence, not only in Fennoscandia but throughout their total area. Especially interesting are three species that were found more or less isolated in the same regions, north of their total area. These are: Chlaenius nigricornis : Panagaeus crux-major. Dromius longiceps It is highly likely that in these cases too the favorable summer temperature has been decisive. Elsewhere (p. 674) reasons are given for considering the occurrence of Agonum thoreyi in these regions as a relict from the postglacial warm period. b. Minus-Districts 1. The thermally unfavorable western and northern coast of Norway. With the exception of the inner parts of the most incisive fjords (Hardanger fjord and Sogne fjord) the entire Norwegian coast from the southern tip of the country lies outside the 15°C July isotherm. Examples of species that avoid these summer-cool coastal regions were provided in the context of the “plus- districts” along the Sogne fjord and the northerly position of many northern limits in Finland. It is easy to draw up a considerably longer list of species that have not crossed at all or have crossed only at isolated points of the main Scandinavian watershed to the west. The latter are in parentheses. 463 464 409 (Agonum ericeti) (Carabus clathratus, in the north) (A. piceum) Chlaenius nigricornis (A. quadripunctatum) Cicindela silvatica (A. sexpunctatum) Dromius sigma A. versutum Dyschirius politus _ (A. viduum) (D. thoracicus) Amara famelica (Harpalus aeneus, in the north) A. fulva Lebia crux-minor A. littorea Metabletus truncatellus A. municipalis (Notiophilus palustris, in the north) (A. ovata) (Pterostichus coerulescens, in the north) (A. tibialis) (P. lepidus, in the north) (Bembidion doris) (P. minor) B. gilvipes (P. vulgaris, in the north) B. guttula (Synuchus nivalis, in the north) (B. lampros, only in the north) (Tachyta nana) (B. obliquum) Trechus quadristriatus. (B. quadrimaculatum) This group is in no way homogeneous. Several of these species—especially those that never reach the actual fjeld areas—could be prevented from enter- ing western Norway “mechanically” by the chain of fjelds (cf. p. 614). Then again, where there may be reason to assume that the animal cannot live on the west Norwegian coast due to special existence requirements, it is not certain at all that the summer temperatures—or the thermal conditions in general—are decisive. The Norwegian coast has other peculiarities, with respect to precipi- tation and insolation and, moreover, there is the condition of the soil (p. 512). Let us therefore postpone a discussion until we can make a comparison with species that are apparently favored by the west Norwegian climate (p. 474). 2. The upland of South Sweden (in the provinces of Smaland and Västergöt- land). As an isolated “minus-district” this is especially interesting. The mean July temperature of the central part does not reach 15°C, a figure that comes down again only in northern Varmland. The fauna of the south Swedish upland is poor. The phenomenon that species of the southern distribution type completely or partially avoid these regions—as proven by the maps in Part II of this book—is so extraordinarily frequent that any enumeration of examples would be superfluous. However, it is not to be assumed that exclusively climatic factors operate in this connection. The condition of the soil, especially the almost complete absence of loam, plays a decisive role for certain species; conversely species that require fine sand and the like, for example Asaphidion pallipes, Bembidion litorale, are favored in these regions. The dearth of limnetic molluscs in these regions seems to be chiefly due to chemical factors (Hubendick, 1947). 465 410 Climatically the south Swedish upland is unfavorably disposed in almost every respect. In every season the temperature is lower than in the adjoining coastal areas (see also p. 474), the precipitation, especially in the west, is abnormally high (Fig. 73) and the insolation consequently low (Fig. 76). It seems scarcely possible to decide in individual cases which climatic factor has been responsible for preventing the entry of a species into the south Swedish upland. In some cases the distribution gap may have arisen under the combined effect of several of these factors. For the climatically dependent northern species a cool region evidently signifies a “plus-district.” In fact it is strange that the south Swedish upland does not show more northern forms, whether “relicts” or not. The only clear case is that of Pterostichus adstrictus, which is a counterpart of Saussurea alpina among the plants (Erlandsson, 1940). Moreover, in these regions Miscodera and Patrobus assimilis show greater frequency and abundance. These facts are undoubtedly climatically determined, but it remains undecided whether the low summer temperature, the short duration of the annual life cycle or some other factor has been effective. 3. The fjeld regions and the Far North. It is obvious that the more or less specifically fjeld and tundra animals, which in Fennoscandia have a dis- tinct southern limit, tolerate low temperatures both in summer and in winter. However, it is not as certain that they also require low temperatures, and hence respond negatively to moderate heat and are thereby adversely affected. Brundin (1934, pp. 159-160), in connection with the most pronounced high alpine species of the Fennoscandian carabid fauna, Nebria nivalis, actually sur- mised that its aimost complete restriction to the Regio alpina is understandable chiefly by the “low competition limit” of the species. Earlier (1935a, p. 616) I took issue with this. The question can be resolved only experimentally. I have not carried out any such experiments, but Krogerus has permitted me to publish here the tem- perature gradient apparatus experiments with Nordic carabids that he carried out during our stay together at Abisko in July, 1939. The readings with the apparatus were taken according to temperature classes, so the results are given differently than those of the other temperature gradient apparatus data. Other experiments with the same species are given separately (Table 31). The humi- dity of the air was regulated with wet cotton at the warm end of the apparatus. Krogerus’ experiments show that Nebria nivalis by far surpasses all other species in respect of marked “thermophoby.” Even at mild temperatures this species showed an alarm response, whereas at the cold end of the apparatus both cleaning response and copulation were observed. It is interesting that among the species investigated Agonum consimile, despite its restricted distribution to the fjelds and to the Far North, has the highest preferendum. However, in contrast with the other species, this carabid 466 lives on wet bog soil (not Sphagnum!), at sites exposed to sun, where the soil 411 465 Table 31. Temperature gradient apparatus experiments with Carabidae of the Far North (Krogerus) Dash (-) means this temperature was not represented in the experiment Temperature, number of individuals See Om LOS 20.129 15° 20° 23% 322 35% Mean Agonum consimile ORGS 4 1 9 6 0 0 = 13:92 Bembidion hasti 1 SA 1 2 2 0 0 0 ISIS B. hyperboraeorum Ze ee 1 2 0 0 0 —<10:02 Ra SE Oia, rts’ 4 3 0 0 0 - 9.98 ehr: SE 7 1 2 0 0 0 0 8772 Dyschirius helleni 07205975 2 3 0 0 0 = 11.9° By: 0, 345 1 0 0 0 0 0 9.4° Nebria gyllenhali f. typ. ABO 0 1 0 0 0 — Sale Ay 10756: 4 1 0 0 0 0 8.0° FR ruf SE 0... 2 0 0 0 0 0 6.0° a a Pe ee 1 1 2 0 0 0 0 0 6.8° N. nivalis N ae U 0 0 0 0 0 = OO a 3.9.0280 0 0 0 0 0 ae In Be AR “Ar. 5 ye ayes Ig pP 469 Fig. 66. September. Mean minimum temperature (1920-1939). Sources: cf. Fig. 68. Bradycellus collaris Olisthopus rotundatus Calathus fuscipes Patrobus atrorufus Carabus coriaceus Pterostichus niger C. hortensis P. strenuus Cicindela campestris (Trechus discus) (Dromius quadrinotatus) (T. micros) Leistus ferrugineus T. obtusus Nebria brevicollis T. secalis N. salina Trichocellus placidus. But the distribution type of these species is not climatically determined to the exclusion of all else. Reasons of immigration history have primarily de- termined the predominantly western (partly also northern) area of Bembidion 416 al ER EU e\ TA, BEN | SSH j CAT a ER ER 469 Fig. 67. October. Mean minimum temperature (1920-1939). Sources: cf. Fig. 68. femoratum (p. 744), B. nitidulum (p. 745), Trechus obtusus (p. 762), and prob- ably also of Bradycellus collaris (p. 397) and Pterostichus strenuus (p. 395). In Dromius quadrinotatus it is possible that its isolated, northernmost occurrence is due to passive dispersal (p. 320). Finally, both species of Trechus are missing from the western part of the country and both were found, totally isolated, only in two locales each in the Trondheim region. 475 The remaining 18 species in the list to a large extent clearly have cli- matically determined areas. Eight of them (Agonum assimile, Amara bifrons, A. ingenua, Carabus hortensis, Patrobus atrorufus, Pterostichus niger, Trechus se- calis, Trichocellus placidus) have a markedly slanting northern limit in Finland. This is in sharp contrast with the “July species” treated above (p. 459). Cor- responding with the lay of the minimum isotherms of the “critical” months, 417 470 Fig. 68. April + October. Mean minimum temperature of air at ground level (1920-1939). From Norsk Meteor. Arbok; Meteor. Jahrb. f. Finn. (until 1937); J. Keränen (in litt.); and journals of Sveriges Meteor.-Hydrol. Inst. Fig. 69. May + September. Mean minimum temperature of air at ground 471 level (1920-1939). Sources: cf. Fig. 68. 472 419 A a mM hu Q 33 En ED ry Fig. 70. April + October. Mean temperature of air at ground level (1901-1930) at same (neighboring) stations as in maps, Figs. 64-69. From Norsk Meteorol. Arbok; Sver. Meteor. Hydrol. Inst. Ärsbok; Mänadsöversikt av vaderleken i Finland (1946); and J. Keranen (in litt.) it runs from southeast to northwest, which means that toward the north these carabid species avoid a continental climate. The gap in the Oslo fjord region concerning the distribution of Nebria salina is also instructive, and to a lesser extent that of N. brevicollis. Any other interpretation than that of a climatically- determined existence limit seems to be erroneous in this case (both species are capable of flight). The Fennoscandian northern limit of Nebria salina even coincides with the + 2.5°C isotherm of the minimum temperature of April + October (Fig. 68). In a similar way Agonum assimile can be compared with the corresponding +3.5°C isotherm, and Calathus fuscipes with the + 4.5°C isotherm for May + September (Fig. 69). If the view expressed here is correct: that at least 18 species of the above 472 420 | | | 4, BS) > st | WZ, N ir? oD. : MGS Os Z a 1928 EN, N | le A u Fig. 71.May + September. Mean temperature ofair at ground level (1901-1930). (Sources: cf. Fig. 70). list are clearly favored by the oceanic west Norwegian climate—the question arises whether the ecology or some other aspect of the biology of these cara- bids does not provide an indication as to which of the climatic factors are the most important, and why these have influences just on these species and so effectively. One is reminded of an observation made by S.G. Larsson (1939, pp. 510 ff.). He found that the number of larval hibernators (“autumn reproducers”) among the carabids in Europe increases toward the west, especially toward the northwest; it is highest on the North Atlantic isiands (see p. 329). He divided the Scandinavian carabid fauna, which according to him includes 25.2% autumn reproducers, into five geographical groups, and gives the following percentages of “autumn insects”: 421 De INHBE: 2 Sy as A fi (Ai = gr Er ” Eas aa DR iS Eee NEES avy PFA, SN PSs 4 N = RN AN = pat Ir i u =) 7 473 Fig. 72. Mean number of days with frost (1881-1911) in May (left) and September (right). After Hamberg, 1914 (p. 40). 1. South Scandinavian group: 23% 2. West Scandinavian group: 89% (possibly 100%) 3. East Scandinavian group: 7% 4. Pan-Scandinavian group: 41% 5. North Scandinavian group: 23%. These figures are of course not very correct with regard to our present knowledge, for some new species have since been added to the Scandinavian 476 fauna, and there are others which cannot be placed in the distribution groups suggested by Larsson. Moreover, it has been shown that a larger number of species, at least in Scandinavia, belong to a different reproduction type than 477 422 Larsson assumes, which is pointed out in Part I against each of the species concerned. The Scandinavian carabid fauna as a whole may include 24.6% species that hibernate exclusively or regularly as larvae. The other figures de- viate insignificantly from Larsson’s*. At this point attention has only to be paid to the detailed consideration of the development types (p. 568). But the principle discovered by Larsson stands established: In Scandinavia the number of larval hibernators increases tremendously toward the west. If this principle is applied to the species grouped above on the basis of the position of their northern limit we find the following: 1. Among the 39 species which in Finland advance farthest north (p. 459) there are 5 species (Amara equestris, Broscus, Cicindela hybrida, Harpalus smaragdinus, Nebria livida), i.e. 13%, that regularly hibernate as larvae. 2. Among the 35 species that advance far into Sweden but avoid the Nor- wegian west coast (at least in the north) (p. 462), 7 species, i.e. 20%, are more or less regular larval hibernators (Amara fulva, A. municipalis, Cicin- dela silvatica, Pterostichus lepidus, P. vulgaris, Synuchus, Trechus quadristria- tus). 3. Among the 26 species whose area in Norway reaches the highest latitudes (p. 474), there are 15 more or less regular larval hibernators, i.e. 58%, (Amara bifrons, Bembidion lunatum, Calathus fuscipes, Carabus coriaceus, C. hortensis, Cicindela campestris, Leistus ferrugineus, Nebria_brevicollis, N. salina, Olisthopus, Patrobus atrorufus, Pterostichus niger, Trechus discus, T. obtusus, T. secalis). If we subtract the 8 species either missing in the western part of the country or having a distribution due to historical reasons (p. 474) the number of larval hibernators increases to 72%. It is scarcely possible to explain these conditions in any way other than as follows (expressed in other words by Larsson, 1939, pp. 526-527): The most important area-limiting climatic factors are those which affect the developmental stages. Hence the chief climatic requirements for adult hibernators are a warm summer, and for larval hibernators mild “critical” months in spring and autumn. The temperature conditions in the dead of winter may have a very secondary role in both cases. As we have seen, there is no lack of exceptions to this rule. But they are mostly understandable. The 12 “eastern” larval hibernators (enumerated above under Points 1 and 2)—with the exception of Nebria livida (see below) and Pterostichus vulgaris—are all more or less pronounced xerophiles. Among the 15 “western” larval hibernators there are only 2 (Amara bifrons, Olisthopus) that can be considered as such. Hence apparently the humidity of the oceanic climate also has a significant—positive or negative—role (see further remarks *Of course the percentage of “autumn insects” in Larsson’s “western group” is too high. His worst mistakes are inclusion of Bembidion nitidulum in the Pan-Scandinavian group, and of Agonum assimile in the eastern group. 478 423 on p. 485). Among the 11 “western” adult hibernators not one is markedly xerophilous. The view that larval hibernation in a continental climate is unsuitable is strongly supported by Palmén’s data (1946) based on records of immature carabids which were found in East Karelia. These are so strikingly late for a series of species considered as larval hibernators in Scandinavia that it must be assumed that the species concerned must have partially or completely switched to adult hibernation in East Fennoscandia. It is especially useful to study the data given by Palmen (mentioned in the Supplement of this book) for the following species: Amara apricaria, A. aulica, A. brunnea, A. consularis, A. fulva, Calathus erratus, C. melanocephalus, Nebria livida (immature beetles, June 18, but also August 22; “larvae in greaier numbers during July”), Trechus secalis. It would be very tempting—a task strongly recommended to some young, energetic entomologist—to undertake a comparative experimental study of the groups of western and eastern Fennoscandian Coleoptera (and other insects). Especially important would be an accurate study of the thermal responses of the larvae. . Further evidence that the difference of the northern limit in Fennoscandia between the “western” and “eastern” type is climatically determined can be provided by a study of the distribution in the rest of Europe. I have preferred to investigate whether or not the species in question occur in Ireland, i.e. a pronounced Oceanic area. If the above three groups are taken we obtain the following results: 1. Among the 39 species that in Finland advance farthest north, 17 species, i.e. 44%, are missing from Ireland (Acupalpus flavicollis, Agonum livens, Amara equestris, Amara montivaga, Badister peltatus, Bembidion andreae, B. articula- tum, B. humerale, B. ruficolle, Cicindela hybrida, Dromius marginellus, Harpalus smaragdinus, Microlestes maurus, M. minutulus, Nebria livida, Oodes helopi- oides, Pterostichus angustatus; Carabus cancellatus and Dyschirius lüdersi are doubtful). However, of these 17 species only 5 (Amara montivaga, Bembidion humerale, B. ruficolle, Dromius marginellus, Microlestes minutulus) are missing from England. Only for these might it be possible to argue that they could be missing from Ireland because of the immigration history, all the more so since none of the remaining 12 species is consistently flightless (Microlestes maurus is dimorphic). 2. Of the 35 species that also advance far north in Sweden but avoid the Norwegian west coast (at least in the north), 11 species, i.e. 33%, are missing from Ireland (Agonum ericeti, A. quadripunctatum, A. sexpunctatum, Amara famelica, A. littorea, A. municipalis, Bembidion obliquum, B. quadrimaculatum, Cicindela silvatica, Dromius sigma, Tachyta nana). Amara littorea, A. munici- palis and Tachyta nana are also missing from the rest of the British Isles. These species, as well as Agonum ericeti and Dromius sigma, which are constantly or almost constantly flightless, may be missing from Ireland for dispersal capa- bility reasons, but scarcely the remaining 6 species. 479 424 3. Among the 26 species whose area in Norway reaches the highest lat- . itudes only 4 species, i.e. 15%, are missing from Ireland (Amara ingenua, Carabus coriaceus, C. hortensis, Trechus secalis). The first 3 species are entirely absent from the British Isles and the fourth species (Trechus) is constantly flightless. The absence of these 4 species from Ireland is certainly due to dy- namic (historical) reasons. Finally it may be mentioned that the 6 species that occur more or less isolated in the “warm region” of southeast Norway (p. 454) are all missing from Ireland, with the exception of Abax ater. The above geographical comparison has shown that the more or less pronounced eastern—one might say continental—element of our “southern fauna” is also characterized in western Europe by avoidance of regions with a markedly oceanic climate. C. Duration and Frequency Figures of Temperature Early on, scientists thought of calculating the period during which the temper- ature exceeds or falls below a definite figure (mean, minimum or maximum), instead of expressing the biologically effective factors of climate by yearly, seasonal or monthly temperatures—.e. the time factor is fixed while the tem- perature factor remains variable. The temperature factor was therefore fixed and a variable time factor was obtained. In this way the isotherms were re- placed on the map by thermo-isochrones (see the historical account in Langlet, 1935). Duration figures are required for a period during which the temperature continuously falls below or exceeds a certain level. If only the number of days in the year with a particular thermal characteristic is summed we get frequency figures. Simple frequency maps are those reproduced above from Hamberg (1914) of the mean number of days with frost. Frequency values of mean temperatures were also compared by Samuelsson (1915), Rubinstein (1924) and Lunelund (1942b), with definite plant limits—especially the area limits of trees. The common isotherm maps for April + October or for May + September remind one of frequency or duration maps in their content, for they appear to provide an idea of the length of the vegetative period. But technically they are isotherm maps. During the last two decades, Enquist (1924, 1929, 1933) seriously followed the frequency method and tested it on plants. He avoids the use of mean temperatures and proceeds consistently from frequency figures (which he calls “duration figures”) of maxima and minima. With these he constructs thermo- isochrones which he compares with the area limits of forest-forming trees. To determine the frequency figure decisive for the species of tree concerned, Enquist (1933, pp. 151 ff.) proceeds as follows: He draws the frequency curves 480 481 425 of the maxima or minima of the year (mean figures for several years) for a number of stations located as close to the area limit of the species as possible. That is, he marks the number of days in the year on which a maximum (or mi- nimum) temperature of + 0,+ 1°,+2°C, etc. is normally exceeded. When two such curves of stations, located along the limit of the plants species, intersect, this indicates that the frequency value represented by this point is common to the two localities. With this procedure Enquist claims to have automatically found the decisive area-limiting factor for the plant species in question. He is misled (1933, p. 207) into such categorical statements as: “For the spruce to grow spontaneously, the maximum temperature must reach at least +12.5°C on altogether 65 days.” No wonder Enquist’s conclusions were seriously contested by a number of biologists. His method has been subjected to severe criticism, foremost by Langlet (1935), whose reasoning, it seems to me, is largely sound. A. Hamberg (1924), Almquist (1929, p. 22), and Lindroth (1939, pp. 244-245) have also been reluctant to accept Enquist’s view. To me the following seem to be the main objections to Enquist’s proce- dure: | 1. The maxima and minima of temperature that he uses are likewise mean values and consequently not “factors,” but constructions. Of course it must be conceded that these represent only “mean values of the first level,” i.e. the figures were obtained by simple calculation from the figures for a number of years. They are therefore less balanced than the mean monthly minima that I used above (Figs. 64-69), which represent “mean values of the second level” whereas the mean temperature for a month (for example, Fig. 63) even signifies a “mean value of the third level”: First the mean was calculated for each day, then for each July, and finally for the sum of all the July months in the series of years studied. But, to arrive at a view as to the extent of variation of the frequency values expressed by thermo-isochrones—the “mean values of the first level’—in dif- ferent years, I selected eight stations in widely separated parts of Sweden and for each year during the period 1919-1943 I expressed the frequency of the days with a minimum temperature > + 0°C (i.e. of frost-free days) by curves (Diagram 49; Table 32). This temperature factor was used by Enquist to ex- plain the area of both Pinus silvestris (1924, p. 207) and Picea excelsa (1929, p. 21). A similar curve is given by Hjelmgvist (1940, p. 195) from Bergen in Norway for maximum temperatures > 18°C over more than 60 years. The 25-year period shows an astonishing variability in the frequency values and the mean value is clearly a construction. The long continuous series of years (up to 8 years) with minus values is especially critical. How does the plant or animal behave if its minimal requirements of the decisive “factor” are not fulfilled over such a long period? It must be conceded that of all organisms, trees must be able to survive such adverse periods the best. It might be possible 426 Table 32. Frequency of days with frost and frost-free days at 8 Swedish stations, 1919-1943. Cf. Diagram 49 Average Amplitude Mean Years with Years with number of days deviation largest largest number of days with with frost from number of plus values minimum average of minus > One values Malmö 29 39.133 21 days 3 3 Halmstad 273 46-145 195 4 4 Visby 212 44-154 Pa 3 5 Vastervik 255, 78-172 16” 8 3 Stockholm 245 83-168 ign 8 6 Harnosand 192 136-209 14” 7 3 Haparanda 170 165-223 12,2 4 4 Karesuando 134 209-253 = 4 3 to assume that the “factor” primarily influences their reproductive capability, seed formation, ability of seed germination, growth of the young plants, etc. They might be able to survive for several years without or with greatly reduced mutiplication. It is known that seed formation in conifers in certain years may fail completely in large areas without seriously endangering the forest stands. But if trees (and other perennial plants) actually occupy a special position, in that they are able to survive long periods with negative deviation from the mean value of the “factor,” it follows that Enquist’s “duration figures” cannot be applied to annual and biennial plants or to animals that must reproduce every 483 year, and so cannot be used to explain the area limits thereof. 2. The procedure described whereby Enquist hopes to be able to find the decisive frequency value for a species of plant is unreliable. He even states emphatically (1933, pp. 152, 195), that the frequency value sought can be “suf- ficiently determined from just two stations,” i.e. that it is directly derived from the point of intersection of only two frequency curves from stations situated at the area limit (in this case at the birch limit). This idea is understandable only on Enquist’s assumption that the factor sought must be found among his frequency values. But the objective researcher would have first sought confirmation of his theory in the intersection of several curves. Langlet (1935, p- 347), who constructed frequency curves for a large number of stations near the timberline, failed to establish the regularity posited by Enquist. 3. It is wrong to take only marginal values as Enquist does. It cannot be inconsequential for organisms whether, for instance, 200 frost-free days include a pronounced or a poor midsummer period. At least as important as the frequency values is the optimum temperature. The longer the part of 484 427 summer close to that temperature, the shorter the time required by the plant or animal to complete the life process necessary before over-wintering. An idea of the optimum temperatures is best obtained from mean temperatures. We learn better the life requirements of the organism if we work both with the minima in spring and autumn and with the midsummer mean temperature. 4. Enquist’s thermo-isochrones show greater detail than is possible from the density of meteorological stations (for example, 1929, p. 21, Fig. 9, North Sweden; 1933, p. 157, Fig. 7). Their correspondence on the map with one or other treeline is thus no proof of a causal relationship (emphasized by Langlet, 1935, for instance, on p. 351). 5. The meteorological stations Enquist uses (for example 1933, p. 151) are not directly comparable because they only partially cover the same periods of time. How important this is in view of the large climatic changes during recent decades is evident from the diagrams given below (pp. 644 ff.). By reverting to some of the frequency values Enquist gives to the same period of the year, Langlet (1935, p. 345) found that they came out reversed for two stations (Kiruna, Storlien). Comparison of the frequency of frost-free days according to the older figures on Enquist’s map (1924, p. 207, Fig. 5) and the figures given above (Diagram 49; Table 32) for the period 1919-1943 likewise reveals considerable differences. 6. All Enquist’s frequency values were taken from meteorological stations. They therefore show the macroclimatic conditions. How little these are to be considered as biologically effective factors will be shown in a subsequent section (p. 498). And, it is just the maxima and minima of temperature that are strongly affected locally, not the media. On the other hand it must be conceded that forest trees, which are En- quist’s main concern, are of all living beings the ones in closest contact with the macroclimate, since they grow high and are less influenced by the win- ter snow cover (Wegener, 1923). But it must be determined whether even for forest trees the germination and growth of young plants do not repre- sent, climatically, the most sensitive time of life. If they do, trees are also influenced lococlimatically and microclimatically to the extent that the usual meteorological data, as they are treated, are completely inadequate.* This must be decided by purely biological observations and experiments on the various Stages of trees. In conclusion, concerning Enquist’s procedure it may be stated that his thermo-isochrones can in some cases—perhaps just for forest trees—provide a better expression of the climatic requirements of a plant than the isotherms of monthly averages, monthly minima, etc. But, and this is a serious objection, they can never express anything more than that, since for one species of plant a *] am unable to judge whether the limits of the pine and spruce area actually climatically determined in all directions, as claimed by Enquist (cf. Langlet, 1935, pp. 357 ff.). 485 487 482 429 mild summer is especially important for life, for another midsummer heat, for a third a long enough vegetative period, etc. Enquist’s big mistake is that he claims to have found the one decisive area-limiting factor, which is impossible. In principle his line of thinking signifies an advancement, but he has caused more damage than good, the more so since his “theory” was received with great enthusiasm not only by geographers and people generally interested in nature, but even by several outstanding botanists, who should have known better. It is accepted as fact in Swedish grammar-school textbooks of geography (for example, Swedberg, 1931). Hjelmgvist (1940), a faithful follower of Enquist, extended the procedure in the right direction by working experimentally with the plant concerned (Fagus silvatica). He expresses himself more cautiously than Enquist, though still not cautiously enough, when he attempts to determine the decisive factor for a certain section of the area limit (for example, pp. 169 ff., 185, 193 ff., 207, 229, 236). But it is strange that Hjelmqvist does not seriously discuss the objections raised by Langlet (1935) and others against Enquist’s method. Precipitation and Humidity The precipitation in Fennoscandia is so unevenly distributed (Fig. 73) that its biological effect, if any, cannot but be evident. Generally speaking, there is a fairly regular decline going from the west, where on the Norwegian west coast several regions exceed 2000 mm per year, to the east, and to a lesser extent from south to north. The lowest figures (< 400 mm) occur in three less extensive regions: in the southernmost parts of Oland and Gotland, in parts closest to the coast of Vbt and Nbt, and in the inland of the Far North. Also of particular interest are two isolated regions: a deficit region in the inner valleys of eastern south Norway, where the annual precipitation remains below 600 mm and an excess region in the western part of the south Swedish highland, which in part receives more than 1000 mm. If we wish to evaluate the zoogeographical consequence from the distribu- tion of precipitation we must be clear whether the yearly (mean or absolute) amount of precipitation can really be considered as a biologically effective factor. This factor might be negatively effective, and in the case of insects in such a way that definite “flying animals” are hindered in the flight effectiveness, essential for life. In this connection the fairly regular decline of Lepidoptera in Diagram 49. Number of frost-free days at 5 Swedish localities, 1919-1943. Compiled from “Sveriges Meteor. u. Hydrol. Inst Arsbok” (1919-1943). a—Malmo; b—Stockholm; c—Harnosand; d—Harparanda; e—Karesuando. Horizontal lines are mean figures at each locality during period. Cf. Table 32. 488 430 Europe toward the west is conspicuous (Pagenstecher, 1909, pp. 15-16; Hey- demann, 1930). It has been partly ascribed to the increased precipitation. But since the phenomenon especially affects butterflies! it is uncertain whether or not the lack of sunshine represents the more important factor (Lindroth, 1931, p. 393). In heavy, continuous rain, soil animals can also naturally be endan- gered by more or less local floods. In our climate, where the precipitation is fairly uniformly distributed over the whole year, this danger is comparatively slight. At any rate the annual precipitation can never be considered as a positive biogeographical factor by itself. It is therefore advisable to take into consid- eration its two most important results (a third, the insolation, is separately dealt with below on p. 495). These are the ground moisture and the humidity of the air. The ground moisture is of course actually an edaphic characteristic. But since it may be primarily dependent on the precipitation and evaporation, and only incidentally (in our region) on the drainage conditions, it seems justified to consider it under climatic factors. In Fennoscandia there are only a few extensive regions uninhabitable for a number of species because of insufficient ground moisture. The most im- portant are the alvar regions on Oland and Gotland and the heaths of the fjelds. But in the latter the cold factor is more detrimental. Nevertheless, a species cannot be excluded from an entire province only because of insufficient ground moisture. The gaps due to that are scarcely noticeable on the maps of Fennoscandia, so this negative factor has on the whole no zoogeographical significance. The condition is different in regions that show an excess of ground mois- ture. We are not thinking here of local moor and other swampy regions formed by poor drainage conditions, but of the regions with the heaviest precipita- tion, meaning Norway west of the main Scandinavian watershed. In the above treatment of the more or less pronounced eastern species, which clearly avoid these regions of Norway (p. 462), we found a large degree of correspondence with the development types, where the species, with few exceptions, are adult hibernators (p. 476). Species that prefer western Scandinavia (p. 474), are predominantly larval hibernators. The exceptions are: a. Eastern species hibernating in the larval stage: Amara equestris C. silvatica A. fulva Harpalus smaragdinus A. municipalis Nebria livida Broscus cephalotes Pterostichus lepidus Cincindela hybrida P. vulgaris t (which are diurnal; suppl. scient. edit.). 489 431 Synuchus nivalis Trechus quadristriatus With the exception of Nebria livida (but also see p. 477) and Pterostichus vulgaris, these species are more or less pronounced xerophilous. b. Western species hibernating in the adult stage: Agonum assimile Bradycellus collaris A. ruficorne Dromius quadrinotatus Amara ingenua Pterostichus strenuus A. lunicollis Trechus micros Bembidion femoratum Trichocellus placidus B. nitidulum None of these species shows xerophily. We may therefore conclude that the faunistic differences between west and east Fennoscandia—apart from factors related to the immigration history—are due not only to the thermic factors discussed above (pp. 474 ff.) but also to the humidity. That is, numerous markedly xerophilous animals cannot survive west of the main Scandinavian watershed. It is not possible to decide without experimental study of the species concerned whether the high moisture of the soil or humidity of the air play the main negative role. Such a study would certainly be worthwhile. In any case a glance at the map, for instance, of Amara fulva shows that the western limit of this flying species cannot be due to dynamics. Humidity of the air can be expressed as an absolute or as a relative quantity. Wallén (1930, p. 33) published maps on the absolute humidity of the air in Sweden for January, July, and the whole year. It is not a biologically effective factor. : Relative humidity of the air, according to the same author (l.c., p. 34), shows differences between different parts of the country that are too small for cartographic representation. On the other hand the daily fluctuations are too large. The great dependence of the humidity of air and moisture of the ground on temperature and the difficulties of expressing the former of these directly on the map have led many researchers to construct climatic indices in which both precipitation and temperature are represented. One understands the motive, for example, to express the reduced evapo- ration of the precipitation in cooler areas. This is done in the simplest way with Martonne’s humidity figures (Hesselman, 1932) = annual precipitation: (mean temperature of the year + 10°C). These figures show the proportion of precipitation that—with identical drainage conditions—reaches the soil (al- though the French “humidité” and the English “humidity” are the same as humidity of the air). In a region the size of Fennoscandia, where the tem- perature differences are not very large, the humidity map (Fig. 74) will on the whole coincide with the precipitation map; in the southern half of the [se See ES SS > N SS er = EZ Ae WS ms “Nedbp- area = > 2000 mm. 432 Fig. 73. Mean distribution of precipitation in Fennoscandia. From 486 riaktt. i Norge” (1938), Ängström (1946), Atlas öfver Finland (1910). Black 491 433 region this is especially evident. The mean—theoretical—ground moisture, as an expression of an area-limiting factor, has little significance. Among other ways of determining the humidity of the climate, mention may be made of A. Meyer’s (1926) “precipitation-saturation deficit quotients,” | i.e. precipitation : saturation deficit. The latter is calculated from the mean va- lues (of the year and of individual months) of temperature and relative humid- ity of the air put out by the meteorological stations. Heydemann (1930) in his study of the dependence of the lepidopteran fauna of Schleswig-Holstein on the Atlantic climate, used Meyer’s N-S quotient maps. But according to Hes- selman (1932, pp. 524 ff.), who tested the method empirically for two Swedish stations, the saturation deficit calculated by Meyer is beset with such large errors that for Sweden he preferred to use Martonne’s humidity figures. As discussed above (pp. 462, 474), the oceanic (Atlantic) climate is marked not only by high humidity (of the air as well as of the ground) but also, equally, by thermal conditions. It is natural to attempt to express the “oceanicity” of a region by a single index value, taking into account both groups of factors. The attempt was made by Kotilainen (1933) with an “oceanicity index” calculated as follows: yearly precipitation x (a —b) 10 x (T, — T,) where a = number of days with mean temperature > +0°C; number of days with mean temperature > +10°C; T, = mean temperature of the warmest month, T, = mean temperature of the coldest month. Oceanicity index = o Il Supplementing Kotilainen’s data (p. 56) with further indices for Norwe- gian and Swedish stations, I have constructed a new “oceanicity map” (Fig. 75). Although it is evidently important to be able to determine whether a particular area limit of an animal or plant species is determined either by thermic or by hygric factors, I believe that Kotilainen’s oceanicity is not without interest. Attention may be drawn to the low figures of the Oslo region (Nebria salina), to the eastern interior of south Norway, and to the comparatively high figures of Visby on Gotland (Nebria salina, Trechus obtusus). At the least such a map, taken with the “pure” maps of temperature, precipitation, etc., can contribute to a judgment as to whether a particular area limit can be existence-ecological or not. Finally mention may be made of Cook’s (1924) Climatograph. It is a simple and practical method of combining into one diagram thermic and hygric factors (mean temperatures and quantities of rain each month) at a number of places within the area of the species of animal concerned. The strong effect of humidity, not only of the ground but also of the air, can be ascertained more easily in the case of p/ants than of animals. Degelius 492 (1935, pp. 244-270) thus clarified the close association of the “oceanic” lichens 434 AESN, N] g EN 1 mat ig N Me u! | [ERK WORK N HA a) EN ? 4 Neel By Wf) IN Dee May: cee La A 2 a x Ti a ur 490 Fig. 74. Martonne’s humidity figures. After Hesselman (1932), supplemented from Kotilainen (1933), “Nedbgriaktt. i Norge,” 42 (1938), and “Norsk Me- teorol. Arbok” (1946). with the Scandinavian regions having the highest precipitation. He rightly considers the hygric factors of climate more important for these plants than the thermic factors. Earlier, Hard (1924, pp. 147, 225) gave a similar account of the western species of the flora of south Sweden. That any comparable dependence of animals is so obscure is due to many factors. First, animals are not so dependent on the occurrence of free water for the intake of food and for their metabolism. Second, unlike the higher plants, they are not bound to a locality. If there is danger of desiccation they can save themselves by hiding in the ground, flying away, etc. And they can search for water, to replace by drinking, any loss caused by transpira- tion. 490 435 NEL ie ASA E N. (3 we 22 Z N L Fig. 75. Kotilainen’s (1933) “oceanicity index”. Supplemented from Ham- berg (1922), “Nedbpriaktt. i Norge”, 42 (1938), “Norsk Meteorol. Arbok” 3 (1946) and official Swedish meteorological data. The experiments with insects in which the humidity factor was opera- tive show the more or less strong response to loss of water in all species, most of all, naturally, in distinctly hygrophilous species. As the transcript of experiments (pp. 78 ff.) shows, in some cases this response is stronger than that to the thermal effect. From this it must be concluded that the humidity of the air, in nature, represents a decisive factor for these animals too, and is probably the most important climatic factor along with temperature. The difficulty in clearly determining its effect in the field, in the habitat of the animal, or from its distribution map, is due not only to its mobility but also to the common micro-stratification of air of diverse humidity in and on the ground (for example, Geiger, 1942, pp. 88 ff., pp. 275 ff.). As long as the conditions of 493 436 relative humidity of the air cannot be represented on a map, it rarely plays a decisive role for the general distribution of insects. In regions where a certain minimum requirement of humidity of the air is nowhere fulfilled—in deserts and steppes—or where the air is almost constantly saturated—in an extreme oceanic climate, in tropical rain forest—certain species can be ruled out for these reasons. Moreover the humidity of the air is a microclimatic factor of the first order. This therefore determines the habitat of the individual rather than the area of the species. For this reason the humidity of the air in our climate is more an ecological than a zoogeographical factor. A phenomenon to which I have drawn attention in connection with the fauna of Iceland (Lindroth, 1931, p. 387) is the tendency of the more or less pronounced forest species in the rainy climate of western Europe to inhabit more open ground, sometimes drier places than normal. The explanation for this “ecological adaptation” must be that in such regions, even in an exposed situation, both air and ground possess sufficient humidity. Of the three carabids (Notiophilus biguttatus, Patrobus atrorufus, Trechus rubens) especially mentioned from. Iceland, the same ecological change applies at least to the Patrobus species in West Jamtland. From west Norway, where this phenomenon is expected to be most pronounced, we unfortunately possess all too little authentic ecological collection data. However, the strikingly high percentage of “forest species” in the fauna of Hitra and the adjacent islands (p. 320), all of which are almost devoid of forest, seems to me explicable only in this way. The combined effect of thermic and hygric factors determines the duration and depth of the snow cover. This plays an important role by insulating the ground thermally against low air temperatures (Geiger, 1942, p. 159). Whether at —20° or at + 0°C, the ground temperature can remain almost unaltered if there is a deep snow cover. Instructive figures were obtained by Keranen (1920, pp. 52-53) from measurements taken in Lk Sodankyla. During one December month the mean temperature of the air was — 24.4°C, but on the ground only — 6.5°C under a snow cover averaging 32 cm. During October 1916, in the absence of snow, the mean temperature was: air 4.3°C, ground — 2.1°C. During December of the same year, with 32 cm mean snow cover, it was respectively — 11.1° and — 1.4°C. Soil organisms susceptible to cold will thus be adversely affected by frosty days when ground is bare. I therefore have attempted to calculate the mean number of such days for different parts of Sweden. Exact data for different sections have never been published, but an isochrone-map on the mean number of frosty days is given by Wallén (1930, p. 21), and I have obtained a corresponding, unpublished map on the average duration of snow cover (1909-1935; in simplified form in Angstrom, 1946, p. 48) from “Sveriges Meteorol.-Hydrol. Institut”. From these two maps it is possible to estimate roughly the mean minimum number of frosty days with bare ground for any locality in Sweden. With the exception 437 494 ofthe region west of Torneträsk in Lapland the number of frosty days is every- 495 where greater than the number of days with snow cover. The difference—i.e. the minimum number of frosty days with bare ground—is largest in the frontier region between Skane and Smaland: about 110 days. The figures show such an irregular distribution that preparation of an isochrone-map on the basis of the available material proved to be impossible. (A map on the average date of the thaw in different parts of Sweden has been published by Edin, 1941, p. 33.) The north, especially the fjeld regions, has the smallest number of frosty days with bare ground. In the south the west coast has somewhat higher figures than the east coast. If we can establish a regionally determined gradation of the number of frosty days with bare grounds on the basis of a thorough consideration of the figures of every station, this undoubtedly represents a significant biogeo- graphical factor. However, it is more probable that it will turn out to be mostly locally (lococlimatically) determined, so that it influences more the local than the general distribution of organisms. Other Climatic Factors For regularly flying insects the wind conditions constitute a decisive environ- mental factor. For such insects the wind is not only an important mode of dispersal but can also be decidedly area-limiting in very exposed regions, especially in regions without forest. This is clear in the case of Lepidoptera (Pagenstecher, 1909, p. 16; Hesse, 1924, p. 555), in Scandinavia especially in the Alvar regions in Oland and Gotland (Wahlgren, 1917, p. 53), in the regio alpina of the fjelds, and on the outer naked rocky skerries, chiefly at the Norwegian coast. The importance of wind as a mode of dispersal for Carabidae is considered elsewhere (p. 573). This factor may possess an area-limiting effect only in exceptional cases. Very few carabids (Cicindela; Bembidion, subgenus Bracteon; see p. 579) fly so regularly as to be seriously endangered at places exposed to strong winds. Probably the fact that Bembidion lapponicum and B. velox in Norway avoid the outer coastal belt can be attributed to the deleterious effect of the wind. However, this does not seem to be so in the case of Cicindela campestris. The above species are also heliophilous, which probably has at least as important an area-limiting role. The main reason for attributing little importance to the effect of wind as an area-limiting factor for carabids is the composition of the insular faunas, considered above (pp. 198 ff.) in a separate section. It was shown that even the fauna of woodless islands such as Helgoland (p. 328), which have been exposed to wind for thousands of years, has not undergone perceptible change in the direction of an increase in flightless species and forms. While some carabids, especially the species with metallic color, are 497 438 markediy heliophilous, it is fairly probable that right from the beginning the insolation, the number of sunny hours, was an effective climatic factor. We are not dealing here only with the effect of light. Apart from the fact that the insolation of a locality is determined indirectly by decrease in fog, precipitation and similar factors, and thus provides a kind of converse index of their extent, it is a thermic factor of the first order. This seems to be all the more significant inasmuch as the usual meteorological measurements take the temperature in the shade. It is actually difficult to decide whether the sun influences heliophilous animals more as a light factor or a thermal factor (and so indirectly as a humidity factor). Probably, for most species, the purely thermic effect is the more important. Anyway, a map of the average number of sunny hours—best taken over the summer half-year—must be important just from the entomogeographical viewpoint (Fig. 76). A comparison with the isothermic map for July (Fig. 63) and for the other relevant months (for example, Hamberg, 1980; Angstrom, 1946) shows that high air temperatures in the shade are not always associated with strong insolation. Instructive in this connection is the Malar region, ther- mally the most favorable part of Scandinavia at midsummer, which is rather poorly insolated at that time (further information for every month was pro- vided by Hamberg, 1909). Some interesting details of the “insolation maps” are considered below: 1. By far the sunniest region of Scandinavia is Gotland (probably, for lack of stations, Oland comes off with low values). As a result the otherwise favorable thermal conditions of the island are considerably enhanced, which further explains the strikingly southern character of the fauna (pp. 289 ff.). 2. The frontier region between Norway and Sweden east of the Oslo fjord. Markedly heliophilous species, such as Amara montivaga and Lebia cyanocephala, might be favored here. Some places in the inner eastern val- leys of south Norway, and at the inner end of the Sogne fjord, also show high figures, which help to explain the thermophilous faunal element considered earlier (p. 454), including some of the xerophiles, here far advanced. 3. Regions of the northern end of the Gulf of Bothnia. At midsummer the number of sunny hours here is relatively even higher than shown on the map. A whole series of more or less heliophilous species here find their northernmost area limit and/or occur with frequency and in abundance, for example: Agonum dolens B. quadrimaculatum A. piceum Carabus clathratus A. versutum Harpalus aeneus A. viduum Pterostichus coerulescens Bembidion obliquum P. lepidus. The same is true of the following non-heliophilous species: Cymindis macularis Dromius longiceps 49 (ee) 439 D. sigma Synuchus nivalis. Lebia crux-minor Nevertheless, they may be thermally favored at sunny places. This is especially probable for Dromius longiceps and the Cymindis species, which live on dry sand slightly warmed by the sun. Likewise we may notice the species mentioned earlier (p. 462) that occur more or less isolated in these regions (Chlaenius nigricornis, Panagaeus crux-major, Pterostichus vernalis). The genus Chlaenius is markedly heliophilous. Three species show a very revealing distribution: Agonum sexpunctatum Metabletus truncatellus. Cicindela silvatica They not only show, like the above, a strikingly rich occurrence north of the Gulf of Bothnia, but they retreat in Jamtland, where they do not reach the region of Storsjon. Also in Norway they are markedly eastern, but two of them (Agonum, Metabletus) reach the inner Sogn (cf. p. 454). It is clear that they avoid a markedly oceanic climate. But which of these factors is decisive: the cool summer, high precipitation and humidity of the air, or lack of sunshine? All three species are pronouncedly heliophilous, two are also clearly xerophilous, but the third species, Agonum sexpunctatum, is hygrophilous, and at least for this species the humidity factor may be excluded. In Lapland the species extends far beyond the 15°C isotherm for July, which in Jamtland falls close to its limit, so its absence from the higher parts of the province does not seem to be explained by the insufficient summer heat. On the contrary Agonum sexpunctatum might represent the best example of a species whose area to the west is limited by deficient insolation. Strikingly, in the British Isles it occurs only in England. Finally it may be mentioned that I tried—of course with a negative result—to test the resistance of individuals of one and the same species (and population) with different intensities of metallic coloration to the effect of direct sunlight. An unusually colorful series (from black to pale green) of 18 individuals of the highly variable Harpalus aeneus (Upl Varmdon, July 3, 1941) was exposed to direct sunlight in a large glass bowl. The carabids were then removed and numbered in the order in which they were affected by thermal paralysis of the hind legs. No correlation could be established between the coloration and resistance (not even according to sex); the longest-lasting specimen was a black male with a very slight blue shade. On Lococlimate and Microclimate Let us come back to the observations at the beginning of this chapter. All the climatic factors mentioned so far and cartographically depicted are of macroclimatic nature. It is of course possible that some isolates, for 440 5 10 3 16" 1 18° 20” 22°. Zi” in 16° bw 12° av © a RB z a [3 2 So | REN REN \ / | 1200! WR Go! \ | | sts Er 5 = / PU ia I Q 33 i EHEN = Ron 496 Fig. 76. Mean number of sunny hours during the summer half-year (April through September). According to Hamberg (1909) and Lunelund (1942). “Isohels” are strongly generalized. Some Norwegian stations with abnormally high figures are indicated. 499 500 441 instance, the small thermal terrains in Norrland on the July isotherm map (Fig. 63), are actually lococlimatically determined, i.e. they reflect the place- ment of the instrument on an extreme surface. This could happen uninten- tionally, since meteorological measurements are generally considered repre- sentative for larger regions. However, in general the usual meteorological figures are clearly macrocli- matic, and in most contexts this is an important advantage. Their dependence on the local conditions at the station is surprisingly slight. In considering the climatic changes of the last decade Angstrom (1938, p. 26; original in Swedish) states: “The very uniform temperature fluctuations of different stations fur- ther support the claim already made that the figures of the monthly averages incorporate no errors arising from location and poor protection from radia- tion, etc. that exceed a few tenths of a degree.” I made the same observation while studying the later climatic variations (p. 643). The minima (chiefly the absolute) and the precipitation figures might show a somewhat higher local variation. The representative character of the meteorological climatic measurements, chiefly due to placement of the apparatuses an average 1.5-2 m off the ground, is an acute disadvantage from the biological point of view. The flora and fauna of the ground are exposed to very different climatic factors, which should be measured in the environment closest to the organism studied. A devastating insight into the unending fluctuations and deviations from the macroclimate of the “climate of the air layer next to the ground” is provided in the well- known contribution by Geiger (1942). The “discoverer” of the microclimate, Krauss (1911), provided plenty of convincing measurements. Concerning the investigation of the lococlimate and microclimate in relation to entomological materials, see Uvarov (1931, pp. 128 ff.), Franz (1931, 1933) and Kuhnelt (1933, 1934). Just one example that I have already given (Lindroth, 1943a) may be mentioned here since it illustrates with unusual clarity how big the thermal difference between adjacent and apparently similar surfaces can be. Two sam- pling plots were taken on and near the shore of a small lake in Up] Djursholm at a distance of 113 m from each other. On a July day the temperature of the air, at the soil surface, and at a depth of 5 cm was continuously measured. The curves of the temperature of the air were as good as identical, but those of the soil (Diagram 50) differed so much that the maximum temperature of the “cold” plot (19.7°C, 1800 hours) did not reach the minimum temperature of the “warm” plot (19.8°C, 0600 hours). It is therefore clear that the characteristics of the microclimate can never be depicted by a cartographic representation. Investigation thereof is a task for ecology. But the biogeographer must never forget that, although the lines on the map that he uses to “explain” the area of distribution all concern the macroclimate, the effective factors are always microclimatic. On the map he is 501 442 working not with factors but with constructions, which suffer from a two-fold error: They are mean values and they do not represent the true environment of the organism studied. Hence even a perfect correspondence between a climate line and an area limit should never be taken to prove more than that the limit is determined by one factor (or more) lying within the climatic domain represented by this line, for example, by the summer warmth, the length of the annual period of life, the humidity of the air, the insolation, etc. To avoid misunderstandings, in the present contribution I offer no area map indicating climate lines. My earlier contention may not have been too pessimistic: “In fact we will never succeed in cartographically representing the decisive climatic factors for an animal or plant species” (Lindroth, 1939a, p. 245). Perhaps in the future we may deepen our understanding of the climatic conditions necessary for the life of animals and plants, apart from the insight obtained by microclimatic measurements, by using the “middle way,” even if not the “ideal way,” of the lococlimate. Here Krogerus (1937) has led the way with his temperature and humidity measurements on the south and north shores of the Paanajarvi Lake (Ks). I believe similar, extended investigations of the so-called southern mountains in northern Fennoscandia can yield good results, as already shown to a limited extent by Frodin (1915). The measure- ments should be taken in sequence with instruments placed close together (100 m or less) in extremely different positions, and the temperature and hu- midity of the air recorded at the usual height (about 1.5 m off the ground), in the air layer next to the ground and in the ground itself. Regional tempe- rature measurements in the water close to the shore of different lakes would also provide a valuable clarification of the thermal conditions, particularly in the case of terrestrial and semiaquatic shore fauna. The limnologists and oceanographers are generally in a more favorable situation. For them the air is replaced by water, which has a much more stable character. Indirect Evidence of Climatic Factors Without directly comparing the area of distribution of animals and plants with different climate lines it would be possible to undertake a parallel study thereof with biological manifestations of another kind which must be climatically de- termined, apparently showing the areas to be climatically determined too. Phenological observations are especially suitable for this purpose. They have been carried out in the Nordic countries since time immemorial, especially on the annual course of development of the more common species of plant. Some observations have been published. Examples of this kind of parallelization are Siivonen’s comparison (1942) 502 between the growth of Bombus species and the arrival time of the cuckoo (Cuculus canorus) in spring, and the calculation of the mean flowering period 443 ‘(REP6l ‘YIOAPUIT 0) Zuıp10asy) "„JOJd PIOD,,—9 :,.JoJd wem ,—qg ‘oyxe] UI—e "Ir6l “ZI-1T “oye uolskgsg ‘wioysinfg Idn Je wo ¢ Jo yIdap je aunye1sdwus] Jo saaino-Aeq “OS welseIq m) = Pe oGl sinoy b C 0 cc 07 8 91 vL cl ol 8 9 sinoy 00S 503 444 of Anemone nemorosa in recent decades in Finland. This was believed to be an indication of climatic improvement and was correlated with enlargement of the areas of a number of birds (Siivonen and Kalela, 1937; Kalela, 1938, p- 239). It seemed possible to arrive at an idea of the annual vegetative period in a similar way, that is, on a phenological basis. Locations with the same length of vegetative period could be joined by isochrones to yield maps that could be compared with those on the mean temperature and the minima of the “critical” months (Figs. 68-71) and with the distribution maps of animals. I have put in a lot of work to produce such phenological maps for Sweden (according to Arnell, 1923, 1927, 1930). The difficulties are enormous. It is easy to find a suitable phenological manifestation that signals the arrival of spring: I chose the first flowering of Anemone nemorosa. But a corresponding indication for autumn is scarcely to be had, at any rate not from wild-growing plants. The events at this time of year tend to proceed more gradually and are hardly noticeable. By way of experimentation I utilized the leaf fall of Fraxinus excelsior and Populus tremula. For every observation point I calculated the mean length of the period between the flowering of Anemone and the leaf fall of Fraxinus and Populus, and these figures were mapped. At any rate the map of “Anemone-Populus” showed a certain regularity in the arrangement of figures of the desired vegetative period. Its length in south Skane is about 180 days, in the Malar region about 160 days, in Mdp- Ang about 140 days; in the south Swedish highland there is a minimum region with <160 days. But the gradations are extremely irregular and it was not possible to draw regular, “reliable” isochrones. Evidently the phenological primary material available to date in Sweden is too erratic for the desired objective. If only series of simultaneous obser- vations of sufficient duration (> 10 years), which would be especially impor- tant in view of the climatic changes (pp. 641 ff.), are utilized, the network of stations is so sparse that no isochrones can be constructed. The decision as to which stage of the observed manifestation should be considered is always left to a certain extent to the subjectivity of the observer. So is the question whether he should or should not consider lococlimatically especially favored (or unfavored) localities. My experience discouraged me from publishing any phenological map on the duration of the annual vegetative period. The available material must be considered inadequate for this purpose. Kaikko (1940) tried to determine the mean annual vegetative period (the time between the development of leaves and leaf fall) of birch (Betula, ex- cluding nana) in Finland, and to depict it cartographically by isochrones. The map shows the most favorable figures in the extreme southwest, then in the isthmus of Karelia and in a small region around the southernmost part of Paijanne. Numerous phenological maps of plants and birds in Sweden are 504 445 given by Edin (1941), but they are very generalized and cannot be used for the present purposes. Soil Animals that are as soil-bound as carabids—with very few exceptions— must be dependent to a great extent On characteristics of the soil. This is especially true of species with burrowing habits even in the adult stage, foremost: Amara infima, Bembidion pallidipenne, Broscus cephalotes, Clivina fossor, Cymindis macularis, species of Dyschirius (at least D. globosus and D. helleni), numer- ous species of Harpalus (for example, A. anxius, H. hirtipes, H. melancholicus, H. neglectus, H. rufitarsis, H. rufus, H. servus, H. smaragdinus, H. tardus), and Omophron limbatum. The larvae have subterranean habits to a still greater extent. Especially familiar are the tunnels of Cicindela larvae. The characteristics of the ground that are biologically effective—often called edaphic factors—are primarily humidity, physical features (especially ther- mal characteristics and particle size) and chemical features. The moisture of the ground is the combined effect of climatic and purely edaphic factors. It was briefly considered above (p. 485). As there empha- sized, the ground moisture is primarily an ecologically effective factor, which has a decisive role in the distribution of every species in the landscape (for example Bro Larsen, 1936, pp. 210 ff.) and in their more or less regular move- ments during the course of the year (see H. Krogerus, 1948, pp. 126 ff.). In an attempt to determine more precisely the ecological character of the “lime- stone species,” and twice in other contexts (pp. 61, 357), various experiments with humidity were carried out. The absence of some xerophiles from west- ern Norway (p. 488) can perhaps be attributed to excessive ground moisture on account of heavy precipitation. Conversely—but also due to the mechan- ical characteristics of limestone gravel (p. 128)—the xerophiles are obviously favored in parts of southeast Sweden, especially on Oland and Gotland. How- ever, deficient ground moisture cannot completely eliminate a species from any larger region of Fennoscandia. Zoogeographically—with respect to area limitation, evident from the distribution map—the ground moisture in our region is therefore a factor of lower rank and will not be further considered here. The thermal characteristics of the ground were also considered elsewhere (pp. 177 ff.) in fair detail (see also p. 513). Hence chiefly two groups of characteristics of the ground remain to be considered here, namely, the importance of particle size and of chemical composition. In Fennoscandia no carabid is associated with outcrop rock (cf. p. 529)—its indirect thermal importance was considered above (p. 186)—and we will therefore focus here exclusively on loose deposits. 505 507 446 Size of Soil Particles The coarser or finer nature of granules determines not only whether an animal can move on the surface but also whether it is capable of burrowing into the ground; it is also secondarily of critical importance for various other characteristics of the ground. Ecologically of greatest importance is the water- holding capacity, which with decreasing particle size—especially around the size 0.2 mm—enormously increases (Atterberg, 1903; Ekstrom, 1927, pp. 17 ff.; Stebutt, 1930, pp. 111 ff). The thermal conductivity also increases thereby. Krogerus (1932, p. 160) concludes that “fine-grained sand provides the or- ganisms with much more favorable ecological conditions than coarse-grained sand,” and that “a grain size averaging 0.2 mm represents a limit between the ecologically favorable and unfavorable types of sand.” In the few experiments I carried out in the substratum gradient apparatus with sand of different particle sizes (Experiments 94-103, p. 83) the above secondary characteristics of sand had no role. The samples were kept as uni- formly moist as possible, and on account of the short duration of exposure (overnight) the water content could not have appreciably altered during the experiment. Uniform room temperature compensated for any thermal differ- ences between the sand types. The species studied all have a more or less burrowing mode of life. It was therefore intended to investigate whether the purely mechanical characteristics of different types of sand based on particle size, exercise any influence on the carabids (on their burrowing ability). Un- fortunately it was not possible to obtain all six kinds of sand from the same native soil (1 and 2, as well as 5 and 6, nevertheless have the same origin), so the mineral characteristics were not exactly the same (1-2 showed traces of limestone, but not the others). Of course the above detailed treatment seems to have revealed that at least the species of Harpalus are extremely insensitive to the composition of the soil (pp. 121 ff.). The distribution of the carabids in these experiments gives no reason for assuming that they are influenced by such differences. The experiments with Harpalus (Experiments 97 ff.; Diagram 51) revealed fairly large differences in the response of the species tested. On account of the usual inclination to gather at the ends of the apparatus (p. 73), the diagrams are of course to be utilized only relatively, not absolutely (not as proof of the most preferred sand type). However, it is clear that all species preferred the finer sand types with the exception of H. serripes. The diagrams are arranged from this viewpoint. The position of H. serripes corresponds to its occurrence in nature. Among the species tested it is the only one that regularly lives on coarse gravel. The sequence of the other species could probably be corrected by using more material. The vacillating behavior of H. neglectus (50 specimens) is surprising; the extreme position of H. rufitarsis (only 35 specimens) is un- certain. On the other hand the sequence H. tardus—H. anxius—H. smaragdinus 508 447 goes well with the ecological characteristics of these species. The two species Dyschirius obscurus and D. thoracicus were studied in a circular gradient apparatus (Experiments 95-96), and the choice of sand type may thus correspond better with reality (Diagram 52). It is obvious that both species avoid the coarsest sand (2-1 mm). That this is less obvious in D. obscurus is undoubtedly due to the fact that this species tolerates the finest sand (< 0.075 mm) better than thoracicus, and that in the apparatus the sectors with the finest and the coarsest sand were adjacent. In general, as compared with D. thoracicus, D. obscurus evinced only slightly the expected stronger attraction to the finer sand types (“mean box” 3.96, as against 3.84 ın D. thoracicus). Unfortunately the carabid material available was too meager. Both species show a distinct maximum with the particle size 0.125-0.075 mm. Krogerus (1932, p. 164) found that D. obscurus in nature prefers sand types < 0.2 mm. The experiments with Harpalus and Dyschirius seem to provide the ex- pected result that the particle size of the substratum has an important influence on the choice of biotope for some carabids, irrespective of secondary differences in moisture, temperature, etc. Ellinor Bro Larsen’s contrary findings (1936, pp. 220 ff.) led the author herself to state that in the region investigated (West Jut- land) there is insufficient variability of particle size with corresponding mois- ture. This state of affairs is actually known to every field entomologist. It is nowhere more evident than along the banks of a large river. Along with usually very smooth transition in the structure of the bank material, from stone and coarse gravel on the rapid upper course through finer and finer sand to the pure loam at the mouth of the river, there is a parallel change in the fauna of the banks. This is true not just of the burrowing forms but also of species that constantly stay on the surface, including numerous carabids. A few examples of riparian species of fresh water whose dependence on shore material of a particular coarseness (categorization after Atterberg, 1905) seems to be clear, are listed here. The classification is based on estimates, since no measurements of particle size were undertaken in nature. a. In rocky and stony places Nebria gyllenhali b. On gravel (pebbles) 20-2 mm in size (Bembidion fellmannı) B. saxatile (B. hasti) B. tibiale B. hirmocoelum B. virens (B. hyperboraeorum) (Perileptus areolatus). B. prasinum Species in parentheses live more or less regularly on coarse sand. ‚un . | 0% 208 Mi = 20ER N oy A 20_ fl = | (@) =e 30 ih 20m ith 10_ 2 a. f ov, Ml aE 300 er " \ S 4 108 -g 0% BEN a 2 3 4 5 6 Box number 1 506 Diagram 51. Distribution of 7 species of Harpalus according to different particle sizes of sand in the substratum gradient apparatus. Types of sand are: 1 = 2-1 mm; 2 = 1-0.5 mm; 3 = 0.5-0.25 mm; 4 = 0.25-0.125 mm; 5 = 0.125-0.075 mm; 6 = < 0.075 mm. a—Harpalus serripes; b—H. neglectus; c—H. hirtipes; d—H. tardus; e—H. anx- ius; {—H. smaragdinus; g—H. rufitarsis,; Experiment 97 ff., p. 84. 449 30 — 20, = 2 1O.e i 7 = | 2 3 4 5 6 Box number 1 507 Diagram 52. Distribution of Dyschirius obscurus (black) and D. thoracicus (white) according to different particle sizes of sand in the circular gradient apparatus. Experiments 95-96, p. 83. Types of sand as in Fig. 51. c. On coarse sand, 2-0.6 mm in size Agonum ruficorne B. velox. Bembidion lapponicum No species is bound to a particular size. Agonum seems to require an admixture of loam, and the two species of Bembidion also occur on common sand. d. On “common” sand, 0.6—0.2 mm in size (Bembidion andreae polonicum) B. ruficolle (B. argenteolum) (Cicindela maritima) (B. pallidipenne) (Dyschirius thoracicus). 2B. repandum 509 Almost all species—those in parentheses—also live on fine sand; Bembid- ion andreae likes places with an admixture of loam. | e. On fine sand (“mo”), 0.2-0.02 mm in size Asaphidion pallipes D. intermedius Bembidion litorale D. nitidus B. semipunctatum D. obscurus B. siebkei D. politus (B. stephensi) (Nebria livida) Dyschirius angustatus (Omophron limbatum). D. impunctipennis The 3 species in parentheses require—or at any rate prefer—soil with a loam admixture. Dyschirius obscurus also lives, in smaller number, in common sand. f. On silt (“mjala”) and loam, particle size less than 0.02 mm 450 (Acupalpus dorsalis) B. nitidulum Agonum marginatum (B. properans) (A. piceum) (B. transparens) Asaphidion flavipes (B. ustulatum) Bembidion aeneum B. varium B. articulatum (Chlaenius nigricornis) (B. assimile) C. vestitus (B. dentellum) (Dyschirius aeneus) (B. femoratum) D. lüdersi (B. guttula) (D. septentrionum) B. illigeri (Pterostichus gracilis) (B. lunatum) (Stenolophus mixtus). (B. lunulatum) Species in parentheses require only a more or less distinct admixture of loam. This may also be true of Agonum dolens and Pelophila borealis, which are not mentioned. Nonriparian species are considered below. The zoogeographical—area-limiting—significance of a species that is bound to ground material of a particular particle size naturally depends on how widely the necessary soil type is distributed in the region concerned. Hence two maps are given here (Fig. 77): one on the distribution of sand and fine sand (particle size 2-0.02 mm; according to Atterberg’s system, 1903, 1905); the other (Fig. 78) on the distribution of silt (“mjala”) and loam (particle size less than 0.02 mm) in Fennoscandia. Sand (including fine sand)—although often locally limited—shows a fairly uniform distribution in the Fennoscandian region. The large gaps in west and north Norway, and on the Kola Peninsula, are undoubtedly partly due to inaccurate data. Regions especially poor in sand are found in East Smaland, in the Malar Lake region, and in west-central Finland. In the north, the sand is largely limited to the middle and lower reaches of the larger rivers. It is scarcely possible to cite a species whose dependence on sand is re- flected in an area corresponding in every detail with the occurrence of sand on the map (Fig. 77). Because of its small scale the map is too generalized, and very small isolated sandy places are not marked at all. Dyschirius politus seems to be the best example of a correspondence, at any rate in principle. In addition to its absence from west Norway this species shows gaps in the above-cited three regions in Sweden and Finland. Roughly the same gaps in distribution are evident in the following species, which are more or less closely bound to sand: a. Missing from west Norway Asaphidion pallipes B. velox Bembidion litorale Cicindela maritima. B. semipunctatum 451 510 Fig. 77. Distribution of sand and fine sand (“mo”) in Fennoscandia. From Atlas öfver Finland (1910), Thunmark (1937), Lundgvist (1942), Sahlström (1944-48), and unpublished data of Lundqvist. Conditions in Norway and in Russian parts are broadly generalized. 511 Fig. 78. Distribution of silt (“mjäla”) and loam (including moraine clay) in Fennoscandia. Sources as in Fig. 77. 513 453 It is nevertheless difficult to decide whether this absence in the west is edaphically determined. For Amara fulva, mentioned above (p. 488), which shows a striking predilection for sand—but is not completely dependent on it—climatic factors are believed responsible for its western limit. Insufficient summer heat, excessive humidity and lack of sunshine can be as effective area- limiting factors for true sand-dwelling animals as lack of sand (probably more correctly: the rare occurrence of sand) in west Norway. The above 5 species are distinctly heat-requiring (for Cicindela see Krogerus, 1932, p. 146; 1937, p. 299) and heliophilous (less distinctly Asaphidion). b. Missing from East Smaland | Asaphidion pallipes Nebria livida Bembidion litorale Omophron limbatum. B. velox Especially for Omophron the restriction to western Smäland is very strik- ing. It is scarcely possible to explain this characteristic in any other way than by the rarity of suitable biotopes in the east. c. Missing or sparse in central Swedish loamy region (especially around Malar Lake) Asaphidion pallipes Pterostichus lepidus. Cicindela silvatica The isolated records of the two last-mentioned species in the region in question were all made, significantly, on sandy diluvial gravel ridges (“rull- stensäsar”), which show up on the “sand map” as narrow bands. d. Missing from west-central Sweden Bembidion nigricorne Cicindela maritima ? B. ruficolle Dyschirius obscurus. However, in all these four cases it is fairly evident that factors of immigra- tion history have also been effective in the Finnish distribution (see pp. 718 ff.). The above-mentioned big gap is not due just to existence factors. But any region where the biotopes suitable for a species are sparsely situated has an obstructive influence by slowing down the dispersal process. Bembidion nigricorne rates special interest because this species occurs as an exclusively sand-dwelling animal only in northern Europe (probably also in the British Isles). In central Europe it also lives on more or less moist bog soil, ie. in the company of B. humerale. It is difficult to explain this behavior in any other way than that in cooler regions the carabid looks for as warm habitable biotopes as possible, which is proof of the thermally favorable characteristics of sand. Its thermal conductivity is much greater than that of bog soil (Geiger, 1942, p. 30), so the temperature in the sun must be higher. Even sand-loving Bradycellus harpalinus is less choosy in central Europe. It 514 454 has already been shown that southern insects in central Europe may become stenotopic sand-dwelling animals near their northern limit (Kühnelt, 1934, p- 122). \ In the same way the peculiar “double” occurrence of Demetrias mono- stigma and Dromius longiceps is perhaps microclimatically determined, and to a lesser extent that of D. sigma. Especially in the case of Dromius longiceps, it is evident that the predilection for sand increases in northern Europe. Like- wise, Dromius melanocephalus and D. nigriventris are less distinctly xerophilous (“psammophilous”) in central Europe. Concerning the association of the ar- boricolous species Dromius angustus with sand, see below (p. 543). Loam (clay, including silt = mjala) in Fennoscandia has a far more char- acteristic occurrence than sand (Fig. 78). It is concentrated in regions covered. by the sea (and the Ancylus Sea) during the postglacial period. In addition, in south Sweden—especially in the southwestern half of Skane—there is “clay- marl” (moraine clay). Otherwise, above the highest shoreline there are only very small, isolated loamy areas. It should therefore be easier to establish a correspondence between the “loam map” and the Fennoscandian distribution of one or other “loam species,” and to deduce a causal connection. This applies especially to the following species (less evident for those in parentheses): (Acupalpus meridianus) B. illigeri Agonum marginatum (B. varium) (Amara famelica) (Dyschirius aeneus) Bembidion articulatum D. ludersi. B. assimile In Sweden the avoidance of the south Swedish upland on the western side is especially characteristic for these species. One should remember that these clay-deficient regions of south Sweden are also climatically very unfavorable, chiefly thermally (pp. 463, 474 above). But the area of Bembidion articula- tum or B. varium, but especially of Dyschirius ludersi, in Finland extending far northward shows that at least in the case of these species, absence from the south Swedish upland can scarcely be due to climate. On the contrary, these represent examples of species whose Scandinavian area has developed primarily due to edaphic factors. In Skane, where the southwestern and northeastern halves of the province are counterparts for the distribution of loam and sand respectively, the influ- ence of the two kinds of soil can be studied in greater detail. Loam-dwelling species are occasionally restricted to the southwestern, loamy half: Bembidion assimile B. properans B. obtusum Chlaenius vestitus. 515 516 455 A larger number of species—not necessarily bound to loam, but not found on pure sand—are missing (as far as is known) at least along the sandy east coast of the province: Anisodactylus binotatus Cicindela campestris Asaphidion flavipes Dichirotrichus pubescens Bembidion aeneum Dyschirius salinus B. minimum Trechus rubens. This is partly true even of species that are otherwise abundant almost everywhere and are more or less ecologically distinct ubiquists: Amara ingenua Pterostichus oblongopunctatus Bembidion gilvipes P. vernalis. Even Carabus granulatus and C. violaceus as well as Pterostichus coer- ulescens and P. cupreus, are poorly represented in eastern Skane. Although these regions have not been thoroughly explored for a long time it is difficult to explain these features otherwise than by the edaphic conditions. Loam exerts an indirect influence, since lakes in loamy ground become more or less markedly eutrophic (s. lat.) (compare the map on the lake districts of south Sweden in Naumann, 1932, p. 45, with the “loam map” above, Fig. 78, p- 511). Several carabids are bound to the shores of such lakes. However, it may not be the loam that is effective (at any rate not the particle size) so much as the thermal or chemical characteristics of the water and of the shore material. The question is partly considered below (p. 528). In loose mineral deposits of nonfluvial nature—primarily in moraine—par- ticles of the most diverse sizes are usually randomly mixed, and any correlation of the animals living there with definite sizes of particles is dubious. One rea- sonis the often low requirements of the animals for material of definite particle size. Especially in more or less pronounced xerophiles, it also happens—even if the soil consists of fairly uniform sized particles—that they may tolerate very diverse particle sizes, if only the humidity is suitable. As examples of such more or less eurytopic xerophiles the ne species may be mentioned: Amara aenea Calathus erratus A. bifrons Cicindela hybrida A. consularis Harpalus fuliginosus A. equestris H. melancholicus A. infima H. tardus A. praetermissa Masoreus wetterhalli. A. quenseli f. typ. Notiophilus pusillus A. tibialis Olisthopus rotundatus Broscus cephalotes Pterostichus lepidus. 517 456 Chemical Properties of the Soil In the section on “limestone species” (pp. 121 ff.) no purely chemical effect of limestone rock on the species studied (chiefly those belonging to the genus Harpalus) was established. However, this does not justify saying that the cara- bids are generally insensitive to the chemical composition of their substratum. Special attention has been drawn to the apparent dependence of some in- sects on common salt (NaCl). A distinction has been drawn between halobiont (salt-requiring) and halophilous (salt-loving) animals (Schaum, 1843, p. 180; Benick, 1926, p. 65; Lengerken, 1929a, b). The source of salt is the sea. Irrespective of this, the conclusive evidence of a species being halobiont is not that it occurs only on the seashore but that inland it is restricted to isolated saline places. It must be experimentally investigated whether Acpus marinus, which throughout its area never occurs away from the seashore, really owes its habits to NaCl. On the other hand the halobiont nature of Dichirotrichus pubescens is evident from the distribution alone. In Fennoscandia there are no inland places with NaCl (on the other hand there are “alaun soils” with SO,, CaO, Al,O,, etc; Aarnio, 1922). Hence in our region halobiont species must live exclusively, and halophiles at least predominantly, on the seashore. There are only 5 quite unambiguous halobionts in our region (in central Europe they also inhabit inland saline places): Anisodactylus poeciloides D. salinus Dichirotrichus pubescens Pogonus luridipennis. Dyschirius chalceus In addition there are Aépus marinus and Trechus fulvus which are bound to the seashore throughout their area (however, there are different subspecies of Trechus living in caves in the Iberian Peninsula), but their dependence on NaCl is not proven. Amara convexiuscula (see below) is doubtful. The following species might be suspected of being halophiles: Bembidion aeneum B. pallidipenne B. fumigatum Dyschirius impunctipennis B. minimum D. obscurus. Doubtful cases are Agonum archangelicum, Bembidion chaudoiri, and B. repandum, which occur within the region only on the shores of the White Sea. Their distribution and ecology are not fully known. Sufficient information has been provided in Part I of this work on further species which have been unjustifiably considered as halophilous or halobiont. To judge the salt requirement of the species in question a map of the salinity of the surface of the Baltic Sea in the Kattegatt and Skagerrak (Fig. 79) 519 457 may serve as the starting point. According to their tolerance of slightly saline water the above 5 halobionts can be arranged in the following order by their saline requirement: 1. Pogonus luridipennis. Also in Germany (aside from inland records) only on the North Sea coast; the old records from the Baltic Sea coast are doubtful (Horion, 1941, p. 186). 2. Dyschirius chalceus. In Sweden—like the following species—as far as the southwestern edge of Skäne. But the species is missing from the southern Danish islands, and is recorded from the German Baltic Sea coast only by a single specimen near Travemünde (Horion, 1941, p. 99). 3. Anisodactylus poeciloides. Also in Lolland and Falster as well as on the German Baltic Sea coast, eastward at least as far as Warnemünde (Horion, 1941, p. 250). i 4. Dichirotrichus pubescens. The record in Ogl may be accidental (trans- ported with ballast ?), but the more or less continuous distribution extends at least to central Gotland. The gap in eastern Skane and Blekinge is fully explicable by the absence of loam. Along the German Baltic Sea coast as far as East Prussia (Horion, 1941, p. 244). In the White Sea the saline content still in Onega Bay is more than 20%,. (Knipovitsch, 1906, pp. 1171-1174) and its occurrence there is thus quite natural. 5. Dyschirius salinus. The only species of the five halobionts that reaches the Finnish Baltic Sea coast. In Sweden it is unknown north of Sma Kalmar and Gotland. This area limit, so unevenly situated on opposite sides of the Baltic Sea, is explained by the course of the isohalines (Fig. 79). Bembidion minimum (see below) is of the same distribution type. A very good correspon- dence between the course of the isohalines and the distribution of Crambe maritima in the same regions of Finland was found by Eklund (1931, pp. 101, 125). Amara convexiuscula has been designated halophilous by all the authors cited above (with the exception of Schaum, 1843; also by Hak. Lindberg, 1931, pp. 148, 164). Inland in central Europe it does not occur only at saline places but also “on scree-slopes close to cities.” Horion (1941, p. 271) assumes that “these ground locations show a certain salt content.” It is quite possible that this species is dependent on substances other than the NaCl present in sea- water, but this can be decided only experimentally. Backlund (1945, pp. 108 ff.) has shown that some animal species of seaweed banks in preferendum experiments respond positively to iodides and nitrates, but this has not so far been established for insects. The 6 suspected “halophiles” mentioned above form a very heterogeneous group. They share the feature that they live chiefly (Bembidion fumigatum exclusively in our region) along the seashore, but there are records from the shores—as far as can be judged—of bodies of fresh water (of the cited Bem- bidion in Denmark and central Europe). ef a a x ae 3 Do } Un UN = 2 ce 518 Fig. 79. lini A of one-half ni er ae es 520 459 Bembidion pallidipenne, Dyschirius impunctipennis, and D. obscurus occupy a special position in that their occurrence on fresh-water bodies is rare, yet so regular that there can be no question of chance (however, D. impunctipennis occurs in our region constantly only on Lake Ladoga; also in Germany, Horion, 1941, p. 99 and Latvia). It must be concluded that for these 3 species what is decisive is not the NaCl content of the soil but the presence of suitable sandy soil of a definite particle size, and possibly also definite hygric and thermal characteristics, chiefly on the seashore. For the species of Dyschirius an important role is played by the frequent mass occurrence of its normal prey, Biedius arenarius Payk., where it lives. Both Hak. Lindberg (1931, pp. 154, 164) and Krogerus (1932, pp. 171, 173) believe that the two species of Dyschirius are not bound to NaCl. Bembidion pallidipenne occupies the same position. Hence in these cases there is no reason to speak of “halophily,” since this word means the species “loves” common salt (NaCl), not the sea. The three remaining “halophiles,” all belonging to the genus Bembidion, behave differently. B. minimum in our region almost always inhabits the seashore (On loam). The only freshwater record, which was not purely accidental, is from Oland, where the species to some extent corresponds to B. aeneum, considered below. It is noteworthy that the northern limit in the Baltic Sea region in principle corresponds with that of the halobiont Dyschirius salinus, which often supersedes it, such that, following the isohalines, in Finland it extends much farther north than in Sweden. Much the same can be said of the halobiont aquatic plant Ruppia spiralis (Samuelsson, 1934, p. 16). The difference between the two carabids is that Bembidion minimum has - everywhere advanced farther—at the Swedish east coast, in Finland both at the west coast and at the Gulf of Finland. The records along fresh water in central Europe are “very sporadic and rare” (Horion, 1941, p. 156); significantly, the species was once found “at a salt-lick for sheep.” We may conclude from this that B. minimum is not indifferent to NaCl. It is probable that Bembidion fumigatum behaves like B. minimum in its response to NaCl. However, the species is very rare in our region and the records from other regions are not precise enough to go beyond conjec- ture. It might be worthwhile, however, to study the distribution of B. aeneum more closely (considered on p. 399 from the dynamic viewpoint). At first sight the distribution of this species in its Fennoscandian area does not suggest “halophily.” There is continuous distribution along the west coast but there is also a broad, slanting belt across central Sweden; inland records are also known from southeastern Norway and from Öland. B. aeneum might thus be presumed to be a loam-dwelling species, and it is in fact bound to loamy soil. But— with the exception of the record of a single specimen near Uppsala—why is it missing from the Malar Lake region? This cannot be a dynamic (temporally determined) limit, since southeastern Dir was reached. 521 522 523 460 The work of S. Johansson (1926, p. 21) led me to venture an explanation. He studied the chlorides in the tapwater of the city of Skara in Vgl and established that the striking amount thereof is due to the presence of NaCl in the loamy soil (found at least at a depth of 1.5-9.5 m; samples from the top layer were not studied). Since Yoldia loam is present here, Johansson surmised that the common salt originated from the seawater in which the loam formed a deposit. The map on the distribution of Yoldia loam (Fig. 80) shows an inter- esting correspondence with the inland distribution of Bembidion aeneum in Scandinavia. This species differs from other loam-dwelling species, such as B. articulatum, B. illigeri, etc. (p. 509) primarily by its almost complete ab- sence from the eastern part of the central Swedish loam region (cf. loam-map, Fig. 78, p. 511), in regions that were submerged even during the Ancylus period, that is, where the saline Yoldia loam was covered with freshwater deposits. * A broadly corresponding distribution in Sweden is shown by Elatine hexan- dra (Samuelsson, 1934, p. 85). It would be worthwhile investigating whether a certain amount of NaCl in the water bodies has a positive role for this plant too. The following details of the map (Fig. 80) are especially striking: 1. In Skane: All 5 records lie below the highest marine boundary, even though extensive loam regions (moraine clay) are also present above it (Fig. 78). 2. In central Sweden: There is one single record (Upl Uppsala, 1 specimen)—possibly of an accidental nature—from the middle of the “Ancylus region.” Constant populations exist partly in the “Yoldia region” and partly on the western limits of the “Ancylus region” (Nke Orebro; Dir Ludvika and Hedemora), where the freshwater deposits must be sparse. 3. The northernmost inland records of Bembidion aeneum (in southeastern Norway, in Vst and Dir) are immediately south of the so-called K-line, which according to Munthe (1940, pp. 79-83) represents the southern limit of inland ice at the end of the Yoldia period. But this line does not signify the limit of loam (see map, Fig. 78), so is not an existence limit for B. aeneum. Given these facts, it seems to be justified to assume that B. aeneum requires a very low NaCl content in the substratum. The only records in the region that contradict this are from the alvar areas of Oland, the southern two-thirds of which was submerged in the postglacial period only during the time of the Baltic ice sheet, thus during a freshwater period. The effect of NaCl on an animal species should of course also be studied experimentally, most simply by preferenda experiments. Such studies have been carried out by Ellinor Bro Larsen (1936, p. 200) and Backlund (1945, p. 105). The first found a distinctly positive response to NaCl in two species of Bledius from the seashore. *The latter Littorina deposits seem to play no positive role for Bembidion aeneum. 521 461 Fig. 80. Bembidion aeneum. Scandinavian inland records. Blank circles—one specimen only. Regions submerged by Yoldia Sea (south of ice edge, k, of that time) are vertically hatched. Broken lines: parts later covered by Ancylus Lake. Geological limits taken from Munthe (1940). It was especially interesting to study Bembidion aeneum experimentally. Unfortunately little material of living carabids was available to me, but the experiments were repeated until 50 observations were complete. B. mini- mum was selected as the species for comparison, for which—as mentioned above—dependence on NaCl was to be expected. The material of both species originated from the seashore near Boh Samstad. To find a suitable saline content for the following alternating experiments, first of all B. minimum alone was studied in a serial experiment with NaCl of 524 462 different concentrations (Experiment 93a, p. 82). Result: Spring water US AIRPORT EZ Nae total No. of specimens 1 3 0 3 8 4 19 In the alternating experiments I used cuvettes containing sand, moistened alternately with spring water and 1% NaCl (Experiments 92a, 93b, p. 82). Both species were tested together: Material from Spring Boh Sämstad water 1% NaCl Total Specimens Bembidion aeneum 16 34 50 B. minimum 17 33 50 Bembidion aeneum apparently shows the same attraction to NaCl as B. min- imum. One source of error was that the material originated from a saline locality (the seashore). It is conceivable that this population would behave in a different way to NaCl than carabids from the inland, either on account of genetically determined “race characteristics” or due to alternative adaptation. The material comprised brachypterous as well as macropterous individuals, whereas in the inland only the macropterous form occurs (see map, Fig. 49, p. 400). Hence it was important to study macropterous inland specimens too, and with difficulty I obtained a small number of B. aeneum from Old Möckelmossen. The experiments (Experiment 92b, p. 82) gave the following results: Material from Spring Öld Möckelmossen water 1% NaCl Total Specimens Bembidion aeneum 9 41 50 Hence the macropterous inland form of B. aeneum shows at least as strongly positive a response to NaCl as the dimorphic coastal form or as B. minimum. This experimental result therefore confirms the “cartographic” concept that the Scandinavian inland distribution of B. aeneum is dependent on Yoldia loam. h The unexplained exception is the occurrence on the Alvar region of Öland, which in the postglacial period—according to geologists—was underwater only in the time of the Baltic ice sheet (a freshwater period). I cannot decide whether physiologically NaCl can be replaced by CaCO, (however, B. min- imum was also found in Öld at a “nonsaline” site); a similar occurrence is shown by the hydrophilid Ochthebius marinus Payk. (Häk. Lindberg, 1948, 525 463 p- 159). If the effect of NaCl on the animals is mainly of osmotic nature, this would be possible. Analysis of a loam sample from the locality Möckelmossen, however, revealed a slight Cl content (1 per mille). The actual physiological significance of NaCl for halobionts can be determined only on the basis of careful experiments. At any rate it seems to me improbable that there would be animals that “love” NaCl without requiring a certain minimum quantity of it. Future de- tailed explorations will doubtless show that the heterogeneous group of the “halophiles” is divisible into two sections: 1. Halobionts, requiring NaCl, which are satisfied with such small quan- tities that they can live away from the seashore at virtually nonsaline sites. Examples: Bembidion minimum, B. aeneum, probably B. fumigatum, possibly Amara convexiuscula. 2. Species indifferent to salt, which find conditions suitable for life on the seashore for other reasons (possibly also at other saline sites). Examples are animals of sandy ground (“psammophilous” species) such as Dyschirius obscu- rus, D. impunctipennis, Bembidion pallidipenne, and Cicindela maritima. Other animals may be associated with halobiont plants or animals for food habit requirements (examples have been provided by Krogerus, 1932, pp. 173-174). At the area limits, the microclimatic effect of the sea also has a role. The oc- currence of Bembidion saxatile in central Europe, and of Nebria gyllenhali on Gotland, can be explained in this way. Finally, it is conceivable that salinity- tolerant animals on the seashore, as has long been assumed of the “halophytes” among the plants, enjoy the reduced competition, and that at localities with saline content they are free from some other natural enemies (animals, fungi, bacteria) (examples have been provided by Ellinor Bro Larsen, 1936, p. 207). Evidently salts and chemical compounds other than NaCl may also exer- cise a positive effect on carabids. Of course it may be pointed out that more detailed analysis of the importance of CaCO,, (pp. 120 ff.) failed to reveal any unambiguous chemical effect on the insects. Concerning the importance of the “nutrient salt content” of the ground—at any rate with regard to the Fennoscandian carabids—I have come to a different conclusion than Holdhaus (1911a, 1911b). In a different context I had the opportunity to carry out a series of experiments which also touched on this complex of problems. A number of clearly synanthropous species have greatly extended their area (See p. 637) during the last decades—chiefly northward—and we cannot attribute it just to enhanced possibilities of passive dispersal. For one of these species, Amara fusca, it was assumed in central Europe that “the more abundant occurrence of this species is associated with the increasing use of certain synthetic fertil- izers” (Nurnberg in Horion, 1941, p. 261). Since I surmised that similar fac- tors would apply to the definite area displacement of Amara ingenua (p. 630) (Part I, p. 140) I decided to investigate the problem experimentally. 526 464 Amara ingenua was studied in the substratum gradient apparatus with the usual synthetic fertilizers, at first in serial experiments, where a suitable con- centration of the substance was tried out for the alternating experiments that followed. As control species—used only in the alternating experiments—A. praetermissa was selected, which is shy of cultivation and seems in no way to be favored by it. The experiments here have been arranged according to the substances tested (pp. 80 ff.; cf. also Experiment 107 and p. 74). 1. Ca(NO,),. Serial experiment (Experiment 84a). Distill- ed water 1/4% 1/2% 1% 2% 3% Amara 23 12 6 3 8 5 ingenua 4% 5% 7 1/2% 10% salt Total 6 8 15 14 100 specimens Ca(NO,),. Alternating experiment (Experiments 84b, 89). Distilled 5% water Ca(NQO3)2 Total Amara ingenua 34 32 66 A. praetermissa 66 34 100 2. Phosphates. a. Thomas phosphate (commercial product). Serial experiment (Experiment 85). Distilled water 0.075% 0.15% 0.3% 0.6% Amara ingenua 9 6 6 9 3 1 1/4% 2 1/2% 5% 7 1/2% 10% salt Total 0 1 2 1 7 44 specimens Thomas phosphate is hard to dissolve and correct gradation of low con- centrations was too uncertain, so superphosphate was used. b. Superphosphate. Serial experiment (Experiment 86a). Distilled water 1/4% 1/2% 1% 2% Amara ingenua 11 10 7 6 8 3% 4% 5% 71/2% 10% salt Total 5 6 8 6 3 des specimens 465 Superphosphate. Alternating experiments (Experiments 86b, 90). Distilled 2% super- Total water phosphate Specimens Amara ingenua 36 36 72 A. praetermissa 48 32 100 3. KCI. Serial experiment (Experiment 87a). Distilled WE AP MPI hl ble NP ce ne Amara ingenua 6 12 14 14 6 8 4% 5% 7 1/2% 10% salt Total 9 6 5) 3 83 specimens KCI. Alternating experiments (Experiments 87b, 91). Distilled 1% water KCl Total Specimens Amara ingenua 56 43 109 A. praetermissa 56 44 100 4. Liquid ammonia (commercial product). Two alternating experiments (Experiments 88a, b). Distilled Distilled Distilled water water water Amara ingenua 7 7 1 1% 2% 3% liquid ammonia Total 0 1 1 17 specimens Distilled water Distilled water Distilled water Amara ingenua 7 4 2 1/8% 1/4% 1/2% liquid ammonia Total 0 0 2 15 specimens Since the species showed such a strong aversion to liquid ammonia no comparative experiment with A. praetermissa was required. The experiments show that Amara ingenua does not respond positively to any of the usual synthetic fertilizers, not even ammonia. The only difference that could be surmised from A. praetermissa, which is not favored by culti- vation, is that A. ingenua tolerates the superphosphate better (but without responding positively to it). It seems to be justified to conclude that A. ingenua is not chemically favored by modern synthetic fertilizers, at least not directly, and that its recent 527 528 466 area expansion cannot be explained on this basis. On the other hand the species shows a distinct predilection for certain synanthropous plants with respect to its food habits, which is dealt with below (p. 539). To the extent that these plants are dependent on fertilizers we can of course speak of the indirect effect of these substances on the insect. Another chemical factor of the soil to which a very important role has been ascribed, at least for the growth of plants, is the soil reaction, which means the hydrogen ion concentration expressed as the pH. How far this factor is crucial for the ground fauna too is a problem that has been taken up for more detailed study by Krogerus in his extensive work on bog fauna, yet to be concluded. In a preliminary communication (Krogerus, 1939, pp. 1222 ff.) he mentions a number of bog species that are “stenoionic,” i.e. they tol- erate only a slight variation in the soil pH. He also mentions 3 carabids, Agonum ericeti, Dyschirius helléni, and Elaphrus lapponicus, the first of which was also experimentally studied, where it was found to give a very strong posi- tive response to the acidic substratum. In a personal communication he added Agonum munsteri (somewhat less pronounced) (on the other hand A. consimile is “acidophobic”). The above species rank among the most stenotopic of all Fennoscandia carabids, since they inhabit only bog soil of a particular type. It would be interesting to investigate Bembidion humerale experimentally for this character. My own investigations in this field are restricted chiefly to an experimental study of some more or less xerophilous “limestone species,” which could be expected to respond positively to alkaline soil conditions. This did not prove to be the case (pp. 122 ff.). The insects were found to be totally indifferent to the soil pH, and I have suggested reasons (p. 125) why xerophilous animals might be expected to be so. The observations on the occurrence of Brachynus crepitans and Agonum dorsale in nature confirm my view. At several localities where these species occur together, I electrometrically determined the pH and found a variation from 4.5 (Ogl Mogata) to 7.9 (Old Greby). Even strongly hygrophilous species can show indifference to pH: Ptero- stichus anthracinus lives in Upl Djursholm with a pH of 4.8, and at a locality in Oland (Halltorp) with a pH of 6.3. Individuals that were reared from the egg onward at a pH of 4.8 even showed a slight preferendum for a pH of 7.5 in the substratum gradient apparatus (Experiment 46, pp. 77, 356). Krogerus’ experimental observations (in litt.) also show that most of the hygrophilous carabids are indifferent to pH. Interestingly, he found that a few species are more or less markedly neutrophilic or alkaliphilic. He mentions Agonum thoreyi, Badister peltatus, Chlaenius tristis, and Pterostichus aterrimus. These species are typical members of the shore fauna of eutrophic bodies of water. It would be interesting to study the remaining ecologically corre- sponding species experimentally for the pH factor, chiefly: 529 467 Agonum lugens Bembidion transparens A. moestum Odacantha melanura Badister dilatatus Pterostichus gracilis. With regard to Oodes gracilis I have proved that thermic factors are decisive (Lindroth, 1943a), and the same may hold true for the almost identi- cally distributed Demetrias imperialis. In conclusion, concerning the area limiting significance of the soil reaction, it may be stated that a clear response to the pH is to be expected only in hygrophilous species, and that even among these only a few species will turn out to be “stenoionic.” But even in such cases it must be ascertained whether the hydrogen ion concentration actually represents the decisive factor, or is more an indicator of other characteristics of water, essential for the life of the animals (p. 196). I am not aware of the reasons for Dahl’s view (1928) that various carabids love “humic acids.” Subjectively, I would be more inclined to accept the soil pH of acidic bogs as a “factor” than the response of species to eutrophic (s.l.) lakes, since these bodies of water are also purely chemically characterized by many other peculiarities (see Naumann, 1932, p. 114; Thunmark, 1937, p. 19). A chemical effect of the substratum is also conceivable for the few species that are ecologically more or less closely associated with burned wood. Fore- most among these are Agonum bogemanni and A. quadripunctatum, which live chiefly under the bark of trees damaged by fire, and Prerostichus angustatus, which—although an exclusively ground insect—tends to appear at places lately devastated by forest fires. Without experiments it is not possible to determine the factor responsible for this choice of biotope by the carabids. Here I merely draw attention to the information provided for the species in Part I of this work (and in the Supplement herewith). There are other examples of insects with a strong positive response to the smoke of burning. Linsley (1943a, p. 342), in his observation on the genus Melanophila, writes: “In summary, the buprestid beetles of the subgenus Melanophila s. str. appear to be attracted over long distances by smoke from a variety of burning materials, including wood, oil, mill refuse, smelter products and possibly tobacco. In nature this attraction leads them to forest fires where they normally oviposit in scorched coniferous wood.”! Swedish foresters are familiar with the fact that the capricorn beetle, Monochamus sutor L., often comes flying in swarms while a forest fire is raging. In this section a few words remain to be said on the term petrophily proposed by Holdhaus (1911a, 1911b, 1927-28). He labels petrophilous “such species as live only on bedrock, i.e. on substrata arising from bedrock. Petrophilous species avoid all loose rocks and are therefore missing from deep sand and rubble deep loamy sediments, and with few exceptions from loess and t (Original citation in English; suppl. scient. edit.). 530 468 tegel subsoil” (1911a, p. 728; 1911b, p. 323). * The phenomenon of petrophily is explained by Holdhaus mainly by the strongly fluctuating nutrient content of the ground: “Both, the chemical and the physical characteristics of the ground or of water exercise an influence on the fauna. However, it appears that greater importance is to be ascribed to the chemical factors” (1911a, p. 742; 1911b, p. 342). Among the physical characteristics he considers chiefly the water- holding capacity, but attaches practically no importance to thermal factors. In Fennoscandia, petrophilous animals would be expected in the fje/ds and on outcropping Cambro-Silurian limestone—especially on Oland and Gotland. Among the “limestone species” of the south, to which a separate section was devoted, it might be possible to designate a small number of carabids as “petrophilous” to the extent that they are apparently favored by outcropping limestone rock, for example, Harpalus azureus, H. rupicola, H. serripes, H. ver- nalis, perhaps also Cymindis humeralis. But it was evident as well, that in these cases it is the physical—chiefly thermal—characteristics that are primarily fa- vorable. In the fjelds only one species, Nebria nivalis, comes to mind. It is the only Fennoscandian carabid clearly and exclusively resident in the Regio alpina, where it lives at the edge of perennial snow drifts, chiefly on firm rock or thin moraine. However, the occurrence of this species, in the Abisko region as elsewhere, shows that it makes no distinction between sedimentary limestone rock, chiefly along the western part of Tornetrask (see map, Fig. 9, p. 113) and limestone-deficient eruptive rock, and generally between different kinds of rock. Nebria nivalis cannot be considered as petrophilous (in Holdhaus’ sense); the area-limiting factor for this species is chiefly thermal: the cold, resulting from the perpetual snow (see p. 465). Holdhaus (1911a, p. 734; 1911b, p. 332) himself believes that petrophilous insects are not to be found in Fennoscandia and are probably completely missing from the region.** He thinks that the cause is the destructive effect of Quaternary glaciation and the generally weak dynamics of petrophiles, such that an interglacial or postglacial recolonization was prevented. In view of the detailed results obtained by Holdhaus one is not justified in considering the term “petrophilous” superfluous, as Brundin does (1934, p. 157). It must therefore be assumed that the condition of the stony ground in central and southern European mountainous regions—chiefly in the Alps—exercises a much greater influence on the ground fauna than in our region. On the other hand, I would be convinced only by the positive results of experiments that even for markedly polyphagous animals, such as almost all carabids, the chemical composition of different kinds of rock is decisive. *Palmén (1946, p. 25) has used the word in a different sense. He designated Bembidion hirmocoelum, which lives on stony river banks, “petrophilous.” **Tt is therefore incorrect for Holdhaus (1927-28, p. 948) to include Pterostichus aethiops among the petrophiles. 332 469 Among other comments, Holdhaus’ remark (1911a, p. 733; 1911b, p. 331), that in the Alps and Carpathians “only such moraines as contain numerous very large block fields” possess a definite petrophilous fauna seems to confirm that even in these regions the thermal characteristics of the ground are decisive. I do not see why material transported by ice (not water!), that is, moraine, should be poorer in nutrient salt than the parent rock and the weathered soil, lying undisturbed over it. According to Kuhnelt (1936, p. 12) even purely hygric conditions of the ground may bring about “petrophily.” Food and Feeding Habits Carabidae, along with some clusely related, chiefly aquatic families, are gen- erally included taxonomically in the Carnivora or Adephaga. Hence they have been labeled predatory animals. This biological image pursues them through- out the entomological literature—pure as well as applied—and contrary ob- servations (mostly in the genera Amara, Harpalus, and Zabrus) are generally considered exceptions. I have not undertaken any systematic investigations on the food of Cara- bidae, but data collected from the literature and my own observations provide a picture which corresponds poorly with the usual concept of the predatory character of Carabidae. The positive response of Carabidae to a vegetarian diet is most easily ob- served with specimens in captivity, which can be mostly (or exclusively) fed on bread. There is definite proof of feeding on bread in the case of the following 2,species: |) Acupalpus consputus Bradycellus collaris A. exiguus Calathus erratus A. meridianus C. mollis Agonum dorsale Carabus* auratus A. ericeti C. cancellatus A. krynicki C. coriaceus A. lugens C. granulatus A. ruficorne C. nemoralis A. sexpunctatum Cymindis angularis Amara aulica C. humeralis A. equestris C. macularis A. ingenua Harpalus aeneus A. lucida H. anxius Bembidion aeneum H. azureus B. minimum H. distinguendus Brachynus crepitans H. hirtipes *The Carabus species according to Jung (1940). 533 470 H. melleti H. smaragdinus H. neglectus H. tardus H. pubescens H. vernalis H. punctatulus Lebia chlorocephala H. puncticeps Olisthopus rotundatus H. rubripes Oodes gracilis H. rufitarsis Panagaeus bipustulatus H. rupicola Pterostichus anthracinus H. seladon P. lepidus H. serripes P. niger. The least fastidious are Brachynus and Harpalus serripes. The former lived in captivity up to 17 months, the latter up to 32 months (!) fed only bread. Most of these observations were made incidentally in connection with other experiments. The preponderance of the genus Harpalus in the list is chiefly due to the fact that I paid special attention to this genus for other purposes. Instances where these insects were observed feeding spontaneously on parts of plants* in nature are of greater interest. Further data on the plant part attacked in each case is superfluous here, since sufficient information can be obtained from Part I of this contribution and the Supplement to this part. The above evidence was obtained in the case of 45 species of our fauna (although some of it abroad). Mostly seeds and fruits were attacked; “v” = only vegetative parts of the plant. Only cases where feeding was proven (or as good as proven) were considered: Amara aenea v B. monticola A. aulica v B. pygmaeum A. bifrons v B. quadrimaculatum v A. convexiuscula Calathus fuscipes A. eurynota C. melanocephalus A. familiaris Calosoma sycophanta A. fulva Carabus auratus v A. infima C. coriaceus A. ovata C. glabratus A. plebeja C. nemoralis A. praetermissa v C. violaceus A. quenseli Clivina fossor A. similata v Dichirotrichus pubescens v Bembidion illigeri Harpalus aeneus v B. lampros H. calceatus *In the few cases where especially juicy plant parts (melon, cherry) were attacked, it was perhaps due partly to thirst (for example in Calosoma sycophanta; but in connection with Carabus species see Jung, 1940). 534 H. distinguendus H. griseus H. pubescens H. punctatulus H. puncticeps H. puncticollis H. seladon H. servus 471 H. tardus Pterostichus cupreus P. lepidus P. madidus P. niger P. vulgaris Zabrus tenebrioides. Ninety-nine species were observed to consume animal food in the adult stage. In cases where the evidence was obtained only in captivity the name of the species is given in brackets: (Agonum dorsale) A. ericeti (A. lugens) (A. ruficorne) (A. sexpunctatum) Amara aenea A. alpina A. aulica (A. communis) A. curta (A. equestris) A. familiaris (A. ingenua) A. ovata A. plebeja Anisodactylus binotatus (Badister unipustulatus) (Bembidion aeneum) Bembidion lampros B. litorale (B. nigricorne) Bembidion nitidulum B. obliquum B. pallidipenne B. rupestre (B. stephensi) B. varium (Brachynus crepitans) (Bradycellus collaris) Broscus cephalotes (Calathus erratus) C. melanocephalus (C. mollis) Calosoma auropunctatum C. inquisitor C. reticulatum C. sycophanta Carabus arvensis . auratus . canceliatus . clathratus . CONVEXUS . coriaceus C. glabratus C. granulatus C. hortensis C. intricatus C. monilis C. nemoralis C. problematicus C. violaceus Cicindela campestris C. maritima C. silvatica Cychrus caraboides (Cymindis macularis) Demetrias imperialis Dichirotrichus pubescens Dromius agilis Dichirotrichus pubescens* erarerske *Repetition in the German original—General Editor. 535 472 Dyschirius impunctipennist D. obscurus D. politus D. thoracicus Elaphrus riparius Harpalus aeneus (H. anxius) H. calceatus Harpalus distinguendus (H. hirtipes) (A. neglectus) H. pubescens (H. punctatulus) H. puncticeps (H. rubripes) (A. rufus) (A. serripes) (H. smaragdinus) (A. tardus) L. crux-minor (Microlestes minutulus) (Nebria livida) Notiophilus germinyi Odacantha melanura (Omophron limbatum) (Oodes gracilis) (O. helopioides) Patrobus septentrionis ? (Pristonychus terricola (Pterostichus anthracinus) P. coerulescens P. cupreus P. punctulatus P. vulgaris Tachyta nana Trechus quadristriatus T. rivularis Zabrus tenebrioides. Lebia chlorocephala It is easy to summarize the above list as follows: Animal diet: 99 species, 31 of which only in captivity. Vegetable diet: 85 species, 40 of which only in captivity. Exclusively animal diet: 53 species. Exclusively vegetable diet: 37 species. Mixed diet: 48 species. It is evident from ihese figures that ihe Carabidae generally take veg- etable diet. Moreover, among the 138 species actually studied here no fewer than 48, i.e. 35%, can eat both animal and vegetable food. | am convinced that this represents the normal situation in carabids, and that future more precise feeding experiments will demonstrate the generally omnivorous character of these insects. Another aspect which is very important—not least for applied ento- mology—may be mentioned: Animal diet is not to be equated with predation. It can be easily observed that most carabids dare not attack healthy, uninjured prey; they flee before animals many times smaller than themselves. However, as soon as an insect, such as a worm, is injured, they seem to be attracted by the oozing body fluid and pounce on it. In all the species I studied in this connection this was true even where individuals of the same species were involved. In cultures with many Harpalus species, Cymindis species, Brachynus, Agonum dorsale, Bradycellus collaris, etc., I noticed that various species of Collembola and nonparasitic acarids always increased in numbers and were apparently not touched by the carabids. Gersdorf (1937, p. 80) mentioned that tGeneric name supplied by us—General Editor. 536 473 even Carabus nemoralis, which readily feeds on trampled snails, is never able to attack living slugs (also according to Jung, 1940). I have seen this species of Carabus in a keen tussle with a small Lumbricus, which it was unable to overpower. The two species of Oodes feed on all kinds of insects, but only after they have been seriously injured (Lindroth, 1943a, p. 115). The response of most carabids to animal food is not really that of a preda- tor but rather of a hungry scavenger. But they mostly avoid putrefying* matter and attack animals that are injured or have just been killed. If this is true, the importance of carabids in agricultural entomology as the generally accepted effective enemies of insect pests of plants, is misplaced. Jung (1940) came to this conclusion with regard to the genus Carabus. There are of course exceptions, among which species of Calosoma and Cicindela are the best known. The former are undoubtedly “useful” insects. Other carabid genera of our fauna whose members apparently consume chiefly animal food, and of which some represent true predators, are: Agonum, Bem- bidion (partly), Calathus, Carabus, Cychrus, Dyschirius, Elaphrus, Notiophilus, Pterostichus (partly). Predominantly plant-eating are mainly species of the genera Amara, Harpalus, and Zabrus, probably also Bradycellus among other “Harpalini.” Some others are considered below. These comments relate only to adults. Our knowledge of the food of the larvae is much poorer. This is regrettable since the presence of suitable food for the early, susceptible stages probably represents a greater problem than for the adult insects. It is therefore possible that the area of a markedly polyphagous beetle is limited by the more strict requirements in the food requirements of its larvae. Carabid larvae of the Fennoscandian species for which definite data on diet were available number 48 species, only 13% of the fauna. These are the following species: Amara aenea C. inquisitor A. consularis C. reticulatum A. curta C. sycophanta A. ingenua Carabus** arvensis A. ovata C. auratus Anisodactylus binotatus C. cancellatus Bembidion bipunctatum C. clathratus Calathus melanocephalus C. coriaceus Calosoma auropunctatum C. granulatus *Only species of Carabus seem to attack putrefying carcasses (Jung, 1940). In some other cases where carabids approach actually decaying carrion—if not just by chance—it might involve hunting for insect eggs, maggots of flies and the like. **All Carabus larvae obtained by rearing from the egg (Bengtsson, 1927) are treated here as carnivorous. It is presumed that if, contrary to expectation, they had been fed on vegetable diet this would have been published. 538 474 C. hortensis C. intricatus C. monilis C. nemoralis C. nitens C. problematicus Carabus violaceus Chlaenius tristis Cicindela campestris C. silvatica Cychrus caraboides Dromius linearis D. quadrinotatus Dyschirius thoracicus Harpalus griseus H. pubescens Lebia cyanocephala Licinus depressus Metabletus truncatellus Nebria brevicollis Oodes gracilis Pelophila borealis Pterostichus anthracinus P. coerulescens P. cupreus P. strenuus P. vulgaris Tachyta nana Trechus quadristriatus Zabrus tenebrioides. No distinction could be made between the feeding experiments and ob- servations made in nature, since there are few indications in the literature. In most cases the results of breeding experiments may have been involved. The food of the larvae in almost every case was evidently animal matter. There are only five exceptions: Amara aenea: “grain.” But aiso designated “chiefly carnivorous.” A. ingenua: “Oats porridge.” Harpalus pubescens: “grain.” Pterostichus strenuus: “decomposed vegetable matter”; in addition, insects (Burmeister, 1939, p. 139). Zabrus tenebrioides: chiefly leaves of cereal plants; in one case an Anisoplia larvae. The feeding habits of carabid larvae—with the exception of the largest forms, such as Calosoma, Carabus, Cicindela—is an unexplored field. It is ex- tremely difficult to draw general conclusions from the above meager list. Of course the idea may be justified that the larvae are far better disposed to an animal diet than the adults. But here too we must be careful not to desig- nate the carabid larvae “predators.” I wish to draw attention to the feeding experiments with the larvae of Oodes gracilis (Lindroth, 1943a, p. 115). Only living lumbricids were attacked, whereas all insects were eaten only when they were dead or seriously injured. Even the smallest living animals chase away carabid larvae. Future precise studies will probably confirm that the larvae of carabids are invariably more dependent on animal food than the adults. However, in my opinion it will show that the larvae—like the adults—primarily attack dead and ailing animals, and also eggs and other defenseless stages. In choice of prey, with few exceptions, the larvae may be as polyphagous as the adults. 475 The feeding habit of a species assumes zoogeographical—area-limi- ting—significance only if there is a more or less considerable specialization. It is important to investigate whether some of the Fennoscandian carabids are dependent on very specific food—-animal or plant—so that perhaps on this ground alone they are excluded from some parts of the Fennoscandian region. An ecological group for which such an assumption could be made is the definite inhabitants of cultivated land. This chiefly includes a number of species of Amara, which are consistently found to occur on “weeds” favored by cul- tivation. The species of Amara, at least in the adult stage, are well known as pronounced plant eaters, chiefly seed eaters. They have been frequently observed nibbling at young pods of crucifers, so it is possible that they are associated with such plants due to their feeding habit. I performed some experiments with Amara ingenua. These were carried out with a view to finding the cause of the rapid dispersal of this species in northern Fennoscandia during recent decades. A number of experiments first revealed that the explanation was not to be found in the increased use of synthetic fertilizers (p. 525). I therefore decided to test the appetite of the insect for different kinds of fruits and seeds. Eight specimens of Amara ingenua, each isolated in a small glass dish, were provided with the same number and the same 13 types of nearly ripe fruits and seeds of the following plants, mainly “weeds.” The experiment lasted 5 days; the diaspores were carefully studied every day and all traces of feeding were recorded. The diet of all 8 specimens is summarized here. The experiment first of all shows the clearly polyphagous character of Amara ingenua. Of the 13 kinds of diaspores only 3 (Anthriscus, Vicia, Viola) were not attacked at all. On the other hand the insects showed a definite predilection for some of them, with Polygonum aviculare in first place. There is a complete correspondence with the natural occurrence of Amara ingenua, since this plant is almost always present. Its low height is especially advanta- geous, since the fruit can be easily reached by the insects. All the Polygonum fruits (with one odd exception) were completely eaten up. None of the other kinds of diaspore was. In the case of Chenopodium album, which according to the table is the next most favored, generally only the soft, juicy outer rind was nibbled, perhaps to satisfy thirst. Since the species of Amara generally feed on young pods of crucifers, the same 8 individuals of A. ingenua were also offered a choice between fruits of Polygonum aviculare and green pods of Capsella and Erysimum: Capsella bursa-pastoris 24 fruits, of which after 1 day 4 were consumed = 17%. Erysimum cheiranthoides 24 fruits, of which after 1 day 3 were consumed = 12.5%. 539 540 476 Table 33. Plants whose diaspores were attacked by 8 specimens of Amara ingenua during a 5-day experiment. F = fruit; S = seed Plant species No. of No. of diaspores Total, diaspores/ consumed on % specimens 1st 2nd 3rd 4th Sth day day day day day Anthriscus silvestris 2F 0 0 0 0 0 0 Cerastium tomentosum 38 8 8 9 10 12 50 Chenopodium album 2F 6 8 9 13 14 88 Cirsium arvense 2F 2 4 8 8 9 56 Galeopsis SP. 24,8 0 0 0 1 1 6 Lapsana communis 3F 1 6 9 9 12 50 Plantago major 5s 1 1 1 1 1 3 Polygonum aviculare 2F 9 15 15 15 15 94 Secale cereale 1F 3 4 5 5 6 75 Sinapis arvensis 38 3 7 9 9 9 38 Thlaspi arvense 38 3 4 6 10 10 42 Vicia sepium 1S 0 0 0 0 0 0 Viola arvensis 38 0 0 0 0 0 0 Polygonum aviculare 24 fruits, of which after 1 day 24 were consumed = 100%. Like all other carabids Amara ingenua readily attacks dead or injured members of its own species. It would be interesting to expose these insects to a choice between animal and vegetable food. So each of 4 specimens was offered another, crushed, specimen along with 3 fruits of Polygonum aviculare. Three of the 4 specimens were observed attacking conspecific individuals during the night. Yet, on the next day 9 of the 12 fruits of Polygonum were found to have been nibbled on. The following conclusion may be drawn: Amara ingenua is markedly poly- phagous, taking both animal and vegetable food, but seems to prefer the latter, and feeds especially on seeds and fruits of “weeds” with a strong preference (as far as tested) for those of Polygonum aviculare. The answer to the question raised elsewhere (p. 630) as to why Amara ingenua has greatly extended its area in recent times, especially northward, may be this: The reclamation of land and the subsequent expansion of “weeds” provided new possibilities of 541 477 existence for the carabid in regions earlier occupied by forest and bog. The improved possibilities of anthropochorous dispersal have probably played only a subordinate role for this carabid which is capable of flight. Attention may be drawn once again to the markedly polyphagous charac- ter of Amara ingenua. The species is in no way bound to Polygonum avicu- lare. Other carabids more or less distinctly associated with particular plants, are the following: Zabrus tenebrioides. The well-known “cereal ground beetle” is undoubt- edly the most pronounced vegetarian among our carabids. It prefers wheat (Triticum), but also readily attacks rye (Secale), occasionally barley (Hordeum). The adults feed chiefly on the grain but the larvae eat young leaves. Occasion- ally the adult beetle, less often the larva, consumes animal food; the former was also observed feeding on a wild Schedonorus. Even if it turns out that Zabrus tenebrioides is compietely dependent on the occurrence of wheat and rye, it is clear that its northern area limit is not dependent on its feeding habits. Some other species also show a more or less clear association with grasses. Burrowing species have been occasicnally found among roots of grass, and it was found in the case of Amara infima and Harpalus rufus that they nibble at them. But there is no ground for assuming that they are bound to a particular type of grass. The tendency of some animals of sandy ground to live in clumps of grass is undoubtedly due chiefly to the more favorabie moisture conditions there. The inclination is enhanced in dry weather. Living in large fascicles of Psamma and Elymus along the seashore we find Demetrias monostigma, Dromius linearis, D. longiceps, D. nigriventris, and occasionally D. sigma. Nevertheless, none of these species is bound to this biotope. The attraction of loose sand for these non-burrowing insects is prob- ably of microclimatic nature (p. 513). Since they cannot burrow themselves, it is natural for them to flee into the big fascicles of grass as the only available hiding place. Possibly they only find suitable prey there. The larva of Dromius linearis is stated to be a polyphagous carnivore; the closely (and ecologicaliy) related Demetrias imperialis was observed feeding on a Collembola. At any rate the distribution of the above mentioned strand grasses cannot represent an area-limiting factor for any of our carabids. In our region Demetrias imperialis and Odacantha melanura seem to have been observed only on shores where Phragmites grows. Whether a causal con- nection exists here is very uncertain. In central Europe Odacantha has been recorded more with Typha. Both species have been observed feeding sponta- neously on Collembola. It is not possible that they are indirectly—through the prey (e.g. aphids)—bound to the plant. Since Phragmites has a wider geo- graphical distribution in all directions than these two carabids, this does not represent an area-limiting factor. 542 478 More constant is the association of Bembidion nigricorne and Bradycel- lus similis with Calluna, which—as far as is known—are found only under this plant throughout their area (apart from accidental swarming flights of the Bradycellus species). Carabus nitens also apparently lives only where Cal- luna grows. Without experiments it is impossible to decide which factor is decisive here. However, the occurrence of Calluna does not determine the area limits of these species in Fennoscandia, since the plant has a much wider distribution. A strong attraction is exercised by Calluna on Amara infima too, and in the southern half of Fennoscandia on Bradycellus collaris, Cymindis vaporariorum and Harpalus fuliginosus. There is no question of the species actually being bound to it, except that one might like to speak of a southern “Calluna-race” in the last three cases (especially in the case of Bradycellus collaris). Amara aulica likes to seek its food in the heads of Compositae, and most members of the Harpalus subgenus Ophonus in the umbels of Umbelliferae. In no case is there a dependence on a particular species of plant, probably not even a complete dependence on these families of plants. In any case the plants are more widely distributed in all directions than the cited carabids. An isolated ecological group is bound to trees, the arboricoles, which play a prominent role in the tropics (for example, Darlington, 1943). In our region the number of carabids that spend their entire lives on living or dead trees is small: only 8 species. With one exception all of them belong to the genus Dromius. Since it must be especially interesting to determine whether one or other of our arboricoles is bound to a particular species of tree, their ecological distribution is tabulated in Table 34. Table 34. Arboreal carabids of Fennoscandia. Only tree species occurring within the region are considered. Big cross = main occurrence; cross in parentheses = outside the region wn Bt 34) 20) n 3 n ll ala ee Dromius agilis —!+1+1-|1-|—-[1++) +1+1+1- IH) + | |(#) + D. anguslus — (+) +11 I1—-1—- | —-|1- |1-|—- | —- 14) - | 1 | —|(+ D. fenestratus —|+1+l1+|-1+|+1-|1+1-1+]-|—- (+) -|-|+ D. marginellus ++ 1+ lH) — I 4+ (+) — 1 — | — | + | ICH) -- (4) - D. quadraticollis — (+) +11—|1-1-1-|- 1-1 |— 1 |—I+)] - | - 1 - D. quadrimaculatus \—\\+))} +++) - I +) —-1 + I + | + lH) + I + || +] + |D. quadrinotatus — (+ +t — 1 + | + 1—I(#)| — ICH) HI HIHI) — I Tachyta nana SS eS ia aS lS 479 This survey shows that two species, Dromius angustus and D. quadraticollis (the latter found only once), have been found in Fennoscandia only on one 543 tree (Pinus silvestris). However, in other parts of their area these species also 544 occur both on spruce (Picea abies) and on deciduous trees. In the case of Dromius angustus, the Fennoscandian localities are probably lococlimatically determined. This species is to be considered a relict, which continues to live only at thermally favored places. As a heat reservoir sand is pre-eminent, and among trees particularly Pinus grows on loose dry sand. It is clear that none of our arboricoles is associated to one species of tree. It is not possible to determine why (with the exception of Dromius agilis) they prefer one particular tree (mostly Pinus). Probably the loose, scalelike structure of the bark of pine has a role here. As carnivores they may perhaps be dependent on a particular prey or may have a predilection for a particular kind. Occasionally, different ipids have been mentioned as species succeeding Dromius (Saalas, 1917, pp. 291 ff.), and Tachyta has regularly been found to inhabit their tunnels. However, in the diet of the latter Collembola have also been mentioned. The only adult of Dromius in which feeding seems to have been observed, D. agilis, is apparently markedly polyphagous (acarids, collem- boles, aphids, “small larvae”). Characteristic is the small beetle “community” which in winter, together with Dromius, lives under the scales of bark at the base of the larger pines: Hylurgops palliatus Gyll., Salpingus castaneus Panz., and certain coccinellids, mainly Adalia bipunctata L. It is possible that these include the normal prey of Dromius, or of their larvae. In summer, when they live on the treetops (see Part I, p. 413), especially aphids seem to be the prey. In conclusion, with regard to the eight arboricoles considered here, it may be stated that none of them is bound to a particular species of tree. All may occur both on conifers and on deciduous trees. It therefore seems to be ruled out that any of these eight species is dependent on a particular prey, since the tree-dwelling phytophagous animais are without exception bound to a more limited number of host plants. Arboricoles of a more transient nature are two species of Calosoma: C. in- quisitor and C. sycophanta, which spend their unusually short period of activity largely on trees, hunting for larvae and pupae. In C. sycophanta the larva also climbs. Otherwise this species holds little interest in the present context, since it is not dependent on any particular species of tree, nor on any particular lepidopteran. On the other hand C. inquisitor, at least in our region*, seems to be bound to oak (Quercus). None of the records was made at places where oaks were not found nearby. Single records have also been made from other trees, which is natural for such a transient carabid. All records in large num- *In central Europe Calosoma inquisitor appears in large numbers on Carpinus too (Holste, 1915). In captivity the larvae was fed with all kinds of lepidopteran larvae. 545 480 bers have been made in stands of oak. It is significant that in North America C. inquisitor is hard to naturalize, whereas this is not difficult in the case of C. sycophanta (Burgess and Collins, 1917). These findings are somewhat enig- matic, since the larvae of Tortrix viridana L. indeed form the chief prey of the carabid. On the other hand, the carabid may appear in fair numbers if there are plenty of geometrid larvae (for example Erannis = Hybernia), which can live on all kinds of deciduous trees. At any rate, Calosoma inquisitor does not quite reach the northern limit of the oak (in Finland almost: map, Fig. 61, p. 437), and not even the area limit of the prey (Tortrix viridana reaches Dir; Benander, 1946, p. 16). Hence it can scarcely be contended that the northern area limit of the beetle is determined by its feeding habits. Indirectly associated with particular trees is Tachys bisulcatus, which in our region lives in wet heaps of spruce bark, but in central Europe shows a broader ecological range. Some species show a marked predilection for beech forest: Abax ater, Calathus piceus, Carabus coriaceus, Leistus rufomarginatus. None of them is bound to the beech: in more southerly regions all of them are more eurytopic (Carabus also in central Scandinavia). But it is symptomatic to some extent that the relict-like northern outposts of Abax ater (southeast- ern Norway) and Leistus rufomarginatus (Vgl Rada) are found in the vicin- ity of the northernmost more extensive occurrence of Fagus (cf. Hjelmqvist, 1940). . This survey of the carabid species that have a more or less distinct associa- tion with one or other species of plant thus shows that an unfailing connection holds only between Calluna and Bembidion nigricorne and Bradycellus similis, possibly also between oak and Calosoma inquisitor. The true nature of these associations is unknown. On the other hand Zabrus tenebrioides is directly bound to species of grass (chiefly to cereals) on account of its feeding habit. However, in no case considered here does the beetle reach the area limit of the plant concerned (the closest is Calosoma inquisitor). It must be concluded that the distribution of a plant has area-limiting significance for none of the Fennoscandian carabids. The association, not with a particular species of plant, but with biotopes with a particular vegetation, is discussed in the section on “stenotopy and eurytopy” (pp. 563 ff.). We now come to the more complicated question whether there are cara- bids that are dependent on quite special prey, whose area may be limited for this reason. The following cases suggest themselves in this connection (with regard to Calosoma inquisitor see above): 1. Genus Dyschirius. The species of this genus are well known as unfail- ing companions and hunters of the species of Bledius. Synopses on this have been provided by Sainte-Claire Deville (1924) and Burmeister (1939). Numer- ous observations have been made by Krogerus (1924, 1925a, 1925b, 1929) and Ellinor Bro Larsen (1936). 546 481 First one must be clear that not all species of Dyschirius are associated with Biedius. D. globosus and D. helleni are not associated with these animals at all. Three other species are more or less “blediophilous,” but are not bound to this prey: D. aeneus, D. lüdersi, D. septentrionum. The last-mentioned species was observed by Krogerus (1924, p. 121) in tunnels of Bledius longulus Er., and the species often lives in association with many other species of the genus. But it occasionally also lives at places quite free from Bledius. The two first- mentioned species occur rarely together with species of Bledius. On the other hand species of Heterocerus, Platystethus or Trogophloeus mostly have locali- ties in common and apparently represent-the normal prey of this Dyschirius species. We will now turn to the remaining species of Dyschirius, which have been observed without Bledius quite sporadically—perhaps only accidentally. These are the following species: Dyschirius angustatus D. obscurus D. chalceus D. politus D. impunctipennis D. salinus D. intermedius D. thoracicus. D. nitidus Nothing further seems to be known about the mode of life of D. neres- heimeri. A simple experiment was carried out separately with D. obscurus and D. thoracicus (Experiments 105, 106, p. 84). In a deep dish with moist sand two small wire gauze cages were placed opposite one another, one empty and the other containing 6-8 specimens of Bledius arenarius Payk. A number of starved Dyschirius were introduced and their distribution in four equal sectors (A-D) was recorded after about 12 hours’ exposure. The sectors were cut out at the time of recording (Fig. 5, p. 85). The following results were obtained: Hao hab 1" alien SELOE GE na Hal Total A B € D specimens Bledius Without Empty Without cage cage cage cage Dyschirius obscurus 14 9 13 14 50 D. thoracicus 13 DD, 15 15 65 It seems to be justified to conclude that Dyschirius is not attracted to Bledius by a particular olfactory stimulus, in other words that its sense of smell is not especially directed toward this prey. The result is surprising because Ellinor Bro Larsen (1936, p. 125) found that a Dyschirius rubbed with the body fluid of a Bledius was at once attacked and eaten by its companions. However, I made a similar observation with Srenus (various species). Non- 547 482 injured individuals of this genus live undisturbed in cultures of Dyschirius thoracicus; but when a Stenus is decapitated, individuals of Dyschirius pounce on it and eat it up. On another occasion I placed 3 healthy individuals of Bledius arenarius and a crushed fly in a culture of Dyschirius obscurus. After 10 hours the fly was found to be largely nibbled at, but the individuals of Bledius were untouched. Like all other carabids, members of the genus Dyschirius also immedi- ately attack an injured conspecific individual and eat it. At any rate it is clear that members of the genus Dyschirius respond especially positively—with equal enthusiasm!—to the oozing body fluids of all kinds of insects (D. obscurus was also tested on chironomids). Even the larvae of Dyschirius are not specially disposed toward Bledius (chiefly its larvae). Both, Schigdte (1867, p. 503) and Ellinor Bro Larsen (1936, p. 126), observed them spontaneously consuming Heterocerus, including its larvae and eggs (in the first case the larva was found to belong to D. thoracicus). This prompts the following conclusions: Members of the genus Dyschirius (at any rate D. obscurus and D. thoracicus) are not exclusively dependent on Bledius for their food. They feed on all kinds of insects, especially injured or dead ones. But where they live, on account of the sparse occurrence or lack of tall vegetation, the fauna is extremely poor in species and consists chiefly of animals that can utilize sand-algae as food, primarily Bledius (Krogerus, 1925a, p. 4). The primary prerequisite for the occurrence of a particular species of Dyschirius is the state of the soil (p. 507). This must be suitable for digging tunnels, for which not only a particular particle size but also a definite, fairly constant moisture is necessary. D. chalceus and D. salinus also require NaCl. Which species of Bledius—or which other soil insects—are present there seems immaterial, provided they occur regularly and in sufficient numbers. I found a good example of the factors effective for the distribution of Dyschirius on the seashore of Vbt Byske (July 15, 1936), where Bledius are- narius was unusually common everywhere on the barren sandy shore. Living along with it, although not associated with it, were found Dyschirius obscurus and D. thoracicus. The former was found alone, closer to the waterline, in finer and moister sand. That D. obscurus prefers fine-grained sand was found both in nature (Krogerus, 1932, p. 164) and experimentally (p. 507). It is also more resistant to exposure to water than D. thoracicus (Krogerus, 1932, p. 237). The almost obligatory association between D. obscurus—and also of D. impunctipennis—and Bledius arenarius is undoubtedly due largely to the identical requirements of these three species in the above-mentioned context with respect to the kind of sand. On the other hand D. thoracicus is more eurytopic (Occurring on coarser and on loam-mixed sand) and was found along with more species of Bledius (for example, Krogerus, 1923, p. 121). It is very significant that the 5 species of Dyschirius not associated with 548 549 483 Bledius (see above), live all at places with a more or less rich and taller vege- tation, where the insect fauna is much richer in species, so that they are not dependent on a particular prey. In conclusion it may be considered inconsequential which factor is primary, sand or Bledius, since both must be present! Strictly speaking, neither of them is actually the decisive area-limiting factor for any species of Dyschirius. More correctly, the disjunct occurrence of shoreline with suitable sand and Bledius is decisive (see sand map, Fig. 77, p. 510). Of prime importance is the ability of Dyschirius to find and to colonize these places, which are so far certainly inhabited only by a small fraction of the possible species. Factors related to the immigration history—in particular, time—hold the key. Since Asaphidion pallipes has also hardly been observed at places free of Biedius it is possible to consider this species as identical with Dyschirius in its feeding habit. But the adult does not burrow. 2. Brachynus crepitans. For three years I spent much time on unsuccessful experiments to clarify the feeding habit of this species. The adult beetle is very easy to please: it can be fed on all kinds of crushed (but not living) insects, on individuals of Lumbricus, etc. Some beetles were given only bread for more than a year. The habits of the /arva are completely unknown. But the attempt has been made to find out why Brachynus crepitans is, so to speak (in our region almost always), consistently in the company of Agonum dorsale, particularly since the larval stage of the North American B. janthinipennis Dej. has been found a parasite of the pupa of a gyrinid species (Dineutes) (Dimmock and Knab, 1904). It was therefore necessary to study the behavior of Brachynus crepitans toward the Agonum species as precisely as possible. At least in captivity Brachynus behaves passively toward the succeeding species, whereas Agonum shows strikingly “friendly” behavior. Often it positions itself next to Brachynus and ardently rubs its back and the sides of the prothorax and elytra fore and aft, like a cat, against various parts of the body of Brachynus. Of course Agonum also shows similar behavior toward its own kind, even toward dead objects. At any rate this does not indicate that the Agonum species is in any way dependent on Brachynus; over wide areas, for example in Skane, it lives alone. During the “cleansing ritual” Brachynus remains passive. Besides, a simple experiment in the substratum gradient apparatus (Experiment 104, p. 84) was used to establish that it is not attracted to Agonum by the olfactory stimulus. In 5 boxes the two species were distributed as follows: Brachynus || Brachynus || Agonum || Brachynus || Brachynus 6 specimens || 6 specimens || 10 specimens || 6 specimens || 6 specimens The Agonum box was covered with thin cloth to prevent entry and exit. After 3 replicates of the experiment and each 3-day exposure the following distribution 550 484 figures were obtained: Brachynus Brachynus Agonum Brachynus Brachynus 27 19 = 17 19 The distribution was thus identical. The two species of beetle were maintained in common culture for several months without attacking each other. On two occasions I also obtained numerous Brachynus larvae in the cul- tures (see Supplement). However, it was impossible to find suitable food; all kinds of crushed insects, individuals of Lumbricus, bread, etc. were left un- touched and the larvae all died without perceptible growth. The larvae behaved passively toward Agonum dorsale. 1 also reared two larvae of Agonum dorsale from the eggs for almost 3 weeks together with Brachynus larvae, without observing any attack. When finally (on July 8) one Agonum larva pupated, the last Brachynus larvae were already dead 3 days. The only combination not tested was Brachynus larva with Agonum pupa. Yet it may be justified in ruling out any parasitism: First, the Agonum pupa is too small (dry body weight of the adult beetle 4.08 mg as against 6.96 mg for Brachynus: mean value for 10 specimens of each from Oland*). Second, it appears too late in the summer, since oviposition by the two species takes place at about the same time. Brachynus and Agonum dorsale apparently do not have parasitic (or feeding habit) relationship with one another. They are just a rare example of two species with nearly identical ecological requirements, and have been considered as such above (pp. 59-64). The larval biology of Brachynus thus remains unsolved. Several biotopes (in Upl, Ogl, Old) where the beetle occurs frequently were carefully dug up and sifted during those three summers without a trace of a larva. The other insect inhabitants of these places were collected and noted, and some, with the right size and time of appearance in the larval or pupal stage, were offered to freshly emerged Brachynus larvae, but without success. Further description of this repeated failure is superfluous. But it seems to me that all this favors the view that the larva of Brachynus crepitans, like that of B. janthinipennis, is a monophagous parasite. If I finally think of ants this is to be considered only a weak, private hypothesis. So at present it is not possible to decide whether the feeding habit of the larva is a decisive area-limiting factor for Brachynus crepitans. Perhaps One must ask which other insect (“the host”) could possess the extremely characteristic distribution of Brachynus. 3. Species of Lebia. Analogous cases exist here. Abroad the larva not only of the notorious Lebia scapularis Fourc. has been found as a monophagous * According to a letter from Darlington, Brachynus janthinipennis Dej. (auct.) is about 7.5 mm long, and its host Dineutes americanus Say (auct.) 10-11 mm. Since the gyrinids are known to possess a sticky body, the difference in body weight in favor of Dineutes is probably still greater. 551 485 ectoparasite (on Galerucella luteola Mull.)* but also of the North American Lebia grandis Hentz (on Leptinotarsa decemlineata Say) (Chaboussou, 1939). The hosts in these cases therefore consist of grown-up larvae or pupae of chrysomelids. The adults of some American and African species of Lebia seem to be monophagous predators on chrysomelids as well (Chaboussou, lie.) : So far, there are no definite observations in this regard on the three species occurring in Fennoscandia. Burmeister’s contention (1939, p. 190) that the larva of Lebia crux-minor feeds on the larvae of Galeruca and Chrysomela is based—as far as I can see—on assumptions by others (for instance, Blunck, 1925, p. 37). However, the assumption is probably correct. In Fennoscandia L. crux-minor is found nearly always at places inhabited by Galeruca tanaceti L. In northern Sweden this species along with Chrysomela marginata L., is the only chrysomelid of a suitable size. On the other hand, the above mentioned Chrysomela species is so rare in parts of southern Sweden that it cannot be the only host. If Lebia crux-minor is bound to a chrysomelid as a monophage this must be Galeruca tanaceti. The presumed host has a much wider distribu- tion than Lebia (especially in Norway). Hence, in Fennoscandia the feeding habit as an area-limiting factor of this species is involved at the most only slightly. The adult Lebia chlorocephala is found to be rather polyphagous in cap- tivity (for instance, it eats aphids and also bread). But in nature it was found regularly associated with species of Chrysomela, especially C. varians Schall. (see also West, 1947, p. 17) and was also observed feeding on its larvae, so that a constant association of the Lebia larva with this prey can be presumed. If the distribution of the two species (Chrysomela according to the Catalogus, 1939) is compared there is a good correspondence inasmuch as C. varians advances northward in Finland much farther than in Sweden. On the other hand this species is so far unknown in southern Osterbotten, in Halsingland and on Bornholm, regions where numerous records of Lebia have been made. C. fastuosa Scop. lives in the first of these regions and the Lebia adult was found feeding on its larva in central Europe. But in His and on Bornholm even this species seems to be missing. We may conclude from this that Lebia chlorocephala, if as larva it is dependent on species of Chrysomela, which seems probable, it nevertheless cannot be monophagously bound to a single species. So the feeding habit can hardly have an area-limiting role for this species either. In the case of the third species of Lebia, L. cyanocephala, no association was established with a particular prey. If Xambeu’s data (1898, p. 175)** are *Burmeister (1939, p. 191) also mentions Galerucella lincola Fbr. as the host of Lebia scapu- laris, without citing his source. Chaboussou (1939) includes only Galerucella luteola. **Erroneously cited by Blunck (1925, p. 37). 3932 553 486 reliable, on the contrary it can be argued that the larva of L. cyanocephala is a thoroughly polyphagous predator. 4. Snail-eating species. Only two species seem to be obligatory snail-eaters, Cychrus caraboides and Licinus depressus, the latter at least in the larval stage. The former of these is said to feed almost exclusively as adult on snails in the shell, for which purpose the slender anterior part of the body with the elongated head is especially suitable (see figure in Burmeister, 1939, p. 47). Jeannel’s assumption (1941-42, p. 989) that the larvae of Licinus feed on snails, which seemed highly probable in view of the strong resemblance with the snail-eating larva of Dicaelus from North America (Dimmock and Knab, 1904, p. 26), has been confirmed. On June 15, 1947, on Oland, Greby alvar, I observed a half-grown larva of Licinus depressus busy in eating a small snail (Vallonia costata Mull., det. N. Hj. Odhner). This snail is widely distributed across the southern half of Sweden inhabiting a much larger area than the Licinus. Until we establish, first, whether the two above-mentioned species (at least as larvae) are able to live exclusively on snails and, second, whether they are associated with a particular species of snail, which I do not believe, it is impossible to evaluate the area-limiting significance of this special mode of life. Moreover, some species of Carabus, especially C. coriaceus, have a strong inclination to feed on snails (both without and with the shell). However, no species is bound to this mode of life. 5. Some other species with special modes of life may be specialized in their diet. Tachyta nana seems to live under the bark of trees only where there are abundant tunnels of ipids, and is said to feed in all stages chiefly (not exclusively) on these insects. However, there is no dependence on any partic- ular species of ipid, so the area-limiting significance of this feeding habit must be slight. For similar tree-dwelling species of Dromius see above (p. 542). The food of Miscodera arctica is not known. However, the constant oc- currence of byrrhids (especially Byrrhus fasciatus Forst. and Cytilus sericeus Forst.) in the habitats of Miscodera is suggestive, and it is possible that these animals represent their normal prey. But the Fennoscandian distribution of these byrrhids is almost universal, so the thinning out of Miscodera in the south and the general limits of the area cannot be explained on the basis of feeding habits. The anthropobiont “domestic animals” Pristonychus terricola and Sphodrus leucophthalmus often live in association with species of Blaps and may be dependent on them or at least favored by them. 6. A very special “association” is formed by the inhabitants of animal nests. Fennoscandian carabids are associated only with the nests of rodents and of Talpa (in England the development of Bembidion harpaloides was observed in a Garrulus nest). 554 487 Trechus discus and T. micros seem to be completely dependent on mam- malian nests (for a similar dependence of other Trechus species on rodents, assumed by Danish entomologists, see Part I, p. 668, footnote). Probably the two species of Bembidion in the subgenus Ocys, B. harpaloides* and B. quin- questriatum, also have an obligate association with animal nests, details of which remain to be clarified. Trechus quadristriatus shows a predilection for rodent dwellings but is not bound to them. For Agonum quadripunctatum see Part I (pp. 86-87) and p. 528 above. How careful one must be in saying that a species of beetle is bound to animal nests purely on the basis of its more or less regular occurrence there, is clearly shown by Sokolowski (1942, for instance, p. 186) for Catopidae. It is not easy to decide whether the dependence on animal nests, chiefly that of Trechus discus and T. micros, is due to feeding habit or not. Possibly these species are secondarily dependent on other nidicolous small animals for food. It should not be forgotten that animal nests also represent a very peculiar biotope with respect to their thermally well-balanced microclimate, as is evident from the detailed studies by Nordberg (1936, pp. 60 ff.). At any rate, not only in Fennoscandia but in the rest of Europe (Borchert, 1938, p. 7, Map 50) the almost identical distribution of the two species of Trechus seems to show clearly that their areas are regulated by one and the same factor. This factor cannot be the distribution of their most important common host, Arvicola amphibius, whose distribution is almost pan-Fennoscandian. On the other hand it would certainly be strange if just the thermal factor had the same effect (an identical distribution) for two species, one of which (T. discus) hibernates as a larva and the other (T. micros) as an adult. The question stands: only this much is established, that the areas of Trechus discus and T. micros are not limited by the occurrence of their presumptive hosts. Summing up the consideration of food as an area-limiting factor for the Fennoscandian carabids, we find that a decisive role can be ascribed to it at the most in exceptional cases. A particular species of plant which could at least indirectly exercise an influence through feeding habits seems indispensable for at most three species, namely Calluna for Bembidion nigricorne and Bradycellus similis, and perhaps also the oak for Calosoma inquisitor. A particular species of animal (Galeruca tanaceti) can be taken, most probably, as necessary only for Lebia crux-minor. For good reasons was a similar, but as yet unknown, dependence also assumed for Brachynus crepitans. Most species of Dyschirius, Lebia chlorocephala and Tachyta nana are more or less “oligophagous” carnivores; for a few others only hypotheses in the same direction are possible. Two species of Trechus are apparently bound to small *Horion (in litt.) is of the view that Bembidion harpaloides is not constantly associated with animal nests (see Supplement). 355 488 mammals, !ikewise possibly two species of Bembidion. In the few cases where a more or less constant feeding habit associa- tion was established between a carabid and a plant or animal species, the “host species”— with Calosoma inquisitor as partial exception—is so much more widely distributed than the carabid dependent on it that the area-limiting factor must be of different nature. One reason that carabids offer an unusually suitable subject for zoogeographical research is that in them the feeding habit factors have a subordinate role, so that other factors, less specialized thus more significant, are isolated and their effect brought out. These are fore- most: the climate and the dynamical characteristics of the animal, and secondly the characteristics of the soil. Competitors and Enemies There is hardly an ecological factor whose effect in nature is more difficult to determine than the competition between organisms having identical or similar requirements of life. In most plants as well as in sessile animals it is to a large extent purely a “struggle for space” (combined with the requirements of food, oxygen, light, etc.). In the case of freely moving animals the struggle for food assumes first place. It is generally contended that competition is strongest among species tax- onomically closely related to one another, since it may be expected that these are also more or less related in their feeding habit. Thus Elton (1946) under- took a Statistical analysis of the animal “communities” of a large number of biotopes of the most diverse types, and found that the number of animal gen- era represented within each biotope (s. 1.) by one single species is larger than what might be expected from a purely accidental distribution of the fauna. Apart from the fact that the figures Elton gives can be interpreted in a dif- ferent way (Williams, 1947; see below), attention may also be drawn to the following: Whether or not taxonomically closely related forms—designated as species or otherwise—are competitors for food, an ecological difference, at the very beginning, must strongly favor the functional separation of a newly originated type from “the old species.” Ecological insolation, rendering hy- bridization difficuit, must help to originate a genetically fixed new form (see, among others, Crombie, 1947, p. 64). If Elton’s view that closely related forms mostly live separately is correct, this need not mean that they compete with one another, but that ecological difference is the primary requirement for their development as physiologically different entities. However, Williams (1947) has opposed Elton’s view (1946) that in iden- tical conditions of competition the mean number of species per genus in each geographical or ecological region remains unaltered, independent of the size of the species stock, and that therefore any decline in the number of species must be ascribed to the effect of increased competition among members of 557 559 489 the same genus. Williams (also 1944) found that the number of genera with 1, 2, etc. species form a logarithmic series, which becomes more distinct, the larger the material in hand. A test of this method on the entire Fennoscandian carabid fauna (Diagram 53) gives a fairiy good correspondence between the empirically and logarithmically calculated figures.” From this it follows that in samples poorer in species a considerable purely mathematical reduction must be expected in the number of species per genus. As an example, to enable us to test statistically the effect of “intrageneric” competition (among members of the same genus), the carabid fauna of some riverside biotopes of Norrbotten (Lindroth and Palm, 1934) and Varmland (Palm and Lindroth, 1937) may be selected. These studies may be especially suitable for the present investigation because banks represent narrowly delim- ited, comparatively small biotopes where the effect of possible competition should be clearly evident. Table 35 gives the calculated mean number of species per genus and the same value for the total of 6 or 12 samples from each riverside biotopes studied, for the entire carabid fauna of Nbt and Vrm respectively, and fi- nally for the carabid fauna of the whole Fennoscandian region. These empir- ical figures were compared with those calculated logarithmically according to Williams’ method. It is evident from this comparison that the mean number of species of carabid genera of these riverside bjotopes is everywhere greater than what might have been expected. If any conclusions can be drawn from this they should rather indicate a selection in favor of the association of tax- onomically closely related forms rather than the opposite! This result agrees with that obtained by Williams (1947). However, mention must be made of a factor that might lead to increase of the mean number of species per genus in special biotopes—such as the river- sides considered. At such places, where more or less extreme factors exist—for instance, sparsity or absence of taller vegetation, big fluctuations in the water level, etc.—a particular constitution (physiology) of the animals is often called for. For instance, species of Dyschirius and Bembidion are especially suited for life on banks. However, there is a larger number of genera that are excluded for these very reasons. In extreme biotopes the number of genera diminishes. Consequently—speaking from the viewpoint of competition—more “space” becomes available to species of the few genera that are represented there. So even with unchanged conditions of competition more species per genus may be expected. It is therefore best not to take the calculations according to Williams’ method as proof for a favored association between members of the same genus, but on the other hand it does not seem possible to confirm the contrary theory * These and all other mathematically calculated figures in this section were worked out by Fil. Lic. Leo Uskila. 25— 6- 4. = I | \ \ \ \ x \ \ \ N N axel Cr ee ee [ eu I x 11,2. ana non VO PT ORB NET LERNENS UL ZUR eCity) OUR TEN CHILE NR RAR 556 Diagram 53. Number of genera with 1, 2, etc. species constituting the Fennoscandian carabid fauna (61 genera, 362 species). Curve based on Williams’ figures (1944, 1947). Outside the diagram one genus each with 30, 32, 39 and 65 species respectively. Bo 560 491 proposed by Elton (1946), using statistical methods. The association of numerous closely related species of carabids is not at all unusual. A good example is the well-known Harpalus locality, visited by many collectors, in Old Stora-Rör (Lindroth, 1948c, p. 44). Here, on the edge of a thick Calluna cover which to the east is fringed by open sand, 12 species were found in one day to occur together (H. aeneus, H. anxius, A. hir- tipes, H. melancholicus, H. neglectus, H. picipennis, H. rufitarsis, H. serripes, H. servus, H. smaragdinus, H. tardus, H. vernalis), i.e. more than 50% of the known Fennoscandian species of Harpalus s. str. Some of them (especially H. anxius and A. smaragdinus) were present in very large numbers. In 1910 all the species were collected by Sandin near Stora-Ror; it must be a fairly stable “community.” Numerous more or less synanthropous species of Amara have also been constantly found together. In many places in southern and central Sweden the following species can be found together in a few square meters of sandy or gravelly soil in a sunny situation, with rich vegetation of Artemisia vul- garis, Arctium and tall species of Rumex, Polygonum aviculare, Stellaria media, etc.; A. apricaria, A. aulica, A. bifrons, A. consularis. A. eurynota, A. famil- iaris, A. fulva, A. ingenua, A. municipalis, A. ovata, A. similata. Also often found living together are Calathus (on Oland and Gotland always C. ambiguus, C. fuscipes, C. mollis), Ophonus (O. melleti, O. punctatulus, O. rupicola, often also O. azureus, occasionally O. seladon). In the winter half-year we find living under the bark of trees several species of Dromius (especially D. fenestratus, D. marginellus, D. quadrimaculatus, D. quadrinotatus). On the basis of both statistical study and the field experience of ento- mologists one gets the impression that the competition for food among different carabids has no significant role. At any rate, the idea that closely related species represent especially serious competitors seems baseless. For this reason I am inclined to ascribe Krogerus’ finding (1947, p. 45), that species of the muscid genus Coenosia occur as biotope dominants, so to speak as vicariads, rather to an undertermined abiotic environmental difference than to a competitive relationship. I know of only one case where there seems to be effective competition among species of carabids. The stronger competitor here is Carabus nemoralis. In the Fennoscandian region it is undoubtedly a comparatively late immigrant (p. 632) which is still in the process of dispersal. To the extent that it has greatly increased in numbers—especially around cities—other species of Carabus have become noticeably scarcer. This is especially true of C. hortensis, which likes the same biotopes as C. nemoralis: sparse woodland, for instance, parks, with moderately moist soil, preferably with an admixture of gravel. It would certainly be interesting to study the competition between Carabus nemoralis and C. hortensis experimentally. No explanation may be forthcoming without that. But attention is drawn here to one aspect that may perhaps = | == | E65 | 19 |29€ | eıpueo9souua;] JO JJoyM oY} Jo eunvy piquies [se | s4¢]rtfels I BoE rf = SE elıjı Tr etme 1| 61 z| S| sx sonıpe907 ZI PDIS|AD YI I DODYSAO +] alad 49WSIO KgsjoH (LET (Seta alps uo1{BupT sol1oH Buvdn A SEEN SDUISDT CNV INYUVA - EROSION aQ Cm ntDO TNO a MONA -ın km N - = = AQ Hm - Le a | a = N saıe907] 9 uhgsaıy VYDY-4IPI N VD -4920 maus 4 DIUAO J -4920 1HunsD YY NALIOIUNON I z 9504 zS'I rosea Kae ie = gor + gv I 92 rom tz 9 9g50+ <5‘1 Re | ae _— Lama! = a =—= A = = AN YH ml Y No x2 MD ee - Oo wın m \ oo mm on N Os = = A SUreIpEM J9}Je) po} -ejnafeg jesusıd “Wy uoljeIAad eloued jo 1oquinN sawads jo Joquiny sn][a90YyoI1 |, eprydojadg snqgoiedg BlIGoN smojqeJaW 1391107] snysloy snjedie] snrwoll] snıy9kJ BUIAN|D RISRUIOTe sniux[y) snyeey snJsolg esıqadg vorpiquisg uoıpıydesy eıewvy snuas Jad saraads jo Joquinu ueaW snyo1jso1a}g snydajlIad snıtydonon snuuy9sÄq wnuosy sndjednoy J[OYA B SB EIPUEISOUUST JO BUN] pIgeies 3) YVIM pUe SadUTAOId OM} BS9q} JO une) PIqe1ed [8903 94} YIM opeuu St uosı1edwo) -PUB[ULIEA PUL 197}0GIION UI Sadojorg apIsIALI 3WOS UI 8.19U33 jUaJayjIp JO Satoeds piqesed Jo SIoquinu Jo uonnginsiq “SE AI4EL 855 561 493 be useful. An important biological difference between the two species is that nemoralis hibernates as an audit, hortensis predominantly as a larva. Moreover, nemoralis is an unusually hardy, insect, which goes hunting on frost-free nights in the winter half-year. It needs to be found out whether the larva of hortensis is not more susceptible to cold: in spring it wakes up to full activity later than the nemoralis adult. Perhaps during this transitional period it is a defenseless prey for the beetle. At any rate it might turn out that quantitative fluctuation in the popula- tions of the two species of Carabus is not due to simple competition for food, but that nemoralis is actually an active enemy of hortensis. Gersdorf (1937, p. 78) surmises that in Germany there is similar com- petition between Carabus auratus and C. cancellatus, in favor of the former. Renkonen (1944, pp. 88 ff.) posits a possible competition between Philonthus quisquiliarius Gyll. and P. micans Gr. Evidently the carabids have numerous enemies. On the other hand there are remarkably few enemies that unilaterally specialize in this prey. The best- known at present is Methoca ichneumonides Latr., belonging to the family Thynnidae, which is an unfailing enemy of Cicindela larvae (Adlerz, 1916, pp- 299 ff.). However, throughout our region this parasite seems to be rare. Larvae of Calosoma and Carabus, as well as those of Nebria brevicollis and Pterostichus vulgaris, are occasionally parasitized by numerous larvae of the proctotrupid genus Phaenoserphus—probably always P. viator Hal. (Kieffer, 1914, p. 29; Holste, 1915, p. 419; Burmeister, 1939, pp. 32, 34, 140). It is possible that the “ichneumonid” mentioned by Lengerken (1921, p. 33) being parasitic on Carabus silvestris Panz. also belongs here. More numerous are the parasites found on carabid adults. In the gut and in the body cavity of the larger species (Carabus, Broscus, Pterostichus niger) gregarines are very Often found, which probably belong to the species Mono- cystis legeri L.F. Blanck (Delkeskamp, 1930, pp. 18-20). They never appear to endanger the life of the host, but if they enter the body cavity they apparently cause castration. Endoparasitic nematodes (genus Rhabditis) have been found both in adults of Calosoma sycophanta and in a larva of Carabus monilis (Burgess, 1911, p- 71). It could not be ascertained whether they are harmful for the host. Gordius worms were observed by Gersdorf (1937, p. 31) once in Prerostichus oblongopunctatus and once in P. vulgaris; according to Burmeister (1939, p. 37) Gordius aquaticus L. lives in Carabus coriaceus. These worms occur in carabids only occasionally. Some tachinid maggots are endoparasites capable of causing the death of various larger carabids. Viviania cinerea Fall. has been found in several species of the genera Amara, Broscus, Calathus, Carabus, Harpalus, Pteros- tichus, and Zabrus (J.C. Nielsen, 1909, p. 72; 1918, p. 253; Lundbeck, 1927, p. 286; Audcent, 1942); Freraea gagathea R.D. (albipennis Zett.) was found 562 494 in Carabus monilis and harpalus pubescens (J.C. Nielsen, 1916; Lundbeck, 1927, p. 120), and finally Weberia pseudofunesta Villen. (curvicauda auct.) in Amara aulica and Harpalus pubescens (Lundbeck, 1927, p. 117), possibly also in A. aeneus (J.C. Nielsen, 1909, p. 76). In North America a braconid (Microctonus carabivorus Mues.) is known to be parasitic on the adult of a carabid (Galerita sp.) (Muesebeck, 1931, pp- 16-17). It is possible that in Europe, too, related species have a similar mode of life. Very often acarids are found attached to carabids, occurring in enormous numbers, especially in cultures. They can cause death at least of larvae, but on the other hand may have no role in nature. These casual parasites may belong to various genera (Burgess, 1911, p. 70; Delkeskamp, 1930, pp. 20-21.) Obligatory parasitic acarids were found on the coxae of Carabus granulatus and Prerostichus niger (Gersdorf, 1937, p. 31), but especially under the elytra of species of Carabus (Family Canestrinidae; Delkeskamp, 1930, p. 22; Gers- dorf, l.c.; Sellnick, 1939, p. 1304) and on hygrophilous species of the genera Agonum, Bembidion and Pterostichus (Parasitiformes: Sellnick, l.c.). taxonom- ically these parasites are very little known. Their effect on the host seems to be insignificant. An idea of the relatively little importance of parasites living on carabids is provided by the fact that in my fairly extensive cultures of adults of numerous species, some of them lasting over long periods, I was never able to attribute the death of an individual to the effect of parasitization. Besides acarids—both casual and obligatory parasites—there were only ectoparasitic fungi of the Family Laboulbeniaceae (Lindroth, 1948b), occurring fairly abundantly. These, too, do not appear to endanger the life of the hosts. We therefore find that—as far as is known—no single parasite (per- haps with the exception of some Laboulbeniaceae) is specific to a particu- lar species of carabid, and of the few species of lethal parasites none seems to occur regularly and in large numbers. The statement is justified that the carabids—especially in comparison with Lepidoptera, Diptera, and other fam- ilies of beetles, such as Chrysomelidae and Curculionidae—are parasite-free to an unusually high degree. On the other hand they undoubtedly have numerous enemies of another kind, designated as predators. The most dangerous are certainly various birds. Often the excreta of birds consist almost entirely of elytra and other hard parts of beetles. Such excreta may originate especially from members of the crow family (Corvidae). Careful studies carried out by Notini (1943) on the diet of the hooded crow (Corvus cornix L.) showed that carabids—especially the diurnal species—constitute a major part of the diet (cf. also Burgess, 1911, p. 70). Other birds feeding on carabids are also mentioned in the literature (Krogerus, 1932, p. 176; Gersdorf, 1937, p. 30, Burmeister, 1939, pp. 27, 32, 129). 563 564 495 Other enemies of carabids are the fox (Canis vulpes; Lengerken, 1921, p- 32; Burmeister, 1939, pp. 32, 43), “mice” (Burmeister, l.c.), the mole (Talpa; Burmeister, l.c.), shrews (Sorex; Lengerken, l.c.), and above all probably the badger (Meles taxus), which has a strong inclination to feed on the larger, chiefly nocturnal species (Notini, 1948, p. 119). Of course bats (Chiroptera) also feed on night-flying species. Toads (Bufo) are also mentioned by Lengerken and Burmeister (l.c.). Gersdorf (1937, pp. 30-31) notes that the larger staphylinids (Staphylinus, Philonthus) hunt for carabids; they seem to possess an important weapon in the form of a paralyzing poison (especially mentioned in Paederus). Carabids also attack one another, especially injured, diseased or still teneral individuals, as well as eggs and larvae (p. 535). This brief survey of the enemies—parasites and predators—of carabids has shown that the number of species involved is high. However, it is striking that these enemies are little specialized. None of them is dependent on one species of carabid for prey. It is evident that these enemies, and I am inclined to put birds in the first place, may bring about a reduction in the size of populations. On the other hand it does not seem possible to consider this as an effective area-limiting factor. The view expressed above (p. 554) in considering the feeding habits is thus confirmed and can be expanded as follows: Biotic factors play a subordinate role in the distribution of carabids; decisive are the abiotic existence factors (climate and soil) and the dynamical factors. Stenotopy and Eurytopy It is a proud moment for the experienced field entomologist when, on an excur- sion, well beforehand—from as far away as possible—he can point out to his younger colleague, in a calm voice, the locality of one or other of his favorite animals. Such prophecies come true in remarkably many cases. Often even “the prophet” is surprised, for he is not always clear why he expected just this or that species there. However, on numerous collecting trips he has learned to associate certain definite environmental factors with some particular animal species, until the psychic channel is so fixed that a particular combination of environmental factors in nature automatically directs the thought to a partic- ular species of animal. In this way the field entomologist gradually and almost unknowingly ac- quires an idea of the conditions essential for life of the more or less stenotopic Species, those more or less closely associated with a particular biotope, such as those identified by the entomologist chiefly from the vegetation. The world “stenotopic” (opposite: “eurytopic”) is nevertheless a highly relative term, not only in the extent to which an obligatory association with a particular biotope is required or not, but chiefly according to whether the 565 496 term “biotope” is interpreted in a narrower or a wider sense. The biotope “comprises sections of the living space, which are identical in the essential conditions of the requirements of life for organisms existing there, the life forms, that are adapted to these conditions, and thereby differ from other locations” (Hesse, 1924, p. 141). It is evident from the examples cited that Hesse interprets the term rather narrowly—with full justification: slimy, sandy, gravelly and rubbly shores (of the sea) are designated as different biotopes. For those terrestrial surfaces that are included under the term “biotope,” therefore, as far-reaching a correspondence as possible may be required in the following most important features: condition of soil (particle size, secondar- ily also the chemistry), moisture, insolation, and—arising chiefly through the combined effect of the above factors—the vegetation. On the other hand it is difficult to judge the extent to which certain requirements of a biotope can be ascribed to the climate (microclimate), since localities that must be consid- ered as belonging to the same biotope (for example, a Sphagnum fuscum bog, a sandy Calluna-Pinus forest) may lie in very different climatic regions. Strictly speaking, then, only species that inhabit a particular biotope exclu- sively and at all times is stenotopic. However, in practice a few restrictions are justified. Accidentally occurring individuals in alien biotopes may be chance migrants, winged insects being especially common. Even if they come to repro- duce at such places, the species need not be denied stenotopy unconditionally, provided this does not result in permanent colonizations in alien biotopes. Species that undertake more or less regular migrations can be stenotopic nevertheless, if the reproduction (larval development) is biotope-bound. Ex- amples of this are to be found chiefly among the riparian carabids, which in winter often live quite far from water (Lindroth, 1943; H. Krogerus, 1948). No species is eurytopic in the broadest sense, i.e. “inhabiting all biotopes.” However, to be justly called eurytopic an animal must not exhibit any predilec- tion for a well-defined biotope. The most eurytopic of all Fennoscandian cara- bids may be Dyschirius globosus. Bembidion lampros could possibly vie with it for this rank. Examples of more limited eurytopes are: Pterostichus diligens (eurytopic hygrophilous species), Bembidion rupestre (eurytopic shore species), Harpalus aeneus (eurytopic dry-meadow species), Calathus micropterus (eury- topic forest species), etc. It is not possible here to enumerate all the more or less distinctly steno- topic carabids. The most pronounced cases, which also deserve detailed study from the purely zoogeographical viewpoint, seem to be following: In hardwood forests Abex ater Leistus rufomarginatus Calathus piceus Nebria brevicollis Calosoma inquisitor (p. 543) 566 .497 In hardwood forest swamps Acupalpus consputus Badister unipustulatus Agonum krynicki Bembidion clarki Badister Sodalis In forest bogs (“Bruchwaldmoore”) Trechus rivularis In open hypnum moss bogs (“Braunmoore”) Agonum consimile Elaphrus lapponicus On Sphagnum quaking land (not S. fuscum) Agonum munsteri In Sphagnum fuscum bogs Agonum ericeti Dyschirius helleni On peat soil Bembidion humerale In dry tundra Diachila polita Trichocellus mannerheimi. Pterostichus middendorffi In sandy fields and “steppe soil” (see “limestone species,” pp. 115 ff.). Furthermore: Amara infima A. spreta A. quenseli Pterostichus lepidus. On Calluna heath (p. 541). On riverbanks with varied sized soil particles (pp. 508 ff.). On the shores of eutrophic lakes (p. 528). It is now clear that the zoogeographical significance of the greater or lesser stenotopy of a species is completely dependent on the more or less frequent occurrence of the required biotope. In a continuously wooded area—such as most of Fennoscandia—a forest-bound species such as Calathus micropterus may occur almost ubiquitously, and for similar reasons, Bradycellus similis, apparently dependent on Calluna, has a wide area in the southwest. Fastidious species such as, among those mentioned above, Agonum ericeti, A. munsteri, Trechus rivularis, must have a correspondingly disjunct area. Clear examples of this kind were found by comparisons between the sand and loam maps (Figs. 77, 78, pp. 510, 511) and in the distribution of some species especially sensitive to the ground material (pp. 512 ff.). Nevertheless it is particularly important that fragmentation of the habit- able area is a great hindrance to dispersal, a question considered below. The cause of stenotopy is doubtful. We can only presume that a markedly stenotopic species has unusually fixed requirements of the environmental fac- tors. In general it must remain undecided whether a single factor or a com- bination of factors is decisive, or whether finally we must look for a complex “holocoenotic standard factor” (Friederichs, 1930, p. 109). 567 498 This may be exemplified by the riparian species. They are undoubtedly ripicolous, since they have definite moisture requirements. The choice between different types of shore seems to be determined chiefly by the mechanical (including hygric) characteristics of the ground material, especially particle size. Moreover, some species inhabit only open shores, exposed to sun. They are to a large extent dependent on the vegetation, which also influences the soil mechanically and chemically. Some riparian species live only at the sea-side, and may be dependent on the saline content of the soil. Thermal contrasts are chiefly represented on the one hand by the banks of springs and rapid brooks (with stable, low temperature) and on the other hand by those of shallow, slightly warmed ponds and lakes. Finally reference should also be made to the availability of sufficient food, which is a particular problem, for instance, for the species of Dyschirius (p. 545). Especially informative are cases where a species inhabits diverse biotopes in different parts of its area. In Part I of this contribution numerous examples of this kind were provided, where the mode of life of a carabid in central Europe may be different from that in our region. A survey of the clearest cases shows the following: a) In central Europe more eurytopic (not only on sand; p. 513): Bembidion nigricorne Dromius angustus. b) In central Europe more hygrophilous: Agonum dorsale Leistus ferrugineus Dromius linearis Stomis pumicatus D. melanocephalus Synuchus nivalis. c) In central Europe more along running waters: Agonum marginatum Bembidion velox. A. ruficorne d) In central Europe more on the seashore: Bembidion saxatile e) In central Europe more in bogs: Agonum obscurum Notiophilus germinyi Amara infima Olisthopus rotundatus Bembidion obliquum Patrobus assimilis B. rupestre Pterostichus diligens Bradycellus similis Trichocellus cognatus. Carabus clathratus In the case of the two last-mentioned species the transition to more humid biotopes is already noticeable in southern Sweden. 568 569 499 f) In central Europe more in woodland: Bembidion schüppeli C. problematicus B. semipunctatum Harpalus rufitarsis Carabus arvensis Neberia salina. C. convexus These 6 seemingly different groups can easily be combined under a common criterion: In the northern parts of their area these species live in open habitats (f), on drier soil (b), also on standing, fresh water (c, d) and not exclusively in bogs (e), i.e. in wamer places. The conclusion, expressed as a general rule, would run as follows: Ar the limits of its area an animal or plant species often tends toward stenotopy, which is microclimatically determined. The best examples are on the one hand the southern carabids, which at their northern limit become sand-inhabiting (p. 513) or limestone insects (pp. 188 ff.), and on the other hand the northern species, which toward their southern limit tend to inhabit bogs. Types of Development Detailed studies on the annual development of the nordic carabids were car- ried out by S.G. Larsson (1939, 1940). He saw that these phenomena have considerable zoogeographical significance. With few exceptions (chiefly, a 2-year development), the nordic carabids can be divided into 2 main types, which Larsson designates spring breeders or spring insects (“F”) and autumn breeders or autumn insects (“H”). For various reasons it seems advantageous to use the designations imago hibernators (I) and larval hibernators (L) respectively. First, June, when most carabids copulate and oviposit, at least in Fennoscandia, is hardly spring; second, the hibernating stage—at any rate from the zoogeographical and ecological viewpoint—is the most important expression of the development type; third, it might be pointless to try to introduce the symbols “F” and “H”— which are abbreviations of German words—into international usage, whereas imago (I) and larva (L) are already in general use. In addition to species that more or less clearly represent either of the two types of development, there are those which are very unstable, regularly hibernating both as imago and in the larval stage. These are here designated O (null). On the other hand in our region it never seems to happen that a carabid has more than one generation per year; the use of the expressions “spring generation” and “autumn generation”—especially for several species of Amara—by Burmeister (1939) is probably due to a misunderstanding (see Part I, p. 128, footnote). The results I obtained in connection with the types of development of a series of species, which differ from Larsson’s are not especially cited here, 569-570 500 but information was provided against the species concerned in Part I (“Spe- cific Part”) of this contribution. Partly our different results are simply due to the fact that the carabids in more northerly regions have different conditions of development (it is uncertain whether genetically determined or not); in other cases, in my opinion, Larsson has based his conclusions on too little observation material. The larval hibernators (L) in Fennoscandia comprise a strikingly small number of (at most) only 72 species (20% of the fauna). Those which also hibernate more or less regularly as imago are in parentheses below. The species of Cicindela—with a 2-year development—are also included. Some doubtful cases are indicated with a question mark. ? .o Aepus marinus (Amara apricaria) A. aulica A. bifrons (A. brunnea) A. consularis A. convexiuscula (A. cursitans) Amara equestris A. fulva A. fusca A. majuscula (A. municipalis) A. praetermissa Asaphidion pallipes Bembidion lunatum (Bradycellus harpalinus) (B. verbasci) (Broscus cephalotes) Calathus ambiguus (C. erratus) (€. fuscipes) (C. melanocephalus) C. mollis C. piceus (Carabus coriaceus) (C. glabratus) C. hortensis C. monilis (C. violaceus) (Cicindela campestris) (C. hybrida) (C. maritima) (C. silvatica) (Cychrus caraboides) (Cymindis angularis) (Cymindis humeralis) C. macularis (€. vaporariorum) Dichirotrichus pubescens Dolichus halensis Harpalus calceatus (H. griseus) (H. melancholicus) (H. pubescens) (H. puncticeps) (H. rufus) (H. smaragdinus) Leistus ferrugineus L. rufescens L. rufomarginatus (Masoreus wetterhalli) (Nebria brevicollis) (N. gyllenhali) N. livida (N. salina) Notiophilus germinyi (N. pusillus) (Olisthopus rotundatus) (Patrobus assimilis) (P. atrorufus) Pterostichus lepidus (P. niger) (P. vulgaris) Sphodrus leucophthalmus Synuchus nivalis Trechus discus 570 group—I include 17 species (4.7 501 ? T. fulvus (T. obtusus) T. quadristriatus T. secalis Zabrus tenebrioides. Among the “O-species” with irregular hibernation—a fairly arbitrary %). For the markedly northern species this is tentative: clarification of their development must be left to future detailed studies. Probably the 2-year development is due to the short summer in the fjeld regions and the far north. It is even possible that the ability of some species (for instance, Notiophilus aquaticus) to switch to this kind of development is a characteristic of great positive selection value (already assumed for Lepidoptera by Valle, 1933, p- 98). The suggested “O-species” are: Abax ater Agonum obscurum Amara alpina A. quenseli A. torrida Bembidion grapei Carabus problematicus Harpalus latus H. punctatulus Sl: H. quadripunctatus H. rubripes Nebria nivalis Notiophilus aquaticus Omophron limbatum Patrobus septentrionis Pristonychus terricola ? Pterostichus madidus. Among species that normally hibernate as imago, there are finally some which in exceptional cases may also hibernate as larvae (possibly as pupae). This has been established in at least the following 29 species, but evidently is also possible in some others: Agonum fuliginosum A. lugens A. thoreyi Amara communis* A. lucida A. nitida Bembidion aeneum B. bipunctatum B. transparens Brachynus crepitans Bradycellus collaris Calathus micropterus Carabus clathratus Dromius agilis D. angustus Dromius linearis Harpalus azureus H. hirtipes H. melleti H. tardus H. vernalis Lebia chlorocephala Loricera pilicornis Pterostichus diligens P. gracilis P. minor *“Irregularities” in Amara communis will easily be understood if it turns out to be a species complex (see Supplement). 572 502 P. nigrita Trechus rubens. P. oblongopunctatus The zoogeographical significance of the two types of hibernation (in addi- tion to the mixed type) may be in the first place the different requirements of climate. The situation was considered above (p. 475) in the relevant section. Thus the type of development of the insect indirectly becomes an important area-limiting factor. Besides, it has a considerable indirect effect on the dynamical character- istics of the insect, namely the ability of the species to utilize the possibilities of hydrochorous transport during the winter half-year. These questions were considered in connection with the insular faunas (pp. 205 ff.). Finally, a small observation which supports the view that the hibernating stage of carabids is not a casual function of external conditions (the prevail- ing autumn conditions), but is probably determined genetically (but see also Uvarov, 1931, pp. 105 ff.). In crossing experiments with macropterous and bra- chypterous Prerostichus vulgaris which emerged on August 31, five larvae, which were fed on pieces of Lumbricus, grew rapidly and soon went through the first molt, whereupon they again ate heartily until their new skin was stretched tight. They were kept indoors throughout. Yet without there being any symp- tom of disease it was impossible to make the larvae undergo the second molt. They stopped eating and each of them made a small cave in the soil, in which they lay motionless—apparently “waiting” for the winter. 572 573 Dynamic Factors The distribution of an animal or a plant species depends on where its require- ments for life are met and on the possibility of the species seeking out such suitable regions and permanently colonizing them. These two groups of factors were Clearly distinguished for the first time by Ekman (1922, p. 308), and were termed respectively existence ecological and dispersal ecological (Swedish: “sprid- ningsekologiska”) factors. In the present contribution—for the reasons given above (p. 13)—the terms existence factors and dynamic factors are preferred. The preceding section was devoted to the existence factors. They operate through the interaction between the characteristics of the insect and those of the environment. At first sight the dynamic factors seem unilaterally dependent on the char- acteristics of the animal—on its capability of dispersal. This is incorrect. Pas- sive modes of dispersal often play a greater role than active ones. And the effect of both is determined by the natural barriers against dispersal. Like the preceding section, this one is also written continually distinguish- ing between the “characteristics of the insect” and the “characteristics of the environment.” Theoretically it could be argued that if only enough time were available, each species would finally colonize the entire habitable area. The point is of no real interest. The duration of (geomorphologically, climatically, etc.) comparatively stable geological periods is too short. However, the effect of the modes of dispersal is at any rate directly proportional to the available time. The last section of this book is devoted to the importance of time as a zoogeographical factor—the history of the Fennoscandian fauna. Flight Capacity and Wind Dispersal One of the most interesting characteristics of carabids as objects of zoogeo- graphical study is the fact that there are both species capable of flight and flightless species, and even species (50) where only certain individuals are capable of flight (p. 337). Hence a study of the dynamical significance of flight capacity and the zoogeographical exploitation of aerial transport by these in- sects should be very fruitful. In our carabid fauna the constantly flightless species form only a small component. They can be considered brachypterous, but on the other hand should not be called apterous (wingless), since at least rudiments of the hind wings are always present. The following 49 species (13.6% of the fauna) are constantly flightless: 573-574 574 504 Abax ater Aépus marinus (Agonum ericett) Badister sodalis (Bembidion dauricum) B. unicolor ? Broscus cephalotes Calathus fuscipes C. micropterus Carabus arvensis C. auratus C. cancellatus C. convexus C. coriaceus C. glabratus (C. granulatus) C. hortensis C. intricatus C. menetriesi C. monilis C. nemoralis C. nitens C. problematicus C. violaceus Cychrus caraboides Cymindis angularis C. humeralis (Demetrias monostigma) Diachila polita Dyschirius globosus D. helleni Elaphrus angusticollis Harpalus vernalis Leistus ferrugineus L. rufescens Licinus depressus Metabletus foveatus Patrobus assimilis P. atrorufus Pristonychus terricola Pterostichus aethiops P. fastidiosus P. madidus P. middendorffi Stomis pumicatus (Trechus fulvus) (T. obtusus) T. secalis Trichocellus mannerheimi. In the case of the 6 species in parentheses (see Part I) macropterous individuals are found in other regions but in our region they do not seem to be dimorphic. This is conceivable for Badister sodalis and Elaphrus angusticollis, where the only moderately reduced wing rudiment shows great variability (see p- 338). Nevertheless it is doubtful whether Broscus is to be considered among the functionally brachypterous species. The wings seem always to be fully devel- oped, with a strong, apical reflexed part; the flight musculature, which was studied in several specimens, shows some degenerative characteristics. On the other hand even the first glimpse gives an impression that the elytra are firmly ankylosed immediately behind the scutellum, possibly even fused. But this im- pression is corrected by a dissection. It has been found (Fig. 81a) that the elytra are not fused and are also not anchored to each other in the way de- scribed by Corset (1931) in some constantly brachypterous species (such as Carabus coriaceus, Fig. 81b). On the other hand, the downward bent sutural margins are anchored in the posteriorly converging sharp longitudinal grooves of the metascutellum still more firmly than in Calathus erratus described above 375 505 (p- 350). Moreover, the elytra, behind the part sketched in the figure, are firmly joined through sharp sutural ridges; these ridges belong alternately to the right elytron (in the middle) and to the left elytron (in front of and behind the middle). All these structures together form an effective locking mechanism, which could be released only by strong lateral movements and by lowering the strongly chitinized longitudinal folds bounding the longitudinal grooves of the metascutellum on the outside. I have not been able to decide whether such movements are possible. Good examples of structurally much simpler sutural margins in carabids capable of flight are provided by some species illustrated by Corset (1931) (Fig. 81c). However, it should be noted that ankylosis of all kinds between the elytra may often look even in constantly brachypter- ous Species (for example Bembidion unicolor, Metabletus foveatus, Patrobus atrorufus). According to Sharp (1913) the elytra of Prerostichus madidus are not constantly ankylosed. If Broscus were actually capable of flight, it would be strange that this has never been observed in such a large, conspicuous and widely distributed beetle. It is also significant that this beetle is missing from the sandy outer islands in the Gulf of Finland, where it would undoubtedly find the best possibilities of existence. The occurrence in Gotska Sandon can perhaps be explained by transport through water (p. 285). In light of the above discussion we may be justified in including Broscus cephalotes—at least provisionally—among the flightless species. The only two other doubtful species are Leistus ferrugineus and rufescens, whose wings are comparatively well developed, but smaller and weaker than Fig. 81. Transverse section through the anterior part of the suture of elytra. a—Broscus cephalotes (about 1 mm behind scutellum, posterior view); b— Carabus coriaceus; c—Penetratus rufipennis Dej. (b and c after Corset, 1931). 576 506 in L. rufomarginatus (see Part I). At any rate it seems to be ruled out that L. ferrugineus could be capable of flight. Both species are considered here as constantly brachypterous. Carabids constantly capable of flight, numbering 263 species (73%), con- stitute the major part of our fauna.* It might have been more prudent to designate these “constantly macropterous,” since the flight capacity can be established beyond doubt only by observations or experiments. Definite records of flight are available for 177 species (for details, ref- erence may be made to Part I and to the Supplement to this part). In the following list, species with flight records only outside the region are given in brackets. On the other hand there is no reason to distinguish between species observed flying spontaneously in nature and those impelled to fly only by exposure to sunlight, warming up, etc. Species demonstrably capable of flight** Acupalpus consputus Amara aenea A. dorsalis A. apricaria A. exiguus (A. aulica) A. flavicollis A. bifrons*** A. meridianus A. communis Agonum assimile A. consularis (A. bogemanni) (A. crenata) A. consimile A. eurynota A. dolens A. familiaris A. dorsale A. fulva A. gracile A. ingenua (A. gracilipes) A. interstitialis (A. livens) A. lunicollis (A. longiventre) A. majuscula A. lugens A. montivaga (A. marginatum) A. municipalis A. micans A. ovata A. mülleri A. plebeja A. piceum A. praetermissa A. quadripunctatum A. similata A. thoreyi (A. spreta) A. versutum A. tibialis A. viduum Anisodactylus binotatus *“Constantly capable of flight” here covers dimorphic species that occur exclusively in the macropterous form in Fennoscandia. **Dimorphic species that have been observed flying are mentioned on p. 337. ***Caught in the flight-apparatus by Ossiannilsson (see Supplement). 577 578 (Asaphidion flavipes) A. pallipes (Badister bipustulatus) B. dilatatus B. peltatus B. unipustulatus Bembidion argenteolum B. articulatum (B. biguttatum) B. bipunctatum (B. dentellum) B. difficile B. doris B. femoratum (B. fumigatum) B. hasti B. hirmocoleum B. hyperboraeorum B. illigeri B. lapponicum B. litorale B. lunatum B. lunulatum B. minimum (B. nitidulum) B. obliquum (B. pallidipenne) B. prasinum B. punctulatum B. quadrimaculatum (B. quinquestriatum) (B. ruficolle) B. rupestre B. siebkei B. stephensi (B. striatum) (B. tibiale) B. tinctum B. varium Bembidion velox Blethisa multipunctata 507 Bradycellus similis B. verbasci Calathus ambiguus (Calosoma auropunctatum) (C. denticolle) (C. inquisitor) C. sycophanta Chlaenius nigricornis C. quadrisulcatus (C. sulcicollis) C. tristis Cicindela campestris C. hybrida C. maritima C. silvatica (Clivina collaris) C. fossor Demetrias imperialis Diachila arctica (Dichirotrichus pubescens) D. rufithorax* (Dolichus halensis) Dromius agilis D. angustus D. fenestratus D. longiceps D. marginellus (D. quadraticollis) D. quadrimaculatus (D. quadrinotatus) (Dyschirius aeneus) ? ( D. impunctipennis) D. ludersi (D. nitidus) (D. obscurus) D. politus D. thoracicus Elaphrus cupreus E. riparius Harpalus aeneus H. anxius *Caught in the flight-apparatus by Ossiannilsson (see Supplement). 579 508 H. calceatus H. distinguendus (A. frölichi) H. griscus (H. hirtipes) HA. latus H. melleti H. pubescens H. punctatulus (H. puncticeps) H. rubripes H. rufitarsis H. rupicola (H. seladon) (H. serripes) (H. smaragdinus) (Omophron limbatum) Oodes gracilis O. helopioides Patrobus septentrionis Pelophila borealis (Perileptus areolatus) Pogonus luridipennis Pterostichus adstrictus (P. aterrimus) P. coerulescens P. cupreus (P. gracilis) (P. niger) P. nigrita (Sphodrus leucophthalmus) Stenolophus mixtus (A. tardus) H. winkleri Lebia chlorocephala (Tachys bistriatus) T. bisulcatus Trechus discus L. crux-minor T. micros L. cyanocephala T. quadristriatus Loricera pilicornis T. rubens Microlestes minutulus Miscodera arctica Nebria brevicollis Trichocellus cognatus (Zabrus tenebrioides). Accordingly 86 constantly macropterous species are left, for which there is no clear proof of flight capacity. Enumeration seems superfluous. Some of them show almost definite proof of flight capacity by their occurrence, for instance, in drift material on the seashore, but there is no such proof in other cases. However, comparison of the relative wing size of these species with their closest relatives provides grounds for a more or less definite verdict on the functional ability of these organs. Perhaps it would be possible to calculate a “flight index” based on the relationship between wing surface and body weight (cf. Prochnow, 1921-24, p. 564), to which at least within each genus some sig- nificance could be attached as a “measure of flight capacity”—assuming the flight mechanism to be identical. But my attempt to establish this index failed due to the great variability in the body weight even of the same individual (on account of uptake of food and water, presence of eggs in the female, etc.); the dry weight cannot be used. The calculated length or surface relationship be- tween the hind wings and the elytra is of no use either; compare the difference between an Odacantha and a Lebia! 580 509 It is left to one’s judgment whether the development of the hind wings ofa particular carabid enables it to fly or not. Actually there are very few cases admit- ting of real doubt: I will mention only Agonum ruficorne, Amara equestris, Oda- cantha melanura and Prerostichus oblongopunctatus, which were nevertheless, probably correctly included among constantly macropterous species. In line with the above, the Fennoscandian carabids are divisible into the following dynamic groups: Constantly macropterous species, demonstrably capable of flight: 177 = 48.9% of the fauna. Constantly macropterous species, without records of flight: 86 = 23.7% of the fauna. Dimorphic species (p. 337): 50 = 13.8% of the fauna. Constantly brachypterous species: 49 = 13.6% of the fauna. The first two groups can naturally be combined, although they were kept separate in considering the dynamics of the insular faunas (pp. 198 ff.). It can be assumed that the first group ( “m,” p. 206) includes the best, most regular fliers, and represents the most mobile elements of the fauna. Otherwise the different carabids capable of flight are by no means dynam- ically on a par. First of all we have a very small group of species for which flying is as common an activity as running, particularly in the sunshine; all of them are markedly heliophilous. These species use their wings chiefly to hunt their prey, to escape from danger, etc. In our region, these obligatory fliers include only members of the genus Cicindela and the subgenera Bracteon and Chryso- bracteon of Bembidion, making the following 9 species: Bembidion argenteolum Cicindela campestris B. lapponicum C. hybrida B. litorale C. maritima B. striatum C. silvatica. B. velox Most of the other carabids capable of flight use their wings only in excep- tional cases. Often one may have to conduct many unsuccessful experiments with exposure to sun, artificial light, heat (in dry and in humid air), before at best, flight is induced. One gets the definite impression that the decisive factor is not the external conditions but the disposition of the insect, “the inclination to take flight.” First of all it is clear that the insects do not fly in all seasons. From this viewpoint I have prepared a synopsis (Diagram 54) of the monthly distribution in the region of all specimens of carabids (excluding Cicindela, Bracteon, and Chrysobracteon) observed spontaneously flying. It was found advantageous to divide the material into two groups: Imago hibernators (which hibernate only exceptionally as larvae), and more or less regular /arval hibernators (see p. 568). 581 510 80 60 40 20 Number of specimens April May June July August September Diagram 54. Number of specimens of Carabidae in Fennoscandia observed spontaneously flying in different months. (Bracteon, Chrysobracteon and Cicindela not included.) Continuous line = Imago hibernators. Broken line = Larval hibernators. From the diagram it is clear that the larval hibernators (s.l.) generally fly less: altogether 39 specimens were observed as against 227 among the imago hibernators. The former included a smaller number of species. In the Fennoscandian fauna the ratio of “L species” to “I species” is 1 : 3. Moreover the former comprise comparatively more constantly flightless species (22 as against 27) as well as dimorphic species (16 as against 34). Bearing this in mind we obtain the following figures. Larval hibernators capable of flight (including dimorphic species): 67 species. Imago hibernators capable of flight (including dimorphic species): 246 species. Ratio 1 : 3.7. Larval hibernators capable of flight (excluding dimorphic species: 51 species. Imago hibernators capable of flight (excluding dimorphic species): 212 species. Ratio 1: 4.2. However, the observations on flight depicted in Diagram 54 and summa- rized above give a ratio of 1 : 5.8 between “L species” and “I species.” Even remembering that the “L species” are less well endowed with flight capacity, we conclude that the larval hibernators are less disposed to flight. We are not far from the hypothesis that the flight of carabids is often associated with the hibernation. At any rate, in the carabids there is no question of a “nuptial flight.” Nor 582 Sit is flight in the males for the purpose of seeking the female: the two sexes participate in flight in about equal numbers.* The flight of carabids is primarily to bring about a rapid, occasionally major change of quarters. This change of quarters using flight is in different seasons for the larval and for the imago hibernators. In the former it comes as soon after emergence as they are hardened enough to use their wings, in midsummer, when the biotopes occasionally dry up and become uninhabitable. The larvae could hardly escape with their slow locomotion. The imago hibernators on the other hand emerge late in the summer half- year and occasionally remain in the pupal stage through the winter. During that season there is seldom any danger of drying up, and the falling temperatures (especially at night) make the flight more difficult. In the next spring, when the insect wakes up, it is usually situated where the larva lived last summer. But in spring the locality may offer very different living conditions, especially with regard to moisture. Less often the imago uses flight to emigrate from a summer biotope markedly inundated in autumn to a drier locality. Hence in the imago hibernators the change in quarters takes place chiefly in spring, and Diagram 54 clearly shows that by far the most numerous observations of spontaneous flight were made in May. During my study of the riparian fauna at Osby Lake (Upl Djursholm; Lindroth, 1943a) I had the opportunity to study this change of quarters more closely (see also Palmen, 1945). H. Krogerus (1948) developed the same theme to a greater extent after extensive study. It was shown, among other things, that on Ösbysjön the riparian fauna is extraordinarily unstable. In winter, only a few species remain on the extreme edge of the shore; several of them gradually move up to hibernate on drier ground, and finally some cover long distances using their wings. Good examples are the Chlaenius species, Oodes gracilis, Pterostichus aterrimus, and also some species of Agonum, Bembidion, etc. In the case of Oodes gracilis it was shown that the inclination for flight is strong in spring, but in midsummer vanishes to the extent that the insects cannot be induced to fly by any means. Yet, there can be no question of a reduction in the flight muscles, etc. In autumn the inclination and capacity to fly reappear, not only in freshly emerged specimens but also in ones at least a year old, which showed incapacity for flight in summer. The significance of flight capacity for riparian species of fresh water mar- gins is therefore quite clear. Darlington (1936, p. 159) correctly emphasizes that the possibility of a rapid change of quarters is not so important for any other ecological group, not only on account of the seasonal changes in shore biotopes, but also because of frequent catastrophes caused by flooding. He points out (p. 160), that the animals of the seashore live under more stable *In this connection unfortunately no statistical data can be given, since the sex of the flying individuals was not noted in enough of cases. 583 584 312 conditions and are therefore more commonly flightless. Examples of this in our fauna are Aépus marinus and Trechus fulvus. The only species bound to fresh water shores that seems to be constantly flightless is Elaphrus angusticollis. Palmén (1944, p. 133) vividly describes how a population of Bembidion doris saved itself from the impending drying up of its biotope by flying away. The importance of flight is of course clearer in the case of riparian species than in other ecological groups. But one must not presume some causation different in principle. The best support for assuming that the flight capacity of the carabids serves mainly (at least in imago hibernators) for change of quarters seems to be the dominance of this phenomenon in spring, a season when a change of quarters is most often necessary. For even in biotopes other than shores this season brings about the most extensive and the most sudden changes, especially with the melting of snow. I cannot judge Palmén’s hypothesis (1944, p. 126; cf. Glick, 1939, p. 129) that electrical disturbances in the atmosphere may act as a strong stimulus to flight. A decisive question from the zoogeographical viewpoint is whether the direction of flight of insects is totally accidentally determined. We pass over the fact that the sense organs must produce some movement toward surfaces that suit the insect. It is not known how this happens in carabids. However, it must be assumed that a hygrophilous carabid, for instance, is capable of sensing the proximity of a body of water during flight, bringing about a change of course if need be. Here we should investigate whether under identical conditions of the earth’s surface other factors, so to speak inconsequential for life, can affect the flight direction of carabids. Two factors suggest themselves: air currents and light, chiefly that of the sun. Most carabids are such poor fliers that atmospheric currents—chiefly those directed horizontally, i.e. winds—cannot but affect the direction of flight. Exceptions are the 9 species of Cicindela, Chrysobracteon and Bracteon men- tioned on p. 579 above. Demoll (1918, p. 6) gives the velocity of flight of Cicindela (tiger beetle) as 1.8-2.3 m/sec. (The frequency of wing beats in C. campestris is 82-87 according to Sotavalta [1947, p. 97].) The others in my experience fly so slowly that they are easily overtaken without running. For Oodes gracilis, a comparatively good flier, I measured the velocity of flight with a stopwatch by making the specimens fly indoors (24°C). They always flew to the window 5 m away. I obtained a mean value of 1.6 m/sec from 20 readings (extreme values: 1.25 and 1.85). The variations may be due to deviations from rectilinear flight. So the flying carabids can be carried along by comparatively weak winds or blown off course. In a very light wind, according to several observations, they mostly fly against it. Interesting results in this field, chiefly with Diptera and Lepidoptera, were 585 586 513 obtained by Lutz (1927), who collected night-flying insects in 8 traps placed in a circle and noted the distribution in them according to the prevailing wind. He established that in a light wind the insects fly chiefly against it (cf. also Glick, 1939, p. 114). As he correctly emphasized, this observation does not contradict the fact that flying insects are carried in the direction of the wind even by a moderate breeze, whichever way they are headed. This experiment certainly applies fully to carabids. Since, with the above- mentioned exceptions, they are slow fliers it is clear that change of place even with a light wind is largely determined by it. The insects are more or less pas- sively carried by strong winds. The complete dependence of the anemochorous transport of insects on the direction of the wind has been best elucidated by Palmen (1944, for instance, pp. 194 ff.). Examples of wind transport of cara- bids were also given above (pp. 282, 287, 295) in connection with the insular faunas. In view of the extensive literature on the transport of insects by wind, reference may also be made to the summaries by Holdhaus (1927-28, pp. 599 ff.) and Uvarov (1931, pp. 116 ff.). In recent years the subject of the anemochorous transport of animals and plants has been greatly extended by collections from aircraft: it has acquired a firm footing. Wind transport of pollen and cryptogam spores, even over long distances, was already known earlier (for example Bror Pettersson, 1940), but the large component of insects in aeroplankton was a surprise. By far the most extensive collections were made in North America, chiefly to a height of 1500 m (5000 feet). On the basis of the data on catches provided by Glick (1939), three diagrams (Diagrams 55-57) are published here to illustrate the distribution and relative abundance of the insect groups collected at different altitudes. Among the flying insects (Diagram 55) the Diptera predominate, consti- tuting 40% of the entire material. They are followed by Hemiptera (nearly 17%), the Coleoptera (a good 14%), and the Hymenoptera (11%). The great- est altitude (more than 4000 m) were attained by Hemiptera (Cicadina), Diptera (Tipulidae, Cecidiomyidae) and Hymenoptera. Division of the collected Coleoptera into the larger families (Diagram 56) shows a substantial dominance of Staphylinidae (39%), which are represented at almost all altitudes up to 3300 m; these are followed by Chrysomelidae (14%), found regularly up to 1500 m, and one Diabrotica at 3300 m. The Carabidae occupy third place, which with 470 specimens—including 30 defi- nitely determined different species—comprise a good 13% of the entire beetle material. They were found regularly up to a height of 1800 m; at 2400 m one Tachys and at 3000 m one Microlestes (Blechrus) were collected. A really big surprise among the results Glick obtained—though fore- shadowed by Berland’s investigations (for example, 1935, 1937)—is the regular occurrence of flightless arthropods in aeroplankton (Diagram 57). With regard to the araneids—which do have functional flight capacity—this was expected: 585 514 4,000 3,000 2,000 1,000 e19}d0y}19 B19}dOs] eIJUIPOL1OT) epııawaydy eJEUOPO e1ajdouesAy J, 219}d019}9}4 e1a}dowoy e19)}d01nIN e¥19}d03]07) e19)d099 | e1a}doyoi 7, e19}d1q e19JdopıdaT e19)doımwÄH Diagram 55. Distribution and relative abundance of flying insects in the atmosphere according to data supplied by Glick (1939). Isolated samples are indicated by crosses. they even reached the greatest measured altitude of 4500 m. But even acrids and wingless insects as well as flightless immature stages of winged species were regularly found at heights of more than 500 m. A Collembola (Bourletiella) was collected at 3300 m. Of particular interest to us is the fact that a coleopteran larva (Trogoderma, family Dermestidae) was collected at 2700 m, and that it 587 was alive. It was the greatest altitude at which any living animal was found in these collections (Glick, 1939, p. 93). The impressive documentary material obtained with these American in- vestigations leads to the following conclusion: Most Carabidae (exceptions given on p. 579) are poor fliers and being easily captured by ascending con- vection currents may be carried into the upper air. Berland (1935, p. 91; 1937, p. 26) notes that only inept fliers let themselves be carried up to great heights. 588 It is obvious that the Lepidoptera and Odonata are less abundant in Glick’s material. In higher layers of the air, the velocity of wind is much stronger than near the ground; and above 500 m it is more stable (Ostman, 1933, p. 16; Angstrom, 1946, p. 86)—even with changing wind conditions close to the ground. At that height (600 m or above) Glick caught 35 specimens of Carabidae. Insects which happen to come in altitudes that high have an unique possibility at their disposal, the long-distance transport. The best-known exam- 586 515 3,000 2,000 ? 1,000 — = aepiqesesy SED N ay, Jeprseue) SepıjawosA1yy 3epı1][2U1DI0) sepıdeydojdA1j 3epıuonndın) SEPIPOlet IEPI4199019}9 Fy sepijiydompky Seppe SBPIAPHIN deplaeqesess aepiurpAydeys Diagram 56. Distribution and relative abundance of larger families of Coleop- tera in the atmosphere, according to data supplied by Glick (1939). Crosses indicate isolated finds, question marks indicate findings made above 1500 m without exact determination of altitude. ple to date is that of the occurrence of large numbers of the aphid Dilachnus piceae Panz. and of Syrphus ribesii L. in the snowfields of northeastern Spitzber- gen, where they could have arrived (Dilachnus is associated with Picea) only by wind transport from over 1300 km away (Elton, 1925, pp. 291 ff.). Does this transportation in the upper air hold any special importance for the dispersal of Carabidae? In this connection the following facts may be considered: 1. The aeroplankton “population” is extremely light. For the stratum just above the 500 m limit (“2000 feet”) Glick calculated the mean volume of air per insect as 877 m’. Per carabid (corresponding with the component of this family in the fauna above 500 m) this means a good 65,000 m?, and for the species (Stenocellus tantillus Dej.) that was most numerous above 500 m (represented by 4 specimens) it means a good 570,000 m? per individual.* It cannot be denied that such an individual can survive long-distance trans- port through the upper air—Glick (1939, p. 93) found living coleopteran adults at least at 600—and that it can accidentally land in a suitable region not earlier colonized by the species. But the chance of its landing sufficiently close to an- * All these calculations are depending on the collecting equipment being 100% efficient, on which I cannot pass judgment. 587 589 516 3,000 x x x 2,000 x x 1,000 mo x x | | x @) T = > = } S eines Diss Zi fw 5) te = > eau ie re eee moe = = = = a For On er’ NES ES a IES wz ing) ae ieee | oO S w 3 > Zi ON EN © et mn oo) = =. Io 5 5 ER DIN sis Geo ame 28 2k Ss Oo. c = Es) ze ae een erw = = = oO INS SOOM 2. So Ena Ss 5 Sa @ , eu a a 2 SM aire N TER > Diagram 57. Distribution and relative abundance of flightless arthropods in the atmosphere according to data supplied by Glick (1939). Isolated samples are indicated by crosses. other individual of the opposite sex, arriving by the same means, is zero. This kind of colonization becomes a real possibility only if it is shown that even impregnated females can stray into the upper air during flight; this question is touched on below (p. 595). On the other hand—in contrast to the carabids—long-distance dispersal through the upper air can play a role in the expansion of the area of species that show periodic or nonperiodic mass multiplication, such as phytophages, including their parasites and more or less monophagous predators, and also parthenogenetic animals. 2. The only unambiguous cases of wind transport of insects that have been precisely analyzed, namely insects washed ashore on the north coast of the Gulf of Finland, show such good correspondence with the wind conditions prevailing, during transport, in the air strata close to the ground that the dependence on them seems indubitable (Palmen, 1944, p. 93). At any rate, wind transport at low heights (about < 500 m) and over moderate distances is 590 S17, the normal phenomenon. Moreover, Glick’s material (Diagrams 55-57) shows how much richer the fauna of the lowest 300 m is. 3. If long-distance dispersal through the upper air were a normal phe- nomenon and regularly led to colonization, then—in my opinion—the dis- tribution of some carabid species would have definitely been different. Fore- most the area displacements, for instance the post-glacial immigration into Fennoscandia, would have involved erratic advances rather than a gradual shift. A species like Amara majuscula, which on account of its pronounced inclina- tion to flight has produced true swarming flights in Poland, and is especially suited for anemochorous transport, certainly would not have had such regular immigration into Fennoscandia as indicated by the map in Fig. 84 (p. 625) if this represented the consequence of long-distance dispersal in the upper air. The only species whose occurrence in Fennoscandia one might be inclined to ascribe to such transport would be Lionychus quadrillum for its isolated locality near Nke Orebro. However, here transport by man seems equally probable. The answer to the question raised above is therefore: /ong-distance trans- port in the upper air, as a mode of dispersal for carabids, plays at the most a minor role. On the other hand winds close to the ground and the speed of flight are of decisive importance. Favorable winds are not only responsible for the insects washed ashore along the Gulf of Finland, but without them Amara majuscula, as well as those immigrants discussed above (pp. 282, 287, 291), probably emigrating from the eastern Baltic, in the fauna of Gotska-Sandon, Faron, and Gotland, could not have covered the much longer aerial route. Hence the Baltic Sea is not an effective barrier against the dispersal of flying forms, which is illustrated below (p. 610) with further examples. It is interesting—but basically to be expected—that a good active flight capacity, which among carabids is possessed only by Cicindela, Bracteon, and Chrysobracteon, actually diminishes the dispersal capacity (similar findings were made by Ekman, 1922, pp. 333-334, by comparing the dispersal tendencies of mammals and birds). This can be explained by the fact that these animals are only rarely transported passively (by especially strong winds). The result is that they often show an inexplicable conservative attachment to small sec- tions of biotopes that do not perceptibly differ from the surroundings—at any rate when viewed with human eyes. Such an observation is made by Krogerus (1932, p. 238) with regard to Cicindela maritima near N| Tvarminne, and Ho- rion (1937, p. 9) reports two localities of Bembidion litorale (which is otherwise stenotopic, bound to river-banks) on a standing body of fresh water. It turned out to be a cut-off old river course so that these were probably relict occur- rences. Moreover, the Fennoscandian area of these two species, and that of Bembidion argenteolum, as a whole has a markedly “conservative” imprint. An exception is provided by the two apparently accidental records of Bembidion (Bracteon) striatum near Nl Tvarminne and in Mgen in Denmark. 591 518 Starting from Glick’s material it is necessary to take up another, very important question: Is it conceivable—whatever the height above ground—that flightless carabids or immature stages of carabids can be transported by the wind? Even apart from the Araneida, which are functionally capable of flight sus- pended from their threads, remarkably many-wingless animals have been found in aerial plankton collections (Diagram 57). Glick’s numerical data (1939) are given below: Acarina 44 specimens Thysanura 40 specimens Collembola 26 specimens Aphaniptera 1 specimen Formicidae 20 specimens Orthoptera larva 1 specimen Hemiptera larvae 29 specimens Coleoptera larva 1 specimen Lepidoptera larvae 5 specimens Diptera larva 1 specimen. This shows that these animals belong to any one (possibly two or more) of the categories below: 1. Very small animals, for example, Acarina, Collembola. 2. Weakly chitinized—hence “light”—animals, for example, immature Hemiptera. 3. Animals with dense pubescence or other appendages, for example, Thysanura, Trogoderma larva (Coleoptera). 4. Animals that spin threads, for example, Lepidoptera larvae, of which 3 were found to belong to the family Gelechiidae. 5. Animals that live in the open and visit the upper parts of the vegetation, for example, workers of Formicidae. Adult carabids can only exceptionally be considered in categories 1 and 5. Undoubtedly they are safe from being captured and carried up by convec- tion currents chiefly because of their marked chitinization—and their “weight” (which can be appreciable in brachypterous species). The earlier stages of cara- bids, especially eggs and pupae, all inhabit very concealed habitats. The larvae have no appendages (the cerci are rather short and weakly developed) nor any pronounced pubescence that could catch the wind. It is significant that the only coleopteran larva in Glick’s material belongs to the family Dermestidae, in which the larvae are characterized by very dense pubescence, arranged in tufts. These theoretical considerations support the view that little significance need be attached to wind transport of flightless carabid adults or of their immature stages. However, there are also facts to prove it. The clearest are provided by maps of the distribution of macropterous and brachypterous forms of some 592 593 319 dimorphic species. We found earlier that the area of a dimorphic species in process of dispersal is characterized by the presence of a more or less broad “belt” of purely macropterous populations at the area limit. Good examples are Calathus mollis (Fig. 28, p. 368), C. erratus (Fig. 35, p. 376), Bembidion aeneum (Fig. 49, p. 400), and B. grapei (Fig. 50, p. 402). This fact is compre- hensible only if regular dispersal through the air involves only carabids capable of flight. In Fennoscandia there is no record of a flightless carabid so isolated and unexpected as to suggest long-distance wind transport. On the other hand Gislén (1940, pp. 19 ff.), for instance, rightly attaches great importance to the dispersal of flightless “very small animals” as aero- plankton. But besides the advantages (also emphasized by Gislén in a later contribution, 1948, p. 121) of the small weight of these animals and their tenacity in the encysted condition (l.c., p. 21) many of them can reproduce asexually (for example, in Protozoa and Rotifera). In conclusion it may be stated, with regard to the wind transport of cara- bids, that it may involve only imagines, and of these only individuals capable of flight. At any rate, other cases seem to be so rare that they are inconsequen- tial for the expansion of area. However, for carabids capable of flight, wind transport has a decisive role, both by extending the distance covered and by determining the direction of the flight. There is no evidence of the effect of long-distance transport in the higher layers of the air. The second factor that can affect the direction of flight of an insect is the light. It is an age-old observation that many night-flying insects, including carabids, are attracted to artificial light. But so far no one has remarked that “natural” light, the sun, could have a similar effect. Two observations led me to these considerations: First, the conditions on many islands of the Baltic Sea, chiefly on Aland and Gotland, including the small neighboring islands. It is clear that the flying element of the carabid fauna of these islands has come more from the east than from the west, even though the prevailing winds—especially in the case of Aland—might favor migration from the west (pp. 254 ff.). Second, the fact that 4 (possibly 5) species in Sweden each have an isolated locality in the Bothnian coastal region far north of the continuous area, which extends much farther north in Finland. See Table on p. 520. Hence these isolated localities are much more closely related to the Finnish than to the Swedish area, which is hard to explain except by migration across the Bothnian Sea from the east. All these species are demonstrably capable of flight. The explanation is confirmed in the case of Anisodactylus by the finding of this species on the tiny island of Oa Norrskar in Kvarken. It is interesting that we know of no species that colonized a new area by flying in the opposite direction across the Bothnian Sea—from Sweden to Finland. And the winds are of course favorable for such transport, even in the higher layers of the air 594 Isolated locality Elsewhere in In Finland Sweden or northward as northward as faras far as Acupalpus dorsalis Vbt Vannas 63° 55 His 61° 15 Om 64° 7 Anisodactylus binotatus Ang Örnsköldsvik 63° 15 Hls 61° 48 Om 64° 50 Asaphidion flavipes Ang Örnsköldsvik 63° 15 Gst 60° 20' Oa 62° 50 Chlaenius tristis Mdp Alnon 62° 25 Upl 60° 7 St 61° 30 Oodes helopioides Mdp Timrä 62° 28 Gst 60° 45 Oa 62° 58! !In the Swedish inland (Dir) Asaphidion flavipes was found as far as 61° 42’, in eastern Finland as far as 61° 20’. Oodes helopioides was also found farther north in eastern Finland. (Ostman, 1933, p. 30). The following conclusion is obvious: The flight direction of these species is affected by some factor independent of the wind conditions. To test how far the sun might be such a factor, I performed the experiments mentioned above (p. 256) in a special “flight direction apparatus” (Experiment 147 ff., p. 109). It was found that the 6 species tested—with the exception of Badister peltatus (for which inadequate material was available)—show a distinct inclination to fly toward the sun. It is interesting that Acupalpus dorsalis is one of the species we mentioned above as an example of a species that arrived in Sweden from Finland across the Bothnian Sea. Especially decisive for the orienting effect of the sun on the flight of an insect species responding positively to it, is the daily time of flight. It is to be assumed that the closer the sun is to the horizon, the greater its effect; hence insects that fly at the time of sunrise or sunset should be particularly affected. In addition to light, a sufficiently high temperature is a key factor for flight activity; the flight of carabids is closely dependent on it (see, for instance, McClure, 1943, p. 38). It can therefore be argued that for species that—exclusively or predominantly—fly in weak daylight the morning hours are rarely warm enough, so they become more or less regular evening fliers (McClure, l.c.).* For the same reason the nocturnal insects fly chiefly in the first half of the night when, for an hour or so, the afterglow is still effective. Accordingly, the flight direction of the species that fly during the hottest mid- day hours should be determined least of all by the sun. Carabids are chiefly evening fliers—the above mentioned Acupalpus dorsalis included, of which mass flight was observed in the evening (Palmen, 1946, p. 32). A considerable *McClure made systematic collections (morning and evening) of flying insects in Kentucky. In his material the carabids were represented by 88 specimens belonging to 20 species, of which 4 were collected in the morning and 84 in the evening. Sp number of carabids fly exclusively at night and can be collected in a light trap. Further information is given against each species in Part I of this contribution. If the carabids use their wings chiefly in the evening or during the first hours of the night, it follows that, to the extent that they generally respond positively to the sun, they tend to fly west. In this way it appears to me that a plausible explanation is provided for the two above-mentioned observations with regard to the flying element on certain Baltic Sea islands and on either side of the Bothnian Sea. It would also be worth testing other species in the “flight direction apparatus,” for exam- ple Amara majuscula, whose unprecedentedly rapid immigration, apparently emanating from the southeast, is discussed later (p. 622), and Bembidion trans- parens, whose macropterous stock in the south Baltic region is unmistakably of eastern origin. The general effect of the “sun rule,” not only on carabids but on in- sects generally, cannot be decided in the present state of our knowledge. First of all it must be established for every species of insect whether it responds positively to the sun and what time of day it usually flies (with regard to night- flying Lepidoptera, in this connection, see Williams, 1939, pp. 119 ff. ; Ellinor Bro Larsen, 1943). It must be pointed out that the two clearest examples of Lepidoptera that have immigrated recently from the east, Phytometra confusa Steph. (Nordstrom, 1945) and Eupithecia sinuosaria Ev. (Wahlgren, 1921)*, are night-flying insects, whereas generally well-known immigrants (some of which are only transgrading species) which, as far as is known have arrived from the south, such as Colias hyale L., C. electo L., Pyrameis atalanta L., P. cardui L., and Phytometra gamma L., and are diurnal insects**. If, as I expect, it turns out that the sun affects the flight direction of a large number of flying insects, this will probably also throw light on the important role played by the Siberian fauna in the post-glacial recolonization of northern and southwestern Europe. Perhaps it will even be possible to show that at high latitudes (i.e. above the Arctic Circle) insect flight is somewhat deflected toward the pole, on account of the summer position of the sun, but near the Equator is more strictly inclined toward the west. The significance of flight capacity and of wind transport for the dispersal of a species of insect can be correctly estimated only if it is established whether impregnated females fly or not.*** It is clear that if they do, the possibility of permanent colonization of a new region is greatly enhanced. *] cannot agree with Wahlgren’s view (1912, p. 161), that the rapid dispersal of Eupithecia sinuosaria westward is chiefly due to transport by man. **Tt must also be investigated whether the direction of migration of the transgrading sphingids (Acherontia, etc.) is governed by particular factors. ***Tn parthenogenetic species, which, as far as is known, do not occur in the family Cara- bidae, this question does not arise. 596 59 SI 522 Experimentally this could be tested by two methods: On the one side by isolating females found flying in nature, to establish the possibility of laying fertilized eggs; on the other side by inducing females after observed copulation to fly, later ascertaining whether they were actually impregnated. I have tried both methods, but unfortunately with very little material and only moderate success. Females of the following species caught during flight were isolated: Amara aenea (caught on May 13), Amara familiaris (3 specimens : May 6, May 13, May 13), A. ovata (May 23), A. similata (May), Bembidion rupestre (May 13), Harpalus distinguendus (3 specimens, May 13). Oviposition could not be established in any of these cases. I tried the other approach with Oodes gracilis. Five pairs of this species were kept separately (May-June 1945) in glass dishes. Copulation was ob- served only in one case, but had probably taken place in all cases. After a few days the females were compelled to fly using strong artificial light, which succeeded in all cases, after which each of them was isolated in its glass dish. In two of them a very small larva was discovered in each after 11 and 13 days respectively, and another after 16 and 19 days respectively. So these two females flew with fertilized eggs. It was thus shown that impregnated females can fly at least in certain cases; it is another matter whether they can do so in nature. As discussed above (p. 581), the flight of carabids in most cases (especially in imago hibernators) serves to change the quarter. Meeting of the sexes and copulation may take place, with few exceptions, only at the summer abode, after the obligatory flight. Thereafter the next occasion for flight is normally only in autumn—assuming the adults, which have already overwintered once, survive. There are exceptions: hygrophilous insects living on very small, particularly standing bodies of water can be threatened in summer with desiccation and compelled to escape through the air (p. 582). In line with this, we cannot ignore the possibility that two dimorphic riparian species (Bembidion assimile and B. transparens, p. 395) have been able to spread in the brachypterous form by means of the flight of impregnated macropterous females. On the other hand this does not seem to happen in the case of the chiefly ripicolous Bembidion aeneum (Fig. 49, p. 400), nor in the case of B. grapei (Fig. 50, p. 402), Calathus erratus (Fig. 35, p. 376) or C. mollis (Fig. 28, p. 368), since these species have a “purely” macropterous stock at the periphery of their area. I believe that the flight of impregnated females in Carabidae is a rather rare exception. But this important question must be investigated carefully with more material, preferably the anemohydrochorously transported insects common on the Finnish south coast (p. 604; Palmén, 1944). For this purpose females still swimming on the water, which have not yet found males, should be isolated to determine to what extent they are impregnated. I believe it will be found that they are not. This is supported by the fact that so many species repeatedly found in this drift material have failed to colonize the new 598 523 region permanently (indicated by an asterisk in Palmén, 1944, pp. 37 ff.). Of these, among the carabids at least for Acupalpus flavicoilis, Bembidion assim- ile and B. illigeri the necessary climatic and other existence requirements in southwest Finland can hardly be lacking (see maps in Part II). Of course the first-mentioned species (according to a personal communication from Palmén) now (since 1946) seems to have actually become resident in the Tvarminne region. The following conclusions can be drawn on the dynamical significance of flight and of wind transport for carabids: 1. The few good fliers (Cicindela, 5 species of Bembidion) are not trans- ported over long distances. They are strikingly “conservative” and show slow dispersal. 2. The other forms capable of flight are largely influenced by the wind and can be transported with it over long distances, for example across the Baltic Sea. 3. Transport in higher layers of the air (> 500 m) has at the most minor significance. 4. The direction of flight is also affected by the sun, such that insects flying in the evening and at the beginning of the night—moving toward the sun—have an inclination to fly west. 5. The area-expanding effect of flight and of wind transport is minimized by the fact that impregnated females rarely fly. However, this problem is offset in the case of anemohydrochorous transport (p. 604), where—on account of the large congregation of individuals in the new region—a meeting of the sexes after transport is facilitated. 6. Despite these important reservations it is clear that carabids capable of flight, are benefited in dynamics in contrast with the flightless carabids. It appears that passive wind transport of the latter can be discounted. The principal difference in the dispersal capacity of species capable of flight and of flightless species, however, is not so clear in Fennoscandia: the devastation of the Quaternary glaciations permitted, with few exceptions, the survival or postglacial immigration of only such completely soil-bound insect species as are markedly eurytopic. Because of this these species also have a great advantage in dynamics, which can give them a strong advantage over species capable of flight but markedly stenotopic. On the other hand, in the central European mountains, chiefly the Alps, where the possibilities of glacial hibernation were far greater, the sharp differ- ence between insects capable of flight and flightless insects regarding the size of the area—i.e. the capability of dispersal—is clearly evident, not least among the carabids. Many non-flying species have very small terrains of habitation, sometimes a single mountain peak. Examples of such carabids, also from other regions, are given by Holdhaus (1927-28, pp. 597 ff.) and by Heberdey (1933). 599 524 Water Dispersal The range of transport of insects by water is a function of the water’s movement and the ability of the animals to survive a sojourn. Experiments on exposure to water—some of them with carabids—have been conducted by numerous researchers, particularly by the following on Nordic beetle material: Mjoberg (1912, pp. 198-199), Lindroth (1931, pp. 484-485), Krogerus (1932, p. 237), Frey (1937, pp. 430-432), Palmen (1944, pp. 154 ff., 1945, p. 29), Backlund (1944, 1945, pp. 140-141). My experiments with the species of Cymindis are described in the present contribution (p. 248). These studies showed that most beetles readily tolerate exposure to water for more than one day. Survival is longest in species that are able to float on the surface (Lindroth, Frey), for example, Cercyon litoralis Gyll.: up to 22 days (Backlund). In this respect the carabids are favored by their large elytra, which are closely juxtaposed and can retain a considerable quantity of air in the subelytral space where the stigmata openings are located (Lindroth and Palmen, 1944). But Dyschirius obscurus, for instance, can even endure complete submersion for 7-8 days (Krogerus) (in winter much longer: see below). This is important, since in nature, especially with rapid water transport in rivers and waves, insects are incessantly splashed by the water when not transported on drifting objects. Some insects—including terrestrial forms—can perform active swimming movements on the surface of water. They are better suited than others to struggle toward a nearby bank, more or less independently of the movement of the water. The carabids are favored in this respect (Joy, 1910, p. 383). In Agonum marginatum, Palmén (1944, p. 78) recorded a swimming speed of up to 4 m per minute (water temperature 18-19°C). In Broscus cephalotes, I obtained a corresponding figure of a good 2 m (water temperature 15°C) and a mean value (from 9 observations) of 1.8 m. According to Joy (l.c.), Agonum ruficorne even has the ability, like Dianous and certain species of Stenus, to secrete a fluid from the abdominal tip which drives the insect forward. Since I assumed Broscus cephalotes to be a flightless species (p. 574) it was interesting to study its resistance to exposure to water (Experiment 146, p. 109). Four specimens were studied, two of which were left swimming undis- turbed on the surface of the water, whereas the other two were vigorously shaken 3 times daily. The room temperature was high (> 20°C). Length of life of 2 specimens undisturbed: 9 3/4 and 11 1/2 days, respectively. Length of life of 2 specimens shaken up in water: 9 3/4 and 17 1/6 days, respectively. Considering the high temperature during the experiment, Broscus must be considered markedly resistant to exposure to water. A big advantage in nature is its indifference to being splashed with water (cf. the species of Cymindis, 600 525 p. 248). This is certainly due primarily to the firm attachment of the elytra along the suture (p. 575), so that a quantity of air essential for respiration can be retained in the subelytral cavity. Hence Broscus is an insect with a good possibility for surviving long-distance transport by water even without the protection of drifting objects. To explain its occurrence on Gotska Sandon in this way (p. 285) may not be too bold. Three “external” factors particularly affect the ability of insects to tolerate exposure to water: First the temperature; a low temperature prolongs survival in water, especially in carabids (Palmen, 1944, p. 169; 1945). It follows that transport by water is easier in the winter half-year, i.e. the imago hibernators are favored in this respect (p. 205). Second, the salt content of the water. Palmén (1944, p. 155) conducted instructive experiments which clearly show that the high saline content of the open sea (> 30 per mille) has a harmful effect on most insects. It causes physi- ological desiccation of the animal, which more or less sharply shortens its life. Baltic Sea water (< 10 per mille) seems not to have such detrimental effect. These experimental results were clearly confirmed by zoogeographical findings. In particular, the fauna of the North Sea islands considered earlier (p. 325) shows, that colonization by hydrochorously transported flightless species has taken place chiefly in winter (in part probably with ice), and that it has been more effective at the mouth of the larger rivers. On the other hand in the Baltic Sea, Gotska Sandon, for instance, evidently obtained the corresponding element to a great extent by purely hydrochorous transport during the summer half-year. Third, the occurrence of all kinds of drifting solid objects in water has a very important role in the possibility of insects to survive long-distance transport. We found above (p. 248) that Cymindis macularis, which has colonized certain Baltic Sea islands probably by hydrochorous transport in summer, has such poor resistance to exposure to water, that long-distance transport might be conceivable only with the help of floating plant material and the like. This species lives chiefly on sandy banks in the immediate vicinity of the sea. It is therefore quite conceivable that some individuals may be set adrift, for instance, by a strong stormytide, along with a fascicle of Psamma or Elymus. In these fascicles even with constant splashing by the waves, there are sufficient air spaces, chiefly under the leaf sheaths, for a safe journey. The same kind of transport may be assumed with still greater certainty for the colonization by Dromius linearis of the Swedish east coast and the Skargard of Aland, where this species is often found on very small skerries. It lives here normally in fascicles of grass, always in the brachypterous form. The most important “object” offering hydrochorous transport is the ice, and there are several reasons for assuming this. First of all, as already mentioned, the low temperature itself increases the resistance of the insects. Second, the insects on the surface of ice-floes can remain as good as dry, 601 602 526 especially when it carries reeds, straw and the like. When such parts of plants, for instance, stems of Phragmites, are present in the solid ice, the insects can survive for months frozen in the ice (Palmen, 1945, p. 32). Moreover, with a sufficiently strong wind, drift ice can move fairly independently of the ocean currents. Ice-floes that enter the sea through river mouths with spring floods are especially advantageous. They are more likely to have been covered with soil, plant parts, etc., and many also harbor other than purely seashore inhabitants and transport them out to sea. Besides, river ice is especially suited for transporting insects sensitive to saline content, even to islands surrounded by highly saline seawater. For colonization by water transport, of all the islands considered, Hailuoto in the mouth of the Ule River is most favored (p. 236); Glomma might have similar significance for the Hvaler islands (p. 318). It may therefore be justified, in our climatic conditions, virtually to equate hydrochorous transport of insects with ice transport. Several concrete examples were given in the chapter on Insular Faunas. We must now inquire which kinds of movement of water bring about long-distance transport of animals within a reasonable period of time. The most effective are rivers. Their importance in promoting the dispersal of insects of course lies not so much in the insects being washed out to sea (or into lakes), but in that—chiefly during flooding (see, Palm, 1945)—species living on the upper course of a river are passively transported down with the water (see also Holdhaus, 1927-28, pp. 603-604). Especially useful modes of transport are provided by very large vegetation-covered clumps of earth, which often fall into the river on account of the erosion caused by water and are carried downstream (Heinze, 1914). In line with this, both along the north- ern Swedish rivers (Lindroth and Palm, 1934, p. 123) and near the Klaralven River in Varmland (Palm and Lindroth, 1936, p. 39), numerous accidental oc- currences of northern or otherwise cold-loving Coleoptera were observed. In the fjelds it is noticed by those with botanical interests that even high alpine plants of the banks of torrential streams in the Regio betulina are carried down- stream. In the same way often even Nebria nivalis is carried to lower altitudes by rivulets of melted snow from its true home on the edge of the perennial snowdrifts. Whether such river transport leads to permanent colonization and area expansion depends on how far the transported animals can survive in lower locations. For species restricted to the Regio alpina and the high altitude forests of the fields, this mode of dispersal has little significance because of their special requirements of cold. The situation is very different for species of the plains, for which a moun- tainous region, such as the Scandinavian chain of fjelds, presents a barrier to dispersal. An excellent example of an animal that was able to overcome this barrier— moving from west to east—partly with the help of rivers, is Bembidion virens, which I have discussed in this connection (Lindroth, 1935a, p. 624). It 603 527 is also possible that the emigration of Agonum piceum and Pterostichus minor from Sweden to the Trondheim region is materially assisted by rivers. An attractive task would be to study experimentally the area-expanding effect of rivers. The task could best be undertaken by “introducing” in suf- ficient numbers, for instance, along the middle course of the Klar River, a carabid from the temperate parts of North America* which is unpretentious in all respects but is associated with the waterside. In subsequent years regular and precise inventories of the riverside above and below this point should be undertaken. Selection of a large, easily recognizable species would be prefer- able. Objections to this deliberate “adulteration” of the endemic fauna would probably be raised in various quarters. However, for the reasons given above (p. 555) I do not believe that the introduction of a polyphagous carabid can disturb “the balance of nature,” and the purely scientific advantage might be enough to justify the procedure. In standing water and in the sea, transport is determined partly by water currents and partly by waves caused by the wind (in the case of drifting objects sticking out of the water, directly by the wind as well). At present most zoo- geographers are rightly skeptical of the possibility of long-distance transport of terrestrial animals in seawater (for example, Holdhaus, 1927-28, pp. 625 ff.). In the Baltic Sea the conditions are more favorable on account of the comparatively short distances and the low salt content of the water (p. 518). Botanists especially have cited the “Baltic drift” for the dispersal of plant diaspores (Sernander, 1901; Eklund, 1931). But it is obvious that living animals are much more susceptible to submersion. Hence transport over such a long distance as from the eastern Baltic or Gotland to Aland has been seriously assumed (p. 248) only for one coleopteran species, Drilus concolor, the female of which lives in the shells of snails. For other cases where hydrochorous transport to various Baltic islands was surmised, reference is made to the chapter on Insular Faunas. There is also a possibility that Carabus clathratus has succeeded in hydrochorously crossing the narrow Strait of Kvarken in the Bothnian Sea (p. 381). The question whether insects are able to cross the Gulf of Finland by purely hydrochorous transport, especially from Estonia to southwestern Fin- land, has been repeatedly discussed. Frey (1937, pp. 423 ff.) surmises that at least part of the drift material he examined from the region of NI Tvarminne arrived by this means, but this is justly contested by Palmen (1944, pp. 81-82). Krogerus (1932, p. 238) thinks such transport with drift ice is conceivable, especially in spring. At least in two cases, Carabus cancellatus and C. convexus (both consistently brachypterous), this assumption seems justified. Both have an isolated occurrence on the mainland in the extreme southwest of Finland. The gap east of it includes the region of Helsinki—the best-explored part of *If an Asiatic species were selected there would be an outside chance of the same species later reaching Fennoscandia “in the natural way.” 604 528 all Finland—so it is out of the question that it is due to insufficient explo- ration. For the same reason it seems inconceivable that the species have been introduced anthropochorously into southwestern Finland, since in that case Helsinki would hardly have been bypassed. The map of C. cancellatus shows that the above-mentioned gap cannot be due to climate. C. convexus was also found on an island in the Skargard of Aland, and this was considered (p. 261) the result of hydrochorous transport (possibly with drift ice). It is not too bold to attribute the occurrence of these two species of Carabus in southwestern Finland entirely to hydrochorous transport from Estonia across the mouth of the Gulf of Finland (concerning the direction of the sea currents in this region, see maps in Fig. 19, p. 247). There may not be any further case where hydrochorous immigration to Fennoscandia has to be assumed. The greatest effect of a factor supporting dispersal is attained by hydrochorous transport if it operates together with anemochorous transport—or more correctly after the latter, i.e. as anemohydrochorous dispersal. Its importance was correctly appreciated only with Palmén’s investigation (1944). The special significance of anemohydrochorous transport and its superior- ity to purely anemochorous transport or purely hydrochorous transport lies in the fact that a long-distance transport over large surfaces is possible without the otherwise inevitable scattering of individuals. The initial phase, i.e. flight and anemochorous transport, determines the distance and the direction of the journey, and the final phase, the washing ashore, provides for a meeting of the sexes in the new region. So it can be argued that for a majority of insects cabable of flight —especially among the Coleoptera and Hemiptera—anemohydrochorous transport represents the most effective passive means of dispersal, which can iead to permanent colonization. Palmen (1944, pp. 206 ff.) is certainly correct when he stresses the effect of a “Baltic direction of immigration” across the Gulf of Finland, for the coleopteran fauna of Finland.. Among the carabids he believes to have immigrated in this way, Demetrias monostigma, which, as far as is known, is always apterous in northern Europe, must be excluded (cf. p. 605), and the transgrading Chlaenius sulcicollis has no permanent area in Fennoscandia. On the other hand some species can be added. Among carabids that have reached Finland (at least partly) by the “Baltic route of immigration” using anemohydrochorous transport, mention may be made of the following: Acupalpus consputus B. schuppeli Agonum marginatum B. varium Amara spreta Calathus ambiguus Bembidion assimile Calosoma inquisitor. B. biguttatum 605 606 329 To a large extent the Baltic islands have also received the flying elements of their fauna in this way, which is evident from the discussion in the section on Insular Faunas (pp. 236 ff.). It is much more difficult to decide whether the fauna of the Scandi- navian mainland has also received a substantial influx by direct anemohy- drochorous transport across the Baltic Sea and the Bothnian Sea. Certainly Dyschirius neresheimeri arrived in this way. The question was touched on else- where (p. 719), but must be left in the main to future investigations. Transport by Animals No carabid in the adult stage leads a parasitic existence, and in cases where the larvae are ectoparasitic on other insects—in our fauna at the most in the case of Brachynus (p. 548) and Lebia (p. 550)—the host (as far as is known) is in an immobile stage (pupa or the last instar larva) and hence cannot actively contribute toward dispersal of the parasite. Actually as well as theoretically only vertebrates can be considered as ac- cidental transporters of carabids (including their developmental stages). Some authors have made particular mention of birds, which can cover long distances in a short time and so can have a role in the colonization especially of isolated islands by all sorts of small animals. In respect of Iceland, I have already (1931, pp- 531 ff.) considered this question in fair detail and concluded that trans- port of fresh-water animals (for example molluscs) on the feet of natatorial birds may not be altogether rare (concerning the dispersal of aquatic plants by birds, see Samuelsson, 1934, pp. 187 ff.), whereas the “atmospheric animals” have little prospect of utilizing such transport. With regard to insects—aside from bird parasites—mostly the immobile stages alone, chiefly the eggs, are involved, and the possibility of permanent colonization of a new region seems realizable at the most in parthenogenetic species. Among the Fennoscandian carabids the possibility of transport by birds perhaps rates consideration in a single case, Demetrias monostigma. Two facts may be cited: First, the quite isolated northernmost occurrences both on Oland and Gotland, as also in Vrm Visnum and in the region of Helsinki on banks rich in vegetation and in birds. Second, on account of the broad fourth segment of all tarsi in this genus, which is provided with strong adhesive hairs that help the insect climb around on grass stems, etc. and hold on, even against strong wind at exposed places. Transport by Man There are very few carabids in the Fennoscandian fauna completely and di- rectly dependent on man. Only two cases are very clear: Pristonychus terricola and Sphodrus leucophthalmus, both of which are found in our region only in 607 530 houses (although one specimen of Pristonychus was collected at the entrance of a cave in Gtl Lilla-Karlso). Ecologically related is Clivina collaris, which in our region occurs more or less accidentally in garden soil or even in glasshouses. All three undoubtedly reached Fennoscandia by human traffic. There are other species not markedly synanthropous elsewhere in their area whose arrival in Fennoscandia is evidently due to transport by man: they occur exclusively in or near cities in strikingly accidental localities. These are Carabus auratus, Pterostichus madidus, probably also Carabus monilis, and pos- sibly Lionychus quadrillum (cf. p. 622). For reasons mentioned elsewhere (p. 632) it is also highly probable that Carabus nemoralis and Dichirotrichus rufithorax, which are benefited by culture, originally arrived along with human beings. With the present (in peacetime!) brisk traffic between most European countries it is strange that apparent cases of insects accidentally introduced by man are not more common. Shipments of potatoes, fruits, and vegetables from southern Europe, of onions from Holland, etc. should provide excellent Opportunities. In other coleopteran families there are examples of species that have been transported to Europe—including Fennoscandia—from afar areas, even from the Southern Hemisphere (p. 638), although not among the Fennoscandian carabids. In general the importance of human traffic for the long-distance trans- port of other clearly synanthropous insects has in my opinion sometimes been overestimated. With regard to Iceland which, on account of the constant im- port of foodstuffs, timber for construction, etc. must be considered unusually suitable for this purpose, I have (Lindroth, 1931, pp. 506 ff.) argued the con- trary in detail. The possibility of an important role for the dispersal of non- synanthropous insects by human beings to this island is contradicted chiefly by the following fact: Iceland and the Faeroes have had identical trade connec- tions from olden times (the same goods are imported from and exported to the same countries). Yet the synanthropous fauna (with clear examples in the genus Cryptophagus and among the Collembola) is largely different in the two insular regions, whereas the non-synanthropous species are remarkably iden- tical. “Whoever is inclined to consider the coleopteran faunas of Iceland and the Faeroes as predominantly imported must argue that distribution [better: dispersal] with cultivation mainly affects species not bound to cultivation!” The faunal exchange between Europe and North America in the historical past with the help of human beings may also be recalled. There are first of all numerous species earlier known from Europe which are believed to have been introduced into America. This undoubtedly holds in the cases of Pristonychus complanatus Dej. and P. terricola, and certainly also for species like Carabus nemoralis. But this kind of interpretation is easily carried too far; for instance, I would no longer assent to the view expressed by Holdhaus and myself (1939, pp. 224, 234) that Barynotus squamosus Germ. was introduced 608 531 to North America. How small a role such traffic has played in the opposite direction is striking; Holdhaus (1927-28, pp. 612-613) is aware of only two cases (Stenopelmus rufinasus Gyll., Neoclytus erythrocephalus Ol.), in which true acclimatization of a North American coleopteran species has taken place in Europe, to which unfortunately the Colorado potato beetle (Leptinotarsa decemlineata Say) must now be added. Of course, just now there are at least as good possibilities for transport eastward as westward across the North Atlantic Ocean. But notably opposing the assumption of frequent long-distance transport is the fact that not a single non-synanthropous carabid species has found its way from the Southern Hemisphere into Europe with human traffic. It cannot be seriously denied that numerous species of the temperate parts of South America and Australia (including New Zealand) would be capable of living in the corresponding parts of Europe. Evidently the possibilities of transport are greater over short distances, for example within the limits of Fennoscandia. It is not impossible that the northernmost, rather isolated Swedish occurrences of Stomis pumicatus orig- inated in this way, all the more so since it is flightless and occurs there as a markedly “cultural species.” The same might hold for the functionally brachypterous Prerostichus vulgaris, which is “favored by culture,” with regard to the records near Lk Pelkosenniemi and Lj Triostrova. Otherwise the “most isolated locality” is indicated in Table 37 (p. 680). At any rate it is striking that so few species in Fennoscandia have a dis- junct area that might be attributed to the erratic effect of anthropochorous transport. Otherwise such a phenomenon might be expected particularly in the case of flightless species, for which generally few modes of dispersal are available. Actually the contrary is true. Markedly disjunct areas chiefly char- acterize winged species, and in most cases these result from active flight and especially wind transport. The role of human culture in assisting dispersal lies more in the new biotopes and new connections between biotopes earlier iso- lated, chiefly in the Nordic woodland region (p. 641), than in the increase of possibilities for passive transport. Other Dynamic Factors Certain characteristics of an animal species, not mentioned earlier, may be dynamically favorable or unfavorable. The soil-bound, flightless carabids show diverse ambulatory activity and the speed of running is highly variable. Unfortunately I have not made any measurements. In general the largest species—calculated absolutely—run faster; certainly no carabid can cover a longer distance per (longer) unit of time than the species of Carabus. Smaller, more or less constantly flightless species, chiefly Calathus, run as fast but have less stamina. Nevertheless, the - 609 610 532 dependence on definite biotopes (the more or less pronounced stenotopy), the daily period of activity, and exertion, etc. might decisively influence the active capability of dispersal of soil-bound species to the extent that simple measurement of the optimal (or average) running speed probably would not support any definite conclusions. We often speak of the especially strong “dispersal urge” of some species. The expression is misleading. No animal has the urge—not even instinctive—to disperse in the geographical sense. The activity of the individual animal is caused by the feeling of discomfort in the momentary milieu or momentary internal state (hunger, etc.), possibly also by attraction toward positive stim- uli, which the insects perceive chiefly by the olfactory sense. However, it is striking that the individuals of certain species show greater migratory activity than those of related species, which is usually impossible to explain. Further- more that on an average they move farther from their “birthplace” and so have greater possibilities of continuous and rapidly expanding their area. The best examples among the Fennoscandian carabids are provided by the “trans- grading” species listed below (p. 621). Even flightless species may undertake more or less long migrations. A mass migratory movement was observed by Gaunitz (1933) in Carabus violaceus (Lyl Sorsele). There were hundreds of individuals which migrated in the same direction toward a river bank, where they entered the water and were mostly drowned. Copulation was repeatedly observed, and it is possible that in this case a strong sexual urge had provided the stimulus for migration. In central Europe similar migrations of Carabus auratus have been observed (among others, by Horion, 1941, pp. 46-47). In one case (Barner, 1937, p. 24) it was emphasized that copulation between the migrating individuals did not take place. The impetus for the locomotion of an individual—for a more or less pronounced “migration”—both in flightless carabids and carabids capable of flight may be mainly due to alterations in the environment. In particular, drying up of a biotope in summer might force markedly hygrophilous inhabitants to emigrate (p. 395). Excessive population density in carabids may rarely, if at all, lead to the same phenomenon (pp. 559, 654). However, it is clear that a high abundance at the biotope can produce a greater multitude of migrating individuals and increase the prospect of a later meeting of the sexes in the new localities. Barriers against Dispersal It is evident that for every species of animals any uninhabitable biotope is a dispersal barrier. This question was touched on in the section on “Steno- topy and Eurytopy” (p. 563) and in connection with “dispersal as a result of cultivation” (p. 641). In northern Fennoscandia the forests have been the prin- cipal barriers—especially in earlier times. They are difficult to cross for all the 611 533 species incapable of living in them. But in this section we will consider only the two dispersal barriers that have played the biggest role in the colonization of Fennoscandia and in the advance of species within the limits of this region, the sea and the mountains. The dry desert, which can be decisive for their dispersal in other parts of the earth, is missing from our region. The Sea The ability of carabids to cover long distances by air and by water has already been considered (on pp. 573 and 598 respectively). The effect of these agencies is best studied by analyzing the insular faunas, of which a series of examples were offered (pp. 198 ff.). Here the area-limiting effect of the sea shall be illustrated by three especially clear cases. We will inquire how far the following seas form area limits for a greater or smaller number of species of carabid (br = brachypterous, d = dimorphic species): 1. The Gulf of Finland a. Species that occur in northern Estonia but have not reached the south coast of Finland. Badister sodalis br Carabus coriaceus br Bembidion argenteolum Pterostichus anthracinus d B. litorale Stomis pumicatus br. b. Species that occur on the south coast of Finland but are not known in northern Estonia. Agonum micans B. ruficolle A. munsteri Harpalus luteicornis Amara interstitialis Patrobus assimilis br A. littorea Pterostichus gracilis A. montivaga Tachys bisulcatus. Bembidion grapei* The discovery of one or other of these species on the north coast of Estonia in a future more thorough exploration of that country might re- strict this latter group. At any rate a basic difference is evident from the first group— which is missing from southern Finland. This group comprises 3 brachypterous and 1 dimorphic (chiefly brachypterous) species, and in ad- dition 2 species of Bembidion, which have good flight capacity and therefore are little affected by anemochorous transport (p. 590). On the other hand, the group of species that have not been found on the southern shore of Gulf of Finland includes, with one exception, only species capable of flight, which *Bembidion grapei is constantly macropterous in the southern part of its area (Fig. 50, p. 402). 612 534 can easily be transported with the wind—also anemohydrochorously. The un- expected fact, that more species have remained on the northern shore than on the southern shore of the Gulf of Finland, which will probably be confirmed by a thorough exploration of Estonia, may be due to favorable wind condi- tions (p. 589) for anemohydrochorous transport toward southern Finland. It is uncertain whether in addition the “positive heliotaxis” of the insects (p. 592) has played a role. 2. The Sea Separating Sweden and Denmark (Skagerrak, Kattegat, Oresund) The barrier formed here by the sea is frequently only partial, which is evident in many cases between Skane and Sjaelland, in others between the Swedish west coast and Jutland. Such species are indicated with an “S” (= obstructed only in the south) and an “N” (= obstructed only in the north). a. Species that occur in Denmark but have not (or have only partially) reached the opposite paris of Sweden: Anisodactylus nemorivagus N Demetrias monostigma br N Bradycellus verbasci N Dromius melanocephalus N Calathus moliis* Harpalus punctatulus N Calathus piceus Patrobus septentrionis australis. b. Species that occur in southern and/or western Sweden but which have not (or have only partially) reached the opposite parts of Denmark: Agonum dolens S Dromius fenestratus Amara littorea N D. marginellus Bembidion dentellum N D. quadrinotatus N B. quadrimaculatum Harpalus distinguendus B. semipunctatum H. luteicornis B.velox _- N Lebia crux-minor S Carabus problematicus br L. cyanocephala S Cymindis angularis br N Licinus depressus br. Dromius angustus It is striking that dispersal of only four species could be impeded by lack of flight capacity. 3. The Channel between the British Isles and Continental Europe I am unable to give a complete list. The following list of species that are missing from the British fauna (at least as constant members) but have reached the opposite side of the Channel has been compiled from Sainte-Claire Deville (1930a, b) and Borchert (1938): *In Jutland the otherwise dimorphic Calathus mollis is constantly macropterous (Fig. 28, - p. 368). 613 535 Abax ovalis Dft. br Diachromus germanus L. A. parallelus Dft. br Dyschirius chalceus Agonum lugens Harpalus calceatus A. viridicupreum Gze H. distinguendus Amara montivaga H. modestus Dej. Bembidion elongatum Dej Leistus piceus Fröl. br Brachynus explodens Dft Licinus cassideus Fbr. br Calosoma auropunctatum L. hoffmannseggi Panz. br Carabus auronitens Fbr. br Molops piceus Panz. br C. cancellatus br Pterostichus interstinctus Sturm C. coriaceus br P. punctulatus Cychrus attenuatus Fbr. br Trichotichnus laevicollis Dft. d*. The number of brachypterous species is strikingly high, but in central Europe they generally constitute a larger component of the fauna than in Fennoscandia. A comparison of three sea straits reveals the following: 1. Gulf of Finland. In the outermost part > 45 km broad. Dispersal limit of 17 species (maximum), 12 of which are winged. 2. The sea between Sweden and Denmark. Width: Öresund > 4 km; Jutland-Sweden > 63 km. Complete or partial dispersal of 25 species, 21 of which are winged. 3. The Channel. Width > 31 km. Dispersal limit of 24 species (minimum), 13 of which are winged. Evidently these species numbers are not directly comparable, since the length of coast, the richness of fauna, the question whether in some cases there may be existence limits rather than dispersal limits, etc. should also be considered. But we may be justified in stating that the Gulf of Finland has been a more easily surmountable barrier for carabids than the straits between Sweden and Denmark and between England and the Continent. And nevertheless, in both these cases firm land connections have existed during the postglacial period! The reason, as far as can be judged, is the much lower salt content of the water in the Gulf of Finland, such that anemohydrochorous—as well as purely hydrochorous—transport was possible to a far greater extent. This favorable circumstance applies to the Baltic Sea as a whole and may be the chief reason for its unexpectedly slight importance as a dispersal barrier. The Mountains Mountains need not be very high to pose a barrier to dispersal. Water- sheds, even if they are geomorphologically poorly marked, represent bound- aries which are difficult to cross for aquatic animals and riparian species without flight capacity. This is vividly illustrated in the case of fish by Ekman (1922, *Dimorphic, according to Horion in litt. 614 615 536 pp. 463 ff.) and by A.M. Hansen (1929, p. 98). An example from the Fennoscan- dian carabid fauna is provided by the constantly brachypterous northeastern stock of Bembidion transparens, which has evidently not succeeded in crossing the divide between the Arctic Ocean (including the White Sea) and the Gulf of Bothnia (p. 392, Fig.-45). For other insects less closely associated with water, as well as for species that can actively or passively disperse through the air, the Regio alpina of the fjelds represents-the most difficult barrier. It forms a nearly unbroken wall along the Scandinavian fjeld-range, which in the stretch extending from the extreme northern tip of Sweden into the province of Harjedalen coincides fairly exactly with the boundary of the kingdom with Norway. Farther south the main watershed bends toward the west and is responsible for the faunistic isolation of the Norwegian “western country.” The barrier formed by the Scandinavian fjeld-range is especially evident from the advantageous effect of the passes, shown in the map in Fig. 61 (p. 437). Let us first consider the passes lying in the Regio coniferina. Such passes are found only in two regions: 5 passes in Jtl and 3 passes in south-central Norway, connecting the Trondheim region and Romsdal with southeastern Norway. Clear examples are (br = brachypterous, d = dimorphic species): 1. Species that have crossed the watershed northward through the central Norwegian passes: Amara tibialis Trechus discus Badister bipustulatus T. micros. To what extent migration through these passes has been in the opposite direction (southward) is discussed elsewhere (p. 770). 2. Species that have crossed the watershed through the passes in the central Jamtland (Storlien region), a. Westward: Agonum dolens B. obliquum A. micans B. quadrimaculatum A. piceum Pterostichus minor d Bembidion doris b. Eastward: Bembidion nitidulum Leistus ferrugineus br B. virens Patrobus atrorufus br. The only passes farther north are in the Regio betulina. These were also utilized by some species incapable of living in the Regio alphina, at least not permanently. Examples are: 3. Species that have used the passes in southern Lapland (Äsl, Lyl); a. Westward: Agonum quadripunctatum Agonum sexpunctatum Amara ovata ? Dyschirius thoracicus. Bembidion lampros d b. Eastward: Bembidion nitidulum Patrobus atrorufus br B. virens Trechus obtusus br* Leistus ferrugineus br 4. Species that have utilized the passes in northern Lapland (Tol) a. Westward: Tachyta nana b. Eastward: Bembidion virens Leistus ferrugineus br. Of course, these examples are only the clearest cases. Without doubt many, more widely distributed species which avoid the Regio alpina have crossed the Scandinavian fjeld-range in one or both directions through these passes. The above “pass ambulators” together comprise 22 species, 17 of which are © constantly macropterous, 3 constantly brachypterous, and 2 dimorphic. Of the two last, Pterostichus minor has crossed the fjeld-range only as macropterous form (Fig. 43, p. 388), whereas Bembidion lampros has done so in both forms (Fig. 40, p. 384). So we have (as minimum value) 18 functionally macropterous and only 4 functionally brachypterous species that have used the passes. An important conclusion can be drawn from this: For dispersal, even the carabids capable of flight, are dependent on the existence of continuous suitable biotopes. Although they are often driven out into such biotopes situated in new regions—chiefly by the wind—this apparently rarely results in permanent new colonies. This is indirect proof of the idea propounded above (p. 589) that long-distance transport of carabids in the higher layers of the air has very little dispersal significance. The discussion on the faunistic effect of the Scandinavian fjeld passes considers only present-day conditions. It is well known, however, that during postglacial time, in the so-called “xerothermic period,” the more favorable climate appreciably raised the timber lines in the fjelds (p. 687) and in this way allowed greater exchange of faunas between the two sides of the Scandinavian fjeld-range. A clear example of a faunal element that apparently utilized the then wooded passes is provided by the partly flightless species of eastern origin considered above (p. 405), which have an isolated relict-like area at the inner part of the Sognefjord in western Norway. The advancement of forests into the fjelds affected the true alpine fauna adversely (p. 755). *Trechus obtusus was found in Hjd just above the timber line, but is certainly not a true inhabitant of the Regio alpina. 617 538 Final Remarks on Area Limits Of prime importance for the correct representation of the area of an animal or plant species is the question whether or not its limits are throughout deter- mined by existence factors. Is there a generalized solution to this? Is it possible to conclude from the mode of occurrence of an animal or plant species at the border of its area whether an existence limit or a dynamic limit is involved? This question is of decisive importance for the discussions in subsequent sections. For if we believe at all we can reconstruct the course of history from the present picture of an area there are two essential preconditions: first, that not all area limits are produced by existence factors alone; second, that it is possible to decide whether in a given case an existence limit or a dynamic limit is involved. The characteristics of a dynamic limit are: 1. The species is demonstrably in the process of dispersal, and the ascer- tainable alterations in area are larger than what can be explained by simulta- neous environmental changes (primarily climatic). The species that are most clearly in process of expanding their area are considered below (p. 621). However, in general it cannot be expected that a species in process of expanding an area should make such great progress that this can be established on the basis of collected material spanning barely a century. Other criteria must be applied. 2. In the case of an existence limit, especially if it is climatically determined, it is to be expected that the biotopes suitable for the species concerned will be scattered. This causes reduced frequency. On the other hand the abundance (the density of individuals) need not necessarily decline. In the case of a dynamic limit, concentration in a few, widely separated surfaces is not to be expected. On the other hand an accelerating spread of individuals toward the periphery must ensue. Hence the consequence is decreased abundance—with the same or slightly reduced frequency. 3. If neither of characteristics 1 and 2 is ascertainable, the dynamic charac- ter of an area limit has to be tested simply by the exclusion method. If the limit is not found to correspond with any climatic line, if no edaphic, food habit fac- tors, etc. can be cited, we can presume at least a dynamic limit. Examples are Amara torrida and other “north-eastern” species which are restricted to north- ern Scandinavia, although they there occupy an area extending from coast to coast with a more or less sharp southern limit; and Bembidion nitidulum and other “western” species, which have crossed over to the Swedish side through the fjeld passes and demonstrated that they are not dependent on the oceanic climate of western Scandinavia. The difference between existence and dynamic limits is especially clear among the dimorphic species. If the macropterous form has advanced farther than the brachypterous form, it can be contended that the area limit of the 618 620 539 latter is a dynamic (provisional) limit (examples: Figs. 28, 35, pp. 368, 376). If both forms have reached the periphery of the area we are justified in speak- ing of a (more or less stable) existence limit (examples: Figs. 40, 41, pp. 383 ff.). To the extent that the dynamic limit can be correctly singled out and judged, we get the most important key to antiquity—to the history of the area. The clearest existence limits in Fennoscandia are the (mostly climatically determined) northern limits of the southern species. To give a general idea of their borderline, I compiled the map in Fig. 82. The species shown are markedly southern, with more or less continuous areas and, as far as can be judged, existence-ecological (climatically) determined northern limits. According to their northern limit on the east coast of Sweden they are divided into the following 6 groups: a) Northern limit in Nbt (south of the Arctic Circle). 14 species (Agonum versutum, Amara bifrons, A. ingenua, A. ovata, A. plebeja, A. tibialis, Calathus erratus, Cicindela campestris, Dromius sigma, Dyschirius politus, Elaphrus uligi- nosus, Lebia crux-minor, Pterostichus minor, Trichocellus placidus). b) Northern limit in Vbt. 9 species (Agonum assimile, Amara consularis, A. famelica, A. similata, Carabus hortensis, Dromius marginellus, Harpalus pubescens, Pterostichus vulgaris, Trechus secalis). c) Northern limit in Ang. 6 species (Bembidion gilvipes, B. unicolor, Carabus granulatus, Chlaenius nigricornis, Dromius fenestratus, Trechus quadristriatus). d) Northern limit in Mdp. 4 species (Bembidion dentellum, B. ustulatum, Carabus nemoralis, Harpalus tardus). e) Northern limit in Hls. 16 species (Acupalpus dorsalis, A. flavicollis, Agonum livens, A. moestum, A. obscurum, Amara aenea, A. curta, A. equestris, Anisodactylus binotatus, Badister bipustulatus s.l., Carabus arvensis, Dromius nigriventris, D. quadrinotatus, Harpalus distinguendus, Lebia chlorocephala, Notiophilus pusillus). f) Northern limit in Gst. 9 species (Asaphidion flavipes, Badister pelta- tus, Bembidion articulatum, Bradycellus similis, Broscus cephalotes, Calathus fuscipes, Harpalus seladon, Oodes helopioides, Pterostichus cupreus). The mean northern limit of all the species in each group was calcu- lated for every second meridian by interpolation. Isolated, probably more or less accidental records (for example of Acupalpus dorsalis, Anisodactylus, Asaphidion, Chlaenius, Harpalus seladon, Oodes, Pterostichus vulgaris) were ig- nored. The “mean” limits drawn are intended to illustrate only the general char- acter of an existence limit and do not justify any other conclusions. How- ever, on an average the northerly position of the limits in Finland is striking (cf. pp. 459 ff.). Only in Sweden is the occurrence (distribution) of the species in each group somewhat uniform. In the two neighboring countries some ap- 619 Fig. 82. “Mean” northern limit of 6 groups of species having southern distribution: a—14 species with Swedish northern limit in Nbt (south of the Arctic Circle); b—9 species with Swedish northern limit in Vbt; c—6 species with Swedish northern limit in Ang; d—4 species with Swedish northern limit in Mdp; e—16 species with Swedish northern limit in His; f—9 species with Swedish northern limit in Gst. 541 pear to be more “Atlantic,” others more “continental” as they extend farthest north in Norway or in Finland. Hence the limits drawn in the map (Fig. 82) primarily reflect the distribution in Sweden. 621 622 Faunal History Faunal Changes in Recent Times There is no doubt that the Fennoscandian fauna is still in the process of change. In reading Gyllenhal’s “Insecta Svecica” (1810-1827) one is struck not only by species that were evidently more abundant and more widely dis- tributed in those days but more especially by the absence of a whole series of species that would certainly not have escaped the attention of this sharp- eyed researcher if they had been as frequent and abundant then as now. In the geologically brief period of one century, several species have in fact immi- grated to Fennoscandia and many have considerably expanded their area. It is of course not very easy to provide evidence for this, since in our region really intensive coleopterological work has been carried out only in recent decades. But in some cases at least the approximate course of the area displacement seems clear, and the total picture gives the distinct impression that, in spite of (partly because of) the increasingly rapid destruction of the natural landscape, the Fennoscandian fauna has undergone more additions than losses, i.e. the fauna has become richer in species. This is the characteristic of a Quarternary glaciated region. This “mobile” element of the Fennoscandian fauna is best divided into the following categories: I. INCREASING SPECIES 1. Transgrading species. These are more or less accidental migrants that do not reproduce within the region, at least not regularly. They are the counterparts of Acherontia atropos, Pyrameis cardui and P. atalanta, Phytometra gamma and other Lepidoptera, although the distances covered by flight are mostly much shorter. Strictly speaking there is no actual area enlargement (on the contrary, in Calosoma sycophanta a postglacial area diminution has taken place, see p. 674), but the occasional advances often give an impression that makes them difficult to distinguish from the actual immigrants. The doubtful species are: Agonum gracilipes Bembidion octomaculatum Amara crenata B. striatum 623 543 ? Calosoma auropunctatum Chlaenius costulatus C. denticolle C. sulcicollis C. investigator ? Dromius quadraticollis C. sycophanta Dyschirius neresheimeri (Carabus auratus ) (Pterostichus madidus). ? (C. monilis) The species that have arrived anthropochorously are given in parentheses. Partially transgrading species are: Harpalus calceatus H. griseus. Both appear to have permanently colonized only southern Skane (see also Landin, 1948), but occasionally undertake long migratory flight. Definite “swarming years” have not been proved. Other species which within the subareas of the region, chiefly on the islands, occur as accidental migrants, are ignored here. An idea of the extent of this phenomenon has been provided by Palmen (1944). 2. Late immigrants. As far as can be judged, these species have immigrated into Fennoscandia during the last hundred years, and have become resident members of its fauna: Amara fusca Harpalus puncticeps (Fig. 85) A. majuscula (Figs. 83, 84) ? H. rupicola Bembidion lunulatum Lionychus quadrillum Bradycellus verbasci Stenolophus mixtus. Clivina collaris Among these immigrants, Amara majuscula represents an especially clear case. The first records within the region (and the whole of Europe) were published in the year 1942 (Stockmann, 1942; Har. Lindberg, 1942). Actually, the species had already been collected in Finland in 1928, in Sweden even in 1917 (Ska Ahus). Harald Lindberg (1942, 1943) regarded Amara majuscula, as also A. crenata and the curculionid Gronops inaequalis Boh., found at the same time in Al Kokar, as old relicts within the Baltic region (as “pseudorelicts” on Kokar). This assumption is certainly erroneous. These are late immigrants, and the record of a single specimen of Amara crenata is evidently the result of an accidental transgression. In the case of Amara majuscula it is possible to trace the immigration almost from year to year (Figs. 83, 84). It occurs as an invasion from the southeast on a broad front, and it is much to be regretted that on account of the political conditions no information on a simultaneous or even earlier occurrence of the species in the Baltic states was available. But Makölski informed me in personal correspondence that near Warsaw on August 3, 1927, A. majuscula abruptly undertook an enormous nocturnal approach flight. Earlier the species was unknown in Poland. In company with it was the likewise alien species Harpalus zabroides Dej., which was not found 625 627 544 again in this region. On the other hand, Amara majuscula stayed on for many years and undertook more mass flights, for instance, in 1947. Concerning a new record of Gronops inaequalis in drift material off the island of Färön, see Palm, 1947 (p. 177). None of the other immigrants shows the clear, Geomlan immigration of Amara majuscula. It is of course clear in the case of Harpalus puncticeps (Fig. 85). In April 1888 several specimens of this species were collected near HIl Saro. It was never found again there even though this narrowly delimi- ted locality was one of the favorite haunts of earlier Goteborg entomologists (Sandin, I.B. Ericson, and others). Apparently the above occurrence represents the result of a single immigration, which probably took place the preceding summer, and did not lead to permanent colonization. The species was found again in Sweden only in 1920, significantly in the extreme southern corner of Skane. Since then, especially during the thirties, it has spread north and east, since 1944 even to Oland. With respect to the other immigrants the following facts may be men- tioned: Amara fusca. The first definite Swedish record on the island of Ven in 1934; eight years earlier it had been discovered on Bornholm (West, 1930, p. 447), but the very first Danish record at all ($rholm) was made as late as 1919 (West, in litt.). Up to 1946, the species was known in Skane from 10 localities. It was found at several of these repeatedly and in fairly large num- bers. It seems out of the question that it should have escaped the attention of earlier entomologists in these well-explored regions. Later, an increase in the Amara fusca population was recorded even in northern Germany (Nurnberg in Horion, 1941, p. 261). Bembidion lunulatum. The first Swedish record (Skane) was made in 1911, and on Oland in 1921. It is conceivable that this comparatively inconspicuous species, leading a solitary and fairly concealed life, was earlier overlooked, but it is striking that during a period of less than 30 years (1911-1938) 6 locali- ties were discovered in Skane. A new immigration seems highly probable. In Denmark this species was discovered in 1909 near Copenhagen (West, in litt.). Bradycellus verbasci. Only two records, in southern Skane, in 1936 and 1939. In Denmark, too, the species seems to have greatly increased in the last 30 years. It was unknown to Schigdte (1841, 1870): the first record (Roden Skov) was made in 1886 (West, in litt.). Stenolophus mixtus. This species was of course repeatedly found in the nineties near Ska Skabersjö, but since then it has considerably spread and has reached northern Skane. 1 specimen was found in 1947 in Oland. The first Danish record (Fyn) goes back to 1871 (West, in litt.). The late immigration is still more evident east of the Baltic Sea, since the species was discovered in Finland only in 1927 (Hellen, 1929, p. 95), and in Estonia in 1933 (Haberman, us fy Sean Sl iS es oe. ee on re R: x ees nl Fig. 83. Amara majuscula. a—Until 1930; b—Until 1940; c—Until 1947. 624 546 er \ Fig. 84. Amara majuscula. The first collecting date for every locality is shown. Black circles—Until 1930; Spotted circles—Until 1935; Crossed circles— Until 1940; Blank circles— Until 1947. 1935, p. 176). In the Leningrad region, however the species was already known since olden days. The two remaining species (without question mark) may be considered with fairly high probability of anthropochorous arrival. In the case of Clivina collaris, which was found to be exclusively synanthropous in our region, even in greenhouses in the environs of Göteborg and Stockholm, the situation seems to be clear.—Lionychus quadrillum, which was first discovered in 1945 in its only north European locality to date near Nke Örebro, where it occurs in astonishing numbers, in my opinion may have been similarly displaced (but see also p. 589). After discussing all the possibilities in this connection, Heinze (1947) leaves it open whether Lionychus should be considered as a relict or as a late immigrant. But the Orebro region, chiefly thanks to the ardent efforts of Anton Jansson, has been so thoroughly explored that the carabid could scarcely have been overlooked earlier. It is just the occurrence in large numbers in all three years 1945-47, which were meteorologically very unusual, that decisively contradicts the idea that Lionychus must have led a concealed life since olden 626 Fig. 85. Harpalus puncticeps. a— Before 1933; b—Before 1937; c—Before 1939; d—Until 1947. 628 Fig. 86. Acupalpus exiguus. The region colonized after 1930 is delimited by broken line. Black circles—Before 1920; Spotted circles—1920—1930; Blank circles—After 1930. times as a relict to appear suddenly in such large numbers. There remains Harpalus rupicola, which is mentioned here with reserva- tions. This species, known within the region only on Oland and Gotland, is now one of the truly characteristic animals of the loam-mixed weathered lime- stone soil on the latter island, where it has wide distribution and is almost frequent at places. Strange, then, that this species was not found by any of the assiduous entomological collectors of the last century (especially Boheman): the first record was in 1905. On Oland it was only found in 1934 (until now only 4 specimens). The distribution of the individuals of A. rupicola collected in different five-year periods, as far as is known in Sweden, seems to be typ- ical of a newly immigrated species as well, which is still in the process of continuous increase: Before 1923 1923-27 1928-32 1933-37 1938-42 1943-47 1 13 20 36 50 99 specimens In Denmark, H. rupicola is at least partially older, since it was found in M@¢en already in the middle of the last century (West, 1930, p. 447). 549 625 Fig. 87. Agonum micans. Black circles—Before 1900; Spotted circles—1900-1914; Crossed circles— 1915-1929; Blank circles—Later finds. The age of records in the Trondheim region is uncertain, but the species was first mentioned there by Lysholm (1937). 628 630 550 Finally, Notiophilus rufipes and Pogonus luridipennis are probably late im- migrants. Concerning Amara montivaga see below. 3. Other species in the process of area expansion. As members of the Fennoscandian fauna these are of course older than the “historical” time for which their arrival could be proved, but within the limits of our region they have expanded their area within a conceivable period, especially north. Their limits are pronouncedly dynamic. Shifting boundaries within the Fennoscandian region are so common that it is impossible to list all the species for which such displacement might be assumed with fair probability. One’s verdict depends on the importance one attaches to the absence of a particular species in older collections from one or other part of its present area. Mostly a pronounced tendency to dispersal is ascertainable only in species that have indubitably demonstrated it in recent decades. | The following list is to be taken as a selection of more or less typical cases: Acupalpus exiguus (Fig. 86). Especially in Sweden. Agonum micans (Fig. 87). Throughout the region (see below). *Amara ingenua (Fig. 88; see also pp. 525, 538). At least in northern Sweden. *A. montivaga (Fig. 89). At any rate in Sweden (see below). *A. municipalis. At least in Sweden. In Norrland only one record from His during 19th century. va \ {N N ur I Fig. 88. Amara ingenua. Swedish localities north of latitude 60° N. a— Until 1910; b—Until 1925; c—Until 1935; d—Until 1947. Blank circles indicate old provincial records. 632 Spl *4. similata. All localities in Sweden north of latitude 61° N, in Finland north of 63° N, were recorded in this century. *Bembidion obtusum. In southeastern Sweden. B. transparens. In Sweden. See map, Fig. 45 (and p. 389). *Carabus nemoralis. Especially in Finland (see below). Dichirotrichus rufithorax. In Sweden (see below). Dromius linearis (see below). D. melanocephalus. All boundary records on Skane were discovered in the thirties. Increased frequency in Denmark, in the last 100 years was also evident (West, in litt.). Harpalus punctatulus (Fig. 90). In Finland (see below). Odacantha melanura (see below). (*)Trechus discus. All Finnish and Swedish records north of 59° were made in this century. (*)T. quadristriatus. All localities in Sweden north of 61°, in Finland north of 62°, were recorded in this century. Species markedly favored by culture are marked with an asterisk (*). Some of these 16 species deserve further consideration. Agonum micans. There is no doubt that this species has sharply increased in Fennoscandia during recent decades and has expanded its area. However, no one-sided new immigration seems to have taken place. On the contrary, there may exist isolated older centers, such as in Denmark and Skane (fossil record from the “warm period”t; p. 666), in the Malar Lake region, in southeastern Norway and in southeastern Finland, whence the species has spread widely, especially in the thirties. But it is impossible at present to decide on the origin of the very isolated stock around the northern end of the Gulf of Bothnia. Amara montivaga. The distribution of this species in Scandinavia seems to be unique. The area seems to be completely isolated, since the species is missing from southern Sweden and from Denmark and occurs very sporadi- cally in northwestern Germany (Horion, 1941, p. 254). Eastward there is no connection with the Finnish area. The species undoubtedly immigrated late into Sweden; the oldest record (Nke Orebro) goes back to 1904, and the oc- currence in Dir is most recent (known since 1935). As surmised by Munster (1927, p. 288), it is possible that the species in Norway is also a late immigrant introduced with commerce; it was of course collected there (near Halden) be- fore 1870, and it is not inconceivable that it was overlooked earlier by the few older Norwegian coleopterists. In Finland several records, not only on the southwest coast but also in the inland (north to 62°), date from the last century, and it is not possible to establish any pronounced dispersal in recent decades. Nevertheless the Finnish subarea is also completely isolated, since there are no records at all of this very conspicuous species from the Baltic t(cf. p. 687; suppl. scient. edit.) D Ren N 631 Fig. 89. Amara montivaga. Black circles = Before 1890; Spotted circles = 1890-1909; Crossed circles = 1910-1930; Hatched circles = 1931-1936; Blank circles = 1937-1947. Fig. 90. Harpalus punctatulus. Black circles = 1919; Spotted circles = 1928-1934; Crossed circles = 1935-1939; Hatched circles = 1940-1944; Blank circles = 1945-1947. states or from the adjoining parts of Russia (not even from eastern Karelia). Carabus nemoralis. This is now by far the commonest species of Carabus in the southern parts of Fennoscandia. But it was not always. This species was not known to Linné, although his contemporary De Geer (1774, p. 89) described it under the name “violaceus,” probably from Swedish material. Dis- tribution data are provided neither by him nor by Paykull (1790, p. 17; 1798, p- 102; “hortensis”), Gyllenhal (1810, p. 59) writes only “minus frequens”. The following interesting information was first provided by Thomson (1857, p. 7; in summarized form 1859, p. 9): “According to a written communication from 633 Prof. Boheman, this species shows a peculiar distribution in that it is common on Skane and near Stockholm, but is not found in the provinces lying in between” (translated from the Swedish). Since Boheman had collected assidu- ously in the parts of Sweden he had in mind (Ogl, Sma, Ble, HIl, etc.) we may accept his judgment. On the other hand the expression “not rare” and “Sk.- Lpl.” in the later contribution by Thomson (1885, p. 2) is not only difficult to 634 553 understand but also rather inconsequential, since it is a notoriously superficial synopsis. Grill (1896, p. 2) writes “Sk.-Gestr.”, and even today the northern limit of the continuous area lies in southern His; farther north there is only the record of a single specimen near Sundsvall around the turn of the century. In Finland the increase of C. nemoralis later is still more pronounced. Of course it was already known from there more than 100 years ago (C.R. Sahlberg, 1834, p. 213), but it was known to J. Sahlberg (1873, p. 60) from only three Finnish localities (Vasa, reported by Wasastjerna, must be excluded; see Part I, p. 13). All of the five northernmost widely separated localities were dis- covered after 1932. In eastern Karelia it was only discovered in 1944 (Palmen, 1946, p. 19). On the other hand the species seems to have remained rather sta- tionary in Norway. In the northernmost known localities (Smola; Trondheim) it was already known in the 1870s. Of course its abundance in this region has perceptibly increased in recent decades (Born, 1926, p. 71; Lysholm, 1937, p. 144). Finally, attention may be drawn to the fact that C. nemoralis was not found in fossil form either in Denmark or in Fennoscandia, as it was in eight other species of the genus. Dichirotrichus rufithorax. The occurrence of this species in Scandinavia exclusively in the central Swedish lake region is striking, as also its exclusively synanthropous occurrence in our region, since it lives only on cultivated land in or around fairly large human settlements. The oldest Swedish record is from Stockholm (before 1868), the most recent records from the western- most regions (Nke Orebro, 1937; Vrm Kristinchamn, 1945; Varpnas, 1947; Ögl Linköping, 1947). Late immigration to Sweden must be assumed from the direction having the only connection, i.e. from Finland. In this country the species does not occur quite so synanthropously and has had a wider distri- bution since earlier times, especially in the southwest (C.R. Sahlberg, 1834, p- 260). Dromius linearis. It is uncertain whether this species actually expanded the limits of its Swedish area during the last century, since even in the northern localities such as Vgl Kinnekulle and the Stockholm region, it was discovered as early as the middle of the preceding century. In any case, the species has greatly increased in frequency and abundance. In older collections it is found in large numbers only from Öland, and in Skäne it was a great rarity for a long time (Thomson, 1859, p. 227). It is possible that the striking increase has resulted from an altered mode of life (or from immigration of an ecologically different stock). At present, D. linearis, namely, shows its maximum abun- dance in Psamma and Elymus hummocks of sandy shores, but this occurrence is mentioned in the literature (also in foreign literature) rather late (as far as I have been able to trace it, first by Jansson, 1913, p. 383; see also Larson, 1939, p. 431; Horion, 1941, p. 336; there is no indication in this connection in Burmeister, 1939, p. 193). Greater abundance also seems to be indicated in Denmark (cf. West, 1940, p. 48, with Schigdte, 1841, p. 96) and in north- 635 554 ern Germany (Horion, l.c.). In Finland D. linearis was found late (1921) and only in the southwestern Skargärd. The doubtful occurrence in Estonia is immigration from Sweden almost certain (p. 239). Harpalus punctatulus. The occurrence of the species in Sweden and Finland is basically different. In our region it has remained restricted to Öland and Gotland constantly and uninterruptedly for more than 100 years. In Finland it is a very late immigrant, having apparently immigrated partly from Estonia and partly across the Isthmus of Karelia (Fig. 90). A certain increased abundance has been noticed on Oland and Gotland over the last two decades, but an area expansion in Sweden has not taken place so far. Odacantha melanura. In this case, too, greater abundance is more evident than any actual area expansion. In some cases there undoubtedly has been new colonization of localities. Thus Odacantha was never found earlier along the Ringsjon in Skane, one of the best explored regions of this province since olden times, but was discovered there in 1939 and later recorded again. In the environs of Halsingborg it was only recorded in 1915. Also in Blekinge and Vastergotland, as well as on Gotland and Aland, all the records date from the last 20 years. However, in addition to Skane, there are older centers on Oland and in the Malar Lake region. In Finland the species occurred at least accidentally more than 100 years ago near Helsinki (C.R. Sahlberg, 1834, p. 268). Odacantha thus seems to represent a case analogous to Agonum micans (see above). Finally, attention may be drawn to two pairs of species for which I have earlier, in another context, surmised a comparatively late area expansion. Amara erratica and A. torrida (Holdhaus and Lindroth, 1939, p. 261; Lin- droth, 1939, p. 246). They possess a nearly identical, markedly northeastern distribution, with a southern limit in Scandinavia running from coast to coast between latitudes 65° and 66° N. This represents not an existence limit but a dynamic (historical) limit. Both are probably to be considered as postglacial immigrants (but see p. 733). It is uncertain whether any evidential importance can be attached to the fact that the southernmost inland records of both species in Sweden were made late (after 1924), since just these parts of Lapland were insufficiently explored earlier. Demetrias imperialis and Oodes gracilis (Lindroth, 1943a, pp. 139-140). In the Stockholm region both species were discovered early, Demetrias already before 1810, Oodes in 1863. In both cases it was a long time before they were rediscovered in Sweden, and indeed in the same region, Demetrias in 1903, Oodes in 1922. During subsequent decades, chiefly in the forties, they were more and more frequently collected, mostly together, and farther away from Stockholm. This noticeable change cannot be explained solely by the less’ intensive earlier collecting activity in the Stockholm region; a sharp increase in the two species is undoubtedly a fact. The unexpected occurrence of the species of Demetrias in 1946, both on the island of Faron (Palm, 1947, p. 171; 636 533 also rediscovered in 1947), and along the south coast of Lolland in Denmark (West, 1947, p. 17) is of course the result of immigration from other directions. But it proves that the external conditions, such as the climate, have become favorable for this insect in recent years. For these two species, see also p. 691. II. DECREASING SPECIES In the past century, on account of natural causes, just those species have become scarcer (or even extinct) that are dependent for their existence on undisturbed “virgin forests” (for example, Saalas, 1939). Typical cases of this kind are not met with among the carabids. It is generally not easy to think of particular species that have unambiguously decreased in numbers or have even become extinct in Fennoscandia in “historical” time (entomologically speaking). The following species may be mentioned with fairly high probability: a. Species avoiding “culture” (with northern distribution): Agonum bogemanni Harpalus nigritarsis. Amara nigricornis b. Species indifferent to culture (with southern distribution): Dolichus halensis Pterostichus punctulatus. c. Species favored by culture (with wider distribution): Agonum quadripunctatum Pristonychus terricola Bembidion quinquestriatum Sphodrus leucophthalmus. An example of Lepidoptera which, possibly for climatic reasons, has ap- parently become extinct in Sweden is Pararge achine Scop. (Kiellander, 1943). Causes of Recent Area Alterations To explain area displacements in recent times the following five reasons may be adduced (cf. also Benick, 1947). 1. These alterations are simply exponents of the normal postglacial immigra- tion of animals into a Quaternary glaciated region. There is no doubt that such immigration has been continuously taking place in Fennoscandia, i.e. that in particular the northern limits of a considerable number of species are deter- mined by Dynamics. The Fennoscandian fauna is actually characterized by a pronounced lability (Ekman, 1922, pp. 579-580). Even with regard to the fauna of Holstein, Warnecke (1929, pp. 51 ff.) thinks, the late, rapid immigration of several lepidopteran species is only a consequence of “normal” postglacial area expansion. But are these “normal” displacements actually perceptible from observations spanning only decades, at the most one and one-half cen- turies? 637 638 556 It can be argued that for an insect that has immigrated from the south and is climatically less susceptible, on an average nearly 15,000 years were at its disposal to reach its actual northern limit in Fennoscandia (Table 36, p- 661) from a glacial refuge in Central Europe. Assuming that this limit is located somewhere in central Scandinavia (near latitude about 63° N), and that the species in question therefore tolerates a high-boreal climate, it can be surmised that it survived the last glaciation not far from the southern edge of Würm ice somewhere in present-day southwestern Germany. The distance from the assumed present northern limit, including the detour around the Danish islands which were once firmly joined with the mainland, might be about 1500 km. A rough calculation gives an average linear area expansion of 100 m per year. The last phase, the stretch from the Stockholm region up to the present northern limit, about 400 km, could have taken about 9500 years. That would mean an average of only 42 m per year. These calculations, pretty far from reality—they ignore the periodic changes in dispersal trends caused by other factors— evidently hold only for ideal cases. Flightless insects in particular come across diverse barriers (the Danish straits after the Ancylus period, the Narke strait during the Yoldia period, etc.) and the immigration assumes a more erratic character. Some of the Ice Age refuges were probably farther away. Nevertheless, we may be justified in arguing that the normal speed of postglacial migration is not of a magnitude to be perceptible by direct observation in a matter of decades. Rapid migrations such as those described above (and partly mapped) must be due to special causes which influence, positively or negatively, either the capability of dispersal or the existence possibilities of an animal. 2. The area alterations are caused by man, either by assisting in dispersal of the animal directly or indirectly, or by cultivation of the land which provided new living conditions. a. That a species was transported by man is beyond doubt only in those cases where a clearly accidental occurrence in an alien region is involved. The following species of our fauna belong to this category: Carabus auratus Pterostichus madidus. Probably falling in this category we also have the following species, which have been resident for a somewhat longer period: Carabus monilis Lionychus quadrillum. Given their strictly synanthropous habits, at any rate in our region, the following species must also have arrived anthropochorously: Clivina collaris Sphodrus leucophthalmus. Pristonychus terricola It is highly probable that Carabus nemoralis and Dichirotrichus rufithorax (see p. 632) also originally arrived with man, although in the case of the 639 387) Carabus species this event occurred far back in time. It was also transported from Europe to North America. Such sudden, erratic faunal alterations take place in no other way than by transportation of species with human traffic. This is due not only to the rapid, convenient possibilities of transport with modern means of travel, even over very long distances, but also to the well-known fact that animals transported to quite new regions have shaken off their natural enemies and may then re- produce undisturbed. Applied entomological literature provides many current examples. The Fennoscandian carabid fauna includes no anthropochorous immi- grants that have multiplied explosively, however, among other Coleoptera mention may be made of: Philonthus rectangulus Sharp (Har. Lindberg, 1937; Benick, 1947; native to East Asia); Lathridius nodifer Westw. (Holdhaus, 1927-28, p. 613; probably native to Australia); Ptinus tectus Boield. (König, 1936; native to Australia); Tribolium destructor Uytt. (Kemner, 1936; according to Hinton, 1948, p. 16, probably native to East Africa); certainly Bohemiellina paradoxa Machulka (Jansson, 1947), whose country of origin is unknown. We have already discussed the significance of anthropochorous dispersal for carabids in general (p. 606). b. With the exception of the fjeld regions and the far north, the human in- fluence has had a decisive impact on the landscape of Fennoscandia. Especially in the south this took place quite early on and to a greater extent than gener- ally assumed, as shown by Iversen (1941) with regard to Denmark. Some carabids were negatively affected. Among the above-mentioned species (p. 636) that have decreased in recent times the 3 species under “a” were adversely affected, possibly directly by cultivation of the land and rational utilization of the forest. For Agonum bogemanni, to a lesser extent for A. quadripunctatum (as surmised in Part I, p. 48) the decline of the old method of clearing forests by burning (Swedish: svedjeburk) may have exercised a great influence. This is supported by the fact that the last two records of Agonum bogemanni in Fennoscandia (1943 and 1944) were made during the war in eastern Karelia and in the Isthmus of Karelia, where the forests were extensively burned down by gunfire. A. quadripunctatum was abundant there in some places. The 3 remaining species included above (p. 636) under “c” are of course strongly synanthropous (in our region even anthropobiont, bound to houses) but they live in old unhealthy buildings whereby their living conditions have become very difficult. It is a puzzle why Dolichus and Ptrostichus punctulatus ( “b”, p. 636) have become scarcer. In Germany they seem to be unaffected by cultivation of the land. Attention should also be drawn to an ecological group whose existence has been more and more seriously threatened in recent decades, namely, the steno- 640 558 topic riparian species. One after another our larger rivers are being regulated for the construction of electrical power stations. The natural water level with its more or less regular fluctuations is terminated, often considerable stretches of the water course are dammed up for water storage and the original banks submerged. All this means a series of catastrophic changes for the fauna asso- ciated with the waterside. Th. Palm told me that the fauna has changed out of recognition along the river Indals near Jtl Bispgärden since our visit to that place in 1930 (Lindroth and Palm, 1934, p. 29). Probably in the near future some of the stenotopic riparian species (in the genera Bembidion, Dyschirius, etc.) will be exterminated along all river systems. As to whether or not some species of insect from larger or smaller regions can be exterminated as a result of the collecting activity of entomologists, those responsible for this activity differ. At the worst this might happen in the case of diurnal especially conspicuous Lepidoptera, particularly if they have a short flying time and are also locally restricted from the very beginning. Given the last stipulation, I doubt even the intensive, quite professional collection of butterflies from time to time for export to foreign insect dealers, especially in the Abisko region in Lapland, has perceptibly influenced the composition of the fauna. My opinion that at least with respect to the carabids the most intensive collection activity is inconsequential from the viewpoint of conserva- tion of the species, was illustrated earlier (1948c, p. 45) with an example: Near Greby (Rapplinge parish) on Oland in June, 1946 all the detectable material of Cymindis humeralis from a restricted biotope (Fig. 11, p. 117) was collected for the purpose of various experiments. Yet the species was found to be still more abundant at that very place the following June. However, it must be con- ceded that “with the best of intentions” it might be possible to exterminate, for example, Carabus intricatus from Skane. Destruction of biotopes is far more disastrous for animals ee direct collecting activity. Conversely it is much more important, where necessary, to protect by law the biotope rather than the insect. It does not seem possible to decide whether, due to the indirect influence of mankind in recent decades, the larger number of insectivorous birds (for example, Beirne, 1947a, p. 41) has brought about a decrease in the carabid fauna. Of course Notini’s study (1943, pp. 33 ff.) showed that, for instance, the hooded crow (Corvus cornix) feeds largely on carabids. Much larger than the group of carabids negatively affected by mankind is the group increased as a result of human culture, with new regions for colonization and increased possibilities of dispersal. It is obvious that among the above-mentioned 16 examples of species in process of area expansion (p. 630), no fewer than seven are markedly favored by culture. Of the 9 “late immigrants” (p. 622)—besides the probably anthro- pochorously transported Clivina collaris and Lionychus—this is true especially of Amara fusca and Harpalus puncticeps. a > = 642 559 Definite promotion due to culture is enjoyed chiefly by species that show a late area expansion in the northern half of Fennoscandia, that is in regions where the cultural impact in the last century was most striking. Of these species I have carried out a more detailed study of Amara in- genua, which has been discussed elsewhere (pp. 525, 538). This species occurs in the north exclusively on civilized land (fields, harbors, refuse dumps, etc.). With regard to species increasing in abundance of similar habitats in Central Europe (for example, Amara fusca; Part I, p. 135) it was surmised that they are promoted by artificial fertilizers. However, the experiments carried out with A. ingenua showed that it reacts positively toward none of the gener- aily used fertilizers. But feeding experiments with the same species revealed that it has a strong predilection for seeds and fruits of certain synanthropous plants, especially Polygonum aviculare. So it must be assumed that cultivated soil attracts Amara ingenua chiefly by virtue of the vegetation. It probably at- tracts other species too, especially of the genus Amara. The character of the cultivated landscape is important taking the form of oases of open grass- and meadowland in the great northern forest region. Artificial drainage of culti- vated fields causes these oases to represent the only inhabitable places, often miles apart, for xerophilous and heliophilous animals. Mention must be made of an important factor facilitating rapid dispersal of the more or less xerophilous species through the great forest regions in recent decades. I mean the construction of road and railroad networks. Soon they criss-cross the entire forest region like a blood vascular system, and their gravelly edges form endless stretches of a remarkably uniform biotope, strips of the dry meadow type. They traverse not only forest but also large bogs, offe- ring excellent uninterrupted migration routes for all more or less xerophilous animals (and plants). Especially, railroad embankments in wooded areas are well known as rich sources of species of Amara and other xerophiles. Finally it is evident that the increasingly brisk traffic, even in the northern wooded region, by rail and road has provided new possibilities of displacement, for example by transport of gravel, hay and potatoes. 3. Area shifts are dependent on climate. This explanation presupposes that apparent alterations in climate in most recent times can be detected. There is already a considerable amount of literature attempting to esti- mate the impact of established or assumed climatic alterations on the distri- bution of animals. In this field the Finnish vertebrate zoologists in particular have been very active (Siivonen and Kalela, 1937; Kalela, 1938, pp. 226 ff.; 1940a, pp. 38 ff.; 1944; 1946; 1947; 1948; Merikallio, 1946, pp. 121 ff.). Va- rious Nordic entomologists share this viewpoint (for example, Munster, 1927, p. 289; H. Krogerus, 1945; Jansson, 1945; Valle, 1946; Ander, 1947; p. 55), in Germany chiefly Horion (1938; 1939). Earlier Ekman (1922, pp. 519-522) was skeptical of such explanations. 643 560 The above authors generally proceed from the meteorologically estab- lished fact that the mean monthly temperatures especially in winter and spring have gradually risen from a minimum in the middle of the last century, and that this rise (with the exception of the three years of severe winters 1940-43) has been especially strong in recent decades (Siivonen and Kalela, 1937; Has- selberg and Birkeland, 1940; Lunelund, 1942b, p. 3; Keränen, 1944; Angstrom, 1946, p. 97, Plate V; Ahlmann, 1948; for a general viewpoint, see A. Wagner, 1940). The zoologists try to correlate these climatic changes with the increase in numbers of several heat-requiring species, and to a lesser extent with the decline of markedly northern species. Although the monthly means reflect the general course of temperature changes over a series of years it is clear that a study of the trend of tempera- ture minima—especially in spring—would be more useful from the biological viewpoint (cf. p. 467). I have collected the figures for the most important months from the Swedish State Meteorological Institute. Since measurement of minimum temperatures started in Sweden around 1880 they do not go as far back in time as the curves of the media, but they should suffice for the present purpose. The figures considered most important (cf. p. 467) were those of the “criti- cal” months of April, May, September, October (Lindroth, 1931, p. 480). Both, the mean and the absolute monthly minima, were taken and, for comparison, the mean temperature of each of the above months. All these figures were calculated for five-year periods and curves for the two spring months were plotted after combining the values from three neighboring stations. Both the thermal effect of purely local conditions and the disturbing effect of moving the apparatus are thus rather compensated. The diagrams (Diagram 58-61) show the following: a. A general rise in temperature during the periods concerned is most marked in April and secondarily in May. In September the trend is utmost weak and in October imperceptible. So for the two autumn months only four stations are cited and they are ignored in the following discussion. With regard to the mean values of the above four months, the similar observations have been made in Finland (Siivonen and Kalela, 1937, p. 615; Keranen, 1944, p- 48). b. The temperature rise of the average minima is almost always greater than that of the mean temperature of the month. This is consistently true in April and in May, especially for southern Sweden. The absolute minima are naturally distributed somewhat irregularly, but the tapering off of really low values over the last decade is distinct, especially during April. c. The extent and general course of the recorded temperature improvement are not arbitrarily distributed all over the country. Certain parts of the country rather seem to be especially favored in this respect, in both April and May: The coastal region of northern Norrland, the inland of southern Norrland, the 652 561 Mälar and Hjälmar regions in central Sweden, and finally Gotland. It was desirable to depict this thermal trend cartographically, since it projects very regularly in respect of the average minima of April and May. But for this, values from more stations were needed, so I preferred to pre- pare maps comparing the mean monthly minima of the periods 1890-99 and 1930-39 (Figs. 91, 92). Norway (after “Norske Meteorol. Inst.” 1890 ff.) is also taken into consideration. These maps show still more strikingly the regular geographical distribution of climatic improvement in spring in terms of temperature minima. It is evident that, climatically speaking, this has primarily affected the continental parts of Scandinavia, whereas western Norway (probably excluding a small region in the Bergen-Sogn region* is virtually unaffected. We may be justified in taking these facts as confirmation of the supposed progressive oceanization of the Fennoscandian climate, repeatedly urged by others (Johansson, 1929; Ahlmann, 1939, p. 57; Angstrom, 1939, p. 65; Hesselberg and Birkeland, 1940, p. 17, and others). If this climatic alteration is correlated with the area shifts established or presumed above (pp. 622 ff.) for a number of carabids in Scandinavia, we seem to be justified in explaining the increase in frequency and abundance at least in part climatically (thermally) in the following cases: a. In eastern North Sweden. Agonum micans, Amara ingenua, A. munici- palis, A. similata, Trechus quadristriatus. In the case of the species of Amara, climatic and “cultural” factors (p. 637) may have combined to exercise their effect. b. In the central Swedish lake region. Acupalpus exiguus, ? Amara monti- vaga (cf. Munster, 1927, p. 288), Demetrias imperialis, ? Odacantha melanura, Oodes gracilis, Trechus discus. c. On Gotland. Harpalus rupicola. It is strange that the three exceptionally severe winters with their delayed spring in the years 1940—42 do not seem to have adversely affected the carabids that had immigrated or moved north in the preceding decades (cf. p. 312). The same observation was made by H. Krogerus (1945, p. 13) with respect to the lepidopteran fauna of southwestern Finland. It is possible that the shortened lifespan was compensated by the above-normal summer temperature in these three years. It is also conceivable that as soon as a species has got established in the (chiefly loco- and microclimatically) most suitable biotopes, it tolerates unfavorable macroclimatic conditions better than during the migration phase, when the individuals are exposed far more to the vagaries of nature. *The different state of affairs in the Bergen-Sogn region is reflected in values from only two stations. It is possible that purely local conditions took effect. (In connection with the spring mean temperature, cf. Hesselberg and Birkeland, 1940, p. 18, Fig. 7, p. 23, Fig. 15.) 1880 90 1900 10 20 30 1940 1880 90 1900 10 20 30 1940 no: Eo Saleen era Dn Sal Er BE he en aL ENE UD ee =5° Ee |= ER a TED! Haparanda, Pited Umea 102 6 N eee % Karesuando, Jockmock . Stensele — 202 1880 90 1900 10 20 30 1940 1880 90 1900 10 1940 ae ee 40° Visby. Kalmar, Karlshamn +0° = 5° Gote borg Halmstad, Lund —5° h 644 1880 90 1900 10 20 30 1940 1880 90 1900 10 20 30 1940 +0% == 5? — 52 Harnésand _ Gale N Uppsala Öslersund. Sveg ‚Bjuräker ~102 _10° 15? _ 1880 90 1900 10 20 30 1940 1880 90 1900 10 20 30 1940152 +5° 2207 + 0% Linkopung, Jonkoping Vaxy6 “ee a .. Stockholm, Vasterds. Orebro iO j © 645 _—— Diagram 58. April. Five-year periods in Sweden of mean temperature (Blank circles), mean daily minima (black circles) and absolute monthiy minima (hatched circles). Three neighboring stations were combined in each case. 1880 90 1900 10 20 30 1940 1880 90 1900 10 20 30 1940 +5° +5° +0° +0* Karesuando Jockmeck., Stensele Havaranda, Piteä Umea —5°. 1880 90 1900 10 20 30 1940 1880 90 1900 10 20 30 1940 gr ’ +5* +5°. Visby, Kalmar, Karlshamn Göteborg, Halmstad, Lund : ; N 646 1880 90 1900 10 20 30 1940 ® +0” Ostersund. Sveg Bjuräker o Linköping Jönköping | Vax, 6 15 647 1880 90 1900 10 20 30 1940 +5" 0% + Y Härnösand Gäule, Uppsala 1880 90 1900 10 20 30 1940 Stockholm. Västeräs, Örebro © +0 Diagram 59. May. Five-year periods in Sweden of mean temperature (blank circles), mean daily minima (black circles) and absolute monthly minima (hatched circles). Three neighboring stations were combined in each case. 566 Or6L 0€ 02 OL 0061 06 0881 oF do punsuagsQ N Oor6Ll O€ 07 OL 0061 06 0881 or6l 0€E 07 OL OO6L 06 0881 567 ‘Sc wEIZEIN 995 uONeueldx3 JO. ‘Uapams Jo suoneIs palesedas Ajapim Ino] ur (SID paysyey) ewiumu Ayyuoru Injosge pue (sajosio yorsq) eunumu Aep uesw ‘(sapolo Yuelg) Jinjeisdwus) uesw Jo sporiod Jeak-aaLj “Joquisaides ‘99 wesgelq I oyosddn 0+ eat RG: 30b+ „oI+ Orel 0€ 074 OL OO6L 06 0881 89 568 puns4esN, OvEL 0€ 07 OL 0061 06 0881 or6l 0€ 07 ol 0061 06 0881 569 ‘gc wEITEIA 298 UONeURIdxa 104 "USPamg Jo suonels payesedas Alopım InoJ ur (So. poyowey) euuu Alyyuow oinjosqe pue (sa[11 yor |q) ewrunu Ajlep weotu “(safosIo YURI_) sınyersduns] ueaw Jo spoldad Ieak-IAL{ ‘199090 “19 wesseld OF Gt or6l 0€ 07 ol 0061 06 0881 Alan Or6l 0€ 07 OL OO6L 06 0881 679 653 654 570 It should be emphasized that the precipitation climate—at any rate in Sweden—during the period (since 1860) has shown no such drastic or conse- quential change as the temperature climate (Bergsten, 1941; Ängström, 1941). Only in northern Sweden—especially in the fjeld regions—is a slight increase in precipitation perceptible. This manifests chiefly as an increase of about 20 days in the annual duration of snow cover. This change may have assisted the hibernation possibilities of the ground fauna. 4. Immigration of new species into Fennoscandia depends on changes having occurred in the original species area, situated outside the region. These alterations may be of a positive or negative nature. Of the former case an excessive pop- ulation density could cause an emigration. In the latter case, emigration could be caused by destruction or deterioration of the original biotopes. a. A hypothesis to that effect was proposed by Lonnberg (1924, pp. 112 ff.) to explain the late, rapid immigration of some aquatic birds into Sweden, chiefly of Nyroca ferina and N. fuligula. He assumes that emigration from their original homeland in West Asia followed general drying up of the lakes there. Later this hypothesis was elaborated, chiefly by Kalela (for example, 1940b; 1946, p. 35). Among carabids there is only one species of this ecological type, Bembidion transparens, which is associated with vegetation-rich shores and immigrated. late from the east. Of course this can apply at the most to its southern area in Fennoscandia (southern Finland, central and southern Sweden), whereas the insect must be long established in the far north, as is evident from the dimorphism map (Fig. 45, pp. 391 ff.). Possibly this interpretation also applies to the two transgrading species, Chlaenius costulatus and C. sulcicollis, which are prominent actively flying bog species. The former clearly shows its eastern origin. Hence the “drying up hypothesis” has at best limited validity for the im- migration of our carabid fauna. At present it can scarcely be judged whether other unfavorable events, such as forest fires, occasional floods, cultivation of the soil, etc., taking place outside the Fennoscandian region have caused a westward or a northward migration of carabids. b. The importance of population density for greater “dispersal urge” has been repeatedly mentioned, especially for birds and mammals (in Fennoscan- dia among others by Ekman, 1922, p. 517; Kalela, 1938, p. 247; 1944, p. 5). It is clear that this factor has been operative in various mass migrations of insects, for example, of Locusta migratoria, and among the Coleoptera, especially of various coccinellids. These examples show some animals with advanced psychic functions, for which a definite territory must be available per family for hun- ting or food intake. Then there are some insects which through periodic excess in numbers, and perhaps specialized food habits (coccinellids on aphids), are easily forced to emigrate as a direct result of lack of food. None of these con- © LY r o 2 Wiis u ee, L ER L 650 Fig. 91. April. Difference between mean minimum temperature of periods 1890-1899 and 1930-1939. Regions with rise of more than 1°C are hatched. I Be | ae \ EL. \ 3 ve Lg i | 3 A N 4) d € \ N = 651 Fig. 92. May. Difference between mean minimum temperature of periods 1890-1899 and 1930-1939. Regions with rise of more than 1°C are hatched. 655 573 ditions applies to carabids. Mass occurrence of carabids is very rare and, as far as | am aware, a normal emigration from a region has never taken place. It must be recalled that the large number of insects washed up on the shore of seas and the larger lakes (Palmen, 1944), often very rich in carabids, results chiefly from accumulation during anemohydrochorous transportation and not from mass emigration. The not very specialized feeding habits of carabids (p. 531) greatly reduce the effect of an accidental increase in population density. It seems improbable that the late immigration of several species to Fennoscandia can be explained by this alone. 5. Area expansion of a species may depend on a change in its ecology, so that new biotopes can be colonized. Often cited as an example among Coleoptera is Oryctes nasicornis L. Both, in Norway (since 1905; Munster, 1927, p. 289; Natvig, in Saalas, 1939, p. 376) and in Finland (since 1919; Saalas, l.c.) it is a late immigrant. It has greatly enlarged its area in Sweden too. Originally the species was undoubtedly an inhabitant of hollows of trees, and its rapid advance north is associated with its switching to “cultural” biotopes, where the beetle now generally breeds in sawdust heaps, in garden compost, etc. Probably a similar explanation applies to the unusually vigorous invasion of Corymbites cupreus aeruginosus Fbr. in Finland (Saalas, 1923, 1939) since the turn of the last century. It has become a pest to cultivated grassland and cereal fields. Among carabids there are only slight indications of such a shift to “cul- tural” biotopes in certain species, for example in Harpalus distinguendus and H. rubripes, Amara nitida and A. montivaga. It is not known whether colonization of new kinds of biotopes is due to changes—hereditary or other—in the insect itself. Nobody will claim that the Colorado potato beetle (Leptinotarsa) must have undergone change after it shifted to Solanum tuberosum; it is only that contact with this plant opened up enormous new areas for colonization. Similarly ornithologists seem inclined to believe that the well-known ecological shift of the blackbird (Turdus merula) to cultural biotopes represents a kind of non-heritable “ecological shift” (for example, Kalela, 1944, pp. 15-17). On the other hand, as an example of a mutation with altered ecological valence, mention may be made of the melanic form of the hamster (Cricetus cricetus) (Timofeeff-Ressovsky, 1940, pp. 97-99). I still cannot decide whether pronounced “ecotypes” (“ecological mu- tants”) also occur among the Fennoscandian carabids. The question must be investigated experimentally. Perhaps it might be worth the effort to study in this connection species like Demetrias monostigma, Dromius longiceps and D. sigma, which are characterized by a peculiar “double ecological occurrence” (see Part I). 656 657 574 This does not mean that any of the late immigrants among the Fennoscan- dian carabids had to emigrate from its original homeland on account of changed ecological requirements, except perhaps Amara majuscula? In conclusion the following two most important causes of recent changes in the Fennoscandian carabid fauna may be established: 1. Alteration of the landscape due to civilization, partly associated with anthropochorous dispersal. 2. Thermal climatic improvement, especially of the spring minima, over recent decades. However, sudden invasions of large numbers, such as those by Amara ma- juscula, which in many respects remind us of the immigration of the geometrid Eupithecia sinuosaria Ev. (Wahlgren, 1921), remain enigmatic. In these cases it is difficult to avoid the conjecture that a physiological alteration of the insect itself must have taken place. The persistent impression of the present section is undoubtedly that of pronounced lability on the part of the Fennoscandian fauna. During the ge- ologically insignificant period of one century, considerable faunal alterations have taken place. One must then ask if it is actually possible to assess, even approximately, faunistic events over the entire postglacial period of which the period considered here constitutes less than 1%, not to mention events even farther back in time. Yet how far the attempt is justified will be evident from the subsequent sections. Nevertheless, it can be safely argued that the last 100 years, chiefly because of drastic influence of civilization, have been the most fateful of the entire postglacial period for fauna and flora. Fossil Records Reconstruction of the postglacial—and older—history of the fauna is possible through the study of fossil and subfossil animals and of the peculiarities of the present-day distribution of each species. The latter method involves much speculation and must be pursued with caution. On the other hand, fossils (in- cluding subfossils) provide exact evidence of the faunal conditions of a particu- lar period in a particular region. Uncertainties are, first, age determination—in which today’s refined pollen analysis is a valuable aid—and, second, the diffi- culty of correctly identifying a more or less reduced, often postmortally altered animal fragment. J have recently (Lindroth, 1948a) considered this last ques- tion and shown that in fossil carabids both the corrugation of the chitin and the formation of more or less regular dots (especially on the intervening spaces of the elytra) may occur as a result of postmortem changes. But the study of microsculpture is often the only certain way to determine a fossil carabid frag- ment. For instance, the elytra of the subgenus Europhilus (Agonum) can be separated only in this way. The standard rule for the identification of a fossil must be: Only specimens that are recognizable with complete certainty should 658 S75 be given a name: new descriptions of “extinct” species should be provided only in unambiguous cases. One must always bear in mind that wrong deter- mination of a fossil can have far worse consequences than that of a recent animal. The study of subfossil insects has been neglected in Fennoscandia, especially because the peat geologists have almost ignored it. The best studied is the Quaternary subfossil fauna of Skane (and Denmark), which was worked out comprehensively by Henriksen (1933). His determinations seem to be thoroughly reliable, and I have only in exceptional cases found it necessary to check his records. However, he has included some of the older records from Skane only from the literature (G. Andersson, 1889; Holst, 1906, 1908, 1911; Kurck, 1917; Kolbe, 1933), and undoubtedly it would be desirable to check some of the determinations. Unfortunately, it was impossible to locate the material in question in any museum, and the records from Skane are given in responsibility of the original author. From the rest of Sweden there are only a few records on subfossil carabids in the literature, and the interglacial material that I recently worked out (Lin- droth, 1948a). Later I obtained various postglacial subfossils from different sources (ML, PU, SG; for abbreviations see below), chiefly from the “warm period.”t These records are included in the catalog below. It is lamentable that (in PU) of the “remains of beetles” collected in a thorough investigation of the peat bogs of Norrland by Post (1906), very few could be traced. With their precise dating they would have been especially interesting. The enormous quantities of peat samples of SG could certainly reveal var- ious subfossils of insects. With regret I have had to forgo this time-consuming work. Perhaps some young Swedish entomologist will devote himself to this fascinating field before too long. In addition to solitary finds, material from two rich Swedish subfossil localities was available to me. These findings deserve to be briefly described: a. Vgl Vargarda, Larke-mosse (Bengt Ekengren). The beetle remains were at a depth of 90 to 110 cm. Two peat samples at this level were kindly subjected to pollen analysis by Dr. Gosta Lundqvist (Sveriges Geol. Undersokn.) and were provisionally dated to the Bronze Age (possibly the earlier part, the Sub-boreal). More definite determination is not possible without a complete series of samples. At any rate, deposition must have taken place before the subatlantic climatic deterioration. Among the carabids, Calosoma inquisitor, Carabus coriaceus and C. canceilatus were present. The material belongs to the Zoological Institute, Lund. b. Dir Stora-Tuna, Skarsjo (E. Bergquist). The subfossils were located in a peat bog at various depths and a precise dating of each stratum remains to be (ef. p. 687; suppl. scient. edit.) 659 576 done. But it is indubitable that all the insect remains originate from the “warm period” (probably chiefly from the late atlantic period), so are older than the subatlantic period. Hazelnuts were abundantly represented in the deposits. A complete pollen analytical series through the entire peat profile was worked out by Gosta Lundqvist. Ten species of carabids were represented (see list below), of which Carabus arvensis and C. coriaceus are most interesting. The material belongs to “Sveriges Geol. Undersokn.”, Stockholm. Also noteworthy are the subfossil records made by Samuelsson at two places in central Dalarna (Alvdalen, Lokbodarne; Evertsberg). He has pre- cisely dated them, though not on the basis of pollen analysis. These findings have been utilized from the botanical viewpoint in a separate publication (Samuelsson, 1906). Of special interest is the occurrence of Trechus rivularis in the atlantic period. The material belongs to the Paleontological Institute, Uppsala. From Finland, Poppius (1911) determined and published the subfossil beetles in a number of postglacial peat samples collected by Harald Lindberg. Unfortunately I obtained only part of this material for examination, and these subfossils, which were rediscovered in the Helsinki Museum, lacked determi- nation labels by Poppius. However, since he lists the insect remains found in each stratum along with a precise description of the various profiles of the deposit, it is generally possible to decide which subfossil he had before him in each case when he named them. Revision—to the extent that the material could be traced—revealed the following: St. Kiukais, Panelia (PPP 1911, pp. 9, 38). Agonum cfr. gracile, 1 elytron = A. thoreyi. Ik Metsapirtti, Viisjoki (pp. 17, 36-38). Calathus micropterus is missing from this locality; however, there is an elytron of the same species bearing the label “Valkjarvi Siipsuo.” Amara cfr. acuminata, 1 damaged elytron = Pterostichus sp. Agonum dolens, 1 very poorly preserved elytron, probably = Europnilus sp. Ik Rautu, Osmina (pp. 18, 37-38). Pterostichus diligens, 1 elytron = P. nigrita. - Agonum cfr. gracile, several elytra = A. fuliginosum 1 specimen, Patrobus assimilis 1 specimen, Pterostichus ? diligens 1 specimen. Ik Rautu, Osmina (pp. 18, 37-38). Agonum cfr. gracile, 2 records, 1 poorly preserved elytron = ? Europhilus sp. A. viduum 2 elytra, poorly preserved = ? Agonum sp. Trichocellus placidus, 1 elytron, poorly preserved = Pterostichus sp. Ik Sakkola, Isosuo (pp. 22-24, 36-38). Agonum cfr. gracile, 1 elytron = Europhilus sp. Pterostichus diligens, 1 elytron = Pterostichus ? diligens. Oodes helopioides, 1 elytron = Oodes helopioides. ST, Oa Ilmola (p. 38). Agonum cfr. gracile, 1 elytron = Agonum gracile. It is regrettable that in the Helsinki Museum no record material was to 660 be had for the three species Bembidion grapei, Pterostichus vermiculosus and 663 P. ? archangelicus Popp. found in the late glacial deposits from Ik Kivennapa (PPP. 1911, pp. 16, 36-37). At any rate the determination of the Bembidion species is to be considered very uncertain. The highly characteristic Prerostichus vermiculosus could hardly have been mistaken. On the whole the determination of subfossils by Poppius does not give the impression of being reliable and generally cannot be accepted without verification. Finnish entomologists who have observed his identification work tell me that it was hasty and superficial. Recently, A.L. Backman has diligently assembled extensive subfossil bee- tle material, especially from Osterbotten, which is now preserved in MH. I studied the Carabidae material. Unfortunately it was possible to obtain pre- cise information, especially on age, only for part of these records, but all of them are postglacial. The records are incorporated in the list below. In the first “special” part of this book the then known subfossil records were given along with each species. Since then considerable contributions have been added. Also I overlooked some of those in the literature. For these reasons, and for greater clarity, it is advisable here to give a complete list of records of subfossil carabids in Fennoscandia. Denmark, whose postglacial history largely tallies with that of Skane, is included. Subfossil records from other regions have not been included. For them, reference may be made to Part I of this book and to the Supplement at the end of this part. As far as possible, along with each species an indication of the time period is given according to the usual scheme of late and postglacial periods (Table 36). There is no consensus among Quaternary geologists, so this scheme is partly a compromise among different opinions (Hogbom and Lundqvist, 1930, p. 157; Gross, 1930, p. 94; Nordhagen, 1933; Henriksen, 1933, p. 92; Wag- ner, 1940, p. 128; Sandegren and Asklund, 1946, p. 96). Dissenting opinions, especially with regard to the Ancylus period, have been expressed by Munthe (1940, p. 63, etc.) and Sauramo (1942, pp. 232 ff.). It would be incorrect to introduce, with Henriksen (l.c.), the hypotheti- cal mean July temperature of the different periods (following A.C. Johansen, 1906; Johansen and Lynge, 1917; according to this the maximum temperature in the sub-boreal). There are still no reliable methods of determining such figures, even approximately. It is also rather audacious to shift the postglacial temperature maximum already to the later part of the Boreal period (Gross, 1930, p. 95; Wagner, 1940, p. 129, puts it still farther back, around about 7000 B.C.). But it gives an idea—certainly correct in principle—of the steep temperature rise at the end of the late-glacial period (before the bipartition of ice in Jamtland). On the other hand, according to Sauramo (1942, p. 263) the thermal optimum set in during the first half, according to v. Post (1924, 664 665 578 p. 114) and Granlund (1932, p. 169) in the later half of the atlantic period; also in Denmark during the Atlanticum (Iversen, 1941, p. 38). The “warm period” certainly extended into the sub-boreal; during this period, the maxi- mum of Trapa natans occurred in Sweden (Samuelsson, 1934, p. 194; see also Sernander, 1908, pp. 213 ff., and the Trapa map by Jessen and Milthers, 1928, p- 349). In the checklist below the following abbreviations are used: For authors BCM—A.L. Backman HNR—K.L. Henriksen LBH—Harald Lindberg LTH—C.H. Lindroth PPP—B. Poppius. For institutions GH—Geological Research Institute, Helsinki. GL—Geological Institute of the University, Lund. MC—Zoological Museum, Copenhagen. MH—Zoological Museum of the University, Helsinki. ML—Zoological Museum of the University, Lund. PU—Paleontological Institute of the University, Uppsala. SG—Sveriges Geolog Undersokning (Swedish Geological Survey), Stockholm. Catalog of Subfossil Carabidae Found in Fennoscandia and Denmark Abax ater Denmark. Sjaelland, Fems¢lyng-Mose, subatlantic (HNR, 1933, p. 140). Agonum consimile Sweden. Ang Langsele, (Riss-Würm-) interglacial (LTH, 1948a, p. 11). Agonum dolens ? Finland. Ik Metsapiriti, late postglacial (PPP, 1911, pp. 17, 38); probably wrongly determined (see p. 659). Denmark. Sjaelland, Vintappermosen, late glacial (HNR, 1933, p. 130). Agonum ericeti Denmark. Sjaelland, Fems¢lyng-Mose, subatlantic (HNR, 1933, p. 130). 579 661 Table 36. Outline of late-glacial and postglacial development of northern Europe. Compiled from Högbom and Lundgvist, 1930; Gross, 1930; Henriksen, 1933; Nord- hagen, 1933; Lindqvist, 1935 (p. 51); Munthe, 1940; Wagner, 1940; Sauramo, 1942; Sandegren and Asklund, 1946 Sea Historical — 1950 period Subatlantic (generally Deteriora- moist). tionof „> climate Present- day Baltic Sea Iron age Bronze Age — 1000 Sub-boreal (generally dry) Late Stone Age 3000 — — 3000 Littorina Sea 4000 Atlantic (generally Warm moist) Early Stone Age — 5000 5000 — 2 > period Warm »> ni a Maxi- | Boreal AN Bipartition in In BE (generally First 7000— Jämtland dry) menin »> —7000 _ Stockholm Sweden? 8000 — region Subarctic Vgl. Billi a= gl. Billingen Baltic Ice Lake Arctic 13000 — ___ Southcoast . [4000— of Skane Fig. 93. Fennoscandian localities of subfossil Carabidae. Sweden: 1—Ska Loddesborg; 2—Akarp; 3—Bara and Nevishög; 4—Ska Svedala; 5—Trälleborg; 6—Lindved; 7—Onnarp; 8-Skä Skurup; 9—Snéres- tad, Sjörup, and V. Nöbbelöv; 10—Oja; 11—Skä Toppeladu gard; 12—Bjarsjo- lagärd; 13—Brösarp; 14—Gtl Havdhem; 15—Sproge; 16—Frojel; 17—Larbro; 581 Agonum fuliginosum Sweden. Ang Langsele, (Riss-Würm-) interglacial (LTH, 1948a, p. 11). Ska Lindved, boreal (HNR, 1933, p. 133). Dir Evertsberg, in sub-boreal Sphag- num peat, elytron without apex (G. Samuelsson, PU!); Stora-Tuna Skarsjo, “Warm period”, 3 elytra (E. Bergquist, SG!). Nbt Pitea, Borgsfors, probably sub-boreal, 1 elytron (G. Nordwall, MH!). Finland. Tb Pihtipudas, Alvajarvi Lake, atlantic or early sub-boreal, 1 elytron (BCM, MH!). Ab Poytya, Pinomaesuo, sub-boreal, 1 elytran (GH!). Om Lappajarvi, Karna, Littorina, 4 elytra (BCM, MH!). Ob Alatornio, post- Littorina, 1 elytron (BCM, MH!). Om Rautio, Karkiskyla, postglacial, 1 elytron (BCM, MH!); Alavieska, Koiraneva, postglacial, 1 elytron (BCM, MH!); Ned- ervetil, postglacial, 1 elytron (BCM, MH!); Perho, postglacial, 1 elytron (BCM, MH!). Ik Rautu. Osmina, late postglacial, 1 elytron (LBH, MH!). The doubtful record from NI Inga, late postglacial (PPP, 1911, pp. 11, 38), is very uncertain. Denmark. Jutland, Esbjerg, late-glacial; Skagen, “Warm period” (HNR, 1933, p. 133). Agonum gracile’ Sweden. Ska Skurup, Saritslov, “Warm period” (Kurck, 1917, p. 52; HNR, 1933, p. 132). Finland. Om Haapajarvi, Haga, atlantic, 1 elytron (BCM, MH!). Sb. Vierema, Kallolampi, Littorina, 1 elytron (BCM, MH!). Om Alavieska, Koiraneva, and Keinola, postglacial, 1 elytron each (BCM, MH!); Oulais, Hirvineva, postglacial, 1 elytron (BCM, MH!). Oa Imola, postglacial, 1 elytron (PPP, 1911, p. 38; MH !). The records of “Agonum cfr. gracile” from various postglacial peat deposits of South Finland by Poppius (1911, p. 38) are to be disregarded (with the exception of Oa Ilmola). 18—Vgl No 1; 19—Vargarda; 20—Dsi Mellerud; 21—Sma Djurs- dala; 22—Osl Jarnlunden; 23—Upl Vittinge; 24—Dir Stora Tuna; 25—Evertsberg; 26—Lokbodarne; 27—Hls Bollnas; 28—Jtl Pilgrimstad; 29—Ragunda; 30—Äng Härnön; 31—Längsele; 32—Bjurholm; 33—Pil Arvidsjaur 34—Nbt Borgfors; Finland: 35—Ob Alatornio; 36—Ranua; 37—Om Pelso; 38—Oulais; 39—Alavieska; 40—Rautio; 41—Haapavesi; 42—Kannus; 43—Nedervetil; 44—Toholampi; 45—Sievi; 46—Haap- ajärvi;, 47—Pyhajarvi; 48—Kaustinen; 49—Reisjarvi; 50—Kärsämäki; 51—Lappajarvi; 52—Om Perho; 53—Tb Pihtipudas; 54—Sb Vieremä; 55—Oa Lapua; 56—Nurmo; 57—Ilmola; 58—Sb Jorois; 59—St Merikarvio; 60—Pomarkku; 61—Kinkais; 62—Ab Poytya; 63—Al Kokar; 64—NI Kyrkslatt; 65—Ik Valkjarvi; 66—Kivennapa; 67—Rautu; 68—Metsapirtti; 69—Sakkola; Norway: 70—35 Tromso. 664 Fig. 94. Calosoma sycophanta in Ancylus clay, Up! Vittinge, Skattmanso. Right elytron (right) and its impression (left). (Photo: C. Larsson). Agonum lugens Denmark. Jutland, Herning (Riss-Wurm-) interglacial (HNR, 1933, pp. 131, 286). Agonum micans 666 Sweden. Ska Skurup, Kallsjö, “Warm period,” (HNR, 1933, p. 132). The determination is probably correct (MC!). Agonum ? moestum Denmark. Jutland (Riss-Wurm-) interglacial (HNR, 1933, pp. 131, 286). The determination is to be considered uncertain. The record from Sweden by Mjoberg (1916, p. 7) is erroneous (LTH, 1948a, p. 5). Agonum mülleri Denmark. Jutland, Kandestederne, “Warm period” (HNR, 1933, p. 130). Agonum piceum Finland. Om Pelso, early Atlantic, “Warmth flora,” 1 elytron; Karsamaki, Kintasahonrame, and Karsamenneva, early atlantic, “Warmth 583 flora,” respectively 5 and 1 elytra; Pyhajarvi, Hoikkaneva, early postglaciai, 1 elytron. Ob Alatornio, early postglacial 1 elytron. Om Rautio, Kärkiskyla, postglacial, 1 elytron; Reisjärvi, postglacial, 1 elytron; Alavieska, Koiraneva, postglacial, 1 elytron; Haapavesi, Piipsanneva, postglacial, 1 elytron. St. Merikarvia, postglacial, 1 elytron. (All BCM, MH!). The late postglacial record from Oa Ilmola (G. Andersson, 1898, p. 143) was doubtfully determined. Denmark. Jutland, 2 localities, one of these records from the “Warm period” (HNR, 1933, p. 132). Agonum thoreyi (see map in Fig. 95, p. 675) Finland. Om Lappajarvi, Karna, Littorina, 2 elytra. Ob Alatornio, late Lit- torina (hence probably sub-boreal). 1 elytron, Om Alavieska, Kotarame, prob- ably sub-boreal, 1 elytron; Alavieska, Keinola, postglacial, 1 elytron; Rautio, Karkiskyla, postglacial, 1 elytron; Kaustinen, postglacial, 1 elytron (all BCM, MH!). Oa Nurmo, Paukaneva, sub-boreal, 1 elytron (GH!). St. Kiukais, Pan- elia, late postglacial (see PPP, 1911, pp. 6 ff.), 1 elytron (LBH, MH!). St. Pomarkku, very fresh elytra of 2 specimens (GH!). Denmark. Sjaelland, Fems¢lyng-Mose, “Warm period” (HNR, 1933, p. 133). Agonum viduum Sweden. Ska Toppeladugärd, late-glacial (Holst, 1906, p. 8; HNR, 1933, p. 131). The determination would require verification. Finland. Ik Sakkola, late postglacial (PPP, 1911, p. 39); determination uncertain (see p. 659). Amara alpina ‘Sweden. Ang Härnön; Ang Längsele; Jtl Pilgrimstad; (Riss-Würm-) inter- glacial (LTH, 1948a, p. 10). Ska Snärestad, late-glacial (G. Andersson, 1889, p- 30; HNR, 1933, p. 140). Denmark. Jutland, Mart¢érv-Bakker, late-glacial (HNR, l.c.). Amara bifrons Sweden. Gtl Sproge, Snoder, Littorina period, 1 elytron (v. Post, 1903, p- 356, “Anchomenus sp.”, PU!). Amara quenseli Sweden. Ska Trälleborg, late-glacial (HNR, 1933, p. 166). 584 Bembidion ? glaciale Heer Sweden. Skä Svedala, late-glacial (HNR, 1933, p. 128; MC!). Today missing from the whole of northern Europe. I have examined the elytron (apex lacking) and will not venture to confirm the identification. The strongly developed 7th band is strange, but too little material of this variable species was available to me. At any rate this species is no longer found in Fennoscandia. Bembidion ? grapei Finland. Ik Kivennapa, Linnamäki, late-glacial (PPP, 1911, pp. 16, 36). The determination is very uncertain. Bembidion hasti Sweden. Jtl Pilgrimstad, (Riss-Würm-) interglacial (LTH, 1948a, p. 9). Ska Bara, late-glacial (HNR, 1933, p. 128; MC!). Bembidion prasinum Sweden. Jtl Pilgrimstad, (Riss-Wurm-) interglacial (LTH, 1948a, p. 8). Bembidion quadrimaculatum Denmark. Sjaelland, Fems¢lyng-Mose, “Warm period” (HNR, p. 128). Bembidion repandum Denmark. Sjaelland, Knabstrup, late-glacial (HNR, 1933, p. 127, “femora- tum’’). The well-preserved elyiron shows the microsculpture very characteristic of repandum (MC). Bembidion stephensi Denmark. Jutland, Skagen, “Warm period” (HNR, 1933, p. 127, “rufi- corne”). Blethisa multipunctata Sweden. Gtl Sproge, Snoder, Littorina period, head and prothorax (v Post, 1903, p. 356, “Feronia sp.”, PU!). Finland. Ni Kyrkslatt, late postglacial (PPP, 1911, pp. 13, 36). Calathus fuscipes Denmark. Jutland, Egtved, “Warm period” (HNR, 1933, p. 129). Calathus melanocephalus Norway. Tromso, undetermined age, 1 elytron (MT!). 668 585 Calathus micropterus Sweden. Dir Evertsberg, sub-boreal-atlantic 2 elytra (G. Samuelsson, PU!). Finland. Om Lapparjarvi, Karna, Littorina, 1 elytron (BCM, MH!). Ik Metsapirtti, Viisjoki, postglacial (PPP, 1911, pp. 17, 38); Valkjarvi, Siipsuo, postglacial, 1 elytron (Har. Lindberg, MH!). Calosoma inquisitor Sweden. Ska Skurup, boreal (HNR, 1933, p. 123). Vgl Vargarda, Larke- mosse, sub-boreal, thorax, both elytra, etc. (B. Ekengren, ML!). Calosoma sycophanta Sweden. Upl Vittinge, Skattmanso, Ancylus clay (boreal) 1 complete elytron (Fig. 94, PU!). Vgl Nol, atlantic, Ds] Mellerud, probably likewise “Warm period” (LTH, 1942b). Carabus arvensis Sweden. Dir Stora-Tuna, Skarsjo, “Warm period,” anterior half of an elytron (E. Bergquist, SG!). Carabus cancellatus Sweden. Vgl Vargarda, Larke-mosse, sub-boreal, 1 elytron without apex (B. Ekengren, ML!). Carabus coriaceus Sweden. Dir Stora-Tuna, Skarsjo, “Warm period” (apparently late at- lantic), abdomen and both elytra (E. Bergquist, SG!). Vgl Vargarda, Larke- mosse, sub-boreal, 2 fused elytra and various other parts of the body (B. Eken- gren, ML !). Carabus glabratus Sweden. Ska Sjorup, Brakna-mosse, boreal (G. Andersson, 1889, p. 24; HNR, 1933, p. 124). Dir Stora-Tuna, Skarsjo, “Warm period,” parts of thorax and 2 elytra (E. Bergquist, SG !). Denmark. Sjaelland, Fems¢lyng-Mose, subatlantic (HNR, 1933, p. 124). Carabus granulatus Sweden. Ska Skurup, Sandakra, “Warm period” (HNR, 1933, p. 124). Gtl Larbro, early postglacial (Sernander, 1894a, p. 64). . 586 Carabus hortensis Denmark. Sjaelland, Fems¢lyng-Mose, both from the “Warm period” and subatlantic (HNR, 1933, p. 124). Carabus nitens Denmark. Sjaelland, Fems¢lyng-Mose, “Warm period” (HNR, 1933, p. 123). Carabus violaceus Sweden. Dir Stora-Tuna, Skarsjo, “Warm period,” well preserved remains of at least 4 specimens (E. Bergquist, SG!). Denmark. Sjaelland, Fems¢lyng-Mose, both from the “Warm period” and subatlantic (HNR, 1933, p. 123). Chlaenius costulatus Sweden. Ska Toppeladugard, late-glacial (Holst, 1906, p. 8, “quadrisulca- tus”; HNR, 1933, p. 141). Chlaenius nigricornis Sweden. Ska Sjorup, Sodra-Vallosa, “Warm oe (G. Andersson, 1880, p. 20, HNR, 1933, p. 141). 669 Cicindela campestris Denmark. Sjaelland, Fems$lyng-Mose, subatlantic (HNR, 1933, p. 122). Clivina fossor Sweden. Ska Toppeladugard, “Warm period” (HNR, 1933, p. 127). Cychrus caraboides Sweden. Sma Djursdala region, late-boreal (Sundelin, 1917, p. 158). Dir Alvdalen, Lokbodarne, late-atlantic, thorax (G. Samuelsson, PU !). Ogl Jarnlunden, sub-boreal (Sundelin, /c.). Dir Stora-Tuna, Skarsjo, “Warm period,” elytra of 2 specimens (E. Bergquist, SG!). Ska Skurup, Saritslov, “Warm period” (Kurck, 1917, p. 51; HNR, 1933, p. 125). Denmark. Sjaelland, Lillemose and Femsélyng-M¢se, “Warm period” (HNR, L.c.). Diachila arctica Sweden. Ang Härnön, (Riss-Würm ?-) interglacial (LTH, 1948a, p. 8). 587 Diachila polita Sweden. Ang Härnön, (Riss-Würm ?-) interglacial (LTH, 1948a, p. 8); Langsele, (Riss-Wurm-) interglacial, fragment of an elytron (Sandgren, SG!). Elaphrus cupreus Finland. N\ Kyrkslatt, late postglacial (PPP, 1911, pp. 13, 36). Elaphrus lapponicus Sweden. Ang Längsele, (Riss-Wirm-) interglacial (LTH, 1948a, 8). Denmark. Jutland, Nérre-Lyngby, late-glacial (HNR, 1933, p. 126). Harpalus nigritarsis Sweden. Ang Langsele (Riss-Würm-) interglacial (LTH, 1948a, Pa): Leistus rufescens Finland. Om Lappajarvi, Karna, Littorina, i elytron (BCM, MH!). Loricera pilicornis Denmark. Sjaelland, Refsnaesgaard, “Warm period” (HNR, 1933, p. 126). tNebria fossilis Kolbe Sweden. Ska Oja, late-glacial (Holst, 1908, p. 5; Kolbe, 1933, p- 212; HNR, 1933, p. 126). Nebria gyllenhali Sweden. Jtl Pilgrimstad (Riss-Wurm-) interglacial (LTH, 1948a, p. 7). Denmark. Sjaelland, Vintappermosen, late-glacial (HNR, 1933, p. 125). Notiophilus aquaticus Sweden. Ang Härnön (Riss-Würm ?-); Jtl Pilgrimstad (Riss-Würm-) in- terglacial (LTH, 1948a, p. 7). Ska Svedala and Västra-Nöbbelöv, Mossby, late- glacial (HNR, 1933, p. 126, “coriaceus”). 670 Oodes helopioides Sweden. Skä Skurup, Munkholms-mosse, “Warm period” (Kurck, 1917, p. 55). The record mentioned from the interglacial deposits near Äng Härnön (Mjoberg, 1916, p. 7) must be erroneous (LTH, 1948a, p. 5). 588 Finland. Om Karsamaki, Kintasahonrame, early atlantic, “Warmth flora,” 1 elytron (BCM, MH!). Ik Sakkola, Isosuo, late postglacial (PPP, 1911, pp. 24, 36; MH!). Denmark. Jutland, Tuesb¢l, (Riss-Wurm-) interglacial (HNR, 1933, p. 141). Patrobus assimilis Sweden. Ang Langsele, (Riss-Würm-) interglacial (LTH, 1948a, p. 9). Ska Skurup, Kyrkomossen, aspenpine-zone, 1 elytron (GL!). Finland. Om Pelso, early atlantic, 1 elytron (BCM, MH!). Oa Lapua, early atlantic, 1 elytron (GH!). Ik Rautu, Osmina, late postglacial, 1 elytron (LBH, MH!). Om Kannus, Hietajärvi, late postglacial, 1 elytron (GH!). Norway. Tromso, undetermined age, 1 elytron (MT!). Patrobus septentrionis Sweden. Ang Längsele, (Riss-Würm-) interglacial (LTH, 1948a, p. 9). Ska Tralleborg; Svedala; Toppeladugärd; late-glacial (Holst, 1908, p. 4; Kolbe, 1933; p. 210; HNR, 1933, p. 129). Jtl Ragunda, late postglacial (Sandegren, 1924, p. 45). Denmark. Sjaelland, Jyderup and Fems¢lyng-Mose; Jutland, Esbjerg; late- glacial (HNR, l.c.). At least in Denmark, the subspecies australis may be present in part. Pterostichus anthracinus Denmark. Sjaelland, Fems¢lyng-Mose, subatlantic (HNR, 1933, p. 137). Pterostichus ? archangelicus Popp. Finland. Ik Kivennapa, Linnamaki, late-glacial (PPP, 1911, pp. 16, 37). This species is today distributed westward as far as the Kanin Peninsula. Pterostichus aterrimus Sweden. Ska Brösarp, Holmana, transition boreal-atlantic; Skurup. Sandakra, “Warm period” (Kurck, 1917, p. 97; HNR, p. 135). Pterostichus diligens Sweden. Ska Akarp region, “Alnarp River,” ? preglacial (Holst, 1911, p. 40; Kolbe, 1933, p. 210), Ang Härnön, (Riss-Würm ?-), and Langsele, (Riss-Würm-) interglacial (LTH, 1948a, p. 10). Ska Toppeladugard, late-glacial (Holst, 1906, p. 8; HNR, 1933, p. 139). Dir Alvdalen, Lokbodarne, atlantic, 2 right elytra (G. Samuelsson, PU!). Pil Arvidsjaur, Langmyren, late postglacial (v. Post, 1906, pp. 254 ff.), 1 elytron (PU!). Dir Stora-Tuna, Skarsjo, “Warm period,” 2 elytra (E. Bergquist, SG!). 589 Finland. Ik Sakkola, Isosuo, and Rautu, Osmina, late postglacial (PPP, 1911, pp. 18, 23, 37); both determinations uncertain (see p. 659). Tb Pihtipu- das, Alvajarvi, atlantic or early sub-boreal, 1 elytron (BCM, MH!). Al Kokar, Tell-mossen, later Littorina, both elytra (BCM, MH!). Denmark. Sjaelland, Fems¢lyng-Mose, “Warm period” and subatlantic (HNR, 1933, p. 138). Pterostichus gracilis Sweden. Ska Loddesborg, “Alnarp River,” ? preglacial (Holst, 1911, p. 40; Kolbe, 1933, p. 210; HNR, 1933, p. 137). tPterostichus (Oreophilus) holsti Kolbe Sweden. Ska Nevishog, Torreberga, “Alnarp River,” ? pre-glacial (Holst, 1911, p. 40; Kolbe, 1933, p. 210; HNR, 1933, p. 139). Pterostichus minor Sweden. Gtl Sproge, Snoder, Littorina, elytron without apex (v. Post, 1903, p- 356, “carabid,”; PU!). Ang Bjurholm, Majamyr, ? sub-boreal (v. Post, 1906, pp. 234 ff.), elytron without apex (PU!). Nbt Alvsbyn, Borgfors, probably sub- boreal, 1 elytron (G. Nordwall, MH!). Finland. Om Lappajarvi, Karna, Littorina, 1 elytron. Sb Vierma, Kallo- lampi, probably Littorina, 2 elytra. Om Toholampi, Littorina, 1 elytron. Ob Alatornio, late postglacial, 1 elytron (all BCM, MH!). Pterostichus niger Sweden. Ska Skurup, boreal; Sjörup, Södra-Vallosa and Hassle-Bosarp, “Warm period” (HNR, 1933, p. 136). Gtl Frojel and Havdhem, early post- glacial (Sernander, 1894a, pp. 43, 62). Dir Stora-Tuna, Skarsjo, “Warm period,” 2 elytra (E. Bergquist, SG!); Evertsberg, sub-boreal, 1 almost complete spec- imen, bordering sub-boreal-subatlantic, 1 elytron (G. Samuelsson, PU!). The determination of a late glacial record from Ska Bjarsjolagard (G. Andersson, 1889, p. 10) is uncertain. Denmark. Sjaelland, 2 localities; Jutland, 3 localities; 4 from the “Warm period,” 1 subatlantic (HNR, 1933, pp. 135-136). Pterostichus nigrita Sweden. Gtl Sproge, Snoder, boreal, elytron without apex (v. Post, PU!). Ska Toppeladugard, “Warm period” (HNR, 1933, p. 136). Dlr Stora-Tuna, Skarsjo, “Warm period,” 4 thoraces, 4 elytra (E. Bergquist, SG!). 672 590 Finland. Om Sievi, Mustalampi, probably Ancylus, 1 elytron. Ob Ranua, Jankalampi, early atlantic, 2 elytra, Om Pelso, early atlantic, 2 elytra; Haa- pajarvi, Pekanneva, early atlantic, 1 elytron; Vierema, Kallolampi, Littorina, 2 elytra, Lappajarvi, Karna, Littorina, 3 elytra; Happajarvi, Murisjärvi, post- glacial, 1 elytron; Sievi, Kukko, postglacial, 1 elytron; Rautio, Iso-oja, post- glacial, 1 elytron (all BCM, MH!). Ik Rautu, Osmina, late postglacial, 1 elytron (LBH, MH!). NI Kyrkslätt and Sb Jorois, late postglacial (PPP, 1911, pp. 13, 2953911): Denmark. Jutland, Maarup, (Riss-Wurm-) interglacial (HNR, 1933, pp. 118, 136, 286). Sjaelland, Fems¢lyng-Mose, “Warm period” and subatlantic, a late-glacial record from Fyen is uncertain (HNR, 1933, p. 136). Pterostichus oblongopunctatus Sweden. Ska Sjorup, Brakna-Mosse, boreal (G. Andersson, 1889, p. 23); Toppeladugard, “Warm period” (HNR, 1933, p. 135). Ang. Bjurholm, Ma- jamyr, ? sub-boreal (v. Post, 1906, pp. 234 ff.), 2 damaged elytra (PU!). Finland. Om Vierema, Kallolampi, Littorina, 1 elytron (BCM, MH!). Denmark. Sjaelland, Fems¢lyng, “Warm period” (HNR, 1.c.). tPrerostichus primarius Kolbe Sweden. Ska Nevishög, Torreberga, “Alnarp River,” ? pre-glacial (Holst, 1911, p. 40; Kolbe, 1933, p. 212; HNR, 1933, p. 139). Closely related to P. aethiops (specifically different?). Pterostichus punctulatus Denmark. Jutland, Kandestederne, “Warm period” (HNR, 1933, p. 134). Pterostichus strenuus Denmark. Sjaelland, Fems¢lyng-Mose, “Warm period” and subatlantic (HNR, 1933, p. 138). Pterostichus (Lyperophorus) vermiculosus Men. Sweden. Ang Härnön, (Riss-Würm ?-) interglacial (LTH, 1948a, p. 10). Finland. Ik Kivennapa, Linnamaki, late-glacial (PPP, 1911, pp. 16, 37). A Siberian species, today distributed westward only as far as Pechora. Pterostichus vernalis Sweden. Ska Skurup, Kallsjö and Sandakra; Önnarp, Sote-mosse, “Warm period” (HNR, 1933, p. 134). Denmark. Sjaelland, Fems¢lyng, subatlantic (HNR, l.c.). 673 Sot Pterostichus vulgaris Sweden. Dir Stora-Tuna, Skarsjo, “Warm period,” 1 elytron and remains of a thorax (E. Bergquist, SG!).* Denmark. Sjaelland, Femsflyng-Mose, Jutland, Skagen; “Warm period” (HNR, 1933, p. 138). Pterostichus (Cryobius) sp. Sweden. Äng Härnön, (Riss-Würm ?-) interglacial (LTH, 1948a, p. 10). Trechus rivularis Sweden. Dir Alvdalen, Lokbodarne, atlantic, 1 elytron (G. Samuelsson, GU!). Trechus secalis Sweden. Dir Älvdalen, Lokbodarne, atlantic, 3, elytra (G. Samuelsson, GU!). So altogether 67 species of the present-day Fennoscandian carabid fauna, i.e. 19%, have been found in subfossil condition within the limits of the region or in Denmark; in addition there are 7 species—including 3 ostensibly extinct species—that no longer occur there. This is not an inconsiderable component, and it can be expected that these records will materially clarify the faunal history. In summary, the following conclusions with regard to the Fennoscandian fauna can be drawn from the subfossil carabids so far available: 1. In the pre-glacial time in Fennoscandia—so far as the age and animal determinations are reliable**—there were some species different from those found today. Their climatic requirements cannot be ascertained. For other reasons as well we must reject the idea that the members of our recent carabid fauna in any part of Fennoscandia represent the direct and in situ surviving descendants of the Fennoscandian Tertiary fauna. 2. During the last (Riss-Wurm-) interglacial period—the only one from which definitely datable records are available—the climate was in part dis- tinctly temperate, even in northern Fennoscandia. The finding of two cur- culionids (Ceuthorrhynchus quadridens Panz., Orobites cyaneus L.) north of the limit of their recent area (Lindroth, 1948a, pp. 18, 26) suggests that the cli- mate was at least not cooler than today. Jessen and Milthers (1928, pp. 334 ff.) have clearly shown that the climate of the last interglacial period in Denmark was in part much warmer than today. It is unlikely that this should not also apply to Fennoscandia. *In the same peat sample 2 elytra of the silphid Preroloma forsstromi Gyll. were found. **Concerning the present concept of the “Alnarp River,” see Brotzen (1948, p. 5) and Sandegren (1948, p. 39). The latter regards the deposits as interglacial. 674 592 3. During the last glaciation (Würm), at the southern edge of the north- ern inland ice, there was a fauna which—like the flora—included many of the characteristic members of our fjeld fauna. Also represenied were species (Amara alpina, Bembidion hasti, Elaphrus lapponicus) today missing from the Central European mainland. Such facts could be cited to support the view that our fjeld fauna, following the ice edge, has immigrated from the south. On the other hand, there are also to be found in the same late-glacial deposits species that have not succeeded in this, such as ;Nebria fossilis, Bembidion (?) glaciale, B. repandum, Pterostichus vermiculosus, and also Simplocaria deubeli Gnglb. (Henriksen, 1933, pp. 109, 174, 296), Thanatophilus trituberculatus Kirby (Sil- pha baicalica, |.c., pp. 165, 290), and !Donacia extincta Kolbe (1933, p. 213). 4. During the postglacial warm period (chiefly in the boreal and atlantic periods, around 7000-2500 B.C.) the more demanding elements of the fauna were certainly favored in the same way as those of the flora. This is evident from the wide distribution of Trapa natans at that time (Malmstrom, 1920; Jessen and Milthers, 1928, Fig. 35) and of the marsh turtle, Emys orbicularis (Isberg, 1930) (concerning the warm period, see also the summary by Nordha- gen, 1933, pp. 155 ff.). Two cases among the subfossil carabids may be cited. First, Calosoma sycophanta. As elucidated above (p. 621) this is presently a transgrading (i.e. nonresident) species in Fennoscandia. The three known Swedish subfossil finds are located north of the present transgression region (if the isolated recent locality in Dir is excluded). Theoretically even the sub- fossil finds could be the remains of accidental immigrants. But if it is taken into account that the chance of discovering a subfossil of Calosoma sycophanta is much more remote than of coming across the magnificent living insect, it must be assumed that the species actually had a wider distribution in Sweden earlier in the postglacial period. And since at one locality (Dsl Mellerud) 2 subfossils were found together it is even possible that the species at that time was actually resident here (Lindroth, 1942). The most interesting fact is that as early as in Ancylus time this species occurred as far north as Upl. The second example is provided by Oodes helopioides. The fossil find near Om Karsamaki, Kintasahonrame, is though not far outside the northern limit of the present area. Found at the same place were subfossils of markedly heat- requiring plants (including Najas flexilis), which undoubtedly originate from the “Warm period.” The sub-boreal finds of Pterostichus niger near Dir Evertsberg seem to lie outside the present area of the species as well. Still more remarkable as testimony of the “Warm period” is the record of the heteromeran Piatydema violacea in Trapa-gyttjat near NI Karis (G. Andersson, 1898, p. 40; Poppius, 1911). This characteristic species—which t(= mud of organic material [limnological deposits] which formed during deficiency of oxygen; suppl. scient. edit.) 676 677 593 can hardly be wrongly identified—is missing from the present-day fauna of Finland. An instructive example of how the present-day area can be supplemented by fossil records is provided by Agonum thoreyi (Fig. 95). They fill the gaps between some of its widely separated localities east and north of the Both- nian Sea. Whether climate influenced its retreat might seem uncertain in view of the records from Lk Muonis, 32 Salten, and 41 Vaggatem. But all three records are still somewhat doubtful; no record material was available from the two Norwegian localities and from Muonio where are also other improbable records (such as Acupalpus dorsalis, Bembidion ustulatum). It looks as though the occurrence of Agonum thoreyi at the northern end of the Gulf of Bothnia is a relict from the postglacial “Warm period” (See p. 692). How important the fossil record of a species of animal is for the history of the animal and for our idea of the environmental conditions then depends on the constancy of the ecological valency of the species. By “ecological valency” Hesse (1924, pp. 16-17) means: “the full scope (amplitude) of the conditions for life within which a species of animal is able to prosper.” It is thus a measure of the “stenotopy” of the animal (pp. 563 ff.). The question is whether the requirements for life determined for the present-day animal (experimentally or in nature) are true for individuals and populations of the same species in past ages. If not, we must be very cautious in judging prehistoric climatic and other conditions from fossil records. The zoologists principally concerned with this problem (for example, Nehring, 1890, p. 134; Warnecke, 1934, 1936a; Reinig, 1938, pp. 13 ff.) are unanimous that it is impossible to postulate the constancy of ecological valency of every species of animal. This orthodox notion would negate the whole concept of evolution. The “species” is often a heterogeneous conglomerate of several biotypes (ecotype; Turesson, 1927, etc.). As a “working hypothesis” (Warnecke, 1936a, p. 6; cf. on the contrary, Schultz, 1930, 1934a, 1934b) the “constancy rule” is very useful. If we work with short periods of time (such as within the limits of the Quaternary) and avoid drawing far-reaching conclusions from the fossil occurrences of a single species, the hypothesis gains a high degree of credibility. Nobody can doubt the arctic character of the climate—as manifested by the subfossil plants and animals—along the late- glacial ice edge in the southern Baltic region, or the wealth of paleontological evidence in favor of a postglacial Warm period (also see p. 687). But one should not posit precise thermal changes especially of the macro- climate, on the basis of the fossil record of a solitary species of animal or plant, as has been done, for instance, by A.C. Johansen (p. 663) with refer- ence to Molluska. Limnetic animals, and also plants (see Samuelsson, 1934, pp. 144-145), are particularly unsuitable for this purpose (Wesenberg-Lund, 1909, pp. 456 ff.), and even the distribution of terrestrial mollusks in Scan- dinavia shows little correspondence with the Johansen’s contentions (®kland, 675 Fig. 95. Agonum thoreyi. Fossil (postglacial) records are marked with crosses. Three northernmost recent localities (Norway, Finland) (cf. map in Part II) are omitted as not being reliable. 678°“ 595 1925, p. 153). Only forest trees may serve as reliable objects for this purpose, as illustrated chiefly by G. Andersson (1902b, 1910) on Corylus. Insects are dependent on microclimate (pp. 498 ff.) to such an extent that their depen- dence on “regional climate” is very indirect. For this reason I do not believe that Henriksen (1933, pp. 286-287) is correct in assuming that Otiorrhynchus dubius Ström is a species whose thermal requirements altered in the direction of “cold” during the Wurm Ice Age, a view already contradicted elsewhere (Lindroth, 1935a, p. 612; Warnecke, 1936, p. 8). However, it seems to me that we can draw the following modest conclusion on the basis of subfossil records of insects: A Quaternary period during which more than one species of insect is found to have occurred north or south of its present area may be presumed to have been warmer oh colder) than the climate there today. Relicts In none of the distribution phenomena is the dependence of the present-day area of an animal or plant species on historical factors and on the condi- tions of distant part so evident as in the case of relicts. These are “living fossils.” Researchers have greatly differed on the true or appropriate meaning of the term “relict.” This is not the place to discuss the whole history, but it is hardly possible to define my own principal concept without discussing earlier views. The term seems to have been used in print for the first time by Peschel (1875, p. 207; “Relictenfauna”). In Nordic literature the expression was in- troduced by botanists, especially by Sernander (1894b) and Nathorst (1895). The best botanical definition of the term may be Th.C.E. Fries’ (1913, p. 392: . .that by this is meant the occurrence of a plant or plant community at a location where the plant or plant community does not occur today, so long as there are no prehistoric, now nonoperative factors in the history of immigra- tion to account for it.” Zoological definitions agreeing in content have been provided by Ekman (1915) and v. Hofsten (1911, pp. 59-60; 1913; p. 34). A particularly clear definition is given by the first of these authors (Ekman, 1915, p. 5; and 1922, p. 278, in Swedish): “A species is a relict in a region where its existence [1935, p. 201, altered to ‘presence’] proves that the species it- self or its primitive form was left there under natural conditions now alien to the region in question.” The slightly altered wording of the definition in his contribution of 1935 (p. 201) adds that a relict species should occur “isolated from its main area of distribution.” As also emphasized by v. Hof- sten (1911, p. 59), the relict concept must be defined both temporally and locally. 679 596 Even if we ignore the very broad idea of the term “relict,” such as that given by Wesenbeg-Lund (for example, 1910, pp. 29 ff.)—which must be re- jected (see, for instance, A.C. Johansen, 1908)—it still admits of varying lim- its. Ekman holds that only those populations of species deserve to be called “relict” which, since becoming relict, have continued to live in situ until present-day, whereas all populations that have emigrated from a former relict center should be called pseudorelicts (Nathorst, 1895; or secundorelicts). In botany this differentiation of terminology is used still oftener, especially by Stoller (1921, p. 420): “From the paleobotanical viewpoint the relict char- acter of a plant for a particular locality can be recognized only if evidence can be produced that the species has lived at this location isolated from its present-day closed district of distribution since its immigration, which took place during an earlier period.” In limnetic organisms it is often possible to provide proof—or evidence supporting high probability—that the species concerned must have lived an uninterrupted life in this or that lake since the original isolation. This is not possible in the case of terrestrial flora; the assumption of this requirement by Stoller signifies “as good as complete elimination of the relict concept from the science of phytogeography” (Wangerin, 1924, p. 61). It is still worse in the case of the mobile elements of the terrestrial fauna (Lindroth, 1943a, p. 140). The following is an attempt to arrive at a consistent division of the relicts (s.i.) among terrestrial organisms into “pseudorelicts” and “true relicts”: a. “Pseudorelicts,” which could not have lived in situ continuously. Ex- amples: Alchemilla alpina and Rhodiola rosea in western Sweden, where they grow largely below the shoreline of the Littorina period (Sernander, 1894b; Nathorst, 1895; Lid and Zachau, 1929); Oodes gracilis in the Stockholm region where the inhabited lakes (with two exceptions) were first formed in the sub- atlantic period (Lindroth, 1943a). b. All remaining “relicts” (s./.), of which it cannot be stated whether they are true relicts (in Stoller’s and Ekman’s sense) or pseudorelicts. I cannot imagine what kind of evidence could prove the uninterrupted occurrence of a terrestrial plant or a terrestrial animal within a narrowly delimited terrain extending, say, over a few square kms. Hence for practical reasons I accept Drude’s (1918, pp. 44-45) and Wan- gerin’s view (1924, p. 67) in phytogeography. The latter (l.c.) writes: “secondary displacements of locality . . . even under the conditions of a definite, locally restricted new dispersal from old places of subsistence . . . in no way detract from the relict nature so long as a new immigration from outside does not take place, or the new dispersal does not assume the level of an extensive expan- sion, a case which probably has yet to be realized.” The term “pseudorelict” should not be rejected, but it must be subordinate to the broad term “relict,” not on a level with it. The above definition by Ekman also fits this broadened idea of a relict, provided his expression “region” is geographically not inter- 397, 680 Table 37. The most isolated localities or more or less close locality-groups of carabids in Fennoscandia. Only cases where the gap is situated within the region (including eastern Baltics and Denmark) are included. Records in the Russian part (before 1939) and on islands are excluded Assumed Species Isolate Nearest Distance cause of locality in km isolation 1 2 3 4 5 Bembidion aeneum Ks Paanajarvi 31 Bodo 706 Relict Agonum mannerheimi 12 Vardal St Ylane 655 Relict Harpalus rubripes 31 Bodö 24 Vägä 655 Import ? Agonum longiventure Up!, Dir, Gst, Estonia, 617 Relict 5 localities Narva Dromius longiceps ObHailuoto Ab Abo 529 Relict ? Dyschirius helléni 24 Vaga Lyl Tarna 517 Relict, also insuffi- ciently explored Chlaenius costulatus Ok Ruthinas- La Lampis 504 Migration salmi (Lammi) Trechus fulvus 26 Hitra 6 Nedstrand 504 Relict, also insuffi- ciently explored Agonum ruficorne 26 Valiersund 7 Bergen region 441 Import ? Bembidion minimum 26, 28, 2 2 Oslo 441 Relict localities Tachys bisulcatus Lyl Sorsele His Los 440 Insufficiently explored Amara convexiuscula Ik Kuokkala Ab Äbo 428 Import ? A. nigricornis 21 Sirdal 24 Skogseter 403 Relict Dromius quadrinota- 26 Hitra 18 Tyssedal 403 Import ? tus Dyschirius obscurus Li Ivalo Ob Uleaborg 403 Migration ? Pristonychus terricola 27 Trondheim 2 Oslo 403 Import ? Pterostichus niger 35, 36, 2 32 Hammernes 403 Relict ? localities Chlaenius nigricornis Nbt Neder- Sb Kuopio 397 Relict ? Kalix Trechus micros 27, 2 locali- 2 Oslo 384 Reiict ties Pterostichus vulgaris Ik Pelkosen- Om Haapavesi 365 Import niemi Bembidion transparens 33, 35, 3 Li Enare 353 Relict localities Agonum mülleri Ks Kuusamo Kb Polvijarvi 347 Import ? Bembidion grapei 6, 27, 2 localities 24 Vaga 346 Relict Amara lucida South Finland North 340 Migration ? 2 localities Kurland Cicindela campestris 35 Tromsdal 31 Bodo 340 Relict ? Species Bembidion nigricorne Calosoma reticulatum Dyschirius angustatus Bembidion hyperborae- orum Trechus discus Bembidion siebkei Calosoma sycophanta Dromius quadri- maculatus Trachypachys zetterstedti Amara curta A. erratica Dromius fene- stratus Brachynus crepitans Dyschirius politus Pterostichus ad- Strictus Dyschirius angustatus Abax ater Agonum gracilipes Isolate His Los Boh Fors Vrm Hoje Ok Sotkamo 27, 2 loca- lities Ks Paanajärvi Dir Korsnäs Äng Docksta Southwestern Finland, 3 localities 28 Tynes, Verdal Ik Metsäpirtti Ok Sotkamo Ni Hango Ks Salla Sma, 2 loca- lities . Ks Paanajarvi Southeastern Norway, 2 loca- lities Norway, Sweden, Finland (several localities) Nearest Nke Laxa Ska Trolle- Ljungby 24 Vaga Ks Sallo His Los Li Kyro Vgl Färdala Upl Alvkarleby Estonia, Narva 20 Geiranger Kn Juustjärvi KI Parikkala Stockholm Ob Uleaborg Vrm Forshaga Lp Lutto Jutland Distance in km 328 328 328 315 309 302 302 302 302 290 283 283 278 277 BAUT, 265 252 >250 Assumed cause of isolation Relict Relict ? Relict ? Migration ? Relict ? Relict Migration ? Migration Dying out spe- cies Relict ? Relict Insufficiently explored Migration Relict ? Relict Relict ? Relict Migration Agonum bogemanni, with several highly isolated, old, mostly poorly determined lo- calities may be added. This species is certainly dying out in our region. preted too narrowly. If one wishes to emphasize an established uninterrupted life in situ the word “eurelict” can be used. On the other hand, on two other counts I wish to modify Ekman’s defini- tion. First, it is not clear why relicts could not be separated under systematic entities other than “species.” Second, the latter qualification “isolated from the main region of distribution” (Ekman, 1935, p. 201) does not seem to be nec- essary, especially since it is not always possible to ascertain the “main region of distribution.” For instance, we may study the total distribution of Preros- 681 682 599 tichus kokeili Mill. which is split into small isolates (Holdhaus and Lindroth, 1939, Plate VIII); even in the copepod Limnocalanus grimaldii (Ekman, 1922, pp. 278-279; 1935, pp. 203-204) a main area is not readily evident. Although this animal in the Baltic Sea basin and in the neighboring lakes was earlier considered as a separate species (L. macrurus), Ekman (1922) pronounced this a relict as a whole, and in principle there is nothing against this view. One should also be able to use the word “relict” for the remnants of a species of animal or plant that is dying out, such as Hatteria (Sphenodon) even if perhaps internal causes have been more responsible than external. In light of these objections the following definition of the term “relict” seems to fill the bill: A relict is a stock of a genetic-taxonomic unit (sub- Species, species, genus, etc.) that is functionally separated from the remaining (as a rule wider, possibly prehistoric) area of the unit and was not able to invade the isolate during present natural conditions. The actual problem—not just one of terminology—is to draw the dis- tinction between relicts and immigrants (autoimmigrants—Ekman), between remnants and advance posts. For this purpose we will now take the carabid material as the basis. The prime characteristic of relicts is their isolated condition. Obviously we must draw up a list of the especially isolated localities or locality-groups of carabids in Fennoscandia in order to assess their possible relict character. Table 37 gives such a list. It omits all insular records, which have been suffi- ciently dealt with in a separate section (pp. 198 ff.), and localities whose nearest occurrence lies outside the limits of Fennoscandia. But to the extent that they are to be considered as relicts, these cases are included in the summary below. Speculative causal explanations for the isolates are suggested by the table itself. A survey shows: The definite isolates in 43 Fennoscandian carabids are divisible as follows: a. Relicts 17 species : 3 ar 23 b. Possible relicts 8 a SD, c. Accidental migrants 4 species| ee d. Probably accidental migrants 4 species P e. Displaced by civilization 3 species 8 2 EU EN — 8 species f. Probably displaced by civilization 5 species g. Insufficiently explored 2 species — 2 species. There remain a considerable number of carabid records in Fennoscandia that may be considered relicts with fairly high probability but are not isolated so far from the “main area” as to be included in Table 37. In the succeeding causally arranged list of relicts such cases are included if they are clear enough, along with species whose Fennoscandian area as a whole is relict-like. In accordance with the definition, all relicts go back to a time when natural conditions differed from today’s. Depending on the decisive environmental factor, at that time more favorable for the species in question, an appropriate 683 600 prefix precedes the word “relict”. We speak of “heat relicts,” “marine relicts,” etc. The possible relicts among the Fennoscandian carabids seem to belong to one of the following four categories (illustrated here by especially good examples):* A. Cold Relicts Geological and paleontological records show that during the first part of the late-glacial period definitely cold-requiring (arctic [see alpine] and subarctic [see subalpine]) flora and fauna lived at the southern edge of the Wurm Ice in northern Central Europe, and in the parts that were the first to become ice- free—at least in Denmark, Skane, and the Isthmus of Karelia. It is thus quite conceivable that even in the more southerly parts of Fennoscandia pronounced Nordic plants and animals survived at microclimatically favorable places as relicts from this distant past. In the later postglacial period the Subatlanticum! signified a climatic deterioration (including a decline in temperature), from which relicts can also be expected south of today’s continuous area. The presumed cold relicts are arranged here geographically for the sake of simplicity. a. In south-central Sweden. Nebria gyllenhah Pterostichus adstrictus The following predominantly Nordic species are more frequent and abun- dant in the southern Swedish highland or its margins: Cymindis vaporariorum Patrobus assimilis Miscodera arctica Trichocellus cognatus. Only the localities of Pterostichus adstrictus and of two other Coleoptera, Otiorrhynchus dubius Strom (maps in Henriksen, 1933, p. 294; Holdhaus and Lindroth, 1939, Plate XVII), and in particular Evodinus interrogationis L. (Fig. 96) are markedly isolated. A counterpart in the flora is Saussurea alpina (Erlandsson, 1940). Actually the south Swedish highland has climatically such a Nordic imprint (pp. 463, 474) that even much more cold-requiring forms might have been expected there. b. Along the coast of Bohuslan. Bembidion virens The counterparts in the flora are Alchemilla alpina and Rhodiola rosea (Lid and Zachau, 1929). *Good examples of relicts attributable to periods of drier or more humid climate are scarcely met with among our carabids (however, see Calosoma reticulatum and Abax ater under “heat relicts”). T(cf. Table 36, p. 661; suppl. scient. edit.). 685 601 c. On Gotland. Miscodera arctica Patrobus assimilis Nebria gyllenhali Only in Nebria is the isolation of course much greater than that due to the location of the island itself, but the two others have a markedly Nordic character as well (concerning Miscodera, see Fig. 56, p. 424). The best counterparts in the flora are Bartschia alpina and Pinguicula alpina. d. In southern and central Finland. Amara erratica Patrobus septentrionis (s. str.) Elaphrus lapponicus Pelophila borealis. Possibly also Bembidion grapei and Trichocellus cognatus. Elaphrus and Pelophila have also been found at one locality each in the eastern Baltic region. A far more isolated relict occurrence in southern Fennoscandia is shown by Simplocaria metallica Sturm (Fig. 106, p. 739). Further information on the isolated record of 1 specimen of Patrobus septentrionis near Kb Juuka, July 5, 1941, is provided by Krogerus (in litt.). The insect was found near Koljunkorpi close to the highest point of Juuanvaara at a height of about 300 m (i.e. a supra-aquatic region) on a steep north-facing slope on the bank of a brook emerging from a spring. The bank was stony with Sphagnum between the stones (S. riparium, S. girgensohnii, S. apiculatum). Nebria gyllenhali was found at the same place. In all the species included under a-d above, the isolation, which could indicate relict character, is not clear. At best it is seen in Pterostichus ad- strictus in Sma and Nebria gyllenhali in Gti. In the former species the nearest localities along the lower reaches of the River Klaralven in Vrm, are probably the result of a water transport by the river (Palm and Lindroth, 1936, p. 40). Cymindis vaporariorum and Patrobus assimilis, which lack any distinct “zone of obliteration,” are functionally brachypterous and therefore colonize new habitable regions relatively slowly. The thermophobic characters of the presumed cold relicts 1s clear in some cases. In the southern sub-areas they are frequently inclined toward stenotopy (cf. p. 567), sometimes on the edge of cold water (rivulets, deep lakes, the sea): Bembidion virens, Nebria gyllenhali (Holdhaus and Lindroth, 1939, p. 268), Patrobus septentrionis and sometimes in bogs; Patrobus assimilis, Trichocellus cognatus. In the case of flying species it is not possible to prove that they did not colonize their southernmost, more or less relict-like isolated localities by ac- cidental migration in recent time. But certain facts discredit that: the marked Scandinavian southern limit of Amara erratica and the distinct northern limit of the southern Finnish stock of Nebria gylienhali, the ciear bicentricity of 602 684 Fig. 96. Evodinus interrogationis L. Northwest European distribution. Compiled from literature and larger col- lections (summarized in Holdhaus and Lindroth, 1939, p. 199). This map is less complete than the maps of Carabidae but clearly shows the relict oceurrence in southern Sweden. 686 603 Elaphrus lapponicus in the fjeld regions; the not yet resolved bicentricity of Cymindis vaporariorum, Miscodera, and Trichocellus cognatus, which is main- tained by a narrow but certainly real gap in Jtl. All these limits are dynamically (dispersal-biologically) determined and therefore show that the species con- cerned at present lacks any pronounced capability of dispersal. Naturally there always occur small displacements of the southern limiting localities, of the “cold-requiring species” treated here (strictly speaking they are at the most “pseudorelicts”). But the first colonization of this region by any of the species in question as far as I can judge, cannot have been re- cent (under present-day climatic conditions). Either they are remnants of the first immigration from the south, or are evidence of numerous cold-requiring species having enlarged their area southward during a late postglacial period, i.e. during the Subatlanticum. Sernander (1894b) considers some isolated occurrences of Nordic plants in Upl (for example, Sceptrum carolinum as subatlantic relicts. Wahlgren (1909; 1935-41, p. 55) similarly interprets the southern Swedish locaiities of Lepi- doptera such as Oeneis jutta Hb. and Argynnis freija Thunb. Henriksen (1933, p- 326) believes that the clumsy flightless curculionid Barynotus squamosus Germ. may have immigrated to Jutland (from the north) since the postglacial warm period, where it represents a subatlantic relict. I have already contested this (Holdhaus and Lindroth, 1939, p. 258; also Sparck, 1940, p. 58). On the other hand I have applied (l.c., pp. 268-269) the same interpretation to the occurrence of Nebria gyllenhali at the shores of lakes Vättern and Vänern. I considered it improbable that these cold-requiring species could have survived the warm period at such advanced localities. My opinion has changed somewhat in recent years since I have come across the manifestations of micro-climate. A macroclimatic rise in temperature, to the extent assumed for the postglacial warm period in Sweden (up to about 2°C in summer), need not necessarily have brought about parallel microclimatic changes in all biotopes. At least those places directly affected by spring water or the sea may have been thermally altered. It is perhaps no accident that of the 11 “cold relicts” mentioned above, no fewer than 7 (Patrobus assimilis and Trichocellus cognatus only in the south) are hygrophilous. It seems possible that Nebria gyllenhali was able to survive the warm period not only near the spring water dripping from the coastal limestone rocks of Gtl but also in the splash zone of the shores of Vattern and Vanern lakes. On Gotland the history of the flightless Tropiphorus obtusus Bonsd. seems to have been the same (Lindroth, 1933, pp. 347-348). We can assume that most of ihe 11 presumed “cold relicts” have persisted in southern Sweden and southern Finland since the first immigration—chiefly during the subarctic period—even in today’s southernmost Fennoscandian habitats. Three of these species (Elaphrus lapponicus, Nebria gyllenhali, Pa- trobus septentrionis) are known from the late-glacial deposits in Denmark 687 604 and/or Skane and are therefore members—at least west of the Baltic Sea—of the original southern immigration group. B. Heat Relicts The assumption that these may occur in Fennoscandia is based on the view that after the last glaciation there was a period with a more favorable climate than at present, an idea which is based on such abundant and unambiguous observation material that it might be considered a fact (p. 673, and Table 36, p- 661). With the beginning of the Subatlanticum (about 500 B.C.) or shortly before that the plants and animals most susceptible to cold would have been compelled to retreat from the northern or high-altitude limits of their areas. Isolated populations would have been left behind in lococlimatically or micro- climatically thermally favored places outside their continuous area. As in the case of “cold relicts,” it is most practical to divide the species in question into geographical groups. Species whose assumed relict area is strikingly isolated are marked with an asterisk (*). a. In Trondelag (Province 27) *Trechus discus *Trechus micros b. Head of the Sogn (Province 19). Cf. p. 454 ? Carabus cancellatus Harpalus tardus Harpalus rubripes Metabletus truncatellus c. In southeastern Norway, chiefly around the Oslofjord *Abax ater Cicindela hybrida ?Amara montivaga ? Lebia cyanocephala ? Bembidion stephensi Licinus depressus d. Along the Swedish west coast ? Calosoma reticulatum *Dromius angustus e. In southeastern Sweden *Calosoma reticulatum Harpalus azureus Carabus intricatus and other “limestone species” (see pp. 112 ff., 289 ff.). f. In the central Swedish lake region. h—hygrophilous; x—xerophilous Agonum dorsale x *Demetrias imperialis h *A. lugens h ? D. monostigma h *Badister sodalis h *Harpalus anxius X *B. unipustulatus h *Fl. rufitarsis x *Brachynus crepitans X *Leistus rufomarginatus 689 605 Licinus depressus x Odacantha melanura h ? Microlestes maurus X *Oodes gracilis h M. minutulus x Panagaeus bipustulatus x. g. In lower Norrland Agonum obscurum ? Carabus arvensis h. At the northern end of the Gulf of Bothnia Agonum thoreyi Pterostichus vernalis (brachypterous form; ? Chlaenius nigricornis ? Trechus rivularis. Fig. 36, p. 378) *Dromius longiceps Panagaeus crux-major i. In the large northern Fennoscandian wooded region some markedly southern species become very rare and occur only at especially suitable places, which probably correspond with the well-known “southern hills” and “south- ern peaks” of the botanists (Andersson and Birger, 1912, pp. 52 ff.). A deter- mination of the isolation of these populations, as well as their actual relict character, must be left for a future more detailed inventory of these places. Attention should be directed to Harpalus latus, Notiophilus biguttatus, Pteros- tichus oblongopunctatus. It is of course impossible to “prove” that these advanced more or less isolated northern occurrences of southern species are in fact relicts. For this reason only certain especially clear cases will be analyzed here. The special situation of the inner Sogn with respect to climate was consid- ered earlier (p. 454) and the species occurring isolated in this area were listed. Naturally they are able to live in the comparatively favorable present-day cli- mate there. Of prime interest here is Metabletus truncatellus, because in Sogn it has been found exclusively in the brachypterous form (p. 454, footnote). In the present time it has not once been observed in the Regio betulina (p. 446), and it is unclear how it could have reached Sogn under the prevailing condi- tions of the plant regions of the surrounding fjelds. Nordhagen’s investigations (1933, pp. 207 ff.) in Sikkilsdal in the Jotunheimen showed, however, that in these regions the timber-line was much higher in the postglacial warm period probably chiefly during the Atlanticum. He concludes that in this period the fjeld passes formed an open route from the eastern valleys to the inner Sogn, even for cold-susceptible plants. Metabletus may also have arrived there in this way. In the case of the likewise flightless Carabus cancellatus the situation is less clear, unless it is found to be actually native to Sogn; so far only 1 specimen is available. The cited elevation of the timber-line in south-central Norway also enhanced the role of the passes leading from the Glomma and Trysil valleys into Trondelag (Province 27), which are still wooded, as faunistic and floristic migration routes. This may well explain the noticeable isolate of 690 606 the two ecologically sister species, Trechus discus and T. micros (p. 553) in the Trondheim region. Southeastern Norway is climatically highly favored, especially in sum- mer (Fig. 63, p. 452), and has a correspondingly rich fauna (p. 454). Some species occur there more or less isolated, most clearly Abax ater and Bem- bidion stephensi. However, it is very uncertain whether the occurrence of the latter species has a relict character. The first finds, not only in the Oslo region but also near Göteborg and in southeastern Finland, were made at a later date (in the first two cases in the same year, 1929). It is therefore possible, even considering the concealed mode of life of the insect, that a late im- migration is involved here. The situation is different in Abax ater, which is constantly flightless and has been found at two widely separated localities. Both are near the most extensive beech forest region of Norway (Hjelmgqvist, 1940, p. 7), which is certainly no coincidence. The total distribution of Abax ater (Borchert, 1938, Plate 40) largely corresponds with that of Fagus sylvat- ica forma typica (Hjelmqvist, 1940, Fig. 9). In Sweden it seems to live only in beech forest (unfortunately there are no ecological data from Norway). It is well known that Fagus earlier had a wider distribution in Scandinavia (Lindqvist, 1931, pp. 427 ff.), and the isolated southernmost stands, if sponta- neous, are io be considered as relicts. It is natural to suppose that Abax ater immigrated simultaneously with beech in a more humid period, and that the Norwegian occurrences are relicts. In Denmark (Sjaelland) the subfossil Abax was found in the “beech zone” (Henriksen, 1933, p. 140). At its completely isolated locality in Vgl, Leistus rufomarginatus may have had a similar history; of course this species is winged. But it is strange that according to Horion (1938, pp. 129, 137; 1941, p. 70) this species represents a late immigration in central and western Germany. Among the remaining possible relicts of the “Oslo region,” Licinus depres- sus is Of special interest, partly because it is constantly brachypterous and partly because it shows a similar isolate in east-central Sweden. Its relict character is fairly evident. However, I am perplexed about the interpretation of Amara montivaga, whose remarkable expansion in Scandinavia in recent decades was discussed earlier (p. 632). The possibility cannot be ruled out that the origi- nal nucleus of this expansion was a relict population in southeastern Norway, and that climatic improvement was responsible for the new area expansion (pp. 641 ff.). Dromius angustus is certainly a relict, not only on the Swedish west coast near Hil Saro, but generally at the Scandinavian record localities (perhaps with the exception of the newly discovered locality near Ska Halsingborg). It is striking that, with the exception of Bornholm, this species is completely missing from Denmark, so that the Scandinavian stock has no direct connec- tion with the south. The reason for the choice of the Saro Peninsula in western Sweden as a refuge is probably not mainly the climate. Only this place had 691 607 a continuous stand of primitive coniferous forest in the immediate vicinity of the sea (see maps in Malstrom, 1939). Dromius angustus in northern Eu- rope is namely associated with pine forest on sand or gravel close to the sea. Otherwise the species has generally a predominantly western distribution as well. It is to be assumed that it immigrated to Scandinavia during the atlantic period, during which Gotska Sandon was formed (p. 280). The question whether Calosoma reticulatum in western Sweden occurs as a relict (or generally as a resident member of the fauna) was discussed earlier (p. 455). The record on Ska is certainly of an accidental nature. Even more distinctly isolated is the area of Oland, where the species lives as a true inhabitant of the open Alvar.* It is therefore a pronounced xerophile, a relict from a warm and at the same time dry period. It might not be too bold to assign its immigration period to the Boreal, when the division of land and water in the southern Baltic Sea region was most favorable. Probably the main part of the “warm element” arrived at this time from Old and Gtl (p. 309). The most pronounced summerlike warm center in all Fennoscandia is the central Swedish lake region (Fig. 63, p. 452). There is no region quite like it, with so many undoubtedly heat-requiring species occurring isolated (p. 455). Let us inquire how far the immigration routes to this region are still open, and whether, and for which species, they were in an earlier era, i.e. let us identify the unambiguous relicts in the fauna of the central Swedish warm region. The isolation of populations there is very different in different species. “It is easy to imagine a series of distribution maps, beginning with Oodes helopioides, which shows a gap already in the south across Smaland, continuing through Odacantha melanura where this gap has grown, and direct contact westward across Lake Vanern has been lost. At the next step, represented by Agonum lugens, south of the central Swedish region there remain only sparse occurrences, on Oland—Gotland and Skane; and in the case of Reichenbachia impressa Panz., Psammoecus bipunctatus Fbr., Silis ruficollis Fbr., Oedemera croceicollis Gyll., and many others exclusively on Skane or Blekinge, the two southernmost provinces of Sweden. The next step brings us to Oodes gracilis and Demetrias imperialis. Somewhere in between is Stenus solutus Er., whose occurrence in Skäne is still uncertain. Nor has the little Euconnus rutilipennis been found so far in southern Sweden. “This series of distribution maps places an odd appearing type like Oodes gracilis and Demetrias imperialis in its organic context at a glance. They are only the most pronounced representatives of the general trend of our ther- mophilous fauna, the retreat after the postglacial warm period. And this an- swers those who, whenever they come across a zoogeographically strange find, say it must have been introduced by man. In the case of Oodes gracilis the *(Plant community consisting typically of mosses and calciphilous herbaceous plants that grow on steppelike shallow alkaline soils overlying Scandinavian limestones; suppl. scient. edit.). 692 693 608 question may be asked: Where from? And: Why Demetrias imperialis as well?” (Lindroth, 1943a, pp. 139-140). The examples selected above are all pronounced hygrophiles, which live on the shores of eutrophic bodies of water. Undoubtedly just these biotopes around the shallow lakes of central Sweden, which are well warmed at mid- summer, are lococlimatically and microclimatically favored. However, the list (“f’) above contains as many, more or less definite xerophiles, from which it is evident that in this part of Sweden other biotopes are favored, i.e. that the locociimate and microclimate are everywhere thermally “tainted” by the macroclimate, as is evident from the map in Fig. 63. Some of these xerophiles were considered earlier (pp. 113 ff.) as presumed “Jimestone species.” Unambiguous relicts are undoubtedly Harpalus anxius and H. rufitarsis. But also the other, less isolated areas, own a more or less dis- tinct “zone of extinction” south of central Sweden, which probably at present function as effective barriers against dispersal. Only Microlestes maurus (Vgl Kinnekulle) arouses doubt, since only one single macropterous specimen was found. It might not be too bold to consider all the species in list “f’ as relicts. Two of these species, a hygrophile (Badister sodalis) and a xerophile (Licinus), are consistently brachypterous and are therefore hardly capable of dispersal by fits and starts. Among species with their northern limit situated farther north, e.g. in lower Norrland, it is far more difficult to establish possible relicts, chiefly because these regions have not been so thoroughly explored that one can trust the gaps shown on the map. However, in one case, Agonum obscurum, this appears to be justified. This species has an unusually well-defined northern limit in Scandinavia, which in Sweden and southeastern Norway corresponds fairly exactly with the isotherms of the “critical months” (Lindroth, 1939a, p- 243), and should be climatically (thermally) determined. The record of seve- ral specimens in southern Hls near Skog, Sodra-Branningen is quite isolated. The gap to the south is undoubtedly real, since this covers the region tho- roughly explored by Palm (1942) on the lower reaches of the Dalalven River. From the entomologist’s point of view there are many suitable biotopes for Agonum obscurum in the boggy forests. With respect to the development of hind wings, the species is dimorphic, but near Hls Skog only brachypterous individuals were found. The places around the northern end of the Gulf of Bothnia, in both Swe- den and Finland, apparently comprises a definite relict region. These are ther- mally very favored in summer (p. 461; Fig. 63, p. 452). The occurrence here of the species listed above (under “h”) is in some cases very isolated, especially of Dromius longiceps. Even in a species like Agonum thoreyi, where the gap on the Finnish side seems insignificant, the subfossil records (Fig. 95, p. 675) show that earlier it enjoyed much wider distribution and was much more fre- quent. In Vbt and Nbt I methodically searched for just this species and was 609 able to confirm its rarity. The only locality where it occurred in large numbers, Nbt Räneä, apparently was strongly favored thermally, since several other pro- nounced southern species also live here (Lindroth and Palm, 1934, p. 17). Of the 7 species enumerated above, doubtful relicts in the region considered here are Chlaenius nigricornis and Trechus rivularis, the former on account of the pronounced general tendency of the Chlaenius species to migrate, the latter because it is a strictly stenotopic animal which easily escapes the attention of the collector. Moreover, in Finland there are no clear examples of heat relicts among the carabids. It is of course possible that Bembidion nigricorne and B. rufi- colle, perhaps also Amara infima and Tachys bistriatus, are such. But the two latter species have a concealed mode of life and may have been overlooked. Throughout the Fennoscandian area the species of Bembidion are perhaps to be considered as relicts from a warmer period. In the case of B. nigricorne this is evident from the map on dimorphism (Fig. 44, p. 390). That 2. ruficolle does not occur even in Sweden because of a late, more or less accidental west- ward advancement (as in western Germany; Horion, 1936; 1937, p. 11; 1941, pp. 123-124) is evident from the fact that the species, found before 1827, was recently rediscovered at its first record locality. The carabids cannot help us answer the question whether in southwestern Finland, especially at the Cape of Hango, there is a relict region for xerother- mal species. Krogerus (1932, p. 250) mentions some species of insects (but no Coleoptera) for which he assumes this, among them Sphingonotus coerulans, which even in Sweden is regarded as a heat relict (p. 455). Palmen (1944, p. 221) is undoubtedly correct when he argues that such isolated records in the Hango Peninsula may represent the result of late anemohydrochorous transport from Estonia. But I think he dismisses the relict problem somewhat superficially. A closer study of the flightless curculionid Scleropterus serratus Germ., which occurs isolated in southwestern Finland, would be very interes- ting in this connection. Har. Lindberg’s view (1942, 1943) that Amara crenata, 694 A. majuscula, and the curculionid Gronops inaequalis Boh. in the Skargard of southwestern Finland are to be considered as Ancylus relicts (“pseudorelicts”) is certainly wrong (see p. 622). Probably, the time of immigration of any of the “heat relicts” consi- dered here cannot be established more precisely. However, on the basis of the Finnish subfossil records the maximal distribution of Agonum thoreyi (p. 674) might fall in the sub-boreal, and the recent finds along the Gulf of Bothnia may be considered as relicts from this period. The markedly xerothermal re- licts (Calosoma reticulatum, many species of Harpalus, and others), which are so characteristic of Oland—Gotland, had probably already immigrated in the Boreal. During that period (Ancylus epoch) the distribution of land and water in the southern Baltic Sea region was especially favorable and some of the species concerned were flightless (pp. 298 ff.). However, among carabids there 695 610 are no such clear examples of xerothermal relicts as Sicista subtilis in central Scandinavia (according to Ekman, 1922, pp. 206 ff., probably a boreal relict) or Stipa pennata in Vgl. (according to Sernander, 1908, a sub-boreal relict). C. Coastal Relicts In Fennoscandia the shorelines were postglacially subjected to continuous, sometimes extensive alteration, chiefly in favor of the land, and the stenotopic riparian fauna has led an extremely unstable existence. The occurrence of relicts in former coastal regions would be easily understandable. Concerning the presumed coastal relicts of the flora, see Elsa Warburg (1910, p. 166). Occurrence along the shore can probably be thermally determined, as in the case of Nebria gylienhali (p. 685) in the negative direction, and in the case of Bembidion assimile probably in the positive direction. The former was therefore described as a “cold relict.” However, in general it is not possible to decide whether thermal, hygric, edaphic, or other factors are decisive for the more or less marked association of a species with the seashore, so that the neutral term “coastal relict” may be appropriate. The following species, whose main occurrence is on the seashore (or at least in coastal regions), also have inland localities—mostly at the larger lakes—which deserve to be investigated from the viewpoint of the relict hy- pothesis: Bembidion assimile Dromius linearis B. pallidipenne Dyschirius impunctipennis Cicindela maritima D. obscurus. It is first important to emphasize, that in none of these species is there any dependence on NaCl of the sea. In the case of Bembidion pallidipenne and the two species of Dyschirius “halophily” was assumed; but this is no doubt purely edaphic (due to the occurrence of suitable sand) and does not deserve any such name (p. 524). Two of the other species are also sand-dwelling insects, but Bembidion assimile is associated with loam. We are justified in arguing that these 6 species have achieved a largely coast-bound distribution, first due to edaphic factors, second due to the highly favored possibilities of immigration along a more or less uniform coastline. Climatic factors may also have been effective. Krogerus (1932, pp. 248, 252) analyzed some Finnish inland localities of Cicindela maritima (Kb Kontiolahti, Sb Kuopio) and Dyschirius obscurus (Kb Kontiolahti)—both stenotopic quicksand species—and provided a map illus- trating the striking position of these localities along the shoreline during the maximal transgression of the Ancylus Lake. There is no doubt of a casual con- nection here. In the case of Dyschirius obscurus, all the Finnish inland localities of records are situated below the highest postglacial shoreline (Magnusson 696 611 and Granlund, 1936, p. 179; Sauramo, 1942, p. 229) and can be considered as coastal relicts. Even in the case of the only Swedish inland locality, Hll Ves- sige, Sjonevadssjon (58.5 m above sea level), this was so at least during the Baltic Ice Lake period; however, it is uncertain whether the species had immi- grated that early. D. impunctipennis, occurring at Lake Ladoga, where the only enduring inland-lake populations in Fennoscandia are found, may likewise be considered a coastal relict. The other 4 species (including Cicinde/a) cannot be coastal relicts through- out northern Europe (including Denmark and the eastern Baltic region). They occur in postglacially constantly supraaquatic regions: Bembidion assimile near Sma Bolmen, in Denmark, and in the eastern Baltic region. B. paliidipenne on two lakes of Jutland (and in Holstein); Cicindela maritima in central and northern Norway; also in the eastern Baltic region; Dromius linearis in Denmark. However, in support of the contention that these species too are partly coastal relicts, the following facts may be added: Bembidion assimile. On the large central Swedish lakes this species also occurs in the brachypterous form—which even predominates. In this form it is not easy to migrate by fits and starts (cf. p. 393). On the other hand this is also true for Bolmen in Sma, a lake which never had a postglacial connection with the open sea. B. pallidipenne. All of the three Swedish inland records (in Ska) lie below the highest shoreline (see Munthe, 1940, Plate II). Cicindela maritima. Even the localities in the northern Finland lie below or at least in the immediate vicinity of the highest shoreline; in Sweden all of the localities are situated below it. Dromius linearis. The only two Swedish localities, situated at a consider- able distance from the sea, are on the shores of lakes Vanern and Vattern. Near Ogl Motala, the species is constantly endemic, here as far as is known, only in the brachypterous form, as also near Vgl Kinnekulle (1 specimen). D. “Anti-culture Relicts” For lack of a better name we here so designate species that seem to have been so badly affected by human culture that they remain as relicts only in part of our region virtually untouched by man. As possible examples, Agonum bogemanni, Amara nigricornis and Harpalus nigritarsis were mentioned above (p- 636) as “decreasing species.” Perhaps the extremely isolated Norwegian locality of Agonum mannerheimi (12 Vardal) is to be similarly interpreted; it is a definite “species of virgin forest.” Agonum bogemanni and Harpalus nigritarsis, probably also Trachypachys zetterstedti, seem to be in process of be- 697 612 coming extinct throughout the Fennoscandian region. It is uncertain whether the restriction of the small, homogeneous, isolated area of Agonum longiventre to the lower reaches of the Dalalven River is partly due to human culture as well (cf. p. 720). I am not in a position to decide whether in any of these cases extinction from “internal causes,” such as “the degeneration of an old species” may be involved. E. Interglacial Relicts Included in this category are animals and plants that lived in Fennoscandia during the last interglacial period and were not completely expelled by the subsequent glaciation (Wurm), but displaced in their distribution—chiefly to- ward the Scandinavian west coast. The best-known example in our fauna is the fjeld lemming (Lemmus lemmus: Ekman, 1922, pp. 397 ff.). Some of the (established or presumed) interglacial relicts have again ex- panded their area in the postglacial period to the extent that their relict cha- racter has been lost, especially when they have merged with the stocks that immigrated from outside the Fennoscandian region. In others, the spatial sepa- ration of the Fennoscandian interglacial stock is still sharply distinct. This phenomenon cannot be categorized with the “cold relicts.” All of the species in question have rather the opposite, no pronounced need of cold and they were not favored, but adversely affected, by the enduring glacial period. Basic to the formulation of our problem is the question how far the present-day Fennoscandian fauna originated from interglacial relicts. Hence the last section of this book is devoted to this question. Perhaps the abiding impression of this section is that I have handled the relict concept somewhat casually, since here the category covers populations not clearly isolated on the map. However, the most important characteristic of a relict is not the space factor but the time factor. 698 699 The Postglacıal Immigration The two preceding sections are designed as useful preliminary exercises for des- cribing postglacial faunal history. The fossil records ought to teach us that no definite conclusions as to the prehistoric areas ofspecies can be drawn from the present-day distribution alöne. The relicts primarily show that the postglacial “migrations” of organisms do not represent any methodical striving in def- inite directions, but pulsating, intermittent area displacements, which—even if verifiable—often remain unexplained. With all their commendable clarity the maps on the recent and prehistoric distribution of Hippophaé rhamnoides (Sandegren, 1943) show how impossible it is in some cases to give a reliable judgment on the postglacial history of an organism based on the present-day area of the species. This is of course an extreme case, the example of an especially noncompetitive plant, which has no counterpart at all among the animals considered in the present contribution. However, the warning reminds us to exercise caution in cases where the recent picture presented by the map is the only available basis of historical discussion and conclusions. An essential precondition for understanding the postglacial history of the fauna and flora of Fennoscandia is a knowledge of nonbiological features, especially within two branches of study: the late Quaternary climate (including the condition of the terrestrial ice) and the distribution of land and water dur- ing different periods, some of which go far back in time. No attempt is made here to provide a cogent glacial and postglacial history of Fennoscandia in re- spect of these. The information in these fields, necessary for an understanding of the faunal development, is given below in the relevant context. However, Table 36 (p. 661) gives a synoptic representation of the development of the late- and postglacial climate. The concurrent changes in the distribution of land and water are illustrated here in four maps taken from Granlund (Mag- nusson and Granlund, 1936) (Figs. 97-100). Right away it must be conceded that the authors are not at all unanimous about the climatic development or the development of different stages of the Baltic Sea. As a layman, with no primary knowledge in these fields confronted with strikingly divergent opinions, it was sometimes very difficult for me to decide which view was the more plausible. And the decision was perhaps too 614 Fig. 97. Baltic Ice Lake (hatched). According to Granlund (1936). often based on the democratic but superficial principle of the majority. As an example of biogeographically important details of Granlund’s maps that are otherwise conceived by other researchers, mention may be made of the following: Munthe (1940, e.g. Plate III) assumes that occasionally salt water forced its way through the Danish straits into the Blatic Ice Lake (also considered probably by Ekman, 1930, p. 239). His view (e.g. Plate XI)— shared 700 by some other authors—that Oland was firmly connected with Smaland du- ring part of the Ancylus period, does not find expression in Granlund’s map. Sauramo (1942) shows Bornholm as completely submarine both during the time of the Baltic Ice Lake (p. 227) and during the Ancylus period (which conflicts with his own isobases). For the Littorina period (pp. 239, 241) he does not assume any open connection between Lake Ladoga and the Gulf of Finland, and substantiates his concept with a shore curve of Ik Ino (p. 242). Concerning the question whether a postglacial sea connection existed between the Gulf of Finland and the White Sea, see below (p. 730). Supra-aquatic land between Öland-Gotland and Pomerania is assigned by Granlund to the Ancylus period, by Munthe (l.c.) to part of the Baltic Ice Lake period and the early Littorina period as well. This question has been considered above (p. 308), where a firm land connection of the two above-mentioned islands southward was considered so to speak a biological necessity. It is a matter of prime importance how far biogeographers are justified 701 702 703 615 Fig. 98. The Yoldia Sea. According to Granlund (1936). in passing judgment on prehistoric land connections, climatic conditions, etc., which have not been clarified by the specialists concerned, perhaps not even discussed. Tanner (1937, pp. 101, 107) thinks that, for instance, the “problem of hibernation” of the flora and fauna of Fennoscandia and other regions must be solved by geologists, and attaches little value to conclusions arrived at purely biogeographically. Typical of his approach is the discussion on the history of the flora of Labrador (1944). With respect to numerous facts cited especially by Fernland in favor of a hibernation of plants in Labrador refuges during the last glaciation, he denies this possibility, “as the whole of Labrador is presumed to have been covered with ice during the Wisconsin [=Wurm] glaciation” (p. 356). He of course earlier (pp. 174-175) conceded that so far it had not been possible to divide the glacial deposits of Labrador into different glacial epochs; all of them—“for logical reasons”—are included in the last glaciation! The ideal researcher would of course be one competent to work in nature both biologically and geologically; Nordhagen comes very close to this ideal. However, it is unreasonable when the geologists deny the right of the biogeo- graphers to interpret the biota history purely on the basis of biological material. They cannot deny that biogeographical conclusions, for instance, with respect to the glacial conditions in the western Scandinavian coastland, can provide and indeed have provided valuable impetus to geological research, as well. On 616 701 Fig. 99. The Ancylus Lake (hatched). According to Granlund (1936). the other hand it is important that the biogeographer familiarizes himself with current geologicai (paleoclimatological, etc.) ideas and—within reason—feels bound by them. He must be able to cite an entire series of facts before ques- tioning the relevant views of another special science. The proclamation of land connections and other geomorphological upheavals on the basis of the distribution of a single species of animal (for example, Ihering, 1927, p. 220) is quite wrong and only discredits the whole of biogeographical science. The position of Fennoscandia on the northwestern corner of a continent 617 702 Fig. 100. Littorina Sea at the time of maximum extension. According to Granlund (1936). shows that its postglacial recolonization was possible from two main directions, south and east. Therefore and on account of the unusually effective destruction of biota even by the last glaciation (Würm), Fennoscandia offers an especially suitable field of research on the most recent history of fauna and flora which has few counterparts anywhere in the world. The problem is complicated by the greater or lesser importance attached to the possibility of “hibernation” of the biota during the Würm period—or even during earlier glaciations; and 704 705 618 of course even these problems in our region are far simpler than, for instance, in the central European mountains. An account of faunal history is therefore logically divisible into three groups of species: I. The southern postglacial immigrants; Il. The eastern postglacial immigrants; II. The hibernation group, which is predominantly western. Doubts may arise in certain cases whether an immigration group is to be considered as eastern or southern. I have preferred to proceed so to speak nationalistically, from Sweden. I call all manifestations of immigration from the other side of the Baltic Sea “eastern,” even the group that has invaded Finland directly from Estonia across the Gulf of Finland, which, seen from Fin- land, is “southern.” Similarly, species that have reached southernmost Norway (southwestern Norway) directly across the sea are to be considered as “wes- tern.” However, no special section is devoted to them, since this involves the important and difficult problem of separating them from the “hibernators.” This “western” group, whatever its assumed origin, is therefore treated in the concluding section devoted to the question of hibernation. I. THE SOUTHERN IMMIGRANTS The first precondition for correctly visualizing the postglacial (s./.) immigration of plants and animals from the south—from Central Europe—is to ascertain the initial stage, the faunistic character of Europe south of the Nordic ice margin during the Wurm period. It is well known that the uninterrupted ice cover during the last glaciation (Wurm) was much less extensive than during the great glaciation preceding it (Riss) (Figs. 101, 102). The difference was greatest in the west and, especially in the east, whereas just south of Sweden, in northern Germany, the distance between the southern edges of the Nordic inland ice during the Wurm and during the Riss scarcely exceeded 150 km. Thanks primarily to the studies by Henriksen (1933) on the subfossil fauna of Denmark and Skane we are well informed about the coleopteran fauna, which during the Wurm Ice Age lived in the immediate vicinity of the south- ern ice edge and hence represented the first immigrated “southern” element in Scandinavia. There are 15 species of carabid from the “late glacial” deposits of these regions (see list of fossil species above, pp. 665 ff.). Of these, 10 live in the present-day Regio alpina, one has apparently become extinct (Nebria fos- silis), only 4 are unknown above—and north—of the Regio coniferina (Agonum dolens, A. viduum, Bembidion repandum, Chlaenius costulatus; Agonum viduum, which was not studied, by Henriksen, is probably doubtful). The high Nordic character of this fauna is best shown by the fact that of the meager 5 carabid species actually native to the Fennoscandian higher Regio alpina, only one 706 705 619 species (Nebria nivalis) is missing from the material (species represented are: Amara alpina, Nebria gyllenhali, Notiophilus aquaticus, Patrobus septentrionis; see Table 30, pp. 440 ff.). The same fauna undoubtedly accompanied the southern edge of the maxi- mal Wurm ice, even in Central Europe. Various subfossil “glacial faunas” are actually known from these regions (Nathorst, 1894; Lomnicki, 1894, 1914; Schille, 1916), but it does not seem possible at present to decide whether they actually lived contemporaneously with the Jast glaciation. In any case, here we are interested in the question in how near to the Wurm ice lived a richer fauna that included such species too which in present-day Fennoscandia do not normally cross the timber-line. In this respect the entomological subfossils seem to leave us in the lurch, and we must have recourse to a more reliable science, the pollen analysis. There are of course few precisely dated samples from Central and Southern Europe, but these give an approximate idea of the pattern of vegetation there at the time of the Wurm maximum (Fig. 103). From these facts, Firbas (1939) concludes that during the Wurm maximum there was only a local Betula-Pinus lide Dem, x PFrennnn Ser) ER, u jes < 2 Y Yoßerlin a , “uu” ae Fig. 101. Maximum ice cover during the penultimate (Riss) and the last, lesser (Wurm) glaciation. Local glaciations in Central European mountains are generalized and indicated only for the Riss. According to Granlund (1936). 620 706 Fig. 102. Last glacial period (Würm) in northern Europe: Position of ice edge during some phases of recession indicated. According to Granlund (1936). 707 flora—corresponding to the present-day Regio betulina and Regio coniferina of Fennoscandia. It was confined only very locally mainly to the west, in the ice- free belt between the Nordic and Alps ice, becoming richer only south of about 48°N. Forests with “heat-requiring trees” (corresponding to the Regio quercina and Regio fagina) seem to have “hibernated” only in the Mediterranean region. These considerable distances in north-south direction between the plant regions of the Wurm period were apparently made up relatively quickly after the melting phase set in. At any rate the forest-forming Berula and Pinus were already in Skäne at the end of the Baltic Ice Lake period (v. Post, 1933, p. 58; T. Nilsson, 1935, pp. 468 ff.; Magnusson, 1936, p. 237), when the ice edge was 708 621 IM i: | N il Fig. 103. European vegetation zones during the last glacial period (Wurm). Black points: glacial flora without trees. Blank points: Betula and Pinus. Crosses: “heat-requiring trees.” According to Firbas (1939). still present in northern Gotland, and “true” deciduous trees already existed at the end of the Yoldia period. Despite the much greater distances from the assumed hibernation regions, in their migration northward to Scandinavia the more heat-requiring elements were thus not far behind the species of the tundra (Regio alpina). However, the first southern immigrants were tundra species. In classifying the Fennoscandian fauna in its larger historical context it is of great interest to establish whether the species of this first immigration group originated from a true “mixed fauna,” (“Mischfauna”)!, which arose in the ice-free Cen- tral European belt during the Wurm period as a conglomerate of Nordic and Central European interglacial montane faunas, or whether the Scandinavian or the alpine fauna emanating from the Alps (and other mountainous re- gions) during the Wurm period, each stayed chiefly along “its” local ice edge. I have already considered the problem (Lindroth, 1935a, p. 626) and cited the following facts. t (Suppl. scient. edit.). 709 622 The late-glacial coleopteran fauna in Denmark-Skäne (according to Hen- riksen, 1933) includes 6 or 7 Nordic species, which are absent from Central Europe today: Amara alpina Colymbetes dolabratus Payk. Bembidion hasti Thanatophilus lapponicus Hbst. Elaphrus lapponicus T. trituberculatus Kirby Agabus serricornis Payk. (= Silpha baicalica Mtsch.). (determination uncertain) Bembidion repandum, which is known in Central Europe only by one spe- cimen from Jutland, should also be included here. The Nordic origin of this group is now more evident since no fewer than 4 of the 7 species (the three carabids plus Colymbetes dolabratus) were found in interglacial deposits in Sweden (Lindroth, 1948a)*. In contrast with this group there are (in Henriksen’s material) only two species from the Central European mountains, which are alien to present-day Fennoscandian fauna: Bembidion ? glaciale Heer Simplocaria deubeli Ganglb. Of these, Bembidion is definitely high-alpine (for example, Heberdey and Meixner, 1933, p. 70), whereas Simplocaria is an animal of subalpine forests (Ganglbauer, 1904, p. 61). Hence from the deposits along the southern edge of the nordic Würm ice we know at most one true “alpine”** species (Bembidion ? glaciale), today missing from Fennoscandia, as against the 6 or 7 nordic species, unknown in Central Europe. And indeed the alpine fauna of the Alps is incomparably richer than that of Fennoscandia! “It is tempting to explicate these facts in the following way. The nor- thern Wurm-ice did not extend southward far enough (viz. the ice of the Alps not northward) to make a ‘Mischfauna’ of northern and southern ele- ments (especially not of their most cold-requesting forms) possible. The alpine and subalpine species therefore, after the Wurm ice had begun to retreat, re- turned mainly the same way by which they had come, to the north or to the south, an exchange taking place to little extent only” (Lindroth, 1935a, p. 627). According to the map by Firbas (1939) (Fig. 103) it cannot be ruled out that a narrow zone of Betula-Pinus forests even at the Wurm maximum represented a kind of barrier for true alpine organisms. The great faunal and floral exchange *This is also an important proof that the occurrence of these species in late-glacial deposits cannot be explained by immigration from the east (from Russia or Siberia) along the Würm ice edge. On the other hand this could be true of Bembidion repandum and, Thanatophilus trituber- culatus, and probably also of Chlaenius costulatus. **The term “alpine” here and in the subsequent discussions indicates an inhabitant of the Regio alpina and does not carry any geographical implication. 710 zul 623 that led to the emergence of the boreo-alpine type of distribution may have occurred by all means not during the Würm but during the Riss Ice Age (as already assumed by Brundin, 1934, p. 172). Let us now consider more closely two questions concerning the southern postglacial immigration to Scandinavia: a. To what extent did the first immigrating alpine and subalpine species manage to reach the Fennoscandian mountains, to become resident there? b. Whatever their climatic requirements and time of their first arrival—how far north have members of the group that immigrated from the south reached? Or: How large a percentage of the area of today’s Fennoscandian fauna is ac- counted for southern immigrants which arrived through Sweden? a. During the initial stage of the melting of ice the possibilities of immig- ration to Skane through Denmark were highly favorable for alpine—eventually also for subalpine—plants and animals, not only on account of the nearly or completely uninterrupted land connection between these regions during the Baltic Ice Lake period but also because the ice retreated here at only moderate speed (about 50 m per year; Granlund, 1936, p. 235), so that fairly stable biotopes were formed. In line with this the localities of subfossil alpine plants in the southern half of Skane are more numerous than anywhere else in Sweden below the fjelds (Granlund, 1936, p. 239). Postglacial subfossil finds of alpine or subalpine insects in Sweden have solely been made in Skane. The extremely sparse subfossil finds of true alpine organisms in Sweden north of Skane have been interpreted by several authors as an indication that in these regions the alpine fauna and flora apparently encountered various bar- riers, blocking or impeding their dispersal northward (Ekman, 1922, p. 374; Brundin, 1934; p. 196; Lindroth, 1935a, p. 625; Nannfeldt, 1935, p. 77; 1947, pp. 75, 77). At first the melting of ice proceeded relatively slowly, which in- dicates only moderate climatic improvement on the average; it took the ice edge more than 4000 years to move from southern Skane to Vgl Billingen (see Table 36, p. 661), so that on the average a strip of just 80 m of land was freed each year. However, hereby the ascertained fluctuations of the ice edge in Skane (for example T. Nilsson, 1935, pp. 471 ff.), are taken into consideration and it is fairly certain that the ice melted faster north of Skane (up to Billin- gen). So we are already led to surmise, especially for the much faster melting phase of the central Swedish moraines (and the Salpausselka of the same age in Finland) up to the point of bipartition in JtI—1.e. the finiglacial period (in De Geer’s sense)—that the forest immediately succeeded the retreating ice, and no space was left for a real alpine vegetation. This view easily leads to mistakes. Even if for climatic reasons the Berula-Pinus forest was able to follow up the ice edge, the receding ice must have been accompanied by a forest-free zone—if only because of the slow growth of young trees; the faster the ice retreated, the broader this zone. According to the above (minimal) shift of the ice edge of southern Sweden, 80 m per year and on the assumption that WZ 624 saplings must be at least 3-4 years old to exercise any influence on the other vegetation, we get a 200-300 m “forest-free” zone along the ice edge. In the Stockholm region, where an average early ice recession of 250 m per year has been established (De Geer, 1940, p. 97), this zone attained a width of almost 1 km. It is somewhat paradoxical: the faster the ice receded, the larger the woodless area available for alpine plants. They were evidently endangered by the forest in a different way: its advancement at the same speed as the ice receded imposed on the organisms dependent on open biotopes a constant, equally fast displacement, which badly affected perennial plants with poor ca- pability of dispersal (like most of the fjeld plants). For the same reason it is conceivable that the high temperature of the macroclimate, already evident during the “Subarctic” period, had this effect on the coldrequiring plants that found microclimatically suitable places for germination close to the ice; but in subsequent years these places had already changed so unfavorably that the perennial plants never reached the stage of bearing fruit. However, this holds only to a very limited extent for actively mobile animals, as well as for annual and biennial plants. I therefore believe that the absence of alpine-subalpine subfossils from central and northern Sweden outside the fjeld regions, and the slight importance we need attach to the “southern” immigration route of such organisms, are not due only to the rapid melting of ice. The standstill period of the ice shortly before and during the Yoldia epoch may have been far more fateful for the alpine fauna and flora. With regard to the duration and the exact conditions of this stage—especially with regard to the isolation of the Baltic Ice Lake—there are of course widely divergent views (Lundqvist, 1946, pp. 281 ff.). But it is clear that the retreat of the nordic inland ice at the latitude of the Billingen mountain in Vgl (and simultaneously near Salpausselka) was interrupted for a period whose duration has been estimated by various authors at 650-800 years (Granlund, 1936, p. 155; Munthe, 1940, pp. 63-64). At first thought one would be inclined to take such a standstill period of the ice as advantageous for alpine plants, during which constantly cold biotopes were formed along the ice edge. But it looks as if such a stage did not arise—at least not exclusively—as a result of a decline in temperature, but as a result of increased winter precipitation (Hyyppa, 1933, pp. 29 ff.; 1936, pp. 446 ff., 458; Sauramo, 1942, p. 281). A general oceanization of the climate could also explain the apparently simultaneous increase of birch at the cost of pine in Central Europe (Firbas, 1939, p. 87). It can easily be visualized that, if this view is correct, the forest in central Sweden would have quite caught up with the ice edge. Of course these questions on climate are not to be considered as de- cided. It is quite possible that the so-called late Dryas period (for example in Thomasson, 1935, p. 615) coincides with the standstill of the ice edge. Still more destructive may have been the effect of the altitude on the alpine organisms that had immigrated from the south. The ice edge was stationary 713 625 for a long time over the Billingen mountain, but as soon as it retreated* from its northern edge, an open connection appeared between the Baltic Sea and the ocean. The Yoldia period set in, and on account of the depressed position of the central Swedish plain, the ice edge lay for 400-500 years milking in the broad Närke Strait, which in the beginning contained only a few, smaller islands. Even if the distances across the Narke Strait were not insurmountable for plant diaspores and flying animals, there was almost no inhabitable land at all on the southern edge of the inland ice. The alpine flora and fauna—to the extent that they had generally followed the ice—remained “Stationary on the southern shore of the Narke Strait,” slow victims of the warmer climate and the advancing forest. It is thus no accident that the northernmost Swedish record of subfossil alpine plants (Arctostaphylos alpina, Dryas octopetala) was made outside the fjelds near Nke Laxa (G. Andersson, 1906, p. 60; v. Post, 1909, p. 694), in the immediate vicinity of some small supra-aquatic regions which emerged as islands in the early stage of the Narke Strait (see map, Plate II, in Munthe, 1940). Ekman (1922, p. 412) thinks that birds, flying insects and possibly small animals with especially strong passive capability of dispersal belonging to the “purely arctic” fauna may have been able to migrate along the Swedish west coast in the late-glacial period and, following the ice edge in southeastern Nor- way, reached the present-day Regio alpina. Wahlgren (1913, p. 143) originally held a similar view with regard to certain alpine Lepidoptera (but he had in mind immigration to Norway from the southwest across the sea). Later (1919, p- 25), however, he classified this element as Wurm-hibernating. Actually the possibilities of a migration or passive transport of alpine organisms during the critical period (at the end of the Baltic Ice Lake period, the beginning of the Yoldia period) across the sea from the Danish-southern Swedish mainland to southeastern Norway were still poorer than across the Narke Strait. The distance was greater and the higher salt content of the sea militated against hydrochorous or anemohydrochorous transport (p. 600). We thus conclude, that the pronounced alpine-subalpine fauna, which demonstrably lived along the southern edge of the Wurm ice at the time of its maximum, and in Denmark-Skane formed the first colonizers, at the most reached the southern shore of the Narke Strait. Relicts from this period, of those mentioned above (p. 682), may be at least Prerostichus adstrictus in Sma and Nebria gyllenhali and Patrobus assimilis on Gtl, the first of these also on the Vatter Lake; from all these regions there are subfossil finds of true alpine plants. Likewise Pinguicula alpina and Bartsia in Gtl, Viscaria alpina and Poa alpina in Old, etc., have generally been considered as late-glacial *It is still contested whether this “isolation” (“Zapfung”)' of the Baltic Ice Lake was at once definitive, or whether it was repeated by oscillations of the ice margin (Lundqvist, 1946, p. 282). *(Suppl. scient. edit.). 714 715 626 relicts. I find this explanation also applicable to the occurrence of Alchemilla alpina, Rhodiola rosea and Viscaria alpina in western Sweden (even if they are “pseudorelicts”; cf. Nannfeldt, 1935, p. 77). The occasional suggestion that they wandered down from Norway during a later cold period (for example Sernander, 1894, p. 200) is most unlikely in light of the southern Swedish inland finds of Alchemilla alpina (map in Lid and Zachau, 1929, p. 97). But none of the above-mentioned species of beetles and plants is markedly alpine or even alpine-subalpine (i.e. with the normal lower limit along the coniferous timberline), so they do not show that the most cold-requiring forms of the late-glacial deposits from Denmark-Skane reached just as far north. We will return (p. 765) to the question whether the alpine and other species were able to immigrate to southern Norway directly across the sea. b. It is often easy to determine how far north the stock that immigrated across southern Sweden has advanced if it is geographically (and function- ally) distinct from the almost obligatory stock on the other side of the Baltic Sea and there is no reason to assume a western (or northern) interglacial element. A series of examples of such species was cited* above (p. 618) and their collective northern limits were cartographically represented (Fig. 82) groupwise—according to the position of this limit in Sweden. Evidently, at least a limited number of species reached the provinces of Nbt and Vbt by immigration across southern Sweden. However, where the Swedish area extends up to the Finnish border on the River Torne-älv, the origin is more uncertain. Practically without exception the area then continues uninterrupted to the other side of the border of the kingdom right across Finland. Simultaneous immigration from the south took place on both sides of the Gulf of Bothnia. The two stocks have merged. Is it possible to decide in such cases whether this happened on the Swedish side or on the Finnish side? In some cases such a reconstruction is in fact possible. We proceed from species in which the Swedish stock and the Finnish stock have still not met in order to ascertain the position of the gap between them. In most cases neither of the two stocks has reached the northern end of the Gulf of Bothnia. Then the northern limit is usually highest in Finland. Doubtless, the situation is often climatically determined and is to be explained (p. 459) by the average higher summer temperature at a given latitude in the inland of Finland. In a series of cases mentioned below (p. 734) the Finnish stock has, how- _ ever, reached the northern end of the Gulf of Bothnia and crossed the border of the kingdom, which the Swedish stock has failed to do, so that on the Swedish side there is a more or less broad gap. This situation can in no way be attributed to present-day climatic conditions. *The selection above (p. 618) was not made from the viewpoint of the immigration route. There some species, whose exclusively southern origin is uncertain, were included among the species that advanced farthest north (as far as Nbt). In the latter context the following “Nbt species” are probably u..ambiguous: Agonum versutum, Amara ingenua, Cicindela campestris. 716 627 The question arises whether—in the name of justification—there is a con- verse group whose members have advanced farther north in Sweden, i.e. have encroached into the Finnish region, and are separated by a more or less dis- tinct gap from the southern Finnish stock. We find at the most 4 species of which this might be true: Agonum micans Carabus violaceus A. versutum Synuchus nivalis. Further analysis reveals: Agonum micans is a very late immigrant (p. 632, Fig. 87) in both countries—at any rate in the north—and the map is to be considered incomplete. If we must explain this it seems more probable that the stock arrived at the northern end of the Gulf of Bothnia from Finland (through Kuusamo), and not the converse. Agonum versutum has just one locality in Finland (Ob Ylitornio) along the River Torne, which is connected with the Swedish area. The encroachment into the Finnish region is thus as meager as possible. Carabus violaceus. That the Finnish population north of latitude 64°N arrived from the west seems very probable from the map. Moreover, in north- ern Finland the “arcticus form” might occur without smooth transition to the southern type (as in Scandinavia) (Hellén, 1934, p. 45). Yet this species does not fit our problem as an example. It undoubtedly belongs to the Wurm hibernators on the Norwegian coast, and the northern Finnish stock certainly did by no means come from southern Sweden. Synuchus nivalis. The locality Ob Kemi, very isolated from the rest of the Finnish area, historically belongs to Sweden beyond doubt. AS as counterpart to the above-mentioned 17 species (p. 734) emanating from the east there seem to be only 2 species (Agonum versutum, Synchus) that reached the northern end of the Gulf of Bothnia apparently by moving up through Sweden, including the nearest part of Finland. In species with wing dimorphism it is often possible, even after two im- migrant stocks merged, to determine the “cicatrice” as a zone with strikingly numerous macropterae. In this way we found above that in Bembidion guttula (p. 387 and Fig. 42), Carabus clathratus (p. 381 and Fig. 38), Pterostichus lepidus (p. 381 and Fig. 37), P. minor (p. 387 and Fig. 43) and P. strenuus (p. 395 and Fig. 47) only the Finnish stock reached the northern end of the Gulf of Bothnia. The only conceivable—but uncertain—contrary case would be Bembidion lam- Pros (p. 382 and Fig. 40), of which it can probably be argued that it also invaded the southern fjeld regions of Sweden by direct immigration from the south. The above discussion therefore confirmed the view earlier expressed (Lin- droth and Palm, 1934, p. 127) that the fauna of the upper Swedish north coast acquired its character mainly by the northeastern immigration group. The un- ambiguous southern immigrants—even among the remaining Coleoptera—are A 628 astonishingly few in these regions. There are corresponding conditions for the flora: “the number of species around the Gulf of Bothnia that have come from Sweden is very small” (Cajander, 1921, p. 6). One may justifiably ask whether on the rugged western Scandinavian coastal region, in Norway, the possibilities for the advancement of a stock that immigrated from the extreme south across Skane have not been still lower, and whether it can be supposed that such immigrants colonized regions north of Trondelag in the postglacial period (Provinces 26-28). However, a verification does not quite draw blank. The following species, which in Norway reach at least latitude 64° N, must be considered as immi- grants across southern Sweden: Agonum assimile Bembidion dentellum A. mülleri Dromius fenestratus A. viduum Harpalus aeneus Amara ingenua H. pubescens. Probably also to be included is Amara familiaris, which in Norway has advanced even beyond latitude 69° N. So has Calathus erratus and the map on dimorphism (Fig. 35, p. 376) clearly shows its southern origin. The migrations were greatly facilitated by the passes between the Oslo and Trondheim re- gions, which were much more favorably situated in the warm period. Namely the flying species could have reached the Norwegian coast from Sweden di- rectly via the northern passes, as was mentioned earlier (pp. 614 ff.). But the small number of these migrants is remarkable. In the Norwegian fauna north of Trondelag the component of the species that immigrated across southern Sweden is of very little importance. It needs to be emphasized that the species that immigrated from the south across Denmark-Skane did not always advance on a broad front along with the northward moving northern boundary of the west-east limit running across southern Sweden. At the beginning even more than today, the southern Swedish highland represented a climatic obstacle. The edaphically fastidious species, for example, loam-bound species, were excluded thereof. Because there is little loam along the east coast of Smaland (Fig. 78, p. 511) for many species it was possible to bypass the highland only in the west. The characteristic animals of the central Swedish warm and loam region immigrated chiefly from the southwest, across the plain of Vgl. This feature is evident from the dimor- phic map of Bembidion assimile (Fig. 46, p. 394) and in B. aeneum (Fig. 49, p- 400; Fig. 80, p. 521). Other examples are Agonum marginatum, A. moestum, A. thoreyi, Bembidion articulatum, B. illigeri, Dyschirius lüdersi. It is quite pos- sible that the isolates interpreted as relicts of a number of species, chiefly in the Malar lake region (p. 691), originated, at least partially, by immigration from the same direction. 718 629 Finally it must be remembered that “southern” immigration into Scan- dinavia also took place by routes other than Denmark-Skane, partly in the southwest across the sea to Norway (see p. 765), partly, and chiefly, to the large Baltic islands. In the latter case the assumption of an early postglacial land connection with northern Germany seems to be necessary (pp. 308 ff.). It is very difficult to judge how far this immigration route also had a role in colonization of the Swedish mainland. Il. THE EASTERN IMMIGRANTS The eastern marginal regions of Fennoscandia, compared with the correspond- ing regions of Scandinavia, were free from Wurm ice astonishingly early. Al- ready at the end of the Baltic Ice Lake period, when the ice edge coincided with the central Swedish moraines and Salpausselka*, the whole of eastern Karelia, the Kola Peninsula and the adjacent parts of Finland were ice-free (Sauramo, 1942, map on p. 227). Immigration of plants and animals on a broad front could have already set in at this time. A purely southern immigration—from the eastern Baltic region across the present-day Gulf of Finland—could have had a role only during the latter part of the Yoldia period, since southern Finland was earlier almost completely under water. Much as in Sweden, subfossil remains of a pronounced alpine flora and fauna have been found only in the extreme south, in the Isthmus of Karelia and in the adjacent parts of Russia (Har. Lindberg, 1916, p. 3; Hyyppa, 1933, pp. 10 ff.), including the carabid Prerostichus vermiculosus (p. 672), which is now distributed westward only as far as Pechora. It must be assumed—as for Sweden (p. 710)—that in Finland the real alpine-subalpine organisms did not reach the fjelds from the south (southeast). Cajander’s surmise (1921, p. 5) that this would have been possible along the supra-aquatic Maanselka has remained unconfirmed. Even in northern Finland the inland ice has apparently been draining directly into the sea or into reservoirs (Sauramo, 1942, pp. 227, 228), so a terrestrial Regio alpina could scarcely come into existence in the immediate vicinity of the ice. The various immigration routes from the south and the east to Finland (and the rest of eastern Fennoscandia)**—with the exception of the north- ernmost one—are not sharply separated from one another. However, the fol- lowing may be appropriately left out of consideration: 1. From the south across the Gulf of Finland. 2. From the southeast across the Isthmus of Karelia, between lakes Ladoga and Onega, or between the latter and the White Sea. cory *Salpausselkä” is a Finnish word which means “end moraine,” but is more correctly described as a “recessional moraine” —General Editor. ** Aland and the rest of the southwestern Finnish Skärgärd are left out here, since a separate section (pp. 236 ff.) has been devoted to them. 719 720 630 3. From the east, especially through Kuusamo and Salla. 4. From the northeast through Kanin to the Kola Peninsula. 1. The “Baltic immigrants” (Krogerus, 1925c) play an important role in . southwestern Finland, which was first shown by Eklund (1931, pp. 88 ff.) with respect to the flora. The importance of this phenomenon for the coleopteran fauna has been elucidated by Palmen (1944 pp. 206 ff.). Postglacially the alti- tudinal conditions in the region of the Gulf of Finland have never been more favorable than today. The large size of the Baltic immigration group is to be ascribed primarily to the unusually favorable situation for anemohydrochorous dispersal in the estuarine region of the Gulf (Palmen, 1944), as discussed above (p. 604), where a list of the clearest examples among the carabids is provided. This immigration route is therefore utilized predominantly by winged insects. However, two constantly flightless species of Carabus (C. cancellatus) and C. convexus; p. 603) could have arrived hydrochorously from the same direction. The important question arises as to how far immigration from the eastern Baltic region—directly or via southwestern Finland (including Aland and the remaining skargard)—has affected Sweden as well. With regard to the Baltic islands, especially Gotland and the neighboring small islands (Sandon, Faron), this question has been dealt with in detail (pp. 282, 287, 291) and the great importance of this migration route for winged forms emphasized. For two functionally brachypterous species (Carabus clathratus, Cymindis macularis) hydrochorous transport from the same direction was assumed. Paim (1942, pp. 49 ff.) investigated the possibility of eastern immigration across the sea directly to the mainland of central Sweden. He gives a list of 27 species of the fauna of the lower Dalalv River region (Upl, Gst, Dir) which in Sweden have a more or less pronounced eastern distribution but seem to lack an area connection to the south (and the north). All of the species considered are winged, and Palm may have been correct in assuming anemochorous immigration from the east for a majority of them. There are only a few carabids with a completely isolated area in east- central Sweden. The best examples are: Agonum longiventre Demetrias imperialis Bembidion humerale Dichirotrichus rufithorax B. transparens Oodes gracilis. Some of these species have been considered earlier. Dischirotrichus rufitho- rax was said to have probably arrived synanthropously from the east (p. 633). The central Swedish stock of the dimorphic Bembidion transparens may have immigrated anemo-(hydro-)chorously in the macropterous form (p. 395). Noth- ing can be said about the immigration route of the two almost identically dis- tributed species, Demetrias imperialis and Oodes gracilis, since their Swedish area is equally isolated to all sides. They were considered as warm period re- licts and represent only the extremes of a whole series of species that have 721 631 found a favorable refuge in the central Swedish warm region (p. 691). Possibly the same interpretation may be applied to Agonum longiventre (Palm, 1942, p. 55). But this is doubtful for three reasons. First, the Swedish record locality of the species—at any rate macroclimatically considered—is not a pronounced warm region. Second, the curculionid Larinus sturnus Schall. in Sweden has a nearly identical area, and this species is distributed in Finland so far north (as far as Oa, Sb, Kb; Catalogus, 1939, p. 107) that it cannot have any definite heat requirement. Third, Agonum longiventre actually has its nearest connection with the east. The species is of course missing from Fin- land, but was found in Estonia near Narva and also in the Leningrad region. In Central Europe it is likewise markedly eastern (Horion, 1941, p. 327), and it is therefore difficult to suppose that the Swedish area represents the result of a southern immigration. But it must be conceded that the last-mentioned objection is also partly applicable to Oodes gracilis, which at any rate in Ger- many is markedly eastern (map in Lindroth, 1943a, p. 119). In the case of this species it is wise to leave the immigration route unresolved, but in the case of Agonum longiventre we are justified in assuming an eastern immigration to Sweden by the aerial route. The present-day restriction to a very small area may be explained (p. 696) by the “virgin forest character” of the region. It seems unambiguous that the two central Swedish records of Bembidion humerale are the result of migration from Finland, where this very active flier has a much wider distribution. It is possible that this is also true of Prerostichus angustatus. It is more difficult to decide whether flightless insects, too, could have reached Sweden from the eastern Baltic region or Finland without the help of human displacement. Crepidodera nigritula Gyll. ought to be considered which Palmén (1944, p. 228) includes among the “Baltic” immigrants. In Sweden this species occurs only in Sdm and Upl (it is also missing from Denmark) and seems to be constantly brachypterous. However, in Finnish drift material (Palmen, 1944, p. 57, and in litt.), 4 macropterous specimens were found. It is thus conceivable that the species originally reached Sweden in the macropte- rous form—as was surmised hypothetically for Bembidion transparens (p. 395). On the other hand, as discussed above (p. 287) the dimorphic Cymindis mac- ularis may actually have colonized the small island of Sandon in the outer skargard of Stockholm by hydrochorous transport in the brachypterous form. 2. The “Karelian” immigration route between the Gulf of Finland and the White Sea, represents so to speak the normal direction of immigration of the Finnish flora and fauna. However, it was fully opened only at the end of the Ancylus period, since the northern part of the Isthmus of Karelia and present- day coastal regions of Finland were earlier submerged (Fig. 99; see also map in Sauramo, 1942, p. 228). This situation has certainly substantially contributed to the fact that the late-glacial “Dryas flora” of the extreme southeast of Finland (Hyyppa, 1933, pp. 10 ff.), as far as is known, has not advanced farther north. - 632 The forest seems to have reached into southeastern Finland as early as the “Salpausselkä” stage (Hyyppä, 1933). The “Karelian” immigration group of flora has been most recently treated by A. Kalela (1943, pp. 37 ff.), that of the coleopteran fauna by Palmen and Platonoff (1943, pp. 171 ff.). An enumeration of individual cases of more or less distinct “Karelian” immigrants has little purpose, since this characteristic is usually evident from the relevant distribution map. On the other hand the same two questions can be raised here that we attempted to answer above with respect to the southern immigration element in Scandinavia, namely: a. Did the pronounced alpine and subalpine species of the present-day fauna of Finland also arrive from the southeast? b. How far north and west has the “Karelian” immigration group ad- vanced? a. The first question has been answered above on the basis of botanical evidence. The late-glacial, pronounced alpine flora and fauna that demon- strably occurred in the Isthmus of Karelia (among the carabids Pterostichus vermiculosus Mén.) were apparently not in a position to follow the retreating ice edge north. Valle (1933, p. 103) is of the same opinion with respect to the immigration of pronounced alpine Lepidoptera. b. Early on the possibility of immigration to Finland was already available to the “Karelian” species. The significance of this fact for the flora has been considered in detail by Hiitonen (1946). Although at first—before the Ancylus period—the land largely lay under water and was fragmented into an enormous skargard, these species had “plenty of time” to colonize a wide region. The southern Swedish group (which arrived through Skane) only later reached comparable latitudes in Scandinavia. Where these two stocks have not merged and lost the imprint of a double provenance, the Finnish stock has almost without exception advanced farther north. In some cases this difference is due to climate (p. 459), in others due to the history of immigration. But it is just the most pronounced examples of the last-mentioned type that are not—or not exclusively—of “Karelian” origin in the sense used here by Palmén and Platonoff (1943, pp. 167 ff.). Another, more northern element is also involved. It is therefore appropriate to postpone further consideration (see p. 724). Let us first take up another question: To what extent are these species with both “Baltic” and “Karelian” immigration? And is the double origin always evident in such cases? This theme has already been considered with respect to Coleoptera as well (Palmen and Platonoff, 1943, p. 183; Palmen, 1944, pp. 217, 227). All of the five examples cited show a more or less pro- nounced gap halfway along the south coast of Finland. Four of them (Aphodius lividus Ol., Aphthona euphorbiae Schrk., Crepidodera nigritula Gyll., Hetero- cerus marginatus Fbr.) are in addition represented in the drift material of the Tvarminne region (Nl), but not the fifth species, Bembidion monticola. In this 723 724 633 case it appears to me somewhat bold to assume a “Baltic” immigration, since all northern limiting localities of Bembidion monticola, as far as I can decide, have a relict character (see map by Netolitzky and Sainte-Claire Deville, 1914). Pterostichus aethiops may occupy a similar position. A distribution gap halfway along the Finnish south coast need not be due to the history of immigration. There are, namely, two other facts to be considered: First, the coastal stretch Helsinki-Viborg (or at any rate Pärnä- Viborg) has been entomologically less explored than the two “corners” of the country. Second, we cannot rule out a climatically determined gap in that region; a lacuna here is also shown by Quercus (Fig. 61, p. 437). Carabids that show a more or less distinct gap on the Finnish south coast are (omitting those found only in drift material in the southwest): 0 Acupalpus dorsalis oO B. transparens w? A. exiguus o B. unicolor w? A. meridianus o Bradycellus collaris o Agonum assimile w? B. similis oO A. dolens k Calathus ambiguus 0. A. ericeti k €. fuscipes 0? A. livens w Carabus cancellatus k A. marginatum k? C. convexus w? A. micans w? C. violaceus kA. ruficorne Cicindela maritima’ oO A. thoreyi o Cychrus caraboides o Amara apricaria o Cymindis macularis oO A. communis o _Dichirotrichus rufithorax o A. consularis o Dyschirius lüdersi o A. famelica D. obscurus' oO A. familiaris o D.z politus oO A. ingenua o Harpalus fuliginosus oO A. montivaga k? AH. rubripes oO A. municipalis o A. tardus oO A. ovata k? Lebia cyanocephala ? A. quenseli K? Metabletus foveatus k A. spreta w? Microlestes minutulus ? Asaphidion pallipes ? Nebria gyllenhali o Badister bipustulatus 0 Notiophilus germinyi oO B. dilatatus w? N. pusillus w Bembidion biguttatum k Odacantha melanura ? B. grapei w? Olisthopus rotundatus OB. properans 0 Panagaeus crux-major w Bembidion schuppeli 0? Pterostichus angustatus ‘Not found in quicksand regions (Krogerous, 1932, p. 26). 725 634 w. P. aterrimus 0? Trechus discus k? P. gracilis 0? T. micros o Tachys bisulcatus o T. rivularis The code indicates the assumed cause of the gap; k—climatic, w—history of immigration, o—due to insufficient exploration, ?—unknown factors. The long list shows how common a distribution gap is along the Finnish south coast, but at the same time that more than one-half of the cases (35 of 64) are very probably attributable to insufficient entomological exploration. Hence great caution must be exercised in proposing a “double” immigration for southern Finland. On the other hand it is obvious that in most species with such a “double” immigration the two stocks soon lost their respective provenance by a merging of the geographical characteristics. So I will not discount the “Baltic” immigration, but merely point out that it is not possible to confirm it in every case. 3. It has already been established that an immigration of plants and animals took place across the eastern border of Finland in the region of the Arc- tic Circle. Cajander (1921, p. 9) has described how he visualizes the migra- tion of Dianthus superbus through Kuusamo to the Gulf of Bothnia. Lists of plants with similar distribution and similar assumed history have been pro- vided by Erlandsson (1939) and A. Kalela (1943, p. 50). A good zoogeograph- ical account—including the history of immigration—of the insect fauna of the Paanajarvi lake region is given by Platonoff (1943, pp. 108 ff.) on the basis of Coleoptera. Krogerus (1932, p. 252) had earlier proposed the Kuusamo route for the immigration of some quicksand insects, including Cicindela maritima and Dyschirius obscurus. Among carabids there is in fact an imposing series of species which, jud- ging from the usual distribution map alone, are to be interpreted as eastern immigrants both above and below the Arctic Circle. Examples (br = function- ally brachypterous): (Agonum micans) 2? (Bradycellus ponderosus) br (Bembidion aeneum) (Cicindela maritima) B. hasti br Dromius sigma B. litorale Dyschirius lüdersi (B. ruficolle) (D. obscurus) B. saxatile (D. septentrionum) br B. schüppeli D. thoracicus B. tinctum Harpalus fuliginosus. br B. transparens The species in parentheses are unknown along the shore of the White Sea, which is unimportant, given the poor exploration of this region. Three species (Bembidion hasti, B. transparens, Dyschirius obscurus) are strikingly isolated along the Gulf of Bothnia. 726 7127 728 635 Even a superficial scrutiny of the list gives a strong impression of the ecological homogeneity of the species concerned. With the exception of two species (Bradycellus, Harpalus; and less regularly Dromius sigma), all of them, 14 of 17! species, are markedly ripicolous. The corresponding group of plants (“Primula sibirica group”; Erlandsson, 1939) distinctly shows the same charac- teristic. Of the 20 species mentioned there is only one (Moehringia lateriflora) that does not grow more or less regularly along the shore—where most of them are found exclusively. It is quite natural that attempts were made to explain this animal and plant group by an earlier aquatic connection between the White Sea and the northern end of the Gulf of Bothnia. Kuusamo and the eastern and southern parts of the parish of Salla (Kuolajarvi lake) are situated east of the main watershed toward the White Sea (see map, Fig. 45, p. 391). This is geomor- phologically a poorly marked area between the Oulankajoki (including its tributaries), emptying into the Paanajarvi, and the water system of the Kemi- joki. It is just the valleys of the Paanajarvi and the Kutsajoki in southern Salla that have been considered as “doors” for eastern flora and fauna (A. Kalela, 1943, p. 50). An idea of the current thinking of Finnish geologists (Hyyppa, Sauramo) concerning the distribution of land and water in the northeastern Finland in the oldest postglacial period (s./.), is given in the maps, Figs. 104 and 105. Noteworthy are first the remarkably early melting of inland ice in these regions, second the low elevation of the land, which in part was responsible for a great extension of the White Sea westward and in part caused the extensive inundation of parts of Finland closest to the ice edge. Erlandsson (l.c.) states the watershed in the Kuusamo-Salla region was breached by two straits (especially in the Baltic Ice Lake period!) and refers to Hyyppa. However, in his contribution of 1936 (pp. 437 ff.), he speaks only of “discharge channels,” and even during the Yoldia period in this region there was apparently no open connection with the White Sea (Fig. 105). The distance between the western bays of the White Sea and the offshoots of the Baltic Ice Lake or of the Yoldia Sea in Kuusamo-Salla was less than 10 km (in the Kutsa River region) or 2-3 miles (west of Paanajarvi), respectively. Especially in the earlier period, when these passes were occupied by enormous rivers, they may have offered excellent “migration routes” for ripicolous (and other) animals and plants coming from the east. Keeping Bembidion aeneum in view, for which species we have assumed a slight halobionty (p. 521), it is important that from time to time Paanjarvi was a western offshoot of the White Sea and so passed through a saltwater stage. The two specimens of the species found here were both brachypterous (Fig. 49, p. 400), and we are therefore justified in assuming that the species had already immigrated in this early period. Such an assumption is not at all improbable. The studies on pollen analysis by Hyyppa (also 1941, p. 609) have, as it seems, unambiguously shown that 726 636 the climate during the latter part of the Baltic Ice Lake period in the eastern border regions of northern Finland was already so favorable that the ice-free and supra-aquatic parts on both sides of the present-day border of the state were covered with Pinus-Betula forests. When the distribution of land and water was favorable (Baltic Ice Lake period and Yoldia period) for an eastern immigration this could have been undertaken by other than purely alpine- subalpine species as well. 500 km M.S-mo 1938. Fig. 104. Final stage of the Baltic Ice Lake. According to Sauramo (1942). Cf. Fig. 97 (p. 699). 729 637 The question must be asked whether the group that immigrated through Kuusamo-Salla is to be considered simply as the “vanguard,” as the forward patrol of the “Karelian” invasion or whether a separate origin—completely or partly—should be ascribed to it. A. Kalela (1943, pp. 46, 49) favors a sort of compromise. He thinks the plant group that arrived in the region of the Finnish state through Kuusamo-Salla belongs partly to the “Karelian” ele- ment, partly to a more northern element, which has spread from the east along the south coast of the Kola Peninsula. Had the wing-dimorphic species not provided a good clue, I could not have ventured an opinion on this issue. However, the dimorphic maps of a number of species, which were considered above from this viewpoint too, bring us a step closer to an explanation: Bembidion guttula (p. 387, Fig. 42) Pterostichus lepidus (p. 381, Fig. 37) B. nigricorne (p. 389, Fig. 44) P. minor (p. 387, Fig. 43) B. transparens (p. 389, Fig. 45) P. strenuus (p. 395, Fig. 47). Carabus clathratus (p. 381, Fig. 38) The distribution of macropterous and brachypterous individuals of these species in eastern Fennoscandia against all expectations shows an increase in the component of flightless individuals in the north. From this it may be concluded that the immigration via Kuusamo-Salla (of Bembidion nigricorne, possibly also of Carabus clathratus, somewhat more southerly) was functionally separated from the normal southeastern “Karelian” immigration route. Is it conceivable, as A. Kalela assumed in respect of several members of the “woodland flora,” that the carabids in question spread from Kanin-Mezen along the south coast of the Kola Peninsula? Of the 24 “Kuusamo-Salla species” in the two lists, 15 species are not found along the south coast of the Kola Peninsula (however, Prerostichus lep- idus has been found near Lm Kantalaks). It must be noted that, thanks to the intensive research by Finnish entomologists of the last century, the sou- thern coastal regions of the Kola Peninsula are among the coleopterologically best-known parts of northern Fennoscandia. Of the same 24 species, 20 are unknown from Kanin-Mezen east of the White Sea (only the following are present: Bembidion hasti, B. saxatile, B. trans- parens, Dyschirius septentrionum). This cannot be considered as conclusive, al- though Poppius (1909a) collected energetically on the Kanin Peninsula. Of these 20 species, as far as is known, 11 or 12 are also missing from the Pe- chora region. Of these latter, 2 of course occur south of the White Sea (in the Archangel region) but without any contact with the stock considered here. Of the same 24 species, as far as is known, 7 species are totally missing from northern European Russia: Bembidion aeneum, B. nigricorne, B. ruficolle, 638 ! 1 1 aly. ent. - S NS 727 Fig. 105. Initial stage of the Yoldia Sea. According to Sauramo (1942) (with corrections by that author, in litt., in respect of position of eastern ice edge). Cf. Fig. 98 (p. 700). Bradycellus ponderosus, Cicindela maritima, Harpalus fuliginosus, Pterostichus minor. Dyschirius lüdersi and D. obscurus are doubtful. These considerations led me to no other conclusion than that most species of the “Kuusamo-Salla group” are functionally (in their history of immigra- tion) separated not only from the normal, southeastern “Karelian” group but also from the “Kanin-Kola group.” Caught in the cross-fire as it were, the element under consideration must have, considered geographically, an inter- mediate origin, somewhere around the White Sea. One is led to the conclusion that somewhere on the western half of the White Sea a Wurm refuge existed, which was isolated both to the south (by ice) and to the easi (by water). If 730 731 639 during the maximum of the last glaciation the seashore—as seems probable (p. 772)—was here considerably lower than today, this assumption is not so implausible as may appear from the map today or, in particular, from a map of the late-glacial and early postglacial period (Figs. 104, 105). Moreover, if the peculiar occurrence of Bembidion chaudoiri on the White Sea is actually as isolated as we at present assume, this would provide further support for the above view. A conclusive answer can be given only after thorough exploration of the fauna along the western and southern shores of the White Sea, which certainly will take time. It may also be mentioned that botanical indices favor a Wurm refuge here, possibly in the Hibina region (Nordhagen, 1935, p. 132). It is possible that the separation between the “Kuusamo-Salla group” and the “true Karelian” group was enhanced from time to time by an aquatic con- nection between the White Sea and the Gulf of Finland (via Onega-Ladoga). Hyyppa (1944, pp. 122 ff.; also Platonoff, 1943, pp. 112-113, footnote) thinks such a connection is indispensable, but it is not generally accepted by others (Tanner, 1930, p. 384; but see also Astrid Cleva-Euler, 1934, p. 92). Poppius’ idea (1909b, pp. 60-61) that some aquatic and riparian species (Coleoptera, Hemiptera) utilized this aquatic route during a dispersal from the Gulf of Finland into the White Sea region is certainly mistaken. 4. The possibility of immigration of plants across the estuarine strait of the White Sea from Kanin-Mezen to Kola has been considered by A. Kalela (1943, pp. 41, 47). He attaches great importance to this route, both for true alpine species (tundra plants) and for woodland flora. He cites Ramsay and Auer, according to whom during the late-glacial period this strait was considerably narrower (see also Tanner, 1930, p. 388), or there was (according to Auer) even a firm land connection between Kola and Kanin. Astrid Cleve-Euler (1934, p. 99) and Har. Lindberg (1938, p. 27) also speak of a fresh-water stage of the White Sea (cf. on the other hand, Tanner, 1930, p. 385). The isolated, certainly old occurrence of Dichirotrichus pubescens in the south on the White Sea seems to show that the salt content in the postglacial period was never below 6 per mille. The possibility of a separate immigration of Coleoptera to the Kola Peninsula has been considered by Palmén and Platonoff (1943, p. 185). The Carabidae that suggest such an immigration route are naturally in the first place those, restricted in Fennoscandia to the Kola Peninsula (br— brachypterous): ** Agonum aldanicum ** br Diachila polita Amara peregrina * br Prerostichus fastidiosus ** Bembidion crenulatum ** br P. middendorffi B. repandum ** br Trichocellus mannerheimi. The species marked with one asterisk (*) is found in the Ponoj region (province of Lj) in the extreme east; those marked with two asterisks (**) occur 732 640 exclusively in this region. At least the last four species are true tundra animals, although they also occur the farthest east in the northern parts of the Taiga. Just these 4 tundra species are wingless and support the assumption that a firm Kola-Kanin land connection in the Ponoj region has existed (A. Kalela, 1943, p. 41). It is namely difficult to see how they could have tolerated hy- drochorous transport in the highly saline water of the estuarine strait of the White Sea (p. 517). Overland immigration via a southern detour around the White Sea seems quite improbable in view of the present-day distribution of the 4 species. Half of the 8 species are known from Kanin (Agonum aldanicum, Bembid- ion repandum, Diachila polita, Pterostichus fastidiosus), a much larger propor- tion than of the “Kuusamo-Salla species” above. But historically Bembidion repandum and Agonum archangelicum (which is distributed farther south along the White Sea coast), both of which are absent from the eastern part of the Kola Peninsula, may belong to the latter group. The distribution of the (7-) 8 “Kanin-Kola species,” restricted mostly to the extreme eastern part of the Kola Peninsula, need not be taken to signify a late (possibly continuing) immigration. It is probable that these continental species have reached their existence limit here in the face of a more oceanic climate. Palmen has rightly emphasized (Paimen and Platonoff, 1934, p. 169; Palmen, 1944, p. 208) that a fixed geographical terminology in respect of the different immigration groups is better than naming them according to the compass bearings, chiefly because the latter is valid only for restricted regions. If one wishes to name the four—from the general Fennoscandian viewpoint—eastern-immigration groups of the fauna (and flora) of Finland, according to this “fixed” terminology the following names may be proposed: 1. Baltic group. 2. Karelian (possibly southern Karelian) group. 3. White Sea group. 4. Kanin-Kola group. It only remains to investigate how far the species (or stocks) that immi- grated across Finland—whatever the route—have reached south and west in Scandinavia. In respect of Coleoptera this question has been considered chiefly by Lindroth and Paim (1934, pp. 120 ff.), Palm and Lindroth (1936, pp. 32 ff.), and Palmen and Platonoff (1943, pp. 171 ff.). These authors are unanimous in that the northeastern group is of decisive importance for the colonization of northern Scandinavia, and that some members thereof have even reached the southernmost parts of Norway and Sweden. Opinions may differ on whether the above-mentioned examples are al- ways correct. For instance, I must presume that Gnypeta coerulea C.R. Sahlb., extending as far as southernmost Norway, considered as extreme by Palmen and Platonoff (1943, p. 176), partially survived the Wurm glaciation in Scandi- 733 641 navia. It is a boreo-British species (Lindroth, 1935a, p. 599), which has at any rate succeeded in doing this in the British Isles. A similar reservation applies to Phytobius velaris Gyll. (Lindroth and Palm, 1934, p. 122). The occurrence on Sma, if it can be confirmed, might be considered as a relict of a southern immigration. The “Karelian” stock seems to have touched only southern Fin- land, and the main Scandinavian range is most probably that of an interglacial relict. A precise mapping of this species would be very rewarding. However, it cannot be denied that some northeastern immigrants to Scan- dinavia reached very far south, and I will try to give some clear instances. It should be expected of them first that they have (as far as is known) uninter- rupted distribution into regions east of the former Wurm ice, second that they are missing from northwestern (or in general from) Central Europe, and third that they are absent from the assumed western and northern Fennoscandian Wurm refuges or occur there only in isolated outposts. Naturally in Scandinavia there are not many widely distributed species that meet all these conditions. The clearest examples are: Bembidion tinctum (Fig. 118, Supplement), in Sweden south to Mdp, and especially Tachyta nana, which in the south has even reached northern Ska. It is missing from Denmark and northwestern Germany (Horion, 1941, p. 171). Agonum mannerheimi, with the peculiar relict occurrence near 12 Vardal, and Tachys bisulcatus (up to Ska) also seem to be eastern immigrants. I must confess I have my doubts about this thesis in the case of species like Notiophilus reitteri and Pterostichus adstrictus, which were earlier (Palm and Lindroth, 1936, p. 34) assumed to be postglacial northeastern immigrants. Both occur within the limits of the assumed Norwegian Wurm refuges. The dimorphic Notiophilus seems to be everywhere functionally brachypterous with a relatively poor capability of dispersalt. Even Amara erratica and A. torrida, which have been repeatedly cited (for instance, Lindroth, 1939a, p. 246; Holdhaus and Lindroth, 1939, p. 261) as typical examples of a northeastern postglacial immigration to Scandinavia could have had a more complicated origin. Of course their almost identical Scandinavian southern limits are undoubtedly dynamicai and not dependent on existence factors (Holdhaus and Lindroth, l.c.). However, especially in the case of A. torrida there seems to be nothing short of a concentration at the assumed northern refuges, which cannot be ascribed to erratic exploration alone. It is possible that today’s seemingly homogeneous area is the result of merging of interglacial and northeastern postglacial stock. On this assumption one might be inclined to argue that “sufficient time” was available for disper- sal farther south along the Scandinavian chain of fjelds. But with this we are touching on the obscure problem of the speed of dispersal, which varies quite inexplicably in different species—and at different times in the same species. tcf. p. 823; suppl. scient. edit.) 734 735 642 Peculiar instances were provided in the earlier section on the recent faunal changes in our region (pp. 621 ff.). The (partially) northeastern origin is especially evident in those Scandi- navian species that have two stocks in the environs of the Bothnian Sea, a south-Scandinavian and an east-Scandinavian, which have still not met. This is the phenomenon of “double” immigration (Lindroth and Palm, 1934, p. 124). As discussed above (p. 714), in the case of double immigration in the Bothnian coastland, the Finnish stock has normally reached farther north than the Swedish stock. In several cases it has even encroached into the region of the Swedish state, where sooner or later a merger with the western stock will follow or has already taken place. If the gap on the Swedish side still exists, the double origin appears readily evident. However, as a precaution it must be remembered that according to the July map (Fig. 63, p. 452) very low summer temperatures are recorded in the border regions Ang-Vbt, where there seems to be a true gap (for in- stance, in the case of Dyschirius thoracicus, Notiophilus palustris) or at least a thinning out of record localities (for instances in the case of Agonum piceum and A. versutum, with reservations on account of insufficient exploration). It seems conceivable that this minus region was strengthened during the subat- lantic climatic deterioration and thus divided an originally homogeneous area in the Bothnian coastland into two parts. In respect of one of the species listed below (Pterostichus niger) the subfossil record near Dir Evertsberg from the sub-boreal-atlantic borderline period (p. 674) seems to show that it was once more widely distributed. The reservation about the significance of the “Bothnian gap” mentioned here, in respect of the history of immigration, nevertheless seems to be largely of theoretical interest. The gap in almost every case has a different (mostly more northerly) position and in my opinion is never climatically determined. In the following list there are also some dimorphic species with no “Both- nian gap,” but in which the limit between the two stocks was determined well enough from the distribution of macropterous individuals. Instances of species with double immigration into the Bothnian coastland, in which the eastern stock has advanced farthest—on Swedish region—are: Agonum dolens Notiophilus palustris A. piceum Panagaeus crux-major A. thoreyi (Fig. 95, p. 675) Pterostichus coerulescens Bembidion guttula (Fig. 42, p. 386) P. lepidus (Fig. 37, p. 379) Carabus clathratus (Fig. 38, p. 380) P. minor (Fig. 43, p. 388) Cymindis macularis Pterostichus niger Dromius sigma P. strenuus (Fig. 47, p. 396) Dyschirius obscurus P. vernalis (Fig. 36, p. 378). D. thoracicus 736 643 An extreme instance among other Coleoptera is provided by Paratinus femoralis Er., which was known in Finland northward as far as Ob, in Swe- den only on Skä. But I discovered it in 1939 on the island of Sandön in the skargard of Nbt Lulea. Finally it may be noted that the eastern postglacial immigrants through Finland, in some cases reached Sweden also by direct immigration across the sea, partly through Aland (p. 719) and partly across the Bothnian Sea, especially at its narrowest part, Kvarken (pp. 381, 593). The peculiar species Amara majuscula (p. 622) has colonized southeastern Fennoscandia on a broad front through the air. III. THE PROBLEM OF “WURM HIBERNATION” When Nathorst (1871) made his famous discovery of the subfossil Dryas and other alpine plants near Alnarp on Skane, it provided the first palpable proof that the Scandinavian fjeld flora was pushed into more southern regions during the Ice Age. Of course the account of the Scandinavian faunal history by Sven Nilsson (1847) anticipated this viewpoint. Gradually the geologists established signs of Quaternary land ice in all parts of Fennoscandia. It was logical to think that the ice in this part of Europe had destroyed all life, and that the present- day Fennoscandian flora and fauna as a whole immigrated in the postglacial period from surrounding land which had remained ice-free. Sernander called this long-held view the tabula rasa theory. It was an important finding for a correct understanding of the biological effect of the Glacial Epoch that the so-called Ice Age was a complex of sev- eral separate glaciations, each of which, geologically as well as biologically, presented a separate problem. A systematic arrangement of the glacial for- mations during different glacial periods was first undertaken for the Alps (Penck and Brückner, 1901-1909), where at least four glaciations were estab- lished. In Fennoscandia their traces could not be consistently separated from one another for understandable reasons; however, since the southern marginal deposits of the northern inland ice in northern Germany were attributed to three or four separate glaciations (Richter, 1937) it was generally assumed by geologists that Fennoscandia had gone through at least three Quaternary glaciations, with long intervening climatically favorable periods (for instance, A. Wagner, 1940, pp. 141 ff.). The estimated durations of the different glacia- tions varied widely according to the different methods of calculation. At any rate they, as well as the interglacial periods, lasted much longer than the entire postglacial period (hence > 20,000 years long). The ice cover was most extensive (at any rate in Europe) during the penul- timate glaciation, generally called the Riss (Fig. 101, p. 705). The subsequent, last interglacial period is known partially also from Swedish deposits (p. 673) and indicates a climate which from time to time was as favorable as today’s. 737 644 During the last glaciation (Wurm) the northern inland ice was less exten- sive in all directions (Fig. 102, p. 706). Among other places, the southwestern part of Jutland lay beyond the ice edge. Since a climatically ameliorative in- fluence on the part of the sea, the Gulf Stream included, can be assumed for the whole of Western Europe, the idea developed early that even on the Scandinavian west coast, ice-free regions would have existed throughout the Wurm Ice Age. At first this problem was purely biogeographical. Blytt (1893, p. 26) tried to explain the occurrence and distribution of some Norwegian fjeld plants on the assumption of Ice Age refuges; he was followed by Sernander (1896, p. 117). On a larger scale, still on the basis of botanical material, the ques- tion was considered by A.M. Hansen (1904, pp. 282 ff.), who visualized the “hibernation” of at least 300 phanerogams in Norway—and indeed along the west coast, but not on nunataks. Wille (1905) assumed a refuge in the envi- rons of 8 Nordfjord. In 1913, two important phytogeographical contributions (Th. Fries, Teng- wall) posited a Wurm hibernation in Norway for many fjeld plants. Fries (p. 314) believed the refuges were chiefly ice-free coastal stretches; Tengwall (p. 269, in respect of southern Norway) visualized them as isolated nunataks. These two studies carried more conviction than the earlier ones, especially because they drew on actual geological material. Since shortly before, Vogt (1912, pp. 6, 47) and Enquist (1913, at greater length in 1918, for instance, pp. 5 ff.) had shown, that parts of the outermost Lofoten islands were located beyond the ice edge even at the Wurm maximum (also according to Ahlmann, 1919, pp. 217, 238). Nordhagen, in a series of resourceful contributions (1933, 1935, 1936), was still more successful in working with a combination of botanical and geological facts—the latter largely from his own observations. The prevailing view of the Nordic biogeographers on the Fennoscandian Wurm refuges has been taken predominantly from his publications. Among other things, he has attempted a more precise localization of these refuges, which is depicted in the map in Fig. 106 (p. 739). Other botanists who have effectively contributed toward a solution of the “hibernation question,” and whose results are considered below, are: Smith (1920, pp. 138 ff); Elfstrand (1927); Degelius (1935, pp. 297 ff.); Nannfeldt (1935, 1947); Holmboe (1937); Faegri (1937, pp. 433 ff.); Bjorkman (1939, pp. 218 ff.); Arwidsson (1943, pp. 98 ff.); Dahl (1946); Lindquist (1948, pp. 319 ff.); AhIner (1948, pp. 140 ff.). The Nordic zoologists first hesitatingly followed in the footsteps of their botanical colleagues. The first of these was Stejneger (1907, 1908). In 1910 Sparre Schneider also clearly expressed the idea of interglacial relicts in the t(= Rocks or mountains which project from glaciers and mass of inland-ice; suppl. scient. edit.). 740 645 northern Norwegian fauna. Wahlgren (1919) has used the “hibernation hy- pothesis” to explain the so-called western Arctic element of our lepidopteran fauna. Ekman (1920; 1922, pp. 397 ff.) considers the lemming (Lemmus lem- mus) to be the only definite Wurm hibernator of the Scandinavian fauna. In recent years it is chiefly the entomologists that have shown interest in these questions: in addition to Wahlgren, Brundin (1934, p. 174), Strand (1935, pp- 67 ff.; 1946, pp. 22 ff.) and myself (especially Lindroth, 1933, 1935a, 1935b, 1939a, 1941, 1948a; Holdhaus and Lindroth, 1939). Among authors who have dealt with similar “hibernation problems” in other regions mention may be made of the following: The Faeroes: Sparck (1924, p. 502), West (1930a, pp. 87 ff.). Iceland: A.M. Hansen (1904, p. 351), Lindroth (1931, pp. 557 ff.) Greenland: Gelting (1934, pp. 250 ff.), Bocher (1938, pp. 312 ff.). Remaining Arctic islands: Wahlgren (1920), Strand (1942), Lynge (1934). North America: Fernald (1925, 1929), Raup (1941). The faunal history of the British Isles is discussed below (p. 793). In the Central European mountains, chiefly in the Alps, these questions center on the so-called “massifs de refuge” (for instance, Heberdey, 1933). Reinig (1937, for instance, Fig. 13, p. 50) attempted to trace back the entire faunal history of the Holarctic region to a series of Ice Age refuges. Criticism of the hibernation theory was expressed by Tanner (1930) from the geological viewpoint and by Wynne-Edwards (1937, 1939) from the botan- ical viewpoint. On the basis of the coleopteran fauna we shall now attempt to take up the problem of Wurm hibernation simply by posing three fundamental questions: I. How is it possible to decide whether a species of animal or plant is a Fennoscandian Wurm hibernator? II. Is it possible to locate the refuges on a purely biogeographical basis? III. Did climatic conditions during the Wurm Ice Age actually permit the presence of a more richer fauna and flora in the Fennoscandian refuges? I. It will be best to take as the basis of our study a species whose charac- teristic as a Fennoscandian Wurm hibernator is clearly evident, namely Sim- plocaria metallica Sturm (Fig. 106, cf. Lindroth, 1948a, p. 19). On the basis of well-known recent and subfossil records, the history of this species during the late-glacial period can be reconstructed as follows: During the last interglacial period it was a frequent species in Scandinavia; it is the only beetle present in all four known Scandinavian subfossil samples contain- ing insects. It was pushed south to Central Europe by the inland Wurm ice; subfossils were found in abundance near Deuben in Saxony (Nathorst, 1894) in “diluvial clay” (but it is uncertain whether this is from the Wurm period). In the final phase of the last glaciation, the species was initially able to follow the ice edge at least in the east; it was found in late-glacial Dryas deposits near Ik Kivennapa (Poppius, 1911). Even today this southeastern postglacial 741 646 stock leads a relict existence as an extreme rarity in southern Finland. The two remaining Fennoscandian subareas are located in the fjeld regions and are completely isolated. They originate from populations that survived the Wurm glaciation on the coast in the west and north. The reliability of this reconstruction, which may also find favor with critics of the hibernation hypothesis, of course depends largely on the availability of subfossils. In other species considered here these are either not available or not so complete. Hence the important question: Is it possible to posit Wurm hibernation of a species of animal or plant without any fossil records? In the present case: Is Simplocaria metallica, even by the recent distribution alone, a definite Wurm hibernator? The last question must be answered in the affirmative, and hence the first, too. Let us here establish more precisely which characteristics of the present-day distribution picture of Simplocaria metallica support a Fennoscan- dian Wurm hibernation (cf. also map in Plate XIII, Holdhaus and Lindroth, 1939). a. The isolation of the Fennoscandian area (or of part of it). First and foremost, Simplocaria metallica lacks any connection to the east. The species is absent even from the Kola Peninsula and is altogether unknown east of Fennoscandia. Off the map (Fig. 106) there are localities only in the higher mountains of Central Europe (most northerly in Riesengebirge) and on Green- land. It is especially significant that this distribution makes a postglacial immi- gration from the east (from Russia and Siberia) impossible, since the in part well-explored Kola Peninsula also lies outside the area limit. b. The occurrence of an isolated stock in a southern (climatically more fa- vorable) locality, in the case of Simplocaria metallica the small isolated subarea in southern Finland. This occurrence namely proves that the Fennoscandian main area was formed under the influence not of climatic factors but of the history of immigration of the species. The postglacial stock has not reached farther than southern Finland. c. The bicentricity in the fjeld regions. Botanists tend to regard this as the most important circumstantial evidence for a glacial hibernation (further dis- cussed below). In the case of Simplocaria metallica this viewpoint is especially relevant because—as a predominantly subalpine species—it cannot be bound to the higher montana regions and the far north on climatic grounds. The central Scandinavian gap is due to historical factors. Not every Wurm hibernator will show all of the three above-mentioned characteristics of Simplocaria metallica, since almost without exception one or more postglacial stocks have “adversely” affected the distribution pattern. Actually among the Fennoscandian carabids there is only one species (Elaphrus lapponicus) that forms a valid counterpart. Its area is completely isolated to the east, the remains of a postglacial stock are found in a relict locality in Latvia, and the bicentricity in Fennoscandia is clear. Here again the fossil records D IS A ee m 739 Fig. 106. Simplocaria metallica Sturm. Assumed Würm refuges are hatched (according to Nordhagen, 1933, 1935). 742 648 (p. 699) show on the one hand that the species was native to Fennoscandia during the interglacial period, on the other hand that during the Würm Ice Age it lived along the southern edge of the Nordic inland ice (in Jutland). Further, to the characteristics mentioned for Simplocaria metallica, it may be added in the case of Elaphrus lapponicus that this is one of the boreo-British species which also “hibernated” in the British Isles (Lindroth, 1935a). In all the other species of the Fennoscandian fauna presumed to be Würm hibernators on the basis of more or less strong circumstantial evidence, we find only one or two pronounced characteristics of the recent area out of those established for Simplocaria metallica. It is advisable to arrange these species according to these characteristics. | a. Species with an isolated Fennoscandian area—chiefly to the east. 1. Total isolation of the entire Fennoscandian stock (less evident cases in parentheses): (Agonum consimile) (B. prasinum) (Bembidion chaudoiri) B. scandicum B. dauricum B. siebkei (B. grapeioides) (Dyschirius helleni) (B. hyperboraeorum) (Nebria gyllenhali balbit) B. lapponicum Trechus obtusus. Part of the European subarea of Agonum archangelicum lying outside (east of) the region seems to be completely isolated as well. 2. Partial isolation, with a discrete subarea in the southwest (W) and/or in the north (N): N Bembidion aeneum N B. transparens W B. argenteolum N Carabus nitens N B. femoratum N C. problematicus W B. litorale WN Cicindela maritima W B. lunatum WN_Dyschirius angustatus W B. minimum W Perieptus areolatus W Bembidion semipunctatum N Pierostichus niger. 3. Isolation only to the northeast, so that the possibility of a postglacial immigration from this direction is eliminated: Amara interstitialis Blethisa multipunctata A. lunicollis Bradycellus collaris Asaphidion pallipes Dyschirius septentrionum Bembidion grapei Harpalus winkleri B. saxatile Notiophilus biguttatus (B. schippelt) (Trachypachys zetterstedti). B. velox 743 649 b. Species whose assumed Fennoscandian interglacial stock is separated from a more southern located subarea (possibly the main area) by a more or less distinct gap (“zone of obliteration’), which cannot be explained on the basis of climatic or other existence factors. In addition to the species listed above under a(2): Amara praetermissa C. violaceus (in Finland) A. quenseli Dyschirius septentrionum (in Finland) Asaphidion pallipes Nebria gyllenhali (in Finland) Bembidion schüppeli Trichocellus cognatus. Carabus coriaceus c. Bicentric species: Amara nigricornis (“tricentric” ?) Dyschirius angustatus (Bembidion fellmanni) D. helleni B. siebkei (Miscodera arctica) Carabus problematicus Nebria nivalis Cicindela maritima (Trichocellus cognatus). (Cymindis vaporariorum) In the case of the 4 species in parentheses the gap may appear insignificant. But it lies in the well-explored province of Jtl and may therefore be real. The bicentricity is most evident in Elaphrus lapponicus (p. 741). It is possible that these peculiarities of the map of one or other species of carabid may not convince the skeptic of a Wurm hibernation in the Fennoscan- dian region. Before I go on, I would like to point out the evidence of wing- dimorphic species (pp. 366 ff.) and draw some inferences. In the above-mentioned chapter (p. 415) it was—if I may say so—proved, that 7 species must be considered as Wurm hibernators in the Fennoscandian region; for 2 more species a refuge along the eastern boundary of the region (on the White Sea) was assumed. The seven “certain” species are: Bembidion aeneum Notiophilus aquaticus B. grapei N. biguttatus B. transparens Pterostichus strenuus. Bradycellus collaris Some of these actually belong to characteristic distribution types, which are also represented by other, non-dimorphic species, and it might not be too bold to assume for these by analogy a largely identical history. The following groups may be excluded: a. Bembidion aeneum type. A more or less isolated “refuge area” in western or northern Norway. Bembidion minimum. The isolate in the Trondheim region cannot have arisen by upward migration into the eastern Norwegian valleys during the 744 745 650 Warm period, as was assumed for Trechus discus and T. micros (p. 689), because the species is bound to the seashore (p. 520). B. femoratum. The homogeneous northern Norwegian subarea is isolated on all sides and represents an unambiguous interglacial relict (Lindroth, 1941, p- 438). Carabus nitens. In Norway between latitude 66° and 69° N there are 7 lo- calities which form a coherent isolate. The species is constantly brachypterous. C. problematicus. Also constantly flightless. The northern Fennoscandian subarea (Subspecies strandi) is also completely isolated to the east, since the species is unknown on that side of the Kola Peninsula. Pterostichus niger. The isolate in northern Norway comprises just 2 locali- ties, one of which is on the quite distant island 35 Hillesoy, where the species cannot have been displaced (p. 324). On the other hand the significance of the isolated localities of Cicindela campestris in the same regions is uncertain. Trechus obtusus. The entire western Scandinavian area of this species, which is constantly brachypterous in our region, represents so to speak an expansion of the western Norwegian Bembidion aeneum area on all sides. The Trechus species has also crossed the Swedish border at four places. b. Bembidion grapei type. The species included here might be considered as postglacial offshoots of a “Siberian” stock. Amara interstitialis. No connection to the northeast. A. nigricornis. Bi- (or tri-) centric. Concentration in the Veranger-Petsamo region. Bembidion difficile and B. prasinum. Slight connection to the northeast. Nebria gyllenhali. In the south there are discrete postglacial stocks, especially in Finland (Holdhaus and Lindroth, 1939, p. 270). Notiophilus reitteri, Patrobus septentrionis, and Pelophila borealis snow con- centrations in one to three assumed refuges on the coast in the north, which are certainly not just apparent (and were discovered by more thorough ex- ploration). Patrobus and Pelophila occur here as far as the outermost islands. Notiophilus is functionally brachypterous (dimorphic, but only one macropte- rous specimen is known). In Pterostichus adstrictus, moreover, the poor connection to the northeast is noticeable. c. Pterostichus strenuus—Bradycellus collaris type, distinguished by lack of or very slight connection to the northeast, also by the distinct separation in the north between the Norwegian and Swedish stocks. Amara aulica Leistus rufescens A. bifrons Pterostichus oblongopunctatus A. communis Trichocellus placidus. A. lunicollis 746 651 Amara familiaris and Harpalus latus are doubtful; in the north they show a tendency toward synanthropy. But there are other distribution types, which are not represented among the dimorphic species but seem to justify the assumption that at least part of the Fennoscandian area is to be attributed to “Würm hibernation” in light of certain cartographic peculiarities. d. The West Scandinavian type. This comprises a few species that occur only in Norway, and in addition a large group whose members reach the highest latitudes along the Scandinavian west coast. This condition in most cases is no doubt climatically determined (pp. 474 ff.), and it is therefore not right to conclude “Würm hibernation” only on the basis of a wide distribution in Norway. On the other hand this is possible in species whose Norwegian stock encroaches into the Swedish region in places they would have reached more easily from the south—if the entire Scandinavian population is due to postglacial immigration. This is a proof that they lack a predominantly western distribution for climatic reasons. Bembidion nitidulum has crossed the Swedish border toward the east, both in Jtl and in southern Lapland. The occurrence east of the Baltic Sea up to latitude 63° N shows still more clearly that it is not an “Atlantic” species. B. virens. 1 earlier (Lindroth, 1935a, p. 624) showed the Norwegian origin of the Swedish stock, which immigrated down the rivers. Carabus coriaceus. The Scandinavian area is very peculiarly split up. Since this largest carabid of our fauna cannot easily escape notice, the gaps on the map must be considered mostly real. This holds especially in its occurrence in the eastern Malar lake region. Here Carabus coriaceus is widely distributed and is even abundant at places in the Stockholm region south of the Malar lake; on the other hand so far not a single individual has been found in southern Upl (it is found again only north of latitude 60°N and in northern Vst). This is a cer- tain similarity with the distribution of Hedera (Froman, 1944, pp. 663 ff.). The enigmatic distribution of this Carabus species seems to be understandable only in light of conditions during the first half of the Littorina period (Munthe, 1940, Plate XII; also Fig. 100, p. 702). The land namely emerged from the sea south of Stockholm much earlier than north of it, and the possibilities of a new col- onization were correspondingly greater for a soil-bound species like Carabus coriaceus. It seems important to establish that these sections of the Littorina period fali in the atlantic period, when the climate was generally very humid. That Carabus coriaceus is favored by a humid climate seems not only to be evident from its advance far north in Norway but namely is related to its mode of life. Its food consists chiefly of large snails (both!. . .) are bound TA line seems to be missing in the original text, and a previous line is repeated in its place. From the biology of Carabus coriaceus (Part J, p. 532), the sentence may be reconstructed as follows: “(both Helix and slugs), which . . .” —Translator. 747 652 to forest, i.e. to a humid microclimate. Like Carabus (for instance, on the island of Hitra), in western Norway they live in more open situations (see, for instance, Arion ater in Pkland, 1925, pp. 15 ff., Plate II). One may therefore surmise that Carabus coriaceus enjoyed good oppor- tunities of dispersal during the atlantic period and at that time colonized, among other places, the region south of Stockholm. The subfossil record near Dir Skarsjo (p. 658) shows that at that time it was already present in central Sweden. During the following, generally drier sub-boreal, Carabus may have retreated deeper into the forests. These were adversely affected in the Bronze Age by the increased reclamation of land, so that the capability of dispersal of the carabid was curtailed. Nevertheless, the most striking gap in the distribution of Carabus cori- aceus is not the one here considered in Upl, but the one that separates the southern Swedish stock (in Ska, Hall! and possibly also Old) from the more northern stock. I find this incomprehensible, unless we are dealing with two immigration groups. This means that the central Swedish stock should have come from the west, from Norway. This direction of migration has already been established for Amara montivaga (p. 632); probably it was also true of Harpalus puncticollis. However, a postglacial immigration from the southwest directly to Norway seems inconceivable for the flightless Carabus coriaceus. Hence it must be assumed that C. coriaceus survived the last glaciation some- where in the southern half of Norway. Carabus hortensis may have had a similar history, but the almost uninter- rupted Scandinavian distribution of this species does not show any separate stocks. Except the striking absence from the Jaeren Peninsula (Province 6) represents a gap. Harpalus winkleri. North of latitude 64° N the western origin of the Scan- dinavian stock is quite evident all the more so since the species is not re- stricted to regions with an oceanic climate. The quite isolated solitary records in northeastern Fennoscandia are of enigmatic origin. Leistus ferrugineus and Patrobus atrorufus are apparently favored by an oceanic climate, which, among other things, is evident from the lie of their northern boundary in Finland (p. 474). Nevertheless, the passes of the Scan- dinavian main watershed were crossed eastward by both species at several, sometimes similar places; they undertook a real invasion through the passes in central Jamtland, which in the Leistus species is distinctly limited to the south. Both are flightless and are undoubted hibernators along the Norwegian coast. Trechus secalis, which is likewise constantly brachypterous, belongs to the same general distribution type and has certainly gone through the same tA misprint in the original. Should be “HIl”— General Editor. 748 653 Fennoscandian history. Only in Jtl has it pushed east through the passes, and its western stock in Sweden has so completely merged with the southern stock that no distinction can be drawn. Nevertheless, it is noteworthy that the carabid is apparently missing from the well-explored region of Los (Hls) (the dot on the map in Part II was inserted out of a misunderstanding: see Supplement). A verification in nature of the ostensible record from 36 Nordreisa in the collection of Embrik Strand would be very interesting. If correct, it would show a hibernation even in the far north. At the same place Nigritella nigra has its only record locality in northern Fennoscandia (Holmboe, 1936). Two particular groups, likewise western Scandinavian, among whose mem- bers Wurm hibernators might be presumed, are the species concentrated in southwestern or southeastern Norway, some of which especially in the former group, have a very small area. Its history is difficult to decide. Further informa- tion will be given in the following section on the disposition of Wurm refuges. e. The alpine-subalpine type, whose members are restricted to fjeld regions (and partly to tundra regions), has no representatives among the dimorphic species that might clearly reveal their history. In this connection, species oc- curring exclusively on the Kola Peninsula have been left out. We thus have three distribution groups: 1. Species distributed along the entire chain of fjelds: Agonum consimile Bembidion hasti. Amara alpina 2. More or less distinctly bicentric species: (Bembidion fellmanni) Dyschirius helléni (B. lapponicum) Nebria nivalis. (Bembidion siebkei) Also Elaphrus lapponicus, considered above (p. 741). Dyschirius angus- tatus belongs partly geographically (but not regionally) to this group. The phenomenon of bicentricity is further discussed below (p. 752). 3. Northern Fennoscandian species: Bembidion dauricum B. scandicum B. grapeioides Diachila arctica B. hyperboraeorum Nebria gyllenhali balbii. These 14 (16) species are undoubtedly Wurm hibernators. Only one species, Amara alpina, seems to have an uninterrupted connection with regions east of the Wurm ice, eastward through the Kola Peninsula. They cannot be southern postglacial immigrants for reasons discussed above (p. 710). Other facts worth mentioning are: There are interglacial subfossils of Agonum consimile, Amara alpina, Bembidion hasti, and Diachila arctica from 749 750 654 Sweden (Lindroth, 1948a). Nebria nivalis (as a boreo-British species) has been found in Scotland—like Amara alpina—and hence must have lived during the interglacial period in Scandinavia (Lindroth, 1935a). Among the 6 strictly northern Fennoscandian species that could most plausibly be considered as postglacial immigrants (from the northeast), two are dimorphic. However, Bembidion dauricum occurs in our region exclusively in the brachypterous form, B. grapeioides predominantly in this form. B. scandicum is unknown outside the region and is the only species of carabid that is probably endemic to Fennoscandia. f. Bembidion schüppeli and Dyschirius septentrionum. It is interesting that two species with such a complicated area map can coincide in all impor- tant features. The only important differences are, on the one hand, that only Bembidion schüppeli occurs in Denmark (Jutland, undoubtedly as an inter- glacial relict) and, on the other hand, that Dyschirius septentrionum is un- known at the White Sea, probably for lack of exploration. The ecology of the two species is largely identical (see Part I) and their history must have been the same. In the southeast both species have a postglacial stock in Estonia and south- ern Finland, north as far as latitude 62°30’ N (Dyschirius) or 63°N. North of this there is a stock extending across northern Finland, which in the case of Bembidion (certainly also in Dyschirius) emanates from the east—from the White Sea. This immigration group, discussed above (p. 729), reached the Bothnian coastland and both species spread on the Swedish side south as far as Mdp; on the Finnish side only Dyschirius extends that far. Of the two stocks described, only in Bembidion is the western population, which hibernated in Norway, sharply separated (the single record of Lyl Gaskelought belongs here); in Dyschirius this western stock merged with the northeastern stock only in the far north. The difference is easily understood, since only Dyschirius is func- tionally macropterous. For the same reason all Swedish localities of Bembidion (with the exception of those in Lyl) are below the highest shoreline; they are so to speak coastal relicts and indicate an early immigration. Dyschirius advanced farther inland. The localities in Vrm certainly also belong to the western area. g. Finally, as Notiophilus aquaticus type may be named the pan- Fennoscandian species that have had the same colorful history as the above-mentioned species. In my opinion this is true of all of them. The area of a Species that is distributed throughout Fennoscandia (considered geographically, not regionally) without perceptible gaps, cannot just be the result of postglacial immigration into the region or of hibernation within the region. We saw earlier what a short distance, with very few exceptions, the southern postglacial stock has advanced even in Sweden (p. 716) and the eastern one even in Finland (p. 718). There is no evidence to show that the 751 655 entire Norwegian coastland could have been colonized by any species that immigrated only postglacially. The most important demonstrable achievement of this kind seems to have been that of Calathus erratus (p. 377, Fig. 35), but it occurs in northern Norway only very locally. Moreover there are astonishingly few “pan-Fennoscandian” species that really deserve the name, having reached the outermost limits of the Fennoscan- dian mainland in all directions. Most of the suitable species are missing along the north coast of the Kola Peninsula and in the farthest north of Nor- way. Besides Notiophilus aquaticus, strictly speaking there are only 2 pan- Fennoscandian carabids (br—brachypterous; in parentheses—dimorphic but macropterous specimens very rare, hence functionally brachypterous): (br) Calathus melanocephalus br Patrobus assimilis. If we relax the stipulation of uninterrupted distribution in the farthest north, the following species may also be called “pan-Fennoscandian”: (br) Agonum fuliginosum Elaphrus cupreus Amara apricaria E. riparius Bembidion bipunctatum Harpalus quadripunctatus B. rupestre Loricera pilicornis br Calathus micropterus (br) Notiophilus germinyi br Carabus glabratus (br) Pterostichus diligens Clivina fossor P. nigrita br Dyschirius globosus Trechus rubens. Other species have of course been found in all the “major regions” of Fennoscandia, but by the pronounced even if small gaps or other irregularities of distribution show that they have a double or multiple origin, and that the various stocks have still not merged. Amara brunnea. Gap in northern Finland (north of the Arctic Circle). A. praetermissa. Gap in central Sweden and on the Finnish west coast. A. quenseli. Large lacunae in southern Sweden and central Finland, partly (but not consistently) as boundary against “forma silvicola”. Bembidion saxatile. See map (Fig. 116, p. 801). B. velox. Large lacunae in southern Sweden and along the Norwegian coast. Concentration in northern Finland. br Carabus violaceus. Gaps in central and northernmost Finland. br Cychrus caraboides. Gaps in the inland of the north and at the Finnish west coast. (br) Cymindis vaporariorum. Small but probably real gap in Jtl. Dichirotrichus pubescens. (Only on sea with high salinity.) Gap on the Kola Peninsula. Miscodera arctica. As in Cymindis vaporariorum. Trichocellus cognatus. As in Cymindis vaporariorum. 752 656 It is most noteworthy that of these 31 widely distributed carabids of Fenno- scandia (including Notiophilus aquaticus and N. biguttatus) 13, i.e. 42%, are functionally brachypterous. The corresponding figure for the Fennoscandian fauna as a whole, even if we consider all dimorphic species as brachypterous, is 99 species, i.e. 27%. The most widely distributed species in Fennoscandia generally have a poorer capability of dispersal! This is in complete contrast with the result (p. 435, Diagram 47) obtained by evaluating the relationship - between flight capacity and total distribution of our species. It is the refuges, with selection in the Wurm glaciation operating in favor of brachypterism, that are responsible for this converse situation in Fennoscandia. I do not claim that every species mentioned in this section (pp. 738 ff.) must represent a Fennoscandian Wurm hibernator. Undoubtedly it will be more prudent and objective to express the situation as follows: For each of these species the problem of a Wurm hibernation should at least be seriously discussed. Table 38 (p. 802) provides a detailed survey not only of the species in question but also of the parts of Fennoscandia where the former refuge of each species might be envisaged. II. Location of the Fennoscandian Wurm refuges has been attempted by Nordhagen in particular (1933, 1935). However, earlier A.M. Hansen (1904, e.g. p. 299), Wille (1905), and especially Th. Fries (1913, p. 312) had already designated definite regions in southern and northern Norway that could be considered as refuges. According to these two Swedish authors, at least in the south the nunatakt regions, which were considered completely free from ice, would have played an important role besides any ice-free coastal stretches. In locating the northern Wurm refuges Th. Fries (1913) was able to draw on geological facts (Vogt, 1912; Enquist, 1913), but his most important source was the peculiar distribution of certain fjeld plants, and especially the marked bicentricity of some of them. As the word signifies, the area of a bicentric species has two “centers”—in the present case in the northern and in the southern part of the Scandinavian fjeld range—with a more or less broad gap or “zone of obliteration” in between. Species that occur in only one of these re- gions may be called “northern or southern unicentric” (Arwidsson, 1928). Col- lectively all the distribution types belonging to any of these categories may be called “centric” (Nannfeldt, 1947, p. 56). Two markedly bicentric species of dif- ferent types are Campanula uniflora (Fig. 107) and Saxifraga aizoon (Fig. 108). Simplocaria metallica is (p. 738; Fig. 106) a clearly bicentric beetle. Among the carabids there is no other case so clear. Nevertheless, bicentricity may be t (cf. p. 736; suppl. scient. edit.). 753 Fig. 107. Campanula uniflora. (According to Arwidsson, 1943). 138 756 658 perceived in the 12 species listed above (p. 743), at least in the form of a small “zone of obliteration” in parts of Jtl. In explanation of bicentricity there can be only one of the two causes: Either the area of the species concerned has split up under the influence of existence factors, or the gap is due to dynamical characteristics of the species, i.e. it is historically determined. a. The view that the gap of a bicentric species may be due to climate (other existence factors scarcely figure), was argued early on (Th. Fries, 1913, p. 317; Tengwall, 1913, p. 268) and thereafter repeatedly cited as a useful reference. The “zone of obliteration” of the bicentric species always more or less coin- cides with the less pronounced high alpine centra! part of the Fennoscandian fjeld range, which, as clearly shown in the map in Fig. 61 (p. 437), is traversed by numerous wooded passes. It is quite possible that a species like Nebria nivalis, which is bound to the margins of perennial snowdrifts, would be ex- cluded from the “gap” for climatic reasons (Lindroth, 1939a, p. 250; cf. also the temperature maps in Figs. 63-72, pp. 452 ff.). These two authors (Th. Fries, Tengwall, l.c.) realized that the postglacial warm period may have fatefully affected the true alpine organisms in the lower central parts of the Scandinavian fjeld range. Later investigations by Hagem (1917) and others, but chiefly by Smith (1920, pp. 120 ff.) showed that the forest—mainly Pinus—extended into the fjelds up to 300 m (in southern Nor- way even up to 500 m, according to Hagem, (1917, p. 167) higher than now, a feature very clearly reconstructed in the central parts concerned (Hjd, Jtl). it is therefore conceivable that high alpine organisms in this region were de- stroyed during the warm period, and that thereafter there was no possibility of a recolonization of them. b. On the other hand the above viewpoint cannot hold for non-high alpine or subalpine organisms or those living at still lower altitudes. This question was taken up by Bjorkman (1939). As a convincing example of a subalpine bicentric species he mapped the distribution of Luzula parviflora (l.c., p. 206). The Scandinavian fauna has only one carabid species, Nebria nivalis, so cold- requiring that the bicentricity may be due to climatic factors (today’s or those of the warm period). That such an interpretation is impossible for the “model species” Simplocaria metallica too is seen at a glance from its distribution map (Fig. 106). We therefore conclude that the more or less pronounced bicentricity in the Scandinavian distribution map of a number of carabids—with the possible exception of only Nebria nivalis—is not due to existence factors but to the history of the species. It would be natural to interpret the subarea south or north of the gap as the result of a southern or northeastern postglacial immigration. But we have already (p. 713) declared a postglacial immigration of alpine-subalpine animals and plants from the south virtually impossible, which applies especially 659 NORGE oc SVERIGE Saxifraga Aizoon A (c Riinpy NS \ Re os 7 3 u g > .d } auf - z = = “ » J re g e N { 4 f = cf ve b ) i ) 5 j \ SB 5 # IT Kg 7 N x 1 q a a on Mat a 8 a = Pale. bm 4 x g 7 RY, yy te. r ß ei LAF x i % ER C A sf = Nn os —— = (67, 5 > | as x 2 pears y 4 Y “ay 2 = N c 4 2 y ee SS = > N 6h = ee a & > ats ay. ” Vy NY oS Fig. 108. Saxifraga aizoon. (According to Holmboe, 1937). 754 TY 660 to Bembidion fellmanni and Nebria nivalis. This is supported by the following facts. Species completely missing from Central Europe: Bembidion siebkei Elaphrus lapponicus B. virens Nebria nivalis. Dyschirius helléni A southern postglacial stock in Scandinavia, separated from the south Scandinavian “center” by a more or less distinct gap, is shown by: Cicindela maritima Trichocellus cognatus. Dyschirius angustatus Amara nigricornis is not bicentric but “tricentric,” as far as can be judged. The distribution of the races of Carabus problematicus (map in Fig. 1, p. 21) shows that the southern postglacial stock is functionally separated from that of the southern “center” (wockei). Only Miscodera arctica is left, whose southern contingent (down to sou- thern Jtl) could well be the result of a southern postglacial immigration. I of course believe that the frequency and abundance of this species in the south Norwegian mountains (Fig. 56, p. 424) reveal a center determined not climatically but in the main historically. Postulation of a northeastern (or southeastern) postglacial origin for the stock in the northern center of a bicentric species is ruled out by the fact that the area of many of these lacks any eastward connection, as far as is known (the Kola Peninsula is well explored!), or it is only poorly formed. (Bembidion fellmanni) D. helleni B. siebkei Elaphrus lapponicus (B. virens) (Miscodera arctica) Carabus problematicus (Nebria nivalıs) Cicindela maritima (Trichocellus cognatus). Dyschirius angustatus Hence one can envisage at the most in Amara nigricornis and Cymindis vaporariorum, that the northern Fennoscandian “center” resulted from post- glacial immigration from the cast (northeast). However, the Cymindis species is functionally brachypterous and hence relatively slow to disperse. We have found that the 12 more or less distinctly bicentric carabids of Fennoscandia cannot have achieved this characteristic of their area by postglacial immigration from the south, and from the northeast at most in two cases. The experience of botanists that bicentricity in the Fennoscandian mountains is to be attributed to a Würm hibernation in two separate refuge regions has general validity for the carabids (and other organisms) as well, a rule with more than solitary exceptions only among organisms that disperse very easily (see section on anemochorous dispersal, pp. 548 ff.). 758 661 Of the 12 carabids considered, 4 have never been found in the Regio alpina (Table 30, p. 440). They can therefore be termed at the most “subalpine;” they form good counterparts to Luzula parviflora (Bjorkman, 1939). For particular conclusions that can be drawn from this fact, see below (pp. 776 ff.). The bicentric species allow us to identify two main refuge regions in Fennoscandia, one in the southwest and the other in the far north. Can these prospective large regions be more precisely located and perhaps divided into definite small refuges? The attempt was made by Nordhagen on the basis of the disjunct distri- bution of the Scandinavian Papaver (1931, 1933, pp. 42 ff.), especially in his contribution of 1935 based on other fjeld plants as well. He thinks a whole series of definite relict species still grow only at localities where they must be assumed to have survived the Wurm glaciation (or in their vicinity). In respect of North America, Fernald earlier drew attention to this “persistence” of some species of plants and to this conservative affinity for old localities. Later Hultén (1937, p. 22) used the term “rigid species” for them. This rigid conservatism of some species in respect of locality is actually a real mystery (also emphasized by Holmboe, 1937, p. 28). It is not explained by characterizing it, as Fernald does (for instance, 1925, p. 336; 1929, p. 1493), as “ancient,” “old,” “conservative,” “unaggressive,” “nearly extinct.” It would be very useful to run accurate biological- and experimental-tests on several such “rigid” species of plant. It can be envisaged that their “conser- vatism” is due to one (or more) of the following four groups of characteristics: 1. Poor capability of dispersal.t The Leguminoseae, for instance, yield diaspores whose transportation is difficult. Even seeds of Papaver are not easily transported passively over long stretches. One would be inclined to en- visage certain difficulties for the postglacial migration argued by Nordhagen, chiefly in the uphill direction from the south Norwegian coastal refuges (for instance, 1933, p. 46). Still, is it not conceivable that the hibernation of Papaver in southern Norway was at least partly on nunataks located farther inland? In Greenland a Papaver was found on the easternmost, most isolated nunatak of the “Jensen group” at a great height (Kornerup, 1890). 2. Strong ecological specialization (stenotopy). In the case of plants a depen- dence on limestone is evident here, as is emphasized repeatedly by Nordhagen (for instance, 1935, pp. 58, 92, 122), Holmboe (1937, p. 28) and others. It has adversely affected not only the “choice” of refuge but also the postglacial dis- persal. According to Nordhagen the Wurm hibernators among plants are gen- erally more or less pronounced ecological specialists. Wynne-Edwards (1939) sees in this the whole explanation of their restricted distribution, which is certainly exaggerated. tcf. p- 823; suppl. scient. edit.). 759 760 662 3. Poor competitiveness. This characteristic is covered by Fernald’s des- ignation “unaggressive species,’ and great importance is attached to it by Nordhagen (for instance, 1935, p. 121). 4. Reduced reproductive ability. This on account of a sharp decrease in isolated populations during unfavorable periods, chiefly during the Wurm hibernation. This purely quantitative variation can be combined with the fix- ation of disadvantageous mutations in Sewall Wright’s sense (see p. 366).* The above observations are not intended to detract from the importance of the distribution of plants as circumstantial evidence of a Wurm hibernation or a more precise fixation of refuges. They only have to indicate that even in the case of species generally recognized as Wurm hibernators, more investigations are necessary for us to formulate a more precise history. In the case of our objects of study, the carabids, some of the conditions are simpler. Of the four above-mentioned groups of factors, competitiveness should of course be discounted (p. 554), and the dependence on limestone has at most a minor role (pp. 195 ff.). How far a decline in populations within the Wurm refuges adversely affected the postglacial capability of dispersal, I certainly do not venture to declare in any particular case, but judge such an effect to be highly probable. Undoubtedly the capability of dispersal through existing or lacking flight capacity has determined the size of the area colonized postglacially from the Wurm refuges, as has been shown chiefly by the study of dimorphic cara- bids (pp. 335 ff.). In this way we can also help to contribute to the botanists knowledge in locating Wurm refuges more precisely. In the far north a refuge was assumed somewhere on the White Sea (p. 729) on the basis of entomological evidence. Conditions in the Finnish- Norwegian border region on the Arctic Sea coast point in the same direction (Petsamo-South Varanger). Nordhagen (1935, p. 130) thinks the assumption of a refuge on (or in the region of) the Fischer Peninsula is (botanically) use- ful and (geologically) possible. The distribution of the dimorphic Bembidion transparens supports this strongly (p. 389; Fig. 45). Moreover there is a whole series of carabids, whose distribution in the far north is most easily explained by postulating a refuge in the Petsamo-South Varanger region. Either they occur there more or less isolatedly or they are particularly frequent, which is not due only to the very thorough exploration of these regions. Examples are [br = brachypterous, (br) = dimorphic, but in the region concerned only or predominantly in the brachypterous form]: Agonum consimile (br) Bembidion grapeioides (br) A. fuliginosum B. saxatile Amara nigricornis B. velox *In this connection M. Fries (1949, p. 47) also speaks of “gene impoverishment” (Swedish: utarmande av anlag). 761 663 br Carabus glabratus Elaphrus cupreus br C. problematicus E. riparius ? Cicindela maritima Nebria gyllenhali balbii br Cychrus caraboides (br) Notiophilus reitteri Diachila arctica (br) Pterostichus diligens Dichirotrichus pubescens Trichocellus cognatus. Dyschirius septentrionum According to data supplied by Tanner (1937, pp. 104 ff.) it seems necessary to assume that the supposed refuge south of the Varanger fjord was located below the present sea level. The existence of a Wurm refuge on the Varanger Peninsula—or mainly on the present-day land below sea level to the north—is well substantiated geo- logically (Holtedal, 1929) and botanically (Nordhagen, 1933, pp. 69 ff.; 1935, pp. 116 ff.). However, entomologically these regions are so poorly explored that the question cannot be further discussed here. The same is true of the assumed small refuges in Mageroy and at the mouth of the Porsanger fjord (Nordhagen, 1935, pp. 84 ff.). According to Nordhagen (1936, p. 113) the Mageroy refuge is now geologically proven as well. The coleopteran fauna here is extremely poor; from Mageröy, for in- stance, only 19 carabid species are known, and only in the case of Carabus problematicus can a Wurm hibernation in situ be assumed. The largest Fennoscandian refuge region during the Wurm period is ge- nerally considered to be the coastal regions (today partly below sea level) from the estuarine zone of the Alta fjord (latitude 71°N), south to about the Arctic Circle (see, for instance, Fig. 106, p. 739). We may cite a whole series of geological evidences to show that the edge of the maximal Wurm ice in some places, like the Lofoten islands, did not even reach the present-day outermost coastline (Vogt, 1912, pp. 6, 47; Enquist, 1913, 1918, pp. 5 ff.; Ahlmann, 1919, pp- 217, 238; Grönlie, 1927, p. 56; Nordhagen, 1933, pp. 20 ff.; 1935, pp. 136 ff.; Undäs, 1939, pp. 181 ff.). On the well-argued assumption that the sea level at the Wurm maximum was much lower than today (see also below), the extensive ice-free coastal regions along the above-mentioned stretch—as also numerous nunataks'—must have been available to the fauna and flora. The carabids that must have survived the Wurm period in this extensive refuge region form an imposing series. The following clear examples may be mentioned: Asaphidion pallipes B. hyperboraeorum br Bembidion dauricum B. lapponicum B. femoratum Bembidion saxatile (br) B. grapei B. scandicum (br) B. grapeioides (br) B. schüppeli tcf. p. 736; suppl. scient. edit.). 762 664 B. siebkei Elaphrus lapponicus (br) B. transparens Harpalus quadripunctatus (br) Bradycellus collaris H. winkleri br Carabus nitens br Leistus ferrugineus br C. violaceus Miscodera arctica br Cychrus caraboides Nebria gyllenhali balbii (br) Cymindis vaporariorum N. nivalis Dichirotrichus pubescens (br) Pterostichus strenuus Dyschirius angustatus br Trechus obtusus br D. helleni Trichocellus cognatus D. septentrionum T. placidus. It seems hardly possible to pass a more precise judgment on which small refuge one or other species had hibernated. In the case of Bembidion dauricum and Trechus obtusus, both of which are constantly brachypterous, it must have been situated (completely or partly) in the region of the Lofoten archipelago. On the other hand the carabids may contribute to fix the southern limit of this northern “major refuge.” Originally Nordhagen (1933, p. 54) did not believe the southern refuges in northern Norway (at any rate not the floristi- cally important refuges) to be situated more south than about the Arctic Circle (Svartis region); later (1935, pp. 139 ff.), in view of the southern record of Are- naria humifusa, he assumed small refuges to the south as far as Leka (latitude 65°N). The hatched region on my map (Fig. 50, p. 402; Fig. 106, p. 739) should therefore have, according to Nordhagen, a more southward extension. But apparently a further adjustment in the same direction must be under- taken. The distribution of Bembidion aeneum (Fig. 49, p. 400), a dimorphic species, is obvious in this connection. lt is known exclusively in the brachypter- ous form in its totally isolated west Norwegian area. Most of these record lo- calities are in the gap between the “major refuges,” and it is incomprehensible what should have caused this coastal species to migrate from the hibernation localities of the Wurm period. It would be far-fetched, for instance, to consider the southern subarea (in Trondelag) as originating from the More refuge. The two isolated west Norwegian record localities of B. minimum are situated on the Trondheim fjord as well. The southernmost localities of B. lapponicum, and to some extent the distribution of the flightless Trechus obtusus, point in the same direction, i.e. the assumption that there were coastal Wurm refuges even in the “gap” between latitudes 63°30’ and 66° N. The geological evi- dence may also be cited in this connection. Both Grönlie (1927, p. 56) and Granlund and Lundqvist (1936, pp. 13-14) think that the outer islands of the Donna group (latitude about 66° N) were not glaciated during the Wurm. According to Undas (1934, pp. 55-57) the terminal moraines of the maximal Würm ice even in Tröndelag are situated on the mainland (Orlandet, north- west of Trondheim). 763 665 In southern Norway—south of the Trondheim fjord—three coastal refuges were assumed by Nordhagen (1933, p. 46); (a) More, (b) Sogn, (c) Ryfylke (Fig. 106, p. 739). In the first case he was able to rely on Kaldhol’s geo- logical findings (1930, pp. 96 ff.; 1931), according to which at least the out- ermost present-day coastal zone was not glaciated during the Würm period. The assumption of the two southern refuges was based on botanical evidence alone. According to Faegri (1940, p. 19) Jaeren seems to have been completely glaciated during the Würm period, only the island of Utsira may have been ice-free. For southern Norway it is more difficult than for the far north to distin- guish the “hibernators” with confidence, since in the case of species of the plains there has often been a secondary merging with postglacial stocks*; or the Scandinavian population as a whole may have immigrated postglacially. Only in exceptional cases is it possible to define the location of the actual refuge more precisely. As examples of hibernators somewhere in southern Norway (south of latitude 64° N) the following may be mentioned: Amara praetermissa Elaphrus lapponicus A. quenseli Miscodera arctica Bembidion fellmanni Nebria nivalis Carabus coriaceus Patrobus septentrionis. Cymindis vaporariorum A few species with characteristic distribution offer an indication as to where their former Wurm refuge (or that of several species) was situated. a) In More, or at any rate in the coastal region between latitude 62° and 64° N. Aépus marinus Carabus problematicus Bembidion grapei (Fig. 50, p. 402) Dyschirius helléni B. lunatum Trechus fulvus. B. siebkei It is possible that a few other species, of the Bembidion argenteolum type, have had the same history of hibernation (see p. 769). In my opinion this possibility must also be seriously considered for the oddly distributed rodent Sicista subtilis (cf. Ekman, 1922, pp. 206 ff.). b) In the outer Sogn. The only species whose north Norwegian distribution points to this refuge is Nebria nivalis. It is noteworthy, on one hand, that it represents the only pronounced high alpine Fennoscandian carabid and, on the other hand, that the prime high alpine Fennoscandian fjeld region, Jotunheimen, is located just at the inner end of the Sogne fjord. * Without race differences, for instance, it would have been difficult to distinguish the south- ern interglacial stock of Carabus problematicus (Fig. 1, p. 21). 764 666 In my opinion there is a causal connection here. Nordhagen assigned his refuges according to the distribution of alpine plants, in the case of Sogn almost exclusively in accordance with the isolated occurrence of Papaver relictum in Valders and along the inner Sogn (1936, p. 110). The essence of the especially disjunct distribution of the species he chose, is their poor capability of dispersal (p. 758). That they continue to live in their isolated relict localities, is the result of a long series of fortuitous circumstances. But especially these two: 1. During the last interglacial they must have lived in the vicinity of a developing refuge. 2. During the postglacial period they must have found a suitable biotope close to their refuge. Both these facts are a matter of course. They lead to an equally simple conclusion: Distinct fjeld plants (of the “rigid” type) “hibernated” close to the present-day fjelds. In Scandinavia they may have already become “cen- tric” (bi- or unicentric) during the last, partly quite warm (p. 673) interglacial period. This means that just as many (or more!) Wurm refuges for pronounced alpine plants and animals are dependent on conditions before and after the hibernation, as from conditions in the glacial epoch itself. We cannot have a thorough understanding of the Wurm refuges by a study of the alpine orga- nisms alone. Species like Bembidion aeneum, which can still live in situ within the limits of the refuge, are independent of the conditions in the adjacent inland. Their distribution is determined by the critical period of the glaciation. Nebria nivalis and Papaver relictum, which must move with every change in the glacial situation, go through critical times during the transitional periods, and their distribution is the result of a favorable interaction between alterna- ting refuges of opposite kinds: glacial refuges against interglacial refuges. If a “Wurm botanist,” equipped with sufficient geological knowledge, had under- taken an expedition along the ice-free stretches of the Norwegian coast, on discovering a rare Papaver or Arenaria humifusa he would have asked himself: “How has the poor plant been able to survive the severe interglacial period?” I thus mean, that probably even along the southern half of the Norwe- gian west coast—as along the north coast—rather than a limited number of large Wurm refuges (three according to Nordhagen) there must have been a whole series of small, ice-free stretches of land, more or less isolated from one another along the outer coastal belt. However, for reasons cited above, only a few were favorably situated in every way for the most pronounced “rigid” alpine plants to survive until today—their hibernation on the nunataks! far- ther inland excluded. In the case of Nebria nivalis at any rate hibernation on the nunataks in Jotunheimen appears to me more probable than in a coastal region at the mouth of the Sogne fjord. 667 765 c) In the extreme southwest, e.g. in Ryfylke. The most eminent botanical index is Saxifraga aizoon (Fig. 108, p. 754; Nordhagen, 1933, p. 53; Holmboe, 1937, P%29). A counterpart among the carabids is Bembidion tibiale, which in the whole of Fennoscandia has been found in only four localities close together in Ry- fylke, where it is encountered constantly and partly abundant. Netolitzky (1929, p. 35) suggested that the isolated Norwegian occurrence of this species resulted from transport down the extended lower course of the Rhine and Elbe across the early postglacial “Dogger-land.” This theory cannot be rejected out of hand, but it loses credibility for the following reasons: 1. “The insect is bound to a particular substratum, just as the phy- tophagous ones are bound to particular plants” (Netolitzky, l.c.), and this substratum consists of coarse material (rubble). For this reason the species is now missing from the lower course of the big West German rivers and nowhere reaches the North Sea coast (Horion, 1941, p. 129). It is still less likely that the species would be able to find suitable biotopes along the banks of the same rivers in the flat Dogger-land, where the gentle current was able to carry and deposit only the finest particles (silt). 2. The exact northward extent of the Dogger-land is not known. But in the postglacial period it cannot have been of the size Lewis pictured it to be (1935, p. 337), following the 85-fathom line. At any rate it was separated from Norway by the present-day Norwegian channel as a broad, deeply incised bay. If, in spite of this, Bembidion tibiale was able to cross this barrier (the insect is capable of flight) it would rather be expected on the banks of the southern Norwegian rivers. 3. Bembidion tibiale also occurs on the British Isles, even in Ireland, and it cannot have originated in the postglacial period. Like the “boreo-British” species (Lindroth, 1935a), B. tibiale indicates an older (interglacial) faunal connection between Scandinavia and the British Isles. On the basis of the above discussion, Bembidion tibiale is to be considered as a Wurm hibernator in southwestern Norway. The situation is very different in the case of Bembidion harpaloides, which is known in Fennoscandia only from two localities: (2 specimens) in the extreme 766 south of Norway, which is exactly what one would expect following the dispersal from the Dogger-land. Besides, this species even today extends north as far as Hamburg on the North Sea coast, it was found even (as a “Dogger-land relict”?) on Helgoland (Netolitzky, 1916). The contrary relationship between the two Bembidion species considered illustrates the great difficulty encountered in judging the specific southwest Norwegian faunal and floral element of Scandinavia. The possibility of a post- glacial immigration across the sea must not be underestimated— whether in the Dogger-land period or later. In Calathus mollis (p. 369; Fig. 28) we had a clear example of this. 767 668 The following species may have reached southwestern Norway as “direct” postglacial immigrants: Agonum marginatum Cymindis macularis Amara lucida Dromius angustus A. quenseli (“silvicola’’) Dyschirius impunctipennis A. spreta D. obscurus Bembidion pallidipenne D. politus Bradycellus harpalinus Nebria livida Calathus ambiguus Pterostichus aterrimus. With the exception of the Pterostichus species, which in our region often occurs as an accidental migrant, all these species live more or less regularly along the sea as littoral or epilittoral species. A common characteristic is fur- thermore that they are winged, with the exception of the dimorphic (normally brachypterous) Cymindis macularis, which despite this we earlier (p. 287) found to be an insect with unusually strong capability of dispersal. Among the Wurm hibernators that survived the last glaciation in more northerly regions many (probably most) have a similar history in southwest- ern Norway. In the following species their long presence in the region seems clear from the recent distribution, partly according to the “dimorphic maps,” (pp. 389 ff.) [br = constantly brachypterous; (br) = dimorphic, in the region concerned also in the brachypterous form]: br Aépus marinus (br) Notiophilus aquaticus Amara nigricornis (br) N. biguttatus Bembidion fellmanni (br) N. germinyi (br) B. grapei Pterostichus adstrictus B. hasti (br) P. strenuus br Carabus glabratus br Trechus fulvus br C. problematicus Trichocellus cognatus. br Leistus rufescens It is striking that two-thirds of the species are functionally brachypterous. Of much greater interest is the question: To what extent have species hi- bernated only in the southernmost Würm refuge of Fennoscandia? It is most probable that these should include the thermally most demanding members of the hibernation group, and that the evidence of their character as interglacial relicts may permit a judgment on climatic conditions during the Würm maxi- mum. That question is discussed below (p. 791), and we will restrict ourselves here to the fairly clear cases. Among the carabids, Bembidion tibiale (see below) is the best example. But the refuge character of the northern European area of Chrysomela crassi- cornis Hell. is still more clear-cut (Fig. 109). It too is restricted to the extreme southwest of Norway. It is a sluggish, soil-bound insect with rudimentary wings, 669 whose total area consists of small, scattered relict occurrences (Holdhaus and Lindroth, 1939, p. 206). It is inclined to the formation of races, since even the form native to the British Isles (including the Shetlands) is, according to Franz (1938), subspecific, different from the Norwegian forma typica. The small, con- centrated Norwegian subarea, where the species is abundant at places, must be ascribed to a Wurm refuge situated in this region. It is interesting to know ' that according to Faegri (1940, p. 19) the island of Utsira (west of Karmoy) may have remained ice-free during the Wurm. Among the carabids the following species may also be involved [br = constantly brachypterous; (br) = dimorphic, in the region concerned also oc- curring in the brachypterous form]: Agonum ruficorne Nebria brevicollis (br) Bembidion assimile (Fig. 46, p. 394) N. salina (br) Calathus erratus (Fig. 35, p. 376) (br) Notiophilus palustris br C._ fuscipes (br) Olisthopus rotundatus (Fig. 32, p. 373) (br) Carabus clathratus (Fig. 38, p. 380) (br) Pterostichus lepidus (Fig. 37, p. 379) 769 br C. granulatus (br) P. minor (Fig. 43, p- 388). In the case of 7 dimorphic species the exclusive or predominant occurrence of the brachypterous form suggests that they occupied a refuge here. The three winged species occur more or less isolated in southwestern Norway. Of course, in connection with Nebria salina it was conceded earlier (p. 475) that present- day climatic factors can operate as in area limiting effect. Before summarizing the results of the attempt to locate the Fennoscandian Würm refuges we must consider a group of species that have so far been ignored, of which Bembidion argenteolum may be taken as a typical example. It is characteristic of this species that the center of its area is situated in the eastern part of the southern Norway. Hence its predominantly western distri- bution in Scandinavia is not the result of a predilection for an oceanic climate, which moreover is clear from its occurrence in the Baltic region southward down to the Karelian Isthmus. Ecologically in Scandinavia the species seems to be associated with the banks of the larger rivers (although also found on Lake Siljan in Dir). But on the German Baltic Sea coast and in the Karelian Isthmus it also lives along the seashore (see map in Netolitzky and Meyer, 1933). On climatic and edaphic grounds it coulc ~o doubt live equally well on the duny shores of Ska, southern HIl or on Jutland. The noteworthy gap south of the Scandinavian area, which includes the whole of Denmark is therefore histor- ically determined and precludes a postglacial immigration from the south. Bembidion argenteolum in Scandinavia is an indubitable Wurm hiberna- tor, much as in the British Isles, where the species occurs exclusively along 770 670 Lough Neagh in Ireland (concerning the erroneous report from England, see Lindroth, 1939a, p. 258). Like the remaining Chrysobracteon species it is a very conservative species (p. 590), which confines itself tenaciously to old habitable regions. But where was its Wurm refuge located? The occurrence in Trondelag and in the upper part of the Gudbrands valley—among the generally accepted refuge regions of Norway—could indicate the coastal stretch between latitudes 62° and 64°N (hence the “More refuge”). B. semipunctatum has a Fennoscandian distribution nearly identical to that of B. argenteolum. The only difference is that with B. semipunctatum a postglacial stock reached Bornholm and the southern half of Ska. However, there the species nevertheless has the character of a vagrant migrant, which does not seem to live permanently at any locality. On the other hand in the case of Bembidion litorale and Cicindela maritima the postglacial stock gained a firm footing in southern Sweden. But it is clear that the occurrence in central Scandinavia cannot be the result of an immigra- tion by this route. The isolated Swedish occurrence in central Norrland must be explained, among other things, by an emigration from Trondelag. Especially in the case of B. litorale any other explanation seems implausible. We earlier (p. 601) emphasized the importance of hydrochorous transport by the large rivers. For instance, we explained the eastward dispersal of Bem- bidion virens in this way. In this species, which lives on coarse gravelly shores, there is a still clear connection between the localities at the Bothnian Sea and the Norwegian main area. On the other hand Bembidion litorale and Cicindela maritima are restricted to sandy and fine-sandy shores, which are not found on the upper course of the rivers concerned (Ljungan, Indals, Angerman). The enclaves in the Bothnian coastland therefore lack all connection with Nor- way today, whence they undoubtedly came, and we get a cartographic picture which in this context largely agrees with that of Arabis petraea (Holmboe, 1937, p. 23), a plant that has had the same history here. It is possible that the iso- lation catastrophes caused by the drainage of the large reservoirs of Jamtland at the end of the Finiglacial period caused the possibility of dispersal. Both for Cicindela maritima and Bembidion litorale we are therefore in- clined to assume a center of dispersal, a Wurm refuge, in the Trondelag-More range. But how are we to envisage the gap in the case of the latter species between Trondelag and the more southern subarea (in southeastern Norway and Varmland)? The Gudbrands valley is one of the best explored parts of Norway and here the species is missing! Edaphically the river banks here are not unfavorable, as is evident from the distribution of the ecologically related species Cicindela maritima, Bembidion semipunctatum, B. lunatum, Asaphidion 771 pallipes, etc. I believe that Bembidion litorale here shows a historically deter- mined gap, and that this stock in southeastern Norway- Varmland has a special origin, separately from the postglacial stock. a 1455 kv-mil ry i=) Ei 1668 kvmil 1888 kvmil 2093 kv-mil > a 2452 kv-mil a a re 2281 kvmil B V. 0 Ö.längd fr.Greenw. 5 768 Fig. 109. Chrysomela crassicornis Helliesen. European distribution and the Norwegian subarea. (According to Franz, 1938, 1943a; Holdhaus and Lind- roth, 1939; and according to Franz, Hinton, and Holgersen, in litt.). Exact position of the locality in northern Scotland ( “Sutherland”) is not known. We 672 This assumption leads us to Perileptus areolatus. It must be conceded that this tiny insect easily escapes attention, and that the record in Hll somewhat strengthens the possibility of a postglacial immigration. But the Norwegian- central Swedish area, with its center in the vicinity of the Oslo fjord, altogether gives the impression of clear homogeneity. And why is the species missing not only from Denmark but also from the entire north German plain (Horion, 1941, p. 173)? Suitable biotopest surely cannot be completely lacking. On the British Isles Perileptus lives exclusively in Scotland and Ireland, where it represents an unambiguous Wurm hibernator; in the latter island it was found on the seashore, among others in the company of Aépus (Halbert, 1937, p. 83). In my opinion Perileptus is a Wurm hibernator in Scandinavia too. But it does not occur within the limits of any of the generally accepted refuges considered above, and it cannot be that its recent area is due to emigration from any of these refuges. It is tempting to assume a refuge of the Wurm period in southeastern Norway, in the vicinity of the outer Oslo fjord (emphasized earlier by Lindroth, 1939a, p. 257). “And it is very tempting to derive support from the opinion of the Swedish geologist Astrid Cleve-Euler [1946, p. 91], according to whom the so-called ‘Raene,’ the great terminal moraines at the mouth of the Oslo fjord (see Nordhagen, 1933, p. 124, and other studies), were actually formed by Wurm ice at the time of its maximum extension” (/.c.). Of course this viewpoint might not be shared by other Swedish geologists. The distribution of the sluggish, flightless curculionid Otiorrhynchus salicis Strom (Lindroth, 1939a, p. 255) would be much easier to understand on this assumption. And in general the plants and animals peculiar to southeastern Norway should be re-examined from this viewpoint, among other carabids those mentioned above (p. 687); also species like Anthicus gracilis Panz. This is no place to do so. The hypothesis must obtain a stronger geological basis before we can begin a more detailed discussion. However, it may be pointed out that during the Wurm period the deep “Norwegian channel” at the coast, where undoubtedly large rivers emptied in these regions, may have caused an upward current of warmer water, which may have had a lococlima- tically ameliorating effect (cf. the remarks by Elfstrand, 1927). In conclusion the following remarks may be made on the location of the Wurm refuges in Fennoscandia: For high alpine plants and other organisms, hibernation is conceivable on the inland nunatakstt surrounded on all sides by ice. In the present-day Fennoscandian higher Regio alpina only 5 carabids are endemic, of which at least in southern Norway nunatak hibernation is possible only for Nebria ni- valis. In the Alps the nunataks have apparently played a much greater biolog- ical role (Lindroth, 1941, p. 439; Franz, 1943, pp. 505 ff.). tbiotypes” in the original, evidently a printing error—Translator. tt(cf. p. 736; suppl. scient. edit.). 773 673 Otherwise the Würm refuges were located in coastal regions. The two easternmost refuges, at the White Sea and in the Petsamo-South Varanger region, are well substantiated by the distribution of carabids, in the former case entirely based on it. In the extreme north botanical evidence indicates refuges in regions of the Varanger Peninsula and Mageröy. Along the entire Norwegian west coast, from latitude 71° N to Jaeren, during the Würm period there seem to have been a whole series of big and small refuges. The difficulty in locating them exactly is that most of them were below the present-day sea level. Using diverse premises the geologists, climatologists, etc. have come to the conclusion that the level of the ocean during the Great Ice Age must have sunk on account of so-called eustatic movement. Attempts have been made to estimate the vertical difference, chiefly by calculating the volume of water stored on the earth as ice (and so removed from the ocean). The figures obtained for the last glaciation are highly variable, for instance, between 88 and 93 m (Antevs, 1928, p. 81) and < 275 m (Ramsay, 1930; but his figures seem to relate to ihe Greatest Ice Age); Enquist (according to Du Rietz, 1935, p- 228) reckons at least 200 m. Most of the estimates are around 100-150 m (Holtedahl, 1929, pp. 7, 9; Tanner, 1930, p. 298; Zeuner, 1945, p. 251; 1946, p. 129). This is meant to indicate the actual regression in comparison with the present-day shoreline and the secondary transgression resulting from relief from the burden and elevation of the sea floor has been subtracted. Farrington (1945) obtained similar figures (120 m) while studying postglacial changes in the depth of the sea in Western Europe. The isostatic movements, which for Fennoscandia as a whole signified a sinking under the pressure of inland ice, exercised little influence on the—in comparison with the ice cover—peripheral coastal regions in the extreme north and west, particularly in the central part of the west coast, in Trondeiag (Fig. 110). Moreover it must be remembered that the isostatic sinking that took place against the eustatic movement, on account of the inertia of the earth’s crust, reached its maximum much later than the greatest extension of the Wurm ice. We may be justified in reckoning about one-half the amount of the total isostatic sinking at that time (i.e. corresponding with the real postglacial elevation). Keeping in view the isostatic and eustatic movements the following esti- mates may be attempted with regard to the lowest position of the coastline at the time of the Wurm maximum: Fischer Peninsula in Petsamo: 75 m. North coast of Varanger Peninsula and Mageroy: 90 m. Outermost coastline of Söröya to Rost: 100 m. Coast between latitudes 67° and 63° N: 50-60 m. More: 75-90 m. Coast between Stad and Jaeren: 100 m. 774 776 674 These figures are probably minimal. Granlund and Lundqvist (1936, p. 14) estimated a level 200 m higher than now even for the Donna region (latitude 66° N). In the hypothetical map (Fig. 111) the outer, seaward limits of the refuges are drawn according to these figures. The inner (southern or eastern) limits (transversely hatched lines) wherever possible are based on the geological facts mentioned earlier. The division of the coastal foreshore into small refuges is based on the geomorphological structures of the land in that way that glacier tongues which reached the sea belt were assumed in the extension of the present-day fjords and valleys. I have not attempted any kind of reconstruction of the White Sea refuge, still less the quite uncertain “Oslo refuge.” II. Climatic conditions during the last glaciation. In the synoptic Table 38, (p. 802) altogether 97 Fennoscandian carabids (including one subspecies) are listed as Wurm hibernators within the region (including the refuge at the White Sea). In addition there are 33 doubtful cases. In the case of 52 carabids (including the subspecies balbii of Nebria gyllenhali) postulation of such a hibernation seems unavoidable. Let us now study the climatic requirements of these species, as evidenced by the present-day distribution of each species in the different plant regions of the region (Table 30, p. 440). Of the 52 “undoubted” hibernators we have: in the Regio alpina: 30* (58%); Regio betulina: 46 (88%); Regio coniferina: 48 (92%). Of the 45 “almost sure” hibernators: Regio alpina: 21 (47%); Regio be- tulina: 29 (64%); Regio coniferina: 45 (100%). Of the 33 “possible” hibernators: Regio alpina: 6 (18%); Regio betulina: 7 (21%); Regio coniferina: 30 (91%). So we find that 42-56% (22-73 species) of the Wurm hibernators at present do not cross the timberline, and apparently do not tolerate an arctic climate. Moreover, a considerable number of these species occur more or less accidentally in the Regio alpina (see Tables 30 and 38). Is this fact sompatihie with the usual idea of the Ice Age climate? The ultimate causes of glaciation need not be considered here. The views of the different authors, all hypothetical, offer glaring contradictions (see, for instance, Woldstedt, 1929, pp. 348 ff.; A. Wagner, 1940, pp. 145 ff; Zeuner, 1946, pp. 134 ff.). But it is natural, at first glance inevitable, that a general decline in the atmospheric temperature should be taken as the secondary cause *Not found in the Fennoscandian Regio alpina are the following “undoubted” Würn hiber- nators: Aépus marinus, Amara interstitialis, (Asaphidion pallipes), Bembidion aeneum, (B. femora- tum), (B. grapet), (B. lapponicum), B. lunatum, B. nitidulum, B. saxatile, B. scandicum, B. schüppeli, B. siebkei, B. transparens, Dyschirius angustatus, (D. helléni), (D. septentrionum), Leistus ferrugineus, Patrobus atrorufus, Pterostichus adstrictus, P. strenuus, Trechus fulvus. Species in parentheses were found in the Regio alpina outside the region. 675 774 Fig. 110. Highest coastline (HK) of the postglacial period (s. I.) and isobases. According to Granlund, 1936. of every ice age. Calculation of the extent of this temperature fall during the Wurm period had to be indirect, using chiefly the earlier position of the snowline. Most estimates envisage a fall of 3-5° C in comparison with the present-day tem- 777 perature of the region concerned (Antevs, 1928, p. 21; Woldstedt, 1929, p. 339), but Penck (1936) and following him Koppen and Wegener (1940, p. 22; also A. Wagner, 1940, p. 140) calculate the temperature fall for the ice-free parts 775 Fig. 111. Hypothetical map of Fennoscandian (and Danish) Würm refuges. Cross-hatching indicates maximum extent of Würm ice from south to north on geological basis. Following sources were used: I—Kaldhol (1930); II—Undas (1934); I1I—Grgn- lie (1927), Granlund and Lundqvist (1936); IV—Vogt (1912); V—Enquist (1918); Grpnlie (1927); VI—Nordhagen (1935, 1936); VII—Holtedahl (1929). 778 677 of Central Europe at 8° C. Mostly the annual mean temperature is cited, but at the same time it is often emphasized that a decline in summer tempera- tures is especially important for the development of glaciation (for example, Woldstedt, 1929, p. 340). ‘ How then would a temperature decline of this order of magnitude in summer affect the fauna of the parts of Fennoscandia that remained ice- free during the Würm period? A study of the July isotherm map (Fig. 63, p. 452) shows that already with the smallest reduction postulated (3°C), the entire Norwegian west coast south to Stad (latitude 62° N) would incur an arctic climate, with the mean July temperature below 10°C. With the 8°C fall calculated by Köppen no part of Fennoscandia would have reached a mean temperature of +10°C in July! Hence one must seriously ask whether the evidence provided by the “hi- bernators” among the carabids that are non-Arctic can be considered so conclusive. Could they not have changed their thermal requirements since the Ice Age? If not, do their relatively high thermal requirements not prove that they were no Würm hibernators? In response to the first question it may be said that in some species we are probably justified in doubting the “constancy of the ecological va- lency,” as Henriksen (1933, pp. 286 ff.) did in the case of Oriorrhynchus du- bius Strom—although in this case perhaps erroneously. However, if we work with whole group of species whose members have identical requirements—in the present context thermal ones—the probability of a concurrent change in all of them can be taken as zero (see p. 676). Besides, it must not be for- gotten that “adaptation” to a new climate does not represent a cumulative “acclimatization” from generation to generation, but must be due to geno- typic changes (mutations), which are then subjected to selection. There is no reason to assume that such processes may take place concurrently in a whole series of species. Furthermore, what kind of selection would have resulted in the postglacial loss of “cold resistance” on the part of the populations that hibernated? The second question, whether there are actually non-Arctic species for which Wurm hibernation within the limits of Fennoscandia must be consi- dered indispensable can be answered by referring to the species having wing dimorphism. As far as I understand it, for some (at any rate for 7) of them hibernation in western and/or northern Fennoscandia is as decisively proven as can be possible using the “evidence” provided by biogeography. And among these very 7 species there are 3 that have never been found above or north of the timberline, which are thus pronouncedly non-Arctic. So it is right to doubt the general validity of the temperature fall calcu- lated for the Wurm period by geologists and paleoclimatologists, asking two questions: 779 678 1. Is it not possible for a glaciation to occur as a result of climatic changes other than a fall in the summer temperature? 2. Can a decline in temperature of the Ice Age macroclimate not be offset at lococlimatically favorable places or at least be strongly moderated? 1. Expressed simply, glaciation occurs in the situation where there is more snowfall in winter than the amount of snow that melts in summer. A state of equilibrium can be shifted toward glaciation not only by low summer tempe- rature but also by increased precipitation in winter. There are some authors (for example, Scharff, 1899, p. 65 ff; Brockmann- Jerosch, 1909, 1919) who consider increased snowfall the most important cli- matic factor responsible for glaciation (see also Antevs, 1928, p. 20). Enquist (1916) also believes the precipitation conditions have played a decisive role. Furthermore it may be pointed out that according to Hyyppa (1933, pp. 29 ff.; 1936, pp. 446 ff., p. 458) and Sauramo (1942, p. 281) the standstill of ice dur- ing the so-called Salpausselka stage was mainly due to an increase in winter precipitation. Assuming that the glacial climatic changes had a global cha- racter, it must be stressed that the glaciations of the northern hemisphere in East Africa and southwest Asia apparently coincided with periods of abundant rainfall (for example, E. Nilsson, 1947, p. 169). Nevertheless, the catastrophic retreat of the glaciers in most parts of the earth in recent decades provided us with important information on the rele- vant climatic determinants. Ahlmann in particular has dealt with this problem in a series of articles. He finds that the regression of glaciers cannot be ex- plained by a rise in the summer temperature: “The most important cause of the melting away. ..seems to have been an increased influx of heat through the atmosphere. In comparison with the accumulation season, the melting season is prolonged due to higher temperatures in autumn and spring” (Ahlmann, 1948, p. 320; translated). This agrees with the changes established for the Scan- dinavian climate of recent times (p. 641), which scarcely involve the summer. In Spitzbergen, where the glaciers have strikingly retreated, the summer tem- perature during the period in question has remained virtually unchanged. Of course, as emphasized by Keranen among others (1944, p. 55) and Ahlmann (1948, p. 307), this is partly due to the fact that “even in summer the heat is largely utilized for the melting of ice and glaciers” (Keranen, l.c.). But con- ditions in Scandinavia show that summer temperatures even in glacier-free regions have risen much less than spring temperatures. The following conclusion seems to be justified: If the large-scale general disappearance of glaciers during the late present epoch is not (or is only slightly) to be ascribed to higher summer temperatures, but to a prolongation of the annual melting season, then it must be assumed that a general augmen- tation of the glaciers—a glaciation—can develop without any substantial fall in Summer temperatures. From biological premises (to use circular reasoning!) it is at any rate clear 780 679 that the decline of 8°C calculated by Köppen and Penck for Central Europe in the last glaciation may not have general validity. The Würm hibernation of plants and animals on Iceland, Greenland, and other Arctic islands (for literature see p. 758), which is supported by sound reasoning, would have been impossible in such conditions. The present-day climatically most favorable parts of Iceland and Greenland would have had a mean July temperature of +2 to 3°C and —1 to +2°C respectively (calculated by Vedrattan, 1925, and Bocher, 1938, p. 10). On biological grounds it seems as if the glacial temperature decline in the north was less than in Central Europe. 2. To understand the loco-climatical conditions at the edge of an inland ‚ice, it is useful to proceed from similarly situated places of the present. There are many good examples around the world of rich flora and the corresponding fauna spreading in the immediate vicinity of a large glacier. The best-known case may be the Malaspina glacier in Alaska (Russel, 1893; Wolff, 1915). There the forest not only approaches close to the ice edge: it even grows on the moraine-covered glacier on a surface of 50-60 km’. A vivid impression of the luxuriant character of this “glacial forest” is provided by Russel’s photograph (1893, p. XIV). In New Zealand and in western Patagonia evergreen rain forest thrives close to large glaciers (Du Rietz, 1935, p. 228). From eastern Greenland, Bocher (1938, pp. 312 ff.) describes several very rich plant localities in the immediate vicinity of ice. One of the richest lo- calities, situated near Wiedemann fjord on a southern slope at an altitude of 300-400 m above sea level is completely surrounded by ice (see his Plate I). Yet, it harbors many species of a pronounced southern distribution type in Greenland. In southern Greenland, Betula also grows at a short distance from ice. Lynge (1934, p. 164) has observed a very rich plant locality in the northern island of Novaya Zemlya, situated between the edge of terrestrial ice and the sea On a southern slope only a few hundred meters broad. The most vivid case I know of, is southeastern Iceland (Lindroth, 1931, pp. 541 ff.). From the sea the coast appears inhospitable. The enormous Vat- najokull—the largest glacier in Europe—projects more than 2000 m almost straight out of the sea, leaving in between only a narrow, gray belt of rock and sand. But inside the valleys, on the southern, protected slopes, there are green meadows and the birch forest thrives. It is of course low (< 3 m), but well covered with an abundant lower stratum, chiefly of the meadow type. Such an oasis of organic life is Skaftafell, surrounded by the glacier on two sides, and on the other two by the totally sterile gravel fields of the glacial rivers, extending right down to the sea. The distance from the place photographed (Fig. 112) to the ice edge is 1200 m. The insect fauna of Skaftafell is very rich, considering Icelandic condi- tions. Six species (1 Collembola, 3 Diptera and 2 Hymenoptera) are known 782 in Iceland exclusively from this place. And what is more important, this fauna "6761 ‘I Ang :1oyJne Aq JOY ‘w 007I SI 38p3 301 ay) 0) »aueIsıp SOUL ‘puejac] Jo Jutod ysaysiy oy) “(stu 66ILZ) Inynuysjepeuueag si punosSyoeq ay) Uy "nyoleweA Je [ofeyeys “pueyooy “ZI SI ISL 783 681 also includes species such as Philonthus trossulus Gr., Quedius umbrinus Er., and the Homoptera, Lausulus pseudocellaris Fall. (Deltocephalus distinguen- dus Fall.), which, as far as can be judged, reached their present-day northern climatic limit in Iceland. It would be a rewarding task to undertake a precise study of the local climate and microclimate of Skaftafell. However, at present, it is not known. The macroclimate, even for Icelandic conditions, is fairly unfavorable. The two nearest meteorological stations, located in rather open country closer to the coast, namely, Fagurhölsmyri and Hölar near Hornafjörthur (at a dis- tance of 23 and 90 km respectively), have mean July temperatures of 9.9° and 10.0°C respectively (Vedrattan, 1925), —which is a summer climate, corres- ponding with that of the furthermost northern peninsula of Norway (Fig. 63, p- 452). In these regions of Scandinavia numerous species of the Skaftafell fauna are absent (among the carabids: Bembidion bipunctatum, Pterostichus adstrictus, Trechus obtusus), at least partly for thermal reasons. It must be assumed that the location of Skaftafell has produced an unusually favorable lococlimate. The fauna and flora of Skaftafell in southeastern Iceland therefore prove not only that the direct vicinity of a large glacier can be without a climatically deleterious effect, but that such a situation even produces a thermally favorable condition in comparison with more open locations farther from the ice. There is no reason to assume, that the proximity of ice in itself would have had a different effect during the Würm period. The loco-climatically advantageous position of Skaftafell and so of simi- larly situated places in a more glaciated region—in Greenland today and along the Fennoscandian coast during the Würm period—seems to be due to the following factors: 1. The landscape is very hilly and therefore has southern slopes with fa- vorable exposure to the sun. The importance of a sunny location on “southern hills” and “southern slopes,” especially in southern Sweden, for the develop- ment of a rich flora has been illustrated by Andersson and Birger (1912, pp. 52 ff.) with a series of apt examples. The temperature measurements by Frodin (1915) and Krogerus (1937) represent attempts to express this thermal advan- tage numerically. It is hoped that someone will take up this question in the future in all its aspects. 2. Marginal mountains and the ice itself provide protection from the wind. The scattered Betula forests of Iceland clearly reveal the decisive role of the wind as a factor detrimental to the vegetation (Lindroth, 1931, p. 445), especially in the thermally determined peripheral regions. Secondarily, a tall, lower stratum thrives under the protection of the forest, and the two together support a rich insect life. We will later return to the question whether it is justified to assume a woodland vegetation in the Fennoscandian Wurm refuges. 3. A glacier situated at a sufficient altitude produces descending winds 784 682 (Foehn-winds), which can bring about a considerable warming of the marginal regions. The frequency of these winds is determined besides the differences in level also by the size of the glacier concerned, i.e. by the constancy of the high pressure that builds up over it. In Greenland such warm Foehn-winds have long been known. It is evident that during the glacial period we have to reckon with an almost constant anticyclone over the ice (Enquist, 1916; Woldstedt, 1929, p. 342). It is on this assumption alone that the loess deposition in Central Europe can be understood (A. Wagner, 1940, p. 141). So the thermal effect of the winds radiating out from nordic inland ice was determined entirely by altitude factors. Along the southern edge, in low-lying Central Europe, the ice was estimated to be only about 100 m thick (Kessler, 1925, p. 153). The northern winds therefore moved almost horizontally and could not show a Foehn character. On the contrary, they had a strongly cooling effect on the ice edge region (Kessler, l.c.; Woldstedt, l.c.). Very different conditions prevailed in western Norway. The fjelds, which form the main Scandinavian watershed, are situated at a short distance from the coast. The mean inland altitude was increased substantially by ice during the Wurm period, and we may reckon with a descent of 1500 to 2000 m over the few miles down to the coast. The air streaming out from above the ice must have regularly achieved the character of a descending wind. It had a warming effect. As far as I can see this is the most important reason why the mean tem- perature decline of 8°C calculated by Penck for the last glaciation in Cen- tral Europe cannot be applied to western Scandinavia. The occasional winds from other directions would not have provided any substantial thermal com- pensation in ice-free parts of Central Europe, especially since these parts were at that time situated farther from the coast (the Dogger-land at any rate was supra-aquatic), and the west winds were therefore cooled down on their way. In Scandinavia the westerly winds moving in straight from the sea were thermally much more favorable as well. The Gulf Stream must have moved along the Norwegian coast even during the Wurm period, for there can be no question of so great a land elevation that the almost 600 m deep Wyville- Thompson ridge between the Shetlands and the Faeroes would have hindered access to the North Sea. On the contrary, the assumed moderate land elevation of 100 m probably signified a thermal advantage for the coastland, since this brought it closer to the edge of the continental shelf and so closer to the warm Gulf Stream water (emphasized by Elfstrand, 1927). The Gulf Stream avoids ocean regions less than 200 m deep (Schott, 1926, p. 184). Conclusion: During the last glaciation those parts of Europe that formed coastal belts below a high inland were climatically much favored, especially if 785 683 they had direct contact with the Gulf Stream. No region meets these require- ments better than the Norwegian west coast (see also Dahl, 1946, p. 239). It is therefore in no way contradictory to the “exact” natural sciences if the biologists take the view that in the Fennoscandian Würm refuges other organisms than the Arctic (alpine) ones have also survived the glaciation. Recently the view has been expressed that the Norwegian Wurm refuges were covered with forest (at any rate with trees of woodland-forming species). Lindquist (1948), in his studies on the variability of Picea abies, not only—so to say incidentally—named two treelike Betula species (B. callosa and B. tor- tuosa) as Scandinavian Wurm hibernators, but also extends his claim to Picea abies, which in the form of his newly described “variety arctica” is said to have hibernated along the Norwegian west coast probably in the south (along Storeggen) and in the north (along the Vesteralseggen). The following objections can be raised against Lindquist’s arguments: 1. The “variety arctica” is not sufficiently distinguished from the Central European “variety germanica Lindq.” Both have smooth annual shoots but are said to differ in the cones. The trees with smooth shoots sporadically occurring also in southern Sweden “probably belong to the variety germanica” (p. 304). How far these indicate a postglacial immigration from the south or originate from plantation trees of Central European origin is left unresolved (p. 321). Is it not more probable that the character of “smooth annual shoots” represents a mutant occurring in different populations independeni of the structure of the cones? 2. The “variety arctica” never occurs as a pure stand. However, the per- centage component (somewhat < 30) increases in Scandinavia northward and westward, and is therefore highest along the area limits. It is then not possible that the smooth shoots are the superficial manifestation of a physiologically ef- fective mutation, which in the critical regions of the area limit have a markedly positive selection value? Lindquist’s argument to the contrary (pp. 307, 332) is not very convincing. 3. The weightiest objection is of a purely phytogeographical nature. As shown by Lindquist’s map (Fig. 3, p. 259) on the localities of the spruce material he investigated, he studied practically no record material from nor- thern Finland. Therefore, the isolation of the Scandinavian “variety arctica,” becomes completely problematic. It is quite conceivable that the northern marginal occurrences of spruce are characterized by a belt with more con- spicuous “variety arctica,” not only in Finland but also in Russia and farther east. Consequently it is impossible to deny that the “variety arctica” may have immigrated as a component of the large postglacial stock emanating from the east. It is not really astonishing, that this eastern type with hairy shoots managed to reach every population of the wintered “variety arctica” in western Norway postglacially, which according to Lindquist (p. 325) has to be assumed because as far as is known there are no pure stands of “variety arctica’? Is 786 787 684 this an example of long-distance aerial transport of pollen? As important evidences supporting the assumption of a hibernation of Picea in Scandinavia, Lindquist (p. 329) cites the distribution of two lichens living on spruce, Cavernularia hultenii and Tholurna dissimilis, which were studied in detail by Ahlner (1948). However, in examining the evidence pro- vided by these lichens the following facts must be borne in mind: 1. Their actual distribution is not sufficiently known. Ahlner himself (p. 94) says of Tholurna that “the map provides only a broad picture of the distribu- tion” (translated). Cavernularia was first (described and) discovered in Scan- dinavia in 1937. Not only was it unknown earlier, it is also absent from the entire older herbarium material (with the exception of an accidental fragment collected along with a specimen of another lichen; Ahlner, p. 171). The best specialists, for instance, Degelius, the author of the species, have not seen this lichen although they have assiduously collected in the regions (e.g. Trondelag) where the species seems to be common. Of the 250 presently known localities of Cavernularia (l.c., p. 33), Ahlner himself discovered more than two-thirds (pp. 169-172). Under these circumstances, to claim that one is even closely acquainted with the distribution of Cavernularia hultenii in Scandinavia, or that conclusions on the history of immigration can be drawn therefrom, is hazardous. 2. Lichens are organisms with unusually strong capability of dispersal. This is evident mainly from their regular occurrence on erratic rocks (for instance, Degelius, 1936). Another example is the occurrence of some oceanic species in Finland (maps in Degelius, 1935). From this follows, that the area limits of the lichens are mostly pronounced existence limits. Hence the lichens are less appropriate as objects of the branch of biogeography that deals with questions of the history of immigration. Ahlner (p. 145) expresses himself more cautiously than Lindquist (e.g. p. 334) concerning the question of hibernation. 3. Lindquist (p. 330) indicates the possibility that Cavernularia probably “hibernated” in western Norway on rock (and not on Picea). His assumption seems indeed to be based on a misunderstanding (Ahlner, p. 143), since the species was observed only in one case, it seems accidentally (pp. 34, 171), ona stone in Trondelag. On the other hand, from the same region there are at least 5 records on Alnus incana and Sorbus aucuparia. | do not know how much importance to attach to this situation. It should of course not be ruled out that in the eastern part of its area Cavernularia has become exclusively an in- habitant of spruce for microclimatic reasons (humidity requirement), whereas it can occasionally live on other trees only in an oceanic climate. If only de- ciduous trees would occur in the refuges, the lichen would perhaps be content with them. These critical remarks on the conclusions drawn by Lindquist (1948) (and Ahlner, 1948) concerning a Wurm hibernation of Picea have been rather de- tailed. However, the question is most important. I do not mean to say cate- 790 685 gorically that spruce cannot have hibernated, but consider it improbable. I do not find it appropriate to draw such definite conclusions from the primary material available to Lindquist. A more precise fixation of the “variety arctica” is necessary—and for this, breeding experiments are indispensable. Lindquist’s material is suitable only for a working hypothesis and nothing more. One can easily imagine how it will be cited by the less critical biogeogra- phers in the near future: “If spruce hibernated in the Norwegian refuges, then it must be assumed that even. ..” etc. They will envisage the refuges overgrown with stately coniferous forest. Actually, so far no definite clue has also been produced to show that birch is a Scandinavian Wurm hibernator. In this field, too, Lindquist is conducting an investigation in connection with which two preliminary maps, on Betula tortuosa and B. callosa, have been published in the Picea contribution (1948, pp. 21-22). Concerning these it can be stated first and foremost that they are incomplete. Among other things, the connection to the east is obscure. However, it must be conceded that these “species” are unusually difficult to map on account of their strong inclination to hybridize. The hibernation of these two Betula species is in itself a strong probability; from the botanical side nothing more can be said at present. Perhaps entomology can be of help in solving this problem. The cara- bids indeed do not include any monophagous species associated (directly or indirectly) with particular trees; but such species do occur among other Coleoptera. Curculio (Balaninus, Balanobius) crux Fbr. is a small but characteristically marked curculionid which does not easily escape the collector’s attention. In Fennoscandia it is restricted to an isolated area in the farthest north (Fig. 113). It is unknown in central and northern Russia as well as in Siberia. The species is widely distributed in Central Europe and is therefore in no way climatically bound to the far north. The southern postglacial stock has reached only Den- mark, not Scandinavia, and in the east Latvia. Curculio crux is an undoubted Wurm hibernator in northern Fennoscandia. It is bound to Salix, where the larva develops inside nematode galls (West, 1940-41, p. 578). However, it does not live on creeping dwarf willows but on tall, smooth-leaved species, in Denmark on S. cuspidata, in the north perhaps chiefly on S. phylicifolia. Cur- culio crux therefore shows that shrub-forming Salix species grew in the Wurm refuges of northern Fennoscandia. Rabocerus (Salpingus) foveolatus Ljungh is a heteromeran with a wider Fennoscandian area (Fig. 114). This species also lacks any connection to the east, being unknown in northern Russia and Siberia. In the south the sou- thern postglacial stock has reached only Skane through Denmark. In Finland it has advanced much farther, but there is a distinct gap in Finnish Lap- land; the northernmost locality, which may belong to the southern area, is Lm Kantalaks. The conditions in Sweden clearly show that the species here is of 291 686 western origin; the east coast has not been reached at any point. The most densely colonized subarea is situated in northern Norway, north of latitude 68° N. Rabocerus foveolatus is an indubitable Würm hibernator in this region, probably also in more southern parts of the Norwegian coast. The species is associated with trees. It lives (as predator or as “commensal”) in the galleries of ipids: in Central Europe, Denmark and southern Sweden on all kinds of deciduous trees, in the north, as far as is known, exclusively on Betula (but not B. nana), where it was discovered by Palm (in litt.) together with Scolytus ratzeburgi Jans. This ipid is likewise found right up to the northernmost part of Norway (Strand, 1946, p. 595). An ipid living on Salix does not occur in north- ern Fennoscandia. If Rabocerus foveolatus is recognized as a Fennoscandian Würm hibernator it can also be cited as evidence for the survival of tree-like Betula. A similar case is mentioned by Strand (1946, pp. 24, 210), namely, of the staphylinid Phyllodrepa vilis Er., which lives in various insect galleries under the bark of trees and the more ragged bushes. In Fennoscandia this species is known exclusively on the Norwegian west coast, but has been found there only from 6 Ryfylke in the extreme south to 35 Tromso. Strand is certainly correct in noting (p. 24) that this range could have been produced neither by the feeding biology of the species nor by climatic factors, especially when the Species in its total area is not bound to an oceanic climate. Phyllodrepa vilis too must be a Norwegian Wurm hibernator. It would be a tempting and ambitious task to monograph the fauna de- pendent on the Fennoscandian fjeld birches, their ecology and history. I also believe that a similar study of spruce insects could provide valuable evidence for or against Lindquist’s view (1948). Here one has the advantage of being able to use Saalas’ valuable study (1917, 1923a) “Die Fichtenkafer Finnlands” (The Spruce Beetles of Finland)t. According to the above discussion, considerable entomological and bota- nical factual material favors the assumption that the Fennoscandian Wurm refuges at least in part had a non-Arctic climate. The findings of paleoclimato- logy do not preclude the idea. The refuges were apparently covered with birch forest in wind-protected places (also assumed by Bjorkman, 1939, p. 207; Du Rietz, 1942, p. 189), and they may have largely corresponded to the present-day Regio betulina of the fjelds. The more or less sure hibernators among the carabids (p. 776) also in- clude species that may be climatically still more fastidious, since they never extend into the Regio betulina. Among them there is one group of species which according to its total distribution seems restricted to the west coast of Europe and has to all appearances already found its climatic northern limit in the southern half of Norway. This is the atlantic (oceanic) element, to which t (suppl. translator). Fig. 113. Curculio (Balaninus, Balanobius) crux Fbr. Fig. 114. Rabocerus (Salpingus) foveolatus Ljungh*. The assumed eastward limit of the postglacial stock is shown. Blank circles depict provincial finds in the Trondheim region. (According to Lysholm, 1937.) *The localities of Rabocerus foveolatus have been compiled from the pertinent literature and from the larger public and private collections. It should be noted that the original description (from Smäland) by Ljungh (1823, p. 269) may be applied equally well to R. gabrieli Gerh. (pointed out already by Seidlitz, 1920, p. 1108); Boheman’s “foveolatus” from Sma (probably Olmestad, Anneberg, situated only 20 km from Ljungh’s residence, Skärsjö in Billaryd) is also identical with gabrieli (RM!). So not only is Ska eliminated as a provincial find but also the name foveolatus. Even the record from Lyl Lycksele (Zetterstedt, 1840, p. 168), based on a damaged specimen (without anterior part of body) in ML, might refer to gabrieli. 792 689 correspond the most heat-requiring species of the so-called Ilex flora of the botanists. Hence in determining whether a species is to be considered as part of the atlantic element, its total distribution must be taken into consideration. Like the curculionid Barynotus squamosus Germ., Trechus obtusus shall therefore rather be called “atlantic-alpine” (Lindroth, 1931, p. 555). Bembidion tibiale (p. 765), found only in Ryfylke, is likewise a montane species in Central Eu- rope, whose occupancy of such a restricted Fennoscandian area is scarcely due to climatic factors. In the present context, Bembidion harpaloides should also be omitted, partly because its occurrence in the extreme south of Norway was ascribed to postglacial immigration (p. 765). As typical atlantic species there remain only Aépus marinus and Trechus fulvus, which have almost identical distribution on the west coast of sou- thern Norway, with the northernmost locality on the island of Hitra. Among other Coleoptera the two curculionids, Otiorrhynchus porcatus Hbst. (only in Province 7; total area less typically atlantic) and Mesites tardyi Curt. (only in Province 6) are the clearest corresponding examples. All 4 of these species occur on the British Isles. It may seem rash to classify these animals (and the corresponding plants) as Würm hibernators, as I have already twice done (Lindroth, 1931, p. 554; 1932). Even the bold A.M. Hansen (1904, p. 57), who was otherwise far ahead of his time on these questions, did not venture to do so. Like Holmboe (1913, p. 90) later in respect of Ilex, he believed that the definite atlantic floral ele- ment had at least partly immigrated postglacially across the sea (probably with birds). In recent years some botanists have been much more daring: they now entertain the idea of at least a partial hibernation of the “atlantic” species in Scandinavia (Degelius, 1935, p. 302; Du Rietz, 1935, pp. 228 ff; see on the other hand Faegri, 1937, p. 437). The following facts may be cited in favor of this assumption in the present case: 1. It is not certain that the species restricted to southwestern Norway have here reached their climatic existence limit. Among the species considered above (p. 765), this cannot be so in the case of Bembidion tibiale, occurring solely in 6 Ryfylke, still less in Corymbites cupreus Fbr., also with an isolated occurrence there (map in Lindroth, 1939a, p. 248), which is in the process of invading Scandinavia from the east. The first specimen was found in the Swedish region in 1944 (Wirén, 1947, p. 191). Small isolated areas are not in themselves evidence for climatic relicts. 2. Of the 4 doubtful Coleoptera, 3 are constantly brachypterous (exception: Mesites). There is thus a limited capability of dispersal. That the terrestrial con- ditions postglacially, even during the Dogger-land period, were not favorable for a southern immigration directly to southwestern Norway has already been shown (p. 765). 793 794 690 3. Trechus fulvus also occurs on the Faeroes (on several islands) (West, 1930, p. 13), where it is undoubtedly to be considered a Wurm hibernator. 4. The two carabids live on the seashore, right at the water, and are affected by the temperature of the seawater at least as much as by the “macroclimatic” atmospheric temperature. For reasons mentioned above (p. 783) the Gulf Stream exercised a strong thermal influence on the Norwegian coastland. 5. In the refuges of northern Norway species with pronounced thermal requirements (non-Arctic species) have overwintered as well (for example Bembidion aeneum, p. 399, B. transparens, p. 389, Pterostichus strenuus, p. 395, Curculio crux, p. 787, Rabocerus foveolatus, p. 790). At any rate the southern Norwegian refuges were climatically better situated; the insolation conditions were more favorable, and the Gulf Stream had still not lost its warmth. The south Norwegian refuges must have had a climate which at lococlimatically advantageous places was much warmer than subarctic (corresponding to the Regio betulina). The two carabids (Aépus, Trechus fulvus) have in my opinion to be consi- dered as Wurm hibernators, probably also at least Otiorrhynchus porcatus. With regard to the problem of the “J/ex flora” I am unable to provide any new point of view. It would perhaps seem strange if the above-mentioned Mesites tardyi, which in Ireland lives largely on J/ex (Munster, 1922, p. 131), had followed its host plant into a new country by postglacial “accidental dispersal.” However, in Norway it was “unfortunately” found on Fraxinus (Munster, l.c.)! A detailed ecological study of the atlantic faunal element in Scandinavia will be a monumental, but certainly very rewarding task for some future zoologist. In the absence of sufficient data from our own region we can turn to the conditions on the British Isles, where a prominent role is played not only by our atlantic element but in addition by a group of species with a more southern imprint, the so-called lusitanic group of species (Scharff, 1899, p. 287), which is native especially to Ireland and shows relationships with the Iberian Peninsula. It is strange that on the British Isles—in many ways the most suitable region of Europe for this purpose—entomologists have shown relatively little interest in biogeographical questions. The best synopsis of the Coleoptera has been provided by a foreigner (Sainte-Claire Deville, 1930a). In recent years, however, the British Lepidoptera have been the subject of a detailed zoogeographical study (Ford, 1945; Beirne, 1947b). Beirne’s contribution is an especially detailed analysis of the history of British macrolepidoptera, and represents an attempt to establish the time of immigration of every species. But his premises are extremely hypotheti- cal; in particular, he rarely makes a serious attempt to explain the area of a species under the influence of the present-day environment. And indeed the Lepidoptera—with a few exceptions—thanks to their relatively strong capa- bility of dispersal are well equipped to reach their existence limits in every climatic period (see also the Summary). 795 691 Especially interesting in the present context is Beirne’s view (1947b, pp. 285-286) that certain “lusitanic” forms (in the families Noctuidae, Geometridae, Arctiidae), which are restricted to Ireland, represent the oldest element of the British lepidopteran fauna. He thinks it possible that they survived (also according to Scharff, 1899, p. 307) the maximum glaciation in Ireland* (or more correctly in a now subaquatic “foreland” south of it). Even the botanists seem to be more and more inclined to assume the hibernation even of definite heatrequiring plants—including the lusitanic species—on the British Isles (Woodhead, 1929; Wilmott, 1930, 1935; Praeger, 1932; Degelius, 1935; Du Rietz, 1935b), at any rate during the last glaciation (according to the definition given by Wright, 1914, p. 76, and Charlesworth, 1929, p. 336, i.e. “New Drift” and Wurm II in Movius, 1942, p. 26). Even the Outer Hebrides (or a foreland to the west of these) have been cited as a Wurm refuge for characteristic southern species (Ford, 1945, p. 320; Harrison, 1947). There are not so many “counsels of caution” now as in the past, but they are still represented by Charlesworth (1930) and Salisbury (1935). Among the relevant Coleoptera first and foremost is the flightless Orior- rhynchus auropunctatus Gyll., which is restricted to eastern Ireland and occurs outside the British Isles only in southern France and the Pyrenees (Scharff, 1907, p. 33). The above-mentioned, peculiar Mesites tardyi, which, outside the small Norwegian area, is endemic to the British Isles and is frequent only in Ireland, also belongs here (map in Scharff, 1907, p. 50). Its closest relatives are distributed from southern France over the western Mediterranean and on the Macronesian islands (Sainte-Claire Deville, 1930a, p. 104). Mesites tardyi cannot be a postglacial immigrant on the British Isles. It lives in decaying wood of //ex and other deciduous trees (but has not been observed on Betula; O’Mahony in litt.) and hence points to the hibernation of “true” deciduous trees. This is imperative in the case of the pronounced American element in Ireland’s flora (Praeger, 1932, p. 128). There is therefore a whole series of facts that support a Wurm hibernation of the pronounced southern species on the British Isles, including (a once larger) Ireland. Among these is the “lusitanic” element (see also Wilmott, 1935, p. 221), whose members, judging from the recent distribution, have greater heatrequirements than the corresponding “atlantic” group of Norway. Of course the present-day climate in Ireland is only slightly more favorable *In respect of chronology, Beirne follows Movius’ opinion (1942, pp. 26 ff.). He implies that the maximum glaciation (“Old Drift’) of the British Isles was contemporaneous not with the Riss, but with the Würm I stage of the Alps and with the Warthe stage of northern Germany. Apart from the fact that the synchronicity of the Warthe stage and Wurm | is quite uncertain (Richter, 1937, pp. 84, 126), it must be conceded that in all probability the maximum glaciation of the British Isles coincided with that of the continent (i.e. Riss and Saale respectively). Flint (1947, p. 343) synchronizes the Warthe stage with the first phase of the “Newer Drift” in Great Britain. However, in the biological context this question is of secondary importance. 796 797 692 than in western Norway (the mean July temperature in western Ireland is only 15°C; cf. Fig. 63, p. 452). If the glacial temperature decline in Western Europe was fairly uniform (which may be assumed at least for the parts af- fected directly by the Gulf Stream), the west Norwegian climate cannot have been much more adverse than that of the British refuges in the southwest and west. In this indirect way we therefore arrive at the conclusion that even the pro- nounced “atlantic” organisms can have hibernated in western Norway during the Wurm. It also needs to be mentioned that during parts of the last interglacial period the land in the present-day North Sea region was apparently situated higher than in any postglacial period (Lindroth, 1935a, p. 628). The land connection, which at that time allowed a dispersal of the “boreo-British” Coleoptera from Scandinavia to the British Isles, could also have been uti- lized by the “atlantic” species in the opposite direction—even if not at the same time. If the climate and vegetation in the western and northern Fennoscan- dian refuges were the same as assumed here—if even in the far north at the most favorable places these refuges had a subalpine character with birch vegetation—then the conditions at any rate in southwestern Norway must have been much more favorable than in present-day Greenland. Undoubtedly all the preconditions were present to enable even man to live in the Fennoscandian Wurm refuges (Ekholm, 1925). The fact that definitely dated remains of such a culture are not so far available is scarcely of decisive importance, since inter- glacial and “refuge” deposits within the region are extremely rare (Sandegren, 1948, p. 41). Moreover the latter—for reasons mentioned above (p. 772)—are to be sought largely on the present-day seabed. The so-called Komsa-culture on the Varanger Peninsula in the far north, considered by various authors (especially Nordhagen, 1933, pp. 82 ff.) as belonging to a Wurm refuge age, was later (Be and Nummedal, 1936, pp. 183 ff.; Munthe, 1940, p. 220) classi- fied as postglacial (s. 1.). Only by definitely datable archeological findings can “the Fennoscandian Ice Age man” become more than a hypothesis. But it is a good and highly probable hypothesis. The question whether some organisms survived more than one glaciation within the Fennoscandian region cannot be brought closer to a solution by a study of the carabids. The so-called west-Arctic element of the fauna and flora is relevant here, i.e. species whose Fennoscandian area indicates a former connection with the west, toward America. As far as is known there is no west- Arctic species among the carabids; even among the remaining Coleoptera there is so far no indubitable case known. On the other hand, the west-Arctic group among the Lepidoptera has a strength of at least 8 species (Wahlgren, 1919; 1935-1941, p. 54). The last list of the west-Arctic phanerogamic plants has been provided by Nordhagen (1935, pp. 143 ff.) and Nannfeldt (1940, p. 39). 798 693 The fauna and flora of the North Atlantic islands, particularly Iceland "and Greenland, show that a direct, biologically effective land connection be- tween Europe and North America did not exist during the last interglacial period (for example, Lindroth, 1931, pp. 567 ff.). It therefore seems we must shift the required “land corridor” into an earlier interglacial or even to the preglacial period. Such a view renders almost unavoidable the assumption that the west-Arctic species survived even the greater, penultimate glaciation (Riss) in Fennoscandia. It is certainly no accident that all the plants and Lepidoptera with pronounced west-Arctic distribution are able to live in the present-day Regio alpina. Concluding Remarks on the History of the Fennoscandian Fauna The most important characteristic of the Fennoscandian fauna is its youth. This characteristic has also been noted by Peyerimhoff (1947) in his lucid survey of the Danish-Fennoscandian coleopteran fauna (according to the Cat- alogus, 1939). Among the carabids there is at the most one species, Bembidion scandicum, which is endemic to the region. The beginning of the formation of an endemic subspecies is best shown by Carabus problematicus. Moreover it is noticeable how even species that have lived in Scandinavia uninterruptedly since the last interglacial period have remained morpholog- ically (taxonomically) unchanged, even when larger or smaller populations were isolated by the Würm Ice in separate refuges (Lindroth, 1941, p. 438). This is concrete evidence that even under strongly fluctuating environmental conditions, 100,000 years is not enough to produce new “species” in the family Carabidae. It is probable that this finding is also valid for other geographical regions. Our fauna may not contain a Tertiary element, and even the possibility of continuous survival in Fennoscandia since the time of the greatest glaciation (Riss) cannot be ascertained for any species of carabid. Hence according to its origin, the Fennoscandian carabid fauna may be appropriately divided into three categories: A. Würm hibernators (within the region). B. Southern immigrants, which entered Scandinavia west of the Baltic Sea in the postglacial period. C. Eastern immigrants, which entered Finland postglacially from regions east of the Baltic Sea (and in many cases advanced into the Scandinavian region). A detailed synopsis on how to imagine the immigration route—or more often routes!—of each species is provided in Table 38. The many question marks betray our lack of certainty, and also the other data contain much that is subjective, especially with respect to the exact route of the postglacial immigration. 799 694 However, the main purpose of the Table is to emphasize the importance of Würm hibernation. Hence in the last four columns relating to this the “un- doubted” cases are indicated by a bold cross, with page references to the text where our views have been substantiated. Table 38 may be summarized as follows: Wurm hibernators*: 97 more or less sure species** = 27.1% of the entire fauna, of which 52 are “sure” species** = 14.5% 45 are “almost sure” species** = 12.6% In addition there are 33 “possible” species** = 9.2%. Southern immigrants: 280 more or less “sure” species = 78.2% of the entire fauna. In addition there are 12 “possible” species = 3.4%. Eastern immigrants: 266 more or less “sure” species = 74.3% of the entire fauna. In addition there are 23 “possible” species = 6.4%. Thus the Würm hibernators comprise at least one-fourth of the Fennoscan- dian carabid fauna. It is striking that most of the species belong to more than one of the three main historical groups, i.e. they have a double or threefold origin. Species with only one origin are: 5 (+1) = 6 Wurm hibernators (1.7%) 55 (+1) = 56 southern immigrants (15.6%) 26 (+3) = 29 eastern immigrants (8.1%) Total: 86 (+5) species = 25.4%. Species with a double origin are: 4 (+3) = 7 species, which are Wurm hibernators as well as southern immigrants (1.97%). 15 (+16) = 31 species, which are Wurm hibernators as well as eastern immigrants (8.7%). 130 (+12) = 142 species, which are southern as well as eastern immigrants (39.7%). Total: 149 (+31) species = 50.3%. Species with a threefold origin, i.e. with Wurm hibernation as well as southern and eastern immigration are: 58 (+29) = 87 species (24.3%). The map (Fig. 115) is an attempt to represent the contents of Table 38 cartographically. *If the “hibernators” of the White Sea refuges (last column) are ignored the figures are: 94 more or less sure species (26.3%), of which 52 (14.5%) are “sure” and 42 (11.7%) “almost sure”; in addition there are 26 “possible” species (7.3%). **Including the subspecies balbii of Nebria gylienhali. 800 Fig. 115. Principal immigration groups of the Fennoscandian carabid fauna. Cf. Table 38. Fig. 116. Bembidion saxatile. Attempted reconstruction of the postglacial 801 history of this species. 697 802 Table 38. Postglacial immigration and migration of Fennoscandian Carabidae from Würm refuges. In the four “hibernation columns,” which are of special interest, “sure” . SEN cases are marked with a bold cross Southern Eastern immigration immigration ay vard as far as Nordkapp) Across the Gulf of Finland Between Gulf of Fin- land and White Sea Through Kuusamo-Salla In southwestern Norway (south of latitude 64°N) In northwestern Norw Through Kanin-Kola (east: Directly to Old-Gtl Through Skagerrak In the northeast (Veranger-Petsamo) On the White Sea! o i=) oN =a Nn a fol?) = ° be = = Abax ater Acupalpus consputus dorsalis dubius exiguus flavicollis meridianus Aépus marinus SER Agonum aldanicum archangelicum assimile bogemanni consimile dolens dorsale criceti fuliginosum gracile gracilipes impressum krynicki livens longiventre le: lugens Tall mannerheimi marginatum 7 micans a moestum a mülleri ui ? + + > > > > DS ++t+++ 4+ | | | | | | | | | | vy + “vu + | | | ++++++ | ++ | et munsteri obscurum piceum : DD RD D> DDO DDE RR ERE RRR ER RR ERD + + ++ ++++++ | 1Perhaps in part immediately east of the Fennoscandian border. To northern Sweden directly from Finland p. 791 p. 741 Probably to Sweden directly from Southeast. Würm non | Remme Eastern immigration Southern immigration Wurm hibernation Remarks Q no) joe c 3 {oe} = = 32125 el. als BO lz 5 ~/si Cres 6 | o flos|oZl © F/O} ofa e |< IZ2ol4e!] él s 2898| | = ieol & | = |EeS|Ee. 1258| a om 8 |: 5 |e =] = E |22/982|55| o x Er ie) 3 ZI 3 Ss nela.|jena| 2 »|o|2 | 2 |52| #4 | es SI ERIE: &|&l2i 3 lez] o| Siselezle S| = | 3 S| 2 || 3 muy © | oulasilias es = zs|l21515 1382| = | e|23|22[=8| = FlEIJA< lass | FlE2|E5|82| © Agonum quadri- punctatum N |. m A. ruficorne | alle? A | 40 42 : | Oly Alice dla Hate three A. sexpunctatum Bag PP 0 4 : | ah | ee Roh ope hern-most A. thoreyı l northe: J Teale Lar ae se lal Sl | localities are A. versutum Sigal ioe ui . : ea | ee uncertain or A. viduum se Pee eee > | ee Weniemaxe, Amara aenea el Sat | auf A. alpina pr R ee In 2 A. apı icarla ee en | OS A. aulica Pile | Aal na an oa aa. A. bifrons ee | Bu Hl le A. brunnea | |) a | le vn || A. communis ZEN, ES Er a | rn eg | fc A. consularis Pr df He) ca, Poe i The finnish A. convexiuscula a | le | fey | eae | lle specimens A. crenata —|— ae big u A. cursitans SI | Fee ro | coal aan War IR | 1A al ehe ; A. curta Fit] ?2|?2|+ A. equestris eile eae te |) ieee es A. erratica rare St yt | till?) 2], A. eurynota II? Pl | ++ i A. famelica Fe) a) sell ee — |||= — | — A. familiaris +1 ?/+(+|+]? ?I?|-|- A. fulva +1 ?| +14 |+]| + = || - a A. fusca 2 al a] | eel eae Waa ee | ern il ome A. infima fale We ea I miele A. ingenua in 5 : A. interstitialis | = lee © Ian] |= I> ga 0 = + + ar! BR aes = foe A. littorea ® Eee . = Zu _ ap PUES | ps a A. lucida I al nen, ol: ae ae ae Par ee se | — || — LITE A. lunicollis Pah edad ae ae ais Shiekh wind: A. majuscula | from east A. montivaga BEN 1 = Pf ] across the sea. 6.6 6 IL — I 2192| + ie a RN gee A. municipalis s 5 3 NE ß er = — 11 u FL b A. nigricornis a | N 744 ae DN) | A. nitida 5 N Ser -- Sa A. ovata : Southern immigration Eastern Würm R h . . . . . T immigration hibernation mans = Qa. = S 8 8 = S2|83 ‚Sn, Sl a lleelleo in A) Spey lez =I=1. las| |< |(Is2|e7| © SS (sel Sell Sess Lets 2a ET ol eesti Re eH | oa (CHa etn | fers za | |e sat Were 3 |£=| 3] = |Io2|0* Se ny) iS :O & how 5 s | bee) 24 ea = a |o| >|» loe|< |=# [288 |E:]| = 51%|2|° |s2]$ |$ |E2j22]28| = s/21212]2=|3|3|a<]s2|es| 2 2/2125 [.8|: |: |231:2[22| = -IF|al| o Amara peregrina — (-/—|+1-|—-|—|— A. plebeja | en a ei BER er) je | ieee | le |e A. praetermissa +|—-/1+]1?|+1+1?|+|j+1?]|?|p- 742 A. quenseli -/+1/+|+|+|+1+|1+!1+1+1?|p. 742 A. similata U ce ||| | | ey | ame || Br een | AP A. spreta BS SEM | eae | re a | | || el A. tibialis Er Sea | ri | es A. torrida Be ri #11 2.119,09 Anısodactylus bino- Toßnorthern ratus A. nemorivagus N | | ae A. poeciloides + Sweden directly from Finland To northern Sweden directly + | ! + RE NN Asaphidion flavipes |+\— | + | + | + : oe FSAURIAInd. A. pallipes +1 -|-|-|+|?2|-|?|+|1—-|—-|p. 742 Badister bipustula- tus = [+ B. bipust lacerto- sus Boll | | ul am B. dilatatus Bel 5 RT = | oe B. peltatus mite | mrad sre |} ean | email rao rl ie re era | B. sodalis A| arco | er bapa | ar) oped | Sara Excl Sere ore B. unipustulatus we | Bembidion aeneum | + |—| ? ie +/+] -|+!|[pp. 399, 762 B. andreae polon. |* | | = |: | + B. argenteolum Sra ber ea ran Ei Gar ia oo | | Saag ee ee Wes B. articulatum Ar rea |e || Se | fe = = B. assimile le ss se ? B. azurescens TAs aa exer at sated aa) ee a ee B. biguttatum Ze | acl B. bipunctatum ay Ae lice telly Bull? 7.50 B. chaudoiri ? B. clarki ae + || - - = B. crenulatum Nr r — B. dauricum a p. 761 B. dentellum Akal seamed hese eae ct Al = B. difficile ?\?|+|+| 4+]? qe 744 B. doris +1? | +l1+|+1|1+ = B. Fellmanni >2|+|+1+1|? | P- 743 B. femoratum #12 | +1? )/+]2 )—? |+)1—|—I|p. 74 B. fumigatum + = Eastern Würm immigration hibernation Southern immigration s| |3| |2<|s® x |< = Ss S a = ze 52 i) S15] 6 Bale |MZol4s! &| s 2) 3/2] 5 |S2] 3 | |53553|22) 2 4|a/2i = lezl o| s/ecies lea) = 2/2/3812 [25] Z| 2 Jas]s 25] = lela < i832) fesleslecl 6 Bembidion gilvipes | +|—|+4¥+ ]+ | | — B. grapei — 1717121? 1? |—q+l] +l +]— p. 401 B. grapeioides ?I-|+1+1|—||p. 74 B. guttula +)—[+ +) +1+ 11-1 1—|+ B. harpaloides | a - ne B. hasti = | - - — | +J + + + an ? p. 748 B. hirmocoelum | | | Bram BASS | SL | a EL ay B. humerale le ee B. hyperborae- orum an ll ln 7 B. illigeri ll | ee — [| — 1 B. lampros +1? | +lJ+|+|? |-1?I-|—-|— B. lapponicum Sb | -J— PIA] 2? ]—llp. 741 B. litorale +{—}--1—| +] +]/—l+]—|—1 ? B. lunatum ee oe p. 763 B. lunulatum Se die) ||| Sa eee | fe | SE B. minimum +? | ++ tp N -- +) — | B. monticola —-|—|—] ?|+ EN B. nigricorne Sree fms (eae e309 fay | B. nitidulum Se Wee er ie TE Rip! 745 B. obliquum a Zur tea ig a | ea ae || Ser B. obtusum au | u IN i B. octomaculatum | + |—| +| + ae = B. pallidipenne Se) || ois y ht { B. prasinum | pe | 5 nn B. properans ler or WR B. punctulatum je Kae N al Bea B. pygmaeum || ly u a N a | | Mar) B. quadrimacula- tum | | — B. quinquestriatum | + ? Sea B. repandum | — - ? se fe ees fat B. ruficolle Pend ers Pa) | el tesa I e B. rupestre +[|?|+1+|+|+|—-|+1[1+1?|? B. saxatile +1? I+1+|+|+1—-|?|I+|+| ?]|p:- 8o1 B. scandicum lee je | ep. 740 B. schüppeli ?|- P+] +s ]—qeitl]els+ip. 740 B. semipunctatun | + |—|—] + | ~ ue a B. siebkei ie i a + eel |) 992 1 Apparently immigrated on a more southern route. Southern Eastern Würm immigration immigration hibernation y ) = a a BE > = ° Z un Ss nu 8 = n ws as) - 3 2 oS n S vo — Across the Gulf of Finland Between Gulf of Fin- land and White Sea In southwestern Norwa Through Kuusamo-Salla (south of latitude 64°N x 2 - ° Zz = In m 2 n o 3 = = In ° [= [= = Through Skane Through Skagerrak Directly to Öld-Gti Through Kanin-Kola In the northeast (Veranger-Petsamo) On the White Sea Bembidion stephensi | + striatum — tibiale ~- tinctum — transparens — unicolor an ustulatum + Sit 2 +I+|+ |p. 389 varium velox ? | + a te ? . virens or a Blethisa multipunc- tata Brachynus crepitans Bradycellus collaris csikit harpalinus ponderosus S| el = similis . verbasci Broscus cephalotes Calathus ambiguus erratus fuscipes melanocephalus micropterus mollis . piceus Calosoma auropunc- tatum denticolle — Inquisitor +h investigator —|—/+]—|— = ut reticulatum ? sycophanta + |— 1 Carabus arvensis + auratus = cancellatus = clathratus + convexus 4 ute by by oy ty by by by by be ? Wp. 745 ll le te In Ie 11,0.3807 +++++ v ++ by by ty by by VU 4- + + | | ! | i ? ?I—-|— | -||(To Finland, ia lialso,fsom Er Sweden + ek 2 AS) | wo + | { | | | j | MENFSES ++ ++++++++ iE v vw | | | | | | | uv | + + | | | | | | | rer Anthropochorous To northern ; | Sweden partly Zi ll Pl ur across the ae en el Bothnian. Sea: AAAN VV yt tu | + +++ Southern Eastern immigration immigration Würm Fr (eastward as far as Nordkapp) In the northeast a) = & = = 62 Set © = =} (6) o a = n n ° ni oO < Through Skagerrak Directly to Old-Gtl Between Gulf of Fin- land and White Sea Through Kuusamo-Salla Through Kanin-Kola In southwestern Norway (south of latitude 64°N) In northwestern Norway (Veranger-Petsamo) On the White Sea oO (S| og a n ie En = ° i oe = Carabus coriaceus +121— C. glabratus ae —| 9 ? C. granulatus ao || || 24 us C. hortensis +/+]? ofl C. intricatus pay | eal | et ||| ae | ee | na || ee | C. menetriest — |e |] | — | ee |] Probacı C. monilis —|+]/ef—]— we Dy | | | ee || Ea { anthro- C. nemoralis ala | > ek |S pochorous C. nitens +1? | +12 |+1+| eleal+tie2] pie 74 C. problematius \+| 2 |-|2|- I -) Pl +I+[+| p|P- 744 C. violaceus SB) || eff ef) A pa | | | — p. 742 Chlaenius costulatus |__| _.|__]} ? |} + | >| — ae C. nigricornis ee P e C. quadrisulcatus \ı |__| 9 a wi | ek || C. suleicollis ra | u 5 See pe cae DREI AR keg | C. tristis 419 hy | ar C. vestitus Ba | | ses | ane | Cicindela campest- ris +) Pah P+]? )—]ele}—i— C. hybrida ur ale | Se | bet na ER C. maritima le] wr) si > C. silvatica or || | | Clivina collaris fase || ae || oy ih, | | | Anthropochorous © er... ala: Kalle ae || ae || Cyhrus caraboides |, | _ a | || Cymindis angularis a le ; C. humeralis ay | Si ni Fa: 0 : C. macularis +|+1+[?|+|?|-| = |< | — (& vaporariorwn ee he tt Demetrias imperi- i : alis 5 4 N 62a unl | io D. monostigma 2 I | >) Diachila arctica a a | | ic D. polita 20 | ar) ar | a, ala Ka un 2 Dichirotrichus pu- bescens +|?} 2] - |? |+1+ #12. zs D. rufithorax AR ati geal ec ca ER a | Dolichus halensis ey | ||| ae aca | ue | ac | || Dromius agilis CON ORE SS be Oe eee Pre Southern Eastern Würm q 2 E : 5 5 2 ; Remarks immigration immigration hibernation y (eastward as far as Nordkapp) Across the Gulf of Finland In the northeast G: Ta! 32 Lo ov Zo = o 53 we nn os Fe Sie on as ° 52 In northwestern Norway Ss os 2 ° B a a” 3 3 na s on 3 S) iu = m Through Kanin-Kola (Veranger-Petsamo) On the White Sea 1 = zs 5 aie on ane s =o Be >> u (de) < o oD Ss : ee o m = 25 m (ea Directly to Old-Gtl oO S ou ee nN < or) 3 ° ~ a m Dromius angustus fenestratus linearis longiceps marginellus melanocephalus To Einland nigriventris wee Sweden quadraticollis || || + = ae eelliines eee quadrimaculatus| + | ? : quadrinotatus | + . sigma af Dyschirius aeneus ln + + + +++++++ | | | 1 from Sweden SSSssyysys + + +0 v + + | | | | angustatus chalceus globosus ? helleni | fen impunctipennis |+ | +|\+|I?| + intermedius ate hs Ip + Re ep lüdersi +? I ++] + neresheimeri 12 E AN Ga nitidus obscurus a ea] | politus Au ae “rufipes’’ ll ai salinus ae we | er = ee Septentrionum Bau | JO | na . thoracicus eta) ge ea | enc le es Elaphrus angusticol- lis — | | — || E. cupreus +1? E. lapponicus —|— E. riparius + E. uliginosus Harpalus aeneus H. anxius H. asureus le H IT. Dp. 742 ote st ++ Vv VU + ae) + + + + v Vv SESSSSsssyyvygygyys + fe + Br + ke) NI 6 | + isst: eo | [ew estate | +0 + -VBEVDENEUEU Vv + + + | | i ++++++ H + | calceatus 441149 distinguendus Te | a |e ee cece 12 ie ea ee | reed ie Eastern Wurm immigration hibernation Remarks Southern immigration (eastward as far as Nordkanp) In the northeast ac) =| & = ee om ° = = (©) o | ~ wn wm ° > iS} < 1 = fy, baal lo} = 3 (do) Ss © o ES Ba © (22) Through Kuusamo-Salla In southwestern Norway (south of latitude 64°N) In northwestern Norwa SJ iS) m [= .S E 3 ni = ot) 3 iS) = „=! = Through Skagerrak Directly to Old-Gtl land and White Sea (Veranger-Petsamo) On the White Sea oO S od ad un oS OD 3 ° i de) = Harpalus frolichi fuliginosus griseus hirtipes latus luteicornis melancholicus melleti neglectus nigritarsis oh ge picipennis sre ha pubescens 9 punctatulus ; puncticeps in puncticollis ie Zu aie | -w iS + + i v vu le =| wv | | | | | | | t+wtt ttt | ba OJ ©) wt tt +++ + | | v | | | v v quadripunctatus rubripes rufitarsis rufus rupicola seladon serripes servus smaragdinus tardus vernalis winkleri Lebia chlorocephala L. crux-minor L. cyanocephala Leistus ferrugineus L. rufescens L. rufomarginatus Licinus depressus Lionychus quadril- lum Loricera pilicornis Masoreus wetter- halli Aa esas | anal | NEST LE (20 ag | | ++ +++ + i | | a 5 | Perhaps introduced | into Ab Abo ‘vu ++ tot + +evvdt tuwt + + SSP SRS STR TERT I I in I ee gl ga ern | + i | oegy i | j Uv + ba ©) SI r NI bh 65 6 552 ©) vu tu + | | b ++ Probably EN aes | AM - | — | — 11% displaced to Sweden } vu = i v ı + v v Southern immigration Metabletus foveatus M. truncatellus Microlestes maurus M. minutulus Miscodera arctica Nebria brevicollis gyllenhali gyll. balbii livida nivalis . salina Notiophilus aquati- cus biguttatus germinyi palustris pusillus reitteri . rufipes Odacantha melanura Olisthopus rotunda- tus Omophron limba- tum Oodes gracilis O. helopioides Panagaeus bipustu- 22222 222222 latus P. crux-major Patrobus assimilis P. atrorufus P. septentrionis P. sept. australis Pelophila borealis Perileptus areolatus Pogonus luridipen- nis Pristonychus terri- cola botee tt + + +++++ oO Es om 1 n = i") 3 fe) De = m Through Skagerrak Directly to Öld-Gtl Eastern immigration Across the Gulf of Finland Between Gulf of Fin- land and White Sea Through Kuusamo-Salla Through Kanin-Kola Würm hibernation n northwestern Norway (eastward as far as Nordkapp) “Jn the northeast In southwestern Norway (south of latitude 64°N) (Veranger-Petsamo) On the White Sea I — — + = — ZA al cl 2 ir | ea een za le ae \ ee | ee. | ears Oe pace a a Oa 4 | 2 2 Aller me ZU ll: let ua, | | (eae | et || a | Vicar | is ae re Im el ae | Bee | | ap) ie are Iran Dank Rd eee Zi tanz le2 2 7a | ar | 16a | um | ee Vis ie | a Nine | ca baat | ea wa | Ps fs 2 ar deo lata leer lisa EA Visas | ee | Wou | eceee| Pe sap | Win Vs anal Ve 0 ? ans || Pee S| Wear lead Gece eae ean ZU Ele, ?2| +1 ?| + — u | ee Alban me m + BER. = A | bt Bes = + = tit \ | +/+|+]| ? 2. +/+)+1? +|? +|+|+ AU | +/+|+|+ ++] 21 - a Wh a N ? x ZU 2 1162 ? ei let 2 le +/+/+ |? | a3 etal aie aaa | To Finland also from | Sweden p. 756 p. 742 p. 742 p. 763 p. 409 p. 405 b. 747 p. 744 Anthropochorous Southern Eastern Würm immigration immigration hibernation Namens ie 3 S E = >33 = = os) Se Eu lese “|- |& |es|$ |< |Isz|ls2| © 215 |%81|22| 8 | = |Fers|_&| § „|81%|= |Se| 3 | 8 |E3|E5|32| 2 3 |=15|& |s8l3 |: |e3|22|82| 2 22 |2 ele | 2 jes1g2jee 5 s|%|>2|, |52|% | 8 |e2]=s]°2|7 21|2|81 |=S| | © osleeiss| = Ele lal i@sle JE tsslsslee| 6 Pterostichus adstric- tus 2\-|-\-|>2|>2|-|+/+|+| 2 | 74 P. aethiops — || =|] 2 + = = P. angustatus | | P. anthracinus +1—|+1-|+[|- P. aterrimus qe) ae ae I} Ae |) ae P. coerulescens +|?I+1+1[-+1|2? P. cupreus 9 2 |) 4] 3 P. diligens HE |) |) Ne a eS |p] ae |p a) P. fastidiosus a eal ese || | Fae NF P. gracilis OT) na cs || Mem ea 2 "als ll 2a |) Canary P. madidus a displaced to P. middendorffi ss | Finland R. minor er 4 2 Re A P. nigrita ale au P. oblongopuncta- tus 4/9) 4+) + 2 2 | tis |e |e) P. punctulatus fe zs I P. strenus ri_—}+ ir} +] e+] — ae l[ +l | + | p. 395 P. vernalis ol | a Vs nt P. vulgaris LD te | @ | iS Sphodrus leucoph- ; thalmus ee le J Anthro- { \ pochorous Stenolophus mixtus | eel eal st S. teutonus = Stomis pumicatus Bea aes tate ||| kha | ire - Synuchus nivalıs = 22 2 | 22 Sn ? a Tachys bistriatus 4 T. bisulcatus BD | al) Tachyta nana ?I—|-|I-|+[|?|I—-|1-|-|-|— Trachypachys zet- terstedti weg rn, — | | || — Trechus discus | ln n T. fulvus ne = [= +1— | —| —|| p. 791 T. micros ed le 7 = T. obtusus tpt canal | an +|+|-|-—.| p. 762 T: quadristriatus Teal) cee tan In ht Fe Southern immigration Trechus rivularis T. TE Trichocellus cogna- T. T. Zabrus tenebrioides 707 Eastern Würm immigration hibernation Remarks (eastward as far as Nordkapp) as) E & i‘ ce — fo) = 3 1) o de = n nn ° bu o < In southwestern Norway (south of latitude 64°N) In northwestern Norway Through Kuusamo-Salla Through Skagerrak Directly to Old-Gtl Between Gulf of Fin- land and White Sea Through Kanin-Kola In the northeast (Veranger-Petsamo) On the White Sea vo i=] om Be un <= lott} = ° bu = ll rubens secalis tus mannerheimi placidus Total*| 2655 | 26 |183|| 130 | 231] 72 | 22 || 72 45 | ıı (+-16)| (+ 224443473) [+ 9)}(+-45)|(+25)}|(+-34) |(+-26)](+-20)|(+55) = 281] = 140 | = 226|= 203] = 240|= 117] = 47 ||= 106| = 95 | — 65 | = 66 nn | | nn | 280 (+ 12) 267 (+ 23) 97 (+ 33) = 292 (81,6%) = 290 (81,0%) = 130 (36,3%) *In the sum and percentage calculation, 7 anthropochores are excluded (Carabus auratus, C. monilis, Clivina collaris, Lionychus, Pristonychus, Pterostichus madidus, Sphodrus), whereas three subspecies (Badister bipustulatus lacertosus, Nebria gyllenhali balbii, Patrobus septentrionis australis) are considered as separate entities. An example of how the multiple origin of a more or less pan-Fennoscan- dian species can be envisaged is provided by the “clarified” map of Bembidion saxatile (Fig. 116). The varied postglacial history of most of the species is easily understand- able. Strictly speaking, there is no question of a new colonization. We are dealing with the interglacial fauna, which was pushed to all sides by the ice and is now moving back into the old region. The influence of the last glaciation on the fauna and flora of Fennoscandia may be simply expressed as follows: This period had a decisive role for the present-day local distribution of every species. But the stock of Fennoscandian species was little changed. English Summary‘ 813 The two first parts of this book, published in 1945, contain a full account of the distribution, ecology, ”biology” and dynamics (i. e. power of dispersal) of every species of the Carabid@, 361 in number, occurring in Fennoscandia. Distribu- tion maps (part II) are given for nearly all of them. The present IlIrd part of the work deals with the material from general points of view with the chief purpose of elucidating the glacial and postglacial history of the Fennoscandian fauna. p. 7. Preface. The necessity of carrying out experiments for solving prob- lems within the domain of ecological zoogeography is maintained. p. 10. Introduction to the IIIrd part. This investigation is no complete survey of all ecological and zoogeographical questions regarding the Fennoscandian Carabidae. Above all the synecology was almost completely neg- lected for reasons given on p. 35 a. f. Any attempt to make a zoogeographical division of the areat treated on the basis of the Carabidae was also abandoned. The species populations in different parts of Fennoscandia seem unusually stable and there are but few examples of clear subspecies; thus — with the exception of the state of the flying-wings (p. 335 a. f.) — no investigation of the variability of each species within the area was realized. The word postglacial was used mainly in a biological sense and thus in- cludes the late glacial time (acc. to the geologists). — For the term ’Ausbrei- tungsökologie”, created by EKMAN (1922), in this book — also in the Ist part — ”Dynamık” was consequently used. In accordance with this ”ausbreitungs- ökologische” and ”existenzokologische” limits of the area of a species are called dynamic limits and existence limits respectively. p. 13. On modern insect taxonomy. Its three main features are: the siplirtting up of genera: the alteration of the names of species; the breaking up of species into smaller cate- gories. Experts are recommended to proceed with the greatest caution in dividing up wellknown genera into smaller ones. The postulated new units usually had better be treated as subgenera or species-groups. The genera — "for practical reasons” (in the widest sense of this expression) — should be too large rather than too small. It is of special importance that a specific name is never transferred from one species to another. In actual cases — in spite of the rules of priority — it is better to let the name in question lapse. The consequences of the priority-rules strictly applied, on the whole, are many unnecessary confusions, especially when old names, long forgotten, are dug out from the literature, ”interpreted’” and used to replace universally prevailing names. It would be worth while to consider re- garding specific names given before, say, 1850 and not (or merely as synonyms) scientificially used after, say, 1900 as invalid in favour of later used names. p. 26. On the consideration of the literature. The zoological — and particularly the entomological — publications during the present century have increased in number and size to such an extent that a complete mastery of its results, even within a restricted branch, is hardly possible. In ”border-sciences”, such as zoogeography, these difficulties are especially severe. The zoogeographer * Reproduced from the German original, pp. 813-842—General Editor. t(= “region” throughout the translation; cf. p. 822; suppl. scient. edit.). 709 814 must know, however, as completely as is possible all the literature concerning the animals constituting his special subject, in order to get fully acquainted with their taxonomy, distribution, ecology etc. He thus produces the foundation for general conclusions. Thereby he comes into contact with several auxiliary sciences, with botany, general geography, geology, climatology etc. in which he usually is but a layman. It therefore seems reasonable that the zoogeographer attacks his problems concerning the history of each species, mainly on the ground of his special know- ledge, i. e. proceeding from the animal studied, and thus that he must not feel himself too much bound by the auxiliary disciplines. His literature studies consequently can be reduced to moderate proportions. Thus a complete knowledge of the case treated is more important than that of the methods used by other investigators on other subjects. p. 29. On the task of the museums. On the basis of 10 questions sent to the curators of the large museums of Fennoscandia and Denmark the present status of the coleopterous collections of each of their own countries and the rules for their organization was put together. Some desiderata for the future were suggested which could be summarized as follows: Ail exactly labelled material is valuable. ”Duplicates” do not exist. It is necessary to strive after long series of specimens collected together in order to know the variability of the population. This means a considerable increase of work on the part of the museum’s personnel, a problem which could be solved, however, if their efforts on the whole were more concentrated upon the investigation of the fauna of their own country and the common ambition to establish a palaearctic or even a world collection was correspondingly reduced. The most practical measure would be to divide the home collection into a concentrated "type collection” (with say 2 go and 2 Q per species, excellently mounted) for rough taxonomical purposes on the one hand, and a large elastic” geographical main collection on the other. Special collections, such as those quoted in mono- graphical publications, from islands or other restricted territories, from certain habitats etc., should not be too much avoided but it is recommended to put with the respective species in the main collection a reference in the form of a label to such separated material. p. 35. On synecology and ”syngeographie” I take an extremely sceptical attitude to all sorts of collective treatment of animal species as members of an ecological community” or of a zoogeographical ”region”. Especially in ter- restrial animals which are non-sedentary the "struggle for space” must play an unimportant réle. Further I am unable to believe that the organic nutriment pro- duced by the green plants as a rule is fully exploited by the “consumers” or even that there exists an exact limit up to which this nutriment may be consumed with- out disturbing the "equilibrium of nature”. In my opinion the terrestrial ”zoo- coenoses” are as a rule non-saturated. The idea of FRIEDERICHS and THIENEMANN of zoocoenoses as something like an organism seems to me quite out of touch with realities. Under all circumstances it must be worthless to create ”zoocoenoses” founded on Carabids alone as these constitute merely a fragment of the consumer group and thus cannot be expected to behave as a constant. It is impossible to maintain, however, that synecology must be a biological branch working with constructions only, lying beyond the reach of exact results 815 710 But the condition precedent is a sufficient knowledge of the autecology of each species constituting the biocoenosis. p. 43. The uniting of organisms into groups with + corresponding distribution or the creating of regions”, each with its characteristic animals and plants, may be termed ”syngeography”. The regional division of greater or smaller parts of the earth becomes quite different according to the animal or plant group used as a starting point. On the other hand it is easy to observe that the proposed animal regions e. g. of Fennoscandia (EKMAN 1922) agree in all important features very well with the botanical, i. e. the forest regions. I have found no reason to propose a new regional system on the basis of Carabidae but feel it satisfactory enough simply to take over the forest regions of the Fennoscandian botanists (fig. 61). Analytic part. p. 48. Experiment journal. The purpose of the experiments made was to isolate such external factors affecting the animals in nature as temperature, moisture and light, and to study the reactions of the animal to them. Above all the preferenda were investigated by means of different sorts of ”orgels”. In many cases also the resistance against extreme temperatures and desiccation was Studied. It is of special importance to point out that the figures obtained are not absolute, partly because of uncontrolled factors influencing the experi- ment, partly owing to the fact that one and the same species — even the same individual in different situations — often shows different reactions. Consequently I tried to avoid these difficulties by making the same experiment — simultane- ously or in the nearest possible identical circumstances — with more than one species. Thus it was possible to obtain at least comparative figures, e. g. for the temperature preferenda of a series of taxonomically and ecologically closely related species, which proved sufficient for the tasks selected. The following experiments are of general importance: — I. 15 spec. of the clearly eurytopic Pterostichus nigrita were individually marked and examined during the 10 following experiments in the temperature ”orgel” apparatus. Only the consecutive order of the individuals in the apparatus was noted. Their distri- bution proved to be regulated by chance only (Diagr. 2). The eurythermous cha- racter of this species thus seems to be founded, not on the presence of several bio- types (ecotypes) within the species, but simply on the relative insensibility of each individual. — 2. The question of stable and labile preferenda, briefly treated in an earlier paper (LINDROTH 1943 a), was attacked on other species. Pierostichus anthracinus, Harpalus serripes and Brachynus crepitans have a stable, Harpalus punctatulus a pronouncedly labile temperature preferendum. In the last- mentioned case the instability of the reaction to humidity seems to be decisive. — 3. Brachynus crepitans and Agonum dorsale regularly occur together without ha- ving any dependence on each other. They represent the rare case of two species with nearly identical ecological demands. It might thus be expected that a comparison between the reactions of these two species in current preferenda and resistance experiments would reveal those of the greatest ecological importance, i. e. those in which they behave in the same manner. This proved to be the case, regarding temperature as to cold-resistance, regarding moisture as to the prefe- rendum. It ought to be justifiable to regard the experimental results on these two ranges as especially important for + xerophilous Carabids. 816 711 All the experimental results, with two exceptions (Diagr. 11, 46), were treated by simple arithmetical methods. Above all it is necessary to point out that the distribution of individuals in the temperature "orgel” is not regulated by chance only as the hot zone of the apparatus constitutes a complete barrier, without an equivalent at the cold end. Therefore the common usc of statistical methods on these figures gives no advantage. In pronouced hygrophilous species the average temperature preferendum, however it may be calculated, is much too low. : The simple apparatuses used in the different experiments appear from fig. 3—8. A new construction is the so called "universal orgel” in which the animals were. exposed to the simultaneous influence of 2—3 factors (temperature, moisture, light). The corresponding diagrams are seen on p. 141 a. f. In the journal (p.69—111), to which reference is made from every experiment treated in the text, complete data are given for each separate experiment. p. 112. The "limestone-species”. It is a well-known fact that many plants are dependent on limestone. Also several insects, among these some Cara- bids, are usually termed calciphilous, in Fennoscandia above all species characte- ristic of the islands of Gland and Gotland, partly occurring also within the iso- lated cambro-silurean districts of central S-Sweden or SE-Norway (fig. 9, 10). The genus Harpalus is especially rich on such species, including the only three found in Scandinavia exclusively on Oland and Gotland. So a series of 15 Harpalus species was selected for closer experimental investigation. p. 119. The Harpali are clearly polyphagous and their, in many cases very restricted, geographical distribution — above all in limestone districts — cannot be due to specialized food-habits. p. 120. In the substratum-”orgel’” all species tested seemed insensible to diffe rent admixture of CaCOs, even if dissolved in water saturated with COs (tab. 2, 3). p. 124. No more the species in question seem to react to the degree of pH of the substratum (tab. 4). p. 125. A series of experiments was made in order to show whether the suggested ”calciphilous” Carabids are able to distinguish between gravel from lime-stone and from siliceous rocks (tab. 5). Actually, 5 species of Harpalus (in addition Panagaeus bipustulatus) seemed to prefer the firstnamed sort. p. 127. With these 6 species the experiments were repeated, but after sub- stituting the siliceous gravel by clay-schist gravel, free from CaCO; but as regards. other qualities practically indentical with the limestone (tab. 6). None of the species showed preference for the latter sort in this case. p. 128. The experiments accordingly seemed to prove that an eventual confine- ment to limestone, shown by the Carabids investigated must be due to demands other than chemical. So their reactions to thermal and hygric factors were tested in the usual manner (diagr. 19--24). A summing-up given in tab. 7 (p. 137) shows that the supposed ”calciphilous” Harpali are to a great degree either thermophi- lous or xerophilous insects, or both. p. 139. An attempt to decide what factor in each species plays the most important röle was made through experiments in the "universal orgel” (vide p. 85) where especially the strength of reactions to the simultaneous influence of thermal and hygric factors was studied (diagr. 25>—35). Among the 11 Harpalus species tested azureus (p. 174) proved to have the greatest demands for warmth, serripes for 817 712 dryness. Contrary behaviour is shown by aeneus, the most ubiquitous species, the thermal preferendum of which not only lies low but is also easily disturbed by the influence of other factors. This may be the criterion of every eurytopic and widely-distributed species. , p. 176. Harpalus Melleti and rupicola are peculiar, not xerophilous and but slightly thermophilous. Evidently, the restricted distribution of these species is due to demands which cannot be elucidated by means of preferenda experiments. The cold-resistance (diagr. 20) — determinating the length of the annual period of activity — seems in these cases more important. p. 177. The thermal qualities of the rock and gravel of cambro-silurean limestone compared with those of granite were first studied by means of laboratory measure- ments on small quantities (diagr. 36—39). All reactions proved to be more rapid in granite which therefore had the highest daily maxima and the lowest minima. p. 181. To a more pronounced degree the same differences were observed in field measurements on each side of a fault fissure (cambro-silurean limestone contra granite) at Dir Rattvik, Sjurberg (fig. 14—16, diagr. 40). It seems justi- fiable to say that the cambro-silurean limestone — as regards temperature — alters the microclimate in an aceanic direction. The com- paratively high minima no doubt have a great biological influence. p. 187. The higher evaporation effect of limestone gravel was demonstrated by a simple experiment (diagr. 41). p. 188. The principle biological advantage created by the cambro-silurean lime- stone no doubt is the lengthening of the annual period of activity. With the sum- mer-temperature during a period of 20 years at Visby on the isle of Gotland as a starting-point, an attempt was made to estimate this prolongation (diagr. 42, tab. 8). One comes to an average value of three weeks. p. 192. A hypothetical map of southern Sweden was drawn, in order to illust- rate the positive thermal influence of limestone on minimum temperature in May and September (fig. 17). This map, however, is valid only for thin soil-layers resting on firm rock. p. 193. Archaic limestone, e. g. marble, or limestone ocurring as moraine only, both widely distributed in central Sweden, seem not to cause any positive bio- logical effect on the ground-fauna, as, no doubt, would have been the case if the influence had been of a chemical character. One feels inclined, therefore, to regard the "petrophilous” insects, especially treated by HoıLpHaus for the Alps, as controlled by thermal and hygric rather than by chemical factors. p. 196. The autotrophic plants are no doubt in many cases dependant on di- rect or indirect chemical qualities of CaCOs. It seems evident, however, that botanists at present show a tendency to underestimate the thermic and hygric importance of limestone. p. 198. The fauna of the isiands. The Carabid fauna of 21 Fenno- scandian island-districts (fig. 18, tab. 9), each with one or two comparison areas on the mainland, was analysed. The species were arranged according to dynamic groups (tab. 10—12), hibernation types (tab. 13) and ecological groups (tab. 14). p. 235. 1. Hailuoto (Karlo) is the youngest island treated (maximum age 2000 years) but, nevertheless, has quite a normal” fauna, no doubt due to its situation — extraordinarily favourable for hydrochorous dispersal — right at the mouth of a large river. 818 713 p. 236. 2. Aland (mainland). According to A. PALMGEN the flora of Aland has immigrated to the island mainly from the west, from Sweden. The composition of its Coleopterous fauna suggests that the path from the east, from Balticum and the mainland of Finland, has played an at least equally important röle. p. 241. The ”western” group of Aland-Coleoptera contains many flightless spe- cies and forms, the eastern” one — with few exceptions — only species capable of flight. The flora shows no similar division and the direction of prevailing winds (fig. 20) gives no explanation. p. 256. The indisputable fact — documented also in other parts of the area (p. 291, 503) — that an anemochorous dispersal of insects in the Baltic area has taken place more from the east toward the west than in the opposite direction, was studied by means of experiments in a flight-direction apparatus (fig. 8, p. 109). It appeared that at least some Carabids, swarming as a rule in the evening, show a tendency to start their flight towards the sun, i. e. in the main towards the west (fig. 21, tab. 19). p. 258. The immigration of flightless beetles from Sweden to Aland may have taken place by means of ice-drift in the spring. p. 260. 3. The "skargard” of SW Finland (excl. Aland). In this archipelago the ice- and waterdrift seems to have been of still greater importance (cit figs 22): p. 263.4. Hogland and5. the other "Outer Isles” in the Gulf of Finland. In accordance with its considerably greater age Hogland possesses a higher percentage of flightless Carabids than the other ”Outer Isles”, the youngest of which, Seiskari, shows a higher figure of macropterous species and forms than any other Fennoscandian island. p. 273. 6. Valamo (in Ladoga). The fresh water may have facilitated a hydrochorous transport. p. 274. 7. Värmdön. In accordance with the position close to the mainland the fauna is completely normal”. p. 276. 8. Ösel and Dago. The natural conditions (incl. the flora) agree in many respects with those of Gotland and Oland. The Carabid faunas are more different. p. 279. 9. Gotska Sandon. This is a standard example of a young and highly isolated island having got its fauna totally by means of over-sea-immi- gration. The percentage of flightless species and forms, in keeping with this, is extremely low. The famous insect species known in N Europa exclusively from Gotska Sandön are all capable of flight. The fauna has no doubt invaded the island by means of anemochorous( incl. anemohydrochorous) transport, of which concrete observations have been made. Introduction by man seems to have been of minor importance. p. 285. 10. Faron. The vicinity of the mainland of Gotland has given the fauna a more normal” stamp but a direct invasion, mainly by air, from the east has given remarkable additions. p. 289 a. f. 11—13. The large Baltic isles, Gotland, Öland and Born- holm, are at first briefly described separately and then (p. 298 a. f.) the common problems (also for Osel — Dago) concerning the origin of their fauna and flora are discussed. It is striking that the relationship of the Oland—Gotland flora — 714 819 in accordance with the similar geology — points to Osel—Dago whilst the co- leopterous fauna points to the south (tab. 23, fig. 23). p. 303. A list of Coleoptera known in Sweden only from Öland and Gotland contains 14 species common to both isles, 5 of which are flightless; among the remaining 46 species only 2 are incapable of flight. Taking the whole Carabid fauna under consideration we find likewise that species common to Öland and Gotland — and also to Bornholm — are characterized in general by a restricted power of dispersal. It is difficult to understand these facts without the assump- tion of a postglacial, + direct land connection between the islands in question. p. 308. For geological reasons such a connection hetween Oland—Gotiand and the present south coast of the Baltic can hardly have existed later than during the early Ancylus Period, Pollenanalytical investigations of recent years have proved, however, that the flora already contained many heatrequiring forms at that time. p. 310. Examples of equivalent Carabids are Harpalus azureus and Calathus mollis. p. 312. 14. Ven. The most characteristic feature of this island is the radical influence of cultivation. Several Carabid species have no doubt been introduced by man. p. 314. 15. The "skargard” off Goteborg. The percentage of bra- chypterous species and forms among Carabids is remarkably high. This ‘seems, nowever, to a great extent to be due to anthropochorous transport. p. 315. 16. Orust and 17. Hvaler show nearly “normal” faunas. p. 318. 18. Hitra (and adjacent islands). This district is characterized by an exceptionally high percentage of functionally brachy- pterous Carabids. This fact is not due to any extraordinary influence of selection but to the former existence of a glacial refuge in the immediate vicinity. p. 321. 19. Donna (and adjacent isles) and 20. Lofoten (and Vesteralen) show the same feature to a somewhat slighter degree. It is significant, however, that in the outer isles (Vzroy, Rost) the brachypterous element is most domi- nant. Among the islands in Troms and Finnmark (p. 324) this holds good for Nordfugloy. This case is especially interesting as it forms a clear proof that the brachypterous element has not been introduced by man. p. 325. An analysis of some other islands of Western Europe (tab. 25) has shown that the isolated and exposed position of an island (e. g. Helgoland) does not in itself cause a selection in favour of brachypterous species and forms, at any rate not within the comparatively short postglacial time. p. 331. A summary of conclusion made on the basis of the island-faunas studied makes it i. a. clear that on moderately isolated islands the immigration of flying insects is rather more favoured than on the continent, whereas, of course, flight- less forms are greatly disfavoured. p. 334. Wing dimorphism. In the fam. Carabidae the wing dimorphism is of the asexual type. In Fennoscandia 50 species (out of 362) occur in both forms (tab. 27); 18 other species have proved to be dimorphic in other parts of their range. Some species are rather polymorphic but all forms with not fully developed flying wings are here termed brachypterous. p. 343. The wing dimorphism of Coleoptera as a whole seems to have a here- TS 820 ditary basis. This has been proved for Sitona hispidulus (JACKSON 1928) and Pterostichus anthracinus. Brachypterous wing is a dominant and the mac- ropterous individuals, consequently, are homozygotes which cannot give brachypterous offspring, unless by means of recurrent mutation. p. 347. In several constantly brachypterous species of Coleoptera — likewise in brachypterous individuals of dimorphic species — the reduction of {lying wings may be correlated with other morphological changes. Of these the iollowing are represented among the dimorphic Carabids investigated: a) Reduction of flight muscles. This is remarkabic also in a macropterous specimen of Pterostichus vulgaris (Dir Lima) with completely reduced muscles. b) Reduction of metathorax. In Calathus melanocephalus there seems to be no difference in this respect between the two forms but brachypterous C. mollis is on the other hand almost constantly characterized by its shorter metepisterna. c) Fastening of elytra along the suture. A real ancylosis seems not to occur among dimorphic Carabids but a suggestion of it is visible in single (as a rule brachypterous) specimens of Calathus erratus and Ptlerostichus lepidus. p. 351. A case of preserved psychical (and partly muscular) functions ordinarily connected with flight in a brachypterous specimen of Bradycellus collaris is described. p. 351. None of the morphological changes commonly correlated with reduction of flying wings seems to be constant. It is therefore probable that they are due to special genes. p. 352. The influence of selection. The best study on the biological importance of flight to Coleoptera (espec. Carabids) is that of Dariincron (1943). He suggests a selection in favour of flightless forms in mountain districts as well as on most islands (exceptions are flat, tropical and young islands). According to him the brachypterous individuals, forms and species are favoured owing to their general higher ”viability” which he finds proved among other things by the breeding experiments made by Jackson on Sitona hispidulus. — A series of experiments on preference and resistance, simultaneously with bra- chypterous and macropterous individuals of the same species, was made (diagr. 43—46). No general “higher viability” of the brachypterous form appeared. Therefore, although the material used was rather small, it hardly seems justi- fiable to ascribe a priori a higher selection value to the brachvpterous form of a dimorphic Carabid species. p. 360. The actual distribution of the two forms in Scandinavia, e. g. in Notio- philus biguttatus, seems likewise to contradict the idea of a general superiority of the f. brachyptera. p. 361. The selective importance of flight powert itself was already maintained by Darwin, especially in the case of Madeira. Dar- LINGTON, on the other hand, stresses the opinion that for species living on islands or in mountains the power of flight must be regarded as “useless but not ne- cessarily harmful”, i. e. that this function in itself has no negative selective effect. He seems to base this view on the fact that flightless Carabids are evi- dently favoured not only in barren wind-exposed but likewise in densely wooded mountains. Also to my mind the wind exposure was overestimated by Darwin t(=“flight capacity” throughout the translation in accordance with Thiele, H. (1977): Carabid beetles in their environments. Springer, New York, 369 pp.; suppl. scient. edit.) 716 821 but there can be no doubt that flight itself in some situations must cause a pro- nounced negative selection, namely for species living in small areas surrounded by unhabitable land or by water. The + stenotopic species of certain altitudinous regions — often quite restricted — of higher mountains form excellent instances. Whether they are wooded or not has little importance; also pronounced forest insects are caught by the ascending convection currents as can be observed abun- dantly on the surface of snow-fields in higher alpine steps. p. 364. Stability, restriction and isolation of habitats favour brachypterous species and forms; variability, extension and moderate splitting up of habitats favour macropterous species and forms. Thus in periods characterized by alternating stability and variability — above all concerning climate — dimorphic species are favoured. This is applic- able to the Quaternary Period with its alternating glacial and interglacial epochs. p. 365. All constantly brachypterous species have no doubt passed through a di- morphic stage (eventually before the origin of the present species) during which they were exposed to factors working in favour of flightless individuals. A com- plete elimination of a recessive gene like "macropterous”, however, is not possible by means of selection alone, it must be due to the action of chance within small populations. In fact, the ”macropterous gene” may be present in many species gene- rally regarded as constantly brachypterous, and accidently the copulation of two heterozygotes may cause the appearance of a single macropterous individual, often termed as ”atavism”. p. 366. The distribution of dimorphic forms. On the basis of 25 maps (fig. 28—-53) an attempt was made to describe the late Quaternary history of some dimorphic Carabids in Fennoscandia. The foremost rule, clearly illu- - strated by Calathus mollis (fig. 28), is that macropterous individuals — "the parachutists’ — are predominant at the periphery of the species; a preponderance of brachypterous indivi- duals indicates an early colonized area. By this method it is possible to suggest for instance that Calathus mollis and Harpalus picipennis have invaded Oland (and Gotland) directly from the south and that Olisthopus must have reached Aland from Sweden. p. 377. Calathus erratus (fig. 35) is of special interest because it shows that se- lection in the coast-land of western Norway does not work as a principle in favour of the brachypterous form. ; p. 381. If two or more originally separate postglacial stocks of a dimorphic species have come in contact and fused together one may be able to detect the ”joint” as a zone of more frequent macropterous individuals, e. g. in Carabus clathratus (ig. 38) and Pterostichus lepidus (fig. 37). p. 383. In cases where the brachypterous form has caught up the macropterous one on all points of the periphery it is justifiable to state that the species in question has actually reached its existence limit, e. g. in Pterostichus vulgaris (fig. 39), Bembidion lampros (fig. 40), B. gilvipes (fig. 41). p. 387. At first sight some species are perplexing which in the north of Finland posses a + separate stem with high percentage of brachypterism, e. g. Bembi- dion guttula (fig. 42), Pterostichus minor (fig. 43), P. strenuus (fig. 47) and, above all, Bembidion transparens (fig. 45). One must assume a separate immi- 77, 822 gration route from the surroundings of the White Sea (in B. transparens also from the north). p. 395. The strongest preponderance of brachypterous forms in Fennoscandia is demonstrated by the western coastland of Norway. In three species, viz. Ptero- stichus strenuus (fig. 47), Bembidion aeneum (fig. 49) and Notiophilus aquaticus (fig. 53), the gene ”macropterous” has apparently been totally lost, in others, e. g. Bradycellus collaris (fig. 48), Notiophilus biguttatus (fig. 52), macropterous indi- viduals are extremely rare. Reasons are brought forward to show that this pheno- menon can by no means be due to selective action in postglacial time. The only acceptable explanation is that the species in question were isolated dur- wnigeact hiey ul a staiiGila cara ton. (Wurm) wal tihrein) ec exk meen. Trek wees along the Norwegian coasts. p. 401. An illustrating example is Bembidion Grapei (fig. 50). Its distribution in itself could give reasons for accepting the species as a postglacial immigrant into Fennoscandia from the east but the occurrence of brachypterous individuals gives a quite opposite picture. It is doubtless a glacial survivor in western Norway. p. 411. During the glacial period the small, isolated coastal refuges formed islands of organic life where selection was strongly active in favour of flightless insects. During the initial and final phase of every glacial period, on the other hand, the enormous and in part also rapid changes of habitats were in favour of flying in- sects. So, in repeatedly glaciated areas wıng-dimorphism must aimoplysia, considerable advantage. to van amsiect: p. 415. Among the dimorphic species treated, at least 9 may be regarded as "win- terers” within the limits of Fennoscandia during the Wurm” period. The present distribution of these 9 species within the plant-regions of the fjelds (tab. 29) shows that only Notiophilus aquaticus is capable of living in high-arctic conditions. Five species have never been recorded above the forest limit. These tacts strongly con- tradict prevalent ideas about the climatic conditions in Fennoscandia during the Last Glaciation. Synthetic part. pudt7. Lhe Cetin tivo nok) are a”? The idea of an animal or plant spe- cies occupying an area is a fiction. The areal limits are constructed lines joining the extreme (northernmost, highest etc.) finding places (or rather populations, if they can be established) of a species. In pronouncedly stenotopic species the area is especially disjunct. A hypothetical example by which one may imagine the real distribution of a stenotopic insect near its northern limit is given (fig. 54). The zoogeographical ”areal limit” — marked with crosses — may be defined as the lintessupaéito, whichwon suitable places: the) species,.in ques- tion occurs permanently. p. 419. The reliability of distribution maps. It is impossible to get a complete picture of the range of a species. But the map must be credible”. A characteristic of this state is whether, during a considerable period of investi- gation, all new added localities organically fit into the former picture without ”spoiling” it. According to this criterion the Carabid maps of this book are ”credible” and thus suited to form a basis for zoogeographical discussions. An idea of the relative exploration of Coleoptera within Fennoscandia is given in fig. 55. tin German = “Areal”; “area” is standing for this meaning throughout the translation (English summary excluded) and the terms “region” and “range” are avoided; cf. pp. 417, 813; suppl. scient. edit.). 718 823 A considerable defect of my distribution maps is that they do not figure the present state; they form the concentrated result of more than 100 years’ collecting. The number of coleopterists, however, is not large enough to make tolerably com- plete maps within much shorter time possible. But the indisputable fact that nearly all recent faunistical changes within the Fennoscandian area have been of a posi- tive nature, i. e. have implied an extension of area, reduces the errors considerably. For many species it would be valuable to make quantitative maps. Fig. 56 is an attempt to realize this on Miscodera arctica and makes the northern cha- racter of this species clearer. It is maintained that the ”point-method” of constructing distribution maps is, owing to its exactness, superior to every other and that not more than one species should be figured on each map. Otherwise the main advantages of mapping in comparison with a descriptive text are lost. Some exemples are given in order to show that, in spite of the quantitatively extremely fragmentary experience of the actual fauna obtained by the collecting entomologist, he catches notwithstanding this fact a surprisingiy high percentage of the actually occurring species. p. 430. The relationships of the Fennoscandian fauna. Only two Carabid species (Bembidion scandicum, Bradycellus ponderosus) are unknown outside Fennoscandia. Otherwise it is clear that maps of the total distribution of each species would be of great importance for determining its history. Owing to the impossibility of checking the frequently conflicting foreign data, I simply have not ventured, however, to construct maps of the total area of any Carabid species. The collective mapping of the total of Fennoscandiar Carabids (362 species) occurring in other parts of Europe (fig. 58) or of the world (fig. 59) is less dangerous. The wide distribution of many species eastward into Siberia is striking, especially against the background of the distribution of species numbers within the limits of Fennoscandia (fig. 60). As would be expected, the powers of dispersal (of flight) are as a rule better developed the wider the distribution of a species is (diagr. 47). p. 436. Species distribution among different plant re- gions. The regions of Fennoscandia established by botanists mainly on the basis of forest vegetation — here partly termed with simplified names — are seen on the map fig. 61, the distribution of Carabid species within these regions in table 30. The summary (p. 448) above all gives an impression of the poor fauna of the Fennoscandian regio alpina. Existence factors. p. 449. Climate. It is generally — and rightly — assumed that climatic factors have the greatest importance in limiting the areas of animal and plant species. Bio- logical and climatological investigations of recent years have shown, however, that the common meteorological figures — as representatives of the "macrocli- mate” — have only a restricted applicability to the factors actually operating, especially for the terrestrial ground fauna. The “microclimate” (the cli- mate of and within square metres) is here decisive. Frequently it is suitable to work with an intermediate conception, the ”\ococlimate”, which could be defined as the climate of hundreds of square metres. p 450. Temperature. The common method is to seek for connections be- t(= “capability of dispersal” throughout the translation; cf. p. 15 of Part I; suppl. scient. edit.). 719 824 tween biogeographical limits and some isotherm of an average monthly tempera- ture. For other than summer months this is rather unsatisfying; of all the months only the warmest one, in Europe July, is — from a biological point of view — tolerably characterized by its mean temperature. An isotherm map of July is given in fig. 63. It shows the following + pro- nounced plus-districts: 1. SE-Norway; 2. Gudbrands and adjacent valleys in eastern C-Norway; 3. the innermost part of Sognefjord; 4. the lake district (espe- cially at Malaren and Hjaimaren) of C-Sweden; 5. the inner parts of S Finland; 6. the surroundings of the northern end of the Gulf of Bothnia. All these districts are, to a + high degree, indicated by the occurrence of southern — apparently heat-requiring — species, Above all the Central Swedish lake district has a markedly rich fauna including species with a + isolated occurrence in this region situated north of the main area. In the cases of Demetrias imperialis and Oodes gracilis the interpretation as relics from a warmer postglacial period is inevitable. Some of these species were shown experimentally to have a high temperature pre- ferendum (diagr. 48). In Finland the favourable thermal conditions in July, mainly in the inner parts and even in comparison with Sweden, find expression in the long series of species (p. 459) which in Finland reach considerably higher latitudes than west of the Gulf of Bothnia. To a lesser degree only this fact is dependent on better opportunities for immigration in Finland. p. 462. Minus-districts, as regards the July temperature, are: 1. the wes- tern and northern coast of Norway; 2. the upland of S-Sweden; 3. the fjeld re- gion. The two first-named districts show so many other climatic pecularities (vide e. g. p. 467) that it is difficult to decide to what extent species are really excluded by the low summer temperature. Concerning the true fjeld-coleoptera especially the preferenda experiments made by KrocErus have shown that they are clearly heat-avoiding and thus may be restricted to alpine and subalpine habitats by the influence of the summer warmth. p. 467. The temperature in spring and autumn. During these critical periods” of insect life the minimum temperature must be of special im- portance. It is an expression for factors controlling the annual period of activity of poikilothermic animals. To a large extent the minima (fig. 64—69) follow the media (fig. 70, 71) but the favoured situation e. g. of the Norwegian coast land is more pronounced. It is therefore exceedingly probable that the numerous species ascending along the Scandinavian westcoast to high latitudes, but wanting in the inland, are dependent mainly on the spring and autumn temperatures. This assump- tion is strongly confirmed by the observation (already made by S. G. Larsson 1939) that Scandinavian Carabids with a pronouncedly western distribution include a strikingly high percentage of species hibernating as larva. It can be shown that some of these species in the easternmost parts of Fennoscandia — in East-Carelia — have in part changed to imaginal hibernation. p. 479. Duration and frequency figures of temperature. In Fennoscandia especially ENguist has worked with "heat-sums” and constructed ”thermo-isochrones” joining places with the same frequencies of days exceeding or falling below a chosen minimum or maximum. He compares these lines with plant-limits for which he pretends thus to find the controlling factor, e. g. in the case of Picea abies (1933, p. 207): ”For the spruce growing spontaneously the 720 825 maximum temperature must amount to at least + ı2.5° C during 65 days alto- gether” (translated from German). Many botanists have enthusiastically accepted the ideas of Engurst, but he has also encountered severe opposition (vide LANGLET 1935). The most important objections against his method seem to be: 1. Even Enguist’s figures are mean values and thus no real factors”; they were calculated from a series of years with considerably varying figures (diagr. 49, table 32). 2. Enquist thinks it possible to find the deciding ”factor” by establishing the point of intersection of the fre- quency curves from two stations situated at the plant-limit in question. Every in- vestigator without preconceived opinions had found a common intersection point of more than two curves necessary. 3. It is absurd that the frequency values alone should determine the distribution of a plant species; the amount of heat during the annual period of activity cannot be insignificant. 4. The thermo- isochrones of ENQuiIsT are drawn with more detail than is allowable from the thinly scattered meteorological stations. 5. The series of years used for different stations are not synchronous. 6. All meteorological figures used by ENQuistT are clearly macroclimatic and thus at most to a quite restricted extent applic- able for biological purposes. p. 485. Precipitation and humidity. The amount of precipitation in Fennoscandia shows a markedly unequal distribution (fig. 73). Its direct biologi- cal influence is mainly negative and unimportant but it works to a greater extent indirectly, causing humidity and ground moisture. It is quite probable that these factors are, at least partly, responsible for the absence of several + clearly xero- philous species in Norway west of the Scandinavian main watershed. The experi- ments especially with the "limestone-species” (p. 128 a. f.) have clearly demon- strated how pronouncedly most Carabids react to moisture and humidity, so there can be no doubt as to their biological importance. In the Fennoscandian climate, however, they play an ecological rather than a geographical röle, i. e. they regulate more the micro- and loco- than the macro-distribution. Above all the capability of otherwise pronouncedly forest species to live in an oceanic climate even on open grounds is remarkable. As the thermic and hygric factors constantly work together many authors have tried to discover some index expressing in one figure for instance the difference between an oceanic and a continental climate. Fig. 74 shows the distribution of ”Humidität”-figures (acc. to MARTONNE) and fig. 75 that of ”oceanity-indices” (acc. to KOTILAINEN) in Fennoscandia. The low figures cf SE Norway in fig. 75 are striking. The persistence of the snow-covering also depends on the combined action of temperature and precipitation. As a thermo-isolator the snow plays an important röle for the hibernating stage but its amount and duration is strongly locally determined. An attempt to illustrate in the form of an isochrone-map the distribution of frosty days without snow gave no regular picture. p. 494. The wind is of great importance to the Carabids as a means of disper- sal (p. 583) but seems to have little influence as an area-limiting factor. In regularly flying forms like the Lepidoptera it is different. p. 495. The numerous heliophilous species among the Carabidae are highly depen- dent for their normal activities on insolation — the number of sunny hours — especially in the summer half-year. Fig. 76 shows pronounced plus-figures i. a. in 721 826 the Baltic coastal districts and the negative influence of the oceanic climate in the west. Agonum sexpunctatum forms the best instance of a species presumably ex- cluded from these parts owing to the lack of sunshine. p. 498. Future investigations on the influence of climate on the organisms will have to work in the field of loco- and microclimate and thus belong to ecology rather than to biogeography. No-one will succeed in mapping exactly a climatic factor deciding the distribution of a plant or an animal. From the common macro- climatic maps it is allowable merely to suggest in what domain (e. g. summer tem- perature, humidity, insolation) this factor has to be searched for. p. 501. It may be possible to study the biological influence of climate also in an indirect way — e. g. by means of phenological! _ observations on the annual aspects of vegetation. One would expect to find in this way a relative expression of the annual period of biological activity in a place or province. I have tried to construct such maps on the basis of plant-phenological material available for Sweden but the figures obtained proved too irregular to make the result sur- veyable without unallowable generalisations. ’ p. 503. Ground. The Carabids, as — with few exceptions — true ”geophiles”, are highly dependent on ground conditions, i. e. on edaphic factors, the hygric, thermic, mechanical and chemical qualities. The hygric and thermic factors were treated above (pp. 177 a. f.; 485). p. 504. The mechanical properties of loose soil are mainly determined by the particle dimensions. Some few experiments (diagr. 51, 52) showed that particularly Carabids with digging habits are in many cases sensitive to this fac- tor, avoiding too coarse sand. In the field it is especially easy te observe how the riparian fauna changes according to the particle dimensions of the shore ma- terial. This may, however; partly be caused by indirect — hygric and thermic — influences. C 1 a yt (fig. 78), and to a restricted degree also different sorts of sand (fig. 77), has so characteristic a distribution in Fennoscandia that a correspond- ingly restricted area of stenotopic species would be expected. In fact, this has proved to be the case, especially for clay-bound Carabids. But clay also affects the organisms indirectly creating the conditions for eutrophic waters (p. 527). p. 516. The chemical properties of the soil were partly dealt with under the heading of "limestone-species” (p. 112 a. f.) which proved to be chemically independent of CaCOg. On the other hand it has long been weil-known that cer- tain insects, among them several Carabids, react positively to NaCl. A distinction was made between halobiontic (salt-demanding) and halophilous (salt-preferring) animals. Only 5 true halobionts occur in Fennoscandia, viz. Anı- sodactylus poeciloides, Dichirotrichus pubescens, Dyschirius chalceus, D. salinus, Pogonus luridipennis; possibly Aépus marinus and Trechus fulvus also belong here. The undeterminated number of ”halophils” includes species confined to sandy shores (independent of NaCl) and 3 species of Bembidion (aeneum, mint- mum, fumigatum) the two first-named of which through experiments proved to react positively to NaCl. B. aeneum has an inland distribution in Scandinavia con- forming very well to the occurrence of Yoldia-clay (fig. 80) supposed still to con- tain small quantities of NaCl. The 3 Bembidion mentioned are probably real halo- bionts but demanding a quite small percentage of salt only. It should be considered whether the term ”halophilous” merits to be maintained at all. Then it ought to mean that a certain species “loves” NaCl — not the seashore — and it is difficult t(= “loam”, in the sense of Lindroth; suppl. scient. edit.). MD 827 to understand how it could ”love” NaCl without demanding it, maybe in very small quantities. f p. 525. It is not so easy to demonstrate an actual demand for other chemicals among Carabidae. It was supposed, however, that some + synanthropous species would be favoured by artificial manure and that this would explain the rapid dis- persal in later years shown by certain among them. As a suitable representative I chose Amara ingenua for experimental treatment and used as ”standard” the culture-indifferent (or rather -avoiding) A. praetermissa. None of the species showed any positive reaction to the commonly used sorts of artificial manure (re- garding A. ingenua cf. p. 539). p. 527. A considerable biological importance is nowadays — especially by the bo- tanists — attached to the soil-acidity constantly expressed as pH-figures. In the case of ”limestone-Carabids”, however, which could be expected to be ”alka- lophilous”, the experiments showed the complete indifference of the insects. On the other hand KrocErus has demonstrated that several hygrophilous in- sects — also Carabids — are clearly ”stenoionous”, i. e. restricted to habitats with either high or low pH-value. But most hygrophilous Carabids are likewise indiffe- rent. The alkalophilous species are bound to eutrophic waters and it is possible that they are in reality dependent on some other factor so that the pH should be estimated merely as an indicator. In acidophilous species (e. g. Agonum cri- ceti a. 0. peat-bog insects) the connection with pH may be firmer. p. 528. 3 Carabid species seem to be chemically dependent on (in any case posi- tively influenced by) burnt wood. p. 529. The term petrophilous was created by Hotpuaus to indicate animals living on firm rock-ground only, or on soil lying in undisturbed position and di- rectly originating from the rock. He thinks this fixed connection to the habitat is mainly due to its special content of nutriment, i. e. haying in the first place a che- mical explanation. In my experience on Fennoscandian Carabids the thermic and hygric qualities of the soil have a far greater influence than the chemi- cal ones and only experiments would convince me that other rules are determining in the Alps. p. 531. Nutriment. Carabids are generally regarded as pronouncedly carni- vorous insects, divergent observations — mainly on species of Amara, Harpalus, av above all on Zabrus — being understood as exceptions. A complete collection of all the data available gives, however. a different picture. Out of 362 Fenno- scandian Carabids: 99 species (incl. 31 under corfinement only) proved able to feed on animal stuffs; 85 species (incl. 40 under confinement only) proved able to feed on vegetables. Of these species (138 in number) 48 were shown to feed on both sorts. It is per- missible to suppose that Carabids are as a rule omnivorous. The zoophagous species, furthermore, generally do not attack living prey but merely wounded or newly dead animals. The réle commonly attributed to Carabids as predators of great importance for controlling injurious insects and on the whole in the natural equilibrium’, seems thus overestimated. On the other hand the larvae, the nor- mal nutriment of which is quite impertectly known, seem to be more pronounced zoophages. p. 538. The food habits of a species have a zoogeographical significance only if it 723 828 can be shown that it lives = monophagous on, or at least clearly prefers, one single animal or plant. I. a. it might be supposed that many + synanthropous species would be dependent on Crucifers and other weeds growing on cultivated soil. As an expe- rimental object I chose Amara ingenua, exposing it to fruits and seeds of several common weeds (table 33). It proved clearly to prefer the fruits of Polygonum avi- culare but was otherwise pronouncedly polyphagous. Other Carabids are = connected with special plants, ec. g. Zabrus tenebrioides with cereals, Bembidion nigricorne and Bradycellus similis with Calluna, Amara aulica with Composites, several Dromius species with different tree species, Calo- some inquisitor with Quercus, etc. But in no case may the plant in question be re- garded as the deciding factor restricting the area of the Carabid species. p. 545. In a few cases there seems to exist a fix connection between a Carabid and its animal prey. The aggregation of most Dyschirius species with different Bledius is wellknown. Experiments on the behaviour of D. obscurus and D. thoracicus towards B. arenarius showed that they exhibit no greater attraction to Bledius than to other insects but react as polyphagous carnivores. A primary factor for the thriving of Dyschirius is no doubt the soil-condition (above all particle dimension and moisture) and their dependence on Bledius is due to the fact that on the suit- able, sterile habitats these insects often constitute the single constant inhabitants fit for prey. p. 548. Brachynus crepitans. It is supposed that the larva lives parasitically on some unknown insect host, as is the case with the American B. janthinipennis. At present the only thing that can be said is that this host is not Agonum dorsale. p. 550. Also in the 3 Fennoscandian species of the genus Lebia — within which parasitic larvae occur abroad — the development is quite imperfectly known. It is probable that at least L. crux-minor and L. chlorocephala are dependent on Chry- somelids but are able to live as larvae on more than one host species. Mollusc-eating species and those + firmly bound to the nests of mammals seem in no case to be dependent on a single animal species. So the food-factor as a whole is of very little importance as restricting the distribution of Carabids. p. 554. Competition. This is the ecological factor most difficult to under- stand concretely. As a proof of the increased competition between closely related animals ELTON has stated that the average number of species per genus in restricted habitats is smaller than in more extended faunal districts. WILLIAMS opposes this conception by demonstrating that an arrangement of genera with I, 2 etc. species forms a logarithmic series. The Fennoscandian Carabid fauna agrees rather well with this (diagr. 53). From this it follows that in small sampies automatically a smaller average number of species per genus must be expected. A treatment of samples from some restricted riparian habitats in Sweden (table 35)) shows that the average figure mentioned is constantly + higher than expected and thus clearly contradicts ELron’s conception that competition is increased between taxo- nomically related species. It is, besides, a common experience of field-entomologists that Carabid genera (e. g. Harpalus, Amara, Dromius) are often abundantly rich in species on quite restricted areas. The competition for food and space ought to play an exceedingly modest role among Carabids. The only case hitherto known, cf supposed competition — or better, of one-sided ”enmity” — between two Cara- bids is that of Carabus nemoralis contra C. hortensts. 724 829 p. 500. Enemies. The number of animals affecting Carabids as predators and parasites is high, but they are little specialized; no enemy seems to be restricted to a single Carabid species. The area-restricting effect of enemies if very unim- portant. p. 563. Stenotopie and eurytopie. The zoogeographical significance of a stenotopic habit is naturally dependant on the = common occurrence of the habitat in question. E. g. a forest-species may be continuously distributed in N- Europe but have a highly disjunct area farther south. A sparse occurrence of suit- able habitats also affects the chance of dispersal. A special interest is connected with species which are + clearly eurytopic in the centre of their area but towards the periphery gradually become stenotopic (e. g. northern species on fens in C. Europe), a very intelligible effect of the "minimum rule”, mainly dependent on microclimate. p. 568. Types of development. Most Carabids hibernate as imago, the exceptions — the # regular larval hibernators — constituting merely about 20 per- cent of the Fennoscandian fauna, It is proposed to designate these types, not as S. G. Larsson (1939) according to the time of propagation with ”F” (spring- insect) and ”H” (autumn-insect) respectively, but according to the hibernation stage as I- and L-type (imago- and larval hibernator respectively) ; quite irregular species were termed as 0-(zero-)type. The zoogeographical significance of the hibernation type mainly concerns the influence of climate and was treated above (p. 475). Dynamic factors. p. 573. Flight power and wind dispersal. The Fennoscandian Cara- bid fauna consists of 3 dynamic elements: constantly flightless 49 species (13.5 percent) ; constantly able to fly 263 species (72.7 percent), among these 177 species observed flying; functionally dimorphic 50 species (13.8 percent; vide p. 335 a. f.). Flight is a normal, daily function only for 9 species, belonging to Cicindela or the subg. Bracteon and Chrysobracteon of gen. Bembidion. The remaining flying forms use their wings only at certain seasons, the imago-hibernators mainly in spring, the larval hibernators — which fly on the whole more seldom — in the sum- mer (diagr. 54). The flight of Carabidae seems to have no sexual function but its main task is to facilitate the change of quarter, especially for the imago-hibernators in spring. Riparian species living by fresh water (of which only Elaphrus angusti- collis is constantly flightless) may be on the wing more frequently owing tu inunda- tion or drying up of their normal habitat also in summer. p. 583. The Carabids — with the exception of the 9 species named above (p. 579) -- are weak flyers, their flight-direction being highly influenced by wind. So they are good objects for an anemochorous transport. For a dispersal over long dis- tances to be realized this transport must take place at high altitudes. In later years the aerial plankton” has been the object of careful studies by means of aeroplanes, especially in U.S.A. (GLick 1939). The figures obtained by these investigations are illustrated by means of 3 diagrams (55—57). Carabids were also present, up to 3000 metres, all flying forms. Nevertheless a dispersal of Carabids through the upper air ought to play a very modest role: they are non-partheno- genetic insects seldom swarming gregariously; they have proved — in connection with anemo-hydrochorous transport — to follow the winds prevailing near the 725 830 earth surface; their actual distribution shows no disjunctions referable to the dis- persal ”by leaps” expected on the assumption of high-air dispersal. The insigni- ficance of all sorts of wind-dispersal of flightless Carabids (incl. earlier stages) is clearly demonstrated by the distribution of dimorphic species (p. 366 a. f.). An in- teresting fact is that the 9 best flyers among Carabids (Cicindela, Bembidion; above, p. 579) show strong geographical conservatism, obviously because they are able to direct their flight actively. p. 592. As previously described (p. 256) the flight-direction of Carabids (and other insects) may also be influenced by sun. Species flying in the evening seem in- clined to direct themselves toward the west and accordingly 4 cases are known of Carabids which have crossed the Bothnian Bay from Finland to Sweden, but none in the opposite direction. p. 595. The question of whether fertilized females do regularly fly is very im- portant, because this would mean an immensely increased chance of colonizing new areas. No females obtained flying in the field proved to be fertilized. But in two cases I succeeded in forcing by means of artificial light a female of Oodes gracilis to fly after copulation, after which she deposited fertile eggs. After all this seems, however, to be a rare exception as is demonstrated 3. a. by the dis- tribution of dimorphic Carabids. p. 598. Water-dispersal. Carabids are comparatively well suited for water-transport as they float high on the surface and thus are able to keep respira- tion-air under their elytra for a long time. Their power of resistance is greater at low temperatures (i. e. in the winter half-year) and especially in fresh and slightly salt water (compared with the ocean water). The Baltic thus makes a fairly surmountable obstacle for hydrochorous insects. The most effective form of water-transport is with drift-ice in the spring, especially with rivers. This would be worth a closer investigation by means of planting out some foreign riparian species on a short bank-stretch on one of the larger Scandinavian rivers. p. 604. The most effective dispersal method is anemo-hydrochorous transport, excellently described by PALMEN (1944). Its main advantage is the con- centration of wind-spread animals within small areas, the shores, and the cor- respondingly increased chance of propagation within the new territory. p. 605. Transport by animals. Only birds have to be seriously regarded in this connection. Carabids, however, are extremely badly adapted to attach themselves to birds, so this method of dispersal seems at the very most to deserve consideration only in the case of Demetrias monostigma, a species with pronoun- cedly disjunct area and possessing effective clinging-organs in its suckerbearing tarsi. p. 606. Transport by man. Merely two Carabids are ”anthropobiont" and therefore must be regarded as introduced into Fennoscandia (Pristonychus, Sphodrus). For some other species, the most interesting of which is Carabus ne- moralis (vide p. 632), the same explanation is at least a high probability. That the importance of introduction by man is often overestimated i. a. becomes evident from the fact that, in spite of the intense traffic, from N. America only 3, from the temperate parts of the southern hemisphere not one, introduced species have been definitely established as members of the non-synanthropous European fauna. p. 609. As a dynamic factor often a proposed desire” of the animal to expand 726 831 its distributional area is maintained, an extremely misleading term. The active movements of the individual are mainly due to feelings of ”discomfort” in its prevailing surroundings and it will certainly stop its migration as soon as it finds a place in all respects suitable. Such a thing as a collective ”want” of the species to colonize the whole habitable area of course do not exist. Nevertheless the result mentioned may be approximately reached by a combination of adequate time, of chance and of sufficient powers of dispersal. p. 610. The barriers against dispersal. In Fennoscandia the sea and the higher mountains form the most evident obstacles. Of course they are most effective for flightless forms. A comparison was made on the area-limiting effect of 3 sea-regions, viz. the Gulf of Finland (> 45 kilo- metres), the sea between Denmark and S-Sweden (> 4. kilometres), the Channel (> 31 kilometres). In spite of the greater distance the first-named of these sea- regions, proved to form the least difficult obstacle to Carabids which is probably mainly due to the considerably lesser amount of salt in its water, facili- tating an anemohydrochorous transport. p. 613. The main Scandinavian watershed from southernmost Norway to N-Finland forms a remarkable barrier for many species. The passes situated below the regio alpina and shown on the map fig. 61 (p. 437) have played a con- siderable positive role, especially during the postglacial climatic optimum. It is interesting to note that also flying forms are to a great extent dependent on the passes for crossing the fjeld-range, a further proof of the small effect of dis- persal through the air in high altitudes. p. 616. Final remarks on area-limits. In practice it is often very difficult to separate existence limits and dynamic limits. The latter may be dis- tinguished by the following characteristics: the species is still in demonstrable expansion (e. g. Amara majuscula, p. 622); the peripheral area is characterized by decreasing abundance but not necessarily by decreasing frequency (existence limits may be expected to demonstrate a contrary picture); after excluding all possibilities of existence-factors forming the limit, this may — provisionally — be regarded as determined by dynamic factors. The clearest existence limits of Fennoscandia are the northern limits of southern species. Fig. 82 gives a general idea of their course in the central parts of the region. The history of the fauna. p. 621. The changes of fauna in recent times. The additions to the Fennoscandian Carabid fauna during the last hundred years are numerous and, naturally, mainly due to intensified investigations. There is, however, quite a series of species which must in fact have invaded the area during the short period of some few decades. Those species may be termed transgrading which occur + regularly but presumably are not permanent inhabitants of Fennoscandia ; a good example is Calosoma sycophanta. The clearest case of a late immigrant, now native, is Amara majuscula (fig. 83, 84), in Sweden also Harpalus puncti- ceps (fig. 85). H. rupicola is doubtful, occurring in Gotland and Oland only. p. 628. Old species with + distinct areal expansion in “historical” time are figured on the maps fig. 86—90. Of special interest also is Carabus nemorals, supposed to be originally introduced by man. Harpalus punctatulus is remarkable Pal 832 in having extended considerably in Finland, but stationary — though possibly with increased abundance — in Sweden. p. 635. Some few species seem to have decreased in recent time, e. g. Agonum Bogemanni and Harpalus nigritarsis. In part they may be culture-avoiding. p. 636. The causes of late faunal changes. It is shown that the — in part very rapid — expansion of species-areas in recent times cannot be an exponent of the normal postglacial immigration to Fennoscandia. There must be special reasons for the remarkable behaviour of these insects. In many cases man is responsible for the changes in fauna, not so much because of direct introduction of new species (vide above, p. 606) as through the revolutionary trans- formation of the landscape — of the habitats — caused by him. Especially the formerly uniform forest districts of the North have been enriched by man with new habitats, and consequently with new insect species. The negative influence, through cultivation of the soil, is more marked in the South. Some habitats, with their fauna, are severely threatened in the whole of Fennoscandia, especially the river banks owing to water-power installations. In comparison with these changes of habitats the reduction of species by the activity of collectors plays quite an unimportant role. p. 641. Changes in fauna may also be due to recent changes of cli- mate. Such have been established beyond all doubt, causing — in the main — an amelioration of the North-European climate during the last decades. The Finnish zoologists (e. g. ©. KALELA) especially have drawn conclusions as to the contemporaneous expansion of area of several animals. It has been stated that the increase of temperature has been greater in winter and spring than during summer and autumn. So far one has worked with average monthly temperatures only, but it is evident that, especially in spring, the minima must be of greater biological importance. The diagrams (diagr. 58—61) show that April is cha- racterized by the grcatest rise of temperature, that this rise is more marked in the minima than in the mean temperature, and finally, that not all parts of Scandinavia have been equally favoured. The last-named fact is more clearly demonstrated by the maps fig. 91 and 92. It seems possible to refer the expansion of area of some species, especially in central and northern Sweden, to the climatic changes mentioned. p. 653. An immigration of insects into Fennoscandia may be caused by environ- ment changes taking place in foreign parts of the species area. Drainage of lakes and fens — caused by man or by climate — has been thought to lead to the immigration into Fennoscandia of some water-birds from the East and it is possible that the same may be applicable to certain hygrophilous Cara- bids. That an increase of population density could give a similar effect seems improbable in the case of Carabidae. p. 654. An expansion of area may be dependent on changed ecology of the species, eventually due to a new mutant. Several Carabids have adapted them- selves to + synanthropous conditions but it is at present impossible to decide whether this premises an altered hereditary constitution. No strict "ecotypes” (biotypes) could hitherto be detected in any Fennoscandian Carabid species. p. 656. The foregoing chapter has no doubt given an impression of the unex- pectedly great changes which have affected the fauna during the short period of a century. One might thus despair as to whether an attempt to reveal the whole 728 833 postglacial period would not be but a series of hypothetical guesses. On the other hand it must be remembered that surely no century during the whole geo- logical evolution has, thanks to man, brought so many biological changes as the last one. p. 656. Fossil finds. The Quaternary insect fossils of Skane and Denmark have been carefully treated by HENRIKSEN (1933) but from the rest of Fenno- scandia there are few finds only. Some new find-localities are described. The fossils reported from Finland by Porrıus (1911) have proved to be unsatis- factorily determined; unfortunately the material (MH) was only partly available. A scheme of the late- and postglacial evolution of Fennoscandia (table 36) and a map of the finding-places for fossils (fig. 93) are given. p. 665. List of fossil Carabids found in Fennoscandia and Denmark. The species follow in alphabetical order. For finds made in other parts of Europe, vide part I of this work and the supplement in this part. Altogether 67 Carabid species (19 percent of the recent Fennoscandian fauna) are found fossil in Fennoscandia or Denmark; in addition 7 species (among these 3 supposed to be totally extinct) now absent from the area. Some conclusions are allowable: In preglacial time there lived in Fennoscandia — if period and insect are correctly determined — in part other species than now. In the last interglacial, on the contrary, the fauna had a quite recent stamp and demonstrates moreover that the climate was in part at least as warm as now (LinpROTH 1948 a). During the last glaciation (Wurm) an alpine fauna, containing many characteristic members of the Fennoscandian regio alpina, but also some few species of a southern (alpine-subalpine) or eastern type now absent from the area, lived at the southern margin of the maximum ice-sheet. In postglacial time the climatic optimum allowed some Carabids to extend, at least slightly, beyond their present northern limit (Calosoma sycophanta, Oodes helopioides, Pterostichus niger). An instructive example of how the fossil finds join the partly scattered recent localities of a species is given by Agonum Thoreyi (fig. 95). p. 676. An extremely important question for judging the value of and the con- clusion made from fossil finds is, to what extent one may calculate upon the constancy of the ecological valence! of a species, i. e. upon the firmness of its ecological demands. In fact it is impossible to deny that a change may occur, and it therefore seems incorrect to make conclusions about ancient climate etc. on the basis of one single fossil species. On the other hand it is quite justifiable to arrive at such results if quite a group of fossil organisms tends in the same direction. p. 677. Relicts. This term by different authors has been used in a rather varying — wider or narrower — sense, the best one formulated by EKMAN (1915, 1922, 1935). Especially in limnic zoogeography, and also in phytogeography, a group called pseudorelicts (secundorelicts) was generally removed from "true relicts” on the plea that its members have not persisted im situ from the time of their isolation but undertaken + extensive migrations from the original relict place. It is here maintained that such a distinction cannot be realized on the terrestrial fauna. The uninterrupted persistence of freely mobile animals in one and the same locality during thousands of years can in no case be proved. Consequently it seems necessary to use the word ”relict” in a wider sense. The t(= “valency”; suppl. scient. edit.). 729 834 following definition is proposed: A population (or stock) of a genetic and taxonomic unit (subspecies, species, genus etc.) is a relict if it is functionally separated from the remaining (as a rule wider, possibly prehistoric) area of the unit and was not able to invade the isolate during present natural con- ditions. In order to present material supposed to include typical relicts a list of the most isolated localities or locality-groups among Fennoscandian Carabid species was put together (table 37). In each case a remark on the supposed cause of iso- lation was added. The different sorts of relicts may be designated according to the factor the change of which in earlier times caused the separation from the remaining area. A. Cold-relicts, e. g. Nebria Gyllenhali in central S-Sweden and Gotland, Pterostichus adstrictus in Smäland; also the Cerambycid beetle Evodinus interro- gationis L. (fig. 96). It is made probable that these relicts — or the main part of them — emanate from the early postglacial southern immigrants and not from a later period of climate-deterioration. — B. Heat-relicts. These are parti- cularly numerous within the lake-district of central Sweden (e. g. Demetrias imperialis, Oodes gracilis), as well as in Oland and Gotland (several Harpalus species), but occur also in S-Norway (e. g. Abax ater), in the inner Sogn district (cf. p. 454), at the northernmost part of the Bothnian Bay etc. The great number of heat-relicts is easily understood as a result of the postglacial climatic op- timum, the far-reaching effect of which in Fennoscandia was clearly demonstrated on fossil plants. — C. Coastal relicts. Owing to the considerable, mainly negative fluctuations of shore-line in Fennoscandia in postglacial time the premises for seashore species being left behind and persisting on fresh-water shores are given. The best examples of such relicts are Cicindela maritima and Dyschirius obscurus in Finland (Kroczrus 1932) and probably Bembidion assi- mile and Dromius linearis by the large Swedish lakes. — D. "Anti-culture- relicts”. Pronounced examples are many forest-beetles, but mainly non-cara- bids. — E. Interglacial relicts. These have surived the last glaciation on isolated refuges at the ice-margin. Their history is treated in the following chapter. p. 698. The postglacial immigration. Reliable conclusions on the faunal history must rest on paleoclimatological and geological facts. Unfortuna- tely the opinions of authors within these domains are often contradictory, e. g. it still remains uncertain whether the Baltic in postglacial time had direct com- munication with the White Sea, or whether Öland-Gotland at any postglacial! period stood in firm land-connection with the Central European continent. In principle traits, the postglacial evolution of Fennoscandia, however, seems clear. It is a question of general importance to what extent the biogeographer must feel himself bound by theories established by geologists etc. At any rate it must be regarded as absurd when TANNER (1937) declares that the problem of floral and faunal hibernation on glacial refuges must be solved by geologist alone. In fact geology has obtained many productive impulses from biogeographers. But the latter have to be fully conscious that they must be able to refer to whole series of biological facts before they are justified in opposing generally ac- cepted geological opinions. 730 835 p. 704. The southern immigrants. The coleopterous fauna at the southern margin of the inland-ice during the last glaciation (i. e. in N-Germany and Jutland) had a northern-alpine character and so had the first postglacial immigrants in Skane. Only two species of Coleoptera otherwise unknown from Fennoscandia are reported as participants in this group, but 6 or 7 which occurred in the north of the Central European continent at the Würm maximum are absent from the Central Europe of to-day. Apparently the northern Wiirm-ice did not extend southward far enough (viz. the ice of the Alps net northward) to make a complete ’Mischfauna” of northern and southern elements possible. The alpine and subalpine species therefore after the Würm-ice began to retreat returned mainly the same way as they had come. The extensive fauna-exchange giving rise to the boreo-alpine type of distribution thus seems to have taken place in the Riss-period (the next-to-last main glaciation). p. 710. There has been much discussion about the question of to what extent the present alpine-subalpine flora and fauna of Fennoscandia has been recruited from the group of the earliest immigrants (i. e. from the south). At present one seems generally inclined to deny any importance to this route, referring i. a. to the almost complete lack of alpine-subalpine fossils in Sweden north of Gotaland. I agree with the conclusion but not always with the arguments. For it is often maintained that, owing to amelioration of climate, the ice-margin north of Skane periodically retreated so rapidly that it was immediately followed by forest ve- getation; — with a yearly retreat of 100 metres or more there must, however, constantly remain a considerable forest-free zone next to the ice, simply owing to the slow growth of trees. A destructive influence, especially on perennial alpine plants, was rather exerted by the permanent change of habitats and — on annual plants and animals — the barrier formed by the Narke-sound, at the southern shore of which the forest may have occupied the whole land before the northern shore became free from ice. p. 714. It is often difficult to decide how far toward the north the tribe which immigrated over southern Sweden has reached in Scandinavia because it may secondarily have been confused with other tribes of the same species which had "hibernated” at the west-coast or, above all, with invaders from the east, over Finland. As a rule we find that in the case of ”double” immigration (over Sweden and Finland respectively) the Finnish tribe almost constantly has reached more northern latitudes and often spread over to Sweden round the Bothnian Bay with a = pronounced gap to the southern Swedish area. In some wing-dimorphic spe- cies it is possible to localize the ”cicatrice” on the Swedish side of the boundary though the two tribes have already fused (p. 381). In two species only (Agonum versutum, Synuchus nivalis) it seems probable that they have reached the northern end of the Bothnian Bay by means of immigration through Sweden. But the fauna of North-Sweden (approximatively N of 64°) obviously has received few contributions by this way alone. Especially in Norway north of Tröndelag (about 65° N) the southern postglacial immigrants are very few. p. 718. The eastern immigrants, Under this heading are included all species (or tribes) which have reached Fennoscandia in postglacial time from centres E of the Baltic. Owing to the rapid disappearance of inland-ice in eastern Fennoscandia this process could begin already before Yoldia Time. It seems obvious that alpine-subalpine organisms — in accordance with the conditions in 731 836 Scandinavia — were not able to reach the present fjeld region of Finland by means of postglacial immigration from the south, partly, however, from the north-east. p. 749. The ”Baltic” path of immigration over the Gulf of Finland was of great importance, mainly because of exclusively favourable conditions for an anemo- hydrochorous transport (PALMEN 1944). Even E Sweden was touched by this immigration group, e. g. probably in the case of Agonum longiventre. p. 721. The ”Carelian” route of immigration, over land E of the Gulf, no doubt has provided Finland with the main part of its fauna. Many species have im- migrated by the Baltic as well as the Carelian route and in these cases a hiatus of area may exist in the middle of the Finnish south-coast. It is important to point out, however, that such a gap may be dependent on climatic factors or on inequal exploration. p. 724. An important immigration route has existed from the White Sea district over Kuusamo-Salla and westward. It has been used especially by rıpicolous plants and insects and it was therefore supposed by some authors that in early postglacial time an uninterrupted sea-connection would have existed between the White Sea and the present Bothnian Bay at the shore of which some of the species in question still exist, often in a rather isolated position. A real breaking through of the watershed in Kuusamo-Salla seems, however, never to have occurred, but the separating isthmus had a breadth of some few kilometres only, thus forming no severe obstacle to dispersal of shore-organisms. p. 728. Some wing-dimorphic species show that the Kuusamo-Salla immigration group was not simply an early vanguard of the ’Carelians” but a functionally se- parated unity. In part its members may emanate from Kanin via the south-coast of Kola but the core of the group can hardly be interpreted except as the descen- dants from a glacial refuge somewhere in the vicinity of the western part of the present White Sea. It may be termed the "White Sea group”. p. 730. The ”Kanin-Kola group”, on the other hand, contains species originating from districts E of the White Sea (i. e. of the maximum Würm-ice limit) and contains mainly tundra forms restricted in present Fennoscandia to the Kola peninsula. Whether a postglacial land-connection Kanin-Kola has really existed is still an open question. p. 732. It is difficult to decide how far to the south members of the eastern (northeastern) immigration group have reached in Scandinavia (cf. above, p. 715). Tachyta nana, however, seems to show the best example of a species which has expanded even to southernmost Sweden (Skane) by this route. A list of species with an unusually distinct ”double” — southern and eastern — origin is given (p. 734). The problem of ”Würm-hibernation”. p. 735. Probably Fennoscandia, like other parts of Europe, has been the object of 4 — or at least 3 — different glaciations; it seems impossible, however, to understand clearly the biological consequences of more than the last one, the Würm. Also during the maximum of this glaciation southwestern Jutland was free from ice, so it was early supposed (BiytT 1893; SERNANDER 1896) that non- glaciated areas might have existed also on the Scandinavian west-coast. The 732 837 theory was developed mainly by botanists (TH. Fries 1913; TENGWALL 1913; and, above all, NORDHAGEN 1933, 1935, 1936); the first zoologist seriously dealing with the problem was WAHLGREN (1919). p. 738. How is it possible to state that an animal or plant species is a "Wuturm-winterer’? It is suitable to start from Simplo- saria metallica, a coleopterous species exposing its history with quite diagrammatic clarity (fig. 106). Its present distribution shows a restricted postglacial tribe in S-Finland but the clearly bicentric Scandinavian area without any doubt emanates from refuges during the Wurm on the west-coast. An important feature is further- more the total absence of a connection eastward into areas in N-Russia outside the boundary of the Würm-glaciation. The only Carabid constituting a similar case of distribution, and likewise an indisputable Würm-hibernator, is Elaphrus lappo- mcus. p. 741. Lists are given of Carabid species the distribution of which agrees in one respect or other with that of the model”, Simplocaria metallica. Attention was paid to the following features: a) isolation of the Fennoscandian area especially toward the East; b) distinct hiatus between a southern postglacial tribe and the supposed hibernated one; c) bicentric Fennoscandian area. p. 743. As previously shown (p. 397 a. f.) a few wing-dimorphic species are likewise indisputable glacial survivors in Fennoscandia. Some of them are representatives of characteristic types of distribution and it therefore seems justi- fied to search for ”hibernators” among non-dimorphic species agreeing in these respects. Thus 6 other Carabids belong to the type of Bembidion aeneum, 9 to that of B. Grapei, 7 to that of Pterostichus strenuus and Bradycellus collaris. p. 745. Supposed glacial survivors may, however, belong to types of distribution not represented by dimorphic species. One is the West-Scandinavian type comprising on the one hand a few species perfectly restricted to western Norway (p. 791), on the other a more numerous group, the members of which reach much further toward the North along the Norwegian coast than in the in- land. The lastnamed type is no doubt to a great extent shaped by climate but as soon as the species area expands into Sweden through the fjeld-passes in a non- oceanic climate (e. g. Bembidion nitidulum, Leistus ferrugineus), this cannot be the case. Carabus coriaceus, treated more in detail, has a South-Swedish tribe, due to postglacial immigration, but the Norwegian and Central Swedish area must emanate from some Norwegian glacial refuge; the latter area, at present very disjunct, probably was more continuous during Atlantic Time. p. 748. Distinctly alpine-subalpine species — 16 in number — certainly all are Würm-hibernators. — Bembidion Schüppeli and Dyschirius septentrionum, demonstrating a striking resemblance in distribution, besides the ”wintered” popu- lation have immigrated in common from the East by two different paths. p. 749. The + panfennoscandian type of distribution, represented among the wing-dimorphic species by Notiophilus aquaticus, have a multiple origin but in all cases ought to include also a hibernated element. Strictly speaking merely 3 species are panfennoscandian, in addition to the Notiophilus mentioned only Calathus melanocephalus and Patrobus assimilis; with somewhat restricted standards for ubiquitous distribution, however, a total of 31 Carabid species may be termed "panfennoscandian”. It is a remarkable fact that no less than 13 of them (i. e. 42 percent) are functionally brachypterous, whereas within the whole 733 838 Fennoscandian fauna this group amounts only to 27 percent (max.). No other explanation seems possible than that species widely distributed (”panfenno- scandian”) are in part descendants of populations living on the Würm refuges where selection worked in favour of flightless species and forms. pou7s2s lesihiut) aploisis ubihkes) from, bioseogwaphical.tacts,, to, „die, termine the exact position of Fennoscandian Wirm re- fuges? An attempt to do so was made, on the basis of the recent distribution of Papaver and other alpine plants, by NORDHAGEN. But already earlier the bicentric type of distribution in Scandinavia was considered by botanists to have emanated from two separate main refuge areas in Norway. For high-alpine organisms, however, the gap between the two ”centres”, situated within the lowest part of the fjeld-range where it is cut through by many passes (vide fig. 61), may be brought about by climate, especially during the postglacial climate optimum when the forest ascended about 300 metres above its present limit. A species like Nebria nivalis may have been severely distavoured at that time. A bicentric dist- ribution in less cold-requiring forms (e. g. Simplocaria metallica, Elaphrus lappo- nicus), on the other hand, must be due to historic factors. It was shown, further- more, that the southern and northern population of bicentric Carabidae at most in exceptional cases may be regarded as postglacial immigrants from the South and the Nordeast respectively. Thus the idea generally entertained by botanists that a pronouncedly bicentric distribution in Fennoscandia is the sign of Wurm- hibernation within two separate main areas holds good also for Carabids. p. 757. The biological theory of localizing in detail the glacial refuges, main- tained in North America by FERNALD, and in Fennoscandia above all by Norp- HAGEN, is based on the assumption that certain plant species (”rigid” species, HULTEN 1937) are extremely conservative and have practically lost their power of dispersal in postglacial time. This idea, in fact, is something of a mystery and it would be highly desirable that some one could give it a more concrete basis by investigating — with experiments, too! — some typical “rigid” species accor- ding to their dispersal of diaspores, their ecological specialization, their competi- tion power, their eventually reduced reproduction owing to population decrease (possibly combined with the "SEwALL WRIGHT effect” within small isolated popu- lations). It seems to me, however, that WyNNE-Epwarps (1939) has clearly overes- timated the stenotopic character of these species as being able to explain their restricted areas. In Carabids it is obvious that the power of dispersal — especially of flight — is the main factor controlling the geographical ”conserva- tism” of species. This was clearly demonstrated by the study of wing-dimorphic species which can therefore contribute towards a localization of Würm-refuges. p. 759. In the High North one refuge was supposed — on entomological grounds — by the White Sea, another in the neighbourhood of Petsamo — S-Va- ranger, probably below present sea level (TANNER 1937). The refuges pro- posed by NorpHaGeN on the Varanger peninsula, on the island of Mageröy, and at the mouth of Porsangerfjord, do not appear clearly from the present entomological material. But the largest of all "ınajor refuges” assumed,.. the, outerwcoastal. districts from, the‘. mouth, ot Alta-fjord south to about the Arctic Circle, including the Lofoten Isles, has played an important röle also in the history of the fauna, no less than 32 Carabids enumerated (p. 761) as proposed ”winterers” here. 734 839 p. 761. Against the facts hitherto presented by botanists Bembidion aeneum and some few other species suggest a refuge also in the immediate vicinity of the Trondheim-fjord, and this is supported also by geological facts. p. 762. In South Norway NORDHAGEN counts upon three separate refuges, in More, at the mouth of Sogne-fjord, and, finally in Ryfylke. Only the first-mentioned is founded also on geological facts. Owing to the possibility of several species having reached these parts of Norway by means af postglacial immigration from the South and Southeast it is usually difficult to distinguish the true hibernators. No doubt, however, they have been more numerous than in any other major refuge”. p. 763. In connection with the Sogn-refuge it is observed that this was es- tablished by NORDHAGEN on the basis of alpine plants the present concent- ration of which in these areas may in part be due to the high-alpine character of Jotunheimen and adjacent fjelds, not only in present but likewise in interglacial time. For pronouncedly high-alpine organisms an interglacial period may be at least as dangerous as a glaciation. In these districts their Würm-hibernation may have taken place in part on nunataks. p. 765. The southernmost refuge assumed by botanists, in the vicinity of Ry- fylke, no doubt is an entomological reality too. I. a. the isolated presence of Bembidion tibiale finds no other acceptable solution. B. harpaloides, on the other hand, is probably a postglacial immigrant from ”Dogger Land”. The hibernators of the South-West are to a great extent constantly brachypterous, or dimorphic species occurring here exclusively or quite prevailingly in their flightless form. An unusually distinct case is formed by Chrysomela crassicornis (fig. 109), constantly brachypterous. p. 769. The idea of a refuge at the mouth of Oslo-fjord is quite hypothetical, founded up to now on no reliable geological facts. On the other hand, the distribution of species like Bembidion litorale and Perileptus areolatus would be easier to understand on the assumption mentioned. p. 772. An attempt to illustrate the possible localization and extension of the Fennoscandian Wurm-refuges is made in map fig. 111. In its construction atten- tion was paid not only to geological and biogeographical facts but also to the generally assumed lower position of the sea-level during the maximum glaciation. p. 774. The climatic conditions during the last glacial period. Table 38 (p. 802) gives a summary of the proposed glacial and post- glacial history of each Carabid species. The number of glacial survivors lies between 52 and 130. The distribution of these species within the Fennoscandian plant regions (acc. to table 30, p. 440) shows that about half the num- ber do not occur in the regio alpina of the present day; even among the ”sure” hibernators — 52 in number — 44 percent (23 species) according to our present knowledge are not able to endure an arc- tic climate. How is it possible to make this fact agree with current con- ceptions of climatic conditions during the Würm-glaciation ? p. 776. It is generally assumed that the temperature — not least in the summer — was lowered during the Wurm to a degree of 3—8° C on an average (the last-named, higher figure in Central Europe acc. to KÖPPEN & WEGENER 1940). Applied to Fennoscandia this would mean that, following the smallest fall pro- posed, the Norwegian coast north of 62° did not reach + 10° C in July; if the 735 840 pessimistic view of KÖPPEN & WEGENER is right, no part of Fennoscandia had attained this figure. p. 777. Under these circumstances the question of the constancy of ecological valence (treated above, p. 676) arises of itself: Is it not possible that the wintered species have altered their ecology? In fact, there seems to exist no reason why such changes should happen contemperanecusly in whole series of species; above all one cannot imagine why an insensibility to cold, hypothetically obtained by selection during a glacial period, should be lost in postglacial time. — It thus seems reasonable to consider whether a glaciation may not arise by factors other than lowering of summer temperature, or whether the loco-climate on favoured places may not have compensated for deterioration of macroclimate. p. 778. Some authors are of the opinion that a glaciation is induced more by means of an increase of snow-fall than by decreasing temperature. Above all the immense regression of glaciers in all part of the globe during the last decades has taught us that such a process can take place without considerable increase of summer temperature, but mainly owing to a leng- thening of the yearly melting period. Mutatis mutandis this may be applicable also to a period with transgression of glaciers. p. 780. The great importance to vegetation and animal life of places situated near inland ice but favoured by loco-climate is shown by examples from many parts of the globe among which the margin area of Vatnajokull on Ice- land is more fully described (LinpRotH 1931). It becomes quite clear that the immediate neighbourhood of inland ice by itself does not necessarily work destructively on organisms. The causes may be: ı. If the landscape is hilly, slopes with favourable exposure to sun may occur. 2. Marginal mountains and the ice itself may form wind shelter. 3. At sufficient niveau differences the winds originating from the high pressure over the ice may receive the character of thermic-favourable descending winds (”Föhn”-winds). Possibly it is due the failure of such winds that an especially severe glacial climate prevailed at the southern ice-margin in Central Europe (cf. p. 776). — It will easily be seen that the West-Norwegian coastal region is very favourably situated regarding the points mentioned above. To these comes the Gulf Stream which, no doubt. touched upon this coast also during the Würm. It is therefore by no means contradictory to ”exact” science to assume that the Fennoscandian Würm-refuges were inhabited also by other than pu- rely arctic (alpne) organisms. p. 784. Quite recently Linnouist (1948) has brought forward the opinion that even the spruce (Picea abies) in a special form, described by him as ”var. arctica”, was a glacial survivor on the Norwegian coast. This conclusion seems to me premature, above all because the incomplete knowledge of the spruce populations in North Finland makes it impossible to decide how isolated the Scandinavian ”var. arctica” is toward the east. Two lichens, brought forward by LINDQUIST as supporting his theory, cannot be of conclusive importance in this connection especially because of the extraordinarily strong power of dispersal characterizing these plants. In summary, a glacial survival of the spruce in Scan- dinavia is, if not quite impossible, still hardly probable; at any rate it is hitherto unproved. 841 736 p. 787. The distribution of some insects — not Carabids — may contribute to the question of whether the glacial refuges were provided with some tree vegeta- tion. Thus Curculio crux (fig. 113) and Rabocerus foveolatus (fig. 114) must be regarded as doubtless ”winterers”, at least in northern Norway, and indicate the survival of shrub-formed Salix and Betula-trees respectively. It would be an ex- tremely interesting task to treat, from ecological and biogeographical points of view, the insects feeding on all sorts of trees in Northern Fennoscandia. p. 791. On the southern Norwegian refuges the glacial climate may be supposed to have been more favourable than in the extreme North — where nevertheless at least the birch survived — and thus the hibernation also of an atlantic (oceanic) faunal and floral element may be seriously considered. Here belong Aépus marinus and Trechus fulvus (both flightless insects), among other Co- leoptera e. g. Otiorrhynchus porcatus and Mesites Tardyı. p. 793. It is of great interest in this connection to study the opinions ofl biogeo- graphers concerning the history of the British fauna and flora. In that country the atlantic element is still more represented and in addition there occurs a ”lusitanic” group of plants and animals with relations to the Iberian Peninsula. It is a strange fact that in Great Britain — in this respect the most suitable country of Europe (Iceland perhaps excluded) — so little interest has been devoted by entomologists to biogeography. The best biogeographical survey of British Coleoptera has been made by a foreigner (SAINTE-CLAIRE DEVILLE 1930 a). (Juite recently, however, some attempts have been made to treat the faunal history of these islands on the basis of Lepidoptera (Forp 1945; BEIRNE 1947 b). The paper of BEIRNE is very thorough but the results obtained by him are founded on extremely hypothetical premises. It is difficult to understand how he could venture to date the arrival of almost every British species of Macrolepi- doptera from its present distribution alone. It seems that he has underestimated the importance of three things: 1. the influence of the environmental factors of the present time, especially the climate, in determining the range limits of species; 2. the changes of area which most species have undregone in postglacial time, mainly due to climatic changes; 3. — above all, the power of dispersal of Lepidoptera. A study of the Lepidopterous fauna of the island of Gotska Sandon in the Baltic (JANSSON 1925, p. 124 a. f.), situated at a distance of 38 kilometres from the nearest land (the island Faron) and totally of young postglacial age (above, p. 279), makes it clearly evident that by no means migratory” species alone are capable of invading new ‘and over sea stretches as wide as the Channel (> 31 kilometres). In fact, Le- pidoptera —- with few exceptions — are not favourable insects for solving the problems of fauna history. Turning back to BEIRNr’s paper it is of special interest to note that, to his mind, some ”lusitanic” forms restricted to Ireland form the oldest element of the British fauna and’ according to him have survived the maximum glaciation within the area. Botanists also are now as a rule inclined to regard the corres- ponding element of the British flora as ”hibernated”. These species — animals and plants — are no doubt less cold-hardy than the trivial atlantic element re- presented in Norway. It may be permissible to conclude, by analogy, that a hiber- nation, as proposed above, was possible also for the major part of atlantic 737 842 anımals and plants in Scandinavia. Probably owing to the distribution of land and sea they had better opportunities during interglacial time than at present to invade Scandinavia from the south. p. 796. Whether man hibernated on the Fennoscandian Würm-refuges is still an open question not to be solved unless indisputably datable archaeological finds are present. The Komsa-culture, suggested by NORDHAGEN (1933, p. 82 a. f.) as glacial, was later estimated as postglacial (BdE & NUMMEDAL 1936). p. 796. It is impossible on the basis of the Carabidae to decide whether some animals have survived more than one glacial period in Fennoscandia. The ”west- arctic” element, represented by Lepidoptera and plants, seems to suggest this eventuality, at least for organisms enduring an arctic climate. p. 798. The contents of table 38 — illustrated also by map fig. 115 — regarding the historical groups of the Fennoscandian Carabid fauna (362 species) may be summarized as follows: Wirm-survivors:t97 + sure species — 27.1 percent of the fauna. (In addition 33 ”possible” species.) Southern immigrants: 280 + sure species = 78.2 percent of the fauna. (In addition 12 ”possible” species.) Eastern immigrants: 267 + sure species = 74.6 percent of the fauna. (In addition 23 ”possible” species.) Only 25.4 percent of the species have a simple origin, belonging to one of the three groups only; 50.3 percent have a double, and no less than 24.3 percent a threefold origin. An example of how one may imagine in detail the history of a + panfenno- scandian species is given by Bembidion saxatile (fig. 116). Rh eı nmel einicres vo fsith ee slaist, elacıation (Wurm) onthe Fennoscandian fauna and flora may simply be expressed thus: This period has played a decisive röle in the de- tauls of the distribution of species. But the stock of Fennoscandian species was only little changed. t(= “Würm hibernators”; cf. p. 798; suppl. scient. edit.). 843 SUPPLEMENT TO PARTS I AND II In the three years since 1945 an extraordinary number of finds of carabids from new localities in Fennoscandia have been added. However, here I give only finds that enlarge the earlier distribution map or significantly add to it*. Only two species are new for the region: Lionychus quadrillum and Perigona nigriceps. There are also numerous important new observations on the ecology, “bi- ology” and dynamics of the species. Also, I had overlooked some observations in the older literature**. On the other hand, the numerous new fossil records—insofar as they concern Fennoscandia—are not taken into consideration in this Supplement. A synoptic account of them is given above (pp. 656 ff.). The following abbreviations of authors’ names may be added: FRD—Axel Ridén; HLD—Nils Hoglund. Acupalpus consputus. Gtl Faron, Ekeviken, in drift material, “May 9, 1948, 2 specimens (HLD !); Sandon, July 6, 1946, 1 specimen on the shore (INS !). In captivity the beetle likes to feed on bread; it has a high predilection for flight, both during sunshine and artificial light (Old Halltorp, June 1946; cf. p. 256). A. dorsalis. His Bergvik, March 29, April 6, 1945, April 1948 (HLD !). Completely isolated: Vbt Vannas, July 7, 1944, 2 specimens (SDH !).—Oa Lappfjard, 1944 (LBH). Ko Jallahti, 1943 (KNG ! KAN !).—Estonia, Osel (SZL, 1942, p. 184). Numerous additional records of spontaneous flight, especially in the after- noon sun (PME, 1946, p. 32). A. dubius. The remarkable record of this species on the small island of Sturkö in Ble dates from April 30, 1947 and May 12, 1948, where it was found *The old account by Ekström (1828) from the Mörkö parish in Sdm was completely new to me. It lists (pp. 50-51) about 80 species of carabids; the nomenclature and serial arrangement follow Gyllenhal (1810). It is not clear from the text whether the determinations were made by Ekström or were at least checked by an entomologist. However, the list gives the impression of being quite reliable. Only the occurrence of Carabus convexus and DIR pubescens (“Harpalus p.” ) may be doubtful. **The interesting contribution by H. Krogerus (1948) is not taken into consideration here. It contains precise data on the development periods and migration of riparian carabids. His ob- servations agree with my own data. 844 139 in great numbers by Nils Hoglund. He gives the following description of the locality: On the northwestern bank of a small, nutrient-rich but also humus- rich pond situated between two stony gravel ridges. Alnus glutinosa on hum- mocks, solitary Betula and Juniperus. Sphagnum in patches, surrounded by firm ground with Myrtillus. The insect was found in moist “forna”t under Alnus at some distance from the waterside. A. exiguus. Gtl Faron, in drift material, June 1946 (Palm). Vrm Visnum- Kil, shore of Lake Vaner, May 17, 1946, 1 specimen (WRN !).—Latvia, 2 localities (LCK, 1942, p. 175). Hibernates as adult, often quite far from the shores (Sv; PME, 1946, p. 32). A. flavicollis. Old Halltorp, Amblystegium swamp, June 11, 1946, 1 speci- men, June 14, 1947, 1 specimen (LTH). Ogl St Anna. Rödskär, May 26, 1947 (WSJ !). His Bergvik, May 21-23, 1947, 8 specimens (HLD ! LBL, RM !). Striking change of habitat, during winter far from the shores, even on heathland (Sv; PME, 1946, p. 32).—1 specimen observed flying upon exposure to sunlight under glass (Old Halltorp, June 1946). A. meridianus. Vrm Kristinehamn, March 27, 1946, numerous specimens (WRN).—NI Dragsvik, September 21, 1947 (NUM); Hango, 1946, numerous specimens on cultivated soil (PME). Numerous immature beetles, July 19 through August 24, 1947 (Ska Häl- singborg, PLQ!).—Both in Latvia (LCK, 1942, p. 175) and in Ska Halsingborg (March 17, 1947, 5:30 p.m., 0°, PLQ!) spontaneous flight—According to HNS and LRS (1941, p. 350), probably in captivity also feeds on vegetable diet. Agonum assimile. Oa Lappfjard, 1944 (LBH). In copulation, May 21, 1946 (Ska Halsingborg, PLQ). Hibernating adults (Sv; PME, 1946, p. 40). A. bogemanni. Ko Kolatselka, June 30, 1944, 1 specimen (Tiensuf!). A. consimile. Jt! Haggenas, July 2, 1944, 1 specimen (KRG!). Lyl, 5 loca- lities in Tarna region, 1945-1948 (FRD!).—Siberia, Yenisei region, Kureika (SBJ 1880, p. 39; MÄ!). A. dolens. Lyl Tarna region, 4 localities, 1945-1948 (FRD!). Tol Silkimuotka, June 23, 1947, July 1, 1948 (HLD!).—According to HOR (in litt.) the report from Slovakia is very doubtful. Spontaneous flight, Ds] Hastfjorden, May 14, 1944 (FRD!). A. dorsale. Ögl Mogata, August 1946, numerous specimens (LTH). Sdm Morko (Ekstrom, 1828, p. 50). Copulation in captivity, May 21, May 26 (Upl). Fed exclusively with bread for several months; also feeds on fresh pieces of Lumbricus and on all kinds of dead insects, but does not attack any living animal.—After many unsuccessful experiments, 1 2, June 25, 1947 (Old Greby), flew repeatedly upon exposure to artificial light. t(= “upper jitter layer of the soil profile”; suppl. scient. edit.). 845 740 A. ericeti. Upl and Gst, 3 localities in the lower Dalalv region, 1945 (ELS!). Dir Grövel Lake, July 6, 1944, 1 specimen (FRD). Jtl Häggenäs, Storflon, August 8, 1945 (Palm!). Lyl Umgransele, 1947, 1 specimen (B. Persson!).— HOR (in litt.) believes the record from northern Spain is doubtful. Copulation in captivity, May 16, and larvae at the end of first instar, June 3, 1947 (Sdm Ricksten, LTH). The beetles fed on bread and a crushed geometrid larva; a living one was not attacked.—In Austria (Kärnten, Kosmatica) sponta- neously flying individuals were seen, June 18, 1939 Hölzel, according to HOR, in litt.). The species thus shows wing dimorphism. A. gracile. Tol Nedre-Soppero, June 4, 1948, 1 specimen (HLD!). One specimen spontaneously flying during sunshine (around 1300 hours), (Upl Angby, Rocksta Lake, May 26, 1946).—Under “fossil records” Skane should be included instead of Denmark. A. gracilipes. Hil Harplinge, 1 specimen (FGQ, RM!). Old Greby-Alvart (for locality, see Fig. 11, p. 117), June 14, 1946, 1 specimen (LTH).—AI Kokar, Ido, seashore, June 1947, 4 specimens (STK, STN!). A. impressum. Hibernating adults observed in November (Sv; PME, 1946, p- 40). A. krynicki. Copulation observed in captivity, July 3; likes to feed on bread (Öld Halltorp, 1946). A. livens. Vrm Visnum-Kil, May 4, 1947 (WRN!).—2 new localities in central Jutland (West, 1947, p. 17). Numerous immature beetles, September 30 (Denmark; West, l.c.), 1 spec- imen also on May 4, 1947 (Vrm, WRN!). A. lugens. Bornholm (West, 1947, p. 17).—Latvia, Papenhof, repeatedly recorded (LCK, 1942, p. 174). Likes to feed on bread, but when offered a fresh piece of Lumbricus, the beetle immediately leaves the bread; also cannibalistic on killed conspecific specimens (Upl). A. mannerheimi. Latvia, Tauerkaln, July 15, 1938, 1 specimen (LCK, 1942, p- 174). A. marginatum. In Mecklenburg under moss, together with Chlaenius tris- tis, hibernating at least 2.5 km distant from the water (NBG, in litt.). A. micans. Tb Pihlajavesi, 1944, 1 specimen (PHJ!). Sv. Kuujarvi and UI- vana, 1943 (KNG! KAN!).—? Osel, 1937 (SZL, 1942, p. 185; “scitulus”). Immature beetle, September 4, 1946 (Jtl, Ragunda, Palm!).—Spontaneous flight (Up! Danderyd, Nora, May 13, 1945, LTH). A. moestum. His Tönnebro, May 26, 1947, 1 specimen (HLD!).—Kg Teru, June 6, 1943, 1 specimen (HLL!).—[Boh Öckerö is to be excluded; the record from this locality pertains to viduum!]. (ef. p- 118; suppl. scient. edit.). 846 741 Several immature beetles, August 23, 1946 (Upl Lovon). Spontaneous flight (Upl Danderyd, Nora, May 13, 1945, LTH). A. mülleri. Small larvae in early June (Denmark; West, 1947, p. 16).—Spon- taneous flight (Dsl Hastfjorden, May 8, 1944, FRD). A. munsteri. Dir Lima. Ostra-Kullstjarn, Sphagnum quaking land, July 21, 1948, 2 specimens (OLS!). Jtl Revsund, July 16, 1944, 1 specimen (BGW!); Haggenas, Storflon, August 8, 1945, several specimens (Palm!). Nbt Kih- langi, Sphagnum quaking land at forest lake, June 1947, numerous specimens (Palm!). According to KRG (in litt.) markedly acidophilic——Immature beetles, August 11, 1945 (NI Hango, Tacktom; THG!). A. obscurum. Up! Fogdö, 1948, several specimens (KLF). Isolated loca- lities: His Skog, Sodra-Brannigen, April 11, 1945, 4 brachypterous specimens (HLD!).—Osel (SZL, 1942, p. 185).—Calabria (SZM, Rivista Sc. Nat. “Natura”, 34, Milano, 1943, p. 94). Spontaneous copulation, July 19, 1945 (Ska Visseltofta, PLQ). Numerous immature beetles, July 16; hibernating adults observed (Sv; PME, 1946, p. 41). A. piceum. Ble Karlskrona, April 7, 1946 (SDH!). Lyl Lycksele, Yttre- Stentrask, September 8, 1946 (FRL!). Nbt Kihlangi, June 1947, 1 specimen (Palm!).—Kc Ontrosenvaara, May 1944. Tiiksa, May 1943 (Laamanen!). In Mecklenburg hibernating under moss, 2.5 km distant from the water (NBG, in litt.). A. quadripunctatum. Kl Valama (Y. Kangas). Near Lyl Umgransele found repeatedly and in great numbers, but exclusive- ly at fresh charcoal kilns (B. Persson, in litt.). In Sv in large numbers on sandy and loamy shores (PME, 1946, p. 39); which could be (?) its primary biotope. In Mk Brandenburg (in 1947) it suddenly appeared in hundreds at a freshly burned location, and also in a house where linseed oil was being carbonized; had not been recorded since 1880 (WGN, in litt.). Immature beetles, August 3 (Ko Santama) and August 9 (Ko Vitele), 1943, (KNG!). Several hibernating beeties in a pine trunk, November 28 va PME, 1946, p. 39). A. ruficorne. Several immature beetles, August 10, 1946 (Ögl Mopatay, -— in captivity the beetle feeds on bread and crushed flies (Ögl Mogata). A. sexpunctatum. Upl Singo, July 16, 1948, 1 specimen on the seashore (KLF). Dir Alvdalen, Mjägen, August 30, 1944 (SVS). Ang Kyrktä Lake, 1946 (R. Jonzon, coll. GTZ!). Lyl Tarna, Ronnback and Gejman, June 1948, 8 specimens (FRD). Tol Nedre-Soppero, June 3, 1948, 1 specimen (HLD!); Karesuando, June 1947 (Paim). Copulation in captivity, June 15, 1947 (Old Hornsjon).—The beetle fed on bread, crushed flies and an uninjured elaterid pupa (Ögl). A. thoreyi. Gtl Sandon, June 30, 1947, 1 specimen (WRN!). His Bergvik, March 24, 1945, May 23, 1947 (HLD! LBL, RM!).—KI Valamo (J. Kan- 742 gas!). Sa Mantyharju (THG, MH!). Sb Kuopio, June 1945, about 10 speci- mens (ELF).—The three northernmost localities (32 Salten; Lk Muonio; 41 Vaggatem) are to be excluded as doubtful. According to KRG (in litt.) very alkalophilic.—Spontaneous flight, Gtl Sandon, June 30, 1947 (WRN!); attracted to light in Hungary (Dorn, 1946). A. versutum. Ke Novinka, May 18, 1943 (Laamanen!). Several spontaneously flying individuals observed (Old, Upl). A. viduum. Impelled to fly upon exposure to sun under glass (Ab Lojo, July 1945, KRH; Up! Angby, May 1946, LTH). Amara aenea. Vrm Ostmark, Rannberg, August 6, 1935 (R. Broberg). Hs Ramsjo, 1943 (LDN). Spontaneous flight: Ska Akarp, May 9, 1947, in sunshine, 3:30 p.m. (CHR); Upl Angby, May 26, 1946, 2 specimens, about 1:00 p.m. (LTH); England (E.M.M., 83, 1947, p. 245). A. alpina. Distribution map also by HNR (1933, p. 297). In Lyl Tarna region up to 1524 msl! (also one pupa), 1945 (FRD).—5 specimens in drift material along Torne-trask, July 1948 (Palm). A. apricaria. Gti Sandon (INS, 2 specimens RM! MJB, 4 specimens, VA!), 1947, 3 specimens (WRN!). In Sv, numerous immature beetles from August through mid-September (PME, 1946, p. 35). Spontaneous copulation, August 20, 1948 (Hll Harp- linge).—Numerous observations of spontaneous flight (Ble, Sma, Upl, Vrm). A. aulica. Lyl Tarna region, several localities, 1945, 1946 (FRD).—Osel (SZL, 1942, p. 184).—Calabria (SZM, Arti Soc. Ital. Sci. Nat., 80, 1941, p. 61). Immature beetles in June and September (Sv; PME, 1946, p. 36). Observed spontaneous feeding on a crushed Harpalus pubescens (Ögl Mogata, August 12, 1946).—Flying to light near Warsaw, 1947 (MAK, in litt.). A. bifrons. Tol Abisko, railroad embankment, September 10, 1947, 1 spec- imen (SJB, coll. LTH).—Om Brahestad, 1944 (LBA).—Ke Ontrosenvaara, September 1943, 2 specimens (Laamanen!). Caught by OSS (!) at three localities (Ska, Sma, Upl) in a flight trap; undoubtedly came flying. A. brunnea. 8 Gloppen, Skjerdal, August 1946 (WSJ!).—Ob Rovaniemi (Y. Kangas!). Lyl Tarna, Morts mountain, 1 specimen from heathland, 1300 msi (July 12, 1945, FRD!); also in northern Norway in the Regio alpina, (STA, 1946, p. 105).—Immature beetle in August (Sv; PME, 1946, p. 35).—Numerous spec- imens in drift material on the shore of Torne Lake, July 1948 (Palm). A. communis. Nbt Kihlangi, June 1947 (Palm).—Osel (SZL, 1942, p. 184). 1 specimen spontaneously flying during sunshine, Up] Angby, May 26, 1946, about 1:00 p.m. t(= “above sea level”; suppl. scient. edit.). 847 743 According to personal communication from J. Makolski, Warsaw, and K. Kult, Prague, communis s. |. (including convexior Steph.) comprises three separate species. I am undecided on this question. A. consularis. Immature beetles, June 15 (Old Vickleby, 1946, WRN!), July 24, August 10 (Sv; PME, 1946, p. 36).—Spontaneous flight in the evening: Sma Skiro, July 8, 1945 (BRC!), Sb Vehmersalmi, July 1946 (HDL). A. convexiuscula. Ska Ven, May 24, 1946 (LLR!). Gtl Rone, Ytterholmen, April 26, 1946, 3 specimens (HLD!). Reports on flight (not taken into consideration above) from Holland (Ent. Berichten, 12, 1949, p. 341). A. crenata. Near Erlangen, in autumn and spring, abundant on open, dry, loamy soil; succeeding species, among others Brachynus (probably explodens Dft.) (Rosenhauer, S. E. Z., 32, 1871). A. crusitans. Ta P.-Pirkkala (Y. Kangas! GBL, coll. HDL!). A. curta. His Bergvik, May 23, 1947, 1 specimen (LBL, RM!). A. equestris. His Los, June 15, 1946; Ljusdal, September 17, 1947 (SJB).—I had overlooked the record from Bornholm (West, 1941, p. 632; 1947, p. 15). Immature beetle, July 2, 1946 (Old Halltorp, WRN!). Pupation on June 3, emergence on June 22 (Denmark; West, 1947, p. 15).—Fed in captivity on bread and crushed flies (Ögl Mogata, August 1946). A. erratica. Lyl Skalmodal, Potato field, June 26, 1947, 1 specimen (FRD!). 1 specimen in drift material at Torne Lake, July 1948, (Palm). A. eurynota. Kc Ontrosenvaara, September 1943, 3 specimens (Laama- nen!). A. femalica. Vgl Fristad, Skalle, May 20, 1944, 1 specimen (SVS!).—Oa Lappfjard, 1944, several specimens (LBH!). Om Lappi, August 13, 1944, 1 spec- imen (LBÄ!). Siberia, Tobol (BGR, MH, as “erratica”!). A. familiaris. 32 Mo in Rana, July 29, 1945 (FRD).—Lyl, Tarna region, 2 localities, several specimens, 1946 (FRD!). Tol Paitasjarvi, June 16, 1947, 1 specimen (HLD!); Abisko, shore of Torne Lake, 1947, 1 specimen (Palm!).— Li Ivalo, July 8, 1939, 1 specimen (THG, coll. PME!). A. fulva. In Sv, immature beetles, August 29, September 5 (PME, 1946, p- 35).—2 specimens spontaneously flying during bright sunshine. June 20, 1945, Vgl Brandstorp (C. Thoren). A. fusca. Ska Hälsingborg, Raus-ınarker, September 8, 1946, October 5, 1947 (PLQ!).—3 new localities (Sjaelland, Falster) in Denmark (West, 1941, p- 632; 1947, p. 15). A. infima. Old Stora-Ror, numerous specimens under Calluna on sand, June 1946, June 1947 (LTH). A. ingenua. The records from Vbt Umea (GTZ) are from 1908 and 1909. Lyl Umgransele, July 18, 1946 (B. Persson!).—Al Eckero, Torp, 1943, 1 spec- imen (LBÄ)—Kc Ontrosenvaara, September 1943, 6 specimens (Laamanen!). 848 744 In captivity fed exclusively with bread for several months (cf. also p. 539); also feeds on dead conspecific specimens. A. interstitialis. In Sv, near Ladoga, strangely a fairly pronounced riparian species (PME, 1946, p. 34). 1 specimen on heathland, 1358 msl (probably got lost in flight), near Lyl Tärna, Daläve (July 19, 1945, FRD!). A. littorea. Old Borgholm, July 9, 1945, 1 specimen (ARV!). Jtl Oviken, 1948, 1 specimen (H. Nyblom through KLF!).—Ok Ruhtinassalmi (SSK, coll. STK!).—Austria, Leitha mountains, 1 specimen, Zurndorf village, 2 specimens (Franz, in litt.). A. lucida. Gtl Sandon (INS, RM, as tibialis!), June 22, 1947, 1 specimen (WRN!). In captivity likes to feed on bread. A. lunicollis. Lyl, Tarna region, 2 localities, 1945, 1946 (FRD!). Tol Karesuando, June 1947, 3 specimens (Palm!).—Lk Pallastunturi (RNK, coll. KNG!).—Calabria, 1 specimen (SZM, Atti Soc. Ital. Sci. Nat., 80, 1941, p- 60). 1 specimen in the lower Regio alpina (790 msl) near Lyl Tarna, Laxfjall, June 25, 1945 (FRD!).—1 specimen observed flying up on exposure to sun in glass (Upl Djursholm, April 4, 1945). A. majuscula. In recent years a large number of localities have been added, illustrating the advancing propagation of the species. Ble Sjoarp, 1945 (HEQ!). Hill Ostra-Karup, 1944 (SJB!); Harplinge, Sardal, 1948 (LTH). Sma Vimmerby, 1944 (LLR!). Old Byerum, 1937 (HNS, coll. LDN!). Gtl Visby, 1945 (LTH). Nke, Orebro, Oset, 1941 (JNS!); Mullhyttemo, 1944 (LLR). Upl Älvkarleby, Längsand, 1944 (WSJ!). His Bergvik, 1945 (HLD!).—Ab Jurmo, 1947 (Lin- navuori!). Ka Tytarsaari, 1938 (HLL, MH!); Seiskari, 1938 (THG, coll. PME!). Kl Parikkala, 1945 (HLL!). Ta Langelmaki, 1943 (Kontuniemi!). Sb Vehmer- salmi, 1943, 1944 (HDL!).—Poland, among others Warsaw (MAK!). PME’s opinion (1946, p. 35) that the species is hygrophilous cannot be correct.—Some immature beetles in June 1947 (Al Kokar, Ido, STK).—Flying to light in the evening (in some cases in large numbers) in July and August, from 7:27 p.m. onward (Warsaw, MAK, in litt.). 1 specimen caught by OSS (Up! Solna, August 1947!) in flight trap. A. montivaga. Dir Siljansnas, Bjorkberget, May 30, 1948, 1 specimen (KLF); Rattvik, June 23, 1945, 1 specimen (Bernell, coll. ARV!), May 30, 1946 (TJT!).—Ta Forssa, 1940 (Kontuniemi!). A. municipalis. Lyl Tarnaby, oats field, May 25, 1946, 1 specimen (FRD!). June 23, 1947 (Old Greby Alvar)t several specimens, all immature.—1 spe- cimen in flight trap, caught by OSS (September 1947, Up! Solna!). t(Plant community consisting typically of mosses and calciphilous herbaceous plants that grow on steppelike shallow alkaline soils overlying Scandinavian limestones; suppl. scient. edit.). 849 745 A. nigricornis. Siberia, Shigansk on the Lena (PPP, “erratica” MH!, 1 2 with light brown-red first antennal segment); Irkutsk (leg. ?, 0°, coll. Kult!). A. nitida. Dir Rättvik, June 24, 1945 (Bernell, coll. ARV!). Kn Ahvenjärvi, 3 specimens (THG, MH!). A. ovata. Ogl St Anna, Stickelskär, July 18, 1946 (WSJ!).—Al Eckero, Torp, 1943, 1 specimen (LBA). Observed flying upon exposure to sunlight, June 14, 1945, 07; spontaneous flight, May 23, 1947, 2 (Up! Djursholm); also May 21, 1947, 9° 2 (Hls, Bergvik, HLD!). A. peregrina. The record from Lt Kola is incorrect (MH = alpina!). A. plebeja. Lyl Tarna region, 1947-48, 2 localities, 3 specimens, (FRD).—Kb Valtimo (Y. Kangas!). Several reports of spontaneous flight (Sma, Vgl, Upl, Dir; Sb). A. praetermissa. Ab Korpo, August 7, 1943 (WLG). NI Parna, August 8, 1940 (Kontuniemi!). Repeated copulation observed in captivity, August 24 through September 5, 1946 (Upl Djursholm).—1 2 observed flying upon exposure to sun in glass (June 24, 1946, Upl Djursholm). A. quenseli. Dir Falun, Österä, May 7, 1944, 1 specimen (KLF!).—Ko Jallahti (Jalguba), June 13, 1943, several specimens (KNG! KAN!). Ke On- trosenvaara, September 1943, 6 specimens (Laamanen!). Solovetsk, Anserek (LEV, MH!). Immature beetle, June 13 (Ko Jallahti, KNG!).—3 specimens in drift material on Torne Lake, July 1948 (Palm). A. similata Lyl Umgransele, May 9, 1948. 1 specimen (B. Persson!).—Ok Ruhtinassalmi (SSK, coll. J. Kangas!). 2 immature beetles, August 13 (Sv; PME, 1946, p. 33).—Spontaneous flight: Upl Danderyd Nora, May 13, 1945; Angby, May 26, 1946 (LTH); Ang Undrom, June 1947 (OSS!). A. tibialis. Vbt Hallnas, June 7, 1946 (HEQ!). Spontaneous flight in the evening, Sb Vehmersalmi, July 1946 (HDL). A. torrida. Lyl Umasjo, 520 msl, August 5, 1945, 1 specimen; Vapstdalen Gransjarvi, Regio betulina (570 msl), June 9, 1946, 1 specimen (FRD!). In Norway never found above the timberline (STA, 1946, p. 107).—1 spec- imen in drift material on Torne Lake, July 1948 (Palm). Anisodactylus binotatus. Ska Sandhammaren, Hagestad-mosse, June 24, 1947 (WSJ!). Dir Svardsjo, Hillersboda, July 23, 1944 (Dubois, according to KLF); Alvdalen, Mjagen, August 30, 1944 (SVS). Hls Bergvik, April 6, 1945, 2 specimens (HLD!).—Sb Vehmersalmi, July 13, 1944, 1 specimen (HDL). Spontaneous flight observed 3 times (Sdm, Upl). A. nemorivagus. Sv Kuujarvi, July 29, 1943, 1 specimen (KNG!); Uslanka, July 25, 1943, 1 specimen (PME, 1946, p. 33).—In Sjaelland 2 instead of 3 localities (West, 1941, p. 632). 746 Asaphidion flavipes. Gtl Sandon, July 1946, (INS). According to PME (1946, p. 23) in eastern Karelia the species shows a distinct change of habitat in that it stays constantly on the shore only in early summer and is far more eurytopic in spring and late summer until autumn. A. pallipes. The Swedish “mo” instead of “mjala” should be used for the ground material. Badister bipustulatus. Harald Lindberg has shown that the form occurring on the Finnish mainland is different from the forma typica occurring in Aland and in Scandinavia, and is apparently identical with B. lacertosus Sturm. The most important differences according to him (in litt.)* are: Larger body size (6.5-7.2 as against 5.5-6.4 mm); more iridescent elytra because of closer and more regular granulation (“Chagrinierung”)'; large, black elytral spot trans- versely divided in front; pale scutellum; first antennal segment more or less darkened at the apex; pale hind tarsi. Besides, the tip of the penis is said to be more slender, without excavation of the dorsal margin at the extreme apex (as in B. bipustulatus s. str.; visible laterally). An examination of most of the Swedish material showed that B. lacertosus also occurs in our region and has a characteristic eastern distribution (Fig. 117), but one of the above-mentioned characteristics was found to be completely consistent, not even the structure of the penis. Adding up the characteristics, it is of course always possible to decide whether an individual should or should not be assigned to B. lacertosus; however, in 1 to 3 of the above-mentioned 6 characters it may be completely identical with B. bipustulatus s. str. B. lacertosus certainly has a completely independent history of immigration. But it is not possible to grant it a higher status than that of a subspecies, especially since the externally similar B. uni- pustulatus has a very different internal structure of the penis (Lindroth, 1943b, p- 18), whereas this seems to be identical in B. lacertosus and B. bipustulatus s. str. Since even the Finnish material of B. /acertosus shows considerable variability, it is uncertain whether the low constancy of characters could have arisen through hybridization. On the other hand it is clear that B. lacertosus is most variable in southern Sweden (up to Ogl). The relatively greater homo- geneity (for instance, in the form of the dark elytral spot) may indicate that the more northern Swedish stock immigrated directly from Finland. | It was not possible to examine the entire Swedish material of “B. bipus- tulatus s. 1.”. However, if the map in Part II holds good for B. bipustulatus Ss. str., at least the following localities must be excluded: all localities on the Finnish mainland; all Swedish localities north of the River Dalalven (Dir, Gst, Hls); Gtl Sandon; Vgl Tived. The northernmost definite record of B. bipustu- latus s. str. in Sweden is Upl Älvkarleö, June 27, 1936 (LTH). In Norway only 851 the forma typica seems to exist (on the basis of 50 specimens in MO!). The *In the meantime this has been published: Notulae Ent, 28, 1949, pp. 96 ff. t (suppl. scient. edit.). 747 subspecies B. lacertosus also occurs in Denmark; I have examined 1 0° from Falster, Korselitse, June 1, 1936 (HSN); according to West (in lite.) also found in Sjaelland and Lolland. The subspecies B. lacertosus seems to be more hygrophilous.—Immature beetle (forma typica), September 16, 1931 (Stockholm, SJB!). B. dilatatus. Sma Oster-Korsberga, Hjartasjon, December 28, 1926 (GTZ!). Gtl Källunge and Hammarsanget in Larbro, June 1946 (Palm); Faron, Eke- viken, drift material, May 9, 1948, 1 specimen (HLD!); Sandon, seashore, July 4, 1946, 2 specimens (JNS!). Copulation in glass, June 16, 1947 (Old Halltorp). Numerous individuals fly in the evening upon exposure to sun, June 1946, June 1947 (Old Halltorp). Badister peltatus. Ab Korpo, June 1946, 1 specimen (WEG!)*. Sb Kuopio, June 1947, 9 specimens (ELF).—Osel (SZL, 1942, p. 183). Several specimens fly in the evening upon exposure to sun, June 1947 (Old Halltorp). B. sodalis. Up! Ekolsund, April 20, 1947, 1 specimen (WSJ, coll. LTH).— Ko Petrozawodsk, at brookside in Aconitum grove, April 23, April 29, 1944, 2 specimens (KRV).—Osel (SZL, 1942, p. 183). Latvia, Alt-Autz, April 19, 1939, 1 specimen (LCK, 1942, p. 175). B. unipustulatus. Ka Hogland, 1 specimen (SRS, MH!); Koivisto, Vasik- kasaari, July 10, 1939, 1 specimen (KNG!). In captivity feeding on a crushed fly (Old Halltorp).—Numerous speci- mens fly in a glass upon exposure to sun (Up! Angby, May 1946; Old Halltorp, June 1947). Bembidion aeneum. In captivity readily feeding on bread (Boh Samstad, August 1946).—Several specimens fly upon exposure to artificial light (Old Mockelmossen, June 1947). B. andreae polonicum. Gti Faron, 1901, 2 specimens (©. Lindbom!). Immature beetle, August 12 (Sv; PME, 1946, p. 26). B. articulatum. Ble Jamjo, Farsksjon, July 28, 1945 (SDH!).—Ka Miehikkala and Virolahti, 1940 (PFF, coll. PME!). According to PME (1946, p. 27) the species shows a clear change of habitat in eastern Karelia and was observed hibernating under bark of trees quite distant from the shore.—Spontaneous flight in sunshine (Upl Adelso, May 25, 1947). B. assimile. Al Finström, 1943, (LBA). NI Ekenäs, Jussarö, 1946 (PME).— Ösel (SZL, 1942, p. 183). B. azurescens. Osel (SZL, 1942, p. 183, “tenellum”). In Sv partly in May, partly September (PME, 1946, p. 26). B. biguttatum. Spontaneous flight observed in England too (E. M. M., 83, 1947, p. 244). *This record was not included while discussing the fauna of the islands. m wo e ce LY r o° r Ly e s rT 10" w ss 20° V F 5 Ay b wel wi ’ nee a GIT BUSS |b A, \ EGE ¥ fy Al \ | N i : AS) N 4 Y R | X / n i = Sc DOA | 5 Neun, 7 F i ASU DT as ae Le i A x R 850 Fig. 117. Badister bipustulatus lacertosus Sturm. Preliminary map. 852 749 B. bipunctatum. Hibernating beeltes observed in pine stumps on a heath (Sv; PME, 1946, p. 24).—Immature beetles, August 14, August 23 (Sv; PME, l.c.). Spontaneous flight: Ds] Frandefors, May 14, 1944 (FRD). B. dauricum. Lul Virihaure, Regio betulina and lower Regio alpina (to 850 msl), July 1944, 8 specimens (BRK!).—Lk Paliastunturi, July 12, 1938, 1 specimen together with Bledius lativentris Janss. (KNG!) and others. On Lake Virihaure (Lul) on dry, heathlike ground with Empetrum, Betula nana, and the like, mostly together with Miscodera and Hypnoides rivularius Gyll. (BRK). B. dentellum. Oa Kauhajoki, August 15, 1939 (KNG!). Kb Jukka, June 30, 1940, 2 specimens (KRG!).—In northeastern Russia at least as far as Archangel (ENW, 2 specimens MH!). B. difficile. Vbt Hallnas, July 30, 1947, 2 specimens (HEQ!). Lyl Tarna region, several localities, 1945, 1946 (FRD!). B. doris. Nbt Kihlangi, June 1947 (Palm). The species shows clear change of habitat in Sv and stays on the shore only from mid-May through early September (PME, 1946, p. 27).—Immature beetle, August 18 (Sv; PME, .c.). B. fellmanni. Tol Karesuando, June 1947 (Palm).—19 Jostedal, Septem- ber 1946 (Jan Lindroth!).—Kanin Peninsula (PPP, 1909, 2 $, MH!). Siberia, Yenisei region, on the islands Nikandrovsk (<") and Briokovsk (2) (SBJ, 1880, p. 20; MA!). Immature beetles, September 8, 1947 (Tol Abisko, SJB). B. femoratum. Ska Ivön, June 1946 (LBL, RM!). Gtl Sandon, June 21, 1947, 1 specimen (WRN!). Upl Forsmark, 1946 (SJB).—Ko Vitele, 4 speci- mens (J. Kangas!). B. fumigatum. Ska Lomma, April 4, 1943, 1 specimen (NYH, O. E., 1945, p. 153), June 3, 1947, 1 specimen (JNS). Hll Harplinge, Sardal, 1 specimen in a small spring fen on sand about 100 m from the sea, August 18, 1948 (LTH). B. gilvipes. His Bergvik, 1945, 1946, numerous specimens (HLD!).—Oa Lappfjard, 1944 (LBH).—Sv, 4 localities (PME, 1946, p. 26). B. grapei. Dir Orsa, Fryksas, June 9, 1937 (TJB!). Hjd Tann valley, July.14, 1944 (FRD!). Lyl Umgransele (B. Persson!); Tarna region, 4 localities, 1945, 1946, numerous specimens (FRD!).—Lt Palaguba (PPP, 1905, p. 90; earlier not identified on the map; situated at the mouth of the Kola fjord). B. grapeioides. Near Tol Bjorkliden, 13 specimens in fissures of flushed soil on moraine ground; succeeding species are Bledius lativentris Janss., Bembidion fellmanni, Hypnoidus algidus J. Sahib. (Palm!). B. guttula. His Bergvik, 1945, numerous specimens (HLD!). Jtl As, June 1947, 1 specimen in flight trap (OSS!).—Oa Lappfjärd, 1944 (LBH).—Osel (SZL, 1942, p. 183).—North Africa (SZM, Atti Soc. Ital. Sci. Nat., 80, 1941, p- 56). Several observations of spontaneous flight (Upl, Jtl). 853 750 B. harpaloides. HOR (in litt.) thinks it has no association with animal nests. He repeatedly found many specimens of the species (in Germany) under bark and in cracks of damp decaying branches (especially of Salix) lying on the shore of a large pond. B. hasti. Lyl Tarna region, several localities (440-840 msl), 1945, 1946 (FRD!). B. hirmocoelum. In Sv, in July and August (PME, 1946, p. 25).—Spontane- ous flight near Sv Vaaseni (KRV, S. H. A., 1945, p. 51). B. humerale. Al Eckero, 1943, 1 specimen (LBA).—Ko Petrosavodsk, lake- side, June 11, 1944 (KRV). B. hyperboraeorum. Lyl Tarna region, 4 localities (460-670 msl, i.e. right up to the timberline, 1945, 1946, several specimens (FRD!). Lul Virihaure, Regio betulina and lower Regio alpina (up to 740 msl), July 1944, several specimens (BRK!). Tol Mäljotjokk, Regio betulina, June 1946, 1 specimen (HLD!).—Siberia, Tolstoinos (SBJ, “virens,” MA‘). Several immature beetles, August 17, 1945 (Lyl Tarna, FRD!). B. illgeri. Osel (SZL, 1942, p. 183). Immature beetles, August 3, 1946 (Ogl Mogata). B. lampros. Hjd Tanndalsjon, Regio betulina (725 msl), July 11, 1944, 1 specimen (FRD!). Lyl Tarna region, several localities and numerous specimens, 1946, 1947 (FRD!).—Kc Rukajarvi Lake, 1942, Tiiksa, 1943 (Laamanen!).—North Africa (SZM, Atti Soc. Ital. Sci. Nat., 80, 1941, p. 54). 3 specimens observed feeding on a fresh, dead Cephus pallipes Klg. (Ska Halsingborg, June 23, 1946, PLQ!). B. lapponicum. Tol Salamasjarvi, 5 specimens, Maljotjokk, 4 specimens, June 1948 (HLD!).—30 Unkervatn, July 2, 1947, 3 specimens (FRD!).—The dot in the map next to Enare Lake has to be deleted. B. litorale. Vrm Ostmark, Rannberg, shore of Rojdalver, July 1, 1935 (R. Broberg).—“Korkeakoski” is situated in the parish of Juupajoki (Ta) and not in Sb.—Ko Jallahti (Jalguba), 1943, 7 specimens (KNG! KAN!). Kn Karhumaki, June 1943 (PRT!).—Jutland, Ngrre Aa (West, 1947, p. 10). B. lunulatum. Spontaneous flight near Ska Lomma (May, 1947, LDN), in Czechoslovakia 3 times (Kult, in /itt.). B. minimum. Gtl Faron, Ava, sea drift, June 1946 (Palm!); Sandon on the sea, July 7, 1946, 1 specimen (JNS!). Sdm Vagnharad, Stensund, June 5, 1944, 2 specimens (Matthiessen!).—Oa Kristinestad, 1944 (LBA). Readily feeding on bread (Boh, August 1946). Numerous specimens fly in captivity during artificial light (Boh Samstad, August 1946). B. monticola. Sv Mattainen, May 14, 1944, 1 specimen (J. Kangas!). B. nigricorne. Ni Lappvik, 1946, 1 specimen (PME). The record from Ik Terijoki is based on a misunderstanding and has to be ignored.—Sv Nurmoila, September 9, 1942, 2 specimens (PME, 1946, p. 24); Aunuksenlinna, July 25, 1943, 1 specimen (KNG!); Ulvana, July 25, 1943, 1 specimen (KAN!). 854 vail Immature beetle, August 13 (Kn Karhumaki, 1943, RNK!). B. nitidulum. Ska Ivö, kaolin pit, September 1945 (NYH). Jtl Are, July 1944 (KRG). Lyl Skalmodal, at brookside on southern slope, 650 msl, Regio betulina, June 26, 1947, 2 specimens (FRD!). Numerous immature beetles, August 18 (Sv; PME, 1946, p. 25). B. obliquum. Lyl Umgransele, July 8, 1948, 1 specimen (B. Persson!).—Im- mature beetles, August 23 (Sv; PME, 1946, p. 25). B. obtusum. Sma Olvingstorp, September 1947, 1 specimen in flight trap (OSS!); Vastervik, September 1947, 1 specimen, collected likewise (OSS!). Old Vickleby, July 11, 1935, 1 specimen (SJB!); Resmo, April 18, 1946, 2 specimens (WRN!). Gtl Faron, Ava, sea drift, June 1946 (Palm!).—In south- eastern Sweden the species has recently expanded its area: The first definite record from Old dates to 1930 (the report by BOH is unsupported by record material); known from Gtl since 1926. The two specimens from Sma Olvingstorp and Vastervik (above) are to be considered as evidence of flight. Spontaneous flight, Boh Lycke, September 16, 1947, in daytime (O. Pehrsson!). B. octomaculatum. Gti Faron, Sudersand, July 1, 1946, 1 specimen on the seashore (JNS!). In Moravia spontaneous flight was repeatedly observed (Kult, in Zitt.). B. prasinum. Upl Alvkarleby, shore of Dal River, July, August, 1945, sev- eral specimens (ELS!). Mdp Indalsliden, Jarkvitsle, August 1, 1945, 4 speci- mens (LBL, RM!). B. properans. Gtl Faron, Ava, sea drift, June 1946 (Paim).—Oa Lappfjärd, 1944, several specimens (Laamanen!).—Kc Ontrosenvaara, 1943, 1 specimen (Laamanen!).—Siberia, Ust-Kut on the Lena (PPP, MH!). B. punctulatum. Kn Karhumaki, June 1943 (PRT!). Immature beetle, August 19 (Sv; PME, 1946, p. 24). B. quadrimaculatum. Ang Kyrktasjo, 1946 (R. Jonzon, coll. GTZ!). Lyl Umgransele, in garden, June 13, 1947 (B. Persson!). Lul Gallivare, July 3, 1944, 2 specimens (HJG!). Tol Nedre-Soppero, July 1, 1948, several specimens (HLD!); Karesuando, June 1947 (Palm).—Ab Korpo, 1945 (WEG). Several records of spontaneous flight (Ska, Upl, Jtl; Sb). B. quinquestriatum. Bornholm (West, 1947, p. 11). B. ruficolle. Ni Kallvik, 1945 (PRT).—Latvia, Halswigshof, 1939, 2 speci- mens (LCK, 1942, p. 173). B. saxatile. Ska Kaseberga, June 26, 1947, 1 specimen (WSJ!). Lyl Tarna, 5 localities (< 554 msl, up to lower Regio betulina), 1945 (FRD). Lul Ku- ouka, July 7, 1948, 2 specimens (HLD!).—Ab Nystad (HLL, coll. PME!).—Kn Karhumaki, 1943 (RNK!). B. scandicum. Near Tol Abisko, 14 specimens were found in an apparently primary biotope (Palm!). June 26, 1947: on the lower course of Nissonjokk (Regio betulina) on a small, barren island of rock and alluvial sand; succeeding 856 132 species: Bledius poppiusi Bernh. and arcticus J. Sahlb., and a few other species of Bembidion (including 1 specimen of B. siebkei). B. schüppeli. According to HOR (in litt.) the species should not be called “boreo-montane.”—Transbaikal (ROU, Folia Ent., 2, Prague, 1938). B. semipunctatum. Ska Lomma, clay pit near the sea, September 16, 1947, 1 specimen (NYH); Sandhammaren, Tyke-a, June 28, 1947, 1 specimen (WSJ!). Gtl Ostergarn. Sandviken, seashore, June 6, 1948, 1 specimen (WSJ!)*. B. siebkei. Near Tol Abisko, September 10, 1947, several specimens, on a sandbank in Nissonjokk where B. scandicum was found by Palm (SJB!); also several immature specimens—Tol Abisko, spontaneously flying, June 26, 1948 (Palm). B. stephensi. Sa Joutseno (THG). B. tinctum. Thanks largely to the revision of the entire Finnish material of B. “dentellum,” the area has been considerably enlarged (Fig. 118).—Mdp Attmar, Lucksta, May 15, 1945, 1 specimen (HLD!). Tol Silkimuotka, June 23, 1947, 2 specimens, Maljotjokk, June 24, 1948, 1 specimen, (Regio betulina) (HLD!).—NI1 Esbo (SBF, MH!). St Nordmark (WKS, MH!). Kb Nurmes (SBJ, MH!). Om Vetil (NSL, Abo Academy!). Ob Oulainen (SDM, MH, according to HLL); Pudasjjarvi (SBJ, ENW, MH!); Kemi (ENW, MH!) Ok Ruhtinassalmi (SSK, 3 specimens, MA!). Li Ivalo, July 22, 1937 (RNK, coll. KNG!). B. transparens. There is no reason to doubt the locality 34 Melbo.—Oa Lappfjard, 1944 (LBH). Sb Jorois (leg ?, coll. STK!). B. unicolor. According to HLL (N. E., 26, 1946, p. 76) the species should again be called B. mannerheimi C.R. Sahlb. B. ustulatum. Osel, (SZL, 1942, p. 183). B. varium. Hil Halmstad, 1 specimen (FGQ, RM!). Ble Jamjo, Farsksjon, July 28, 1945 (SDH!). Gtl Sandon, seashore, July 7, 1946, 1 specimen (JNS!). — The locality Ni Aggelby (KNG) has to be excluded (= B. obliquum!). B. velox. Lyl Tarnafors (440 msl) and Umasjo (520 msl, up to Regio be- tulina), 1945, numerous specimens (FRD!).—Kc Ontrosenvaara (Laamanen!). B. virens. His Granon, July 17, 1947, 1 specimen (HLD!). Lyl, several localities in the Tarna region (up to 715 m, lowest Regio betulina), sometimes numerous, 1945-1948 (FRD!). Nbt Kihlangi, June 1947 (Palm). Tol Nedre- Soppero region, 3 localities, June 1948, several specimens (HLD!).—Ks Salla, Kusta, June 23, 1937, 2 specimens (KNG!). Blethisa multipunctata. Lyl Vapstdalen, Granssjon, Regio betulina (570 msl), June 30, 1946, 1 specimen (FRD). Tol Silkimuotka, Maljotjokk, Regio betulina, June 25, 1948, 1 specimen (HLD!). Spontaneous flight, Up] Angby, May 26, 1946, during bright sunshine (LTH); upon exposure to sun in glass, Ab Lojo, July 1945 (KRH). *This record was not included while discussing the fauna of the islands. / TF % 855 Fig. 118. Bembidion tinctum Zett. 754 Brachynus crepitans. Ögl Mogata, numerous specimens in company with Agonum dorsale, August 1946 (LTH). Sdm Mörkö (Ekström, 1828, p. 50). —NI Hangö, Tulludden, 1 specimen, strayed to the seashore, August 27, 1945 (PME, S. H. A., 11, 1945, p. 185). According to HOR (in litt.) the species often occurs in Germany without Agonum dorsale.—Spontaneous copulation, June 9, twice (Gtl), June 10, June 19 (Old), in captivity again on June 21, June 25. 1 specimen, not fully sclero- tized, August 8, 1946 (Ögl Mogata).—In captivity fed exclusively with bread up to 18 months; also feeds on dead conspecific specimens and all kinds of crushed insects. Bradycellus collaris. Ke Solovetsk, Anserek (LEV, MH!). Near Hjd Tanndalen, 1 specimen in Regio alpina (July 16, 1944, FRD!).— likes to feed on crushed conspecific specimens in captivity. B. harpalinus. Ik Koivisto, July 1935, 2 specimens (Paulomo, N. E., 28, 1948, p. 62; MH!). Moreover in MH there is 1 specimen, Ok Ruhtinassalmi (SSK!), which may be wrongly labeled. In the section on “Ecology,” there is a lapsus calami: “bog regions” (“Moorgebiete”)t instead of moss regions (Moosgebiete”)!. B. similis. Ble Torhamn, August 7, 1946 (SDH!). 1 specimen caught in flight trap by OSS (Vgl Skara, September 1947!). B. verbasci. Numerous specimens on flowers, especially thistles (West, 1947, p. 14). Broscus cephalotes. Notwithstanding the observations above (p. 574) this species is capable of flight at least in some parts of its area or at certain times. There is evidence of flight from Holland (Tijdschr. v. Ent., 70, 1927, p. XII; Ent. Berichten, 12, 1948, p. 312). Calathus ambiguus. Sdm Morko (Ekstrom, 1828, p. 50).—Al Kokar, August 7, 1941, 1 specimen (LBÄ!). Om Nedervetil, September 17, 1943, 1 specimen (LBA!); if there is not a mix-up of labels here this may be a case of acciden- tal anthropochorous transport.—Sv Pisi, August 19, 1942, 1 specimen (PME, 1946, p. 38). Numerous immature beetles, June 11-14, 1947 (Old Greby Alvar)tt. C. erratus. 20 Sundalsoren (coll. KLF!). 36 Malselv, Solvang, June 20, 1930, 1 specimen (JEN!). In captivity likes to feed not only on bread but also on crushed conspecific specimens, other species of Calathus and flies.—In Sv immature beetles as late as July 20, August 14 (PME, 1946, p. 38). C. fuscipes. Sv Kuuttilahti, June 19, 1942, 3 specimens (PME, 1946, p. 38). Fossil record: Ireland, late-glacial (Jessen and Farrington, 1938, p. 241). t(suppl. translator). t(Plant community consisting typically of mosses and calciphilous harbaceous plants that grow on steppelike shallow alkaline soils overlying Scandinavian limestones; suppl. scient. edit.). 857 755 C. melanocephalus. Kl Valamo (Y. Kangas). In Sv immature beetles as late as September 14, September 22, September 28 (PME, 1946, p. 38). C. micropterus. In Sv several immature beetles from mid-August through September (PME, 1946, p. 38). C. mollis. In northern Europe the species is divisible into two well-defined subspecies, which are also geographically separated: Forma typica is the western form, which is generally somewhat larger with more slender body and longer legs. The dorsal side is almost uniformly pitch brown or yellowish-brown (only on Bornholm almost constantly paler brownish-yellow), always without sharp. contrast in color between pronotum and elytra. The right paramere always with a distinct tooth (LTH, E. T., 1943, p. 53, Fig. 25C). To date I have seen only macropterous specimens. Record specimens were studied from Norway (numerous specimens), Jutland (in large quantities), Bornholm (Slusegaard, numerous specimens), and Eng- land (5 specimens). Subspecies erythroderus Gaut. is the eastern form, which in Sweden occurs exclusively. It is stouter, with shorter legs; there is a definite color contrast between the pale pronotum and the more or less darkened elytra and head. The females are more dull. The right paramere of the male has at the most the hint of a dorsal tooth, but mostly it is quite unarmed (LTH, l.c.; Fig. 25a, b). This form shows wing dimorphism (Fig. 25, p. 340). Outside Sweden only record specimens from Bornholm and Sjaelland were seen (Jaegerspris, Kulhus, 1862, 2, Copenhagen Museum!). Within the region both forms occur only on Bornholm. According to the records from Mecklenburg (GRD, 1937, pp. 81-82) the structure of the parameters goes for that region too. i Several immature beetles, June 11-16, 1946 (Old Greby).—Fed with bread in captivity and also with a crushed C. erratus (Old). Calosoma inquisitor. Distribution map also given by HNR, 1933, p. 313.— Sma Visingsö, June 1947, numerous specimens (HEQ). C. sycophanta. Ska Barkakra, Vejby-strand, 1 dead specimen on the sea- shore, July 8, 1946 (WSL). Vgl Horred, July 1941, one flying specimen (L. v. Post, jr., coll. LTH).—Jutland Rye, 1942 (West, 1947, p. 9).—According to Benick (1947), in northwestern Germany too C. sycophanta is a transmi- grating (not native) species. Carabus arvensis. Upl Alvkarleö, May 1946 (Rapp). Gst Gävle, 1946 (Rapp). —Osel (SZL, 1942, p. 182). C. auratus. In captivity the larva is partly cannibalistic (LNG, 1921, p. 76). The beetle also feeds on fungi (l.c., p. 47) and fruit, in captivity it likes to feed on bread (Jung, 1940). C. cancellatus. Ko Petrosavodsk, 1944, 1 specimen (PHJ). Immature beetle, August 22, 1948 (Hll Harplinge).—In captivity the beetle 858 756 likes to feed on bread (Jung, 1940). —Near Ta Juupa River in 1945 KNG found numerous specimens, but had never found the species from 1922 through 1931 despite intensive collecting. He holds that a new immigration has taken place (in litt.). i C. clathratus. Osel (SZL, 1942, p. 182). C. coriaceus. Dir Enviken, Overtanger, May 14, 1941 (K. Danielsson, according to KLF); Lima, Tandberget (600 msi), dry coniferous forest soil with tall Calluna July 3, 1941, 1 specimen (OLS!). Hls Marmaverken, 1944, 1 specimen (HLD!). The beetle likes to feed on bread in captivity (Jung, 1940). C. glabratus. Kn Maaselka (Peltonen!). C. granulatus. Vrm Ostmark, Rannberg, July 27, 1935 (R. Broberg). Dir Enviken, Marnas, 1945 (Ruth Skogblad, according to KLF). Hls Bergvik, April 5, 1945, 2 specimens (HLD!). Copulation, June 10, 1947 (Ska Halsingborg, PLQ).—The beetle iikes to feed on bread in captivity (Jung, 1940). C. hortensis. Distribution map also in HNR, 1933, p. 317. C. menetriesi. Especially “in very boggy, sparse pine forests with a thick layer of moss”; Sv (PME, 1946, p. 18).—Imago, May through mid-June and mid-August through September (Sv; PME, l.c.). C. nemoralis. Vrm Ostmark, Rannberg, since 1935, more frequent in recent years (R. Broberg). His Bergvik, 1945, frequent (HLD!).—Sv Pisi (PME, 1946, p. 19); Mjatusova, Lodeinoje-Pole, 1944, 1 specimen (PHJ). Spontaneous copulation, April 17 (Upl), May 27 (Denmark; West, 1947, p. 9), June 10 (Ska, PLQ).—Seen fighting with a small (uninjured) Lumbricus (Up! Djursholm, May 1947). The beetle likes to feed on bread in captivity (Jung, 1940). C. nitens. Hjd Lillharjean, 1947, 1 specimen (HLD). Jtl Anjan, Steuker, Regio alpina (650 msl), Cailuna-Empetrum heath, June 19, 1945, 1 specimen (H. Kauri). “A species of sunny fine-sandy ground, and accordingly a characteristic animal of the flood-bank of Lake Ladoga”; Sv (PME, 1946, p. 19).—In eastern Karelia also found in late summer and autumn (September, October) (PME, 1946, p. 19); NI Tvarminne, August 11, 1947 (PME!). C. violaceus. Sb Vehmersalmi, July 1942, 1 specimen (HDL). Chlaenius nigricornis. Kontiolahti (Finland): Tb instead of Kb. According to PME (1946, p. 29), the species in Sv may have the same change of habitat as in C. ¢ristis, etc.—Immature beetle, August 3, 1946 (Ogl Mogata). C. quadrisulcatus. Undoubtedly this species at: Sma Balaryd, Skarsjo (Ljungh, 1823, p. 271, “Harpalus sulcicollis”).—Sv Obzha, fragment (PME; 1946, p. 29). 859 750 C. tristis. Ble Torskors, July 1947, numerous specimens (NYH!). Old Hall- torp, Carex-Amblystegium swamp, June 11, 1946, 1 specimen (LTH). Found under moss, at least 2.5 km distant from water, in its winter habitat (NBG, in litt.).—A full-grown larva (Up! Djursholm, August 1942) in captivity fed on freshly killed Lumbricus, Charaeas, Serica, tachinids. C. vestitus. Jutland Hadsten (West, 1947, p. 12). The hypothesis of adult hibernation is being confirmed by a find of 4 immature beetles on September 8, 1946 (Ska Lomma, NYH). Cicindela campestris. Ska St Olof, June 28, 1943 (FRL). Still frequent on June 19-21, 1947 on Old Greby Alvart, spontaneous copulation observed on June 19 and 21. C. silvatica. Sim Morko (Ekstrom, 1828, p. 50). Nbt Pajala; Kihlangi, Muodoslompolo; June 1947, numerous specimens (Palm). Tol Abisko, dead specimens, along railroad tract, August 1944 (LDN!). In Sv already found on April 29 (PME, 1946, p. 17).—1 specimen was ob- served catching and eating an individual of Formica fusca (Vbt Kulbacksliden, June 18, 1939, FRL). The larva even attacks large butterflies (Satyrus) and Odonata (Old; ADZ, E. T., 1912, p. 159). Clivina collaris. In Bohemia always on more or less sandy soil (Kult, in litt.).—Attracted to light, June 1942 (Bohemia, Kult). C. fossor. Tol Silkimuotka, June 23, 1947 (HLD!).—Osel (SZL, 1942, p- 182). Spontaneous flight, Lyl Tarna, July 3, 1945 (FRD). Cychrus caraboides. His Bergvik region, 3 localities (HLD). Lul Virihaure, Regio betulina, July, August 1945, 2 specimens (BRK). Immature beetle, July 9 (Sv; PME, 1946, p. 18). Cymindis angularis. Sim Morko (Ekstrom, 1828, p. 50).—Sa Joutseno, 1936 through 1944, 3 specimens (THG! BLQ).—Sv Gumbaritsa, July 14, 1942, 1 specimen (PME, 1946, p. 42). a Copulation in captivity, June 25, 1947 (Old). —The beetle likes to feed on bread. x C. humeralis. Immature beetle, June 13, 1947 (Old Halltorp).—In captivity likes to feed on bread. C. macularis. Al Eckerö, Torp, 1943 (LBA). In captivity feeding on bread as well as on a crushed Calathus (Old). C. vaporariorum. Upl Alvkarleby, September 4, 1945 (ELS!).—Osel (SZL, 1942, p. 186). Demetrias imperialis. Gtl Faron, Ava, on the seashore (permanently?), June 16-17, 1946, 6 specimens (Palm, E. T., 1947, p. 171!). Sudersand, June 20, 1947, 1 specimen (Aberg!). Sdm Oster Malma, June 1946, 2 specimens t(Plant community consisting typically of mosses and calciphilous herbaceous plants that grow on steppelike shallow alkaline soils overlying Scandinavian limestones; suppl. scient. edit.). 758 (HLD!).—Denmark, south coast of Lolland, 1946, 1 specimen, together with D. monostigma (West, 1947, p. 17). D. monostigma. Öld Högsrum, Gladvattnet, June 20, 1947, numerous spec- imens (LTH). At the locality Vrm Visnum-Kil several specimens in spring of 1946 and 1947 (WRN). Diachila arctica. Tol Karesuando, Maljotjokk, June 24, 1948, 1 specimen, in the stomach of a trout (HLD!). In Norway there seem to be no records from the Regio alpina (STA, 1946, 5 7S) Dichirotrichus pubescens. Doubtful: Sdm Morko (Ekstrom, 1828, p. 50. “Harpalus pubescens”: ? = Trichocellus placidus!). In Schleswig-Holstein the beetle attacked the heart shoots of young Beta plants (HOR, in litt). : D. rufithorax. Ogi Linköping, September 1947, 1 specimen (OSS!). Vrm Varpnas, September 1947, 1 specimen (OSS!); Kristinehamn, among weeds in a garden, 1945, 1946, several specimens (WRN). Both specimens were collected by OSS in his flight trap. Dolichus halensis. Ska Ven, 1942, 1 specimen (P. Prytz, according to LDN). ‘Numerous specimens flying to artificial light near Temesvar in Hungary (Dorn, 1946). Dromius agilis. Tol Abisko, drift material on the seashore, June 1947 (Palm).—Ko Petrosavodsk, June 11, 1944 (Kontuniemi! KRV). Flying to light, June 28, 1944 (Ble Rodeby, SDH!). D. angustus. Ska Halsingborg, Raus-marker, under bark at the foot of an old Pinus, April 14, 1947, February 1, 1948, numerous specimens (PLQ!). Boh Lycke, under bark of Pinus, February 2, February 17, March 1, 1949, 3 specimens (O. Pehrsson!). Gtl Stenkyrka, July 1944, 1 specimen (OTT!); rediscovered in Sandon, June 1947, 2 specimens (WRN). D. fenestratus. Dir Orsa, 1947 (SJB). Jtl Bispfors, 1946 (Palm).—Oa Lappf- jard, 1944 (LBH); Vasa (Kontuniemi!). D. linearis. Sma Vissefjarda, June 13, 1944 (LBA). Gtl Faron, July 11, 1947, 2 specimens (KMK!).—AI Eckerö, Skog, 1943 (LBA).—Sv Kuujärvi, August 10, 1942, 1 specimen (KAN!). In West Germany not xerophilous, but found mostly at moist places (HOR, in litt.).—Immature beetles, June 9 (Gtl), September 21 (Ska). Spon- taneous copulation, June 17 (Gtl), August 22 (HIl). D. marginellus. His Bergvik, 1944-1945, several specimens (HLD!). Jtl Bispfors, 1946 (Palm).—Al Eckerö, Torp, 1943 (LBA).—Sv Gumbaritsa (PME, 1946, p. 41). Spontaneous flight, NI Tvarminne, July 25, 1948 (PME). D. nigriventris. Dir Hagge, May 20, 1944, 1 specimen (OTT!). His Bergvik, April 19, April 22, 1947, 2 specimens; Tonnebro, April 16, 1947, 1 specimen (HLD!).—6 Kvitsoy, July 1930 (MID, coll. JEN!). 860 759 Spontaneous copulation, June 12, 1946 (Öld Greby). D. quadraticollis. Sv Ulvana, June 5, 1944, 1 specimen (KAN!). D. quadrimaculatus. Vrm Visnum, September 1944, (JNS); Ostmark, Rannberg, June 19, 1936 (R. Broberg). Upl Alvkarleby, May 1945, 1 specimen (ELS!).—Ab Korpo, 1945 (WEG).—Osel (SZL, 1942, p. 186). Also near Goteborg repeatedly found under bark of Pinus (SDN, LTH).— Flying to light, July 25, 1947, 10:00 p.m. (Up! Djursholm). Collected in flight trap by OSS (Ogl, Upl). D. quadrinotatus. Hls Bergvik, November 20, 1945, 2 specimens (HLD!).— Several new localities in Jutland, 2 in Fyen (West, 1947, p. 18). In the top of a pine, August 17, 1945 (Upl Stocksund, FRL!).—Immature beetle, August 21, 1945 (Ble Sjoarp, HEQ!). D. sigma. Boh Samstad, August 2, 1946, 2 specimens (HLD!). Dyschirius aeneus. Gtl Faron, Ava, sea drift, June 16, 1946, 1 specimen (Palm!); Sandon, June 23, 1947, 1 specimen (WRN!).—NI Helsinge, July 9, 1941, 1 specimen, together with D. lüdersi (STK, coll. PME!). KI Parikkala, 1945, 1 specimen (HLL!).—Sv Ulvana, July 25, 1943, 1 specimen (KNG!).—Lat- via, 2 localities (LCK, 1942, p. 173). Flying upon exposure to sun in glass (Bohemia, Kult, in Jitt.). D. angustatus. Kn Karhumaki, Kumsa River, July 28, 1942, 1 specimen (PRT!). D. helleni. Lyl Tarnaby, Betula nana moors, June 1, 3 specimens, July 11, 2 specimens, 1946; Strimasund, Kätaviken, moor, 525 msl, July 18, 1946, 1 specimen, Abelvattnet and Virisen, July 13 to August 2, 1947, 4 localities, 8 specimens, all in Betula nana moors; only in Regio betulina (FRD!). Tol Karesuando, Sphagnum fuscum moors, June 1947 (Palm).—Lk Pallastunturi, July 12, 1938, 1 specimen (KNG!). Le Ounastunturi, July 22, 1932, 1 specimen (RNK, coll. HDL!). D. impunctipennis. In 1931 rediscovered by JNS in Gtl Sandon (E: T., 1935, p- 62). D. ludersi. Latvia (LCK, 1942, p. 173). D. nitidus. Sv Segezha, May 25, 1943, 2 specimens (PME, 1946, p. 22).—Lat- via Libau, 1938 (LCK, 1942, p. 171). Spontaneous flight observed in Bohemia, May 1942 (Kult, in Jitt.). D. obscurus. Ko Soutjarvi, June 9, 1943, 10 specimens (KNG!). Flying in large numbers during hot sunshine. May (Germany, Memmert; Alfken, 1924, p. 386). Food, see p. 545. D. politus. Gtl Faron, Ava, sea drift, June 16 to 17, 1946 (Palm). Spontaneous flight near Prague (Kult, in Jitt.). D. “rufipes”. Comparison with D. rufipes identified by G. Muller showed that the specimens from Ik and Kl belong to a different species, which is closest to D. importunus Schaum or is to be considered as its subspecies. A definite conclusion is awaited. 861 760 The Fennoscandian specimens have fully developed wings. D. septentrionum. Nbt Kihlangi, June 1947 (Palm). Tol Nedre-Soppero, July 1, 1948, 1 specimen (HLD!); Karesuando, June 1947 (Palm). D. thoracicus. His Tonnebro, 1947 (HLD!); Storjungfrun, 1945 (HLD!); Hudiksvall, Maln, 1943 (WSJ!). Food, see p. 545. According to SDT (1867, p. 503) the larva feeds on adults and larvae of Bledius as well as Heterocerus.—1 specimen caught flying upon exposure to artificial light (Old Hornsjon, June 15, 1947). Elaphrus angusticollis. Sv, bank of the River Segezha, 2 specimens (PME, 1946, p. 21). The size of the wings varies but attains hardly more than 60% of the surface area of the wing of E. riparius. At least in Fennoscandia the species is to be considered flightless. E. cupreus. Lyl Tarna region, 5 localities, 1945 to 1946 (FRD!), among them 2 specimens (Nuolptrask) in the lower Regio alpina (825-835 msl). Numerous young larvae, July 12 (Denmark; West, 1947, p. 10). E. lapponicus. Lyl Tarna, Laxfjäll, Regio betulina (730 msl)t, June 25, 1945, 1 specimen (FRD!). Lul Virihaure, Staloluokta, Regio betulina (600 msl), July 15, 1945, 1 specimen (BRK). 1 specimen in winter abode in a Sphagnum hummock (Tol Abisko, Septem- ber 1947, SJB). E. riparius. Spontaneous flight, May 29, 1946, 6:30 p.m. (Ska Halsingborg, PLQ). E. uliginosus. Ögl Mogata, on the seashore, August 1946, 1 specimen (LTH). Harpalus aeneus. Spontaneous copulation, August 5, 1946 (Ogi Soderko- ping). Likes to feed on bread and crushed conspecific specimens (Old Greby). H. anxius. Hil Halmstad, etc., May 1944, several specimens (FGQ, RM!). Copulation in captivity, June 21, 1946 (Old Stora-Rör). Likes to feed on bread as well as crushed conspecific specimens (Old). In captivity numerous flying males and females at artificial light (Old). H. azureus. Spontaneous copulation, June 10, 1946 (Old Greby); later, in captivity, again on June 22 and 23. Immature beetles: Gtl Fardume, August 10, 1923; Visby, August 26, 1928.—In captivity likes to feed on bread (Old).—Six specimens flying to artificial light near Temesvar in Hungary (Dorn, 1946). H. calceatus. Ska Loderup, August 1948, numerous specimens (LDN); Vitemolla, July 22, 1947, 1 specimen (TJB); Ivo, August 20, 1946 (SDH). Hil Harplinge, Sardal, August 19-21, 1948, 2 living specimens, 1 dead specimen, on sandy soil, close to the sea (LTH). Boh Lycke, August 20, 1946, 1 specimen (Olof Pehrsson!). Gtl Faron (MJB, 1 specimen VA‘); Sandon, June 1947, 1 specimen (WRN). Upl Varmdon, Vaster-Skagga, 1945, 1 dead specimen (LTH).—NI Kottby, 1945, 1 specimen (HLQ, coll. KNG). Ta T (cf. p. 846; suppl. scient. edit.). 862 761 Ruovesi, Siikakangas, August 15, 1943, 1 specimen (PME!). Numerous specimens flying to light near Ska Loderup, August 1948 (LDN, 1948). 1 specimen near NI Kottby (HLQ). H. distinguendus. Im Muolaa, 1 specimen (PME!).—Osel (SZL, 1942, p. 184).—According to SZM not known in North Africa (but known in Madeira and Azores) (Atti Soc. Ital. Sci. Nat., 80, 1941, p. 58). Immature beetle, July 29, 1945 (Upl Varmdon).—In captivity feeding on bread as well as pieces of Lumbricus and a crushed elaterid larva (Upl Djur- sholm). H. frolichi. Ska Södra-Sandby, Skatteberga, August 20, 1944 (CHR); Fagel- sang, April 4, 1924 (Kullander, coll. LDN!). Old Vickleby, Stora-Fro, August 1943, 1 specimen (KMN; NYH, ©. E., 1945, p. 154; ML!); Resmo Alvar,! September 16, 1946, 1 specimen (ML, according to NYH). Flew to artificial light, as well near Warsaw (MAK, in litt.) as in Bohemia (Acta Soc. Ent. Boh., 1910, p. 16, according to Kult, in litt.). H. fuliginosus. Dir Malingsbo, 1941, 1 specimen (FRL!); Norrbarke, Bjorsjo, May 29, 1947, 1 specimen (OTT!).—Latvia Libau, 1936, 1 specimen (LCK, 1924, p. 175). H. griseus. Ska Vitemölla, July 20, 1947, 1 specimen (TJB!); Löderup, August 1948, frequent (LDN). Ble Brakne-Hoby, Sjoarp, August 6, 1939, 1 specimen (KNG!). August 30, 1945, 1 specimen (HEQ!). Hll Ostera-Karup, August 7, 1944, 1 specimen (SJB!); Harplinge, Sardal, August 18, 1948, 1 spec- imen (LTH). Vgl Skovde, August 23, 1944, 1 specimen (FRD!). Old Kastlosa, Vaderstad, August 16, 1946, 1 specimen (HZE!). Gtl Faron, Norsta-aura, June 21, 1947, 1 specimen (Aberg!); Sandön, July 7, July 8, 1946, 2 speci- mens (JNS!).—Ik Muolaa, June 28, 1938 (MER, coll. HDL!). Jutland, Mols (West, 1947, p. 13). Copulation in captivity, August 2, 1948 (Ska Loderup, LDN).—Mass flight to light, August 1948 (Ska Loderup, LDN, 1948). | H. hirtipes. Ska Hassleholm July 17, 1947, 1 specimen (PLQ). Old Ottenby, July 23, 1941 (LDN!). ; In captivity feeding on bread as well as on a crushed fly (Old Stora-Rör). Spontaneous flight to artificial light, observed in Bohemia (Acta Soc. Ent. Boh., 1910, p. 16; according to Kult, in Zitt.). H. latus. Immature beetle, April 12 (Sv; PME, 1946, p. 31).—Spontaneous flight in the evening. Sb Vehmersalmi, July 1946 (HDL). H. luteicornis. Hil Halmstad, Eketanga, July 26, 1943, 1 specimen (LDN!). Gtl Farosund, June 30, 1946, 1 specimen (JNS!); Faron, 1901, 1 specimen (Lindbom!), Broa, June 15, 1946, 3 specimens, Ava (sea drift), June 16, 1946, 1 specimen (Palm!).—Ab Korpo, June 13, 1944, 1 specimen (WEG!). Ta Hat- tula, May 27, 1940, 1 specimen (WEG!). tcf. p. 866; suppl. scient. edit.) 762 In Sv on sandy soil with sparse vegetation, partly found in open terrain and partly in sparse mixed forest (PME, 1946, p. 31). H. melancholichus. Jutland Mols (West, 1947, p. 13). 1 specimen flying to light (Ska Loderup, August 2, 1948, LDN, 1948)* H. melleti. Gti Faron, Broa, June 15, 1946, 2 specimens (Palm). Near Gtl Visby partly in the company with A. rupicola and A. punctatulus at the locality described for the former and partly in a downtown ruin on much more shady, somewhat moister, humus-rich soil.—In captivity fed with bread (GTL).—Numerous specimens observed spontaneously flying to artificial light (Bohemia; Kult, in Jitt.). H. neglectus. Hil Sondrum and Ovraby, 1 specimen each (FGQ, RM!). Gtl Eksta, Djupvik, May 31, 1948, 6 macropterous specimens (WSJ!)*. Several new Danish localities (West, 1947, p. 13). In captivity feeding on bread as well as on crushed flies (Old). H. picipennis. Ska Vitemölla, July 29, 1947, 1 specimen (TJB!). In Bohemia 1 specimen observed during spontaneous flight, May 17, 1936 (Kult, in litt.). H. pubescens. Immature beetles, May 14 and August 12-20 (Sv; PME, 1946, p. 30). —Mass flight to light in Hungary (Dorn, 1946). H. punctatulus. Gtl Sandon, June 1947, 2 specimens (WRN).—Near Ka Viborg, probably in 1928. Ab Korpo, 1945 (WEG). Sa Villmanstrand, August 21, 1945, 2 specimens (THG!); Joutseno, August 23, 1945 (BLQ).—Osel, 1934, repeatedly found (LCK, 1942, p. 175; SZL, 1942, p. 183). Latvia, 2 localities (LCK, l.c.). Numerous specimens (more than 100), April 28, 1945, June 18, 1946, at the locality near Gtl Visby described under A. rupicola (strangely, not observed here earlier).—Numerous specimens in umbels of Daucus (Denmark; West, 1947, p. 12). In captivity fed as well with bread (some exclusively) as with crushed carabids and flies (Gtl).—After a two-day stay on wet substratum one male flew in the evening during artificial light, July 1, 1946 (Gtl). H. puncticeps. Ska Farslov, July 9, 1947, 1 specimen (Matthiessen!). Old Vickleby, 1944, 1 specimen (LBA!), 1946, 1 specimen (WRN!); Halltorp, 1946, 12 specimens, 1947, 2 specimens (LTH); Greby, 1946, 1 specimen (LTH). Ik Muolaa, May 28, 1938, 1 specimen (PME!).—Several localities in Jutland and in Fyen (West, 1947, p. 13, and in litt.). Calabria (SZM, Atti Soc. Ital. Sci. Nat., 80, 1941, p. 57). In captivity fed with bread (Old). H. puncticollis. Sdm Vagnharad, Stensund, April 10, 1944, 1 specimen (Matthiessen!). Osel (SZL, 1942, p. 183). H. quadripunctatus. Hjd Tanndalen, July 18, 1944 (FRD). *This record could not be taken into consideration above. 863 763 In northern Norway right into highest part of the Regio berulina (STA, 1946, p. 96). H. rubripes. Vrm Vase, 1945, 1 specimen (WRN). In captivity feeding as well on bread as on crushed conspecific specimens (Old). One male flying in captivity (Old). Spontaneous flight of a male during hot sunshine, June 21, 1947, 10:30 a.m. (Old Greby). H. rufitarsis. Old Vickleby, June 15, 1946, 1 specimen (WRN!).—Calabria (SZM, Atti Soc. Ital. Sci. Nat., 80, 1941, p. 58). In captivity fed with bread (Old). One female flying during artificial light (Old). H. rupicola. In captivity fed with bread (Gtl).—One male flying during artificial light (Gtl). H. seladon. Dir Soderbarke, Glafse, June 20, 1947, 1 specimen, Larsbo, July 5, 1947, 2 specimens (OTT!).—Ka Hogland, July 2, 1939, 1 specimen (THG!).—Sv Pisi, June 14, 1942, 1 specimen (PME, 1946, p. 30).—Osel (SZL, 1942, p. 183, A. “brevicollis”). In captivity fed exclusively with bread (Old, Upl). Flying to light near Warsaw, 1947 (MAK). H. serripes. Spontaneous copulation, June 10, 1946 (Old Greby), in cap- tivity on May 4, 1945 (Gtl). Immature beetle, July 27, 1946 (Old Borgholm, SDH!). One specimen lived in captivity from April 1945 through January 1948.—The beetle can be fed exclusively with bread; fed on a crushed Tene- orio too. Spontaneous flight near Prague, May 16, 1946 (Kult, in /itt.). H. servus. Hil Harplinge, Sardal, in loose sand under Calluna, August 19, 1948, 6 specimens (LTH).—Old Stora- Ror, in loose sand under Calluna, June 12, 1947, 3 specimens (LBL, LTH). H. smaragdinus. Ble Torhamn, August 7, 1946, 4 specimens (SDH!). Gtl Faron, July 10, 1947 (KMK!). Vrm Vase, 1945, 1 specimen (WRN); Torsby, Sorbo, August 10, 1944, 1 specimen (SVS!). Copulation in captivity, July 28, 1946 (Old). Likes to feed on bread and crushed Harpalus (Old).—Spontaneous flight of one individual to artificial light (Ska Loderup, August 2, 1948, LDN, 1948). H. tardus. Ke Tiiksa, June 1943, 1 specimen (Laamanen!). In captivity feeds on bread and crushed conspecific specimens (Gtl). H. vernalis. Gtl Visby, on the eastern city wall, April 30, 1945, 1 specimen (LTH). In captivity fed with bread (Old). H. winkleri. Ta Lammi, June 18, 1946, 2 specimens (J. Kangas!). Ik Muolaa, 1938, 1 specimen (MER, coll. HDL!).—Kn Karhumaki (CRP, coll. STK!). Latvia (LCK, 1942, p. 175). Lebia chlorocephala. Vrm Ostmark, Rannberg, 1935, 1936 (R. Broberg). Dir Enviken, Overtanger, 1945 (Karin Eriksson, according to KLF). Hls Bergvik, 1945. Langvind, 1947 (HLD!); Farila (LBL). 764 Copulation in captivity, August 19 (Ögl). Immature beetle, August 6, 1946 (Ogl Mogata).—Numerous specimens on Hypericum in company with Chrysomela varians larvae in June (Denmark; West, 1947, p. 17). In captivity the beetle also feeds on bread.—Spontaneous flight in the evening in July 1946, Sb Vehmersalmi (HDL). L. crux-minor. Vrm Ostmark, Rännberg, 1945, (R. Broberg). Repeatedly flying in glass upon exposure to sun, May 27, May 31, 1946 (Upl). L. cyanocephala. sl Mogata, August 11, 1946, 1 specimen (LTH); St Anna, April 7, 1947, 1 specimen (WSJ!).—Osel, (SZL, 1942, p. 186). In captivity, 1 specimen flying during high room temperature, June 26, 1946 (Old Stora-Rör). Leistus ferrugineus. Lyl Tarna, Rövattsliden, Regio betulina (685 msl). July 1, 1945, 1 specimen Regio coniferina, May 25, 1946, 1 specimen, Vapstdalen, Grans Lake, Regio betulina (570 msl), June 7, 1946, 2 specimens, Abelvattens- dal, several localities, all in Regio betulina, 1947 to 1948 (FRD!). L. rufescens. Lyl Umgransele, July 8, 1945, 1 specimen (B. Persson!); Skalmodal (540 msl), June 26, 1947, 1 specimen (FRD).—Ko Petrosavodsk, 1943 (HIl!).—Osel (SZL, 1942, p. 182). Licinus depressus. Old Greby, June 11 to 24, 1947, numerous larvae (most of them half-grown), only 2 adults (both old and damaged); one larva observed feeding on a Vallonia costata Mull. (det. N. Hj. Odhner) (LTH). Lionychus quadrillum Dft. Distribution Sweden. The totally unexpected discovery of this species in northern Europe was made by HZE on May 27, 1945 near Nke Orebro, Oset. Subsequently (even in 1948) it was found again in large numbers (E. T., 1947, p. 120). Absent from the rest of Fennoscandia and neighboring regions, except England, where the species occurs in the south at the south coast (Joy, 1932, p- 373). Total area: Euro-Mediterranean species. In Europe northward as far as northeastern France (DEV, 1935-38, p. 58), Belgium and Holland (EVS, 1898, p- 101; 1922, p. 39), central Germany to Berlin region, but not east of the Oder (HOR, 1941, p. 343), eastward as far as Transylvania (PTI, 1912, p. 40) and Greece (OTZ, 1896, p. 213). Southward as far as northeastern Spain (FUE, 1921, p. 22) and southern Italy, also Sicily (LUI, 1929, p. 141).—Asia Minor (SZM, Atti. Soc. Ital. Sci. Nat., 80, 1941, p. 64), Cyprus (BUR, 1939, p. 197). Ecology The habits of this species in Sweden have been described in detail by HZE 865 765 (E. T., 1947, p. 120). Here it lives in the dry, barren gravel of a railroad embankment constructed in 1943 in boggy terrain, especially on the southern side, strongly exposed to the sun. In the rest of Europe this species occurs at gravelly or sandy, more or less moist places mostly on riverbanks, less often if there is no connection to open water (EVS, 1898, p. 101; Arnold, E.B., 1929, p. 56). In Germany “in mountainous and hilly terrain” (HOR, 1941, p. 343). Evidently the insect is very thermophilous: it tolerates strong, continuous exposure to the sun (August 1947) with no perceptible distress. Biology Since copulation was observed in Italy as carly as April (E. B., 1929, p. 56), it may be assumed that the species hibernates as imago. Near Orebro imagines were found even in late October (E. T., 1947, p. 121), immature beetles in late August (!). Dynamics The wings (in the Swedish specimens) are fully developed. Hence the insect may be capable of flight, although as far as I am aware there are no observa- tions. Repeated exposure to sunlight in a glass to induce flight (August 1947) was unsuccessful.—The species probably reached Sweden by traffic. Metabletus truncatellus. Lyl Umgransele, 1946 (B. Persson!). A second instar larva was found on July 7 (Upl Djursholm): molting on July 10, pupation on July 20, emergence on July 30 (LTH). It was fed on small pieces of Lumbricus. Three specimens observed flying spontaneously in the evening, Sb Vehmersalmi, June, July 1946 (HDL). Microlestes maurus. Gtl Faron, Broa, June 15, 1946 (Palm).—Osel (LCK, 1942, p. 176; SZL, 1942, p. 186). Latvia (LCK, l.c.). M. minutulus. Sma Överum, July 1945 (JNS!).—Al Jomala, July 2, 1942 (STN, coll. WEG!); Finström, 1943 (LBÄ). Hattula is located in Ta (not “Ka”). Miscodera arctica. One specimen flying spontaneously during sunshine, May 22, 1946, 4:20 p.m. (Jtl Bispfors, Palm). Nebria brevicollis. Gtl Sandön, June 28, 1947, 1 specimen (WRN!). Sdm Mörkö (Ekström, 1828, p. 50). Nebria gylienhali. In Lyl Tarna region up to 1300 ms] (FRD).—2 specimens in drift material in the Torne Lake, July 1948 (Palm).—An interglacial fossil record on Iceland (Thorkelsson, 1935, p. 5). N. livida. Kl Parikkala, 1945 (HLL). In Sv immature beetles, June 18 and August 22, larvae during July (PME, 1946, p. 19). N. nivalis. In drift material on the shore of Torne Lake, June 1947 (Palm). 866 766 N. salina. Boh Ljung, Direhuvud, June 24, 1945, 1 specimen (KLF). I had overlooked the record (3 specimens, Dublin County, OMH, Dublin Museum!) from Ireland (Donisthorpe, E. M. M., 65, 1929, p. 186). Notiophilus aquaticus. In Lyl Tarna region as far as 1538 msl (FRD). Fossil records: Scotland, interglacial (Movius, 1942, p. 268). Dogger Bank, postglacial “moorlog” (Bell, 1922). (Last line: “in this” should be deleted). N. biguttatus. The numerous new records called for a new map (Fig. 51, p. 406). In northern Finland three localities have been deleted: Lp Pitkajarvi (STA; 4 specimens = N. reitteri!); Lk Muonio (PPP; no record specimen); Kittila, Aakenustunturi (SAA, 1917, p. 281; only one larva, which may very well belong to N. reitteri).—Osel (SZL, 1942, p. 182).—In North Africa according to Koch (Mitt. Minch. Ent. Ges., 1939, p. 234). N. germinyi. Ska Ven, May 1934 (Palm, coll. wanes —Osel (SZL, 1942, p- 182). Regio alpina (950 msl), Lyl Umfors, Norra-Storfjället, July 16, 1945 (FRD!). N. palustris. The doubtful record from Jtl Are (AND) pertains to N. germinyi!—Kc Ontrosenvaara (Laamanen!). N. pusillus. Gtl Faron, Broa, June 15, 1946 (Palm). Bornholm, 2 localities (West, 1947, p. 9). Osel (SZL, 1942, p. 182). Several immature beetles, June 11-15, 1947 (Old Greby Alvart). N. reitteri. Lyl Tarna, Yttervik, mixed forest (440 msl), August 23, 1945, 1 specimen; Gejman, June 20-21, 1948, 3 specimens, all in Regio coniferina (FRD).—Tol Tjarro (north of Nedre-Soppero), June 17, 1948, 1 specimen (HLD!).—24 Dovre, Kongsvoll, 1920, 1 specimen (LYS, MD!). 30 Unkersvatn, July 2, 1947, 1 specimen (FRD). H. rufipes. Ska Borringekloster, November 1, 1923, 1 specimen (KMN; NYH, O. E., 1945, p. 153; ML). Odacantha melanura. Gti Faron, Ava, sea drift, June 16-17, 1946 (Palm). Up! Älvkarleby, Gardskar, seashore, 1942, several specimens (ELS).—Sv Gum- baritsa, Ladoga shore, June 1942, 1 specimen (PME, 1946, p. 42!). Osel (SZL, 1942, p. 186). In Denmark oviposition in mid-May, larvae half-grown at the end of June (West, 1941, p. 632). In Holland there is a report of flight (Tijdschr. v. Ent., 70, 1927, p. XII) that was not taken into consideration above. Olisthopus rotundatus. “8 Vadheim” (KLF) should be excluded, since it actually pertains to O. synuchus!—Al Sottunga, 1943, 1 specimen (LBA). Kl Harlu, 1 specimen (PME!). In captivity likes to feed on bread (Ogl).—The flight capacity of the t(Plant community consisting typically of mosses and calciphilous herbaceous plants that grow on steppelike shallow alkaline soils overlying Scandinavian limestones; suppl. scient. edit.). 867 767 macropterous form was established by the discovery of one macropterous spec- imen “on Sphagnum acutifolium quaking land” near Ab Sammatti, 925 (KRG!). Omophron limbatum. Ble Jamjo, July 28, 1945, several specimens (SDH!); Torhamn, Färsksjon, September 26, 1947, several specimens (BRK). Oodes gracilis. Sdm Oster-Malma, June 1946, 1 specimen (HLD!). The frequency of the species at Upl Osbysjon is evident from the fact that at the small sampling plot I (LTH, 1943a) 50 specimens were collected in 1 1/2 hours on June 3, 1947. j In spring 1947 near Upl Osbysjon, appearing as early as May 7 (3 speci- mens) after an unusually hot day (Matthiessen).—The beetle likes to feed on bread. O. helopicides. Gst Mardangsjon, May 29, 1947, 8 specimens (LBL, RM!). —Osel (SZL, 1942, p. 182). In contrast with O. gracilis, O. helopioides is not a definite migrant. More than 10 specimens were found on April 22, 1945, right at the edge of “Norra- karr,” Upl Danderyd, among moss and in turf, apparently at the place of hibernation (LTH). In Sv hibernating under Salix foliage close to the shore (PME, 1946, p. 30). Panagaeus bipustulatus. Ögl Mogata, together with Brachynus; larva August 14, adult August 17, 1946 (LTH).—KI Salmi, June 23, 1944, 1 specimen (PME!). u The beetle likes to feed on bread (Old). P. crux-major. Ble Jämjö, Färsksjön, July 28, 1945 (SDH!). Up! Älvkarleby, Ostana, June 22, Hyttön, July 1, 1945 (ELS!). Hls Bergvik, December 30, 1944, 3 specimens (HLD!).—Osel (SZL, 1942, p. 183). Patrobus assimilis. Oa Lappfjärd, 1944 (LBA).—Latvia, Tauerkaln, 1938, 1 specimen (LCK, 1942, p. 173).—Pechora region, Pjoscha River and Kuloj River (KLM, MH!). Northern Ural, Kondinsk (BGR, according to SBJ, 1880, p- 21; MH!). In Sv immature beetle, August 13 (PME, 1946, p. 29). P. atrorufus. Vbt Hallnas, July 25, 1947, 2 specimens (HEO!). Lyl Skal- modal, June 24-28, 1947, several specimens (as far as 540 msl, Regio betulina) (FRD!); Strimasund (525 msl), July 18, 1946, 1 specimen (FRD!).—Om Haapavesi (Arppe, coll. Y. Kangas!). P. septentrionis. 8 Gloppen, Gjengalundsbraeen (1000 msl), August 9, 1946, 2 specimens, (WSJ!). Near Lyl Tärna, Rufjället, as far as 1400 msl, August 22, 1945 (FRD). —Interglacial fossil record on Iceland (Thorkelsson, 1935, p. 5). P. septentrionis australis. Sb Vehmersalmi (HDL!).—In Sjaelland 2 new localities (West, 1947, p. 11). Pelophila borealis. Vbt Vannas, July 7, 1944 (SDH!) In Lyl Tarna region (northern Storfjallet, Stropialke) as far as 1040 msl, August 6, 1945 (FRD). 868 768 Perigona nigriceps Dej.* Sweden. 1 specimen of this cosmopolitan species was collected by S. Berden on July 17, 1948, in a park near Ska Alnarp in the evening by sweeping (O. E., 13, 1948, p. 167!). Not known elsewhere in Fennoscandia or neighboring regions. Total Area: The species originates from the environs of the Indian Ocean (JEA, 1941-42, p. 581), but was transported with traffic to West Africa, North and Central America, as well as displaced to Europe. Northward only as far as Moravia and Rhineland (HOR, 1941, p. 190); later there are several records northward as far as Lubeck (Kol. zeitschr., 1, Frankfurt, 1949, p. 83). Perhaps the Swedish record is the forerunner of an impending invasion. Ecology The insect lives in compost and all kinds of modern herbs, partly subterranean. Biology I know nothing about the development periods and the feeding habits. Dynamics The wings are well developed, and the species has been repeatedly observed in flight, especially in the evening (JEA, l.c. ). Pristonychus terricola. Ab Korpo, June 12, 1944, 1 specimen (WEG). NI Helsinki, 3 specimens in a cellar, April 1949 (LBG).—Osel (SZL, 1942, p- 185). Pterostichus adstrictus. In Lyl Tarna region as far as 680 msl (Regio be- tulina), 1945 (FRD). P. angustatus. From Sv no association of the species with burned wood is reported by PME (1946, pp. 36-37) but he regards it as “a very characteristic species of boggy spruce and pine forests,” which has also been found on drier soil.—Immature beetles, August 20-22, 1942, found in large quantities (Sv; PME, l.c.). P. anthracinus. Gtl Sandon, June 28, 1947 (WRN!). In connection with crossing experiments (LTH, 1946) some observations were made on the development periods (under optimal conditions, at room temperature): copulation, April 16 through May 22; oviposition, May 12 through June 1; larvae hatching from May 19 through June 9; first molt, *Since Perigona nigriceps occurs only accidentally in our region it has been ignored in calcu- lating the different percentages of the carabid fauna of Fennoscandia as a whole. 769 May 28 through June 19; second molt, June 5 through July 7; pupation June 26 through Juiy 12; emergence of adults, July 4 through 17. The shortest total period of development from oviposition to emergence of the adult was 9+9+8+ 12+ 7 days = 45 days. Found near Old Halltorp under moss of fallen, decaying, wet oak trunks in very shady location on June 11, 1946: one pair in copulation, numerous females with eggs in their cavities, and two batches of freshly hatched larvae.—In captivity the beetle feeds on bread but prefers animal food; the larvae were fed exclusively with fresh pieces of Lumbricus.—A reared specimen flew in the evening during artificial light. 2 specimens flying to light in Hungary (Dorn, 1946). P. aterrimus. Old Högsrum, Gladvattnet, June 20, 1947, 1 specimen (LBL!). P. coerulescens. Copulation on June 10, 1947 (Ska Halsingborg, PLQ).—In captivity the larva was partly cannibalistic (LNG, 1921, p. 76). P. cupreus. Oa Lappfjard, 1944 (LBH). Kl Valamo (Y. Kangas). Spontaneous flight during sunshine, May 14, 1948, 2.00 p.m. (Stockholm, LTH). P. diligens. Interglacial fossil record on Iceland (Thorkelsson, 1935, p. 5). P. gracilis. Gtl Sandon, June 30, 1947, 1 specimen (WRN!).—AI Eckero, Torp, 1943, 1 specimen (LBA). Ta Tavastehus, May 1940 (NUM). P. lepidus. Sim Morko (Ekstrom, 1828, p. 50). Upl Adelso, May 25, 1947, 1 specimen (LTH); Roslags-Nasby, 1943 (OLS!). Lyl Umgransele, numerous (B. Persson!).—30 Grong (LYS, MD!).—Record specimen on Lm Kantalaks in MH (!). Copulation on August 22 (Sv; PME, 1946, p. 36), in captivity on August 11 (NI Tvarminne, LTH).—In captivity likes to feed on bread (NI). P. minor. On the British Isles, too, the species shows wing dimorphism, with about the same ratio of the two forms (Sharp, 1913). P. niger. Dir Bingsjo, 1945 (F. Mansson, according to KLF). In Central Europe apparently dimorphic (Maran, 1927). On the British Isles the species may be constantly macropterous, as in our region, but accor- ding to Sharp (1913) not all specimens are capable of flight. Spontaneous flight of 1 specimen to light near Temesvar in Hungary (Dorn, 1946). P. nigrita. North Africa (SZM, Atti Soc. Ital. Sci. Nat., 80, 1941, p. 61). On Gautavardo near Lyl Tarna, 1 specimen in the lower Regio alpina (825 msl), June 23, 1945 (FRD!).—Spontaneous flight (during sunshine); Ska Silvakra, April 20, 1947 (CHR); Upl Danderyd, Nora, May 13, 1945 (LTH); Angby, May 27, 1946, 3 specimens (LTH). P. oblongopunctatus. Distribution map in HNR (1933, p. 308).—Ska St Oloi, May 23, 1945, 1 specimen (HEQ!). In Lyl Tarna region as far as the lower Regio betulina (550 msl) (FRD!). P. punctulatus. The locality Ska Ystad (Palm) should be deleted. P. strenuus. Spontaneous flight of 1 specimen during sunshine near Sb Vehmersalmi, June 1946 (HDL!). 869 770 P. vernalis. 16 Bandak (MST, MO!).—Ka Virolahti (PFF, MH!). Oa Lap- pfjard, 1944 (LBH); Vasa (Kontuniemi!). Ob Hailuoto (WUO, 2 specimens, MH!).—Kn Kontupohja (Vaartaja, MH!). The considerable variability of wing size was noted by Sharp for the British Isles (1913, p. 86). P. vulgaris. Hls Ramsjo (LDN).—Kr Kontschosero (SBJ, MH!); Suma (LEV, MH!). Late glacial fossil record from Ireland Coser and Farrington, 1938, p. 241). Sphodrus leucopthalmus. Sb Vehmersalmi, 2 specimens under the floor- boards of a flour warehouse, 1946 (HDL). Stenolophus mixtus. Old, southern end of Möckelmossen, June 15, 1947, 1 specimen (LTH). Gtl Ostergarn, Sandviken, on the seashore, June 6, 1948, 1 specimen (WSJ!)*.—Ta Kuusankoski, Voikka, July 1947, 1 specimen (NUM). In Sv numerous beetles as early as June and July (PME, 1946, p. 31). Stomis pumicatus. Dir Leksand, Leksandsnoret, June 1948 (G. Enlund, according to KLF). Synuchus nivalis. 8 Vadheim, August 11, 1937 (KLF!). —Osel (SZL, 1942, p. 185). Tachys bisulcatus. Ska Lomma, 1 specimen on the wall of a house, June 13, 1947 (S. Berdén!). Sma Overum, July 7, 1945, June 10, 1946, numerous specimens (JNS, E. T., 1947, p. 4). Vrm Lundsberg, 1945, 2 specimens, flying in the evening (WRN). Tachyta nana. His Bergvik, May 3, 1947 (HLD!). Observed repeatedly feeding on Collembola (Sv; PME, 1946, p. 28). Tachypachys zetterstedti. Sv Olonets (PME, 1946, p. 20). In Sv found on bog soil or in boggy mixed and coniferous forests, alto- gether 6 specimens; thus probably a hygrophilous species (PMF, l.c.). Trechus discus. Vrm Lungsund, Kungsskogen, July 23, 1947, 1 specimen (WRN); Varpnas, August 1947, flying specimens (OSS!). His Los, August 7, 1947, flying specimens (SJB).—Sa Villmanstrand, 1942 (HDL).—Also in northern Jutland (West, 1947, p. 11). HOR (in litt.) does not assume any association of this species with animal nests. Immature beetle, June 5 (southern Germany, HOR, in litt.). T. fulvus. 6 Sande, May 24, 1928; Sandebukten, May 2, 1933; Hafsjord, May 17, 1942; each time only 1 specimen, always right at the seawater (JEN). T. micros. Svir region without specific locality data, 1944 (PME, 1946, p- 28). T. obtusus. Lyl Tarnaby, Abelvattensdal and Vapsdal, several localities also in Regio betulina (as far as 750 msl), 1946, 1947 (FRD).—8 Gloppen, Skjerdal, August 1946, 2 specimens (WSJ!). In Hjd (Hamrafjäll), also collected by FRD in the lower Regio alpina next *This record was not considered in the treatment of the fauna of the islands. ATA to a snowdrift, July 1944.—Immature beetle, August 18-20 (8 Gloppen, WSJ!). T. quadristriatus. Vrm Torsby, Sorbo, August 10, 1944 (SVS).—Om Ned- ervetil, 1943 (LBA). T. rivularis. Upl Fiby, September 1947 (LBL). T. rubens. Lyl Tarna region, 2 localities, 1945, 1946 (FRD). T. secalis. The locality His Los (SJB) should be excluded.—Osel (SZL, 1942, p. 193). In Sv the beetle appears around July 1, but immature specimens are ob- served as late as in August (PME, 1946, p. 28). Trichocellus cognatus. Vgl Gardsjo, June 30, 1942, 2 specimens (BRC!). Lyl Tarna region, 2 localities (as far as Regio betulina), 1946 (FRD).—Oa Lappfjard, 1944 (LBH).—Denmark, southern Fyen (West, 1947, p. 14). T. placidus. Gtl Faron, Ekeviken, sea drift, May 9, 1948, 2 specimens (HLD!). His Ramsjo (LDN). Oa Lappfjard, 1944 (LBH). The locality Lk Muo- nio is doubtful. Bibliography’ 870 AARNIo, B. 1922: Ueber Salzboden (Alaunböden) des humiden Klimas in Finn- land. — Internat. Mitt. f. Bodenkunde, 12. Berlin. Apamovic, L. 1909: Die Vegetationsverhältnisse der Balkanlander &c. — A. Enc- LER & O. DrupeE: Die Vegetation der Erde. XI. Leipzig. ADLERZ, G. 1916: Gravsteklarnas liv. — Sthlm. AGRELL, I. 1941: Zur Okologie der Collembolen. — Opusc. Ent., Suppl. Lund. — — 1947: The effect on the physiological state of insects (Carabides) on their thermal preference. — Opusc. Ent. 12. Lund. AHLMANN, H. W:son 1919: Geomorphological studies in Norway. II. The Nord- land. — Geogr. Annaler. 1. Sthlm. — — 1939: Glaciarerna som uttryck for den pagaende klimatandringen. — Ymer. 59. Sthlm. — — 1948: Den nutida klimatfluktuationen och dess utforskande. — Norsk Geogr. Tidsskr. 11 (1947). Oslo. AHLNER, S. 1948: Utbredningstyper bland nordiska barrtradslavar. — Acta Phyto- geogr. Suec. 22. Uppsala. ALFKEN, D. 1891: Erster Beitrag zur Insekten-Fauna der Nordsee-Insel Juist. — Abh, Naturw. Ver. 12. Bremen. — — 1924: Die Insekten des Memmert. Zum Problem der Besiedelung einer neuentstehenden Insel. — Ibidem. 25. ALMmauiıst, E. 1929: Upplands vegetation och flora. — Acta Phytogeogr. Suec. 1. Uppsala. — — 1937: In „Stockholmstraktens Växter”. 2, Aufl. — Sthlm. ANDER, Kj. 1942: Ett par sällsynta gaddsteklar (Hym. Aculeata). — Opusc. Ent. 7. Lund. — — 1947 a: Flygformagan hos vara hopprätvingar. — Fauna och Flora. Uppsala. — — 1947 b: Forandringar i Skanes fjarilfauna, — Skanes Natur. 34. Lund. — — 1947 c: Notiser om gaddsteklar. 1. — Opusc. Ent. 12. Lund. ANDERSSON, G. 1889: Studier öfver torfmossar i sodra Skane. — Bih. K. Vet. Ak. Handl. 15. III, 3. Sthlm. — — 1898: Studier öfver Finlands torfmossar och fossila kvartarflora. — Fennia. 15. Hfors. — — 1902 a: En stenälders-boplats pa Hven. — Ymer. 22. Sthlm. — — 1902 b: Hasseln i Sverige fordom och nu. — Sver. Geol. Unders. Ser: Ca. 3. Sthlm. — — 1906: Die Entwicklungsgeschichte der skandinavischen Flora. — Rés. Scient. Congr. Intern. Botan. Wien (1905). — Jena. tReproduced from the German original, pp. 870-901 (Not edited by scientific editor). —General Editor. 773 871 — — 1910: The climate of Sweden in the late-quaternary period. — Sver. Geol. Unders. Ser. C. 218. Sthlm. ANDERSSON, G. & BiIRGER, S. 1912: Den norrlandska florans geografiska fordelning och invandringshistoria &c. — Norrländskt Handbibl. 5. Uppsala. ÄNGSTRÖM, A. 1938: Lufttemperatur och temperaturanomalier i Sverige 1901—1930. — Medd. Statens Meteor.-Hydrogr. Anst. 7. Sthlm. — — 1939: Temperaturklimatets andring i nuvarande tid och dess orsak. — Ymer. 59. Sthlm. (Auch in: Medd. Statens Meteor.-Hydrogr. Anst., Uppsala. 26.). — — 1941: Nederbördsklimatets andring i nuvarande tid. — Medd. Statens Me- teor.-Hydrogr. Anst., Uppsala. 37. Sthlm. — — 1946: Sveriges klimat. — Sthlm. ANTEvS, E. 1928: The last glaciation. With special reference to the ice retreat in northeastern North America. — Amer. Geogr. Soc., Research Ser. 17. New York. Anvisningar till byggnadsstadgan. 1946. — Kgl. Byggnadsstyrelsens Anvisn. 1. Sthlm. ARNELL, H. W. 1923: Vegetationens ärliga utvecklingsgang i Svealand. — Medd. Statens Meteor.-Hydrogr. Anst. 2. Sthlm. ARNELL, K. 1927: Vegetationens utvecklingsgang i Norrland. — Ibidem. 4. ARNELL, K. & S. 1930: Vegetationens utveckling i Gotaland. — Ibidem. 6. ARRHENIUS, O. 1926: Kalkfrage, Bodenreaktion und Pflanzenwachstum. — Leipzig. Arwıpsson, TH. 1928: Bizentrische Arten in Skandinavien — eine terminologische Erörterung. — Bot. Notiser. Lund. — — 1938: Studier over floran och vegetationen pa Gotska Sandon med särskild hänsyn till nationalparken. — K. Vet. Ak. Avh. Naturskyddsär. 1. Sthlm. — — 1943: Studien über die Gefässpflanzen in den Hochgebirgen der Pite Lapp- mark. — Acta Phytogeogr. Suec. 17. Uppsala. Atlas öfver Finland. — Sällsk. Finl. Geogr. — Helsingfors 1910. ATTERBERG, A. 1903: Sandslagens klassifikation och terminologi. — Geol. Foren. Förh. 25. Sthlm. — — 1905: De klastiska jordbeständsdelarnas terminologi. — Ibidem. 27. AUDCENT, H. 1942: A preliminary list of the hosts of some British Tachinidae (Dipt.) — Trans. Soc. Brit. Ent. 8. Southampton. BAckLuns, H. O. 1944: The ability of the shore beetle, Cafius xantholoma Grav. to evade seadrift. — K. Fysiogr. Sallsk. Förh. 14. Lund. — — 1945: Wrack fauna of Sweden and Finland. Ecology and Chorology. — Opusc. Ent., Suppl. Lund. Baitey, J. H. & Britten, H. 1943, 1946: The Coleoptera of the Isle of Man. — The North Western Naturalist. 18. 21. Arbroath. Barner, K. 1937: Die Cicindeliden und Carabiden der Umgebung von Minden und Bielefeld. I. — Abh. Landesmus. Westfalen, Mus. f. Naturk. 8. Munster i, W. Beirne, B. P. 1047 a: The effect of human activities on the distribution and abundance of the Lepidoptera. — Ent. Record. London. — — 1947 b: The origin and history of the British Macro-Lepidoptera. — Trans. R. Ent. Soc. 98. London. BENANDER, P. 1946: Microlepidoptera Sueciae. — Opusc. Ent. 11. Lund. 774 872 BEnıcK, L. 1926: Die Käfer der Oldesloer Salzstellen. — Mitt. Geogr. Ges. u. Naturh. Mus. 31. Lübeck. — — 1947: Über den Massenwechsel und die Verbreitung norddeutscher Kafer. — Forsch. Geogr. Ges. u. Naturh. Mus. Lübeck. BERGROTH, O. 1894: Anteckningar om vegetationen i gränstrakterna mellan Aland och Abo-omradet. — Acta Soc. Fauna Fl. Fenn. 11. Hfors. BERGSTEDT, N. H. 1883: Bernholms Flora. — Bot. Tidsskr. 13. Kbhvn. BERGSTEN, F. 1941: Undersokningar rorande sekulara andringar i avrinningen i vissa svenska vattendrag &c. — Medd. Statens Meteor.-Hydrogr. Anst., Uppsala. 36. Sthlm. BERLAND, L. 1935: Premiers résultats de mes recherches en avion sur la faune et la flore atmosphériques. — Ann. Soc. Ent. France. 104. Paris. — — 1937: Données recentes sur le transport aérien d’animaux et de plantes, d’apres des recherches faites en avion. — C. R. Soc. Biogeogr. 14. Paris. BJÖRKMAN, G. 1939: Kärlväxtfloran inom Stora Sjöfallets nationalpark &c. — _K. Vet. Ak. Aivh. i Naturskyddsär. 2. Sthlm. Biair, K. G. 1931: The beetles of the Scilly Islands. — Proc. Zool. Soc. 4. London. Brunck, H. 1925: Syllabus der Insektenbiologie. Coleoptera: 1. — Berlin. Biytt, A. 1893: Zur Geschichte der Nordeuropäischen, besonders der Nor- wegischen Flora. — Englers Bot. Jahrb. 17. Beiblatt 41. Leipzig. BÖCHER, T. W. 1938: Biological distributional types in the flora of Greenland. — Medd. om Grgnl. 106. Kbhvn. BODENHEIMER, F. S. 1931: Uber die Temperaturabhängigkeiten der Insekten. III. Die Beziehungen der Vorzugstemperatur zur Luftfeuchtigkeit der Um- gebung. — Zeitschr. Vergl. Physiol. 13. Berlin. — — 1938: Problems of animal ecology. — Oxford. BODENHEIMER, F. S. & SCHENKIN, D. 1928: Über die Temperaturabhängigkeit von Insekten. I. — Zeitschr. Vergl. Physiol. 8. Berlin. Bge, J. & Nummepat, A. 1936: Le Finnmarkien. Les origines de la civilisation dans l’extreme-nord de l’Europe — Inst. Sammenlign. Kulturforskn. B (Skrifter). 32. Oslo. Bonnier, G. & Tepın, O. 1940: Biologisk variationsanalys. — Sthlm. BoRCHERT, W. 1938: Die Verbreitung der Käfer Deutschlands. Tatsachen und Folgerungen. — Schönebeck. Born, P. 1926: Die Carabenfauna Norwegens. — Norsk Ent. Tidsskr. 2. Oslo. BRAENDEGAARD, J. 1932: Isländische Spinnentiere. 3. Araneae. — K. Vet. Vitt. Samh. Handl. (5) B. 2. Göteborg. BREITENBECHER, J. K. 1925: An apterous mutation in Bruchus. — Biol. Bull. Marine Biol. Labor. 48. Woods Hole, Mass. — — 1926: Variation and heredity in Bruchus quadrimaculatus, Fabr. — Canad. Ent. 58. Orillia. BRENNER, M. 1871: Bidrag till kannedomen af Finska vikens ovegetation. — Notiser Sallsk. Fauna Fl. Fenn. 11. Hfors. BRENNER, W. 1930: Beitrage zur edaphischen Okologie der Vegetation Finnlands. I. Kalkbegiinstigte Moore, Wiesen und Wiesenwalder. — Acta Bot. Fenn. 7. Hfors. 775 873 BREUNING, S. 1932-36: Monographie der Gattung Carabus L. — Best.-Tab, europ. Coleopt. 104-110. Troppau. Brinck, P. 1948: Coleoptera of Tristan da Cunha. — Res. Norw. Sci. Exp. Tristan da C. 1937—1938. 17. (Norske Vid.-Ak.). Oslo. BROCKMANN-JEROSCH, H. 1909: Neue Fossilfunde aus dem Quartar und deren Bedeutung für die Auffassung des Wesens der Eiszeit. — Vierteljahrsschr. Naturf. Gesellsch. 54. Zürich. — — 1919: Weitere Gesichtspunkte zur Beurteilung der Dryasflora. — Ibidem. 04. BROTZEN, F. 1948: The Swedish Paleocene and its foraminiferal fauna. — Sver. Geol. Unders. C 493 (Arsbok 42). Sthim. BrueEs, T. 1903: The structure and significance of vestigial wings among insects. — Biol. Bull. Marine Biol. Lab. Woods Hole, Mass. Bryk, F. 1942: Zur Kenntnis der Grossschmetterlinge der Kurilen. — Deutsch. Ent. Zeitschr. „Iris”. 56. Dresden. Burcess, A. F. 1911: Calosoma sycophanta: Its life history, behavior, and success- ful colonization in New England. — U. S. Dept. Agric., Bureau of Ent., Bull. 101. Washington. Burcess, A. F. & Corrins, C. W. 1917: The Genus Calosoma &c. — Ibidem. 417. BURMEISTER, F. 1939: Biologie, Ökologie und Verbreitung der europäischen Käfer. — Krefeld. CAJANDER, A. K. 1921: Zur Kenntnis der Einwanderungswege der Pflanzen- arten nach Finnland. — Acta Forest. Fenn. 19. Hfors. Caspers, H. 1942: Die Landfauna der Insel Helgoland. — Zoogeographica. 4. Jena. Catalogus Coleopterorum Daniae et Fennoscandiae, 1939. — (Soc. Fauna Fl. Fenn.) Hfors. CEDERCREUTZ, C. 1937: Eine pflanzengeographische Einteilung der Seen Alands und die regionale Verbreitung der verschiedenen Seetypen. — Acta Soc. Fauna Fl. Fenn. 60. Hfors. CuHasoussou, F. 1939: Contribution a |’ étude biologique de Lebia grandis Hentz., prédateur americain du Doryphore. — Ann. Epiphyt. Phytogén. 5. Paris. CHARLESWORTH, J. K. 1929: The South Wales end-moraine. — Quart. Journ. Geol. Soc. 85. London. — — 1930: Some geological observations on the origin of the Irish Fauna and Flora. — Proc. R. Irish Acad. 39 (B). Dublin. , CHOoPARD, L. 1932: Un cas de microphthalmie liée & |’ atrophie des ailes chez une blatte cavernicole. — Soc. Ent. France, Livre de Centen. Paris. (Zitiert nach DARLINGTON 1930.) CHOPARD, L. & BELLECROIXx, R. 1928: Dimorphisme alaire chez les Gryllides &c. — Bull. Biol. France et Belg. 62. Paris. CLEvE-EULER, ASTRID 1934: Det gotiglaciala havets utbredning samt maximihojd i Nordfinland och i trakterna omkring Vita Havet. — Terra. 46. Hfors. — — 1946: Om den sista landisens bortsmältning frän Södra Sverige, den s. k. Baltiska issjön, tappningarna vid Billingen och Degerfors samt Vatterns historia. — Sthlm. CoLtet, R. 1866: Zoologisk botaniske Observationer fra Hvalgerne. — (Nyt Mag. f. Naturv. 15. 1868.) Oslo. Cook, W. C. 1924: The distribution of the pale western cutworm, Porosagrotis 874 776 orthogonia, Morr., a study in physical ecology. — Ecology. 5. Brook- lyn, N. Y. CoRSET, J. 1931: Les coaptations chez les insectes, — Bull. Biol. France et Belg., Suppl. 13. Paris. CROMBIE, A. C. 1947: Interspecific competition. — Journ. Animal Ecol. 16. Cam- bridge. Csıkı, E. 1927-33: Carabinae und Harpalinae. — Junk & SCHENKLING, Coleopt. Catal. 91. 92. 97. 98. 104. 112. 115. 121. 124. 126. 127. — Berlin. Curnot, L. & Mercier, L. 1923: Les muscles du vol chez les mutants alaires des Drosophiles. — C. R. Acad. Sciences. 176 (p. 1112). Paris. Daur, EıLır 1946: On different types of unglaciated areas during the Ice Ages and their significance to phytogeography. — New Phytologist. 45. Cambridge. Daur, T. (MrozEk-) 1928: Coleoptera oder Käfer. I: Carabidae (Laufkäfer). — DaHL: Tierwelt Deutschlands &c. 7. — Jena. DARLINGToN, P. J. Jr 1936: Variation and atrophy of flying wings of some Carabid beetles. — Ann. Ent. Soc. America. 29. Brooklyn, N. Y. — — 1938: The American Patrobini (Coleoptera, Carabidae). — Ent. Americ. (N. F.) 18. Brooklyn, N. Y. — — 1943: Carabidae of mountains and islands : Data on the evolution of isolated faunas, and on atrophy of wings. — Ecol. Monogr. 13. Durham, NG: Darwin, CH. 1859: Origin of Species. — London. DE Geer, CH. 1774: Memoires pour servir a l’histoire des Insectes. 4. — Sthlm. DE Geer, G. 1940: Geochronologia Suecica Principles. — K. Vet. Ak. Handl. (III) 18. 6. Sthim. DEGELIUS, G. 1935: Das ozeanische Element der Strauch- und Laubflechtenflora von Skandinavien. — Acta Phytogeogr. Suec. 7. Uppsala. — — 1936: Till kannedomen om lavfloran pa bark, lignum och urbergsblock pa Gotland. — Bot. Notiser. Lund. DELKESKAMP, K. 1930: Biologische Studien über Carabus nemoralis Mull. — Zeitschr. Morph. u. Ökol. d. Tiere. 19. Berlin. DEmoLL, R. 1918: Der Flug der Insekten und der Vogel. — Jena. Dewızz, J. 1917: Uber die Entstehung rudimentarer Organe bei den Tieren. — Zool. Jahrb., Abt. allg. Zool. u. Physiol. 36. Jena. Dietze, H. 1939: Die Käfer — insonderheit die Carabidae — der Insel Helgo- land. — Zool. Jahrb., Abt. f. Syst. &c. 73. Jena. Dimmock, G. & Knar, F. 1904: Early stages of Carabidae. — Springfield Mus. Nat.. Hist. 1. Springfield, Mass. Dorn, K. 1946: Eine Käferausbeute an elektrischem Licht in Temesvar, — Addita- menta Faun. Col. 4. Hamburg. Drupe, ©. 1918: Die Pflanzengeographie und die Naturschutzpflege. — Abh. Naturw. Gesellsch. Isis. 4 (1917). Dresden. (Zitiert nach WANGERIN 1924.) Duranoo, S. 1946: Faro, naturkontrasternas 6. — Natur pa Gotland. (Svensk Natur.) Goteborg. Du Retz, G. E. 1925: Die regionale Gliederung der skandinavischen Vegetation. — Sv. Vaxtsoc. Sällsk. Handl. 8. Uppsala. — — 1931: Vegetationsforschung auf soziationsanalytischer Grundlage. — Ap- DERHALDEN: Handb. biol. Arbeitsmeth. 5, 1. Wien. Td 875 — — 1935 a: Klimat och jordman. — Hantverkets Bok. 3, 1. Sthlm. — — 1935 b: Glacial survival of plants in Scandinavia and the British Isles. — Discuss. Orig. & Relationship Brit. Flora. — Proc. R. Soc. (B) 118. London. — — 1942: De svenska fjällens vaxtvarld. — Norrland (Geogr. Förbundet & Industriens Utredn.-inst.). Sthlm. EDER, ROBERTA 1940: Die kutikuläre Transpiration der Insekten und ihre Ab- hängigkeit vom Aufbau des Integumentes. — Zool. Jahrb., Abt. allg. Zool. u. Physiol. 60. Jena. Epin, G. (Lunpguist, G. & SODERBERG, R.) 1941: Det svenska äret. En ärstids- kronika. — Sv. Turist.-Foren. 760. Sthlm. EKBLoOM, T. 1927-28: Vererbungsbiologische Studien über Hemiptera-Heteroptera. 1. Gerris asper Fieb. — Hereditas. 10. Lund. — — 1941: Untersuchungen über den Flügeldimorphismus bei Gerris asper L. — Notulae Ent. 21. Hfors. EKHOLM, GUNNAR 1925: Manniskor i Skandinavien under sista interglacialen? — Ymer. 45. Sthlm. EKLUND, O. 1927 a: Versuche über das Keimungs- und Schwimmvermogen einiger Samen und Früchte im Ostseewasser. — Mem. Soc. Fauna Fl. Fenn. 2. Hfors. — — 1927 b: Weitere Versuche über Keimung in Meerwasser. — Ibidem. 3. — — 1928: Notizen über die Flora des nördlichen und westlichen Dagö (Hiiumaa) in Estland. — Ibidem. 4. — — 1929: On the resistibility of some seeds against seasalt. — Ibidem. 5. — — 1931: Über die Ursachen der regionalen Verteilung der Schärenflora Süd- west-Finnlands. — Acta Bot. Fenn. 8. Hfors. — — 1933: Urkalken i skärgärdshavet och dess betydelse för växterna. — Terra. 45, 2. Hfors. — — 1937: Klimabedingte Artenareale. — Acta Soc. Fauna Fl. Fenn. 60. Hfors. Exman, S. 1915: Vorschläge und Erorterungen zur Reliktenfrage in der Hydro- biologie. — Arkiv f. zool. 9. Sthlm. — — 1920: Der skandinavische Lemming (Lemmus lemmus) als Uberrest einer interglazialen skandinavischen Fauna. — Festschr. f. Zschokke. Basel. — — 1922: Djurvärldens utbredningshistoria pa skandinaviska halvön. — Sthlm. — — 1930: Die südbaltischen maringlazialen Relikte und die Stauseetheorie. — Internat. Revue gesamt. Hydrobiol. u. Hydrogr. 24. Leipzig. — — 1935: Tiergeographie des Meeres. — Leipzig. — — 1940: Begründung eıner statistischen Methode in der regionalen Tiergeo- graphie nebst einer Analyse der paläarktischen Steppen- und Wüstenfauna. — Nova Acta R. Soc. Sc. (IV) ı2. Uppsala. — — 1944: Djur i de svenska fjällen. — Sv. Turist-Fören. handb. sv. fjallet. 3. Sthlm. Exstrom, C. U. 1828: Beskrifning öfver Mörkö Socken i Södermanland. — Sthlm. EKSTRÖM, G. 1927: Klassifikation av svenska äkerjordar. — Sver. Geol. Unders. C 345. Sthlm. ELFSTRAND, M. 1927: Var hava fanerogama växter övcrlevat sista istiden i Skandi- navien? — Svensk Bot. Tidskr. 21. Sthlm. 778 876 ELton, C. S. 1925: The dispersal of insects to Spitsbergen. — Trans. Ent. Soc. London. Eton, CH. 1930: Animal ecology and evolution. — Oxford. — — 1946: Competition and the structure of ecological communities. — Journ. Animal Ecol. 15. Cambridge. ENGsTROM, A. 1926: Gotska Sandon. — Stockholm, ENQUIST, F. 1913: Om den senglaciala utvecklingen inom norra Skandinaviens fjall- trakter. — Geol. Foren. Forh. 35. Sthlm. — — 1916: Der Einfluss des Windes auf die Verteilung der Gletscher. — Bull. Geol. Inst. 14. Uppsala. — — 1918: Die glaziale Entwicklungsgeschichte Nordwestskandinaviens. — Sver. Geol. Unders. C. 285. Sthlm. — — 1924: Sambandet mellan klimat och vaxtgranser. — Geol. Foren. Förh. 46. Sthim. — — 1929: Studier over samtidiga vaxlingar i klimat och växtlighet. — Sv. Geogr. Arsbok. Lund. — — 1933: Tradgransundersokningar. — Sv. Skogsvardsforen. Tidskr. 31. Sthlm. EPPELSHEIM, E. 1879: Ueber geflügelte und ungeflügelte Lathrobien. — Deutsch. Ent. Zeitschr. 23. Berlin. ERDTMAN, G. 1943: Pollenspektra fran svenska vaxtsamhallen jamte pollenanaly- tiska markstudier i sodra Lappland. — Geol. Foren. Förh. 65. Sthlm. — — 1946: Palynologiska synpunkter pa primarfasen i den svenska florans in- vandringshistoria. — Sv. Bot. Tidskr. 40. Uppsala. ERLANDSSON, S. 1939: Arktiska kustvaxters invandring fran Ishavskusten till Bott- niska viken. — Naturen och Vi, Hft. 11. Sthlm. — — 1940: Saussurea alpina (L.) DC. funnen i Smäland. — Bot. Notiser. Lund. ESCHERICH, K. 1940: Nachwort zu dem Aufsatz von Heikertinger. Zugleich ein Aufruf. — Zeitschr. angew. Ent. 27. Berlin. Esxora, P., Hackman, V., LAITAKARI, A. & WILKMAN, W. W. 1929: Kalkstenen i Finland. — Geol. Komm. Finl., Geotekn. Medd. 21. Hfors. Feori, K. 1937: Some fundimental problems of taxonomy and phylogenetics. — Bot. Review. 3. Lancaster Pa. — — 1940: Quartärgeologische Untersuchungen im westlichen Norwegen. II. Zur spatquartären Geschichte Jzrens. — Bergens Mus. Arbok (1939-40). Ber- gen. FARRINGTON, A. 1945: The level of the ocean in glacial and late-glacial times. — Proc. R. Irish Acad. 50. Dublin. FERNALD, M. L. 1925: Persistence of plants in unglaciated areas of boreal America. — Mem. Amer. Ac. Arts & Sc. 15.3. Lancaster, Pa — — 1929: Some relationships of the floras of the Northern Hemisphere. — Proc. Internat. Congr. Plant Scie. Ithaca (1926). 2. Menaska, Wisc. FırBAs, F. 1939: Vegetationsentwicklung und Klimawandel in der mitteleuropäischen Spät- und Nacheiszeit. — Die Naturwissenschaften. 27,6. Berlin. FLInT, R. F. 1947: Glacial Geology and the Pleistocene Epoch. — New York. Forp, E. B. 1945: Butterflies. — The New Naturalist. — London. FoRSSLUND, K.-H. 1943: Studier över det lagre djurlivet i nordsvensk skogsmark. — Medd. Statens Skogsforsoksanst. 34. Sthlm. 779 877 FRANCK, P. 1928: Nachtrag zu den Lauf- und Schwimmkäfern unserer Fauna. — Verh. Ver. Naturw. Unterhalt. 19. Hamburg. Franz, H. 1931: Uber die Bedeutung des Mikroklimas für die Faunazusammen- setzung auf kleinem Raum &c. — Zeitschr. Morph. Ökol. d. Tiere. 22. Berlin. — -- 1933: Auswirkungen des Mikroklimas auf die Verbreitung mitteleuropaischer xerophiler Orthopteren. — Zoogeographica. 1. Jena. — — 1938: Revision der Verwandtschaftsgruppe der Chrysomela gypsophilae Küst. (Coleopt. Chrysomel.) — Ent. Blätter. 34. Krefeld. — — 1939: Grundsätzliches über tiersoziologische Aufnahmsmethoden mit beson- derer Berücksichtigung der Landbiotope. — Biol. Reviews. 14. Cambridge. — — 1943 a: Die Landtierwelt der mittleren Hohen Tauern. Ein Beitrag zur tier- geographischen und soziologischen Erforschung der Alpen. — Denkschr. Ak. d. Wiss., Math.-Naturw. Kl. 107. Wien. — — 1943 b: Die Tätigkeit der Kleintiere im Boden und Wirtschaftsdünger und ihre Bedeutung für das Dauergrünland. — Pflanzenbau. 19. Leipzig. Frey, R. 1937: Einige Massenvorkomnisse von Insekten an der Südküste Finn- lands während des Sommers 1935. — Acta Soc. Fauna Fl. Fenn. 60. Hfors. FRIEDERICHS, K. 1930: Die Grundfragen und Gesetzmässigkeiten der land- och forstwissenschaftlichen Zoologie. 1. — Berlin. Fries, M. 1948: Limes norrlandicus-studier. En vaxtgeografisk gränsfräga histo- riskt belyst och exemplifierad. — Sv. Bot. Tidskr. 42. Uppsala. — — 1949: Den nordiska utbredningen av Lactuca alpina, Aconitum septentrio- nale, Ranunculus platanifolius och Polygonatum verticillatum. — Acta Phy- togeogr. Suec. 24. Uppsala. Fries, THoRE 1913: Botanische Untersuchungen im nördlichsten Schweden &c. — Diss. — Uppsala. — — 1925: Die Rolle des Gesteinsgrundes bei der Verbreitung der Gebirgspflanzen in Skandinavien. — Sv. Växtsoc. Sallsk. Handl. 6. Uppsala. — — 1926: En vaxtsociologisk huvudfraga. — Ibidem. 10. Fropin, J. 1915: Nagra märkliga sydberg i Lule lappmark &c. — Sv. Bot. Tidskr. 9. Sthlm. — — 1919: Om förhällandet mellan berggrundens kalkhalt och de nordsvenska vaxtarternas utbredning. — Bot. Notiser. Lund. FromaN, I. 1944: De senkvartära strandforskjutningarna som vaxtgeografisk fak- tor i belysning av murgronans geografi i Skandinavien—Baltikum. — Geol. Foren. Förh. 66. Sthim. Fics, B. 1919: Einwanderung von Insekten auf einer entstehenden Insel unter Be- rücksichtigung der gesammelten Coleopteren. — Zeitschr. wiss. Insektenbiol. 14. Berlin. GANGLBAUER, L. 1892, 1904: Die Käfer von Mitteleuropa. 1, 4. — Wien. Gaunitz, C. B. 1933: Nagra anteckningar fran Sorsele socken om arterna Carabus violaceus L. och Carabus glabratus Payk. — Ent. Tidskr. 54. Sthlm. GEIGER, R. 1942: Das Klima der bodennahen Luftschicht. — Die Wissenschaft. 78. Braunschweig. Ge.tinc, P. 1934: Studies on the vascular plants of East Greenland &c. — Medd. om Grgnl. 101. Kbhvn. 780 878 GERSDORF, E. 1937: Ökologisch-faunistische Untersuchungen über die Carabiden der mecklenburgischen Landschaft. — Zool. Jahrb., Abt. Syst. 70. Jena. Gistn, H. 1947: Analyses et synthéses biocénotiques. — Arch. Sc. Phys. et Nat. (5) 29. Genéve. GIsLEN, T. 1940: The number of animal species in Sweden with remarks on some rules of distribution especially of the microfauna. — K. Fysiogr. Sallsk. Handl. (N. F.). 51. (Lunds Univ. Arsskr. [N. F.] Avd. 2. 36.) Lund. — — 1948: Aerial plankton and its conditions of life. — Biol. Reviews. 23. Cam- bridge. GLick, P. A. 1939: The distribution of insects, spiders and mites in the air. — U. S. Dept. Agric., Techn. Bull. 673. Washington. ' GOLDSCHMIDT, R. 1938: Physiological genetics. — New York & London. GRANLUND, E. 1932: De svenska hogmossarnas geologi &c. — Sver. Geol. Unders. C 373 (Arsbok 26). Sthlm. — — 1936, s. MAGNUSSON & GRANLUND. GRANLUND, E. & Lunpgvist, G. 1936: Nagra iakttagelser fran en resa i Helge- land sommaren 1935. — Norsk Geogr. Tidsskr. 6. Oslo. GriLL, C. 1896: Catalogus Coleopterorum Scandinaviae, Daniae et Fenniae &c. — Sthim. Gr@NLiE, O. T. 1927: The Folden Fiord. Contributions to the quaternary geology of the Folla district. — Tromsg Mus. Skrifter. 1.2. Tromso. Gross, H. 1930: Das Problem der nacheiszeitlichen Klima- und Floraentwicklung in Nord- und Mitteleuropa. — Beihefte Bot. Centralbl. 47. II,1. Dresden. Gunn, D. L. & KEnnepy, J. S. 1936: Apparatus for investigating the reactions of land arthropods to humidity. — Journ. Exp. Biol. 13. Cambridge. GYLLENHAL, L. 1810— 27: Insecta Svecica. II—IV. — Skara. HABERMAN, H. 1935: Mardikate esmasleide Eestis, III. — Eesti Lodus (Estonian Nature). 3. Tartu. Hacem, O. 1917: Furuens og granens frgsetning i Norge. — Medd. Vestlandets Forstl. Forsgksstation. 2. Bergen. HALBERT, J. N. 1937: In Ent. Monthly Mag. — London. HALDANE, J. B. S. 1932: The causes of evolution. — London. HAMBERG, A. 1924: Till frägan om förhällandet mellan vaxtgranser och klimat. — Geol. Foren. Forh. 46. Sthlm. HAMBERG, H. E. 1908: Medeltal och extremer af lufttemperaturen i Sverige. — Bih. Meteor. Iaktt. i Sverige. 49 (1907). Uppsala. — — 1909: Molnighet och solsken pa den skandinaviska halvon. — Ibidem. 50. — — 1914: Storleken av temperaturens dagliga variation pa den skandinaviska halvon. — Ibidem. 54 (1912). HANDLIRSCH, A. 1913 a: Die systematischen Grundbegriffe. — ScHROpER: Handb. d. Ent. 3. Jena: — — 1913 b: Nomenklatur, Typen und Zitate. — Ibidem. Hansen, A. M. 1904: Landnam i Norge. En utsigt over bosztningens historie. — Oslo. — — 1929: Bre og biota. — Norske Vid. Ak. Skr., Mat.-naturv. Kl. 5. Oslo. Hansson, C. A. 1902: Spridda anteckningar om skandinaviska ratvingar. — Ent. Tidskr. 23. Sthlm. 781 879 HArp ar SEGERSTAD, F. 1924: Sydsvenska florans vaxtgeografiska huvudgrupper. — Malmo. — — 1939: Vaxtgeografisk karta over Sverige. — (Norstedt) Sthlm. Harnty, M. H. 1930: A critical temperature for lengthening of the vestigial wings of Drosophila melanogaster with sexually dimorphic effects. — Journ. Exp. Zool. 56. Philadelphia. Harrison, J. W. Hestop 1947: The pleistocene races of certain British insects and distributional overlappings. — Ent. Record. London. Hausen, H. 1910 a: De gamla strandbildningarna pa Aland och deras forhallande till stenaldersboplatserna. — Fennia. 28. Hfors. — — 1910 b: Älands och Abo skargard. — Atlas öfver Finland. Text. (Sallsk. Finl. Geogr.) Hfors (1911). — — 1946: Alands natur. En handledning for naturvannen, alanningen och som- margasten. — Abo. Heperbey, R. F. 1933: Die Bedeutung der Eiszeit fur die Fauna der Alpen. — Zoogeographica. 1. Jena. — — 1939: Die Bedeutung tiergeographischer Befunde für die Altersbestimmung rezenter Tierformen. — VII. Intern. Kongr. Ent. Berlin. 1. Weimar. HEBERDEY, R. & MEIXNER, J. 1933: Die Adephagen der östlichen Hälfte der Ost- alpen. — Verh. Zool.-Bot. Gesellsch. 73. Wien. Heerpt, P. F. van 1946: Eenige physiologische en oecologische problemen bij Forficula auricularia. — Utrecht. HEIKERTINGER, F. 1914: Noch ein Gedenkblatt für Ludwig Ganglbauer. — Wiener Ent. Zeit. 33. Wien. — — 1935: Ein neuer Beweis für die Nichteignung des Prioritätsprinzips in der Tierbenennung. — Koleopt. Rundschau. 21. Wien. — — 1939 a: Über das Namenändern in der Zoologie. Der Fall Chalcoides (Col. Haltic.) als Schulbeispiel. — Ent. Blätter. 35. Krefeld. — — 1939 b: Über den heutigen Stand des Nomenklaturproblems. Die Alltagsfor- mel als Lösung. — VII. Intern. Kongr. Ent. Berlin. 1. Weimar. — — 1941: Welche Einwände werden gegen das Kontinuitatsprinzip in der No- menklatur erhoben? — Zeitschr. angew. Ent. 28. Berlin. — — 1942 a: Das Nomenklaturproblem der Gegenwart. Zugleich ein Aufruf an alle Biologen. — Der Biologe. 11. Berlin. — — 1942 b: Wie arbeitet das Kontinuitätsprinzip in der Nomenklatur? — Ent. Zeitschr. 56. Guben. (Enthält ein vollständiges Verzeichnis seiner nomenkla- torischen Schriften.) Heinze, A. 1914: Om hydrokor spridning af vegetationsklädda tufvor. — Sv. Bot. Tidskr. 8. Uppsala. — — 1932—35: Handbuch der Verbreitungsökologie der Pflanzen. 1. 2. — Sthlm. HEINZE, V. 1947: Lionychus quadrillum Duft. (Col. Carabidae) funnen talrikt vid Örebro. — Ent. Tidskr. 68. Sthim. HELLAND, A. 1808, 1907, 1911: Norges land og folk. 15. 18.1 u. 3. 16. — Oslo. HELLEN, W. 1925: Die nordeuropaischen Arten der Gattung Mycetoporus Mann. (Col.). — Notulae Ent. 5. Hfors. — — 1929: For Finland nya eller sallsynta Coleoptera. — Mem. Soc. Fauna Fl. Fenn. 5. Hfors. — — 1934: Die Carabenfauna Finnlands. — Notulae Ent. 14. Hfors. 880 782 HELMS, ANNA & JORGENSEN, C. A. 1924: Temperaturforholdene paa Maglemose og deres Betydning for Vegetationen. — Bot. Tidsskr. 38. Kbhvn. HEMMENDORFF, E. 1897: Om Olands vegetation. — Uppsala. HENRIKSEN, K. L. 1933: Undersggelser over Danmark—Skanes kvartzre Insekt- fauna. —- Vid. Medd. Dansk Naturh. Foren. 96 (Festskrift II). Kbhvn. Herter, K. 1923: Untersuchung über den Temperatursinn der Hausgrille (Acheta domestica) und der roten Waldameise (Formica rufa). — Biol. Centralbl. 43. Leipzig. — — 1924: Untersuchungen über den Temperatursinn einiger Insekten. — Zeit- schr. vergl. Physiol. ı. Berlin. — 1934: Eine verbesserte Temperaturorgel und ihre Anwendung auf Insekten und Säugetiere. — Biol. Centralbl. 54. Leipzig. — — 1939: Über den Temperatursinn der Insekten. — VII. Intern. Kongr. Ent. Berlin. 2. Weimar. L'HERITIER, Pu., Neers, Y. & Teıssier, G. 1937: Apterism des Insectes et selec- tion naturelle. — C. R. Ac. Sci. 204 (p. 907). Paris. Hesse, R. 1924: Tiergeographie auf ökologischer Grundlage. — Jena. HESSELBERG, TH. & BirkELAND, B. J. 1940: Säkulare Schwankungen des Klimas von Norwegen. — Geofys. Publ. 14. Oslo. HESSELMAnN, H. 1932: Om klimatets humiditet i värt land och dess inverkan pä mark, vegetation och skog. — Medd. Statens Skogsförsöksanst. 26. Sthlm. HEYDEMANN, F. 1930: Der Einfluss des atlantischen Klimas auf die Lepidopteren- Fauna Nordwest-Europas, insbesondere Schleswig-Holsteins, — 4. Wander- vers. Deutsch. Ent. Kiel. Berlin-Dahlem. — — 1931: Selid. ericetaria Vill. subsp. scandinaviaria Stdgr. — Intern. Ent. Zeitschr. 25. Guben. Hirtonen, I. 1946: Die Karelische Landenge als Einwanderungsweg der Pflanzen- arten ım Lichte ihrer heutigen Verbreitung. Ann. Bot. Soc. Zool.-Bot. Fenn. Vanamo. 22. Hfors. Hinton, H. E. 1948: A synopsis of the genus Tribolium Macleay, with some re- marks on the evolution of its species-groups. — Bull. Ent. Research. 39. London. HjJeLmovıst, H. 1940: Studien über die Abhängigkeit der Baumgrenzen von den Temperaturverhältnissen unter besonderer Berücksichtigung der Buche und ihre Klimarassen. — Lund. Hoesten, N. v. 1911: Zur Kenntnis der Tiefenfauna des Brienzer und des Thuner Sees. — Arch. f. Hydrobiol. u. Planktonkunde. 7. Stuttgart. — — 1913: Glaciala och subarktiska relikter i den svenska faunan. — Pop. Naturv. Revy. 3. Sthlm. — — 1920: Planaria alpina som glacialrelikt pa Bornholm jämte nägra ord om landfaunans invandring till Bornholm. — Vid. Medd. Dansk Naturh. Foren. 71. Odense. Höcsom, A. & Lunpovist, G. 1930: Beskrivning till Kartbladet Malingsbo. — Sver. Geol. Unders. Aa. 168. Sthlm. HoroHaus, K. 1911 a: Uber die Abhängigkeit der Fauna vom Gestein. — Verh. VIII. Intern. Zool.-Kongr. Graz (1910). Jena. 783 881 — — ıgıı b: Ueber die Abhängigkeit der Fauna vom Gestein. — I. Congr. Intern. d’Ent. (1910). 1. Bruxelles. — — 1929: Die geographische Verbreitung der Insekten. — SCHRÖDER: Handb. d. Ent. 2. Jena. — — 1932: Die europäische Höhlenfauna in ihren Beziehungen zur Eiszeit. — Zoogeographica. 1. Jena. HorpHAus, K & LIiNDROTH, C. H. 1939: Die europäischen Koleopteren mit boreo- alpiner Verbreitung. — Ann. Naturh. Mus. 50. Wien. HoLMBoE, J. 1913: Kristtornen i Norge. En plantegeografisk undersokelse. — Bergens Mus. Aarb. 7. Bergen. — — 1936: Über Nigritella nigra (L.) Rchb., ihre Verbreitung und Geschichte in Skandinavien. — Ber. Schweiz. Bot. Gesellsch. 46. Bern. — — 1937: The Trondheim district as a centre of late glacial and postglacial migrations. — Avh. Norske Vid.-Ak. (I. Mat.-Naturv. K1.). Oslo. Horst, N. O. 1906: De senglaciala lagren vid Toppeladugard. — Sver. Geol. Unders. C 200. Sthlm. — — 1908: Efterskord fran de senglaciala lagren vid Toppeladugard. — Ibidem. 210 (Arsbok 2). — — 1911: Alnarps-floden, en svensk „Cromer-flod”. — Ibidem. 237 (Arsbok 4). HoLste, G. 1915: Calosoma sycophanta L &c. — Zeitschr. angew. Ent. 2. Berlin. HoLTEDAHL, O. 1929: Some remarkable features of the sub-marine relief on the north coast of the Varanger Peninsula, northern Norway. — Avh. Norske Vid.-Ak. (I. Mat.-Naturv. KlI.). Oslo. Hore-Sımson, J. F. 1938: A chalk flora on the Lower Greensands: its use in interpreting the calcicole habit. — Journ. Ecol. 26. Cambridge. Horıon, A. 1936: Studien zur deutschen Kaferfauna. 1. — Ent. Blatter. 32. Krefeld. — — 1937: Die rheinischen Arten der Tribus Bembidiini. — Decheniana. 95 B. Bonn. — — 1938: Studien zur deutschen Käferfauna. II. Die periodischen Klima- schwankungen und ihr Einfluss auf die thermophilen Käfer in Deutsch- land. — Ent. Blätter. 34. Krefeld. — — 1939: Studien zur deutschen Käferfauna. III. Weitere Beispiele für das sporadische und periodische Auftreten thermophiler Käfer in Deutschland. — Ibidem. 35. — — 1941: Faunistik der deutschen Kafer. 1. — Krefeld. Horn, W. 1907: Brullés „Odontochila aus dem baltischen Bernstein” und die Phylogenie der Cicindeliden. (Col.) — Deutsch. Ent. Zeitschr. Berlin. — — 1938: Uber 11 Thesen zwecks Reform der Entomo-Systematik. — ,,Memi- nisse-Serie”, 41. Berlin-Dahlem. Horn, W. & KAHLE, ILsE 1935-37: Über entomologische Sammlungen, Entomologen und Entomo-Museologi. — Berlin. Husenpick, B. 1947: Die Verbreitungsverhaltnisse der limnischen Gastropoden in Südschweden. — Zool. Bidr. 24. Uppsala. Hurten, E. 1937: Outline of the history of arctic and boreal biota during the quaternary period. — Sthlm. — — Mskr. (Atlas über die nordische Verbreitung der Gefässpflanzen). Im Druck. „Hütte”, des Ingenieurs Taschenbuch. I. 1931. 26. Aufl. — Berlin. 784 882 Huxıey, J. 1939: Clines: An auxiliary method of taxonomy. — Bijdr. Dierkunde. 27. Leiden. — — 1940: The new systematics. — Oxford. — — 1942, 1945: Evolution. The modern synthesis. (1. und 4. Aufl.) — London. Hyyepd, E. 1933: Das Klima und die Walder der spatglazialen Zeit im Bereich der Karelischen Landenge. — Acta Forest. Fenn. 39. Hfors. — — 1936: Uber die spatquartare Entwicklung Nordfinnlands mit Ergänzungen zur Kenntnis des spatglazialen Klimas. — Bull. Comm. Géol. de Finlande. 115. Hfors. — — 1941: Uber das spatglaziale Klima in Finnland. — Geol. Rundschau. 32. Berlin. — — 1944: In Terra. 55. Hfors (Vortragsreferat, p. 122 u. f.). InerING, H. v. 1927: Die Geschichte des Atlantischen Ozeans. — Jena. IsBEerc, ©. 1930: Das ehemalige Vorkommen der Sumpfschildkrote (Emys orbi- cularis L.) in Schweden &c. — Arkiv f. Zool. 21. A. Sthlm. IveRSEN, J. 1941: Landnam i Danmarks Stenalder. En pollenanalytisk Under- ségelse over det fgrste Landbrugs Indvirking paa Vegetationsudviklingen. — Danm. Geol. Unders. (II) 66. Kbhvn. — — 1945: Helianthemum som fossil Glacialplante i Danmark. — Geol. Foren. Förh. 66. Sthlm. Jackson, Dorotuy, J. 1928: The inheritance of long and short wings in the weevil, Sitona hispidula, with a discussion of wing reduction among beetles. — Trans. R. Soc. 55. Edinburgh. — — 1933: Observations on the flight muscles of Sitona weevils. — Ann. Appl. Biol. 20. Cambridge. JacoBson, G. 1899: Über den äusseren Bau flügelloser Käfer. — Ann. Mus. Zool. Ac. Imp. Sci. Leningrad. Jansson, A. 1913: Frän exkursioner i Helsingborgs- och Engelholmstrakten. — Ent. Tidskr. 34. Sthlm. — — 1922:Faunistiska och biologiska studier Over insektlivet vid Hornsjon pa norra Öland. — Ark. f. zool. 14. Sthlm. — — 1925: Die Insekten-, Myriopoden- und Isopodenfauna der Gotska Sandon. — Örebro. — — 1927: Coleopterologiska bidrag. 16—18. — Ent. Tidskr. 48. Sthlm. — — 1933: Förteckning över Bornholms Coleoptera enligt litteratur samt H. Lohmanders och E. Klefbecks insamlingar. — Ibidem. 54. — — 1935: Supplement till Die Insekten-, Myriopoden- und Isopodenfauna der Gotska Sandon. — Ibidem. 56. — — 1945: Ceuthorrhynchus inaffectatus Gyll., en nykomling till Sveriges fauna eller en förbisedd art? — Ibidem. 66. — — 1947: En märklig nyhet för var staphylinidfauna. — Opusc. Ent. 12. Lund. JEANNEL, R. 1925: L’apterisme chez les Insectes insulaires. — C. R. Ac. Sci. 180 (p. 1222). Paris. -— — 1926-28: Monographie des Trechinae. — L’Abeille. 32. 33. 35. Paris. — — 1936: Les Trechinae des Iles Canaries. — Revue Franc. Ent. 3. Paris. — — 1937: Notes sur les Carabiques. 1. Le genre Nebria Latreille. — Ibidem. 4. — — 1940: L’aptérisme des Coléoptéres des iles australes et sa répercussion hérédi- taire. — C. R. Ac. Sci. 211. Paris. 785 883 — — 1941-42: Coléoptéres Carabiques. — Faune de France. 39. 40. Paris. Jessen, K. & Farrincton, A. 1938: The bogs at Ballybetagh, near Dublin, with remarks on late-glacial conditions in Ireland. — Proc. R. Irish Acad. 44 B. Dublin. Jessen, K. & MILTHERS, V. 1928: Stratigraphical and paleontological studies of interglacial fresh-water deposits in Jutland and northwest Germany. — Danm. Geol. Unders. (II) 48. Kbhvn. JoHansen, A. C. 1906: Om Temperaturen i Danmark og det sydlige Sverige i den senglaciale Tid. — Medd. Dansk Geol. Foren. 12. Kbhvn. — — 1908: Om Brugen of Betegnelsen ,,Relikt” i Naturhistorien. — Ibidem. 14. JoHANSEN, A. C. & Lynce, H. 1917: Om Land- og Ferskvandsmolluskerne SSO for Kalundborg, og deres Vidnesbyrd om Klimaforendringer. — Ibidem. 5. Jouansson, K. 1897: Hufvuddragen af Gotlands vaxttopografi och vaxtgeografi. — K. Vet. Ak. Handl. 29. Sthlm. Jouansson, O. V. 1910: Meteorologi. — Atlas öfver Finland, Text. (Sallsk. Finl. Geogr.) Hfors. — — 1929: Blir Europas klimat mera oceaniskt? — Terra. 41. Hfors. Jonansson, S. 1926: Hydrogeologisk undersökning av ett leromräde vid Skara. — Sver. Geol. Unders. C. 332. Sthlm. Joy, N. H. 1910: The behaviour of coleoptera in time of floods. — Trans. Ent. Soc. London. Jung, W. 1940: Ernahrungsversuche an Carabus-Arten. — Ent. Blatter. 36. Krefeld. KAIKKo, J. 1940: The bursting into leaf, defoliation and duration of foliage of birches in Finland. (Finnisch.) — Terra. 52. Hfors. KaLpHoL, H. 1930: Sunmgres kvartcergeologi. — Norsk Geol. Tidsskr. 11. Oslo. — — 1931: Har vi spor efter flere istider i Norge. — Ibid. 12. KALELA, A. 1943: Die Ostgrenze Fennoskandiens in pflanzengeographischer Be- ziehung. — Veroff. Geobot. Inst. Rübel. 20. Zurich. KALELA, O. 1938: Über die regionale Verteilung der Brutvogelfauna im Fluss- gebiet des Kokemäenjoki. — Ann. Zool. Soc. Zool.-Bot. Fenn. Vanamo. 5. Hfors. — — 1940 a: Uber die Einwanderung und Verbreitung des Iltis, Putorius putorius (L.), in Finnland. — Ann. Ac. Sci. Fenn. A. 54. Hfors. — — 1940 b: Zur Frage der neuzeitlichen Anreicherung der Brutvogelfauna in Fennoskandien mit besonderer Berücksichtigung der Austrocknung in den früheren Wohngebieten der Arten. — Ornis Fenn. 17. Hfors. — — 1944: Zur Frage der Ausbreitungstendenz der Tiere. — Ann. Zool. Soc. Zool.-Bot. Fenn. Vanamo. ı0. Hfors. — — 1946: Zur Ausbreitungsgeschichte der Vogel vegetationsreicher Seen. — Ann. Ac. Sci. Fenn. A. IV: 12. Hfors. — — 1947: Zur Charakteristik der neuzeitlichen Veränderungen in der Vogelfauna Mittel- und Nordeuropas. — Ornis Fenn. 23 (1946). Hfors. — — 1948: Changes in the distribution of the Polecat in Finland. — Suomen Riista. 2. Hfors. „Kalkutredningen” 1931. Utredning rörande det svenska jordbrukets kalkforsorjning. Sver. Geol. Unders. (stenciliert). Sthlm. 786 884 KEMNER, N. A. 1936: Tva nya skadedjur inomhus, Tribolium destructor Uyttenb. och Sitotroga cerealella Oliv. — Opusc. Ent. 1. Lund. Kempers, K. J. W. 1897: Bijdrage tot de kennis der Coleopterafauna van het eiland Texel. — Tijdschr. v. Ent. 40. Haag. KERANEN, J. 1920: Uber die Temperatur des Bodens und der Schneedecke in Sodankyla. — Ann. Ac Scient. Fenn. A. 13. Hfors. — — 1944: Uber die Temperaturschwankungen in Finnland und Nordeuropa in den letzten hundert Jahren. — Sitz.-ber. Finn. Ak. d. Wiss. (1941). Hfors. Kesster, P. 1925: Das eiszeitliche Klima und seine geologischen Wirkungen im nicht vereisten Gebiet. — Stuttgart. KiAErR, A, N, 1885: Norges land og folk. 1. — Oslo. KiIEFFER, J. J. 1914: Serphidae et Calliceratidae. — Das Tierreich. 42. Berlin. KiıHLmAan, A. O. 1890: Pflanzenbiologische Studien aus Russisch Lappland. — Acta Soc. Fauna Fl. Fenn. 6. Hfors. KIRIAKoFF, S. G. 1948: On the so-called ”lower” taxonomic categories. — Lepid. News. 2 (p. 3). Cambridge, Mass KJELLANDER, E. 1943: Pararge achine Scop. (Deianira L.), en fjarıl som kanske försvunnit fran värt land, jamte anknytande djurgeografiska reflexioner. — Opusc. Ent. 8. Lund. KLINGSTEDT, H. 1939: Die Uvarovsche Theorie der Wanderheuschreckenphasen und ihre Bedeutung für die Zoologie. — Notulae Ent. 19. Hfors. KniıpovitscHh, N. M. 1906: Grundzüge der Hydrologie des europäischen Eis- meeres. (Russisch.) — Russk. Obsc., Otdel Geogr, Zapiski 42. Leningrad. KoıseE, H. 1921: Über Mutationen und Aberrationen deutscher und auswärtiger Coleopteren und anderer Insekten. — Deutsch. Ent. Zeitschr. 40. Berlin. — — 1933: Fossile Coleopteren aus praeglazialer und spätglazialer Zeit Süd- Schwedens. — Ent. Medd. 18. Kbhvn. Konic, W. 1936: Biologische Studien über Ptinus tectus Boield. — Zeitschr. Wiss. Zool. 148. Leipzig. Kopren, W. &. WEGENER, A. 1940: Die Klimate der geologischen Vorzeit. Er- gänzungen und Berichtigungen. — Berlin. KoRrNERUP, A. 1890: Det organiske Liv paa den gstligste Nunatak. — Medd. om Grenl. 1 (2. Aufl.). Kbhvn. KoRSCHELT, E. 1923: Bearbeitung einheimischer Tiere. 1. Der Gelbrand, Dytiscus marginalis L. 1. — Leipzig. KOoTILAINEn, M. J. 1933: Zur Frage der Verbreitung des atlantischen Floren- elementes Fennoskandias. — Ann. Bot. Soc. Zool.-Bot. Fenn. Vanamo. 4. Hfors. Kraus, G. 1911: Boden und Klima auf kleinstem Raum. — Jena. Krocerus, H. 1945: Die Veränderungen der Schmetterlingsfauna des Lojo- Gebietes in Sudwestfinniand wahrend der letzten Jahrzehnte. — Acta Soc. Fauna Fl. Fenn. 65. Hfors. — — 1948: Ökologische Untersuchungen über Uferinsekten. — Acta Zool. Fenn. 53. Hfors. Krocerus, R. 1924 a: Insekter som befordrare av flygsandsbildning. — Geogr. Sallsk. i Finl. Tidskr. Hfors. — — 1924 b: Skalbaggsfaunan vid Vammeljoki älv (Nykyrka, Ik). — Notulae Ent. 3. Hfors. 787 885 -— — 1925 a: Studien über Lebensweise und Entwicklung einiger Bledius-Arten. — Acta Soc. Fauna Fl. Fenn. 56. Hfors. — — 1925 b: Weitere Beiträge zur Kenntnis der Biologie und Verbreitung einiger Bledius- und Dyschirius-Arten. — Notulae Ent. 5. Hfors. — — 1925 c: In Mem. Soc. Fauna Fl. Fenn. 1 (p. 61). Hfors. — — 1929: Beobachtungen über Bledius- und Dyschirius-Arten auf Degersand (Al Eckerö). — Ibidem. 5. — — 1932: Über die Ökologie und Verbreitung der Arthropoden der Triebsand- gebiete an den Küsten Finnlands. — Acta Zool. Fenn. 12, Hfors. — — 1937: Mikroklima und Artverteilung. — Acta Soc. Fauna Fl. Fenn. 60. Hfors. — — 1939: Zur Ökologie nordischer Moortiere. — VII. Intern. Kongr. Ent. Berlin. 2. Weimar. — — 1947: Nagot om torvmarksdjurens ekologi. — Svensk Faun. Revy. 9. Sthim. KRUMBIEGEL, I. 1932: Untersuchungen über physiologische Rassenbildung. Ein Beitrag zum Problem der Artbildung und der geographischen Variation. — Zool. Jahrb., Abt. f. Syst. 63. Jena. KUuHnett, W. 1933: Kleinklima und Landtierwelt. — Zoogeographica. 1. Jena. — — 1934: Die Bedeutung des Klimas für die Tierwelt. — Bioklim. Beibl. 1. (Meteor. Zeitschr.) Braunschweig. — — 1936: Der Einfluss des Klimas auf den Wasserhaushalt der Tiere. — Ibidem. 3. — — 1943: Uber Beziehungen zwischen Tier- und Pflanzengeseilschaften. — Biol. Generalis. 17. Wien. Kuprrer, K. R. 1925: Grundzüge der Pflanzengeographie des Ostbaltischen Gebietes. — Abh. Herder-Inst. ı. 6. Riga. Kurck, C. 1917: Den forntida utbredningen af kärrsköldpaddan, Emys orbicularis &c. — Lunds Univ. Ärsskr. (N. F.) 2. 13.9. Lund. Kvirte, G. 1941: Om Orthoptera i Sgr-Norge. — Norsk Ent. Tidsskr. 6. Oslo. Kyrın, H. 1923: Vaxtsociologiska randanmärkningar. — Bot. Notiser. Lund. LACKSCHEWITZ, TH. 1942: Zur Koleopterenfauna des Ostbaltischen Gebietes, III. — Korr.-blatt Naturf. Ver. Riga. 64. Posen. LAGERBERG, T. 1937-39: Vilda vaxter i Norden. 1—3. — Sthlm. Lanow, B.-O. 1948: Nagra iakttagelser Over Harpalus (Pseudophonus) griseus Panz. (Col. Carab.) i Skane 1948. — Opusc. Ent. 13. Lund. LANGLET, O. 1935: Till frägan om sambandet mellan temperatur och vaxtgranser. — Medd. Statens Skogsförsöksanst. 28. Sthlm. Larsen, ELtinor Bro 1936: Biologische Studien über die tunnelgrabenden Kafer auf Skallingen. — Vid. Medd. Dansk Naturh. Foren. 100. Kbhyn. Käfer auf Skallingen. — Vid. Medd. Dansk Naturh. Foren. roo. Kbhvn. — — 1943: The importance of master factors for the activity of noctuids. — Ent. Medd. 23. Kbhvn. LarsEn, ©. 1930: Welche Bedeutung hat ein Verlust des Flugvermogens für die lokale Verbreitung der Wasserhemipteren. — Zeitschr. wiss. Insektenbiol. 25. Berlin. — — 1931: Beitrag zur Kenntnis des Pterygopolymorphismus bei den Wasser- Fysiogr. Sällsk. Handl. (N. F.) 42. Lund. hemipteren. 788 — — 1947: Ranatra linearis L. in Schweden. — Opusc. Ent. 12. Lund. Larsson, S. G. 1939: Entwicklungstypen und Entwicklungszeiten der danischen Carabiden. — Ent. Medd. 20. Kbhvn. — — 1940: Sandspringernes Forekomsttider i Danmark (Cicindelidae). — Ibidem. Leece, O. 1913: Der Memmert. Eine entstehende Insel und ihre Besiedelung durch Pflanzenwuchs. — Abh. Naturw. Ver. 21. Bremen. LENGERKEN, H. von 1921: Carabus auratus L. und seine Larve. — Arch. f. Naturgesch. A. 87. 3. Berlin. — — 1929 a: Die Salzkafer der Nord- und Ostseeküste &c. — Leipzig. — — 1929 b: Halophile und halobionte Coleopteren. — Tierwelt Nord- und Ostsee. 16 (XI, e 1). Leipzig. LETZNER, K. 1847-52: Systematische Beschreibung der Laufkafer Schlesiens. — Zeitschr. f. Ent. Breslau. Lewis, R. G. 1935: The orography of the North Sea bed. — Geogr. Journ. 86. London. Lip, J. & ZacHau, A. R. 1929: Utbredningen av Viscaria alpina (L.) G. Don, Alchemilla alpina L. och Rhodiola rosea L. i Skandinavien. — Medd. Gbgs Bot. Tradg 4 (1928). Göteborg. LinpBERG, HARALD 1916: Hvilka vittnesbörd lämnar fytopaleontologin om värt lands och dess floras utvecklingshistoria sedan istiden samt rörande tiden for manniskans första upptradande i landet? — Öfvers. Finska Vet.-Soc. Forhandl. 58 (C). Hfors. — — 1937: Nykomlingar till Finlands skalbaggsfauna. — Notulae Ent. 17. Hfors. — — 1938: De till Phyllodecta s. str. (Col., Chrysomelidae) horande arternas utbredning i Fennoscandia orientalis. — Notulae Ent. 18. Hfors. — — 1942: Ibidem. 21 (1941, p. 144). — — 1943: Markliga vaxt- och skalbaggsfynd belysande Finlands utvecklings- historia. — Mem. Soc. Fauna Fl. Fenn. 18. Hfors. LinpBERG, HAKAN 1924: Bidrag till kannedomen om äländska ögruppens skal- baggsfauna. — Medd. Soc. Fauna FI. Fenn. 48. Hfors. — — 1925: Insektekologiska iakttagelser pa Aland. — Ibidem. 49. — — 1929: Vingdimorfismen hos Hemiptera Heteroptera. — Beretn. 18. Skand. Naturforskermgde. Kbhvn. — — 1931: Die Küstenkäfer Finnlands &c. — Mem. Soc. Fauna Fl. Fenn. 7. Hfors. — — 1933: Untersuchungen in N.-Petsamo tiber die Kaferfauna hochnordischer Biotopen. — Mem. Soc. Fauna Fl. Fenn. 9. Hfors. — — 1948: Zur Kenntnis der Insektenfauna im Brackwasser des Baltischen Meeres. — Comment. Biol. X. 9. Hfors. Linvguist, B. 1931: Den skandinaviska bokskogens biologi. — Diss. — Svenska Skogsvardsforen. Tidskr. Sthlm. — — 1948: The main varieties of Picea abies (L.) Karst. in Europe &c. — Acta Horti Berg. 14. Uppsala. Lınpovist, S. 1935: Svenskarna i heden tid. — ”Orientering i aktuella ämnen” (Bonniers). Sthlm. LinprotH, C. H. 1931: Die Insektenfauna Islands und ihre Probleme. — Zool. Bidr. 13. Uppsala. 887 789 1932: Über die Geschichte der färöischen und isländischen Fauna. — Arch. f. Naturgesch. 1. Leipzig. 1933: Die Gattung Tropiphorus Schönh. (Col. Curculionidae) in Nord- europa und ihre Verbreitung. — Zoogeographica. ı. Jena. 1935 a: The Boreo-British Coleoptera. A study of the faunistical connec- tions between the British Isles and Scandinavia. — Zoogeographica. 2. Jena. 1935 b: Die Coleopterenfauna am See Pjeskejaure im Schwedischen Lapp- land. — Ark. f. Zool. 28 A. Sthlm. 1938: Vita fläckar pa Sveriges skalbaggskarta, -— Entomologbl. 2. (Beil. zu Ent. Tidskr. 59.) Sthlm. 1939 a: Die skandinavische Käferfauna als Ergebnis der letzten Vereisung. — VII. Intern. Kongr. Ent. Berlin. 1. Weimar. 1939 b: Zur Systematik fennoskandischer Carabiden. 2—3. — Ent. Tidskr. 60. Sthlm. 1941: Endemismus in der fennoskandischen Fauna, erlautert an der Fam. Carabidae. — Zool. Bidr. 20. Uppsala. 1942: Ett subfossilfynd af Calosoma sycophanta L. — Pop. Biol. Revy. Sthlm. 1943 a: Oodes gracilis Villa. Eine thermophile Carabide Schwedens. — Notulae Ent. 22 (1942). Hfors. 1943 b: Zur Systematik fennoskandischer Carabiden. 13—33. — Ent. Tidskr. 64. Sthlm. 1945 a: Die fennoskandischen Carabidae. I. — K. Vet. Vitt. Samh. Handl. (B) 4. 1. Goteborg. 1945 b: Die fennoskandischen Carabidae. II. — Ibidem. 4. 2. 1946: Inheritance of wing dimorphism in Pterostichus anthracinus Ill. — Hereditas. 32. Lund. 1948 a: Interglacial insect remains from Sweden. — Sver. Geol. Unders. C. 492 (Arsbok 42). Sthlm. 1948 b: Notes on the ecology of Laboulbeniaceae infesting Carabid beetles, — Sv. Bot. Tidskr. 42. Sthlm. 1948 c: Halltorpstrakten pa Oland. Ett entomologiskt eldorado. — Sv. Faun. Revy. 10. Sthlm. 1948 d: Influence of the sun on the flight direction of some Coleoptera. — Ent. Tidskr. 69. Sthlm. 1948 e: Insektresterna i Längselefyndet. — In Sunpıus & SANDEGREN: Inter- glacialfyndet vid Längsele. — Sver. Geol. Unders. C. 495 (Ärsbok 42). Sthlm. 1948 f: Vingdimorfismen inom familjen Carabidae. — Arch. Soc. Zool.-Bot. Fenn. Vanamo. 1 (1946). Hfors. LinprotH, C. H. & Patm, TH. 1934: Bidrag till kannedomen om coleopterfaunan i Ovre Norrlands kustland. — K. Vet. Vitt. Samh. Handl. (B). 4. 3. Goteborg. Liystey, E. G. 1943: Attraction of Melanophila beetles by fire and smoke. — Journ. Econ. Ent. 36. Menasha, Wisc. Lyuncu, S. I, 1823: Nya’ Insecter, utur egen Samling. — Vet. Ak. Nya Hand. Sthlm. 790 LoHMANDER, H. 1946: Vad landmolluskerna be-atta om faunans invandring till Gotland. — Natur pa Gotland. (Svensk Natur.) Goteborg. LomnickI, A. M. 1894: Fauna pleistocenica insectorum boryslaviensium. (Polnisch.) — Muz. Im. Dzieduczyckich. 4. Lwow. — — 1914, s. SCHILLE. LONNBERG, E. 1924: Ett bidrag till den svenska faunans invandringshistoria, — Fauna och Flora. Uppsala. LuNDBEcK, W. 1927: Diptera Danica. 7. — Kbhvn. Lunpsvap, OÖ. 1921: Mezira tremulae (Germar) och dess forekomst i Sverige. — Fauna och Flora. Uppsala. LUNDEGÄRDH, H. 1930: Klima und Boden in ihrer Wirkung auf das Pflanzenleben. 2. Aufl. — Jena. Lunpevist, G. 1925: Utvecklingshistoriska insjöstudier i Sydsverige. — Sver Geol. Unders. Ärsbok 18 (1924). Sthim. — — 1928: Studier i Ölands myrmarker. — Ibidem. 22. — — 1942: Jordarterna och deras fördelning inom Norrland och Dalarna. — Norrland. Natur, befolkning och naringar. (Geogr. Forb.) Sthlm. — — 1943: Norrlands jordarter. — Sver. Geol. Unders. Arsbok 37. Sthlm. — — 1946: Kvartargeologien i Sverige under aren 19211945. — Geol. Foren. Forh. 68. Sthlm. Lunerunp, H. 1942 a: Uber die Sonnenscheindauer in Finnland. — Soc. Sci. Fenn.. Comm. Phys.-Math. 11. Hfors. — — 1942 b: Uber Klima und Vegetation in Finnland. — Ibidem. Lutz, F. E. 1927: Wind and the direction of insect flight. — Amer. Mus. Nov. 291. New York. Lynce, B. 1934: Some general results of recent Norwegian research work on Arctic Lichens. — Rhodora. 36. Lancaster, Pa. LysHoLm, B. 1937: Coleopterfaunaen i Trgndelag. — Norsk Ent. Tidsskr. 4. Oslo. Mac Gittavry, D. 1914: De entomologische fauna van het eiland Terschelling voor zoover zij tot nu toe bekend is. — Tijdschr. v. Ent. 57. Haag. Maanusson, N. H. & GRANLUND, E. 1936: Sveriges Geologi. — Sthlm. MALAISE, R. 1945: Tenthredinoidea of South-Eastern Asia. — Opusc. Ent., Suppl. 4. Lund. MALMSTRÖM, C. 1920: Trapa natans L. i Sverige. — Svensk Bot. Tidskr. 14. Sthlm. — — 1939: Hallands skogar under de senaste 300 ären. — Medd. Statens Skogs- forsoksanst. 31. Sthlm. Mänadsöversikt av vaderleken i Finland. jo. 1946. — Meteor. Centralanst. Hfors. Maran, J. 1927: The study of the rudiments of the wings by the genus Pte- rostichus &c. (Tscheckisch.) — Sbornik Ent. Odd. Nar. Mus. 5. Praha. Mast, S. O. 1911: Light and the behaviour of organisms. — New York. Mc CrurE, H. E. 1943: Further notes on aero-plankton of Kentucky. — Ent. News. 54. Philadelphia. MEIXNER, J. 1934: Coleopteroidea. — KÜKENTHAL: Handb. d. Zool. 4. 2. Berlin & Leipzig. MERIKALLIO, E. 1946: Uber regionale Verbreitung und Anzahl der Landvogel in Süd- und Mittelfinnland, besonders in deren östlichen Teilen, im Lichte von 791 889 quantitativen Untersuchungen. 1. — Ann. Zool. Soc. Zool.-Bot. Vanamo. 12. Hfors. Meteorol. Jahrb. fur Finnland. 1920—37. — Hfors. METZGER, A. A. TH. 1927: Zur Kenntnis des Nordbaltischen Kambro-Silurs auf Aland und im südwestlichen Küstengebiet Finnlands. — Fennia. 47. 12. Hfors. Mevius, W. 1921: Beiträge zur Physiologie „kalkfeindlicher” Gewächse. — Jahrb. wiss. Bot. 60. Leipzig. — — 1924: H-ion-Konzentration und Permeablitat bei ,,kalkfeindlicher” Ge- wachse. — Zeitschr. Bot. 16. Jena. MEYER, A. 1926: Ueber einige Zusammenhänge zwischen Klima und Boden in Europa. — Jena. MJÖBERG, E. 1905: Biologiska och morfologiska studier Ofer Färöns insektfauna, grundade pa tvenne mänaders vistelse pa ön. — Ark. f. zool. 2. Uppsala. — — 1912: Om en syd- och mellaneuropeisk relikt insektfauna pa Gottland och Oland jamte en del allmännare insektgeografiska spörsmäl. — Ent. Tidskr. 33. Uppsala. Moun, H. 1895: Klimat-tabeller for Norge. I. Lufters temperatur. — Vid.-selsk. Skr., I. Math.-naturv. Kl. 10. Oslo. Movius, H. L. 1942: The Irish Stone Age. Its chronology, development & rela- tionships. — Cambridge. MUESEBECK, C. F. W. 1931: The genera of parasitic wasps of the Braconid sub- family Euphorinae &c. — Misc. Publ. U. S. Dept. Agric. 241. Washington. Munster, T. 1922: Tilleg til Norges Koleopterfauna. — Norsk Ent. Tidsskr. 1. Oslo. — — 1927: To bidrag til Norges koleopterfauna. — Nyt Mag. f. Naturv. 65. Oslo. — — 1933: Arpedium (Col. Staphylinidae). A revision of the norwegian species. — Norsk Ent. Tidsski. 3. Oslo. Muntue, H. 1910: Studier ofver Gotlands senkvartara historia. — Sver. Geol. Unders. Ca. 4. Sthlm. — — 1911: Hufvuddragen af Gotlands fysiskt-geografiska utveckling efter istiden. — Ymer. 31. Sthlm. — — 1940: Om Nordens, framst Baltikums, senkvartära utveckling och sten- äldersbebyggelse. — K. Vet. Ak. Handl. (3) 19. Sthlm. MUNTHE, H., Hens, J. E. & Lunngvist, G. 1936: Beskrivning till kartbladet Faro. — Sver. Geol. Unders. Aa. 180. Sthlm. MUNTHE, H. & LOHMANDER, H. 1946: Ur Gotlands geologiska utvecklingshistoria. — Natur pa Gotl. (Sveriges Natur). Goteborg. NANNFELDT, J. A. 1935: Taxonomical and plant-geographical studies in the Poa laxa group. — Symb. Bot. Ups. 1. Uppsala. — — 1940: On the polymorphy of Poa arctica R. Br., with special reference to its scandinavian forms. — Symb. Bot. 4. Uppsala. — — 1947: Nagra synpunkter pa den skandinaviska fjallflorans alder. — Kgl. Vet. Soc. Arsbok. Uppsala. Natuorst, A. G. 1871: Om nägra arktiska vaxtlamningar i en sotvattenslera vid Alnarp i Skane. — Lunds Univ. Ars-skrift (Acta Univ. Lund.). 7. (1870). Lund. — — 1894: Die Entdeckung einer fossilen Glacialflora in Sachsen &c. — Ofvers. K. Vet. Ak. Förh. 51. Sthlm. 890 792 — — 1895: Ett par glaciala ”pseudorelikter”. — Bot. Notiser. Lund. NAUMANN, E. 1932: Grundzüge der regionalen Limnologie. — THIENEMANN: Die Binnengewasser. 11. Stuttgart. Ned bgriakttagelser i Norge. 42. 1938. — Norsk Meteor. Inst. Oslo (1939). NEHRING, A. 1890: Ueber Tundren und Steppen der Jetzt- und Vorzeit, mit be- sonderer Berücksichtigung ihrer Fauna. — Berlin. NETOLITZKY, F. 1916: Die Verbreitung von Ocys harpaloides Serv. — Ent. Blätter. 12. Krefeld. — — 1929: Zoogeographische Ueberraschungen in der Carabidengruppe Bembi- diini. — Koleopt. Rundschau. 15. Wien. NETOLITZKY, F. & MEYER, P. 1933: Die Verbreitung des Bembidion argenteolum Ahr, — Ent. Blätter. 29. Krefeld. NETOLITZKY, F. & SAINTE-CLAIRE DEVILLE, J. 1914: In Ent. Blätter. 10. Berlin. NiIELSEN, J. C. 1909: lakttagelser over entoparasitiske Muscide-larver hos Arthropoder. — Ent. Medd. (2) 4. Kbhvn. — — 1916: Om Gymnopeza-Arternes Biologi. — Vid. Medd. Dansk Naturh. Foren. 67. Kbhvn. — — 1918: Tachin-studier. — Ibidem. 60. NIELSEN, Y. 1897: Valamo i Ladoga. — Norsk Geogr. Selsk. Aarbog. 8. Oslo. Nırsson, Erik 1947: Den nutida klimatforandringen i Ostafrika. — Ymer. 67. Sthlm. NıLsson, SvEn 1847: Skandinavisk Fauna. I. Däggdjuren. 2. Aufl. — Lund. Nırsson, T. 1935: Die pollenanalytische Zonengliederung der spät- und post- glazialen Bildungen Schonens. -— Geol. Foren. Förh. 57. Sthlm. NorpBerc, S. 1936: Biologisch-ökologische Untersuchungen über die Vogelnidi- colen. — Acta Zool. Fenn. 21. Hfors. NorDHAGEN, R. 1931: Studien über die skandinavischen Rassen des Papaver radicatum Rottb. &c. — Bergens Mus. Ärbok. Bergen. ~~ — 1933: De senkvartzre klimavekslinger i Nordeuropa og deres betydning for kulturforskningen. — Inst. Sammenlign. Kulturforskn. A. 12. Oslo. — — 1035: Om Arenaria humifusa Wg. og dens betydning for utforskningen av Skandinavias eldste floraelement. — Bergens Mus. Ärbok. Bergen. — — 1936: Skandinavias fjellflora og dens relasjoner til den siste istid. — Nord. (19. Skand.) Naturf.-motet. Hfors. Norpuinc, E. 1904: En resa i ornitologiskt syfte till nagra utöar i Finska viken. — Medd. Soc. Fauna Fl. Fenn. 30. Hfors. Norpman, A. F. 1943: Till kannedomen om fjarilfaunan i ett lundomrade 1 det centrala Skärgärdshavet i SW-Finland (Foglo, Bano). — Mem. Soc. Fauna Fl. Fenn, 18. Hfors. — — 1946: In Notulae Ent. 26 (p. 111). — Hfors. NORDSTRÖM, F. 1945: Phytometra confusa Steph. i Sverige. — Ent. Tidskr. 64 Sthlm. — — 1946: Nägot om Gotlands storfjärilar. — Natur pa Gotl. (Svensk Natur). Goteborg. Nortnun, N. 1946: Gotlands högre ryggradsdjur, en översikt. — Natur pa Gotl. (Svensk Natur). Goteborg. 793 891 Norske Meteorol. Inst, Jahrbuch. 1890-99, 1930—39, 1945. — Oslo. Norinı, G. 1943: Den gra kräkan (Corvus cornix L.). Naringsekologiska studier. — Sv. Jagarforb. Medd. 7. Uppsala. — — 1948: Biologiska undersokningar över grävlingen (Meles meles). — Ibidem. 13. OERTEL, R. 1924: Studien über Rudimentation, ausgeführt an den Flügelrudimenten der Gattung Carabus. — Zeitschr. Morph. Ökol. d. Tiere. ı. Berlin. OKLAND, FR. 1925: Die Verbreitung der Landgastropoden Norwegens. — Skrifter Norske Vid.-Ak. (I. Mat.-Naturv. Klasse) 8. Oslo. OLSEN, C. 1921: Studier over Jordbundens Brintionkoncentration og dens Betyd- ning for Plantefordelingen i Naturen. — Medd. Carlsberg Labor. 15. 1. Kbhvn. Otsont, B. 1927: Vaxtvarlden pa Tytärsaarı och Sayvi. — Mem. Soc. Fauna FI. Fenn. 2. Hfors. OSSIANNILSSON, F. 1937: Zur Kenntnis der schwedischen Homopterenfauna &c. — Opusc. Ent. 2. Lund. — — 1946—47: Stritar, Homoptera Auchenorrhyncha. — Svensk Insektfauna. 36. 37. Sthlm. Ostman, C. J. 1933: Vinden i Sveriges hogre luftlager. — Medd. Statens Meteor.- Hydrogr. Anst. 6. 3. Sthlm. PAASIVIRTA, PAULA 1936: Kulturgeographische Skizze der Insei Hailuoto. (Finnisch.) — Terra. 48. Hfors. PAGENSTECHER, A. 1909: Die geographische Verbreitung der Schmetterlinge. — Jena. Patm, TH. 1935: Coleopterologiska studier pa ön Ven. — K. Vet. Vitt. Samh. Handl. (B) 4. 9. Göteborg. — — 1942: Coleopterfaunan vid Nedre Dalalven. — Ent. Tidskr. 63. Sthlm. — — 1945: Vattnet som insektspridande faktor. — Ibidem. 66. — — 1947 a: För Sverige nya Coleoptera. IX. — Ibidem. 68. -— — 1947 b: Anteckningar om svenska skalbaggar. III. — Ibidem. 68. PALm, TH. & LinprotH, C. H. 1936: Coleopterfaunan vid Klarälven. I. Allman del. — Ark. f. zool. 28 A, 19. Sthlm. — — & — — 1037: Coleopterfaunan vid Klaralven. II. Speciell del. — Ent. Tidskr. 58. Sthlm. PAaLm£n, ERNST 1944: Die anemohydrochore Ausbreitung der Insekten als zoo- geographischer Faktor &c. — Ann. Zool. Soc. Zool.-Bot. Fenn. Vanamo. 10. Hfors. — — 1945 a: Über Quartierwechsel und submerse Überwinterung einiger ter- restrischen Uferkäfer. — Ann. Ent. Fenn. 11. Hfors. — — 1945 b: Zur Systematik finnischer Chrysomeliden. ı. Gattung Galerucella Crotch. — Ann. Ent. Fenn. 11. Hfors. — — 1945 c: Ibidem. (p. 174). — — 1946: Materialien zur Kenntnis der Kaferfauna im westlichen Swir-Gebiet (Sowjet-Karelien). — Acta Soc. Fauna Fl. Fenn. 65. Hfors. Paım£n, E. & PLATonorr, S. 1943: Zur Autökologie und Verbreitung der Ost- fennoskandischen Flussuferkäfer. — Ann. Ent. Fenn. 9. Hfors. Patmén, E. & SUOMALAINEN, H. 1945: Experimentelle Untersuchungen über die 892 794 Transpiration bei einigen Arthropoden, insbesondere Käfern. — Ann. Zool. Soc. Zool.-Bot. Fenn. Vanamo. 11. 2. Hfors. PALMGREN, A. 1915— 17: Studier öfver lofangsomradena pa Aland. I—III. — Acta Soc. Fauna Fl. Fenn. 42. Hfors. — — 1921: Die Entfernung als pflanzengeographischer Faktor. — Ibidem. 49. — — 1925: Die Artenzahl als pflanzengeographischer Charakter &c. — Fennia. 46. Hfors. — — 1927: Die Einwanderungswege der Flora nach den Alandinseln. — Acta Bot. Fenn. 2. Hfors. — — 1943—44: Naturskyddet pa Aland och dess framtida utveckling. — Acta Soc. Fauna Fl. Fenn. 63. Hfors. PALMGREN, P. 1930: Quantitative Untersuchungen über die Vogelfauna ın den Wäldern Südfinnlands mit besonderer Berücksickiigung Alands. — Acta Zool. Fenn. 7. Hfors. — — 1939, 1943: Die Spinnenfauna Finnlands. 1. 2. — Ibidem. 25. 36. PayKULL, G. de 1790: Monographia Caraborum. — Uppsala. — — 1798: Fauna Svecica. I. — Uppsala. PENcK, A. 1936: Das Klima der Eiszeit. — Verh. Internat. Quartärkonferenz. 3. Wien. (Zitiert nach A. WAGNER 1940.) Penck, A. & BRUCKNER, E. 190I—09: Die Alpen im Eiszeitalter. I-III. — Leipzig. PESCHEL, O. 1875: Neue Probleme der vergleichenden Erdkunde. 14. Die Ent- wicklungsgeschichte der stehenden Wasser auf der Erde. — Das Ausland. Ueberschau der neuesten Forschungen auf dem Gebiete der Natur-, Erd- und Völkerkunde. 48. ıı (15. 3. 1875). Stuttgart. PEsoLa, V. A. 1928: Calcium carbonate as a factor in the distribution of plants in Finland. (Finnisch.) — Ann. Soc. Zool.-Bot. Fenn. Vanamo. 9. 1. Hfors. PETERSEN, B. 1947: Die geographische Variation einiger fennoskandischer Le- pidopteren. — Zool. Bidr. 26. Uppsala. PETTERSSON, BENGT 1946: In Natur pa Gotland. (Svensk Natur.) Göteborg. PETTERSSON, BROR 1940: Experimentelle Untersuchungen über die euanemochore Verbreitung der Sporenpflanzen. — Acta Bot. Fenn. 25. Hfors. PEYERIMHOFF, P. DE 1947: Les Coléoptéres de l’Europe boréale d’apres le Catalogus Coleopterorum Daniae et Fennoscandiae. — Revue Franc. d’Ent. 13. Paris. PLATONOFF, S. 1943: Zur Kenntnis der Kaferfauna um den See Paanajarvi in Kuusamo, Nordfinnland. — Notulae Ent. 23. Hfors. Poisson, R. 1924: Contribution a l’etude des Hémiptéres Aquatiques. — Bull. Biol. France et Belg. 58. Paris. — — 1925: Sur la persistance des muscles vibrateurs du vol chez un Hémiptére heteroptere brachyptére: Chorosoma schillingi Schum. — C. R. Soc. Biol. 92. Paris. Porrıus, B. 1905: Kola-Halfons och Enare Lappmarks Coleoptera. —~Festschr. f. Palmén. Hfors. — — 1909 a: Die Coleopteren-Fauna der Halbinsel Kanin. — Acta Soc. Fauna Fl. Fenn. 31. Hfors. — — 1909 b: Uber die Verbreitung einiger an den Meeres-Küsten lebenden Cole- opteren und Hemipteren in Fennoscandia. — Medd. Soc. Fauna Fl. Fenn. 35. Hfors. 795 893 — — 1911: Beiträge zur postglazialen Einwanderung der Käfer-Fauna Finn- lands. — Acta Soc. Fauna Fl. Fenn. 34. Hfors. Post, L. von 1903: En profil genom högsta Litorinavallen pa sodra Gotland. — Geol. Foren. Forh. 25. Sthlm. — — 1906: Norrländska torfmossestudier. 1. — Ibidem. 28. — — 1909: Stratigraphische Studien über einige Torfmoore in Narke. — Ibidem. 31. — — 1924: Ur de sydsvenska skogarnas regionala historia under postarktisk tid. — Ibidem. 46. — — 1933: Den svenska skogen efter istiden. — Verdandis Smaskrifter. 357. Sthlm. PRAEGER, R. L. 1932: Recent views bearing on the problem of the Irish flora and fauna. — Proc. R. Irish Acad. 41. Dublin. PRATJE, O. 1923: Helgoland. — Sammlung geologischer Führer (E. KRENKEL). 23. — Berlin. ProcHnow, O. 1921—24: Mechanik des Insektenfluges. — SCHRÖDER: Handb. d. Ent. 1. Jena. Ramsay, W. 1896: Nagra upplysningar om Hogland. — Geogr. Foren. Tidskr. 8. Hfors. — — 1930: Changes of sea-level resulting from the increase and decrease of glaciations. — Fennia. 52. Hfors. Rapp, ©. 1933: Die Kafer Thüringens unter besonderer Berücksichtigung der faunistisch-oekologischen Geographie. 1. — Erfurt. Rave, H. M. 1941: Botanical problems in boreal America. — Bot. Review. 7. Lan- caster, Pa. RECLAIRE, A. 1926: Korte -mededeeling omtrent eenige op Terschelling waar- genomen kevers, wantsen en mieren. — Ent. Berichten. 7. Haag. — — 1930: Korte mededeeling omtrent eenige op Vlieland waargenomen insecten. — Ibidem. 8. Régles Internationales de la Nomenclature Zoologique adoptées par les Congrés Internationaux de Zoologie. — Paris 1905. Reınıc, W. F. 1937: Die Holarktis. Ein Beitrag zur diluvialen und alluvialen Geschichte der zirkumpolaren Faunen- und Florengebiete. — Jena. — — 1938: Elimination und Selektion. — Jena. RENKONEN, O. 1938: Statistisch-okologische Untersuchungen über die terrestrische Kaferwelt der finnischen Bruchmoore. — Ann. Zool. Soc. Zool.-Bot. Fenn. Vanamo. 6. Hfors. — — 1944: Die Carabiden- und Staphylinidenbestände eines Secufers in SW-Finn- land. — Ann. Ent. Fenn. 10. Hfors. Renscu, B. 1929: Das Prinzip geographischer Rassenkreise und das Problem der Artbildung. — Berlin. Ricuter, K. 1937: Die Eiszeit in Norddeutschland. — Deutscher Boden. 4. Berlin. Roserts, E. 1918: Fluctuations in a recessive mendelian character and selection. — Journ. Exper. Zool. 27. Philadelphia. Rome Lt, L.-G. 1938: Vaxternas spridningsmöjligheter. — SKoTTSBERG: Vaxternas liv. 4. Sthlm. Rosperc, J. E. 1893: Nagot om Karlö. — Geogr. Foren. Tidskr. 5. Hfors. — — 1898: En landstigning pa Tytarsaari. — Ibidem. 10. 796 894 RosBErc, J. E., GROTENFELT, K, HıLDEn, K. & WeECKSELL, J. A. 1923: Suomenmaa. V. — Hfors. Rostrup, E. 1943: Den Danske Flora. — Kbhvn. RUBINSTEIN, EUGENIE 1924: Beziehungen zwischen dem Klima und dem Pflanzen- reiche. — Meteor. Zeitschr. 41. Braunschweig. — — 1927: Klima der Union der Sozialistischen Sowjet-Republiken. I, 1. — Lenin- grad. RÜSCHKAMP, F. 1927: Der Flugapparat der Kafer. Vorbedingung, Ursache und Verlauf seiner Rückbildung. — Zoologica. 28:75. Stuttgart. Russe, I. C. 1893: Second expedition to Mount Saint Elias in 1891. — 13. Ann. Rep. (1891—92) U. S. Geol. Survey. II (Geology). Washington. SAALAS, U. 1917, 1923 a: Die Fichtenkäfer Finnlands. I. II. — Ann. Ac. Sci. Fenn. (A). 8. 22. Hfors. —-— 1923 b: Studien über die Elateriden Finnlands. I. Corymbites cupreus Fabr. subsp. aeruginosus Fabr. &c. — Ann. Zool. Soc. Zool.-Bot. Fenn. Vanamo. 2. Hiors. — — 1939: Über einige kulturbeeintrachtigte Käferarten sowie einige Neuan- kömmlinge in der Käferfauna Finnlands. — VII. Intern. Kongr. Ent. Berlin. 1. Weimar. SAELAN, TH. 1900: Nagra anteckningar om floran pa Hogland. — Medd. Soc. Fauna Fl. Fenn. 25. Hfors. SAHLBERG, C. R. 1834: Insecta Fennica. I. — Hfors. SAHLBERG, J. 1868: Bidrag till kännedomen om Finlands Dimorpha Insekt-arter. — Notiser Sällsk. Fauna Fl. Fenn. 9 (1867). Hfors. — — 1873: Enumeratio Coleopterorum Carnivorum Fenniae. — Ibidem. 14. SAHLSTRÖM, K. E. 1944—48: Jordartskarta över södra och mellersta Sverige. — Sver. Geol. Unders. Ba. 14. Sthlm. SAINTE-CLAIRE DEVILLE, J. 1924: Kleinere Mitteilungen über die Dyschirius- Arten Europas. — Koleopt. Rundschau. 11. Wien. — — 1930 a: Quelques aspects du peuplement des Iles britanniques. (Cole- opteres). — Peuplem. Iles Brit. — Soc. Biogéogr. III. Paris. — — 1930 b: Als die Nordsee noch trocken war — — — Eine Skizze zur Zoogeo- graphie. — 4. Wandervers. Deutsch. Ent. Kiel. Berlin-Dahlem. SALISBURY, E. J. 1920: The significance of the calcicolous habit. — Journ. Ecol. 8. Cambridge. -— — 1935: Are most of the present British plants post-glacial immigrants from extra-British regions with possibly some human introductions? — Discuss. Brit. Flora. Proc. R. Soc. 118. London. SAMUELSSON, G. 1900: Om de ädla loftradens forna utbredning i Ofre Oster- Dalarne. — Bot. Stud. tillagn. F. R. Kjellman. Uppsala. — — 1915: Über den Rtickgang der Haselgrenze und anderer pflanzengeogra- phischer Grenzlinien in Skandinavien. — Bull. Geol. Inst. 13. Uppsala. — — 1934: Die Verbreitung der höheren Wasserpflanzen in Nordeuropa (Fenno- skandien und Dänemark). — Acta Phytogeogr. Suec. 6. Uppsala. SANDEGREN, R. 1943: Hippophaé rhamnoides L. i Sverige under senkvartär tid. — Svensk Bot. Tidskr. 37. Uppsala. 797 895 -— — 1948: In SUNDIUS & SANDEGREN: Interglacialfyndet vid Längsele. — Sver. Geol. Unders. C 495. (Arsbok 42.) Sthlm. SANDEGREN, R. & ASKLUND, B. 1946: Beskrivning till kartbladet Möklinta. — Sver. Geol. Unders. Aa. 186. Sthlm. SANDEGREN, R. & JOHANSSON, H. E. 1931: Beskrivning till kartbladet Goteborg. — Sver. Geol. Unders. Aa. 173. Sthlm. SANDMAN, J. A. 1892: Fagelfaunan pa Karlo och kringliggande skär. — Medd. Soc. Fauna Fl. Fenn. 17. Hfors. SAURAMO, M. 1942: Kvartargeologiska studier i östra Fennoskandia. — Geol. ‘Foren. Forh. 64. Sthlm. SaxeEn, M. 1928: Über den Einfluss des Gesteinsgrundes auf die Vegetation &c. — Fennia. 50. Hfors. SCHARFF, R. F. 1899: The history of the European fauna. — London. — — 1907: European Animals: their geological history and geographical distribu- tion. — London. SCHAUM, H. 1843: Beitrag zur Kenntniss der norddeutschen Salzkafer. — Zeitschr. f. Ent. (GERMAR). 4. Leipzig. SCHILLE, FR. 1916: Entomologie aus der Mammut- und Rhinoceros-Zeit Gali- ziens. (Übersetz. d. poln. Arbeit „Wykopaliska Starünskie” von A. M. LomMnickI 1914). — Ent. Zeitschr. 30. Frankfurt a. M SCHI@DTE, J. C. 1841: Genera og Species af Danmarks Eleutherata. 1. — Kbhvn. — — 1867, 1872: De Metamorphosi Eleutheratorum Observationes. 3. 6. — Nathurhist. Tidsskr. (3). 4. 8. Kbhvn. — — 1870: Tilleg til Danmarks Karaber og Dytisker. — Ibidem. 8. SCHNEIDER, J. SPARRE 1879: Coleoptera iagttagne ved Tromsg og 1 nermeste omegn. — Tromsg Mus. Aarsh. 2. Tromso. — — 1885: Nordfuglg, en zoologisk skizze. — Tromso Mus, Aarsberetn. (1884). Tromso. — — 1888: Dyrlivet pa vore havskj&r. En zoologisk udflugt til Hillesg juni 1887. — Ibidem (1887). — — 1889: Entomologiske udflugter i Tromso omegn. — Ent. Tidskr. 10 Sthlm. — — 1899: Insektfaunaen paa Kvalgen. — Tromsg Mus. Aarsh. 20. Tromso. — — 1910: Hillesg, et litet supplement. — Ibidem. 31. 32. Tromso. SCHNEIDER, ©. 1898: Die Tierwelt der Nordsee-Insel Borkum unter Berücksichti- gung der von den übrigen ostfriesischen Inseln bekannten Arten. — Abh. Naturw. Ver. 16. Bremen. ScHoTT, G. 1926: Geographie des Atlantischen Ozeans. — 2. Aufl. Hamburg. SCHUBART, O. 1920: Die Coleopterenfauna einer neu entstehenden Nordseeinsel. -- Ent. Mitt. 9. Berlin-Dahlem. ScHuLtz, V. G. M. 1930: Ein Einwand gegen die auf „Relikt”arten beruhende Annahme einer postglazialen Steppenzeit. — Internat. Ent. Zeitschr. 23. Guben. — — 1934 a: Kritisches zur Xerothermrelikt-Hypothese. — Ibidem. 28. — — 1934 b: Nochmals die Xerothermrelikt-Hypothese. — Ibidem. ScHÜTTE, H. 1927: Krustenbewegungen an der deutschen Nordseeküste. — Stutt- gart. (Zitiert nach WARNECKE 1936 b.) 798 896 SEIDLITZ, G. von 1920: In EricHson: Naturgesch. d. Insekten Deutschl. 5. — Berlin. SELLNICK, M. 1939: Milben als Parasiten von Insekten. — VII. Intern. Kongr. Ent. Berlin. 2. Weimar. SEMENOV (-TIAN-SHANSKIJ), A. 1910: Die taxonomischen Grenzen der Art und ihrer Unterabteilungen. — Berlin. (Zitiert nach RENscH 1929 und BRrEU- NING 1932—36.) -— — 1935: Les limites et les subdivisions zoogeographiques de la région pale- arctique pour les animaux terrestres, basées sur la distribution géographique des insectes Coléoptéres. — Trav. Inst. Zool. 2. Leningrad. (Russisch. Mir nur durch die schwed. Zusammenfassung von STOCKMANN 1942a_ bekannt.) SERNANDER, R. 1894 a: Studier öfver den gotländska vegetationens utvecklings- historia. — Diss. — Uppsala. — — 1894 b: Om s. k. glaciala relikter. — Bot. Notiser. Lund. — — 1896: Nagra ord med anledning af Gunnar Andersson, Svenska växt- världens historia. — Ibidem. — — 1901: Den skandinaviska vegetationens spridningsbiologi. — Uppsala. — — 1908: Stipa pennata i Vastergotland. En studie öfver den subboreala periodens inflytande pa den nordiska vegetationens utvecklingshistoria. — Svensk Bot. Tidskr. 2. Sthlm. SHARP, D. 1913: Notes on the wings of Coleoptera. I. The British species of Pterostichus. — The Entomologist. 46. London. SHvoNEN, L. 1942: Zur Phanologie des Frühjahrsauftretens der Hummel (Bom- bus, Hym., Apidae). — Ann. Ent. Fenn. 8. Hfors. SHVoNEN, L. & KALELA, O. 1937: Über die Veränderungen in der Vogelfauna Finnlands während der letzten Jahrzehnte und die darauf einwirkenden Faktoren. — Acta Soc. Fauna Fl. Fenn. 60. Hfors. SKOTTSBERG, C. & VESTERGREN, T. 1901: Zur Kenntnis der Vegetation der Insel Oesel. — Bih. K. Vet. Ak. Handl. 27, III. Sthim. SmitH, H. 1920: Vegetationen och dess utvecklingshistoria i det centralsvenska hogfjallsomradet. (Norrl. Handbibl. 9.) — Uppsala. SOKOLOWSKI, K. 1942: Die Catopiden der Nordmark. (Col., Catopidae). — Ent. Blatter. 38. Krefeld. Sgmme, S. 1937: Zoogeographische Studien über norwegische Odonaten. — (Contr. Biol. Norw. Fish Food Animals. 3.) Avh. Norske Vid. Ak. I. Mat.-Naturv. KI. (1936) 12. Oslo. SOTAVALTA, O. 1947: The flight-tone (wing-stroke-frequency) of insects. — Acta Ent. Fenn. 4. Hfors. SpArck, R. 1929: Den zoologiske Undersggelse af Fergerne og dens forelgbige Resultat. — Beretn. 18. Skand. Naturf.-mgde. Kgbenhavn. — — 1940: Den Danske Dyrerverden dyrgeografisk og invandringshistorisk belyst. — Univers. Festskr. Christian X. Kbhvn. STEBUTT, A. 1930: Lehrbuch der allgemeinen Bodenkunde. — Berlin. STEJNEGER, L. 1907: The origin of the so-called atlantic animals and plants of western Norway. — Smiths. Misc. Coll. 48. Washington. — — 1908: Hvorfra kom Vestnorges eiendommelige dyr og planter? — Naturen. 32. Bergen. ie 799 897 STENSTROM, INGA & RoMELL, L.-G. 1945: Till det sydgotlandska angets minne. — Ymer. Sthlm. STERN, C. 1926: Die Käfer der Umgegend von Hamburg-Altona. Carabidae. — Verh. naturw. Unterhalt. 18. Hamburg. STERNER, R. 1938: Flora der Insel Gland. — Acta Phytogeogr. Suec. 9. Uppsala. — — 1946 a: In Geol. Foren. Förhandl. 68. — Sthim. — — 1946 b: Veronica praecox All. pa Öland. — Bot. Notiser. Lund. Stiny, J. 1929: Technische Gesteinkunde. — 2, Aufl. Wien. Stock, C. 1914: Zur Coleopterenfauna der Nordseeinsel Sylt. — Ent. Blatter. 10. Berlin. : Stockholmstraktens Vaxter, 1937. — Ausgegeben von „Botaniska Sallsk. 1 Sthlm”. 2. Aufl. Sthlm. STOCKMANN, S. 1942 a: Ett hos oss föga kant djurgeografiskt arbete. — Notulae Ent. 21. Hfors. (S. SEMENOV 1935.) — — 1942 b: Ibidem (p. 144). STOLLER, J. 1921: Die Pflanzenwelt des Quartars. In PoroxıE-GoOTHAN: Lehrbuch der Palaobotanik. 2. Aufl. Berlin. STRAND, A. 1935: De fennoskandiske former av Carabus problematicus Hbst. — Norsk Ent. Tidsskr. 4. Oslo. — — 1942: Die Kaferfauna von Svalbard. — Ibidem. 6. Oslo. - — — 1946 a: Nord-Norges Coleoptera. — Tromsg Mus. Arsh. 67 (1944). Tromso. — — 1946 b: Koleopterologiske bidrag IV. — Norsk Ent. Tidsskr. 7. Oslo. STURTEVANT, A. H. 1940: Genetic data on Drosophila affinis, with a. discussion of the relationships in the subgenus Sophophora. — Genetics. 25. Brooklyn, New York. Suenson, E. 1942: Byggematerialer, III. Natursten. 3. Aufl. — Kbhvn. SUENSON, EıGIn 1934: Om en Indsamlings Effektivitet. — Ent. Medd. 18. Kbhvn. SUKATSCHEW, W. 1928: Einige experimentelle Untersuchungen über den Kampf ums Dasein zwischen Biotypen derselben Art. — Zeitschr. Indukt. Abst.- Lehre u. Vererb. 47. Leipzig. SUNDELIN, U. 1917: Fornsjöstudier inom Stängäns och Svartäns vattenomräden. — Sver. Geol. Unders. Ca. 16. Sthlm. SWEDBERG, S. 1931: Geografi for det differentierade gymnasiet. 1. — Sthlm. Sveriges Meteorol. & Hydrol. Institut, Ärsbok, 1919—1943. — Sthlm. SZELIGA-MIERZEYEWSKT, W. vON 1942: Verzeichnis der Käfer Osels in meiner Insektensammlung. — Korr.-blatt Naturf. Ver. Riga. 64. Posen. Tamm, O. 1921: Om berggrundens inverkan pa skogsmarken. — Medd. Statens Skogsforsoksanst. 18. 3.. Sthlm. TANNER, V. 1930: Studier over kvartarsystemet i Fennoskandias nordliga delar. 1V. — Bull. Comm. Geol. Finl. 88. Hfors. — — 1937: Nagra ord i fragan om den sista landisens utbredningsgrans inom Fennoskandias nordligaste delar. — Geol. Foren. Förh. 59. Sthim. — — 1944: Outlines of the geography, life and customs of Newfoundland-Labra- dor. — Acta Geogr. 8. Hfors. TANSLEY, A. G. 1946: Introduction to plant ecology. A guide for beginners in the study of plant communities. — London. 898 800 Tayror, W. P. 1935: Significance of the biotic community in ecological studies. — Quat. Review Biol. 10. Baltimore. TENGWALL, T. A. 1913: De sydliga skandinaviska fjallvaxterna och deras in- vandringshistoria. — Sv. Bot. Tidskr. 7. Sthlm. THIENEMANN, A. 1939: Grundzüge einer allgemeinen Ökologie. — Arch. f. Hydrobiol. 25. Stuttgart. — — 1941: Leben und Umwelt. — Bios. 12. Leipzig. TuHomasson, H. 1935: Äldre baltiska skeden. — Geol. Foren. Förh. 57. Sthim. THomsEN, E. & M. 1937: Über das Thermopräferendum der Larven einiger Fliegenarten. — Zeitschr. vergl. Physiol. 24. Berlin. Thomson, C. G. 1857: Skandinaviens Coleoptera. I (unvollendet). — Lund. — — 1859: Skandinaviens Coleoptera. I. — Lund. — — 1885: Skandinaviens insecter, en handbok i entomologi till allmänna läro- verkens tjenst. 1. Coleoptera. — Lund. THORKELSSON, TH. 1935: A fossiliferous interglacial layer at Ellidaärvogur, Reykja- vik. — Greinar. I. Reykjavik. THorsLunv, P. 1936: Siljansomrädets brännkalkstenar och kalkindustri. — Sver. Geol. Unders. C 398. Sthim. THUNMARK, S. 1937: Uber die regionale Limnologie von Südschweden. — Sver. Geol. Unders. C 410. Sthlm. THURMANN, J. 1849: Essai de phytostatique appliqué a la chaine du Jura &c. I. — Bern. TIMOFEEFF-RESSovsKY, N. W. 1940: Mutations and geographical variation. — In Huxtey: The new systematics. — Oxford. TJEDER, B. 1936: Studies on Psectra diptera Burm. (Neuroptera, Hemerobiidae). — Notulae Ent. 16. Hfors. Trap, J. P. 1921: Kongeriget Danmark. 3. — 4. Aufl. (H. WEITEMEYER). — Kbhvn. Turesson, G. 1927: Contributions to the genecology of glacial relics. — Here- ditas. 9. Lund. Tuxen, S. L. 1944: The hot springs of Iceland: Their animal communities and their zoogeographical significance. — Zool. of Iceland. 1. Kbhvn. UnpAs, J. 1934: Fra istiden i Trgndelag. I. — Kgl. Norske Vid. Selsk. Skr. Trondheim. — — 1939: Kvarterstudier i Vestfinnmark og Vesteralen. — Norsk Geol. Tidsskr. 18. Oslo. Uvarov, B. P. 1921: A revision of the genus Locusta, L. (Pachytulus, Fieb.), with a new theorie as to the periodicity and migration of locusts. — Bull. Ent. Research. 12. London. — — 1931: Insects and climate. — Trans. Ent. Soc. 79. London. WAGNER, A. 1940: Klimaänderungen und Klimaschwankungen. — Die Wissenschaft. 92. Braunschweig. ; WAGNER, H. 1938 (u. f.): Monographie der polaarktischen Ceuthorrhynchinae (Cureul.). — Ent. Blätter. 34 (u. f.). Krefeld. WAHLGREN, E. 1909: Glaciala relikter och pseudorelikter bland vara dagfjärilar. — Fauna och Flora. 4. Uppsala. — — 1910: Xeroterma relikter i Olands alvarfauna. — Ibidem. — — 1913: Sveriges insektgeografiska indelning pa grundval av makrolepidopterer- nas utbredning, — Ent. Tidskr. 34. Uppsala. 801 899 — — 1915, 1917: Det öländska alvarets djurvarld. I. II. — Ark. f. zool. 9. 11. Sthlm. — — 1919: Vastarktiska element i Skandinaviens fjarilfauna. — Ent. Tidskr. 40 Uppsala. — — 1920: De europeiska polaröarnas insektfauna, dess sammansättning och harkomst. — Ibidem. 41. — — 1921: Brunbandade malmatarens (Eupithecia sinuosaria Ev.) nordiska ut- bredning. — Ibidem. 42. — — 1935—41: De svenska fjärilarnas utbredning och harkomst. — In Norp- STROM-WAHLGREN-TULLGREN: Svenska fjarilar. — Sthlm. — — 1948: Eriophyidcecidier av biogeografiskt intresse. — Ent. Tidskr. 69. Sthlm. VALIKANGAS, I. 1930: Der Gartenschlafer, Eliomys quercinus (L.), auf der Insel Tytarsaari im Finnischen Meerbusen. — Mem. Soc. Fauna Fl. Fenn. 6. Hfors. ' — — 1936: Über die Bedeutung der existenzökologischen und verbreitungs- ökologischen Faktoren fiir die Zusammensetzung der Vertebratenfauna auf zwei Inseln mitten im Finnischen Meerbusen. — Nord. (19. Skand.) Naturf.-motet. Hfors. VALLE, K. J. 1933: Die Lepidopterenfauna des Petsamogebietes. — Ann. Zool. Soc. Zool.-Bot. Fenn. Vanamo. 1. Hfors. — — 1946: Zur Invasion von Libellula depressa L. (Odon., Libellulidae) nach Finnland. — Ann. Ent. Fenn. 12. Hfors. Wa ten, A. 1930: Climate of Sweden. — Statens Meteor.-Hydrogr. Anst. 279. Sthlm. VALOVIRTA, E. J. 1937: Untersuchungen über die sakulare Landhebung als pflanzengeographischer Faktor. — Acta Bot. Fenn. 20. Hfors. WANGERIN, W. 1924: Beiträge zur Frage der pflanzengeographischen Relikte, unter besonderer Berücksichtigung des nordostdeutschen Flachlandes. — Abh. Naturf. Gesellsch. 1. Danzig. (Mitt. Mus. f. Naturgesch. u. Vorkunde.) WARBURG, ELSA 1910: On relics in the Swedish Flora. — Bull. Geol. Inst. 9 (1908—og). Uppsala. WARMING, E. & GRAEBNER, P. 1918: Eug. Warmings Lehrbuch der ökologischen Pflanzengeographie. 3. Aufl. — Berlin. WARNECKE, G. 1929: Über Einwanderung und Grenzvorkommen von Gross- schmetterlingen in Ostholstein. — Schr. Naturw. Ver. Schlesw.-Holst. 19. Kiel. — — 1931: Einige kritische Bemerkungen über die Frage der Verwendbarkeit meteorologischer Klimamessungen für zoogeographische Untersuchungen. — Intern. Ent. Zeitschr. 25. Guben. — — 1934: Grundsätzliches zur Methodik zoogeographischer Untersuchungen in der Entomologie. — Ibidem. 28. — — 1936 a: Über die Konstanz der ökologischen Valenz einer Tierart als Voraussetzung für zoogeographische Untersuchungen. — Ent. Rundschau. 53. Stuttgart. — — 1936 b: Über lusitanisch-atlantische Elemente in der Lepidopteren-Fauna der deutschen Nordsecküste. — C. R. XII. Congr. Intern. Zool. (1935). Lisboa. 802 900 Vavitov, N. I. 1922: The law of homologous series in variation. — Journ. of Genetics. 12. Cambridge. Veörattan 1925. Manadaryfirlit samio a. vedurstofunni, — Reykjavik. WEGENER, A. 1923: Das Wesen der Baumgrenze. — Meteor. Zeitschr. 40. — Braunschweig. VERHOEFF, C. 1891 a: Ein Beitrag zur eolkonteren Fauna der Insel Norderney. — Ent. Nachrichten. 17. Berlin. — — 1891 b: Einige Bemerkungen über die dsniächen Calathus- Arten — Ibidern. WESENBERG-LuNp, C. 1909: Om Limnologiens betydning for kvartergeologien &c. — Geol. Foren. Forh. 31. Sthlm. — — 1910: Grundzüge der Biologie und Geographie des Süsswasserplanktons, nebst Bemerkungen über Hauptprobleme zukünftiger limnologilscher Forschungen. — Internat. Revue gesamt. Hydrobiol. u. Hydrogr. 3 (Biol. Suppl.-Heft ı). Leipzig. WEST, A. 1930 a: Coleoptera. — Zool. of the Faroes II. 1, Kbhvn. — — 1930 b: Tilleg og Rettelser til Fortegnelserne over de danske Coleoptera — Ent. Medd. 16. Kbhvn. — — 1940— 41: Fortegnelse over Danmarks Biller. — Ibidem. 21, — — 1947: Tilleg til „Fortegnelse over Danmarks Biller” &c. I. — Ibidem. 25 WuitinG, P. W. 1926: Two wing mutations in Habrobracon and their method of inheritance. — Americ. Naturalist. 60. New York. Witte, N. 1905: Om Invandringen af det arktiske Floraelement til Norge. — Nyt Mag. f. Naturv. 43. Oslo. WıuLıams, C. B. 1939: An analysis of four years captures of insects in a light trap. I. — Trans. Ent. Soc. 89. London. — —.1944: Some. applications of the logarithmic series and the index of diversity to ecological problems. — Journ. Ecol. 32. Cambridge. — — 1947 a: The logarithmic series and its application to biological problems. — Ibidem. 34. — — 1947 b: The generic relations of species in small ecological communities. — Journ. Animal Ecol. 16. Cambridge. WILMOTT, A. J. 1930: Concerning the history of the British flora. — Peuplement . des Iles Britann. — Soc. de Biogéogr, (III). Paris. — — 1935: Evidence in favour of survival of the British flora in glacial times. — Discuss. British Flora. Proc. R. Soc. 118, London. Wiren, E. 1947:. Bidrag till kannedomen om coleopterfaunan i norra delen av det nordsvenska barrskogsomradet — fran insamlingar vid Pälkem. — Ent. Tidskr. 68. Sthlm. WittinGc, R. 1911: Omgifvande haf. — Atlas öfver Finland, Text. (Sallsk, Finl. Geogr.) Hfors. — — 1943:. Landhojningen utmed Baltiska Havet under ären 1898—1927. — Fennia. 68. Hfors. Voct, TH. 1912: Landskapsformerne i det ytterste av Lofoten. — Norsk, Geogr. Selsk. Aarbok. 23. Oslo. Wo.pstept, P. 1929: Das Eiszeitalter. Grondiinien einer Geologie des Diluviums. _— Stuttgart. WoopHEAD, T. W. 1929: History of the vegetation of the southern Pennines, — Journ. of Ecol. 18. Cambridge. 803 901 Wo rr, W. 1915: Uber die Grossgletscher von Alaska und die diluviale Vereisung von Nordamerika. — Geogr. Zeitschr. 21. Leipzig. WRIGHT, SEWALL 1931: Evolution in mendelian populations. — Genetics. 16, Brook- lyn, N. Y. — — 1932: The roles of mutation, inbreeding, crossbreeding and selection in evolution. — Proc. 6. Intern. Congr. Genetics. 1. Brooklyn, N. Y. WRIGHT, W. B. 1914: The Quaternary Ice Age. — London. WyNNE-Epwarps, V. C, 1937. Isolated arctic-alpine floras in Eastern North America: a discussion of their glacial and recent history. — Trans. R. Soc. Canada (V). 3. Ser. 31. Ottawa. — — 1939: Some factors in the isolation of rare alpine plants. — Ibidem. 33. XAMBEU, P. J. V. 1898: Moeurs et métamorphoses d’insectes. — Ann, Soc. Linn 45. Lyon. ZETTERSTEDT, J. W. 1840: Insecta Lapponica descripta. — Leipzig. ZEUNER, F. E. 1945: The Pleistocene Period, its climate, chronology and faunal successions. — Ray Soc. 130. London. — — 1946: Dating the past. An introduction to geochronology. — London. ZIMMERMANN, K. 1935: Zur Fauna von Sylt. — Schr. Naturw. Ver. Schlesw - Holst. 21. Kiel. SLE AS : a A IR ee SB CAM a ge f I RAs in > 5 nr & : ve SE Index to Part III 902-911 Only the carabids found in Fennoscandia, Recent or fossil, are included here. There is no reference to entries in the Supplement (p. 843). Abax ater 16, 113, 440, 454, 478, 544,