HARVARD UNIVERSITY Library of the Museum of Comparative Zoology .^^ LIBRARY H_ MAY 2 9 1990 ^ HARVARD UNIVERSITY GREAT BASIN NATURAUST VOLUME 50 NO 1 - MARCH 1990 BRIGHAM YOUNG UNIVERSITY GREAT BASIN NATURALIST Editor James R.Barnes 290 MLBM Brigham Young University Provo, Utah 84602 Associate Editors Brian A. Maurer Jimmie R. Parrish Department of Zoology Department of Zoology Brigham Young University Brigham Young University Provo, Utah 84602 Provo, Utah 84602 Other Associate Editors are in the process of being selected. Editorial Board. Richard W. Baumann, Chairman, Zoology; H. Duane Smith, Zoology; Clayton M. White, Zoology; Jerran T. Flinders, Botany and Range Science; William Hess, Botany and Range Science. All are at Brigham Young University. Ex Officio Editorial Board members include Clayton S. Huber, Dean, College of Biological and Agricultural Sciences; Norman A. Darais, University Editor, University Publications; James R. Barnes, Editor, Great Basin Naturalist. The Great Basin Naturalist, founded in 1939, is published quarterly by Brigham Young University. Unpublished manuscripts that further our biological understanding of the Great Basin and surrounding areas in western North America are accepted for publication. Subscriptions. Annual subscriptions to the Great Basin Naturalist for 1990 are $25 for individual subscribers, $15 for student and emeritus subscriptions, and $40 for institutions (outside the United States, $30, $20, and $45, respectively). The price of single issues is $12. All back issues are in print and available for sale. All matters pertaining to subscriptions, back issues, or other business should be directed to the Editor, Great Basin Naturalist, 290 MLBM, Brigham Young University, Provo, UT 84602. Scholarly Exchanges. Libraries or other organizations interested in obtaining the Great Basin Naturalist through a continuing exchange of scholarly publications should contact the Exchange Librarian, Harold B. Lee Library, Brigham Young University, Provo, UT 84602. Editorial Production Staff JoAnne Abel Technical Editor Carolyn Backman Assistant to the Editor Heidi Larsen Production Assistant ISSN 017-3614 3-90 650 4.5221 The Great Basin Naturalist Published AT Phovo, Utah, by Bhi(;ham Younc Univkhsity ISSN 0017-3614 Volume 50 31 March 1990 No. 1 A PLASMA PROTEIN MARKER FOR POPULATION GENETIC STUDIES OF THE DESERT TORTOISE (XEROBATES AGASSIZI) James L. Glenn' ", Richard C. Straight', and Jack W. Sites, Jr.^ Abstract. — Fifty-seven individual plasma samples from desert tortoises (Xerohates a^,assizi) representing 10 separate populations were analyzed by polyacr\lamide gel electrophoresis using alkaline buffers. An albumin-like protein was found to be polymorphic for two electromorphs in northern populations inhabiting the Mohave Desert Province, while Sonoran Desert populations to the south were monomorphic. The genetic divergence demonstrated in this survey is similar to earlier studies and provides evidence for the Colorado River as a potential barrier to gene flow among tortoise populations. These data suggest that tortoise plasma, examined by various electrophoretic methods, may provide a nondestructive means of determining the broad regional origin of desert tortoises. Desert tortoises {Xerohates agassizi) pres- ently inhabit two regions of southwestern Utah, separated east and west by the Beaver Dam Mountains. The population dynamics of the Beaver Dam Slope tortoises (west of the moimtains) have been severely impacted by both human and animal activities during the past several decades. Despite both fed- eral and state protective regulations, tortoise numbers in the slope region are probably at an all-time low (Mike CofiPeen, UDWR, personal communication). Two stable populations near St. George, east of the Beaver Dam Moun- tains, face new human development projects that threaten the future of these previously isolated populations. The eastern populations are found in Paradise and City Creek canyons, and relocation of some tortoises from these populations is presently under consideration by the Utah Division of Wildlife Resources. However, several relocation issues remain unresolved, including methods of collection, conditioning, transport, sex ratios, and speci- men numbers involved. One problem to be addressed is the question of genetic compati- bility of Utah's tortoise populations, especially if any of the tortoises east of the Beaver Dam Mountains are translocated to the western slope. Few data are available comparing the physiology or morphology of Utah's separate tortoise populations. Rainboth et al. (1989) electrophoretically analyzed whole blood homogenates of 146 desert tortoises from two separate localities in California for allozyme expression at 23 loci. The two populations were quite similar, as each locality contained unique elements only when allozyme combi- nations were used and there was a consider- able degree of overlap in allozyme frec^uency. Lamb et al. (1989) included five tortoises from Paradise Canyon in an analysis of phyloge- ographic patterns in mitochondrial DNA (mtDNA) of the gopher tortoise complex. Their data indicated that these individuals fit within an eastern Mohave "clone" represent- ing the northern Arizona and eastern Nevada region. Also, Jennings (1985) included five tortoises from the Beaver Dam Slope region Venom Research Laboratory, Veterans Administration Medical Center, Salt Lake City, Utah 84148. ^Hogle Zoological Gardens, Salt Lake City, Utah 84148. ■ Department of Zoolog\. Brigham Young University, Provo, Utah 84602. J. L. Glenn ETAL. [Volume 50 Table 1. Genotypes and variability estimates for the polymorphic GP-1 locus resolved in 10 localities oi Xerobates agassizi. Number of GP-1 Locality N genotypes A* H** L Pima Co., AZ .3 3AA 1.0 0.0 2. Tucson, AZ 6 6AA 1.0 0.0 3. Pinal Co., AZ 4 4AA 1.0 0.0 4. Maricopa Co., AZ 4 4AA 1.0 0.0 5. Kingman, AZ 3 2AA:1BB 1.3 0.0 6. Beaver Dam Slope, AZ 12 9AA:2AB;1BB 1.5 0.083 7. Paradise Canvon, UT 12 3AA:6AB:3BB 1.5 0.208 8. Lincoln Co., NV 2 1AA;1AB 1.5 0.250 9. Riverside Co., CA 4 1AA:3AB 1.5 0.375 10. San Bernardino Co ,CA 7 .3AA:4AB 1.5 0.214 - mean number ol alleles/locus. - mean heterozygosity-direct count. in biogeographic investigations comparing blood and tissue enzymes using horizontal starch gel electrophoresis. His study included no specimens from east of the Beaver Dam Mountains in Utah but showed that a north- to-south variation was evident and that the Arizona Beaver Dam Slope specimens fit within the northern (Mohavean) group. Mitochondrial DNA and isozyme studies are costly and may involve traumatic (to the tortoise) and labor-intensive biopsies of inter- nal tissue or the sacrificing of specimens for necropsy. The purpose of this study was to determine whether general proteins in tor- toise plasma could be used to detect geo- graphical differences in tortoise populations. Our primary interest was to compare plasma proteins among Utah's aforementioned tor- toise populations using alkaline polyacry- lamide gel electrophoresis. However, as elec- trophoretic profiles of plasma or serum from Xerobates agassizi have not been previously reported, we examined plasma profiles from tortoises from Arizona, California, and Ne- vada as well. This report follows the taxo- nomic grouping of gopher tortoises by Bram- ble (1982) and the nomenclature revision of Bour and Dubois (1984), applying Xerobates as the genus for the desert tortoise. Materials and Methods Study Area and Sampling A total of 68 desert tortoises were collected from 11 localities in Arizona (AZ), California (CA), Nevada (NV), and Utah (UT), repre- senting most of the range of this species in the United States. Plasma samples were collected in heparinized tubes (3 ml) by venapuncture (jugular vein or antebrachial sinus) from 15 tortoises from the Beaver Dam Slope, AZ, and Paradise Canyon, UT. Additional plasma sam- ples were donated by colleagues involved in desert tortoise projects in other parts of AZ, and in CA and NV. Localities and sample sizes for the populations sampled are listed in Table 1, and their geographic locations are plotted in Fig. 1. Electrophoresis Plasma samples were lyophilized and stored at 4 C. Polyacrylamide gel electrophoresis (PAGE) was run in a Bio-Rad vertical-slab electrophoresis cell. Model 220. The gel was 1.5 mm thick. Sample wells 10 mm long were formed with Canalco stacking gel (2.5% acry- lamide). The 7% acrylamide separating gel was 80 mm in length. Samples were elec- trophoresed toward the anode ( + ) at 20 mA constant current and were stopped 10 mm from the bottom of the gel. The electrophore- sis buffer used was 0.025 M Tris/0.192 M glycine, pH 8.3; bromophenol blue was used as tracking dye. The gel was fixed in 10% acetic acid/40% isopropanol/50% water for 60 min, stained with coomassie blue (0.05%) in 10% acetic acid/10% isopropanol/80% water for 60 min, and destained in 10% acetic acid/ 10% isopropanol/80% water, using three or four changes of solution over a 24-hr period. Only qualitative differences were exam- ined in this study, and only the faster migrat- ing proteins were used, since alkaline PAGE is not the preferred method for resolving differences in basic (globulin-like) proteins, due to their short migration distances. The 1990] Desert Tortoise Genetic Studies Fig. 1. Distribution of collection localities of tortoise genotypes AA, BB, and AB listed in Table 1. Numbers indicate localities. electrophoretic mobilities of rattlesnake {Crotalus atrox) plasma albumin and bovine serum albumin were used as markers and compared with the albumin-like protein(s) of the desert tortoise samples. Both albumin markers migrated at faster rates than the tortoise albumin-like proteins analyzed in this study. Genetic Analysis Two protein loci, identified in order of de- creasing anodal mobility, were designated "general proteins" (GP) 1 and 2. Allelic data at both loci were recorded as individual geno- types for analysis with the BIOSYS-1 program of SwofFord and Selander (1981). Measures of genetic variability computed for each popula- tion sample included average locus hetero- zygosity (H, direct count) and the mean num- ber of alleles per locus (A). The genetic dis- tance and similarity coefficients of Nei (1972, 1978) and Rogers (1972) were calculated for all pairwise combinations of samples (corrected for small samples sizes as described by Levene [1949]), and all such matrices were clustered by the UPGMA algorithm of Sneath and Sokal (1973). Genotype ratios from the largest samples (localities 6 and 7, n = 12 for both) were tested for conformance to Hardy- Weinberg proportions by the X" goodness-of- fit option of BIOS YS, again corrected for small sample sizes (Levene 1949). Results Si.xty-eight individual samples were ana- lyzed from four states, including 11 localities J. L. Glenn ETAL, [Volume 50 1 2 i * 3 H GP-2-1-*- GP-1 t! I AA BB AB BB AB (+) Fig. 2. Alkaline polyacrylamide gel illustrating GP-1 and GP-2 loci (enclosed in dotted rectangle), with allelic variation at GP-1 locus for plasma samples from three individuals (run in duplicate). Animals 1, 2, and 3 were consistently scored as genotypes AA, BB, and AB, respectively. Table 2. Matrix of genetic distance coefficients of Nei (1978, above diagonal) and Rogers (1972, below diagonal) for all pairwise combinations of Xerobates agassizi localities; locality numbers are as listed in Table 1 and D values are rounded off to two decimals. Locality 1 2 3 4 5 6 7 8 9 10 1 0.00 0,00 0.00 0,03 0.01 0,17 0.00 0.05 0.06 2 0.00 — 0,00 0.00 0.03 0.01 0.17 0.00 0.05 0.06 3 0.00 0.00 — 0,00 0.03 0.01 0.17 0.00 0.05 0.06 4 0,00 0.00 0,00 — 0.03 0.01 0.17 0.00 0.05 0.06 5 0.17 0.17 0,17 0,17 — 0,00 0.00 0.00 0.00 0.00 6 o.os 0.08 0,08 0,08 0.08 — 0.08 0.00 0.00 0.01 7 0.27 0.27 0,27 0,27 0.10 0,19 — 0.01 0.00 0.00 8 0,13 0.13 0.13 0,13 0.04 0,04 0.15 — 0.00 0.00 9 0.19 0,19 0.19 0,19 0.02 0,10 0,08 0,06 — 0.00 10 0.18 0,18 0.18 0,18 0.01 0, 10 0,09 0,05 0,01 — (considered as separate populations). The pro- tein profiles present in 7 of 9 plasma samples from the Desert Tortoise Natural Area (CA) and 2 samples from Utah were uni(}ue but were very likely artifacts due to their badly hemolyzed condition. These 11 samples were excluded from the data analysis. This reduced the number of samples to 57 and the number of localities to 10 (Table 1). General protein- 1 was polymorphic for two electromorphs (des- ignated A and B) in several samples (Fig. 2). Table 1 summarizes the ratios of GP-1 geno- types across these 10 localities and the esti- mates of variability across both loci. Geno- type ratios at localities 6 and 7 conformed to Hardy-Weinberg expectations (X" = 2.789, P = .095; X' .503, P .478, respectively. df = 1 in both cases), suggesting that this is a simple Mendelian co-dominant system with two alleles segregating in some populations. Table 2 summarizes pairwise comparisons of two genetic distance coefficients (Rogers 1972, Nei 1978) and shows that between- sample divergence was minimal. Nei's D val- ues, for example, range from 0.00 to 0.17. Four of the Arizona samples (Maricopa Co., Pinal Co. , Pima Co. , and Tucson) are identical (Nei's D = 0.00). Five localities having the B allele at GP-1 formed a distinct separate clus- ter, albeit the total degree of divergence irom the monomorphic populations was slight (D = 0.05). Within this group, the Arizona and California populations were nearly identical, while the Paradise Canyon (UT) samples were 1990] Desert ToHToisE Genetic Studies 1-4 10 ^ 6 8 7 I — \ — \ — \ — \ — I \ \ \ \ I I r~ .18 .16 .14 .12 .10 .08 .06 ~\ \ I \ — 1 .04 .02 .00 Rogers' Distance Fig. 3. Dendrogram based on Rogers' (1972) genetic distance values for 10 samples ofXerobates agassizi (see Table 1, Fig. 1). Clustering was by the UPGMA algorithm of Sneath and Sokal (1973), and the cophenetic correlation value was 0.826. the most divergent. These relationships are also visually displayed in the UPGMA den- drogram presented in Figure 3, using the statistical analysis method of Rogers (1972). All other UPGMA dendrograms gave identi- cal or nearly identical topologies (data not shown). Discussion Dessauer (1970) reported that while some blood proteins in reptiles are relatively conservative, others are quite polymorphic. Many species can be readily distinguished by the electrophoretic mobilities of blood pro- teins, and certain subspecies and populations can also be distinguished by differences in plasma albumin-like proteins (Dessauer and Fox 1958, Masat and Dessauer 1968). Masat and Dessauer (1968) also found that the albumin-like proteins of the Testudines have slower migration rates in alkaline buffers than do these same proteins in most other reptiles (and mammals). We also found that under alkaline conditions the desert tortoise albumin-like proteins migrate at slower rates than rattlesnake plasma albumin and bovine serum albumin. The results of this investigation suggest geographical differences in genetic variability of the albumin-like protein (GP-1) of desert tortoises. The northern (Mohavean) popula- tions were polymorphic, whereas the south- ern (Sonoran) populations were monomorphic at the GP-1 locus. An east-west Mohave dif- ference was observed due to the eastern isola- tion of the BB genotype (Table 1, Fig. 2) in populations from the eastern Mohave region of Utah and northwestern Arizona (Fig. 1). The B allele was not present in any of the central and southern Arizona samples. Those samples expressing the B allele may differ in the frequency of this allele, as suggested by the differentiation between Paradise Canyon and the Beaver Dam Slope, but the present sample sizes are too small for accurate deter- mination. Despite the small sample sizes, the heterozygosity estimates are more similar for the Paradise Canyon population and the three Mohavean populations (8, 9, and 10) than for the Paradise Canyon and Beaver Dam Slope populations (Table 1). If the Paradise Canyon tortoises differ from the Beaver Dam Slope population(s) in frequency of the B allele and are in fact more similar to California tortoises, this could reflect: (a) a divergence of allele frequencies between the slope and Paradise populations in allopatry, with allele frequen- cies at Paradise simply drifting to values similar to California populations (similarity by convergence); (b) transport of tortoises from California or Nevada to St. George and "dumping" into Paradise Canyon, but not at the slope (i.e., human-induced gene flow between California, Nevada, and Paradise); or J, L Glenn ETAL. [Volume 50 (c) transport and release of Arizona tortoises (most with AA genotypes at the GP-1 locus) on the Beaver Dam Slope but not Paradise Canyon, which would cause the A allele at the slope to increase in frequency at the expense of the B allele and "push" the slope population away from Paradise Canyon and California allele frequencies (see genotype differences in Table 1). These are not mutually exclusive hypotheses since a certain amount of "mixing" may have occurred at all localities of Califor- nia, Nevada, Beaver Dam Slope, and Paradise Canyon populations. The long history of human collection and translocation of desert tortoises between states constitutes a variable that could influ- ence the genetic structure of desert tortoise populations, especially when comparing al- lele frequencies between populations well known as captive release sites. Hundreds, perhaps thousands, of tortoises have been picked up along roadways and released in dif- ferent regions over the past several decades (Mike Coffeen and Eric Coombs, UDWR, personal communication). This logistical dis- placement of tortoises continues at present. Certain localities have been popular release sites, e.g., regions of southern California (Desert Tortoise Natural Area), Arizona (McDowell Mountain region near Phoenix), Nevada (near Las Vegas and recreation areas), and Utah (St. George). The most common avenues of translocation by motorists crossing the Mohave Desert are east to west and vice versa. Favorable habitats for tortoises exist from Washington County in southwestern Utah to southern California — a region heavily traveled over a major interstate highway for decades. In addition to the release of tortoises in this region by motorists and local citizenry, Utah's Beaver Dam Slope population(s) have been the site of approximately 200 captive release tortoises, regulated by the Utah Divi- sion of Wildlife Resources since 1970 (Mike Coffeen, UDWR, personal communication). Although Utah's tortoise populations lo- cated east of the Beaver Dam Mountains are often regarded as "captive escapees" in a nonindigenous setting, there are no scientific data confirming this view. The hypothesis seems to have originated from several sources, e.g., magazine and newspaper arti- cles, and the opinions of a few naturalists and herpetologists. Support for the intro- duced status is based on three general obser- vations. One is the fact that some tortoises found in the St. George region obviously have been captive specimens, exhibiting rope or chain holes in their shells or having painted areas on their shells. Second, residents of Washington County often keep tortoises as pets, and some have escaped or were pur- posely released. Third, the Beaver Dam and Virgin Mountains presently form an east-west barrier between natural assemblages, and a few reptiles found west of the mountains are not present east of them (e.g. , Crotalus scutu- latus, PhyllorhyncJius deciirtus, Dipsosaurus dorsalis). The possibility still remains that some of Utah's tortoises found east of the mountains may be derived from ancestral stock of naturally occurring tortoise popula- tions that have since mixed with captive re- leased specimens. Support for the natural population relies on the fact that the Mohave Desert Province extends into this region, and many other Sonoran life-zone animals found on the western slope are also found east of the mountains and are natural assemblages. Like the desert tortoise, some of the reptiles are life-zone specific and occur on both sides, for example, the banded gecko {Coleonyx varie- gatus), Gila monster {Heloderma suspectum), and sidewinder {Crotalus cerastes). The geographical differences observed in this investigation are similar to those found in the allozyme survey by Jennings (1985), the mtDNA survey by Lamb et al. (1989), and the morphometric analysis of tortoise remains by Weinstein and Berry (1987). Specifically, the present study supports the earlier molecular investigations of Lamb et al. (1989), which showed divergence between tortoise popula- tions north and west of the Colorado River and those to the south and east. Their report pro- vided good evidence that tortoise populations now isolated on opposite sides of the Colorado River have likely been separated from each other for several million years. The mtDNA lineages from central and southern Arizona formed a single haplotype that differed from the northern haplotypes in CA, NV, UT, and extreme northwestern AZ by a minimum of 17 restriction site changes (see Fig. 2 in Lamb et al. [1989]). This is one of the highest levels of intraspecific genetic divergence reported for any animal species and exceeds that reported for man\ interspecific comparisons. 1990] DesehtTohtoise Genetic Studies The exception in our study was the small sample fi-om near Kingman, AZ (locality 5 in Fig. 1), which genotypically grouped with the Mohavean populations north and west of the Colorado River. Conse(}uently, the genotypic composition for population 5 must be inter- preted with caution. The single BB homo- zygote in a sample of three individuals would not be expected unless the B allele was segre- gating at a high frec^uency. These results may be due to several factors, e.g., the transloca- tion of this specimen by humans, a sampling error for a low-frequency allele, or a degrada- tional artifact in this sample. If future sam- pling verifies the presence of a high- frequency B allele at this locality, it could represent an ancestral polymorphism shared with Mohavean populations north of the Colo- rado River. This anomalous result under- scores the need for statistically adequate sam- ple sizes in all future genetic studies of the desert tortoise. Therefore, identifying the specific origin of any individual tortoise on the basis of nuclear gene markers may be difficult. Weinstein and Berry (1987) suggest using a combination of physiological and mor- phometric screening methods to designate re- gional types. They analyzed shell morphology of adult (> 180 mm) desert tortoises of both sexes by using morphometric data gathered by the Bureau of Land Management at 31 different localities. These measurements were collected from tortoise remains by several persons over a 48-year period. These authors noted that live tortoises were not used in their analysis and that shell morphology of tortoise remains does incur some shrinkage over time following death. They recommend further studies comparing live tortoises. None of these authors suggested that the differences observed in their investigations justify sub- specific designation for any of the regional populations, and thus Xerobates agassizi re- mains a monotypic species. Alkaline PAGE was useful in examining the albumin-like proteins in tortoise plasma but does not resolve slight differences in the elec- trophoretic mobility among the majority of the plasma proteins. However, this method did detect the polymorphic nature of the albu- min-like (GP-1) protein, and this protein may be one "marker" that could be used for desig- nation of broad regional types. Since the re- sults of this investigation are, with the possi- ble exception of locality 5 noted above, very similar to the findings of others in that the Ijroad regional genotypes in desert tortoises are approximately concordant, the PAGE screening of the plasma protein marker may provide one inexpensive method of objec- tively determining the regional origin of tortoises. If allele frequency data are to be used, additional specimens from Paradise Canyon and Beaver Dam Slope populations are needed to determine the significance of the allele frequency differences between these two localities. Additional samples from throughout Nevada and California would also be required. Also, some variations observed in the slower-migrating proteins could be ex- amined with more high-resolution techniques (e.g., isoelectric focusing, two-dimensional electrophoresis) combined with bio-image an- alytical instrumentation that can quickly and accurately scan and record qualitative differ- ences in electrophoretic profiles. Morpho- metric data from live tortoises should be col- lected to compare Paradise Canyon and western slope populations in conjunction with future molecular analyses. In fact, external morphology (such as shell shape) could be a more functional criterion for the survivability of Paradise Canyon tortoises on the Beaver Dam Slope (see Weinstein and Berry [1987]), since genetic differences between these pop- ulations appear to be slight. Acknowledgments We greatly appreciate the cooperation of the following individuals for their efforts in supplying plasma samples used in this study: Trip Lamb, Savannah River Ecology Labora- tory, Aiken, South Carolina; Randy Jennings, University of New Mexico, Albuquerque; Jim Jarchow, Neglected Fauna International, Tucson, Arizona; Brian Henen, Ken Nagy, and Charles Peterson, Laboratory of Bio- medical and Environmental Sciences, Uni- versity of California, Los Angeles; and Paula Acer, Kingman Animal Hospital, Kingman, Arizona. Trip Lamb and Ken Nagy also reviewed this manuscript and offered valu- able comments, for which we are grateful. Michael CoflFeen, Utah Division of Wildlife Resources, was instrumental in facilitating the fieldwork involved in this study. We also thank Martha Wolfe, Veterans Administration J. L. Glenn ETAL. [Volume 50 Medical Center, Salt Lake City, Utah, for her technical skills, and appreciate the coopera- tion of Cecil Schwalbe and Terry Johnson, Arizona Game and Fish Department, and Randy RadantandTim Provan, Utah Division of Wildlife Resources. This research was sup- ported in part by the Utah Division of Wildlife Resources and Hogle Zoological Gardens, Salt Lake City, Utah. Literature Cited BouR, R.,andA Dubois 19S4. Xcrohates agassiz, 1857, synonyme plus ancien de Scaptochclys Bramble, 1982 {Reptilia. Chelonii, Testudinidae). Bulletin Mensuel de la Societe Linneenne de Lyon .dS: 30-32. Bramble, D. M 1982. ScaptocJwhjs: generic revision and evolution of gopher tortoises. Copeia 1982: 852-867. Dessauer. H C 1970. Bloodchemistry of reptiles: physi- ological and evolutionary aspects. Pp. 1-54 in C. Cans and T. S. Parsons, eds.. Biology of the Rep- tilia. Vol. 3. Academic Press, New York. Dessauer, H. C . andW. Fox 19.58. Geographic variation in plasma protein patterns of snakes. Proceedings of the Society of E.xperimental Biology and Medicine 98: 101-105. Jennings, R D. 1985. Biochemical variation of the desert tortoise, Gopherus agassizi. Unpublished thesis. University of New Mexico, Albucjuerque. 72 pp. Lamb, T, J. C.AviSE, AND J. W, Gibbons 1989. Phylogeo- graphic patterns in mitochondrial DN.\ of the desert tortoise (Xerobates agassizi), and evolu- tionary relationships among the North American gopher tortoises. Evolution 4.3: 76-87. Le\ene. H 1949. On a matching problem arising in genetics, .\nnals of Mathematical Statistics 20: 91-94. Masat. R J , AND H C Dessauer 1968. Plasma albumins of reptiles. Comparative Biochemistrv and Phvsi- ology 25: 119-128. Nei. M 1972. Genetic distance between populations. American Naturalist 106: 283-292. 1978. Estimation of average heterozygosity and genetic distance from a small number of individu- als. Genetics 89: .583-.590. Rainboth, W J , D G BuTH. and F B Turner 1989. Allozyme variation in Mojave populations of the desert tortoise, Gophcnts agassizi. Copeia 1989: 11.5-123. Rogers, J. S. 1972. Measures of genetic similarity and genetic distance. Studies in Genetics, University of Texas Publications 72: 145-153. Sne,\th, p. H. A., AND R R SOKAL. 1973. Numerical tax- onomy. W. H. Freeman, San Francisco. SwoEFORD, D. L. andR B Selander 1981. BIOSYS-1: a FORTRAN program for the comprehensive anal- ysis of electrophoretic data in population genetics and systematics. Journal of Heredit\' 72: 281-283. Weinstein, M N . and K H Berry 1987. Morphometric analvsis of desert tortoise populations. Bureau of Land Management Report CA9.50-CT7-003. Riv- erside, California. 39 pp. Received 5 Novem])er 1989 Accepted 16 January 1990 Cireat Basin Naturalist ",()( 1 1. 19«), pp 9-19 EFFECTS OF NITROGEN AVAILABILITY ON GROWTH AND PHOTOSYNTHESIS OF Afi7EA//S/A TRIDENTATA SSP. WYOMINGENSIS Paul S. DoesclicM-', Hichaid F. M illcr , Kill mio Wans ', iiiicl Ji'H Hose" Abstract. — This study examined the effects of aherations in soil nitrogen on the growth oi Artemisia tridentata ssp. wyomingensis Niitt. Soil nitrogen content was altered by applying sugar (45 g/m"), nitrate (4.5 g/m^), or ammonium (4.5 g/m"), and the results were compared with a control treatment (no soil amendments). Addition of either form of nitrogen significantly increased leaf nitrogen content, mean maximum length of ephemeral leaves, number of ephemeral leaves per terminal shoot, and current year's vegetative stem length over the control and sugar treatments. Both soil water and predawn xylem potentials during active growth were lower in the nitrogen-treated plots. The higher growth acti\it\ and greater leaf mass of A. tridentata in the nitrogen treatments may have been responsible for this result. Higher photosynthetic rates observed in the nitrogen treatments during an early June sampling period also lend support to this observation. This study suggests A. tridentata ssp. ivyomingensis would opportunistically take advantage of increased availability of soil nitrogen. The ability of this species to respond positively to increased soil nitrogen may enhance its competitiveness over associated perennial species. Artemisia tridentata Nutt. is a semidecidu- ous perennial shrub occupying 44.8 million ha in the western Intermountain sagebrush- steppe. It is the most abundant shrub in this ecosystem. During the past century, in- creases of A. tridentata have been attributed to overgrazing of perennial grasses by domes- tic livestock, cultivation of lands too arid to produce crops, and alterations in fire fre- quency (Hironaka and Tisdale 1963, Tisdale et al. 1969, Tisdale and Hironaka 1981). As A. tridentata has increased in the Great Basin, both production and diversity of herba- ceous understory species have declined. Nu- merous physiological and morphological char- acteristics of A. tridentata have been shown to enhance its effectiveness as a competitor with native bunchgrasses, especially for soil mois- ture (DePuit and Caldwell 1973, Eissenstat and Caldwell 1988, Miller and Shultz 1987, Miller 1988). Among these, the ability of A. tridentata to maximize leaf area early in the growing season by overwintering one-third of its leaf biomass and by developing ephemeral leaves early in the spring strongly enhances its ability to photosynthesize during favorable growth periods (Depuit and Caldwell 1973, Miller and Shultz 1987, Miller 1988). A deep, well-developed root system also allows A. tri- dentata to capture soil moisture from a soil volume much larger than that of perennial grasses (Sturges 1977). Relatively little research has examined the response of this species to soil nutrients such as nitrogen. Limited work, however, indi- cates A. tridentata to be an effective competi- tor for soil nutrients. Caldwell et al. (1985) demonstrated that this species successfully competes for soil phosphorus with the native perennial grass Afi,ropyron spicatum (Pursh) Scribn. & Smith. The accumulation of nutri- ents and higher soil nitrogen mineralization rates in surface soils beneath A. tridentata canopies may also convey an ecological advan- tage to plants during active growth periods (Charley and West 1975, 1977, Doescher et al. 1984). Few studies, however, have eval- uated the response of A. tridentata to in- creased or decreased amounts of available soil nitrogen. Carpenter (1972), working in the Colorado Plateau, reported that 134 kg N/ha applied to A. tridentata yielded an 81% in- crease in total leafy material compared with a nontreated control. However, Carpenter and West (1987) found little response to nitrogen in A. tridentata grown on mine spoils. The form of nitrogen, whether NH4 or NO3, may also be an important factor in the mineral nutrition of aridland shrubs (Wallace et al. 1978). Department of Rangeland Resources. Oregon State University. Corvallis, Oregon 97.331, Eastern Oregon Agricultural Research Center, Burns, Oregon 97720. 10 p. S. DOESCHERETAL. [Volume 50 Our experiment was designed to determine how depletion or addition of different forms of nitrogen affects A. tridentata growth and car- bon-assimilation rates. Our hypothesis was that A. tridentata responds favorably to in- creases in soil nitrogen. Materials and Methods The study was conducted at the Squaw Butte Experimental Range in southeastern Oregon (119°43'W longitude, 43°29'N lati- tude), 67 km west of Burns, on the northern fringe of the Great Basin. The 37-year mean annual precipitation for this area is 284 mm. Precipitation during the 1987 crop year (Sep- tember-August) was 296 mm. The Squaw Butte Experimental Range typically receives most of its moisture between October and June, generally as snow, with little precipita- tion received in July and August. The mean temperature in winter is —0.6 C, with the daily minimum averaging —4.8 C, and the mean temperature in summer is 17.6 C, with the daily maximum averaging 26.8 C. The study site is located in an Artemisia tridentata spp. wyomingensislStipa thurberiana habitat type, at an elevation of 1,372 m (Doescher et al. 1984). This site has been excluded from grazing by domestic herbivores for the past 40 years. Soil texture is gravelly fine sandy loam and classified as Xerollic Durothids (Lentz and Simonson 1986). Soils vary in depth from 35 to 45 cm and are underlain by an indurated duripan 5-20 cm thick, which is underlain by unweathered basalt. A detailed description of soil nutrient levels is provided by Doescher et al. (1984). Experimental Procedures A completely randomized plot design was used with 10 replications of each treatment. Plots 5 X 5 m were laid out with an A. triden- tata located in the center of each plot. To maximize uniformity, we selected plots that had vigorous-appearing A. tridentata plants of similar growth form and size. Plant measure- ments were recorded on the center A. triden- tata plants, and soil measurements were col- lected within 1.5 m of the stem base. The remainder of the plot was used as a buffer. Treatments were applied both in March and late November of 1986. Treatments were (1) control (no amendments added), (2) grami- lated sugar (45 g/m"), (3) ammonium — (NH4)oS04 (nitrogen = 4.5 g/m"), and (4) nitrate — HNO3 (nitrogen = 4.5 g/m^). Sugar addition was assumed to increase the C:N ratio to decrease availability of soil nitrogen (Baathetal. 1978). Both ammonium and sugar were broadcast onto the 5 x 5-m plots. Ni- trate was diluted in water (1 part HNO3 to 5 parts water) and applied with a backpack sprayer. All herbaceous plants were dormant at the time of application. Soil and plant growth measurements were recorded during the following 1987 growing season. Soils were analyzed for ammonium and ni- trate concentration in the A and B horizons in five plots per treatment on 14 April, 26 May, and 25 July. Soil analysis was performed using a KCl extracting technique (Horneck et al. , in press). Soil water content was measured from 1 April to 15 September once every two weeks in each of the A and B horizons. One soil sample was collected for each of the two depths within each plot for all treatments. Soil water was measured gravimetrically, and soil water release curves were developed for each depth to define soil water potential. Ephemeral leaf number and maximum length, and vegetative stem elongation were measured on five randomly selected terminal branchlets of the single Artemisia located within each plot. Leaf measurements were recorded on three dates during initiation and expansion of ephemeral leaves (15 April to 5 May). Vegetative stem elongation was mea- sured on five dates from initiation to termina- tion of growth. Leaf nitrogen content was measured on current year's leaves (both ephemeral and persistent) collected from veg- etative stems on 15 and 21 April, 1 June, and 1 August on all plots. Collections represented three phenological stages: initial leaf elonga- tion, rapid leaf and stem growth, and early flowering. The Semimicro-Kjeldahl method was used to determine total leal nitrogen con- tent (Bremmer 1965). Specific leaf weight (g/ m") was obtained by measuring leaf area on current year s green leaves on 12 dates during the growing season. Leaves were removed from one randomly selected terminal branch in each plot and placed in a damp cooler. Several hours later leaf area was measiued on a leaf area meter, and weight was determined b\ o\ (Mi-drying the leaves at 60 C for 48 hr. 1990] Effects OF NiTRociKNAvAiLABiiJiY 11 Table 1. Soil nitrate and amiiioiiiiiin coTitt'iit (pi)m) at soil dcpllis ol 0-20 cm and 20-40 cm. Sampling date (Control Treatment Sugar Nitrate Ammonium April 14 Mav26 July 25 April 14 May 26 July 25 April 14 May 26 July 25 April 14 Mav26 July 25 0.50^* 1.16^ 1.82' 0.74' 1.56' 1.62' 6.32" 4.92" 0.80" 7.88" 8.96" 1.04" NO, 0-20 cm 0.48" 12.42'' 1.14" 6.84" 3.28" 20-40 cm 11.64'' 0.84" 20.60'' 1.10" 8.46'' 1.90" NHj 0-20 cm 9.02'' 4.18" 5. 14" 4.82" 7.22" 0.70" 20-40 cm 3.00" 6.82" 10.18" 7.68" 11.04" 1.10" 2.84" 1.20" 1.16" 5.02"'' 1.64" 1.46" 3.46"'' 7.78" 4.92" 0.70" 12.32" 11.96" 9.30'' *Niinibers followed by the same letters are not significantly different (P < .0.5) between treatments for each soil depth and date Xylem potentials (Scholander et al. 1965, Waring and Cleary 1967) on current year's vegetative branchlets were measured during the 1987 growing season with a pressure chamber (PMS Corporation, CorvaUis, Ore- gon). Predawn (19 May, 3 June, and 21 July) and midday (15 April and 3 June) measure- ments were recorded between 0430 and 0630, 1130 and 1230 hr, respectively. Five branch- lets were measured in each treatment. Sam- ples were selected at random, removed from the shrub, and immediately measured in the pressure chamber. Photosynthesis was measured on one ran- domly selected vegetative branchlet of an Artemisia plant in 5 of the 10 plots for each treatment on eight dates from mid-April through early August. Measurements were recorded between 1200 and 1300 hr using a LI-6000 (LI-COR, Lincoln, Nebraska) por- table photosynthesis meter with a quarter- liter chamber. To attain an adequate amount of leaf area in the chamber, we recorded measurements on both previous and current season vegetative branchlets. Initial photo- synthesis values were used and corrected using the formula developed by LI-COR (McDermitt 1987). Statistically significant treatment effects for variables measured on Artemisia were identi- fied using analysis of variance procedures. Time was set as a variable, in addition to treat- ment. Least significant differences (LSD) (P < .05) were calculated only when the F value was significant (P < .05) (Steel and Tor- rie 1980). The General Linear Model Proce- dure of SAS was used to evaluate treatment differences for specific leaf weight and photo- synthesis (SAS 1988). This procedure permit- ted statistical analysis of unbalanced data sets. Both interactions and main effect means were separated using Fischer's least significant difference test. Only statistically significant results are reported in the Results and Dis- cussion. Results Soil Nitrogen The addition of nitrate increased nitrate levels 25- and 28-fold in the upper and lower soil depths, respectively, at the beginning of the growing season (Table 1). Nitrate levels remained high compared with the other three treatments at both depths during the growing season. The shallow character of the soil prob- ably limited nitrate losses to leaching, maxi- mizing the amount available for plant uptake. The addition of ammonium increased ammo- nium 188% in the B horizon. The increase in 12 P. S. DOESCHERETAL. [Volume 50 25 c _J E E 'x D 20-- 15- 10 5- 0 April 15 April 20 April 25 Date April 30 May 5 Fig. 1. Ephemeral leaf maximum length oi Ai-temisia tridentata ssp. ivtjomingensis during the early growing season. Standard errors are presented for each mean. the A horizon was not significant. The sugar treatment did not appear to change soil nitro- gen levels when compared with the control, possibly indicating that carbon was not limit- ing decomposer microflora (McGarity 1961). Growth In the early spring both mean maximum length of ephemeral leaves and number of ephemeral leaves per terminal shoot were greater in the ammonium and nitrate treat- ments than in the sugar and control treat- ments (Figs. 1, 2). Ephemeral leaf lengths during April in the nitrogen-treated plots did not differ from one another but were greater than in the nonnitrogen treated plants. Con- trol and nitrate-treated plants did not differ from one another in leaf length for the May sampling period. Analysis of main effect means for ephemeral leaf numbers revealed that plants in nitrogen-treated plots did not differ from one another, and that their values were greater than similar values found for the control and sugar treatments. In addition, the development of new leaves on terminal shoots in both nitrate and ammonium treatments appeared to increase at a greater rate in early May than in sugar and control plots. The addition of nitrogen increased current year's vegetative stem length compared with no nitrogen addition throughout the growing season (Fig. 3). Stem length was similar be- tween both nitrogen forms in April and May but continued at a more rapid rate in the nitrate treatment in June. At the termination of vegetative stem elongation, stems in the nitrate and ammonium treatments were 175 and 140% longer, respectively, than shoots in the control treatment. Stem elongation in the control and sugar treatments was similar. The addition of either form of nitrogen did not increase specific leaf weight averaged across the growing season as compared with the control treatment (Table 2). The addition of sugar, however, reduced specific leaf weight across dates on the average by 12% compared with the control. The major differ- ence in specific leaf weight between control and sugar treatments occurred during abscis- sion of ephemeral leaves in late July and early August (Fig. 4). Specific leaf weight increased approximately 180 to 200% when ephemeral 1990] Effects OK NrrH()(;i':NA\'Ai I. Mil III V 13 E 16-- 14- 12 T 10-j- 8 6 4-- 2-- 0 D -n Control -• Sugar ■ Nitrate A Amrnonium April 15 April 20 April 25 Date April 30 May 5 Fig. 2. Ephemeral leaf numbers per terminal bud of Ai-temisia tridcntata ssp. wijomingensis during the early growing season. Standard errors are presented for each mean. cn c 0} _J E -M 0) > > 100-- D 80 60" 40 20-- 0 D- — D Control • - — • Sugar ■ - — ■ Nitrate A- A Ammonium May 5 May 15 May 25 June 4 June 14 June 24 Date Fig. 3. Effects of four different nitrogen treatments on current year vegetative stem length of Ar/PHiis/a tridcntata ssp. wyomingensis. Standard errors are presented for each mean. 14 P. S. DOESCHERETAL [Volume 50 0.250 (N £ O JC Oi o *♦-; *o Q) Ql 0.200 -- 0.150 -- 0.100 0.050 -- 0.000 Control I \ I Sugar ■I Nitrate I / I Ammonia i Fig. 4. Apr15 Apr21 Apr22 Jun03 Jun09 Junll Jun18 Jul13 Jul27 Aug05 AugIB SepOl Date Specific leaf weights oi Artemisia tridentata ssp. wyomin^ensis during the course of the 1987 growing season. Table 2. Specific leaf weights of Ar-fcmisia tridentata ssp. wijomingensis growing under four nitrogen treat- ments. Mean values are averaged over the 19S7 growing season. Treatment Specific leaf weight (g/m") Control Sugar Ammonium Nitrate 145' 132'' 141" 145' *Means followed by similar letters are not significantly different at P £ .05. T\BLE 3. Leaf nitrogen concentration (mg/g) of Artemisia tridentata ssp. wyomingensis growing under four nitrogen treatments. Means are averaged over the 1987 growing season. Treatment Leaf nitrogen concentration (mg/g) Control Sugar Ammonium Nitrate 19.6" 19.6" 25. l'' 27.0'^ *Means followed In similar letter not signifieantly different at P = leaves abscised in August for all treatments. By late August, when most ephemeral leaves had abscised, specific leaf weight between treatments was similar. Application of both forms of nitrogen re- sulted in greater leaf nitrogen contents of Artemisia plants (Table 3). Analysis of main effect means revealed that plants in the nitrate and ammonium treatments had greater con- centrations than did the control and sugar treatments. Water Relations and Photosynthesis Soil water depletion rates were generally more rapid in the nitrogen-treated plots at both soil depths than in the sugar and control plots (Figs. 5, 6). Differences were greatest during rapid growth in mid- May in the lower 20 cm. In the upper 20 cm, the sugar-treated plots maintained the highest soil water con- tent from April through May. Predawn plant water potentials were simi- lar among treatments in mid-May and late July (Table 4). During rapid growth in June, however, predawn plant water potentials in nitrogen-treated plots were lower than both control and sugar treatments. Midday water potentials were also lower in June in both nitrogen treatments compared with sugar and control. Both predawn and midday readings declined in all treatments as the season pro- gressed. Photosynthesis was significanth different between treatments on only the 15 June 1987 1990] Effects of Nitrogen Availability 15 c 0) c o o Q) o o CO April June July Months Fig. 5. Seasonal pattern of soil water content in the upper soil profile (2-20 cm) for four different nitrogen treatments in 1987. Vertical bars are 95% confidence limits. Field capacity (-0.03 MPa) = 18% soil water content. C 0) -t-J c o o ■+-' D O 15 — 13—- 11 -- 9-- 7- D- □ Contro • Sugar ■ Nitrate A Ammonium April May June July Months Aug Fig. 6. Seasonal pattern of soil water content in the lower soil profile (20-40 cm) for four different nitrogen treatments in 1987. Vertical bars are 95% confidence levels. 16 P. S DOESCHERETAL. [Volume 50 Table 4. Predawn and midday plant water potentials (MPa) in Ai-tcmisia tridcntata ssp. ivijomin^ensis through the growing season in 1987. Date Treatments Control Sugar Nitrate Ammonium Predawn May 19 June 3 July 21 Midday April 15 June 3 1.24' i.3r 1.62'' 1.85"' -2.09'" 1.05" 1.44'' 1.62'' -1.67' 2.06'' 1.04-" 1.77''^ 1.74'''' 1.65-' 2.24*' 1.21" ■1.77'' ■1.77'' -1.62" ^.Sl*" 'Numbers followed by the same lowercase letters are not significantly different (P < .05) between dates for each treatment, numbers followed by the same uppercase letters are not significantly different (P < .05) between treatments tor each site. Table 5. Photosynthesis (mg COo m ' sec ^) of Ai-temisia tridcntata ssp. uyominocnsis under four nitrogen treatments during 1987. Date Control Sugar Nitrate Ammonium April 17 June 15 June 20 July 30 August 1 0.30 0.25'' 0.20'' 0.42" 0.43" 0.00' 0.33'' 0.16" 0.22" 0.43" 0.36' 1.73" 0.28" 0.22' 0.20" 0.54 2.18" 0.25" 0.20" 0.51" *Means followed bv the same letter ; lit significantK ditTerent between treatments at the same date date (Table 5). Increased photosynthesis on this date corresponded to the period of rapid leaf and stem growth in early June. Discussion Nitrogen availability is probably second only to water as the most limiting factor to biomass production in Great Basin plant com- munities (James and Jurinak 1978). At our study site, addition of nitrogen was also found to significantly promote aboveground growth of A. tridentata ssp. wyomingensis. Fertiliza- tion increased ephemeral leaf size, number of ephemeral leaves, and stem elongation rates. Fertilization also increased nitrogen amounts in the subsoil and the leaves. These responses lend support to our original hypothesis, namely, that A. tridcntata responds favorably to increases in soil nitrogen. Although our results showed a positive growth response in A. tridcntata to nitrogen, the role nitrogen plays in intlucncing growth of arid species is poorly imderstood. In the desert Southwest, researchers have report- ed a variable response of Larrea tridentata (DC.) Cov.— dominated communities to water and nitrogen inputs (Gutierrez and Whitford 1987, Fischer et al. 1987). During certain years, application of water and/or nitrogen lesulted in enhanced growth, while in other years no response was observed. Fischer et al. (1987) concluded this variable response was due to yearly changes in available soil nitro- gen. In the Great Basin, biomass production of forage species has not always been related to precipitation (Charley 1972, Sneva and Britton 1983, Miller et al. 1990). Sneva and Britton (1983) and Miller et al. (1990) reported reduced herbaceous production in the third of three consecutive wet years. It has been spec- ulated that nitrogen may be limiting following successive wet years, as prolonged plant growth depletes soil nutrients and poor (jual- itv organic matter is slow to decompose (Parker et al. 1984, Fischer et al. 1988). On a reclaimed mine spoil. Carpenter and West (1987) indicated no response to nitrogen addi- tions for the species Artemisia tridcntata ssp. vaseyana. Available nitrogen was probably not limiting due to stockpiling of topsoil and the lack of site occupation In an establishing plant community. In Carpenter and West's (1987) stud\ leaf nitrogen concentration of control plants was 3.2%. In contrast, our study had leaf nitrogen concentrations of 2.0, 2.7, and 2.5%, tor control, nitrate, and am- nioniuin, respecti\ el\ . Site and subspecies 19901 Effkctsok Nithockn A\aii,\iuli it 17 differences ina\ ha\e also eontiihuted to tlie diflercMit response. A positi\e u;ro\vth response was shown lor the amnioninni plots in spite o( enhanced ani- nioninni le\ els heinsi; fonnd onK in the 20-40 cm soil depth. Tlu> nonsignificant concentra- tions of amnioninni in the snrface soils may have l:)een cansed 1)\ losses dne to volatiliza- tion, absorption of annnoninm on soil colloids, high biological activit>', and/or ammoninm be- ing hydrolyzed to nitrate in the lower soil depth. Also, the growth response may have been related to the application of ammoninm in the snlfate form. Whether or not A. triden- tata responds to snlfnr additions has yet to be determined. Despite the lack of enhancement in the upper soil depth of the ammonium- treated plots, the extensive root system of A. trident at a would allow this species to readily utilize ammonium in the lower soil depths. Xylem potential and soil moisture readings found in this study would, at first, appear contradictory to the observed growth re- sponses. The more negative soil moisture and plant water potentials reported for A. triden- tata growing in the nitrogen plots suggest that these plants were more water stressed, and thus should have reduced aboveground growth. However, the opposite was found. We feel that a possible explanation for these results centers on the greater growth and physiological activity of plants in the nitrogen plots. Since specific leaf weights were simi- lar between treatments during active leaf growth, we can assume that differences in leaf area between treatments are approximately proportional to biomass. Leaf biomass of A. tridentata in the nitrate and ammonium plots was found to have increased 520 and 230%, respectively, over control (Wang 1989). Greater growth and leaf area of these plants may mean that more soil water was used by plants in the nitrogen treatments. This observation is supported by the research of Svejcar and Browning (1988). They re- ported a greater leaf area, higher physiologi- cal activity, and a subsequently more rapid soil water depletion in burned versus un- burned stands of Androf)Ofi,()n ^erardii Vit- man. The greater photosynthetic rates of A. tridentata in the nitrogen plots during early June also further support this observation. In addition, an increase in shoot-to-root ratio could have influenced plant-water relations. Nitiogen additions have been reported to in- crease shoot-to-root ratios in Larrea trideti- tata (Fischer et al. 1988, Lajtha and Klein 1988). Specihc leal weights averaged across the growing season were significantly influenced by the sugar treatment. Prior to leaf senes- cence, specific leaf weights were similar be- tween treatments on seven of eight dates mea- sured. Specific leaf weights in the sugar treatment, however, were significantly less than in the control during senescence and abscission of current year's ephemeral leaves. Once the majority of ephemeral leaves had abscised in mid-August for all treatments, specific leaf weights were again similar. The decrease in specific leaf weights in the sugar treatment was probably a function of delayed leaf senescence and abscission rather than leaves being lighter per unit of surface area. Although not indicated by the soil nitrogen data, some reduction in available soil nitrogen may have occurred on the sugar plots. Marschner (1986) reported an increase in leaf area indices in plant populations with an increase in nutrient supply. Increased leaf de- velopment during the early part of the grow- ing season and a larger leaf area index in the years when mineralizable nitrogen levels are relatively high may increase Artemisia's com- petitive advantage for nutrient resources over associated species. Miller (1988) reported that Artemisia maintained a relatively high leaf area early in the spring compared with associ- ated species, allowing it to capture soil water resources early in the growing season. Early increased leaf area also enhances its ability to maximize photosynthesis when environmen- tal conditions are favorable (DePuit and Cald- well 1973). In warm desert shrubs, rate of leaf area development was the primary factor lim- iting whole-plant carbon gain during the early portion of the growing season (Comstock et al. 1988). In conclusion, application of both nitrate and ammonium increased growth response of A. tridentata over control and sugar treat- ments. Apparently this species can oppor- tunistically take advantage of increased soil nitrogen by increasing amount of leaf area available for growth. Increases in avail- able soil nitrogen might occur following events such as several consecutive years of below-average precipitation or weakening of 18 P. S. DOESCHERETAL. [Volume 50 perennial grasses through overgrazing. A. tri- dentata may enhance its competitiveness by responding favorably to increased levels of soil nitrogen. Acknowledgments We thank T. J. Svejcar, W. C. Krueger, R. H. Waring, and A. R. Tiedemann for their helpful reviews on earlier versions of this manuscript. This is Technical Paper No. 9006 of the Oregon Agricultural Experiment Station. Literature Cited BaATH, E., V LOHM, B LUNDREM, T ROSSVALL. B SODER- STROM. B SOHLEUIUS, AND A WiREN 1978. The effect of nitrogen and carbon supply on the devel- opment of soil organism populations and pine seedlings: microcosm experiment. Oikos 31; 153-163. Bremmer, J. M 1965. Total nitrogen. Pp. 1149-1178 in C. A. Black, ed.. Methods of soil analysis. Part 2. American Society of Agronomy, Madison, Wis- consin. Caldwell, M M , D M Eissenstat, J H Richards, and M. F Allen. 1985. Competition for phosphorous: differential uptake from dual-isotope labeled soil interspace between shrub and grass. Science 229: 384-386. Carpenter, A T., and N E We.st 1987. Indifference of mountain big sagebrush growth to supplemental water and nitrogen. Journal of Range Manage- ment 40: 448-451. Carpenter, L H 1972. Middle Park deer study: range fertilization. Game Research Report. Part II. Col- orado Division of Wildlife, Denver. Charley. J. L 1972. The role of shrubs in nutrient cv- cling. Pp. 182-203 in C. M. McKell, J. P. Blais- dell, and J. R. Goodin, eds., Wildland shrubs — their biology and utilization. United States Department of Agriculture, Forest Service Gen- eral Technical Report lNT-1. Charley, J L., and N E West 1975. Plant induced soil chemical patterns in some desert shrub ecosys- tems in Utah. Journal of Ecology 63: 945-964. 1977. Micro-patterns of nitrogen mineralization activity in soils of some shrub dominated ecosys- tems in Utah. Soil Biological Biochemistry 9: 357-366. CoMSTOCK, J. P , T A Cooper, and J R Ehleringer 1988. Seasonal patterns of canopy development and carbon gain in nineteen warm desert shrub species. Oecology 75: 327-335. Depuit, E J , and mm Caldwell 1973. Seasonal pat- tern of net photosynthesis oi Artemisia tridentata. American Journal of Botany 60: 426-435. 1975. Gas exchange of three cool semi-desert spe- cies in relation to temperature and water stress. Journal of Ecology 63: 835-858. Doescher, P S , R F Miller, and A H. Winward 1984. Soil chemical patterns under eastern Oregon plant communities dominated by big sagebrush. Soil Science Society of America Journal 48; 659-663. Eissenstat. D. M., and M M Caldwell 1988. Compet- itive ability is linked to rates of water extraction. A field study of two aridland tussock grasses. Oe- cologia 75; 1-7. Fischer, F M , L W Parker, J P Anderson, and W G. Whitford 1987. Nitrogen mineralization in a desert soil: interacting effects of soil moisture and N fertilizer. Soil Science Society of America Jour- nal 51; 1033-1041. Fischer. F M . J C Zak. G L Cunningham, and W G Whitford 1988. Water and nitrogen effects on growth and allocation patterns of creosotebush in the northern Chihuahuan Desert. Journal of Range Management 41: .387-391. Gutierre.x, J. R , AND W G Whitford 1987. Responses of Chihuahuan Desert herbaceous annuals to rain- fall augmentation. Journal of Arid Environment 12; 127-139. HORNECK, D A . J M Hart, and K Topper 1989. Meth- ods of soil analysis used in the Soil Test Laboratory at Oregon State University, Corvallis. SM 89:4. HiRONARA, M , AND E W TiSDALE 1963. Secondary suc- cession in annual vegetation in southern Idaho. Ecology 44: 810-812. James. D W , and J. J Jurinak 1978, Nitrogen fertiliza- tion of dominant plants in the northeastern Great Basin desert. Pp. 219-231 in N. E. West and J. Skujins, eds.. Nitrogen in desert ecosystems. Dowden, Hutchingson and Ross, Inc., Strouds- burg, Pennsylvania. Lajtha, K . AND M Klein 1988. The effect of varying nitrogen and phosphorus availability on nutrient use by Lanea tridentata. a desert evergreen shrub. Oecology 75; .348-353. Lentz, R. D,. and J H Simonson 1986. A detailed soils inventory and associated vegetation of Squaw Butte Range Experimental Station. Oregon State University Agricultural Experiment Station Spe- cial Report 760. Marschner, H. 1986. Mineral nutrition in higher plants. Academic Press, New York. 674 pp. McDermitt, D K 1987. Photosynthesis measurement systems; performance comparison of the LI-6200 to the LI-6000. LI-COR, Inc. Application Note 6200-1. McGarity, J W 1961. Denitrification studies on some south Australian soils. Plant Soil 14: 1-21. Miller, R F 1988, Comparison of water use by Arte- misia tridentata ssp, ivyomingensis and Chryso- thamntis viscidiflonts ssp. viscidiflorus. Journal of Range Management 41: 58-62. Miller. R F., M. R Haferkamp, andR F Angell. 1990, Effects of season of defoliation on soil water and plant growth: crested wheatgrass. Journal of Range Management (in press). Miller, R F, and L M Shultz 1987. Development and longevity of ephemeral and peremiial leaves on Artemisia tridentata Nutt. ssp, tridentata. Great Basin Naturalist 47; 227-230. Parker. L W , P F Santos, J Phillips, and W G Whit- ford 1984. Carbon and nitrogen dynamics during the decomposition of litter and roots of a Chi- huahuan Desert annual, Lepidium lasiocarpum. Ecological Monographs 54; 339-360. 1990] Efffx;ts ok NniuK;KN Availability 19 SAS. 1988. (General liiifar models [)r()ti'cliirt's. I'p. 549-640 ill SAS users guide: statisties. Release 6.03 edition. SAS Institute, Gary, North C^arolina. SCHOLANDER, P. F.. H T Hammel. E D Bhadsthekt. AND E. A. Hemmincsen 1965. Sap pressure in va.scular plants. Seience 148:339-346. Sneva, F a . AND C M Brhton 1983. Adjustinj^ and forecastinjj; lierhage yields in the interniountain big sagebrush region of the steppe province. Ore- gon State L'liiversity Agricultural Experiment Station Bulletin 659. 61 pp. Steel, R D (J., and J H Torkie 1980. Principles and procedures of statistics. A bionietrical approach. 2d ed. McGraw-Hill, Inc., Hightown, New Jersey. Sturges, D. L. 1977. Soil water withdrawal and root char- acteristics of big sagebrush. American Midland Naturalist 98: 257-273. SVEICAR, T. J., AND J. A. BROWNING. 1988. Growth and gas exchange of Andropogoii gerardii as influ- enced by burning. Journal of Range Management 41: 239-244. TiSDALE, D., AND M HiRONAKA 1981. The sagebrush- grass region: a review of the ecological literature. University of Idaho, Forestry, Wildlife, and Range Experiment Station Bulletin 209. TiSDAI.i;, E \V , M lilRONAKA, AND M A Fo.SRERc; 1969. The sagebrush region in Idaho — problem in range resource management. Idaho Agricultural Experi- ment Station Bulletin 512. 15 pp. Wallace, A. E M Romney, G E. Kleinkopf, and S M SouEL 1978. Uptake of mineral forms of nitrogen by desert plants. In: N. E. West and J. Skujins, eds.. Nitrogen in desert ecosystems. Dowden, Hutchingson and Ross, Inc., Strouds- burg, Pennsylvania. West, N. E. and] Skujins. eds., 1978. Nitrogen in desert ecosystems. Dowden, Hutchinson and Ross, Inc., Stroudsburg, Pennsylvania. Wang, J 1989. Response of Artemisia tridentata ssp. ivtjoiningetisis and Stipa thtirberiana to nitrogen amendments. Unpublished thesis, Oregon State University, Corvallis. Waring, R H , and B. D. Cleary 1967. Plant moisture stress: evaluation by pressure bomb. Science 155: 1252-1254. Received 10 October 1989 Accepted 1 November 1989 Crt-at Basin NatiiralisI 50(11 19W). pp 21-31 FORM AND DISPERSION OF MIMA MOUNDS IN RELATION TO SLOPE STEEPNESS AND ASPECT ON THE COLUMBIA PLATEAU Ceorge VV. Cox Abstract. — Patterned ground consisting of Mima-type earth mounds and associated sorted stone circles and nets is widespread on the Cohnnhia Plateau of western North America. Studies of the geometric relationships of mounds and stone nets to slope aspect and steepness were carried out at the Lawrence Memorial Grassland Preserve, north central Oregon, in June 1987. Mound and moundfield characteristics were sampled on randomly chosen 1-ha plots on slopes of different aspect and steepness. Mounds were largest, most circular and symmetrical in form, and most fully encircled by beds of size-sorted stones on level sites. On slopes of increasing steepness, mounds decreased in size, showed increasing asymmetry and downslope elongation, and became connected into lines oriented up- and downslope. Encircling stone beds became more weakly developed or disappeared on upslope and downslope sides of the moimds, and the lateral beds developed downslope extensions that eventually merged with those of adjacent upslope and downslope mounds. These patterns are interpreted as reflecting changes in the manner of soil translocation by northern pocket gophers, Thomomys talpoidcs, due to their responses to tvuineling on slopes and to the modification of the flow of water across the slope because of the presence of moimds. Mima-type earth mounds are a characteris- tic feature of grasslands with shallow soils or poor drainage in western North America (Cox 1984). These mounds, containing stones up to about 50 mm in diameter, commonly range up to 2 m in height, 20 m in diameter, and 50 ha in density. Recent investigations at several locations have supported the hypothesis that Mima-type mounds are formed over long pe- riods of time by the centripetal translocation of soil toward centers of activity of geomyid pocket gophers. These centers, located ini- tially in the deepest, best drained sites avail- able, are gradually transformed into mounds by soil translocation (Co.x 1984, Cox and Allen 1987b, Cox and Gakahu 1987, Cox et al. 1987). Mima mounds are an extensive and promi- nent feature of the shrub steppe of the Colum- bia Plateau in eastern Washington, northern Oregon, and southwestern Idaho, USA. Here the mounds are frequently encircled by beds of sorted stones, and intermound flats often exhibit polygonal networks of sorted stone beds (Waters and Flagler 1929, Kaatz 1959, Malde 1961, 1964, Fosberg 1965). These fea- tures, formerly interpreted as periglacial fea- tures, have also been interpreted as a result of soil translocation bv pocket gophers (Cox and Allen 1987a). Our previous studies of Mima mounds and sorted stone beds have been conducted on level areas where mounds are circular in form and regular in spacing. Observations by previ- ous workers (cited above) and patterns evi- dent on aerial photographs indicate that mound form and moundfield geometry are modified considerably on slopes. The objec- tive of this study was to define variation in mound form and moundfield geometry with slope aspect and steepness, and to determine if the activities of pocket gophers can account for the variation. Methods Studies were conducted at the Lawrence Memorial Grassland Preserve (LMGP) and on adjacent ranch land of the Friday Brothers Corporation, southern Wasco County, Ore- gon (44°57'N, 120°48'W), 1-11 June 1987. This was the site of previous studies of the structure of mounds and associated beds of sorted stones (Cox and Allen 1987a, Cox et al. 1987). The LMGP, a registered national land- mark owned by the Nature Conservancy, lies Department ol Biology, San Diego State University, San Diego, California 92182-0057 21 22 G. W Cox [Volume 50 at an elevation of 1,036-1,060 m on the Sha- niko Plateau, formed of Columbia River basalts, and includes several ravines that fall steeply northward into the valley of Ward Creek, 122 m below. The preserve has a cold, semidesert climate with an average annual precipitation of 280 mm. The surface of the plateau is mounded "biscuit scabland" with Mima mounds that range up to about 2 m in height and 20 m in diameter. The mound soils are classed as Condon eolian silt loams, and the intermound soils as Bakeoven residual very cobbly loams. The vegetation of mounds and deeper upland soils is dominated by Idaho fescue (Festuca idahoensis) and blue- bunch wheatgrass {Agropyron spicatimi). The shallow intermound soils are dominated by scabland sagebrush {Artemisia rigida). Sand- berg bluegrass (Poa scahrella), several spe- cies of biscuitroot {Lomatium spp.), and bit- terroot {Lewisia rediviva). The northern pocket gopher {Thomomys talpoides ) is abun- dant throughout the preserve. A comprehen- sive physical and biotic inventory of the LMGP is given by Copeland (1980). Relationships of mound and moundfield characteristics to slope and aspect were ex- plored by sampling mounds throughout the LMCP and on a small area of adjacent ranch land. For this sampling, an aerial photo with a superimposed 100 x 100-m grid was used. Some grid units that included very steep slopes or canyon bottoms were largely or entirely unmounded. Since the objective of the study was to examine mound characteris- tics in relation to slope, only mounded grid units (> 50% of the surface showing mound- intermound topography) were considered for sampling. This criterion also assured that the grid units selected were internally uniform in their slope. These grid units, with the aid of a topographic map, were also grouped tenta- tively into five aspect classes: level (with an overall slope less than 2.5°), or north-, east-, south-, or west-facing. All grid units within the 153-ha LMGP were considered. In addi- tion, to allow ade(}uate representation of south-facing slopes, a 10-ha area of land north of Ward Creek was also included. Sets of 7 grid units were chosen by random coordinates in each of the five aspect classes (a total of 35 grid units). The aerial photo was then used to locate these grid units (hereafter termed plots) in the field. The overall aspect of each plot was de- termined with a compass and the slope steep- ness measured with an inclinometer. A count of the mounds in each plot was obtained, and the mound nearest the plot center was desig- nated for detailed measurements. Maximum mound height was measured with a meter stick and line level. The orientation of the long axis of the mound (downslope direction) was measured with a compass. Maximum and minimum diameters were measured with a meter tape, and the components of these di- ameters relative to the highest point on the mound were also recorded (mound top to up- slope edge, top to downslope edge, etc.). Dis- tances to the first and second nearest neigh- bors (between mound high points) were also measured with a meter tape, as was the mini- mum distance between the highest points of the two mounds that were furthest apart in this three-mound set. The fraction of the mound encircled by beds of bare, size-sorted stones (Cox and Allen 1987a) was recorded for the upslope and downslope halves of the mound. The length of stone beds diverging from that surrounding the mound and extend- ing downslope ("tails") was recorded. The maximum length of this measurement was the point at which this "tail" reached another mound. Finally, the number of discrete pocket gopher activity areas was recorded as an estimate of the number of animals occupy- ing the mound. Areas with surface heaps or plugged tunnel openings were considered separate activity areas when they were sepa- rated by more than 5 m. From these measurements a number of descriptive characteristics were calculated. Mound area was calculated assuming that the base was a circle or ellipse, and volume was computed on the assumption that the mound was a segment of a sphere or prolate spheroid. Elongation (El) was computed from the fol- lowing relationship: El = V(a' - b')/a where a and h are the major and minor radii. A second measure of elongation (E2) was also calculated as the ratio of the major to minor diameters of the mound. Asymmetry (AS) was calculated AS = [(1 - SI)- + {l- Ss)~T' where Si and S.v are the asymmetries on the 1990] Mima Mounds 23 Tabi.K 1. C^harac'trristics oi Mima iiiomuls and iiioiincUickls on plots dillerinj;; in slope steepness at the Lawrence Memorial Grassland Preserve, north central Oregon. Values are means ± standard errors. Overall Slope steep ness (°) 0-2.5 2.5 -5.0 5.0-7.5 7..5- -10.0 Characteristic (n - 35) (u - 6) (u 15) (n ^ 8) (n -6) Mound features Hei,uht(cm) 68.7 ± 2.8 81.7 ± 6.8 67.9 ± 4.4 62.9 ± 5.1 65.7 ± 6.5 Area (m') 145.9 ± 14.6 188.0 ± 25.1 174.9 ± 24.7 118.2 ± 26.2 68.0 ± 8.9 Volinue (m ) 52.8 ± 6.2 78.6 ± 13.7 60.6 ± 10.1 40.6 ± 11.2 23.5 ± 5.3 Elongation 1.42 ± 0.11 1.23 ± 0.18 1.46 ± 0.20 1.73 ± 0.20 1.07 ± 0.11 Asymmetry 0.59 ± 0.08 0.14 ± 0.04 0.59 ± 0.08 0.78 ± 0.23 0.77 ± 0.27 Sorted stone beds Upslope side (%) 36.6 ± 5.6 59.2 ± 17.1 35.3 ± 7.8 34.4 ± 13.1 20.0 ± 8.2 Downslope side (%) 25.8 ± 4.9 45.8 ± 16.6 28.0 ± 7.5 17.5 ± 7.0 11.7 ± 7.9 Overall (%) 31.5 ± 4.7 52.5 ± 15.6 31.7 ± 6.3 25.9 ± 9.5 17.5 ± 7.7 Downslope tails (m) 2.17 ± 0.64 0 1.65 ± 0.73 4.89 ± 2.13 2.02 ± 1.11 Moundfield features Density (ha ') 21.6 ± 1.2 18.8 ± 2.3 21.5 ± 1.9 24.4 ± 3.1 21.2 ± 2.9 Connection 0.11 ± 0.03 0.05 ± 0.05 0.18 ± 0.06 0.04 ± 0.04 0.09 ± 0.06 Linearity 0.84 ± 0.02 0.74 ± 0.06 0.88 ± 0.03 0,77 ± 0.06 0.91 ± 0.03 Mounded surface (%) 31.5 35.3 37.6 28.8 14.4 long and short axes, respectively, calculated SI = a'/a Ss = bVb where a ' and b ' are the longer, and a and h the shorter distances from mound edge to peak along the major and minor axes. Connec- tion (C) of the sampled mound to the two nearest mounds on opposite sides was cal- culated as the mean of ratios of heights of between-mound divides to maximum mound height. Linearity (L) of alignment of the sam- pled mound and its two nearest neighbors was determined as the ratio of the sum of distances measured from center to center for a series of mounds to the single straight-line distance between the first and last in the sequence. Data on mound and moundfield character- istics were analyzed by BMDP statistical soft- ware procedures (Dixon 1983). Logarithmic and arc-sine transformations were employed to achieve normality for some variables. Means and standard errors were computed for plot data grouped by slope aspect and slope steepness classes, and these classes were then compared by t-tests and ANOVA (BMDP ID and 7D). Correlations among all combinations of variables and stepwise linear regressions using selected variables as the dependent variable were also performed (BMDP 2R). Plant names follow Hitchcock and Cron- quist (1973). Results Measurements of the overall slope aspect in the field revealed that several of the plots tentatively placed in particular aspect classes actually fell in other classes. As a result, the final set of samples comprised 10 N-facing, 5 E-facing, 5 S-facing, and 8 W-facing plots (Table 1). Mound and moundfield characteristics showed a complex relationship to slope aspect and steepness. With respect to slope steep- ness, mounds at the LMGP were confined to slopes less than about 10° in steepness. Mounds exhibited greatest height, basal area, and volume on level areas or very gentle (< 2.5°) slopes (Table 1). Basal area and vol- ume declined progressively with increased slope steepness (ANOVA, F33, = 3.62, P < .05 for area; F = 3.29, P < .05 for vol- ume). The mean height of mounds on level to very gently sloping areas was significantly greater than on steep (5.0-7.5°) slopes (t = 2.26, DF = 12, P < .05). Volume, the best overall indicator of mound size, showed a strong negative correlation (r = —.496, P < .01) with slope steepness (Fig. 1). Mound asymmetry and elongation were both greatest on steep (5.0-7.5°) slopes (Table 1). Asymmetry was significantly lower for mounds on level to very gently sloping 24 G. W. Cox [Volume 50 rr Nearest Neighbor Distance Volume Slope Steepness Elongation Overall Down- slope Up- slope Rock Ring Development Connection Linearity Slope Aspect Density Mound Axis Direction Asymmetry P<0.05 P'0.01 P<0.001 Fig. 1. Correlation relationships between major \arial)les of mound and moundlield geometr> tor plots sampled on slopes of differing steepness and aspect at the Lawrence Memorial (Grassland Preserve, north central Oregon, June 1987. (0-2.5°) areas than on gentle (2.5-5.0°) slopes (t = 3.30, DF - 19, "P < .01), steep slopes (t = 2.35, DF = 12, P < .05), or very steep (7.5-10.0°) slopes (t = 2.33, DF = 10, P < .05). Asymmetry was also strongly corre- lated (r .449, P < .01) with slope steepness (Fig. 1). Elongation, defined as the ratio of long to short mound axis, was significantly greater on steep slopes than on level to very gently sloping areas (t 3.07, DF 12, P < .01). Elongation was significantly greater on steep than on very steep slopes (t —- 2.56, DF= 12, P< .05). The development of encircling beds of bare, size-sorted stones was strongest on level to very gently sloping plots (Table 1), the de- gree of overall development being negati\el\ correlated (r = -.343, P < .05) with slope steepness (Fig. 1). The negative correlation between the development of sorted stone beds and slope steepness was stronger on the downslope side of mounds (r = —.367, P < .05) than on the upslope side (r = .304, P> .05). Moimdfield characteristics showed weaker patterns. Density of mounds showed little variation with slope steepness (Table 1). Con- nection and linearity, which were positively correlated (r - .343, P < .05), exhibited highest values on steeper slopes. Linearity was significantly lower on le\ el to very gently sloping (0-2.5°) plots than on plots with gentle (2.5-5.0°) slopes (t 2.29, DF 19, P < .05) or very steep (7.5-10.0°) slopes (t - 2.28, DF ^ 10, P < .05). The combination of mound size (basal area) and densit\ resulted in a greater surface coverage i)\ mounds on very 1990] Mima Mounds 25 TaIU.K 2. (^hiirac'tL'ristics oiMiiiia iiiouiuls and iiiouiulliclcis on plots diili'iin^i in aspect at the l^awrt'ntc Mt'niorial Grassland Preser\t', north central OTctjon. Valui's are means ± standard errors. Overall Aspect Le\el North East South \V est Charaeteristie (n = = 35) (n 7) (n 10) (n 5) (■> 5) (n -H) Mounds Height (cm) 68.7 ± 2.8 80.7 ± 5.8 75.6 ± 4.4 60.0 ± 8.0 54.2 ± 5.4 63.9 ± 5.2 Area (m') 145,9 ± 14.6 172.3 ± 26.4 146.6 ± 24,5 117,1 ± 42,3 121.0 ± 42.4 155.2 ± 39.2 Volume (m ) 58,2 ± 6.2 71.5 ± 13.6 57.1 -+- 11,1 40.7 ± 18.3 30.7 ± 8.4 52.1 ± 15.6 Elongation 1.42 ± 0. 1 1 1.10 ± 0.20 1.32 ± 0.20 1.72 ± 0.32 1.60 ± 0.18 1.51 ± 0.28 As\iniuetr\' 0.59 ± 0.08 0. 15 ± 0.03 0.51 + 0.08 1.00 ± 0.34 0.66 ± 0..30 0.75 ± 0.14 Rock circles llpslope side (%) .36.6 ± 5.6 50.7 ± 16.7 .33.5 -¥■ 11,5 41,0 ± 18,6 38,0 ± 10.7 24.4 ± 6.2 Downslope side {%) 25.8 ± 4.9 39.3 ± 15.4 20.5 + 10,5 20.0 ± 10.5 25.0 ± 12.2 25.0 ± 11.3 Overall (%) 31.5 ± 4.7 45.0 ± 15.2 27.0 + 9.1 30.5 ± 13.5 31.5 ± 10.3 25.9 ± 6.1 Rock tails (m) 2.17 ± 0.64 0 1.95 ± 1.06 4.64 ± 3.25 3.28 ± 1.63 2.11 ± 1.03 Moundfield features Density (ha ') 21.6 ± 1.2 17.8 ± 2.2 20.0 ± 1.3 23.6 ± 4.1 26.8 ± 5.0 22.5 ± 2.6 Connection 0.11 ± 0.03 0.09 ± 0.06 0.20 ± 0,09 0,06 ± 0,06 0.08 ± 0.08 0.06 ± 0.04 Linearit) 0.84 ± 0.02 0.78 ± 0.06 0.91 ± 0,03 0,75 ± 0,09 0.82 ± 0.04 0.86 ± 0.04 Mounded surface (%) 31.5 .30.7 29.3 27,6 .32,4 .34.9 gentle to gentle slopes than on steep to very steep slopes. Data for mounds grouped by slope aspeet (Table 2) indicate that moinid size (height, basal area, volume) was greatest on level plots and ne.xt greatest in height and volume on north-facing slopes. Mounds were smallest in both height and volume on south-facing slopes and next smallest on east-facing slopes. The variation among heights on plots differing in aspect was significant (ANOVA, F4 3,, = 3.48, P < .05). Mean volume on south-focing slopes was significantly less than on level sites (t = 2.31, DF = 10, P < .05). Mounds were least elongate or asymmetric on level and north-facing plots, and most elon- gate and asymmetric on east-facing slopes (Table 2). Elongation, expressed as the ratio of longer to shorter axis, was significantly greater for east- and south-facing plots than for level plots (t = 3. 14 and 2. 76, respectively, for east- and south-facing plots, DF = 10, P < .05). Variation in asymmetry was significant among aspect groups (ANOVA, F430 = 3.29, P < .05). Elongation of mounds was closely parallel to slope, the long axis of the mound being very highly correlated with the slope direction (Fig. 3,' r = .822, P < .001). The development of sorted stone circles differed little for slopes of differing aspect (Table 2). Downslope tails of sorted stones were noted on slopes of all aspects. Mound density was positively correlated with slope aspect, expressed as deviation in degrees from north (r = .405, P < .05). Den- sity ranged from 17.8 mounds ha ' on level plots to 26.8 mounds ha on south-facing slopes (Table 2). Linearity of a sample mound and its two nearest neighbors was greatest for north-facing and least for east-facing slopes. Linearity was significantly greater for north- facing slopes than for level plots (t = 2.26, DF = 15, P<.05). Several other important relationships were not directly linked to either slope steepness or slope aspect (Fig. 1). A number of these cen- tered on nearest neighbor distance and stone circle development. Nearest neighbor dis- tance was positively correlated with mound volume (r = .349, P < .05). In addition, nearest neighbor distance showed a strong, direct correlation with mound elongation (r = .449, P < .01) and stone circle develop- ment (r = .468, P < .01). Furthermore, the correlation of nearest neighbor distance to development of the stone circle on the downslope side of mounds was very strong (r = .641, P < .001). Mound volume also showed a direct relationship to stone circle development, both overall (r = .346, P < .05) and on the downslope side (r ^ .415, P < .05). The more elongate a moimd, the greater was the development of the stone circle on its downslope side (r == .379, P < .05). The 26 G W. Cox [Volume 50 greater the density of mounds, however, the poorer was the development of the stone cir- cle on the downslope side of the mound (r = — .423, P < .05). Linearity of mound arrange- ment, on the other hand, was negatively re- lated to development of the stone circle, both overall (r = —.339, P < .05) and on the up- slope side (r = —.335, P < .05). Finally, dens- ity showed a positive correlation with asym- metry (r = .364, P < .05), and mound a.xis direction was negatively correlated with linearity of mound arrangement (r = —.335, P< .05). Discussion The major patterns of variation of mound and moundfield characteristics with slope steepness and aspect are listed below. 1. Mounds show maximum size, circular and symmetrical form, low connection and linearity, and well-developed stone circles on level sites. 2. On slopes, mounds become smaller and more elongate and asymmetrical, with the long axis parallel to the slope, and show greater connection and linearity of arrange- ment. 3. On slopes, stone circles become weaker, especially on the downslope side of mounds, and stone beds diverge to form downslope tails. 4. Slope effects are, in general, more in- tense on south- and east-facing slopes than on north- and west-facing slopes (except for con- nection, which tends to peak on gentle north- facing slopes). Much variation exists in the literature con- cerning the steepness of slopes on which mounds occur. Waters and Flagler's (1929) data on the Cohmibia Plateau and nearby areas record mounds on slopes up to only 6° in steepness, and Kaatz (1959) stated that mounds occur on slopes up to about 7° in steepness. Brown (1951), however, reported that mounds in this region occur on slopes up to 35-45°. Vitek (1973) reported mounds in southern Colorado on mountain slopes up to 20° in steepness, and in southern California, Cox (1984) found mounds on slopes up to 30° in steepness. Price (1949) stated that mounds occur in the western states in mountain mead- ows with slopes up to 20-30°. This variation in maximum steepness of mounded slopes may be real and may relate to soil texture and other factors affecting vulnerability to erosion. Mounds may occur only on slopes of 20° steep- ness or greater when soils are rich in clays, as they are in many locations in southern Califor- nia (Cox 1984). Researchers also offer diverse statements on how mound shape varies with slope steep- ness. ScheflFer (1958) states that Mima mounds are "generally circular in shape as seen from above, regardless of slope. Vitek (1973), in southern Colorado, found that mounds tend to be nearly circular even on slopes up to 10.4° in steepness. On the Columbia Plateau, how- ever, mounds are usually described as being more elongate on slopes. In southern Wash- ington and northern Oregon, Waters and Flagler (1929) reported that the mean ratio of major to minor axes increases with increasing slope to a value of 1.43 on slopes of 6° steep- ness. The long axis of these mounds is said to be parallel to the slope. Malde (1964) noted that mounds in southwestern Idaho are typi- cally circular, but that on hillsides they are noticeably elliptical, with the long axis di- rected downslope. Kaatz (1959), in central Washington, found that the typical mound is elliptical in shape, with a ratio of major to minor axis of about 1.41, the long axis being parallel to the slope. Olmsted (1963) found that mounds in eastern Washington are often elongate, the ratio of major to minor axis being 1.1-1.5. He also stated that the long axis is aligned with prevailing winds and is some- times across slopes rather than parallel to them. Fosberg (1965) stated that in Twin Falls County, Idaho, mounds elongate into down- slope stripes. A degree of connection, or confluence, of mounds and their alignment into rows parallel to the slope has been noted b\' several work- ers. Waters and Flagler (1929)', Malde (1964), and Fosberg (1965) describe mounds on the Cohnnbia Plateau as forming beadlike rows along small drainage di\ ides or between stone stripes on steeper slopes. Perhaps the best overall description of this pattern, together with that of mound form, is given by Brown (1951) for sites near Maupin, Oregon: On the stft'iicr slopes they are oriented in more or less l^arallt'l lines aloni^ the rill divides, tend to be elongate and coalesce and are not as high nor as perfectly kept up as on the level. Looking at these slopes from a distance 1990] Mima Mounds 27 or stiuKing afiial photos, one ijains tlu' iiiipri'ssion ol a continuous uiouiul clown the slope as thounh it consti- tuted the entire rill divide. A close inspection, how- ever, reveals that the crest of tlie strip is not even, that it is divided 1)\' well iornied mounds rising ai)()ve the level of the siurounding soil. Fewer authors de.scrihe changes in the arrangement of sorted stone beds as slope steepness increases. Malde(1961, 1964) states that in southwestern Idaho the stone pave- ments surrounding mounds change progres- sively to parallel stone stripes running up- and downslope, implying the disappearance of pavement sections on the up- and downslope edges of mounds. Kaatz (1959), stating that stone circles and networks change to sorted stone stripes on steep slopes, also notes that sorted stripes may occur without any upslope connection to the former features. Brunn- schweiler (1962) describes a similar pattern and diagrams a configuration in which stone "tails" arise from stone rings encircling mounds, or from intermound polygonal net- works, to extend downslope. Pyrch (1973) noted that sorted stone stripes occur on slopes up to a maximum steepness of 15-33°. At the LMGP, Cox and Allen (1987a) found that on level areas the development of stone circles is directly correlated with mound size, and that on slopes the initial pattern of modification is the weakening of the bordering bed on the downslope side of the mound and the diver- gence of downslope "tails." Are these mound features compatible with the basic hypothesis of origin of both mounds and associated stone circles, polygonal nets, and stripes by the soil translocation activities of pocket gophers? And if so, how does this mechanism interact with site characteristics related to steepness and aspect to yield the observed patterns? Let us first consider the implications of the relationship of the intensity of moundward soil translocation by pocket gophers to dis- tance from the center of a small mound and elevation below its top, as observed by Cox and Allen (1987b). For a mound on a level site, average moundward translocation increased with distance from the mound center, and average upward translocation increased with elevation below the mound top. On a level site these tendencies would be distributed symmetrically, other factors being equal, and the mound would tend to enlarge symmetri- cally, maintaining a circular shape. For a similar small mound on a slope, how- ever, differences in translocation would result even if the amount of tunneling activity re- mained the same in terms of distance and direction Irom the movmd center. On the sides of the mound lying on the slope contour, moundward and upward translocation will be similar to soil movements on a moimd on level ground. On the downslope side of the mound, however, an animal must move soil a greater vertical distance to achieve the same horizon- tal displacement. Since this requires greater energy expenditure, horizontal displacement will probably often be less than expected. On the upslope side of the mound, in contrast, a given horizontal displacement will occur with less vertical displacement. In some cases, of course, much of the actual horizontal dis- placement will be downslope. Thus, expendi- ture of the same energy will lead to a greater than expected horizontal displacement. The consequence of differences in mean displacement distance is that more soil will be translocated onto the upslope side of the mound than onto the downslope or lateral sides. The mound should thus grow most in a lateral and upslope fashion. However, this growth could permit a circular form to be retained as long as the average of upslope and downslope addition rates equals the addition rates to the lateral edges of the mound. Such a pattern will prevail whenever the mean hori- zontal translocation distance of soil at a given distance from the mound center is linearly related to the mean slope of the translocation path (Fig. 2). If the relationship of mean displacement distance to slope is curvilinear and convex, however, then additions to the lateral edges will be greater than expected, and to the up- slope and downslope edges less than ex- pected; thus, the mound will expand in width (across the slope). If the relationship is curvi- linear and concave, additions to the upslope and downslope edges will be greater than ex- pected, and those to the lateral edges less than expected; the mound will elongate up- and downslope. In the latter case, the total amount of soil translocated onto the down- slope side of the mound will still be much less than that moved onto the upslope surface. As the mound grows in height, addition to the downslope side of the mound will also de- cline. At the same time, slope conditions on 28 G W Cox [Volume 50 140 120 r 100 80 60 40 20 I - * 1 \ 1 \ I 1 - > 1 1 1 1 !--• 1 10 20 30 40 50 Angle of Slope (°) 60 70 Fig. 2. Horizontal displacement distance (mound- ward) for soil mined by Thomomys hottac in southern California in relation to slope for data from Cox and Allen (1987b). the upslope side oi the mound will still permit heavy translocation onto the mound surface. Thus, slope relationships should cause ero- sional loss of soil to balance moundward translocation sooner on the downslope side of the mound than on the upslope side. Upslope growth should therefore continue after down- slope growth has stopped. Data from soil translocation studies of Thomomijs bottae in southern California (Cox and Allen 1987b) suggest that soil transloca- tion by pocket gophers of this genus varies with steepness in a curvilinear, concave fash- ion over a range from less than 5° to more than 50° (Fig. 3). These data were obtained in stud- ies of soil translocation on level sites, where the only slope was that of the mounds them- selves. Elongation of mounds on the Columbia Plateau should be coupled with an upslope movement of the mound high point. At maxi- mum size, the highest point of an elongated mound should thus be nearest its upslope end. Elongation of mounds should lead to con- nection with adjacent moimds up- and downslope when the soil mantle is deep enough to permit the development of large mounds. Such connection may create a linear, "beaded" arrangement of mounds parallel to the slope. This arrangement will give a strong measure of linearity if the two nearest neigh- bors of a given mound are upslope and downslope in the connected line. Strong lin- earity of arrangement on a slope would thus imply that the mean distance between mounds in the same line is less than that be- tween mounds in different lines. This should not be the case if uniformly spaced mounds develop on a slope and those directly up- and downslope from each other become con- nected. Our data show only a weak relation- ship between slope steepness and linearity, suggesting that little more than the connec- tion of mounds lying up- and downslope from each other has occurred. The presence of a mound on a slope would modify the flow of water across the slope, concentrating it along the upper and lateral sides of the mound and producing a dry shadow downslope (Cox and Allen 1987a). The concentrated flow of water along the sides of a mound would tend to continue di- rectly downslope. These areas of maximum wetness favor the formation of sorted stone beds extending downslope. Several possible mechanisms may contribute to the transfor- mation of beds encircling the mounds into elongate stripes paralleling them. Growth of flesh\ -rooted plants may lead to extensive tunneling by pocket gophers in these areas. The collapse of deep tunnels and the down- ward settling of soil and small stones during the wet season may thus sort and expose stones near the surface (Cox and Allen 1987a). Erosion may also play an important role, par- ticularly as slope steepness increases. Figure 4 diagrams a hypothesis of how cir- cular, isolated mounds surrounded by sorted stone nets become transformed into elongate, interconnected mounds bordered by linear beds of sorted stones. This hypothesis pre- dicts that lines of mounds on slopes should be separated by two stone stripes, one represent- ing the fusion of downslope extensions of the encircling beds of each line of mounds. Exam- ination of aerial photos of sloping areas on the LMGP shows that this is generally true. These observations strongly support the over- all hypothesis that pocket gophers interact with ph\ sical conditions and processes to pro- duce the dislinctixc patterns of mounds and sorted stone beds on the (>olumbia Plateau. 1990J MimaMoi'nds 29 ++ C 0 "O c i2 c o ■co o o o w iS c w CO /-\ CO ■o c o + ^■e/( 0 Downslope T Slope Steepness Steep Upslope Downslope Edge Lateral Edges Upslope Edge Fig 3 (A) Possible relationships between mean horizontal translocation distance (moundward) and slope steepness. (B) Volume of soil moved onto mound surface at various points around mound perimeter by pocket gopher translocation in relation to slope steepness, based on the relationships outlined in (A) for a mound oi a given slope. 30 G W. Cox [Volume 50 Fig. 4. Diagrammatic model of the transition of unconiu'ctfcl, circular mounds with encircling stone rings on level sites to linearly connected, elongate mouTids bordered by stoTie stripes on steeji slopes. 1990] Mima Mounds 31 Ackn{)\vij<:d(;mknts I thank Elizahetli Lucas for assistance' in field studies and in laboratory analyses oi mound and nioundiield geometry and Annan Friday for permission to conduct much of the work on areas of the Friday Brothers Ranch adjacent to the LMGF. Donald B. Lawrence gave advice and encouragement throughout the study. J. Hunt, J. R. Mackay, and V. B. Scheffer provided valuable comments on an earlier draft of the manuscript. This study was carried out under a contract with the Oregon Chapter of the Nature Conservancy and was supported in part by grant INT-8420336 from the U.S. National Science Foundation. Literature Cited Brown. H. C 1951. Mound inierorelief of tlie Columbia Plateau and adjacent areas. Unpublished manu- script. .36 pp. Brunnschweiler, D 1962. The periglacial realm in North America during the Wisconsin glaciation. Biuletyn Peryglacjalny 11: 15-27. CoPELAND. W N 1980. The Lawrence Memorial Grass- land Preserve: a biophysical inventory with man- agement recommendations. Unpublished report (revised 1983), Oregon Chapter, Nature Conser- vancy, Portland. Cox. G. W. 1984. The distribution and origin of Mima mound grasslands in San Diego Countv, Califor- nia. Ecology 65: 1.397-1405. Cox. G. W.. AND D W Allen. 1987a. Sorted stone nets and circles of the Columbia Plateau: a hypothesis. Northwest Science 61: 179-185. 1987b. Soil translocation by pocket gophers in a Mima moundfield. Oecologia 72: 207-210. Cox, G. W.. andC G. Gakahu 1986. A latitudinal test of the fossorial rodent hypothesis of Mima mound origin in western North America. Zeitschrift fiir Geomorphologie .30: 485-501. Cox, G. W , C. G. Gakahu, and D W Allen 1987. The small stone content of Mima mounds in the Co- lumbia Plateau and Rocky Mountain regions: im- plications for mound origin. Great Basin Natural- ist 47: 609-619. Dalouest, VV W . and V B Sciikkkkk 1942. The origin ol the Mima mountls of western Washington. Jour- nal of (icology ,50: 68-84. Dixon. \V J 198.3. BM DP statistical software. University oi (^aliiorTiia Press, Berkeley. FosBERC M A 1965. Characteristics and genesis of pat- terned ground in Wisconsin time in a chestnut soil in southern Idaiu). Soil Science 99: 30-37. llnciicucK, C. L , AND A. CRONQt'iST. 1973. Flora of the Pacific Northwest. University of Washington Press, Seattle. Kaai:z;, M R 1959. Patterned ground in central Washington; a preliminary report. Northwest Science ,33: 145-1.56. Malde, H E 1961. Patterned ground of possible solifluc- tion origin at low altitude in the western Snake River Plain, Idaho. United States Geological Sur- vey Professional Paper 424-B, 170-173. 1964. Patterned ground on the western Snake River Plain, Idaho, and its possible cold-climate origin. Geological Society of America Bulletin 75: 191-207. Olmsted, R K 1963. Silt mounds of Missoula flood sur- faces. Geological Society of America Bulletin 74: 47-.54. Price, W. A. 1949. Pocket gophers as architects of Mima (pimple) mounds of the western United States. Texas Journal of Science 1: 1-17. Pyrch. J B 1973. The characteristics and genesis of stone stripes in north central Oregon. Unpublished the- sis, Portland State University, Portland, Oregon. Scheffer, V B 1947. The mystery of the Mima mounds. Scientific Monthly 65: 283-294. 1958. Do fossorial rodents originate Mima-type microrelief? American Midland Naturalist 59: ,505-510. ViTEK, J D. 1973. The mounds of south-central Colorado: an investigation of geographic and geomorphic characteristics. Unpublished dissertation, Okla- homa State University, Stillwater. Waters, A C , andC. W Flagler 1929. Origin of small mounds on the Colimibia River Plateau. American Journal of Science 18: 209-224. Zar, J H 1974. Biostatistical analysis. Prentice-Hall, Englewood Cliffs, New Jersey. Received 10 June 1989 Accepted 30 November 1989 Great Basin Naturalist .WU), 1990, pp. 33-39 ESOX LUCIUS (ESOCIDAE) AND STIZOSTEDION VITREUM (PERCIDAE) IN THE GREEN RIVER BASIN, COLORADO AND UTAH Harold M. Tyiis' and Jiiincs M. Bfard' " Abstract. — Northern pike, Esoxlucius, stocked in the Yampa River in 1977, invaded the mainstream Cheen River by 1981 and subsequently increased in range and abundance. The speed of this invasion is inthcated by two recaptured pike that moved 78 and 110 km, respectively, downstream in about one year. Pike stomachs (n = 123) were usually empty (54.5%), but some contained fish (43%) and nonhsh items (2.4%). Red shiner, Notropis lutrensis, and fathead minnow, Pimephales promdas, predominated among the 12 fish species eaten. Walleye, Stizostccliun vitreum, presumably introduced to the Green River drainage in the 196()s, was widely distributed but low in abundance. Most of 61 adult walleye stomachs contained food (60.7%); of 6 fish species eaten, channel catfish, Ictalurus punctatus, and fathead minnow were most frequently consumed. Northern pike and walleye were captured in habitats occupied by endangered Colorado River fishes, particidarly Colorado squawfish, Ptijchocheilus lucius. Fredation on endangered fishes was not detected, but northern pike and walleye consmned at least three other native fishes. The northern pike may pose a threat to endangered fishes due to its population expansion, piscivory, and resource sharing. Diets of northern pike and walleye species should be further evaluated if their abundance increases. Northern pike were introduced into Elk- head Reservoir, an impoundment on the Yampa River drainage, in 1977 (P. J. Mar- tinez, personal communication) and collected in the mainstream Yampa River as early as 1979 (E. J. Wick, personal communication). Their numbers increased in the upper Yampa River in the early 1980s (Wick et al. 1985), and a downstream movement into the Green River was subsequently documented in 1981 (Tyus et al. 1982, Green River fishery investi- gations). Northern pike reproduction has been reported in the upper Yampa River drainage, where it has access to the main- stream river (T. P. Nesler, personal commu- nication). Walleye presumably accessed the main- stream Green River by moving downstream from various tributaries. The fish was first reported in Utah in 1951 (Sigler and Miller 1963), and reproducing populations of wall- eye were established by fish stockings in Duchesne River reservoirs (Fig. 1) in the 1960s and 1970s (G. M. Davis, personal com- munication). The Green River basin of Colorado and Utah is an important recovery area for four rare and endangered Colorado River fishes (reviewed by Joseph et al. 1977, Carlson and Carlson 1982, U.S. Fish and Wildlife Service 1987). However, over 20 nonnative fishes have been introduced into the basin for sport, forage, food, or by accident (Tyus et al. 1982, Fishes of upper Colorado). Impacts of these introduced fishes on the native fauna are not well understood, but the presence of two large piscivores, northern pike, £.sc».v lucius, and walleye, Stizostedion vitreum, in areas presently occupied by endangered fishes, is cause for concern. Control of nonnative fishes has been identified as a recovery measure under provisions of an interagency recovery program for endangered fish species in the upper Colorado River basin (U.S. Fish and Wildlife Service 1987). Fish introductions in other locations have eliminated or partially extirpated native fish faunas, and the instabil- ity of resultant communities has caused man- agement problems (Moyle et al. 1986). The purpose of this study was to determine diets of northern pike and walleye in the Green River, and to evaluate the degree of predation on native and endangered fishes. We also doc- ument the recent invasion of northern pike into the Green River basin, and the abun- dance and distribution of northern pike and walleye in the mainstream Green River. The results ol this study are interpreted relative 'U.S. Fish ami Wildlitc Service. 1(>S()W. Hiijliwav 40, Vernal. Utah 84078. -Present address; 1.361 \ernon St., Eureka, California 9.5.501, 33 34 H. M.TyusandJ. M. Beard [Volume 50 t\ N WYOMING Flaming Gorge Reservoir "^4 '^^. *»«/. UTAH «. 75 km/year) in the Yampa River. Long-distance upstream and downstream movement of radiotagged northern pike has also been reported by T. P. Nesler (personal communication). The majority of fishes consumed by north- ern pike in this study were soft-rayed forms (Table 2), as previously noted by others (Beverle and Williams 1968, Weithman and Anderson 1977, Wolfert and Miller 1978). Channel catfish, the only spiny-rayed fish consumed, was found in two stomachs. We could not positively identify roundtail chub, Gila robusta, in northern pike stomachs taken from the Yampa River, but presumably one Gila spp. was a roundtail chub. T. P. Nesler (personal communication) reported that roundtail chub were present in northern pike stomachs he examined from the Yampa River. Most of the pike we examined were from the Green River where roundtail chub are rare (Tyus et al. 1982, Fishes of upper Colorado), and this may have resulted in the relative absence of roundtail chub as prey in pike stomachs we examined. Northern pike may spawn in the main- stream Green River, but if so, recruitment is low. We did not capture small northern pike (< 321 mm TL) in this study, and, to our knowledge, pike reproduction has not been noted by others. However, one 115-mm-TL specimen was seined by HMT and others from a shoreline area of the Green River in Dinosaur National Monument on 8 July 1988. It is not known whether this fish hatched in the Green River or was transported there from another location. Also, we captured sev- eral ripe female pike, and it is possible that some of these fish spawned in the Green River. Most ripe female pike (76%) had empty stomachs, suggesting a reduction in feeding activity with increasing water temperatures and ripening ovaries (Frost 1954, Lawler 1965). Walleye were rare in the Green River, and their long period of residency suggests that their numbers will probably not increase. Walleye were easily captured by electrofish- ing, and very few fish that we sighted escaped capture. However, it was difficult to capture northern pike with electrofishing, and many fish escaped. A direct comparison of the rela- tive abundance of walleye with that of north- ern pike could be somewhat misleading, and it is noted that walleye were more rare, and northern pike more abundant, than indicated by electrofishing catch rates. We captured 38 H M TyusandJ M Beard [Volume 50 only one female walleye with developed ovaries, and that was in May at a water tem- perature of 13 C. Walleye in other locations usually spawn at cooler water temperatures (3.3-7.2 C, Sigler and Miller 1963; 5.6-11.1 C, Scott and Grossman 1973). No small wall- eye (< 395 mm TL) were captured in this study. Young of the endangered humpback chub, Gila cypha; bonytail chub, G. elegans- razor- back sucker, Xyrauchen texaniis; and Colo- rado squawfish, Ptychocheilus luciiis, may be potential prey for northern pike and walleye. None of these fishes were identified in stom- achs of northern pike or walleye, but our abil- ity to detect such predation was constrained by a small sample size of stomachs that con- tained food, rarity of endangered fishes, and inability to identify all of the fishes eaten. Sympatry of adults of northern pike, wall- eye, and endangered fishes is a cause for con- cern, particularly if resource sharing occurs during periods of limited availability. We col- lected northern pike, walleye, and Colorado squawfish in similar shoreline habitats in the mainstream Green River; in addition, radio- tagged northern pike and Colorado squawfish were syntopic in the Green and Yampa rivers (Valdezand Masslich 1989, Wick and Hawkins 1989). Northern pike were captured in shal- low, flooded habitats also utilized by razor- back sucker. Stocking programs for northern pike and walleye have been discontinued by state agencies in Colorado and Utah (G. M. Davis and P. J. Martinez, personal communication), and the relative absence of small fish of both species suggests that reproduction in the mainstream Green River is low or nonexistent during most years. However, the continuing invasion of northern pike and walleye into the Green River from established, reproducing stocks should be monitored, and their interac- tions with endangered fishes further evalu- ated until it can be more clearly demonstrated that competition or predation on endangered fishes does not occur or pose a serious threat. The increasing abundance and spread of northern pike, the diversity of fishes con- sumed, and its syntopy with endangered fishes make this voracious piscivore a poten- tial threat to endangered Colorado River fishes. Acknowledgments This work was funded by the Bureau of Reclamation and the Fish and Wildlife Ser- vice. Numerous Fish and Wildlife Service employees assisted with data collection. We thank S. R. Cranny and M. Moretti of the Utah Division of Wildlife Resources and R. A. Valdez of BIO/WEST Incorporated for pro- viding northern pike and walleye stomachs. G. B. Haines and D. Moses aided in data summarization and word processing. C. A. Karp, P. Pister, and an anonymous reviewer improved an earlier draft manuscript. Literature Cited Beyerle, G B , AND J E Williams 1968. Some observa- tions of food selectivity by northern pike in aquaria. Transactions of the American Fisheries Society 97; 28-.31. Carlander, K D 1969. Handbook of freshwater fishery biology. Vol. 1. Iowa State University Press, Ames. 7.52 pp. Carlson, C A . and E. M Carlson 1982. Review of selected literature on the upper Colorado River system and its fishes. Pages 1-8 in W. H. Miller, H. M. Tyus, andC. A. Carlson, eds., Fishes of the upper Colorado River system: present and future. American Fisheries Society, Bethesda, Maryland. 131 pp. Cook. M F , and E P Bergersen 1988. Movements, habitat selection, and activity periods of northern pike in Eleven Mile Reservoir, Colorado. Transac- tions of the American Fisheries Society 117: 495-502. Frost, W. E 19.54. The food of pike, Esox lucius L., in Windermere. Journal of Animal Ecology 23: 339-360. HOLDEN, P B , AND C B Stalnaker. 1975. Distribution and abundance of mainstream fishes of the middle and upper Colorado River basins, 1967-1973. Transactions of the American Fisheries Society 104:217-231. Joseph, T W , J A Sinning, R J Behnke, and P B HoLDEN 1977. An evaluation of the status, life history, and habitat requirements of endangered and threatened fishes of the upper Colorado River .system. FWS/OBS-77-62. U.S. Fish and Wildlife Service, OfTice of Biological Services, Fort Collins, Colorado. 194 pp. Lawler, G H 1965. The food of the pike, Esox lucius, in Hemming Lake, Manitoba. Journal of the Fish- eries Resource Board of Canada 22: 1357-1377. Miller, R B 1948. A note on the movement of the pike, Esox lucius. Copeia 1948: 62. MOYLE. P B , H W Li, and B A Barton 1986. The Frankenstein effect: impact of introduced fishes on native fishes in North America. Pages 415-426 in R. H. Stroud, ed.. Fish culture in fisheries management. American Fisheries Society, Be- thesda, Maryland. 481 pp. 1990] Green Riveh Fish Ecology 39 Ross, M J , andJ D WiNTKK 1981. Winter movements of four fish species near a thermal plume in northern Minnesota. Transactions of the American Fish- eries Society 10: 14-18. Scott, W B . and E J Grossman 1973. Freshwater fishes of Canada. Bulletin of the Fisheries Re- search Board of Canada 184: 1-966. SiGi.ER, W F , AND R R MiiXF.R 1963. Fishes of Utah, Utah State Department of F'ish and Game, Salt Lake City. 203 pp. Tyus, H M , B D BuRDiCK, R A Valdez, G M Haynes, T A Lyti.e. and G R Berry 1982, Fishes of the upper Colorado River Basin: distribution, abun- dance, and status. Pages 12-70 in W. H. Miller, H. M. Tyus, andC. A. Carlson, eds., Fishes of the upper Colorado River system: present and future. American Fisheries Society, Bethesda, Maryland. 131 pp. Tyus, H M., G. W McAda, and B D Burdick 1982, Green River fishery investigations: 1979-1981. Pages 1-99 in W. H. Miller, ]. J. Valentine, D. L. Archer, H. M. Tyus, R. A. Valdez, and L. Kaed- ing, Colorado River Fishery Project, final report, field investigations. Part 2. U.S. Fish and Wildlife Service and U.S. Bureau of Reclamation, Salt Lake City, Utah. 365 pp. US Fish and Wii.ni.iiK, Service 1987, Recovery imple- mentation program for endangered fish species in the upper (Jolorado River Basin, U,S. Depart- ment of the Interior, Fish and Wildlife Service, Division of Endangered Species, Denver, Colo- rado. [Six sections, various pagination, ] Valdez, R A , and W J Masslich 1989, Winter habitat studv of endangered fish — (Jreen River. Report 136-2, BIO/WEST Incorporated, Logan, Utah. 184 pp, Weitiiman, A S , AND R O Anderson 1977. Survival, growth, and prey of Esocidae in experimental sys- tems. Transactions of the American Fisheries So- ciety 106: 424-430, Wick, E J , and J A Hawkins 1989, Colorado scjuawfish winter haliitat study, Yampa River, Colorado, 1986-1988, Final report. Contribution 43, Larval Fish Laboratory, Colorado State University, Fort Collins, Colorado. 96 pp, WoLEERT, D R , andT J MiLLER 1978. Age, growth, and food of northern pike in eastern Lake Ontario. Transactions of the American Fisheries Society 107: 696-702. Received 25 July 1989 Accepted 15 ISlovemher 1989 Great Basin Naturalist 50(1). 1990. |ip 11 40 INFLUENCE OF SOIL FROST ON INFILTRATION OF SHRUB COPPICE DUNE AND DUNE INTERSPACE SOILS IN SOUTHEASTERN NEVADA Wilhcrt II. Hhukl )iirn aiu 1 M. Kail W (xkI- Abstract. — The influence of soil frost on the intiUiation rate of shruli coppice dune and dune interspace soils was evaluated near Crystal Springs, Nevada, using siinulati-d rainfall. The infiltration rate of the coppice dune soil was greater than the dime interspace soil under frozen or inifro/.en conditions. Because of different vegetation cover and surface soil characteristics, coppice dime and dune interspace soils responded differently to freezing, thus imposing a spatial and temporal response to infiltration rate. Infiltration rate of soils with porous concrete frost increased as the soils thawed during simulated rainfall, hut soils with nonporous concrete frost allowed very little infiltration to occur. Both coppice dune and dime interspace soils that were classified in Jamiar\ as having granular frost had a higher infiltration rate than the same unfrozen soils in March. Soil frost influences water infiltration rates and is often a major influence on runoff^ (Grav and Granger 1987, Harris 1972, Haupt 1967, Kane and Stein 1983, Klock 1972, Kuzick and Bezmenov 1963, Wilcox et al. 1989, and Zuzel and Pikul 1987). Areas where soil frost strongly influences infiltration are character- ized by cold winters, transient snow cover, and soils that may freeze and thaw several times each winter, in addition to a diurnal freeze-thaw cycle (Pikul and Allmaras 1986, Zuzel et al. 1986). Infiltration rate of frozen soil is strongly influenced by the structure of the soil frost, which is determined in part by the soil water content at the time of freez- ing. Concrete frost has been identified as hav- ing the greatest impact on infiltration rates (Haupt 1967, Lee and Molnau 1982, Storv 1955). The spatial influence of shrub coppice dune and dune interspace soils on infiltration of unfrozen soil was originally established by Blackburn (1975) and verified by numerous other investigators (Johnson and Gordon 1988, Swanson and Buckhouse 1984, Thurow et al. 1986, Wood and Blackburn 1981, Wood et al. 1987). Because shrub coppice dune and dune interspace soils have different vegeta- tion cover and surface soil characteristics, we hypothesized that they will respond differ- ently to soil freezing and thawing, thus impos- ing a spatial and temporal response to infiltra tion during winter. The study objective was to determine the spatial variation in infiltration rates of frozen and unfrozen shrub coppice dune and dune interspace soils. Study Area The study area is located in southeastern Nevada about 7 km west of Crystal Springs, 37°20' N latitude, 115°20' W longitude at 1,200 m elevation, and 9 km east of the 1,850-2,100 m ridgeline of the Pahranagat Range. The normal annual precipitation of 330 mm occurs mostly during winter as snow or during Jifly and August as thundershowers. Winters are characterized by periodically cold temperatures with frequent diurnal freeze- thaw cycles. Blackbrush {Coleogijne ramosissimum) is the dominant shrub, with 22% crown cover; associated species are joint fir {Ephedra nevadensis) and box thorn {Lijcium ander- sonii), each with 2% crown cover. Herba- ceous cover is sparse. The study site is located on the tops of long, narrow alluvial fans with 5% slope to the east, and the soils are loamy-skeletal, mixed, thermic, shallow, Typic Durorthids. Two major types of surface soils are found on the study site: the coppice dune soil under shrubs and the barren interspace soil between shrubs. The coppice dune soil covers 35% of 'l'.SD.\, .\,i;ritiiltiiral Research Service, Northwest Watershed Research Center, 270 South Orchard. Boise, Idaho S3705. "Department olWninial and Rani;e .Science. New Mexico State University, Las Cruces, New Mexico 8S003. 41 42 W. H. Blackburn and M K Wood [Volume 50 the surface, whereas interspace soil covers 65%. The A horizon of the coppice soil is characterized by a weakly subangular blocky structure and a gravelly sandy loam texture. Interspace soils have gravel pavement several pebbles thick over the mineral soil. The 50-mm-thick, loamy, crusted A horizon is massive, has vesicular pores, and is broken into 80-150-mm-diameter polygons. The crusted interspace soil slakes and disperses readily when wetted. Methods Infiltration rates were determined using a drip-type rainfall simulator (Blackburn et al. 1974) with simulated raindrops 2.5 mm in diameter. Drops falling 2.1 m reach 5.25 m sec^ or 71% of the terminal velocity achieved by raindrops in an unlimited fall (Laws 1941). Simulated rainfall was applied on frozen soil in January and on unfrozen soil in March 1974 at a rate of 76 mm hr^ for 30 min. The rainfall intensity was chosen to assure runoff from each study plot. RunoflF plots (280 x 500 mm) were randomly placed on 16 and 6 shrub cop- pice dune soils and 13 and 5 interspace soils during the January and March sample dates, respectively. This sample size is considered adequate for rangeland conditions (Wood 1987). Shrubs were cut at ground level and removed from the coppice dunes to reduce rainfall interception losses. Volume of runoflPwas measured every 5 min for 30 min. Infiltration rates were determined as the difference between simulated rainfall and runoff volumes. Soil frost was character- ized adjacent to each runoff plot prior to the simulated rainfall event. Three structural forms of soil frost were observed and subjec- tively classified according to criteria by Hale (1951) and Haupt (1967). Granular frost con- sisted of scattered granules of ice binding min- eral soil together. Nonporous concrete frost was characterized by dense, thin ice lenses and ice crystals. Porous concrete frost was less dense than concrete frost, but frozen chunks of soil were harder to break. Porous concrete frost was further defined by resistance to re- peated thrusts of a pick before being punc- tured. Water used for rainfall simulation aver- aged 4 C in January and 9 C in March. Analysis of variance and least significant dif- ference mean separation tests (Snedecor and Cochran 1971) were used to test for differ- ences between infiltration rates of the coppice dune and dune interspace soils for the January and March sample dates. Results and Discussion All plots during January were classified as having soil frost 100 mm thick located about 50 mm below the surface. Three of the cop- pice dune and five of the interspace plots were classified as granular frost. The remaining plots of both soils were characterized as porous or nonporous concrete frost, of which 12 coppice dune and 6 interspace plots were classified as porous concrete frost. Infiltration rate after 25 min was signifi- cantly greater for coppice dunes than for dune interspaces under both frozen and unfrozen soil conditions (Fig. 1, Table 1). Similar rela- tionships between unfrozen shrub coppice dune and dune interspace soils have been reported by Blackburn (1975), Johnson and Gordon (1988), Thurow et al. (1986), and Wood and Blackburn (1981). The differences in infiltration of coppice dune and dune inter- space soils have been attributed to differences in vegetation and surface soil characteristics (Blackburn 1975, Johnson and Gordon 1988). Infiltration of unfrozen rangeland soil is usu- ally characterized by a high initial rate that decreases rapidly with time and stabilizes at some constant rate within 30 to 60 min (Fig. 1). However, mean infiltration rates in Janu- ary declined within the first 15 min for the coppice dune soils and within 20 min for inter- space soils; in neither case did they stabilize at a constant rate. Infiltration rates for both soils increased during the latter part of the rainfall due to thawing of the porous concrete soil frost layer of some plots (Figs. 2, 3). As a result, there was no significant difference in 30-min infiltration rates between frozen and unfrozen dune interspace soils. However, 30- min infiltration rates tended to be lower in the frozen coppice dune soils in January than in unfrozen soils in March (Fig. 1). Infiltration rates of coppice dune plots classified as granular frost were similar to the rainfall application rate and 10 mm hr~ greater than when the soils were unfrozen in March (Figs. 1, 2). Coppice dune plots classified as porous concrete frost and thawing during the rainfall event reached a minimum 1990] Influence of Soil Frost 43 80 I X UJ < < CXI 60 - 40- 20- COPPICE DUNE ■ JANUARY ♦ MARCH DUNE INTERSPACE + JANUARY A MARCH -T" 10 -r- 15 20 25 30 35 TIME (MIN) Fig. I. Nevada. Mean infiltration rates for all coppice dune and dune interspace soils in January and March, Crystal Springs, Table L — Significant difference of five-minute interval infiltration rates for coppice dune and dune interspace soils for the January and March sample dates'. Crystal Springs, Nevada. Soil/Sample date Time (minutes) 10 15 20 25 30 Coppice March vs. coppice January Coppice March vs. interspace March Coppice March vs. interspace January Coppice January vs. interspace March Coppice January vs. interspace January Interspace January vs. interspace March ns ns ns Sample size: coppice, January n 16, March n 6- interspace, January n 13, March n 5 Level of significance (P s .01 **; P s 0.5 *; ns - nonsignificant at P > .05)cleterniined with a one-way analysis ofvariance and a least significant difference (Isd) mean separation test. infiltration rate after 20 min of 44 mm hr~', but 30-min rates increased to a rate similar to that of the granular frost plots (Fig. 2). The one porous concrete frost plot that remained frozen during the rainfall event reached a minimum rate after 15 min and then increased slightly during the remainder of the event to 33 mm hr . Infiltration rate at 30 min of interspace soil with granular frost was 14 mm hr~ greater than when the soil was unfrozen in March. Interspace plots that were initially classified as porous concrete frost and thawing during the rainfall event reached a minimum infiltra- tion rate of 12 mm hr^ after 20 min, after which rates increased and were similar to un- frozen soils in March. Infiltration rate of the interspace nonporous concrete frost plot de- creased to 2 mm hr at 30 min, 21 mm hr~ lower than the 30-min rate of unfrozen soils in March. Other researchers have reported sim- ilar infiltration response caused by soil frost 44 W. H. Blackburn and M. K. Wood [Volume 50 80 I a: < a: < q: 60 - 40- 20 - ■ GRANULAR FROST + POROUS CONCRETE FROST/THAWING ♦ POROUS CONCRETE FROST 10 -I— 15 20 25 30 35 TIME (MIN) Fig. 2. Intiltnition rates in January for coppice dune soils classified as having granular host, porous concrete frost that was thawing during the rainfall exent, and porous concrete frost. Crystal Springs, Ne\ada. 80 I X 3: 60 H < on < 40 - 20 - ■ GRANULAR FROST + POROUS CONCRETE FROST/THAWING ♦ NON-POROUS CONCRETE FROST 35 TIME (MIN) Fig. 3. Inhltration rates in January for dune interspace soils classilietl as ha\ ing granular frost, porous concrete frost that was thawing during tlu' rainfall event, and nonporous concrete frost, C^rx stal Springs, Ne\ada. 1990] InfluknckoI' Soil, Kiu)sr 45 structure. Trinible et al. (1958), iu N(>\v Hampshire, reported iufiltration rate ol soils with granular frost to he higher than that ol untrozen soils, llaupt (1967) lound, for the eastern slope of the Sierra Nevada, the infil- tration rate of porous concrete frost to in- crease as the soil thawed. Trimble et al. (1958) and Stoeckeler and Weitzman (1960) found infiltration rates of nonporous concrete frost in northern Minnesota to be very low. Conclusions The infiltration rate oi shrub coppice dune soils was greater than dune interspace soils under frozen and unfrozen conditions. Con- crete frost located 50 mm below the surface had a pronounced effect on infiltration rates. Infiltration rates of soils with porous concrete frost increased as the soils thawed during the simulated rainfall, but soils with nonporous concrete frost allowed very little infiltration to occur. Both coppice dune and dune inter- space soils classified as having granular frost had a higher infiltration rate than the same unfrozen soils in March. Due to different veg- etation cover and surface soil characteristics, shrub coppice dune and dune interspace soils responded differently to soil freezing and thus imposed a spatial and temporal response to infiltration rate. Literature Cited Blackburn, W. H 197.5. Factors influencing infiltra- tion and sediment production of semi-arid range- lands in Nevada. Water Resources Research 11; 929-937. Blackburn, W H , R O Meeuwig, and C M Skau 1974. A mobile infiltrometer for use on rangeland. Journal of Range Management 27: 322-323. Gray, D. M., and R. J. Granger 1987. Frozen soil: the problem of snow melt infiltration. In: Y. S. Fok, ed., Proceedings, International Conference on Infiltration Development and Application, Water Resources Research Center. University of Hawaii, Honolulu. Hale, C E 1951. Further observations on soil freezing in the Pacific Northwest. Pacific Northwest Forest and Range Experiment Station Research Note No. 74, Portland, Oregon. Harris, A R 1972. Infiltration rate as affected by soil freezing under three cover types. Soil Science Society of America Proceedings 36: 489-492. Haupt. H F 1967. Infiltration, overland flow and soil movement on frozen and snow-covered plot. Wa- ter Resources Research 3: 145-161. Johnson, (] W, and N E (Gordon 1988. HunoU and erosion from rainlall simulator plots on sagebrush rangeland. Transactions oi the American Society ol .\gricultiual Engineers 31: 421-427. Kank, D L, and J Stein. 1983. Water movement into frozen soils. Water Resources Research 19: 1547- 15.57. Klock, G O 1972. Snow niclt temperature iiinuence on inliltration and soil water retention. Journal ol Soil aTid Water Conservation 27: 12-14. Kuzk;k, I A . and A I. Bezmenov 1963. Inhltration of meltwater into frozen soil. Soviet Soil Science 6: 665-670. Laws, J D 1941, Measurements of the fall-velocity of water-drops and raindrops. Transactions of the American Geophysical Union 22: 709-721. Lee, R W , and M. P. Molnau, 1982. Infiltration into frozen soils using simulated rainfall. Paper No. 82-2048, American Society of Agricultural Engi- neers, St. Joseph, Michigan. Pikul, J. L.. Jr., and R R Allmaras 1986. Physical and chemical properties of a Haplo.xeroll after 50 years of residue management. Soil Science Society of America Journal .50: 214-219. Snedecor, G W., and W. G. Cochran, 1971. Statistical methods. Iowa State University Press, Ames. Stoeckeler, J. H., and S. Weitzman. 1960. Infiltration rates in frozen soils in northern Minnesota. Soil Science Society of America Proceedings 24: 1.37-139. Story, H C 19.55, Frozen soil and spring and winter floods. In: Yearbook of Agriculture 19.55. U.S. Department of Agriculture, Washington, D.C. SwANSON, S. R., AND J. C BucKHOUSE. 1984. Soil and nitrogen loss from Oregon lands occupied by three subspecies of big sagebrush. Journal of Range Management 37: 298-302. Thurow, T L., W H. Blackburn, and C. A Taylor, Jr. 1986. Hydrologic characteristics of vegetation types as affected by livestock grazing systems, Edwards Plateau, Texas. Journal of Range Man- agement .39: .505-509. Trimble, G R., R S. Sartz, and R S Pierce 19.58. How types of soil frost affect infiltration. Journal of Soil and Water Conservation 13; 81-82. Wilco.x, B P , C L Hanson, J, R Wight, and W H Blackburn 1989. Sagebrush rangeland hydrol- ogy and evaluation of the SPLTR hydrology model. Water Resources Bulletin 25; 653-666. Wood, J C , M. K, Wood, and J. M Tromble. 1987. Im- portant factors influencing water infiltration and sediment production on arid lands in New Mexico. Journal of Arid En\ironment 12: 111-118. Wood, M K 1987. Plot numbers recjuired to determine infiltration rates and sediment production on rangelands in southcentral New Mexico. Journal of Range Management 40: 259-263. Wood, M K,andW. H. Blackburn 1981. Grazing sys- tems: their influence on infiltration rates in the rolling plains of Texas. Journal of Range Manage- ment 34; .331-335. 46 W. H. Blackburn and M. K. Wood [Volume 50 ZUZEL. J. F , AND J. L. PiKUL, Jr 1987. Infiltration into a Journal of Climate and Applied Meteorology 25; seasonably frozen agricultural soil. Journal of Soil 1681-1686. and Water Conservation 42: 447-450. , ,^ ^ 7 -,r,ar, ZUZEL J F J L P.KUL, jR , AND R N Greenwalt 1986. Received 12 October 1989 Point probability distributions of frozen soil. Accepted 12 December 1989 Great Basin Naturalist 50(1), 1990, pp 47,56 SEED PRODUCTION AND SEEDLING ESTABLISHMENT OF A SOUTHWEST RIPARIAN TREE, ARIZONA WALNUT (JUGLANS MAJOR ) Juliet C. Stromherg' and Duncan T. Patten' Abstract. — A four-year study of five populations has revealed influences on seed production and seedling establish- ment of the Southwest riparian tree Julians major. Germination is abundant after production of large seed crops (masts), but masts are produced infrecjuently. Within years, germination is stimulated by summer rains, enabling seedlings to establish on riparian terraces as well as streambanks. Traits such as capacity for dormancy during summer drought allow some seedlings to survive on terraces, but abundant rainfall is essential for high rates of seedling success. Ranges of moisture tolerance vary among seeds collected from different populations, suggesting that ecotypes may exist between riparian sites with dissimilar moisture regimes. Population-based differences are associated, in part, with differences in seed size. Arizona walnut (Juglans major [Torr. ] Hel- ler) is a member of the "Interior riparian de- ciduous forest," an assemblage of trees that grow along streams in the Interior South- west (Brown 1982). Walnut sometimes domi- nates this community (Szaro 1989), but more often it occurs at relatively low densities and frequencies. This distribution pattern, while indicating that establishment occurs infre- quently, does not reveal the stage where re- generation is limited. Sudworth (1934) sug- gested low rates of seed production and high rates of seed predation by tree squirrels as possible natural causes of infrequent estab- lishment; seedling mortality from cattle graz- ing also may play a role (Rucks 1984). Lack of suitable germination and establish- ment sites may limit recruitment, but little is known regarding the relationship between these requirements and /. major. Certain generalizations have been made about the dis- tribution of walnut trees, but the habitat char- acteristics of trees may differ from those in which the seedlings established. A study that specifically addressed riparian seedlings re- vealed that seedlings of/, major, a facultative riparian species, establish in many microsites throughout the riparian zone. This is in con- trast to seedlings of obligate riparian trees such as Arizona alder (Alnus oblongifolia Torr.) that occur only immediately adjacent to the stream (Larkin 1987). Larkin conducted her study, however, in a wet, montane area of central Arizona; germination "safe sites " for walnut may be more restricted in drier ripar- ian sites such as along ephemeral streams. For example, seed burial may become a prerequi- site of germination as soil moisture decreases, as is true for certain oaks (Barrett 1931). Germination and establishment of walnut or other riparian species may also vary as a function of genetic differences between popu- lations. The existence of riparian ecotypes is not a new idea (e.g.. Hook and Stubbs 1967), and riparian populations may differ between isolated Southwest watersheds with different hydrologic characteristics. Other studies in this series on walnut reproduction have iden- tified differences between populations in flo- ral ratios and seed weight (Stromberg 1988), indicating that /. major shows either plastic responses or genetically based responses to environmental differences within and be- tween riparian sites. The objectives of this study were to: (1) identify factors that limit seed production and seedling recruitment of /. major between sites and between years; (2) determine how germination requirements vary between microsites; and (3) determine how seedling growth response to soil moisture differs between populations. Methods Seed production, germination, and seed- ling survival of/, major were studied through Center for Environmental Studies, Arizona State University. Tempe. Arizona 85287-1201. 47 48 J. C. Stromberg and D. T. Patten [Volume 50 field and greenhouse experiments and field observations. Five sites were selected in cen- tral Arizona: Hitt Wash, an ephemeral stream near Prescott National Forest; Rock Creek, an intermittent stream in the Mazatzal Moun- tains; and perennial streams (South Fork Walnut Creek in Prescott National Forest and two sites along Workman Creek in the Sierra Ancha Mountains) (Stromberg 1988). Eleva- tions vary from 1,100 m at Rock Creek to 2, 100 mat Aztec Peak. Field Observations Twenty trees were selected at each site, and samples of 20 shoots per tree were se- lected for monitoring of seed production from 1983 to 1986. Average values per tree were calculated for seed production per shoot and for flower production, flower abortion, seed weight, and seed viability (Stromberg 1988). A mast (large seed crop) is operationally de- fined in this study as a crop 33% larger than average for the study. Demography of natural seedlings was stud- ied to determine when, where, and how many seeds germinate, and to compare survival rates between microsites. One walnut tree near the stream and one on the adjacent terrace with nearby seedlings were selected at each site. Each marked the center of an 8-m-radius circular plot. All seedlings in the plots were tagged and scored monthlv from May to October of 1983, 1984, and 1985 for stem height, leaf length and number, and mortality. Twenty seedlings in a plot at Hitt Wash were excavated to measure depth of the nut in the soil profile. Field Experiments Seeds were sown in field e.xclosures in fall 1982 and 1983 to test for influence of microsite (soil moisture and canopy cover) and burial depth on germination and survival. Exclo- sures protected seedlings from trampling and predation. Four microsites (streamside, open canopy; terrace, open canopy; streamside, closed canopy; terrace, closed canopy) were selected per site, with three replicate exclo- sures per microsite. Exclosures were con- structed from 1.3-cm hardware cloth, and were 0.6 x 0.6 X 1 m high, with 15 cm of mesh below ground. Seeds were planted 36 per exclosure in 6 rows in 3 planting treatments: surface sown, partially buried, or buried 2 cm deep. Sites were planted with their own seeds (except for Workman Creek and Rock Creek), and seedling size and sur- vival were recorded through 1985. Microsite soil moisture was recorded monthly with a dew-point psychrometer, and canopy cover was estimated visually (Stromberg 1988). Influences on germination and seedling sur- vival were analyzed through multiple regres- sion analysis of germination and survival per- centages with average seasonal values for microsite moisture level, canopy cover, and herbaceous cover. Greenhouse Experiments Nuts were collected in 1983 from trees at Rock Creek, Hitt Wash, and Walnut Creek to test for influence of seed weight and seed source on germination and seedling growth. Nuts were sown in the greenhouse in loam soil in polyethylene-lined pots under four water- ing regimes. One regime was watered every other day to maintain saturated conditions. Others were watered at less-frequent inter- vals, producing average soil moisture con- tents by dry weight of 80%, 40%, and 20%. Seeds were sown on the surface and buried 2 cm deep. Seven nuts from each of 24 parent trees were sown per treatment. Germination percent and speed and seedling survival were monitored. Eight weeks after emergence, plants were harvested for measurement of dry weight, stem height, and root length. At each soil moisture level, regression analysis was used to test for influence of seed weight, a continuous variable. Duncan's multiple range test was used to test for effects of seed source on germination and seedling growth rates. Results Seed Production Seed production differed substantially be- tween sites and years. All sites had large an- nual variation in seed production, with aimual coefficients of variation ranging from 100 to 140 among sites. Masts were produced in one or two of the four years among sites (Fig. 1). Large crops of seeds, irrespective of vial)ility, were produced more frequently. Some excep- tional trees produced four consecutive, large seed crops. Number of large flower crops also \aried substantialK between sites, ranging from one at Walnut Creek to three at Hitt Wash. 1990] Kii'AHiAN Sei:i) Phoduction 49 Fig. 1. Site means for female flowers and viable nuts produced per shoot iorjughins major from 1983 through 1986 at Aztec Peak (A), Workman Creek (K), Walnut Creek (W), Hitt Wash (H), and Rock Creek (R). 50 J C Stromberg AND D T Patten [Volume 50 Table 1. Average rates of seed production (expressed as viable seeds produced per 100 shoots) and abundance of Juglans major seedlings per 200 m" at five sites in Arizona. Seeds Seedlings Seeds Seedlings Seeds Seedlings 1983 1984 1984 1985 1985 1986 Aztec Peak 11 0 0 0 2 0 Workman Creek 9 2 3 1 3 0 Walnut Creek 38 52 0 0 14 23 Hitt Wash 46 207 10 48 6 24 Rock Creek 10 3 10 1 2 1 Masts were produced at all sites in 1983, with abundant viable nuts maturing on trees (Fig. 1). Seed production patterns in following years differed among sites. Seed crops were low in diflFerent years among sites, and crops failed at different reproductive stages. The most common causes of crop failure were re- ductions in numbers of female flowers and production of inviable, low-weight seeds; high rates of abortion were a less frequent cause. At Walnut Creek, for example, viable nut production was lowest in 1984, a result of few flowers and inviable seeds (3% viable). Hitt Wash trees produced a large crop of vi- able seeds only in 1983; seed production was limited at sequentially earlier reproductive stages from 1984 to 1986. In 1984 many flow- ers were produced and matined, but seed viabilities were low (28% viable). In 1985 flowers were again abundant, but many (96%) were aborted. In 1986 Hitt Wash finally had a sharp decline in female flowering, resulting in a low crop size. Rock Creek trees showed a third pattern. Following the mast of 1983, moderate numbers of flowers and viable seeds were produced in 1984. This was followed by large declines in flower number in 1985. Mod- erate numbers of viable seeds were produced in 1986. Field Germination Natural seedling abundance differed sub- stantially between sites and years. These dif- ferences were related, in part, to seed produc- tion rates. Seedlings were abimdant only at sites that produced many seeds (Hitt Wash and Walnut Creek) (Table 1). Few seed- lings were present at Rock Creek, despite moderate seed production, because of seed pn^dation by Arizona gray squirrels (Scitinis arizoiiensis; see Stromberg 1988). Annual seedling abundance was influenced by size of the prior year's crop of viable seeds, since seeds generally germinated the year after they were produced. The mast year of 1983 was followed by abundant seedlings in 1984, whereas the largely inviable seeds produced during dry 1984 resulted in low or no recruit- ment in 1985. Rainfall in the year of germi- nation may also have increased seedling numbers. This is suggested by the greater abundance of seedlings in wet 1983 than in dry 1984 (e.g., 132 vs. 52 at Walnut Creek), despite production of masts in 1982 and 1983. Within years, seeds germinated during wet seasons (Table 2). Most germinated during the late summer (August-September) rains. Seeds germinated infrequently in May and only during wet springs or in wet streambank microsites; none germinated during the July dry period. Germination increased with microsite soil moisture, although moisture did not exclu- sively regulate germination (r" = .24, df = 14, P < .04; values for exclosures). Few seeds germinated on open terraces, particularly at the low-elevation sites (Rock Creek, Hitt Wash) (Table 3) where soils were drier (Table 4). Canopy cover was positively associated with germination percentages within these drier microsites, and dense herbaceous cover was negatively associated with germination rates. Together, soil moisture, canopy cover, and herbaceous cover explained 45% of the \ ariation in germination percentages between microsites. Few surface-sown nuts germinated in any microsite (Table 3). Burial increased germina- tion rates within all microsites at all sites except Workman Creek, where burial in satu- rated soil decreased germination. Partially buried seeds germinated in moderate mun- bers in streambanks and imder canop)'; com- plete burial was necessarx for germination in open terraces. Lack of burial limited germina- tion of natural seed populations. For example, man\ seeds were imgerminated on the soil surface at Hitt Wash terrace. Excavation of 1990] l\ll'AKIAN SKKI) PhoDI'CTION 51 Tari.K 2. Iii^ldii.s inujor si'cdliiiii rc'ciuitiiK'iil 1)\ laoutli l,.Ma\ tlii()uji;h 0(jti)l)ei) in streuiiisiclL' and ttMiact' i)l()ts at sites in Arizona. I'lots arc 200 uT. 1983 1984 1985 Site Plots M J J S () .1 .1 S () M J J S O Workman Creek Walnut Creek Hitt Wash Rock Creek Terrace Streaniside Terrace Streaniside Terrace Streaniside Terrace Streanisitle 0 0 0 0 6 14 0 0 1 0 ■1 0 53 44 22 13 0 92 43 0 35 13 0 0 0 0 3 0 18 12 14 8 0 105 48 0 47 25 0 0 0 0 1 0 0 0 0 0 0 0 0 0 13 5 0 0 1 0 Table 3. Germination percentages oijuglans major sown at tliree pkmting deptlis in tour microsites. Average seed viability was 60%. Values are means and standard deviations for 3 groups of 12 nuts. Planting depths are: buried at 2 cm, partially buried, and surface sown. Sites are: Aztec Peak (A), Workman Creek (K), Walnut Creek (W), Hitt Wash (H), and Rock Creek (R). NA = not available. Streambank, canopy Streambank, open Site Buried Pa tially bu ied Surface Buried Part ially buried Surface A 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] K 4 [6] 25 [0] 8 [0] NA ■ NA NA W NA NA NA 29 [6] 13 [6] 8 [0] H 45 [30] 3 [5] 0 [0] 46 [6] 0 [0] 0 [0] R 50 [18] 5 [5] 0 [0] .50 [15] 0 [0] 0 [0] Te race, canopy Tei race, open Site Bmied Pa ■tially buried Surface Buried Part iaih Iiuried Surface A 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] 0 [0] K 21 [6] 25 [0] 8 [0] 50 [.33] 0 [0] 0 [0] W 35 [19] 16 [11] 3 [5] 28 [16] 0 [0] 0 [0] H 53 [5] 25 [14] 8 [9] 0 [0] 0 [0] 0 [0] R 21 [12] 0 [0] 0 [0] 4 [8] 0 [0] 0 [0] Table 4. Soil water potential at 30 cm ( — M Pa) for terrace (T) and streaniside (S) open canopy microsites during May, July, and September of 1983, 1984, and 1985. Sites are: Aztec Peak (A), Workman Creek (K), Walnut Creek (W), Hitt Wash (H), and Rock Creek (R). Microsite 1983 1984 1985 M can Site May Jul Sep Ma> Jul Sep Ma\- .Jul Sep [±SD] A T 0.02 0. 15 0.01 0.68 0.83 0.06 0.11 0.79 0.06 0.30 [0.,33] S 0.01 0.16 0.00 0.54 0.63 0.03 0.06 0.54 0.03 0.22 [0.25] K T 0.03 0.19 0.01 1.03 1.36 0.06 0.24 1.16 0.06 0.46 [0.52] S 0.01 0.17 0.00 0.78 0.96 0.02 0.12 0.93 0.04 0..34 [0.40] W T 0.03 0.24 0.02 1.18 1.26 0.05 0.42 1.33 0.04 0.51 [0.54] S 0.02 0.18 0.00 0.84 0.96 0.02 0.18 1.02 0.02 0.36 [0.42] H T 0.03 0..32 0.01 1.65 1.84 0.15 0.48 1.80 0.12 0.71 [0.76] S 0.03 0.22 0.02 1.45 1.63 0.05 0.26 1.58 0.06 0..59 [0.69] R T 0.04 0.43 0.20 1.86 2.04 0. 13 0.64 1.83 0.14 0.81 [0.80] S 0.03 0..32 0.02 1.45 1.63 0.05 0.26 1.58 0.06 0.60 [0.68] Mean 0.03 0.26 0.04 1.27 1.44 0.07 0..32 1..39 0.07 ± SD 0.01 0.09 0.06 0.38 0..39 0.04 0. 16 0..35 0.04 52 J. C. Stromberg and D. T. Patten [Volume 50 Table 5. Germination percentages (or Julians major buried in soil or sown on the surface at three soil water contents (by weight) in the greenhouse. No seeds germinated at 20% soil moisture. Seeds are from streambanks (S) or terraces (T) from Walnut Creek (W), Hitt Wash (H), or Rock Creek (R). Values are means and standard deviations. Satui ated 80% moisture 40% moisture Seed source Buried Surface Buried Surface Buried Surface W S 21 [21] 50 [7] 50 [7] 0 [0] 7 [7] 0 [0] T 0 [0] 50 [29] 42 [18] 0 [0] 21 [21] 0 [0] H S 5 [6] 28 [31] 63 [29] 0 [0] 24 [33] 0 [0] T 0 [0] 18 [12] 61 [16] 0 [0] 19 [27] 0 [0] R S 0 [0] 22 [9] 39 [11] 0 [0] 13 [0] 0 [0] T 0 [0] 14 [7] 48 [15] 0 [0] 16 [5] 0 [0] seedlings revealed that only 5% of the seed- lings were from surface nuts, whereas 75% were from partially buried nuts and 20% were buried at depths of 9 to 15 cm in pocket gopher {Tlwmomys spp.) caches. Greenhouse Germination Jiiglans major germinated abundantly in soil with a moisture content of 80-100% by weight (Table 5). Few seeds germinated in the dry soil or saturated soil. Planting depth influ- enced germination at all soil moisture con- tents. In the saturated soil, surface-sown seeds germinated in substantially higher per- centages than buried seeds. In contrast, seeds required burial for germination at all other moisture contents. Germination rates in saturated soil varied between populations and with seed weight. Seeds collected from trees along the banks of perennial Walnut Creek germinated in higher percentages than seeds from sites with ephemeral (Hitt Wash) or intermittent (Rock Creek) stream flow (Table 5). Many cohorts from these latter sites did not germinate in saturated soil. Germination percentage in sat- urated soil also tended to increase as seed weight (g) per cohort decreased (y = 31.8 — 5.3x, r' - .15, df = 22, P < .05). Seed weight was not related to germination in dry or moist soil. Speed of germination varied with soil mois- ture level and with seed weight. Seeds in moist soil (80-100%) germinated rapidly (16 ± 10 days), whereas seeds in drier soil (40% by weight) germinated 28 ± 12 days after planting. Although variance in germination speed was high within a cohort, median ger- mination speed increased significantly with seed weight for a cohort (v -0.5 + 5.5.\, r" = .34, df = 22, P < .01, in 80% moisture). For example, seeds weighing 3 g germinated in 16 days, compared with 38 days for those weigh- ing 7 g. Seedling Survival in the Field Seedling mortality had a major role in limit- ing regeneration of/, major. Only 1 of 374 natural 1983 seedlings among the study sites was alive as of fall 1985. Mortality was high in the year of germination and in the two years following (Fig. 2). Most seedlings died in June, typically a dry month, or in winter. Precipitation was sparse in early 1984 (Table 6), and seedlings that germinated in 1983 had high winter mortalitv: onlv 7% of Hitt Wash and 9% of Walnut Creek 1983 natural fall co- horts survived until spring 1984. Many that did survive remained dormant through the spring and summer drought of 1984, with- holding leaf-out until the late summer rains. In contrast, more than 20% of 1984 cohorts at both sites survived to spring of 1985. Seedling survivorship varied between microsites (Table 7) as well as between years. A large part of the variation between mi- crosites was attributable to soil moisture. Sur- vivorship after two years increased with soil moisture between exclosures (r" = .31, df = 14, P < .05), with seedlings on open stream- banks having highest sur\'ival. Seedlings that did survive in dry microsites had high mor- tality of buds and shoots during their first and second winters, and grew slowly. On terraces, 42% of year-old surviNors originated new spring growth from basal buds, 49% from lat- eral buds, and only 9% from the terminal bud. On streambanks, in contrast, 15% regrew from basal buds, 58% from lateral, and 31% from terminal buds. Second-\'ear seedlings on the open terrace that regenerated from a basal bud had average stem height of 2.5 cm, with 1990] KirVKlAN SEE13 FHODL'CTION 53 Zi 2H Q LjJ LU CO 1H ° HUT WASH + WALNUT CREEK -I — I — I — r A S 0 N D 1983 T — r S 0 T — r N D 1 — r M A "i 1 — r A S 0 1984 1985 Fig. 2. Seedling survivorship ior Juglaufi major that germinated in fall 1983 at Hitt Wash (HW) and Walnut Creek (WC). Values are logp of seedlings remaining each month. T.ABLE 6. Precipitation (cm) from 1982 to 1986 at a climatic station near the Walnut Creek study site. Aver- age annual precipitation is 39 cm. 198:; 1983 1984 1985 1986 January-June July— December Total 21 19 40 31 53 3 37 40 14 31 44 16 30 46 as few as two leaves 1 cm long. These seed- lings did not markedly increase in size from 1983 to 1985. Seedlings in moist, open areas grew rapidly, as evidenced by a two-year streambank seedling in full sim that reached 44 cm tall. Flooding killed some streambank seedlings but did not impact terrace seedlings. Fall floods killed 10% of Walnut Creek and 5% of Hitt Wash 1983 streambank seedlings; mor- tality rates from winter/spring floods were not quantified. Physical impacts of floods in- cluded stem breakage, coverage with debris, and scouring of seedlings. Several seedlings resprouted after stem topping. Herbivory damage from insects also con- tributed to seedling mortality. Interestingly, seedlings did not have greatest herbivory at sites where adult herbivory was high (see Renaud 1986). Rather, seedlings in specific microsites had greatest leaf loss from herbi- vores. Seedlings on terraces, under canopy, had greatest herbivory damage; leaf area con- sumed by October 1983 was 33% ± 15, com- pared to < 9% for all other microsites. Effects of cattle grazing were included in the study out of necessity because of the al- most ubicjuitous presence of cattle in South- west riparian areas. Two sites, Hitt Wash and Walnut Creek, had heavy to moderate cattle grazing. Seedlings in exclosures at both sites had substantially higher survival rates than those in similar tmprotected areas; however, this was also true for the ungrazed sites (Table 7). Adverse impacts of cattle on seedlings in- cluded trampling and grazing. At Hitt Wash, 22% of 213 natural seedlings had broken or eaten stems, as did 13% of 130 seedlings at Walnut Creek. Ability of seedlings to recover from trampling and grazing varied between sites; 40% of all stem-damaged seedlings at 54 J C Stromberg and D T. Patten [Volume 50 Table 7. Survival percentages one, two, and four years after germination iov fu^,l(ins major seedlings in exelosures and natural areas in four microsites. Microsite Stream, open Stream, canopy Terrace, open Terrace, canop)' Exelosures lyr 2yr 4yr 51 28 16 22 15 0 6 0 0 14 0 0 Natural areas lyr 2 yr 4 yr 0 0 0 7 0 0 5 0 0 3 0 0 Table 8. Dry weight, root length, shoot height, root:shoot weight ratio, and mortality of eight-week y«^/flji.v major seedlings in the greenhouse in soil at three moisture contents. Asterisk (*) indicates significant difference between Hitt Wash (H) and VVahiut Creek (W). Values are means and standard deviations. Moisture Seed Dr\- weight Root Shoot R:S Mortality treatment source (g) (cm) (cm) ratio (%) 40% H 0.35 [0.05] 25 [2] 9 [2] 1.4 [0.3] 0 [0] W 0.44 [0.05] 29 [3] 11 [2] 1.1 [0.3] 0 [0] 80% H 0.83 [0.14] 34 [7] 16 [2] 0.9 [0.2] 0 [0] W 0.76 [0.12] 34 [6] 15 [2] 0.9 [0.2] 0 [0] Saturated H 0.12 [0.02]* 4 [2]* 7 [2]* 0.4 [0.1]* 80 [22]* W 0.27 [0.04] 13 [3] 10 [3] 0.7 [0.2] 13 [6] Hitt Wash regenerated a new stem during the same year of breakage, whereas none did so at Wahiut Creek. Greenhouse Seedhng Survival All seedling cohorts had greatest growth in intermediate soil moistures and poorest in sat- urated soil (Table 8). Root growth in particular was low in saturated soil, and root-to-shoot ratios were low compared to drier soils. Some cohorts, however, grew better than others in saturated soil. Similar to results for germina- tion rates, seedling growth and siuvival in saturated soil were related to seed weight and seed source. Size and weight of seedlings in saturated soil increased significantly with de- creasing seed weight among cohorts (e.g., seed weight in g = 26.5 — 4.5 * root length in cm, r' = .49, df = 11, P < .01). With regard to seed source, cohorts from the perennially saturated streambanks of Walnut Creek had greater root development and sinvi\'orship in saturated soil than did cohorts from drier Hitt Wash. Disc:ussiON Although a survey of five populations is not representative of a species as a whole, this study has highlighted factors influencing re- cruitment of /. major. Low and fluctuating availability of seeds plays a large role in limit- ing abundance of seedlings between years and sites. The extent of fluctuation in annual seed production by/, major is similar to values for other mast-cropping trees (Silvertown 1980), and freciuency of mast production is similar to other Juglandaceae (Nixon et al. 1980, Sork 1983, Waller 1979). Seedlings were abundant only after mast years, substantiating the view that infrequent production of viable crops lim- its regeneration (Sudworth 1934). Rainfall, which varies considerably between years in the Southwest, appears to ha\ e an important influence on mast production. The evidence for this, although based on a limited number of years of observation, comes from associa- tions detected between rainfall and the repro- ductive stages that are critical to production of successful masts — flower production, associ- ated with abundant prior and present year rainfall, and seed weight, which increases with abundant spring rainfall (Stromberg 1988). Seed number between sites is limited variously by low soil moisture and high pre- dispersal seed predation (e.g.. Rock Creek) and insect herbivory (e.g., Workman Creek) (Stromberg 1988). Low rates of germination also limit recruit- ment to some degree. Whereas moisture for germination and establishment of some obligate riparian trees is provided by stream flow (Fenner et al. 1985), these processes in /. major, and perhaps other facultative 1990] Kii'AHiAN Si;i;i) Fiu)1)1'(:ti()n 55 riparian trees, are iiilliieneed in larue part 1)\ rainfall. Recent doennuMitation of fairly higli regeneration rates for waliuit in the Sontli- west (Larkin 1987, Medina 1986) may l)e a result of long-term moisture cycles, Arizona being in an above-average cycle at the time of these studies. Although moist areas pro\'ided optimum safe sites for germination and seedling estab- lishment, some/. 7nq/or seedlings established on terraces and along ephemeral streams. Seedlings were somewhat drought-tolerant, as indicated b>' high survival in greenhouse drought conditions, high rootishoot ratios in drier conditions, and ability to survive sum- mer drought via dormancy. Nevertheless, seedling numbers in drier riparian sites were low. Recruitment may be abundant only after a sequence of several wet years — two for pro- duction of a large viable seed crop, another for abundant germination, and one or two more for high survivorship. The low frequency of such a sequence may explain the uniform age structure of adult walnut populations at some low-elevation sites (Stromberg 1988). An additional factor that may limit estab- lishment of seedlings in dry sites is lack of burial. Processes that bury seeds include deposition of flood debris (rare on terraces), caching by pocket gophers (rare except in sandy soils), trampling by large animals, and possibly caching by squirrels. Although tree squirrels commonly cache walnuts (Stapanian and Smith 1978), there is conflicting evidence about the frec^uency at which they cache nuts of /. major. In parts of their range where winters are mild, squirrels immediately con- sume gathered nuts (Brown 1984). This be- havior may contribute to the decline in abun- dance of walnut at low elevations. This study suggests that germination and establishment requirements of/, major differ between populations, as well as between microsites. Specifically, populations from pe- rennial stream sites appear to be more toler- ant of saturated soil at the seed and seedling stages; this should be verified on a larger sam- ple of populations. The greater germination and seedling survival in saturated soil for seeds from such sites may be a consequence of lower oxygen demands of their smaller seeds (Stromberg 1988) or of physiological adapta- tions (Hook and Crawford 1978). Tolerance of moisture level is known to vary among seeds and seedlings as a result of ecotypic differen- tiation in morphology or physiology (Hook and Slnbbs 1907), and differences in tolerance ol Hooding and saturated soil are common among tices that grow on sites with different flooding histories (McGee ct al. 1981). Isola- tion of Southwest riparian populations within "mountain islands" with distinct moisture regimes may have led to development of phvsiological and morphological ecot\pes (Little 1950, Thornber 1915). Whereas small seed size and tolerance of saturated soil are associated with wet riparian sites, the large seeds produced by walnuts on drier sites (Stromberg 1988) may increase sur- vival of seedlings stressed by factors such as drought or grazing. Although purely specula- tive, the greater abilit\' of Hitt Wash seedlings to recover from trampling and grazing com- pared with Walnut Creek seedlings may have been a consequence of larger seed size. Large seeds have large cotyledons that remain at- tached to seedlings for up to a year after ger- mination (Stromberg 1988), allowing young seedlings to regenerate stems (Wetzstein et al. 1983). In any case, differences in seedling responses between walnut populations high- light the need for study of many populations to thoroughly understand reproductive dynam- ics of any riparian species. Acknowledgments This study was supported in part by a coop- erative agreement with the U.S. Forest Ser- vice (RM-80-133-CA). Literature Cited Barketi, L I 1931. Iniluence of forest litter on the germi- nation and early survival of chestnut oak, Ouercus montana Willd. Ecology 12; 476-484. Brown, D E .ed 1982. Biotic communities of the Ameri- can Southwest — United States and Mexico. Des- ert Plants 4: l-,342. 1984. Arizona's tree scjuirrels. Arizona Game and Fish Department, Phoenix. 114 pp. Fenner, p. W. W Brady, and D R Patton 1985. Effects of regulated water flows on regeneration of Fremont cottonwood. Journal of Range Manage- ment ,38: 1.3.5-1.38. Hook, D D , and R M Crawford, eds 1978. Plant life in anaerobic environments. Ann Arbor Science Publishers, Ann Arbor, Michigan. 564 pp. Hook, D D. and J Stubbs. 1967. Physiographic seed source variation in tupelo gums grown in various water regimes. Pages 61-67 in Proceedings of 56 J. C. Stromberg and D T. Patten [Volume 50 the Ninth Southern Conference on Forest Tree Improvement, Knoxville, Tennessee. Committee on Southern Forest Tree Improvement, Macon, Georgia. Larkin, G. J. 1987. Factors influencing distribution and regeneration of riparian species along mountain streams in central Arizona. Unpublished thesis, Arizona State University, Tempe. 84 pp. Little, E. L., Jr 1950. Southwestern trees: a guide to the native species of New Mexico and Arizona. Agri- cultural Handbook 9. United States Department of Agricultiue, Washington, D.C. McGeE, A B , M R SCHMIERB.ACH, .\ND F A Bazzaz 1981. Photosynthesis and growth in populations of Popiilus dcltoides from contrasting habitats. American Midland Naturalist 105: 305-311. Medina. A. L. 1986. Riparian plant communities of the Fort Bayard watershed in southwest New Mexico. Southwestern Naturalist 31: 345-359. Nixon, C. M.. M W McClain. and L P Hansen 1980. Six years of hickor\ seed yields in southeastern Ohio. Journal of Wildlife Management 44: .534-539. Renaud, D. 1986. The impact of herbivor> on the repro- ductive and defense allocation of the Arizona wal- nut. Unpublished thesis, Arizona State Univer- sity, Tempe. 54 pp. Rucks, M. G. 1984. Composition and trend of riparian vegetation on five perennial streams in southeast- ern Arizona. Pages 97-107 in R. E. Warner and K. M. Hendrix, eds., California riparian systems: ecology, conservation, and producti\e manage- ment. University of California Press, Berkeley. SiL\ ERTOWN, J W 1980. The e\'olutionary ecology of mast seeding in trees. Linnean Societ\' Biological Join- nal 14: 235-250, Sonk \' L 1983. Mast fruiting in hickories and availa- bilitv of nuts. American Midland Naturalist 109; 81-88. Stapanian, M. a, andC C Smith 1978. A model for seed scatterhoarding: coevolution of fox squirrels and black walnut. Ecology 59: 884-896. Stromberg, J C. 1988. Reproduction and establishment of Arizona walnut (Julians major). Unpublished dissertation, Arizona State Universitv, Tempe. 183 pp. Sl DWORTH, G B 1934. Poplars, principal tree willows, and walnuts t)f the R()ck\' Mountains region. Tech- nical Bidletin 420. United States Department of Agriculture, Washington, D.C. SZARO. R. C 1989. Riparian forest and scrubland commu- nity tvpes of Arizona and New Mexico. Desert Plants 9: I- 138. Thornber, J J 1915. Walnut culture in Arizona. Univer- sit\ of Arizona Agricultural Experiment Station Bulletin 76: 469-503. Waller, D M 1979. Models of mast fruiting in trees. Journal of Theoretical Biology 80: 223-232. Wetzstein, H, Y,, D, Sparks, and G, A. Lang 1983. Cotyledon detachment and growth of pecan seed- lings. HortScience 18: 331-333, Received 30 J line 1989 Accepted 1 January 1990 Cn-al Basin Naturalist 50( 1), 1990. pp. 57-62 FORAGE QUALITY OF RILLSCALE {ATRIPLEX SUCKLEYl) GROWN ON AMENDED BENTONITE MINE SPOIL MaitjMci itf Iv XOorlu'cs Abstract. — At peak .staiKlin^ wop. lill.scalc (Atriplcx .sucklciii) loliagc mown on aiiU'iKk'd hciitoiiitc tiiiiK' spoil contained adeejuato digi'.stil)le energy, eriide protein, and all niineral element.s e.xeept pliosphonis neee.ssary tor cattle, sheep, antelope, and deer. Ainendnient.s (sawdust, NPK, gypsum) generally did not afleet forage ciuality. Iron, manganese, alumininn, sodium, and potassium concentrations were high and ma\ luuc ad\('rsely alfeeted forage quality. Forage utilit\ would be limited to a few months during the growing season. Atriplex siickleyi (Torrey) Rytlb-, eoiii- monly called rillscale, is the dominant native invader on hentonite mine spoil (Sieg et al. 1983). Rillscale is a spreading annual plant, usually less than 30 cm in height, that flowers from early June to mid-August, bearing ma- ture seed before the end of July. The plant is found only in southern Saskatchewan, south- ern Alberta, Montana, Wyoming, North Da- kota, South Dakota, and Nebraska. It has been observed that the plant grows in saline, clayey, and alkaline land "where nothing else seems to grow" (Frankton and Bassett 1970). Little in known about the biology of this species. Forage quality is an important consider- ation in the selection of species for use in revegetation of hentonite mine spoil, since grazing is the major postmining land use in regions where hentonite is mined. Wildlife forage and habitat are also emphasized in reclamation efforts. Twenty-two species of wildlife are known to use Atriplex species for food and cover (Robinette 1971, Martin et al. 1951). Atriplex species are valued by range managers because of their high protein con- tent (Bidwell and Wooton 1925). The peak forage value of rillscale, as with most annual forbs, is in all likelihood limited to spring and earlv summer months (Cook 1972, Stoddartetal'. 1975). When available, it may make an important contribution to the nutrition of livestock and wildlife. The objec- tive of this study was twofold: (1) to examine chemical properties of rillscale foliage col lected from plots on raw hentonite spoil and on hentonite spoil that had been amended with various combinations of gypsum, fertil- izer (NPK), and sawdust during the year prior to harvest; and (2) to assess the effects of treat- ments on growth of rillscale during the year following treatment. Methods Study Area and Treatments The study area is located just west of the central Black Hills near Upton, Wyoming, on the Mowry shale formation. Sagebrush {Arte- misia tridentata) is the predominant vegeta- tion on this grassland, with scattered stands of ponderosa pine (Pimis ponderosa). Annual precipitation averages 350 mm (National Oceanic and Atmospheric Administration 1981), falling mostly during the growing sea- son from May to September. Soils are gener- ally shallow and poorly developed. An area was selected on unreclaimed hen- tonite mine spoil that was mined before 1968 on the property of American Colloid. The ex- perimental design was that of a 2 factorial arrangement of treatments with each of three spoil amendments at two levels (Voorhees et al. 1987). One level was the absence of each amendment, while the other level was the presence of the amendment. The study site was rototilled to a depth of approximately 5 cm, and gypsum was applied at a level of 31 metric tons per hectare. Intro- duction of Ca^^ in the form of CaSOj was USDA Forest Service, Rocky Mountain Forest and Range E.xperiment Station. Rapid Cit\ , South Dakota 57701. 57 58 M. E. VOORHEES [Volume 50 intended to facilitate exchange with monova- lent sodium, which would encourage floccula- tion and water penetration (Brady 1974) and discourage surface crust formation. Fertilizer was added at the rate of 1 14 kg nitrogen, 23 kg phosphorus, and 50 kg potassium per hectare. Nitrogen and phosphorus were added as ammonium nitrate (NH4NO3) and diammo- nium phosphate ((NHJoHPO^). Potassium was added as potassium chloride (KCl). Saw- dust was added at the ratio of one part sawdust to two parts spoil (by volume). Inorganic ni- trogen (NH4NO3) corresponding to 0.6% of sawdust (by weight) was added to the sawdust before mixing with spoil to prevent a large increase in the carbon-to-nitrogen ratio and subsequent tie-up of soil nitrogen by micro- organisms (Allison 1965). This amount of nitrogen corresponded to 6 kg nitrogen per metric ton of sawdust. The sawdust amend- ment was intended to increase structural sta- bility and tilth of spoil as well as air and water permeability (Voorhees et al. 1983, 1987). The effects of organic matter additions in the form of sawdust might be expected to increase the stability of the substrate where organic matter is less than 2% (Marshall and Holmes 1979) as in bentonite mine spoil (Uresk and Yamamoto 1986). Gypsum and sawdust amendments were manually incorporated into tilled spoil, whereas the fertilizer amendment was ap- plied to the surface. All eight combinations of the three amendments, including control, were replicated twice to give a total of 16 plots, each 60 x 150 cm. The plots were tilled, amended, and seeded on 8 May 1982. Plots were self-seeded in 1983 as no seed was planted that year. Rillscale seed, obtained from sites along the Montana-Wyoming border during late sum- mer of 1980, was planted in each plot so that seed weights corresponded to approximately three live seeds per cm". This weight of seed was calculated from total percentage germina- tion and seed density determinations made within six weeks of planting. Seed was broad- cast on the surface and raked (1 cm) into spoil. One-half of each plot was harvested for chemical analysis by manually cutting off stems at ground level during estimated peak standing crop (7 July 1983). The other half was harvested approximately six weeks after estimated peak standing crop (17 August 1983) to determine the rate of decline in standing crop resulting from drying and shat- tering of foliage as the season progressed. It was assumed that harvesting one-half of each plot had an insignificant effect on plants on the remaining half. All harvested biomass was oven-dried at 55 C, weighed, and ground through a 20-mesh screen. Plant Tissue Analyses Plant tissue analyses included total nitro- gen by conventional micro-Kjeldahl, in-vitro, dry-matter digestibility, and percentage ash (Church and Pond 1978). Duplicate samples of plant tissue were analyzed to determine nitrogen, ash, and dry-matter digestibility. Dry-matter digestibility was determined with acid pepsin using two 48-hour digestions in a rumen buflPer solution taken from cattle eat- ing grass hay (Tilley and Terry 1963). Crude protein percentage was estimated from Kjel- dahl nitrogen (CP = N% x 6.25). Digestible energy (DE) was estimated from dry-matter digestibility values (Rittenhouse et al. 1971) and converted to metabolizable energy (ME) (Mcal/kg dry matter) using the following for- mula (Swift 1957): DE(Mcal/kg) X 0.79 = ME(Mcal/kg). Elemental concentrations of nitric acid- extractable aluminum, arsenic, barium, bor- on, cadmium, calcium, chromium, copper, iron, magnesium, manganese, moKbdenum, nickel, phosphorus, potassium, selenium, sodium, strontiinn, titanium, and zinc were determined for the plant tissue. Samples were analyzed in duplicate; checks (standards) and blanks were included. Elemental con- centrations of nitric acid extracts were mea- siued using inductively coupled plasma atomic emission spectrometr\- (ICP-AES) (Fassel and Knisely 1974, Jones 1977) on the nitric acid digestion (Havlin and Soltanpour 1980, Gestringand Soltanpour 1981). Statistical Analysis A three-wax factorial anaK sis of \ariance was used to determine the effects of spoil amendments (gxpsum, sawdust, and fertil- izer) and associated interactions on each fo- liage property. Significant differences were accepted at the .05 probability level. 19901 1{| 1 .1 ,s( A 1 ,1", F( )K.\( ;!■: Qr Ai.i I'v 59 TahlI', 1. CMieiiiical fonipositioii oi tlic loliauc ol rillscalc urown on hcnloiiitc iiiiiic s|)()il averaged across (rratiuciit that did or did not include amendment \\ itti sawdust, NPK fertilizer, or gypsum. nits) Sawdust NPK Gy psnni Property (u without with without with without with Standing en Dp (kg/ ha) 955*' 1,973 1,297 1,631 1,534 1,394 Dr\-matter digest ihilitv (%) 73 72 73 72 73 73 Digestible ( ■nergy (keal'/kgDM)- 2,968 2,923 2,970 2,920 2,954 2,955 Metaholizal )le energy (Meal/kgDM) 2.35 2.31 2.35 2.31 2.32 2.33 Kjeldalil nit rogen i {%) 1.70* 1.87 1.74* 1.83 1.74 1.83 Crude protein (%) IJ* 12 11 12 11 12 Ash (%) 42* 35 40 37 38 39 Ca (%) 0.47 0.43 0.45 0.45 0.43 0.47 Mg (%) 0.99 0,92 0.98 0.94 0.96 0.95 ?(%) 0.19 0.17 0.18 0.18 0.18 0.18 Ca:P 2.55 2.57 2.65 2.47 2.47 2.64 Na (%) 8.55 8.18 8.30 8.43 8.63 8,11 K(%) 1.13 1.10 1.11 1.12 1.12 1.11 Zn (fxg/g) 66 63 61 68 55 74 Fe (|JLg/g) 10,775 11,128 9.924 8,188 3,999 10,775 Mn (M-g/g) 496 297 450 343 332 461 N (jig/g) 8 6 8 7 7 8 Cr (|jLg/g) 5 5 5 5 5 5 CU (fJLg/g) 7* 6 6 6 6 7 Mo (|xg/g) 16 17 21 13 17 16 Cu:Mo 0.5 0.5 0.5 0.4 0.4 0.5 B (|xg/g) 35* 41 37 39 39 37 AKfig/g) 1,296* 1,006 1,176 1,126 938* 1,365 Ba(^lg/g) 37 32 35 34 28 41 Sr (|xg/g) 69 73 72 69 68 73 Ti (|xg/g) 10 8 9 9 8* 10 Means for each propert\ that are followed In "Based on in-\itro, dr\ -matter dijjestiliility. an asterisk (*) are siunihrantly different {p > .0.5). Results Peak standing crop averaged 1,464 kg dry matter per hectare (Table 1) and ranged from 267 to 2,913 kg dry matter per hectare. Peak standing crop was 107% greater and ash aver- aged 17% lower on plots that had been amended with sawdust (alone or in combina- tion with other amendments) than on plots that had not been amended with sawdust. Standing crop decreased without grazing by 21% from early July to mid-August. Digestible energy of rillscale foliage at peak of standing crop (based on IVDMD) was 2,948 kcal/kg dry matter (Table 1). Digestibility of dry matter and estimates of digestible and metabolizable energy levels were all signifi- cantly decreased when sawdust and fertilizer were used in combination (with or without the gypsum amendment) relative to the use of other combinations of amendments. Crude protein of rillscale foliage ranged from 9 to 14%. Amendment of spoil with saw- dust (alone or in combination with other amendments) significantly increased the con- centration of nitrogen in foliage from 1.70 to 1.87% and increased the crude protein rating from 11 to 12% relative to foliage on spoil not amended with sawdust (Table 1). Calcium and magnesium levels in rillscale foliage averaged 0.45 and 0.96%, respectively (Table 1). The level of phosphorus was 0. 18%, and the ratio of calcium to phosphorus was 2.6:1. Levels of sodium and potassium in rillscale foliage were 8.37 and 1. 12%, respec- tively. When sawdust and fertilizer amend- ments were used in combination (with or without the gypsum amendment), the foliar content of magnesium and potassium de- creased relative to concentrations in foliage on spoils amended with either of these two amendments alone. Zinc levels in rillscale foliage averaged about 65 |JLg/g (Table 1). Iron and manganese levels were 9,131 and 397 |Jig/g, respectively. Nickel levels averaged 7 fxg/g, whereas chromium levels were 5 |xg/g. The concentration of copper in foliage was 6 jxg/g, while molybdenum concentration was 17 |xg/g (Table 1). The sawdust amendment 60 M. E. VOORHEES [Volume 50 (alone or in combination with other amend- ments) decreased foliar copper from 7 to 6 |JLg/g but did not significantly alter the ratio of copper to molybdenum. High levels of aluminum and iron in rillscale foliage caused severe spectral inter- ferences for arsenic and selenium; thus, it was not possible to determine the concentrations of these elements. Foliar aluminum levels ranged from 1,000 to 1,300 (xg/g. The gypsum amendment (alone or in combination with other amendments) significantly increased foliar aluminum levels from an average of 938 to 1,365 fxg/g (Table 1) relative to foliage on spoil that had not been amended with gypsum. Amendment of spoil with sawdust (with or without other amend- ments) resulted in a decrease in the concen- tration of aluminum in foliage compared with foliage from spoil not amended with sawdust. Cadmium levels in rillscale foliage were below detection limits (1.0 |xg/g) for the ICP-AES procedure. Boron concentrations in foliage averaged 38 |Jig/g and were signifi- cantly greater when sawdust was added to spoil (with or without other amendments) than they were in foliage grown on spoil not amended with sawdust (Table 1). Barium levels in foliage averaged 35 |xg/g, while strontium concentrations averaged 71 |jLg/g. When sawdust and fertilizer amend- ments were used in combination (with or without the gypsum amendment), strontium levels were greater than when either of these amendments was used alone. Titanium concentrations in foliage aver- aged 9 |xg/g and increased by 25% when gyp- sum was added (with or without other amend- ments) relative to foliage from spoil not treated with gypsum (Table 1). The fertilizer amendment (with or without other amendments) had little effect on foliar composition (Table 1) except through interac- tion with the sawdust amendment. Dlsc:ussi()N The forage utility of rillscale as an annual forb is probably limited to a few months din- ing the growing season, culminating with peak of growth in late June and rapidly declin- ing thereafter. Late in the growing season most species of forbs fail to meet the protein and energy needs of gestating animals and are considered inadequate as forage after the fruiting stage (Cook 1972, Stoddart et al. 1975). The (juantities of forage available from growth of rillscale on spoil would be inade- quate for most grazing uses except during a few months of the year. Standing crop de- clined by 21% without grazing from early July to mid-August. This decline following matu- rity was attril)uted to drying and shattering of foliage. No evidence of grazing by insects was observed. However, rillscale could make an important contribution to the nutrition of live- stock and wildlife during the short period of time it is growing and available. Other plant species could be planted with rillscale (Uresk and Yamamoto 1986, Welch 1989) to help meet the nutritional requirements of herbi- vores. The sawdust amendment increased stand- ing crop and decreased ash by improving plant-water relations and increasing the availability of nitrogen. Increased availability of water reduces plant requirements for salts; conversely, plants under water stress accu- mulate other nutrients when nitrogen is limit- ing (Mengel and Kirkby 1982). Decreases in plant ash as a result of the sawdust amend- ment would account for significant decreases in plant uptake of copper and alumimun. The foliage of rillscale at peak standing crop contained adequate digestible energy (based on R'DMD), crude protein percentage, and concentrations of all mineral elements except phosphorus for cattle, sheep, and wild rumi- nants (National Research Council 1975, 1984, Dean 1980, Stone et al. 1983). The sawdust amendment (either alone or in combination with other amendments) resulted in an in- crease in the concentrations of nitrogen and crude protein, while addition of both sawdust and fertilizer (with or without the gypsum amendment) decreased dr\ -matter digestibil- ity and estimated digestible and metaboliz- able energy. Phosphorus supplementation would be advisable for animals foraging on bentonite- mined lands revegetated with rillscale. Cal- cium levels exceeded most dietar\ require- ments of livestock (National Research Council 1975, 1984, W(4ch 1989) but were marginal for deer (Di>an 1980). Ade(iuate fresh water at low salinity levels would also be necessary, since rillscale contains high (quantities of sodium and potassium. Toxicities of electrolytes are 19901 Rii.i.scALK lM)KA(;i:()rAi,irv 61 considered unlikeK unless \\atc>r inlake is re- stricted or water is highly saHne ((Church and Pond 1978). Iron, manganese, and ahnninuin were also present in \er\' high concentrations in the fohage of rillscal(\ wliich may depress cellulose digestion (National Research Coun- cil 1984, Martinez and Church 1970, Grace 1973). Amendment of spoil with gypsum in- creased, whereas addition of sawdust de- creased, the concentration of foliar alu- minum, an important consideration because aluminum can cause gastrointestinal irritation or produce rickets by inteifering with phos- phate absorption if present in large (|uantities in the diets of some animals (Underwood 1977). Finally, the copper-to-molybdennm ratio of the foliage of rillscale was low at 0.7 and could cause molybdenum-induced cop- per deficiencies in livestock and wildlife that do not have access to copper supplements or forages high in copper concentration (Milt- more and Mason 1971, Stone et al. 1983). Other micro- and macro-minerals were ade- quate to meet the requirements of most herbivores. The fertilizer treatment (with or without other amendments) had little effect on foliage composition, except through an interaction with the sawdust amendment. The fertilizer amendment may have been ineffective for in- creasing the availability of nitrogen, phospho- rus, and potassium in spoils, or other condi- tions may have inhibited uptake of these ions. Added nutrients were probably not leached below rooting depth since permeability of unamended spoil is extremely low. Loss of fertilizer as runoflp may have been a factor and thus would explain the interaction between fertilizer and sawdust amendments, since sawdust amendment has been shown to in- crease infiltration and decrease runoff (Voor- hees 1986). Alternately, these elements might not have been limiting to plant growth. The latter hypothesis seems unlikely because the sawdust amendment was effective for increas- ing the level of foliar nitrogen. Rillscale would be a good choice to consider in revegetating bentonite mine spoils because it provides substantial quantities of forage and nutritional qualities generally ade(|uate to meet requirements of livestock and wildlife. For the few nutritional inadecjuacies and toxi- cities of rillscale, introduction of other plants on bentonite spoils may be feasible. Also, graz- ing natixc xcgetation of the surrounding area should be encouraged. A(:kn()Wi,I',d(;mknts This study was conducted in cooperation with the Department of Range Science, Colo- rado State University, Fort Collins. Thanks are extended to Dan Uresk and joe Trlica for their assistance and encouragement through- out the study. Literature Cited Allison. F. E. 1965. Decomposition of wood and liark sawdusts in soil, nitrogen recjuirements, and effects on plants. USDA, Agricultural Research Service Technical Bulletin 1332. 57 pp. BiDW ELL. G L . AND E O WooTON. 1925. Saltl)ushes and their allies in the United States. USDA Bulletin 1245. 40 pp. Brady, N. C 1974. The nature and properties of soils. 8th ed. Macmillan Puhl. Co. Inc., New York. 639 pp. CHURCfL D C . AND W G. Pond. 1978. Basic animal nutrition and feeding. O & B Books, Corvallis, Oregon. 300 pp. Cook, C. VV 1972. Comparative nutritive values of forbs, grasses, and shrubs. Pp. 303-310 in C. M. McKeil, J. P. Blaisdell, and J. R. Goodin, eds., Wildland shrubs — their biology and utilization. Proceedings of a symposium, July 1971, Logan, Utah. Intermountain Forest and Range Experi- ment Station, Ogden, Utah. Dean, R E 1980. The nutrition of wild ruminants. Pp. 278-305 JJi D. C. Church, ed., Digestive physiol- ogy and nutrition of ruminants. 2d ed. O & B Books Inc., Corvallis, Oregon. Fassel, V A , AND R. N Knlsely 1974. ICP-optical emission spectroscopy. Analvtical Chemistrv 46: UIOA. Frankton, C, AND I. J. Bassett 1970. The genus Atriplex (Chenopodiaceae) in Canada. II. Four native western annuals: A. argentea, A. truncata, A. powelli. and A dioica. Canadian Journal of Botany 48: 981-989. Cestring, W D., and p. N. Soltanpour 1981. Boron analysis in soil extracts and plant tissue by plasma emission spectroscopy. Communities in Soil Science and Plant Analysis 12: 7.33-742. Grace. N D. 1973. Effect of high dietary Mn levels on the growth rate and the level of mineral elements in the plasma and soft tissues of sheep. New Zealand Journal of Agricultural Research 16; 177-184. Hanlin. J L . AND P N Soltanpour 1980. A nitric acid plant tissue digest method for use with inductively coupled plasma spectrometry. Communities in Soil Science and Plant Analysis II: 969-980. Jones. J B , Jr 1977. Elemental analysis of soil extracts and plant tissue ash b\' plasma emission spec- troscopy- Communities in Soil Science and Plant Analysis 8: 349-365. Marshall, T J , and J \V Holmes. 1979. Soil physics. Cambridge University Press, New York. 345 pp. 62 M. E. VOORHEES [Volume 50 Martin, A C , H S Zim. and A L Arnold 1951. Ameri- can wildlife and plants: a guide to wildlife food habits. McGraw-Hill Book Co., New York. 500 pp. Martinez, A., and D. C. Church 1970. Effect of various mineral elements on in-vitro rumen cellulose di- gestion. Journal of Animal Science 31: 982-990. Maynard, L. A, and J K Loosu 1969. Animal nutrition. McGraw-Hill, New York. 613 pp. Mengel, K., and E a. Kirkbv 1982. Principles of plant nutrition. 3rd ed. International Potash Institute, Worblaufen-Bern, Switzerland. 655 pp. MiLTMORE.J. E.andJ L. Mason 1971. Copper to molyb- denum ratio and molybdenum and copper concen- trations in ruminant feeds. Canadian Journal of Annual Science 51: 193-200. National Research Council 1975. Nutrient require- ments of domestic animals. No. 5. Sheep. 5th ed. National Academy of Science, Washington, D.C. 72 pp. 1984. Nutrient requirements of domestic animals. No. 4. Beef cattle. 6th ed. National Academy of Science, Washington, D.C. 90pp. National Oceanic and Atmospheric Administration 1981. Wyoming climatoiogical data: annual simi- mary. Environmental Data and Information Cen- ter, National Climatic Center, Asheville, North Carolina. Rittenhouse, L. R., C. L Streeter, and D C Clanton 1971. Estimating digestible energy from di- gestible dry and organic matter in diets of grazing cattle. Journal of Range Management 24: 73-75. ROBINETTE. W L 1971. Browse and cover for wildlife. Pp. 69-70 in C. M. McKell, J. P. Blaisdell. and J. R. Goodin, eds., Wildland shrubs — their biol- ogy and utilization. Proceedings of a symposiimi, July 1971, Logan, Utah. Intermountain Forest and Range Experiment Station, Ogden, Utah. SiEcC H , D W Uresk, andR M Hansen 1983. Plant- soil relationships on bentonite mine spoils and sageljrush-grassland in the northern high plains. Journal of Range Management 37: 289-294. Stoddart, L a, a. D. Smith, andT W Box 1975. Range management. 3rd ed. McGraw-Hill Book Co., New York. 532 pp. Stone, L. R , J A Erdman, G. L Feder, and H D Holland. 1983. Molybdenosis in an area under- lain by uranium-bearing lignites in the northern Great Plains. Journal of Range Management 36: 280-285. Swift, R W 1957. The nutritive evaluation of forages. Pacific Agricultural E.xperiment Station Bulletin 615. 34 pp. Tilley, J M . and R a. Terry 1963. A two-stage tech- nicjue for the in vitro digestion of forage crops. Journal of British Grassland Society 18: 104-111. Underwood, E J 1977. Trace elements in human and animal nutrition. 4th ed. Academic Press, New York. 545 pp. Uresk, D W . andT Ya.mamoto 1986. Growth of forbs, shrubs and trees on bentonite mine spoil under greenhouse conditions. Journal of Range Manage- ment 39: 113-117. VooRHEES, M E 1986. Infiltration rate ofbentonite mine spoil as affected b\ amendments of gypsum, saw- dust and inorganic fertilizer. Reclamation and Revegetation Research 5: 483-490. \'()()RHEEs, M E , D. W. Uresk, andR M Hansen 1983. Atriplcx sttckleyi (Torrey) Rydb.: a native annual plant for revegetating bentonite mine spoils. In: Arthur R. Tiedcmann, E. Durant McArthur, Howard C. Stutz, Richard Stevens, and Kendall L. Johnson, compilers. Proceedings — symposium on the biology ol Atriplex and related chenopods, 2-6 May 1983, Provo, Utah. USDA Forest Ser- vice General Technical Report INT-172, U.S. De- partment of Agriculture, Forest Service, Inter- moimtain Forest and Range Experiment Station, Ogden, Utah. .309 pp. N'ooRHEEs, M E , M J Trlica, and D W Uresk 1987. Growth of rillscale on bentonite mine spoil as influenced by amendments. Journal of Environ- mental Quality 16: 411-416. Welch, B 1989. Nutritive value of shrubs. In: C. M. McKell, ed. , The biology and utilization of shrubs. Academic Press, Inc., New York. Received 15 May 1989 Accepted 15 December 1989 :,ieat Basin Naturalist 50(1 1. 1990, i)i). 63-fi5 SUMMER FOOD HABITS OF COYOTES IN IDAHO'S KIN'EK OF NO RETURN WILDERNESS AREA Cliai L. Kllu.lt' l^itliaid (;iK'tiji;~ Abstiucc. — SuiiiiiUM- food habits of fONotcs {C'linis latiiiii.s) in tfu' llixtTof No Hetiiin WiKfcnifss Aix^a, Idaho, were determined. Anal\'.si.s of 51 seats (feeal samples) revealed tliat (^oluiiihiaii y;roiiiid sijuirrels {S))('ntio))liilus coluin- biantis ), mule deer (Odocoiletis hemionus ), and deer miee (Pcraniijsctis inaniculatus ) exhibited the greatest f reciuency ofoeeurrenee for identified food items, Iieing deteeted in 57%, 27%, and 16%, respeetively, of seats examined. One of the most iihiciuitous and adaptable predators of the American West is the coyote (Canis latrans). As man altered habitats in the western states, the coyote adapted its behav- ior and diet to take advantage of these new environments. Being generally dietary oppor- tunists (Johnson and Hansen 1977), coyotes have found prev to their liking on man's range- land (Murie 1951, Short 1979, Green and Flinders 1981) and farms (Gipson 1974), and in his cities (MacCracken 1982). Although many aspects of coyote ecology in man-altered or man-impacted areas of the West have been investigated, less is known of the role of the coyote in relatively undisturbed wilderness. The objective of this study was to determine the summer food habits of coyotes in Idaho's River of No Return Wilderness Area (RNRWA). Study Area and Methods The study was conducted in the Big Creek Ranger District, RNRWA (formerly the Idaho Primitive Area). A description of the RNRWA and Big Creek area has been provided by Hornocker (1970). Canid scats were collected from trails located in the Big Creek drainage of the RNRWA. Trails were surveyed the beginning of May 1977 and 1978, and all scats encoun- tered were removed. After the initial clear- ing, trails were surveyed at least once a month for newly deposited scats. Scat collection con- cluded at the end of August 1977 and 1978. Collected scats were air-dried and weighed, Department of Botany and Range Sciences, Biighani Young University, P Kentucky University, Richmond, Kentucky 40475-09.50 "Department ol Biological Sciences, Eastern Kentucky I'niversity, Ricliniond, KentMck\ 4047.5-09.50 and diameter at the widest point was deter- mined. Using criteria established in other western studies (Weaver and Fritts 1979, Green and Flinders 1981, Danner and Dodd 1982), we classified all scats > 20 mm in di- ameter as coyote. Scats were washed, sepa- rated, and prepared for analysis in a manner similar to that described by Johnson and Hansen (1979). Prepared scats were analyzed following the procedure of Green and Flinders (1981). Hair was identified by medullary characters (Moore et al. 1974). Teeth were also used to verify the animal species consumed. Each coyote scat was treated as an individual observation. No at- tempt was made to determine the density of potential prey items in the Big Creek area; hence, it was not possible to determine pref- erence indices for the items identified in the scats examined. Results and Discussion Fifty-one scats collected met the > 20-mm- diameter criterion and were classified as coy- ote. The average dry weight (± SD) of indi- vidual coyote scats was 15.3 ± 5.9 g. Soluble endogenous material accounted for an aver- age 3.9 ± 2.3 g (25%) of dry weight/scat. Thirteen mammal species were identified as food items consumed by coyotes during the summer in the RNRWA (Table 1). Percent occurrence of identified food categories was as follows: rodents 100%, Cervidae 41.4%, insects 39.2%, birds 27.4%, reptiles 3.9%, .o, Utah 84602. Present address; Department of Biological Sciences, Eastern 63 64 C. L. Elliott AND R. Guetig [Volume 50 Table 1. Percent occurrence of material identified in 51 coyote scats. River of No Return Wilderness Area, Idaho, May-August 1977 and 1978. Montli Species identified May (9)' June (14) Juh (15) August (13) Total (51) Columbian ground squirrel (Spermuphiltis columbiautis ) 42.9 73.3 92.3 .56.8 Mule deer (Odocoilcus hcmionus ) 33.3 35.7 13.3 30.7 27.4 Deer mouse 22.2 14.2 20.0 7.6 15.6 {Peromijsciis maniculatns ) Moose (Alces alces) 33.3 21.4 7.6 13.7 Northern pocket gopher (Thomomys talpuides ) 11.1 14.2 20.0 11.7 Montane vole {Microtus montaniis) 22.2 21.4 6.6 11.7 Golden-mantled ground squirrel {Spennophihts lateralis ) 22.2 6.6 5.8 Audubon's cottontail {Sylvila^us auduhonii) 11.1 7.6 3.9 Horse (Equiis caballus) 11.1 7.1 3.9 Long-tail weasel {Mustela frenata) 7.1 1.9 Northern water shrew {Sorex palustris) 7.1 1.9 Water vole {Arvicola richardsom) 7.1 1.9 Snowshoe hare 7.1 1.9 (Lcpii.s amcricanii.s) Unknown Mammals Reptiles Arthropods Birds Plant matter 22.2 11.1 11.1 11.1 7.1 21.4 7.1 7.1 13.3 53.3 33.3 40.0 61.5 53.8 23.0 5.8 3.9 39.2 27.4 21.5 Number of scats exaininecl domestic livestock 3.9%, and other carnivores 1.9%. The results of this study correspond favorably with those of Ribic (1978), Johnson and Hansen (1979), and Short (1979), in that rodents were the most frequently identified food category in the summer diet of western coyotes. The Columbian ground scjuirrel (Sper- mopJiilus columhianus) was the most fre- ((uently occurring food item identified in siun- mer coyote scats, being found in 29 (56.8%) of the 51 scats examined. The percent occiu- rence of Columbian ground scjuirrel remains in summer scats reflects the seasonal availabil- ity of the s{}uirrel as a prey item. Sper- mophilus columhianus within the RNRWA emerge from hibernation in late May and re- main active above ground until late August- early September (Elliott and Flinders 1980). The seasonal importance of Columbian ground scjuirrels as a prey species for other predators (i.e., mountain lions [Fclis con- color]) in the RNRWA was noted by Seiden- sticker et al. (1973). Increased mountain lion activity dining the day in summer was felt to be related to the availability of Columbian groimd s(iuirrels as a food item (Seidensticker et al. 1973). The presence of lesser species in the summer diet of lions was thought to hold down an\ increases o\ er the lion s winter kill rate of elk (Cervus i'l(iphti.s) and mtile deer (Odocoilcus Jicmionus) (Hornocker 1970). Although coNotes and mountain lions uti- lize a common food resource, it is doubtful that they are serious summer dietary competi- tors. Elk and mide deer are the major food 1990] HiLLscALK F()ha(;k Quality 65 items consumed I)\ niouiitain lions cknin^ the summer in the HNHW'A (Hornoeker 1970), whereas rodents comprise the hulk of sununer items consiuned In coyotes (see Tahle I). In the hierarchy of predators in the RNRWA, the coyote appears to occupy a trophic level below that of the mountain lion. Acknowledgments The senior author thanks Dr. Jerran Flin- ders, Brigham Young University, for use of laboratory facilities in the preparation of scat samples, and the University of Idaho s Wilderness Research Committee for permis- sion to use the facilities at the Taylor Ranch Field Station, River of No Return Wilderness Area. Literature Cited Danner, D. a . AND N DoDD 1982. Comparison of coy- ote and gray fox scat diameters. Journal of Wildlife Management 46; 240-241. Elliott, C. L., and J T. Flinders 1980. Seasonal activity patterns of Columbian groimd squirrels in the Idaho Primitive Area. Great Basin Naturalist 40: 175-177. GiPSON, P. S. 1974. Food habits of coyotes in Arkansas. Journal of Wildlife Management 38: 848-853. Green, J. S , and J T. Flinders 1981. Diets ofsympatric foxes and coyotes in southeastern Idaho. Great Basin Naturalist 41: 251-254. lloUNOCKKH. M C 1970. .\n anahsis ol inouiilain lion predation ni^on mule deer in the Idaho Primitive Area. Wildlife Monograph 21. Johnson, M K , and H. M. Hansen. 1977. Foods of coy- otes in the Lower Grand Canyon, Arizona. Ari- zona Academy of Science 12: 81-83. 1979. (Coyote food habits on the Idaho National Engineering Laboratorv. loiunal of Wildlife Man- agement 43: 951-956. MacCracken, J. G. 1982. Coyote foods in a southern California suburb. Wildlife Socictx BulietiTi 10; 280-281. Moore, T. D., L E Si'ence. andC. E. Ducnolle. 1974. Identification of the dorsal guard hairs of some mannnals of Wyoming. Wyoming Game and Fish Department Bulletin 14. 177 pp. MURIE. A 1951. Coyote food habits on a southwestern cattle range. Journal of Manuualogy .32; 291-295. RiBic, C. A. 1978. Summer foods of coyotes at Rocky Flats, Colorado. Southwestern Naturalist 23; 152-1.53. SEIDEN.ST1CKEH, J C IV, M G HORNOCKEH, W. V WiLES, and J P Messick 1973. Mountain lion social orga- nization in the Idaho Primitive Area. Wildlife Monograph 35. Short, H L 1979, Food habits of coyotes in a semidesert grass-shrul) habitat. U.S. Forest Service, Rocky Mountain Forest and Range Experiment Station, Research Note RM-364. 4 pp. We.wer, J L.. AND S H Fritts. 1979. Comparison of coyote and wolf scat diameters. Journal of Wildlife Management 43; 786-788. Received 1 J tme 1989 Accepted 10 October 1989 Cn'.it Hasi.i Naturalist 50(1 1, mx). pp. H7-72 INFECTION OF YOUNG DOUGLAS-FIRS BY DWARF MISTLETOE IN THE SOUTHWEST Rohcrt L. Mathiascn , Clarlcton B. luliiiiiistcr", and I'raiik ('.. Ilawkswortlr Abstract. — Stuclies in several areas in Arizona and New Mexico show that dwarf mistletoe {Arcetithohium doug- lasii) is rare in yonng Douglas-firs growing under infected overstories. Less than 5% of the Douglas-firs under 26 years old and less than (Wc of those under 1.4 m tall were infected in 77 mistletoe-infested stands. Both percetit infection and mean dwarf niistletoi- rating of young Douglas-firs increased as tree age, height, and stand dwarf mistletoe ratings increased. Douglas-fir dwarf nii.stletoe {ArceiitJiobium (lou^Iasii Engelni.) is tlie most prevalent and damaging disease agent in southwestern mixed-conifer forests (Andrews and Daniels 1960, Hawksworth and Wiens 1972, Jones 1974). This parasitic flowering plant occurs throughout the range of its principal host, Douglas-fir {Pseudotsu^^a menzicsii [Mirb.] Franco), in the Southwest. Andrews and Daniels (1960) estimated that approximately 50% of the Southwest's Douglas-fir type was infested by dwarf mistletoe. Douglas-fir regeneration is a frequent com- ponent of the understory of southwestern mixed-conifer stands (Moir and Ludwig 1979, Gottfried and Embry 1977, Fitzhugh et al. 1987). When overstories are infested with dwarf mistletoe, spread to young and advance regeneration perpetuates the infestation over time. Therefore, management of mixed- conifer forests should attempt to minimize the infection of new and established regeneration from alreadv infested overstories (Jones 1974, Gottfried and Embry 1977). Mathiasen (1986) summarized previous re- search on this problem and the factors that influence dwarf mistletoe infection; he also provided some preliminary information on in- fection of young Douglas-firs and spruces in the Southwest. He found that little infection of Douglas-fir occurs before saplings are 26 years old. Only 6% of the Douglas-firs he sampled that were less than 26 years old were infected, whereas infection of older Douglas- fir reproduction averaged 83%. Mathiasen (1986) also related infection of Douglas-firs less than 26 years old to three factors affecting infection ot young trees listed by Wicker (1967). These included exposure time, over- story inoculum levels, and sapling density. During a study designed to collect growth data for the development of a regeneration model for southwestern mixed-conifer stands, additional data on the infection of young Douglas-firs were collected from 13 mistletoe- infested stands in the White Mountains, Ari- zona. These data were combined with the original data collected by Mathiasen (1986), and the results are reported here. In addition, the entire data set was summarized using the heights of sampled Douglas-firs because pre- vious investigators have suggested that height may be a critical factor influencing infection of young trees bv dwarf mistletoes (Graham i960, Hawksworth 1961, Childs 1963, Wicker and Shaw 1967, Scharpf 1969). Methods During 1980-81 Douglas-fir regeneration was sampled in 64 mistletoe-infested mixed- conifer stands in four national forests in Ari- zona and New Mexico. A total of 364 Douglas- fir saplings were sampled for total age, height, and height to live crown. In addition, each Douglas-fir was examined for dwarf mistletoe infection and assigned a dwarf mistletoe rating (DMR) using the 6-class system (Hawksworth 1977). This rating system divides the live crown of a tree into thirds, and each third is 'U.S. Forest Servicf, Fore.st Pest Management. .324 2.5tli St., Ogden. Utali 84401. 'U.S. Forest Service, Rocky Mountain Forest and Range Experiment Station, 240 W. Prospect St., Fort Collins, Colorado 80.526. 67 68 R. L. Mathiasen etal. [Volume 50 rated separately as: 0, no mistletoe infection; 1, less than 50% of live branches infected; 2, more than 50% of live branches infected. The ratings for each third are totaled to obtain the DMR for a tree. Mean stand DMR and mean sapling DMR are calculated by adding the DMRs for all live overstory trees or saplings and dividing by the total number of live trees or saplings, respectively. Infection intensity is defined here as the mean DMR of the over- story or saplings in a stand. Overstory data collected for the 1980-81 stands were from rectangular plots ranging from 0.04 to 0.36 ha. For each live tree over 1.4 m in height the species, diameter at breast height (dbh) to the nearest 2.54 cm, DMR, and crown class (dominant, co-dominant, in- termediate, or suppressed) were recorded. These data provided information on overstory dwarf mistletoe infection intensity, species composition, and stand structure. In 1988 an additional 334 Douglas-fir sap- lings were sampled in 13 mistletoe-infested, mixed-conifer stands in the White Mountains, Arizona. Data were collected as in 1980-81. Overstory data collected were the same as in 1980-81 but 0.04-ha circular plots were used. Stand dwarf mistletoe ratings were calcu- lated using all live Douglas-firs greater than 2.54 cm dbh for 1980-81 plots and greater than 5.08 cm dbh for 1988 plots. Sapling crown ratios were calculated by subtracting height to live crown from total height and then dividing by total height. Percent infection and mean DMR for saplings were calculated by five-year age classes and .3-m height classes for each of three stand DMR classes (0. 1-1.5, 1.6-3.0, and greater than 3.0). Sapling den- sities were determined for the number of Douglas-fir saplings in 0.04-ha circular sub- plots nested in the center of larger plots in each stand. Results Both the number of infected saplings (per- cent infection) and infection intensitv (mean DMR) increased as total age, total height, and stand DMR increased (Tables 1 and 2). No mistletoe infection was found on saplings un- der 21 years old in stands with a stand DMH less than 3.0, and only five saplings under 21 years old were infected in stands with a stand DMR greater than 3.0 (Table 1). The five infected saplings represent less than 4% of saplings under 21 years old sampled. Infec- tion of saplings less than 16 years old was only 2% in stands with a stand DMR greater than 3.0. Also, very little infection of saplings less than 26 years old was found (Table 1). Only 10% of saplings 21-25 years old were infected, and all were in moderately infested (stand DMR 1.6-3.0) or severely infested stands (stand DMR greater than 3.0). Infection of 26-30-year-old saplings in- creased to 30% in lightly infested stands (stand DMR 0. 1-1.5) and to over 65% in both moderately and severely infested stands (Table 1). Generally, infection continued to increase as sapling age increased (Table 1). A total of 14 infected saplings under 26 years old were sampled. These saplings were in severely infested stands, were over 1.4 m in height, had high crown ratios (greater than 0.70), or were in stands with over 740 saplings per ha. Many of these 14 saplings had more than one of the above factors contributing to their infection potential. Percent infection and mean DMR for saplings demonstrated the same pattern for height classes as for age classes (Tables 1, 2). Little infection (10% or less) was found in saplings less than 1.4 m in height, except in the most severely infested stands, where we found 27% infection in saplings 1.09-1.4 m tall. However, saplings over 1.4 m in height had much higher infection levels (percent infection) and intensities (mean DMR) than smaller saplings (Table 2). Discussion Wicker (1967), Wicker and Shaw (1967), and Mathiasen (1986) discussed several of the factors influencing the infection of young trees by dwarf mistletoes, including duration of ex- posure to inoculum, amount of inoculum, target area, density of regeneration, and re- moval of seeds by wind, snow, and other envi- ronmental factors. Infection of susceptible N'oung trees is largeK influenced b\ a complex interaction of the above factors. Nhithiasen (1986) presented information on the influence of exposure dination to inoculum (as ex- pressed 1)\ tree age), amoimt of inoculum (as expressed by stand DMR), and regeneration density (as expressed b\' number of saplings per ha). Additional information is reported 1990] DWAKl" Ml.SI LKIOI', IM'KCIIOX 69 T.\B1,K 1. Iiiii'c'tioii ol Doiitila.s-lii- sai)liiiiis l)\ at;e cla.ssfs and .stand DM K classt'.s. Stand DMR class 0.1-1.5 1.6-3.0 > 3.0 Total Age class % Mean % Mean % Mean % Mean (years) N Inf DMR N Iiil DMH N Inf DMH N Inf DMH < 16 49 0 0.0 58 0 0.0 108 2 <0.1 215 1 <0.1 16-20 30 0 0.0 29 0 0.0 21 14 0.1 80 4 <0.1 21-25 15 0 0.0 43 9 <(). 1 35 14 0.1 9.3 10 <0.1 26-30 27 30 0.3 38 79 0.8 35 69 0.7 100 62 0.6 31-35 8 25 0.3 39 80 1.1 41 88 0.9 88 78 0.9 36-40 12 66 0.9 40 90 1.4 36 97 1.8 88 90 1.5 41-45 " 71 0.9 15 80 2.0 12 100 2.8 34 85 1.9 Total 148 16 0.2 262 43 0.6 288 41 0.6 698 36 0.5 'Percent inleitioii T.VBLK 2. Infection of Dotiglas-fif saplint^s 1)\- height classes and stand DMR classes. Stand DMR class 0.1-1. 5 1.6-3.0 > 3.0 Total Height class % Mean % M can % Mean % Mean (m) N Inf DMR N Inf DMR N Inf DMR N Inf DMR .15-. 45 23 0 0.0 20 0 0.0 45 0 0.0 88 0 0.0 .46-. 76 14 0 0.0 13 0 0.0 35 3 <0.1 62 2 <0.1 .77-1.07 21 5 <0.1 33 0 0.0 31 6 0.1 85 4 <0.1 1.10-1.37 21 5 0.2 29 10 0. 1 22 27 0.2 72 14 0.2 1.40-1.68 15 27 0.4 44 61 1.2 32 66 1.3 91 57 1.0 1.71-1.98 15 40 0.7 48 65 1.5 41 71 1.6 104 64 1.4 2.01-2.29 9 44 0.6 23 70 1.4 24 75 2.0 56 68 1.3 <2.29 30 23 0.6 52 70 1.2 58 69 1.8 140 59 1.4 Total 148 16 0.2 262 43 0.6 288 41 0.6 698 36 0.5 Percent infectii here for these factors as well as for the in- fluence of target area as expressed by the total height and crown ratio of young infected Douglas-firs. In most situations where in- fected young Douglas-firs (less than 26 years old) were found, a severely infested overstory (stand DMR greater than 3.0) was present. Infected young Douglas-firs in stands with a stand DMR less than 3.0 either had high crown ratios (greater than 0.70), were over 1.4 m tall, or were in stands with regeneration densities over 740 saplings per ha. All three factors would increase the potential for infec- tion of young trees in dwarf mistletoe- infested stands because of greater available target area. Because severe infection by dwarf mistle- toe significantly reduces the growth of mer- chantable-size Douglas-firs in the Southwest and increases mortality of all size classes (An- drews and Daniels 1960, Mathiasen et al. 1990), its control is an important consider- ation for resource managers. Vegetation man- agement plans that do not successfully pre- vent or significantly reduce the infection of Douglas-fir regeneration only serve to perpet- uate mistletoe infestations. Although the total heights, crown ratios, and densities of re- generation contribute to the potential for reinfection of young understories, these fac- tors cannot be managed on a practical basis. However, management plans that remove the most severely infected trees, followed with intermediate sanitation removals, can effec- tively reduce the level of dwarf mistletoe in- fection in stands (Hawksworth 1978), thereby reducing the potential for infection of new or advance regeneration. In addition, because Douglas-fir regeneration is not frequently in- fected before it reaches ages over 25 years in the Southwest, cutting cycles of 20 years or less allow managers at least two management entries for reducing the level of mistletoe in a stand before new Douglas-fir regeneration will be affected. Because little or no infection will occur until Douglas-firs are over 20 years 70 R. L. Mathiasen etal [Volume 50 old in lightly infested stands, the removal of severely infected overstory trees will signifi- cantly reduce the potential for infection of new and advance Douglas-fir regeneration. The age at which Douglas-fir regeneration becomes infected by dwarf mistletoe in the Southwest contrasts sharply with results re- ported for other tree species and regions. Weir (1918) found that the average age of 50 naturally infected Douglas-fir seedlings, used for assessing the effects of dwarf mistletoe on seedling growth in the Northwest, was 18 years. Hawksworth and Graham (1963) found very little infection in lodgepole pine (Piuus contorta Dougl. ex Loud.) reproduction un- der 10 years old, but infection increased markedly in older stands: 9% at age 15, 18% at age 20, and 32% at age 25. Some infection of ponderosa pine (Pinus ponderosa Laws.) by southwestern dwarf mistletoe {Arceuthobium vaginotum subsp. cnjptopodiim [Engelm.] Hawksw. & Wiens) has been foimd in 10-year- old seedlings (Gill and Hawksworth 1954, Hawksworth 1961). Based on these findings for pines, Johnson and Hawksworth (1985) recommended that mistletoe-infected resid- ual trees be removed before the young stand is 10 years old. However, the results of this study indicate that for southwestern Douglas- fir the infected overstory trees could be left for up to 20 years because of the very slight chance of infection. There is less published data for the relation- ship of regeneration height and dwarf mistle- toe infection, but the general recommenda- tion is that mistletoe-infected residual trees should be removed before the yoiuig stand is 0.9 m tall (Johnson and Hawksworth 1985). Graham (1960) found that dwarf mistletoe infection in Douglas-fir increased as size class increased in northern Idaho: Only 15% of the saplings sampled by Graham were infected, whereas 25 and 39% of the small and large poles, respectively, were infected. Hawks- worth (1961) reported that 19% of the pon- derosa pines in the 2.54-cm-diameter class were infected in stands infested by southwest- ern dwarf mistletoe in northern Arizona, but infection increased to 57% in the 12.7-cm- diameter class. Ghilds (1963), working in the Pacific Northwest, found that uninfected ponderosa pines averaged 1.5 and 1.1 m in Graham did not spccih tlic diaiiictfrs ol tlu' si. height in lightK and heavily mistletoe- infested stands, respectively, and infected pines averaged 2.3 and 2.0 m in the same stands. Scharpf (1969) reported that only 7% of true firs under 0.9 m tall were infected in severely infested stands in California but that infection intensified rapidly in taller regeneration. The residts of infection of Douglas-firs by height classes reported here indicate that little infection can be expected until the trees reach heights greater than 1.4 m in the Southwest. Because these findings have important implications in managing dwarf mistletoe- infested stands, similar studies should be con- ducted for other dwarf mistletoe-host combi- nations in other regions of the western United States. The results show that the generally accepted recommendation that infected over- story pines and true firs be removed before the young stand is 10 years old or 0.9 m tall is more restrictive than need be for Douglas- fir in the Southwest, where little infection occurred in stands under 20 years old or less than 1.4 m tall. Literature Cited Andrews, S. R.. and J P Daniels 1960. A survey of dwarf mistletoes in Arizona and New Mexico. USDA Forest Service, Rocky Mountain Forest and Range Experiment Station Paper 49. 17 pp. Childs. T W. 1963. Dwarfmistletoe control opportuni- ties in ponderosa pine reproduction. USDA Forest Service, Pacific Northwest Forest and Range Experiment Station, unnumbered report. 20 pp. FiTzmcu. E. L.. \\ H Moik | A Lldwig. and F RoNCO 19(S7. Forest habitat types in the .\pache, Gila, and part of the Cibola National Forests, Arizona and New Mexico. USDA For- est Ser\ice, General Technical Report RM-145. 116 pp. Gill. L S . and F C; IIawkswohth 1954. Dwarfmistle- toe control in soutliwestcrn ponderosa pine forests under management. Journal of Forestry 52: 347-353. Gottfried. G. J . and R S Embrv 1977. Distriimtion of Douglas-fir and ponderosa pine dwarf mistletoes in a virgin .\rizona mixed conifer stand. USDA Forest Ser\ ice. Research Paper RM-192. 16 pp. Graham. D P 1960. Dwaifmistletoe sur\ey in Nezperce National Forest. USDA Forest Service, Research Note INT-75. 7 pp. 11 \\\ KsuoRTH F (; 1961. Dwarfmistletoe of ponderosa pine in the Southwest. USDA Forest Service, Technical Bulletin 1246. 112 pp. 1977. The 6-class dwarfmistletoe rating s\stem. I'SD.V I-'orest Ser\ ice. General Technical Report RM-48. 7 pp. 1990] D\\ \Hi' Misn.Ki'oK Infkciion 1 197S. liitcriiR'diatc ciittiniis in iiiistlctof-iiiffstcil lodgepok' pint" and soiitliwestcrii lioiickTosa pinr stands, //i; R. F. Scharpland J. H. FarnietiT, tctli. c'Odrdinators, Froceediniis oi the s\inpc)siuni on dwarf mistletoe control throngh forest inanatje- nient. USDA Forest Service, General 'leclniical Report PS\V-31: 86-92. H.WVKSWORTH, F G , AM) D P Gr.uiam 1963. Spread and intensification of dwarfinistletoe in lod^epole pine reprotluction. [oiiriial of Forestr\ 6f: 587-591. H.WVKSWORTH, F G.. AND D W'lENS 1972. Biology and classification of dwarf mistletoes {Arceuthohiiim ). USDA Forest Service, Agriculture Handbook 401.234 pp. Johnson. D. W., and F G. Ha\\k;s\\orth. 1985. Dwarf mistletoes. Candidates for control through cul- tural management. USDA Forest Service. Gen- eral Technical Report \\"0-46; 48-55. Jones. J. R. 1974. Silviculture of southwestern mixed conifers and aspen: the status of our knowledge. USDA Forest SeiAice, Research Paper RM-122. 44 pp. M.\THI.\SEN. R L 1986. Infection of young Douglas-firs and spruces by dwari mistletoes in the Southwest. (Jreat Basin Naturalist 46: 528-534. M MiiiASKN R L F (; Hawksworth. ANuG B Fdmin- STKR 1990. Kffects of dwarf mistletoe on Douglas- fir in the Sontiiwest. (In preparation.) Moiu \\ II \M) I A Lidwk; 1979. Classification of sprnce-tir and mi.xed conifer habitat types of .Ari- zona and New Mexico. USD.A Forest Service, Research Paper RM-207, 47 pp. S( HARPK, R. F. 1969. Dwarf mistletoe on red fir: infection and control in understorx' stands. USDA Forest Service, Research Paper PS\\'-50. 8 pp. Wicker, E. F 1967. Seed destiny as a klendusic factor of infection and its impact upon propagation of Ar- ceuthohitim spp. Phytopathology 57: 1164-1168. \\i( KER. E. F., AND C G Shaw 1967. Target area as a klendusic factor in dwarf mistletoe infection. Phy- topathology 57: 1161-1163. W'lER, J R 1918. Effects of mistletoe on \oung conifers. Journal of Agricultural Research 12: 715-718. Received 31 January 1989 Accepted 16 January 1990 Great Basin Natiinilisl 50(1), UM). pp. 73-H2 NEW MEXICO GRASS TYPES AND A SELECTED BIBLIOGRAPHY OF NEW MEXICO GRASS TAXONOMY' Kelly W. Allred- Abstract. — Collection data, bibliographic citations, and curatorial iiiiorinalioTi on 52 names of New Mexico grass types are compiled. A bil)liography of taxonomic research pertinent to the study of New Mexico grasses is cross-refer- enced with genera known to occur in the state. Bibliographic and historical information are an essential, l)nt often neglected, resoince for the student of plant systematics. The cor- rect application of plant names requires accu- rate information concerning nomenclatural types, and precise floristic and identification work demands access to reliable monographic or revisionary literature. This is especially so when changes are made in traditional systematic alignments; reference literature allows others to understand and evaluate the revisions. Nearly every botanist engaged in the taxon- omy of grasses (Gramineae) dining the past century described at least one novelty from New Mexico material. The list of grass types presented here includes 52 taxa known to have been described from specimens gath- ered in New Mexico. As a point of compari- son, 19 grass taxa have been named from Utah material (Welsh 1982). Twenty-one different authors contributed new taxa; but three, George Vasey (14 names), John Torrey (6 names), and Ernst Steudel (5 names), ac- counted for nearly 50% of the plant names (Table 1). Botanical publication of grass taxa from New Mexico began in 1854 with species of Aristida, Muhlenbergia, Onjzopsis, and Foa (Steudel 1854) and has continued for well over a century, the latest being in 1986 from the genus Andropogon (Campbell 1986). Of the authors, only A. S. Hitchcock, Paul Standley, George Thurber, George Vasey, and E. O. Woo ton also participated as field collectors of new grasses from New Mexico (Table 1). George Vasey heads the list with collections of 9 new taxa from New Mexico. Grant and Santa Fe counties contain the most localities of new grasses (Table 2). Santa Fe is one of the oldest towns in the United States and was visited by many collectors early in the 1800s. William Gambel passed through in 1841 or 1842 on his way to California; his collections were described by Thomas Nut- tall. Wislizenus followed in 1846. August Fendler made extensive collections there in the spring of 1847, sending them to Asa Gray. Most of his collections came from the Santa Fe Creek area and within 10-12 miles of Santa Fe. A. A. Heller, G. R. Vasey, S. M. Tracy, and T. D. A. Cockerell were other botanists who collected near Santa Fe in the late 1800s or early 1900s; the collections of Vasey and Tracy contributed new grasses. Many of the collections from Grant Coimty came from the mining camp of Santa Rita, 15 miles east of Silver City. Charles Wright, J. M. Bigelow, and George Thurber collected there in the 1850s. Mangas Springs, also in Grant County, was visited by O. B. Metcalfe in 1903, who collected several hundred sets of plants. C. G. Pringle, H. H. Rusby, and J. G. Smith also collected new grasses from Grant Countv (Standley 1910). The ensuing list attempts to include all grass names based on New Mexico material. The author of the name, publication data, collector and number, locality of collection, deposition of type material, and current taxo- nomic disposition of the name are given for each type. Following the list of types is a list of the Journal Article 1474, New Mexico Agricultural Experiment Station, Las Cruces. "Department of Animal and Range Sciences, Box 3-1, New Mexico State University, Las Cruces, New Mexico 88003. 73 74 K. W. Allred [Volume 50 Table 1. Authors and collt'ctors of New Mexico grass types. Table 2. Counties of collection of New Mexico grass types. Name Authored CJollected Countv Niunber of types Thomas Antisell 0 William J. Beal 1 John M . Bigelow 0 W. S. Boyle 1 Samuel B. Buckley 2 Christopher S. Campbell 1 Karel Domin 2 William H. Emory 0 August Fendler 0 William Gambel 0 Eduard Hackel 1 Alberts. Hitchcock 2 Edwin James 0 Ivan M. Johnston 1 Marcus E. Jones 1 O. C. Louis-Marie 1 Edgar A. Mearns 0 Elmer D. Merrill 1 Orrick B. Metcalfe 0 George Nash 3 Thomas Nuttall 2 Cyrus G. Pringle 0 H. H. Rusby 0 Frank L. Scribner 4 Cornelius L. Shear 1 Jared G. Smith 1 Paul C. Standley 1 Ernst G. Steudel 5 Jason Swallen 3 George Thurber 1 John Torrey 6 Samuel M. Tracy 0 George Vasey 14 Wilkins 0 S. W. Woodhouse 0 Elmer O. Wooton 1 Charles Wright 0 gras.s genera of New Me.xico cross-referenced to a selected bibliography. The bibliography is not intended to be exhaustive; rather, only significant revisionary or summary papers pertinent to New Mexico grass taxonomy are listed. Further references may be obtained by consulting the works listed here, particularly Gould and Shaw (1983) and Soderstrom et al. (1987). New Mexico Grass Types The acronym in parentheses (Holmgren et al. 1981) refers to the locality of tyjic material: holotype, isotype, fragment, or other dupli- cate material. Agrostis minufis.sjm« Steudel, S\n. I'l. (duni. 1: 171. 1854. Fendler 986, in 1847, Santa Fe Co., near Santa Fc (US). = Muhlcnher^ia mintttissiiiia (Steud.) Swallen Bernalillo Colfax Doiia Ana Eddy Grant Hidalgo Lincoln Otero Rio Arriba San Miguel Santa Fe Socorro 1 1 4 3 12 1 1 2 2 3 11 1 Andropoiion ' Mount. 408. 1885. Thurlier 269, probably Grant Go., "Rio Mimbres. = Stipa neomexicana (Thurb.)Seribn. Stipa scribneri Vasey, Bull. Torr. Bot. Glub 11: 125. 1884. Vasey s.n., in 1881?, Santa Ee Go., Santa Ee (US). Tricuspis mutica Torrey, U.S. Expl. Miss. Pacif Rpt. 4: 156. 1856. Bigelow s.n.', 22 Sep 1853, San Miguel Go., Laguna, Colorado (NY). = Tridens muticus (Torr.) Nash Trisetum montanum Vasey, Bull. Torr. Bot. Glub 13: 118. 1886. G. R. Vasevs.n., in 1881, San Miguel Co., Las Vegas (US). Trisetum montanum Vasev var. pilosum Louis-Marie, Rhodora 30: 212. 1928. P. C. Standley 4536, Aug 1908, San Miguel Co., Gowles. = Trisetum montanumVasey Uralepis composita Bucklev, Proc. Acad. Nat. Sci. Phila. 1862: 94. 1862. Woodh'ouse s.n., probably 1851, "New Mexico" (PH). - Leptochloa fascicularis (Lam.) Gray 76 K. W Allred [Volume 50 Uralepis poaeoides Bucklev, Pioc. Acad. Nat. Sci. Phila. 1862: 94. 1862. Fendler 932, in 1847, probably Santa Fe Co. near Santa Fe (PH). = Poa fendleriana (Steud.) Vasey Vilf a tricholepis Torrey, U.S. E.xpl. Miss. Pacif. Rpt. 4: 155. 1857. Bigelows.n., Oct 1853, Bernalillo Co., Sandia Mountains (NY). = Blcpharonciiron tricholepis (Torr.) Nash Selected Bibliography of New Mexico Grass Taxonomy Generalliterature 3, 20, 46, 48, 58, 59, 60, 102, 106, 116, 118, 119, 150, 189, 199, 204, 205, 227, 229, 236, 237, 238, 239 Triticeae tribe 18, 19, 23, 38, 39, 41, 77, 78, 79, 151 Agropyron 51, 65, 75, 76, 78, 80, 151, 167, 185, 216 Agro.sfi.s- 32, 165,214,233 Alopcctirus 194 Androi)ogon 47, 49 Anthoxanthiim 107, 242 Apera 118 Aristida 2, 102 Arrhcnatheruiu 118 A r undo 118 Avena 21, 22 Bechnannia 90, 177 Blepharoncuron 118 Boiitcloiui l{)3. 115, 180 Brachiaria 229, 240 BromusM, 84, 113, 146, 157, 166, 191, 195, 198, 207, 208, 224 Biichloe 171 Calamagrostis 108, 131, 206 Calamovilfd 222 Catahrosa 118 Cenchrus 67 Chloris 6, 138 Cinna 118 Coi.rll8 Coriadcria 63 Cottea 118 Cynodon 74 Dactylis 209, 243 Dactyloctenium 102, 118 Danfhonia 56, 94 Deschampsitt 130, 135 Dichanthelium 105 Digitaria 228 Distichlis 31 Echinochloa 104 Eleusine 102, 118 xElyhordeum 18, 19, 110 xEhjmotrigia 18, 19 Elymus 75, 76, 78, 98, 100, 151, 158, 196, 234 £/|/on»ru.v 102, 118 Elytrigia 18, 19, 51, 65, 99, 151, 216 Enneapogon 53, 118 Eragro.stis 59, 97, 134, 136, 164, 179, 235 Eremopyrum 18, 19, 78 Eriocldoa 190 Erioneuron 186 Feshica%, 95, 96, 124, 148, 160, 161, 162, 163, 217, 219 Glyccria 54 Hackelochloa 118 Helictotrichon 118 Heteropogon 102, 118 Hierochlo'e 230 Hilaria 200, 232 HoIcusUH Hordeum 24, 25, 26, 27, 28, 29. 30, 35, 36, 38, 64, 152, 153 Imperata 118 Koeleria 109, 192, 193 Leersia 172 Leptochloa 147 LcymusW, 17,78, 151 Lolium 83, 124, 218, 220 Lycuriis 175 Melica 40 Muhlenhcrgia 92, 155, 156, 169, 170, 174, 197 Miinroa 8 Oryzopsis 15, 125 Paniciim 37, 44, 52, 117, 142, 143, 144, 145, 225, 240 Pappophorum 53, 182 Paspalum 1, 12, HI, 159 Pennisetiim 52, 221 Phalaris5 Phleiim 121 Phragmites 57 , 91 Phyllostachys {">() Piptocliactium 15, 118 Poa 10, 33, 112, 132, 133. 141, 149, 201, 202 Polypogon 118 PsatJtyrostachys 78, 81 Puccincllia 54. 61. 66. 93 Redfiddia 50, 178 Rhynchclytrum 118 SaccJuiniin 60 Schedonnardus 118 Schismus 62, 89 Schizachnc 137, 212 Schiz(ichyritn)i 102, 118 Sclerocldod 213 Scleropogon 181 Secfl/f'38, 211 S('fr;ri« 85, 86, 88, 129, 168, 176, 184, 221 Sorghastnim 223 Sorghum 69, 72, 82, 226 Spat-tina 154 Spheuopholis 87 Sporobolu.s 128, 183 Stenotaphrum 45, 187, 241 Sfi/Jrt 13, 14, 15, 16, 125, 188 Torreyochloa 54, 55, 61 Trachypogon 102, 118 Tragus 7 Tridcns 215 Triplasis 102, 118 Tripsannii 71 Trisctum 120. 135. 140, 173 xTriticosccalc 42, 203 rriticum 38, 43, 126, 127 C;Y)r/i/(W 229 \'i///)iV/ 139 Z<'rt 122, 123 Zoiy.virt 60, 118 1990] Ne\n Mexkx) Grass Taxonomy 77 6. 16 17 Litkhati'rkCitki) Ai.i.iU'.n. K W 19.S2. Faspdlum (li.\tirltiiiii L. \ar. (»yron ]. Dewey, D R, wdC Hsiao 1983, A cytogenetic basis loi- transferring Russian wildr\e from Ehpnus U)Psalhyrostachys. Crop Sci, 23; 123-126, DocciriT 11 1970. Sorglium, Longmans, London. 403 pp 19901 New Mexico (iHAss Taxonomy 79 83. Dork. W. C 1950. Persian danicl in Canada. Sii. Agric. 30: 157-164 {Loliimi |. 84. Elliot, F. C. 1949. Brointif iiicrmi.s ami li. pnmpcl- /iflMMs in North America. Evolution 3; 142-149. 85. Emery, W. H. P. 1957a. A cytotaxonomic .stncly of Setaria iiuicrustaclujd (C.raniint'af) and it.s rela- tives in the .southwe.stern United State.s and Mex- ico. Bull. Torrey Bot. Cluh 84: 95-l(ri. 86. 19571). A stud\' of reproduction in Setaria nuicrostachya and its relatives in southvvestein United States and Mexico. Bull. Torres' Bot. Club 84: 106-121. 87. Ekdm.an, K S 1965. Taxononn of the .sjenus Sphcuopliolis ((iraniineae). Iowa State ). Sci. 39: 289-336. 88. Fairbrothers, D. E. 1959. Morphological variation of Setaria faheri and S. viridis. Brittonia 11: 44-48. 89. FARUgL S A., and H B ^l haish 1979. Studies on Libyan grasses. V. Popidation \ariabilit\ and dis- tribution of Schisjutis arahicus and S. harbatus in Libya. Pak. J. Bot. 11(2): 167-172. 90. Fernald. M 1 1928. The American and eastern Asiatic Beckmamiia. Rhodora30: 24-27. 91. 1932. Phrap,mites communis Trin. var. herlandieri (Fournier) comb. nov. Rhodora 34: 211-212. 1943. Five common rhizomatous species of 92. 93. 94. 95. 96. 102. 103. 104. 105. Muhlenberp.ia. Rhodora 45: 222-239. Fernald. M I , and C A. We.atherby 1916. The genus Puccinellia in eastern North America. Rhodora 18: 1-23. FlNDCAY. J N . AND B R Bai'M 1974. The nomen- clatural implications of the taxonomy oiDanthonia in Canada. Canad. J. Bot. 52: 1573-1581. Frederiksen. S 1979. Festuca minutijlora Rydb., a neglected species. Bot. Notiser 132: 315-318. 1982. Festuca brachypliyUa. F. saximon- tana and related species in North America. Nord. J. Bot. 2: 525-536. 97. Freeman. D. 1979. Lehmann lovegrass. Range- lands 1: 162-163 [Erag,rostis]. 98. Gable, M. L. 1984. a biosystematic study of the genus Elymus (Gramineae: Triticeae) in Iowa. Proc. Iowa Acad. Sci. 9: 140-146. 99. GlLLETT, J. M. andH a. Senn 1960. Cytotaxonomy and infraspecific variation of Agropyron sinithii Rydb. Canad. J. Bot. 38: 747-760. 100. Gould, F. W. 1947. Nomenclatural changes in Elijmus with a key to the California species. Madrono 9: 120-128. 101. 1957. New North American andropogons of subgenus AmpJiiloplius and a key to those spe- cies occurring in the United States. Madrono 14: 18~29 [Bothriuchloa]. 1975. Grasses of Texas. Texas A&M Uni- versity Press, College Station. 653 pp. 1979. The genus Bouteloua. Ann. Missoiu Bot. Card. 66: 348-416. Gould, F. W., M. A All and D E Fairbrothers 1972. A revision of Echinochloa in the United States. Amer. Midi. Naturalist 87: .36-59. Gould, F. W., and C. A Clark. 1978. Dichanthe- liuin (Poaceae) of the United States and Canada. Ann. Missouri Bot. Card. 65: 1088-11.32. 106. (ioi LD F W ANDli B Sii\u 1983. Grass system- atics. 2dcd. Texas AiyM Uiiiversit\' Press, (College Station. 397 pp. 107. GuANi-, M (; , andJ Antonovicks 1978. Biology of ecologically marginal populations of Aiillioxatt- thum odoratum. I. Phenetics and d\namics. Evo- lution .32: 822-8,38. 108. (;ki;i:ne. C. W. 1984. Sexual and apomictic repro- duction in Calamuf^rostis (Gramineae) from east- ern North America. Amer. J. Bot. 71: 285-293. 109. Ckkuteh. W 1968. Notulaenomenclaturaleset bit)- liographicae 1-4. Candollea 23: 81-108 [Koele- ria \ . 110. Gross. A. T. H 1960. Distribution and ecology of Elymus macouniiVasey . (>an. J. Bot. 38: 63-67. 1 1 1. Guzman, R.. and F J. Santana M 1987. Las espe- cies mexicanas del genero Paspalum L. (Grami- neae). Univ. Guadalajara P'olleto Tecnico No. 1. 112. Halperin, M 1933. The taxonomy and morphol- ogv of bulbous bluegrass, Pua bulbosa vivi])ara. ]. Amer. Soc. Agron. 25: 408-413. 113. Harlan, J R 1945. Cleistogamy and chasmogamy in Bromus carimitus Hook. & Arn. Amer. J. Bot. 32: 66-72. 114. Harvey, L 1975. Eragrostis. Pp. 177-201 in F. W. Gould, Grasses of Texas. Texas A&M University Press, College Station. 653 pp. 115. Hill, S R. 1982. Vegetative apomixis ("Vivipary") in Boutcloua hirsufa Lag. (Poaceae). Sida 9: 355-357. 1 16. Hitchcock:, A. S. 1920. The genera of grasses of the United States, with special reference to the eco- nomic species. USDA Bull. 772. 117. HiTCHc:ocK. A S., and M A Chase 1910. The North American species o{ Panicum. Contr. U.S. Natl. Herb. 15: 1-396. 118. 1951. Manual of the grasses of the United States. 2ded. U.S.D.A. Misc. Publ. 200. 1,051pp. 119. Holmgren, P. K., W Keuken, and E. K, Scho- FIELD. 1981. Index Herbariorum, part I. The herbaria of the world. Bohn, Scheltema & Hol- kema, Utrecht. 120. Hulten. E 1959. The Trisetum spicatum complex. Sv. Bot. Tidskr. 53: 203-228. 121. HiMPHRlES, C J 1978. Notes on the genus Phlcum L. Bot. J. Linn. Soc. 76: 337-339. 122. Iltis, H H 1983. From teosinte to maize: the catastrophic sexual transmutation theorv. Science 222: 886-894 [Ztv;]. 123. Iltis, H H , and J F Doebley 1980. Taxonomy of Zea (Gramineae). II. Subspecific categories in the Zea mays complex and a generic synopsis. Amer. J. Bot. 67: 994-1004. 124. Jaukar, P P 1975. Chromosome relationships be- tween Lolium and Festuca (Gramineae). Chromo- soma 52: 103-121. 125. Johnson. B L. 1945. Natural hybrids between Orij- zapsis hymenoides and several species of Stipa. Amer. J." Bot. 32:599-608. 126. Johnson, B L., D. Barnhart, and O. Hall 1967. Analysis of genome and species relationships in the polyploid wheats by protein electrophoresis. Amer. J. Bot. 54: 1089-1098 [Tn7ic(/»i]. 127. Johnson, B L , and H S Dhaliwal 1976. Repro- ductive isolation ofTriticum boeoficum and Triti- cum urartu and the origin of the tetraploid wheats. Amer. J. Bot. 63: 1088-1094. 80 K. W. Allred [Volume 50 128. 129, 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. JONE.S, E. K , ,\ND N C. Fassett 1950. Suljspecific variation in Sporoholus cniptandrus. Rhodora 52: 125-126. Karlsson, T. 1987. Setaria adliaerans and S. fabcri new to Sweden. Svensk Bot. Tidskr. 81: 305-311. Kawano, W. 1963. Cytogeography and evolution of the Deschampsia caespitosa complex. Can. J. Bot. 41:719-742. 1965. Calamagro.^tis purpurascens R.Br. and its identitv. Acta Phvtotax. Geobot. 21: 73-89. Kellogg, E 1985a. A biosystematic study of the Poa secunda complex. J. Arnold Arbor. 66: 201-242. 1985b. Variation and names in the Poa secunda complex. J. Range Management 38: 516-521. Koch, S D 1974. The Eragrostis pcctinacea-pilosa complex in North and Central America (Grami- neae: Eragrostoideae). Illinois Biol. Monogr. 48: 1-74. 1979. The relationships of three Mexican Aveneae, and some new characters for distin- guishing Deschampsia and Trisetum. Taxon 28: 225-235. Koch, S. D,, and I. Sanchez \' 1985. Eragrostis mexicana, E. neomexicana, E. orciittiana and E. virescens: the resolution of a taxonomic problem. Phytologia 58: 377-381. Koyama. T , AND S Kawano. 1964. Critical taxa of grasses with North American and eastern Asiatic distribution. Can. J. Bot. 42: 859-884 [Schiz- achne]. L.VZARlDES, M 1972. A revision of Australian Chlo- rideae(Gramineae). Australian). Bot. (siippl. ser.) Snppl. No. 5[C/i/om]. LONARD, R I., AND F W. Goi'LD 1974. The North American species of Vulpia (Gramineae). Madrono 22: 217-230. Louis-Marie, O. C 1928. The genus Trisetum in America. Rhodora 30: 209-228, 231-245. Lush, W M 1989. Adaptation and differentiation of golf course populations of annual bluegrass (Poa annua). Weed Sci. 37: 54- 59. McGregor. R L 1984a. Panicum capillare L. var. occidentale Rydb. (Poaceae): an illegitimate name. Contr. Univ. Kansas Herb. No. 10. 1984b. Panicum capillare L. var. harbipul- vinatum (Nash) comb. nov. Phytologia 55(4): 256. 1984c. Panicum dichotomiflorum (Poaceae) variation in Kansas including notes on the sterile palea. Contr. Univ. Kansas Herb. No. 8. 3 p. 1985. Panicum hillmanii Chase (Poaceae): vailiditv and distribution. Contr. Univ. Kansas Herb. No. 13. 9 pp. McNeill, J. 1976. Nomenclature of four perennial species of Broinus in eastern North American, with a proposal for the listing of 6. purgans L. as a rejected name imder Article 69. Taxon 25: 611-616. 1979. Diplachne dx^d Leptochloa (PoAcede) in North America. Brittonia 31: 399-404. Malik, C P 1967. H\bridization of Fe.y^i/r« species. Can. J. Bot. 45: 1025-1029. Marsh, V. L. 1952. A taxonomic revision oi liic genus Poa of United States and southiin Canada. Amer. Midi. Naturalist 47: 202-250. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. Martin, W C . and C R Hutchins 1980. A flora of New Mexico. Vol. 1. Vaduz, West Germany, J. Cramer. 1,276 pp. Melderis. a 1978. Ta.xonomic notes on the tribe Triticeae (Gramineae), with special reference to the genera Ehpnus L. seiisu lata, and Agropyron Gaertner sensu lato. Bot. J. Linn. Soc. 76: 369-384. Mitchell, W W 1967. On the Hordeum juba- tum-H. braclu/antherum question. Madrono 19: 108-110. Mitchell, W W , and A. C. Wilson 1964. The Hordeum jubatum - caespitosimi - brachijantherum complex in Alaska. Madrono 17: 269-280. Mobberley, D G 1956. Taxonomy and distribution of the genus Spartina. Iowa State Coll. J. Sci. 30: 471-574. Morden, C W , and S L H.\tch 1986. Vegetative apomixis in Muhlenbergia repens (Poaceae: Era- grostideae). Sida 11: 282-285. 1987. Anatomical study of the Muhlenber- gia repens complex (Poaceae: Chloridoideae: Eragrostideae). Sida 12: 347-359. Morrow, L A . and P W Stahlman 1984. The history and distribution of downy brome {Bromus tectorum) in North America. Weed Sci. 32 (suppl.): 2-6. Nelson, E. N . and R. J Tyrl 1978. Hybridization and introgression between Ehjmus canadensis and Ehpnus virginicus (Poaceae). Proc. Okla. Acad. Sci. 58; 32-34. Nomencl.ature Committee. IAPT 1983. Report on Proposal 528: rejection oiPaspalum distichum L. (Gramineae). Taxon 32; 281 [name stands]. Pa\ LICK, L E 1983a. Notes on the taxonomy and nomenclature of Festuca occidentalis and F. ida- hocnsis. Can. J. Bot. 61; 337-344. 1983b. The taxonomy and distribution of Festuca idahoensis in British Columbia and north- western Washington. Can. J. Bot. 61; 345-353. 1984. Studies on the Festuca ovina com- plex in the Canadian Cordillera. Can. J. Bot. 62: 2448-2462. 1985. A new taxonomic smvev of the Festuca rul^ra complex in northwestern North America, with emphasis on British Columbia. Phytologia 57: 1-17. Perry. G . and J. McNeill 1986. The nomencla- ture of Eragrostis cilianensis (Poaceae) and the contribution of Bellardi to .\llioni's Flora Pede- montana. Taxon 35: 696-701. PiilLiPSON, W R 1937. A revision of the British species of the genus Af^ro.sfi.v. (. Linn. Soc. Bot. 51:73-151. PiNTo-EscoBAR, P 1976. Notasobreelejamplartipo de "Bromus catharticus" VaW. Caldasia 11: 9-16. PoilL. R. W. 1962a. Agropyron hybrids and the status of Agropip'on pscudorepens. Rhodora 64:143-147. 1962b. Notes on Setaria liridis and S. faberi (Gramineae). Brittonia 14: 210-213. 1969. Muhlenbergia subgenus Muhlenber- gia (Gramineae) in North America. .\nu'r. Midi. Naturalist 82: 512-542. PoiiL R W . andW W MnciiELL 1965. Cytogeog- rapln of the rhizomatous American species of Muhlenbergia. Brittonia 17: 107-112. 1990] New MkxicoGhass Taxonomy 81 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. PoZ.\KN,SKV, T 198.1. liiitlalouiass: home on llic range, but also a turf i^ras.s. Hanuelaiul.s .5; 214— 216 [Buchloc]. Pyrah. G. L. 1969. Ta.xoiioinic and cli.strihiitional studies in Lt't'r.vfV/ ((waniincac). Iowa St. J. Sei. 44: 215-270. Randall. J L AND K W lin.i 1986. Hios\stiinatic studies of North American Trisetiiin siiiciiliim (Poaceae). Syst. Bot. 11:567-578. Reeder. C G 1949. Mulilciihcr' of Natural History 7: 375-428. Hall, E R 1946. Mammals of Nevada. University of California Press, Berkeley. 710 pp, 1951. American weasels. University of Kansas Publications, Museum of Natural History 4: 1-466. 1981. The mammals of North America. John Wiley & Sons, New York. 2 vols. 1,181 pp. Jorgensen, C D , andC L Hayward 1965. Mammals of the Nevada Test Site. Brigham Young Uni\ ersity Science Bulletin 6: 1—81. OFarrell, T P., and L. a Emery. 1976, Ecology of the Nevada Test Site: a narrative summary and anno- tated bibliography, U,S, Energy Research Devel- opment .'\dministration Report. NVO-167. 249 pp. RoMNEY, E M , V Q Hale, A Wallace, O R Lunt, J. D. Childress, H. Kaaz, G V Ale.xander, J. E. Kinnear, and T. L. Ackerman 1973. Some characteristics of soil and perennial vegetation in northern Mojave desert areas of the Nevada Test Site. University of California, Los.\ngeles. Report 12-916. .340 pp'. Received 10 Aug,ust 1989 Accepted 2 November 1989 Cleat Basin Naturalist 50(1), 199(1, pp. 85-87 FORMATION OF PISOLITHUS TINCTORIUS ECTOMYCORRHIZAE ON CALIFORNIA WHITE FIR IN AN EASTERN SIERRA NEVADA MINESOIL R, F, Walker' The Gasteroinycete Pisolitlin.s tiiictorius (Pers.) Coker & Couch occurs in temperate, subtropical, and tropical zones worldwide and in ectomycorrhizal association with niunerous conifer and hardwood hosts (Marx 1977). Fre- quent reports of the occurrence of its basidio- carps near various pine species on harsh sites in the eastern United States (Lampky and Peterson 1963, Schramm 1966, Hileand Hen- nen 1969, Lampky and Lampky 1973, Marx 1975, Medve et al. 1977) have prompted extensive efforts to inoculate seedlings in forest nurseries with this mvcobiont (Marx etal. 1976, 1982, 1984, 1989a,' 1989b). Subse- quently, the improved survival and growth of pine seedlings with P. tinctoriiis ectomycor- rhizae after outplanting on surface mines has been attributed to enhanced uptake of nutri- ents (Marx and Artman 1979) and water (Walker et al. 1989). Current research is focused on development of more effective in- ocula and inoculation procedures, discovery of locally adapted P. tinctoriiis isolates, and identification of new host species. A recent report (Walker 1989) disclosed P. tinctoriiis occurring in ectomycorrhizal association with Jeffrey pine {Piniis jeffreyi Grev. & Balf) and Sierra lodgepole pine {Pinus contorta var. miirraijana [Grev. & Balf] Engelm.) on spoils of the Leviathan Mine in Alpine County, California. Located on the eastern slope of the Sierra Nevada (38°42'30"N, I19°39'15"W) at an elevation of 2,200 m and consisting of approximately 100 ha, this open-pit suhur mine has been inactive since 1962. The average annual precipitation of approximately 50 cm is primarily snowfall, and the minesoil has a pH of 4.0 to 4.5, a deficiency of plant-available N, and a poten- tially phytotoxic concentration of Al (Butter field and Tueller 1980). Vegetation is sparse on most of the spoils, but in addition to the two pine species mentioned previously, Cali- fornia white fir (Abies concolor var. lowiana [Gord.] Lemm.), singleleaf pinyon [Pinus monophi/Ua Torr. &Frem.), Utah juniper (/?/- niperiis osteospenna [Torr. ] Little), and quak- ing aspen {Popiilus trcmuloides Michx.) have become reestablished on the periphery of the mine near adjoining undisturbed forest and woodland. Walker's (1989) report concerning examinations made in September 1988 of the probable hosts of P. tinctorius in Leviathan Mine noted that basidiocarps of this symbiont were absent in the immediate vicinity of the latter four tree species. Reexamination of Leviathan Mine spoils in August and September 1989, however, re- vealed numerous P. tinctorius basidiocarps near seedlings and saplings of California white fir. Typically, one or two dark yellow to brown basidiocarps (Fig. lA), matching the descrip- tion of Coker and Couch (1928), were ob- served around solitary white fir seedlings, while as many as five encircled individual white fir saplings. Stipitate, substipitate, and sessile forms were observed, varying in size from 9 to 17 cm in length and from 3 to 7 cm in diameter. Approximately 100 basidiocarps were found associated with white fir, and these were rarely more than 2 m from the host. Strands of mycelia with gold-yellow pig- mentation, which compare favorably with the P. tinctorius rhizomorphs described by Schramm (1966), were traced through the spoils from basidiocarps to the root systems of white fir seedlings and saplings. These mycelia were connected to ectomycorrhizae of similar pigmentation that matched the 'University ol'Nevada, Reno. Department of Hange, Wildlife and Forestr.'. 1000 Valley Road, Reno. Nevada 89512. 85 86 Notes [Volume 50 Fig. 1. Pisolithus tinctorius associated with California white fir on an eastern Sierra Nevada mine spoil: A, hasidiocarps (bar represents 5 cm); B, ectomycorrhizae on roots (bar represents 1 cm). description of those formed by P. tinctorius reported Iw Marx and Bryan (1975). Examina- tion of the complete root system of an isolated white fir seedling with a single associated ba- sidiocarp revealed numerous P. tinctorius ec- tomycorrhizae with approximately 35% of the roots exhibiting the bifurcate or coralloid form (Fig. IB) or the fungal mantle characteristic of these mycorrhizae. An earlier attempt to inoculate white fir seedlings at outplanting with P. tinctorius ba- sidiospores was largely unsuccessful (Alvarez and Trappe 1983). The evidence presented here, however, indicates that this host and 19901 NOTKS 87 synibiont association occurs naturally in the Sierra Nevada. Efforts to monitor the in\ tor- rhizal dexelopincnt of Sierra Ncxada and In- terniountain woocK flora will continue in or- der to further ascertain the host ranges of this and other ectomycorrhizal funi^i. Acknowledgments This paper contains results of the Nevada Agricultural Experiment Station Research Project 612 funded by the Mclntire-Steunis Cooperative Forestry Research Program. The author is indebted to P. M. Murphy of the Division of Forestry, Nevada Department of Conservation and Natural Resources, and to D. C. Prusso of the Department of Biology, University of Nevada, Reno, for their invalu- able assistance. LlTERATlfRE CiTED Alvarez, I. F , and J. M Trappe 1983. Du.sting roots of Abies concolor and other conifer.s with Pisolithus tinctorius spores at outplantinsi; time proves inef- fective. Canadian Jomiial of Forest Research 13; 1021-1023. BUTTERFIELD. R I., AND P T. TuELLER. 1980. Revegeta- tion potential of acid mine wastes in northeastern California. Reclamation Review 3: 21-31. CoKER. W. C, AND J N Couch. 1928. The Gastero- mycetes of the eastern United States and Canada. University of North Carolina Press, Chapel Hill. 201 pp. HiLE, N . AND J F. Hennen. 1969. In vitro culture of Pisolithus tinctorius mycelium. Mvcologia 61: 195-198. Lampky.J R , andJ H Peterson. 1963. Pisolithus tincto- rius associated with pines in Missouri. Mvcologia 55: 675-678. La.mpky, S. a., and J R Lampkv 1973. Pisolithus in cen- tral Florida. Mvcologia 65: 1210-1212. Marx. D. H. 1975. Mycorrhizae and establishment of trees on strip-mined land. Ohio Journal of Science 75: 288-297. 1977. Tree host range and world distribution of the ectomycorrhizae fimgus Pisolithus tinctorius. Canadian Journal of Microbiology 23: 217-223. Marx. D H.. and J. D. Artman. 1979. Pisolithus tincto- rius ectomycorrhizae improve survival and growth oi pine seedlings on acid coal spoils in Kentucky and \irginia. Reclamation Review 2: 23-31. M\n\ D II and VV C Bryan 1975. Growth and ecto- nu'corrhizal development ol loblolly pine seed- lings in lumigated soil inli'sted with the fungal sxinbiont Pisolithus tinctorius. Forest Science 21: 245-254. Marx. D H., W. C Bryan, and C E Cordele. 1976. Growth and ectomycorrhizal development of pine seedlings in nvusery soils infested with the fungal symbiont Pisolithus tinctorius. Forest Science 22: 91-100. Mahx, D H , C E CoRDELL, D S Kenney. J G Me.xal. J. D. Artman, J. W Riefee, and R. J Moeina 1984. Commercial vegetative inoculum of Piso- lithus tinctorius and inoculation technicjues for development of ectomycorrhizae on bare-root tree seedlings. Forest Science Monograph 25. 101 pp. Marx. D H , C E Cordeel, S B Maul, and J L Ri'EiiEE 1989a. Ectomycorrhizal development on pine by Pisolithus tinctorius in bare-root and con- tainer seedling niuseries. I. Efficacy of various vegetative inoculum formulations. New Forests 3: 45-56. 1989b. Ectomycorrhizal development on pine by Pisolithus tinctorius in bare-root and container seedling nurseries. II. Efficacy of various vegeta- tive and spore inocula. New Forests 3: 57-66. Marx, D H , J L Ruehle, D S Kenney. C, E Cordell, J W Riffle, R J Molina, W. H. Pawuk. S. Navratil. R W. Tinus, and O. C. Goodwin 1982. Commercial vegetative inoculum of Piso- lithus tinctorius and inoculation techniques for development of ectomycorrhizae on container- grown tree seedlings. Forest Science 28: 373-400. Medve, R J , F M Hoffman, andT W Gaither 1977. The effects of mycorrhizal-forming amendments on the revegetation of bituminous stripmine spoils. Bulletin of the Torrev Botanical Club 104: 218-225. Schramm, J R 1966. Plant colonization studies on black wastes from anthracite mining in Pennsylvania. Transactions of the American Philosophical Soci- ety 56: 1-194. Walker. R F 1989. Pi.solithus tinctorius, a Gastero- mycete, associated with Jeffrey and Sierra lodge- pole pine on acid mine spoils in the Sierra Nevada. Great Basin Naturalist 49: 111-112. Walker, R F , D C West, S B McLaughlin, and C C. Amundsen 1989. Growth, xylem pressure potential, and nutrient absorption of loblolly pine on a reclaimed surface mine as affected by an induced Pisolithus tinctorius infection. Forest Science 35: .569-581. Received 14 October 1989 Accepted 15 November 1989 Crcat liasMi N.ihiralisI 50( 1 1, IWK). p. W) BONECIIEWINC BY liOCKY MOUNTAIN Bl(;ilORN SHEEP K. A. Kcaliiii!; Bone clu'wini:; has not, to my know Icduc been reported in wild Noitli Anicricaii l)o\ids. Herein I deserihe an instanei' ol hone chewing hy a Koeky Monnlain bighorn sheep (Ovis canadensis). An adult female eonsnmed two bones on Mt. Everts winter range in Yellowstone National Park during winter 19(S()-(S1. The bones apparently were hom a small ungidate, probably l)ighoru sheep, uuile deer, or pronghorn antelope. The first bone was eonsnmed in 5-10 minutes; 1 inter- rupted consumption ot the second. Sekidic and Estes (1977) report that sable anteloj)e frequently spend < 1 hr chewing a bone and described a case of a yearling male chewing on the same bone for 5.5 hr. The rapid consump- tion observed here was likely related to the weathered, brittle condition of the bones. Sekulic and Estes (1977) also report at least two instances in which l)one possession by sable antelojie led to aggressi\(" displacement of younger animals by adult females. During my observations, several other bighoin sheep continued to feed nearby but showed no inter- est in the bone. Bone chewing was observed during the second of two unusually mild win- ters in which forage was generally free of snow and range use by elk was minimal. In- dices of population (juality are believed to reflect the nutritional status of a population (Geist 1971). In this study, high young:adult lemale ratios, long suckling times, male matu- ration rates, and low concentrations of lung- worm {Protoslron^iilnssp]).) larvae in bighorn leces were all indicative of a high-(iuality, expanding population (Keating 1982). This suggests that bone chewing was not a result of general nutritional deficiency in the popula- tion, though deliciencies in individual animals or in specilic dietary components cannot be discounted. A(:KN()VVLi:i)(;MK,NTS This research was supported by the l^ob and Bessie Welder Wildlife Foundation, the National Kille Association, the Montana De- partment of Fish, Wildlife and Parks, and Yellowstone National l\uk. LllKliAllHi; GlIKI) C.Kisi. V 1971. Mountain .sheep; a .study in behavior and evolution. Uuivetsitv olCihieajjo j-'ress, Chiea^o. :3.S3 pp. KlvMiNC, K A 1982. PopulaUoH ixolot^y ollioeky Moiui- taiu l)ijj;lioni sheei) in the upper Yellowstone iiiver drainajiie, Montauii/Wyoming. Unpublished the- sis, Montana State University, Bo/enian. 79 pp. SKkULlc. IV, andH I) ICsiKS 1977. A noteon boneehew- ing in the sable antelope in Ken\ a. Manuiialia 4 1 : 537-5,39. Received 1 Se))tenih('r I9HH Accepted 17 October 19HfJ Glacier Natiuiial Park, Science Ceiitui , \\ csl (,1a Muiitaiia .599.36. 89 RKOUCED PRICES O N BACK ISSUES Great Basin Naturalist and Great Basin Naturalist Memoirs The prices listed below are in effect through December 1990. GREAT BASIN NATURALIST: regularly $12 per back issue The Great Basin Naturalist is published quarterly. Volume 1 (1939) through Volume 34 (1974): $ 4 per issue or $10 per volume Volume 35 (1975) through Vokune 48 (1988): $ 5 per issue or $15 per volume Complete sets: A few complete sets are available. For information contact James R. Barnes, Editor, (801) 378-5053. GREAT BASIN NATURALIST MEMOIRS: selected Memoirs 50% off the prices listed below. No. 1 The birds of Utah. $10 No. 2 Intermountain biogeography: a symposium. $15 No. 3 The endangered species: a symposium. $6 No. 4 Soil-plant-animal relationships bearing on revegetation and land reclama- tion in Nevada deserts. $6 No. 5 Utah Lake monograph. $8 No. 6 The bark and ambrosia beetles of North and Central America (Coleoptera: Scolytidae), a taxonomic monograph. $60 No. 7 Biology of desert rodents. $8 No. 8 The black-footed ferret. $10 No. 12 Research in the Auchenorrhyncha, Homoptera: a tribute to Paul W. Oman. $30 Send orders for back issues to: Editorial Office Attn: Carolyn Backman Great Basin Naturalist 290MLBM Brigham Young University Provo, UT 84602 USA Please include $2 handling and postage for each Great Basin Naturalist \olume or Meinoir. Great Hasiii Nati.ralisI 5()i 1 1, UJ^X), pp. 01-92 DISTRIBUTION OF LIMBER PINE DWARF MISTLETOE IN NEVADA |{()l)('rt L. Miithiast'H aiul l*'iank(;. I lawksworth" Liinlier pine dwarf mistletoe, Arceufho- hium cyanocarpum (A. Nels. ex Rydh.) A. Nels. (Viscaceae), parasitizes several species of white pine (subgenus Strobus) in the west- ern United States from Montana and Colo- rado to Oregon and California (Hawksworth and Wiens 1972, Mathiasen and Hawksworth 1988). The principal hosts of this mistletoe are limber pine {Piniis flexilis James), whitebark pine (P. alhicuulis Engelm.), Great Basin bris- tlecone pine (F. lon^aeva D. K. Bailey), and Rocky Mountain bristlecone pine (P. aristata Engelm.). It also occurs on western white pine (P. monticola Dougl.) and foxtail pine (P. baljouriana Grev. & Balf) in northern California (Hawksworth and Wiens 1972, Mathiasen and Hawksworth 1988). The lim- ber pine dwarf mistletoe causes local, severe mortality in several areas (Hawksworth and Wiens 1972, Mathiasen and Hawksworth 1988). The distribution of limber pine dwarf mistletoe in Nevada is poorly documented (Hawksworth and Wiens 1972, Hawksworth 1988, Kartez 1988). First collected in Nevada in 1881 on Great Basin bristlecone pine in the Spring Creek (Charleston) Mountains in Clark County, it was not collected elsewhere in Nevada until 1958 when it was found in the Ruby Mountains near Elko (Elko County) (Hawksworth and Wiens 1972). It has since been reported from the Sheep Mountains (Clark County), Copper and East Humboldt mountains (Elko County) (Hawksworth and Wiens 1984), and the South Snake Mountains (White Pine County) (collection by D. K. Bai- ley). Hawksworth (1988) suggests that the mistletoe probably occurred in the Toiyabe Mountains of Nye County because Linsdale et al. (1952) illustrated a limber pine that ap- peared to be infected by this mistletoe. This note confirms Hawksworth's suspicion that limber pine dwarf mistletoe occurs in the Toiyabe Mountains. In 1989 the senior author found Arceiitho- hium cyanocarpum at several new localities in Nevada (Fig. 1): on limber pine in the North Snake Moimtains (White Pine County) near Mount Moriah, in the Bull Run Mountains (Elko County) near Porter Peak, in the Santa Rosa Mountains (Humboldt County) south of Windy Gap, at three locations in the Toiyabe Mountains (North Toiyabe Peak and Bunker Hill, Lander County, and near Arc Dome, Nye County), and on the southeast slopes of Boundary Peak in the White Mountains (Es- meralda County) just east of the California state line. Limber pine dwarf mistletoe was also collected on limber pine and Great Basin bristlecone pine near Mount Washington in the South Snake Mountains, probably near the same location where D. K. Bailey col- lected it on Great Basin bristlecone pine. In addition, a population of limber pine dwarf mistletoe was found in the Warner Mountains of northeastern California (Modoc County) on whitebark pine. Specimens of dwarf mistle- toes collected are deposited at the US DA Forest Service Pathology Herbarium (FPF), Rocky Mountain Forest and Range Experi- ment Station, Fort Collins, Colorado. Four additional mountain ranges in Nevada were visited: the Toquima Mountains (Mount Jef- ferson area) and the Monitor Range (Monitor Peak area) in Lander County, the Indepen- dence Mountains (Jack Peak area) in Elko County, and the Pine Forest Mountains (Duf- fer Peak area) in Humboldt County. Although extensive populations of limber pine and/or whitebark pine were observed in those areas, no dwarf mistletoe was found. However, it is probable that the parasite occurs in other 'us. Forest Service, Forest Pest Management, .324 2.5tli Street, Ogden, Utah 84401. U.S. Forest Service, Rocky Mountain Forest and Range E.xperinient Station, Fort Collins, Colorado 80.526. 91 92 Notes [Volume 50 Fig. 1. Distribution o{ Arcetithohium cyanocarpum in Nevada and adjacent areas. Solid dots denote previously reported locations (Hawksworth and Wiens 1972, 1984), and open circles indicate locations first reported in this studv. parts of these ranges as well as other isolated mountain ranges in Nevada that support size- able populations of limber pine, Great Basin bristlecone pine, or whitebark pine (Criteh- field and Allenbaugh 1969, Critchfield and Little 1971). Considering that limber pine dwarf mistle- toe is noted for its extremely disjunct distribu- tion in the western United States (Hawks- worth and Wiens 1972, 1984) and its occur- rence in adjacent areas in western Utah (Deep Creek Mountains) and eastern California (Panamint Mountains and Warner Moun- tains), it is not surprising to lind it surviving in widely separated mountain ranges in Nevada. Both limber and Great Basin bristlecone pines occurred some 800 to 1,000 m lower than at present during the late Quaternary (Thompson and Mead 1982). Thus, their dwarf mistletoe was presumably much more widely distributed then also. However, as the climate warmed, the pines and their associ- ated dwarf mistletoe receded to the higher elevations of major ranges and the dwarf mistletoe became restricted to scattered relictual populations. Literature Cited Critchfield, W B , .^nd G L Allenbaugh. 1969. The distribution of Pinaceae in and near northern Ne- vada. Madrono 20: 12-26. Cbitchfield, VV B . and E L. Little 1971. Geographic distribution of the pines of the world. USDA Forest Service Miscellaneous Publication 991. 97 pp. Hawksworth. F G. 1988. Mistletoes of Nevada. North- ern Nevada Native Plant Society Newsletter 14: 3. Hawksworth, F. G , and D. Wiens 1972. Biology and classification of dwarf mistletoes (Arcetithobium). USDA Forest Service Agricultural Handbook 401. 234 pp. 1984. Biology and classification oi Arccuthobittni: an update. Pp. 2-17 in Biology of dwarf mistle- toes: proceedings of the symposium. USDA For- est Service General Technical Report RM-111. Kartesz, J. T. 1988. A flora of Nevada. Unpublished dis- sertation. University of Nevada, Reno. 1,729 pp. (Dissertation Abstracts International B 49: 5119. 1989). LiNSDALE, M A , J T Howell, and] M Linsdale. 1952. Plants of the Toivabe Mountains area, Nevada. Wasmannjournalof Biology 10: 129-200. Mathiasen. R L , AND F G Hawksworth 1988. Dwarf mistletoes on western white pine and whitebark pine in northern California and southern Oregon. Forest Science 34: 429-440. Thompson, R. S., and J. I Mead 1982. Late Quaternary environments and biogeography in the Great Basin. Quaternary Research 17: 39-55. Received 25 November 1989 Accepted 30 November 1989 (;ival Hasin N.itiM.ilisI "i()( I > !')'«), pp. 93-95 SOREX PREBLEI IN THE NORTHERN GREAT BASIN Mark A. Forts' and Saiali B. George" Sorex prchlci has becMi described as a rare shrew of tlie Cohiinbia basin, with a distril:)u- tion extending as far east as the nortliwestern Great Plains (Junge and Hoffmann 1981). Its presence in the Great Basin desert is delin- eated by specimens that have been collected primarily on the periphery of this region (Zeveloff^ 1988). Other specimens have been collected from the northern, eastern, and western edges of the Great Basin. These in- clude shrews from the northeastern corner of California (Williams 1984), the northwestern corner of Nevada (Hoffmann and Fisher 1978), eastern Oregon (Jackson 1922, 1928, Hansen 1964, Verts 1975), southeastern Washington (Armstrong 1957), west central Idaho (Larrison and Johnson 1981), Montana (Hoffinann et al. 1969, Hoffmann and Fisher 1978), western Wyoming (Hoffmann et al. 1969), and the south shore of the Great Salt Lake in Utah (Tomasi and HoflPmann 1984). There are no records from the Snake River Plain of southern Idaho or from the bulk of the Great Basin Desert in Nevada or Utah south of the 40th meridian. Herein we report records of Sorex preblei from Elko County, Nevada, in the northern Great Basin. These records fill in a major gap in the distribution of the species and offer additional information on their habitat and sympatric associations with other species of Sorex. The northern Great Basin is included in the Great Basin Division of the Intermountain Floristic Region (Holmgren 1972). This area includes the closed river basins of the Hum- boldt River and the Mary's River as well as the seasonally wet Bonneville Flats and the Snake River Plain. The region has a continental climate of fairly hot summers and cold, snowy winters. Approximately 130 fault-block moun- tain ranges following a north-to-northeast pro- gression, and high valley floors are the prom- inent physiographic features. Within the higher valleys and river drainages are mesic to xeric shrublands dominated by Artemisia and Chrysothamntis. The river bottoms are domi- nated by more xeric halophytes such as Sarco- batus and Chnjsothamnus . Collection of S. preblei specimens from northern Elko County suggests that this shrew is more common and widespread in the northern Great Basin than previously supposed. A total of 12 specimens of S. preblei were collected near Sheep Creek, 10 km west of Haystack Ranch (55 km N of Elko); Sheep Creek drains the Independence Range. Seventy-five sunken pitfall traps were placed in a grid and shrews were sampled from this area from June to October 1984 (Ports and McAdoo 1986). Other Sorex include 31 S. vafirons, 13 S. monticolus, and 7 S. inerri- ami. These specimens are housed at the Natu- ral History Museum of Los Angeles and Northern Nevada Community College. S. preblei was also collected approximately 49 km east of Sheep Creek in the perennial riparian willow and wild rose of the Mary's River. At this locality, approximately 87 km NE of Elko, a single specimen was taken in a snap trap on 12 June 1986. Also collected here were three specimens of S. vagrans and one specimen of S. monticolus. These records fill in a distributional gap for S. preblei in the Great Basin. The only other record for this species in Nevada is from Washoe County (Hoffmann and Fisher 1978). To the east, S. preblei has been collected from the southern shore of the Great Salt Lake, Tooele Co., Utah (Tomasi and Hoffmann 1984). The specimen from the NW corner of Nevada, Washoe County, is approximately 'Life Science Department, Northern Nevada Community College, 901 Elm .St., Elko. Nevada S9801. -Section of Mammalogy, Natural History Museum of Los Angeles County, 900 E.xposition BKd , Los Angeles, California 90007. 93 94 Notes [Volume 50 345 km from the Sheep Creek locahty, whereas the Mary's River locahty is approxi- mately 233 km from the Great Salt Lake record. To the north, the nearest record of S. preblei is from the Jordan Valley, Malheur County, Oregon (Jackson 1922), approxi- mately 223 km away from the north edge of the Great Basin. Both the Independence Range and the Mary's River are on the border of the floristic regions of the Great Basin to the south and the Columbia basin to the north (Holmgren 1972). Some authors have suggested that S. pre- blei is found in marshy and riparian habitats (Larrison and Johnson 1981), whereas others have stated that this species prefers "arid to semiarid shrub-grass associations or openings in montane coniferous forest dominated by sagebrush" (Tomasi and Hoffmann 1984). Our habitat observations in northeastern Nevada onlv partiallv agree with those of Tomasi and Hoffmann (1984). At the Sheep Creek locality we collected S. pre/;»/cnn a seasonally wet, sagebrush-dom- inated community. The stream exhibits a snowmelt spring rimoff from early spring until early summer, and then it dries up. Big sage- brush (Artemisia trident at a), rubber rabbit- brush (ChnjsotJunnnus naiiseosns), and ante- lope bitterbrush (Purshia tridentata) provide a dense overstory, and a variety of bunchgrass species and forbs provide a stable understory. By the end of summer, the grasses and forbs have seeded and the area is very dry. The elevation at this locality is 2,150 m. The Mary's River locality has a perennial source of water from the Jarbidge Mountains to the north. The soils here tend to be more fertile with extensive meadows, and the dominant vegetation includes willows (Salix sp.). Wood's rose (Rosa woodsii), greasewood (Sarcobatiis sp.), and Great Basin wildrye (Ehpnns cana- densis). Other grasses and forbs provide an abundant cover in adjacent meadows that are cut seasonally for hay. Williams (1984) has collected S. preblei in similar mixed sagebrush, aspen, and willow riparian habitats in the Warner Mountains of northeastern California. Six years of shrew trapping by the senior author in wet meadows, montane coniferous forest, and high-elevation mountain brush of northeastern and central Nevada have failed to turn up a specimen of S. preblei (Ports and McAdoo 1986). The most likely habitat associ- ations for S. preblei in this area seem to be ephemeral and perennial streams dominated by shrubs, primarily below 2,500 m in eleva- tion. Although sympatry among species of long- tailed shrews is common in the western United States (Spencer and Pettus 1966, Williams 1984), S. preblei has rarely been captured in association with other shrews. Junge and Hoffmann (1981) found S. preblei associated with its congener, S. cinereus Juiydeni, in coniferous forest-mountain shrub habitats in Yellowstone County, Montana. In California's Warner Mountains, S. preblei is sympatric with S. va^rans and S. merriami in ecotonal habitats of forested riparian and drv shrublands (Williams 1984). Herein we document the association of S. preblei with three other species oi Sorex. It is difficult to explain why the plant commu- nit\' at Sheep Creek, which is a xeric, ephem- eral stream habitat, can support four species of shrews, whereas the more mesic, complex mountain habitat in the Warner Mountains and the perennial riparian habitat on the Marys River support only three shrew spe- cies. Certainly it is unusual to find S. montico- liis in a low-elevation, sagebrush community; this species is normally associated with more mesic habitats in Nevada (Hall 1946). Churchfield (in press) suggests that shrews are very flexible in their foraging habits and will decrease their dietar\' overlap when the number of shrew species in a community in- creases. Dietary generalists usually are found in greater numbers than dietary specialists in multi-species communities. This may be the case with shrew communities in the Great Basin, but these species assemblages will be explained only with detailed ecological stud- ies examining microhabitats, diets, and life- histor> patterns, plus comprehensive trap- ping to insiue that all shrew species in a particular area are well documented. Acknowledgments We wish to thank Marcus "Pete" Rawlins and the Ne\ada Department of Wildlife for their contributions of shrew specimens and habitat information from the Mary's River. We also thank Lois Ports who assisted in the field. 1990] NOTKS 95 LitkkatuhkCJitkd Armstkonc. F II 1957. Notes on Sorcx })r(l)hi in Wasliington Statf. MurrrK't 3(S; (i. Cm HCHKIKI.I), S In press. Niclic cK TiaTuics, food rc- sourc'fS, and leedinj:; strategics in midti-spccics connminities of shrews. Musenni of Soiitliwesteni Biolog) Special Publication. II,\1J„ E R 1946. The maiinnals of'Nexada. Uni\ersit\()i Caliiornia Press, Berkelc\'. 710 pp. II.\NSEN. A, 1964. Ectoparasites of maniinals from Ore- gon. Great Basin Naturalist 24: 75-81. I1()IKM.\NN, R. S . AND R D FisnER. 1978. Additional distributional records of Preble's shrew {Sorcx prcbU'i). Journal of Mannnalog\ 59: 883-884. HoFKM.ANN. R. S . P L. Wright, .-\nd F E. Newby. 1969. The distribution of some mammals in Montana. I. Mammals other than bats. Jomnal of Mammalogv 50: 579-604. Holmgren, N. H 1972. Plant geography of the Inter- mountain Region. Pp. 77-161 ;;; A. Cronquist, A. H. Holmgren, N. H. Holmgren, and J. L. Reveal, Intermountain flora. Vol. I. Hafner Pub- lishing Co. , New York. Jac;ks()N. H H. T 1922. New species and subspecies of Sorer from western America. Journal of Washing- ton Academy of Science 12: 262-264. 1928. A ta.xonomic review of the American long- tailed shrews. North American Fauna 51: 1-238. Jl'NGE.J A \\l)ii S Hoi'KMANN 1981 . An annotated key to tlu- long-tailed shrews (genus Sorex) of the United States and (>'anada, witli notes on middle American Sorcx. Occasional l^apcrs, Museum of Natural History, University of Kansas 94: 1-48. Larri.son. E. J., AND D R Johnson 1981. Manniials of Idaho. University of Idaho Press, Moscow. 166 pp. Poms, M. A., AND J K McAdoo 1986. Sorex merriami (Insectivora: Soricidae) in eastern Nevada. South- western Naturalist 31: 415-416. Si'Encer, a. W., and D. Pettus. 1966. Habitat prefer- ences in five sympatric species of long-tailed shrews. Ecology 47: 677-683. ToMASi, T. E.. and R. S. Hoffmann. 1984. Sorcx prchlci in Utah and Wyoming. Journal of Mammalogy 65: 708. Verts, B J 1975. New records for three uncommon mam- mals in Oregon. Murrelet .56: 22-23. Wh.liams, D F. 1984. Habitat associations of some rare shrews (Sorcx) from California. Journal of Mam- malogy 65: 325-328. Zexeloff, S I 1988. Mammals of the Intermountain West. University of Utah Press, Salt Lake City. 365 pp. Received 10 February 1989 Accepted 12 December 1989 GREAT BASIN NATURALIST MEMOIRS SERIES The Great Basin Naturalist Memoirs series was established in 1976 for scholarly works in biological natinal history of greater length than can be accommodated in the Great Basin Naturalist. The Memoirs series appears irregularly. Approval for the production is the respon- sibility of the Great Basin Naturalist Editorial Board. Information concerning the production of the Memoirs series can be obtained from James R. Barnes, Editor, Great Basin Naturalist. 290 MLBM, Brigham Young University, Provo, UT 84602. GREAT BASIN NATURALIST MEMOIRS All Memoirs are available for purchase. Direct inquiries to: Editorial OflPice, Attn: Carolyn Backman, 290 MLBM, Brigham Young University, Provo, UT 84602. Only checks or money orders drawn on U.S. banks or U.S. currency are accepted. Please include $2 per copy for mailing and handling charges. No. 1 The birds of Utah. $10 No. 2 Intermountain biogeography: a symposium. $15 No. 3 The endangered species: a symposium. $6 No. 4 Soil-plant-animal relationships bearing on revegetation and land reclamation in Nevada deserts. $6 No. 5 Utah Lake monograph. $8 No. 6 The bark and ambrosia beetles of North and Central America (Coleoptera: Scolyti- dae), a taxonomic monograph. $60 No. 7 Biology of desert rodents. $8 No. 8 The black-footed ferret. $10 No. 9 A Utah flora. $40 No. 10 A reclassification of the genera of Scolytidae (Coleoptera). $10 No. 11 A catalog of Scolytidae and Platypodidae (Coleoptera), part 1: bibliography. $30 No. 12 Research in the Auchenorrhyncha, Homoptera: a tribute to Paul W. Oman. $30 INFORMATION FOR AUTHORS The Great Basin Naturalist welcomes previously unpublished manuscripts pertaining to the biologi- cal natural history of Western North America. Pref- erence will be given to concise manuscripts of up to 12, OOO words. MANUSCRIPTS SHOULD BE SUBMITTED to James R. Barnes, Editor, Great Basin Naturalist, 290 MLBM, Brigham Young University, Provo, Utah 84602. A cover letter accompanying the manu- script must include phone number(s) of the author submitting the manuscript; it must also provide information describing the extent to which data, text, or illustrations have been used in other papers or books that are published, in press, submitted, or soon to be submitted elsewhere. Authors should adhere to the following guidelines; manuscripts not so prepared may be returned for revision. CONSULT THE MOST RECENT ISSUE of this journal for general style and format. Also refer to the CBE Style Manual, 5th edition (Council of Biology Edi- tors, 9650 Rockville Pike, Bethesda, MD 20814 USA; $24). TYPE AND DOUBLE SPACE all materials, including literature cited, table headings, and figure legends. Avoid hyphenated words at right-hand margins. Underline words to be printed in italics. Use stan- dard bond (22 x 28 cm), leaving 2.5-cm margins on all sides. SUBMIT 3 COPIES of the manuscript. Number all pages and assemble each copy separately: title page, abstract and key words, text, acknowledg- ments, literature cited, appendices, tables, figure legends, figures. TITLE PAGE includes an informative title no lon- ger than 15 words, names and addresses of authors, a running head of fewer than 40 letters and spaces, footnotes to indicate change of address and author to whom correspondence should be addressed if other than the first author. ABSTRACT states the purpose, methods, results, and conclusions of the research. It is followed by 6-12 key words, listed in order of decreasing im- portance, to be used for indexing. TEXT has centered main headings printed in all capital letters; second-level headings are centered in upper- and lower-case letters; third-level head- ings begin paragraphs. VOUCHER SPECIMENS. Authors are encouraged to designate, properly prepare, label, and deposit high-quality voucher specimens and cultures docu- menting their research in an established perma- nent collection, and to cite the repository in publi- cation. REFERENCES IN THE TEXT are cited by author and date; e.g., Martin (1989) or (Martin 1989). Multiple citations should be separated by commas and listed in chronological order. Use "et al. ' after name of first author for citations having more than two au- thors. ACKNOWLEDGMENTS, under a centered main heading, include special publication numbers when appropriate. LITERATURE CITED, also under a centered main heading, lists references alphabetically in the fol- lowing formats: Mack, G. D., and L. D. Flake. 1980. Habitat rela- tionships of waterfowl broods on South Da- kota stock ponds. Journal of Wildlife Man- agement 44: 695-700. Sousa, W. P. 1985. Disturbance and patch dynam- ics on rocky intertidal shores. Pages 101-124 in S. T. A. Pickett and P. S. White, eds. , The ecology of natural disturbance and patch dy- namics. Academic Press, New York. Coulson, R. N., and J. A. Witter. 1984. Forest entomology: ecology and management. John Wiley and Sons, Inc., New York. 669 pp. TABLES are double spaced on separate sheets and designed to fit the width of either a single column or a page. Use lower-case letters to indicate foot- notes. PHOTOCOPIES OF FIGURES are submitted ini- tially with the manuscript; editors may suggest changes. Lettering on figures should be large enough to withstand reduction to one- or two- column width. Originals must be no larger than 22 X 28 cm. NOTES. If your manuscript would be more ap- propriate as a short communication or note, follow the above instructions but do not include an ab- stract. A CHARGE of $45 per page is made for articles published; the rate for subscribers will be $40 per page. However, manuscripts with complex tables and/or numerous half-tones will be assessed an additional charge. Reprints may be purchased at the time of publication (an order form is sent with the proofs). FINAL CHECK: • Cover letter explains any duplication of infor- mation and provides phone number(s) • 3 copies of the manuscript • Conformity with instructions • Photocopies of illustrations (ISSN 017-3614) GREAT BASIN NATURALIST voi so. no i. March 1990 CONTENTS Articles A plasma protein marker for population genetic studies of the desert tortoise (Xerobates agassizi) James L. Glenn, Richard C. Straight, and Jack W. Sites, Jr. 1 Effects of nitrogen availability on growth and photosynthesis of Artemisia tridentata ssp. wyomingensis Paul S. Doescher, Richard F. Miller, Jianguo Wang, and Jeff Rose 9 Form and dispersion of Mima mounds in relation to slope steepness and aspect on the Columbia Plateau George W. Cox 21 Esox lucius (Esocidae) and Stizostedion vitreum (Percidae) in the Green River basin, Colorado and Utah Harold M. Tyus and James M. Beard 33 Influence of soil frost on infiltration of shrub coppice dune and dune interspace soils in southeastern Nevada Wilbert H. Blackburn and M. Karl Wood 41 Seed production and seedling establishment of a Southwest riparian tree, Arizona walnut (Juglans major) Juliet C. Strombergand Duncan T. Patten 47 Forage quality of rillscale (Atriplex suckleyi) grown on amended bentonite mine spoil Marguerite E. Voorhees 57 Summer food habits of coyotes in Idaho's River of No Return Wilderness Area Charles L. Elliott and Richard Guetig 63 Infection of young Douglas-firs by dwarf mistletoe in the Southwest Robert L. Mathiasen, Carleton B. Edminster, and Frank G. Hawksworth 67 New Mexico grass types and a selected bibliography of New Mexico grass taxonomy Kelly W. Allred 73 Notes Noteworthy mammal distribution records for the Nevada Test Site. . . . Philip A. Medica 83 Formation of Pisolithiis tinctorius ectomycorrhizae on California white fir in an eastern Sierra Nevada minesoil R. F. Walker 85 Bone chewing by Rocky Mountain bighorn sheep K. A. Keating 89 Distribution of limber pine dwarf mistletoe in Nevada Robert L. Mathiasen and Frank G. Hawksworth 91 Sorex preblei in the northern Great Basin Mark A. Ports and Sarah B. George 93 MCZ 115RARY H gEP 2« 1990 MARVARO UNIVER^ri Y GREAT BASIN NiffURAUST VOLUME 50 NO 2 - JUNE 1990 BRIGHAM YOUNG UNIVERSITY GREAT BASIN NATURALIST Editor James R. Barnes 290MLBM Brigham Young University Provo, Utah 84602 Associate Editors MichaelA. Bowers Blandy Experimental Farm University of Virginia Box 175 Boyce, Virginia 22620 Brian A. Maurer Department of Zoology Brigham Young University Provo, Utah 84602 PaulC. Marsh Center for Environmental Studies Arizona State University Tempe, Arizona 85287 JimmieR. Parrish Department of Zoology Brigham Young University Provo, Utah 84602 Other Associate Editors are in the process of being selected. Editorial Board. Richard W. Baumann, Chairman, Zoology; H. Duane Smith, Zoology; Clayton M. White, Zoology; Jerran T. Flinders, Botany and Range Science; William Hess, Botany and Range Science. All are at Brigham Young University. Ex Officio Editorial Board members include Clayton S. Huber, Dean, College of Biological and Agricultural Sciences; Norman A. Darais, University Editor, University Publications; James R. Barnes, Editor, Great Basin Naturalist. The Great Basin Naturalist, founded in 1939, is published quarterly by Brigham Young University. Unpublished manuscripts that further our biological understanding of the Great Basin and surrounding areas in western North America are accepted for publication. Subscriptions. Annual subscriptions to the Great Basin Naturalist for 1990 are $25 for individual subscribers, $15 for student and emeritus subscriptions, and $40 for institutions (outside the United States, $30, $20, and $45, respectively). The price of single issues is $12. All back issues are in print and available for sale. All matters pertaining to subscriptions, back issues, or other business should be directed to the Editor, Great Basin Naturalist, 290 MLBM, Brigham Young University, Provo, UT 84602. Scholarly Exchanges. Libraries or other organizations interested in obtaining the Great Basin Naturalist through a continuing exchange of scholarly publications should contact the Exchange Librarian, Harold B. Lee Library, Brigham Young University, Provo, UT 84602. Editorial Production Staff JoAnne Abel Technical Editor Carolyn Backman Assistant to the Editor Heidi Larsen Production Assistant ISSN 017-3614 8-90 6.50 46829 The Great Basin Naturalist PUHLISIIEDAI PHOVO, UlAII, hV Brig HAM Young University ISSN 0017-3614 Volume 50 30 June 1990 No. 2 MAYFLY GROWTH AND POPULATION DENSITY IN CONSTANT AND VARIABLE TEMPERATURE REGIMES Russfll B. Rader' -and James V. Ward' Abstract. — The tliernial ecjuilihriiim hypothesis predicts that acjiiatic inseet body size/fecundity and, consequently, population densit\- and biomass will be maximized in geographic areas or along altitudinal gradients where the thermal regime is optimal with respect to growth and development. Seasonal growth analyses of three mayfly species, combined with detailed thermal descriptions, were used to explore differences in body size and fecimditv at three sites with similar elevations but different temperature regimes. Site 1 was located near the upper altitudinal distribution for each species, whereas sites 2 and 3 were located below a deep-release storage reservoir. The temperature pattern at site 1 had rapid seasonal changes, with a short summer and a long, freezing winter. Site 2 demonstrated gradual seasonal changes combined with winter warm and sinnmer cool temperatures. Site 3 was intermediate with respect to seasonal change and winter harshness but had the highest maximum and mean annual temperatures. Mayfly develop- ment at site 1 was characterized by slow growth during the summer-autumn period, no growth during the winter, and a rapid increase during the spring-summer period. In contrast, growth at site 2 was continuous throughout the year, including the winter. Growth at site 3 was either continuous across sites or rapid during the spring-simimer period, depending on the species. Based upon the interactions among temperature, body size, and metabolic costs, the thermal equilibrium hypothesis was successfid at predicting body size and fecundity differences among sites. It was less successful at predicting variation in population density and biomass. Density-dependent and density-independent sources of mortality, including temperature, may interrupt the translation of higher fecundity into higher population density and biomass. Temperature, because of its influence on optimal with respect to growth, development, metabolism, growth, and reproductive sue- and body size. Fecundity, an essential compo- cess, is a dominant ecological determinant of nent, but not the only factor defining repro- the geographical and altitudinal distributions ductive success, should decrease in tempera- of aquatic insects (e.g., Vannote and Sweeney ture regimes warmer or cooler than optimal. 1980, Ward and Stanford 1982, Sweeney Other factors, which may influence fecundity 1984, Ward 1986). The thermal equilibrium and may or may not be influenced by temper- hypothesis (TE) is a conceptual model of the ature, can determine population size and dis- effects of temperature on aquatic insect tribution (e.g., egg-hatching success, emer- metabolism, growth, body size, and therefore gence success, mating success, feeding rates, fecundity (Sweeney and Vannote 1978, Van- assimilation efficiency, food quality and quan- note and Sweeney 1980). It predicts that pop- tity, biotic interactions). The TE hypothesis, ulation density, distribution, and stability however, attempts to define the influence of (Connell and Sousa 1983) are determined liy temperature on population size and distribu- individual reproductive success and will be tion based only on the effect of temperature ma.ximized in geographic areas or along altitu- on metabolism, growth, and therefore body dinal gradients where the thermal regime is size/fecundity. ' Department of Biolog\'. Colorado State Universit>. Fort Collins, Colorado 8(62.3. Present address: Savannah River Eeology Laboratory. Drawer E. Aiken, South Carolina 2980.3. 97 98 R. B. RA.DERAND J. V. WaRD [Volume 50 The objective of this study was to analyze the influence of temperature on the growth and body size/fecundity of three mayfly spe- cies and compare these results with popula- tion size (density and biomass) data and the predictions of the thermal equilibrium hypothesis. Three sites were chosen on the same river, all with similar elevations but dif- ferent temperature regimes. Site 1 was lo- cated near the upper altitudinal limit for each species, whereas sites 2 and 3 were located in a warmer, more constant thermal regime downstream from a deep-release reservoir. Specifically, we sought to test the following hypotheses: (1) population size and body size will be smaller for each species at site 1 com- pared with sites 2 and 3; (2) population size will correspond with body size at the same sites (i.e., site-specific ranks of body size and population size should be the same); (3) sea- sonal patterns of growth will parallel seasonal temperature patterns. Body size/fecundity was compared to examine its ability to explain among-site differences in population size. Several studies have demonstrated a posi- tive correlation between body size and fecun- dity in aquatic insects (see Clifford and Boer- ger 1974, Kondratieff and Voshell 1980, Sweeney and Vannote 1981). Therefore, we assume that larger mayflies produce more eggs compared with smaller mayflies of the same species. We did not attempt to deter- mine, as predicted by the TE hypothesis and numerous other authors, whether an increase in population size is positively correlated with an increase in population stability. Consider- able evidence, however, indicates that larger populations are more stable than smaller pop- ulations (e.g., smaller populations are more susceptible to extinction). Study Sites and Insect Altitudinal Distributions The first hypothesis recjuired us to a priori rank the study sites according to how closely they approximated the optimal temperature conditions for the mayflies imder investiga- tion. Our ranks were based on the altitudinal distributions of the insects. Because the study sites were located near their upper limits, the unaltered temperature regime at site 1 was considered cooler than the optimum neces- sary to maximize body size/fecundity. There- fore, the warmer, dam-impacted sites 2 and 3 were assumed to be nearer the insects' opti- mal temperature regime. There was no a pri- ori reason to separate sites 2 and 3 with re- spect to their influence on growth, body size, and population size even though they had very different temperature regimes. The study was conducted in the Upper Col- orado River on the western slope of the Rocky Mountains in the vicinity of Granby Reser- voir, a large (666 km ), deep-release storage impoundment. Granby Reservoir is located 38 km northwest of Denver, Colorado. Site 1 was located in a third-order, free-flowing sec- tion of the river 4.0 km above the reservoir; sites 2 and 3 were located 0.4 and 4.0 km, respectively, downstream from the dam. Al- though differentially influenced by stream regulation, all three sites had similar gradi- ents (0.006-0.009), canopy cover, geology, riparian vegetation, and elevation (2,593 m, 2,454 m, and 2,426 m, respectively. (For com- plete site descriptions see Rader and Ward [1988].) Three mayfly species were analyzed in this study: DruneUa grandis (Eaton), Ephemerella infrequens (McDunnough), and Baetis tricau- datus (Dodds). Ward (1980, 1986) determined the altitudinal distribution of macroinverte- brates, including the mayflies of this study, in the St. Vrain River, a free-flowing stream run- ning from the alpine tundra to the plains on the eastern slope of the Rocky Mountains. Based on his results, site 1 of this study (upper montane zone) was above the altitudinal dis- tribution for D. grondis and very near the upper limits for E. infrequens and B. tricaii- dafiis. All three species exhibited maximum densities at lower elevations in the foothills or plains zones. Even though their altitudinal upper limits appear to be somewhat higher in the Colorado River, probably because of its larger size compared with the St. Vrain River, we concluded that all three study sites were located near the upper altitudinal limit for each of the three mayfly species. Methods Temperature, Growth, and Body Size Water temperature was measured continu- oush' at each site for 18 months using Ryan 9()-day thermographs. Each thermograph was checked against a Weksler hand-held ther- mometer on a monthly basis and calibrated 1990] Mayfly (Jkowi II andTemi'kkatuhe Constancy 99 prior to placement and following retrieval . Daily mean temperatures were used to ealeu- late annual mean temperatures, annual coeffi- cient of variation, annual degree days, num- ber of days less than 3 C, number of days e(iual to 0 C, length of spring-summer and summer- autumn periods, and rate of spring-summer increase and summer-autumn decline. Annual growth rate analyses and general temperature descriptors (e.g., accumulated degree days, mean annual temperature, etc.) cannot explain site-specific variation in aquatic insect body size and fecundity be- cause they average over important seasonal information. Seasonal growth rate analyses combined with seasonal temperature profiles can, however, provide insights into the rela- tionships between temperature, growth, and body size. Temperature profiles for each site were separated into three periods: (1) spring-sum- mer, (2) summer-autumn, and (3) winter (Fig. 1). The winter period was defined by mean daily temperatures <3 C in order to include the winter warm temperatures at site 2. The end of the spring-summer period/ beginning of the summer-autumn period was set at 15 August, based on temperature peaks apparent at sites 1 and 3 (Fig. 1). Therefore, the spring-summer period began when the mean daily temperature exceeded 3 C and ended 15 August. The summer-autumn pe- riod began 15 August and ended when the mean daily temperature dropped below 3 C. Growth, defined as the monthly increase in mean biomass of individuals collected per sampling date, was determined for D. grandis and £. infreqiiens. Growth for B. tricaudatiis was not analyzed because of difficulty in as- signing intermediate-sized instars to the cor- rect generation. Site-specific differences in seasonal growth rates were determined by regressing monthly mean individual biomass estimates against the numl:)er of Julian days accumulated over the three separate growth periods (spring-sum- mer, summer-autumn, and winter). A slopes comparison test (analysis of covariance) was used to determine among-site differences in seasonal growth rates. No transformation was necessary because growth was linear over the short seasonal periods. Population Size and Body Size Estimates of population density and bio- mass were based on four Surber samples (().()9"m each, 240 jjim mesh) collected monthly across the width of the stream at each site and four artificial sjibstrates. Artificial substrates consisted ofclay bricks (23 x 19 x 9.5 cm) that had been in the streams for one month before being sampled. Most of the Surber samples enclosed natural substrate particles similar in size to the artificial brick substrates. There- fore, sampling units from both techni(jues were combined because the sampling areas were approximately et^ual. A simple t test indicated that the mayfly population means based on Surber samples of natural substrate were not significantly (P ^ .63) different from population means based on the artificial sub- strate samples. Following identification and enumeration, all nymphs were divided into 0. 1-mm size classes, based on maximum head capsule width, and dried at 60 C for 48 hours. Mean annual population biomass was deter- mined by summing biomass estimates for all size classes across all sampling dates. Mean annual population density was determined after summing the number of individuals in each size class at each site. Head-capsule measurements were also used to construct size-frequency plots for life-history determi- nations, including the number of annual generations produced. Complete life-history information for these species at each site can be found in Rader and Ward (1989). Female body size (dry weight l^iomass) of late instars was estimated by taking the mean of the three largest size classes collected from each site. Late mayfly instars have a full com- plement of mature eggs. Results Temperature, Growth, and Body Size A progressive increase in accumulated de- gree days and mean annual temperatures was found from site 1 to site 3 (Table 1). Site 1 was characterized by rapid seasonal changes in temperature and a long, freezing (0 C) winter (Table 1, Fig. 1). In contrast, site 2 demon- strated gradual seasonal changes combined with winter warm and summer cool tempera- tures. Site 3 was intermediate with respect to rates of seasonal change but had the greatest amount of thermal energy (largest number of accumulated degree days, largest maximum temperature, and largest annual mean tem- perature). 100 R. B. R\DER AND J. V. Ward [Volume 50 Site 1 Summer - 14 •< Autumn - Spring - Summer -,">> Winter 12 - .'!." 'V'.. 10 - 1 \ ,', t 8 - 1 1*' 1 , 1 ' * . ' . . ."' .'', • ,v.. ■' • ■ 1 ■ 6 7. 1 4 '-•'. v' '\, 2 - «.' -^ 1 0 " ' • , 1 A, , , , , , -V"" MAY JUN JUL AUG SEPT OCT NOV DEC JAN FEB MAR APR M/ 14 - _^ o Site 2 o 12 - lU . Spring - Summer Summer - Autumn Winter DC 3 - h- < GC 8 . .. .>^VV..'''''''''''''^"'^A^V. 1 LU ,':'r'\:M^ • • - \ CL . •Vs 1 2 UJ 6 ■\ H- ..''^' tf cc 4 "^^\ LU / |A 1 2 0 - 1 1 1 r 1 1 1 1 J 1 14 _MAY JUN 1111 1 1 1 ! JUL AU . 1 1 ■ . ■ 1 T I 1 1 1 P G SEPT OCT T I T T TT-T T T T r I 1 1 1 1 1 | 1 1 1 1 1 M 1 1 1 1 ' 1 1 1 T' 1 1 ' MOV DEC JAN FEB MAR APR MAY Site 3 , Summer - 12 Spring - Summer ,'l ^ , Autumn r Winter • 1 •V-' \: *" f 10 J ,1 I" , 'l- 8 w .' '• 6 -•A \ ',,,1 1 II 4 2 ■.y 0 1 1 ' 1 1 1 1 '' 1 ( -ea MAQ ADD K>IAV II IM ' ' 1 1 1 ■ 1 11 ri A 1 in ccDT r>nT M n\/ ' npr. ' ' i anj ' ccn Fis- 1. Tt'inpt'iatmc piolilcs ior cath site (liiriiii:; 19.S1-1(-)S2. Iii(li\ idii.il |i()ints rt'|iirsi'iit daiK im'ans. (See text for explanation of criteria used to determine seasonal separations. Nott' that the \-a\es lia\i' !)een adjusted to facilitate a comparison of the length of each period at each site.) 1990J Mavi'ia Ghowi II andTkmi'kkatuhi-: Constancy 101 Tabi.K 1. Tciiiiit'ratmr characteristics ol cacli site ("haractcMistic Site 1 Site 2 Site 3 Annual clei^ii'c cla\s (C) 1130 1729 2()S2 Mean annual temperature (C) 3.6 4,7 5.7 C.V. (%) 119 53 64 Minimum (C) 0.0 i.S 0.0 Maximum (C) 18.0 9,.S 18.2 Davs <3 C 191 155 101 Days = 0 C 153 0 17 Rate of spring-summer | (°C/clay) 0.14 0.05 0.06 Length of spring-summer (clays) 99 94 177 Rate of summer-autumn [ (°C/clay) 0.20 0.05 0. 10 Length of summer-autumn (clays) 70 115 87 Seasonal growth rates were analyzed to ex- plain among-site differences in body size for each species. The seasonal pattern of growth for D. ^ranclis and E. infrcqucns at site 1 was characterized by slow growth dining snmmer- autumn, no growth during winter, and a rapid increase during the spring-summer period (Figs. 2, 3). In contrast, growth at site 2 was comparatively fast and continuous through each seasonal period including winter. At site 3, D. grandis demonstrated a continuous growth pattern similar to that observed at site 2, whereas E. infrcqucns demonstrated a spring-summer pulsed pattern of growth more similar to that at site 1. The seasonal growth rate (0.0108 mg/day) for the early instars of £. infrcqucns during the summer-autumn period was significantly (P = .0001) faster at site 2 than at sites 1 and 3 (0.004 and 0.003 mg/day, respectively), which were not significantly different (Fig. 2). This trend continued into the winter period when the growth rate at site 2 (0.022 mg/day) was again significantly (F = .0001) faster than at sites 1 and 3, where growth rate was not differ- ent from zero. By the end of winter, the site 2 population had completed over 76% of its growth, and larvae were over five times larger than those at sites 1 and 3 (Fig. 2). Therefore, the larger body sizes at site 2 can be attributed to rapid growth starting at egg hatch and con- tinuing through the winter. During the spring-summer growth period, individuals at sites 1 and 3 grew significantly (P = .0001) faster (0.034 and 0.033 mg/day, respectively) than individuals at site 2 (0.014 mg/day). However, the body sizes of late instar larvae at sites 1 and 3 were still considerably smaller than those at site 2. The growth rate (0.064 mg/day) of early instars ol D. grandis dining the summer- autunni period was significantly different among sites (P = .0001), being greatest at site 2 and slowest at site 1 (0.035 mg/day). Winter growth was also faster at site 2, with an aver- age rate of 0.064 mg/day, followed by site 3 (0.044 mg/day) and tlien site 1 (0.015 mg/day). Winter growth at site 1 was not significantly different from zero. Spring-summer growth at site 1, however, was significantly (P .0001) faster than at sites 2 or 3, with the fastest seasonal rate of increase for this study (0.119 mg/day). Spring-summer growth rates at sites 2 and 3 (0.076 and 0.055 mg/day, respec- tively) were not significantly different. Bactis tricaudatus was univoltine at site 1, but bivoltine with slow and fast seasonal gen- erations at sites 2 and 3. In contrast, D. gran- dis and E. infrcqucns had univoltine, slow seasonal life cycles at each site. Complete life- history data for each species can be found in Rader and Ward (1989). As predicted by the first hypothesis, mean annual density and biomass of each species were much greater at the warmer sites below the dam than at site 1 (Table 2). Population densit) and biomass of Bactis tricaudatus and E. infrcqucns were largest at site 2, followed by sites 3 and 1. Maximum density and biomass of D. grandis were greatest at site 3, followed by sites 2 and 1. Contrary to the predictions of the second hypothesis, dry weights of the largest instars, and population sizes of each species, did not correspond when ranked by sites (Table 2). Although the largest late instars of E. infrc- qucns occurred at site 2, where the density was greatest, the largest instars of B. tricauda- tus and D. grandis did not occiu- in the largest population. Body size and population size cor- responded in only two other instances; the smallest late instars of B. tricaudatus and the intermediate-sized late instars of D. grandis occurred in the small- and intermediate-sized populations, respectively, at site 1. The largest B. tricaudatus larvae occurred at site 3, whereas its maximum population density and biomass occurred at site 2. Late B. tricaudatus instars at sites 2 and 3 were over two times larger than late instars at site 1 (Table 2). The largest-sized larvae of D. gran- dis occurred at site 2. Site 3, which had the largest population of D. grandis, had the smallest late instars (Table 2). 102 R. B. Rader AND J. V. Ward [Volume 50 1.5 T 1.0 -- SUMMER-AUTUMN C/) if) < < 9 > < 4.U - WINTER 0 3.0 - - ...g- ■ ■■' 2.0 - 8 . ■ • ■ ■ o ■ 6 8 1.0 - .^ —St ^ 0.0 - 1 \ 1 \ 160 190 220 250 280 o.u- SPRING- -SUMMER 5.0- 4.0- . 8 . • • • • •• o •■•«•■■■ .. .8 •■••••■■ ■ o O ■ o o t 3.0- . 1 o t^^'^'^^ 2.0- «l^'' ^-^^ A 1.0- - A 0.0- 1 1 1 280 310 340 JULIAN DAYS 370 Fig. 2. St'asonal iijrovvth rate patterns ior Ephciucrclld iiifrciiucn.s. Site 1 is rrpiesented In a solid liiu' and diamonds, sites 2 and 3 l)y a dotted line and eireles, and a dashed line and tiian,e;les, respeetix ely. Each s\nil)ol (diamond, circle, and triangle) represents the mean hiomass for a single sample. 1990] May FLY (iRovvTH andTempekatuke Constancy 103 180 170 20.0 T 200 SPRING-SUMMER 230 260 290 260 290 320 JULIAN DAYS 350 380 Fig. 3. Seasonal growth rate patterns for DnineUa grundis. (For further description, see Fig. 2 legend. 104 R. B. R\DERAND J V. Ward [Volume 50 T.ABLE 2. Mean annual population densit\ (#'.s m " ) and biomass (mg m " ), plus mean individual size (nig dr\- \vt. ) for late instars of each species. Values in parentheses for the population parameters are the percentage of mean represented by the standard error. Values in parentheses for size estimates indicate the number of individuals used to determine each mean. Species Site 1 Site 2 Site 3 Popula Density tion size Biomass Bod\' size PopuL DensitN ition size Biomass Body size PopuL DensitN ition size Biomass Body size B. tricaudatiifi 754 (21%) 142.6 (17%) 0.730 (31) 7720 (18%) 1501.7 (21%) 0.550 (35) 4018 (18%) 1162.8 (27%) 1.960 (35) E. infreqiiens 367 (25%) 502.6 (23%) 3.500 (23) 1034 (23%) 4964.7 (25%) 4.930 (35) 644 (24%) 1188.5 (26%) 2.800 (25) D iirondis 102 (31%) 1585.9 (18%) 14.930 (16) 80 (10%) 2368.4 (22%) 18.110 (9) 278 (14%) 3355.4 (20%) 14.220 (17) Discussion Temperature, Growth, aud Body Size The winter warm and summer cool condi- tions of site 2 allowed rapid continuous growth of £. infreqiiens and D. (irandis, which pro- duced larger instars and greater fecundity. Site-specific explanations of growth patterns and body-size diflferences for D. grandis and E. infrequens are consistent with the TE hy- pothesis. Vannote and Sweeney (1980) pro- posed that the seasonal pattern of growth for aquatic insects, as for other small ectotherms (e.g., Phillipson 1981), may be determined by the interaction between temperature and body size. Smaller instars, which have a large surface-to-volume ratio, will have a higher metabolic rate than larger instars at the same temperature. At site 1 the smallest D. grandis and E. infrequens instars appeared during the warmest months of the year (July, August, and September). High summer-autumn tem- peratures coincident with small instars at site 1 likely resulted in large metabolic costs and, therefore, slow growth rates and possibly high mortality rates. Both species ceased to grow during the freezing winter tempera- tures. In the spring, winter survivors experi- enced a rapid increase in temperature and thus a relatively short period (99 da\s or less) to complete growth and maturation. All else being equal, the magnitude and length of summer-autumn temperatures when coinci- dent with early instars, plus the rate of vernal rise, limit growth and bocK size/lecundity and probably have an important intluence on the geographic distribution and upper altitudinal limits of aquatic insects. This may be espe- cially applicable to cool-adapted boreal spe- cies (see Edmunds 1982). Early instars at site 2 began growth in much cooler summer-autumn temperatures; meta- bolic costs were low and growth rates fast compared to those at sites 1 and 3. Winter temperatures, which varied slightly around 2 C, were not sufficiently cold to inhibit growth, which continued at a rapid pace. Growth appeared near completion before the vernal rise in temperature, thus leaving plent\ of time for maturation and emergence. The rapid completion of growth probabh' re- sulted in the extended emergence of £. infre- quens and the addition of a second generation of B. tricaudatus at site 2. The earl\ instars of D. grandis (JuK) and £. infrequens (August and September) began growth at site 3 during the warmest months of the year. August and September were, on average, 7-8 C warmer at site 3 than site 2. Although E. infrequens did not grow, D. grandis early instars grew rapidly during the siunmer-autumn growth period. Because D. grandis earl\- instars were approximately two times larger than E. infrequens earh' instars, they probabh had lower metabolic costs. This allowed them to grow at the warm summer- autinnn temperatures. Although both species stopped growing during winter at site 1 but continued to grow during winter at site 2, only D. grandis continued winter growth at site 3. The fact that D. grandis was over three times larger than E. itifrequens at the beginning of winter ma\ explain its ability to grow in winter at site 3 in contrast to E. infrequens. These site-specific explanations of growth are consistent with the TE Inpothesis 1990] Mayfly Ghowiii andTemfkhai urk Constancy 105 suggesting that growth rate and e()nse(|uently body size and iecundity are determined by the length of time individuals are exposed to a speeific optimal range of temperatures. Other factors, however, that may also influence mayfly growth rates (e.g., food abundance; Sweeney et al. 1986) were altered by the ef- fects of stream regulation. For example, con- stant flow conditions and the addition of planktonic diatoms from the reservoir en- hanced food qualitv and quantitv at sites 2 and 3(RaderandWardl989). If summer temperatures increase meta- bolic costs, causing growth to slow or stop, then the rate of vernal rise and autimin de- cline determines the amount of time individu- als are exposed to optimal temperatures and, therefore, the amount of time available for growth. Growth of both D. grandis and E. infrequens continued as long as temperatures remained between 2 C and 10 C. However, when temperatures exceeded this range, growth slowed or stopped. Where growth was continuous (site 2), temperatures were always within this range. The optimal temperature range for these two species appears to lie be- tween 2 C and 10 C. Population Size and Body Size Winter warm and summer cool conditions at site 2 and the long spring-summer period at site 3 probably accounted for the multivoltine life cycle of B. tricaiidatus at these sites. Stan- ley and Short (1988) suggested that population size may remain unaltered or even increase in warmer than optimal conditions if faster growth rates and shorter generation times compensate for smaller body sizes and lower fecundity. For some aquatic insects, warmer than optimal temperatures may offer a trade- oflP between body size/fecundity and genera- tion time. Will they maximize reproductive effort by producing fewer, larger individuals (slow growth and a univoltine life cycle) or many, smaller individuals (fast growth and multivoltine life cycle)? These data demon- strated that Baetis may have the genetic plas- ticity necessary to respond to such tradeoffs. As temperatures approached optimality (site 2) from cooler conditions (site 1), both voltinism and body size increased within the same population. When comparing site 1 with the much warmer and very different tempera- ture regimes downstream from the reservoir, w(> found that our data support the predictions of the TF hypothesis. The largest body size and i)()pulati()n size of each species occurred downstream from the reservoir. However, site-specihc comparisons of body size and population size did not correspond as pre- dicted by the TE hypothesis. The thermal regimes at sites 2 and 3 were apparently suffi- ciently similar that temperature did not have an overriding influence on population sizes. The TE hypothesis assumes that higher fecun- dity is equivalent to larger population size. Sources of mortality at every stage of develop- ment, eggs, nymphs, and adults (Sweeney and Vannote 1982, Butler 1984, Peckarsky 1984, Gilliam et al. 1989), which may vary across sites, may interrupt the translation of higher fecundity into higher population den- sity and biomass. Numerous factors, in addi- tion to the influence of temperature on body size/fecundity, will undoubtedly influence the geographic or altitudinal variation in aquatic insect population size and stability. Toward the center, and probably over most of a species range, other sources of mortality and determinants of reproductive success should have a greater influence on aquatic insect pop- ulation size. Temperature at the edge of a species range may, however, be more limit- ing, compared to other factors, in determin- ing the extent of a species distribution. Acknowledgments J V. McArthur made helpful comments on an earlier draft of this paper. Data analysis and manuscript preparation were supported by contract DE-AC09-76SR00-819 between the U.S. Department of Energy and the Univer- sity of Georgia. Data collection was supported by a research grant to J. V. Ward, Colorado State University, from the Colorado Experi- ment Station. References Butler, M. G. 1984. Life histories of aquatic insects. Pages 24-55 in V. H. Resh and D. M. Rosenburg, eds.. The ecology of aquatic insects. Praeger Pub- hshers. New York. Clifford. H F., and H. Boerger 1974. Fecundity of mayflies (Ephemeroptera), with species reference to mayflies of a brown-water stream of Alberta, Canada. Canadian Entomologist 106: 1111-1119. 106 R B. RaderandJ. V. Ward [Volume 50 CONNELL, J H . andW P, Sousa 1983. On the evidence needed to judge ecological stability or persistence. American Naturalist 121: 789-824. Edmunds, G F 1982. Historical and life history factors in the hiogeography of mayflies. American Zoologist 22: 371-374. GiLMAM. J F , D F. Fraskk, and A M Sabat 1989. Strong effects of foraging minnows on a stream benthic community. Ecology 70: 445-452. KONDRATIEFF, B C.. AND J R Vo.SHELL, Jk. 1980. Life history and ecology of Stenonema inodesftim (Banks) (Ephemeroptera: Heptageniidae) in Vir- ginia, USA. Aciuatic Insects 2: 177-189. Peckarsky, B L. 1984. Predator-prey interactions among acjuatic insects. Pages 196-254 in V. H. Hcsh and I). M. Rosenburg, eds.. The ecology of aejuatic insects. Praeger Pui)lishers, New York. PuiM.lPSON, J 1981. Bioenergetic options and phylogeny. Pages 20-45 in C. R. Townscnd and P. Galow, eds., Physiological ecology. Blackwell Scientific Puiilishers, Sunderland, Massachusetts. Rader. R B . AND J V Ward 1988. Influence of regula- tion on environmental conditions and the macroinvertebrate community in the upper Colo- rado River. Regulated Rivers 2: 597-618. 1989. Influence of impoundments on mayfly diets, life histories, and production. Journal of the North American Benthological Society 8: 64-73. Stanley. E. H.. and R. A Short 1988. Temperature ef- fects on warmwater stream insects: a test of the thermal e(iuilibrium hypothesis. Oikos 52; 313-320. Sweeney, B W 1984. Factors influencing life-history patterns of aquatic insects. Pages 56-100 in V. H. Resh and D. M. Rosenburg, eds.. The ecology of acjuatic insects. Praeger Publishers, New York. Sweeney, B W., and R L. Vannote 1978. Size variation and the distribution of hcTuiiuetabolous acjuatic insects: two thermal ec|uilii)rium hypotheses. Sci- ence 200; 444-446. 1981. EpJunncrcUa mayflies of White Clay Creek: bioenergetic and ecological relationships among si,\ coexisting species. Ecology 62; 1353-1369. 1982. Population synchrony in mayflies: a preda- tor satiation hypothesis. Evolution 36: 810-821. Sweeney, B W , R L Vannote, and P J Dodds 1986. Effects of temperature and food ciuality on growth and development of a mayfly, Leptophlchia inter- nicdici. (Canadian Journal of Fisheries and Acjuatic Sciences 43; 12-18. Vannote, R. L., and B W Sweeney 1980. Geographic analysis of thermal ec|uilibria; a conceptual model for evaluating the effect of natural and modified thermal regimes on acjuatic insect communities. American Naturalist 115: 667-695. Ward, J V 1980. Abundance and altitudinal distribution of Ejihemerojitera in a Rockv Mountain stream. Pages 169-177 in J. F. Flannagan and K. E. Mar- shall, eds.. Advances in Ejjhemeroptera biology. Plenum Publishers, New York. 1986. Altitudinal zonation in a Rocky Mountain stream. Archiv fur Ilydrobiologie, Suj^jilement Band 74; 133-199. Ward. J V , and J A Stanford 1982. Thermal resjionse in the evolutionary ecology of acjuatic insects. An- nual Review of Entomology 27; 97-117. Received 1 May 1990 Accepted 15 June 1990 Gii-at Basil! Natur.ilist 50(2). !')')(), 1)]) 107 11,1 BLACK-TAILED PHAIKIL DOC LOPULA TIONS ONE YLAK AFTEH IKEA TMENT Wnil KODENTICIDES Antlioiu' I). A\y.i\ DaTiicl VV. I'lcsk", Kayiiioiid I,. Liiidci' AUSTHACI. — TlucH' roclciititiclc ticatiiu'iits, zinc pliosphidc \vi(h |)r(l)ail, sliycliiiinc wllli pichait, and slryclininc without prchait, were applied to black-tailed ])rairie do^ (Cijnomtis liuloviciaiius) colonies in west central South Dakota. Hesiilts were compared immediately ])()sttreatment and ibr one year alter application. Zinc phosphide was the most eHective for reducing prairie doji numbers immediately. When burrow activity levels ol prairie doys were initially reduced by 45% with strychnine only, they returned to untreated levels within ten months. When initial reductions were 95% with zinc phosijhide, however, the number of active burrows was still reduced 77% in Se|)teml)er the following year. Strychnine with prebait treatment showed initial reductions ol 83% in burrow activity. Bail consump- tion by prairie dogs was highest lor zinc iihosphide. Rodenticide-treatcd oal.s have been and arc the primary tool for control of l)lac k-tailed prairie dogs {(Jynoinys ludoviciamis) to pre- vent expansion of colonies on the plains. Widespread control programs for prairie dogs have been common on the (ireat Plains for over 100 years (Merriam 1902) and are still common practices (Schenbeck 1982). Strych- nine, first introdnced into the United States about 1847, has had varied success as a roden- ticide (Crabtree 1962). The alkaloid form on grain was recommended by the U.S. Depart- ment of Agriculture at the beginning of the century. Two characteristics that may have impeded its acceptance by rodents are its bitter taste and the noxious effect of sublethal doses; attempts to circumvent these charac- teristics have failed. Strychnine also is consid- ered hazardous to many non target species (Tietjen 1976a). Zinc phosphide was introduced as a verte- brate pest-control agent in 1943 because of strychnine shortages during World War II (Crabtree 1962). However, use of zinc phos- phide as a field rodenticide was limited imtil 1976, when it was developed specifically for black-tailed prairie dog control (Tietjen 1976a). Since then, zinc phosphide has been the only rodenticide federally approved for prairie dog control. Bioassays have shown that zinc phosphide causes no secondary poi- soning of predatory or scavenging wildlife (Crabtree 1962, Tietjen 1976a). The objc^ctives ol this study were to deter- mine (1) seasonal activity of prairie dogs and (2) short- and long-term effects of zinc phos- ])hide (with preliait), strychnine with prebait, and strychnine without prei)ait on prairie dog colonies during the one-year period following rodenticide application. Immediate effects of the three rodenticide treatments have been reported by Uresk et al. (1986). Study Ahka The study was conducted in Badlands Na- tional Park and Buffalo (iap National (Grass- land in west central South Dakota. The cli- mate is considered semiarid, with a 12-year average annual precipitation of 40 cm at the (vcdar Pass Visitors Center, Badlands National Park. Approximately 80% of the total precipitation faffs as thundershowers from April to September. Temperatures range from -5 C in January to 43 C in July, with an average animal temperature of 10 C Raymond and King (1976) described the soils on the study area as sedimentary deposits of clay, silt, gravel, and volcanic ash. Topo- graphic features consist of rugged |)imiacles, vegetated tabletop buttes, creek gullies, and grassland basins. Gently rolling grasslands in the northern portion of the study area ranged from 700 to 1,000 in in elevation. 'South Dakota Cooperative Ki-,li and WlldlilcHcscarcli Unit, Sontli Dakota Slate University, Uro()kiri«s, Sontli Dakota 57(K)() ^United States Department of Auricnltore, Forest Service, Hoeky Monntain Forest and HanKe Fxperirnent Station, Soiitii Dakota School City, South Dakota 57701. Kapirl 107 108 A. D. Apaetal. [Volume 50 The vegetation comprises a mosaic of native grasses, forbs, shrubs, and isolated trees. Dominant grasses include blue grama (Boute- loua gracilis), buffalograss (Biichloe dacty- loides), needleleaf sedge (Carex eleocharis), and western wheatgrass (Agropyron smitJiii) (Uresk 1984). Common forbs include scarlet mallow (Sphaeralcea coccinea), American vetch {Vicia americana), dogweed (Dyssodia papposa), sage {Scdvia rejlexa), and prairie sunflower (Helianthus petiolaris). The domi- nant shrub is pasture sagebrush {Artemisia frigida), and nonnative grasses include cheat- grass {Bromiis tectoriim) and Japanese chess {B.japo7iicus). Native herbivores inhabiting the Badlands region are the black-tailed prairie dog, mule deer {Odocodeus hcmioniis). Rocky Mountain bighorn sheep (Ovis canadensis), American bison (Bison bison), pronghorn {Antdocapra americana), black-tailed jackrabbit (Lepus calif ornicus), white-tailed jackrabbit (L. townsendii), and eastern cottontail {Sylvila- gtis floridanus). Small rodents include the deer mouse (Peromyscus manicidatus) and grasshopper mouse {Onychouiys Icucogas- ter). Livestock are not present in the Park, but bison graze the area all year. Cattle are allowed to graze on the National Grassland for six months during the growing season each year. Methods Eighteen sites on 15 prairie dog colonies were sampled in 1983 and 1984, with 9 desig- nated as treatments sites and 9 as controls (Uresk et al. 1986). Sites were clustered into three major areas, one for each rodenticide treatment, containing three treatment and three control sites. Zinc phosphide was ap- plied to the area within Badlands National Park because administrative restraints forbid the use of strychnine: four sites were clus- tered and paired on an approximately 600-ha prairie dog colony. The other two sites were located on prairie dog colonies northwest of the larger colony and northeast on a colony in the Buffalo Gap National Grassland. The other two treatments, strychnine with and without prebait, were randomly assignetl to the two remaining major areas on the National Grassland. All treated and control sites were randomly assigned in the clusters. The area with prebaited strychnine was located in Conata Basin, and the area treated with strychnine only was located east and south of Scenic. All treatment and control sites were on isolated towns ranging from 12 to 283 ha. Within each treatment regime, treat- ment or control designation was assigned ran- domly except where administrative restric- tions applied. The open-burrow technique used to deter- mine the effectiveness of the rodenticide treatments evaluated the number of active burrows (Tietjen and Matschke 1982). Burrow entrances in a 100 x 100-m area (1-ha) were filled (plugged) with soil to prevent egress/ ingress by prairie dogs. Forty-eight hours later the reopened burrows large enough for prairie dogs to pass through were counted. Burrow activity for pretreatment periods was recorded in June, July, and early September 1983. Posttreatment counts were taken in late September 1983 (four days after poisoning) and in June, July, August, and earlv Septem- ber 1984. Treated and untreated steam-rolled oats were obtained from the U.S. Fish and Wildlife Service (USFWS), Pocatello Idaho Supply Depot. Poisons were applied in the field, in accordance with federal label instruc- tions, when proper environmental conditions existed to insure optimum consumption of oats by prairie dogs (Tietjen 1976a, 1976b, Tietjen and Matschke 1982). Untreated oats (prebait) and the poisoned oats were applied on large areas from 3-wheel-dri\e, all-terrain cycles fitted with bait dispensers (Schenbeck 1982), and by hand with teaspoons on smaller areas. At the six sites requiring prebaiting, 4 g of high-(|uality, untreated steam-rolled oats was applied as prebait at a minimum of 95% of the burrows. Three sites were prebaited on 20 and 21 September 1983. Prebait was applied (<0.01-nr area) at the edges of prairie dog moimds. Prebaited areas were examined before poi- soned oat treatment to assiue that the prebait was consumed by prairie dogs. Three days after prebait application (22 September 1983), three sites were treated with 4 g of 2.0% active zinc jihosphide steam-rolled oats. Three addi- tional sites were treated with 8 g of 0.5% strychnine alkaloid steam-rolled oats per bur- row on 23 September 1983. The last three 1990] Praihie Dck; Con tkolwi'i 11 HoDiiNTiciDEs 109 Tablk 1. Average number of black-tailed prairie dog burrows/ha, active burrows/ha, and percent active burrows/ha (± standard error of the mciui) on untreated areas for four sampling periods in 1983 and 1984 in west central South Dakota. Sampling period Total l)urro\\s/ha N lunber activt ■/ha Percent active 1983 June' 121 ± 9 98 ± 8 81 ± 3 July' 117 ± 9 87 ± 8 74 ± 3 Early September" 113 ± 8 48 ± 5 43 ± 3 Late September ' 104 ± 13 34 ± 4 35 ± 4 1984'' Jime 103 ± 13 82 ± 12 77 ± 4 July 103 ± 14 66 ± 10 64 ± 3 August 97 ± 15 54 ± 11 55 ± 4 Earl\' September 86 ± 15 66 ± 13 75 ± 3 "n - 18 sites, n - 9 sites. sites, which were not prebaited, were treated with strychnine oats on 24 September 1983. Three days after apphcation the percentage of poisoned oats remaining on each burrow in a 1-ha grid on each treated site was estimated visually. Statistical Evaluation Analysis of covariance was used to compare each treated group (cluster) of sites with its respective control group. Applications of re- peated measures were examined but required constant response through time — no interac- tion between time and treatment. These data did not show a constant response through time and had interactions; therefore, we used covariance adjustments. Pretreatment obser- vations were used as covariates. Effect of ro- denticide treatment for each time point was estimated as the covariance-adjusted differ- ence between treated and control sites for each rodenticide. After obtaining an overall rejection of the hypothesis of no treatment effect, contrasts for each rodenticide treat- ment were evaluated for significance based on a variance estimated only from the sites in each cluster (because variance was heteroge- neous among clusters). If the correlation between pretreatment and posttreatment observations was not significant (a < .20), then the change was estimated as posttreat- ment minus pretreatment observation (re- peated measures). This analysis uses the interaction between time and treatment as the indicator of a significant change due to treatment (Green 1979). Rodenticides were compared by forming pairwise contrasts of the contrasts obtained for the individual ro denticide treatments. Randomization proce- dures (Edgington 1980, Romesburg 1981) based on 10,000 random permutations of the data pairs among treatment groups were used to estimate statistical significance of the vari- ous contrasts. Because omission of any effect due to poi- soning was considered more serious than the potential incorrect declaration of a significant treatment effect, Type II error protection was produced by testing each contrast individu- ally. However, some Type I error protection was afforded by testing individual contrasts only after first observing a significant (P = . 10) overall test of treatment differences using analysis of covariance (Carmer and Swanson 1973). Individual contrasts were considered biologically significant at P = .20. Although admittedly unconventional for the number of sites available for study, this significance criterion produces a power (probability of de- tecting a true difference) of approximately 0. 80 for a contrast twice as large as its standard error. This was considered a reasonable com- bination of Type I and Type II error protec- tion for this study (Carmer 1976). Results Prairie dog burrow activity declined during the summer months both years (Table 1). In June 1983 the number of active burrows was high (81%) and decreased steadily until late September, when 35% of the burrows were active. In June 1984 activity of prairie dog burrows was high (77%) and decreased through July and August, but increased to 75% in September. no A. D. Apa et al. [Volume 50 POISON POST-POISON <120 n I 100 - O 80 H ac £ 60 H m 40 g 20 H < S CONTROL D TREATED SEPT JUNE JULY SEPT 1983 1984 Fig. 1. Seasonal comparisons of active black-tailed prairie dog burrows on zinc phosphide-treated and control sites from initial treatment in September 1983 through September 1984. Means followed by the same letter by date are not significant at a = .20 after F-protection at a = .10 using analysis of covariance. (Data on September 1983 for initial poison are adapted from Uresk et al. 1986.) In September 1983, immediately after treatment with zinc phosphide, the number of active prairie dog burrows was reduced 95% from the number on the 76 control sites (F = .017, Fig. 1). The reduction in the number of active burrows was maintained (96%) in June 1984 (P = .002). Reductions of active burrows in July and September were 92% and 77%, respectively (F = .006 and .014, respectively). Treatment with strychnine only immedi- ately reduced active burrows by 45% (F = .164, Fig. 2). In June 1984 active burrows on the treated sites remained 45% below the strychnine control sites (F = .177). By July, however, the number of active burrows on the treated sites was not different from the control sites (F = .706). The treated and control sites also showed similar burrow activity levels in September (F = .637). Treatment with prebaited strychnine im- mediately reduced the number of active bur- rows by 83% (F - .035, Fig. 3). Burrow activ- ity remained 85% below controls (F = .019) in June 1984. This reduced level of prairie dog activity compared with controls reached 99% in July and 95% in September 1984 (F = .083 and .057, respectively). A comparison of the effectiveness of roden- ticide treatments at initial poisoning of prairie dogs in 1983 showed that number of active burrows was reduced more with zinc phos- phide than with strychnine alone (F = .034, Table 2). Burrow counts in June 1984 showed that towns with zinc phosphide treatment had fewer (F = .006) active burrows than those with strychnine only treatment. Similar re- sults continued through Julv (F ^- .035) and September (F = .039) 1984. ' Zinc phosphide had a greater initial effect than prebaited strychnine in reducing num- bers of active burrows (F = .075, Table 2). There were no differences between the effects of the two rodenticides by 1984, however (F --- .20). When the two strychnine treatments were compared, reduction in active burrows was not different (F .391) in September 1983 (Table 2). Strychnine compared with pre- baited strychnine treatment in June 1984 showed a significant difference of 60 more 1990] Phaihie Doc; Control vvnii Kodenticides HI <120 ^ ilOO H o 80 H QC oc 60 H m 40 H I 20 H < POISON POST-POISON 0 CONTROL D TREATED SEPT JUNE JULY SEPT 1983 1984 Fig. 2. Seasonal comparisons of active black-tailed prairie dog bnrrows on strychnine-only-treated and control sites from initial treatment in September 1983 through September 1984. Means followed by the same letter by date are not significant at a = .20 after F-protection at a = .10 using analysis of covariance. (Data on September 1983 for initial poison are adapted from Uresk et al. 1986.) POISON POST-POISON <120 n X Jo 100 - O 80 - DC g^ 60 - g uj 40 - a > ^ 20- b < ^ • National Agrieultural Pesti- cide Impact Assessment Program (NAPIAP) and Nebraska National Forest (Interagency Agreement IAG-56). LiTERATi'RE Cited B.MLKV. y 1926. A biological .sur\ey of North Dakota. United States Department of Agriculture, North American Faima, No. 49. C.JiRMER, S G 1976. Optimal significance levels for appli- cation of the least significant difference in crop performance trials. Crop Science 16: 9.5-99. Carmer, S G . and M R Swan.son. 1973. An evaluation of ten pairwise multiple comparison procedures by Monte Carlo methods. Journal of American Statis- tics Association 68; 66-74. Crabtree. D. G 1962. Review of current vertebrate pes- ticides. In Vertebrate Pest Control Conference Proceedings, California Vertebrate Pest Control Tech., Sacramento. .390pp. Edgington. E. S 1980. Randomization tests. Marcel Dekker, Inc., New York. 287 pp. Green, R H 1979. Sampling design and statistical meth- ods for environmental biologists. Sec. 4.1. John Wiley and Sons, New York. 257 pp. Knowles, C. J 1982. Habitat affinity, populations, and control of black-tailed prairie dogs on the Charles M. Russell National Wildlife Refuge. Unpub- lished doctoral dissertation. University of Mon- tana, Missoula. 171 pp. Koford, C. B 1958. Prairie dogs, whitefaces, and blue grama. Wildlife Monograph 3. 78 pp. Merriam, D 1902. The prairie dog of the Great Plains. Pages 257-270 in Yearbook of the United States Department of Agriculture. Government Printing Office, Washington, D.C. O'Meuja, M, E . F. L. Knopf, and J C Lewis. 1982. Some consequences of competition between prairie dogs and beef cattle. Journal of Range Management 35: 580-585. Raymond, W H . and R. U King 1976. Cieologic map of the Badlands National Monument and vicinity, west-central South Dakota. United States Cleolog- ical Survey. Map 1-934. 1U)MESIU'RG, C 198L Randomization test. Resource Evaluation Newsletter. Pages 1-3. Technical Article 1. USDI-BLM, Federal Center, Denver, Colorado. ScuENBEGK. G L. 1982. Management of black-tailed prairie dogs on the National Grasslands. Pages 207-217 in R. M. Timm and R. J. Johnson, eds., Fifth Great Plains Wildlife 13amage Control Workshop Proceedings, 13-15 October 1981, Lhiiversity of Nebraska, Lincoln. TlETjEN. H P. 1976a. Zinc phosphide — its development as a control agent for black-tailed prairie dogs. United States Department of Interior, Fish and Wildlife Service. Special Science Report Wildlife No. 195. 14 pp. 1976b. Zinc phosphide — a control agent for black- tailed prairie dogs. United States Department of Interior, Fish and Wildlife Service. Wildlife Leaflet No. 509. 4 pp. TiETjEN, H P , and G. H. Matschke. 1982. Aerial pre- baiting for management of prairie dogs with zinc phosphide. Journal of Wildlife Management 46: 1108-1112. Uresk. D W 1984. Black-tailed prairie dog food habits and forage relationships in western South Dakota. Journal of Range Management 37: 325-329. Ure,sk, D W., J G M,\cCracken. and A. J Bjugstad. 1982. Prairie dog density and cattle grazing rela- tionships. Pages 199-201 in Fifth Great Plains Wildlife Damage Control Workshop Proceedings, 13-15 October 1981, University of Nebraska, Lincoln. Uresk, D W.. and A. J Bjugstad. 1983. Prairie dogs as ecosystem regulators on the Northern High Plains. Pages 91-94 in Seventh North American Prairie Conference Proceedings, 4-6 August 1980, Southwest Missouri State University, Springfield. Uresk, D W , R M King, A. D Apa, and R. L Linder. 1986. Efficacy of zinc phosphide and strychnine for black- tailed prairie dog control. Journal of Range Management 39: 298-299. Received 5 May 1990 Accepted 15 June 1990 Great Basin NaUualist r)0(2), KWO. pp. 115 120 EFFECTS OF BURNING AND CLIPPING ON FIVE BUNCHGRASSES IN EASTERN OREGON Carlton M. Hiittoii , Cux H. McPhcrsoii' '. and Forrest A. Sneva' Abstract. — Response of five perennial Ininehgrasses following clipping and binning was evaluated in eastern Oregon. Burned plants were compared with clipped plants on several dates from spring to fall with respect to mortality and change in basal area. Basal area was generalK reduced for one year but did not change the second year after defoliation. Treatments rareh' affected yield. Burning in May was luost detrimental, reducing basal area of all species. Fall clipping was least harmful, producing little or no change in basal area. Plant mortality was significant only for burned Thurber needlegrass {Stipa thurheriana). Bunchgrasses comprise a major proportion of herbaceous vegetation in the Great Basin; yet httle information is available regarding their response to defohation. Furthermore, reports of bunchgrass response to fire and/or chpping are variable within and between spe- cies. Differences in defoliation effects are probably largely due to differences in growth form, phenology, season of treatment, and favorabilitv of studv vears (Wright and Bailev 1982). Bluebunch wheatgrass {Agropyron spica- tum), Idaho fescue (Festiica idahoensis), junegrass (Koeleria cristata), bottlebrush squirreltail (Sitanion Jiystrix), and Thurber needlegrass {Stipa thurberiana}are dominant herbaceous species in eastern Oregon (bo- tanical nomenclature follows Hitchcock and Cronquist [1973]). Published data concerning response of these species parallel that of most bunchgrasses in scarcity and variability. Ap- propriate management of these and other bunchgrass communities requires knowledge of their response to defoliation. The objective of this study was to evaluate the effects of fire on these five perennial bunchgrasses on east- ern Oregon rangeland. Methods The study area is located on Squaw Butte Experiment Station, 65 km west of Bin-ns, Oregon. Elevation is 1,370 m, and average annual precipitation is 29.4 cm. Precipitation during the study was 24.6 cm in 1976, 27.5 cm in 1977, and 28.0 cm in 1978. Soil on the study area is a fine-loamy, mixed, frigid Aridic Durixeroll. A 1-ha area was fenced to exclude livestock, and 90 plants each of five species were marked with wire stakes. Ten plants received no defo- liation treatment and were treated as controls for mortality assessment. Ten randomly se- lected plants were burned with an individual plant burner (Britton and Wright 1979) on each of three dates: 15 May, 15 June, and 11 November 1976. Time-temperature curves peaked at 200 C at 30 sec. Ten randomly selected plants were clipped to 1-cm stubble height on each of the three dates. These clipped plants served as controls for burning treatments to evaluate the effects of fire sepa- rately from the effects of aboveground bio- mass removal. An additional 10 plants were clipped on 27 August and 12 October to com- pare effects of defoliation during late summer and early fall with other defoliation dates. Treatment effects were measured as per- centage changes in basal area and yield. After treatment each plant was photographed to determine initial basal area (cm"). A wire grid (2.5 X 2.5 cm) was placed atop each plant base before photographing to provide a permanent record of basal area. Aboveground biomass Department ol Range and W'ildlite Management, Texas Tech Uni\eisit\, Lnhbutk, Texas 79409. "Forest-Watershed Sciences Program, School of Renewable Natural Resources, University of Arizona, Tucson. Arizona 8.5721. Squaw Butte Experiment Station, Oregon State University, Burns, Oregon 97720. Address all correspondence to Guy R. McPherson, Forest-Watershed Sciences Program. School of Renewable Natural Resources, LI niversitv of Arizona, Tucson, Arizona 85721, (602) 621-5389, FAX: (602) 621-7196. 115 116 C. M. BmrroNETAL. [Volume 50 O O UJ < LU < -J < CO < Agropyron spicatum n d Festuca idahoensis Sitanion hystrix E D t Stipa thurberiana b L Koeleria cristata L ■ Burned U Clipped b 1 b L May June Aug Oct Nov TREATMENT DATE Fi^. 1. Mean reduction in basal area f)f' five hunchgrasses one year after hurniniz; or clipping in eastern Oregon. Plants were treated on various dates in 1976; basal area was measured inunediately following treatment and in JuK of 1977 and 1978. Basal area reduction two years after burning rarely differed (P > .05) from 0%. W'itliin a species, histograms marked with the same letter are not significantly different (P > .05). 1990] DKFOI.I ATION Fl' !■ KCTS ()\ Bl'NCllCRASSFS 117 was clipped in late Jul\ one and two growing seasons after treatment, and each plant was rephotographed at that time. Percentage change in basal area was calc-nlated, and anaK- sis of co\ ariance (adjnsted to initial plant basal area) was nsed to test for differences in basal area among treatments one \ ear and two years posttreatment. Yield was expressed as grams per scjiiare decimeter of basal area to adjust for different plant sizes. Plants with no live tillers two years after treatment were assumed to be dead. Means were separated using Fisher's protected LSD test (F .05). Results Basal Area Basal area generally declined the first year after treatment (Fig. 1). Bluebunch wheat- grass basal area decreased 45% one year after burning and 22% after clipping. May burning was especially damaging to basal area ( — 69%); by contrast, November clipping had no effect on basal area. Other treatment-date combina- tions were intermediate (mean = —32%) and did not differ. Effects of August and October clipping treatments (mean = —13%), were in- termediate between effects of May and June clipping (mean = -28%) and November clip- ping (no change). Idaho fescue basal area was affected by date of defoliation but did not differ between treat- ments. Defoliation in May and June reduced basal area by an average of 48%. Other treat- ment-date combinations did not significantly (F > .05) reduce basal area. Basal area of junegrass was reduced 42% by burning in May. No other treatment-date combinations differed significantly from 0%. Squirreltail basal area was reduced 71% by burning in May, which was a greater reduc- tion than other ti-eatment-date combinations (mean ^ -24%). Basal area change of plants clipped in August, October, and November did not differ significantly from 0%, indicating that squirreltail was resistant to late-season clipping. Needlegrass basal area was reduced 93% by May and June burning, largely due to high mortality associated with early burns (May mortality = 50%, June mortality ^ 70%). August and November clipping treat- ments did not significantly affect basal area. Change in basal area did not differ between other defoliation treatments (mean - —33%), which were intermediate between early- season burning and late-season (August, November) clipping treatments. Subse(juent (second-year) decreases in basal area of all species were slight and gener- alK not significant (F > .05) and will not be discussed. Exceptions were (1) junegrass, which decreased the second year following burning (mean = —21% from first year to second) or clipping (mean = —18%) on all dates; (2) scjuirreltail, which decreased follow- ing burning in May (53%) and June (42%); and (3) needlegrass, which decreased follow- ing clipping in August (-27%). Yield Yield (adjusted for basal area) of bluebunch wheatgrass was not affected by method or date of defoliation. First-year yield was 4.2 g/dm ; second-year yield was 7.3 g/dm (Fig. 2). First-year yield of Idaho fescue was re- duced by date of defoliation, but clipping and burning did not differentially influence yield. Plants defoliated in May produced 6. 1 g/dm" dry matter, compared with a mean yield of 3. 1 g/dm" for all other dates. Yields from sup- plementary August and October clipping treatments did not differ from June and November yields. Second-year yield (mean ^ 8.0 g/dm") was not affected by date or method of defoliation. First-year yield of junegrass was affected by defoliation date in a manner similar to that of Idaho fescue. Defoliation in May resulted in 2.7 g/dm" dry matter production, com- pared with 1.5 g/dm" following defoliation in November. Yield following clipping in Octo- ber was 1.8 g/dm", which was lower than yield following May defoliations. Yields from June and August treatments were intermediate (mean =- 2.3 g/dm") and did not differ from yields on other dates. Second-year yields were unaffected by date or method of defolia- tion (mean = 7.0 g/dm"). Squirreltail yield was lower one year after burning (1.6 g/dm") than after clipping (5.8 g/dm") in May. A significant treatment x date interaction precluded comparisons across all dates. Yields following burning and clipping did not differ on other dates, and date of defo- liation did not affect yield of clipped plants (mean = 4.7 g/dm"). Second-year yield was 118 C. M. BRITTON ET AL. [Volume 50 CNJ E 3 z o O o o Q. Agropyron spicatum / a a 1 5 ■ 1 0 ■ 5 ■ a a 36 ■ 15 " -^ 10 ■ 5 ■ 1 5 10 ■ 5 ■ bcB Festuca idahoensis Koeleria cristata a ¥ a ab ab ab Sit an ion hystrix a be ab ab Stipa thurberiana ab n / u be be be a i ab ab May June Aug Oct Nov TREATMENT DATE 1 yr after burning 2 yr after burning □ 1 yr after clipping 2 yr after clipping Fig. 2. Mean production oi'live hundignisscs one and two years after l)urninK or eliiiiiinu in eastern Orei^on. Plants were treated on various dates in 1976; abovcground hioniass was clipped in Jnl\ oi 1977 antl 197S. Within a species, histograms marked with the saini' letter are not signiheantK dilVerent (P > .05). 1990] DKFOLIAHON EFKIX ISON Bl'NCIICIiASSKS 119 not affected l)> date or method ol defoliation (mean 12.3 ^/dm"). First-year yield ot needle<2;ra.s.s was aileet- ed by date and method ol defoliation; the date X treatment interaction was also siunilicant. Yield following clippiiii!; in Ma\ was hi,e;lu>r than any other date-treatment combination (6.4 g/dm"). Mean yield following November burning was higher than after June burning (2.6 vs. 0.6 g/dm"), largely due to higher sur- vival of" fall-burned plants (90 vs. 30%). First- year yields of other date-treatment combina- tions were intermediate (1.9-2.5 g/dm") and did not differ. Needlegrass was the only spe- cies in which second-year yield was affected by treatment. Mean second-year yields of burned and clipped plants were 2.0 and 4.2 g/dm", respectively. Increased mortality of burned plants (mean = 57%) accounted for the diffeience. Second-year yield was not affected by date of defoliation. Needlegrass was the only species in which significant mor- tality occurred. Burning in May, June, and November resulted in 50%, 70%, and 10% mortality, respectively. No control or clipped plants died. Discussion Bluebunch wheatgrass was relatively toler- ant to all treatments except May burning. Basal area declined more following burn- ing than clipping, but yield responses were similar between treatments. Similarly, Uresk et al. (1976) reported decreased basal area and increased yield one year after a wildfire in eastern Washington. Conrad and Poulton (1966) found a 29% reduction in basal area one year after a wildfire in the same area. Wright (1985) summarized the literature on grass re- sponse to fire in sagebrush-grass communities and concluded that bluebunch wheatgrass is slightly affected by burning, with yield re- turning to preburn levels in one to three years. Idaho fescue appeared to be less suscepti- ble to defoliation than has been reported pre- viously. Basal area and mortality were not affected by late summer or fall defoliation. Furthermore, burning and clipping had simi- lar effects on plants. Wright et al. (1979) sum- marized studies bv Pechanec and Stewart (1944), Blaisdell(1953), Countryman and Cor- nelius (1957), Conrad and Poulton (1966), and llarniss and Murra\ (1973), stating that "the majority of evidence indicates that Idaho fes- cue is severely damaged regardless of when or where it is burned." However, Wright (1971) lound increased resistance to burn damage from late Jul) through late September, and attributed the altered resistance to low energy reserves and high respiration demands during late summer. Daubenmire (1970, 1987) re- ported mixed results for wildfires in eastern Washington. Wright and Klemmedson (1965) reported minimal damage to Idaho fescue af- ter late-summer and fall fires; data from this study indicate that late-season defoliation may not l)e harmful at all. Higher yields of plants defoliated in May are attributable to (1) de- creased basal area with no decrease in yield per plant, and (2) decreased growing period for plants treated later in the growing season. Plants treated after 15 May had begun growth at the time of defoliation. Second-year yield was not affected by date or method of defolia- tion, further indicating that the additional growing period associated with early defolia- tion produced the observed differences in ffrst-year yield. Britton et al. (1983) found that first-year yields of Idaho fescue following August burning were 25% lower than yields following burning in October. Junegrass was tolerant to all treatments except May burning. Basal area decreased in the second year after defoliation, although production increased. Wright et al. (1979) attributed junegrass's resistance to the rela- tively small size of typical junegrass plants. Relatively high yields following early defolia- tion and low yields following late defoliation support the hypothesis that yield is influenced primarily by length of growing season follow- ing defoliation. Further support was provided by second-year yields that were unaffected by date or method of defoliation. Squirreltail was moderately affected by de- foliation, with significant basal area decreases in all treatments except late-season clipping. May burning was most detrimental to basal area. Wright (1971) reported similar results in southern Idaho and attributed squirreltails late-season tolerance to clipping to summer dormancy (Wright 1967). Young and Miller (1985) found no change in basal area, but increases in above- and belowground yield, following July burning. Scjuirreltail's resis- tance to fire derives from plant growth form 120 C. M. Brittonetal. [Volume 50 (coarse stems with little leafy material) and small size (which precludes development of dead centers) (Wright and Bailey 1982, Daubenmire 1987). Since basal area de- creased by an average of 47% the second year after early-season burns, increased abun- dance on burned areas (Blaisdell 1953, Barney and Frischknecht 1974) probably results from squirreltail's ability to survive and subse- quently invade sites previously occupied by other perennial plants, and not from in- creased size of individual plants. Lack of dif- ferences in yield two years after defoliation indicated that response of squirreltail, along with bluebunch wheatgrass, Idaho fescue, and junegrass, varied independently of the method of defoliation. Needlegrass was severely damaged by all defoliation treatments. Burning was particu- larly harmful, increasing mortality and reduc- ing mean basal area and yield. Uresk et al. (1976) reported a 53% reduction in basal area following an August wildfire; recovery was not complete three years later (Uresk et al. 1980). Wright et al. (1979) concluded that Thurber needlegrass is probably the least resistant needlegrass. Early-season clipping was more damaging to needlegrass than late-season clipping in this study. Plants responded simi- larly to May, June, and August clipping and November burning, but later clipping treat- ments had no measurable effect. Acknowledgments This study is a contribution of the College of Agricultural Sciences, Texas Tech Univer- sity, Publication No. T-508, and Oregon Agricultural Experiment Station, Publication No. 9057. Literature Cited Barnky, M. a., and N C. Frischknecht. 1974. Vegeta- tion changes following fire in the pinyon-juniper type of west-central Utah. Journal of Range Man- agement 27: 91-96. Blai.sdell, J P. 195.3. Ecological effects of planned burn- ing of sagebrush-grass range on the upper Snake River Plains. USDA Technical Bulletin 1075. BRirroN. C. M., R C, Clark, and F A Snk\a 1983. Effects of soil moisture on burned and clijipcd Idaho fescue. Journal of Range Management 36; 708-710. Britton, C. M., and H a Wright 1979. A portable burner for evaluating effects offire on plants. Jour- nal of Range Management 32: 475-476. Conrad. E C , and C E Poulton 1966. Effect of a wildfire in Idaho fescue and bluebunch wheat- grass. Journal of Range Management 19: 138-141. Countryman, C M., and D R Cornelius 1957. Some effects offire on a perennial range type. Journal of Range Management 10: 39-41. Daubenmire, R 1970. Steppe vegetation of Washington. Washington State Agricultural E.xperinient Sta- tion, Technical Bulletin 62. 1987. Some effects ot tire on perennial grasses in the steppe of eastern Washington. Phvtocoenolo- gia 15: 145-148. Harniss, R. O., and R B Murr.\y 1973. Thirty years of vegetal change following burning of sagebrush- grass range. Journal of Range Management 26: 322-325. Hitchcock, C L., and A Cronquist 1973. Flora of the Pacific Northwest. University of Washington Press, Seattle. Pechanec, J F., andG. Stewart 1944. Sagebrush burn- ing— good and bad. USDA, Farmer's Bulletin 1948. Uresk, D W , J F Cline, and W H Rickard 1976. Impact ot wildtire on three perennial grasses of south-central Washington. Journal of Range Man- agement .33: 309-310. Uresk, D. W , W K Rickard, and J F Cline 1980. Perennial grasses and their response to a wildfire in south-central Washington. Journal of Range Management .33: 111-114. Wright, H A 1967. Why squirreltail is more tolerant to burning than needle-and-thread. Jovirnal of Range Management 24: 277-284. 1971. Contrasting responses of squirreltail and needle-and-thread to herbage removal. Journal of Range Management 20: 398-400. 1985. Etlects of fire on grasses and forbs in sagebrush-grass communities. Pages 12—21 in K. Sanders and J. Durham eds., Rangeland fire effects. Proceedings of a symposiiun held 27-29 November 1984, Boise, Idaho. USDI, Bureau of Land Management. Wright, H. A , and A W Bailey 1982. Fire ecology: Lhiited States and southern Canada. John Wiley & Sons, New York. Wright, H. A., and J. O Klemmedson 1965. Effects of fire on bunchgrasses of the sagebrush-grass region in southern Idaho. Ecology 46: 680-688. Wright, H A , L F Neuenschwander, andC M Brit- ton. 1979. The role and use offire in sagebrush- grass and pinyon-juniper plant communities. USDA, Forest Service General Technical Report lNT-58. Younc;, R, P , and R. F Miller. 1985. Response oiSitun- ion hi/strix (Nntt.) J. G. to prescribed burning. American Midland Naturalist 113: 182-187. Received 29 J a una n/ 1 990 Accepted 22 March 1990 Great Basin Naturalist 50(2), 1990, pp 121-134 FOLIAGE BIOMASS AND COVER RELATIONSHIPS BETWEEN TREE- AND SHRUB-DOMINATED COMMUNITIES IN PINYON-JUNIPER WOODLANDS H. J.Tau.sth'andF. T. Ttu'ller Abstract. — Woodlands dominated bv singk'leuf pinyon (Piniis moiu)))hyll(i Torr. and Fn-ni.) and I'tali jiniiper (Juiiiperits osteospenna [Ton.] Little) cover e.xten,sive area.s in the Creat Basin and Southwest. Both species are aggressive and can nearly eliminate the previous shrub-dominated community. Successional pathways from shrub- dominated conununities before tree establishment to the tree-dominated communities that follow are known onl\' for a few specific sites. How site growing conditions affect successional patterns needs further study. We compared the relationship of foliage biomass and percentage of cover between paired shrub-dominated and tree-dominated plots o\er several sites. Sites studied are from dif!^erent elevation and topographic conditions on one mountain range. Foliage biomass in shrub-dominated plots had about a three-to-one variation over the range of site conditions sampled. Tree-dominated plots varied In about two to one. Cover in shrub-dominated plots had a four-to-one variation; cover in the tree-dominated plots varied by about two to one. Total foliage biomass in both tree- and shrub-dominated plots correlated best with the site index of height at 200 years of age. Variation in percentage of cover in both tree- and shrub-dominated plots correlated best with elevation. Foliage biomass variation in shrub-dominated plots was proportional to the variation in the paired tree-dominated plots. A similar proportional relationship was present for percentage of cover between paired tree- and shrub-dominated plots. Foliage biomass was more sensitive to topo- graphic differences than to cover. Variation in plant species sampled in the shrub-dominated plots correlated with total foliage biomass of the same plots. Species sampled also correlated with pinyon height at 200 years of age and total foliage biomass in the paired tree-dominated plots. Singleleaf pinyon (Pmus monopJiylla Torr. and Frem.) and Utah juniper {Junipcriis os- teospenna [Torr.] Little) woodlands cover more than 72,000 km"" (18 million acres) in the Great Basin, coverage greater than it was before European settlement (Tausch et al. 1981). Both species are successionally aggressive and, once established, can nearly eliminate the understory. Loss of forage and increased soil erosion can result from domi- nance by the trees (Doughty 1987). Estab- lished woodlands provide wood products, pine nuts, and habitat for many wildlife species. Successional pathways from shrub-domi- nated communities before tree establishment to the resulting tree-dominated communities that follow are known from only a few specific sites (Barney and Frischknecht 1974, Tausch et al. 1981, Young and Evans 1981, Everett and Ward 1984, Everett 1987). Variability in both tree- and shrub-dominated communities (Ronco 1987) complicates extrapolation of these results to sites of different growing con- ditions. Comparisons of biomass and cover relationships between shrub- and tree-domi- nated communities on the same sites are needed for more locations. Woodlands have a higher percentage of cover at higher than at lower elevations and on north than on south aspects (West et al. 1978, Tueller et al. 1979). Both tree- and shrub- dominated communities appear to show an increase in cover, and in biomass, on the bet- ter sites. The potential three-dimensional form of these relationships is illustrated in Figure 1. Orientation of the X, Y, and Z axes in Figure 1 is for clarity of presentation of the three-dimensional representation. The vertical X axis represents improving site conditions. Increasing cover or biomass in tree-dominated communities is represented by the Y axis. The Z axis represents increasing cover or biomass in shrub-dominated commu- nities. The line a-e (Fig. 1) represents the relationship between site and shrub cover or biomass. The line a'-e' represents the same relationship with site for biomass or cover of tree-dominated communities. If the relation- Intermountain Research Station, Forest Service, U.S. Department of Agriculture, Reno, Nevada 89512. Department of Range, Wildlife, and Forestry, University of Nevada- Reno, Reno, Nevada 89512. 121 122 R, J. Tausch and p. T. Tueller [Volume 50 \ SHRUB \ (Z) TREE (Y) Fig. 1. Three-dimensional representation of h> potliesized relationships hctween site (luality (.\) and eover or biomass in tree- (Y) or shrnh-doniinated (Z) eonimunities in pinyon-jnniper woodlands. The lini's a-e, a'-e', and a"-e" represent hypothesized relationships among the respeetive axes. ships of the X-Y and Y-Z phmes hold true, a relationship also exists on the Y-Z plane. This is a proportional relationship between the quantities in the shrub- and tree-dominated commimities represented by the a"-e" line. The a-e-e'-a' plane (Fig. 1) represents the faniih of sueeessional path\va\ s for these eom- numities for the site eonditions represented. Succession in these woodlands without distur- bance proceeds from shrub to tree domination 1990] FOLI ACK HlOMASS IN FlNYON-Jl'Ml'KH W^OODLANDS 123 (Tiiuschetal. 1981). Dotted lines a-a' througli e-e' estimate specific pathways for eacli site class. These pathways are drawn linearly only for visibility. They usually follow various types of curvilinear patterns (Tausch et al. 1981). This study investigated the hypothesized three-dimensional relationship between cover or biomass of tree- and shrub-dominated commimities and site. The X-Y, X-Z, and Y-Z planes in Figure 1 represent these rela- tionships. Analyses used the total foliage biomass and total percentage of cover of tree- dominated and shrub-dominated communi- ties of several sites on one mountain range. Sampled sites cover a range of elevational and topographic conditions. Percentage of total vegetal cover has broad use in many other studies in these communi- ties. Total foliage biomass (which can be di- rectly related to leaf area) was included be- cause it is a community dimension that reaches an equilibrium level in many forest types (Moller 1947, Marks and Borman 1972, Long and Turner 1975, Long and Smith 1984). More mesic sites than drier sites sup- port higher ecjuilibrium biomass (Waring et al. 1978). Equilibrium leaf biomass levels can be directly related to the hydrologic environ- ment (Nemani and Running 1989). Other studies have also shown equilibrium levels of leaf biomass (or area) in relationship to site moisture conditions (Whittaker and Niering 1975, Grier and Running 1977) and nutrient stress (Waring et al. 1978). Only the end- points of the potential sere on each site were sampled to increase the number of sites avail- able. Methods Data Collection This study used six sites on the Sweet- water Mountains, Nevada and California (Table 1). We sampled a tree-dominated and a shrub-dominated plot at each site. The tree- dominated plots were fully stocked or fully tree-occupied as defined by Meeuwig and Cooper (1981). Shrub-dominated plots did not have trees larger than seedlings. These seedlings were less than 3 dm tall. Tree- and shrub-dominated plots were paired on each site on the same slope, aspect, and elevation. Plots were as close as physically possible while Iabi.k 1. Aspect, slope, and elevation (m) for the six sample sites on the Sweetwater Mountains, Nevada and (California. Plot identiOeations for tree-dominated plots Irom Meeiiwit;; (1979) are in parentheses next to the site numhcr. Aspect Slope Elevation Site (degrees) (degrees) (m) 1 81 3 2.120 2 75 2 2,0.30 3 90 4 2,280 4(S1) 80 3 2,210 5(S3) 120 9 2,300 6(S4) 345 20 2,020 still meeting the criteria for tree or shrub dominance. Tree plot data. — Tree data for tree-domi- nated plots for three sites (4-6, Table 1) are from Meeuwig (1979) and Meeuwig and Budy (1979). We sampled additional tree-domi- nated plots on sites 1-3 (Table 1) to extend the elevational and topographic range of the data. All tree-dominated plots had only pinyon, ex- cept site 6, which had some juniper. Sites 2, 1, 4, and 5 represent a transect up the east side on the main alluvial fan and mountain slope of the Sweetwater Mountains. The sites cover the width of the woodland belt at about 100- m -elevation intervals. Site 3 is on the flat top of a foothill away from the main mountain mass, site 6 on a north-facing slope in a narrow canyon. Tree-dominated plots for sites 1, 2, and 3 (Table 1) were 20 x 50 m in size (0. 1 ha). We measiued all trees in each plot for average crown diameter, tree height, and basal diame- ter about 15 cm above the ground surface. Where multiple trunks were present, we indi- vidually measured each trunk and deter- mined a geometric average basal diameter (Meeuwig and Budy 1979). Tree foliage biomass and trunk cross sections were col- lected from a random sample of 12-14 trees in each plot (Tausch and Tueller 1988, 1989). These trees were aged by ring counts on two radii of their cross sections. Tree-dominated plots for sites 4-6 from Meeuwig (1979) and Meeuwig and Budy (1979) were 30 x 30 m in size. All trees in each plot were measured using the methods de- scribed above and harvested. A random sam- ple of the harvested trees was weighed to determine total wet and dry biomass for bole, bark, branch, twig, and foliage. Multi- 124 R J. Tausch and P T. Tueller [Volume 50 pie regression techniques were used to derive the total dry biomass values for each part and the total of the remaining trees on each plot from their measurements. Meeuwig and Budy (1979) aged all trees by ring counts. We extrapolated their leaf biomass data to a 0. 1-ha plot size. As a part of this study, we collected addi- tional tree foliage biomass data from a random sample of trees adjacent to the plots from Meeuwig (1979) and Meeuwig and Budy (1979). These data were collected by the same techniques used for the tree-dominated plots on sites 1-3. Analysis results from these trees were used as an independent test of the fo- liage biomass predictions (Tausch and Tueller 1988). Shrub plot data. — Suitable shrub-domi- nated areas without mature trees varied in size between sites. Shrub-dominated plots on sites 1, 2, and 4 were 20 x 50 m (0.1 ha) in size. The largest shrub-dominated area on site 3 permitted a 15 x 30-m plot. Adjacent shrub-dominated areas of the same environ- mental conditions were not present for sites 5 and 6. A strong recovery by the understory was present in the plot areas originally cleared by Meeuwig (1979) and Meeuwig and Budy (1979) seven years earlier. Shrub-dominated plots 20 X 20 m in size were centered in their former tree plots. We used five transects to sample plant spe- cies data on all shrub-dominated plots except site 3. These transects were 20 m long and randomly located perpendicular to the plot axis. Each transect contained 10 contiguous 1 X 2-m microplots, for a total of 50 micro- plots. Site 3 was sampled with seven ran- domly located transects 14 m long. Each tran- sect was divided into seven 1 X 2-m micro- plots, for a total of 49 microplots. Although the overall plot size varied, the mnnbcr of micro- plots sampled was ec^uivalent for all the shrub- dominated plots. The same techniques used in the 20 x 50-m shrub-dominated plots were used to collect understory data in the tree-dominated plots for sites 1, 2, and 3. Understory data were not available for the tree-dominated plots from Meeuwig (1979) and Meeuwig and Budv (1979). We measured three crown dimensions on all shrub and perennial grass species in each microplot: (1) longest crown diameter, (2) diameter perpendicular to the longest, and (3) height of the foliage-bearing portion of the crown. A random selection of the sampled microplots was used to collect foliage biomass for the more common measured species. Fo- liage biomass was collected from 24 random individuals of the dominant and co-dominant shrubs and 12 random individuals of the sub- dominant species in each plot. We collected foliage biomass of infrequently occurring spe- cies on both tree- and shrub-dominated plots whenever they were present in any micro- plot. All measured species in the understory samples of the tree-dominated plots, except the dominant shrubs, were sampled when- ever present in a microplot. We estimated the foliage biomass of forb and annual grass species in each microplot using the reference unit method (Andrew et ai. 1979, 1981, Kirmse and Norton 1985, Cabaral and West 1986, Carpenter and West 1987). Actual foliage biomass of reference unit species was also collected in the random sam- ple of microplots. This collected foliage bio- mass data provided a double sampling correc- tion on the reference unit estimates. Foliage biomass of infrequently occurring forb species was collected whenever such species were present in any microplot. Percentage of each plot covered by each species of forb and an- nual grass was estimated for each microplot and averaged. Data Analysis Tree plot data. — We determined rela- tionships of basal area to tree foliage biomass of the randomly sampled trees in each plot by nonlinear allometric regression analvses (Tausch and Tueller 1988, Tausch 1989). Analysis re- sults were used to estimate the foliage bio- mass of the remaining trees in each plot from their basal diameters. Individual tree foliage biomass values were summed, for a total tree foliage biomass in each plot. The process was repeated for the trees sampled next to the three plots from Meeuwig (1979) and Meeu- wig and Budy (1979). We used oiu- tree data to predict their total foliage biomass values as a check on the methodologv (Tausch and Tueller 1988). Five indices of site class were used for this study: (1) Site Index I, height at an age of 200 years; (2) Site Index II, height at a basal diameter of 25.4 cm; (3) tallest tree, height of 1990J FoLlACiK BlOMASS IN PlNVON-JUNIPER WOOULANDS 125 tlie tallest tree on the plot; (4) averaiie tree height, average height ol doiniiiant and eo- dominant trees; (5) elevation, in meters of the sample site. Site Index I was determined by the teehni({ues deserihed 1)>- Agnirre-Bravo and Smith (1986). Their methods were sne- cessfnlK' applied to pinyon in the New Mex- ico, Colorado, and Arizona area by Smith and Schnler (1988). This method uses the Chapman-Riehards ecjuation to ht the guide curve for a family of anamorphic site index curves. The equation is: H = 61(1 -e.xp( OoA))' :i] where H = tree height, A -^ tree age, K = a constant equal to 1/(1-63), '^"tl G,, 60, 63 = parameters of the Chapman-Richards equa- tion. Equation 1 was fitted to the combined tree height and age data for all the sampled tree-dominated plots by an iterative, non- linear regression procedure (Caceci and Cacheris 1984). The average age and height of the domi- nant and co-dominant trees were based on the entire plot for sites 4-6. On sites 1-3 these averages were based on the ran- domly sampled trees that were dominant or co-dominant. We determined Site Index I for each tree-dominated plot, using these av- erages in a site-prediction ecjuation based on equation 1 (Aguirre-Bravo and Smith 1986). H (l-exp(-aAJ) (l-exp(-eoA)) (2) where Aq = the age of reference (200 years), A = the average age of the dominant and co- dominant trees, H = the average height of the dominant and co-dominant trees, and S = Site Index I. The second measure of site class, tree height at a constant basal diameter of 25.4 cm (Site Index II), was from work in Nevada pinyon-juniper woodlands by Chojnacky (1986). Nonlinear regression (Caceci and Cacheris 1984) was used to fit the allometric equation (height = a(diameter)') to the di- ameter and height data for all trees in each plot. We determined Site Index Class II height from the equation for each plot for the diameter of 25.4 cm. The last three site in- dices, average height of dominant and co- dominant trees, height of the tallest tree, and elevation, were used directlv. Tabi.f, 2. Nonlinear regros.sion results for hasal area to foliage hioniass ri'lationships for trees sampled in tree- dominated areas of six sample sites. Data for sites 4, 5, and 6 are also discussed in Tausch and Tueller (1988). Site designations arc from 'I'ahlc I. Site Sample Standard number size /"" error (kg) 1 12 .97 3.15 2 12 .85 7.89 3 14 .98 1.03 4(S1) 12 .88 7.81 .5(S2) 12 .92 3.93 6(S4) 12 .93 4.19 Shrub plot data. — Crown volumes for the measured shrub species are based on the equation for one-half of an ellipsoid. A cylin- der was used for the perennial grasses (Tausch 1980, Johnson et al. 1988). We used a sum of crown areas to compute percentage of cover of the measured species on each plot. Allometric ec^uations were derived from crown volume and foliage biomass data randomly collected for each measured species, using nonlinear regression (Johnson et al. 1988, Tausch and Tueller 1988, Tausch 1989). These equations were used with crown volume data for the remaining plants in each plot to estimate foli- age biomass by species. Foliage biomass data from crown measure- ment and reference unit methods were summed for individual species total leaf biomass in each shrub-dominated plot. We extrapolated all data to a 0.1-ha plot size. Understory data from the tree-dominated plots on sites 1, 2, and 3 were similarly treated. Tree/shrub/site comparisons. — We used regression and correlation analyses to com- pare all relationships among total foliage biomass, total percentage of cover, and the five indices of site class. Total foliage biomass, cover, and the five site indices were also com- pared with the number of species sampled in the shrub-dominated plots. A foliage biomass ratio (percentage) of the total in the tree-dominated plots divided by the total in the paired shrub-dominated plots was computed for each site. We com- puted a similar ratio (percentage) for total per- centage cover. These ratios were compared with the number of species sampled in shrub- dominated plots and with the five site class indices by correlation analysis. 126 R. J. Tausch and p. T. Tueller [Volume 50 Table 3. Coefficient of determination values (r") for equations to predict ratio of crown volume to foliage biomass for the sampled shrub and perennial grass species. The letter a indicates plots where foliage biomass was collected from all individuals occurring in the sampled microplots. Site numbers are from Tal)le 1. Site numbei Shrub-dominated plots Plant species 3 Tree-dominated 1 3 Artemisia tridcntata vaseijana A. tridcntata icijominoensis Ceratoides lanata Chnjsothamnus vicidiflorits Ephedra viridis Erio^onum umbellatiim Opuntia sp. Pruniis andersonii Purshia tridcntata Ribes vehitinum Sym))]ioricarpos sp. Elymiis cinereus Ortjzopsis hynicnoidcs Poa sandhcr^ii Sitanion hi/strix Stipa thurheriana .88 .83 a .99 .85 .90 .76 .87 .81 .87 .96 .88 .95 .98 .98 .86 a .96 .98 96 .98 .78 a a ,98 87 a 81 .72 .95 99 .81 .82 87 .86 a .95 .86 .91 .85 Table 4. Standard error of the estimate (g) for equations to predict ratio of crown volume to foliage biomass for the sampled shrub and perennial grass species. The letter a indicates plots where foliage biomass was collected from all individuals occurring in the sample microplots. Site numbers are from Table 1. Site number Shrub-dominated plots Plant species 1 3 Tree-dominated 3 A>-temisia tridcntata vaseijana 33.2 A. tridcntata wyomin^ensis Ceratoides lanata 1 . 28 Clirysothamnus vicidiflonis a EpJiedra viridis Erio^onum iinU)cIlatum Opuntia sp. Prtintis andersonii a Pit rshia triden tata 1.51 Ribes vehitinum 46.6 Symphoricarpos sp. Ehjmus cinereus Oryzo))sis hynicnoidcs Poa sandl)er^,ii .09 Sitanion hystrix .21 Stipa thurheriana .29 15.5 1.79 1.38 .18 7.74 31.3 3.64 1.70 1.25 3.68 1.76 .59 41.8 19.4 07 a 07 .08 .03 20 .06 .14 29 .06 a 1.30 .18 7. 13 a 1.60 Relationships between the foliage biomass components of the shrub-dominated plots were determined by correlation analysis. The components used were the total shrub, total perennial grass, total cheatgrass, and total fori) leaf biomass, and the ninnber of species sampled. These five shrub-dominated plot components were similarly compared with the total foliage biomass in the tree- and shrub-dominated plots and with the five site indices. Results and Discussion Foliage Biomass Predictions Tree data. — Prediction of piin on total leaf biomass in Meeuwig s (1979) plots (4, 5, and 6, Table 1) using e(juations trom trees we col- lected adjacent to those plots had an average error of +0.5% (Tausch and Tueller 1988). Ecjuations for tree data on sites 1, 2, and 3 had coefficient of determination and standard error \'alues ver\ similar to those for sites 4, 5, 1990] F()lia(;k, Biomass in PiNYON-jUNiPKH Woodlands 127 (Y.Z) 75 1- lu 45 O q: LU LU > O o 15 - " ATREE- -DOMINATED PLOTS Y = -170.4+ 0.102 X A r2 = 0.85 P< 0.01 ^^^^^ 2 ^^^ .^^^^ ^ ^ ^ ^^^^ * ^ 6 _ •' ^ • SHRUB-DOMINATED PLOTS ^•^ Y = -218.4+ 0.115 X ^^ 2 J r2= 0.84 P<0.01 I.I.I (X) 2000 2100 2200 2300 ELEVATION (m) Fig. 2. Regression analyses between elevation and total percentage of cover in both tree-dominated plots and shrub-dominated plots. Axis designations follow those in Figure 1. Site numbers follow Table 1. and 6 (Table 2). From these results we consid- ered our tree data from sites 1, 2, and 3 to be similar enough to Meeuwig's (1979) tree data for sites 4, 5, and 6 for the data to be com- bined. Shrub, grass, and forb data. — Based on coefficient of determination (Table 3) and standard error of the estimate values (Table 4), prediction equations for the measured spe- cies have similar precision. Precision is also similar to the tree results (Table 2) and to other test results for sagebrush and bunch- grass foliage biomass (Tausch 1989). The mea- sured shrub and perennial grass species aver- aged 98% of the total foliage biomass on the shrub-dominated plots. This combination also averaged more than 99% of the total foliage biomass of the understory in the three tree- dominated plots we sampled. Total under- story foliage biomass on the tree-dominated plots averaged less than 0.50% of the total plot foliage biomass. We considered the error re- sulting from the lack of understory data for the three tree-dominated plots from Meeuwig (1979) and Meeuwig and Budy (1979) to be minimal. 128 R. J. Tausch and p. T. Tueller [Volume 50 Cheatgrass (Bromus tectorum) occurred on all sites and plots. Common forbs sampled included Colinsia parviflora and Arahis hol- bolii on all but site 2, and Phlox longifolia and Descuriana pinnata on all but sites 2 and 3. Crepis accuminata, Lupinus caudatus, Lii^ ^^ >^ 1*^ 4 - / / / /•' ./^ ^^ ^^ / / ^^ •>< /^ ^ ^^ •SHRUB -DOMINATED PLOTS - ^^ ^^ Y=- 92.49 + 19.05 X *2^^^ r^= 0.97 P<0.01 ■ -"^ 120 - 90 - 60 2: - 30 10 CO ^ CO ■^~ \- o o —I O) Q. j«: Q CO CO < < Z O ^ CD u Q lU C3 < m _| O 3 (£ X CO 6 7 8 9 SITE INDEX I (HEIGHT, m AT 200 YR) Fig. 4. Regression analyses between Site Index I and total foliage biomass in both tree-dominated plots and shrub-dominated plots. Axis designations follow those in Figure 1. Site numbers follow Table 1. The number of species sampled in the shrub-dominated plots correlated with both the total foliage biomass in those plots and with Site Index I (Table 5). Species sampled also positively correlated with total foliage biomass of tree-dominated plots (Table 5) and negatively correlated with the foliage biomass ratio (r = —.86, P < .05). But the species sampled were not significantly correlated with the vegetal cover in either the shrub- or tree-dominated plots, with the percentage of cover ratio, or with the other four site indices. A positive relationship occurred between the percentage of cover ratio and total tree foliage biomass (r" = .81, F < .025), but not between it and the foliage biomass ratio. Cheatgrass was negatively correlated with all the other components of the shrub- dominated plots (Table 5). The highest nega- tive correlation for cheatgrass was with the total tree foliage biomass in the paired tree- dominated plots. Tree- and shrub-dominated plots had sufficiently similar environmental conditions for many relationships to exist be- tween them. A larger effect of topography on foliage bio- mass than on vegetal cover was evident in the data. Sites 2 and 6 were at nearly the same elevation and less than 200 m apart. Percent- age of cover (Fig. 2) did not reflect the envi- ronmental differences between a steep north slope (site 6) and a flat alluvial fan surface (site 2). Topography strongly affected both tree and shrub plot foliage biomass data. Foliage biomass on the north slope (site 6) was about one-third more than on the fan (site 2). Differ- ences in species composition may have also affected foliage biomass more than cover. Sites 3 and 5 are a similar comparison. Site 5, high on the side of the main mountain mass, 130 R. J. Tausch and p. T. Tueller [Volume 50 CO x: ^ CO O) _| •^ Q. CO Q CO !f] CD LU < o o I CD ■D DC I CO (Z) 100 60 20 Y= - 38.25 + 0.0933 X r^= 0.78 P< 0.025 • 5 ^•2 (Y) 800 1000 1200 1400 TOTAL TREE FOLIAGE BIOMASS (kg/0.1 ha) IN TREE- DOMINATED PLOTS Fig. 5. Regression analysis between the total tree iolia^e l)ioniass in tree-dominated plots and total foliage bioniass in shrub-dominated plots over six sites. Axis designations follow those in Figure 1. Site numbers follow Table 1. had Jeffrey pine (Piiuis jcffrcyi) in the vieinity. Site 3, on top of a foothill, appeared drier but had slightly higher cover (Fig. 2). The higher cover on site 3 appeared to result from a higher density of smaller plants. For total fo- liage hiomass, the situation was reversed, with site 5 about one-third higher than site 3. The paired tree-dominated and shrub- dominated plots on each site were connected by dashed lines (Fig. 7) to appro.ximate the a-e-e'-a' successional plane in Figme 1. The site-to-site connections between shrub- and tree-dominated plots were, with one excep- tion, regular over the range of foliage biomass values sampled. At least on the mountain range sampled, the tradeoffs iinoKed are generally consistent with the hypotheses of Figure 1. Site Index I and elevation did not signih- cantlv correlate with each other or with the other three site indices. Site Index II, tallest tree, and average tree height were signifi- cantK- correlated onK with each other (Table 6). Conclusions Foliage biomass and percentage of cover xariation in both shrub- and tree-dominated communities had significant responses to en- vironmental ditYerences. Responses reflected the h\ potheses of Figure I but were not the same for foliage biomass or cover. Total foliage biomass in both tree- and shrub-dominated plots was correlated with Site Index I (height at 200 years of age). They also correlated with each other but not with percentage of coxer. Percentage of cover correlated best with ele- vation. Total foliage biomass was more vari- able in response to topographic differences between sites than total percentage of cover. 19901 Foi.i Aci'. BioMASs IN PiNvoN-ji'NiniH Woodlands 131 30 I- co m D cc I CO *». LU LU CC Q 20 I- < cc CO CO < o LU o < 10 8 10 SITE INDEX I (HEIGHT, m AT 200 YR) Fig. 6. Regression analysis between tlie Site Index I \ alnes and the ratio oi total foliage bioniass in tree-dominated plots divided by that in shrub-dominated plots. Site numbers follow Table 1. T.-\BLE 5. Correlation eoeillicients between four foliage biomass eomponents and the number oi plant species sampled in shrub-dominated plots, among those components and the total foliage biomass in tree- and shrub- dominated plots, and with Site Index I. Relationships between foliage biomass and Site Index I are in Figure 4. Total foliage iiiomass Shrub Perennial grass Cheatgrass Forb Species sampled Shrub Perennial grass Cheatgrass Forbs Species sampled Total foliage biomass (tree-dominated plots) Total foliage biomass (shrub-dominated plots Site Index I ■"f s . 10. ^'P^ .05. 'Ps.Ol. 1.00 .79" .92^ .34 1.00 74-' .60 .W) -.80' -.51 1.00 -.87' -.79^' -.71 .67 .14 -.49 1.00 .28 .59 .66 .82'^ .76" -.72 .66 1.00 .81'' .94'' .96' 132 R. J. Tausch and p. T. Tueller [Volume 50 (0 2 o ^ Q. CO HI CO I- < < 2 z Q2 CQ O < CQ 5i CO < O (e) 100.- 60 20 0 - 4* (a) ^-^V (a') I . 4 3|1 (e') 0 400 800 1200 TOTAL TREE FOLIAGE BIOMASS (kg/0.1 ha) IN TREE- DOMINATED PLOTS Fiji;. 7. ('oinpaiisoiis ol the iclationsliips Ix-twccn total loliai^c hioinass of tree- and slinil)-(l()iniTiatccl plots on six sites on the Sweetwater Mountains. .\\is desiiinations follow those in I'iuuic I. Site numbers follow Table 1. TaHI.I-' (i. (^oirelation eoellieients anioni; three site in- diees of tree heiuht at 25.4 eni basal diameter (Site Index II), the height of the tallest tree, and the average heitiht of dominant and eo-dominant trees in six tre<'-(l()miiiated ]ll()tS. Sit e Ind ex Tallest .■\\('ray;e 11 tree tree heijiht Site Index II 1.00 .97'' or' Tall \st tree 1. 00 0,-V' Ave •age tr i'C lu ight 1 00 ^PS .01. Foliage hioinas.s i.s also clo.seK related to i)i"i- mary procliietioii (W'hittakt'f and Nieiiiiu; 1975) and would appear to he a more .sen.sitive measme for monitoring management residt.s. I'^oliage hioma.ss and vegetal eo\t'r repre- sented (liHerent indicators of en\ iroinnental variation among sites. I'liis appears to he re- lated to the considerable size/density varia- tion among individual plants and species pos- sible when two or more sites are compared. A connniinit) of man\ small plants and/or many small species has higher \ egetal cover than a connmmitx with the same total foliage hiomass hut with fewer, larger plants (Tansch 1980). Foliage hiomass on tree-dominated plots was about 12-25 times higher than on shrub- dominated plots (Fig. 6). This difference may be related to a more efficient use of site re- soinx-es b\ the trcx\s (Doiight\ 19(S7). Foliage hiomass ratios also had imcrse r(dationships to both total foliage hiomass and increasing ele\ ation. Total leaf hiomass, and possibK' an- nual prodnctix it\ , in shrub-dominated com- muniti(\s increast\s more with better site conditions than in tree-dominated comnm- nities. The primary resource iuNoKcxl with impi()\ ing site conditions appears to be mois- ture a\ailal)ilit\ , as described In Nemani and Hmming(1989'). Our loliage hiomass data for shrub- and Iree-domiuali'tl conummities are from onl\ 1990] FoLiACF BioMAss IN Pi\V()\-|rMi'i: H Woodlands 133 six sites on one iiiomitain range. Tlie\ do iiol tulK represent llie lange of variation present on that nioinitain range. In many areas of this and otlier nionntain ranges the speeies eoni- position of the tree- and shrnh-doniinated sites ean ha\'e large variations Ironi the sili-s nsed heri'. Speeilie tohage hioniass le\ cfs and ratios eonid thns (hfler lor other sites. Addi- tional stndies, paitienlarly on a regional basis, will he needed to better establish the varia- tion in the foliage biomass levels and ratios involved. Height versns age emves, widely nsed in eoniniercial forestry, appear to be nsehil in determining site elass on Great Basin sites with pinyon, at least in the Sweetwater Moun- tains. For our data this site index most elosely eorrelated with total toliage biomass and, therefore, potentially with primary produe- tion. A height versus age site index also ap- pears to work equally well for both tree- and adjaeent shrub-dominated eommunities on the same sites. An available inde.x for site could potentially increase ease and accuracy of determining site potential for management of shrub-dominated eonununities, particu- larly in association with pinyon-juniper wood- lands. Additional verification is required to determine the suitability of a site index method for this and other areas of the Great Basin. Acknowledgments The authors thank Suzanne Stone, Donna Peters, and James Gilleard for their help in collecting and processing the field data. Funds for this study were provided by the US DA Forest Service Intermountain Re- search Station agreement number 22-C-4- INT-26. Literature Cited Agiikki:-Bha\(), C , and F W Smith 1986. Site index and volume equations for Piniis pattila in Mexieo. Commonwealth Fore.stiy Review 6.5:51-60. Andkew, M H , I R Noble, and R T Lange. 1979. A nondestruetive method for estimating the weight of forage on shrubs. Australian iiange [ournal 1:22.5-231 . Andrew. M. H.. I. R. Noble, R T. Lange, and A. W. Johnson. 1981. The measurement of shrub forage weight: three methods eompared. Australian Range Journal ,3:74-82. Bahnev. M. a., and N C Frischkneciit 1974. Vegeta- tion changes following fire in the pinyon-juniper t\pe 1)1 west-eenlra! I'tali. Journal oi Range Man- agement 27:91-96. Cahakal. D R . AND N E. VVe.st. 1986. Reference unit- based estimates of winterfat browse. Journal of Range Management 27:187-189. Cacixm. M S . AND W P (;a(:hi;hls 1984. JMlting curves to (lata. Byte 9:340-,362. Caiu'KNTEH. a T . AND N E West 1987. Validating the relerence unit method ol abovegroimd phytomass estimation on shrubs and herbs. X'egetatio 72:7,5-79. (JIOJNACKY, D. C 1986. Pin\()n-junii)(r site (|ualit\ and volume growth ecjuatioTis for Nevada. USDA Forest Service, Intermountain Research Station Research Pajier INT-372. 7 pp. Doughty. J VV 1987. '["he problems with custodial man- agement ol pinyon-juniper woodlands. Pages 29-.33 in K. L. Everett, ed.. Proceedings— pinyon-juniper conferc-nce. I'SDA Forest Ser- vice, Intermountain Research Station General Technical Report INT-215. FvEBE'n. R. L 1987. Plant response to fire in the pinyon- juniper zone. Pages 1.52-1.57 (M R. L. Everett, ed., Proceedings — pinyon-juniper conference. LIS DA Forest Service, Intermountain Research Station General Technical Report INT-215. EvEHErr. R. L.. and K O Ward 1984. Early plant succes- sion in pinyon-juniper controlled burns. North- west Science 58:57-68. (iRiER. C. C , AND S W. Running. 1977. Leaf area of ma- ture northwestern coniferous forests: relation to site water balance. Ecology 58:893-899. Johnson. P S . C L. Johnson, and N. E. West. 1988. Estimation ol phytomass for ungrazed crested wheatgrass plants using allometric ecjuations. Journal of Range Management 41:421-425. KiR.MSE, R D., AND B. E Norton 1985. Comparison of the reference unit method and dimensional analy- sis methods lor two large shrubby species in the Caatinga woodlands. Journal of Range Manage- ment 38:425-428. Long. J N . and F W Smith 1984. Relation between size and density in developing stands: a description and possible mechanisms. Forest Ecology and Management 7:191-206. Long, J N . and J Turner. 1975. Aboveground biomass ol understory and overstory in an age seciuence of four Douglas-fir stands. Journal of Applied Biologv 12:179-188. Marks, P, L., and F II. Borman. 1972. Revegetation fol- lowing forest clearing: mechanisms for return to steady nutrient cycling. Science 176:914-915. Meeuwig, R O. 1979. Growth characteristics of pinyon- juniper stands in the western Great Basin. L'SDA Forest Service, Intermountain Research Station Research Paper lNT-238. 22 pp. Meeuwig, R O, and J. D Budy 1979. Pinyon growth characteristics in the Sweetwater Mountains. USDA Forest Service, Intermountain Research Station Research Paper INT-227. 26 pp. Meeuwig, R. O., and S. V. Cooper 1981. Site ciuality and growth of pinvon-juniptM' stands in Nevada. Forest Science 27:.59.3-601. MoLLER. C M. 1947. The efiect of thinning, age and site on foliage, increment and loss of dry matter. Jour- nal of Forestry 45:393-404. 134 R. J. Tausch and p. T. Tueller [Volume 50 Nemani, R. R., and S. W. Running 1989. Testing a theo- retical climate-soil-leaf area hydrologic ecjui- librium of forests using satellite data and ecosys- tem simulation. Agric. For. Meteor. 44:245-260. Ronco, F., Jr. 1987. Stand structure and function of pinyon-juniper woodlands. Pages 14-22 in R. L. Everett, ed.. Proceedings — pinyon-juniper con- ference. USDA Forest Service, Intermountain Research Station General Technical Report INT- 215. Smith, F. W.. andT. Schuler 1988. Yields of southwest- ern pinyon-juniper woodlands. Western Jouinal of Applied Forestry 3:70-74. Tausch, R J 1980. Allometric analysis of plant growth in woodland communities. Unpublished disserta- tion, Utah State University, Logan. 142 pp. 1989. Comparison of regression methods for biomass estimation of sagebrush and bunchgrass. Great Basin Naturalist 49:373-380. Tausch, R. J., N E West, and A A Nabi 1981. Tree and age dominance patterns in Great Basin pinyon- juniper woodlands. Journal of Range Management 34:259-264. Tausch, R. J , and P T Tueller 1988. Comparison of regression methods for predicting singleleaf pin- yon phytomass. Great Basin Naturalist 48:39-45. 1989. Evaluation of pinyon sapwood to phytomass relationships over different site conditions. Jour- nal of Range Management 42:209-212. Tueller, P T . C D Beeson, R J Tausch, N. E West, andK H Rae 1979, Pin>'on-juniper woodlands of the Great Basin: distribution, flora, vegetal cover, USDA Forest Service, Intermountain Research Station Research Paper INT-229, 22 pp. Waring, R H , W H Emmingham, H L Goltz, and C C Crier. 1978, Variation in ma.ximuni leaf area of coniferous forests in Oregon and its ecological significance. Forest Science 24:131-140, We.st, N E , R. J Tausch, K H Rae, .\nd P T Tueller. 1978, Phytogeographical variation within juniper- pinyon woodlands of the Great Basin, Great Basin Naturalist Memoirs 2:119-136, Whittaker, R. H , and W A Niering 1975. Vegetation of the Santa Catalina Mountains, Arizona. \'. Bio- mass, production and diversit\' along the elevation gradient. Ecolog\- 56:771-790. YouNc;, J A , and R. A. Evans. 1981. Demography and fire history of a western juniper stand. Journal of Range Management 34:501-506. Received 30 March 1990 Accepted 10 June 1990 (;rcat Basin N.iliM.ilist 50i2l. I>)H(I, pp 1:1')- 151 TAXONOMY AND X'ARIATION OF THE LOriDKA MCIUDIA COMPLEX OF WESTERN NORTH AMERICA (HETEROPTERA: MIRIDAE: ORTHOTYLINAE) Adam As(|iiitli Abstkact. — Extt'inal inorpholoiiiical xariatimi in the Iai])1(I((1 iii5.0 mm), brightly colored plant bugs displaying some pattern of contrasting red- black or yellow-black coloration. There is no ta.xonomic revision of the genus, but most species were described in a series of papers bv Knight (1917, 1918a, 1918b, 1923, 1962, 1965) and Knight and Schaffner (1968, 1972). Many species are superficially very similar in habitus, and most have been distinguished b\ the form of the right paramere. This struc- ture is relatively uniform in any given species but extremely variable in size and form among different species of Lopidea. It appears that this is the most valuable diagnostic character available for distinguishing different species o{ Lopidea., aside from the vesica. External and internal male genitalia are now widely used to differentiate taxa in cer- tain groups of Heteroptera, but detailed stud- ies of the variation in these structures are lacking. In the Orthotylini, males often have elaborate parameres and vesicae, and differ- ences in these structures are used to define species (Kelton 1959, Stonedahl and Schwartz 1986). The limits of the variation of these structures in populations and throughout the range of species need to be defined. Several species of Lopidea described from western North America have parameres very similar if not identical to an earlier described species, Lopidea iiifiridia Uhler. I undertook the present study to resolve the taxonomy of this group, which I refer to as the nigridia "complex. In this paper I describe the mor- phological, genitalic, and color variation within this complex and document the charac- ters that unite it as a single taxonomic unit. Material AND Methods Over 3,000 specimens from throughout the range of Lopidea were examined during the course of this study. Male specimens with 'Systematic Entomologx Laboratory, Department of Entomology, Oregon State Universit\. Corvallis, Oregon 97.3.31, Present address; University of Hawaii, College of Tropical Agriculture, Kauai Branch Station, 7370-A Kuamoo Road, Kapaa. Hawaii 96746. 135 136 A. AsguiTH [Volume 50 '^nigridia type" paramere morphology and the associated females were sorted by grouping series that displayed common patterns of color, size, and paramere morphology. Local- ity data from all specimens examined are available in the author's doctoral dissertation, Oregon State University. Male genitalia of specimens from different geographic localities within each group were examined. Techniques for the dissections generally followed Kelton (1959). To deter- mine the infraspecific variation in the struc- tures, I compared variation within and among the populations with the closely related spe- cies marginata Uhler. I had previously deter- mined that the female genitalia are too uni- form throughout the genus to provide information at the specific and subspecific lev- els. Morphological variation in this complex was examined by recording metric data from 139 males from the following localities (N fol- lows each locality): Mexico: Baja California Norte, Parque San Pedro (7); California: Los Angeles Co., El Segundo (10); Mono Co., Leavitt Meadow (10); Trinity Co., Buckhorn Mt. (14); Tuolumne Co., Yosemite Park (10); Colorado: Elbert Co., Kiowa (11); Nevada: Elko Co., (7); Oregon: Polk Co., Dallas (2); Crook Co., Ochoco Summit (10); Deschutes Co., Metolius River (5); Harney Co., Pike Creek (5); Jackson Co., Pinehurst (10); Wash- ington: Pierce Co., Mt. Adams (10); Pierce Co., Mt. Rainier (10); Wyoming: Carbon Co. (14). Specimens from Mt. Adams and Mt. Rainier are topotypes of L. rolfsi Knight and rainieri Knight, respectively. Samples from the rest of the populations were selected to cover the range of type localities as well as color and paramere variation of the nominal species in the nigridia complex. An ocular micrometer was used to measure eight external characters: rostral length (RL) (because the rostrum was often bent at the joints, making its total length difficult to ascer- tain, only the length of the last three segments was measured); hind tibial length (HTL); length of antennal segment 1 (AD; length of antennal segment 2 (A2); width of head across eyes (HW); maximum length of the pronotinn (PL); anterior width of the pronotum (APW); posterior width of the jironotum (PPW). To examine the multidimensional morphological Table 1. Correlations between the first two principal components and the niorphometric measurements of male Lopidca niiiridia. Cluiracter PCI PC II Rostral length 0,636 -0.731 Hind tibial length 0.896 0,129 Antennal segment 1 0.912 0. 173 Antennal segment 2 0.870 0.288 Head width 0.892 -0.135 Pronotal length 0,941 0.032 Anterior pronotal \\i dth 0.847 0.002 Posterior pronotal width ().S92 0.031 variation in these populations, I applied prin- cipal component analysis to the measure- ments (PCA; Morrison 1976) using SYSTAT (Wilkinson 1986). Although a logarithmic transformation usually results in a more nearly normal distribution of the data (Sokal and Rohlf 1981), it can also distort the multi- variate space described by the measurements (Ricklefs and Travis 1980). Analyses using both raw and log-transformed data produced almost identical restilts; therefore, only re- sults using raw data are presented here. Because most of the described species in the nigridia complex were based on differ- ences in color and male paramere morphol- ogy, I recorded eight characters of color and paramere morphology from the 139 speci- mens used in the niorphometric analysis. Color characters were calli, scutellum, em- bolium, and cuneus, and they were coded for black, red, or white. Paramere characters in- cluded angle of the dorsal spine (CA), straight, slightK angled, acute; nimiber of serrations on apex of paramere (SER); number of spines/ bifurcations at apex of dorsal spine (SPIN); development of secondary spine on body of paramere (SECSPIN). These data were standardized and analyzed by SPSS/PC Hierarchical Cluster Analysis using UPGMA on distance matrices of scjuared euclidean distances. Results Principal ('omponent Anai\sis Tlu' first two principal components ac- coimted lor 84% of the morphological \ aria- tion among indi\iduals. The first component (P(> I, 76.1%) reflects the general size varia- tion among indi\ iduals; all variables were pos- iti\ely correlated with PC I (Table 1). PC II 1990] Western N. American Orthotylinae Taxonomy 137 2 •- Q. -1 -1 PC I Fig. 1. Morphological variation o{ Lupidca ni^ridia Uhler based on principal component analysis. Populations are plotted on the first (PC I) and second (PC II) principal components, enclosed in polygons connecting the outlying individuals of each sample. Abbreviations: R = Mt. Rainier, WA (L. n. nigridia); A = Mt. Adams, WA(L. n. nigridia); Ca = Carbon Co., WY (L. n. scrica); N = Elko Co., NV (L. n. nigridia); B = Baja California Norte (L. n. aculeata); J = Jackson Co., OR(L. n. actdcata); Cr = Crook Co., OR(L. n. acideata); LA = Los Angeles Co., CA(L. n. nigridia); M ^ Mono Co., CA (L. n. nigridia); E = Elbert Co., CO (L. n. serica); T - Trinity Co., CA (L. n. acideata); Y = Yosemite Park, CA (L. n. acideata); D = Deschutes Co., OR(L. n. acideata). (7.9%) reflects an inverse relationship be- tween RL and A2. To illustrate the distribu- tion of populations in the morphological space described by the principal components, indi- viduals were plotted on axes described by PC I and PC II and populations were enclosed in polygons by connecting the outlying indi- viduals with lines (Fig. 1). This analysis illustrates some oi the mor- phological differences among populations. For example, the Yosemite population (Y) is composed of large individuals with relatively long antennae and short rostra. The Mono County population (M) is composed of rela- tively small individuals with short antennae and long rostra. These two populations exam- ined separately are quite distinct; they do not overlap in overall size and have differently proportioned antennae and rostra. However, both populations overlap other groups to 138 A. ASQUITH [\^olume 50 E :3 o O cc Q. U. O C5 Z. UJ ~i S o QC u. O £ 2: ^\H 1.6 - 1.5 - 1.4 - 1.3 - J' I" »^ ■■ ■■ ■ i^ ■ - , 1.2 - . \ ■■"-. 1.1 - ■ ■ ■ 1 - '■ I I 1 1 1 1 B ^ 0.6 0.8 1 LENGTH OF PRONOTUM (mm) 1.2 Fig. 2. Relationship between relative length of the rostrum (rostrinn length/pronotuni length) and pronotuni length in Lopidea nigridia; ij = 2.1777 - 0.884x, r" = 0.757, N = 128. some degree, creating a continuum of mor- phological variation in all dimensions. This pattern makes it difficult to clearly segregate a population or groups of populations based on external morphology alone. There was no clear pattern of morphologi- cal variation with regard to geography. The largest individuals were found in two Califor- nia populations (Y, T), the Wyoming popula- tion (Ca), and the Colorado population (E). Individuals with short antennae and long ros- tra were found in the Wyoming population (Ca) and a California population (M). The two most morphologically similar populations were Wvoming (Ca) and Los Angeles Countv (LA). Not all coefficients of variables in the F('A analysis were of ecjiial magnitude, suggesting allometric relationships among the varial)les. For example, PC I represents general si/e variation among individuals, and rostral length has the lowest correlation with PC I (Table 1). This suggests that as size increases rostral length increases more slowly than other characters. The significance of this pattern can be seen by examining the relationship of rostral length to the best single measure of size, pronotal length. The relative length of the rostrum (RL/PL) decreases with increasing size (Fig. 2). Very small individuals have rostra that are 1.5 times the length of the pronotum, whereas very large indiv iduals hav e rostra that are only equal to the length of the pronotum. This has important implications regarding the taxo- nomic \alue of these and similar characters, such as the distance the rostrum extends pos- teriorly on the sternum. In very small individ- uals of the nigridia complex the rostrum ex- tends to or slightK be\()nd the metacoxae, whereas in huge intlix iduals the rostrum may not reach the mesocoxae. Color Pattern Dorsal coloration of indix iduals from any one series was usualK uniform, but color 1990] Wkstern N. Ami:rk:an Ohthotvunae Taxonomy 139 Fig. 3. Variation in dorsal color pattern ofLopidea nigridia Uhler: A, fuscous-white color pattern characteristic of L. n. nigridia Uhler; B, fuscous-red-white color pattern characteristic of L. n. aculcata Van Duzee; C, solid red color pattern characteristic of L. n. scrica Knight. Stippled areas represent fuscous coloration; gray areas represent red coloration. varied dramatically among collection.s. At one extreme is a red form that is uniformly brick red with slight to moderate infuscation on the clavus. At the other extreme is a fuscous- white form with the clavus and corium pre- dominantly to completely reddish fuscous and the embolium and cuneus pale white (Fig. 3). Color variants intermediate of the two ex- tremes also occur. The color patterns of the nigridia complex also occur in several related sympatric spe- cies. Lopidea marginata Uhler displays very similar color variation, with some populations composed of solid red individuals, while in other populations the clavus and corium are infuscated and the embolium and cuneus pale white. The different color forms in the nigridia complex do not appear to be segregated with regard to host plant west of the Rocky Moun- tains, and both color extremes have been collected from near sea level in southern Cali- fornia to >5,000 ft. elevation in the Sierra Nevada and Cascade Mountain ranges. The most conspicuous geographic patterns are the absence of the red form from the Intermoun- tain Sagebrush Province and the absence of the fuscous-white form from the Great Plains short-grass prairie (Fig. 4). This latter pattern also seems to correspond to a switch in pre- ferred host plants from Lupinus to Astragalus (see Biology). Paramere Structure There were few correlations between color and paramere variables and morphology. PC I, representing size, was negatively corre- lated with the mmiber of serrations on the paramere and all color variables (Table 2). In general, populations of large individuals also 140 A. ASQUITH [Volume 50 Fitj;. 4. Distribution of siilisptcies and color forms of L. ni^ridia L'iiler in WL'Stcrn iNcjrth America: triangles -- fuscous white form of L. n. niaridia Uhler; open circles = more reddish color form of L. n. ni^ridia Uhler; solid circles - solid red color form of L. n. acideata \'an Duzee; half solid circles = red-white color form of L. n. aculeata \'an Duzee; solid squares = L. n. scrica Kniglit. tend to have more serrations and to he sohd red with no white on the eml){)hum or cuneus, and smaller individuals have fewer serrations and are more fuscous with a white embolium and cuneus. Although this trend was apparent for most specimens I examined, it was not always true; individuals from Deschutes Co. , Oregon (D), are relatively small and yet are solid red in color, and I have seen very large specimens from Santa Barbara Co., Califor- nia, that ha\ e a light emholiinn and cuneus. Many characters of the right paramere for- merly used to distinguish species within the ni^hdia complex vary among individuals within a population. For example, wilcoxi Knight was distinguished from rainieri Knight h\ the absence of a secondary spine in icilcoxi. In onl\- two populations examined 1990J Western N. Amekican ORriioivLiNAE'rAXoNOMv 141 Table 2. Pearson cont'latioii coefTicients between the first two principal components and paraniere and color characters of male Lopidca iii .05). Character PCI PC 11 CA -0.157NS 0.121 NS SER -0.361 ** -0.066 NS SPIN -0.020 NS -0.054 NS SECSPIN -0.067 NS -O.llONS CALLI -0.308 ** 0. 195 * SCUT -0.207* O.ISO* EMBOL -0.594** 0.072 NS CUN -0.564** 0. 1 16 NS did all individuals either have or completely lack this structure. Some populations in the Siskiyou Mountains of California and Oregon contain indi\iduals with a distinct toothed hook ventrally on the apex, used by Knight (1965) to distinguish calcaria Knight and eri- ogoni Knight. Other individuals from the same series lack this structure and display parameres more similar to other described species in the complex. Figure 5 illustrates the extent of variation of the right paramere seen in the nighdia complex. The only aspect of the right paramere common to all popula- tions and absent in other species ofLopidea is the presence of the elongate dorsal spine at the apex. Examination of the left paramere and inter- nal genitalia corroborated the patterns seen in the right paramere. The left paramere is structurally less complex than its counterpart and thus shows less variation. The medial flange is digitiform, with its distal end usually slightly clavate and free from the main body of the paramere. The vesica bears a slender, slightly curved ventral spicula, toothed at the apex and with a slight swelling at its midpoint. The dorsal spicula is short, broadly lanceolate, toothed, and slightly curved. The variation in these structures between color forms of the nighdia complex is no greater than the in- fraspecific variation seen in other species. This is illustrated in Figure 6, where genitalic structures of a fuscous-white and a red form of nighdia, both from Wyoming, are compared with the same structures from individuals of marginata Uhler from Oregon and Baja Cali- fornia. The dorsal spicula is usually shorter and straighter in the fuscous-white color form. The dorsal spicula, however, varies in shape from straight and blimt to curved and evenly pointed {V\g. 7); it also shows considerable \ ariation in other species of Lopidca. Cluster Analysis This analx sis demonstrates thc> diniculty of separating groups within the nighdia complex based on color and paramere characters. In no case were all individuals from one population found to be most similar to each other; at least one individual was always grouped with those irom another population. In most cases, individuals from any given population were scattered throughout the dendrogram. For example, the Mt. Adams population (A) had individuals placed in four of the five major clusters (Fig. 8A). The cluster analysis did not identify groups composed of individuals that I determined as being the same color form. For example, all individuals from Crook Co. (Cr), Trinity (T) (Fig. 8B), and Jackson Co. (J) (Fig. 8A) repre- sent the solid red form; however, Cr speci- mens were grouped in the uppermost cluster, T specimens in the next lower cluster, and J specimens in the middle three clusters. Sim- ilarly, specimens representing the fuscous- white form were also found in all of the major clusters. This analysis further suggests that grouping specimens within the nighdia com- plex based on color and paramere morphology gives equivocal results. Ta.xonomy All specimens examined in this study clearly belong to a monophyletic group. They are united by the presence of an elongate dorsal spine on the apex of the right paramere; a free, digit-shaped medial flange on the left paramere; and a slender, slightly spindle- shaped ventral spicula. These are derived characters found in no other species of Lopi- dea. In addition, all specimens are believed to be conspecific for the following reasons. Pop- ulations or groups of populations cannot be distinguished by combinations of external morphological measurements. Although pop- idations display considerable color variation, color is not highly correlated with external or paramere morphology, and similar color vari- ation is seen in related species. Most charac- ters of the right paramere vary among individ- uals from any population. Only characters 142 A. ASQUITH [Volume 50 Fig. 5. \'ariation ot li^ht paraiiieie in Lopidca iii^hdia Uhler. Drawn in posterolateral view. common to all population.s, such as the elon- gated spine on the dorsal apex of the right paramere and the digit-shaped medial flange on the left paramere, also corresponded with unique characters of the male vesica. I have also examined the type specimens of all nominal species in the nigridia complex and have determined, using the above crite- ria, that they also are conspecific with nigridia Uhler. I interpret nigridia as being a poly- typic species comprising three subspecies segregated to some degree by geography and/ or habitat. I have elected to use the subspe- cies category for these taxa because, based on the available data, it adequately describes the broad geographic patterns of the color forms. I have retained the subspecies L. n. nigridia Uhler for the Intermountain, fuscous-white form and L. n. serica Knight foi- the solid red, eastern Rocky Mountain and prairie form. I also recognize L. n. aculeata Van Duzee as a polymorphic form of the Pacific Coast states. Below I provide a complete synonymy for nigridia and its subspecies. All lectotype and holotype label data are given verl)atim. Lopidca nigridia Uhler Lopidca iii (loisalK and laterally, reddish fuscous distalK with hiack apex. ANTENNAE: black, fuscous, or red; I, length 0.40-0.64, with two huge, stiff setae distallv on the medial surface; II, 1.34-2.28; III, 0.81-1.50; IV, 0.35-0.51. Pronoti'M: length 0.65-1.29, posterior width 1.25-1.96, broadly convex, siuface smooth, anterior angles rounded, lateral margins cari- nate, slightK- arcuate in dorsal view, lined with erect, black setae, posterior margin straight or slightK sinuate; calli lighth infus- cate to piceous, posterior angles broadly rounded, surrounded by fulvus or yellowish white; disc brick red to gray fuscous; pro- pleura smooth, glabrous, episternum fulvus to white, sternum black. Legs: black, testa- ceous, or fulvus; coxae and trochanters pale or fulvus; femora black on dorsum, paler on ante- rior and ventral surfaces, often spotted with fuscous, pale at apex; tibiae black or dark red, tarsi black. Genitalia: Tergal process: rela- tively long compared with other species of Lopidea, evenly narrowed to a sharp point, slightly curved medialK'. Riglit paramerc: roughly rhomboidal in outline, apex with long, erect spine; spine pointed or bifurcate at tip, straight or inclined toward base of paramere (Fig. 5). Apical edge of paramere slightly curved medially, usually with two vertical rows of small teeth; number and posi- tion of teeth variable. Small secondary spine occasionally present on dorsal edge near base of apical spine. Basal arm long, thick, curved medioventrally, apex variable, usually bifur- cate (Fig. 6). Left paramere: sharply angled with apical lobe oval in lateral view. Medial flange distinct, separate from lateral flange for most of its length; narrow, elongate with distal end usually slightly expanded. Vesica: Dorsal spicida: short, lanceolate, straight or slightly curved, both margins of distal third serrate (Fig. 7). Ventral spicula: long, slender, slightly curved, a small swelling present near middle, apex with small teeth (Fig. 6). Vesti- TURE: head and pronotum with short, stiff, erect, black setae, black setae on hemelytra variable in length, suberect to erect, occasion- ally pale on light-colored area of corium, pronotum and hemelytra also with flattened sericeous setae, venter moderately covered with short, suberect pale setae. Female. — Similar in structure, color, and vestiture, but larger, broader, and more ro- bust; frons more protuberant and broadly con- vex than in male, vertex fkit, basal carina less distinct, lateral margins of pronotum less cari- nate, hemelytra arcuate laterally. Length 4.82-7.46. Head: width across eyes 1.12- 1.30, vertex 0.69-0.82. Ro.strum: length 1.22-1.55. Antennae: I, length 0.51-0.76; II, 1.48-2.49; III, 1.01-1.47; IV, 0.41-0.52. Pronotum: length 0.91-1.50, posterior width 1.42-2.17. Lopidea nigridia nigridia Uhler Lo])'ulca iti^ridia Uhler, 1895:30 (n. sp., desc). Lopidea niiiridca ninridea: Van Duzee, 1921:128. Henry and Wheeler, 1988:42,3 (cat.). Lopidea rainieri Knight, 196.5:8-9 (n. sp.). Henry and Wheeler, 1988:423 (eat.). Neiv synonymy Lopidea scuUeni Knight, 1965:9 (n. sp.). Henry and Wheeler, 1988:424 (cat.). Neiv synonymy Lopidea rolfsi Knight, 1965:9 (n. sp.); Akingbohungbe, 1972:842 (note). Henry and Wheeler, 1988:424 (cat.). New synonymy Lopidea wilcoxi Knight, 1965:11-12 (n. sp.). Henry and Wheeler, 1988:425 (cat.). New synonymy Diagnosis. — L. n. nigridia Uhler is small to moderate in size, parallel sided, with a contrasting dorsal color pattern of smoky fus- cous on the pronotum, scutellum, clavus, and most of the corium and pale white on the outer corium, embolium, and cuneus (Fig. 3A). Distribution. — L. n. nigridia occurs along the western slopes of the Rocky Mountains, throughout the Great Basin from southern Nevada and Utah to southern British Colum- bia. It is the common form along the eastern slopes of the Cascade Mountains and northern Sierra Nevada and occurs west of these ranges through xeric, low-elevation passes and river basins in California. L. n. nigridia also occurs throughout the coastal chaparral of southern California and into Baja California Norte. This subspecies inhabits the sagebrush-steppe habitat of the Great Basin, xeric mountain slopes, and dry lowlands. Its range appears to interdigitate with and superimpose on the ranges of the other two subspecies in some areas. However, the subspecies appear to be segregated by habitat in areas of sympatry, with n. nigridia inhabiting xeric shrub steppe or chaparral habitats and the other subspecies occurring in more mesic conditions, usually at higher elevations. Lopidea nigridia aculeata Van Duzee, new status Lopidea aeuleata Van Duzee, 1917:271 (n. sp.). Carvalho, 19.58:83 (cat.). Knight, 1965:11 (color, dist.). Henry and Wheeler, 1988:417 (cat.). 146 A. AsyuiTH [Volume 50 R(9) A(4) Ca( 1 ) N(2) B(1) A( 1) Ca( 1 0) Ca(3) A(1) N(1) N(1) J(2) N(2) B(1) J(1) B(1) J(1) R(1) A(3) 8(2) A(1) N(1) J(3) B( 1) J(3) B(1) H Figs. 8A-B. Results of UPGMA cluster analysis of color and paramere characters of 12 populations of L. ni'^hdia. Letters represent populations; numbers represent the number of incli\i(hials from that population placed in that cluster. Both dendrograms are identical; to facilitate viewing and discussion, half the samples are shown on dendrogram A and the other half on dendrogram B. A, R Mt. Rainier, WA {L. ii. ni'^ridia ); A - Mt. Adams, WA (L. n. ni^ridia ); Ca = Carbon Co., VVY (L. n. scrica); N =- Elko Co., N\' (L. n. nicridia). B Baja California Norte (L. n. aculeata),] = Jackson Co., OR (L. n. acideata). B, Cr ^= Crook Co., OR (L. n. acidcata); LA = Los Angeles Co., CA (L. n. nifiridia); M = Mono Co., CA (L. n. nii:,ridia); E = Elbert Co., (X) (L. n. scrica); T Trinity Co., CA (L. n. acidcata). Y - Yosemite Park, CA (L. n. aculcala ). A scale of distance \ alues is not included because this analysis was not jierformed to measure morphological dillcrcuces among OIT's but to illusti;itc groupinus of OITs usinu conxcntiouiil taxonomic characters (see text). 1990] Wkstekn N. Amehican Oh i no iveinae Taxonomy 147 B Crd) LA(6) M( 1) _I}n E(6) T(2) Cr(5) M(2) E(3) Y(3) Cr(4) M(3) E(1) Yd) Ed) T(3) Md) Y(4) T(9) M( 1) Md) Yd) LAd) LA(2) LAd) IF Lopide Lopidc Lopide Lopidc Lopide adiscreta\'imDuzee, 1921:127 (n. sp.). Carvalho, 1958:84 (cat.). Henry and Wheeler, 1988:419 (cat.). New synonymy Lopide a nifiridea hiiia Van Duzee, 1921:128 (n. subsp.). Carvalho, 1958:87 (cat.). Henry and Wheeler, Lopide 1988:42.3 (eat.). Neiv synonomy a fallax Knight, 192.3:69 (n. sp.). Van Duzee, Lopide 1933:96 (note). Carvalho, 1958:84 (cat.). Henry and Wlieeler, 1988:420 (cat.). New synonymy a yakima Knight, 1923:69-70 (n. sp.). Carvalho, Lopide 1958:88 (cat.). Henry and Wheeler, 1988:425 (cat.). New synonymy ■a usingeri Van Duzee, 1933:96 (n. sp.). Carvalho, Lopide 1958:88 (cat.). Henry and Wheeler, 1988:425 (cat.). New synonymy a audeni Knight, 1965:9-10 (n. sp.) Henry and Wheeler, 1988:417 (cat.). New synonymy (I eriogoni Knight, 1965:10 (n. sp.). Henry and Wheeler, 1988:420 (cat.). Neic synonymy a calcaria Knight, 1965:11-12 (n. sp., note). Henr\ and Wheeler, 1988:418 (cat.). New syn- onymy a ch(ind)erlini Knight, 1965:12-13 (n. sp., note). Henr\' and Wheeler, 1988:418 (cat.). Neiv syn- onymy a angustata Knight, 1965:12 (n. sp.). Henry and 148 A. ASQUITH [Volume 50 FiK. 9. Gcneiali/rcl (listrihiitioii oi'/.. uii^riditi I'lilcr. Sniall dots /.. n. scrica Kiiitilit; laruc dots /.. ii. uculcata Van Diizee; diagonal lines L. ii. niiiridia rlilcr; dark circles additional localities lor /.. /(. /i(i,'/»/((/ ; dark s((uares = additional localities [or L. n. scrica. 1990] Wkstki{\ N Amkkicw OhiiiotylinakTaxoxomy 149 Wliccler, 19S.S:417 (cat.). New si/nuiu/niij Lopiclcd nihrofusca Knight, 1965:13 (n. sp.). IIimiin and Wliccler, 19SS:424 (t'at.). \cusijiioiup)iij LopicU'd fhivicostdtd Kniglit and SchaOher, 1968:75 (n. sp.). Hi'nn and Wlu't^lcr, I9,SS:42() (cat.). New si/noniji)iy Di.\c;nosis. — L. n. aculeata Van Duzee i.s highly variable in size and coloration (Fig. 3B). It is iisnalK larger than n. nighdia and often larger than n. serica, bnt it is always more linear than the latter. In the movnitains of British Columbia, Washington, and Ore- gon it is solid red in dorsal coloration, with more \ ellow ish indi\ iduals found at lower ele- vations. Northern California individuals show some white along the embolium and cuneus, this pattern increasing in distinctness and fre- quency in southern populations. This subspecies is itself highly variable, and several distinct color forms can be distin- guished as follows: (1) The type specimens of aculeata from Seattle, Washington, are yel- lowish with a dark head and a large hook at the posterior angle of the apex of the right paramere. The type material is representative of populations found at low elevations in the Willamette-Puget Lowland area of Washing- ton and Oregon. (2) L. n. Jiirta Van Duzee was described from San Miguel Island off the coast of southern California. These specimens are solid red, small, and distinctly arcuate later- ally. I have seen four males from San Miguel Island in the USNM. These specimens are larger and slightly less arcuate than the type specimens of n. Jiirfa, but are still different from mainland populations at that latitude. (3) Specimens from the mainland of southern California are large and linear; most have a noticeably pale embolium and cuneus. Some populations from the southern Sierra Nevada, the San Gabriel and Santa Rosa mountains of southern California, are very distinct. The hemelytra are darker, almost fuscous, the disc of the pronotum is deep red and always shiny, and the setae, especially on the pronotum, are shorter and more decumbent. The type speci- mens ofdiscreta Van Duzee are of this form. Distribution. — L. n. aculeata occurs in the Cascade Mountains of British Columbia, Washington, and Oregon, the eastern slopes of the coastal mountain ranges in these areas, and in the Blue and Wallawa mountains of Oregon and Washington. It occurs through- out the Coastal and Sierra Nevada ranges of California. In southern California, however, the ranges of n. aculeata and u. ni^ridia over- lap, and specimens intermediate and distinct in color pattern occur. Detailed studies of the local distributions of the color forms in this area are needed to clarify the problem. Lopidea id^ridia serica Knight, new status Lopidea serica KnighU 192.3:69 (n. .sp.). Kclton, 1980:235 (dist., ho.sts, fig., key). Akin5i;l)<)linnKbe, 1972:842 (note). Henry and VVheeler, 1988:424 (cat.). Loiiidea inedleri Akin^hohungbe, 1972:840-842 (n. sp.). Henry and Wheeler, 1988:422 (cat.). New sijn- onipuij DiACNOsis. — L. n. serica Knight is larger, more robust, with the lateral margins usually arcuate and solid red in dorsal coloration, ex- cept for black on the calli and light infuscation on the clavus (Fig. 3C). Females are usually submacropterous, with the membrane of the hemelytra reduced and barely reaching the end of the abdomen. Although this is the most morphologically distinct of the subspecies, it did not appear as such in the PCA because I did not use characters such as total length and maximinn width of hemelytra. Distribution. — L. n. serica occurs along the eastern slopes of the Rocky Mountains from Alberta to Colorado and east across the northern Great Plains to southern Manitoba. It appears to inhabit the mesic grasslands of the eastern Rocky Mountains and short-grass prairie systems. There are two interesting disjunct localities for n. serica in western Wisconsin and south- western Yukon Territory and adjacent Alaska (Fig. 9). Although u. serica might be expected to occur in the relictual prairies of Wisconsin, the Wisconsin record comes from an area of scrub oak savannah. The Yukon records are from an area along the western edge of the Yukon Plateau and at the southern edge of the Alaska- Yukon glacial refugium. This record may represent a relictual population from the refugium or the tip of the post-Pleistocene northern migration along the Interior Plateau of British Columbia, although there are no other localities north of southern British Columbia. The host plants Lupinus and Astragalus are common to both the disjunct localities. Discussion of Species Lopidea nigridia is the original spelling used in the description by Uhler (1895). This 150 A. ASQUITH [\'olume 50 clearly was not a lapsus, as I have seen Uhler determination labels using this spelling. The next citation to the species is Van Duzee (1914), who used the incorrect spelling of ni- gridea. All subsequent citations have also used the incorrect spelling. There is confusion concerning the true identity of the species that Uhler referred to as nighdia. In his description (Uhler 1895), he described the color as brownish black with the outer border of the corium and cuneus rufo- fulvous or rufous, with no mention of white on the embolium or cuneus. However, this is clearly a contrasting dark-light pattern like that of the fuscous-white color form (n. ni- gridia). In addition, Uhler describes the ante- rior border of the pronotum as white, a pat- tern that occurs only in the iuscous-white form (n. nigridia) and not the red form (/i. serica). I located a fuscous-white specimen in the USNM bearing the label Colo. 1387. This number, 1387, corresponds with the follow- ing information in the C. F. Baker catalog: Steamboat Springs, Colo., July, C. F. Baker, ex. Delphinium occidcntalc (I attached a label with those data on the specimen). This infor- mation matches that given by Uhler for one of the specimens he examined for his original description. Knight (1923) illustrated the right paramere of another specimen from the type locality, and it is this concept of nigridia that has been used by all subsequent authors. Therefore, I have selected the former speci- men as the lectotype of Lopidca nigridia Uh- ler and indicated such b) attaching a label. I have also seen specimens oi nigridia with Uhler determination labels bearing the name Lopidea ohscura Uhler, a Uhler manuscript name. It is possible that this is the name Uhler used for L. n. nigridia, and his description of nigridia referred to some other species with a contrasting light-dark color pattern. In addi- tion, I have seen diflerent specimens from the same locality identified by Uhler as both ni- gridia and ohscura. It is likely, however, that the specimen I have selected as the lectotype was examined by Uhler in his description of nigridia. Discussion of Subspp:(:ies Lopidea nigridia aeuleafa is highK variable and remains confusing to taxonomists. When discussing aculeata. Knight (1965) noted that specimens collected from diflerent areas in Oregon had identical parameres but varied from \ ellow fuscous to red fuscous and con- cluded that this species was variable in color. Van Duzee (1921), when describing discreta, commented, "It might be best to consider this a race or variety ofnigridea. In his discussion ofusingeri (Van Duzee 1933), he stated, "This species, like ohscura exhibits considerable variation in the depth of coloration. ' Knight distinguished serica from nigridia by the presence of golden sericeous pubes- cence in serica, but all specimens of nigridia (all North American Lopidea, in fact) have this pubescence if it is not rubbed off. I have seen specimens that are topotypes of rolfsi Knight and rainieri Knight that Knight originally determined as nigridia Uhler and other fuscous and white specimens from Idaho determined as nigridia Uhler. Several specimens of intermediate color pattern from California have also been determined as ni- gridia by Van Duzee. Lopidea ruhrofusca Knight was described from a single male from Monticello, Utah, and is somewhat enigmatic. It is almost solid red, typical of n. serica. In size and development of the hemelytral membrane, however, it is more similar to n. nigridia: thus, I have s\n- onymized it with n. nigridia. Analysis of the ecology, behavior, habitat, and host preference in areas of sympatry may prove that the subspecies of L. nigridia are actually distinct species, but morphologically they do not dispku differences as great as those seen between other species of Lopidea. In addition, more detailed studies of the pop- ulations in some areas may suggest that some of the color forms within the subspecies de- serve taxonomic recognition. With the avail- able information, however, it is more prudent to recognize the structural similarit\ between these populations and the rest of nigridia and detail the geographic variation, rather than assign names to populations with distinct color patterns. Genitalia I have weighted genitalic characters heavily in forming a species concept for L. nigridia. rhis is based on examinations of these struc- tures throughout the genus and in related 1990] Wkstkhn N. Amkhicw OhtiiotvlinakTwonomv 151 Orthotylini. M\' anal\ ses of paranuMc striic- ture show no geographic pattern or (hstinclioii among snhspeeies. It is possible that incipient speciation has occurred in this complex and that it is not reflected in ixuamere morphol- ()g\'. This is most plausible lor L. //. /i/^/(V//V/ and L. n. scrica in the northern and eastern parts of the range, where the\ retain distinct color patterns and exhibit the greatest difVer- ences in the shape of the dorsal spicula. Other species of Miridae also display geographic \ariation in size, \estiture, or color, including Irbisia brachijccra (Uhler) (Schwartz 1984) and Pilophorus tibialis Van Duzee (Schuh and Schwartz 1988). Although the parameres and vesicae ha\ e been used as ta.xonomic characters in the Miridae for at least 40 years, few studies haxe described the within-species variation of these structures. Stonedahl and Schwartz (1986) illustrate the variation in paramere structure for some species of PscuclopsaUus. Stonedahl (1988) described clinal \ariation in the size and shape of the vesica of PJiytocoris yuIIaboUac Bliven and recognized two bio- t\ pes of P. fraterciihis Van Duzee based on geographic differences in male genital struc- tures. He found that other species o( Phijto- coris such as P. tenuis Van Duzee are highly variable in size, color, and genital structure; yet none of these variables were correlated with each other, nor did any show clear pat- terns of geographic variation. Detailed docu- mentaion of variation in genitalic structures is rare for any group of Heteroptera. Several examples are available for the auchenorrhyn- chous Homoptera, however. Euscelis incisus (Kirschbaum) exhibits seasonal variation (Muller 1954), and E. incisus Brulle shows temperature-induced variation (Muller 1957) of the aedeagus. Wagner (1955) illustrated ex- treme clinal geographic variation in the aedeagus of Philaenus spumarius (L.). Other studies have documented the intra- and inter- populational variation of aedeagal characters in this group (Wagner 1967, Le Quesne and WoodroflFe 1976, Oman 1987). Studies of the infraspecific variation in spicula shape in the orthotyline Miridae are greatly needed. In L. ni^ridia the dorsal spicula varies from straight and bhmt to curved and pointed (Fig. 7). The ventral spicula can also be twisted and varied in its curvature and dentation. (>()1.()K The distinction between the subspecies in some areas and their discrete distributions probabK reflects some degree of genetic seg- regation. This pattern might be interpreted as a species-levt'l phenomenon; however, the subspecies are almost identical morphologi- cally and do not appear to be segregated by host plant, as are otlier species oi Lupiclea. L. n. aculcata, however, shows inter- and intra- populational variation in color pattern from fuscous-red or solid red to red-white. Although I have placed subspecies deter- mination labels on all specimens I examined for this study, the assignment of some popula- tions to L. /(. nifiridia or n. aculeata is equivo- cal. For example, I have examined two series of specimens both collected from Mokelunnie Hill, Calaveras Co., California, but from dif- ferent years. One series exhibits the fuscous- white color pattern typical of n. ni oi Utali, Salt Lake City, Utah 84112. 155 156 R K \'ICKEKY, Jr [Volume 50 stigma of the next flower. Mimuhis lewisii has textbook-typical bee flowers (Faegri and \an der Pijl 1979). Minudus cardinalis ranges from southern Oregon south to central Baja California, and from the California Coast Range inland to mid-elevations in the Sierra Nevada (Vickerx' and Wullstein 1987). The lavender-flowered race of M. lewisii occurs at elevations higher than M. cardinalis in the Sierra Nevada. The magenta-flowered race ranges from the north- ern Sierra Nevada north to Alaska and east to the Rocky Mountains (Viekery and Wullstein 1987). The two species rarely overlap and then only when seeds of M. lewisii wash down into the range of M. cardinalis and become established as ephemeral populations on streamsides, principally in the central Sierra Nevada (Hiesey et al. 1971). The sympatric populations flower at the same time, which heightens the importance of their reproduc- tive isolation by different pollinators. Both species produce nectar throughout the day, although the nectar production of lA/. cardinalis is far more copious than that of A/. lewisii. Before the main, long-range (|uestion of the effect of differences in flower color and/or shape on the pollinators can be investigated, it is necessary to establish some basic facts. First, do M. cardinalis and M. lewisii require the service of pollinators? Or, do they self-pol- linate, at least to some extent? Second, if polli- nators are required, which ones normally visit the flowers of the two species? Once the norms are ascertained, then the effect of dif- ferent colors and/or shapes can be deter- mined. Third, are the pollinators faithful to their species? Or, does cross-pollination occiu' between the two species? That is, would a difference in pollinators isolate the two spe- cies reproductively? Or, only partially? Or, would the differences between the species tend to swamp out? The purpose of this study is to answer these intrinsically interesting ba- sic questions and, in addition, to prox ide the necessary foundation data for the long-range study. Matkiuals AM) MirmoDs Plants of t\'pical red-floweied M. cardinalis Douglas (culture 13313 from ('edros island, Baja California) were grown from seed in the 30cm Fiif. 1. Arraiigfiiifiit of potted plants in the experi- mental sets. The reciprocal arrangement was ot A/. leuisii in tlie center surrounded by six M. cardincilis plants. The dottt'd line indicates the location ot the screen cage in the pollinator exclusion trials. University of Utah greenhouse, as were plants of magenta-flowered M. lewisii Pursh (culture 5875 from Alta, Utah), t>'pical of the Rocky Moimtain race. The seedlings were trans- planted first into 4" pots and then, when large enough, into deep 8" pots. The bigger pots allowed the plants to grow larger (20-60 cm high) and produce man\ flowers for the field studies. The field tests were carried out at two sites in the Wasatch Mountains of Utah. The first location was in the Red Butte Canyon Natural Area, Salt Lake County, and the second, at Silver Fork in Big (>ottonwood Canyon, also Salt Lake County. In Red Butte Cauxon the pots of plants were placed on the wet delta at the head of the reservoir, elevation 5,360 feet, so they could be watered naturally. At Silver Fork the pots of plants were placed in the meadow, eknation 7,800 feet, below Silver I^'ork Lodge, and were watered daih' by Luther Light. The plants were arranged in experimental sets of seven plants. In each set the center pot contained a plant of one species, e.g., M. c(irdin(dis: and a w horl of six pots snrroimding it I'acli contained one plant of the other spe- cies, e.g., M. leuisii (Fig. 1). This arrange- 1990] POLLIN.VIION IN MiMULUS 157 intMit was designed to f'acilitale cioss-iiollina- tion, sliould it oeeur. At the Red Butte Cainoii site, lour exjieri- inental sets were exposed to the pollinators. Two sets had M. canlindlis as the eentcM" plant surrounded by M. lewisii plants, and two sets had M. lewisii in the eenter surrounded by M. cardiiuilis. In addition, tour eorresponding sets were plaeed in 1 X 1 X 1-ni screen cages (plastic mesh, 20 threads per inch, pore size 1x1 mm) designed to exclude poHinators. The same experimental design was repeated at the Big Cottonwood Canyon study site. The first stud)' site was in a streamside, partially shaded, maple-box elder forest; the second was in an open meadow in the aspen-spruce forest. Two contrasting sites were employed as controls in case different pollinators oc- curred in different habitats and at different elevations in the canyons. At the beginning of the experiments all cap- sules and flowers were removed. New flowers began opening the next day. The plants were observed to note pollinator visits for a total of 20 hours for each experimental set. The obser- vations were one-hoiu- periods scattered from dawn to dusk on different days. Experiments were run for one month, by which time new flowers had opened on most plants; they had been exposed to pollinators (that is, the uncaged sets); and capsules had formed and were starting to ripen. Plants were then re- turned to the greenhouse, and capsules on plants of both exposed and shielded sets were harvested as they ripened. Seeds set were not counted inasmuch as the number of seedlings produced seemed a more meaningful mea- sure of pollinator success or selfing rate. In the summer of 1984 all seeds produced by the peripheral whorl of plants in each ex- perimental set were sown together in one pot, and seeds produced by the plant in the central pot were sown in another. Resulting seedlings were scored as to whether they were of parental type, indicative of pollinator faithful- ness, or hybrids, indicative of pollinator promiscuousness, that is, pollinators visiting both species. The Fj hybrids, which have leaves intermediate in width between the broad leaves of M. cardinalis (13013) and the narrow leaves of M. lewisii (5875), can be dis- tinguished at an early stage. Nevertheless, the seedlings were grown until they flowered and exhibited either the unambiguous F^ pink color or the jiarental red (A/, ((irdiudlis) or magenta (A/, lewisii). RKsn.rs and Disci'ssion Are pollinators necessary? Results of this research indicate a resounding yes! All plants in cages set a total of only one seed that germi- nated and grew into a seedling (Table 1). It was a vigorous M. cardinalis plant from the central plant in one of the Red Butte Canyon sets. In contrast, plants in the sets exposed to pollinators produced a total of 1,535 seeds that germinated and grew into seedlings. Of these, 1,047 were M. cardinalis and 488 were M. lewisii. While there were equal numbers of plants, there were more M. cardinalis flow- ers. Hybridizations were possible in three of the eight experimental sets. The results are very clear despite the heavy depredations by deer and the lack of flowering in the other sets (Table 1). Pollinator observations revealed the pres- ence of Broad-tailed Hummingbirds and bumble bees at both sites and syrphid flies at the Red Butte Canyon site. Hummingbirds and bumble bees flew near the Mimulus plants at both sites but, surprisingly, were not observed visiting the flowers. However, in the Red Butte Canyon experiments, small syrphid flies visited both species occasionally, but not on the same foraging bout (1-5 min- utes, 1-3 flowers) nor often enough to account for the observed seed sets. There were only five total visits (at scattered times), and the only pattern revealed was that syrphids vis- ited the lower-elevation experiments of Red Butte Canyon but not the higher-elevation experiments of Big Cottonwood Canyon. The flies appeared to be foraging for pollen inas- much as they walked all over the flowers, including the anthers and pistils. Of the 1,535 seedlings produced, not one was a hybrid. This was true also in the progeny grown from plants of a natural, sympatric pop- ulation of both species in the Yosemite Valley by Hiesey et al. (1971). Apparently the polli- nators are effectively faithful to each species both in the Wasatch Mountains and the Sierra Nevada. The study raises some intriguing questions. Why were hummingbirds and bees not ob- served pollinating the flowers when the Carnegie study (Hiesey et al. 1971) showed 158 R K. ViCKERY, Jr. [Volume 50 Table 1. Seedlings produced from the seeds set by A/. carcHnalis and M. lewisii plants in Red Butte Canyon and Big Cottonwood Canyon (1) when exposed to poUinators and (2) when shielded from pollinators b\' cages. Plants were arranged in sets consisting of a center plant of one species surrounded In a whorl of six plants of the other species (see Fig. 1). Set number Composition of set Number of seedlings resulting Exposure to Shielded from pollinators pollinators 1 cardintilis 0 0 0 Q** Q** 0 0 1 0 0 Red Butte Canyon experiments #1 1 central cardinalis 0* 6 peripheral lewisii 0* #2 1 central cardinalis 0* 6 peripheral Icicisii 71 lewisii #3 1 central lewisii 190 lewisii 6 peripheral cardinalis 350 cardinalis #4 1 central lewisii 0** 6 peripheral cardinalis 420 cardinalis Total cardinalis seedlings 770 Total lewisii seedlings 261 Total F| hybrid seedlings 0 Big Cottonwood Canyon experiments #5 1 central cardinalis 184 cardinalis 6 peripheral leicisii 137 leieisii #6 1 central carf/i'nfl/i.s 93 cardinalis 6 peripheral lewisii 90 leicisii #7 1 central lewisii 0* 6 peripheral cardinalis 0* #8 1 central lewisii 0* 6 peripheral cflr(^//»i«/K 0* Total cardinalis seedlings 277 Total lewisii seedlings 227 Total F| In brid seedlings 0 Grand total cardinalis seedlings 1,047 Grand total lewisii seedlings 488 Grand total F, hybrid seedlings 0 0 0*=" 0 0 0 0 0 0 0 0 1 0 0 *Capsiiles on experimental plants eaten In ile 'Failed to (lower during experiment them to be the main polhnators of A/, cardi- nalis and M. lewisii? What would their visits show about temporal partitionini:;? Or, per- haps, morphologieal partitioning tor pollen transfer on different parts of the pollinator s body? Are there significant differences in (quantity and sugar content of the nectar pro- duced by the flowers that might affect pollina- tor preferences and visits? In conclusion, despite the questions raised for future studies, these experiments demon- strated that neither M. cardinalis nor M. lewisii self-pollinates under natural condi- tions; at least, the rate is less than .1%. Clearly, pollinators are recpiired for seed set. Only syrphid thes were observ ed actualK pol- linating the flowers, although lumnningbirds and bumble bees are probable pollinators also (Hiesey et al. 1971). The experiments showed that the pollinators (seen and unseen) are ef- fectively faithful to their own Mimulus spe- cies. So, (1) pollinators are recjuired, (2) the only observed pollinators are the small sxr- phid flies, and (3) the pollinators are effec- tively faithful to their species, either on each foraging bout or by using species-specific parts of their bodies for pollen transfer. Ac:knowledgments I thank jeri Ann Ernstsen and Stephen Sutherland for help in setting up the experi- mtMits, and Luther Light for faithfully water- ing the Silver Fork, Big Cottonwood Canyon, experiments. I am grateful to Carl Freeman and Burton I^endleton for their critical read- ing of the manuscript and percepti\e, helpful suggestions. 1990] POI.IJNATION IN MlMUlUS 159 LiTKHAruKi: Cited I'\m;(:iu K and L. \an dkk Piji. 1979. Tlu' jiriiKiplcs oi pollination. Ecolo,U\ . 3rcl cd.. vc\ . I'tiiianion Press, Oxford. 244 pp. Grant. K A., and V Grant. 1968. Hinnmingbirds and their flowers, ("oluinbia University Press. 115 pp. IIirsKV, W M., M. A. Nobs, andO Bjorkman. 1971. Ex- perimental studies on the nature of species. V. Bios\'stematics, genetics, and physiological ecol- ogy of tlie Enjtliranthc section of Minuthi.s. Carnegie Institute of Washington, Washington, D. C., Publication No. 628. 213 pp. \i( Ki'.RV. R K , Jh 1978. Case studies in the evolution of species complexes in MiDitilus. Evolntionarv Biol- ogy 1 1 : 404-506. \l< kinn \{ K , Jr . AND H L Oi.son 1956. Flower color inheritance in the Miinulus canlindlis complex. Journal of Heredity 47; 194-199. ViCKF.RV. R. K , Jr . andB M Wui.i.sikin 1987. Compari- son of six approaches to the classification oi Mitnu- his sect. Erythranthe (Scrophulariaceac). System- atic Botany 12: 339-364. Received 24J(imuin/ 1990 Revised 28 March 1990 Accepted 15 April 1990 Great B.iMM Naturalist 50121, IWX). pp l(il-165 OBSERX'ATIONS ON THE DWARF SHREW {SOHEX NANUS) IN NORTHERN ARIZONA 1 lowaid J. Hc'i 11,1 Absthact. — ()h.ser\ati()iis of 23 dwari sinews (Sorcx luimts) at Fracas Lake in Arizona cxtc-iul the ian,m- ol this uncommon shrew northward on the Kaihah Plateau and provide finther information regarding the eeologv and liahitat re(}uirements of this species. Slirews \\ ere captured in a previousl\ unreported liahitat type (Hocky Moimtain montane conifer forest; Brown 1982). This stiuK illustrates the usefidness of intensi\e, long-term studies and taunal surveys using pitfall traps. Since Meriiam discovered the dwarf shrew in 1895, it has been considered a rare species. For 70 years after it was named, Sorex nanus was known from only 18 specimens (Hoff- mann and Owen 1980). With the recent use of pitfall traps this number has increased greatly (e.g., 81 S. nanus in Colorado [Armstrong et al. 1973], 48 in Wyoming [Brown 1967], and 16 in Arizona [Marshall and Weisenberger 1971]). The dwarf shrew is one of many mammal species inhabiting an archipelago of forested montane islands in the western United States (Lomolino et al. 1989). The species is cur- rently known from Montana, Wyoming, South Dakota, Colorado, Utah, New Mexico, and Arizona (Hoffmann and Owen 1980). Sorex nanus is known from reports of only 20 specimens from three areas in Arizona (Fig. 1; Hoffineister 1986); no new Arizona localities have been reported for 15 years. The first Sorex nanus in Arizona was collected on 17 September 1937 from the Kaibab Plateau, Coconino Co., 14.5 km east of Swamp Point within the Grand Canyon National Park (GCNP, North Rim) at an ele- vation of about 2,439 m (Schellbach 1948). On 28 August 1973 another specimen was taken 5.6 km from the first record, near Kan- abownits Springs within the GCNP (Ruffher and Carothers 1975). Both areas typically con- tain mixed-conifer forest, Picea pun^ens, Picea engehnannii, Abies lasiocarpa. Abies concolor, Pseudotsuga menziesii, and Populus tremuloides (subalpine conifer forest; Brown 1982). Another specimen of the dwarf shrew was reported from the Kaibab Plateau by HoflPmeister (1955). It was found in the Kaibab Lodge, VT Ranch, when the lodge was opened in April 1944. The lodge is sur- rounded by extensive grassy meadows to the east and subalpine conifer forest to the west. A single specimen was collected on 14 August 1959 in the White Mountains of Greenlee Co., near Hannagan Meadows, in spruce-fir forest habitat (subalpine conifer forest; Brown 1982), at an elevation of 2,805 m, extending the range into the second area of the state (Bradshaw 1961). To my knowledge, there have been no recent records of Sorex nanus from this area. Marshall and Weisenberger (1971) trapped Sorex nanus in Arizona in a third area, near Flagstaff, in the Inner Basin of the San Fran- cisco Mountains at elevations between 2,865 and 3,293 m. During the summer of 1969 eight specimens were taken in rocky talus and eight from mesic subalpine meadows and surrounding spruce-fir forest. The specimens reported in this article extend the known range of Sorex nanus in Arizona northward on the Kaibab Plateau and describe a new habitat for this species in Arizona. Study Area The study area was Fracas Lake, Coconino Co., 9.6 km south-southwest of Jacob Lake (36°37'52"N, 112°14'20"W, elev. 2,514 m). Fracas Lake is a permanent, natural limestone Department ot Zoology, Arizona State University, Tempe, Arizona S.5287-1501, Present address; University of Georgia, Franklin College of Arts and Science, Department of Zoology, 724 Biological Sciences Building, Athens, Georgia 30602. 161 162 H.J Berna [\'olume 50 Colorado River Fi.^ 1 ( :olU-cti.,n sU.s oiSorcx nanus in An/.ona. Boxed area ..t state enlarsed. C.n^Unn hue outhnes ka.hah Plateau at eU^ation 2.195 ,n: A. th.s stucK : B. Hoffine.ster 1955. C:. Sehelll,aeh 194S. D Huftner and Carothers 19.o; L. San Francisco Peaks (Marsl.all and We.senberger 1971V F, llannauan Mead.ms ^Bradslum 1961). 1990] SoKtx .\ AM b IN Arizona 163 sinkhole basin less than 1 lia in area. \ar\ in an alimiinum drift fence 214.6 m in circumference. The 40-cm-high fence was buried 7-10 cm below ground level. Twci 4.5- gallon buckets were buried as pitfall traps at each of 27 stations, approximateh 7 m apart. The pitfall traps were adjacent to the fence, with one bucket on each side of the fence per station. The average distance of the drift fence to the water was S m. Pitfall traps were checked daily from 17 Ma\ to 15 September 19SS. The\ were checked e\er\ two weeks from 15 September until 19 \o\ ember 19SS and then daily from 27 .\pril 1989 until 14 September 19S9. After 14 September the\ w ere checked once e\"ery w eek until IS Octo- ber 1989. Therefore, trapping occurred for 186 days in 1988 (10,044 trap nights^ and 174 days in 1989 ^9.396 trap nights). T.\BLE L Monthly capture.s ot Sorc.v nanus at Fracas Lake. Coconino Co.. .\nzona, in 19S8and 1989. Montli 19SS 1989 April — 0 Mav 0 0 lime 1 3 Jiilv 1 3 .\ngust 6 6 Septenil l)cr 3 0 October 0 0 XoXCMll ler 0 — Standard measurements of specimens were taken, and shrews were aged b\" tooth wear J3iersing and Hoflmeister 1977) and placed into two categories. Minimum and maximiun temperatures and precipitation were monitored daily through- out the study periods. Air temperatures \aried from —9 to 42 C, and shrews were caught on e\"enings when minimum tempera- tures \ aried from 1 to 10 C. The total number of da\s with rain during the studx" in 1988 was 58. with a total rainfall of 249 mm. It rained duriuii 34 stud\ da\s in 1989, with a total rainfall of 130.5 mm. Results and Discussion 1 collected 23 Sorcx nanus during this study (Table L. Specimens were positive!) identi- fied as Sorcx )uinus based on small bod\ size dess than 52 nun', upper third unicuspid smaller than fourth, and presence of medial tines on first incisor Junge and Hoffmann 1981. Hoftmeister 1986). One specimen was kept as found in a mummified state, six were captured ali\ e and released, se\ en were pre- served as skins plus skulls, and nine were fixed in 109f formalin and later stored in 65^^ ethanol. Specimens haxe been deposited in the mammal collection at Arizona State I'ni- \ersit\ . Standard bod\" measurements, sex, and age of se\eral dwarf shrews are reported in Table 2. Some shrews were partialK eaten b\" carrion beetles in pitfall traps, making sex determination impossible. Mean body mea- surements (in mm, range in parentheses) were as follows (n = 16): total length 75 ul-77^ tail length 36 ^34-39), hindlbot 9 (8-10), ear from notch 6 (4-7). Mean body mass was equal to 2 g in = 12). Reproducti\e condition w as unambiguous for onl\ one of the 164 H J Berna [\ olume 50 Table 2. Standard measurements (length in mm, mass in g) of body size, sex, and age (Ad = adult, Juv = juvenile) of So7-ex nanus captured at Fracas Lake during 1988 and 1989. TL = total length. T = tail length, HF = hindfoot length. E = length of ear from notch. Date Age Sex TL T HF F Mass 198S 23 June Ad male 82 36 9 7 2.5 16 July Ad female 81 37 9 5 2.0 1 .August Ad male 79 36 9 7 2.0 6 .August Juv female 74 36 9 6 2.0 17 August* — 0 — — — — — 21 August Juv male 73 34 7 6 2.0 23 August Juv male 72 38 9 6 1.8 30 August Jux male 78 38 10 7 2.0 2 September'' — 0 — — — — — 7 September — ? (munnnil :ied specimen. no measurements) 1989 8 June Ad 0 76 37 10 6 — 10 June Ad female 74 38 10 i — 11 June Ad 0 76 36 9 6 — 5Julv Juv male 74 36 8 5 2.0 21 Julv Juv male 71 36 8 4 1.8 27 July Juv ? 71 35 9 "* — 10 August' — ? — — — — — 12 .\ugust'' — 0 — — — — — 21 August Juv male 72 38 10 6 2.0 25 August Tuv male 77 39 10 6 l.S 26 .\ugust Juv male ~- 3fi 10 (S \.S ■"Captured ali\e and released unharmed. 'Two live specimens caiiaht on this date and released unharmed. three female.s captured. Tlie temale captured on 16 Jul\- 1988 had three developing foUicles greater than 2 mm in diameter. All shrews were collected in pitfall traps outside the drift fence, suggesting that the shrews do not reside within the fenced area but use the area only as a foraging site. Ju\e- niles were caught in Jul\ and August (Table 2"!. and this acti\ it\ ma\ correspond to the period of juvenile dispersal. All shrews were col- lected in the morning, although pitfall traps were checked twice dail\ . Shrews were more fre(iiientl\ caught on da\s with measiuable rainfall (17 of 23. or 747c caught on da\s with rain). This ma\' be due to increased foraging range when the stress of water requirements of the shrew is lessened, or it ma\ siinpK correspond to increased prey acti\ it> . Potential prey for Sorex nanus at this local- ity include spiders, which were abundant early in the season (April-Iune); tenebrionid beetles, connnon throughout the stud\ period (May-September); and the following families and orders of in\ ertebrates that appeared in pitfall traps at \ arious times during the study: Formicidae, Carabidae, Scarabidae, C'urculi- onidae, Coccinelidae, lar\al lepidopterans. orthopterans, and a few hemipteran and ho- mopteran species. Observations of three li\e shrews caught in pitfall traps suggested that dwarf shrews axoid tenebrionid beetles. In two instances the shrew and a few tenebrionid beetles were all that remained ali\ e in the pitfall trap. Shrews seemed to favor carabid beetles, which they attacked voraciousK with bites to the head and thora.\ followed b\' a quick retreat, and then repeated until the beetle was subdued and consumed. I also observed a shrew feed- ing on ants while it was in a pitfall trap. Shrews are known to take a \ ariet\ of small \ertebrates as prey (Hoflfmeister 1986). On two occasions in this stud\ a Sorcx nanus was foimd in the same pitfall trap as an adult tiger salamander {Ambystoma tighiiuui nchulo- suin). The salamanders (snout-\ent length S5-9() mm) were unharmed on both occa- sions, but in one instance the shrew w as found dead. This may indicate an unwillingness to consume these large \ertebrates with toxic skin secretions, or it may suggest that adult salamanders are much too large to be consid- ered part of the ordinar\ diet of e\en the lumgriest Sorcx nanus. 1990] SoREX \am s IN Arizona 165 These recent observations extend the known range of So/r.v nanus nortliward on the Kaibah Plateau in Arizona 25 km From the pre\ious record ot Ihiftmeister (1955'. The habitat at this locahtx is dt)minated b\ Pinus pondcrosa. which is a new habitat t\pe recorded for this species in Arizona. In this stud\ Sorcx )ianits was collected within S m of water, unlike pre\ious records tor Arizona ,Hoffmeisterl9S6'. Future surveys using pitfall traps else- where on the Kaibab Plateau, and on other southwestern montane "islands, would be beneficial in determining the abundance and distribution of Sorex nanus in the southern portions of its range. They would also pro\ ide a clearer definition of the habitat rec}uire- ments of this species. Acknowledgments I thank T. R. Jones and B. Spicer for en- couraging this manuscript, and D. F. Hoff- meister for pro\ iding all researchers with an excellent guide to mammals in Arizona. Man\" thanks for field assistance go to C. A. Schmidt. S. J. Berna. and S. L. Steele. This article benefited from comments by J. P. Collins, C. A. Schmidt. S. George, and J. A. Junge. This research indirectly benefited from funds pro\ ided b\ the Arizona State Universitx De- partment of Zoology Graduate Student Re- search Fund, and Sigma Xi. the Scientific Re- search Society". iiloiiii a cross-sectional transect throui;h the .\r- kansas Ri\ er watershed. Colorado. Southwestern Naturalist 17: 31.5-.326. Bkkna H J I99(). First record of the ermine .Mustela crminca^ in .\rizona. Southwestern Naturalist lin press*. Bradsh AW G \" 1961. New Arizona locality for the dwarf shrew. Journal of Manmialog) 42: 96. Bkown D E 1982. Biotic communities of the .American Southwest — United States and Mexico. Desert Plants 4: 1-342. Bkown. L N 1967. Ecological distribution of si.x species of shrews and comparison of sampling methods in the central Rock> Mountains. Journal of Mammal- og\ 48: 617-623. DiERSlNG. V E . AND D F HoFFMElSTER 1977. Revision of the shrews Sorcx merricimi and a description of a new species of the subgenus Sorcx. Journal of Mammalog) 58: 321-333. Hoffmann R S . and J G Owen 1980. Sorcx tcnellus and Sorex nanus. Mammalian Species 131: 1—4. HoFFMElSTER D F 19.55. Mammals new to Grand CauNon Natiouid Park, .\rizona. Plateau 18: 1-17. 1986. Mammals of .\rizona. Universit> of.\rizona Press. Tucson. 602 pp. Ji NCE J .\ AND R S Hoffmann 1981. An annotated key to the long-tailed shrews i^genus Sorex^ of the United States and Canada, with notes on middle .American Sorex. Uni\ersit\ of Kansas Museum of Natural Histon. Occasional Papers 94: 1-4S. LoMOUNO M \" . J H Brown andR Davis 1989. Island biogeograph> of montane forest mammals in the American Southwest. Ecologx 70: 180-194. -Marshall. L G.. and G J Weisenbercer 1971. A new dwarf shrew localit\ for .Arizona. Plateau 43: 132-137. RiFFNER G .a . AND S ^^" C.\ROTHERS 1975. Recent notes on the distribution of some mammals of the Grand Canyon region. Plateau 47: 1.54-160. SCHELLB.ACH. L. III. 194S. .A record of the shrew Sorex nanus for Arizona. Journal of Mammalogx 29: 295. Literature Cited -Armstrong D M B H Banta, and E J Pokrofis. 1973. .Altitudinal distribution of small mammals Received 16 Xoveinbcr i9S9 Revised 26 January 1990 Accepted 15 April 1990 Cri-at K.isiii Naturalist 5(H2), UWO, pp. 167-172 FUNGI ASSOCIATED WITH SOILS COLLECTED BENEATH AND BETWEEN PINYON AND JUNIPER CANOPIES IN NEW MEXICO P. li. Fiesciiu'z Abstract. — The soil tiingal community beneath piuNon (Piiiiis edulis Enu;elm.) and one-seeded juniper (Junipertis luoHospenna [Engehii.] Saru;.) tree canopies is described and com]:)are(l with luufji from adjacent interspace soils dominated b\' blue grama (Boiitcloud gracilis [ H. Fi. K. ] Lag. ). Signihcanth' higher organic matter contents and fungal propaguie levels were foimd in soils beneath pinxon and juniper trees than in interspace soils. Soils under pinyon and juniper trees contained similar chemical, physical, and biological properties and, consecjucntly, many groups of fungi in common (64% of the species isolated were common to both). In contrast, soil fungi in adjacent interspace soils were \ astl\ different from those collected in soils beneath pinyon and juniper canopies (44% and 48% species in common, respectively). Soil fimgi that were isolated more often from pinyon-juniper soils than from interspace soils included Ahsidia spp., Bcauvarki spp., Gliodadium spp., A/((forspp., Pcnicilliiim cijclopiiim, P. fasciculata, P. frequentans, P. rcstricttiin. Tluinmiditim spp, and TricliDdvntui spp. Soil fungi that were isolated more often in interspace soils than in pinyon or jvmiper soils included Aspergillus (iliitaceus spp., A. fiimi^attis, some Fitsdiimu spp., Pcnicillium liitciim, and P. talaroniijci'S. Pinyon {Piniis edulis Engelm.) and one- seeded juniper ijiiniperus monospenna [En- gelm.] Sarg.) tree.s have been reported to aceinnulate significant amounts of organic matter and nutrients beneath their canopies (Barth 1980, Tiedemann 1987, Klopatek 1987). Although the soil nutrient content directly under pinyon and juniper trees is higher than in interspace soils, growth of other plant spe- cies has been reported to be severely inhib- ited. Armentrout and Pieper (1988), for exam- ple, reported that the mean basal cover of grasses located directly beneath pinyon and juniper trees was 3.7% and 0.7%, respec- tively, while adjacent interspaces contained over 10% grass cover. Low grass cover under pinyon and juniper canopies is caused by a number of foctors, including severe shading (Johnsen 1962), litter accumulation that in- hibits seed germination (Jameson 1966), in- terception of precipitation (Gilford 1970), and allelopathv effects (Jameson 1961, Lavin et al. 1968). Soil fungal populations and composition are related to soil properties as inilucnced by veg- etation (Christensen 1981). Fungal popula- tions, for example, generally increase as or- ganic matter increases (Alexander 1977). Also, Badurowa and Badura (1967) demonstrated that the composition of litter in a given plant community is the decisive factor in determin- ing the dominant groups of fungi. To date, only limited observations have been reported concerning the changes in the fungal commu- nity (i.e., vesicular-arbuscular endomycor- rhizae) associated with plant litter under pin- yon or juniper trees (Klopatek and Klopatek 1987). The objective of this study was to evalu- ate the quantitative and qualitative differ- ences in the saprophytic fungal community associated with soils collected beneath pinyon and juniper canopies and to compare them with the fungal community in adjacent blue grama {Boiitelotia graci/is)-dominated inter- space soils. Materials and Methods Four pinyon and ioiu" jimiper trees (4-6 m in height) located just east of Santa Fe, New Mexico, were randomly selected for study in August 1989. Soil samples were collected from two of the three zones described by Ar- mentrout and Pieper (1988); zone 1 was lo- cated beneath the tree canopy, and zone 2 was well outside the canopy. This soil area (inter- space) was dominated by blue grama. At each of the eight trees, four soil subsamples were Rocky Mountain Forest and Range Experiment Station, 220.5 Columbia SE, Aliniqneniue. New Mexico 87106. Present address: Los Alamos National Laboratory, K490, Los Alamos, New Mexico 87.545. 167 168 P. R. Fresquez [Volume 50 collected randomly from beneath the canopy to a depth of 15 cm with a 5-cm-diameter bucket auger. The four subsamples were mixed (composited) to represent one sample per tree. Four subsamples were also ran- domly collected and composited from inter- space soils occurring adjacent to the tree spe- cies. Thus, a total of 16 composite samples were transported in an ice cooler back to the laboratory. At the laboratory, the samples were passed through a 2-mm sieve and stored at 4 C prior to fungal analysis. An aliquot of approximately 500 grams from each soil sam- ple was taken for laboratory chemical (soluble Na, Ca, Mg, and K; available P; total Kjeldahl N [TKN]; organic matter [OM]; pH; and elec- trical conductivity [EC]) and physical (per- cent sand, silt, and clay) analysis at the New Mexico State University Soil and Water Test- ing Laboratory. All methods of chemical and physical analyses have been described previ- ously (Fresquez and Lindemann 1983). Numbers of soil fungal propagules were es- timated by the dilution and plating technique described by Wollum (1982). For population estimates, dilutions were plated in triplicate on rose bengal-streptomycin agar (Martin 1950) and incubated at 27 C for seven days. The numbers of fungal propagules per gram of oven-dry soil are reported. Fmigal groups were isolated by placing 1 mL oi a 10 ' dilu- tion from a 10-g over-dry weight equivalent soil sample in a petri dish, adding cooled rose bengal-streptomycin agar, and swirling for an even distribution. Ten plates were inoculated for each composited soil sample and incu- bated at 27 C for seven days (Fresquez and King 1989). After incubation, a portion of every colony was transferred to a carrot agar medium by removing a portion oi the agar containing hyphal tips. The colony was subcultured on a carrot agar plate to allow for maximum fruiting potential and identifica- tion. After a four-day incubation period, the colonies were identified using the taxonomic guides of Barnett and Hunter (1972), Barron (1968), Oilman (1968), and Domsch et al. (1980). Diversity of fiuigi may give some insight into the physiochemical conditions of the soil environment, in terms of identifying more productive soils (Fres(|uez et al. 1990) and/or soils that are physiochemically stressed (Fres- quez etal. 1986). Thus, the diversity of fungal groups among the treatments was estimated using Shannon s index of species diversity, H (Zar 1974): k H = - S Pi log p, i= 1 where p, ^ the proportion of fungal group / in the sample, and k ^ the number of groups. The corresponding test for evenness is where //„ sitv. the maximum possible diver- An estimate of the similarity in the fungal composition among the sample populations was calculated with Sorensen's presence com- munity coefficient (SPCC), which was de- scribed bv Mueller-Dombois and Ellenberg (1974) as ' SPCC - 200 C /(A + B) where C is the total number of groups com- mon to two habitats, A is the total number of groups in sample A, and B is the total number of groups in sample B. If the same groups were found in both samples, then the commu- nity coefficient would be 100; if the sample had no groups in common, the coefficient would be 0. Variations in soil chemical properties and in the populations of soil fungi between trees and adjacent interspaces were analyzed using paired t tests at the .05 level. Unpaired t tests were used to compare soil properties and fun- gal population means between the two tree species at the .05 level. Results and Discussion Significantly higher fungal propagules were found in soils from beneath pinyon and ju- niper canopies than from interspace soils (Table 1). Other soil fungal groups, such as vesicular-arbuscular endomycorrhizae, have been reported to be higher in soils from pinyon and juniper trees than from interspace soils (Klopatc^k and Klopatek 1987). The highei- numbiM- of soil fungal propagules from pinyon and juniper trees was probabK due to diffei-ences in organic matter (OM) contents, as signihcantly higher soil OM levels were found under canopies of both pin\ on and ju- niper trees than in interspace soils (Table 2). 19901 FuNci FHOM PiNYON-ji'Nii'KH Soils 169 'I'ahi.K 1. Soil lungal propajjiilt-s and Sorciiscn s prtsiiitc (.■oiiniiiiiuly coclliiiciils associated uitli a ])iiiy()ii-jniiipfr plant foiniiiiiiiit)' in Now Mexico. Soil site SimilaritN coeflicii'iits Fim.ual |)roi)amiles ( ■ 10') Piinon UiulerstoiN Interspace Inniper Understorv Interspace Pin\()n undi'rstor\ SlSa'A' (123) interspace 43b (17) luniper iniderstorv 393\ A (119) interspace 39z (15) 44 64 35 38 56 48 'Means within the same column and tree species tollowed by the same letter are not sii;intieaiitl\ dillerent at the ,05 level usiny paired ( tests (standard deviation). "Means within the same column followed by the same uppercase letter are not sit;nilicantl\ diilerent .it the ,0.5 level using unpaired t tests. Tablf-: 2. Soil (sand\- loam) chemical properties associated with a pinyon-jnniper plant commnnit\' in New Mexico. Solul)le c ations Phosphorus ; uid nitroi ien Soil (u.ti.U ') (UKR ') OM' EC' (gkg ') (dSm ') pH site Na Ca Msj; K P NHj N NO3-N TKN' Pinvon understorv 6a-A^ 188aA 23aA 32aA 9aA 3. 7a A 9.0aA 2143aA 49a A 0.83aA 7.4aB (1.8) (18.8) (4.9) (5.08) (3.63) (1.74) (5.85) (649) (13.3) (0.15) (0.17) interspace 12a 62b lib lb 3b 1.6a 5.0a 525b 14b 0.44b 7.3a (8.2) (17.5) (5.1) (0.27) (0.83) (0.85) (0.56) (271) (2.8) (0.08) (0.22) Juniper understorv UvA 238vA 35vA 45vA 13yA 2.8vA 10.6vA 1989vA 49vA 1.13vA 7.8yA (2.8) (24.1) (9.5) (3.94) (8.50) (1.68) (3.82) (300) (4.7) (0.21) (0.21) interspace 9y 76z Hz 3z 4v 2.0v 5.6v 598z 16z 0.44z 7.4z (4.2) (13.6) (2.1) (0.86) (0.66) (0.67) (0.74) (208) (2.5) (0.07) (0. 16) 'TKN (total Kjeldahl nitrogen), OM (organic matter), EC (electrical conductivity). -Means within the same column and tree species followed by the same letter are not significantK different at the ,05 level using paired ( tests deviation). ■'Means within the same column followed by the same uppercase letter are not significantK different at the 05 level using unpaired ( tests. (standard Soils with high OM usually have higher Ringal populations than soils with low OM concen- trations (Alexander 1977). Moreover, the low fungal populations in the soil from the inter- spaces may be attributed not only to low OM content but also to the more severe environ- ment at the surface of the more exposed inter- spaces. For example, lower amounts of plant litter mulch lead to higher amounts of exposed soil, higher surface temperatures, and lower soil moisture contents that reduce soil fungal populations (Wicklow 1973). Other soil properties, such as Ca, Mg, K, and TKN, were significantly higher under pinyon and juniper trees than in interspace soils (Table 2). Many studies have shown that soils under pinyon and juniper canopies con- tain higher soil nutrient levels, including N, than soils in adjacent interspaces (Barth 1980, Tiedemann 1987, Thran and Everett 1987, Klopatek 1987). Soil NH4-N and NO3-N levels were higher, although not significantly different, in pinyon and juniper soils than in interspace soils. In any case, the conversion of NH4-N to NO3-N (nitrification) does not ap- pear to be inhibited. These data agree with Klopatek and Klopatek (1987), who found that nitrification, despite low nitrifier counts, oc- curred under pinyon and juniper trees. Thus, since soil fungi were found in significantly higher populations directly under pinyon and juniper trees than in interspace soils (Table 1), and since nitrifying bacteria have been re- ported to be partially inhibited by allelopathic substances directly beneath pinyon and ju- niper canopies (Klopatek and Klopatek 1987), soil NO3— N produced under pinyon and ju- niper trees as compared to interspace soils 170 P R. Fresquez [Volume 50 may be more a result of heterotrophic rather than autotrophic nitrification. Some soil lungi have been shown to produce si 1 4 lutciiin 1 49 2 5 nionovei't 0 11 29 8 niiXi'icdiis 0 0 25 0 rcslrictnni 74 30 13 0 tdldronn/ccs 0 (i 0 6 r. sp. 1 0 0 ■''2 0 r. sp. 2 4 0 0 0 /'. sp. 3 0 1 0 I Tlidinnidiuni 3 0 2 0 Trichndcinid 50 0 2 6 Unidentified 1 2 0 0 No. ot isolates 26S 213 33,3 107 No. oi groups 21 20 2(S 16 Diversit\ 0.97 1.04 1.16 1. 14 E.Ncnness 0.73 O.SO ().S2 0.95 was dillerent Iroin that of interspace soils. Piu\()n and juniper soils, for example, con- tained oiiK 449f and 48%, respectively, of 1990] FuNC'.i I'HOM PiNYON-jUNiPKH Soils 171 all species in cominoii with the interspace soils. Soil fnnual oiuanisnis isolated more of- ten from piinon and jnniper soils than from interspace soils were Ahsidia spp., Bcauvaria spp., Gliodadium spp., A/uro/- spp., Peiiicil- liuni cijclopium, P. fasciculafa, P. jvequen- tdiis, P. rcstrictintK Tlidninidiuiit spp., and Trichodcnna spp. Nh)st of these fnn,u;al organ- isms are common soil saproplntic tnngi. Bcauvana is an insect parasite (Moore-Lan- decker 1972) and is probably associated with the hitiher microarthropod popnlation inhab- iting the litter la\ er under pin\ on and juniper canopies (Whitford 1987). In contrast, fungal organisms isolated more often in blue grama- dominated interspace soils than in pinyon or juniper soils were Aspcr'^illus alutaceiis spp., A. fumi^atus, some Fiisariuin spp., Penicil- lium lutciim. and P. talaromyces. Aspergillus fumigatus and especially Fitsariiim spp. are characteristic of grassland soils (Christensen 1981) and blue grama-dominated soils in the semiarid Southwest (Fresquez et al. 1988). Literature Cited ALE.XANDER. M 1977. Introduction to soil niici()liiolog\ . John Wiley, New York. Armentrout, S. M.. and R D Pieper 1988. Plant distri- bution surrounding Rocky Moinitain pin\on pine and one-seed juniper in south-central New Mex- ico. Journal of Range Management 41; 1.39-14.3. Arnold, J F . D A Jameson, and E H Reid 1964. The pinyon-juniper type of Arizona: effects of grazing, fire, and tree control. USDA, Production Re- search Report. 84 pp. Badurowa, M , AND L Baduka 1967, Further investiga- tions on the relationship between soil fungi and the macroflora. Acta Societies Botanicorum Pola- nia 35: 515-529, Barnett. H. L., and B B Hunter 1972. Illustrated genera of imperfect Fiuigi. Burgess, Minneapolis, Minnesota. Barron. G L 1968. The genera of Ihphoniycetes from soil. Williams and Wilkins, Baltimore, Maryland. Barth, R. C. 1980. Influence of pinyon pine trees on soil chemical and physical properties. Soil Science Society of America Journal 44: 112-114. Chrlstensen. M. 1981, Species diversity and dominance in fungal communities. Pages 201-232 in D, T, Wicklowand G, C, Carroll, eds,. The fimgal com- munity. Dekker, New York. Dennis, G. L., and P R Fresquez. 1988. The soil micro- bial community in a sewage-sludge-amended semi-arid grassland. Biologv and Fertilitv of Soils 7:310-317. DoMSCH, K. H., W Gams, and T 11 .Anderson 1980, Compendium of soil fungi. Academic Press, New York. Doxtader, K G \nd M ,\i,i;\\\i)EH 1966. Role of .3-nitropr()panoic acid in nitrate lormation l)y AsjX'r^ilhis fhivus. lournal nl Baclcriolog\ 91: 1186-1191. l3()\r\i)i;R K (i \\i).\ I) Box ira 1968. Nitrification by :\.sj)cr^illii.s Ihit us in a sterilized soil. Australian journal of Soil Rcsearcli 6: 141-147. JMUisguEZ, P R , E. F Ai.don, and W C Lindemann. 1986, Microbial re-establishment and the diver- sity of finigal genera in reclaimed coal mine spoils and soils. Reclamation and Revegetation Research 4: 245-258, Fresquez. P. R. AND (; L Dennis 1990. (,'omposilioii ol fimgal groups associated with sewage sludge auRMuled grassland soils. Arid Soil Research and Rehabilitation 4: 19-32. Fresquez, P R.. R E. Francis, andG L Dennis 1988, Fungal communities associated with phyto- edaphic communities in the semiarid Southwest, Arid Soil Research and Rehal)ilitation 2: 187-202. Fresquez, P R . and R King 1989. Number of fungal isolates required to describe species differences on reclaimed coal mine areas in New Mexico. USDA Forest Service Research Note RM-491. Fresquez. P R. and W C. Lindemann 1983. Green- house and laboratory e\ aluations of amended coal- mine spoils. Reclamation and Revegetation Re- search 2: 205-215. GiFFORD. G F 1970, Some water movement patterns over and through pinyon-juniper litter, Joiunal of Range Management 23: 365-366, Gilman. J C. 1968. A manual of soil Fungi. Iowa State L'niversity Press, Ames. Jameson. D A 1961. Growth inhibitors in native plants of Arizona. USDA Forest Service Research Note RM-61. 1966. Pinyon-juniper litter reduces growth of blue grama. Journal of Range Management 19: 214-217. JoHNSEN.T. N. 1962. One-seed juniper invasion of north- ern Arizona grasslands. Ecological Monographs 32: 187-207. ' Klopatek. C, C, AND J. M Klop.'\tek. 1987. Mycorrhizae, microbes and nutrient cycling processes in pinyon-jimiper systems. Pages 360-362 in Pro- ceedings— pinvon-juniper conference, Reno, Ne- vada, 13-16 January 1986. USDA Forest Service General Technical Report INT-215. Klop.^tek, J M 1987, Nutrient patterns and succession in pinyon-juniper ecosystems of northern Arizona, Pages .391 -.396 in Proceedings — pinxon-juniper conference, Reno, Nevada, 13-16 January 1986. L^SDA Forest Service General Technical Report INT-215. L,\viN, F . D A Jameson, and F. B. Gomm. 1968, Juniper extract and deficient aeration affects germination of six range species, [ournal of Range Management 21:262-263. Martin, J P. 1950. Use of acid, rose bengal and strepto- mycin in tfie plant count method for estimating soil fungi. Soil Science 69: 215-233. Moore-Lan DECKER. E 1972. Fundamentals of the Fungi. Prentice-Hall, Englewood Cliffs, New- Jersey. 172 P. R. Fresquez [Volume 50 Thran, D. F., and R. L. Everett 1987. Impact of tree harvest on soil nutrients accumulated under singleleaf pinyon. Pages 387-390 in Proceed- ings— pinyon-juniper conference, Reno, Nevada, 13-16 January 1986. USDA Forest Service Gen- eral Technical Report INT-21.5. Whitford, W G 1987. Soil fauna as regulators of decom- position and nitrogen cycling in pinyon-juniper ecosystems. Pages 365-368 in Proceedings— pinyon-juniper conference, Reno, Nevada, 13-16 January 1986. USDA Forest Service General Technical Report INT-21.5. WiCKLOu, D T 1973. Microfungal populations in surface soils of manipulated prairie stands. Ecologv 54- 1302-1310. WOLLUM, A. G. 1982. Gultural methods for soil micro- organisms. Pages 781-802 in Methods of soil analysis. American Society of Agronomy, Madi- son, Wisconsin. Zar, J H 1974. Biostatistical analysis. Prentice-Hall Englewood Cliffs, New Jersey. Received 2 November 1989 Revised 22 February 1990 Accepted 5 April 1990 C;tval H.isiH Nalviialist 5()i2!, 1>W0, pp 173-179 EFFECTS OF DWARF MISTLETOE ON (;UOVVTH AND MORTALTIY OF DOUGLAS-FIR IN THE SOUTHWEST KoliiTt L. Matliiascn', Frank C. Hawkswortlr, arid (^ailcton B. Edminster^ Abstkact. — The effects ol clwari mistletoe (Arcetithohiuiii (l()ii 16.0 inches) (Table 1). Little if any effect of dwarf mistletoe was found in DMR classes 1 or 2; overall, the percent reduction was about 176 R. L. Mathiasen etal. [Volume 50 Table 2. Ten-year periodic annual volume increment (PAI) bv dwarf mistletoe rating (DMR) and diameter class for Douglas-fir (1980 and 1981 data onK 0. Diameter class (inches) Small Large Poles sawtimber savvtimber (6.1-10.0) (10.1-16.0) (16.0-^) Table 3. Total cubic foot volume, infected live vol- ume, and dead volume for Douglas-fir by stand dwarf mistletoe rating (DMR) class (1980 and 1981 data only). DMR PAF 0.11 A'' 0..35A 0.84 A 0 N 1,.391 1,148 699 DMR PAI 0.12A 0.34A 0.83A 1 N^ 166 173 134 % change + 9 -3 -1 DMR PAI 0.12A 0.32A 0.81A 2 N 143 110 109 % change + 9 -9 -4 DMR PAI 0. lOA 0..33A 0.75B 3 N 197 172 167 % change -9 -6 -11 DMR PAI 0.()8B 0.24B 0.67B 4 N 141 139 122 % change -27 -31 -20 DMR PAI 0.07B 0.18C 0.45C 5 N 148 192 140 % change -36 -49 -46 DMR PAI 0.05C 0.13D 0.28D 6 N 296 229 105 % change -.55 -63 -67 ■^Ten-Near periodic annual increment (cubic ieet/\'ear), trees lart^er than 6.0 inches dbh only. Numbers followed by different letters are significantly different within each diameter class; oneway ANOVA. p .05, Student-Newman-Kuels. 'Percent change from DMR 0 (PAD 10% for DMR class 3, 30% for class 4, 50% for class 5, and 60% for class 6. Ten-year periodic annual (cubic) voliune in- crement (Hann and Bare 1978) was deter- mined by DMR class for pole-size trees (di- ameter ranges as above), small sawtimber, and large sawtimber size classes (Table 2). Percent change in 10-year periodic annual volume increment shows a pattern similar to that for 10-year periodic radial growth. Much variation was encountered for all size classes in DMR classes 1 and 2 but much less in classes 3-6. Class 3, 4, 5, and 6 trees had decreases in periodic annual volume of 6-11%, 20-31%, 36-49%, and 55-67%, re- spectively, when compared with the growth of healthy trees of the same size classes (Table 2). However, when a one-way analysis of vari- ance (p = .05, Student-Newman-Kuels) was applied to the results, only large sawtimber- size trees with DMR greater than 2 and trees with DMR greater than 3 for the smaller size classes had statisticallv significant xohmie Percent Stand Number Total of live Percent DMR of cubic volume volume class plots feet/acre infected dead 0 105 5,. 590 0 1.3 0.1- -0.5 68 9,1.30 23 2.5 0.6- -1.0 37 6,6.30 56 2.6 1.1- -1.5 20 9,060 61 2.6 1.6- -2.0 29 6,660 80 3.4 2.1- -2.5 26 8,830 91 3.6 2.6- -3.0 32 5,.300 94 4.6 3. 1- -3.5 27 7,680 93 5.0 3.6- -4.0 22 9,. 560 95 4.1 >4.0 21 5,8.50 99 3.8 growth reductions when compared with the growth of healthy trees (Table 2). Percent re- ductions in 10-year periodic annual volume increment for all sawtimber-size trees aver- aged approximately 10%, 25%, 45%, and 65% for infection classes 3, 4, 5, and 6, respec- tively. Although site index affected the rate of 10- year periodic annual volume increment (lower sites had lower growth rates), the re- ductions in volume increment associated with dwarf mistletoe infection followed the same pattern as above regardless of the site index class considered. Consistent differences in the growth loss patterns associated with dwarf mistletoe infection have been demonstrated for different habitat types using the data col- lected in this study (Mathiasen and Blake 1984). However, no differences in growth loss could be associated with other stand at- tributes such as slope, aspect, or elevation based on these data. Total cubic volume, infected live volume, and dead volume were calculated for trees larger than 6.0 inches dbh by stand DMR classes (Table 3). The xoliune of dead trees doubled in plots with stand DMR of 0.1-1.5 and was two and one-half to almost four times greater in plots with stand DMR greater than 1.5 when compared with the percentage of dead voliune in healthy plots. Percent mortalitv was calculated tor the fol- lowing size classes: small saplings (dbh 0.1- 1.0 inch), large saplings (dbh 1. 1-6.0 inches), poles, small sawtimber, and large sawtimber (diameter ranges as above) by stand DMR classes (Table 4). Percent mortality ranged 1990] Effects OF DwAHiM IS ri.KTOF. 177 Tabi.k 4. Percent mortality of Douglas-fir by stand tKsart inistlctoe rating (DMR) and diameter class. Diameter class (inches) Stand DM H class Seedlings (0.1-1.0) Saplings (1.1-0.0) Poles (0.1-10.0) Small saw tiniher Large savvtimhei (10.1-16.0) (16.0+) 0 0.1-1.0 1.1-2.0 2.1-3.0 3.1-4.0 4.0+ .\11 plots 10 (i IS 19 24 10 12 19 17 26 30 44 20 6 10 13 16 15 35 12 2 5 7 14 13 22 3 3 5 12 21 16 6 T.-^BLE 5. Percent mortalit\ b\- stand histor\- and diameter class for Douglas-fir Diameter class (inches) Stand histor\- 0.1-1 Cutover (vears) 12-20 5 21-30 5 >30 15 \'irgin 11 1.1-6.0 6.1-10.0 10.1-16.0 16.0 17 19 18 23 15 9 10 15 9 4 6 10 4 3 10 8 Table 6. Percentage of dead Douglas-fir with dwarf mistletoe ratings 0 and 2-6 by diameter class. Total dead Dwarf mistletoe rating Diameter class 0'' 2 3 4 5 6 (inches) trees Percentage of dead Douglas-fir 0.1-1.0 443 70 4 4 7 3 12 1.1-5.0 1,238 68 3 3 6 5 15 5.1-10.0 354 58 1 2 6 12 21 10.1-16.0 227 49 1 1 9 14 26 16.0+ 125 63 1 1 11 15 9 Total 2,387 65 3 3 7 / 15 ■■Does nut include dwarf mistletoe ratings of 1 because this class nicluded dead trees tliat could not be assigned an accurate rating- Trees with ain indication of past mistletoe intection that could not be accurately rated were assigned a 1 to indicate they had been infected. May include infected old dead trees not having signs of past dwarf mistletoe infection. This category includes any mortality observed in plots that was probably not related to mistletoe infection. from 1.3% to 5.0% but generally increased as stand DMR increased, particularly when the stand DMR was greater than 2.0. Mortality of small sawtimber demonstrated the largest increase in percent mortality as stand DMR increased (from 2% in healthy plots to 22% in plots with a stand DMR greater than 4.0). The percentage of dead trees was greatest for large saplings, compared with other size classes, in all stand DMR classes. Mortality in most size classes was generally greatest in virgin stands compared to cut- over stands (Table 5), but mortality in stands cutover more than 30 years prior to data col- lection was higher than in virgin stands for the small sapling and large sawtimber-size classes. Forty-three percent of the dead Douglas-fir that could be accurately assigned a dwarf mistletoe rating were rated as class 6 (Table 6). Most infected dead trees in each size class were rated as class 6 except for the large saw- timber, where a higher percentage of dead Douglas-fir were rated as class 5. More dead trees in the small and large sapling-size classes were rated as class 4 than class 5. The percent- 178 R, L. Mathiasen etal. fX'olume 50 age of dead Douglas-fir rated as classes 2 and 3 decreased as size class increased (Table 6). Discussion The relationship between percentage of trees infected and stand DMR is similar to that reported for stands infected with lodge- pole pine dwarf mistletoe (A. amehcanum Nutt. ex Engelm.) (Hawksworth 1978). Severe dwarf mistletoe infection greatly re- duces volume increment of Douglas-fir in the Southwest. Ten-year annual radial growth re- ductions for trees in DMR classes 3, 4, 5, and 6 averaged about 10%. 30%, 50%, and 60%, respectively. Statistically significant growth losses occur for sawtimber-size trees with in- fection levels greater than 3. Pierce (1960) in western Montana and Shea (1963) in Oregon also showed that growth rates of severely in- fected Douglas-fir were markedly reduced. In addition, Filip et al. (1990) found significant reductions in 10-year mean diameter incre- ment in dwarf mistletoe-infested Douglas-fir stands in eastern Oregon and Washington. Wicker and Hawksworth (1988) gave general loss estimates for growth reduction for all dwarf mistletoes as about 10%, 25%>, and 50% or more for trees in classes 4, 5, and 6, respectively. However, our results indicate that growth reductions for Douglas-fir in the Southwest are greater than these general esti- mates. Our estimates of mortality in Douglas-lir dwarf mistletoe-infested stands are similar to those reported in the Southwest by Hawks- worth and Lusher (1956) and Andrews and Daniels (1960). Although not tested statisti- cally, mortality was generally higher in mistletoe-infested virgin stands than in cut- over stands in this study. Hawksworth and Lusher (1956) reported similar findings for Douglas-fir stands in southern New Mexico, but Andrews and Daniels (1960) found higher mortality rates in cutover stands. Increases in mortality are generally related to increases in stand DMR. Mortality was highest for large saplings in each stand DMR class. The reasons for the higher mortality rate in the large sapling class are unknown, but they may be related to more severe competition for light, moisture, and nutrients, combined with in- creased stress related to mistletoe infection. Small saplings are subjected to severe compe- tition, too, but usualK have lower levels of mistletoe infection (Mathiasen 1986). The high mortality rates we observed in class 4 and 5 trees for Douglas-fir are in contrast to mor- tality patterns in mistletoe-infected pon- derosa pines, where mortality is predomi- nantlv in class 6 trees (Hawksworth and Lusher 1956). Douglas-fir dwarf mistletoe is widespread and common in southwestern mixed-conifer forests (Andrews and Daniels 1960, Hawks- worth and Wiens 1972, Jones 1974, Gottfried and Embry 1977). This study demonstrates that the damage caused by Douglas-fir dwarf mistletoe in unmanaged forests can be signifi- cant in terms of increased mortality and re- duced growth of Douglas-fir in heavily in- fested stands. Therefore, reducing population levels of this parasite through sil\ icultural management should be a high priority for re- source managers in the Southwest. Acknowledgments We are grateful to Jerome Beatty, Dick Bas- sett, Greg Filip, and John Schwandt, US DA Forest Service, for review of the original manuscript. We also acknowledge the many indix'iduals who assisted with field data collec- tion, particularh' Da\ id Conklin and Kathr\'n Kennedx . The stud\' was funded b> the Rock\' Mountain Forest and Range Experiment Sta- tion and the Universitv of Arizona. Literature Cited Andrews. S R., and J. P. Daniels I960. A survey of dwaifniistletoes in Arizona and New Mexico. US DA Forest Service, Rocky Mountain Forest and Range Experiment Station Paper 49. 17 pp. Dooi.iNc;, O J . R R Johnson, and R G Eder 1986. CJrowtli, impact, spread, and intensification of dwarf mistletoe in Douglas-fir and lodgepole pine in Montana. USDA Forest Ser\ice, Northern Re- gion, Poorest Pest Management Report 6-6. 11 pp. Fdminster. C. B., andF". G. Hawksworth 1984. Model- ing growth and yield of southwestern mixed conifer stands including effects of dwarf mistletoe. Pages 5-11 ;;i Proceedings. 32nd Western Inter- national Forest Disease Work (conference, Taos, New Mexico, 2.5-28 September. Fdminster. C B . and L H. Jump. 1976. Site index curves for Douglas-fir in New Mexico. USDA Forest Ser- vice Research Note RM-326. 3 pp. FONUNSTER. C. B . H T MOWRER. R L M.\THIASEN. AND 1'^ G Hawksworth 1990. GENGYM; a variable density stand table projection system calibrated 1990 J Effechs oi D\\ AHF Mistletoe 179 for inixc-d conifer stands in the Soiitliucst. I'SDA Forest Service Hesearcli l^apei- (in press). Fiiir C M J .1 Cloi.HKin. C C Sii\\\ ill \\ F. Hess- lu H(. K I' llosMW AM)(; A i'MiKs 1990. Some ii'lations ainoni^ dwari mistletoe, western sprnce hndworm, and Douglas-fir: modeling; and man- afjement implieatit)ns. In: Proceedings of tlie Sym- posiinn on Interior Douglas-fir: the species and its management, Spokane, Washington, 27 F'ehru- ar>-lMarch 1990 (in ])ress). FiLlP. G M.,.andC a Parks 1987. Simultaneous infesta- tion by dwarf mistletoe and western spruce bud- worm decreases growth of Donglas-fir in the Blue Moimtains of Oregon. F'orest Science 33: 767- 773. CorrFHlEi:). C J , and R S Embkv 1977. Distribution of Douglas-fir and ponderosa pine dwarf mistletoe in a virgin Arizona mixed conifer stand. USDA Forest Ser\ ice Research Paper RM-192. 16 pp. Graham, D P 1961. Dwarfmistletoe of Douglas- fir. USDA Forest Service, Forest Pest Leaflet 54, 4 pp. Haglund, S a., andO J DooLiNC 1972. Observations on the impact of dwarf mistletoe on Douglas-fir in western Montana. USDA Forest Service, North- ern Region, Forest Insect and Disease Report D-72-1. 6 pp. Hann, D. W , AND B B Bare 1978. Comprehensive tree volume equations for major species of New Mexico and Arizona: I. Results and methodology. USDA F'orest Service Research Paper INT-209. 43 pp. Hawkswurth. F G 1961. Dwarfmistletoe of ponderosa pine in the Southwest. USDA F'orest Service Technical Bulletin 1246. 112 pp. 1977. The 6-class dwarfmistletoe rating system. USDA Forest Service General Technical Report RM-48. 7 pp. 1978. Intermediate cuttings in mistletoe-infested lodgepole pine and southwestern ponderosa pine stands. Pages 86-92 in USDA Forest Service Gen- eral Technical Report PSW-31. Hawksworth, F. G , and A A Lusher 1956. Dwarf mistletoe survey and control on the Mescalero Apache Reservation, New Mexico. Journal of Forestr\ 54: 384-390. II \w kswoiirii V G WD D VVikns 1972. Biology and classification oi dwarf niistletoes (Arceuthobiiim). USDA Agriculture Ilamibook 401. 234 pp. JoNl'.s, J R. 1974. Sii\iculturi' of southwestern Tiiixed conifers and aspen: the status of Our knowledge. USDA Forest Serxice Research Paper RM-122. 44 pp. Maiiiiasen. R L f986. inlection oivoung Douglas-firs and spruces by dwarf niistletoes in the Southwest. Great Basin Naturalist 46: 528-534. Mathiasen, R L , AND E. A Beake 1984. Relationships I)etween dwarf mistletoes and habitat types in western coniferous forests. Pages 111-116 in USDA Forest Service General Technical Report RM-111. MoiK, \V II., AND J. A. LuDWic 1979. A classification of spruce-fir and mixed conifer iiabitat types of Ari- zona and New Mexico. USDA Forest Service Re- search Paper RM-207. 47 pp. Pearson, G. A. 1950. Management of ponderosa pine in the Southwest. USDA Forest Service Agriculture Monograph 6. 218 pp. PlEHCE, VV R 1960. Dwarfmistletoe and its effect upon the larch and Donglas-fir of western Montana. Montana State Universitv, School of F'orestrv Bul- letin 10. 38 pp. Shea. K R 1963. Diameter increment in old-growth Douglas-fir infected by dwarf mistletoe. Weyer- hauser Company, Forestry Research Note 50. 11 pp. USDA Forest Service 1962. Annual report, 1961. In- termountain Poorest and Range Experiment Sta- tion. 46 pp. Wicker, E. F., and F G. Hawksworth 1988. Relation- ships of dwarf mistletoes and intermediate stand cultural practices in the Northern Rockies. Pages 298-302 in USDA Forest Service General Techni- cal Report INT-243. Received 7 Febrnari/ 1990 Revised 28 March 1990 Accepted 22 April 1990 Great Basin Naturalist 5()i2'. I99(), pp IM- 191 USING THE ORIGINAL LAND SURVEY NOTES TO REGONSTRUCT PRESETTLEMENT LANDSCAPES IN THE AMERIGAN WEST S. M. Calatowitsch' Abstract. — Rectan5j;ular survexs toinpleted between 1796 and 1925 1)\ the General Land Ortice have tretjuently been used in the eastern and central U.S. to determine land cover prior to European settlement. These survey notes are less often used in the western U.S., although they are the only site-specific presettlement records available in many areas. Recent efforts to restore riparian and grassland habitats recjuire an understanding of the conditions of these sites before settlement. General Land Office Sur\e\ notes provide a description of each township, including water supplies, timber resources, and agricultmal potential. The width and course of rivers and streams were recorded on survey lines, along with notes on topography, vegetation, wetlands, mineral deposits, and soils. The township and section descriptions ma\' be used with other historic information to reconstruct presettlement landscapes. Incomplete or vague descriptions, land use before sur\e\\ bias in recording data, and contract fraud limit the usefulness of some survey notes. However, survey notes have proved useful in establishing baseline conditions of riparian habitats in Colorado and Oregon and grasslands in Colorado and New Mexico. Information from historic photographs, ex- pedition journals, and original land survey notes have been used to reconstruct vegeta- tion at the time of European settlement (Hutchison 1988, Noss 1984). The General Land Office (GLO) notes have been consid- ered the most reliable source of historic land- scape data because of standardized data col- lection and systematic coverage of most of the United States (e.g., Bourdo 1956). Many of the published studies using survey notes de- scribed regional patterns in upland forests of the north central and northeast U.S. (e.g., Grimm 1984, Gottam 1949). Land survey notes ha\ e been used to assist in determining fire return intervals (Lorimer 1977), to sub- stantiate early explorers records (Grimm 1984, Rankin and Davis 1971), and to assess range trends (BuflFington and Herbal 1965). Few studies have used earlier metes and bounds survey notes available in the eastern states for vegetation characterization because of the lack of standardized data (Siccama 1971). Use of survey notes for site-specific studies, especially in the landscape of the American West, has not been evaluated. This review discusses the methods used by field survey crews, limitations of interpreting sur- vey data, and site-specific applications in the western U.S. General Land Office Surveys Surveys east of Ohio were conducted at the local political level and did not use standard- ized techniques (Siccama 1971). The rectan- gular survey was initiated at the western boundary of Penns\l\ania when the Land Ordinance of 1785 was passed by Gongress. The Northwest Ordinance of 1787 encour- aged the rapid settlement of new territories and states, creating the need for surveys. The Office of Surveyor General was created by the Land Act of 1796, when public lands were offered for disposal and further escalated the need tor surve\ s. Several configurations of the rectangular surve\" were used between 1785 and 1796. EventualK- the survey was stan- dardized to partition the land into townships ot thirty-six square miles that included one- mile-square sections (Fig. 1). Townships were aligned along north-south principal meridians and east-west baselines. The General Land Office was formed in 1812 to oversee the national survey. Surveys were contracted to the lowest bidder until 1908 (Senti 1988, personal communication), the sur\ e> or being compensated for each mile of line completed while also being account- able for errors in the survey. The contract holder hired the sur\'ey crew. Although each Colorado Natural .\reas Program. Colorado Department of Natural Resources, Den\ Iowa State University, Ames, Iowa .50010 Colorado 80203 Present address; Department of Botany, 181 182 SM Galatowitsch [Volume 50 RECTANGULAR SURVEY SYSTEM TOWNSHIPS T1S R2E NW 1/4 SW 1/4 NE 1/4 NW1/4 SE 1/4 SE1/4 NE1/4 SE1/4 SW1/4 SE1/4 SE1/4 6 5 4 3 2 1 7 8 9 10 11 12 18 17 16 15 14 13 19 20 21 22 23 24 29 28 27 26 25 31 32 33 34 35 36 SECTIONS SUBDIVISIONS Fig. 1. The rectangular survey partitioned land into townships of thirt\-six square miles that included one-niile- square sections. Section suhdixisions were generalK' not sur\ e\ etl during the original fieldwork. crew member and the surveyor took oatlis to perform their duties faithfully (Cazier 1976), frequent fraudulent sinveys caused the Gen- eral Land Office to abandon contracting in 1908. Since 1908, salaried federal emploNces have conducted the surveys. The surveys notes were transfered to each state as the survey was completed, but rec- ords for states with incomplete surveys in 1925 were retained by the General Land Of- fice. The Office of Surveyor General was abol- ished in 1925 when most of the suitable public land had been surveved, and duties were then reassigned within the General Land Offices. However, some remote areas were not sur- veyed by that time. In addition, privately held Spanish Land Grants, common in the south- western U.S., were never part of the public domain lands of the United States and were not included in the rectangular survey sys- tem. Areas rich in locatable minerals, such as gold, siKer, and lead, were usually not suit- al)le for agricultural use and often were not surveyed in the rectangular survey system. Mining claims cotild be located on mineral dc^posits uiuUm- the General Mining Law of 1990] Si'H\F,v Notes in thk Amkhican West 183 ND I — ^U^ 925 1861-1908 \^ MN A^ZX^ -^ • 1857- J ^^^/^ notes can onl\' represent data from a point in time rather than a "pristine baseline. The effects of pre-European land use should be considered in many parts of the western U.S. Bias in vegetation description. — Bias in field notes for forest studies has been well documented. Bearing trees were selected to be easily relocated, not necessarily the closest to the section or half-section post. Bearing trees were selected b\' size, age, species longevity, distance from the corner, and con- spicuousness (Grimm 1984). Statistical analy- sis of (}uantitati\'e tree data from the field notes is flawed because certain sizes and spe- cies were favored and because the sample is systematic, not random (Grimm 1984). Bias in vegetation descriptions of nonforested habi- tats is difficult to assess. General instructions to surveyors re(|uired information on avail- able forage; thus, descriptions of shrubs and forbs may be underrepresented. Errors in species identification. — Spe- cies identifications are not standardized among surveys. For example, "bunch grass 1990] Sum i:y Noiks in riiK American Wfst 187 Tahi.K 1. Plant names iiscil in tlic li'tritorial sur\(\ oi New Mexico ((iross 1973) Name used in IS.SO (Common name tocla\' SeicTitilie Tiam Buffalo grass Buneli grass (Jrama grass Sand grass Salt grass Buekhrusii t^liami/.a Chieo Greasewood Labina Locust Manzanita Sabinos Sage Cedar Oak Pine Pinon Bulhiio grass Bine grama Salt grass Deer biier Four-wing saltbiiish Iodine bush Greasewood Locust Sage Juniper Gambel s oak Ponderosa pine Pinon Ihichldc (Itiityloidcs lioulcloiia ' Mountain states (Colorado, Wyoniing, Montana) have an at- tenuated fauna of spider wasps. For the most part, this is a result of the fact that until recently there had been little systematic col- lecting in these states. In this report, distribu- tion records are presented for 25 species and subspecies not previously reported from these states. The following list includes records not reported by Townes (1957) or Evans (1950, 1951) and not reported or implied in the Cata- log of Hymenoptera north of Mexico (Krom- bein et al. 1979). Arrangement follows the catalog. Specimens are in the collections of Colorado State University, Fort Collins (CSU), or the University of Colorado, Boulder (UC), except as othei-wise noted. Cryptocheilus hesperus (Banks). CO: 1 9, Otero Co., Higbee, 14 June 1966 (pit trap, J. Brookhart) [UC]; WY: 1 9, Fremont Co., 3 mi E Moneta, 7 July 1963 (B. Vogel) [UC]. Not previously reported east of Utah. Priocnemis minorata Banks. CO: 1 9 , Larimer Co., 13 mi W Livermore, 12 June 1987 (H. E. Evans) [CSU], Previously re- ported from forests east of the 100th meridian. Priocnemis kevini Wasbauer. CO: 2 6 6, Larimer Co., 13 mi W Livermore, July, Sept 1983, 1987 (malaise trap, H. E. Evans) [CSU]. Described from Idaho, with a paratype from Michigan, by Wasbauer (1986); identification confirmed by Wasbauer. Priocnemis scitula relicta Banks. CO: 1 9, 1 (5, Larimer Co., on alfalfa, 9 Aug 1985 (F. Peairs) [CSU]; 1 6, Larimer Co., Fort Collins, 4 July 1982 (W. J. Pulawski) [Calif. Acad. Sci.]. Not previously reported west of Wisconsin. Calicurgus hijalinatiis excoctus Townes. CO: 1 9, Larimer Co., 13 mi W Livermore, 23 July 1985 (H. E. Evans) [CSU]; 3 9 9, Larimer Co., Hewlett Gulch, Aug-Sept 1978 (H. E. Evans) [CSU]. Previously recorded only from New Mexico and Arizona. Dipogon lignicolus Evans. CO: 5 9 9, 2 6 6 , Larimer Co., 13 mi W Livermore, April 1986, 1989 (reared from trap nests containing fragments of prey, salticid spiders, probably Phklippus species; H. E. Evans) [Mus. Comp. Zool., CSU]. Described bv Evans, 1987. Dipogon sericeus Banks. CO: 4 9 9, 1 6, Larimer Co., Hewlett Gulch, Aug, Sept 1987 (malaise trap, H. E. Evans) [CSU]. Previously reported only from Oregon and California. Dipogon iracundus Townes. CO: 1 9, Boulder Co., Nederland, 25 Aug 1961 (U. N. Lanham) [UC]. Previously recorded only from Arizona. Auplopus mellipes variitarsatus (Dalla Torre). CO: 2 9 9, Larimer Co., Fort Collins, 20 June 1986, 15 July 1987 (black light trap, W. Cranshaw) [CSU]. Widely distributed in northeastern states. Ageniella rufescens (Banks). CO: 1 9 , Boul- der Co., Hygiene (J. Polhemus) [UC]; 1 9, Larimer Co., 20 mi N Fort Collins, 30 Aug 1974 (H. E. Evans) [CSU]. WY: 2 9 9, Platte Co., Glendo, 22 Aug 1957 (D. R. Tyndall) [Univ. Wyoming]. Previously reported from Kansas, Texas, New Mexico, and Arizona. Aporus luxus (Banks). CO: 6 6 6, Mon- tezuma Co., Arriola, Sept 1975 (malaise trap, T. Marquardt) [CSU]. A West Coast species; also reported from Utah and New Mexico. Psorthaspis nigriceps (Banks). CO: 1 9, Montrose Co., Uravan, 28 Aug 1947 (H. G. Rodeck) [UC]. Previously known only from Department of Entomology, Colorado State University, Fort Collins, Colorado 80523. 193 194 Notes [Volume 50 New Mexico, Arizona, and Utah. Evagetes c. crassicornis (Shuckard). CO: 1 9, Larimer Co., Fort Collins, 21 May 1982 (H. E. Evans) [CSU]; 1 9, Larimer Co., 13 mi W Livermore, 20 July 1989 (H. E. Evans) [CSU]. WY: 1 9, Platte Co., Wheatland, 24 Aug 1974 [Univ. Wyoming]. A Canadian Zone subspecies. E. crassicornis consimilis (Banks) is widely distributed in Colorado and Wyoming. Sericopompilus neotropicalis (Cameron). CO: 1 9, Otero Co., Hawley, 8 Aug 1978 (H. E. Evans) [CSU]. Occurs in Kansas but otherwise not previously reported north of Texas and Arizona. Episyron quinqiienotatus hurdi Evans. CO: 1 9, 25 6 6, Alamosa Co., San Luis Lake, 19 Aug 1981 (H. E. Evans) [CSU]; 2 9 9, Mesa Co., Fruita, 21 May 1963 (B. Vogel) [UC]. Not previouslv reported east of Utah. Tachypompilus u. unicolor (Banks). CO: 1 6 , Larimer Co. , 20 mi N Fort Collins, 14 July 1975 (H. E. Evans) [CSU]. I have also seen 1 9 from Badlands National Monument, SD, 24 July 1953 (F. Rindge) [Amer. Mus. Nat. Hist.]. Previously reported from the West Coast states east to LUah, Wvoming, and Idaho. Tachypompilus unicolor cerinus Evans. CO: 1 9, 1 c5. Bent Co., Hastv, 1 July 1982 (W. J. Pulawski) [Calif. Acad. Sci.]; 3 9 9, 3 6 6, Bent Co., 23 mi S Las Animas, 10 Aug 1986 (H. E. Evans) [CSU]. Previously re- ported from LItah, New Mexico, and Texas. Anoplius lepidiis atramentarius (Dahl- bom). CO: 1 (5, Baca Co., Two Buttes Reser- voir, 9 July 1986 (H. E. Evans) [CSU]; 1 9, Larimer Co., 13 mi W Livermore, 27 Julv 1989 (H. E. Evans) [CSU]. Occurs primari- ly east of the lOOtli meridian but extending westward. Anoplius acapulcoensis (Cameron). CO: 2 9 9, 5 c5 c5, Prowers Co., 10 mi W Lamar, 19 July 1974 (H. E. Evans) [CSU]; 1 9,3 6 6, Bent Co., Hasty, 26 June 1974 (H. E. Evans) [CSU]; 2 6 6, Cheyenne Co., Aroya, 25 June 1974 (H. E. Evans) [CSU]. These represent the northernmost records for this species, which ranges widely in Mexico. Anoplius percitus E\ans. (X): 1 9, Douglas Co., Castle Rock, 12 Aug 1962 (S. M. Sutton) [UC]; 1 9, Boulder Co., 4 mi NE Lvons, 15 Sept 1962 (U. N. Lanham) [UC]; 19,16, Larimer Co., 13 mi W Livermore, July, Aug 1986, 1989 (H. E. Evans); 4 9 9, 6 66, Larimer Co., Fort Collins and vicinitv, June-Aug 1974, 1977, 1985 [CSU]; 1 9 , Eagle Co., Dotsero, 21 June 1977 (H. E. Evans) [CSU]. An eastern species not previously re- ported from Colorado. Evans (1970) reported it from Teton Co., Wyoming. Pompilus silvivagus Evans. MT: 1 9, Car- bon Co., 5 mi NE Belfry, 24 Aug 1965 (M. Vogel) [UC]. Considerably north of the known range for this species. Minagenia congrua (Cresson). CO: 3 6 6, Larimer Co., Fort Collins, July 1978, 1989 (H. E. Evans) [CSU]. Not previously recorded west of Michigan. Minagenia montisdorsa Dreisbach. CO: 12 6 6, Bent Co., Hasty, 17 July 1974 (malaise trap, H. E. Evans) [CSU]. Previously re- ported west of the 100th meridian from Texas, Arizona, and Mexico. Ceropales elegans aquilonia Townes. CO: 1 6 , Larimer Co. , 14 mi N Fort Collins, 7 fulv 1974 (H. E. Evans) [CSU]; 3 66, Weld Co.', 12 mi NE Nunn, July 1982, 1987 (H. E. Evans) [CSU]. Previously recorded from Minnesota and Alberta. C. c. elegans Cresson is more widespread in Colorado. Ceropales r. robinsonii Cresson. CO: 3 9 9, 1 d Larimer Co., 13 mi W Livermore, 7-28 July 1985, 1989 (H. E. Evans) [CSU]. Not previously recorded west of Illinois, al- though C robinsonii stigmatica Banks is known from Kansas. In summary, range extensions for spider wasps in the U.S. are verified for the following areas: eight species from the East, eight from the Southwest, five from the Pacific Coast, and three from the North. The remaining spe- cies is presently known only from Colorado. LiTERATU HE Cited Enans, H E 1950-1951. A taxonomic stucK of thr Nearc- tic spider wasps belonging to the tribe Poiiipilini (Hymenoptera: Pompilidae). Transaetions of the American Entomological Societ\ 75: 133-270, 76: 207-361,77:203-340. 1970. Ecological-beha\ ioral studies ofthe wasps of Jackson Hole, Wyoming. Bulletin of the Museum of Compaiatixf Zoologx , IlarxanI l'ni\ersit\ 140: 451-511. 1987. The genus Di])05()m high. Small caves and solution pockets were common, and perma- nent seeps contributed to a perennial water supply. Pellets were cleaned in 2% aqueous solution of sodium hydroxide to permit identi- fication of skeletal contents (Longland 1985). Unpublished data were obtained from E. D. Forsman (personal communication), who collected 409 prey items in summer 1977 from pellets of 2 pairs of Spotted Owls in the Chiricahua Mountains (Fig. 1). For compari- son, J. L. Ganey's (1988 and personal commu- nication) data are included here on bat speci- mens among 1,193 prey items collected from 29 pairs of Spotted Owls throughout Arizona during 1984-1987. Pellets that we collected contained skeletal remains of 39 white-footed mice (Peromysciis spp.), 34 woodrats {Neotoma spp.), 1 cotton- tail {Sylvilagus sp.), 3 Northern Pygmy Owls {Glaucidium gnoma), 1 White-throated Swift {Aerotiautes saxatalis), 1 unknown bird, 1 mountain spiny lizard {Sceloporus jarrovi), and 11 bats (Table 1). Bats comprised 12% of the total prey items. Thirty-five bats, 8.6% of prey items, were identified in Forsman s (personal communica- tion) sample from the Chiricahua Mountains (Table 1). Ganey (1988) listed 24 bats from Spotted Owl pellets (11 from pellets in south- eastern Arizona, Table 1) representing 8% of prey items in southern Arizona but only 2% of total prey items of these owls statewide. The presence of three species of molossids, Tadarida hrasiliensis, T. macrotis, and Eu- mops perotis (Table 1), in pellets provides new records for two of the mountain ranges (see Hoffmeister 1986). A band and skeletal remains of one Eptesi- ctis fusctis (a juvenile male loanded by R. Department of Ecology and Evolutionary Biology. University ot Arizona, Tucson. Arizona 85721 , "Present address; 607 North Sixth Avenue, Tucson, Arizona 85705. 197 198 Notes [Volume 50 O BANDED • RECOVERED MOUNTAIN RANGE ARIZO Pima County SANTA RITA MTS Cochise County | NA ' CHIRICAHUA Santa Cruz ^. County 1^^ Tombstone HUACHUCA MTS MTS NEW'MEXICO 0 100 1 ' ' I I kilometers MEXICO Fig. 1. Iluachuca Mountains and surrounding area. Mountain ranges depict relative area above 1,525 m. Locations ol banding and recovery sites in Cochise Co. , Arizona, are shown for a big brown bat found in a spotted owl pellet. Sidner and R. Davis on 10 July 1988 at a ma- ternity roost near Tombstone, Arizona) were found in a sample of pellets (dated 7-24 February 1989) from the Huachuea Moun- tains (Fig. 1). The specimen was found 44 km southwest of the banding site (occupied by bats from May to September only) and is the only off-site recovery of 1,535 banded E. fiis- cus. Because mean home range of mated pairs of Mexican Spotted Owls in northern Arizona is only 847 ha (Ganey and Balda 1989b), recov- ery of this bat in a Spotted Owl pellet pr()\ ides information about natural mortality and win- ter dispersal of £. fiiscus. It may also suggest the presence of a hibernaculiun in the Hua- chuea Mountains where none is known for this species (Hoffmeister 1986). Ruprecht (1979) proposed that a high per- centage of bat remains occurs among prey items when owl territories overlap home ranges of bats. Pellet analyses have show n that barn owls roosting in the same building \\ itii E. fiiscus consumed a high percentage of these bats (Kunz 1974), while Long-eared Owls (Asio otus) roosting in an isolated patch of trees among sand dunes caught only one bat (Antrozous pallidus) and 1,365 rodents (Kotler 1985, personal communication). The highest percentage of bats as pre\' of Spotted Owls was found in oiu" winter sample and reflects the abundance of bat species pre- sumed to have winter ranges in southeastern Arizona (Hoffmeister 1986). In this study, bats contributed little to total prey biomass of Spotted Owls and simply may have been taken opportunistically. The three species of bats, E. fitscus, A. pallidus. and T. brasiliensis\ that were the most munerous in pellets are relatively abundant, colonial species. Spotted Owls normalK^ employ a sit-and- wait (perch-and-poimce) himting strategy (Forsman et al. 1984) and are thus unlikely to pursue bats in fhght. Mexican Spotted Owls roost and forage in forest adjacent to steep- sided canxons (Ganex' and Balda 1989a, 1989b), which proxide cool, shaded roosts in tre(>s, cliff ledges, and caves (also used by bats). Owls max' take active bats entering or exiting roosts or torpid bats from the interior of roosts (Beer 1953). .\11 bats found in Spotted Owl pellets thus far are species that become 1990] Notes 199 TvBi I 1 . Bat spt'cics t'roni spotted owl pellets in southeastern Arizona. Species are listed for mountain ran^t^s from wlutii tiic eontrilnitors eolleeted pellets. New speeies records for a mountain ranue are indicated hy *. Nmnhers in ixncntheses are indi\ idual hats recorded. Species Mijotis spp. Myotis cdlifoniicus (California nnotis^ Mijotis ciliolahniin (western small-footed myotis) Laskmycteris noctivagani! (silver-haired hat) Lasiuni.s cine reus (lioar\' hat) Pipistrellus Iwsperus (western pipistrelle) Eptesictts fusciis (bighrown hat) Antruzoiis paUidus (pallid hat) Tadarida spp. Tada rida brasiliensis (Brazilian free-tailed hat) Tadarida femorosacca (pocketed free-tailed hat) Tadarida macrotis (hig free-tailed hat) Eumops perotis (western mastiffhat) Unidentified hat Chiricaluias Forsman (3) Caney (1) Forsman (1) Ganey (1) Forsman (8), Ganey (2) Forsman (9), Ganey (1) Forsman (I) Forsman (3), Ganey (2) * Forsman (2) Forsman (4) * Ganey (1) Forsman (4) Mountain range Santa liitas Ganey (1) lluaclmcas this study (1) Ganey (1), this study (1) Ganey (1), this study (7) nhis studv (2) torpid during roosting. Bats ol this type may benefit beyond energy conservation by select- ing darker or less accessible roost sites (Erkert 1982). By comparison, bats tliat remain alert may stay in the outer, lighted portions of roosts without excessive risk. Two species of bats, Sanborn's long-nosed bat {Leptotiyctcris sanborni) and the Mexican long-tongued bat (Choeromjctcris mexicana ), occur in the three mountain ranges where pellets were collected but do not use torpor or hibernation. Each species hangs alert, making use of hghted por- tions of roosts (HofFmeister 1986) and has not been reported from owl pellets collected dur- ing any season in Arizona. Oin- findings demonstrate tliat in Arizona, Mexican Spotted Owls utilize a wide verte- brate prey base, suggesting opportunistic for- aging as occurs in the northern subspecies (Forsman et al. 1984). Diets may contain a considerable diversity of bats (Tal)le 1), which may be an important component of the winter diet of individual Mexican Spotted Owls in southeastern Arizona. A(;knc)\vlp:dgments We gratefully acknowledge E. D. Forsman (USES) and J. L. Ganey (Northern Arizona Universitv), who generously provided their unpublished data. M. A. Bogan (USEWS) identified bat skulls in the Eorsman sample. We thank R. Davis, E. D. Eorsman, J. L. Ganey, K. C. Green, B. J. Verts, the Eort Huachuca Game Management Branch, R. T. Smith, and G. Ernstein for various assistance. Litkhati'RkGited Allen, G M 1939. Bats. Harvard University Press, Cambridge, Massachusetts. 368 pp. Akmstkonc, D M . AND J K Jones, Jr. 1972. Notiosorex crawfordi. Mammalian Species 17; 1-5. Barrows, C. 1987. Diet shifts in breeding and noni)reed- ing spotted owls. Journal of Raptor Research 21: 95-97. 200 Notes [Volume 50 Beer. J R 1953. The screech owl a.s a predator on tlie big brown bat. Journal of Mammalogy 34: 384. Brown, D. E , ED. 1982. Biotic communities of the Ameri- can Southwest — United States and Mexico. Desert Plants 4: 1-342. Erkert, H G 1982. Ecological aspects of bat activity rhythms. Pages 201-242 mT. H. Kunz, ed.. Ecol- ogy of bats. Plenum Press, New York. 425 pp. FoRSMAN, E. D . E C Meslow, and H M Wight 1984. Distribution and biology of the spotted owl in Oregon. Journal of Wildlife Management 48(87) (Wildlife Monographs suppl.): 1-64. Ganey. J. L 1988. Distribution and habitat ecology of Mexican Spotted Owls in Arizona. Unpublished thesis. Northern Arizona University, Flagstaff. 229 pp. Ganey, J. L.. and R P Balda 1989a. Distribution and habitat use of Mexican Spotted Owls in Arizona. Condor 91: 355-361. 1989b. Home range characteristics of spotted owls in northern Arizona. Journal of Wildlife Manage- ment 53: 1159-1165. Gillette, D D. and J D Kimbrough 1970. Chirop- teran mortality. Pages 262-283 ;/i B. H. Slaughter and D. W. Walton, eds.. About bats, achiropteran biology symposium. Southern Methodist Uni\er- sity Press, Dallas, Texas. 339 pp. HOFFMEISTER. D F 1986. Mammals of Arizona. Univer- sity of Arizona Press and Arizona Game and Fish Department. 602 pp. JOHNSGARD, P A 1988. North American owls: biology and natural history. Smithsonian Institute Press, Washington, D.C.'295pp. KOTLER, B P 1985. Owl predation on desert rodents which differ in morphology and behavior. Journal of Mammalogy 66: 824-828. KuNZ. T H 1974. Reproduction, growth, and mortality of the vespertilionid bat, Eptesicus ftisctts. in Kan- sas. Journal of Mammalogy 55: 1-13. Long. C A , and W C Kerfoot 1963. Mammalian re- mains from owl pellets in eastern Wyoming. Jour- nal of Mammalogy 44: 129-131. LONGLAND, W S 1985. Comments on preparing owl pel- lets bv boiling in NaOH. Journal of Field Ornithol- ogy 56: 277. Marshall, J T , Jr 1942. Food and habitat of the spotted owl. Condor 44: 66-67. Marti. C D 1974. Feeding ecology of four sympatric owls. Condor 76: 45-61. RUPREGHT, A J. 1979. Bats (Chiroptera) as constituents of the food of barn owls, Tyto alha. in Poland. Ibis 121:489-495. Wagner. P.W.CD Marti. andTC Boner 1982. Food of the spotted owl in Utah. Journal of Raptor Re- search 16: 27-28. Received 1 December 1989 Revised 12 Fchruanj 1990 Accepted 22 March 1990 INFORMATION FOR AUTHORS The Great Basin Naturalist welcomes previously unpublished manuscripts pertaining to the biologi- cal natural history of western North America. Pref- erence will be given to concise manuscripts of up to 12,000 words. MANUSCRIPTS SHOULD BE SUBMITTED to James R. Barnes, Editor, Great Basin Naturalist, 290 MLBM, Brigham Young University, Provo, Utah 84602. A cover letter accompanying the manu- script must include phone number(s) of the author submitting the manuscript; it must also provide information describing the extent to which data, text, or illustrations have been used in other papers or books that are published, in press, submitted, or soon to be submitted elsewhere. Authors should adhere to the following guidelines; manuscripts not so prepared may be returned for revision. CONSULT THE MOST RECENT ISSUE of this journal for general style and format. Also refer to the CBE Style Manual, 5th edition (Council of Biology Edi- tors, 9650 Rockville Pike, Bethesda, MD 20814 USA; $24). TYPE AND DOUBLE SPACE all materials, including hterature cited, table headings, and figure legends. Avoid hyphenated words at right-hand margins. Underline words to be printed in italics. Use stan- dard bond (22 x 28 cm), leaving 2.5-cm margins on all sides. SUBMIT 3 COPIES of the manuscript. Number all pages and assemble each copy separately: title page, abstract and key words, text, acknowledg- ments, literature cited, appendices, tables, figure legends, figures. TITLE PAGE includes an informative title no lon- ger than 15 words, names and addresses of authors, a running head of fewer than 40 letters and spaces, footnotes to indicate change of address and author to whom correspondence should be addressed if other than the first author. ABSTRACT states the purpose, methods, results, and conclusions of the research. It is followed by 6-12 key words, listed in order of decreasing im- portance, to be used for indexing. TEXT has centered main headings printed in all capital letters; second-level headings are centered in upper- and lower-case letters; third-level head- ings begin paragraphs. VOUCHER SPECIMENS. Authors are encouraged to designate, properly prepare, label, and deposit high-quality voucher specimens and cultures docu- menting their research in an established perma- nent collection, and to cite the repository in publi- cation. REFERENCES IN THE TEXT are cited by author and date: e.g., Martin (1989) or (Martin 1989). Multiple citations should be separated by commas and listed in chronological order. Use "et al." after name of first author for citations having more than two au- thors. ACKNOWLEDGMENTS, under a centered main heading, include special publication numbers when appropriate. LITERATURE CITED, also under a centered main heading, lists references alphabetically in the fol- lowing formats: Mack, G. D., and L. D. Flake. 1980. Habitat rela- tionships of waterfowl broods on South Da- kota stock ponds. Journal of Wildlife Man- agement 44: 695-700. Sousa, W. P. 1985. Disturbance and patch dynam- ics on rocky intertidal shores. Pages 101-124 in S. T. A. Pickett and P. S. White, eds. , The ecology of natural disturbance and patch dy- namics. Academic Press, New York. Coulson, R. N., and J. A. Witter. 1984. Forest entomology: ecology and management. John Wiley and Sons, Inc., New York. 669 pp. TABLES are double spaced on separate sheets and designed to fit the width of either a single column or a page. Use lower-case letters to indicate foot- notes. PHOTOCOPIES OF FIGURES are submitted ini- tially with the manuscript; editors may suggest changes. Lettering on figures should be large enough to withstand reduction to one- or two- column width. Originals must be no larger than 22 X 28 cm. NOTES. If your manuscript would be more ap- propriate as a short communication or note, follow the above instructions but do not include an ab- stract. A CHARGE of $45 per page is made for articles published; the rate for subscribers will be $40 per page. However, manuscripts with complex tables and/or numerous half-tones will be assessed an additional charge. Reprints may be purchased at the time of publication (an order form is sent with the proofs). FINAL CHECK: • Cover letter explains any duplication of infor- mation and provides phone number(s) • 3 copies of the manuscript • Conformity with instructions • Photocopies of illustrations GREAT BASIN NATURALIST v., a. no .'Z',^ CONTENTS Articles Mayfly^ gmwth^ and population density in constant and variable temperature '^^'"'^■^ Russell B. Rader and James V. Ward 97 Black-tailed prairie'dog populations one year after treatment with rodenticides • • ^"^'^"">' ^ AP^^' I^^^"i^l W. Uresk, and Raymond L. Linder 107 Effects of burning and clippmg on fix e bunchgrasses in eastern Oregon ^^"'^^^^^ ^^ ^'■>«""' Guy R. McPherson, and Forrest A. Sneva 115 Fohage biomass and cover relationships between tree- and shrub-domn.ated com- munities in pmyon-juniper woodlands R. j. Tausch and P T Tueller 121 TaxoiKm.y and variation of the LopUlea ni.ruUa complex of western North America (Heteroptera: Miridae, Orthotylinae) ^^.^^ ^^^^^;^ ,3^ Pollination experiments in the Mumdus carclinalis-M. letcmi complex Robert K. Vickerv, Jr. 155 Observations on the dwarf shrew (Sorex nanus ) in northern Arizona Howard J. Berna 161 Fungi associated with soils collected beneath and between pinvon and .uniper canopies m New Mexico p u ' ''""1^^' ^^ r . K. Jh resquez 167 Effects of dwarf „mtle.„e <„, growth and ,n«nali,> of Douglas-fir in the Sonthwest . Robert L. Ma.hia.sen, Frank G. Hawksworth, and Carleton B. Edminster 173 ""'AnreZrWe:;'' '"''"' """■' '" '■"""*■"'■' "-«■"'--" l-Kl«-apes in the S. M. Galatowitsch 181 Notes ""'" tTT" ;"""^^ "^^ •^^"''"" '"''''' (H>--'-Ptera, Pompilidae) from the Rocky Mountam states Lvard E. Eva." 193 Bats in Spotted Owl pellets in southern Arizona Russell B. Duncan and Ronnie Sidner 197 H E DEC » GREAT BASIN TY NiffURAUST VOLUME 50 NO 3 - OCTOBER 1990 BRIGHAM YOUNG UNIVERSITY GREAT BASIN NATURALIST Editor James R. Barnes 290 MLBM Brigham Young University Provo, Utah 84602 Associate Editors Michael A. Bowers Blandy Experimental Farm University of Virginia Box 175 Boyce, Virginia 22620 PaulC Marsh Center for Environmental Studies Arizona State University Tempe, Arizona 85287 Jeanne C. Chambers US DA Forest Service Research 860 North 12th East Logan, Utah 84322-8000 Brian A. Maurer Department of Zoology Brigham Young University Provo, Utah 84602 Jeffrey R. Johansen Department of Biology John Carroll University Cleveland, Ohio 44118 JimmieR. Parrish Department of Zoology Brigham Young University Provo, Utah 84602 Other Associate Editors are in the process of being selected. Editorial Board. Richard W. Baumann, Chairman, Zoology; H. Duane Smith, Zoology; Clayton M. White, Zoology; Jerran T. Flinders, Botany and Range Science; William Hess, Botany and Range Science. All are at Brigham Young University. Ex Officio Editorial Board members include Clayton S. Huber, Dean, College of Biological and Agricultural Sciences; Norman A. Darais, University Editor, University Publications; James R. Barnes, Editor, Great Basin Naturalist. The Great Basin Naturalist, founded in 1939, is published quarterly by Brigham Young University. Unpublished manuscripts that further our biological understanding of the Great Basin and surrounding areas in western North America are accepted for publication. Subscriptions. Annual subscriptions to the Great Basin Naturalist for 1990 are $25 for individual subscribers, $15 for student and emeritus subscriptions, and $40 for institutions (outside the United States, $30, $20, and $45, respectively). The price of single issues is $12. All back issues are in print and available for sale. All matters pertaining to subscriptions, back issues, or other business should be directed to the Editor, Great Basin Naturalist, 290 MLBM, Brigham Young University, Provo, UT 84602. Scholarly Exchanges. Libraries or other organizations interested in obtaining the Great Basin Naturalist through a continuing exchange of scholarly publications should contact the Exchange Librarian, Harold B. Lee Library, Brigham Young University, Provo, UT 84602. Editorial Production Staff JoAnne Abel Technical Editor Carolyn Backman Assistant to the Editor Heidi Larsen Production Assistant ISSN 017-3614 11-9065048375 The Great Basin Naturalist Published AT Provo, Utah, by Bricham Youncj University ISSN 0017-3614 Volume 50 31 October 1990 No. 3 SPROUTING AND SEEDLING ESTABLISHMENT IN PLAINS SILVER SAGEBRUSH {ARTEMISIA CAN A PURSH. SSP. CAN A) C. L. Wamholt', T. P. Walton', and R. S. White" Abstract. — The importance and nature of vegetative reproduction was compared with seedhng estabUshment in plains silver sagebrush (Aftcniisia cana Pursh. ssp. cuna). Sixty-three percent of plants excavated originated from rhizomes. Sites that experienced habitat disturbance did not have a significantly different number of plants originating from vegetative reproduction than did undisturbed sites. Parent rhizomes were significantly older than taproots, which were significantly older than aboveground stems. Rhizome systems were spread 3.3 times that of plant height. Seventy-nine percent of rhizomatous daughter plants were 100 cm or less from parent plants. Lip to 52 sprouts were found on one rhizome. Seedling establishment was greatest during wet growing season.s, and vegetative reproduction was greatest during dry years. Sagebrush {Artemisia L.) taxa are among the most important plants on rangelands of the western United States (Beetle 1977). Plains silver sagebrush {Artemisia cana Pursh. ssp. cana) is a major consideration in the management of rangelands in the north- ern Great Plains. This is due to the taxon s competitive nature with livestock forages and its importance as a habitat component for several wildlife species. Together with the other two subspecies of silver sagebrush, mountain silver sagebrush (A. cana ssp. vis- cidula [Osterhout]) and Bolander silver sage- brush (A. cana ssp. holanderi [Gray] Ward), this complex is encountered on millions of hectares in 13 western states and 2 Canadian provinces (Harvey 1981). The literature (Young and Evans 1972, Bostock and Benton 1979, Went 1979) pro- vides contradictory evidence as to whether seed or vegetative reproduction is more im- portant for survival of plants displaying both habits in arid and semiarid environments. Accordingly, this will have to be determined for each taxon individually. The most perni- cious weeds generally grow from under- ground roots, rhizomes, and buds (Cook 1983); thus, these are important traits to un- derstand in successful rangeland taxa. Para- doxically, vegetative reproduction in sage- brush taxa has not been previously studied in detail despite the importance of these taxa (Beetle 1977, McArthur and Plummer 1978) and their obvious reproductive success (Har- vey 1981). Understanding sagebrush repro- ductive success would provide insight into plant population dynamics throughout west- ern North America (Mott 1979). Our objective was to assess the importance and nature of vegetative reproduction (sprouting) versus seedling establishment in plains silver sage- brush. Study Sites Description Six study sites were selected in drainages of the Tongue and Yellowstone rivers near Miles City in southeastern Montana. The sites Department of Animal and Range Sciences, Montana Slate University, Bozeman, Montana 59717. "USD.A, Agricultural Research Service, Fort Keogh Livestock and Range Research Station, Miles City, Montana .59.301. 201 202 C. L. Wamboltetal. [Volume 50 were considered typical in climatic and edaphic relationships of plains silver sage- brush habitats in the northern Great Plains (Harvey 1981). Soils at each study site are texturally heterogeneous but are largely a mo- saic of loams, with silty clay loams predomi- nant. Three sites (Yellowstone River — frigid Ustic Torrifluvents, Lower Black Springs — mixed [calcareous] frigid Ustic Torrifluvents, and Lower Flood — fine montmorillonitic Bor- ollic Camborthids) had experienced fire or ice scraping and shearing in the preceding five years. No evidence of such disturbance was found at the remaining study locations (Lig- nite Creek — Caniborthid Torrifluvents, Paddy Faye — fine montmorillonitic Borollic Cam- borthids, and Moon Creek — fine montmoril- lonitic Borollic Natriargids). All sites have received periodic cattle grazing. The area has an average annual precipitation of 340 mm, with peak precipitation received in May and June. Methods Transect Excavations At each study site a plains silver sagebrush plant 16 to 40 cm in height was located at each 5-m interval along 25-m transects (4). Estab- lished plants of this size were selected be- cause rhizomatous connections to parent plants, if present, were still readily apparent. The 20 plants located at each site were exca- vated to determine whether they had sexual or asexual origins. Roots were carefully excavated b>' hand so that fragile rhizome connections remained in- tact to determine if plants were of indepen- dent origin or connected to another plant. Rhizomes were generally found in the top 10 cm of soil, while taproots were excavated to a depth of 1 m or an impenetrable layer. Plants without connecting rhizomes were consid- ered to have originated from seed. Plant height, length of rhizomes, and stem and rhi- zome diameters were measured on all origi- nally located plants and those plants to which they were directly connected. Root distribu- tion from each excavation was mapped within a grid, and line sketches of each plant were drawn. Samples for age determination (Fer- guson 1964) were taken from stem, root, rhi- zome, and connecting rhizome sections of each excavated plant. Aging of sagebrush was feasible despite the difficulty created by com- mon stem splitting and lavering (Ferguson 1964). To determine whether differences existed in the number of sprouts to seedlings over the six study sites, we conducted a paired Stu- dent's t test (P < .025). A chi-square analysis (P < .05) was performed to learn if fire or ice action on three of the sites was significant in determining the ratio of sprouts to seed- lings compared with three undisturbed sites. Analysis of variance (ANOVA) was used for comparison of age and growth means among plant parts. Duncan's multiple range test pro- tected by a prior F-test was used for compar- ing treatment means. Isolated Plant Excavations Two large, well-established plants at the Lower Flood site were the subject of a com- plete root excavation. Plants were subjec- ti\ eh' selected based upon two criteria: (1) the plant had to be relatively isolated from other large plains silver sagebrush plants to mini- mize major competitive influences, and (2) there had to be an abundance of small plains silver sagebrush plants surrounding the po- tential parent plant. The two plants selected for excavation were slightK' more than 1 m in height. A 5-m area around each of the large plants was excavated so that all roots, including rhizomes of all smaller plants, were exposed. The size of each plant and the root distribution from the excavations were mapped to differentiate seedlings and sprouts. Although we did not analyze the data statistically, our direct observation of the root networks facilitated interpretation of the transect data. Results Transect Excavations Plants arising from rhizomes were more abundant than those that grew from seedlings (P < .025) (Tal)le 1). Approximately 63% of the excavated plants were connected by rhizomes to an estalilished plant or rhizome system. Counts of annual rings established that the rhizomatous connections were one to four \ears old. UsualK , a large, established parent plant was the soiuce of rhizomes connect- ing either single plants or a series of sprouts 1990] Sfhoutinc andSeedlin(;Estahlisiimentin Sacehrush 203 Fig. 1. Graphic example of an excavated sprout connected to a parent plant and other offspring. Original or oldest material is on far right. Table 1. A comparison of the number of rhizomatoiis to nonrhizomatous plains silver sagebrush plants found at each studv site. Stud\ site' 1 2 3 4 5 6 Total Rhizomatons' Nonrhizomatous 13 7 13 15 7 13 7 5 13 7 14 6 75a3 45'' Sites are numbered as follows; 1 — Yellowstone River, 2 — Lower Black Springs, 3 — Lower Flood, 4 — Lignite Creek, .5 — Paddy Faye, 6 — Moon Creek. "Plants with rhizome connections (alive or dead! to other plants. Significant (P < .025) differences between rhizomatous and nonrhizomatous plant totals b\ Student's t test are followed by different letters. (Fig. 1). However, some plants were from a series of sprouts along a rhizome presently terminated with a dead or decadent stump. One site, Lignite Creek, was different in that it had a majority of nonconnected individuals (Table 1). Reduction of available soil moisture due to clay pan soils overlying extensive gravel at Lignite Creek might explain the difference. Plant water use would be less favorable with this condition at the surface. atlording an advantage to taprooting plants in reaching deeper, more favorable conditions. The three sites disturbed by fire or ice action were compared with the three undisturbed sites to learn whether the ratio of sprouts to seedlings changed with disturbance. No sig- nificant differences (P < .05) in numbers of plants arising from rhizomes due to distur- bance were found. Generally, an elaborate subsurface rhizome system was found that was older than above- ground stems. There were significant (P < .05) age differences among plant stems, tap- roots, and parent rhizomes (Table 2) over all six study sites. Aboveground stems were three to five years younger than taproots and associated rhizomes. Parent rhizomes with di- rectly connected sprouts were significantly older than taproots. Taproots and rhizomes without direct connections to a parent plant were not significantly different in age from each other. 204 C. L. Wamboltetal. [W^lunie 50 0-50 51-100 101-150 151-200 Centimeters 201 and over Fig. 2. Number of distances encountered between parents and traceable sprouts of rhizomatous plains siher sagebrush over six stud\- sites. T.\ble2. Age relationships of abo\e- and below groinid Table 3. Growth relationships of abo\e- and below- parts of plains siKer sagebrush plants fioin the si.x stud\' groiuid parts of rhizomatous plains silver sagebrush plants sites. from the six stud\ areas. Mean age St anda ■d Number of Plant part (\ears) de\iation samples Stems 3.4^^ 2.0 204 Taproots 6.9'' 3.1 28 Parent rhizome" S.8" 3.7 68 Rhizome s\ stem' 6.0'' 2.. 5 128 Significant [P < .05) mean differences by Duncan s nuiltiple range test ; followed b\ different letters. "Rliizonie originating from a parent plant (or dead stnnip ' ti) u liic h sprout v directly connected. Rhizome sections other than in 2 abo\e. Rhizome e.xtension was greater than abo\e- ground heights (P < .05) (Table 3), even in older plants that had the largest aerial por- tions. Rhizome length from the selected plant to the parent plant averaged 2.4 times that of plant height. Total lateral spread of the rhizome system averaged 3.3 times that of plant height. Fignre 2 summarizes the number of dis- tances encountered between parent plants and traceable sprouts. The largest proportion (59%) of these connections were from 50 to Mean Range Niunber of Plant part (cm^ (cm) samples Plant (sprout) height 32"' 11-59 155 Lateral distance to parent connection" 78'' 14-277 61 Lateral spread of rhizome s\stem' 105^ 11-369 90 Signilicant \P < 1151 mean differences b\ Duncan s multiple range test are toUowed b\ different letters. "Lateral distance from parent plant or rhizome to nearest sprout on rhizome system expressed as a mean of all plants with this growth habit. ■TTotal lateral extent of all rhizomes in an exca\ ation expressed as a mean of all plants with rhizome systems. 100 cm in length, followed b\- the 0-50-cm distance (207f ). Isolated Plant Exca\ ations The extensive sprouting natiue of plains siKer sagebrush was apparent after exca\a- tions had been completed in areas surround- ing two large, isolated plants. Most roots were part of a shallow, complex imdergroimd net- work of interconnected rhizomes that often included several smaller, nearly independent 1990] Sprouting AND SekdlingEstabi.isiimkntin Sa(;ebhush 205 HEIGHT 5 45+ cm 4 35 — 45cm 3 25-35cm 2 l5-25cm 1 0-15 cm O NUMBERS IN CATEGORY 2 14 13 17 26 46 "^ ' rhizome aboveground shoots canopy cover of main plant Fig. 3. Diagram of an isolated plant excavation. All aboveground shoot.s are shown individually and numbered according to height. systems (Fig. 3). This was most apparent with older, well-established plants fiom which the network originated (Fig. 3). Characteristic of the horizontal rhizome expansion were size classes decreasing in con- centric circles away from the parent plant. There was considerable variation in rhizome complexity within these individual systems. Excavations established that rhizomes can sprout at least 3 m from the parent plant. Therefore, a large number of progeny may arise asexually from one individual. Individ- ual rhizomes had from 1 to 52 sprouts. No evidence indicated that all individual systems were of the same origin. That is, no common root connections could be traced. However, some might have been connected and later separated after mortality of connec- tive rhizomes. Discussion Vegetative reproduction is prevalent in plains silver sagebrush, and the causal agents are of interest and importance to rangeland management. Benefits of vegetative repro- duction include (1) an enhanced ability to uti- lize unevenly distributed resources and (2) an increased competitive ability to occupy adja- cent areas (Harper 1977, Cook 1983). In addi- tion, sprouts are better able to resist invasions of seedlings from other species while reducing the probability of extinction. This is accom- plished by spreading the risk among many genetically identical individuals (Cook 1983). An evolutionary strategy that employs asexual mechanisms is consistent with the findings of Abrahamson (1980), who reported that in- creased environmental severity generally shifted emphasis to vegetative reproduction. Generally, vegetative reproduction is most important where fire, weather phenomena, and other disturbances are common (Bostock and Benton 1979, Went 1979, Abrahamson 1980, Legere and Payette 1981). The sprout- ing nature of plains silver sagebrush is likely an adaptation to its northern Great Plains habitat. Flooding with associated deposition, 206 C. L. Wamboltetal. [Volume 50 along with ice scraping during winter events, is common. Plant production and subse- quently fuel loads for fires are relatively high in bottomlands inhabited by the taxon. Consequently, fires are common in plains silver sagebrush habitats. It is logical that plains silver sagebrush is a vigorous sprouter in response to the evolutionary influences of recurring disturbances. This taxon is re- ported to produce only 18% as many achenes as big sagebrush (Artemisia tridentata Nutt.) (Harvey 1981, Tisdale and Hironaka 1981), which likely reflects a reliance on asexual re- production. Abundant herbaceous vegetation in mesic flood plains produces substantial competition for seedlings. Vegetative sprouts may com- pensate through more rapid morphological development. Because sprouts have the ad- vantage of a nutrient reserve from established plants, the sprouting strategy increases sur- vival (Abrahamson 1980). Although not rare in the communities stud- ied, seedling establishment was found in only one-third of the plants excavated (Table 1). This may be attributable to the inconsistency of specific environmental conditions required for germination and seedling establishment. Environmental factors, especially drought, might best explain differences in ratios of sprouts and seedlings found in 1983. For ex- ample, a three-year drought at the study area occurred between 1979 and 1981 when the mean annual precipitation was 23.0 cm and preceded the wet year of 1982 with 41.6 cm of precipitation. The long-term average precipi- tation is 34.8 cm. This drought coincided with the ages of most plants examined in this study. The relatively moist years preceding (1978 with 44.7 cm) and following (1982 with 41.6 cm) this drought provided the periods of es- tablishment for seedlings at the study sites. However, just as seedlings appear favored during wet years, sprouts were found to have the advantage in establishing during rela- tively dry periods. The cool, wet growing sea- son of 1982 was followed by a warm, dry (22.3 cm) growing season in 1983. Subseciuently, numerous seedlings and few sprouts were produced during 1982, and few seedlings with an abundance of sprouts were produced in 1983. Few seedlings from 1982 survived beyond the dry second season. Therefore, it appears that both the mode and the success of plains silver sagebrush reproduction is strongly related to available moisture as in- dicated by Salisbury (1942) for wild garlic {Allium carinatum L.). Perhaps this influence of climate on reproduction might mask differ- ences of sprout-to-seedling ratios expected between disturbed and undisturbed sites. In our study, these ratios did not vary signifi- cantly (Table 1). The reproductive strategies of plains silver sagebrush partially explain the taxon's success and require consideration in managing its habitats for optimum balance between livestock forage production and suit- able wildlife habitat. Conclusions We conclude that vegetative reproduc- tion in plains silver sagebrush is the pri- mary means of plant establishment. Although sprouting in this taxon has likely evolved with habitat disturbances, this study did not estab- lish that a greater percentage of plants arising from sprouts should be expected on disturbed than on undisturbed sites. However, annual precipitation does appear to be related to the relative success in initiation and survival of seedlings and sprouts. Seedlings apparently require more moisture for both germination and survival than do sprouts. Acknowledgments This paper was published with the approval of the Montana Agricultural Experiment Sta- tion as Journal Article J-2431. Literature Cited Abrahamson. W. G. 19namics of ph\tomass and minerals in two salt desert shrub communities. Great Basin Naturalist 44: 327-3.37. 78. Blackburn, W H 1967. Plant succession on se- lected habitat types in Nevada. Unpublished the- sis. University of Nevada, Reno. 162 pp. 79. Bl.^ckburn, W. H , R. E Eckert, Jr., and P. T. Tueller. 1969. Vegetation and soils of the Crane Springs Watershed. Ne\'ada Agricultural Experi- ment Station Bulletin R-55. Reno, Nevada. 63 pp. 80. Blackburn, W H. AND P. T. Tueller 1970. Pinyon and juniper invasion in black sagebrush communi- ties in east-central Nevada. Ecology 51: 841-848. 81. Bl.\ckburn, W. H.. P T Tueller, and R. E. Eckert. Jr. 1968a. Vegetation and soils of the Mill Creek Watershed. Nevada Agricultural Experi- ment Station Bulletin R-43. Reno, Nevada. 69 pp. 82. Bl.\ckburn, W H . P T Tueller, and R E. Ec;kert, Jr 1968b. Vegetation and soils of the Crowley Creek Watershed. Nevada Agricultural Experiment Station Bulletin R-42. Reno, Nevada. 60 pp. 83. Bl,u:kburn, W H , P T. Tueller, and R. E. Eckert, Jr. 1968c. Vegetation and soils of the Duckwater Watershed. Nevada Agricultural Experiment Station Bulletin R-40. Reno, Nevada. 76 pp. 84 Hl\( KHUKN. W H , P T Tueller, and R. E. f,( Ki HI, Jr 1969a. \'egetation and soils of the Cow CJrcck Watershed. Nevada Agricultural Experi- ment Station Bulletin R-49. Reno, Nevada. 80 pp. 85. Bl.\ckburn. W H . P T Tueller, and R E Eckert, Jr 1969b. Vegetation and soils of the Coils C^reek Watershed. Nevada Agricultural Ex- periment Station Bulletin R-48. Reno, Nevada. 81pp. 1990] Nevada-Utah Vkcktation Bihi.iochai'iiv 213 86. Blackbuhn W II V T Tuellek. and H K ECKERT, Jk 1969c'. X't'Uctation and soils oi' tlu- Churchill Canyon Watershed. Nesada Aj^riciil- tural Experiment Station Bulletin U-45. lU'iio, Nevada. 157 p]). 87. Blackburn, VV. 11 , 1' T Ti kli.kr, and H E ECKERT, Jr 1969d. Vegetation and soils of the Pine and Mathevv Canyon watersheds. Nevada Agricul- tural E.xperinient Station Bulletin R-46. Reno, Nevada. Ill pp. 88. Bl.^ckburn, W. H., P. T. Tueller, and R E ECKERT, Jr. 1971. Vegetation and soils of the Rock Springs Watershed. Nevada Agricultural Experiment Station Bulletin R-S3. Reno, Nevada. 116 pp. 89. Blackhawk Coal Company. 1981. Vegetation re- sources. Chapter 9, section 9.2, pages 9-1-9-27 in Mining and Reclamation Plan for Willow Creek Mine, Blackhawk Coal Company. Utah Division of Oil, Gas and Mining Numher ACT/()()7/0()2. Salt Lake City, Utah. 90. Blaisdell. J. P., \ND R. C. Holmgren 1984. Managing intermoiuitain rangeland.s— salt-desert shrub ranges. US DA Forest Service General Technical Report INT- 163. Intermountain Forest and Range Experiment Station, Ogden, Utah. 91. Blaisdell, J. P., R. B. Murray, and E D McArthur 1982. Managing intermountain range- lands: sagebrush-grass ranges. USDA Poorest Ser- vice General Technical Report lNT-134. Inter- mountain Forest and Range Experiment Station, Ogden, Utah. 41 pp. 92. Blaksley, J. A. 1987. Avian habitat relationships in riparian zones of northern Utah. Unpublished thesis. University of Idaho, Moscow. 93. Blank, D L. 1984. Forage quality comparison of burned and unburned aspen communities. Un- published thesis, Utah State University, Logan. 94. Bleak, A. T , N. C Frischknecht, A P. Plummer, and R. E. Eckert, Jr. 1965. Problems in artificial and natural revegetation of the arid shadscale veg- etation zone of Utah and Nevada. Joiunal of Range Management 18: 59-65. 95. BOLEN, E. G. 1962. Plant ecology of spring-fed salt marshes. Unpublished thesis, Utah State Univer- sity, Logan. 96. BoLEN, E G. 1964. Plant ecology of spring-fed salt marshes in western Utah. Ecological Monographs 34: 143-166. 97. BosTiCK, V B , W E Niles, W A McClellan, E. H Oakes. and J R Wilbanks. 1975. Inventory of natural landmarks of the Great Basin, Vols. I and II. Unpublished report prepared for the USDI, National Park Service, by University of Nevada, Las Vegas. 690 pp. 98. Boucek. M M 1986. Vegetation survey at the Sum- mit No. 1 Coal Mine, Summit County, Utah. Vol- ume 1, section 783.19, appendix 783.19, pages 1-15 in Mining and Reclamation Plan for Summit No. 1 Mine, Summit Minerals Incoroporated. Utah Division of Oil, Gas and Mining Number ACT/043/001. Salt Lake City, Utah. 99. Bowers, J. E. 1982. The plant ecology of inland dunes in western North America. Journal of Arid Environments 5: 199-220. 100. liowNS, J F , and N. E. We.st. 1976. Blackbush {Colcofiitijiic ramosissima Torr. ) on southwestern Utah rangelands. I'tah Agricultural Experiment Station Research Report 27. Logan, Utah. 27 pp. 101. Bracken, A. F. 1940. Dry fanning as developed in the sagebrush zone. Proceedings of Utah Academy of Science, Arts and Letters 17: 25-32. 102. Bradley, W. G. 1964. The vegetation of the desert game range with special reference to the desert bighorn. Pages 43-67 in Transactions of the Desert Bighorn Council. Las Vegas, Nevada. 103. Bradley. W. G. 1965. A study of blackbrush com- munity on the desert game range. Pages 56-61 in Transactions of the Desert Bighorn Council, 9th annual meeting, 6-8 April 1965, Redlands, California. 104. Bradley. W G. 1966. Populations of two Sonoran Desert plants and deductions as to factors limiting their northward extension. Southwestern Natural- ist 11: 395-401. 105. Bradley, W. G. 1967. A geographical analysis of the flora of Clark County, Nevada. Journal of the Ari- zona Academy of Science 4: 151-162. 106. Bradley, W G 1973. Ecological study of Zion Na- tional Park. Unpublished report prepared for the USDI, National Park Service. Springdale, Utah. 183 pp. 107. Bradley, W G , and J E. Deacon. 1965. The biotic communities of southern Nevada. University of Nevada Prepublication Series 9. Desert Research Institute, Las Vegas, Nevada. 86 pp. 108. Bradley, W. G., and J. E. Deacon. 1967. The biotic communities of southern Nevada. Nevada State Museum Anthropological Papers 13, Part IV. Car- son City, Nevada. 69 pp. 109. Branson, F. A 1966. Geographic distribution and factors affecting the distribution of salt desert shrubs in the United States. Pages 13-43 in Pro- ceedings from the Salt Desert Shrub Symposium, 1-4 August 1966. USDI, Bureau of Land Manage- ment, Cedar City, Utah. 110. Branson, F. A. 1985. Vegetation changes on west- ern rangelands. Range Monograph 2. Society for Range Management, Denver, Colorado. 76 pp. 111. Branson, F. A , R F. Miller, and I S. McQueen 1967. Geographic distribution and factors affect- ing the distribution of salt desert shrubs in the United States. Journal of Range Management 29: 287-296. 112. Breternitz, D. a., and D W. Martin 1973. Report of Dolores River project reconnaissance, 1972- 1973. Unpublished report submitted to the USDI, National Park Service, Midwest Archaeological Center, Lincoln, Nebraska. 113. Briggs. G. M. 1978. A study of sedge-dominated areas in the Uinta Mountains. Unpublished thesis, Utah State University, Logan. 89 pp. 114. Briggs. G. M., and J. A. MacMahon. 1982. Struc- ture of alpine plant communities near Kings Peak, Uinta Mountains, Utah. Great Basin Naturalist 42: 50-59. 115. Brk;gs, G M , and J. A. MacMahon. 1983. Alpine and subalpine wetland plant communities of the Uinta Mountains, Utah. Great Basin Naturalist 43: 523-530. 214 P. S. BOURGERON ET AL. [Volume 50 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 1.30. 131. Britton, C. M., and M H Ralphs 1979. U.se of fire as a management tool in sagebrush ecosystesni. Pages 101-109 »i The Sagebrush Ecosystem: A Symposium, April 1978, Utah State University, Logan. Brotherson, J. D. 1974. A vegetation analysis of areas involved in the proposed Four Seasons pro- ject. Pages 60-104 in Environmental report on proposed Four Seasons Project. Unpublished report prepared by Brighani Young University Center for Health and Environmental Studies for Four Seasons, Inc. Brotherson, J D. 1978. Vegetative and wildlife assessment of the Jordan Aqueduct E.\tension, Bonneville Unit, Central Utah Project. Unpub- lished report prepared for the Rocky Mountain Research Co. and USDI, Bureau of Reclamation, Provo, Utah. 45 pp. Brotherson, J D 1981. Aquatic and semiaquatic vegetation of Utah Lake and its bays. Great Basin Naturalist Memoirs 5: 68-84. Brotherson, J. D 1987. Plant community zonation in response to soil gradients in a saline meadow near Utah Lake, Utah County, Utah. Great Basin Naturalist 47: 322-333. Brotherson, J D , D L Anderson, and L. A. SzvsKA 1984. Habitat relations of Cercocarpus montcitiiis (true mountain mahogany) in central Utah. Journal of Range Management 37: 321-324. Brotherson, J D . and S J Barnes. 1984. Habitat relationships of Glmix maritima in central Utah. Great Basin Naturalist 44: 299-.3()9. Brotherson, J. D,, and K J Brotherson 1979. Ecological and community relationships of Eri- o^onum corymbosum (Polygonaceae) in the Uinta Basin, Utah. Great Basin Naturalist 39: 177-191. Brotherson, J D.. and W. T Brotherson 1981. Grazing impacts on the sagebrush communities ol central Utah. Great Basin Naturalist 41: 335-340. Brotherson. J D., J G Carman, and L. A. Szyska. 1984. Stem-diameter age relationships ofTamarix ramo.sissima in central Utah. Journal of Range Management 37: 362-364. Brotherson, J D , J. N Davies, and L Green- wood. 1980. Diameter-age relationships of two species of mountain mahogany. Journal of Range Management 33: 367-370. Brotherson, J D., andW. E. Evenson. 1983. Veg- etation communities surrounding Utah Lake and its bays. Utah Lake Vegetation Studies. Unpub- lished report prepared for the Utah Division of Wildlife Resources and USDI, Bureau of Recla- mation, Provo, Utah. 401 pp. Brotherson, J. D., and D. Field 1987. Tamarix: impacts of a successful weed. Rangelands 9; 110-112. Brotherson, J. D. andW J Masslich 1985. Veg- etation patterns in relation to slope position in the Castle Cliffs area of southern Utah. Great Basin Naturalist 45: 535-541. Brotherson, J. D , and C Morden 1979. Vegeta- tion of Utah Lake shorelini- and bays. Map in four parts. Press Publishing Limited, Provo, Utah. Brotherson, J D , G Nereker, M Skol'Gard, and J. Fairchhj:). 1978. Plants of Navajo National Monument. Great Basin Naturalist 33: 19-30. 132. Brotherson, J D . and S T. Os.wande. 1980. Min- eral concentrations in true mountain mahogany and Utah juniper and in associated soils. Journal of Range Management 33: 182-185. 133. Brotherson. J D , and K P. Price 1984. Natural- ization and habitat relationships of bitter night- shade (Solanum dulcamara) in central Utah. Great Basin Naturalist 44: 317-323. 134. Brotherson, J. D . K P Price, and L ORourke. 1987. Age in relationship to stem circumference and diameter in cliffrose {Coivania mexicana var. stansbiiriana) in central Utah. Great Basin Natu- ralist 47: 334-338. 1.35. Brotherson. J. D., L. L R\smussen, and R. D. Blac;k 1986. Comparative habitat and community relationships of Atriplex confeiiifolia and Sarco- battis icnniculattis in central Utah. Great Basin Naturalist 46: 348-357. 136. Brotherson, J D., S R Rushforth, and J R. Johansen. 1983. Effects of long-term grazing on cryptogam crust cover in Navajo National Monu- ment, Arizona. Journal of Range Management 36: 579-581. 1.37. Brotherson. J. D., and V. Winkel. 1986. Habitat relationships of saltcedar (Tamarix ramosis- sima) in central Utah. Great Basin Naturalist 46: 535-541. 1.38. Brown, D E , C H Lowe, and C P Pase. 1979. A digitized classification system for the biotic com- munities of North America, with community (se- ries) and association examples for the Southwest. Journal of the Arizona-Nevada Academy of Sci- ence 14: 1-16. 1.39. Brown. D E . and C H Lowe 1980. Biotic com- munities of the Southwest. USDA Forest Service General Technical Report RM-78. Rocky Moun- tain Forest and Range Experiment Station, Fort Collins, Colorado. 140. Brown, L L 1962. Establishment of forage plants on difficult salt-desert range sites. Unpublished thesis. University of Nevada, Reno. 141. Brun, J M 1962. A comparison of the line-inter- ception, point frame and distance measurement methods for analyzing desert shrub vegetation. Unpublished thesis, Utah State University, Lo- gan. 142. Bruner, a. D 1979. Predicting success of pre- scribed fires in pinyon-juniper woodland in Ne- vada. USDA Forest Service Research Paper INT- 219. Intermountain Forest and Range Experiment Station, Ogden, Utah. 11 pp. 143. Buchanan, B. A., K. T Harper, and N. C. Frischknecht. 1978. Allelopathic effects of bur I)iittercup tissue on the germination and growth of \arious grasses and forbs in vitro and in soil. Great Basin Naturalist 38: 90-96. 144. Bi CHANAN, H 1960. The plant ecology of Bryce Canyon National Park, Utah. Unpublished disser- tation, Uni\ersit\ of Utah, Salt Lake City. 145. BrcHANAN. H., and K T Harper 1981. Forests of Bryce Canyon National Park: conseciuencesofpast management policies and recommendations for thchiture. Unpublished final report, part III, Pro- ject CX-1200-8-B016 (t\ pcscript). 146. Bi'DY, J D . AND j A Voi'N(;. 1979. Historical use of Nevadas pinvon-juniper woodlands. Journal of Forest History 23: 112-121. 1990] Nenada-Utah Ve(;etatk)N Bibliockaimiy 215 147. IkNOKHstJN, E D . D J \\ i;ni,n. and D L Ni:lson. 1986. Diseases associated witli Jutiipcrus ostco- spcnna and a model for predicting their oeenr- 163. reneewith environmental site laetois. (lieat liasin Natnralist 46: 427-440. I4S. Hi HKl., M H 1934. Plant distrihntion studies in the Well.sville Range, Utah. Unpublished thesis, Utah State Agricultural C^ollege, Logan. 33 pp. 149. BURH, G. O 1931. Native vc-getation of the pre- historic Lake Bonneville Basin. Journal of the 164. American Society of Agronomv' 23: 407-413. 150. Bi'TLKR, J R. .\ndJ L Engl.'VND. 1979. Vegetation map of the southeastern Uinta Basin, Utah and 165. Colorado. USDl Geological Survev' Miscellaneous Investigative Series 1-1141. Washington, D.C. 151. Buzz.\RD. R F . Jr 1981. Comparison of two tradi- 166. tional techniques for determining range trend and recomendations for improved procedures. Un- published thesis, Utah State University, Logan. 152. BvE. R A 1972. Ethnobotany of the southern Paiute 167, Indians in the 1870's. In: Great Basin Cultural Ecology, a Symposium, Reno, Nevada. Research Publications in the Social Sciences 8. 168. 153. C.-\LDVVELL, M. M., .-VND N. E We.st. 1972. The plant ecologv of Utah's desert rangelands. Utah Science 33: 14-15. 154. Call, C.A., and J. R. Barker. 1980. Vegetation sur- 169. vey of U.S. Fuel Company property, Hiawatha, Utah. Volume 3, chapter 9, section 9.9, appendix IX-1, pages 1-80 in Mining and Reclamation Plan for Hiawatha Mines Complex, U.S. Fuel Com- 170. panv. Utah Division of Oil, Gas and Mining Num- ber ACT/007/011. Salt Lake City, Utah. 155. Callison.J, Jr., andJ. D. Brotherson 1985. Habi- tat relationships of the blackbush community 171. {Coleogyne ramosissima ) of southwestern Utah. Great Basin Naturalist 45: 321-326. 156. Callison. J . Jr.. J D Brotherson, and J. E. BovvNS. 1985. The effects of fire on the blackbrush 172. (Coleogyne ramosissima) community of south- western Utah. Journal of Range Management 38: 535-538. 157. Campbell, V O. 1977. Certain edaphic and biotic 173. factors affecting vegetation in the shadscale community of the Kaiparowits area. Unpublished thesis, Brigham Young Universitv, Provo, Utah. 59 pp. 158. Cannon, H. L. 1960. The development of botanical 174. methods of prospecting for uranium on the Colo- rado Plateau. USDI Geological Survey Bulletin 1085-A. Washington, D.C. 50 pp. 159. Canon, S. K. 1985. Habitat selection, foraging 175. behavior, and dietary nutrition of elk in burned vs. unburned aspen forests. Unpublished thesis, Utah State LIniversity, Logan. 160. Carman, J. G, andJ D Brotherson. 1982. Com- parisons of sites infested and not infested with 176. salt cedar (Tamarix pentandra) and Russian olive [Elaeagnus an gust if alia). Weed Science 30: 360-364. 161. Carter, K S 1981. Analysis of trend and methods 177. used to determine trend on southern I'tah Na- tional Forest. Unpublished thesis, Utah State L'niversity, Logan. 178. 162. Castle, E, S. 1955. The vegetation and its relation- ship to the dune soils of the pink sand dunes of Kani- County, Utah. Uni)ublished thesis, Brigham Young University, Provo, Utah. Cedar CIreek Associates Incorpor.ated 1987. Draft vegetation information for the Alton coal mining project. Volume 6, cha])tcr 3, appendix 3.6-B, pages 1-41 in Miningand Reclamation Plan for Alton Mine, I'tah International Incorporated. Utah Division of Oil, Gas and Mining Number ACT/()25/003. Salt Lake City, Utah. Charoi, B. F, and W, D Bh.linos 1972. Origins and ecology of the Sierran alpine flora and vegeta- tion. Ecological Monographs 42: 163-199. Chamberlain. R V 1911. The ethnobotany of the Gosiute Indians, Proceedings of the Academy of Science of Philadelphia 63: 24-99, Chambers. J C 1979. The effects of grazing on salt desert shrub species survival during a period of below average precipitation. Unpublished thesis, Utah State University, Logan. Chapman. V. J 1960, Salt marshes and salt deserts of the world. Plant Science Monograph. Inter- science Publishers, Inc., NewY'ork. .392pp. Choate, G. a. 1965, Forests in Utah. USDA Forest Service Resource Bulletin INT-4. Interniountain Forest and Range Experiment Station, Ogden, Utah. 61 pp. Christensen, E M 1949. The ecology and geo- graphic distribution of oakbrush (Qtiercus gam- helii) in Utah, Unpublished thesis. University of Utah, Salt Lake City. 70 pp. Christensen, E M 1950. Distributional observa- tions of oakbrush (Querciis gamhelii Nutt.) in Utah. Proceedings of Utah Academy of Science, Arts and Letters 27: 22-25. Christensen. E. M. 1955. Ecological notes on the mountain brush in Utah. Proceedings of Utah Academy of Science, Arts and Letters 32: 107-111. Christensen, E. M. 1957, Photographic history of the mountain brush on "Y" Mountain, central Utah. Proceedings of Utah Academy of Science, Arts and Letters .34; 154-155 (abstract). Christensen, E, M, 1958. Growth rates and vege- tation changes in the oak-maple brush in lower Provo Canyon, Utah. Proceedings of Utah Academy of Science, Arts and Letters 35: 167-168. Christensen, E. M, 1959a. A climatic study of some northern desert shrub communities. Proceedings of Utah Academy of Science, Arts and Letters 36: 174 (abstract). Christensen, E. M, 1959b. A comparative study of the climates of mountain brush, pinyon-juniper, and sagebrush communities in LUah, Proceedings of Utah Academy of Science, Arts and Letters 36: 174-175 (alxstract,) Christensen, E M 1961a, The Agropijron-Poa bunchgrass vegetation in central Utah. Proceed- ings of Utah Academy of Science, Arts and Letters ,38: 113 (abstract.) Christensen, E M 1961b. The deciduous tree communities of central Utah. Bulletin of the Eco- logical Society of America 42: 62-63 (abstract.) Christensen. E. M 1962. The rate of naturalization oiTamarix in Utah. Americal Midland Naturalist 68: 51-57. 216 P. S. BOURGERON ET AL. [Volume 50 179. Christensen, E. M 1963a. Naturalization of Rus- sian olive {Elaeagmis angustifolia L.) in Utah. American Midland Naturalist 70: 133-137. 180. Christensen. E. M. 1963b. The foothill bunchgrass vegetation of central Utah. Ecology 44: 156-158. 181. Christensen, E. M. 1964a. Changes in composition of a Bromus tectortim-Sporobolus cnjptandnis- Aristida longiseta community following fire. Pro- ceedings of Utah Academy of Science, Arts and Letters 41: 53-57. 182. Christensen, E. M 1964b. Succession in a moun- tain brush communitv' in central Utah. Proceed- ings of Utah Acadeni\- of Science, Arts and Letters 41; 10-13 (Erratum 41[2];ii). 183. Christensen, E. M 1967a. Bibliography of Utah botany and vvildland conservation. Brigham Young University Science Bulletin, Biological Series 9(1). Provo, Utah. 135 pp. 184. Christensen, E. M. 1967b. Bibliography of Utah botany and wildland conservation, No. II. Pro- ceedings of Utah Academy of Science, Arts and Letters 44: 545-566. 185. Christensen, E. M., and J. D Brotherson. 1972. Preliminary botanical survey of the Bonneville Unit of the Central Utah Project. Pages 5-134 in E. M. Christensen etal., eds.. Preliminary survey of the biota of the Bonneville Unit ot the Central Utah Project. Unpublished report prepared for the USDI, Bureau of Reclamation, Provo, Utah, by Brigham Young University Center for Health and Environmental Studies, Provo, Utah. 186. Christensen. E M., and B F Harrison 1961. Ecological study area at Lily Lake in the Uinta Mountains, Utah. Proceedings of Utah Academ\- of Science, Arts and Letters 38: 36-49. 187. Christensen, E. M., and M. A. Hutchinson. 1965. Historical observations of the ecology of Rush and Tooele Valleys, Utah. Proceedings of Utah Academy of Science, Arts and Letters 42: 90-105. 188. Christensen, E. M . and H B Johnson. 1964. Pre- settlement vegetation and vegetational change in three valleys in central Utah. Brigham Young Uni- versitv Science Bulletin, Biological Series 4(4): 1-16.' 189. Christensen, E M.. and S L. Welsh 1963. Pre- settlement vegetation of the valleys of western Summit and Wasatch Counties, Utah. Proceed- ings of Utah Academv of Science, Arts and Letters 40: 163-174. 190. Christian, R 1962. Plant geography of the Beaver Dam Mountains. Unpublished thesis. University of Utah, Salt Lake City. 153 pp. 191. Clement, R. C 1979. Culture and species endan- germent. Great Basin Naturalist Memoirs 3: 11-16. 192. Clokev, I W 1951. Flora of the Charleston Moun- tains, Clark County, Nevada. University of Cali- fornia Press, Berkeley. 193. Clo\er, E. U , and L Jotter 1944. Floristic studies in the canyon of the Colorado and tributaries. American Midland Naturalist 32: 591-642. 194. Cole, N. J. 1968. Mule deer utilization of rehabili- tated Nevada rangelands. Unpublished thesis. University of Nevada, Reno. 195. Coles, F H., and J. C. Peder.son 1967. Utah big game range inventory, 1966. Utah Di\ison ol Wildlife Resources Publication 67-1. Salt Lake City, Utah. 171 pp. 196. Coles, F. H., and J. C. Pederson. 1968. Utah big game range inventory, 1967. Utah Division of Wildlife Resources Publication 68-2. Salt Lake City, Utah. 120 pp. 197. Coles, F. H . and J C Pederson 1969. Utah big game range inventory, 1968. Utah Division of Wildlife Resources Publication 69-2. Salt Lake City, Utah. 164 pp. 198. Coles, F H., and J. C. Pederson. 1970. Utah big game range inventory, 1969. Utah Division of Wildlife Resources Publication 70-1. Salt Lake City, Utah. 127 pp. 199. Collins. P. D. 1980. Comparative life history and floral characteristics of desert and montane plant commimities in Utah. Unpublished thesis, Brig- ham Young University, Provo, Utah. 163 pp. 200. Collins, P D 1983. Vegetation and reclamation of the Wellington Coal Cleaning Plant. Volume 3, appendix H, pages 1-44 in Mining and Reclama- tion Plan for Wellington Prep Plant, Kaiser Coal Corporation. L^tah Division of Oil, Gas and Min- ing Number ACT/007/012. Salt Lake City, Utah. 201. Collins. P D., and K T Harper 1980. Vegetation map of the Mt. Nebo complex and San Pitch Mountains, Utah. Brigham Young University Press, Provo, Utah. 202. Collins, P. D. , K. T. Harper, and B. K. Pendleton. 1983. Comparative life history and floral charac- teristics of desert and mountain floras in Utah. Great Basin Naturalist 43: 385-393. 203. Collins, W B 1977. Diet composition and activi- ties of elk on different habitat segments in the lodgepole pine type, Uinta Mountains. Unpub- lished thesis, Utah State University, Logan. 204. CoLTHARP. G. B.. and N. E West 1966. Effects of surface soil treatments on soil, water, and vegeta- tion in Utah's east desert area. Pages 88-97 in Proceedings from Salt Desert Shrub S\mposium, 1-4 August 1966. USDI, Bureau of Land Manage- ment, Cedar City, Utah. 205. Committee on En\ironmental Quality. 1970. Preserving Nevada s environmental heritage. Un- published final report prepared for the Governor's Natural Resources Council by the Department of Conservation and Natiual Resources, Carson City, Nevada. 206. Cook, C. W. 1977. Eft'ects of season and intensity of use on desert vegetation. Utah Agricultural Experiment Station Bulletin 483. Logan, I'tah. 57 pp. 207. Cook, C. W , and L. E. Harris 1950. The nutritive content of the grazing sheep s diet on the summer and winter ranges of Utah. Utah Agricultural Experiment Station Bulletin 342. Logan, I'tah. 66 pp. 208. Cook. C W , and R Hurst 1962. A (juantitative measure oi plant association on ranges in good and poor condition. Journal of liange Management 15: 266-273. 209. Cook. C. W . and C. E Lewis 1963. ("ompetition between big sagebrush and seeded grasses on foothill ranges in I'tah. Joiunal of Range Manage- ment 16: 245-250. 1990] Nevada-Utah Vegetation Bibmochaphy 217 210. Co-Oi' Mining Company 1988. Vegt-tation. Vol- ume 2, chapter 9, .section.s 9.1-9.3, pages 9-1- 9-10, plusappen;es 9-1-9-19, pins tahle 9-14, pages 9-38-9-41 in Mining and lieclaniation Plan for Black Jack Mine No. 1, New Tech Mining Corporation. Utah Division of Oil, Gas and Min- ing NnmberACT/019/()()4. Salt LakeCJity, Utah. 278. England. J. L. 1979. Measnred and inferred mois- ture gradient relationships: a stud\' of a montane steppe in central Utah. Unpnl)lished thesis, Brigham Young Unixersity, Provo, Utah. 36 pp. 279. Encle. D M , C D Bonham, and L E. Bartel. 1983. Ecological characteristics and control of Gambel oak. Journal of Range Management 36: 363-365. 280. Ekdman, K S 1961. Classification and distribution of the nati\'e trees of Utah. Unpublished thesis, Brigham Young University, Provo, Utah. 221 pp. 281. EsPLiN. A. C . J. E Greanes. and L A Stoddart 1937. A study of Utah's winter range. Utah Agri- cultural Experiment Station Bulletin 277. Logan, Utah. 49 pp. 282. Evans, C. C. 1986. The relationship of cattle grazing to sage grouse use of meadow habitat on the Shel- don National Wildlife Refuge. Unpublished the- sis. University of Nevada, Reno. 283. Evans, F. R. 1936. A comparative study of the vege- tation of a grazed and an ungrazed canyon of the Wasatch Range. Unpul)lished thesis. University of Utah, Salt Lake City. 284. Evans, P. A 1925. A plant ecology study in Utah. Unpublished dissertation. University of Chicago. 285. E\'ANS, P. A. 1926. An ecological study in Utah. Botanical Gazette 82: 253-285. 286. Evenson, W E . J D Brotherson, and R B Wilcox. 1980. Relationship between environ- mental and vegetational parameters for under- storv and open-area communities. Great Basin Naturalist 40: 167-174. 287. Enerett, R L 1968. Use of the cottonwood in an investigation of the recent history of a flood plain. American Journal of Science 266: 417-439. 288. Everett, R. L. 1983. Understory seed rain on tree- harvested and unharvested pinyon-juniper sites. Journal of Environmental Management 17: 349-358. 289. Everett, R L 1985, Great Basin pinyon and ju- niper communities and their response to manage- ment. Pages 53-61 in Proceedings of the Cultural, Physical and Biological Characteristics of Range Livestock Industry in the Great Basin Sympo- sium, 38th annual meeting of the Society for Range Management, 11-14 February 1985, Salt Lake City, Utah. 290. Everett, R. L. 1986. Understory seed rain in har- vested pinyon-juniper woodlands. Great Basin Naturalist 46: 706-710. 291. Everett, R. L., and S. Konlak. 1981. Understory vegetation in fully stocked pinyon-juniper stands. Great Basin Naturalist 41: 467-475. 292. Everett. R, L, and S. H Sharrow. 1983. Response of understory species to tree harvesting and fire in pinyon-juniper woodlands. USDA Forest Service CJi'ui'ral Ti-ihnical liepoit lNT-i57. Intcrmonn- tain Forest and Range I'lxperinicnt Station, Og- den, Utah. 293. E\ ERErr, R L , S II Smarrovv , and R () Meeuwk;. 1983. Pinyon-juniper woodland understory distri- bution patterns and species associations. Bulletin of the Torrey Botanical Club 110: 454-463. 294. EvERETr. R. L.. and S H Sharrow. 1985. Under- story response to tree harvesting in singlcleaf pinyon and Utah juniper. (Ireat Basin Naturalist 45: 105-112. 295. Everett, R. L , S 11 Sharrow. and D. Thran. 1986. Soil nutrient distribution under and adjacent to singlcleaf pinyon crowns. Soil Science Society of America Journal 50: 788-792. 296. Everett, R. L,, and K. Ward. 1984. Early plant succession on pinyon-juniper controlled burns. Northwest Science 58: 57-68. 297. Everitt, B L. 1980. Vegetation and sediment mi- gration in the Henry Moimtains region, Utah. Pages 208-239 in M. Picard, ed.. Proceedings Utah Geological and Mineral Survey — 1980 Henry Mountains Symposium. Geological Associ- ation, Salt Lake City, Utah. 298. Fautin, R. W. 1941. Biotic communities of the northern desert shrub biome in western Utah. Unpublished dissertation. University of Illinois, Urbana. 299. Fautin, R. W 1946. Biotic communities of the northern desert shrub biome in western Utah. Ecological Monographs 16; 251-310. 300. Fears. R. D. 1966. The effects of intensity and season of use on recovery of desert range plants. Unpublished thesis, Utah State University, Logan. 301. Fenemore. R M . Jr. 1970. Plant succession in a receding lake bed in the western Great Basin. Unpublished thesis. University of Nevada, Reno. 302. Ferguson. C. W. 1962. Annual ring studies of desert shrubs. Unpublished progress report financed by National Science Foimdation Grant G-5568. 5 pp. 303. Fernandez, D. A., and M. M. Caldwell 1975. Phenology and dynamics of root growth of three cool semi-desert shrubs under field conditions. Journal of Ecology 63; 703-714. 304. Folliott, P. F., and W P Clary. 1982. Under- story-overstory vegetation relationships: an anno- tated bibliography. USDA Forest Service General Technical Report INT-136. Intermountain Forest and Range Experiment Station, Ogden, Utah. 39 pp. 305. Fireman, M., and H. E. Hayward. 1952. Indicator significance of some shrubs in the Escalante Desert, Utah. Botanical Gazette 114; 143-155. 306. Fisher. M 1978. A survey of plants and animals of Hill and Wendover Bombing Ranges, western Utah. Unpublished report prepared for the De- partment of Wildlife Science, Utah State Univer- sity, Logan. 70 pp. 307. Flowers, S. 1934. Vegetation of the Great Salt Lake region. Botanical Gazette 95: 353-418. 308. Flowers. S. 1942. Plant life of the Great Salt Lake region. News Bulletin of the Mineralogical Society ofUtah 3(3): 36-56. 220 P. S. BOURGERON ET AL. [Volume 50 309. Flowers, S. 1959. Vegetation of Glen Canyon. Uni- 326. versity of Utah Anthropological Papers 40: 21-61. 310. Flowers, S. 1960. Vegetation of Flaming Gorge Reservoir Basin. Pages 1-98 in Studies of the flora 327. and fauna of Flaming Gorge Reservoir Basin, Utah and Wyoming. University of Utah Anthropological Papers 48. Salt Lake City, Utah. 328. 311. Folks, F. N. 1969. Evaluation of five methods for sampling desert and wetland vegetation. Unpub- lished thesis, Utah State University, Logan. 329. 312. FoRSLiNG, C. L., AND E. V. Storm. 1929. The utiliza- tion of browse forage as summer range for cattle in southwestern Utah. USDA Circular 62. Washing- ton, D.C. 29 pp. 313. FoRSLiNG, C. L. 1931. A study of the influence of 330. herbaceous plant cover on surface run-off and soil erosion in relation to grazing on the Wasatch Plateau in Utah. USDA Agricultural Technical Bulletin 220. 71 pp. 331. 314. FOSBERG. F. R. 1938. The Lower Sonoran in Utah. Science 87; 39-40. 315. Foster. R. H. 1968. Distribution of the major plant communities in Utah. Unpublished dissertation, 332. Brigham Young University, Provo, Utah. 124 pp. 316. Foster, R. M. 1980. Threatened and endangered plant inventory: Wells Resource Area, Nevada. Unpublished final report prepared for USDI, 333. Bureau of Land Management, Elko, Nevada, un- der contract YA-512-CT9-125, by Inter-Mountain Research. 317. Freeman, J. AND J L. Mahoney 1977. Geothermal areas in Nevada. The distribution of vascular 334. plants near the thermal springs surveyed. Mentzelia3:7-14. 318. Frischknecht, N C 1975. Native faunal relation- 335. ships within the pinyon-juniper ecosystem. Pages 55-65 in Pinyon-juniper Ecosystem: A Sympo- sium, Utah State University, Logan. 319. Frischknecht, N C. and L. E. Harris 1968. Graz- .336. ing intensities and systems on crested wheatgrass in central Utah: response of vegetation and cattle. U.S. Department of Agriculture Technical Bul- letin 1388. Washington, D.C. 47 pp. 337. 320. Gardiner, H G 1984. Dynamics of arid land, perennial plant populations with an examination of potential causal agents. Unpublished dissertation, Utah State University, Logan. 321. Gates, D H. 1956. Ecology of plant distribution on .338. the salt-deserts of Utah. Unpublished disserta- tion, Utah State University, Logan. 322. Gates, D. H., L. A Stoddart, and C. W Cook 339. 1956. Soil as a factor influencing plant distribution on salt-deserts of Utah. Ecological Monographs 26: 155-175. 323. Genwal Coal Company. 1988. Vegetation refer- ence area. Volume 3, chapter 9, item 9-2, pages 340. 1-8 in Mining and Reclamation Plan for Crandall Canyon Mine, Genwal Coal Company. Utah Divi- sion of Oil, Gas and Mining Number ACT/015/ 032. Salt Lake City, Utah. 341. 324. Gehmano, D J., and D N. Law head. 1986. Species diversity and habitat complexity: Does vegetation organize vertebrate conununities in tlie Great 342. Basin? Great Basin Naturalist 46: 711-720. 325. GiFKORD, G. F., and F. E. Busby, eds. 1975. The pinyon-juniper ecosystem: a symposium. Utah State University, Logan. 194 pp. Giunta, B. C. 1979. Utah big game range inventory, 1977. Utah Division of Wildlife Resources Publi- cation Number 79-3. Salt Lake City, Utah. 257 pp. Giunta, B C 1980. Utah big game range inventory, 1978. Utah Division of Wildlife Resources Publi- cation Number 80-9. Salt Lake City, Utah. 333 pp. Giunta, B. C 1982. Utah big game range inventory, 1980. Utah Division of Wildlife Resources Publi- cation Number 82-5. Salt Lake City, Utah. 262 pp. Giunta, B., D. Barnhur.st, B Hosea, T. Howick, and J. Teegarden 1986. Utah big game range trend studies, 1984. Utah Division of Wildlife Re- sources Publication Number 86-6. Salt Lake City, Utah. 443 pp. Giunta. B C . and R Musclow. 1983a. Utah big game range trend studies, 1981. Utah Division of Wildlife Resources Publication Number 83-1. Salt Lake City, Utah. 189 pp. Giunta, B. C . and R. Musclow. 1983b. Utah big game range trend studies, 1982. Utah Division of Wildlife Resources Publication Number 83-2. Salt Lake City, Utah. 421 pp. Giunta, B. C , and R. Musclow. 1984. Utah big game range trend studies, 1983. Utah Division of Wildlife Resources Publication Number 84-3. Salt Lake City, Utah. 399 pp. Giunta, B. C., R Ste\ens, K. R Jorgensen, and A. P Plummer 1978. Antelope bitterbrush, an important wildland shrub. Utah Division of Wildlife Research Publication 78-12. Salt Lake City, Utah. 47 pp. Gleason, H. A., AND A. Cronquist 1964. The natu- ral geography of plants. Columbia University Press, New York. 420 pp. Goebel, C J 1960. Effect of range condition and plant vigor upon the production and nutriti\e \alue of forage. Lhipublished dissertation, Utah State University, Logan. GoETZ. H. 1968. Vegetation and soil responses to nitrogen fertilization on different range sites. Unpublished dissertation, Utah State University, Logan . Gonzales, M. H. 1964. Patterns of livestock behav- ior and forage utilization as influenced by environ- mental factors on a summer mountain range. Unpublished dissertation, L'tah State University, Logan. Goodrich. S. 1981. A floristic study of central Ne- vada. Unpublished thesis, Brigham Young Uni- versity, Pro\ o, Utah. 400 pp. Goodwin, D L 1963. Vegetation and soils relation- ships of the shadscale zone in Utah. Pages 24-25 in Proceedings of the American Society of Range Management, 12-15 February 1963, Rapid City, South Dakota. Graham, E H 1937. Botanical studies in the Uinta Basin of Utah and Colorado. Annals of the Carnegie Museum, Vol. 26. Carnegie Institute, Pittsburgh, Pennsylvania. 432 pp. Graybosch, R. A , AND II Buchanan. 1983. Vegeta- ti\e t\pes and endemic plants of the Bryce Canxon Breaks, (ireat Basin Naturalist 43: 701-712. (Greenwood. L. R , and J D Brotherson. 1978. Ecological relationships between pinyon-juniper and true mountain mahogan\ stands in the Uintah Basin, Utah. Journal of Range Management 31: 164-167. 1990] Nevada-Utah Vi'XiKTATioN Bibliociufhy 221 343. Grikkitiis, D. 1902. Forage conditions in the north- ern border oi the Great Ba.sin. USDA Bureau ol Phint Industry Bulletin 15. 344. Gri'Kli,. G E., S. Buntinc:, and L. Nkuhnscuvvan- DKH 1985. Influence of hre on curlleaf mountaiu- nialiogau\' in the Interniountain West. Pages 58-72 ill J. E. Lotau and J. K. Brown, compilers. Fire's effects on wildhle hal)itat — symposium pro- ceedings. USDA Forest Service Cieneral Techni- cal Report INT- 186. Intermountain Forest and Range Experiment Station, Ogden, Utah. 345. Haa.s, R. H . H L Mokton. and P. J. Torell. 1962. Influence of soil salinity and 2,4-D treatments on establishment ol desert wheatgrass and control oi halogeton and other annual weeds. Joiunal of Range Management 15; 205-210. 346. Hall, H H 1954. The impact of man on the vegeta- tion and soil of the Upper Valley Allotment, Garfield County, Utah. Unpublished thesis, Uni- versity of Utah, Salt Lake City. 347. Hall, H H 1971. Ecology of a subalpine meadow of the Aquarius Plateau, Garfield and Wayne Coun- ties, Utah. Unpublished dissertation, Brigham Young University, Provo, Utah. .348. Halverson. R M 1974. ERTS-1 imagery for vege- tation mapping and phenology detection in Ne- vada. Unpublished thesis. University of Nevada, Reno. 349. Hanks, L M 1982. The pollinator community of a sand dune ecosystem. Unpublished thesis. Uni- versity of Nevada, Reno. 350. Hansen, D. J 1977. Interrelations of valley vegeta- tion, stream regimen, soils, and solar irradiation along the Rock Creek in the Uinta Mountains of Utah. Unpublished dissertation. University of Michigan, Ann Arbor. 351. Hansen, D. S., P. Dayanandan, P. B. Kaufman, and J. D. Brotherson 1976. Ecological adaptations of salt marsh grass, Distichlis spicata (Gramineae), and environmental factors affecting its growth and distribution. American Journal of Botany 63: 635- 650. 352. Hansen. H. P. 1947. Postglacial vegetation of the northern Great Basin. American Journal of Botanv 34: 164-171. 353. Hanson, C A. 1962. Perennial Atriplex of Utah and the northern deserts. Unpublished thesis. Brig- ham Young University, Provo, Utah. 354. Hanson, H C 1950. Ecology of the grassland. II. Botanical Review 16; 283-360. 355. Hanson. W R. 1939. The ecology of Agropijron inerme on protected and heavily grazed range land in Cache Valley, Utah. Unpublished thesis, Utah State University, Logan. 33 pp. 356. Hanson. W R, and L. A. Stoddart 1940. Effects of grazing upon bluebunch wheat grass. Journal of the American Society of Agronomy 32; 279-289. 357. Harner, R. F., and K. T. Harper. 1973. Mineral composition of grassland species of the eastern Great Basin in relation to stand productivity. Canadian Journal of Botany 51; 2037-2046. 358. Harner, R. F., and K. T. H.'\rfer 1976. The role of area, heterogeneity and favorability in plant spe- cies diversity of pinyon-juniper ecosystems. Ecol- ogy 57; 1254-1263.' 359. Harnlss, R O , and K T Harrer. 1982. Tree dy- namics in serai and stable aspen stands in central Utah. USDA Forest Service Research Paper INT- 297. Intermountain Forest and Range Experiment Station, Ogden, Utah. 7 pp. 360. Harrer. K T 1959. Vegetational changes in a shad- scale-winterfat plant association during twenty- three years of controlled grazing. Unpublished thesis, Brigham Young University, Provo, Utah. 72 pp. 361. llARi'ER, K. T 1967. The vegetational environment of the Bear River Number 2 archaeological site. University of Utah Anthropological Paper 87. Salt Lake City, Utah. 362. Harper, K. T. 1976a. Permanent plot native vegeta- tion studies. Pages 20-43 in A. C. Hill et al., eds.. Vegetation-air pollution investigations in the vicinity of Four Corners and San Juan power plants. New Mexico. University of Utah Research Institute Environmental Studies Laboratory, Salt Lake City. 363. Harper, K T 1976b. Work plan; ecological impact study of weather modification in the Uinta Moun- tains, Utah. Unpublished report prepared for Utah Division of Water Resources, Salt Lake City, and the USDI, Bureau of Reclamation Office of Atmospheric Resources Management, Denver, Colorado. 46 pp. 364. Harper, K. T. 1977. The Utah ecology project: eco- logical impact of weather modification studies in the Uinta Mountains, Utah. Unpublished interim report prepared for Utah Division of Water Re- sources, Salt Lake City, and the USDI, Bureau of Reclamation, Denver, Colorado. 365. Harper. K. T 1978. The Uinta ecology study; bi- ennial report. Unpublished report prepared for Utah Divison of Water Resources, Salt Lake City, and USDI, Bureau of Reclamation, Denver, Colo- rado. 24 pp. 366. Harper, K. T 1979. Some reproductive and life history characteristics of rare plants and implica- tions for management. Great Basin Naturalist Memoirs 3; 129-1.38. 367. Harper, K. T. 1985. Predicting successional rates in Utah aspen forests. Pages 96-100 in Proceedings of the 1985 Society of American Foresters National Convention, 28-31 July 1985, Bethesda, Mary- land. 445 pp. 368. Harper, K T., and C. S. Climer. 1985. Factors af- fecting productivity and compositional stability of Artemisia steppes in Idaho and Utah, U.S.A. Pages 592-594 in Proceedings of XVth Inter- national Grassland Conference, August, Kyoto, Japan. 369. Harper, K T , D C Freeman. W K Ostler, and L G. Klikoff. 1978. The flora of Great Basin mountain ranges: diversity, sources and disper- sal ecology. Great Basin Naturalist Memoirs 2: 81-103. 370. Harper. K T , and R A. Jaynes. 1986. Some edaphic and compositional characteristics oi Artemisia tri- dentata and associated plant communities in southeastern Utah. Pages 265-272 in E. D. McArthur and B. L. Welch, compilers, Proceed- ings— Symposium on the Biology oi Artemisia and Chrysuthamnits, 9-13 July 1984, Provo, Utah. 222 P. S. BOURGERON ETAL. [Volume 50 USDA Forest Service General Technical Report INT-200. Intermountain Research Station, Og- den, Utah. 371. H.ARPER, K T , J R MuRDOCK. AND S L. Welsh. 1974. Botanical evaluation of proposed power plant sites in Garfield and Wayne Counties, Utah. Unpublished report prepared for Environmental Systems Department of Westinghouse Corpora- tion, Salt Lake City, Utah. 372. Harper, K T., J R Murdock, and S L. Welsh. 1975. The vegetation of Salt Wash Area, Wayne County, Utah. Unpublished report prepared for Environmental Systems Department of Westing- house Corporation, Salt Lake City, Utah. 30 pp. 373. Harper. K T.. W K Ostler, and DC Anderson. 1978. The Utah ecology project; ecological impact of weather modification studies in the Uinta Mountains, Utah. Unpublished special report prepared for Utah Division of Water Resources, Salt Lake City, and USDI, Bureau of Reclamation, Denver, Colorado. .374. Harper. K T , W K. Ostler, and K. B McKnic.ht 1979. The Uinta ecology study, interim report. Unpublished report to Utah Division of Water Resources, Salt Lake City, and USDI, Bureau of Reclamation, Denver, Colorado. 76 pp. 375. Harper, K. T , W K Ostler. K B McKnight. and D. L. Hunter. 1981a. Potential ecological impacts of snow pack augmentation in the Uinta Moun- tains, Utah. Unpublished report prepared for USDI, Bureau of Reclamation, Denver, Colo- rado. 291 pp. 376. Harper, K T., W. K Ostler. K B McKnight, and D L HlNTER. 1981b. Effects oflate-lyingsnowon an alpine herbland in the Uinta Mountains. Pages 179-203 in K. T. Harper, W. K. Ostler, K. B. McKnight, and D. L. Hunter, eds.. Potential eco- logical impact of snow pack augmentation in the Uinta Mountains of Utah. Unpublished report prepared for USDI, Bureau of Reclamation, Den- ver, Colorado. 377. Harper. K T. and J L Reveal, eds 1978. Liter- mountain biogeography; a symposiinn. Great Basin Naturalist Memoirs 2: 1-268. .378. Harper, K. T, S C Sanderson, and E D McArthur 1987. Vegetation communities of Zion National Park. Pages 191-196 in University of Wyoming National Park Service Research Cen- ter eleventh annual report. Laramie, Wyoming. .379. Harper, K. T., F. J. W.\gstaff. and L. M. Kunzler. 1985. Biology and management of the Gambel oak vegetative type: a literature review. USDA Forest Service General Technical Report INT-179. Inter- mountain Forest and Range E.vperiment Station, Ogden, Utah. 380. Harper, K. T., S. Waite, andC. Lamb. 1973. Vegeta- tion of the proposed power generating plant sites. Pages 1-33 in Evaluation of three alternative power plant sites in Utah and Wyoming for the Utah Power and Light Companv, Salt Lake Citv, Utah. 381. Harper, K T . B W Wood, and J D Brotherson 1974. A vegetational analysis of areas inxolved in the proposed Foiu" Seasons Project. Pages 60-104 in unpublished environmental rept)rt on proposed Four Seasons Project prepared for Brighani Young University Center lor Health and Environmental Studies, Provo, Utah. 382. Harper, K. T., R. A. Woodward, and K. B McKnight 1980. Interrelationships among pre- cipitation, vegetation, and streamflow in the Uinta Mountains, Utah. Encyclia 57: .56-86. .383. Harris. M L 1926. An ecological study of Tim- panogos Creek, from Aspen Grove to Wildwood. Unpublished thesis, Brigham Young University, Provo, Utah. 82 pp. 384. Harrison, B.F 1980. Botanical survey, threatened, endangered and other rare plants of the Schell Resource Area, Ely District, Nevada. Unpub- lished report prepared for USDI, Bureau of Land Management, by Inter-Mountain Research Cor- poration, Provo, Utah. 101 pp. 385. Hassan. M. A. 1983. Effects of wildfire on seed- rain and soil. Seed reserve dynamics of a good condition sagebrush-grass rangeland in central Utah. Unpublished thesis, Utah State University, Logan. .386. Hassan, M A , and N E West. 1984. Soil seed reserve dynamics in burned and unburned sage- brush-grass vegetation of central Utah. Bulletin of Ecological Society America 65: 180. 387. Havward, C. L. 1943. Biotic communities of the coniferous forests of the southern Wasatch and Uinta Mountains, Utah. Unpublished thesis, Uni- versity of Illinois, Urbana. 388. Hayward, C. L. 1945. Biotic communities of the southern Wasatch and Uinta Mountains, Utah. Great Basin NaturaHst 6: 1-124. 389. Havward, C L 1948. Biotic communities of the Wasatch Chaparral, Utah. Ecological Monographs 18: 473-506. 390. Havward. C. L 1952. Alpine biotic communities of the Uinta Moimtains, Utah. Ecological Mono- graphs 22: 9.3-120. .391. Havward, C L., D E. Beck, and W. W. Tanner. 1958. Zoology of the Upper Colorado River Basin. Part I, The biotic communities. Brigham Young University Science Bulletin, Biological Series 1(3): 1-74. .392. Heinze, D H , R E Eckert, and P T. Tueller. 1962. The vegetation and soils of the Steptoe Watershed. Unpublished report prepared for USDI, Bureau of Land Management. 40 pp. .393. Henderson.] A 1978. ECOSYM— potential vege- tation classification. Appendi.\ report 7 in J. A. Henderson, L. S. Davis, and E. M. Ryberg, eds., ECOSYM: an ecosystem classification and data storage system for natural resources management. Unpublished report prepared for Utah State Uni- \ ersity, Logan. 12 pp. .394. Henderson. J A , R L Mauk. D. L Anderson. T a. D.wis. and T J Keck. 1977. Preliminary forest habitat t\ pes of the Uinta Moimtains, Utah. Unpublished report prepared for Department of F"orestr\'and Outdoor Recreation, I' tab State Uni- versity, Logan. 94 pp. .395. Henderson, J. A., S. A. Simon, and S B Hart- \ ICSEN. 1977. Plant community types and habitat types of the Price District Manti-La Sal National Forest. Unpublished report prepared for Depart- ment of Forestry and Outdoor Recreation, Utah State University, Logan. 1990] Nevada-Utah Vec;k'1"atk)n Biblioghaphv 223 396. 397. 398. 399. 400. 401. 402. 403. 404. 405. 406. 407. 408. 409. Henderson, J A., and N. E West 1977. 410. ECOSYM — vej:;ctati()ii cla.ssifkation. Appendix report 6 in J. A. Ilender.son and L. S. Davis, eds., ECOSYM : a cla.ssification and data .storage systcTn 411. for natural resources management. Unpublished report prepared for I'tali State Uni\ ersit\ , Logan. 89 pp. Henningson, DtiRH.\.M, & RICHARDSON Inc.: 1981a. Preliminary FEIS, grazing. M-X Enviroiniiental 412. Technical Report 40. Unpublished report pre- pared for USAF, Norton Air Force Base, Cahfor- nia. 168 pp. 413. Henningson. Durham, & Richardson Inc 1981b. Preliminary FEIS, native vegetation. M-X Envi- ronmental Technical Report 14. Unpublished re- port prepared for USAF, Norton Air Force Base, Cahfornia. 197 pp. 414. Henningson, Durham, & Richardson, Inc. 1981c. Preliminary FEIS, aquatic habitats and biota. M-X Environmental Technical Report 16. Unpub- lished report prepared for USAF, Norton Air Force Base, California. 70 pp. 415. Herman, F 1953. A growth record of the Utah juniper. Journal of Forestry 51: 193-195. Hessing, M B , C D Johnson, and R P Balda 1982. Early secondary succession of a pinyon- juniper woodland in a northern Arizona powerline 416. corridor (Junipenis monospcnna, Pintis cdtilis). Southwestern Naturalist 27: 1-9. HiEBERT, R D., AND J L Hamrick 1984. An ecolog- ical study of bristlecone pine (Piniis lon^aeva) in Utah and eastern Nevada. Great Basin Naturalist 44: 487-494. 417. Higgins, L. C. 1967. A flora of the Beaver Dam Mountains. Unpublished thesis, Brigham Young University, Provo, Utah. 305 pp. Hill, A. C, K. T Harper. .\nd W H. Edwards 418. 1972. Ecological evaluation of proposed routes for the Huntington-Sigiud power transmission line. Unpublished report prepared for Utah Power and Light Company, Salt Lake City, Utah. 419. Hilton, J. W. 1941. Effects of certain micro-ecologi- cal factors on the germinability and early develop- ment of Eiirota laiiata. Northwest Science 15: 86-92. Hoffman, W. J. 1877. The distribution of vegetation 420. in portions of Nevada and Arizona. American Nat- uralist 11: 336-343. Holland, J 1982. A floristic and vegetation analysis 421. of the Newberry Mountains, Clark County, Ne- vada. Unpublished thesis. University of Nevada, Las Vegas. Reproduced in the Lake Mead Report Series by Lhiiversity of Nevada Cooperative Na- 422. tional Park Resources Studies Unit, Las Vegas, Nevada. Holland. J. S,. W. E Niles, and P. J. Leary. 1979. Vascular plants of the Lake Mead National Recre- 423. ation Area. Lake Mead Technical Report 3. Uni- versity of Nevada Cooperative National Park Re- sources Studies Unit, Las Vegas, Nevada. 223 pp. Holland, J. S., W E Niles, and D. R Schram 424. 1980. A guide to the threatened and endangered vascular plants of the Lake Mead National Recre- ation Area. Lake Mead Technical Report 4. Uni- versity of Nevada Cooperative National Park Re- 425. sources Studies Unit, Las Vegas, Nevada. 75 pp. Holmgren. A II 1948. Handbook of the \a.scular plants of the northern Wasatch. Lithotype Process (Company, San Francisco. 202 pp. Holmgren, A. H 1962. The vascular plants of the Green River from the Flaming Gorge to Split Moimtain (Jorge. Unpublished report prepared for the' National Park Service by Utah State Uni- versity, Logan. 40 pp. Holmgren, A. H 1979. Strategies for preservation of rare plants. Great Basin Naturalist Memoirs 3: 95-100. Holmgren, A H., and J. L Reveal. 1966. Checklist of the vascular plants of the Intermountain Re- gion. USDA Forest Service Research Paper INT- 32. Intermountain Forest and Range Experiment Station, Ogden, Utah. 160 pp. Holmgren, R C, 1975. The desert experimental range description history and program. Pages 18-22 /;i Arid Shrubland Proceedings of the Third Workshop of the United States/Australia Range- lands Panel, Tucson, Arizona, 1973. Holmgren, R. C, and S F Brewster, Jr 1972. Distribution of organic matter reserve in a desert shrub community. USDA Forest Service Re- search Paper INT-130. Intermountain Forest and Range Experiment Station, Ogden, Utah. Holmgren, R. C, and S S. Hutchings. 1972. Salt desert shrub response to grazing use. Pages 153-164 in Wildland shrubs — their biology and utilization. USDA Forest Service General Techni- cal Report INT-GTR-1. Intermountain Forest and Range Experiment Station, Ogden, Utah. Hosea, B , and D Barnhurst. 1988. Utah big game range trend studies, 1986. Utah Divison of Wild- life Resources Publication Number 88-5. Salt Lake City, Utah. 305 pp. Houston, W R 1951. A preliminary study of some factors affecting forage production in the aspen type of central Utah. Unpublished thesis. Univer- sity of Utah, Salt Lake City. Houston. W R. 1954. A condition guide for aspen ranges of Utah, Nevada, southern Idaho, and western Wyoming. USDA Forest Service Re- search Paper 32. Intermountain Forest and Range Experiment Station, Ogden, Utah. 25 pp. Hubbard, K G. 1980. Relating fire occurrence to weather conditions on the Great Basin rangelands. Journal of Range Management 33: 360-362. Huff, C. L 1962. Utah big game range inventory, 1961. Utah Division of Wildlife Resources Depart- mental Information Bulletin Number 62-5. Salt Lake City, Utah. 45 pp. Hi'FF, C L 1963. L^tah big game range inventory, 1962. Utah Division of Wildlife Resources Depart- mental Information Bulletin Niuiiber 63-2. Salt Lake City, Utah. 161 pp. Huff. C. L , and R. Blotter 1964. Utah big game range inventory, 1963. Utah Division of Wildlife Resources Departmental Information Bulletin Number 64-2. Salt Lake City, Utah. 110 pp. Huff, C L , and J. E. Bowns, Jr 1965. Utah big game range inventory, 1964. Utah Division of Wildlife Resources Publication Number 65-1. Salt Lake City, Utah. 111pp. Huff, C. L., and F H. Coles. 1966. Utah big game range inventory, 1965. Utah Division of Wildlife 224 P. S. BOURGERON ET AL. [Volume 50 Resources Publication Number 66-3. Salt Lake City, Utah. 96 pp. 426. Hull, A. C. Jr., and M. K. Hull. 1974. Presettle- ment vegetation of Cache Valley, Utah and Idaho. Journal of Range Management 27: 27-29. 427. Hunt, C B , P. Averitt, and R L. Miller. 1953. Geology and geography of the Henry Mountains region, Utah. USDI Geological Survey Profes- sional Paper 228. Washington, D.C. 428. HuTCHENS, S K, AND K T Harper 1965. A method for determining grazing indicators in aspen com- munities. Proceedings of Colorado-Wyoming Academy of Science, Arts and Letters 42; 319. 429. Ibrahim, K. 1962. Plant communities of the shad- scale zone in Grand County as related to soil and geology. Proceedings of Utah Academy of Sci- ence, Arts and Letters 37-39: 191-193. 430. Ibrahim, K M 1963. Ecological factors influencing plant distribution in the shadscale zone of south- eastern Utah. Unpublished dissertation, Utah State University, Logan. 431. Ibrahim, K. M. , N E. West, and D. L. Goodwin. 1972. Phytosociological characteristics of peren- nial Afnp/c.t -dominated vegetation of southeast- ern Utah. Vegetatio24: 13-22. 432. Irvine, J R 1970. Riparian environmental-vegeta- tion interrelationships along the Escalante River, Glen Canyon National Recreation Area, Utah. Unpublished thesis, Utah State University, Lo- gan. 82 pp. 433. Irvine, J. R., and N. E West 1976. Riparian envi- ronmental-vegetation interrelationships along the lower Escalante River, Glen Canyon National Recreation Area, Utah. Unpublished report pre- pared for the National Park Service, Utah. 82 pp. 434. Irvine, J. R., and N. E. West. 1979. Riparian tree species distribution and succession along the lower Escalante River, Utah. Southwestern Natu- ralist 24: 331-346. 435. Isaacson, H E 1967. Ecological provinces within the pinyon-juniper type of the Great Basin and Colorado Plateau. Unpublished thesis, Utah State University, Logan. 436. Ives, R. L. 1942. Atypical subalpine environments. Ecology 23: 89-96. 437. Jackson, L E 1985. Floristic analysis of the distri- bution of ephemeral plants in treeline areas of the western U.S. Arctic and Alpine Research 17: 251-260. 438. Jackson, L. E , and L. C. Bliss. 1984. Phenology and water relations of three plant life forms in a dry tree-lined meadow. Ecology 65: 1302-1314. 439. Jarek, L B 1986. A photographic history of vegeta- tion changes on the Humbolt National Forest. Unpublished thesis. University of Nevada, Reno. 440. Jatkar, S. a., S. R Rushforth, and J. D. Bhotheh- SON. 1979. Diatom floristics and succession in a peat bog near Lily Lake, Summit County, Utah. Great Basin Naturalist 39: 15-43. 441. Jaynes, R. a , AND K. T Harper. 1978. Patterns of natural revegetation in arid southeastern Utah. Joiunal of Range Management 31: 407-41 1 . 442. Jensen, G H 1940. The relation of some ph\sical and clu'mical factors of the soil to the producti\it\ and distribution ol certain waterfowl food plants at Bear River Migratory Waterfowl Refuge. Unpub- lished thesis, Utah State University, Logan. 443. 444. 445. 446. 447. 448. 449. 450. 451. 452. 4.53. 454. 455. 456. Jensen, M E 1989. Soil characteristics of moun- tainous northeastern Nevada sagebrush commu- nity types. Great Basin Naturalist 49: 469-481. Jensen, M. E., L. S. Peck, and M V Wilson, 1988. A sagebrush community type classification for mountainous northeastern Nevada rangelands. Great Basin Naturalist 48: 422-433. Jensen, S. E., and J. S. Tuhy. 1981. Soils investiga- tion of riparian communities of East Smiths Fork and Henrys Fork drainages. North Slope Uinta Mountains, Utah. Unpublished report presented to USDA Forest Service, Ogden, Utah. 35 pp. Johnson, C M. 1970. Common native trees of Utah. USDA Forest Service Special Report 22. Utah Agricultural E.xperiment Station, Logan, Utah. 109 pp. Johnson, H. B 1964. Changes in the vegetation of two restricted areas of the Wasatch Plateau as related to reduced grazing and complete protec- tion. Unpublished thesis, Brigham Young Univer- sity, Provo, Utah. 123 pp. Johnson, J. L., and R. D. Pfister 1982. A survey of potential ecological natural landmarks of the Mid- dle Rocky Mountains. Unpublished report pre- pared for USDI, National Park Service by USDA, Forest Service Intermountain Forest and Range E.xperiment Station, Ogden, Utah. 197 pp. Johnson, K L 1986. Sagebrush types as ecological indicators to integrated pest management (IPM) in the sagebrush ecosystem of western North Amer- ica. Pages 1-10 in J. A. Onsager, ed.. Integrated pest management in rangeland: state of the art — in sagebrush ecosystem, 27-28 March 1984, Univer- sity of Nevada, Reno. Johnson, K, L., ed. 1989. Rangeland resources of Utah. Utah Rangeland Committee report. Coop- erative Extension Service, Utah State University, Logan. 103 pp. Johnson, N. K 1978, Patterns of Asian geography and specialism in the Intermountain Region. Great Basin Naturalist Memoirs 2: L37-160. Johnson, R V 1929, A taxonomic-ecological survey of the Truckee Meadows. Unpublished thesis, University of Nevada, Reno. Kaiser Coal CoRPOR,\TiON. 1985a. Vegetation. Vol- ume 3, chapter 8, sections 8.0-8.3, pages VIII- l-VIII-21, plus appendices VIII- 1, VIII-3, VIII-5 in Mining and Reclamation Plan for Horse Canyon Mine, Kaiser Coal Corporation. Utah Divison of Oil, Gas and Mining Number ACT/()07/013. Salt Lake City, Utah. Kaiser Coal Corporation. 1985b. Vegetation re- sources. Book 7, chapter 9, sections 9.1-9.3, pages 1-8, plus tables IX-l-IX-38, plus appendix IX-1 in Mining and Reclamation Plan for Sunny- side Mine, Kaiser Coal Corporation. Utah Divi- sion of Oil, Gas and Mining Num!u>r ACT/007/ 007. Salt Lake City, Utah. Kay, P a., andC. G. Oviatl 1978. Pinuslonanyonlands National Park, Utah. Unpublished dissertation, Utah State University, Logan. 505. LooPE, W. L., AND N. E. West. 1979. Vegetation in relation to environments of Canyonlands Na- tional Park. Pages 195-199 in R. M. Linn, ed.. Proceedings, First Conference of Scientific Re- search in the National Parks, Vol. I, 9-13 Novem- ber 1976, New Orleans, Lousiana. LISDI, Na- tional Park Service Transactions and Proceeding Series Number 5. 506. LoR.\lN, G. E 1970. 70 mm large scale aerial photog- raphy for range resources evaluation. Unpub- lished thesis. University of Nevada, Reno. 93 pp. 507. LoTAN, J E 1975. The role of cone serotiny in lodge- pole pine forests. Pages 471-495 in Proceedings, Symposium on Management of Lodgepole Pine, Washington State University, Pullman. 508. Louderbough, E. T, and L. D Potter. 1982. Man- cos shale: some physical and chemical properties which affect vegetative communities on shale out- crops. Pages 547-550 in Proceedings, Symposium on Surface Mining Hydrology, Sedimentology and Reclamation, University of Kentucky, Lexington. 509. LOVEJOY, T. E. 1979. The epoch of biotic impover- ishment. Great Basin Naturalist Memoirs 3: 5-10. 510. LUDWIG, J. A 1965. Relationships of grasslands to edaphic gradients of the foothill zone, Salt Lake County, Utah. Unpublished thesis. University of Utah, Salt Lake City. 511. LuDWiG, J A. 1969. Environmental interpretation of foothill grassland communities of northern Utah. Unpublished dissertation. University of Utah, Salt Lake City. 100 pp. 512. M.ACDOL'GAL, D T 1908. Sagebrush deserts of Ne- vada and Utah. Pages 28-30 in Botanical features of North American deserts. Carnegie Institute of Washington Publication 99. 513. MacMahon. J A , andT F Wieboldt. 1978. Ap- plying biogeographic principles to resource man- agement: a case stud\ e\ aluating Holdridge s Life Zone Model. Great Basin Naturalist Memoirs 2: 245-257. 514. Madany. M H 1980. Petersboro Prairie. Unpub- lished report on field project prepared for Depart- ment of Range Science, Utah State University, Logan, Utah. 19 pp. 515. Madany, M. H 1981. Land use-fire regime interac- tions with vegetation structure of several montane forest areas of Zion National Park. Unpublished thesis, Utah State University, Logan. 103 pp. 516. Madany. M. H No date. Ponderosa pine forests of the Colorado Plateau. Unpublished report pre- pared for Dr. N. E. West, Department of Range Science, Utah State University, Logan. 517. Madany, M H , and N E West 1980a. Fire history of two montane forest areas of Zion National Park. Pages 50-56 in Proceedings of the Fire His- tor\' Workshop. USDA Forest Service General Technical Report RM-81. Rocky Mountain Forest and Range Experiment Station, Fort Collins, Col- orado. 518. Madany, M H., and N E West 1980b. Fire regime and land use interactions in ponderosa pine forests of Zion National Park, Utah. Bulletin of the Ecological Societ\' of America 61: 122. 519. Madany, M H , and N E. West 1982. Montane vegetation of several relict mesas in Zion National Park, Utah. Abstracts of the Annual Meetings of the Society for Range Management 35: 35. 1990] Nevada-Utah Vecetation Bibi.i()(;hai'iiy 227 520. Madany. M H . AND N E VVkst 1983. Livrstock grazinji-firc rcuiinc intfiactions witliin montam' forests of Zion National I'ark, I'tali. EcoiotiN 64: 661-667. 521. Madany, M H . and N E VVe.st 19.S4. Ncm-tation of two relict mesas in Zion National I'ark. journal of Range Mangenient 27: 456-461. 522. Major, J. 1948. Elements of a range condition classi- fication of tile aspen t\pe in L'tah. Proceedings ol the Utah Acadenu of Science, ,\rts and Letters 25: 173-174. 523. Malanson, G. P. 1980. Habitat and plant distribu- tions in hanging gardens of the Narrones, Zion National Park, Utah. Great Basin Naturalist 40: 178-182. 524. Malanson, G. P. 1982. The assembly of hanging gardens: effects of age, area, and location. Ameri- can Naturahst 119: 145-150. 525. Mariah Associates 1981. Vegetation baseline data analysis, Alton coal lease study area. Volimie 6, chapter 3, appendi.x 3. 6-A ()i M ining and Reclama- tion Plan for Alton Mine, Utah International In- corporated. Utah Division of Oil, Gas and Mining Number ACT/025/003. Salt Lake City, Utah. 526. Mariah Associates. 1989. Vegetation data report. Volume 4, chapter 9, section 9.1, pages 1-47 in Mining and Reclamation Plan for Castle Gate Coal Mine, Castle Gate Coal Company. Utah Division of Oil, Gas and Mining Number ACT/007/004. 527. Markham, B. S 1939. A preliminary study of the vegetation cover in Spanish Fork Canyon, Utah. Unpublished thesis, Brigham Young University, Provo, Utah. 79 pp. 528. Marr, J. W., and D Buckner. 1973. Ecological analyses of potential shale oil products pipeline corridors in Colorado and Utah. Unpublished re- port to Colony Development Operation, Atlantic Richfield Company, Denver, Colorado, by Thorne Ecological Institute and the University of Colorado, Boulder. 529. M.\rr, J. W , D L. Buckner. and C Mutel. 1974. Reaction of vegetation and soils to impact from construction and operation of pipelines in western Colorado and eastern Utah. Unpublished report prepared for Colony Development Operation, Atlantic Richfield Company, Denver, Colorado. 138 pp. 530. Marston. R B 1963. Some comparative hydrologic characteristics of aspen and mountain brush com- munities on steep mountain watersheds in north- ern Utah. Unpublished dissertation, Utah State University, Logan. 531. Martin, W. E 1963. Close-in effects of an under- ground nuclear detonation on vegetation. I. Im- mediate effects of cratering, throw-out and blast (Project SEDAN). U.S. Atomic Energy Commis- sion Report PNE-228F. 532. Mason, L. R 1963. Using historical records to de- termine clima.x vegetation. Journal of Soil and Water Conservation 18: 190-194. 533. Mason, L. R 1971. Yield and composition of Utah's range sites. USDA Soil Conservation Service, Portland. Oregon. 534. Mason, L. R., H M. Andrews, J A Carley, and E D Haacke. 1967. Vegetation and soils of No Man's Mesa relict area. Journal of Range Manage- ment 20: 45-59. 535. Mason, L R , and V Mortensen 1977. Vegetation and soils of Zion National Park. Unpublished re- port prepared for USDA, Soil (Conservation Ser- vice, Salt Lake City, Utah. 536. M.\soN, L. R . and N. E West. 1970. Timber Top Mesa: a relict area in Zion National Park. Proceed- ings of Utah Academy of Science, Arts, and Let- ters 47: 284-285. 537. Mauk, R. L., and J. A Henderson. 1984. Conifer- ous forest habitat types of northern Utah. USDA Forest Service General Technical Report INT- 170. Intermountain Forest and Range Experiment Station, Ogden, Utah. 89 pp. 538. McDonald, J. E. 1946. Measuring the forage crop of range lands by vegetation analysis. Un- published thesis, Utah State Agricultural College, Logan. 539. McGraw, J. F 1980. Landsat computer-aided analysis techni(jues for range vegetation mapping. Unpublished thesis. University of Nevada, Reno. 540. McKell, C M. 1950. A study of plant succession in the oak brush {Querciis gamhelii) zone after fire. Unpublished thesis. University of Utah, Salt Lake City. 79 pp. 541. McLaughlin, S P 1986. Floristic analysis of the southwestern United States. Great Basin Natural- ist 46: 46-65. 542. McMillan, C. 1948. A taxonomic and ecological study of the flora of the Deep Creek Mountains of central western Utah. Unpublished thesis, Uni- versity of Utah, Salt Lake City. 99 pp. 543. McNuLTY. 1 B 1947. The ecology of bitterbrush in Utah. Unpublished thesis. University of Utah, Salt Lake City. 544. Meeuwig, R O , and S. Cooper 1981. Site quality and growth of pinyon-juniper stands in Nevada. Journal of Forestry Science 27: 593-601. 545. Meeuwig. R O , R B Murray, P. T Tueller, and S V. Cooper. 1977. Stratifying the pinyon-juniper woodlands according to biotic potentials. Unpub- lished report prepared for Nevada Agricultural Experiment Station, Reno. 546. Meiners. W R. 1965. Some geologic and edaphic characteristics useful to management programing within pinyon-juniper type woodlands. Unpub- lished thesis, Utah State University, Logan. 547. Meinke, R. J 1975. A preliminary ecological and historical survey of north and south Caineville Mesas, Wayne County, Utah. Unpublished report prepared for USDl, Bureau of Land Management, Richfield, Utah. 127 pp. 548. Menzies. C. W. 1935. Effects of overgrazing on the mortality of desert browse on the Utah West Desert. L^npublished thesis, Brigham Young Uni- versity, Provo, Utah. 549. Merria.m. C. H 1893. Notes on the distribution of trees and shrubs in the deserts and desert ranges of southern California, southern Nevada, north- western Arizona, and southwestern Utah. North American Fauna 7: 285-394. 550. Meyer, J. C, and R D. Comer. 1985. Reclamation investigations on disturbed lands at Glen Canyon National Recreation Area. Unpublished report prepared for USDI, National Park Service, by Thorne Ecological Institute, Boulder, Colorado. 10 pp. 228 P. S. BOURGERON ET AL. [Volume 50 551. Meyer, S. E. 1978. Some factors governing plant distributions in the Mojave-Intermountain transi- tion zone. Great Basin Naturalist Memoirs 2: 197-207. 552. MiLLER.R. F, FA Branson! S McQueen, AND C. 567. T. Snyder. 1982. Water relations in soils as related to plant communities in Ruby Valley, Nevada. Journal of Range Management 35: 462-468. 553. Miller, W. M. 1888. The investigation of the inter- dependence of plant life and climatic conditions of Nevada. Nevada Agricultural E.xperiment Station Bulletin 2. Reno, Nevada. 568. 554. Milton. N M, andT. L. Purdy 1983. Plant and soil relationships in two hydrothermally altered areas of the Great Basin. Great Basin Naturalist 43; 457-469. 569. 555. Mitchell, J E 1965. Relationship between soil mineralogy and plant distribution within two com- munities of the shadscale zone in Utah. Unpub- lished thesis, Utah State University, Logan. 556. Mitchell, J. E., and N. E. West 1965. Effects of physical soil properties on the distribution of 570. two shadscale zone species in northern Utah. Bul- letin of the Ecological Society of America 46: 46 (abstract). 557. Mitchell, J E , N E. West, and R W Miller. 1966. Soil physical properties in relation to plant 571. community patterns in the shadscale zone of northwestern Utah. Ecology 47; 627-630. 558. MOONEY, H. H 1973. Plant communities and vege- tation of the White Mountains. Pages 7-17 in R. 572. M. Lloyd, and R. S. Mitchell, A flora of the White Mountains, California and Nevada. University of California Press, Berkley. 559. MooNEY, H H., G. St. Andre, and R D Wright 1962. Alpine and subalpine vegetation of the 573. White Mountains of California. American Mid- land Naturalist 68: 257-273. 560. MooNEY, M.J. 1985. A preliminary classification of 574. high elevation sagebrush-grass plant vegetation in northern and central Nevada. Unpublished thesis. University of Nevada, Reno. 118 pp. 561. Moore, R R. 1964. A study of stand structure in the 575. uneven-aged stands in the Engehnann spruce- subalpine fir types on the Utah State University forest. Unpublished dissertation, Utah State Uni- versity, Logan. 562. MoRDEN. C , J D Brotherson, .and B N Smith. 1986. Ecological differences of C3 and C4 plant 576. species from central Utah in habitats and mineral composition. Great Basin Natiualist 46; 140-147. 563. Moretti, M. C. 1979. Vegetation and soil factors in relation to slope position: a study of plant commu- nities on foothill knolls in the Uinta Basin of Utah. 577. Unpublished thesis, Brigham Young University, Provo, Utah. 31 pp. 564. MoRE'iTi, MC, AND J. D Brotherson 1982. Vege- 578. tation and soil factors in relation to slope position of foothill knolls in the Uinta Basin of Utah. Great Basin Naturalist 42; 81-90. 565. MoziNCO, H 1987. Shrubs of the Great Basin: a natural history. University of Nevada Press, Las Vegas. 342 pp. 579. 566. MuE(;gler, W. F. 1976. Ecological role ol hre in western woodland and range ecosystems. Pages 1-9 in Proceedings: Use of Prescribed Burning in Western Woodlands and Range Ecosystems Sym- posium, 18-19 March 1976. USDA Forest Ser- vice, Utah Agricultural E.xperiment Station, Lo- gan, Utah. Mueggler, W. F. 1985. Vegetation associations. Pages 45-55 in Debyle and Winokur, eds. , Aspen: ecology and management in the western United States. USDA Forest Service General Technical Report RM-119. Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colo- rado. Mueggler, W F 1988. Aspen community types of the Intermountain Region. USDA Forest Service General Technical Report INT-250. Intermoun- tain Research Station, Ogden, Utah. 135 pp. Mueggler, W F.. and D. C Bartos. 1977. Grind- stone Flat and Big Flat exclosures; a 41 year record of changes in clearcut aspen communities. USDA Forest Service Research Paper INT-195. Rocky Mountain Forest and Range Experiment Station, Ogden, Utah. Mueggler, W. F., and R. B Campbell, Jr. 1986. Aspen community types of Utah. USDA Forest Service Research Paper INT-362. Intermountain Forest and Range Experiment Station, Ogden, Utah. MURDOCK, J R. 1951. Alpine plant succession near Mount Emmons, Uinta Mountains, Utah. Lhipub- lished thesis, Brigham Young Lhiiversitv, Provo, Utah. MuRDOCK. J. R , S L Welsh, AND B W.Wood, eds. 1975. Navajo-Kaiparowits environmental baseline studies 1971-1974. Unpublished report prepared for Brigham Young University, Center for Health and Environmental Studies, Provo, Utah. 864 pp. Nabi, a. a 1978. Variation in successional status of pin\ on-jiuiiper woodlands in the Great Basin. Un- published thesis, Utah State University, Logan. Nachlinger, J. L. 1985. The ecology of subalpine meadows in the Lake Tahoe region, California and Nevada. Unpublished thesis. University of Ne- vada, Reno. N.\Ti\'E Plants lNCORPOR.vrED. 1983. J. B. King Mine, soil and vegetation data. Volume 1, section 783.19 in Mining and Reclamation Plan for J. B. King Mine, Western States Minerals Corporation. Utah Division of Oil, Gas and Mining Number ACT/015/002. Salt Lake City, Utah. Neely, E. E , .AND M E Barkworth 1984. Vegeta- tion on soils derived from dolomite and quartzite in the Bear River Range, Utah: a comparative studv. Bulletin of the Torre\' Botanical Club 3: 179-192. Neese, E. 1980. Vegetation of the Henry Moun- tains. Pages 219-236 in Proceedings u'gA-1980 Henry Mountains Symposium. Neilson, R P 1986. On the interface between cur- rent ecological studies and the paleobotany of pinyon-juniper woodlands. Unpublished paper presented to Pinyon-Jimiper (conference, 13-16 Januar\ 1986, USDA F"orest Service, Reno, Ne- vada. Neilson. R. P.. and L H Wullstein. 1985. Com- parative drought phvsiology and biogeography of Qucrctis clii and Qticrcus turhinella. Ameri- can Midland Naturalist 114: 259-271. 1990] Nevada-Utah Vecjetation Bihi.i()(;i togeographical notes on the Rock\ Moimtain region N'lII. Distribution of the montane plants. Bulletin of the Torre\- Botani- cal Club 46: 295-327. Rydberg. P A 1920. Ph> togeographical notes on the Rock\' Mountain region IX. Wooded forma- tions of the montane zone of the southern Rockies. Bulletin of the Torrey Botanical Club 47: 441-454. Rydberg. P A 1921. PhNtogcographical notes on tlu' U()ck\ Mountain region X. (Grasslands and 1990] Nevada-Utah Vfx.etation Bihliochai'iiv 233 other open formations of the montane zone of the southern Roekies. Bulletin ()ftheTorre\' BotuTiieal Ch.1) 48; 315-326. 709. Salisbukv, F. B 1952. iManl-soil lelations on male- rial derived from IndrothermalK altered roek in the Marysvale retjion of lUah. Unpublished the- sis, University of Utah, Salt Lake Cit> . 710. SalISBI'HV, F B. 1964. Soil formation and vegetation on h\clrothermallv altered roek material in Utah. Ecology 45: 1-9. ' 711. Sampson, A. W. 1925. The foothill-montane-alpine flora and its environment. Pages 24-31 in 1. Tide- strom, ed.. Flora of Utah and Nevada. U.S. Na- tional Herbarium Contributions 25. Washington, D.C. 665 pp. 712. Sargent, C. S. 1879. The forests of central Nevada. American Journal of Science and Arts 17: 417-426. 713. Savage, D. E 1968. The relationship of sage grouse to upland meadows in Nevada. Unpublished the- sis. University of Nevada, Reno. 714. SCHIER. G A 1975. Deterioration of aspen clones in the middle Rocky Mountains. USDA Forest Ser- vice Research Paper INT-710. Intermountain Forest and Range E.vperiment Station, Ogden, Utah. 14 pp. 715. ScHiMPF, D. J., J. A. Henderson, and J. A Mac- Mahon 1980. Some aspects of succession in the spruce-fir forests of northern Utah. Great Basin Naturalist 40: 1-26. 716. Schramm, D. R. 1982. Floristics and vegetation of the Black Mountains, Death Valley National Mon- ument. Unpublished report prepared for Univer- sity of Nevada, Las Vegas. 717. SCHULTZ, B. 1987. Ecology of curlleaf mountain mahogany (Cercocarptis IcdifoUus) in western and central Nevada: population structure and dy- namics. Unpublished thesis. University of Ne- vada, Reno. 718. ScHULTY, L. M., E E Neelv, and J. S Tuhy 1987. Flora of the Orange Cliffs of Utah. Great Basin Naturalist 47: 287-298. 719. Science Applications Incorporated. 1980. Vege- tation survey at the Knight Mine, Sevier County, Utah. Volume 1, section 783.19, appendix 783-H in Mining and Reclamation Plan for Knight Mine, Utah International Incorporated. Utah Division of Oil, Gas and Mining Number ACT/041/005. Salt Lake City, Utah. 720. Shantz, h'. L 1916. Plant succession in Tooele Val- ley, Utah. Pages 233-236 /« F. E. Clements, Carnegie Institute Publication 242. Washington, D.C. 721. Shantz, H. L. 1925. Plant communities in Utah and Nevada. Pages 15-23 in 1. Tidestrom, Flora of Utah and Nevada. Contributions of the U.S. Na- tional Herbarium 25. Washington, D.C. 665 pp. 722. Shantz, H. L., and R. L. Piemeisel. 1940. Types of vegetation in Escalante Valley, Utah, as indicators of soil conditions. USDA Technical Bulletin 713. Washington, D.C. 46pp. 723. Sheeter, G. R. 1968. Secondary succession and range improvements after wildfire in northeastern Nevada. Unpublished thesis. University of Ne- vada, Reno. 724. Shields, L. M., and P. V. Wells. 1962. Effects of nuclear testing on desert vegetation. Science 135: 38-40. 725. 726. 727. 728. 729. 730. 731. 732. 733. 734. 735. 736. 737. 738. 739. 740. Shields, L M , and P V Wells 1963. Recovery of vegetation atomic target areas at the Nevada Test Site. Pages .307-3 10i»i iiadioccology. Proceedings of the First National Symposium on Radioeeology. Rinhold PublisiuTig (Corporation, New York. Shields, L M , P V Wells, and W II Rickard 1963. Vegetational recovery on atomic target areas in Nevada. Ecology 44: 697-705. Shreve. F. 1942. The desert vegetation of North America. Botanical Review 8: 195-246. Siii'PE.J B , andJ D Brotherson 1985. Differen- tial effects of cattle and sheep grazing on high mountain meadows in the Strawberry Valley of central Utah. Great Basin Naturalist 45: 142-149. Shupe, J B , J D BRcrPHERSON, and S R. Rush- forth. 1986. Patterns of vegetation surrounding springs in Goshen Bay, Utah Countv, Utah, U.S.A. Hydrobiologia 139: 97-107. Shute, D , AND N E West 1978. The application of ECOSYM vegetation classifications to rangelands near Price, LJtah. Appendi.x reports 14 and 16 in J. A. Henderson and L. S. Davis, eds., ECOSYM — an ecosystem classification and data storage system for natural resources management. Unpublished report prepared for Utah State Uni- versity, Logan. 53 pp. SiGLER, J. W., AND J S TuHY, EDS. 1982. Ecological surveys — Altex Oil Leases — southern Utah, Octo- ber 1981-September 1982. Unpublished report prepared for Altex Oil by W. F. Sigler & Associ- ates Inc., Logan, Utah. SlGLER, J. W., AND J S. TuHY. 1983. Resource sur- veys: Tar Sands Triangle, Wayne and Garfield Counties, Utah. Unpublished report prepared for Santa Fe Energy Company, Houston, Texas. 506 pp. SiLKER, A. R. 1972. Maintenance and survival of vegetation on the Sunrise Campground, Cache National Forest. Unpublished thesis, Utah State University, Logan. 41 pp. SIMMS, P, L. 1967. The effect of stand density on characteristics of wheatgrasses. Unpublished dis- sertation, Utah State Lhiiversity, Logan. Simpson,] H. 1859. Report of exploration across the Great Basin of the territory of Utah for a direct wagon route from Camp Floyd to Genoa in Carson Valley. U.S. Government Printing Office, Wash- ington, D.C. Singh, T. 1967. Quantitative analysis of perennial A^rip/c.Y -dominated vegetation of southeastern Utah. Unpublished dissertation, Utah State Uni- versity, Logan. 178 pp. Singh, T., and N. E. West. 1965. Ordination of shadscale zone vegetation of southeastern Utah. Bulletin of the Ecological Societv of America 46: 198. Singh, T, and N. E West 1971. Compari.son of some multivariate analyses of perennial Atriplcx vegetation in southeastern Utah. Vegetatio 23: 289-313. Skougard, M G 1976. Vegetational response to three environmental gradients in a salt playa near Goshen, Utah County. Unpublished thesis, Brig- ham Young University, Provo, Utah. 75 pp. Skougard, M. G., and J. D. Brotherson. 1979. Vegetational response to three environmental 234 P. S. BOURGERON ET AL. [Volume 50 gradients in the Salt Playa near Goshen, Utah County, Utah. Great Basin Naturahst 39; 44-58. 741. Skousen, J., J. N. Davis, .\nd J D Brotherson. 1986. Comparison of vegetation patterns resulting from bulldozing and two-way chaining on a Utah pinvon-juniper big game range. Great Basin Natu- ralist 46: .508-512'. 742. Smiley, F. J. 1915. Alpine and subalpine vegetation of the Lake Tahoe region. Botanical Gazette .59: 265-286. 743. Smith, C. B 1986. Grazing effects and site factors in relation to grazing use of salt desert shrub, desert e.xperimental range. Unpublished thesis, Utah State University, Logan. 744. Smith, G L 1973, A flora of the Tahoe Basin and neighboring areas, Wasman Journal of Biology 31: 1-231. 745. Soldier Creek Co.al Comp.\ny 1986. Vegetation. Volume 2, part 3, section 3.7, pages 3-161-3-192 in Mining and Reclamation Plan for Soldier Canyon Mine, Soldier Creek Coal Company. Utah Division of Oil, Gas and Mining Number ACT/007/018. Salt Lake City, Utah. 746. Southard, A. R. 1958. Some characteristics of fine soil profiles under quaking aspen in Cache Na- tional Forest. Unpublished thesis, Utah State University, Logan. 747. Spence.J R 1990. A preliminary review of various vegetation classification schemes for use on the Colorado Plateau, with suggestions for future con- sideration. Unpublished report prepared for Capi- tol Reef National Park, Torrey. Utah. 29 pp. 748. St John, C P 1970, Establishment report for Red Butte Canyon Research Natural Area within Wasatch National Forest, Salt Lake County, Utah. Unpublished report prepared for USDA Forest Service, Washington, D,C. 6 pp. 749. St. Clair, L. L , S R. Rushforth, and J D Broth- erson. 1986, The influence of microhabitat on diversity, distribution, and abundance of corti- colous lichens in Zion National Park, Utah, and Navajo National Monument, Arizona. Mycota.xon 26: 253-262. 750. Stager, D W 1977. Mule deer response to succes- sional changes in the pin\'on-juniper vegetation type after wildfire. Unpublished thesis. Univer- sity of Nevada, Reno. 751. Stansbury. H. 1851. E.xploration and survey of the valley of the Great Salt Lake of Utah. Senate Doc- ument Volume 2, Number 3. U.S. Government Printing Office, Washington, D.C. 752. Stanton, W. D. 1933. The ecology and floristics of the Henry Mountains. Proceedings of the Utah Academy of Science, Arts and Letters 10: 25. 753. Steele, P. L. 1958. The comparative mineral re- quirements of pioneer and climax range grasses. Unpublished thesis, Brigham Young University, Provo, Utah, 754. Steele, R , R D Pfister, R A Ryker. and J A KiTTAMS 1974. Preliminary habitat types of the Challis, Salmon and Sawtooth National Forests. Unpublished co<)perati\'e habitat-type classihca- tion report prepared for USDA Forest Service Region IV and Intennountain F"orest and Range Experiment Station, Ogden, Utah. 72 pp. 755. 756. 758. 759. 760. 761. 762. 763. 764. 765. 766. 767. 768. 769. Stein. B A , and S. F Warrick, eds 1979. Granite Mountains resource survey. University of Califor- nia Environmental Field Program Publication 1. Santa Cruz, California. 361 pp. Stevens, G. L 1963, Appraisal of the timber re- source potential of the Dixie National Forest, Un- published thesis, Utah State University, Logan, 75 pp. Ste\ens, R 1983. Description and discussion of field tour sites. Pages 47-49 in Proceedings of the First Utah Shrub Ecology Workshop, Utah State Universitv College of Natural Resources, Logan. 50 pp. Stewart, G. 19.35. Plant cover and forage conditions on spring-fall and winter ranges, largely on public domain lands of the Intermountain Region. Utah Juniper 6: 9-13. Stewart, G 1941. Historic records bearing on agri- cultural and grazing ecology in Utah. Journal of Forestry 39: 362-375. Stew.\rt, G. 1948. Utah's biological heritage and the need for its conservation. Proceedings of Utah Academy of Science, Arts and Letters 25: 5—22. Stewart, G , W. P. Cottam, and S. S. Hutchings, 1940, Influence of unrestricted grazing on north- ern salt desert plant associations in western Utah. Journal of Agricultural Research 60: 289-317. Stewart, G , and W. Keller. 1936. A correlation method for ecology as exemplified by studies of native desert vegetation. Ecology 17; 500-514. Stewart, G , and J Widtsoe 1943. Contribution of forest land resources to the settlement and devel- opment of the Mormon-occupied West. Journal of Forestry 41: 633-640. Stoddart. L a 1941. The palouse grassland associ- ation in northern Utah. Ecology 22: 158-163. Stoddart, L A 1945. Range lands of Utah County, Utah, and their utilization. Utah Agricultural Ex- periment Station Bulletin 317. Utah State Agricul- tural College, Logan. 32 pp. Stoddart. L A , A H Holmgren, andC W. Cook. 1949. Important poisonous plants of Utah. Utah Agricultural Experiment Station Special Report 2. Utah State Agricultural College, Logan, 21 pp. Stoddart, L A , P B. Lister, G Stewart. T. D. Phinney, and L W. Darson 19.38, Range condi- tions in the Uinta Basin, Utah. Utah .Agricultural Experiment Station Bulletin 283. Utah State Agri- cultural College, Logan. 34 pp. Stokes. W L 1977. Subdi\isions of the major ph\s- iographic proxinces in I'tah. Utah Geology 4; 1-17. Stutz, H. C. 1951. An ecological study of a sphag- num lake in the subalpine forest of the Uinta Mountains of Utah. Unpublished thesis, Brigham Young University, Provo, Utah, 53 pp. Stutz, H C. 1979, The meaning of "rare" and "endangered" in the evolution of western shrubs. Great Basin Naturalist Memoirs 3: 119-128. SuMMERFiELD. H. B., Jr 1969, Pedological factors related to the occurrence of big and low sagebrush in northern Washoe County, Nevada, Unpub- lished thesis. University of Nevada, Reno. Summit Coal Company. 1987. Vegetation re- sources. Volume 1, chapter 9, sections 9.1-9.3, 1990] Nevada-Utah Vkc;etati()n Bibliockafiiy 235 pages 1-6, plus apix'iulix 9-1 in Miiiingaiul lictla- iiiatioii I'laii lor Hover Mine, Siiniinit CJoal (^om- paii\'. L'tah Di\isioii of Oil, (Jas aiul Miiiiiiu; Niim- 789. ber'ACT/()43/()()S. Salt Lake City, Utali. 773. T.\NNKK, V M 194()a. A eiiapter on the natural iiis- torv ofthe Great Basin, I-SOO to 1855. (ireat Basin 790. Naturalist 1:33-61. 774. T.\nm:r, V M. 1940b. A hiotie study of the Kai- parowitz region of Utah. Great Basin Natinalist 1: 97-126. 791. 775. T.^NNER, V. M, andC L H.ww ahd 1934. A I)i<)logi- cal study ofthe La Sal Moiuitains, Utah. Report 1. Utah Academy of Sciences, Arts and Letters 11: 209-235. 776. T.\u.S(:h, R J 1973. Plant succession and mule deer 792. utilization on pinyon-juniperehainings in Nevada. Unpublished thesis, Lhii\ersity of Nevada, Reno. 777. T.\l'SCH, R J 1980. Allometrie analysis of plant growth in woodland commimities. Unpublished 793. dissertation, LUah State University, Logan. 778. Tausch, R. J. and P T. Tuellkr 1977. Plant succes- sion following chaining of pinvon-jimiper wood- lands in eastern Nevada. Joiunal of Range Man- agement 30: 44-49. 779. Tausch. R J., and P T Tueleer 1989. Evaluation 794. of pinyon sapwood to phytomass relationships over difTerent site conditions. Journal of Range Management 42; 209-212. 780. TauschR J.N E VVest.andA.A. N.\BI 1981.Tree 795. age and dominance patterns in Great Basin pinyon-juniper woodlands. Journal of Range Man- agement 34: 259-264. 781. Tausch, R. J., and N. E West 1985. Patterns of pinyon and juniper establishment on a southwest- ern Utah site. Abstracts ofthe annual meetings of 796. the Society for Range Management, #78. 782. Tausch, R. J., and N E West 1986. Morphological variation/precipitation relationships of Great Basin single-needled pinyon. Pages 86-91 in 797. Pinyon-juniper Conference Proceedings, LISDA Forest Service General Technical Report INT- 215. Intermountain Forest and Range Experiment Station, Ogden, Utah. 91 pp. 783. Tausch. R J , and N. E. West. 1988. Differential 798. establishment of pinyon and juniper following fire. American Midland Naturalist 119: 174-184. 784. Technical Committee. 1971. Vegetation of the Lake Tahoe region. A guide for planning. Unpub- lished report prepared for the Tahoe Regional Planning Agency and USDA Forest Service Plan- 799. ningTeam, South Lake Tahoe, California. 43 pp. 785. Tepedino. V. J 1979. The importance of bees and other insect pollinators in maintaining floral spe- cies composition. Great Basin Naturalist Memoirs 800. 3: 139-150. 786. Tew, R K.J T Lorr, andT. M. Bliss. 1988. Ecosys- tem stratification on the Fishlake National Forest. Unpublished report prepared for USDA Forest Service Fishlake National Forest, Richfield, Utah. 787. Th.^tcher, a P.. and V L. Hart. 1974. Spy Mesa 801. yields better understanding of pinyon-juniper in range ecosystems. Journal of Range Management 27: 354-357. 788. Thompson. R. 1979. Life zones, plant associations and habitat types, Ferron-Price planning units. 802. Unpublished report prepared for USDA Forest Service, Manti-LaSal National Forest, Moab, Utah. TllORNE. K F 1982. The desert and other transmon- tane plant eonnnunities of southern (California. Aliso 10: 219-257. Thorne. R F. B. a Pric(;e, and J IIenrickson. 1981. Flora of the higher ranges and the Kelso Dunes of the eastern Mojave Desert in (California. Aliso 10: 71-186. TiSDALE, E, W., AND M. H1RONAK.A. 1981. The sage- brush-grass region: a review ofthe ecological liter- ature. University of Idaho Forest, Wildlife, and Range E.xperiment Station Bulletin .33. Moscow, Idaho. TowEK. J D 1970. Vegetation stagnation in three- phase big game e.xclosures and its effects in deter- mining forage utilization. Unpublished thesis, University of Nevada, Reno. TovvNSEND, T. W. 1966. Plant characteristics re- lating to the desirability of rehabilitating the Arctostaphylos patiila - Ceanothtts velutinus - Ceanotlttts prostratus association on the east slope ofthe Sierra Nevada. Unpublished thesis. Univer- sity of Nevada, Reno. Trueblood, R. W. 1954. The effect of grass reseed- ing in sagebrush lands on sage grouse populations. Unpublished thesis, Utah State Agricultural Col- lege, Logan. 77 pp. TUELLER, P. T. 1973. Secondary succession, discli- max, and range condition standards in desert shrub vegetation. Pages 57-65 in D. N. Hyders, ed.. Arid Shrublands, Proceedings ofthe Third Workshop of the U.S. /Australia Rangelands Panel, 26 March-5 April 1973, Tucson, Arizona. TuELLER. P T. 1975. The natural vegetation of Ne- vada. Low deserts, high mountains provide con- trasts from shadscale shrub to ancient bristle- cones. Mentzelia 1: 3-6, 23-28. TuELLER. P T 1985. Sagebrush-dominated vegeta- tion ofthe Great Basin. Pages 24-30 in Proceed- ings, .38th annual meeting ofthe Society for Range Management, 11-15 February 1985, Salt Lake City, Utah. TUELLER. P. T.. C. D. Beeson, R. J. Tausch, N. E. West, and K H. Rea. 1979. Pinyon-juniper wood- lands of the Great Basin: distribution, flora and vegetal cover. USDA Forest Service Research Pa- per INT-229. Intermountain Forest and Range Experiment Station, Ogden, Utah. 22 pp. TUELLER, P. T, AND W H. Blackburn 1974. Condi- tion and trend of the big sagebrush/needle- andthread habitat type in Nevada. Journal of Range Management 27: 36-40. Tueller, p. T., AND J. E. Clark 1975. Autecology of pinyon-juniper species of the Great Basin and Colorado Plateau. Pages 27-39 in G. F. Gifford, and F. E. Busby, eds.. The Pinyon-juniper Ecosystem: A Symposium. Utah State University, Logan. Tueller. P. T.. and R. E Eckert, Jr 1987. Big sagebrush (Artemisia tridentata vaseijana) and longleaf snowberry (Sijmphoricarpos oreophilus) plant associations in northeastern Nevada. Great Basin Naturalist 47: 117-131. Tueller. P. T., D. H. Heinze, and R. E. Eckert. 1966. A tentative list of existing Nevada plant 236 P. S. BOURGERON ET AL. [Volume 50 804. 805. 806. 807. communities (a third approximation). Unpub- lished report prepared for University of Nevada, Department of Range, WildUfe and Forestry, Reno. 14 pp. 803. TuELLER. P T.. AND J D TowER 1979. Vegetation stagnation in three-phase big game exclosures. Journal of Range Management 32: 250-263. TuHY, J. S.. AND J. A M./^cMahon. 1988. Vegetation and relict communities of Glen Canyon National Recreation Area. Unpublished final report pre- pared by Utah State University for USDl National Park Service Rocky Mountain Region, Lakewood, Colorado. 299 pp.' USDA 1871. The Great Salt Lake. Pages 559-569 in 1870 Report of the Commission on Agriculture. Unpublished report prepared for USDA, Wash- ington, D.C. USDA Soil Conservation Service. 1981. Land re- source regions and major land resource areas of the United States. USDA Soil Conservation Service Agriculture Handbook 296. Washington, DC. 156 pp. USDA Soil Conservation Service. 1988a. Major land resource area 28A — Great Salt Lake area — Nevada site descriptions. Unpublished internal technical guide, section HE. Reno, Nevada. USDA Soil Conservation Service. 1988b. Major land resource area 28B. Central Nevada Basin and Range. Nevada site descriptions. Unpublished internal technical guide, section IIB, Reno, Nevada. USDI Bureau of Land Manacement. 1972. His- tory of range fires in the Uinta Basin, Utah. Un- published report. USDl National Park Service. 1974. Wilderness recommendation; Lake Mead National Recreation Area: Arizona-Nevada. Unpublished report pre- pared for USDI National Park Service, Denver Service Center, Denver, Colorado. 80 pp. US. Fuel Company. 1980. Vegetation. Volume 3, chapter 9, sections 9.1-9.3, pages IX-l-lX-76 in Mining and Reclamation Plan for Hiawatha Mines Complex, U.S. Fuel Companv. Utah Division of Oil, Gas and Mining Number ACT/007/011. Salt Lake City, Utah. Utah Division of Wildlife Resources and Ecol- ogy Consultants, Inc. 1977. Annotated bibliog- raphv of natural resource information: southern Utah. USDI Fish and Wildlife Service, Biological Services Program FWS/OBS-77/34. Fort Collins, Colorado. 265 pp. Utah Environmental and Agricultural Consul- tants. 1973. Environmental setting, impact, miti- gation, and recommendations for a proposed oil products pipeline between Lisbon Valley, L'tah, and Parachute Creek, Colorado. Unpublished re- port prepared for Colony Development Opera- tions, Denver, Colorado. Utah Intern.ationalIncorpoh.jiTED. 1987. Vegeta- tion. Volume 14, section 6, pages 6-1-6-53 in Mining and Reclamation Plan for Alton Mine, Utah International Incorporated. Utah Division oi Oil, Gas and Mining Number ACT/025/003. Salt Lake City, Utah. Utah State University. 1978. Rehabilitation po- tential for the Henrv Mountain coal lield EMRIA 809. 810. 811. 812. 813. 814. 815. Report 15. Pages 102-108 and 199-203 in Vol. I, Energy mineral rehabilitation inventory and anal- ysis in the mountain coal fields of Garfield County, Utah. Unpublished report prepared for USDI Bu- reau of Land Management by Utah State Univer- sity, Logan. 203 pp. 816. Vale. T R 1973. The sagebrush landscape of the Intermountain West. Unpublished dissertation. University of California, Berkeley. 508 pp. 817. Vale, T R. 1975. Presettlement vegetation in the sagebrush-grass area of the Intermountain West. Journal of Range Management 28: 32-36. 818. Valentine, J. F 1961. Important Utah range grasses. Utah State Universitv Extension Circular 281. Logan, Utah. 819. Valley Camp of Utah Incorporated 1984. Vege- tation. Volume 6, section 783.19, pages 15D- 15N32 in Mining and Reclamation Plan for Belina Mines Complex, Valley Camp of Utah Incorpo- rated. Utah Divison of Oil, Gas and Mining Num- ber ACT/007/001. Salt Lake City, Utah. 820. Valley Engineering Incorporated. 1988. Vege- tation and terrestrial wildlife report. Volume 3, chapter 9, item 9-1, pages 1-38 in Mining and Reclamation Plan for Crandall Canyon Mine, Genwal Coal Companv. Utah Division of Oil, Gas and Mining Number ACT/015/032. Salt Lake City, Utah. 821. Van Balen, J A 1973. Plant cover types of the upper Bear River drainage, Utah. Unpublished thesis, Universitv of Utah, Salt Lake Citv. 822. Van Pelt, N. S. 1978. Woodland parks in southeast- ern Utah. Unpublished thesis. University of Utah, Salt Lake City. 823. Van Pelt. N. S 1988. Baseline description of a na- tive sagebrush/grass reference area at Wildcat Loadout. Volume 1, appendix I, pages 1-5 in Mining and Reclamation Plan for Wildcat Loadout Facility, Andalex Resources. Utah Division of Oil, Gas and Mining Number ACT/007/033. Salt Lake City, Utah. 824. Van Pelt, N S., R, Stevens, and N. E. West. 1989. Comparative survival and growth of immature Utah jiuiiper after chaining in central Utah. Ab- stracts of the annual meeting of the Society for Range Management #287. 825. Vest. E. D. 1962a. Biotic comnumities in the Great Salt Lake Desert. University of Utah Institute of Environmental Biological Research, Ecology and Epizoology Series 73. Salt Lake City, Utah. 122 pp. 826. Vest, E, D 1962b. The plant conununities and asso- ciated fauna of Dugvva\- \'alle\- in western Utah. Unpublished thesis. University of Utah, Salt Lake City. 827. Wac:ner, W W 1973. Cesium- 137 accumulation and distribution in plants and soils of an alpine communitv' of Utah. Unpublished dissertation. University of Utah, Salt Lake City. 267 pp. 828. Wakefield. H. 1933. A study of the plant ecology of Salt Lake and Utah \alle\s before the Mormon immigration. Unpublished thesis, Brigham Yoiuig University, Provo, Utah. 54 pp. 829. Wakefield. H. 1936. A study of the native vegeta- tion of Salt Lake and Utah valleys as determined by historical evidence. Proceedings of Utah Academv of Science, Arts and Letters 13: 11-18. 1990] Nevada-Utah Vkc;etatk)n Bibliogiufhy 237 830. Walkek. C R . AND J D Rrothf.hson 1982. Habi- tat relationships ot basin wiklryo in tlir hi,tj;h nionn- tain valleys olct-ntral I'tali. Joninal olRan^i' Man- agement 35; 628-633. 831. \Vall.\ce, a , AND E. M Romnky 1980. The role of pioneer species in revegctation of clistnrl)ed desert areas. Great Rasin Natnralist Memoirs 4; 31-33. 832. Wall.-vce. a , E M Romney, anu R R Hunteh 1980. The challenge of a desert: revegctation of disturbed desert lands. Great Rasin Naturalist Memoirs 4: 216-225. 833. VVallis. C. A 1982. An overview of the mixed grass- lands of North America. Pages 195-208 in A. C. Nicholson, A. Mclean, and T. E. Raker, eds., Proceedings of the Grassland Ecology and Classifi- cation Symposium, 2-4 June 1982, Kamloops, Rritish Columbia. Province of Rritish Columbia Ministry of Forests, Victoria. 834. Ward, K. V 1977. Two-year vegetation response and successional trends for spring burns in the pinyon-juniper woodland. Unpublished thesis, Universitv of Nevada, Reno. 835. Warner,] H, and K.TH.ARPER 1972. Understory characteristics related to site quality for aspen in Utah. Rrighani Young University Science Rul- letin, Riological Series 16: 1-20. 836. Watson. S. 1871. Rotany. In: C. King, ed.. Report of the geological exploration of the Fortieth Paral- lel, Volume 5. U.S. Army Corps of Engineers Professional Paper 18. 525 pp. 837. Webb, G. M , and J. D. Rrotherson. 1988. Eleva- tional changes in woody vegetation along three streams in Washington County, Utah. Great Rasin Naturalist 48: 512-529. 838. Webb.R H, J. WSteiger,.\ndR M Turner. 1987. Dynamics of Mojave Desert shrub assemblages in the Panamint Mountains, California. Ecology 68: 478-490. 839. Wefirlie, L J 1973. Vegetation analysis of a Sierra Nevada montane coniferous forest. Unpublished thesis. University of Nevada, Reno. 840. Weir, G H 1976. Palynology, flora and vegetation of Hovenweep National Monimient: implications for aboriginal plant use on Cajon Mesa, Colorado and Utah. Rotany 3749-R (abstract). 841. Wells, P V 1960. Physiognomic intergradation of vegetation on the Pine Valley Mountains in south- western Utah. Ecology 41: 553-556. 842. Wells. P. V. 1961. Succession in desert vegetation on streets of a Nevada ghost town. Science 134: 670-671. 843. Wells, P. V 1983. Paleobiogeography of montane islands in the Great Rasin since the last glacioplu- vial. Ecological Monographs 53; 341-382. 844. Wells, P. V, and L M. Shields. 1964. Distribution oi Larrea divaricata in relation to a temperature inversion at Yucca Flat, southern Nevada. South- western Naturalist 9; 51-55. 845. Welsh, S. L. 1957. An ecological survey of the vege- tation of the Dinosaur National Monument, Utah. Unpublished thesis, Rrighani Young University, Provo. 86 pp. 846. Welsh, S L. 1970. Vegetation of the Colorado Plateau, its phytogeographic distinction from the Great Rasin. Unpublished manuscript prepared for Rrighani Young University, Provo, Utah. 29 pp. 847. Welsh. S L 1979. Endangered and threatened plants of Utah: a case study. Great Rasin Naturalist Memoirs 3: 64-80. 848. Welsh, S L., N D Atwood, L C, Hig(;ins, and S. Goodrich 1987. A Utah flora. Great Rasin Natu- ralist Memoirs 9. 894 pp. 849. Welsh, S L., and E. M. (^ihustenskn. 1957. An ecological study of the vegetation of the Dinosaur National Monument. Proceedings of the Utah Academy of Science, Arts and Letters .34: 55-56. 850. Welsh, S. L. and R. J. Kass, 1985. Vegetation moni- toring of reclaimed slopes and corresponding ref- erence areas at the Skyline Mine, 1985. Appendix volume A2, pages 1-29 in Mining and Reclama- tion Plan for Skyline Mine, Utah Fuel Company. Utah Division of Oil, Gas and Mining Number ACT/007/005. Salt Lake City, Utah. 851. Welsh. S L., and J. R Murdock. 1980. Report of vegetation, plant community analysis, threatened and endangered plant species, soils, and reclama- tion plans for Coastal States Energy Company, McKinnon Properties, Skyline Project, Carbon- Emery Counties, Utah. Appendix vohune A2, pages 1-135 in Mining and Reclamation Plan for Skyline Mine, Utah Fuel Company. Utah Divi- sion of Oil, Gas and Mining Nunii)er ACT/007/ 005. Salt Lake City, Utah. 852. Welsh, S. L:, and J. R. Murdock. 1982. Report of revegetation success and vegetation monitoring of reclaimed and reference areas at Skyline Mine during the 1982 field season. Appendix volume A2, pages 1-27, plus appendices I, II, III in Min- ing and Reclamation Plan for Skyline Mine, Utah Fuel Company. Utah Division of Oil, Gas and Mining Number ACT/007/005. Salt Lake City, Utah. 853. Welsh. S. L., J. R. Murdock, L. Juarros. and E. Neese. 1981. Report of studies of vegetation and soils for Coastal States Energy Company Con- vulsion Canyon Mine of Southern Utah Fuel Com- pany (SUFCO), Sevier County, Utah. Volume 5, pages 1-11 and 39-78 in Mining and Reclamation Plan for Convulsion Canyon Mine, Southern Utah Fuel Company. Utah Division of Oil, Gas and Mining Number ACT/041/002. Salt Lake City, Utah. 854. Welsh, S L., andC A. Toft 1981. Riotic commu- nities of hanging gardens in southeastern Utah. National Geographic Societv Research Reports 13; 663-681. 855. West, N. E. 1966a. Ecology and management of salt desert shrub ranges: a bibliography. Wyoming Range Management 228; 147-167. Revised and reprinted 1968, Utah Agricultural Experiment Station Mimeograph Series 505. 30 pp. 856. West, N. E 1966b. Research on the ecology of the Utah salt desert areas. Pages 160-169 in Proceed- ings of the Salt Desert Shrub Symposium. USDI, Rureau of Land Management, Cedar City, Utah. 857. West, N E 1971. Pinyon-juniper ranges. Unpub- lished report prepared for Utah State University, Department of Natural Resources, Logan. 28 pp. 858. West, N. E. 1974. The shrublands of Utah. Utah Science 35; 4-6. 859. West, N E 1977. An ecologist's thoughts toward determining the recreational carrying capacity of 238 P. S. BOURGERON ETAL. [Volume 50 860. 861. 862. 863. 864. 865. 866. 867. 868. 869. 870. 871 white water sections of the Upper Colorado River drainage. Pages 125-145 in L. E. Royer, W. H. Becker, and R. Schreyer, eds., Managing Colo- rado River whitewater: the carrying capacity strat- egy. Institute for the Study of Outdoor Recreation and Tourism, Utah State University, Logan. West, N E 1978a. Basic synecological relationships of sagebrush-dominated lands in the Great Basin and Colorado Plateau. Pages 33-41 in G. F. Gif- ford and F. E. Busby, eds., The Sagebrush Eco- system, a Symposium. Utah State University, Col- lege of Natural Resources, Logan. 251 pp. West, N E. 1978b. The changing woodlands of the Great Basin. Utah State University, College of Natural Resources. Edge: Natural Resources/Peo- ple 1: 13-16. West, N. E. 1979. Survival patterns of major peren- nials in salt desert shrub communities of south- western Utah. Journal of Range Management 32: 442-445. West, N. E. 1982. Approaches to synecological char- acterization of wildlands in the Intermountain West. Pages 633-641 in T. C. Brann, ed. , In-place resource inventories — principles and practices. A national workshop. University of Maine, Orono. Society of American Foresters, 9-14 August 1981, McClean, Virginia. West, N E. 1983a. Great Basin-Colorado Plateau sagebrush semi-desert. Pages 331-349 in N. E. West, ed.. Temperate deserts and semi-deserts. Ecosystems of the world, Vol. 5. Elsevier Publish- ing Company, Amsterdam. West, N. E 1983b. Western Intermountain sage- brush steppe. Pages 351-374 in N. E. West, ed., Temperate deserts and semi-deserts. Ecosystems of the world. Vol. 5. Elsevier Publishing Com- pany, Amsterdam. West, N. E 1983c. Intermountain salt desert shrub- lands. Pages 375-397 in N. E. West, ed.. Temper- ate deserts and semi-deserts. Ecosystems of the world. Vol. 5. Elsevier Publishing Company, Amsterdam. West, N. E. 1983d. Colorado Plateau-Mohavian blackbrush semi-desert. Pages 399-412 in N. E. West, ed.. Temperate deserts and semi-deserts. Ecosystemsof the world. Vol. 5. Elsevier Publish- ing Company, Amsterdam. West, N. E. 1983e. Southeastern Utah galleta- three awn shrub steppe. Pages 413-421 in N. E. West, ed.. Temperate deserts and semi-deserts. Ecosystems of the world. Vol. 5. Elsevier Publish- ing Company, Amsterdam. West, N E 1984. Successional patterns and pro- ductivity potentials of pinyon-juniper woodlands. Pages 1301-1332 in Natural Resources Council/ National Academy of Sciences, Developing strate- gies for rangeland management. Westview Press, Boulder, Colorado. West, N. E. 1987. Human impacts on vegetation of the semi-arid temperate intermountain region of western North America. Page 398 in Abstract oi the XIV International Botanical Congress, Berlin (West), Germany. West, N. E. 1988. Intermountain deserts, shru!) steppes and woodlands. Pages 209-230 in M. G. Barbour and W. D. Billings, eds., North Ameri- can terrestrial vegetation. Cambridge University Press, New York. 872. West, N. E. 1989. Vegetation types of Utah. Pages 18-51 in K. Johnson, ed., Rangelands of Utah. Utah State L^niversity Press, Logan. 873. West, N. E. and M. M. Baasher. 1968. Determina- tion of adequate plot size by use of mean distance between salt desert shrubs. Southwestern Natu- ralist 13: 61-74. 874. West, N. E., and J. E. Bowns 1988. Shrubland, grassland, shrubsteppe and woodland vegetation of the northern Mohave Desert, southwestern Utah and northwestern Arizona. Pages 73-81 iti J. L. Vankat, ed.. Vegetation of the southwest- ern United States. A handbook for the 1988 excur- sion of the International Association for Vegetation Science. Special Publication 1. North American Section International Association for Vegetation Science. 875. West, N. E , D. Cain, and G Gifford 1973. Biol- ogy, ecology and renewable resource manage- ment of the pigmy conifer woodlands of western North America; a bibliography. LItah Agricultural E.xperiment Station Research Report 12. Logan, Utah. 36 pp. 876. West, N E , and M. M. Caldwell 1983. Snow as a factor in salt desert shrub vegetation patterns in Curlew Vallev, Utah. American Midland Natural- ist 109: 376-379. 877. West, N. E., G. B. Coltharp, J P. Worknl\n, R. W. Wein, V. B. Hancock;, K Simonson, and J. Keith. 1973. Vegetational micro-climate, edaphic, hy- drologic and economic effects of contour furrows and gully plugs on frail lands near Cisco, Utah. LUah Agricultural Experiment Station Technical Bulletin (manuscript). Logan, Utah. 878. West, N. E., and J. Gasto. 1978. Phenology of the aerial portions of shadscale and winterfat in Curlew Valley, Utah. Journal of Range Manage- ment 31: 43-45. 879. West, N E.. and K I. Ibrahim. 1968. Soil-vegeta- tion relationships in the shadscale zone of south- eastern Utah. Ecology 49: 445-456. 880. West. N. E., and W L. Loope, 1980. Occurrence of wildfires in Zion National Park in relation to ecosystem and record-taking \ariables. Pages 21-28 in Proceedings of Second Conference on Scientific Research in the National Parks. Vol. X, Fire ecology. USDA, National Park Service Trans- actions and Proceedings Series 6. Washington, D.C. West, N E., and M. H. Madany 1981. Fire history of the Horse Pasture Plateau, Zion National Park. University of Utah-National Park Service Cooper- ative Studies Unit F'inal Report. 225 pp. West, N E , R T Moore, K A. Valentine. L. W. Law, p. R. Ogden. F. C Pinkney, P. T Tueller, AND A. A. Beetle. 1972. Galleta: taxonomy, ecol- ogy, and management of Hilaria jamesii on west- ern rangelands. Utah Agricultural Experiment Station Bulletin 487. Logan, Utah. S83. West, N. E., F. D. Provenza, P S. Johnson, and M. K. Owens. 1984. Vegetation change ;ifter 13 years of livestock grazing exclusion on sagebrush scmidesert in west central Utah. Journal of Range Management 37: 262-264. 881, 882. 1990] Nenada-Utah Vegetation Bibliochaimiv 239 884. West, N. E . K H Rka, and R. J Taiscm 1975. Basic syiH'fological rclatioii.ship.s in jiiniptM- pinyon woocUaiuLs. Pages 41-53 in C. F. (iifford and F. E. Bushy, eels.. Proceedings of the Pinxon- Juniper Ecosystem Syniposiiini, L'tah Atiiienl- tural Experiment Station, Logan. 885. West, N. E., R. J Tausch, and A A Nabi. 1979. Patterns and rates of pinyon-juniper invasion and degree of suppression of imderstory vegetation in the Great Basin. USDA Forest Service Inter- mountain Region Range Improvement Notes, Og- den, Utah. 14 pp. 886. West, N E , R J Tausch, K H Rea, and A R Southard 1978a. Soils associated with pinyon- juniper woodlands of the Great Basin. Pages 68-88 in C. T. Youngberg, ed.. Forest soils and land use. Proceedings of the Fifth North American Forest and Soils Conference, Colorado State Uni- versity, Fort Collins, Colorado. 887. West, N. E.. R. J. Tausch, K. H. Rea, and P T Tueller. 1978b. Taxonomic determination, dis- tribution, and ecological indicator values of sage- brush within the pinyon-juniper woodlands of the Great Basin. Journal of Range Management 31: 87-92. 888. West, N. E , R J Tausch, K. H Rea, and P. T. Tueller. 1979. Phytogeographical variation with- in juniper-pinyon woodlands of the Great Basin. In: K. T. Harper and J. L. Reveal, coords.. Inter- mountain biogeography: a symposium. Great Basin Naturalist Memoirs 2: 119-136. 889. West, N. E., and P T. Tueller. 1972. Special approaches to studies of competition and succes- sion in shrub communities. Pages 165-171 in C. M. McKell, J. P. Blaisdell, and J. R. Goodin, eds., Wildland shrubs: their biology and utiliza- tion. USDA Forest Service General Technical Re- port INT-1. Intermountain Forest and Range Ex- periment Station, Ogden, Utah. 890. West, N. E , and N. S. Van Pelt. 1987. Successional patterns in pinyon-juniper woodlands. Pages 43-52 in R. L. Everett, compiler. Proceedings — Pinyon-juniper Conference. USDA Forest Ser- vice General Technical Report lNT-215. Inter- mountain Research Station, Ogden, Utah. 891. Wheeler N. C, and P. Curies. 1982. Biogeogra- phv of lodgepole pine. Canadian Journal of Botanv 60;' 1805-1814. 892. Whipple, A. W 1854. Report of exploration for a railway route near the 35th Parallel of latitude from the Mississippi River to the Pacific Ocean. 37th Congress, 2nd Session, HoiLse Document 129, Volume 4. Washington, D.C. 154 pp. 893. White, J A. 1971. Climate, plant, soil, and water relationships to a wildlife plan. Unpublished the- sis, University of Nevada, Reno. 894. White, W N. 1932. A method of estimating ground- water supplies based on discharge by plants and evaporation from soil — results of investigations in Escalante Valley, Utah. USDl Geological Survey Water Supply Paper 659(A). Denver, Colorado. 105 pp. 895. Wieslander, A E 1932. Vegetation type maps of California and western Nevada. University of California Press, Berkeley. 896. WiLco.x, R B 1977. Canopy inilucncc as a factor in determining imderstory comnnniity composition. Unpui)lishcd thesis, Brigham Young University, Provo, Utah. 42 pp. 897. Wn.co.x, R. B., J. D Bhothkrson. and W. E. EvENSON. 1981. Canopy influence on understory community composition. Northwest Science 55: 194-201. ' 898. Williams, G. 1969. Analysis of hydrologic, edaphic, and vegetative factors affecting infiltration and erosion on certain treated and untreated pinyon- juniper sites. Unpul)lished dissertation, Utah State University, Logan. 899. Williams, J. E., D. Bowman, J. Brooks, A. EcHELLE, R. Edwards, D. A Hendrickson, and J. J. Landye. 1985. Endangered aquatic ecosys- tems in North American deserts with a list of vanishing fishes of the region. Journal of Arizona- Nevada Academy of Science 20: 1-62. 900. Wilson, R. O. 1985. Aerial and near surface spectral characteristics of eight Nevada rangeland commu- nities. Unpublished thesis. University of Nevada, Reno. 901. Windell, J T., B. E Willard, D. J. Cooper, S. Q. Foster. C F Knud-Hansen, L P Rink, and G N KlLADls. 1986. An ecological characterization of Rocky Mountain montane and subalpine wet- lands. USDl Fish and Wildlife Service Biological Report 86(11). 298 pp. 902. Winn, D S. 1976. Terrestrial vertebrate fauna and selected coniferous forest habitat types on the north slope of the Uinta Mountains. Unpublished report prepared for USDA Forest Service, Wasatch National Forest, Ogden, Utah. 145 pp. 903. Wood, B. W. 1966. An ecological life history of budsage in western Utah. Unpublished thesis, Brigham Young University, Provo, Utah. 85 pp. 904. Wood, B W, and J D Brotherson. 1984. Ecologi- cal adaptation and grazing response of budsage {Artemisia spinescens) in southwestern Utah. Pages 75-92 /;i Proceedings — Symposium on the Biology oi Artcnnisia and Chnjsothamnus. LISDA Forest Service General Technical Report INT- 200. Intermountain Research Station, Ogden, Utah. 905. Wood, B W , S L Welsh, and J R. Murdock 1975. The vegetation of the Kaiparowits region: an overview. Pages 249-302 in J. R. Murdock, S. L. Welsh, and B. W. Wood, eds., Navajo- Kaiparowits environmental baseline studies 1971-1974. LInpublished report prepared for Brigham Young University, Center for Health and Environmental Studies, Provo, Utah. 864 pp. 906. Woodbury, A. M. 1933. Biotic relationship of Zion Canyon, Utah, with special reference to succes- sion. Ecological Monographs 3: 148-245. 907. Woodbury, A. M. 1938. The lower Sonoran in southwestern Utah. Science 87: 484-485. 908. Woodbury, A M 1947. Distribution of pigmy conifers in Utah and northeastern Arizona. Ecol- ogy 28: 113-126. 909. Woodbury, A. M. 1956. Ecological checklists. Great Salt Lake Desert Series. Lhipublished re- port prepared for University of Utah, Salt Lake Citv. 240 P. S. BOURGERON ET AL. [Volume 50 910. Woodbury, A. M., S. D. Durrant, and S. Flowers. 1959. Survey of vegetation in the Glen Canyon Reservoir Basin. University of Utah Anthropologi- cal Papers 36. Salt Lake City, Utah. 53 pp. 911. Woodbury, A. M., S. D. Durrant, S Flowers. 1960. A survey of the vegetation in the Flaming Gorge Reservoir Basin. University of Utah An- thropological Papers 45. Salt Lake City, Utah. 121 pp. 912. Woodbury, W. V. 1966. The history and present status of the biota of Anaho Island, Pyramid Lake, Nevada. Unpublished thesis, Lhiiversity of Ne- vada, Reno. 913. Workman, G. W, N E West, B H.wvs, and C. Hurst. 1983. A(juatic and riparian ecological relationships of the North Fork of the Virgin River and its Deep Creek-Crystal Creek Tributary, Zion National Park, Utah. Unpublished report prepared for Department of Fisheries and Wildlife, Utah State University, Logan. 272 pp. 914. 915. 916. Wright. H A 1964. An evaluation of several factors to determine why Sitanion hijstix is more resistant to burning than Stipa comata. L^npuhlished dis- sertation, Utah State University, Logan. Wright. H. A. 1980. The role and use of fire in the semi-desert grass-shrub type. USDA Forest Ser- vice General Technical Report INT-85. Inter- mountain Forest and Range E.xperiment Station, Ogden, Utah. 23 pp. Wright. H A., L. F Neuenschwander, and C. M. Britton 1981. The role and use of fire in sage- brush-grass and pinyon-juniper plant communi- ties. USDA Forest Service General Technical Re- port lNT-58. Intermountain Forest and Range E.xperiment Station, Ogden, Utah. 917. Wyckoff, J W. 1973. The effects of soil texture on species diversity in an ariel grassland of the east- ern Great Basin. Great Basin Naturalist 33: 163-168. 918. Yake. S.ANDj D Brotherson 1979. Differentia- tion of serviceberry habitats in the Wasatch Moun- tains of Utah. Journal of Range Management 32: 379-386. 919. Young, J. A 1981. Demography and fire history of a western juniper stand. Journal of Range Manage- ment 34: 501-505. Young, J. A, AND J D BUDY 1979. Historical use of Nevada's pinyon-juniper woodlands. Journal of Forestry History 23: 112-121. Young, J. A., and R. A. Evans. 1970. Invasion of medusahead into the Great Basin. Weed Science 18: 89-97. 922. Young, J. A., and R. A Evans. 1973. Downy brome — intruder in the plant succession of big sagebrush communities in the Great Basin. Journal of Range Management 26: 410-415. 923. Young, J. A., and R. A. E\ans 1974. Population dynamics of green rabbitbrush in disturbed big sagebrush communities. Journal of Range Man- agement 27: 127-132. 924. Young, J. A., and R. A. Evans 1978. Population dynamics after wildfires in sagebrush grasslands. Journal of Range Management 31: 283-289. 925. Young, J. A , R. A. Evans, and R E Egkert. 1985. Successional patterns and productivity potentials of the sagebrush and salt desert ecosystems. In: 920. 921. Developing strategies for rangeland management. National Academy of Science. Westview Press, Boulder, Colorado. 926. Young, J. A., R. A. Evans, B A Roundy. and J A. Brown 1986. Dynamic landforms and plant com- munities in a pluvial lake basin. Great Basin Natu- ralist 46: 1-21. 927. YouNG,J.A.,R. A. Evans.P.T TuELLER. 1976. Great Basin plant communities — pristine and grazed. Pages 186-215 in R. Elston, and P. Headrick, eds., Holocene environmental change in the Great Basin. Nevada Archeological Survey Re- search Paper 6. Reno, Nevada. 928. Young, S 1958. E.xclosures in big game manage- ment in Utah. Journal of Range Management 11: 186-190. 929. Youngblood, A P , and R. L. Mauk 1985. Conifer- ous forest habitat types of central and southern Utah. USDA Forest Service General Technical Report lNT-187. Intermountain Research Station, Ogden, Utah. 89 pp. 930. Zamora, B 1968. Ai-femisia arhuscula. Ai-teinisia long,iloh(u and A;t("»ii.si« nova plant associations in central and northern Nevada. Unpublished thesis, University of Nevada, Reno. 931. Zamora, B , and P T. Tueller 1973. Ai-temisia arhuscula, A. longiloha, and A. /uha habitat types in northern Nevada. Great Basin Naturalist 33: 225-242. 932. Zan, M 1968. Evaluation of the effects of reduced transpiration upon soil moisture retentions in an aspen stand throughout the growing season in northern Utah. Unpublished thesis, Utah State University, Logan. 49 pp. 933. Zarn, M. 1977. Ecological characteristics of pinyon- juniper woodlands on the Colorado Plateau. A literature survey. USDI Bureau of Land Man- agement Technical Note 310. Denver, Colorado. 183 pp. 934. Zohner, K D 1967, A guide to the biogeographic literature of Utah. Unpublished thesis, University of Utah, Salt Lake City. 83 pp. Keyword-cit.xtion Index age-size structure 1, 125, 126, 134, 561 alpine 3, 62, 63, 74, 114, 115, 164, 376, 390, 480, 495, 503, 559, 571, 698, 699, 700, 711, 742, 827 autecologv 21, 22, 41, 46, 48, 61, 63, 71, 100, 104, 121, 122, 123, 133, 135, 137, 169, 176, 219, 231, 250, 279, 333, 351, 353, 355, 366, 379, 400, 402, 405, 438, 455, 494, 507, 543, 580, 616, 620, 649, 656, 687, 688, 717, 734, 766, 797, 800, 818, 830, 833, 844, 859, 882, 891, 903, 904, 918, 921 baseline studv 7, 8, 10, 11, 16, 18, 19, 20, 30, 31, 32, 33, 36, 46, 53, 54, 55, 56, 57, 65, 75, 76, 89, 98, 112, 117, 118, 119, 143, 151, 154, 156, 160, 163, 173, 174, 175, 185, 186, 193, 195, 196, 197, 198, 200, 204, 205, 206, 208, 210, 211, 221, 222, 227, 229, 236, 242, 243, 244, 246, 247, 248, 249, 251, 252, 257, 258, 262, 266, 267, 277, 281, 302, 303, 305, 320, 323, 326, 327, 328, 329, 330, 331, 332, 335, 337, 357, 362, 363, 364, 365, 371, 372, 373, 374, 375, 380, 381, 382, 384, 397, 398, 399, 404, 408, 417, 421, 422, 423, 424, 425, 433, 448, 453, 454, 1990] Nevada-Utah Vecktation Bibmochafhy 241 461, 462, 475, 486, 490, 504, 525, 526, 527, 529, 533, 538, 562, 572, 575, 595. 602, 603, 604, 605, 606, 611, 623, 628, 636, 648, 651, 685, 711, 719, 731, 732, 733, 745. 758, 769, 772, 809, 811. 813, 814, 819, 820, 823, 850, 851, 852, 853, 856, 873, 877, 885, 894, 906, 912, 932 hihliounipln 183, 184, 304, 600. 791, 812. 855, 875 hiogeograpliv 9, 49, 64, 66. 67, 68, 72, 74. 76, 104, 105, 106, 107, 109, HI, 123. 127, 129. 131, 144, 145, 148, 150, 155, 164, 167. 168, 169, 170, 180, 186, 189, 190, 192, 202, 213, 215, 225, 228, 232, 238, 242, 244, 245, 246, 249, 263, 264, 268, 271, 280, 284, 285, 297, 306, 307, 308, 309, 310, 315. 317. 321. 334, 338, 340, 341, 348, 351, 352, 361. 365. 371, 372. 377, 403, 406, 409, 411, 426, 427, 430, 431. 434, 437, 444, 451, 456, 458, 478, 479, 482, 485, 487, 491, 495, 496, 497, 498, 499, 500, 501, 502, 503, 505, 512, 514, 518, 523, 527, 542, 547, 549, 551, 556, 558, 559, 565, 577, 579, 583, 584, 596. 598. 608, 615, 620, 627, 631, 635, 647, 650, 651, 658, 661, 666, 667, 670, 672, 685, 689, 690, 695, 698, 699, 700, 701, 702, 703, 704, 705, 706, 707, 708, 716, 718, 721, 727, 729, 739, 742, 744, 749, 752, 755, 760, 771, 775, 789, 790, 796, 798, 802, 805, 806, 807, 808, 810, 816, 825, 840, 843, 845, 846, 864, 866, 867, 872. 878, 887, 891, 899, 907, 908, 910, 925, 934 classification 13, 51, 69, 138, 212, 280, 293, 393, 396, 444, 457, 459, 460, 522, 546, 560, 613, 614, 730, 747, 754, 786, 788 communitv analysis 3, 4, 7, 8, 15, 18, 20, 23, 25, 27, 29, 30, 35. 40, 42, 48, 49. 52. 60. 65, 71, 95, 96, 99, 103, 105, 107, 108, 113, 114, 115, 118, 119, 127, 129, 130, 131, 135, 139, 140, 141, 144, 150, 153, 155, 157, 160, 171. 174. 175, 177, 181, 182, 188, 189, 199, 208, 217, 233, 235, 241, 253, 258, 261, 276, 283, 288, 289, 290. 292. 294, 298, 299, 315, 321. 324, 325, 342, 344, 347, 349, 350, 354, 365, 370, 375, 378, 395, 407, 415, 439, 442, 449, 452, 456, 463, 465, 466, 469, 477, 481, 488, 493, 498, 499, 504, 511, 523, 524, 541, 544, 555, 557. 558. 563, 567, 570, 573, 574, 590, 607, 609, 617, 641, 642, 649, 653, 655, 673, 674, 675, 676, 696, 713, 748, 756, 761, 764, 773, 777, 779, 780, 782, 785, 792, 793, 801, 815, 828, 835, 838, 839, 849, 860, 862, 863, 865, 868, 870. 876, 884, 887, 896, 897, 900, 920, 923, 924, 926, 927, 928, 930, 933 desert 30, 46, 47, 48, 49, 51, 58, 69, 75, 90, 99, 102, 104, 109, 140, 149, 153, 162, 167, 199, 202, 204, 206, 227, 263, 264, 265, 266, 299, 300, 302, 305, 311, 321, 322, 349. 479, 502, 512, 531, 548, 549, 597, 622, 653, 675, 676, 690, 691, 695, 724, 725, 726, 727, 743, 761, 762, 789, 790, 825, 831, 832, 842, 871, 899, 907 disturbance 26, 35, 37, 45, 47, 50. 51, 58, 93, 124, 142, 156, 181, 221, 224, 248, 287, 288, 292, 294, 296, 344, 346, 385, 386, 416, 420, 464, 473, 475, 476. 492, 515, 517, 520, 528, 529, 531, 540, 550. 554, 566, 569, 582, 600, 609, 633, 668, 674, 676. 691, 723, 724, 725, 726, 739, 750, 763, 776, 778, 783, 809, 813, 824, 826, 834, 870, 880, 881, 914, 915, 916, 919, 923, 924 diversity pattern 291, 293, 324, 360, 369, 468, 610, 662, 749, 917 early exploration 34, 59, 152, 165, 239, 254, 259. 502, 549. 553, 589, 618, 644, 645, 646, 664, 697, 712, 735, 751, 774, 805, 836, 892 fire 35, 93, 116, 142, 156, 159, 181, 248, 292, 296, 344, 385. 386, 420, 464, 473, 476, 492, 515, 517, 518, 520, 540, 566, 633. 723, 750, 783, 809, 834, 880, 881,914,915,916,919,924 forest 1,3, 4, 13, 27, 28, 37, 39, 41, 93, 125, 145, 159, 160, 168, 177, 178, 179, 203, 219, 228, 230, 240, 260, 276, 280, 286, 304, 359, 367, 383, 387, 394, 402, 418, 419, 428, 446, 459, 460, 475, 481, 485, 488, 496, 507, 515, 516, 517, 518, 520, 522, 537, 561, 567, 568, 569, 570, 584, 608, 630, 655, 668, 702, 707, 712, 714, 715, 746, 754, 756, 763, 835, 839, 902, 929, 932 gradient analysis 120, 265, 278, 286, 510, 530, 545, 551, 563, 564, 637, 660, 736, 737, 738, 740, 741, 839 grassland 29, 34, 91, 110, 113, 176, 180, 214, 224, 262, 274, 278, 282, 283, 319, 340, 345, 347, 354, 355, 356, 357, 368, 438, 466, 467, 468, 469, 470, 510, 511, 514, 574. 598. 611, 612, 616, 624, 637, 682, 701, 708, 713, 728, 734, 753, 764, 817, 818, 833, 874, 916, 917, 924 grazing 24, 34, 40, 90, 91, 102, 124, 136, 159, 166, 194, 203, 206, 207. 208, 222, 223, 224, 240, 253, 269, 270, 273, 275, 282, 283, 300, 312, 313, 318, 319, 335. 337. 343, 355, 356, 360, 416, 418, 428, 447, 450, 463, 489, 490. 520, 548, 569, 591, 594, 597, 623, 633, 652, 657, 663, 668, 669, 681, 682, 683, 684, 728, 743, 750. 759, 761, 765. 767, 776, 792, 883, 904, 927 habitat type 3, 4, 13. 78. 92, 122, 137, 155, 212, 394, 395, 488, 537, 684, 754, 788, 799, 902, 929. 931 inventory 5, 9, 10, 11, 12, 16, 17, 19, 31, 32, 33, 36, 40, 50, 53, 54, 55, 56, 57, 62, 64, 66, 75, 79, 81, 82, 83, 84, 85, 86, 87, 88, 89, 97, 98, 102, 105, 106, 107, 108, 112, 117, 127, 131, 139, 148, 149, 154, 163, 164, 166, 167, 168, 176, 177, 180, 185, 186, 187, 188, 189, 192, 193, 195, 196, 197, 198, 200, 201, 205, 210, 211, 213, 215, 216, 226, 232, 236, 244, 245, 251, 252, 256, 257, 268, 271, 277, 280, 291, 306, 307, 309, 310, 316, 323, 326, 327, 328, 329, 330, 331, 332. 338, 340, 341, 352, 361, 369, 371, 372, 378, 380, 381, 384, 385, 387. 388, 389, 390, 391, 394, 395, 398, 399, 403, 404, 407, 408, 409, 410, 411, 413, 417, 421, 422, 423, 424, 425, 426, 430, 446, 448, 452, 453, 454, 461, 462. 465. 478, 479. 481. 482, 485, 487, 496, 497, 498, 499, 500, 503, 504, 512, 516, 525, 526, 527, 529, 534. 535, 537, 539, 541, 542, 547, 558, 559. 565, 568, 570, 575, 577, 583, 587, 595, 601, 602, 603, 604, 605, 606, 615, 616, 622, 626, 627, 628, 631, 635, 636, 647, 650, 654, 661. 665, 677, 678, 689. 699, 701, 702, 705, 707, 708, 712, 716, 718, 719, 721, 727, 731, 732, 742, 744, 745, 752, 755, 766, 772, 773, 775, 784, 789, 790, 796, 797, 802, 804, 807, 808, 810, 811, 813, 814, 815, 818, 820, 821, 822, 823, 825, 826, 829, 833, 840, 845, 846, 847, 848, 850, 851, 852, 853, 854, 858, 866, 871, 872, 874, 905, 909,910,911,926,929,931 map 25, 130, 150, 201, 459, 460, 539, 677, 895 model 38, 141, 142, 147, 161, 311, 367, 428, 513, 693, 762 relict vegetation 250, 483, 519, 521, 534, 536, 619, 804, 843 shrubland6, 23, 24, 45, 52, 61, 70, 77, 80, 90, 91, 94, 100, 101, 103, 109, 111, 116, 123, 124, 126, 132, 133, 135, 141, 155, 156, 166, 169, 170, 171, 172, 173, 174, 175. 182, 225, 243, 258, 264, 279, 294, 298, 302, 303, 305, 312, 333, 340, 353, 358, 360, 379, 242 P. S. BOURGERON ET AL. [Volume 50 385, 386, 389, 415, 416, 429, 430, 431, 443, 444, 449, 472, 477, 486, 530, 540, 555, 556, 557, 565, 579, 582, 583, 588, 591, 592, 594, 597, 598, 599, 619, 637, 657, 660, 669, 673, 674, 679, 680, 682, 723, 736, 737, 738, 743, 757, 770, 771, 791. 793, 794, 795, 799, 801, 816, 817, 838, 844, 855, 856, 858, 860, 862, 864, 865, 866, 867, 868, 871, 873, 874, 876, 878, 879, 882, 883, 889, 916, 918. 922, 923, 925, 930, 931 soil-vegetation relationship 14, 26, 42, 43, 44, 70, 71, 77, 79, 81, 82, 83. 84. 85, 86, 87, 88, 94, 120, 129, 132, 140, 143, 157, 158, 162, 204, 229, 230, 234, 237, 249, 254, 255, 257, 260, 263, 278, 286, 295, 297, 313. 317, 322, 336, 339, 345, 346, 350, 370, 376, 382, 392, 429, 442, 443, 445, 470, 471, 472, 479, 491, 508, 529, 534, 535, 546, 552, 554, 555, 556, 557, 563, 564, 576, 588, 621. 626, 630, 642, 643, 679, 686, 687, 691, 709, 710, 722, 746, 749, 771, 827, 853, 877, 879, 886, 893, 894, 898, 917 succession 23, 35, 37, 38, 47, 51, 65, 78, 80, 94, 110, 128. 151, 161, 172, 173, 178, 179, 181, 182, 188. 191, 209. 214, 224, 233, 248, 255, 258, 261. 262, 269, 270, 272, 273, 275, 282, 287, 288, 296, 300. 301, 358, 359, 360, 366. 367, 368, 385, 401. 434, 439, 440, 441, 447, 457, 467, 472, 473. 474. 476, 480, 484, 489, 490, 492, 515, 517, 520, 531, 532, 540, 550. 566, 569, 571, 573, 591. 594, 597, 633, 637, 638, 639, 640, 668. 669, 676, 680, 682, 683, 690, 693, 714, 715, 720, 723. 724. 725. 726, 739, 750, 753, 770, 774, 776, 778, 781. 783. 794, 795, 799, 803, 809, 817, 824, 831, 832, 834, 842, 843, 861, 869, 880, 881, 883. 885, 889, 890, 906, 915, 916, 919, 921, 922, 923, 924, 925, 927 wetland 95, 96, 115, 119, 120, 130, 301. 311, 340, 351, 399, 440, 442, 528, 585, 586, 629, 641, 659, 729, 769, 901 woodland 2, 7, 8, 14, 15, 23, 35, 39, 52, 60, 68, 73, 137, 142, 146, 147, 175, 212, 228, 235, 288, 289, 290, 291, 292, 293, 294, 295, 296, 318, 344, 400, 401, 435, 464, 473, 474, 482, 483, 544, 545, 546, 566, 573, 578, 583, 592. 593. 599, 617, 632, 638, 648, 665, 739, 750, 776, 777, 778, 779, 780, 781, 782, 783, 787, 798, 800, 822, 824, 834, 857, 858, 861, 869, 871, 874, 875, 884, 885, 886, 887, 888, 890, 898, 908, 916, 919. 920, 933 zonation 67, 72, 120, 435, 551. 581, 625, 653, 671, 704, 768. 788, 837, 841, 888 Received 1 July 1990 Accepted 15 September 1990 C;re;it Basin Naturalist 50(3). IWK), pp. 243-218 CONSERVATION STATUS OF THREATENED FISHES IN WARNER BASIN, OREGON Jack E. Williams', Mark A. Storir, Alan V. Mimliall', and Cary A. Anderson' Abstr.'\c:t. — Two federalK' listed fishes, tlie Foskett speckled dace and Warner stickler, are endemic to Warner Basin in south central Oregon. The Foskett speckled dace is native only to a single s]5ring in (Coleman Valley. A nearby spring was stocked with dace in 1979 and 1980, and now provides a second population. The present nimihers oi dace prohahly are at their highest levels since settlement of the region. The Warner sucker historically occurred throughout much of the Warner X'alley, but its distribution and abimdance have been reduced by construction of reservoirs and irrigation dams and the introduction of predator) game fishes. Lentic habitats have become dominated by introduced fishes, particular!)- white crappie, black crappie, and brown bullhead. The largest remaining population of Warner suckers occurs in Hart Lake, where successful reproduction was documented but there is no evidence of recruitment to the adult population. Two threatened fishes inhabit separate val- leys in Warner Basin, Oregon. In Coleman Valley the only native fish is the Foskett speckled dace, Rhinichthijs osciilus ssp., which occurs in Foskett Spring along the west margin of the Coleman Lake bed. The lake is dry except during years of exceptional rainfall. The dace was listed as threatened because of small population size, trampling of its re- stricted habitat by cattle, and subsequent degradation of the springpool area (U.S. Fish and Wildlife Service 1985a). To provide a refuge population free of the effects of intense livestock grazing, 50 dace from Foskett Spring were transplanted on 14 November 1979 into an unnamed spring (now known as "Dace Spring") on Bureau of Land Management (BLM) land approxi- mately 1.5 km south of Foskett Spring. An- other 50 dace were transferred into the spring on 26 August 1980. A reproducing population subsequently established in Dace Spring, and more than 300 dace of three size classes were observed there in 1986 (BLM Lakeview Dis- trict, unpublished data). The presumed historical range of the War- ner sucker, Catostomus ivarnerensis, con- sisted of the main Warner lakes (Pelican, Crump, and Hart) and other accessible lakes and sloughs in Warner Valley, and low- to moderate-gradient reaches of tributary streams. The species description by Snyder (1908) was based on specimens collected from Deep (= Warner) Creek near Adel. The Warner sucker was listed as threatened pri- marily because of fragmentation of stream habitats by irrigation diversion dams and the establishment of large populations of intro- duced piscivorous fishes in lentic habitats (U.S. Fish and Wildlife Service 1985b). Long-time residents recalled that during the 1930s large numbers of spawning Warner suckers (referred to as "redhorse") ascended Honey Creek far into upstream canyon areas (Andreasen 1975). By the 1970s the species range was fragmented by numerous irriga- tion diversion dams on the lower reaches of streams tributary to Pelican, Crump, and Hart lakes (Andreasen 1975, Kobetich 1977, Swensen 1978, Coombs et al. 1979, Hayes 1980), which block spawning runs from the lakes into streams. Coombs et al. (1979) found that although habitats had been fragmented resident stream populations still persisted. Nearly two-thirds of all adult suckers (198 of 300) were captured by Coombs et al. (1979) in the canal between Anderson and Hart lakes, immediately north of the Hart Lake spillway. Adult and larval suckers also were captured in Snyder Creek, in Honey Creek above the dam at Plush, at the mouth of Honey Creek in Hart Lake, at the south end of Warner Valley in Twentymile Creek between the south end of the valley floor and the confluence with Twelvemile Creek, and in Twelvemile Creek immediately above and below the O'Keefe Diversion Dam. 'Division of Wildlife and Fisheries, Bureau of Land Management, 18th 6c C Streets, N.W., Washington, U. C. 20240. ^he Nature Conservancy. 12()5 Northwest 25th Avenue. Portland, Oregon 97210. Bureau of Land Management. Box 151. Lakeview. Oregon 97630. ■•Oregon Department of Fish and Wildlife, Bo.x 1214, Lakeview, Oregon 97630. 243 244 J. E. Williams ETAL. [Volume 50 In 1980 Coombs and Bond (1980) sampled 22 sites throughout the basin, capturing 46 Warner suckers at 4 localities: Honey Creek between Hart Lake and the dam at Plush, canals of Deep Creek at the east end of Pelican Lake, the spillway immediately north of Hart Lake, and Swamp Lake. In 1983 Smith et al. (1984) captured 1 adult Warner sucker in Crump Lake and 2 juveniles (approximately 130 mm total length [TL]) in Deep Creek between Adel and the falls. In 1987 an adult Warner sucker was caught by an angler along the slough just south of Flagstaff Lake (J. E. Williams, personal observation). This paper summarizes the current status of these two threatened fishes, as determined by surveys conducted from 1987 to 1989. Other native fishes of Warner Valley include a local form of redband trout (OncorJiynchus mykiss ssp.), tui chub (Gila bicolor), and the common form of speckled dace (R. osculus). The Warner Valley redband trout largely has been displaced by introduced trout and is listed as "of special concern" by the American Fisheries Society (Williams et al. 1989). Habitat Description and Survey Methods The Warner Basin comprises 6858 sq km in south central Oregon and small portions of northeastern California and northwestern Ne- vada (Fig. 1). Drainage is internal and is di- vided between Coleman Valley and the much larger Warner Valley. Coleman Valley is a separate drainage in the southeastern part of the basin and receives sparse runoff. Observa- tions of the Foskett speckled dace and its habi- tat in Coleman Valley were made from 1987 to 1989. Standardized transects were estab- lished along Foskett Spring and its outflow to monitor vegetation recovery following cessa- tion of grazing, and to (juantify amounts of open-water habitat. In Warner Valley all water flows into a se- ries of north-south oriented shallow lakes, sloughs, and potholes. During periods with above-average precipitation, as occurred dur- ing the early 198()s and again in 1989, these lakes fill from the south and eventually over- flow into the northern part of the valley. Only the three most southerly lakes, Pelican, Crump, and Hart, are permanent. Fish col- lections in Warner Valley were made from 1987 to 1989. Samples were collected from lakes by use of traps, gill nets, and seines, and from streams with dip nets, trap nets, elec- troshocker, kick nets, and seines. Most fishes were identified, measured, and returned to their habitat. Voucher specimens or those ac- cidentally killed during collecting are housed at the Wildlife and Fisheries Museum, Uni- versity of California, Davis. Opercles from five suckers were aged according to the meth- ods described by Scoppettone (1988). Visual observations were made of spawning Warner suckers in Honey Creek. Foskett Speckled Dace In 1987 the BLM acquired Foskett Spring and the surrounding 65 ha, of which approxi- mately 28 ha were fenced to exclude cattle. The dace population at Foskett Spring has since expanded to the spring pool, its outflow, and downstream marsh. Baseline water qual- ity and vegetation monitoring at Foskett and Dace springs were initiated by BLM in 1987. The following data collected on 28 September 1988 from Foskett Spring and Dace Spring, respectively, exemplify the two habitat simi- larities: air temperature 19 and 17 C, water temperature 17 and 16 C, dissolved oxygen 5.3 and 5.9 mg/1, conductivity 350 and 250 mohs/cm, pH 8.1 and 8.2, alkalinitv 114 and 99 mg/1 CaCO.j, hardness 40.0 and 24.7 mg/1, and turbidity 1.4 and 1.8 NTU. The dace population maintains itself at Dace Spring despite a tendency for vegetation to choke out most open water. The introduced population has expanded by movement offish through a connecting pipe into a livestock watering trough just east of the spring. No other fish occur in Coleman Valley. Warner Sucker Surveys on Twentymile Creek above and below the Dyke Diversion Dam located 1 adult and 2 larval Warner suckers in 1988. Additional 1987 and 1988 surveys failed to locate Warner suckers elsewhere in Twelve- mile Creek (including sections in Nevada and Oregon upstream of the Nevada border), the canal north of Hart Lake, the slough between Flagstaff Lake and Mugwump Lake, the slough between Lower Campbell and Camp- bell lakes, or Stone Corral Lake. In April 1989, 28 adult suckers were captured at the 1990] Threatenkd Warner Basin Fishes 245 Fig. 1. The Warner Basin of south central Oregon. 246 J. E. Williams ETAL, [Volume 50 Table L Frequency of fishes collected in Warner Basin during 1987-89. All collection sites are in Lake County, Oregon, unless otherwise noted. Collections made at the same habitat are combined. Location Warner sucker Tui chub Speckled dace Trout' Large- White Black mouth Spotted Brown crappie crappie bass bass bullhead Twelvemile Cr. (Washoe Co. , NV) Twelvemile Cr. Dyke Diversion Canal Irrigation canal along Twentymile Cr. Twentymile Cr. Greaser Reservoir Deep Creek Hart Lake lower Honey Creek upper Honey Creek canal north of Hart Lake Anderson Lake Flagstaff Lake slough slough between Lower Campbell and Campbell Campbell Lake Stone Corral Lake Total caught Relative catch (%) 404 591 51 1 1 6 25 476 854 5 400 40 70 12 19 1620 2 1 14 4 2 449 7 2 7 27 82 17 31 10 107 39 371 5 30 59 1 I 40 95 152 649 22 69 96 2183 106 2 1 607 2.5 10.7 37 '.4 1.6 26.0 1.7 <0.1 <0.1 10.0 ■"May include native redhand trout and/or introduced rainbow trout. mouth of Honey Creek in Hart Lake, and 42 were captured along the east side of Hart Lake. Fish ranged from 311 to 440 mm TL (avg. 385.2, n = 70), with most 350 to 410 mm. Approximately 80-100 other adult Warner suckers were observed in Honey Creek be- tween the most downstream diversion dam and Hart Lake. These fish were in breeding condition and migrating upstream, where they were visible because flow in the creek was reduced by upstream diversions. In mid- May 1989 water began spilling from Hart Lake into the canal toward Anderson Lake. Suckers dispersed into the canal, and 7 spawners were collected there in June. Stan- dard length, TL, and age of 5 of these were 331, 357, 7; 307, 361, 7; 333, 387, 7; 335, 390, 9; and 340, 397, 8. Larval suckers also were collected from Honey Creek just above the downstream-most diversion dam, indicating at least limited spawning upstream. Overall, Warner suckers constituted onlv 2.5% of all fishes collected during 1987-89 (Table 1). Nearly all suckers were found in Hart Lake, Honey Creek just upstream of Hart Lake, or the canal immediately north of the lake. Introduced fishes dominated the fauna of Hart Lake and other lakes and sloughs in the valley. White crappie {Pumoxis annu- laris) and brown bullhead {Ictalurus nebitlo- siis ) outnumbered native fishes in our collec- tions from Hart Lake by slightly more than 25:1. Tui chub, which historically was the most abundant fish in lentic habitats, largely has been replaced by white crappie. The Warner sucker population appears to be largest in Hart Lake, but no recent recruit- ment could be documented. Except for a small number of larvae in lower Honey Creek, no suckers smaller than 310 mm TL were found. White crappie were alnmdant at the mouth of Honey Creek during June and may have preyed on sucker larvae as they drifted into Hart Lake. A single trap net set there in June collected 1530 white crappie and 20 brown bullliead. Discussion The Foskett speckled dace appears to be near recovery. No exotic species are present 1990] Threatened Warner Basin Fishes 247 in either spring, and the priniar\' threats ha\c' been ehminated. Some vegetation needs to l)e eleared fioin the pool at Daee Spring in order to pro\ ide sufheient opcMi water. Also, feneing along tlie honndar\' of Daee Spring shonld be extended to the east to inehide additional habitat. Continued habitat and population monitoring are neeessary at both springs beeause the small habitats are vulner- able to slight disturbanees. The largest remaining population of War- ner suckers appears to be in Hart Lake, where spawning fish ascend lower Honey Creek and the canal north of the Hart Lake spillway. Populations also may exist in Crump and Peli- can lakes. Successful recruitment of young into the Hart Lake population is limited by reduced spawning habitat in Honey Creek and large populations of crappie. White crappie were introduced into Hart Lake in 197L and white plus black {P. nigromaculatus) crappie were introduced into Crump Lake during 1972 and 1973 (Oregon Department of Fish and Wild- life, unpublished data). Subsequent collec- tions of the Oregon Department of Fish and Wildlife indicated that white crappie, black crappie, and brown bullhead were common in Crump Lake by 1978 (K. Daily, unpublished data) and presumably in Hart Lake as well. Adult white crappie commonly feed on small fishes (Pflieger 1975); thus, their abundance at the mouth of Honey Creek during the same time that larval suckers were collected from the creek increases the likelihood of predation on young-of-year suckers. Seven irrigation dams on Honey Creek be- tween the lake and Plush result in limited access by adults to upstream spawning areas. During 1989 only two riffles between Hart Lake and the first diversion dam contained suitable gravel for spawning. Depending on stream flows, water-diversion boards may be placed in the irrigation structures before, during, or after the spawning run. Swenson (1978) reported that during 1978 adult suckers migrated as far as the seventh irrigation dam at Plush before boards were installed and wa- ter diverted for irrigation. A remnant population of Warner suckers may persist in Crump Lake, as indicated by collection in 1989 of young-of-year in Twenty-mile Slough below Greaser Dam. Ad- ditional surveys of Crump and Pelican lakes are needed to determine the extent of any remaining sucker populations. If present, however, recruitment may be prevented by populations of crappie. In conclusion, Warner suckers once were common throughout the basin but gradually declined from about 1900 until the early 1970s as a result of agricultural development and placement of irrigation structures in spawning streams. Despite habitat fragmentation and lack offish passage, recruitment to lake popu- lations continued until the late 1970s, when large populations of piscivorous fishes became established. Recruitment of Warner suckers continues in stream habitats but appears from our observations to be greatlv curtailed since 1979. Control of introduced fishes in Hart and Crump lakes may be impractical because of habitat size (2928- and 3108-ha area, respec- tively) and large populations. Recovery of the Warner sucker in Hart Lake therefore at least requires increased spawning sites and rearing habitat. Acknowledgments Our fieldwork was aided by W. J. Berg, R. G. Bolton, C. A. Macdonald, J. F. Morawski, G. A. Rosenberg, L. M. Swinney, and R. K. White. Mark Warner of the Nevada Department of Wildlife facilitated our collec- tions in Nevada. Earlier drafts of this manuscript benefited from reviews by J. K. Andreasen, W. J. Berg, and C. D. Williams. Age analysis of the Warner suckers was kindly confirmed by G. G. Scoppettone. This work was completed while the senior author was on an Intergovernmental Personnel Act appoint- ment with the U.S. Fish and Wildlife Service at the University of California, Davis. Our sincere appreciation to Mr. Joe Flynn for ac- cess to his property along Honey Creek and for his interest in the fishes therein. Literature Cited Andreasen. J. K. 1975. Systematics and status of the family Catostomidae in southern Oregon. Unpub- Ushed dissertation, Oregon State University, Cor- valhs. 76 pp. Coombs, C I , andC. E Bond 1980. Report of investiga- tions on Catostomiis wamerensi.s, fall 1979 and spring 1980. Report to U.S. Fish and Wildlife Service, Sacramento, California. 248 J. E. Williams ETAL. [Volume 50 Coombs, C. I.. C. E. Bond, and S. F. Drohan. 1979. Spawning and early life history of the Warner sucker (Catostomus warnerensis). Report to U.S. Fish and Wildlife Service, Sacramento, Califor- nia. Hayes, J. P. 1980. Fish of Warner Valley. Pages 131-137 in C. Oilman and J. W. Feminella, eds.. Plants and animals associated with aquatic habitats of Warner Valley. Oregon State University, Corvallis. KOBETICH. G. C. 1977. Report on survey of Warner Valley Lakes for Warner suckers, Catostomus warneren- sis. Report to U.S. Fish and Wildlife Service, Sacramento, California. 6 pp. Pflieger, W. L. 1975. The fishes of Missouri. Missouri Department of Conservation. 343 pp. SCOPPETTONE, G. G. 1988. Growth and longevity of the cui-ui and longevity of other catostomids and cyprinids in western North America. Transactions of the American Fisheries Society 117: 301-307. Smith, M , T Steinback, and G. Pampush 1984. Distri- bution, foraging relationships and colony dynam- ics of the American White Pelican (Pclecanus erythrorhynchos) in southern Oregon and north- eastern California. Oregon Department of Fish and Wildlife Nongame Technical Report 83-0-04. Snyder. J O. 1908. Relationships of the fish fauna of the lakes of southeastern Oregon. Bulletin of the Bu- reau of Fisheries 27(1907); 69-102. Swenson, S. C. 1978. Report of investigations on Cato- stomus warnerensis during spring 1978. Report to LT.S. Fish and Wildlife Service, Sacramento, Cali- fornia. 27 pp. US Fish and Wildlife Service. 1985a. Determination of threatened status for Hutton tui chub and Fos- kett speckled dace. Federal Register 50: 12302- 12306. 1985b. Determination that the Warner sucker is a threatened species and designation of its critical habitat. Federal Register 50: 39117-39123. Williams, J E , J E Johnson, D. A. Hendrickson, S. Contrer.\s-Balder.\s, J. D. Williams, M Navarro-Mendoza, D E. McAllister, and J E De.\con 1989. Fishes of North America en- dangered, threatened, or of special concern: 1989. Fisheries (Bethesda) 14(6): 2-20. Received 30 'November 1989 Revised 11 August 1990 Accepted 6 September 1990 Great Basin Naturalist 50(3), 199(), pp 249-256 HOME RANGE AND ACTIVITY PATTERNS OF BLACK-TAILED JACKRABBITS C;raliaiii W. Sniitli' Abstract. — Home range use and activity patterns ol black-tailed jackiahhits (Lcpus californicus) in northern Utah were studied using telenietr\'. Home range sizes ranged Irom <1 km" to 3 km" and did not differ between sexes or among seasons. Jackralihits were inactive during daylight, became active at dusk, and remained active throughout the night. Animals often traversed their home ranges in a tew horns. During the I)reeding season, males were more active than females. Jaekrabbits were most aeti\ e dining well-lit nights, and high winds decreased jaekrabbit activity. The black-tailed jaekrabbit occupies a wide geographic area and is an important compo- nent of the biota throughout its range. In the Great Basin the jaekrabbit is the most abun- dant large herbivore (Wagner 1981) and serves as an important prey item for many predators. Considering the central role of the jaekrabbit in many ecosystems, little research on the species has been reported. Detailed quantitative information regarding activity patterns and home range use is lacking. Home range use varies with the patterns of food, cover, and water distribution (Dunn et al. 1982). I examined patterns of jaekrabbit activ- ity and home range use throughout a calendar vear in shrub-steppe vegetation in northern Utah. Study Area The study was conducted in northern Utah near the Wildcat Hills in Curlew Valley, about 10-35 km north of the Great Salt Lake. Topography and vegetation of the area are described in detail by Gross et al. (1974). Four major vegetation types occur in Curlew Valley: (1) open stands of juniper (Juniperus osteosperina) at higher elevations, (2) big sagebrush {Artemisia tridetitata) in the north- ern portions of the study area, (3) greasewood {Sarcobatus vermiculatus) in more saline soils closer to the Great Salt Lake, and (4) expanses of salt-desert vegetation, primarily shadscale {Atriplex conferti folia) and saltbush {Atriplex falcata), scattered throughout the study area. Southern portions of the valley are typically more xeric than northern portions, and pre- cipitation is most abundant during winter and spring. Acciunulated snowfall in 1983-1984 was 69 cm at Snowville, Utah, 15 km east of the study area, with snowcover persisting from mid-November through mid- March (National Oceanic and Atmospheric Adminis- tration 1984). Methods Black-tailed jaekrabbit activity and home range use were monitored via telemetry. Inci- dental direct observations of jaekrabbits also aided in describing activity patterns. Black-tailed jaekrabbits were captured by night-lighting and netting (Griffiths and Evans 1970); they were then equipped with radio-transmitters and released. Periodically during each season additional animals were caught and instrumented to replace those that had died. During winter some jaekrabbits were captured in live traps and handled simi- larly to those captured by netting. Sex was determined from external examination of gen- italia, and age class was estimated from body size, color, and relative eye size (L. C. Stod- dart, unpublished data). Transmitter collars were designed to mini- mize chafing of the jaekrabbits neck (Wywi- alowski and Knowlton 1983). I assume the transmitters had little discernible eflFect on jaekrabbit behavior (Stoddart 1970, Donoho 1972, Brand et al. 1975, Keith et al. 1984). Telemetry stations on the Wildcat Hills overlooked areas with instrumented jack- rabbits. Each station was equipped with two horizontally stacked 5-element yagi anten- nas, coupled out of phase with a sum-and- diflference hybrid junction. A compass rose 'Department of Fisheries and Wildlife, Utah State University, Logan, Utah 84.322, Present address: Office of Migratory Bird Management, U.S. Fish and Wildlife Service, Laurel, Maryland 20708. 249 250 G. W. Smith [Volume 50 mounted on the antenna mast indicated the directional orientation of the antennas. A transmitter placed at a known azimuth fiom each station was used as a beacon to orient the compass rose to true north. Home range use was assessed by repeat- edly recording azimuths of animals simulta- neously from two tracking stations. Four-hour tracking sessions were distributed throughout the 24-hour day with most occurring between dusk and dawn. Locations were recorded every 20 minutes for a 4-hour period on all animals whose transmitter signals could be detected. For each reading it was noted whether the signal varied in amplitude, sug- gesting movement of the transmitter antenna, which was interpreted as movement by the animal. The amount of night light was classi- fied into one of three categories (low, me- dium, or high), depending upon the phase of the moon and cloud cover. Wind intensity during the 4-hour period was classified as low or high by noting the wind conditions at the tracking shelters. Periods of no wind were included in the low category. Home range use was assessed using pro- gram HOME RANGE (Samuel et al. 1985b). This program offers a series of statistical tests to derive the appropriate home range estima- tor (Samuel and Garton 1985). Within each data set locations outside the home range were identified statistically (Samuel and Gar- ton 1985) and discarded (given a weight of 0) if they appeared to be errors or excursions out- side the normal home range area (Burt 1943). "Core areas" were identified (Samuel et al. 1985a). As recommended by Samuel and Gar- ton (1985), only data sets with >50 locations were analyzed. Comparisons of the sizes of areas used by subsets of the jackrabbit popula- tion were made using a two-way analysis of variance (SAS Institute Inc. 1985). Because of the model-selection criteria of the program, I used the harmonic mean esti- mator (Dixon and Chapman 1980) for all area- of-use analyses. Dixon et al. (1981) recom- mended this technique to analyze lagomorph spatial use because it eliminates many prob- lems associated with other analyses. Choice of a contour isopleth to represent the home range is somewhat arbitrary (Ander- son 1982). I chose the 80% contour for jack- rabbits because it appears to reflect observed patterns of land use by these animals. To allow comparisons with other published accounts, I also report the 95% contour interval, although it probably overestimates home range size. No statistical rationale exists for the choice of the 95% level, and its use may result from biologists confusing utilization distributions with alpha levels in statistical tests (White and Garrott 1990). The relationships of time of day, season, sex, amount of moonlight, and wind intensity to jackrabbit activity were assessed using log- linear analyses (Sokal and Rohlf 1981). Terms included in the resulting models reflect signif- icant relationships within the data. Seasons were defined by Curlew Valley weather pat- terns. Winter ended with the melting of snow in March. Summer began in late June 1984 with the onset of hot temperatures and ended in early September with the arrival of fall rains. Results The daily movements of 16 jackrabbits were monitored from February through April 1984, and those of an additional 44 jackrabbits were monitored from June through Novem- ber 1984. I determined the sizes of areas used by 30 jackrabbits, with 5 animals having 2 areas each, for a total of 35 areas of use (Table 1). The time periods for which areas were measured ranged from eight days to five months. Home range has been defined as the area used by an animal on a day-to-day basis (Burt 1943). How an animal uses its home range affects how long the animal must be monitored before it traverses its entire home range. Typical home range use by black-tailed jackrabbits involved extended use of an area measuring <1 km". Animals often traversed the whole area of activity in less than four hours. By dawn a jackrabbit was usually back near the previous day s resting location. Some animals maintained this pattern of space use for up to three months. Periodically, jackrabbits changed their areas of use. These changes involved exten- sions of the areas of use into previously un- used areas and abandonment of portions of the previously used areas. New areas were then generally used for extended periods. These shifts in areas of use enlarged overall home ranges to 1.5-3 km". No differences in the patterns of shifts in areas of use among differ- ent sex and age segments of the population 1990] Black-tailkd Jacki^mum r Home Rance 251 Table L Black-tailed jatkrahhit home range sizes (km") (and standard errors) in (anlew Valley, Utah, 198.3-1984. Nund)er of animals Harmonic contour are a Season 9.5% SE 80% SE core SE Adults Winter Male Female 2 6 2.51 1.25 0.77 0.45 1.66 0.83 0.64 0.28 0.96 0.47 0.33 0.15 Spring Male Female 5 4 1.08 0.85 0.25 0.16 0.73 0.,55 0.18 0. 13 0.43 0..32 0.13 0.06 Summer Male Female 5 4 2.88 1.63 0.72 0.49 1.83 1.05 0.45 0.29 1.07 0.62 0.31 0.16 Fall Male Female 1 2 2.21 0.87 0.14 l.,30 0.,52 0.10 0.72 0.26 0.06 Juveniles Summer-fall Male Female 1 5 1.13 1.70 0.,35 0.73 0.92 0.21 0.43 0..57 0. 15 Two-way analysis of variance df using the 80% contour SS F P Season Sex Season X sex Error 3 1 .3 21 3.10 2.14 0.,56 1.5,41 2.35 4.85 0.42 .10 .04 .74 were apparent. Core-areas, as identified by program HOME RANGE, encompassed areas slightly larger than half the size of home ranges (the 80% contour) (Table 1) and tended to be near the center of the home range areas. Jackrabbits changed home range areas sea- sonally. Where more than one home range was recorded for an animal, each was analyzed independently (Table 1). Areas traversed dur- ing seasonal migrations were not considered part of either home range (Burt 1943). In March, just after the winters snowcover melted, jackrabbits left wintering areas and moved to new home ranges. In the fall many animals abandoned their summer home range areas and moved to wintering areas (Smith 1987). Wintering areas were located in stands of tall vegetation, primarily greasewood and sagebrush, although a few juniper stands were also used. Wintering areas encompassed only a small portion of the available habitat as jackrabbits concentrated in groups; areas of low vegetation were not used during winter. With spring shifts in home range areas, jackrabbits reoccupied much of the valley (Smith 1987). Sizes of areas appeared to change season- ally, with areas of use being smaller in spring and fall than in winter and summer. Patterns of home range use also appeared to be similar throughout the year. Males tended to have slightly larger home ranges than females (Table 1). Too few juvenile jackrabbits were instrumented to adequately determine their pattern of home range use. However, sizes of areas used by juveniles and adults did not differ (Table 1). Only juveniles large enough to wear a radio-transmitter (>3 months of age) were instrumented. Jackrabbit activity changed daily and sea- sonally (Fig. 1). Significant differences oc- curred during the day and among the seasons time, r = season; In f(ij|,, = |x - a3„ + ar,, + pr,, + apr,,; (a = active, (3 = + a, + p, + r, ap- G = 179.5, 18 df, P < .001; aV,^: G 594.3, 20 df, P < .001; pr,^: G = 856.3, 30 df, P < .001; apr.jk: G = 128.8, 15 df, P < .001). Jackrabbits were least active from 0800-1700 hours during the spring, summer, and fall, when they often rested in shallow depressions under shrubs, big sagebrush and grease- wood being the most common. In addition, some jackrabbits used badger (Taxidea taxus) 252 G. W. Smith [Volume 50 U.On .0.7. -^ o Winter ~ 0.6 i \ • Spring < 0.5- o 0.4- A Summer ^ Fall S 0.3, TV ^ \/^^^ \ . Q. C o 0.2^ C\_^ \ /^ '^ 0.1- x^ \ ^yx \ , 4 ^^^^^ 1800 0000 0600 Time 1200 1800 Fig. 1. Frequency of black-tailed jackrabbit activity atid lelationships with time of day and season of year in Curlew Valley, Utah, 1983-1984.' Table 2. Frequency of black-tailed jackrabbit activity and relationships with sex and season of year in Curlew Valley, Utah, 1983-1984. The data are the number of telemetry readings under the categorical conditions. The proportion active is in parentheses. Active (a) Se.\ (P) Season (F) Female Male Winter yes no 129 (.24) 409 100 (.29) 240 Spring yes no 108 (.16) 572 88 (.20) 336 Summer yes no 147 (.24) 456 156 (.22) 555 Fall yes 21 4 (.17) (.04) 101 87 In f(„k) fx + a, + p, + r, + ap„ + aF,t + pF,, + apr,,^ = 17.6, 4df, P<.005 = 94.3, 6df P<.001 = 52.3, 6df F< .001 = 17.0, 3df P< .001 burrows during winter, especially when deep snows reduced mobility. Jackrabbit activity changed seasonally, with jackrabbits least active during fall and most active in winter and summer (Fig. 1). Daily activity patterns also changed seasonally, with jackrabbits less active during morning hours (after midnight) in the fall. The sexes showed different patterns of activity (Table 2), with males slightly more active than females during winter and spring. Females were slightly more active than males during summer and much more so during fall. The proportion of observations classified as active was lower during fall (Table 2), consis- tent with the results of the previous analysis (Fig. 1). Jackrabbit activity was apparently influ- enced by the amoimt of light dining night hours (Table 3). The proportion of observa- tions in which jackrabbits were active at night changed seasonally, with jackrabbits most ac- tive in summer and least active in fall. The relationship of activity and night light also changed seasonally (Table 3). During fall, winter, and spring, jackrabbits were most ac- tive when night light was greatest (a fidl moon and little cloud cover). During summer the amount of night light did not appear to influ- ence jackrabbit activity. Jackrabbits were most active when there was little wind; high winds were associated with decreased jackrabbit activity (Table 4). Activity in relation to wind intensity changed seasonally (Table 4), with jackrabbits less likely to be active during the winter when winds were high. Sampling of jackrabbit activ- ity during high winds in the fall was insuffi- cient for basing conclusions. Discussion Home Range Shape and Size The shape of most jackrabbit home ranges tended to be elliptical. The shape is not a result of the locations of the tracking shelters in relation to radio-collared jackrabbits, as very acute or obtuse telemetry bearings were excluded from the analysis. A similar elliptical shape was noted by Rusch (1965), who ob- tained many locations from snow-tracking. The sizes of black-tailed jackrabbit home ranges determined in this study (Table 1) are larger than those reported for the species in other studies. Rusch (1965) and Nelson and Wagner (1973), also working in Curlew Val- ley, reported that jackrabbits used areas <0.2 km" (minimum convex polygon) for periods of one to two months in the fall and winter. It is not clear why the home ranges I report are greater in size than those described in these earlier studies. Two factors, however, may have had some influence. My relocation effort was more intensive than those conducted pre- viously, resulting in a greater number of re- locations per animal. Moreover, advances in 1990] Black-tailkd Jackkahbii Home Ran(;e 253 Table 3. Freciuency of black-tailed jatkiahhit activity and relationships with ainouiit ol lij^ht at night and season of year in Curlew Valley, Utah, 1983-1984. Tlie data are the nninher of telemetry readings observed under the categorical conditions. The proportioti active is in parentheses. Season (D Active (a) Amount of'light (p) Mi'diui High Winter \es no Spring yes no Summer yes no Fall yes no In f(„|„ = |x + a, + p, + r^ + ap„ + af,, ap,- G = 76.0, 8df, P< .001 al\: G = 103.2, 9df, P<.001 pr,^: G = 721.5, 12df', F< .001 apr,,^: G = 56.1, 6 df, P < .001 pr. 133 ( 477 33 ( 138 234 ( 449 12 (.21) (.19) (.34) (.09) 126 37 ( 25 81 ( 223 77 ( 175 21 ( 104 (.60) .27) .31) ,17) 23 ( 21 48 I 66 143 273 6 ( 16 (.52) (.42) (.34) (.27) Table 4. Frequency of black-tailed jackrabbit activity and relationships with wind intensity and season of year in Curlew Valley, Utah, 1983-1984. The data are the num- ber of telemetry readings under the categorical condi- tions. The proportion active is in parentheses. Wind i ntensity (p) Season (F) Active (a) Low High Winter yes no 214 (.32) 459 14 (.09) 134 Spring yes no 112 (.18) 512 52 (.28) 211 Summer yes no 392 (.26) 1125 107 (.21) 396 Fall yes 33 6 (.09) .27) 318 16 Inf (ijk) apr.j,: jx + a, + Pj + r^ + ap,^ + af.^ + pf^^ + apF.^^ = 252.3, 4df, P< .001 = 704.1,6df, P< .001 = 759.7, 6df, P< .001 = 198.8, 3df, P< .001 radiotelemetric and analytical procedures since the early 1970s may have played an im- portant role. Lechleitner (1958a) in California, French et al. (1965) in Idaho, and Tiemeier (1965) in Kansas also reported seasonal ranges of <0.2 km". Areas of use were determined in these three studies from visual observations of marked animals. Differences in sizes of home ranges may also reflect differences in habitats among study areas. Habitat in Lechleitner's and Tiemeier's studies consisted of pastures and cultivated land. Plant cover on French et al.'s (1965) study area was similar to the native Great Basin shrub-steppe vegetation of Curlew Valley. Two limitations of the home range analysis may have influenced the results. These are the precision of the tracking system and serial correlation of the location data. I believe that the precision of the tracking system averaged 200 m (Smith 1987, Mills and Knowlton 1989). Sequential locations had to be at least 200 m apart before I could be certain that movement had occurred. Locations closer to the tracking shelters had greater precision, probably to a minimum of 100 m, while locations more dis- tant from the shelters had less precision, probably less than 300 m. The most precise locations were located along the arc created using the baseline between the two towers as the diameter of a circle. The reduced preci- sion in locating distant jackrabbits may have produced overestimates of home range sizes. Home range estimation methods assume that locations are serially independent (Swihart and Slade 1984). When serially correlated data are 254 G, W. Smith [Volume 50 analyzed, these methods underestimate home range size. All the location data sets showed some degree of serial correlation. I used all locations obtained for each animal and did not subsample to decrease serial correlation be- cause, for many of the animals, too few loca- tions were obtained to allow me to discard data points and still have a minimum sample size of 50 locations per analysis (Table 1). Be- cause home range size is a statistic, the great- est value of which lies in comparisons among subsets of a population (White and Garrott 1990), I opted for the procedure that gave me larger samples. Home Range Use Resting in forms during daylight hours is a behavior observed in virtually all black-tailed jackrabbit populations (Vorhies and Taylor 1933, Lechleitner 1958b, Rusch 1965, Haug 1969, Costa et al. 1976, and Flinders and El- liot 1979). Forms are shallow depressions in or under bushes (Vorhies and Taylor 1933). Form use by Curlew Valley jackrabbits appears typical for the species. The use of burrows during the winter in Curlew Val- ley, however, appears unusual (Lechleitner 1958b). Jackrabbits use burrows to evade predators (Vorhies and Taylor 1933, personal observation) and construct shallow burrows to escape summer heat in the Mohave Desert (Costa et al. 1976), but daily use of deep bur- rows has not been reported previously. I be- lieve jackrabbits use burrows during winter in Curlew Vallev to reduce the risk of predation (Smith 1987).' Jackrabbits have been reported to use sys- tems of trails to travel about their home ranges (Vorhies and Taylor 1933, French et al. 1965, Rusch 1965). Although trails used by jackrab- bits in Curlew Valley were obvious in the snow and vegetation, I was unable to study trail use because the tracking system could not locate jackrabbits with suflicient accuracy. Home ranges of individual instrumented jackrabbits overlapped extensively. I have no data that suggest individual jackrabbits in- fluenced the home range use by other jack- rabbits from spring through fall, although such intraspecific interactions may have oc- curred. My study suggests that during winter jackrabbits were social and gathered in groups (Smith 1987). Similar winter behavior was reported earlier from Curlew Valley (Rusch 1965) and has also been observed in southern Idaho (personal observation). Jackrabbits appeared to traverse the entire home range in short periods of time. Similar home range use was reported by Lechleitner (1958a), who observed both female and male jackrabbits covering their ranges in about an hour. French et al. (1965) also reported that jackrabbits traversed home ranges in a matter of hours. Factors Governing Home Range Shape and Size Seasonal variation and weather effects. — The greater activity of males compared with females (Table 2) during winter and spring is probably related to reproductive activity. The reproductive season for jackrabbits in Curlew Valley usually begins in January and lasts through May or June (Gross et al. 1974). Black-tailed jackrabbits have a complex mat- ing behavior in which males seek out females (Dunn et al. 1982). Males would thus be ex- pected to be more active than females during the breeding season. Lechleitner (1958a) and Haug (1969) also report greater activity by males during the breeding season. Other researchers have reported seasonal changes in the daily patterns of jackrabbit activity. Donoho (1972) and Costa etal. (1976) reported that jackrabbits were generally less active during winter. The timing of daily ac- tivity observed in this study changed season- ally and was probably related to changes in day length and times of sunset and sunrise. Similar results were reported by Donoho (1972). The evening activity peak I observed is similar to that reported bv Lechleitner (1958b) and Haug (1969) (Fig.' 1). However, Haug (1969) also described a period of height- ened activity just before sunrise. Blackburn (1968) and Knowlton et al. (1968) reported that jackrabbit activity was influ- enced by ambient air temperature. I was un- able to record air temperatures at jackrabbit locations, but m\' finding that jackrabbits were less active during high winds tends to support the idea that temperature influences jackrabbit activity. Lechleitner (1958a) and Tiemeier (1965) also reported decreased ac- ti\ity during high winds and inclement weather. Lechleitner (1958a) also noted that jackrabbits were more active during bright 1990] Black-tailf.d Iackhabiut Home Range 255 moonlit nights, a finding consistent with ni> results. The sizes of home ranges measured by this study appeared to change with the seasons, with spring and fall ranges being slightly smaller. Males used larger areas than fe- males. Tiemeier (1965) and Donoho (1972) found no significant differences in home range size between sexes. Other researchers, how- ever, reported that female jackrabbits used larger areas than males in sunuuer, fall (Lech- leitner 1958a), and winter (Nelson and Wag- ner 1973). Differences in procedures, espe- cially analytical, make comparisons among studies difficult. Many jackrabbits changed home range areas on a seasonal basis, with animals moving to wintering areas in the fall and early winter, and leaving in the spring. Rusch (1965), working in Curlew Valley, and Tiemeier (1965), in Kansas, also reported sea- sonal shifts in home range areas, with animals moving to areas with larger shrubs. Age. — As juvenile jackrabbits mature, one would expect their home ranges to increase in size. It appears that juvenile jackrabbits in- crease the size of their areas of use to roughly that of adults within the first six months post- partum. Young jackrabbits are precocial, are usually weaned by six weeks of age, before the arrival of the next litter, and are independent of their dams at a verv young age (Drake 1969). Population density. — Jackrabbit popula- tions in Curlew Valley undergo changes in density on a 10-year cycle (L. C. Stoddart and F. F. Knowlton, Mathematical model of coyote-jackrabbit demographic interactions, northern Utah. Poster presented at 4th In- ternational Theriological Congress, Edmon- ton, Alberta, Canada, 1985). My study was conducted during a population low, i.e. , < 30 jackrabbits/km" (Smith 1987). 1 do not know whether patterns of jackrabbit home range use change with population density. How- ever, I noted changes in the pattern of use of wintering areas (with fewer wintering areas used during low densities) and habitat types (Smith 1987), suggesting the possibility of other changes in home range use with chang- ing density. Differences between my findings and others reported in the literature may be a function of differing jackrabbit densities. Habitat. — Jackrabbit home ranges in this study were contiguous, and separate resting and feeding areas were not used. This reflects the availability of resources within home ranges of Curlew Valley jackrabbits. Where feeding and resting resources are available in the same area, jackrabbits do not need to travel far from daytime forms to nocturnal feeding sites (Vorhies and Taylor 1933, Nel- son and Wagner 1973). In areas where feeding resources are separated from cover, jackrab- bits have been reported to travel distances > 1 km nightly (Vorhies and Taylor 1933, Hang 1969). Jackrabbits have also been reported to shift feeding sites to feed in agricultural fields (Bronson and Tiemeier 1959). Acknowledgments This study was a part of the Predator Ecology and Behavior Project of the Den- ver Wildlife Research Center of the U.S. Fish and Wildlife Service. The Center transferred to the Animal and Plant Health Inspection Service on 3 March 1986. I thank F. F. Knowlton and L. C. Stoddart for their support and guidance. K. Corts, L. S. Mills, and K. Paulin helped with field research. I thank the many persons, especially E. Hanson and W. Johnson, who helped catch jackrabbits. F. F. Knowlton, J. P. Gionfriddo, and A. P. Wywialowski reviewed the manuscript. F. A. Johnson prepared the figure. Literature Cited AndeR-SON. D J 1982. The home range; a new nonparainet- ric estimation technique. Ecology 63: 103-112. Blackburn, D. F. 1968. Behavior of white-tailed and black-tailed jackrabbits of mideastern Oregon. Unpublished thesis. University of Idaho, Mos- cow. 47 pp. Brand. C. J. R. H.VowLES, AND L. B. Keith. 1975. Snow- shoe hare mortality monitored by telemetry. Jour- nal of Wildlife Management 39; 741-747. Bronson. F. H., and O. T Tiemeier. 19.59. The relation- ship of precipitation and black-tailed jackrabbit populations in Kansas. Ecology 40; 194-198. Burt, W H 1943. Territoriality and home range concepts as applied to mammals. Journal of Mammalogy 24: 346-3.52. Costa, R., K. A Nagy, and V. H. Shoemaker 1976. Ob- servations of behavior on black-tailed jackrabbits in the Mohave Desert. Journal of Mammalogy 57; 399-402. Dixon, K. R., and J A. Chapman 1980. Harmonic mean measure of animal activity areas. Ecology 61: 1040-1044. Di.xoN, K R , O J. Rongstad, AND K M Orhelein 1981. A comparison of home range size in Sylvilagus 256 G. W. Smith [Volume 50 floridamis and S. bachmani. Pages 541-548 in K. Myers and C. D. Maclnnes, eds.. Proceedings of the World Lagomorph Conference, 12-16 August 1979. University of Guelph Press, Guelph, On- tario, Canada. DONOHO, H. S. 1972. Dispersion and dispersal of white- tailed and black-tailed jackrabbits. Pawnee Na- tional Grasslands. Unpublished thesis, Colorado State University, Fort Collins. 83 pp. Drake, E. 1969. Maintenance and social behavior of cap- tured jackrabbits, Lepus californictis Gray. Un- published thesis, Utah State University, Logan. 45 pp. Dunn, J. P . J. A Chapman, and R E Marsh 1982. Jackrabbits, Lepus californicits and allies. Pages 124-145 in J. A. Chapman and G. A. Feldhani- mer, eds.. Wild mammals of North America; biol- ogy, management, and economics. Johns Hopkins University Press, Baltimore, Maryland. Flinders, J. T.. and C. L Elliott 1979. Abiotic charac- teristics of black-tailed jackrabbit forms and a hy- pothesis concerning form function. Encyclia 56: 34-38. French. N R., R McBride, and J Detmer 1965. Fertil- ity and population density of the black-tailed jackrabbit. Journal of Wildlife Management 29: 14-26. Griffiths, R. E . and J Evans 1970. Capturing jackrab- bits by night-lighting. Journal of Wildlife Manage- ment 34: 637-639. Gross, J. E . L C Stoddart, and F W Wagner 1974. Demographic analysis of a northern Utah jack- rabbit population. Wildlife Monographs 40. 68 pp. Haug, J. C 1969. Activity and reproduction of the black- tailed jackrabbit in the coastal cordgrass prairie of Texas. Unpublished thesis, Texas A&M Univer- sity, College Station. 115 pp. Keith, L B , J. R Gary, O. J Rongst.^d, and M C Brit TINGHAM 1984. Demography and ecology of a de- clining snowshoe hare population. Wildlife Mono- graphs 90. 43 pp. Knowlton, F F , P E Martin, and J C Haug 1968. A telemetric monitor for determining animal activ- ity. Journal of Wildlife Management 32: 943-948. Lechleitner, R R 1958a. Movements, density and mor- tality in a black-tailed jackrabbit population. Jour- nal of Wildlife Management 22: 371-384. 1958b. Certain aspects of behavior of the black- tailed jackrabbit. American Midland Naturalist 60: 145-155. Mills, L. S. and F. F. Knowlton. 1989. Observer perfor- mance in known and blind radio-telemetry accu- racy tests. Journal of Wildlife Management 53: 340-342. National Oceanic and Atmospheric Administration 1984. Climatological data: Utah. National Oceanic and Atmospheric Administration, Asheville, North Carolina. Nelson, L.. Jr , and F. H. W.'VGNER 1973. Effects of sub- lethal, cerebral x-irradiation on movement, activ- ity and home-range patterns of black-tailed jack- rabbits. Health Physics 25: 507-514. RUSCH, D 1965. Some movements of black-tailed jackrabbits in northern Utah. Unpublished thesis, Utah State University, Logan 43 pp. Samuel. M D . and E. O. Carton. 1985. Home range: a weighted normal estimate and tests of underlying assumptions. Journal of Wildlife Management 49: 513-519. Samuel. M D , D J Pierce, and E O Carton 1985a. Identifying areas of concentrated use within the home range. Journal of Animal Ecology 54: 711-719. Samuel, M D., D. J. Pierce, E O. Carton, L. J. Nelson, AND K R. Dlxon 1985b. User's manual for pro- gram HOME R.\NGE. University of Idaho Forest, Wildlife and Range Experiment Station Technical Report 15. 69 pp. SAS Institute Inc 1985. SAS/STAT guide for personal computers. 6th ed. Gary, North Carolina. 378 pp. Snhth, G W 1987. Mortality and movement within a black-tailed jackrabbit population. Unpublished dissertation, Utah State University, Logan. 101 pp. SoKAL, R R , and F. J ROHLF 1981. Biometry. 2d ed. W. H. Freeman and Co., San Francisco, Califor- nia 859 pp. Stoddart, L. C 1970. A telemetric method for detecting jackrabbit mortalitv. Journal of Wildlife Manage- ment .34: 501-507. ' Svvthart, R K., and N. A. Slade 1985. Testing for inde- pendence of observations in animal movements. Ecology 66: 1176-1184. TiEMElER. O W 1965. Bionomics. Pages 5-37 in The black-tailed jackrabbit in Kansas. Kansas State University Agricultural Experiment Station Tech- nical Bulletin 140. 75 pp. VORHIES, C. T.. and W. P Taylor 1933. The life histories and ecology of jack rabbits Lepus alleni and Lepus californicus spp. , in relation to grazing in Arizona. University of Arizona Agricultural Experiment Station Technical Bulletin 49: 467-587. Wagner, F. H. 1981. Role of lagomorphs in ecosystems. Pages 668-694 in K. Myers and C. D. Maclnnes, eds.. Proceedings of the World Lagomorph Con- ference, 12-16 August 1979. University of Guelph Press, Guelph, Ontario, Canada. White, G C , and R A C.\rr()TT 1990. Analysis of wild- life radio-tracking data. Academic Press, Inc., San Diego, California. 383 pp. Wywlalowskl a P , AND F F Knowlton 1983. Effects of simulated radio-transmitters on captive black- tailed jackrabbits. Proceedings of the Interna- tional Conference on Wildlife Biotelemetrv 4: 1-11. Recciv (-(11 April 1990 Accepted 11 September 1990 C;rcat Basin Naliiralist 50(3), UM). p|V 25T-2(il HUMPBACK CHUB (GILA CYFHA) IN THE YAMPA AND CREEN RIVERS, DINOSAUR NATIONAL MONUMENT, WITH OBSERVATIONS ON ROUNDTAIL CHUB (G. fiOBL/STA) AND OTHER SYMPATRIC FISHES Catheriiu^ A. Karp' and Harold M. Tyus' Abstract — We e\aluatrd distribution, habitat use, spawning, and species associations of the endangered humpback chub {Gild cyplid ) in the Vaiupa and Green rivers. Dinosaur National Monument, from 1986 to 1989. Adult and juvenile humpback chub were captured in high-gradient reaches of Yampa and Whirlpool canyons where they were rare (n = 133, 230 mm TL; based on capture of the smallest ripe fish, a 232-mm-TL male) were consistently captured in, and apparently se- lected, seasonally flooded shoreline eddies (i.e., formed and maintained by spring runoff). These habitats were dominated by low or negative water velocities and influ- enced by river surges (i.e. , water velocities at any particular point varied in magnitude of up- and downstream currents). Substrate con- sisted mostly of sand and boulders, and water depth averaged 1.3 m at the estimated point of capture. Humpback chub were not collected in riffles and rapids. Eleven of 76 Carlin-tagged humpback chub (x = 312 mm TL, SD = 19) were recaptured one week to two years after initial capture (5 within a year, 6 from one to two years). Ten fish were recaptured in the immediate vicin- ity of their original capture, and one was col- lected about 8 km downstream from its initial capture site. Eight fish (73%, n = 11) were recaptured in breeding condition on at least one occasion. We detected no growth in re- captured fish. About 22% (n = 29) of humpback chub were juveniles (88-228 mm TL). These were most often captured by electrofishing in rocky shoreline runs and small shoreline eddies. One juvenile (122 mm TL) was taken from the stomach of a 61-cm-TL garter snake (Thamnophis species) caught at the conflu- ence of the Yampa and Green rivers. 260 C. A. KarpandH M.Tyus [Volume 50 ROUNDTAIL CHUB. — A total of 1482 round- tail chub were captured in all reaches of DNM except Split Mountain Canyon and the upper 29 km of Lodore Canyon. The fish constituted 37% (n = 256) of the standardized angling and 15% (n = 1016) of the standardized electro- fishing catch. Roundtail chub were at least three times more abundant in Yampa Canyon than in the DNM portion of the Green River (Tables 1, 2) and were most prevalent in the upper 44.8 km of Yampa Canyon (73% of all roundtail chub captures, n = 1085). The fish was incidental in Lodore Canyon (<1%, n = 3). Adults and juveniles were most often captured in eddies, pools, and shoreline runs, but they were also taken in riffles and lower portions of rapids. Species Associations of Humpback Chub Humpback chub were captured in associa- tion with 7 native and 12 nonnative fish spe- cies (numbers of native sculpins and non- native redside shiners not recorded). Species that dominated the standardized catch in- cluded flannelmouth sucker {Catostomits latipinnis), bluehead sucker (C. discobolus), roundtail chub, common carp (Cypriniis car- pio), and channel catfish {Ictalurus punc- tatiis)iTab\esl,2). A total of 350 fish were captured by angling in eddies occupied by humpback chub. Roundtail chub composed about 45%, chan- nel catfish 35%, and humpback chub 15% of this catch. More channel catfish were cap- tured by angling than was any other species (n = 328, 47% of angling catch), and it was the most abundant nonnative fish in eddies that also yielded humpback chub. Other species, including Colorado squawfish (Ptychocheilus luciiis), flannelmouth sucker, common carp, black bullhead {Ameiurus melas), and rain- bow trout (Oncorhijnchus mykiss), composed less than 5% of the angling catch. Electrofish- ing catch was dominated by flannelmouth (n = 2049, 29%) and bluehead (n = 1801, 26%) suckers, and these fishes were common in canyon habitats (Table 1) and open parks (Table 2). The most abundant introduced fishes in DNM were common carp {n 1321) and channel catfish (n = 1153). These species were relatively common in canyon-bound Whitewater reaches and lower-gradient slow- water sections. Standardized C/f data indi- cated both were most abundant in Split Mountain Canyon (Tables 1, 2). During September 1989, flows in Yampa Canyon were reduced to less than 2.83 mVs, and fish habitat was limited to shallow riffles (about 15-cm depth) and deeper pools and runs (about 1-m depth). On September 7 we collected five chubs (four roundtail and one suspected roundtail x humpback chub hy- brid) and seven channel catfish in pools and eddies (about 1 m deep) in Big Joe Rapid (km 38.4). Other chubs, including a suspected humpback chub, were observed about 0.8 km upstream in a 1.1-m-deep pool created by shoreline boulders. No fish were observed or collected in the vicinity of Warm Springs Rapid (km 6.4) on September 14. Spawning of Humpback Chub and Roundtail Chub Thirty-nine humpback chub (16 ripe males, 5 ripe females, and 18 tuberculate but nonripe fish) were captured in shoreline eddy habitats in a 48-km reach (km 20.8-68.8) in Yampa Canyon {n = 37) and in a 2-km reach (km 545.6-547.2) in Whirlpool Canyon (n = 2). Turbidity precluded direct observation of the fish; thus, spawning behavior and microhabi- tat use were not documented. All ripe fish were silvery colored with "gold flecks" on the dorsum. Ripe males always had some orange coloration on the lower side of the head, opercles, abdomen, and paired and anal fin bases. Ripe males and females usually bore light tuberculation on portions of the head, nuchal hump, opercles, and paired fins. This tuberculation was more robust in males. Ripe males averaged 311 mm TL (n = 16, SD = 35, range 232-370 mm) and 229 g (n = 14, SD = 67, range 130-348 g), ripe females aver- aged 300 mm TL (n = 5, SD = 20, range 280-333 nun) and 230 g {n - 4, SD = 75, range 160-336 g), and nonripe tuberculate fish averaged 303 mm TL (n = 18, SD = 35, range 232-382 mm) and 203 g (n = 17, SD = 62, range 92-356 g). Ripe humpback chub were collected following highest spring discharges from mid-May to late June 1987 to 1989 (Table 3, Fig. 2). Captures of nonripe but tuberculate fish also occurred within this 5-6 week period (Table 3). Although sampling in 1986 did not include prerunoff conditions and thus was ex- cluded from Figure 2, tour humpback chub in 1990] Humpback Chub in Dinosaur National Monument 261 Tablk 1. Total catch (N) aiul catcli per unit of clloit oi iislics collcclcd In staiulaiclizcd aiif^ling (AN) and clectrofish- ing(EL), 1987-1989, Yainpa, Lodorc, Whirlpool, and Split Mountain canxons. Dinosaur National Monument. Total eflfort in hoius spent angling (angler hours) and electrofishing. Split Y; mi pa Lodore Whi rlpool Mountain N Ci myon Canyon Canyon (Canyon Species AN EL EL AN EL EL Native species Flannelmouth sucker 2,159 0.30 28.94 24.72 ().()() 27.82 20.92 Bluehead sucker 1,812 0.01 22.35 14.43 0.00 31.96 83.67 Roinidtail chub 1,238 3.25 17.09 0.28 1.02 2.66 0.00 Humpback chub 109 0.65 1.03 0.00 0.23 0.00 0.00 Colorado squawfish 27 0.01 0.30 0.28 0.00 0.53 1.02 Razorback sucker 4 0.00 0.07 0.00 0.00 0.00 0.00 Mountain vvhitefish 2 0.00 0.02 0.00 0.00 0.13 0.00 Introduced species Common carp 1,100 0.24 15.06 10.94 0.23 9.19 25.51 Channel catfish 1,091 4.01 9.64 1.79 2.61 8.26 71.94 Trout' 277 0.01 0.16 22.26 0.00 3.86 1.02 Black bullhead 31 0.11 0.21 0.09 0.68 0.4 0.51 Northern pike 15 0.00 0.18 0.00 0.00 0.13 2.04 White sucker 13 0.00 0.14 0.18 0.00 1.73 0.00 Smallmouth bass 6 0.00 0.11 0.00 0.00 0.00 0.00 Creen simfish 1 0.00 0.02 0.00 0.00 0.00 0.00 Total fish 7,885 653 5,349 795 42 641 405 Total effort'' 76 56 11 9 8 2 ^Includes rainbow, cutthroat, brown, and lake trouts. ^Rounded T.\BLE 2. Total catch (N) and catch per unit of effort of fishes collected by standardized electrofishing (EL), 1987-1989, Island and Echo parks. Dinosaur National Monument. Total effort in hours spent electrofishing. Island Park Species N EL Common carp (Channel catfish Trout' Black bullhead Northern pike White sucker Green sunfish Total fish Total effort'' Introduced species 125 55 16 2 1 1 2 575 19.29 11.9 0.64 0.00 0.00 0.32 0.32 261 3 Echo Park EL Native species Flannelmouth sucker 185 26.37 Bluehead sucker 145 21.54 Roundtail chub 42 3.22 Mountain vvhitefish 1 0.32 26.34 19.95 8.18 0.00 16.62 4.6 3.58 0.51 0.26 0.00 0.26 314 4 ^Includes rainbow, cutthroat, brown, and lake trouts. ''Rounded Table 3. Capture dates of humpback and roundtail chubs in reproductive condition, Yampa and Green riv- ers. Dinosaur National Monument, 1986-1989. Ripe males Ripe females Tuberculate fish'' Humpback chub Jul 5-15 Jul 5 — May 20-Jun 29 May 18-Jun 16 May 18-Jun 22 1986' 1987 1988 1989 Jun 7-28 Jun7 Jun 15 Jun 6-15 May 27-Jun 6 Roundtail chub 1986' Jul 6-29 — — 1987 May 18-Jun 20 May 17-Jun 23 May 17-Jun 29 1988 Jun7-Jul5 Jun 16 Jun 7-29 1989 May 27-Jun 7 Jun 20 May22-Jun20 ''No sampling prior to July .5 in 19H6, tuberculate hsh were not ripe but exhibited secondary sex characters. breeding condition (two of each sex) were col- lected in July of that year. Ripe fish were captured at water temperatures of about 19.5 C (range 14.5-23 C). Roundtail chub in reproductive condition (n = 242: 117 males, 6 females, and 119 tuber- culate but nonripe fish) were darker than 262 C A KarpandH. M Tyus [Volume 50 300 - 1 , 1 ; 1 ' ;i ' 1 ■ - 200 . - pTuJ - 100 - ■ 1 - 0 1,1. : 1 1 ii 1 Fig. 2. Relationship between average distribution hy- drograph and spawning period for humpback and round- tail chubs, Yampa River, 1987-1989. Dashed vertical lines delineate first and last capture of ripe humpback chub; solid vertical lines delineate first and last capture ol ripe roundtail chub; 1986 not included because sampling was initiated late in spring runoff. humpback chub and exhibited more robust tuberculation and more brilHant orange col- oration. Patterns of tubercles and breeding coloration were similar between the two chubs. Ripe male roundtail chub averaged 344 mm TL (n = 117, SD = 24, range 292-419 mm TL) and 329 g (n = 100, SD = 84, range 190-652 g), and ripe females averaged 363 mm TL (n = 6, SD = 15, range 343-380 mm TL) and 363 g (n = 3, SD = 104, range 276-478 g). Nonripe tuberculate fish aver- aged 351 mm TL (n = 119, SD = 29, range 264-447 mm TL) and weighed about 364 g (n = 77, SD = 123, range 140-844 g). Ripe roundtail chub were captured in pools and shoreline runs and eddies during the period of declining spring runoff (Fig. 2). Humpback and roundtail chubs in breeding condition were collected syntopically on 13 occasions. Although this indicated overlap in use of shoreline eddies during spring runoff, ripe females of both species were not syntopic. Discussion Humpback chub and roundtail chub were sympatric in DNM in the reach from upper Yampa Canyon to upper Whirlpool Canyon, although hiunpback chub were rare (<1% of total catch and only 8% of the two Gila species combined). Humpback chub were most prevalent in, and presumably selected, eddy habitats in moderate- to steep-gradient reaches, whereas roundtail chub were ubiqui- tous in parks and most canyons in eddies, riffles, and runs. Both fishes were most abun- dant in Yampa Canyon; neither was captured in Split Mountain Canyon, and the humpback chub was absent and the roundtail chub rare in Lodore Canyon. The paucity of Colorado River chubs in Split Mountain and Lodore Canyon reaches indicates a general decline of G//fl species rel- ative to earlier decades (e.g.. Banks 1964, Vanicek et al. 1970, Holden and Stalnaker 1975a). This may be related to the loss of historic temperature and flow regimes due to regulated flow releases from Flaming Gorge Dam, and to the proliferation of nonnative fishes, particularly channel catfish and com- mon carp. The current rarity of Colorado River chubs in Split Mountain Canyon was also noted by the authors in 10 hours of oppor- tunistic sampling and by the State of Utah during their 1988-89 studies (T. Chart, Utah Division of Wildlife Resources, personal com- munication). Capture of 133 humpback chub, including 39 breeding adults and 29 juveniles, indicates that a reproducing population exists in Yampa Canyon. However, only one ripe fish, a male, was collected in the Green River (i.e.. Whirl- pool Canyon), and it is unknown whether it spawned there or was a stray from the Yampa River. Collection of ripe roundtail chub in canyon reaches yielding ripe humpback chub indicates some temporal and spatial overlap in habitat use during the spawning period, as observed bv others in the upper Colorado River (Kaed'ing et al. 1990). Ripe humpback and roundtail chubs were collected during declining spring flows and increasing river temperatures after highest spring rimoflf. This occurred in May and June in low- (e.g., 1987, 1989) and average- (e.g., 1988) flow years but extended into July in the 1986 high-flow year. No hiunpback chub in breeding condition were captiued during pre- nmotl and late postrunoft periods, and we presume the fish spawned only during the 5-6 week period following highest spring flows. Captiue of only a few ripe female chubs (fixe humpback and six roundtail chubs, 4% of all breeding captures) suggested that females may be ripe for a limited time. Ripe 1990] Humpback Chub IN Dinosaur National Monument 263 humpback chub were captured at teuipera- tures (x - 19.5, range = 14.5 -23 C) that approximate optimum egg incubation couth- tions (i.e., 20 C; Marsh 1985). These tempera- tures are similar to the 14-24 C range noted by Kaeding et al. (1990) but slightly higher than the 11.5-16 C temperatures noted by Valdez and Clemmer (1982), both in the up- per Colorado River. All humpback chub and most roundtail chub in breeding condition were captured in shoreline eddies. Our recapture data indicate that adult humpback chub remain in or near specific eddies for extended periods and that they return to the same eddy during the spawning season in different years (i.e., they exhibit a fidelity to a specific site). Ten of the 11 recaptures were captured in the same eddy as the initial capture (50% in two different spawning seasons), and 73% were captured in breeding condition at least once. We do not know whether these fishes deposited eggs in these eddies or used such habitats only for staging, resting, or feeding. However, we consider the use of such habitats as part of the breeding requirements of humpback chub in the Yampa River. Shoreline eddy habitats in Yampa Canyon were ephemeral (i.e., disap- peared with declining summer flows), and it was obvious that the fish moved elsewhere after the spawning period. Our observations of Gila species in pools near Big Joe Rapid in September 1989 suggest that some fish re- main in nearby deep habitats during low-flow periods. Feeding habits of humpback chub are not well known and were unknown in DNM . Cap- ture of some fish in the interfaces between shoreline eddies and adjacent runs suggests that chubs use these areas for feeding on drift. Stomachs of two humpback chub that died in trammel nets contained hymenopterans and plant debris; and gross examination of fecal material taken from live fish indicates exten- sive use of hymenopterans and other terres- trial insects (e.g., Mormon crickets) as food. We observed humpback chub and other fishes (e.g., roundtail chub, common carp) feeding on Mormon crickets at the water surface in eddies. The high numbers of channel catfish in habitats used by humpback chub and round- tail chub and the gross overlap in foods consumed and in feeding habits (Banks 1964, llolden and Stalnaker 1975a, Tyus and Minckley 1988, Tyus and Nikirk 1990) indi- cate a potential for negative interactions be- tween these fishes. Although the incidence of predation by channel catfish on native fishes is unknown, observations of bitelike abrasions on some chubs collected in DNM suggest channel catfish predation because no other piscivorous fish in that system could have caused such damage. Humpback chub re- mains were found in channel catfish stomachs from the Little Colorado River (W. L. Minck- ley, personal communication), and channel catfish are known to consume fish, fish parts, and eggs in DNM (Tyus and Nikirk 1990). Only a few common carp were captured syn- topically with humpback chub. However, we speculate that their abundance may also have some negative impact on the native fishes, due perhaps to predation on eggs. The humpback chub persists in only a few canyons in the Colorado River basin, and planned water development projects may fur- ther jeopardize its survival. The Yampa River in DNM supports all native fishes known to have occurred there, including the endan- gered humpback chub, Colorado squawfish, and razorback sucker (Xyrauchen texanus). Existing flows of the Yampa River may be singularly responsible for enabling the persis- tence of chubs in the Yampa and Green rivers. Alteration of Yampa River flows could reduce the availability or character of chub spawning habitat and presumably adversely affect their reproduction, aid in further proliferation of introduced competitors and predators, and reduce the quality and quantity of usable habitats. Dinosaur National Monument should be considered a refugium for native fishes, and efforts should be made to protect flows of the Yampa River. Acknowledgments This study was funded in part by U.S. Fish and Wildlife Service, Bureau of Reclamation, National Park Service, and the Northern Col- orado Water Conservancy District. J. Beard, P. Clevenger, and L. Trinca were among several who assisted with data collection. P. B. Marsh, C. O. Minckley, and W. L. Minckley improved an earlier draft of the manuscript. 264 C. A. KarpandH. M.Tyus [Volume 50 Literature Cited Banks. J. L 1964. Fish species distribution in Dinosaur National Monument during 1961-1962. Unpub- lished thesis, Colorado State University, Fort Collins. 96 pp. Douglas, M. E., W. L. Minckley, and H. M. Tyus. 1989. Qualitative characters, identification of Colorado River chubs (Cvprinidae, genus Gila), and "The Art of Seeing Well." Copeia 1989: 653-662. HOLDEN, P. B., AND L. W. Crist 1981. Documentation of changes in the macroinvertebrate and fish popula- tions in the Green River due to inlet modification of Flaming Gorge Dam. BIO/WEST PR-16-5. Logan, Utah. HoLDEN, P. B , AND C B Stalnaker. 1970. Systematic studies of the cyprinid genus Gila, in the Upper Colorado River Basin. Copeia 1970: 109-120. 1975a. Distribution and abundance of mainstream fishes of the Middle and Upper Colorado River Basin, 1967-1973. Transactions of the American Fisheries Society 104: 217-231. 1975b. Distribution of fishes in the Dolores and Yampa River systems of the Upper Colorado Basin. Southwestern Naturalist 19: 403-412. Kaeding, L. R., and M. a. Zimmerman. 1983. Life history of the humpback chub in the Little Colorado and Colorado rivers of the Grand Canyon. Transac- tions of the American Fisheries Society 112: 577-594. Kaeding, L. R., B. D. Burdick, P A. Schrader, and C W. McAda. 1990. Temporal and spatial relations be- tween the spawning of humpback chub and round- tail chub in the upper Colorado River. Trans- actions of the American Fisheries Society 119: 135-144. Leach, L L. 1970. Archaeological investigations at De- luge Shelter in Dinosaur National Monument. Unpublished dissertation. University of Colorado, Boulder. 336 pp. Marsh, P C 1985. Effect of incubation temperature on survival of embryos of nati\'e Colorado River fishes. Southwestern Naturalist 30: 129-140. Miller, W H , D L Archer. H M Tyus, and R M McNatt. 1982. Yampa River fishes study. Final report. U.S. Fish and Wildlife Service and Na- tional Park Service Cooperative Agreement 14- 16-0006-81-931. Salt Lake City, Utah. 107 pp. Nielsen, L. A., and D. L. Johnson 1983. Fisheries tech- niques. American Fisheries Society, Bethesda, Maryland. 468 pp. Seethaler. K. H , C. W McAda, .\nd R. S. Wydowski. 1979. Endangered and threatened fish in the Yampa and Green rivers of Dinosaur National Monument. Pages 605-612 in Proceedings of the First Conference on Scientific Research in the National Parks. Vol. 1. National Park Service Transactions and Proceedings Series 5. Smith, G R , R. R Miller, .\nd W. D. Sable 1979. Spe- cies relationships among fishes of the genus Gila in the upper Colorado River drainage. U.S. National Park Service Transactions Proceedings Series 5; 613-623. Tyus, H M , and C. A Karp 1989. Habitat use and streamflow needs of rare and endangered fishes, Yampa River, Colorado. U.S. Fish and Wildlife Service, Biological Report 89(14). 27 pp. Tyus, H. M., and W L. Minckley 1988. Migrating Mor- mon crickets, Anabrus simplex (Orthoptera: Tetti- goniidae), as food for stream fishes. Great Basin Naturalist 48: 25-30. Tyus, H. M , and N J Nikirk 1990. Abundance, growth, and diet of channel catfish, Icfaltirus punctatus, in the Green and Yampa rivers, Colorado and Utah. Southwestern Naturalist 35: 188-198. Tyus, H M , B D Burdick, R A Valdez, C M H.^ynes, T A. L'iTLE, andC R. Berry 1982. Fishes of the Upper Colorado River basin: distribution, abun- dance and status. Pages 12-70 in W. H. Miller, H. M. Tyus, andC. A. Carlson, eds.. Fishes of the Upper Colorado River system; present and future. American Fisheries Society, Bethesda, Maryland. Valdez, R. A. 1990. The endangered fish of Cataract Canyon. Final report. Bio/West, Inc., Logan, Utali. Valdez, R A , andG H Clemmer 1982. Life history and prospects for recoverv of the humpback chub and bonytail chub. Pages 109-119 in W. H. Miller, H. M. Tyus, andC. A. Carlson, eds.. Fishes of the Upper Colorado River system: present and future. American Fisheries Society, Bethesda, Maryland. Vanicek, C D , R H Kramer, and D R Franklin 1970. Distribution of Green River fishes in Utah and Colorado following closure of Flaming Gorge Dam. Southwestern Naturalist 14: 297-315. Received 3 April 1990 Revised 1 7 August 1990 Accepted 6 September 1990 Great Basin Naturalist 50(3), 1990, pp, 2ri5-272 MANAGEMENT OE ENDANCEKED SONORAN TOPMINNOW ATBYLAS SPRINGS, ARIZONA: DESGRIPTION, GRITIQUE, AND REGOMMENDATIONS Panic, Marsh' and VV, L, Mincklcy' Abstract. — Efforts hctwct'ii 1982 and UM) lia\c lailed to recover and setinc three natural populations of endan- gered Sonoran topniinnow (Pueciliopsis o. ocridoitalis) at B\las Springs, Arizona, Flooding in the Gila River in 1977-78 allowed ingress hy predatory inosciuitofish (Clamhusia affinis), anti topininnows began to decline. Since that time (1) one stock has l)een replaced twice and is again nearly gone because of depredations by mosfjuitohsh that resisted two eradication attempts; (2) topniinnows at a second spring were extirpated through vegetation encroachment after fencing to protect the habitat from li\estock; and (3) a third population was lost to mosquitofish, restocked after the nonnative was removed, and the restocked population is again in jeopardy, or extirpated, since mos(juitofish re- invaded. Recommendations for a more intensive program of recovery are based on reassessments of past efforts and new suggestions for eradication and exclusion of mos(juitofish. The Sonoran (Gila) topminnow, Poeciliop- sis o. occidentalis , is a poeciliid fish endemic to and once widespread in the Gila River basin of Arizona and New Mexico, USA, and the rios Gila, Concepcion, and Sonora basins of Sonora, Mexico (Hubbs and Miller 1941, Minckley 1973, Vrijenhoek et al. 1985). It was listed in 1967 as an endangered species (U.S. Fish and Wildlife Service [USFWS] 1989) because of predation by introduced mosquitofish {Gambusia affinis) and habitat degradation in the Gila River basin (Miller 1961, Mefife et al. 1983, MeflPe 1985). Most efforts to recover the Sonoran topminnow have emphasized reintroductions within its native range (USFWS 1984). This paper deals only with attempts to maintain its natural pop- ulations, which in the USA are now restricted to fewer than 10 sites, all in Arizona. Study Area Three natural topminnow populations oc- cupied a series of small springs adjacent to the Gila River on San Carlos Apache Indian Tribal lands near Bylas, Graham County, Arizona. Two were discovered in 1968 (Johnson and Kobetich 1970) and the third in 1981 (Meffe et al. 1983). No other fishes were present. Al- though collectively known as Bylas Springs, the habitats gained various coined names in the literature. For brevity, we term them S-1 (with west and east sources) through S-III, west to east, and provide a synonymy below. Descriptions and/or mention of the springs have appeared in papers noted above and others by MeflPe (1983a, 1983b), Meffe and Marsh (1983), Hendrickson and Mincklev (1985), Williams et al. (1985), Taylor (1987), and Hershler and Landye (1988). All the springs are small, with base flows of a few liters (S-II = Middle Spring, which dried temporarily in 1990) to a few tens of liters per minute (S-I, with west and east sources, = Bos and Medicine springs, respec- tively; S-Ill = Salt Creek). S-lII has the great- est base flow. S-I and S-II rise through alluvial fill along a stony escarpment; they have essen- tially no surface watersheds other than their immediate surroundings. S-III rises in the channel of an otherwise ephemeral watershed of >50 km" (Burkham 1976a). All are intermit- tent in their lower reaches, originally isolated from the Gila River either by desiccating on or percolating into the terrace that parallels the river's broad floodplain (S-I and S-II), or by falling over an alluvial escarpment >2.5 m high (S-III). Unusually high flooding in the Gila River in winter 1977-78 aflForded an opportunity for mosquitofish to invade S-I, and that spe- cies plus red shiner {Cyprinella hitrensis; recorded only once) also colonized S-III. Sub- sequent declines in topminnow populations 'Center for Environmental Studies, Arizona State University, Tempe. Arizona 85287-1201. Department of Zoology, .Arizona State University, Tempe, Arizona 85287-1501. 265 266 P. C. Marsh AND W L Minckley [Volume 50 were immediate and dramatic in both sys- tems, and various management strategies were planned and implemented in attempts to eradicate the introduced fish and restore topminnows. This paper recollects those ef- forts, provides a status update on topminnows at Bylas Springs, and offers recommendations to perpetuate the stocks. The Sequence of Events Quantitative data are lacking on the abun- dance of topminnows at the times of their discovery in Bylas Springs, but these fish were present in substantial numbers, espe- cially in summer throughout the systems and in winter in and near the springheads. After invasion of mosquitofish, topminnow in S-I between summer 1980 and spring 1982 re- mained in the upper reaches of runs (62-72% of all fishes present) but declined to rare or absent downstream where mosquitofish were numerous. A parallel situation existed in S-III. S-II was never invaded by mo- squitofish, and topminnows remained com- mon until 1988. S-I. — In March 1982 S-I was poisoned with antimycin-A (MeflFe 1983b). No live fish were found three weeks later, and >150 topmin- nows, along with large numbers of indigenous invertebrates salvaged prior to poisoning, were reintroduced in the springheads. By July 1982 substantial populations of both top- minnows and invertebrates were reestab- lished. At the same time, however, mosquito- fish were discovered downstream. Since the Gila River had not again flooded, poisoning must have failed to remove them from the lower part of the system. Meffe (1983b) rec- ommended massive, repeated poisoning com- bined with construction of barriers as a possi- ble method of maintaining the topminnow. Both strategies were embraced in the species' recovery plan (USFWS 1984). Barriers intended to prevent upstream movements of mosquitofish were constructed on the runs of all three springs during winter 1983-84. Each consisted of low (0.7-0.8 m), concrete, V-notch weirs with earthen wings at S-I and S-III extending laterally and up- stream for 3.5-6.6 m. No berms were con- structed lateral to the barrier at S-II. In addi- tion, the western source of S-I and the single source of S-II, plus the areas surrounding each new barrier, were fenced to exclude domestic livestock. S-I was again poisoned with antimycin-A in April and June 1984 and restocked in July with >200 topminnows and uncounted inverte- brates removed prior to poisoning. All native animals soon reestablished in large numbers. However, mosquitofish again survived down- stream, to reinvade above the barrier when it was bypassed by spate-augmented flows in late summer. The introduced species consti- tuted 24% of total fishes caught in December 1985, 69% in September 1986, and 98% by July 1987 (Simons 1987). In January 1989 topminnows (n = 9) in the fenced (west) source of S-I were accompanied by mosquitofish {n = 6), and only mosquitofish were taken downstream; the unfenced (east) source was not sampled. Three topminnows and one mosquitofish were collected from the fenced source in April 1990, while topmin- nows composed 44% of 189 fish taken from the unfenced source. No topminnows were found in the pool above the barrier or in the stream below, where mosquitofish were abundant. In summer 1989, in anticipation of another poisoning (which has not yet occurred), the channel of S-I was realigned to again flow over its barrier, and low (<0.5 m high) gabions of wire-constrained and loose rock were ex- tended for 19-24 m on either side to replace the earthen berms. In summer 1990 notches in all the barriers were reshaped from the original "V to approximately rectangular, doubling their areas to 1500 cm^ to accommo- date larger flows. The upstream base of the barrier at S-I was also sealed with bentonite to prevent seepage under the structure, which had become substantial, and a large saltcedar (Tamarix diincnsis) that threatened damage was removed. S-II. — This spring experienced no dis- charge that passed over the barrier and re- mained secure from mosquitofish; neverthe- less, topminnows disappeared in 1990 due to surface water depletion. Invasion by cattail {Typlia anfi,iistifoIi(i) after livestock exclusion, discussed below, resulted in increased evapo- transpiration and acciunulation ot \egetative debris and entrained sediment that dried the system. S-III. — Topminnows were extirpated from S-III by 1984. A Hood in the Salt Creek water- shed in 1983 incised the lower part of the run 1990] Toi'MiNNow Management 267 to destroy the iornier vvateHall. The intermit- tent, lower part of the spring run now led into the river over a gentle eataract that should have posed no reasonable deterrent to fish passage. Despite this, the reach al)o\'e the artificial barrier remained fishless from the time mos(iuitofish were poisoned in April 1984 until 300 topmiunows fiom S-Il were stocked in 1986 (Brooks 1986). Topmiunows became abundant, and no exotic fish was cap- tured in 1986, 1987, or winter 1988. The run flooded around the barrier sometime in 1988, and mosquitofish again appeared in spring- summer 1989. By April 1990 mosquitofish constituted 76% of all fishes collected above the barrier and were abundant downstream where only a single topmiunow was captured. No topmiunows were observed in S-III in July 1990. In 1989, in anticipation of future renova- tion and as at S-I, lateral gabions of wire- constrained and loose rock were extended 10-11 m on either side of the structure, and the channel of S-III was realigned to again pass over its artificial barrier. Discussion The inability to secure populations of the endangered Sonoran topmiunow at Bylas Springs is disturbing, given that nearly a decade has passed since management efforts began. Moreover, of the 10 known natural stocks of Sonoran topmiunows remaining in the Gila River basin (including S-I and S-II, the latter now extirpated), 6 were sympatric with mosquitofish in 1987 (Simons 1987) and expected to disappear in the near future. There is little doubt that two of the three populations at Bylas Springs would already be gone had mosquitofish not been partially con- trolled, but the facts remain that (1) a topmiu- now stock at S-I, although removed and re- placed twice, is again nearing extinction through depredations by mosquitofish that re- sisted two attempts to remove them; (2) a native stock at S-II has been extirpated through encroachment of vegetation after fencing to protect it from livestock; and (3) one population (S-III) was lost to mosquitofish, necessitating restocking from S-II after the nonnative was removed, and the restocked population is again in jeopardy, or extirpated, since mosquitofish reinvaded. Major habitat changes occurred as a result of barrier construction and fencing in the By- las Springs system. All the runs begin at a relatively low gradient, which is reduced even further as they pass onto the gently slojiing surface of the Cila River floodplain. This pre- cluded construction of high barriers without extensive concrete work or excavation for long, lateral berms. The structures decided upon were placed as far downflow as practica- ble (ca 400 and 200 m downstream from the respective west and east sources in S-I, 60 m in S-II, and 575 m in S-III) to protect maxi- mum lengths of spring runs. All barriers are 700 m or more from the Gila floodplain. None spanned the entire flood channel, however, and those at S-I and S-III failed to direct high discharges over the concrete weirs, since sur- face runoff from one or more precipitation events cut around them. Berm replacement by longer rock gabions in 1989 again failed at S-I, where surface flow bypassed the barrier in July 1990. The barriers also created small ponds on the low-gradient runs. Goncern existed that such ponds might enhance mosquitofish, and the exotic species did, in fact, quickly expand its population as soon as the lentic habitat was achieved. However, because they had already invaded prior to the presence of ponds, the question was moot. The pond habitats were transitory anyway. Sedimentation was exten- sive, and all three were quickly invaded by cattails. By 1990, emergent vegetation was so dense above the barriers on S-I and S-III that open water scarcely existed. Gattail stands had trapped even more sediment, so that by- passing by floodwaters may have been forced in part by mounding of vegetative debris and silt upstream from the weirs. Given sufficient time, accumulations of cattails became so ex- tensive that we are convinced the barriers would have been clogged and breached with- out high water. In January 1989 the barrier pool at S-II held only small pockets of water, 20-30 cm in diameter and scarcely a centimeter deep. Several dozen topminnows had died in these pockets, likely due to combined low tempera- ture and oxygen depletion in the thin layer of water overlying organic sediments. The popu- lation had dwindled to a few individuals, and a single adult male was caught. In April and July 1990 the habitat was only moist; no fish were found despite exhaustive examination. 268 P. C. Marsh andW. L. Minckley [Volume 50 A similar sequence occurred after fencing to exclude cattle from the source of S-II. The headspring and its outflow were rapidly in- vaded by cattails, and by January 1989 the plants had formed a virtual mound of living and dead vegetative materials, with surface water only along the margins. We speculated that the site would be uninhabitable by fishes the next growing season, and by April 1990 no surface water was present and the topminnow population was gone. Succession had pro- ceeded to include invasion of a large bulrush (Scirpus sp.). In July 1990 water was present but fishless, and the cattail stands here and at the downstream barrier pool were dead for unknown reasons. Vegetation in the western source of S-I re- sponded differently to livestock exclusion. The marsh and springhead, although <5.0 m in diameter, were avoided by livestock, pre- sumably due to its dangerously spongy, "quaking," organic deposits overgrown by "cienega" vegetation of small sedges (Eleo- charis sp. ) and grasses. Only its periphery was heavily grazed. Over the years Minckley (un- published data) found three dead cattle mired in the center of this tiny marsh. It remains similar today inside its protective fence, al- though there is a slightly more luxuriant growth of the same low vegetation but no invasion of cattails. Invasion by other than low, especially adapted sedges and grasses may be precluded by water-saturated, reduc- ing hydrosoils, as suggested by Hendrick- son and Minckley (1985). Interestingly, the spongy marsh has now solidified and no longer appears dangerous to livestock. Currently, the fish habitat comprises only a limnocrene, ca 50 cm long, 25 cm wide, and 25 cm deep, and its outflow. The eastern source of S-I was always larger than any other spring but Salt Creek, and remains so. It was not fenced and remains an open, flowing limnocrene I.O m across, 2.0 m long, and 25 cm deep. The combination of intense watering/grazing/trampling by live- stock plus human activities, although un- sightly and outwardly appearing to damage the system, precludes overgrowth by semi- aquatic vegetation. Recommendations Our recommendation is that the U.S. Fish and Wildlife Service make a firm, appropri- ately funded commitment to maintain the Bylas Springs topminnow stocks. Piecemeal efforts to date have largely failed because hy- drologic and vegetational dynamics and com- plexity were either not understood or taken into account. Habitat responses to manage- ment prescriptions were thus unpredicted. A formal plan for recovery must be imple- mented, followed by programmed, event- responsive monitoring for the foreseeable future. Next, it is appropriate to define the degree of isolation of the three Bylas Springs. The presence of endemic hydrobiid snails, mem- bers of a specialized group restricted to springs in the American West (Taylor 1987, Hershler and Landye 1988), indicates consid- erable antiquity of the habitat. These animals are essentially unknown in streams, and their presence in large, erosive rivers like the Gila is even less probable. Their presence in S-III, which rises within a channel that floods on occasion, is unusual. Topminnows could have colonized the springs at any time. Mosquitofish were not locally available to invade Bylas Springs until perhaps the 1930s. Chamberlain (1904) collected none in the Saf- ford area, and the species was found nowhere else in Arizona until 1926, when it appeared in the Colorado River at Yuma and Salt River in Tempe (Miller 1961, Miller and Lowe 1967). Invasion progressed rapidly, and mosquito- fish were abundant statewide at low eleva- tions by the 1940s (Minckley 1973). Topmin- nows were thus protected for about 40 years, until flooding in 1977-78 was sufficient to per- mit ingress by the nonnative species. Recon- struction of the history of these habitats helps understand how and why mosquitofish were originally excluded. The Gila River channel, <90 m in width in the period 1875-1903, was eroded to an aver- age of 600 m in width in 1905-17 (Burkham 1972, Turner 1974). The terrace on which Bylas Springs now occur was not present in the latter period, and the springs were much nearer or could ha\ e flowed directly into the river. During 1905-1906 a cone of coarse allu- vium was washed into the Gila River channel by flash flooding in Salt Creek (the chaimel in which S-III rises). By 1914 the river was being deflected southward and threatening the town of Bylas (U.S. Army Corps of Engineers 1990] TOFMINNOW MaNACEMKNT 269 1914, Olinsteacl 1919). This was accompanied !)>' deposition on the north side of the channel, downstream of the Salt Creek allnvial cone, as docmnented by photographic evidence (Bnrkham 1972) and indirectly by the sizes of mes(inite trees (Prosopis sp.; see Gavin 1973) that conld only have colonized after the ter- race was formed (see also Minckley and Clark 1984). In part because of saltcedar invasion (Burkhani 1976b), the channel again nar- rowed, to average <12() m by 1964-68. The river remained on the south side of its flood- plain; thus Bylas Springs were isolated. Flooding in 1977-78 and again in 1983 contin- ued to erode southward, stimulating major engineering attempts to stabilize and control the channel (personal observation). Since it was unknown whether invasion by mosquitofish in 1977-78 was an isolated event or whether the system was changed enough to assure continued access by exotic fishes, we reexamined it in July 1990. S-I ended, as it did before 1977-78, in a variably wetted sump formed within saltcedar thickets on the ter- race. This sump may exceed 0.5 ha in area and was heavily utilized by livestock. Also, as be- fore, there was no apparent outlet to the Gila River, which lay at least a kilometer farther south and considerably lower in elevation (ca 5.0 m). S-II similarly remained equally as iso- lated in 1990 as it was in the recorded past. As noted before, the channel into which S-III rises passed unimpeded into the Gila River and was thus accessible to mosquitofish dur- ing flood. As long as mosquitofish exist in the sump of S-1, an artificial barrier will be required, but it must be designed to function under all but the most severe conditions. S-Il seems sufficiently isolated without a barrier, assum- ing fish habitat can again be established. S-111 will require a barrier for the foreseeable fu- ture, since mosquitofish will continue to oc- cupy the Gila River. Existing weirs, especially those for S-III, will require further modifica- tions to accommodate high discharges or must be replaced with structures that will do so. The last is difficult because of the gently slop- ing surrounding terrain, which may necessi- tate berms extending tens of meters on each side. Each should be equipped with wide spillways of nonerodable material. We also recommend installation of soil pipe or some other means of passing base flows through all barriers so that upstream ponds and their associated problems will not exist. Once mos(juitofish are eradicated, a barrier should not be needed on S-II. Forty years of protection by natural isolation would seem an acceptable period of time for management of an endangered, short-lived species such as the Sonoran topminnow. We cannot predict the recurrence interval of major floods, but such events, some of which have exceeded 4000 m • sec in a river that averages 14 m^ • sec ' at Saffbrd (Burkham 1970, U.S. Geologi- cal Survey 1989), cannot be engineered against. Such a flood, directed against the north side of the Gila River floodplain, would destroy the system as it now exists. Despite these data and pronouncements, mapping of the entire spring complex is clearly in order for future management refer- ence. The effort should include aerial pho- tography at the time of minimum vegetative development in winter, accompanied by ex- tensive ground truth to confirm intricacies of the aquatic system. The extent of aquatic habitat should be determined under both drought and wet conditions to assure an un- derstanding of actual and potential intercon- nections. Next, mosquitofish must be eradicated, an operation which must be preceded by re- moval of substantial numbers of native fish and other animals from each spring to secure refugia. Fortunately, fish from S-II have al- ready been transplanted to a spring in Roper Lake State Park near Saffbrd, Arizona, where they established a large, reproducing popula- tion (Arizona Game and Fish Department [AZGFD], unpublished data). A stock from S-I is being similarly maintained at Arizona State University. The stock originally inhabit- ing S-IIl is extinct. Invertebrates may present a problem, es- pecially the two endemic hydrobiid snails, Tnjonia gilae and the monotypic Apachecoc- ciis arizonae (Taylor 1987, see also Hershler and Landye 1988), which have previously been held on-site for no longer than a few days and in aquaria for about three months. If these animals are to be held captive for a longer period, special treatment or facilities may be required. Typically, invertebrates may be reintroduced soon after the poison dissipates, generally a few days after the final application. Numbers of all animals retained for restocking 270 P. C. Marsh AND W. L. Minckley [Volume 50 should be large enough to assure maintenance of genetic variability and a reasonable proba- bility for representation of rare alleles (Meffe 1986, 1987, Meffe and Vrijenhoek 1988). Considering the negative results of previ- ous attempts, piscicide should be applied re- peatedly, perhaps three or more times at weekly or longer intervals. Application should be accompanied by physical dewatering of runs, lateral pools, and the downstream sump of S-I, if possible. Footprints of livestock in muds of the sump provide tiny, but effective, life-support sites in which mosquitofish may survive immediately adjacent to toxic water. Temporary exclusion of livestock might well alleviate this problem. Alternatively, water could be retained upstream by temporarily damming the spring run, so the sump could be repeatedly dewatered, at least to a degree, and then flooded to inundate such refugia with poisonous water. A fishless period of at least a year should be required for the entire system before topmin- nows are restocked. Presence of mosquitofish in any of the springs allows children (or adults) to inadvertently or intentionally move them from place to place. Humans are attracted to springs in otherwise arid lands, and the Bylas Springs, although reasonably isolated, are pe- riodically used by local residents for recre- ation. In order to assure long-term success, the area must be inspected frequently and man- aged to assure maintenance of habitat in- tegrity and continued exclusion of mosquito- fish, and to detect and interdict local land uses that may prove detrimental. We recommend a cooperative agreement with the San Carlos Apache Tribe to perform a quarterly or more frequent schedule of surveillance and moni- toring. Biannual, intensive surveys should en- list the assistance of a professional biologist. The entire system is small and may be thor- oughly examined in one day. The activities of domestic livestock, which must have precluded overgrowth by cattail in the past, may be a necessary part of the ecol- ogy of these springs. Encroachment by semi- aquatic vegetation destroys small, isolated habitats, and this must be avoided. The habi- tat disruption and apparent degradation by livestock is preferable to loss of the habitats and topminnows. If reliable, close-order surveillance is developed, the existing fences could be gated and opened periodically to allow removal of vegetation by livestock. If not, we recommend the fences be removed from headsprings and barrier pools alike. Cattail and other emergent plants could also be controlled by cutting, burning, or chemical herbicides; but these methods are labor-intensive and may be more habitat dis- ruptive than livestock, a situation that should be avoided if possible. Experimental manipu- lations to determine reasonable vegetation control measures might be attempted at S-II, which is now fishless. Once tested, ac- ceptable control techniques could be applied to other sites. Conclusions The current and past recovery efforts in behalf of Sonoran topminnows in the Gila River basin (Brooks 1985, 1986, Simons, 1987, Simons et al. 1989) have emphasized introductions and reintroductions within his- toric range far more than dealing with natural populations, resulting in continuing jeopardy to natural populations. In fact, according to the current recovery plan (USFWS 1984), all natural populations could disappear without influencing the down-listing or delisting crite- ria. And, since the criteria have been satisfied (Simons et al. 1989), the species could con- ceivably be down-listed to threatened status. In 1990 the USFWS Desert Fishes Recov- ery Team recommended against such action (Minckley, unpublished data). We do not underestimate the difficulties associated with habitat renovation for main- tenance of native topminnow stocks, but we suggest a concerted effort be directed toward accomplishing that end at Bylas Springs. If topminnows cannot be secured here, which undoubtedly represents some of the least complex habitats occupied by nat- ural populations of topminnows in the USA, it seems highly unlikeh that an\ natural popu- lation threatened by moscjuitofish has much hope ofa persisting. Yet, repeated attempts to eradicate moscjuitofish ha\e failed, and other efforts to manage the habitat ha\ e had unde- sirable results. It is more efficient to devote necessary planning, manpower, and material to initial operations, even if the assurance of succt^ss is costh , than to expend lesser amounts on repeated, unsnccesstiil operations 1990] T( )1'M I NN( )\\ M ANA( ;KM KNT 271 over time, whereby euinulative eosts heeoine exorbitant and the populations are still lost. A new roiuid of effort is clearly needed, and quickly. Acknowledgments Many individuals have been involved in research and management ellorts at Bylas Springs. USFWS personnel have included, among others, J. Brooks (formerly AZGFD), J. Hansen, S. Jacks, D. Parker, B. Robertson (deceased), and S. Stefferud. B. Bagley, D. Hendrickson, and L. Simons, AZGFD, provided data and assistance, as did a number of students and others from Arizona State University; D. Langhorst, G. Meffe, R. Tim- mons, and S. Vives deserve special mention. The San Carlos Apache Indian Tribe permit- ted access through their Wildlife Depart- ment, and tribal wardens physically assisted in many operations. All deserve special thanks. We, of course, take full responsibility for any errors in fact or interpretations. Literature Cited Brooks, J. E. 1985. Factors affecting the success of Gila topminnow introductions on four Arizona national forests. Report to the U.S. Fish and Wildlife Ser- vice, Albuquerque, New Mexico. Arizona Game and Fish Department, Phoenix. 35 pp. Brooks, J. E. 1986. Status of natural and introduced Sono- ran topminnow {Poeciliopsis o. occidcntalis) popu- lations in Arizona through 1985. Report to the U.S. Fish and Wildlife Service, Albuquerque, New Mexico. Arizona Game and Fish Depart- ment, Phoenix. 43 pp. BURKHAM, D E. 1970. Precipitation, stream flow, and major floods at selected sites in the Gila River drainage above Coolidge Dam, Arizona. U.S. Ge- ological Survey Professional Paper 655-B: 1-33. 1972. Ghannel changes of the Gila River in Safford Valley, Arizona, 1848-1970. U.S. Geological Sur- vey Professional Paper 655-G: 1-23. 1976a. Flow from small watersheds adjacent to the study reach of the Gila River Phreatophyte Pro- ject. U.S. Geological Survey Professional Paper 655-1; 1-19. 1976b. Hydraulic effects of changes in bottom- land vegetation on three major floods, Gila River in southeastern Arizona. U.S. Geological Survey Professional Paper 6.55-J: 1-14. Chamberlain, F. W. 1904. Unpublished Arizona field notes. Filed at the Smithsonian Institution, Washington, D.C. 52 pp. (handwritten). Gavin, T. A. 1973. An ecological survey of a mesquite bosque. Unpublished thesis, University of Ari- zona, Tucson. IlKNDRicKsoN, D A , AND W L MiNCKi.KY. 1985. Cicne- gas — vanishing a(|uatic climax cotnnumities of the American Southwest. Desert Plants 6: 131-175. IlKH.siil.KH, R., AND J J Landyk 1988. Arizona llydro- biidae (Prosobranchia: Rissoacea). Smithsonian Contributions in Zoology 459: 1-63. HUBBS.C. L.,andR R Mll.i.EH. 1941. Studies of the hshes of the order Cyprinodoutes. XVII. (ienera and species of the Colorado River system. Occasional Papers of the University of Michigan Museum of Zoology 442: 1-9. Johnson, J. E., and G. Kobetich. 1970. A new locality for the Gila topminnow, Poeciliopsis occidentalis (Poeciliidae). Southwestern Naturalist 14: .368. Meffe, G K. 1983a. Ecology of species replacement in the Sonoran topmiimow {Poeciliopsis occidentalis) and the moscjuitofish (Gamhtisia affinis) in Ari- zona. Unpublished dissertation, Arizona State University, Tempe. 143 pp. 1983b. Attempted chemical renovation of an Arizona springbrook for management of the en- dangered Sonoran topminnow. North American Journal of Fisheries Management 3; 315-321. 1985. Predation and species replacement in Amer- ican southwestern fishes: a case study. South- western Naturalist .30: 17.3-187. 1986. Conservation genetics and the management of endangered fishes. Fisheries 11: 14-23. 1987. Conserving fish genomes: philosophies and practices. Environmental Biology of Fishes 18: 3-9. Meffe, G. K., D. A. Hendrickson, W. L. Minckley and J. N Rinne 1983. Factors resulting in decline of the endangered Sonoran topminnow Poeciliop- sis occidentalis (Atheriniformes: Poeciliidae) in the United States. Biological Conservation 25: 135-159. Meffe, G. K., and P. C. Marsh. 1983. Distribution of aquatic macroinvertebrates in three Sonoran Desert springbrooks. Journal of Arid Environ- ments 6: 363-371. Meffe. G. K., and R. C. Vrijenhoek. 1988. Conservation genetics in the management of desert fishes. Con- servation Biology 2: 157-169. Miller, R. R. 1961. Man and the changing fish fauna of the American Southwest. Papers of the Michi- gan Academy of Science, Arts and Letters 46: 365-404. Miller, R R., and C H. Lowe 1987. Part 2. Fishes of Arizona. Pages 133-151 in C. H. Lowe, ed.. Ver- tebrates of Arizona. 2d printing. Lhiiversity of Ari- zona Press, Tucson. Minckley, W L. 1973. Fishes of Arizona. Arizona Game and Fish Department, Phoenix. 293 pp. Minckley. W L , andT O Clark 1984. Formation and destruction of a Gila River mesquite bosque com- munity. Desert Plants 6: 23-.30. Olm,stead, F. H 1919. Gila River flood control — a report on flood control of the Gila River in Graham County, Arizona. U.S. 65th Congress, 3rd Ses- sion, Senate Document 4.36; 1-94. Simons, L. H. 1987. Status of the Gila topminnow (Poe- ciliopsis occidentalis occidentalis) in the United States. Report to the U.S. Fish and Wildlife Ser- vice, Albuquerque, New Mexico. Arizona Game and Fish Department, Phoenix. 36 pp. 272 P. C. Marsh and W. L. Minckley [Volume 50 Simons, L H , D. A. Hendrickson, and D. Papoulias. 1989. Recovery of the Gila topminnow: a success story? Conservation Biology 3: 1-5. Taylor, D. W. 1987. Fresh-water molluscs from New Mexico and vicinity. Bulletin of the New Mexico Bureau of Mines and Mineral Resources 116: 1-50. Turner, R M 1974. Quantitative and historical evidence for vegetation changes along the upper Gila River, Arizona. U.S. Geological Survey Professional Paper 655-H: 1-20. US Ar.my Corps OF Engineers 1914. San Carlos Irriga- tion Project, Arizona. U.S. 63rd Congress, 2nd Session, House Document 791: 1-168. U.S. Fish and Wildlife Service 1984. Recovery plan for Gila and Yaqui topminnow (Pocciliopsis occi- dentals Baird and Girard). U.S. Fish and Wildlife Service, Region 2, Albuquerque, New Mexico. 67 pp. U.S. Fish and Wildlife Service 1989. Title 50— Wildlife and fisheries. Part 17 — Endangered and threatened wildlife and plants. Subpart B — Lists, endangered and threatened wildlife (CFR 17.11 and 17.12). U.S. Government Printing Office, Washington, D.C. 25 pp. US Geological Survey 1989. Water resources data for Arizona — water year October 1987 to September 1988. U.S. Government Printing Office, Washing- ton, D. C. 358 pp. Vrijenhoek, R C , M E Douglas, and G K Meffe. 1985. Conservation genetics of endangered fish populations in Arizona. Science 229: 400-402. Williams, J E., D B Bowman, J E. Brooks, A. A. Echelle. R J. Edwards, D A. Hendrickson, and J J Landye. 1985. Endangered aquatic ecosystems in North American deserts with a list of vanishing fishes of the region. Journal of the Arizona-Nevada Academy of Science 20: 1-62. Received 26 ]uhj 1990 Accepted 6 September 1990 Great Basin Naturalist .5()(3t. 199(), p|>. 273-281 DAM-SITE SELECTION BY BEAVERS IN AN EASTERN OREGON BASIN William C. McConih', James K. Sfdcll", and Todd 1). IJiitliliol/.' Abstkact. — Wc compared pli\ sical and vegctatix e habitat characteristics at 14 dam sites occupied by beaver {Castor canadensis) with those at 41 random imocciipied reaches to identify leatiires important to dam-site selection in the Long Creek basin. Grant County, Oregon. Stream reaches with dams were shallower and had a lower gradient than imoccupied reaches. Beaver did not build dams at sites with a rock substrate. Bank slopes at occupied reaches were not as steep as those at unoccupied reaches; and occupied stream reaches had greater tree canopy cover, especially of thinleai alder {Aintis fcniiifolia), than did unused reaches. A discriminant model using transformations of bank slope, stream gradient, and hardwood cover classified all beaver dam sites correctly and 35 of 41 random sites as unoccupied sites. The 6 misdassified sites had rock substrates. We also tested four habitat suitability models for beaver in this basin. Three models produced significantly different (P < .05) scores between occupied and random unoccupied reaches, suggesting that they might have some utility for this region. Beaver {Castor canadensis) have long been recognized as having a significant effect on riparian ecosystems. Through alteration of stream flow, they impact soil moisture, biomass distribution, soil redox potential, pH, and plant-available nitrogen in riparian areas (Naiman et al. 1988). Creation of pool habitat is important to some salmonids (Card 1961) and other pool-inhabiting animals, par- ticularly in areas lacking pools formed by nat- urally occurring, coarse woody debris. Pool habitats can be particularly important for some species in arid regions where water lev- els decrease substantially during the summer. As central-place foragers, beaver also create early seral-stage patches that add to habitat complexity and may influence the diversity of terrestrial organisms (Naiman et al. 1988). Beaver management represents a low-cost al- ternative to intensive riparian rehabilitation activities, such as cabling coarse woody debris in streams, but its success depends on the ability of land managers to predict where beaver are likely to build dams and thus create pools. Not all portions of all streams are suit- able beaver habitat. Allen (1983) developed a habitat suitability index (HSI) model for evaluating lacustrine, riverine, and palu- strine habitats for beaver. A similar model was developed by Urich et al. (1984) in Missouri. Howard and Larson (1985) in Massachusetts and Beier and Barrett (1987) in northern Cali- fornia used multivariate techniques to iden- tify habitat features associated with beaver- occupied reaches. Slough and Sadlier (1977) developed a land capability classification sys- tem for beaver in British Columbia based on regression relationships. However, no mod- els have been developed for beaver in arid habitats, and none of the existing models have been tested on independent data from arid habitats. Our objectives were (1) to locate all beaver dams in a third-order basin representative of arid habitat in eastern Oregon, (2) to iden- tify habitat features potentially important to beaver, (3) to develop a habitat classification model for beaver in the basin, and (4) to test four existing habitat classification models. Study Area The Long Creek basin drains approxi- mately 490 km^ of Grant County, Oregon (Fig. 1). Elevations range from 760 to 1900 m. Average annual precipitation is 30-35 cm with most of that occurring in the winter. Temperatures range from about — 10 to +30 C (Franklin and Dyrness 1973). The area is dominated by shrub-steppe vegetation typical of arid eastern Oregon in the Blue Mountains physiographic region (Franklin and Dyrness 1973). Sagebrush {Artemisia spp.) dominates, with junipers (Juniperus spp.) and ponderosa pine {Pinus 'Department of Forest Science. Oregon State University, Corvallis, Oregon 97331 USA. ^USDA Forest Service, Pacific Northwest Research Station, Forestry Sciences Laboratory, .3200 S W Jefferson, Corvallis. Oregon 97.331 USA. 273 274 W. C. MCCOMB ETAL. [Volume 50 JU u b BEAVER DAM SITE u UNOCCUPIED SITE Fig. 1. Location of Long Creek basin. Grant County, Oregon, and distribution of beaver dams (b) and random unoccupied reaches (u) in tlie basin. ponderosa) occurring in the higher eleva- tions. Riparian vegetation is primarily thinleaf alder {Alnus tenuifolia), willow {Salix spp.), hawthorn (Crataegus spp.), and cottonwood (Populus trichocarpa). The dominant land use is grazing, and the land is privately owned except for the portion of the upper basin in the Ochoco National Forest. Methods On 2 September 1988 we examined 98 km of perennial streams in the Long Creek basin from the air at an altitude of 200-300 m. This included 48 km of Long Creek, 21 km of Pass Creek, 1 1 km of Pine Creek, 15 km of Basin Creek, and 3 km of unnamed streams. Thirty sites showing signs of possible beaver activity 1990] Dam-Sitk SKi.EtrnoN by Beavers 275 (ponds, pools, or Idled trees) were marked on a topograpliie map and then visited on the gronnd. Fourteen of the possible heaver sites were aetnully occupied h\ heaver. The others were either natural pools or human-induced disturbances or structures. In September 1988 we recorded habitat characteristics at the occupied sites and at 16 randomly selected imoccupied reaches. Random reaches were selected by drawing random numbers to iden- tify points that corresponded to distances in meters from the mouths of the streams. These reaches happened to be skewed toward the lower basin; so an additional 25 randomly se- lected unoccupied reaches were visited in March 1989 to obtain a better representation of riparian habitat available throughout the basin, resulting in a total of 41 unoccupied reaches. Twenty-two habitat characteristics, includ- ing those used in previous studies, existing models, and some that were potentially im- portant in this basin, were measured at each dam site (n = 14) and each unoccupied reach (n = 41) (Table 1). Stream variables were mea- sured immediately below the dam at occupied sites or at the randomly selected point on unoccupied reaches. Terrestrial habitat was measured at two 40-m-diameter plots per site. Plots were established on both sides of the stream and were immediately adjacent to the dam at occupied sites or to the streambank at unoccupied reaches. Values for the two plots were averaged to characterize each site. Hall (1970) found that 90% of woody food was cut within 30 m of the stream edge, and Johnston and Naiman (in press) reported that most for- aging occurred within approximately 35 m of the stream. Therefore, we assumed that 40- m-diameter plots adequately sampled terres- trial habitat for beaver. Additional variables were measured to characterize dam sites: dam height (cm), pond surface area (m"), average basal diameters of woody stems (by species) cut by beaver, and percentage of available woody stems (by species) that had been cut by beaver. Univariate comparisons were made be- tween occupied and unoccupied reaches with a t test. Linear correlation between all combinations of pairs of variables was con- ducted. For pairs with r > .80, only the vari- able that seemed most biologically meaning- ful to beaver dam building in this basin was 'rABLK I. Vaiialilc's iiicasurccl at 14 hcavi-r dam silt's and 41 unoccupied random stream reaclu-s in the Long Creek basin, Crant County, Oregon, 1988-1989. \'arial)le Method Stream gradient (%) Stream width (m) Stream depth (em) Floodplain width (m) Bank slope (%) Bank type Distances (m) Drainage area (km") Plant cover (%) Hardwood Shrub Total canopy Grazing pressure Average of gradient upstream and downstream from dam or at a random point on unoccupied reaches measiu'ed with a clinometer. Migh-water widtli immediately l)elow dam or random point. High-water depth immediately below dam or random point. Width of area dominated by alluvial soils at the dam or random site. Average of bank entrance angle on both sides of the stream mea- sured with a clinometer. Classified as predominantly dirt or small cobble (<20 cm diame- ter), cobble (>21 cm diameter), or solid rock. Distance to nearest road, liuild- ing, or bridge. Area drained above a dam or random point. Ocular estimates averaged over two 40-m-diameter plots (see text) for grasses and sedges, forbs, thinleaf alder, willow, hawthorn, cottonwood, juniper, and other conifers (mostly pon- derosa pine). The sum of alder, willow, hawthorn, and cottonwood covers. The cover of all stems <1 cm diameter. The sum of hardwood and conifer covers. Classed as low (<25% stems eaten), medium (25-50% stems eaten), high (50-75% stems eaten), or very high (>75% stems eaten). retained for subsequent analysis. Continuous variables were examined for normality using the W statistic (SAS Institute, Inc. 1982: 580). Nonnormal data were subjected to square root or logarithmic transformations to address assumptions behind parametric analysis. Any variables, either raw or transformed, with W < 0.7 (max = 1.0) were excluded from multivariate analyses. Based on these criteria, 10 of the original 20 continuous variables were retained for analysis. The subset of these 276 W. C. MCCOMB ET AL. [Volume 50 Table 2. Average (SE) habitat characteristics measured at beaver dam sites and unoccupied reaches. Long Creek basin. Grant County, Oregon, 1988-1989. Habitat Occupied Unoccupied characteristic Transformation W ('1 - 14) (" = 41) P .2). Although we did not detect any association between grazing and dam-site selection, vegetation re- sponses may have been obscured by historic cutting patterns of beaver, length of pond occupancy, and previous grazing practices (Kindschy 1985, Johnston and Naiman, in press). Many of the preferred food species may have been eliminated from the area prior to this study. 1 labitat Classification Bank slope, stream gradient, and hardwood canopy cover best separated (P < .0001, Pil- lai's trace = 0.62) occupied from unoccupied reaches. The model was: Response Variable = 3.753 - [(VBank slope * 0.272) + (logio Stream gradient * 5.239) - (log),) Hardwood cover * 1.273)]. With zero as a decision level, negative val- ues of the response variable were classified as beaver dam sites, and positive values were classified as random unoccupied reaches. Low values for bank slope and stream gradient and high values for hardwood cover produced negative values. The model correctly classi- fied all dam sites and 35 of the 41 (85%) unoc- cupied reaches. Misclassified unoccupied reaches were dominated by either bedrock or cobble. Therefore, when all sites except those with dirt banks were deleted from the data set prior to running the model, classification was 100%. The accuracy of this model in other drainage basins of this size in eastern Oregon is unknown, but it seems likely that these habitat characteristics would influence beaver dam building elsewhere in the region. Assessment of Existing Models The only model that produced scores that did not differ significantly between occupied and unoccupied reaches was the Massachu- setts model (Table 3). This model was de- signed for use in small watersheds (<750 ha) in the northeastern United States and in- cluded variables that did not pertain to condi- tions in eastern Oregon (soil-drainage class and abandoned-field proximity). The other three models produced scores that differed between occupied and unoccupied reaches (F < .006), suggesting that they can provide an index to beaver habitat quality in this basin. Beier and Barrett (1987) used stream depth (a classificatory variable in their study) and stream gradient to identify beaver-occupied and unoccupied reaches in the Truckee River basin, California. When we assessed these 278 W. C. MCCOMB ETAL. [Volume 50 Table 3. Average (SE) scores for four models tested with data from beaver-occupied and random unoccupied reaches. Long Creek basin. Grant County, Oregon, 1988-1989. Mode Occupied Unoccupied in = 14) (n = 41) 0.56(0.14) 0.52(0.06) 1.44(0.05) 0.. 39(0. 14) 0.67(0.03) 0.55(0.02) 0.69(0.03) 0.54(0.02) 0.39(0.06) 0.20(0.03) 1.46(0.23) 0.49(0.09) 0.50(0.00) 0.43(0.02) 0.79(0.11) 0.29(0.05) 1.46(0.23) 0.49(0.09) 1.00(0.00) 0.78(0.05) 1.51(0.12) 1,34(0.23) p12% could facilitate identification of suit- able dam-building segments along Long Creek and its tributaries. At most sites, gradi- ents >7% are probably only of marginal value (Retzer et al. 1956). However, gradient alone is probabK' not the best indicator of dam-site 1990] Dam-Site Selection by Bea\ eus 279 14 m 7-\ a < DC O b BEAVER DAM SITE u UNOCCUPIED SITE B MEAN VALUE OF USED SITES, BEIER AND BARRETT (1987) U u u uu u u u u u 0.8 2.9 5.0 7.1 STREAM AREA (m ) Fig. 2. Relative stream gradient diagram (stream gradient relative to stream cross-sectional area). Five random unoccupied reaches fell below O.S-m" cross-sectional area. Values from random unoccupied reaches below the diagonal line were classified as unusable beaver habitat because of stream substrate or food availability (see text). suitability. The relationship between gradient and dam building is influenced by the cross- sectional area of the stream because small, high-gradient streams can be dammed (up to a point), but large, high-gradient streams can- not. Similarly, large streams of low gradient can be dammed, but again only up to a point (~5-m" cross-sectional area on Long Creek). Our data support this concept, as does the mean value from active colony sites (B, Fig. 2) in the Truckee River basin (Beier and Barrett 1987). Although stream depth, width, and drainage area above the dam were important features in other studies (Howard and Larson 1985, Beier and Barrett 1987), the degree to which these variables indicate habitat qual- ity for beaver is largely dependent on the length of stream sampled and the location of sampling in the watershed. In first- and second-order streams, these variables must 280 W. C. MCCOMBETAL. [Volume 50 be sufficiently large to provide adequate wa- ter for beaver (Howard and Larson 1985). In large streams, depth and width have a nega- tive association with dam building because the force of the water can prevent dam persis- tence during high flows. Sampling a wide range of stream sizes resulted in a Gaussian distribution of these factors with similar means for occupied and unoccupied reaches (due to the location of beaver dams in the central basin), but the range of values for width and depth is narrower for occupied than for unoccupied reaches. Using relative stream gradient (cross-sectional stream area at a given gradient) overcomes this problem. Substrate type can also be used to further refine selection of potential dam sites. Ap- proximately 63% of Long Creek and its tribu- taries passes through substrates of rock or large cobble that seem to restrict dam con- struction. Slough and Sadlier (1977) reported that beaver in their study area did not use lakes with rocky margins. Bank slope is another physical feature that seems important to dam-site selection. Urich et al. (1984) considered steep banks important to beaver in Missouri, probably because they offer suitable locations for dens along large streams. In our study and that of Beier and Barrett (1987) beaver were associated with gentle bank slopes. The influence of bank slope on habitat suitability may be a locally important variable and should not be univer- sally included in habitat models. An adequate and accessible supply of food and dam-construction materials must be present for establishment of a beaver colony (Slough and Sadlier 1977). On our study area, sites with <7% hardwood tree cover were unlikely to be dam sites (based on a 95% confi- dence interval). Denney (1952) summarized the food preferences of beaver in North Amer- ica and reported that aspen {Populus tremu- loides), willow, cottonwood, and alder were most often selected. The food species present may be less important in determining habitat quality than are physiographic and hydrologic factors (Jenkins 1981, Allen 1983). If food is not adequate, but the geomorphic features already described for dam placement arc met, then the land manager can encourage the growth of food and dam-construction materi- als by restricting grazing of the riparian area, by artificial regeneration of the trees and shrubs, or both. Once a dam is built, forb abundance will probably increase (Table 2), resulting in improved food quantity and qual- ity in the summer (Jenkins 1981). Conclusions For streams similar to those in the Long Creek basin, we suggest that land managers may evaluate the potential for beaver dam establishment using either the Allen (1983) HSI model modified for eastern Oregon con- ditions or the Beier and Barrett (1987) model. The discriminant model that we developed provided excellent classification of the origi- nal data and used habitat features identified by other investigators as important to bea- vers, but it has at least two weaknesses. First, variable transformations obscure direct rela- tionships between beaver and the habitat characteristic (the square root or logarithm of a variable may not be as meaningful as the original value). Second, the model has not been tested on an independent data set. An alternative to using the Allen (1983) or Beier and Barrett (1987) models is to use the following logic-based decision tree. A stream segment may support beaver: (1) if the rela- tive stream gradient falls in the domain below the diagonal line in Figure 2, (2) if the stream substrate is not rock or cobble, and (3) if the hardwood cover is >7%. If hardwood cover is <7%, then the land manager has the option of improving the section of stream habitat by encouraging woody plant growth. To increase the volume of pool habitat in a stream by encouraging beaver, the land manager should identify reaches with adequate geomorphic characteristics, reestablish hardwoods (if nec- essary), and minimize trapping of beaver until the population is well established. For suit- able stream sections, this approach would be more economical than adding logs or similar instream structures that could be better used elsewhere. ACKNOWLEDCiMENTS B. H. Smith (Salmon National Forest, Salmon, Idaho), C. Dahm (Department of Biology, Univcrsit\ of New Mexico, Albu- querque), and K. J. Naiman (Center for Streamside Studies, University of Washing- ton, Seattle) reviewed and improved the 1990] Dam-Sitp: Selection by Beavers 281 manuscript. This research was supported by funds provided through tlie Cooperative Ex- tension Service, College of" Forestry, Oregon State University. This is Paper No. 2622 of the Forest Research Laboratory, Oregon State University. Literature Cited Allen, A. W. 1983. Habitat suitability index models: beaver. Western Energy and Land Use Team, Division of Biological Services, LLS. F"ish and Wildlife Service, Fort Collins, Colorado. 20 pp. Beier, p., and R H Barrett 1987. Beaver habitat use and impact in the Truckee River basin, California. Journal of Wildlife Management 51: 794-799. Denney. R. N. 1952. A summary of North American beaver management. Colorado Fish and Game Department Report 28, Colorado Division of Wildlife, Denver. 14 pp. FiNLEY, W. L. 1937. The beaver — conserver of soil and water. Transactions of the North American Wildlife and Natural Resources Conference 2: 295-297. Franklin, J. F , andC T Dyrness 1973. Natural vegeta- tion of Oregon and Washington. U.S. Department of Agriculture General Technical Report PN W-8, U.S. Forest Service, Portland, Oregon. 417 pp. Card, R. 1961. Effects of beaver on trout in Sagehen Creek, California. Journal of Wildlife Manage- ment 25: 221-242. Hall, J. G. 1970. Willow and aspen in the ecology of beaver in Sagehen Creek, California. Ecologv 41: 484-494. Howard. R J . and J S Larson 1985. A stream habitat classification svstem for beaver. Journal of Wild- life Management 49: 19-25. Jknkin.s. S H. 1981. Problems, progress, and prospects in studies of food selection by beavers. Pages 559-579 in J. A. Chapman and D. Pursley, eds., Worldwide Furbearer Conference Proceedings, Vol. 1. University of Maryland, Frostburg. Johnston, C. A., and R. J. Naiman. In press. Browse se- lection by beaver: effects on riparian forest compo- sition. Canadian Joinnal of Forest Research. KlNDSCHY, R. R. 1985. Response of red willow to beaver use in southeastern Oregon. Journal of Wildlife Management 49: 26-28. Naiman. R J., C. A. John.ston, and J. C. Kelley. 1988. Alteration of North American streams bv beaver. BioScience 38: 753-762. Retzer, J L , H M Swope. J D Remington, and W H Rutherford. 19.56. Suitability of physical factors for beaver management in the Rocky Mountains of Colorado. Colorado Department of Game, Fish and Parks, Technical Bulletin No. 2, Denver. 33 pp. SAS Institute, Inc. 1982. SAS user's guide: basics. SAS Institute, Inc., Cary, North Carolina. 921 pp. Slough. B. G., and R M. F S Sadlier. 1977. A land capability classification system for beaver {Castor canadensis Kuhl.). Canadian Journal of Zoology .55: 1324-1.335. Urich. D L , J P GR.AHAM, and E a Gaskins. 1984. Habitat appraisal of private lands in Missouri. Wildlife Society Bulletin 12: .3.50-,356. Williams. R. M. 1965. Beaver habitat and management. Idaho Wildlife Review 17: 3-7. Received 12 April 1990 Accepted 23 June 1990 Great Basin Naturalist 50(3), 199(). pp. 283-285 SMALL MAMMAL RECORDS FROM DOLPHIN ISLAND, THE GREAT SALT LAKE, AND OTHER LOCALriTES IN THE BONNEVILLE BASIN, UTAH Keiiiit'th L. C^ranier', A. Lee Foote', and Joseph A. Clliapmaii' Collections made during 1985 and 1986 re- sulted in the following notes on reproduction, extensions of geographic ranges, and speci- mens of rare and uncommon small mammals from the Bonneville Basin in northwestern Utah. Collapsible Sherman live traps and Vic- tor snap traps baited with a mixture of rolled oats, peanut butter, chopped raisins, and ba- con fat were used for collections. Exact locali- ties and dates of capture are reported under each species description. Vagrant shrew (Sorex vagrans vagrans). — Three individuals were captured in June 1986 at Twin Springs in Tooele Co., a small spring dominated by saltgrass {Distichlis spicata) ap- proximately 35 km south of Wendover, Utah (T9S, R16W). One specimen was found in an insect pitfall trap. Two additional specimens were caught 21 March 1986 in the Grassy Mountains (T3N, RllW, S26) in a shallow, narrow, dry ravine. The female contained six embryos 8 mm in length. These records ex- tend the known range of this subspecies 35 km to the north (previous record, Durrant 1952, Ibapah, Utah) and substantiate the occur- rence of this subspecies in this area of the Bonneville Basin. Sagebrush vole (Lagurus cur'tatus inter- mediiis). — Two females were recorded from the Grassy Mountains, near the area in which the vagrant shrews were captured. One cap- tured 23 February 1986 was lactating and had four placental scars; the other was captured in September 1985. The latter specimen was prepared and deposited in the Department of Fisheries and Wildlife teaching collection at Utah State University. These records support the general distribution of this subspecies in northwestern Utah postulated by Durrant (1952) and establish the occurrence of the sagebrush vole in this western-central range of the Bonneville Basin. In addition, we feel it noteworthy to mention a siting of a sagebrush vole on the extreme northern Newfoundland Mountains (T6N, R13W, S17), because of the isolated nature of this range, which is surrounded by barren salt flats. The vole, observed one afternoon in June 1985, was clearly identified by its short tail and very light pelage. Little pocket mouse {Perognathus longi- memhris gulosus). — ^Thirteen specimens were collected in May 1986 from the western edge of Floating Island, Tooele Co., Utah (T2N, R16W, S22), approximately 50 km northeast of Wendover, Utah, near the end of Silver Island Mountains. The site had fine sandy soil, and the dominant shrub was desert milk- wort. Poll/gala intermontana. Three speci- mens were also collected from the north end of the Newfoundland Mountains (also re- ported there by H. Egoscue, personal com- munication). These records confirm Durrant's (1952) hypothesized distribution for this sub- species in the western deserts of Utah. Small Mammals of Dolphin Island We trapped for two nights in August 1986 when the Great Salt Lake was at a peak level of 4,212 feet above sea level. The high lake lev- els reduced this island to an area of <25 ha (area calculated based on the 4,210-foot con- tour line). A drop in lake level to 4,200 feet expands the island area to 210 ha, although much of this area is unvegetated mud flats. In 750 trap nights on the island (T9N, RlOW) only Dipodomijs ordii and P. longimembris were captured. This contrasts markedly with Goldman's (1939) and Marshall's (1940) cen- suses of the island 50 years ago. Goldman spent two nights on the island and found a Department ofFisheries and Wildlife, Utah State University. Logan, Utah 84322-5210. 283 284 Notes [Volume 50 much more diverse small mammal fauna in only 37 trap nights. At that time the Great Salt Lake was at a historic low, and the island was connected to the mainland by a low sand- bar. Goldman reported capturing deer mice {Peromijsciis moniculatiis), ground squirrels {Spennophilus toivnsendi), and both Ord's (D. ordii) and chisel-toothed (D. microps) kangaroo rats. Also, he recorded evidence of desert woodrat (Neotoma lepida), coyotes [Canis latrans), and a carcass of a porcupine {Erethizon dorsatum). Goldman (1939) named a new subspecies of chisel-toothed kangaroo rat (D. m. russeohis) and Ord's kangaroo rat (D. o. cineraceus), based on specimens he captured on the island. We saw sign of runways of S. toivnsendi through dense stands of cheatgrass, although we saw no aboveground activity in August when we visited the island. In addition, we saw droppings and weathered nests of Neo- toma, but none of them were recent, suggest- ing that there may be no woodrats left on this island. While no live lagomorphs were ob- served on the island, two weathered, disartic- ulated skeletons of jackrabbits (Lepus sp.) were also found, but these could have been carried there by raptors. The island almost certainly has no Peromyscus remaining. In other west desert areas we would normally catch a minimum of 10-15 deer mice for 750 trap nights of effort even in very poor habitat such as the cheatgrass {Bromus tectoriim) monoculture dominating the island. We caught no specimens of D. microps and be- lieve that the subspecies named for its occur- rence on the island, D. m. russeohis, is ex- tinct. Ord's kangaroo rat {Dipodonu/s ordii mar- shalli). — We captured 11 individuals of this subspecies. Five specimens were deposited in the National Museum of Natural History and another five reside in the University of Utah Museum of Natural History. The specimens of Ord s kangaroo rat do not appear to fit within the range of variation for D. o. cineraceus, the subspecies first de- scribed by Goldman (1939) as endemic to Dol- phin Island. Our specimens are much darker than cineraceus, particularly the tails. In addi- tion, all of our specimens have black facial markings like the mainland subspecies, D. o. niarshalli. Only one specimen from the origi- nal series of cineraceus has these markings (personal communication, Don Wilson, U.S. Biological Survey). However, our specimens are not identical to niarshalli; they are slightly paler and the tails are darker than marshalli. The skulls of all specimens are very similar. These comparisons of our specimens with the original series (collected by Goldman) were confirmed by comparisons to specimens of D. o. marshalli at the University of Utah Museum of Natural History. Thus, we feel that our specimens of D. ordii collected on Dolphin Island are more closely related to the subspecies D. o. marshalli than to the original subspecies D. o. cineraceus described by Goldman (1939). Durrant (1952) earlier ques- tioned the validity of subspecific status for cineraceus, noting frequent connection of Dolphin Island with the mainland and a lack of nearby mainland specimens. Little pocket mouse (Perognathus longi- membris gulosus). — Six specimens of the lit- tle pocket mouse were collected on Dolphin Island. This species has not been recorded previouslv from any island in the Great Salt Lake (Goldman 1939, Marshall 1940, Bowers 1982). Few records are available for this spe- cies in the Bonneville Basin (Durrant 1952, Shippee and Egoscue 1968), the nearest from Kelton, Utah, on the north shore of the lake. Trapping on the nearby mainland at higher elevations (5,500 feet) in the Hogup Moun- tains failed to produce any individuals of this species. This may have been due to the ab- sence of habitats usually preferred by this species. The P. longimembris specimens col- lected on Dolphin Island are much darker overall than gulosus (although still within the range of variation of this subspecies) but ap- pear identical in skull morphology. Speci- mens examined from Dolphin Island are de- posited in the National Museum of Natural History (3) and the Unixersity of Utah Mu- seum of Natural History (3). The complete isolation of the island from the mainland for several \ears probabK' ex- plains these faunal changes. High lake levels have inundated formerly choice dime habitats occupied by the heteromyids that still persist on the island. It is likely that the island fauna has changed repeatedly over the years as a result of lake lexel fluctuations that alternately isolated it from and connected it with the mainhmd. In thi' 19()()s alone, the island has been isolatt^l from and reconnected to the 1990] Notes 285 mainland on at least three separate oeeasions (Gwynn 19(S()). This eoukl aeeount tor the ap- parent reinvasion ot the island 1)\' D. o. nuir- slialli and possible swamping of variation found in the subspecies cineraceus. P'reciuent and periodic invasions and subseciuent isola- tion make Dolphin Island a very dynamic sys- tem whose mammalian fauna could change dramatically as often as lake levels fluctuate with varying precipitation patterns. Major changes include the extinction of a imicjue subspecies (D. in. russeohis) iind the potential creation of a new subspecies of little pocket mouse. Whereas Marshall (1940) recorded seven species of mammals on Dolphin Island when it was connected by a narrow sandbar to the mainland, now apparently only two small mammal species, Dipodomijs ordii and Per- ognathus longimembris, survive, with possi- bly a third species, Spennophilus toivnsendii, surviving as well. Acknowledgments This study was funded in part by U.S. Air Force Department of Defense Contract No. F42650-84-C3559 through the Utah State University Foundation. The Department of Fisheries and Wildlife and the Ecology Cen- ter of Utah State University provided vehicles and otluM- logistical and technical support. The Utah Department of Wildlife Resources granted the necessary collecting permits. Ad- ditional thanks are due to Don Wilson of the U.S. Biological Survey for his comparison of our specimens of D. ordii with the topotypes from Dolphin Island. Literature Cited BowKRS, M. A 1982. In.sulaihiogi'o^raphy ot'inammals in the Great Salt Lake. Great Basin Naturalist 42: .589-596. DliRRANT, S D 1952. Mammals of Utah: taxonomy and distribution. University of Kansas Publications, Lawrence. Goldman, E. A 1939. Nine new mammals from islands in the Great Salt Lake, Utah. Journal of Mammalogy 20: 351-357. GwYNN. J. W. 1980. Great Salt Lake, a. scientific, historical and economic overview. Utah Geological and Mineral Survey, Utah Department of Natural Re- sources, Bulletin 116. Marshall. W H. 1940. A survey of the mammals of the islands in Great Salt Lake, Utah. Journal of Mammalogy 21: 144-159. Shippee, E. a , and H J Egoscue. 1958. Additional mam- mal records from the Bonneville Basin, Utah. Journal of Mammalogy 39: 275-277. Received 20 February 1990 Revised 24 May 1990 Accepted 24 May 1990 Great Basin Naturalist 50(3). 199(), p. 287 TWO PRONGHORN ANTELOPE FOUND LOCKED TOGETHER DURING THE RUT IN WEST CENTIUL UTAH David H. Smith' American pronghorn antelope (Antilocapra americana americana Ord 1818) bucks begin showing signs of entering the rutting period in late August in western Utah (Smith and Beale 1980). Bucks exhibit territorial dominance by marking vegetation with subaricular gland secretions, urine and feces, and by direct in- teractions with other bucks (Kitchen 1974, Yoakum 1978). This territorial behavior as- sures that the larger, healthier males do the majority of the breeding (Kitchen 1974). Of- ten fights between bucks result in injury (Kitchen 1974); on rare occasions these fights have resulted in the deaths of male mule deer (Geist 1981) and white-tailed deer (Marchin- ton and Hirth 1984) when their horns have become locked together. The occurrence of pronghorn bucks locking together as a result of fighting has been documented very few times. Spencer (1942) reported a case in South Park, Colorado, in which the right horn of one buck pierced the underside of the second buck s jaw while its left horn locked behind the second buck's right horn. Yoakum (personal communication, 1988) reported that two bucks with interlocked horns were found dead on the Hart Mountain National Wildlife Refuge in Oregon during the late 1940s. On 17 September 1986 a rancher from Sutherland, Utah, found two dead pronghorn bucks locked together. The pair were found approximately 10 km west of Sand Mountain in Juab County, Utah (T14S, R6W, Sec. 9.). This area is flat saltbrush desert, and domi- nant vegetation species include black grease- wood {Sarcohatus vcrmiculatus) and shad- scale {Atriplex confertifolia). One buck had heart-shaped horns, 39.4 cm in length, with inward curving tips 3.8 cm apart. During the fight this buck evidently thrust its horns up- ward on the underside of the second buck's neck; the horn tips flexed far enough apart to allow the second buck's neck to pass through. The horns were then locked around the sec- ond buck's neck. The second buck s neck was rubbed raw and heavily scabbed, indicating that the two animals may have remained locked together for some time before dying. LiTER.\TURE Cited Geist, V. 1981. Behavior: adaptive strategies in mule deer. Pages 157-223 in O. C. Wallmo, ed.. Mule and black-tailed deer of North America. Univer- sity of Nebraska Press, Lincoln. Kitchen, D. W 1974. Social behavior and ecology of the pronghorn. Wildlife Monograph No. 38. 96 pp. Marchinton. R L , AND D H Hirth. 1984. Behavior. Pages 129-168 in L. K. Halls, ed.. White-tailed deer: ecology and management. Stakepole Books, Harrisburg, Pennsylvania. Smith, A. D., and D M Beale 1980. Antelope in Utah: some research and observations. Utah Division of Wildlife Resources Publication 80-13. 87 pp. Spencer, C. C. 1942. Antelope with locked horns. Journal of Mammalogy 23: 92. Yoakum, J. D. 1978. Pronghorn. Pages 103-121 in J. L. Schmidt and D. L. Gilbert, eds.. Big game of North America. Stackpole Books, Harrisburg, Pennsylvania. Received 7 March 1990 Accepted ISJutie 1990 'Utah Division of Wildlife Resources, 1950 W. 1120 S., Delta, Utah 84624. 287 Creat Basin Naturalist 50(3), 199(). pp. 289-294 MICROBIOLOGY AND WATER CHEMISTRY OF TWO NATURAL SPRINGS IMPACTED BY GRAZING IN SOUTH CENTRAL NEVADA Deborah A. Hall' and Ptimy S. Amy' This study was initiated to monitor the water chemistry and microbial populations at two sites in southern Nevada: Ash Springs and Condor Canyon. Cattle impact was suspected to be a causative factor in increased mortality of two endangered fish species: White River springfish (C renicthys baileyi bailey i) in Ash Springs and Big Spring spindace (Lepidomeda moUispitiis pratensis) in Condor Canyon. Condor Canyon, located at the northern end of the Meadow Valley Wash in south cen- tral Nevada, is approximately four miles long and contains a stream system that runs alter- nately through Bureau of Land Management and private land. Site 1 is furthest down- stream near the bottom of the canyon. Site 2 is within the canyon, and site 3 is near the mouth of the canyon, adjacent to the spring source. Site 4 is closest to another spring source on Delmues's Ranch, a private ranch where cattle currently graze. Cattle are also occasionally present near site 3. Ash Springs, in Pahranagat Valley, is a warm-water spring with temperatures vary- ing from 33 C to 35 C. Between summer of 1986 and spring of 1987 cattle were present at the headpool, and there was a marked de- cline in springfish and other endemic species. Removal of the cattle by fencing (initiated by the BLM in 1987) allowed fish numbers to increase to the same levels as prior to the decline (Taylor et al. 1990). Because of the recovery, this area served somewhat as a con- trol, but residual manure continues to influ- ence the spring whenever precipitation oc- curs. The headpool is still utilized by the public as a "hot tub. Collection trips were planned for both spring sites on a monthly basis beginning in September 1988 and continuing until August 1989. We gathered water samples from four sites in Condor Canvon and two sites at Ash Springs with duplicate samples taken from one site each month on a rotational basis. Weather conditions, water levels, and flow were monitored by ocular estimation. Water and air temperatures were taken using a cali- brated thermometer. Conductivity (Corning, model PS-17), pH (Hanna Instruments, model 0624-00), and dissolved oxygen (Hach) were tested in the field using calibrated equipment. Water was collected in sterile Nalgene bottles by hand-dipping the bottle, rinsing, and refilling without sampler-related contamination. Total bacterial counts were evaluated by dilution and spread plating on R2A agar (Difco) as well as by membrane filtration (Gel- man GN6) during the winter months when counts were low. Plates were incubated at room temperature for five to seven days. Total coliforms were measured using the most probable number method (MPN) (American Public Health Association 1985) and membrane filtration followed by growth on mEndo agar (Difco) at 37 C. Fecal coli- forms were also cultured after membrane fil- tration and support on mFC agar (Difco) at 44 C. Each coliform colony and positive MPN tube was confirmed by inoculation into Bril- liant Green Bile Broth (Difco) tubes, and each was scored positive when gas and acid were produced. Pseudomonas aeruginosa and Aeromonas hydrophila were also evaluated by membrane filtration. Bacteria on the filters were grown on mPA agar (American Public Health Associ- ation 1985) at 41 C for isolation of P. aerugi- nosa. Positive colonies were confirmed by streaking on skim milk agar with clearing of the milk by colonies. MacConkey agar (Difco) was originally used for enumeration of A. hy- drophila, but it did not clearly select for that organism. A comparison of several selective Department of Biological Sciences, University of Nevada, Las Vegas, 4505 Maryland Parkway, Las Vegas, Nevada 89154. 289 290 Notes [Volume 50 WATER TEMPERATURE o o Q- E 40 35 30 + 25 20-- 15-- 10- 5- 0 SITE O O A- ▲ - • 2 A 3 ▲ 4 ■D 5 '■ 6 AMMONIA LEVELS Fig. 1. Water temperatures in Ash Springs and Condor Canyon. Sites 1-4 were in Condor Canyon and sites 5 and 6 at Ash Springs. media for enumeration of A. hydropJiila by Arcos et al. (1988) found mA agar (Rippey and Cabelli 1979) to be the most effective; there- fore, we replaced the MacConkey agar with mA agar beginning with the November sam- pling. Incubation was carried out at 37 C, and suspected positive colonies (denoted by a yel- low color) were inoculated into AH semisoft agar tubes (Kaper et al. 1979). A positive reac- tion was scored in tubes exhibiting alkaline conditions at the top and acid production at the butt of the tube. Confirmed organisms were motile, produced indole, and did not produce hydrogen sulfide. Water temperatures in Condor Canyon ex- hibited a gradient with higher temperatures at sites 3 and 4 near the source of the springs and cooler temperatures within the canyon. Temperatures in the canyon dropped during the winter months and increased again during the summer, while in Ash Springs they re- mained stable throughout the year (Fig. 1). The ammonia levels in Condor Canyon fluctuated throughout the year, but a general increase was observed during December and January, along with a less significant increase in June (Fig. 2). Sites 5 and 6 at Ash Springs exhibited a similar pattern during these months. Fig. 2. Ammonia levels in parts per billion. Top frame depicts Condor Canyon, bottom frame Ash Springs. 1990] Notes 291 When the nitrate phis nitrite levels were graphed together, a pattern similar to the NH3 data was seen, bnt there was a lag time of nearly a month in the peaks for both C'ondor Canyon and Ash Springs (Fig. 3). There was a nitrate phis nitrite gradient in the Condor Canyon sites, with site 4 at the top of the eanyon being highest and site 1 at the bottom lowest. A large peak in organic phosphorus (OP) levels during March correlated with NH3 peaks observed during that month. Ash Springs and Condor Canyon both exhibited this phenomenon (Fig. 4). Total viable counts in Condor Canyon were dramatically similar to OP levels during March (Fig. 5). The peak at site 2 in January was notable. At Ash Springs total counts varied little. A slight peak was observed in March, but this may not be statistically valid. Pseiidomonas aeruginosa, an opportunistic fish pathogen, was found on a regular basis only at site 5 in Ash Springs, with rare colonies appearing at site 6. A pattern similar to that exhibited in the total viable counts can be seen, with a dramatic peak in March and also with higher numbers during the summer and early fall (Fig. 6). NITRATE PLUS NITRITE LEVELS 300- ■ 225 150- \x\.^-.^,JA 75' 0- 0 N D J SITE — n 5 Fig. 3. Nitrate plus nitrite levels in parts per billion. Top frame depicts Condor Canyon, bottom frame Ash Springs. ORGANIC PHOSPHORUS LEVELS Fig. 4. Organic phosphorus levels in parts per billion. Sites 1-4 were in Condor Canyon, sites 5 and 6 at Ash Springs. 292 Notes [Volume 50 TOTAL VIABLE COUNT 15.000 12.500-- o o CD Q_ o X Z) o 2.500 SITE o- A- A- A 3 -A ■D S 0 N D J E M MONTH Fig. 5. Total viable counts in colony-forming units per 100 ml. Sites 1-4 were in Condor Canyon, sites 5 and 6 at Ash Springs. total viable counts (Fie. 5). The peak at site 2 Pseudomonas aeruginosa Counts i i in January appeared also. The increase of ammonia levels in Condor Canyon during December, January, and June correlated with precipitation and cattle pres- ence in the canyon. Rain and snow were abun- dant in December and January, with the addi- tional summer showers typical of the Mojave Desert occurring in June. Cattle were present at site 3 in December, and the streambanks were trampled in January. When cattle were present, they deposited NH3 in the form of urine and feces, and runoflFduring winter rain- storms washed additional nitrogen into the system from residual feces. Because nitrogen is often a limiting nutrient in ecosystems, monitoring nitrogen-containing chemical spe- cies is important. The peaks in nitrogen spe- cies at sites 5 and 6 (Ash Springs) in December and June were probably due to precipitation rather than cattle presence (Fig. 2). A peak was also noted there dining Nhirch. (It was raining during this sampling trip.) The lag time observed between the higher ammonia and nitrite plus nitrate levels was most likely due to the o.xidation of ammonia through nitrite to nitrate (Fig. 3). This most often occurs as a metabolic process of nitrify- ing bacteria, and this conversion can become slower in cool temperatures and when low numbers of nitrifying bacteria are present. Fig. 6. Pseudomonas aeruginosa counts in colony- forming units per 100 ml at site 5 in Ash Springs. Aeromonas hydrophila, another oppor- tunistic fish pathogen, was found at all sites. Colony-forming units (CFU) in Ash Springs followed the pattern seen with Pseudomonas aeruginosa levels, showing peaks in March and July (Fig. 7). In Condor Canyon a slight peak was also observed in March, with in- creased numbers during the summer months when total microbial counts were also higher. The shape of the graph of total coliform levels in Condor Canyon (Fig. 8) is similar to the graph of total viable bacterial CFU (Fig. 5). Fecal coliforms were elevated in March, and Figure 9 markedly resembles the graph of 1990] Notes 293 AEROMONAS HYDROPHILA COUNTS TOTAL COLIFORMS D- ■ - -D 5 -■ 6 ■ ; \ ^Ji= — ■- \ NDJFMAMJJA MONTH Fig. 7. Aeromonas hijclrophila counts in colony-form- ing units per 100 ml. Top frame depicts Condor Canyon, and bottom frame depicts Ash Springs. Fig. 8. Total coliform counts in colony-forming units per 100 ml. Top frame depicts Condor Canyon, and bot- tom frame depicts Ash Springs. The elevated level of total viable bacteria during March was probably due to the availability of OP and other nutrients washed in by the rain from both feces and any fertiliz- ers used by the nearby rancher (Figs. 5, 6). At other times bacterial growth was probably limited by lack of phosphate. The higher counts in summer and early fall are likely due to the warmer water temperatures in Condor Canyon. At Ash Springs, where water tem- perature remained stable, total counts varied little. Pseudomonas aeruginosa levels fol- lowed a similar pattern although they were found only in Ash Springs. Figure 7 depicts Aeromonas hijclrophila data beginning with November and extending through August only because of the change in the isolation medium explained above. On the total coliform graph (Fig. 8) the peaks from February through March in Con- dor Canyon again probably reflect precipita- tion. The significantly higher numbers at site 4 during the warmer months correlate with the time periods when cattle were most prevalent and water was slowest and warmest. At Ash Springs the numbers of coliforms re- flected precipitation rather than cattle pres- ence, as exhibited by the peaks in March and July when precipitation occurred. There had been no cattle access to Ash Springs for six months prior to the beginning of this study, but the effect was still seen during times of sufficient precipitation. Increased fecal coli- form numbers were similarly observed during the rainy season and when water temperature was warmest in Condor Canyon (Fig. 9). As expected prior to this study, bacterial levels were influenced by water tempera- tures, with higher counts correlating with warmer water. In Condor Canyon increased numbers reflected these changes during the warmer months, while in Ash Springs they remained fairly stable throughout the year. Bacterial levels also reflected increased pre- cipitation and cattle presence because of the influx of nutrients necessary for growth of microorganisms. Influence of cattle could be seen months after their physical presence when precipitation allowed an influx of nitro- gen and phosphorus. 294 Notes [Volume 50 FECAL COLIFORMS D- ■ - SITE -D 5 — ■ 6 ■- D \ a s*- =»= =«= =9- ^ — ■ — ^ — ■ — 1 F M MONTH Fig. 9. Fecal coliforni counts in colony-forming units per 100 ml. Top frame depicts Condor Canyon, and bot- tom frame depicts Ash Springs. Acknowledgments We greatly appreciate the help of Hermi D. Hiatt in sampling and microbiological test- ing. Thanks also go to Dr. John Unrue for the purchase of a piece of equipment critical to this study. This research was funded by the U.S. Bureau of Land Management under Contract UV053-R09-11 to the University of Nevada, Las Vegas. Literature Cited Arcos, M L. a. de Vicente, MA. Morinigo, P Romero, AND J. J. BORREGO. 1988. Evaluation of several selective media for recovery of Aeromonas hij- drophila from polluted waters. Applied and Envi- ronmental Microbiology 54: 2786-2792. American Public Health Associ.\tion 1985. Standard methods for the examination of water and waste water. 16th ed. American Public Health Associa- tion, Washington, D.C. Kaper. J , R J Seidler. H Lockman, andR R Colvvell. 1979. Medium for the presumptive identification of Aero77W7}as htjdrophila and Enterobacteri- aceae. Applied and Environmental Microbiology 38: 1023-1026. RiPPEY, S R , and V J Cabelll 1979. Membrane filter procedure for enumeration of Aeroiijonas hij- drophila in fresh waters. Applied and Environ- mental Microbiology 38: 108-113. Taylor, F R, L P Gillman, and J W Pedretti 1989. Impact of cattle on two isolated fish populations in Pahranagat Vallev, Nevada. Great Basin Natural- ist 49: 491-495. Received 5 February 1990 Revised 22 August 1990 Accepted 9 September 1990 Great Basin Naturalist 50(3), 199<). pp 295-298 BIRDS OF A SHADSCALE (ARTRIPLEX CONFERTIFOLIA HABITAT IN EAST CENTRAL NEVADA Dean E. Mcdin' Despite widespread distribution of shad- scale (Atriph'x confcr-tifolia) habitat in the Great Basin Desert (Fowler and Koch 1982), it has been largely ignored by avian ecologists. There are few quantitative assessments of breeding bird populations in these vast areas used primarily for livestock grazing (but see Fautin 1946 for western Utah, Smith et al. 1984 for southwestern Idaho). This informa- tion is basic to understanding the ecology of desert birds and the stewardship of their habi- tats. In this paper I describe breeding bird densities of a shadscale community in the Snake Valley of east central Nevada and com- pare them with other quantitative studies from shadscale habitats. Study Area The study area is located 4 km north of Baker in southeastern White Pine County, Nevada, at a median elevation of approxi- mately 1600 m. The study area is a flat valley bottom bounded by foothills and mountains; there are no seeps, springs, or live streams on the site, although dry washes cross the valley floor. Climatically, the area is a cold desert with cold winters and hot, dry summers. Max- imum temperatures in summer frequently ex- ceed 35 C, and minimum temperatures in winter often drop to — 29 C (Houghton et al. 1975). Annual precipitation ranges from 10 to 20 cm (Houghton et al. 1975). The area is grazed lightly by cattle trailing to and from spring-fall ranges (R. Jenson, personal com- munication). Vegetation in the study area comprises a mixture of low shrubs with a sparse herbaceous component. Dominant shrubs are shadscale, green molly (Kochia americana), common winterfat (Eiirotia lanata), bud sagebrush (Arteinisia spinescens), and spiny hopsage (Grayia spinosa). Fourwing saltbush (Atriplex canescens), black greasewood (Sar- cobatus venniculatus), and rubber rabbit- brush {Chrysothamnus nauseosus) occasion- ally occur along shallow washes. Three perennial grasses, Indian ricegrass (Oryzopsis hymenoides), galleta {Hilaria jainesii), and squirreltail (SitanUm hystrix), occur through- out the site. Cheatgrass (Bromus tectorum), an annual, is a frequent associate. Plant names follow Holmgren and Reveal (1966). Methods A 20-ha plot was censused for breeding birds using the spot-map method (Interna- tional Bird Census Committee 1970). A cen- sus plot, chosen as the best representative of the shadscale community, was selected by ex- amining the vegetation and topography of the general area. A square plot was surveyed and gridded with points numbered and marked with stakes at 75-m intervals. Ten census vis- its to the plot were made annually from 29 March to 1 June from 1981 to 1983. Most spot mapping was done from sunrise to early after- noon when birds were most active. Different census routes through the plot were used, with different starting and ending points dis- tributed as evenly as practicable among the visits. To ensure complete coverage, the plot was censused by walking within 50 m of all points on the grid. Recorded bird observa- tions extended a minimum of 75 m beyond plot boundaries. At the end of the sampling period, clusters of observations and coded activity patterns on species maps were circled, indicating areas of activity or approximate territories. Fractional parts of boundary territories were determined by estimating the portion of each edge cluster that fell within the study plot. Oelke (1981) Intermountain Research Station, Forest Service, U.S. Department of Agriculture, Boise, Idaho 83702. 295 296 Notes [Volume 50 Table 1. Passerine breeding bird densities (individuals/ha) in shadscale vegetation, east central Nevada, 1981-1983. Foraging category" Nesti ng Breedi ng bird densit\' Species sulistr; ite'' 1981 1982 1983 Horned Lark GGO G 1.28 1.52 1.32 (Eremophila alpestris) Brewer's Sparrow GGI B 0.08 0.10 0.08 (Spizella breweri) Sage Thrasher GGI B 0.02 +" 0.05 {Oreoscoptes montanus) Total individuals/ha 1.38 1.62 1.45 Bioniass (g/ha; )'' 42 49 44 Species richness (n) 3 2 3 ^After DeGraaf et al. (1985): GGO -= ground gleaning oninivore, GGI = ground gleaning msectivor ''After Harrison (1979): G = ground nester, B - bush nester. '^+ indicates the species was observed infrequently (less than three registrations). ■'Species weights from Dunning (1984). and Verner (1985) summarized methodologi- cal and other special problems of the mapping method. Total bird biomass was calculated annually by summing the products of breeding bird species densities and average bird species body weights (Dunning 1984). Bird nomen- clature is from the 1983 AOU check-list (American Ornithologists' Union 1983). Results and Discussion Three passerine bird species bred on the study site (Table 1). By far the most common breeder was the Horned Lark {Eremophila alpestris). A permanent resident, this broadly distributed bird occurred throughout the study plot. Less common, and in more re- stricted locations, were two summer resi- dents, the Brewer's Sparrow (Spizella brew- eri) and the Sage Thrasher {Oreoscoptes montanus). Other species, observed as occasional visi- tors on or over the study plot during the breeding season, included Northern Harrier {Circus cyaneus). Red-tailed Hawk {Buteoja- maicensis). Ferruginous Hawk {Buteo re- galis). Golden Eagle {Aquila chrysaetos), American Kestrel (Falco sparverius), Prairie Falcon {Falco mexicanus). Mourning Dove {Zenaida macroura). Burrowing Owl {Athene cunicularia). Short-eared Owl {Asio flam- tneus), Violet-green Swallow {Tachycineta thalassina). Cliff Swallow {Hirundo pyrrho- nota). Barn Swallow (Htr«n- O 1— 4000 • LT) Z UJ 2000 Q 200 300 400 500 600 700 800 900 1000 SCRAPERS OJ E >- in z LjJ Q SHREDDERS UJ 100 200 300 400 500 700 800 900 1000 PREDATORS 200 300 DISTANC 400 500 600 700 800 E from OUTLET (m) 200 300 400 500 600 700 800 900 1000 DISTANCE from OUTLET (m) Fig. 2. Macroinvertebrate density by functional feeding group in three lake outlet streams in June 1986. Open circles = Pettit Lake, closed circles = Stanley Lake, and open triangles = Yellowbelly Lake. Bars represent ± 1 standard deviation. to determine the amount of hentliic organie matter (AFDM). The sample was dried at 60 C, weighed, ashed at 550 C, rehydrated, redried at 60 C, and reweighed. Results Community Analysis Macroinvertebrate density and hiomass increased rapidly immediately below the out- lets and then plateaued or, as in the case of Pettit and Stanley, decreased before stabiliz- ing (Fig. 1). Total density and biomass at Stan- ley and Yellowbelly Lake outlets plateaued within 40 m. Yellowbelly Lake outlet had den- sities twice those of Pettit and Stanley Lake outlets, although biomass was similar among sites. This was probabK in response to great^f food availabilit}' as reflected in differences in organic matter standing crops between the two locations (Fig. 1). Macroinvertebrate density in Pettit Lake outlet was lower than that in the two other outlets at 80 m, but showed a relativcK' rapid increase to levels exceeding those of Stanley Lake outlet at transect 7. 1990] Community Development Below Lake Oitlets 307 MINERS 100 200 300 400 500 600 700 800 900 1000 SHREDDERS 100 200 300 400 500 600 700 BOO 900 1000 0.800 ■ ^^ 0.700- FILTERERS £ 0.600- Q-, 0.500 BIOMASS o o o o o o o o o o o o o o o . o 5 ' — 1 -H iP- 1 1 1 -T-? PREDATORS 200 300 400 500 600 700 800 900 1000 DISTANCE from OUTLET (m) 100 200 300 400 500 600 700 800 900 1000 DISTANCE from OUTLET (m) Fig. 3. Macroin\ertebrate bioinass b\ functional feeding group in three lake outlet streams in June 1986. Open circles = Pettit Lake, closed circles = Stanley Lake, and open triangles = Yellowbelly Lake. Bars represent ± 1 standard deviation. Species richness increased immediately downstream from each lake outlet (Fig. 1). Pettit and Stanley Lake outlets showed slight declines in richness 20-80 m downstream, although there was a tendency, best seen at Pettit, to progressively add species with in- creasing distance from the lake. Functional Feeding Group Analysis Stanley Lake outlet. — The density and biomass of gatherers, scrapers, filterers, and predators each showed patterns comparable to that of total density and biomass (Figs. 2, 3). An exception was the extended high abun- dance of predators at 40-160 m. Shredder density downstream of 160 m showed a resur- gence to high values observed at 20 m rather than a maintenance of values comparable to those found at 80 m as occurred for total density. Miners did not show the marked peak at 20 m seen for total numbers and for other functional feeding groups. Miners, such as chironomids, have been found to be abun- dant in lentic sediments, which may explain their lack of response immediately below lake outlets. Pettit Lake outlet. — Gatherer, scraper, and miner densitv and biomass all showed 308 C. T. Robinson and G. W. Minshall [Volume 50 oPETTIT 6/86 • PETTIT8/86 200 300 400 500 600 700 800 900 1000 0 10O 200 300 4-00 500 600 700 800 900 1000 0 100 200 300 400 500 600 700 800 900 1000 DISTANCE from OUTLET (m) Fig. 4. Macroinvertebrate species richness, density, and biomass in Pettit Creek in June and August 1986. Bars represent ± 1 standard de\iation. patterns similar to those of total density and biomass. Filterer density and biomass peaked at 40 and 160 m (Figs. 2, 3). Predators showed an accentuated recovery in numbers at 160 m and continued high levels at 400 m in contrast to the pattern for total numbers. Predator biomass followed the pattern oliserved for filterer biomass (Fig. 3). Yellowbelly Lake outlet. — The density and biomass of gatherers, filterers, and preda- tors showed patterns similar to those of total density and biomass. However, the jiredator biomass deviated irom the general trend 1)\ decreasing downstream of the 8()-m transect (Figs. 2, 3). Filterer density and biomass peaked shortly below the outlet as was found at Pettit Lake outlet. The high filterer density and biomass were at a single location (80 m) rather than over an extended stretch (40-160 m) as at Pettit. Greater current velocity and substrate size may have facilitated coloniza- tion by filterers at 80 m at Yellowbellv Lake outlet (Table 1). In general, the density and biomass of shredders followed the pattern seen for ben- thic organic matter at all three lakes. This was the "expected pattern for all functional groups, based on the assumption of a lake outlet stream gradually accruing food down- stream from allochthonous sources. Devia- tions from this pattern, especially by filter feeders, suggest "contamination" of the water by lake plankton. This was least evident at Stanley Lake outlet and most pronounced at Pettit Lake outlet. However, even at Pettit Lake outlet filter feeder populations declined rapidly within 160-400 m, indicating deple- tion ot this material (Figs. 2, 3). Scraper den- sity and biomass suggest that, for the most part, autochthonous sources of food were low, as would be expected for the headwater streams we were attempting to simulate. Yellowbelly Lake outlet at the 20-m transect is a notable exception. Although there were some minor deviations, data for density and biomass of functional feeding groups showed similar patterns (Figs. 2, 3). Seasonal Study of Pettit Lake Outlet Longitudinal patterns of total density, biomass, and species richness were somewhat different in August from those found in June (Fig. 4). Animal density, biomass, and species richness peaked sooner in August than in June and were not significantly different down- stream of the 400 m transect. Total density and biomass increased downstream to 160 m, declined markedly for the next 240 m, and then stabilized in August (Fig. 3). The peak in al)undance 40—200 m downstream of the lake outlets in August suggests greater production occurring at this time of year, possibK due to increases in stream temperature, solar radia- tion, and leutic inputs. Longitudinal distributions of all functional feeding groups except filterers and scrapers differed in August from those in June (Figs. 5, 6). Filterer density and biomass peaked early 1990] Community Dk\ elopmknt Bklow Lakk Outlets 309 z LjJ Q MINERS 700 800 900 1000 SHREDDERS >- (^ en 2 LJ Q 100 200 300 400 500 600 700 800 900 1000 100 200 300 400 500 600 700 BOO 900 1000 PREDATORS 200 300 400 500 600 700 800 900 1000 DISTANCE from OUTLET (m) 100 200 300 400 500 600 700 800 900 1000 DISTANCE from OUTLET (m) Fig. 5. Macroinvertebrate density by functional feeding group in Pettit Lake outlet in June and August 1986. Open circles = June, closed circles = August. Bars represent ± 1 standard deviation. and then virtually disappeared from tlie com- munity downstream for both sampling dates. Gatherers, miners, and shredders increased in abundance (density and biomass) down- stream of the outlet in June, whereas gather- ers, miners, and shredders had high densities through 160 m and then decreased to low values at 400 and 900 m in August. Gatherer, miner, and shredder biomass was similar among transects in August. Scrapers peaked in biomass at 80 m in June but displayed simi- lar biomass values among transects in August (Fig. 6). The main difference in predator abundance between the two dates was the reduced peak at 40 m and the decrease at 80 m in June that was absent in August (Figs. 5, 6). Discussion The results support our hypothesis of a gradually developing stream community (greater numbers/m" and taxonomic complex- ity) with progressive distance downstream of a lake outlet. The distance required for the development of full community potential (i.e., the recovery distance following com- plete interception of incoming drift) could not be determined precisely and seems to vary 310 C. T. Robinson and G. W. Minshall [Volume 50 MINERS 100 200 JOG 400 500 600 700 800 900 1000 100 200 300 400 500 600 700 800 900 1000 SHREDDERS 100 200 300 400 500 600 700 800 900 1000 PREDATORS 100 200 300 400 500 600 700 800 900 1000 DISTANCE from OUTLET (m) 100 200 300 400 500 600 700 800 900 1000 DISTANCE from OUTLET (m) Fig. 6. Macroinvertebrate l)ioinass by functional feedinsj; group in Pettit Lake outlet in Jiuie and August 1986. Open circle.s = June, closed circles = August. Bars represent ± 1 standard deviation. widely depending on the particular stream and time of year. In June, during a period of relatively high discharge, "recovery," mea- sured in terms of species richness and total density, ranged from 20 m at Stanley Lake to over 900 m at Pettit Lake. During near base flow conditions in August, community devel- opment in Pettit Lake seemed to be much more rapid than in Jime, peaking somewhere between 160 and 400 m. These data suggest that community development is impeded im- der conditions of high flow. Additional mea- surements should be made in several outlet streams having unaltered flows and channels so that the full distances required for recovery during each season and the factors responsible for the different rates of community develop- ment among streams can be established. Our results also show a restricted distribu- tion by filter feeders. The decline from peak numbers below the outlet was more rapid than reported by Sheldon and Oswood (1977), thus supporting our prediction that oligotrophic lakes will show more limited sup- plies of seston and consecjuently a more re- stricted distribution of filter feeders in their outlet streams than meso- or eutrophic lakes. In addition, we foimd that filter feeder abun- dances increased from low numbers immedi- ately below the outlet to peak numbers some 1990] (>()MMUM rV Dk\ Kl.OrMKNI Bl'I.OW l.AKK Ol'TLFTS 311 distance (40-80 m) downstream. Tliis dillers from the progressive downstream decrease in filter feeder abnndance modeled 1)> Sheldon and Oswood (1977) and ma\' have been over- looked by them becanse they sampled no closer than 25 m below the lake ontlet. A parabolic relationship of filterer density with distance rather than a negativ e linear regres- sion may be due to snboptimal environmental (e.g., velocity) or biotic conditions near the ontlet. Current velocities in Pettit Creek were less near the outlet (26-29 cm/s) than fmther downstream (33-40 cm/s) and may not haxe met the needs of filterers for feeding or respiration. Further, changes in substratum characteristics occinred within 40 m of the lake outlet (Table 1). Mackay and Waters (1986) suggest that changes in filterer abun- dances between the impoundments they studied may be due to a greater abundance of attachment sites. Our data contribute a spatial dimension to the recolonization of stream benthos by macroinvertebrates. These data suggest the importance of the habitat templet in the structuring of benthic communities. This implies faster recovery or community devel- opment in streams below lake outlets in which adequate structural habitat is present. These data suggest that low-head hydro installations can impact macroinvertebrate communities by reducing the structural attributes of the habitat templet. Acknowledgments We are grateful to B. Jamison, J. Mann, and C. Nelson for their help in the field and in the laboratory throughout the course of this pro- ject. We also thank D. Bosen, S. Danamraj, D. C()etseh(\ S. Hart, E. Hitchcock, S. Min- shall, D. Misner, 1.. Heed, C. Richards, and T. Tiersch for their assistance. Funding for the project was provided by the U.S. Fish and Wildlife Service. Literature Cited Chandler. D. C. 1937, Fate- of typical lakr plankton in streams. Ecological Monographs 7: 445-479. CusHiNG, C. E. 1963. Filter-feeding insect distribution and planktonic food in the Montreal River. Trans- actions of the American Fisheries Society 92: 216-219. Maciolek, J A , AND M. G TuNZi. 1968. Microseston dynamics in a simple Sierra Nevada lake-stream system. Ecology 49; 60-75. Mackav, R. M , andT'f VV.atehs 1986. Effects of small impoundments of hydropsychid caddisfly produc- tion in Valley Creek, Minnesota. Ecology 67: 1680-1686. Morin, a.. andR. H Peters 1988. Effects of microhabi- tat features, seston quality, and periphyton on abundance of oveiAvintering Ijlack fly larvae in southern Quebec. Limnology and Oceanography 33: 431-446. Reif, C B 19.39. The effect of stream conditions on lake plankton. Transactions of the American Micro- scopical Society 58: 398-403. Sheldon, A L, and M W. Oswood 1977, Blackfly (Diptera: Simuliidae) abundance in a lake outlet: test of a predictive model. Hvdrobiologia 56: 113-120. St.^tzner. B. 1978. Factors that determine the benthic secondary production in two lake outflows — a cybernetic model. Verhandlungen der Interna- tionalen Vereinigung fiir Theoretische und Ange- wandte Limnologie 20: 1517-1522. Received 15 September 1990 Revised 20 November 1990 Accepted 28 January 1991 Cn-at Basin Naturalist 50(41, 1990, pp. 313-319 SPATIAL PATTERN AND INTERFERENCE IN PINON-JUNIPER WOODLANDS OF NORTHWEST COLORADO Charles W, WVldcii' ", William L, .Slaii.soii', and KichaidT. Ward' Ab.STHACT. — Till' local sjiatial aiiaiiiicmciit oi the tonitcioiis dees I'iiiii.s cdiilis and }iitii])crus ostcospcniui wa.s mapped in two woodland stands and measured in two shrnh-dominated stands in the semiarid Pieeance Basin of northwest Colorado. In the woodlands, small trees were often clumped, while mediimi and large trees were either randomly or uniforniK dispersed. Significant regressions were obtained between a tree s basal area or canopy area and the area of its Dirichlet domain (the region closer to it than to any other tree). Both findings Irom the woodland stands accord with results obtained b\' other workers in other \ egetation. Like earlier workers, we interpret these patterns to indicate density-dependent mortality and density-dependent depression of growth rates among the trees in the woodlands. In contrast, the trees in the shrub-dominated stands are located at random with respect to each other. However, they are strongly associated with shrub cover. Apparently, tree seeds arrive in these stands primarily by long-distance dispersal, and the establishment of seedlings is more likeK in the shade of shrubs. Since plants are sessile and their growth is plastic, their arrangement in space and their sizes can reflect the history of their interac- tions with each other and with the environ- ment. With long-lived, slow-growing plants, studying pattern may be the only feasible way to discover which processes and inter- actions are important in determining commu- nity structure. We used some of the methods compared by Goodall and West (1979) to study the local spatial arrangement (pattern) of the small co- niferous trees Piniis edulis and Jiinipenis osteospenna in four stands in the semiarid Pieeance Basin of northwest Colorado. Our goals were twofold. First, we wished to de- termine whether the differences between methods Goodall and West (1979) detected in artificial populations are borne out in more complex real populations. Second, we wished to infer the processes that influence the estab- lishment of seedlings and the growth and mor- tality of plants. Study Area The Pieeance Basin occupies about 3000 km" in Garfield and Rio Blanco counties of northwest Colorado. Elevations range from 1707 to 2743 m (Tiedeman and TeiAvilliger 1978). The climate is semiarid with average annual precipitation ranging from 28 cm in the northwest to 63.5 cm in the southeast. About half of the annual total falls as snow and most of the remainder as rain in late-summer thun- derstorms. In the short term, precipitation is unpredictable and variable (Wymore 1974). The average annual temperature is 7 C at 1825 m (the elevation of the only permanent weather station in the basin), with a minimum monthly average in January of —5.9 C and a maximum monthly average in July of 20.3 C. The average annual temperature decreases by approximately 0. 85 C for every 100 m increase in elevation. Both temperature and precipita- tion are stronglv influenced bv local topogra- phy (Wymore 1974). We studied the spatial patterns of Pinus edulis Engelm. and Junipenis osteospenna (Torr.) Little (pinon and Utah juniper). Nomenclature follows Goodrich and Neese (1986). P. edulis and/, osteospenna are small coniferous trees common throughout the western United States, where they form mixed stands, often with an understory of scattered grasses, forbs, and shrubs. They commonly attain heights of 6-8 m, and both reproduce by seed. P. edulis usually possesses a single stem, while /. osteospenna is often multistemmed. The vegetation of the basin includes shrublands and woodlands of various floristic 'Depaitiiient of Bioloij) , Colorado .State Universits , Foit Collins, Colorado, L'S.^ 80.523. -Present address: Department ol'Bioiogv , Southern Oregon State College, 12.50 Siskiyou Boule •<1, Ashland, Oregon, USA 97.520-5071. 313 314 C. W. Welden etal. [Volume 50 compositions. Pinon-jimiper woodlands (as described in Tiedeman and Tei-williger 1978) have open canopies dominated by P. ediilis and J. osteospenna and occm- on broad, flat ridge tops at elevations between 1890 m and 2170 m, where soils are shallow, rocky, light brown, sandy loams (Entisols). Shrnblands dominated by Ai'temisia tridcntata Nutt. (sagebrush flats) oft:en occur on the same ridges as do piiion-juniper woodlands, at roughly the same elevations, but where soils are finer and deeper. Where pinon- juniper woodlands abut sagebrush flats, zones of in- termediate vegetation are often found. In these intermediate areas, the vegetation is dominated by Artemisia, with small, scat- tered individuals of P. edulis and /. osteo- spenna. Few of the trees overtop the shrubs. We studied two piiion-juniper woodlands (stands A and B), which were dominated by mature P. edulis and /. osteospenna, with little shrub understory. The canopies in these stands are not closed, but individual cano- pies sometimes abut or overlap. It is known from others (Powells 1965) and from personal observation that the roots of these trees usu- ally extend beyond the canopy. Thus, neigh- boring trees which do not seem to be compet- ing for light may nonetheless be competing belowground for water or nutrients. These stands lie at elevations of 2164 m and 1890 m, which approximate the elevational limits of this vegetation in the basin. Stand A slopes 1.5° and feces to the northwest (N62°W). Stand B slopes 3.0°, facing to the north-north- west (N22°W). Stands C and D are intermediate between pinon-juniper woodlands and sagebrush flats. None of the trees in these stands is as large as the largest trees in the piiion-juniper wood- lands, although many bear cones and are thus sexually mature. These stands occupy ridge tops at elevations of 2164 m and 1981 m. Stand C slopes 4.5°, facing west (N80°W), and stand D slopes 6.5°, facing north (N5°W). Methods Goodall and West (1979) reviewed pattern methods based on analyses of artificial popula- tions. They compared the statistical powers of the methods, that is, the probabilities of rejecting a false null hypothesis. With large samples, all the tested methods gave results reflecting the true dispersion pattern of artifi- cial populations, with powers approaching 100%. With smaller samples, however, methods differed in power. We used those having the greatest power with small samples: the variance/mean ratio (Clapham 1936) among quadrat methods, and the indices of Hopkins (1954) and Pielou (1959, 1960, 1961) among distance methods (see descriptions be- low). We also compared the frequencies of (quadrats containing exactly 0, 1, 2, . . . plants with the expected Poisson distribution by a chi-squared goodness-of-fit test. In addition to these methods, we included a measure of pattern that uses information not only about the locations of plants but also about their sizes. The Dirichlet domain (or Thiessen or Voronoi polygon) of a plant com- prises all the points closer to that plant than to any other (Honda 1978, Jack 1967, Mead 1971, Mithen, Harper, and Weiner 1984). Its size thus represents the area more easily ac- cessible to the plant than to its neighbors and may represent the amount of resources cap- tured or sequestered by a plant, or potentially more available to it than to its neighbors. This in turn may influence the plant s growth and fitness and indicate what effect, if any, its neighbors have on it. To detect whether this is the case, we regressed the areas of plants' Dirichlet domains on the sizes of the plants. The variance/mean ratio test (Clapham 1936) is based on the expectation that, in a randomly dispersed population, the fre- quency distribution of quadrats containing ex- actly 0, 1, 2, 3, . . . individuals approximates the Poisson distribution. One property of this distribution is that its mean and variance are equal, and their ratio therefore imity. The distribution of this ratio in large samples is approximately normal, with a mean of 1 and a standard deviation of (2/n — 1) " (Goodall and West 1979), where n is the sample size (num- ber of quadrats). In regularly dispersed popu- lations the ratio is less than 1, in aggregated ones greater. Hopkins's (1954) index A is based on the expectation that, in a randomh dispersed population, the average distance from ran- domly located points to the nearest plant ecjuals the average distance between plants and their nearest neighbors. Hopkins pro- posed the ratio of these two a\erages as his index: 1990] Patiern ak\) Im kiu khknck 315 A - (^ P,-)/(i' 1,-) where P, and I, are tlie sums of ecjual iiuinhers ol distances from random points to the nearest plant and from randomh selected plants to their nearest neighbors, respectivcK. In a randomly dispersed population, the expected \alue of A is 1, and for large samples its fre- (luencx distribution is approximately normal. N'alues of A larger than 1 indicate aggregation, less than 1 regularity Pielou (1959, 1960, 1961) developed two distance methods to measure pattern. The first uses a sample of distances from randomly located points to the nearest plant and an independent estimate of plant density. From these a statistic, alpha^,, can be calculated as follows: alpha,, = pi(D)omegap where D is the density of the plants, omega,, is the mean squared point-to-plant distance, and pi is the trigonometric constant. The second method (Pielou 1960) uses a sample of distances from randomly chosen plants to their nearest neighbors. A statistic alpha^ is calculated in the same way as alpha,,, substituting the mean squared plant-to-plant distance for the mean squared point-to-point distance. Pielou (1959) provides tables of con- fidence intervals and significance levels for values of alpha, and shows how they may be used to interpret alpha,, (Pielou 1960). We mapped the location of each Piiiiis echdis and Jiinipcnis ostcospcnna 10 cm tall or taller in parts of stands A and B. The mapped area in stand A was 2250 m"; in stand B it was 2500 m". We checked the accuracy of the maps by comparing plant-to-plant distances calcu- lated from map coordinates to the same dis- tances measured in the field. The greatest difference was about 10 cm. We classified plants into three height- classes. Small plants were 10 cm to 1 m tall, medium plants between 1 m and 3 m tall, and tall plants were taller than 3 m. The tallest trees in our stands were about 5 m tall. Small plants were not mapped in about one-third of stand A. For each P. echdis in these stands we mea- sured one canopy diameter in an arbitrary direction and estimated the area of its canopy as if it were circular. The living canopies of /. osteospenna were often interrupted by dead branches. We measured the living portions of their canopies and sununed the areas estimated from these. Basal areas were calculated lor both species from stem diame- ters measured at ground level. For multi- stemmed plants, the basal areas of all living stems were summed. We measured the dispersion patterns of the plants on these maps, using l)oth (juadrat and distance methods. Small plants were sampled with (juadrats 2.5 m on a side (in map scale), medium and tall plants with (}uadrats 5 m on a side. Quadrats were placed at the intersec- tions of a regular grid of lines 5 scale-meters apart; thus every point on the map was in- cluded in exactly one quadrat of a given size. There were 100 large and 400 small quadrats in stand B. Stand A was more irregular, en- compassing 90 large and 230 small quadrats. The spatial dispersions of each size class and species were measured separately and pooled. That is, the null hypothesis of random spatial dispersion was tested by five indices for small P. edidis, small /. osteospenna , all small plants, medium P. edulis, medium /. osteospenna, all medium plants, tall P. ediilis, tall/, osteospenna, all tall plants, medium and tall P. edidis combined, medium and tall /. osteospenna combined, and all medium and tall plants combined. We constructed Dirichlet domains (Honda 1978, Jack 1967, Mead 1971, Mithen, Harper, and Weiner 1984) for the plants by drawing lines connecting each plant to its immediate neighbors, and then constructing perpendic- ular bisectors of these lines (Fig. 1). Note that we did not weight the distance from a plant to the bisector by the size of the plant, and thus there is no necessary correlation between the size of a plant and the size of its Dirichlet domain. We estimated the areas of the Dirich- let domains by cutting the polygons from the maps and weighing them. We regressed the areas of the Dirichlet domains on the basal areas, and separately on the canopy areas, of their plants. Regressions on basal areas were compared to regressions on canopy areas, with and without logarithmic transformation, by graphical analysis of residuals. In stands C and D we located every P. edidis and /. osteospenna 10 cm or more in height within a square 50 ni on a side, noting whether it had become established under a plant canopy or in the open, based on observations of each 316 C. W. Welden etal. [Volume 50 'rk w ^V)C \ M Fig. 1. Construction of Dirichlet domains. a. Draw line segments connecting a focal plant to its neighbors. b. Draw the perpendicular bisector of each line segment. c. The Dirichlet domain is the region closer to the focal plant than all the perpendicular bisectors. d. Repeat for each plant. The Dirichlet domain of each plant is the region closer to it than to any other plant. tree's association with living or dead shrubs. We measured total plant cover of all species with two 50-m line intercepts. The association of P. edulis and /. osteosperma with plant cover was tested by a chi-squared test. We did not map these stands but measured distances between neighboring trees in the field. We used Pielou's alphaj to describe the spatial dispersion of the two tree species. Results Table 1 shows the number of P. edulis and /. osteosperma in each stand and the corre- sponding numbers per hectare. Table 2 shows the five dispersion indices for the trees in stands A and B, and Table 3 the interpreta- tions of these values. In the woodland stands (A and B) small plants tend to be clumped, and larger plants tend to be randomly or uniformly dispersed. The secjuence from clumped to random to uniform is violated in only three instances (asterisks in Table 3). These viola- tions may be the result of chance, since the tests for significance were all set at the 5% level and some spurious results are expected among such a large number of separate tests. All log-log transformed regressions of Dirichlet domain areas on plant canopy areas and basal areas in stands A and B are significant at the 5% level, except for that of /. osteosperma in stand A (Table 4, Fig. 2). These regressions show that, on average, larger plants have larger Dirichlet domains and are correspondingly farther from their neighbors. The Dirichlet domains of small plants are more variable in area than those of larger plants. Logarithmic transformation of both variates improves the distribution of variates and residuals and produces reason- able conformity with the assumptions of re- gression, but it does not change the signifi- cance of the regressions. These results are similar to those of regressing the distance be- tween a pair of neighboring plants on the sum oftheir sizes (Welden 1984, Welden, Slauson, and Ward 1988, cf Fuentes and Gutierrez 1981, Gutierrez and Fuentes 1979, Nobel 1981, Phillips and MacNhihon 1981, Pielou 1960, Table 1. Stand censuses, divided by height categories (10 cm < small < 1 m < medium < 3 m < tall) and by species. In parentheses are numbers per hectare Small plants Medimn plants Tall plants Total Stand A Piniis edulis 1 It ni perns o.steospeniui 61 (432) 26 (184) 56 (229) 12 (49) 67 (274) 39 ( 160) 184 (753) 77 (315) Stand B V. edulis J. osteosperma 88 (352) 86 (344) 32(128) 9 (36) 11 (44) 41 (164) 131 (524) 136 (544) Stand C /'. edulis J. osteosperma 56 (224) 7 (28) 22 (88) 7 (28) 1 (4) 0 79 (316) 14 (56) Stand D I', edulis J. osteosperma 47(188) 34(1.36) 30 (120) 46(184) 4 (16) 17 (68) 81 (324) 97 (388) 1990] P A'l'IKHN AND InTIsKI' I'.HI^NCK 317 Tahi.K, 2. N'aliics ol ilisix'isioii imliccs in slaiuls A and B. Indices arc idcntiiicd in llic' lc\t and these \aliics are inteipreted in lahle 4. A dash inchcates that tlu' test eonid not he |)ei'l()inicd. Distanei' methods Quadrat methods Stand alp l>a„ alpl la, A c hi- var/ mean A B A B A B .\ B A B F. rclulis Small 0.9.3 1.04 1.12 0.84 0.83 1.21 12.2 25.0 1.27 1.98 MecUiMn ().9S 1.23 1.15 0.66 0.85 1.83 5.97 — 1.72 1.60 Mcdinm and t all !.()() 0.74 1.04 0.97 0.96 1.38 2.90 2.16 1.48 1.28 Tall 0.S4 0.38 o.ss 0.91 0.95 0.42 0.17 — 1.10 0.89 /. (isfcospcnnii Small 1.03 1.36 1.04 1.07 0.99 1.27 — 29.4 1.52 1.15 Medium 0.61 1.06 0.84 0.64 0.72 1.65 — — 1.04 0.86 Medium and t all 0.73 0.93 1.09 1.29 0.78 0.73 0.24 2.12 0.96 0.86 Tall 0.73 0.73 LIS 1.39 0.75 0.54 0.24 1.10 0.97 0.82 Species coiuhint •d Small 1,25 1.29 0.79 0.86 1.58 1.50 18.1 8.18 1.62 1.65 Mediimi 1.17 1.58 0.96 0.54 1.22 2.94 2.28 10.9 1.44 2.22 Medium and t all 1.06 0.S6 0.94 1.14 1.12 0.92 4.35 0.91 1.34 1.19 Tall 0.92 0.74 0.98 1.13 0.78 0.59 0.60 3.76 1.04 0.95 Tabli-: 3. Pattern analyses ol stands A aud B. C indicates that the plants are clumped, R that they are randomly dispersed. U that they are uniformly dispersed. All indicated nonrandom dispersions are sii^nificant at the 5% level. A dash indicates that the test could not be performed. Asterisks denote contradictions to the general trend of C - R - U with increasing plant size. Distance methods Q uadr at methods Stand a Ipha,, alpha, A chi- \ar/i nean A B A B A B A B A B P. ediilis Small R R R R* R R* C C C C Medium R R R C R C c; — C C Medium and t all R U R R R R R R C R Tall R U R R R R R — R R /. ostcospenna Small R C R R R R — C C C Medium U R R R R R R C Mediiun and t; all U R R R R R R R R R Tall U U R U R U R R R R Species combined Small c C R R* R C C C C C Medium R C R C R C R C C C Medium and t; M R R R R R R R R C R Tall R I' R R R U l\ R R R 1961, Yeaton and Cody 1976, and Yeaton, Travis, and Gilinsky 1977). Plant cover (primarily of Artemisia ) in .stand C was approximately 20%, and al)out 96% of the P. edtdis and about 71% of the /. osteo- spenna had become established under plant canopy. Plant cover in stand D was about 18%, and about 93% of the P. edtdis and about 87% of the/, osteospenna had become estab- lished under plant canopy. The probability that establishment of P. edtdis or /. osteo- spenna is random with respect to plant cover is less than .001 in every case. The pattern statistic alphaj (Pielou 1960) showed no sig- nificant deviations from random dispersion among P. edtdis or /. osteosperma in stands CandD. Discussion Pielou (1959) and Goodall and West (1979) show that distance methods are more sensi- tive to uniformity and quadrat methods are 318 C. W. Weldenetal. [Volume 50 Table 4. Coefficit-nts of lou-log traiisiormed regressions of Diriclilet cloinain area on canopy and Ijasal areas. Significance is the probal)ility of sucli data if tlie true slope and r ecjual zero. Species Stand Y-intercept Slope Significance P. edulis Independent variable In (canopy area) A 98 0.056 10.12 0.13 0.019 In (basal area) 0.052 11.05 0.11 0.024 In (canopy area) B 33 0.272 5. 15 0.26 0,002 In (basal area) 0.191 6.81 0.21 0,011 /. osteosperma In (canopy area) A 27 0.031 12.50 -0.06 0.377 In (basal area) 0.020 12.04 -0.04 0.479 In (canopy area) B 31 0.367 5.41 0.23 0.000 In (basal area) 0.,352 6.84 0.18 0.000 Species combined In (canopy area) A 125 0.039 10.49 0.10 0.027 In (basal area) 0.047 11.15 0.09 0.015 In (canopy area) B 64 0..333 5.27 0.25 0.000 In (basal area) 0.26.S 6.88 0.18 0.000 5 10 ln(Canopy Area) (cirT) Fig. 2. Regression of Diriclilet domain area on canop\ area of pinons in stand A, Both variates lia\i' been trans- formed to their natural logarithms. more sensitive to clumping. This is borne out by Table 4, where it can be seen that the (juadrat methods ne\er detected uniform dis- persion while the distance methods did. The distance methods, on the other hand, tailed to detect clinnping in several cases where it was detected by the (}uadrat methods. The trees in the woodland stands (A and B) appear to be interfering (sensii Harper 1961, 1977) with one another, either by com- petition or by allelopatin . The trend Ironi chnnped to random to imiform dispersion with increasing plant size suggests density- dependent mortality. Density-independent mortalit) in a clumped popidation might con- ceivably reduce sample sizes in successively larger size-classes until the clumping is no longer detectably different fiom a random dis- persion, but it seems unlikely that it could produce a uniform dispersion (Phillips and MacMahon 19 and D. However, both tree species are significanth associated with plant cover. We presume that these trees became established after long-distance dis- persal (> 100 m) from nearby woodlands. The 1990] Pattern and Ini i-:ufeuenc:e 319 significant interaction in these stands is e\'i- clentK not interference between neighhorinii trees, but amelioration of abiotic stress under the canopies of preexistinii plants. Fowells (1965) reports that P. vihdis re(iuires shade earl\ in its de\elopnient. Our exidence for these interpretations is circumstantial. However, given the long lives and slow growth of these plants, and the vary- ing physical environment of the study area, such evidence may be the most informative. These pattern methods integrate the effects of environment and biotic interactions over the life spans of the plants, a time scale not usually accessible to more mechanistic methods. All our inferences of processes leading to the present pattern recjuire further examina- tion. Although ]. ostcosperma has been re- ported to produce allelochemicals (Jameson 1971), experiments should be done to deter- mine whether allelopathic effects occur under the conditions and in the soils of the Piceance Basin, and more field studies are needed to determine whether establishment occurs more often near neighbors or far from them. The dynamic behavior of the various pattern indices and regressions should be explored under conditions of density-dependent and density-independent mortality. Literature Cited Clapham, a R 1936. Over-dispersion in grassland com- munities and the use of statistical methods in plant ecology. Journal of Ecology 24: 232-251. Fowells, H A., ed 1965. Silvics of forest trees of the United States. Agriculture Handbook 271. 1965. FuENTES, E. R., AND J. R. GUTIERREZ. 1981. Intra- and interspecific competition between matorral shrubs. Oecologia Plantarum 16: 283-289. GooDALL, D W , AND N. E. WEST, 1979. A comparison of techniques for assessing dispersion patterns. Vegetatio 40: 15-27. Goodrich, S , and E Neese 1986. Uinta Basin flora. United States Department of Agriculture Forest Service — Intermountain Region. Ogden, Utah. Gutierrez, J R. and E R. Fuentes 1979. Evidence for intraspecific competition in the Acacia caven (Leguminosae) savanna of Chile. Oecologia Plan- tarum 14: 151-158. Harper, J L. 1961. Approaches to the study of plant competition. Pages 1-39 in Mechanisms in biolog- ical competition. Academic Press, New York. 1977, Population biology of plants. Academic Press, New York. Honda, H. 1978. Description of cellular patterns by Dirichlet domains: the two-dimensional case. Journal of Theoretical Biology 72: 523-543. Hopkins, B 1954. A new method for determining the t\ pe oldistribution of plant individuals. Annals of Botany (London) 18: 213-227. Jack, W H 1967. Single tree sampling in even-aged plantations for sinvey and experimentation. Pro- ceedings, 14th I.U.F.R.O. Congress, Section 25, Munich. Jameson. D A 1971, Degradation and accuinulation of inhibitory substances irom J ttnifX'nis ostcosperma (Torr.) Little. Pages 121-127 in Biochemical interactions among plants. National Academy of Science. Mead, R I97I. Models for interj)lant competition in irregularly distributed jjopulations. Pages 13-.32 in G. P. Patil, E. C. Pielou, and W. E. Waters, eds.. Statistical ecology. Vol. 2, Sampling and modeling biological populations and population dynamics. PcTinsylvania State University Press, University Park. Mithen, R,, J L Harper, and J. Weiner, 1984. Growth and mortality of individual plants as a function of "available area. Oecologia 62: 57-60. Nobel, P. S 1981. Spacing and transpiration of various sized clumps of a desert grass, Hilaria ri<:,i(la. Journal of Ecology 69: 735-742. Phillips, D L , and J A McMahon. 1981. Competition and spacing patterns in desert shrubs. Journal of Ecology 69: 97-115. Pielou. E C 19.59. The use of point-to-plant distances in the studv of pattern in plant populations. Journal ofEcology 47: 607-613. I960. A single mechanism to account for regular, random and aggregated populations. Journal of Ecology 48: .575-584. 1961. Segregation and symmetry in two-species popidations as studied by nearest-neighbour rela- tionships. Journal of Ecology 49: 2.55-269. TiEDEMAN, J A, AND C Tervvilliger, Jr 1978. A phytoedaphic classification of the Piceance Basin. Colorado State University Range Science Depart- ment Science Series 31. 265 pp. Welden, C. W 1984. Stress and competition among trees and shrubs of the Piceance Basin, Colorado. LInpublished dissertation, Colorado State Uni- versity, Fort Collins. Welden, C W , W L Slauson, and R T Ward 1988. Competition and abiotic stress among trees and shrubs in northwest Colorado. Ecology 69: 1566-1577. Wymore. I F. 1974. Estimated average annual water balance of Piceance and Yellow Creek water- sheds. Colorado State Lhiiversity Environmental Resources Center Technical Report Series No. 2. Yeaton, R, I,, AND M L Cody 1976. Competition and spacing in plant communities: the northern Mohave Desert. Journal ofEcology 64: 689-696. Yeaton, R. I , J. Travis, and E. Gilinsky 1977. Com- petition and spacing in plant communities: the Arizona upland association. Journal of Ecology 65: 587-595. Received 20 May 1990 Revised 5 November 1990 Accepted 8 January 1991 (;ri-at Basin Natuialisl 50(4), 1990. pp. 321-325 TRAMPLING DISTURBANCE AND RECOVERY OF CRYPTOGAMIC SOIL CRUSTS IN GRAND CANYON NATIONAL PARK David N.Cole' Abstract. — Cryptogamif soil ciiists in (Iraiul (>'anyoii National Park were trampled by hiker.s, imdi'r eontrolled condition.s, to determine how rapidh they were pulverized and how rapidly they reeovered. Only 15 trampling passes were required to destroy the structure of the crusts; visual evidence ol bacteria and cryptogam cover was reduced to near zero after 50 passes. Soil crusts redeveloped in just one to three years, and after five years the extensive bacteria and crvptogam cover left little visual evidence of disturbance. Surface irregularity remained low after five years, however, suggesting that reco\ ery was incomplete. Cryptogamic soil crusts are common and functionally significant features of arid eco- systems. Bacteria, algae, fungi, lichens, and mosses bind surface soil particles together, creating a highly irregular surface crust of raised pedestals (typically black and several cm tall) and intervening cracks. Crusts pro- vide favorable sites for the germination of vascular plants (St. Clair et al. 1984) and play important roles in water conservation (Brotherson and Rushforth 1983) and nitrogen fixation (Snyder and Wullstein 1973). These crusts are particularly significant in reducing soil erosion. Soil aggregation raises the wind and water velocities required to detach soil particles, while the irregular soil surface tends to reduce wind and soil velocities (Brotherson and Rushforth 1983). Increased water infiltra- tion in crusted soils also reduces runoff and erosion. Increased soil stability is highly sig- nificant in arid environments where sparse vegetation and surface soil organic matter as well as sporadic torrential rainfall contribute to a high erosion hazard. A number of recent studies have examined the response of cryptogamic soil crusts to dis- turbance by grazing and by fire (Anderson et al. 1982, Johansen et al. 1984, Johansen and St. Clair 1986, Marble and Harper 1989). The results of these studies suggest that crusts are unusually fragile and can be seriously disrupted by low levels of disturbance that have no noticeable effect on vascular plants (Kleiner and Harper 1972). The fragility of crusts presents unique chal- lenges to land managers attempting to avoid adverse impacts on desert lands. This is particularly true in the many national parks located in the arid lands of the southwestern United States. The popularity of these desert parks has made it increasingly difficult for managers to meet management objectives that stress the maintenance of natural con- ditions and processes. Many hikers now visit places that a decade or two ago had few visi- tors. These backcountry users can signifi- cantly impact cryptogamic soil crusts if they wander off the trail or set up camp in crusted areas. The purpose of this study was to examine the effect of trampling disturbance on soil crusts to better understand how rapidly they are disturbed and how quickly they can re- cover. It was conducted in the backcountry of Grand Canyon National Park on a study site located close to the Bass Trail, at an eleva- tion of about 1,650 m. The site is flat, and during the study the soil crusts exhibited well- developed pinnacles and were conspicuously blackened with lichens. The vegetation type is a Coleogyne ramosissima-Pinus edulis— Junipcrus osteosperma woodland (Warren et al. 1982). Soils, derived from sandstones of the Supai Group, are shallow and highly sandy. The climate can be characterized as that of a cold desert; annual precipitation is about 25 cm with a bimodal occurrence in winter and summer. IiiternKnintain Re L'h Station. Forest Service, U.S. Department of Agriculture, Forestry Sciences Laboratory, Missoula. Montana .59807 321 322 D N.Cole [Volume 50 Table 1. Changes in the cryptogam cover, vertical distance, and coefficient of variation of vertical distance in response to different levels of trampling/ Fig. 1. The two trampling lanes immediately after 50 passes in tennis shoes (left) and 250 passes in lug-soled boots (right). Note horizontal bar for measuring vertical distances. Methods Two lanes about 6 m long and 0.4 m wide were delineated with lengths of PVC pipe in an area of well-developed, undisturbed soil crust (Fig. 1). The lanes were separated by a path that was trampled during the period when the treatments were applied and then allowed to recover aftei'ward. One lane was trampled by a 75-kg person in tennis shoes, the other by an 86-kg person in lug-soled boots. Measurements were taken prior to trampling and after 5, 15, 25, and 50 passes, a pass being one walk down the lane at a nor- mal gait. The lane trampled with lug-soled boots was trampled another 200 times, for a total of 250 passes. Subsequent measure- ments were taken one, three, and five years after the treatments were administered. Treatments and measurements occurred in late spring — April or May 1984. Each lane was sam])!ed along five transects oriented perpendicular to the lane and lo- Number Crvptogam Wrtical Coefficient of passes cover distance of variation (%) (mm) (%) 0 89 a 492 a 2,7 a 5 69 b 497 ab 1.9 b 15 45 c 505 be 1.5 b 25 36 c 502 abc 1.4 b 50 9d 511c 1.4b 250 Od 511c 1.4b ''Any two values in the same column tbllovved b\ the same letter are not significantly different (Duncan s multiple range test, p - .05). cated 1 m apart. Each transect consisted of 10 measurement points 2 cm apart in the cen- tral part of the lane. At each point along the transect the vertical distance between a hori- zontal pipe, temporarily connecting the pipe at each end of the transect, and the ground surface was measured. Then the ground sur- face at that point was categorized as either bare soil or cryptogam. These data provide three measures to eval- uate disturbance. First, the vertical distances, a mean of 50 observations per lane, provide a measine of the degree to which crusts have been compressed by trampling. The variabil- ity of vertical distances across each transect provides an indication of surface roughness, which should decline with trampling. Rough- ness increases with crustal development and is important in reducing soil erosion. The measure used is the coeflPicient of variation of the vertical distances. CoeflFicients were cal- culated for each of the five transects across each lane and then averaged. The third mea- sure is cryptogam cover, expressed as a per- centage of the 50 gromid surface observations for each lane. The significance of differences, between treatments and between years, was tested with analysis of variance and Duncan's multiple range test. Results Cryptogamic crusts were immediately pul- verized by trampling. Pedestals were flat- tened, and the black veneer of bacteria and cryptogams was obliterated. Changes in cryp- togam co\ (M-, \ ertical distance, and the index of surface roughness were all statistically sig- nificant (Table 1). Differences between the effects of trampling with tennis shoes and boots were not significant, however. 1990] Thamfunc; DisruHBANCE OF Ckyit()(;ams 323 'g 480 -1 E uj 490- O z H 500- sic- ca) -^ (b) _J iA^ 1 50 100 250 PASSES ^ 3 YEARS Fig. 2. Mean vertical distance from a horizontal transect to the ground surface (a) after different levels of trampling in lug-soled boots and (b) after one, three, and five years of recovery. Standards errors were all 2.2-2.8 mm. Table 2. Cryptogam cover, vertical distance, and co- eff^icient of variation of vertical distance 0, 1, 3, and 5 years following trampling.' Years since C ryptogam \'ertical C oefficient trainpl »g cover (%) distance (mm) o \ariation 0 3a 511a 1.3 ab 1 20 b 499 b 1.0 a 3 71c 491c 1.9b 5 85 d 490 c 1.9b Pre- 89 d 492 c 2.7 c trampl ng *Any two values in the same column tollowed by the same letter are Tiut significantly different (Duncan s multiple range test, p - .0.5). Cryptogam cover was reduced by 50% after 15 passes and was reduced to zero after 250 passes (Table 1). At this point the organisms were so widely dispersed that all visual evi- dence of their existence disappeared (Fig. 1). Destruction of pedestals also occurred rapidly (Fig. 2a). The vertical distance below the transect increased 13 mm following 15 passes. Additional trampling caused no significant further compression; the pedestals were already destroyed. Surface roughness, as measured by the mean coefficient of variation of the vertical distances, declined as the pedestals were pulverized (Table 1). All treat- ments were significantly different from the control, but not from each other. A black- ened, irregular, aggregated soil surface was replaced after trampling by a flat surface of unconsolidated sands, which was much more vulnerable to erosion. Substantial recovery occurred in the first Fig. 3. The lane that received 250 passes in lug-soled boots after five years of recovery. View is from the end opposite that in Figure 1. year after trampling ceased. After one year of recovery, cryptogam cover had increased sig- nificantly (Table 2), and vertical distance had decreased significantly (Fig. 2b); however, surface roughness had not increased (Table 2). The unconsolidated sands left by trampling had reaggregated into a smooth, raised crust, but neither pedestals nor the blackened ve- neer of organisms had reformed. After three years of recovery, vertical distances were sim- ilar to pre-trampling levels. Cryptogam cover had increased dramatically, as had surface roughness, although both were still below pre-trampling values (Table 2). After five years of recovery, cryptogam cover had re- turned to pre-trampling levels. At this point all visual evidence of damage was gone (Fig. 3). Surface roughness values remained depressed (Table 2), however, suggesting that pedestals had not redeveloped fully. The typical pattern of structural destruction and recovery is illustrated in Figure 4, which 324 D. N. Cole [Volume 50 ^ 450- 1 ^ 470- UJ o < 490 UNDISTURBED 50 PASSES — \ -z. AFTER 5 YEARS I ' I ' I ' I I I 4 8 12 16 20 t I ' I ' 1 I I ' I 4 8 12 16 20 DISTANCE ACROSS HORIZONTAL TRANSECT (cm) Fig. 4. Vertical distance from a horizontal transect to the ground surface (a) before trampling, (b) after 50 passes, (c) after one year of recovery, and (d) after five years of recovery. Data are for one of five transects across the lane trampled in tennis shoes. shows the changes that occurred under one of the transects. Fifty passes with tennis shoes increased mean vertical distance and de- creased variations between adjacent sample points. The redevelopment of a soil crust din- ing the first year of recovery reduced vertical distance (i.e., the ground surface apparently rose), but surface irregularity remained low. After five years of recovery, the surface was more irregular than after trampling, but less irregular than before trampling. Discussion These results illustrate the damage hikers can do to cryptogamic soil. The structure of these crusts was destroyed by only 15 passes, and cryptogam cover was negligible after only 50 passes. Compared with the response of vascular plants to similar levels of trampling disturbance, cryptogamic crusts are highly fragile but moderately resilient (Cole 1985, 1988). No other experimentally trampled veg- etated surfaces have been denuded by such low levels of trampling. Recovery was surprisingly rapid, however. This conclusion agrees with that of studies of recovery after grazing and fire (johansen et al. 1984, Johansen and St. Clair 1986), which report more rapid and extensive recoxery than anticipated. In this study recovery rates were probably increased by the close proxim- ity of inoculum to the disturbed lanes and b>' the fact that disturbance occurred only once and was then removed. This study and previ- ous ones rely primarily on visual criteria foi- evaluating recovery. The depressed surface roughness values five years after trampling suggest that complete recovery will take lon- ger than five years. On disturbed sites at Canyonlands National Park such parameters as chlorophyll content, species diversity, and the thickness of the subsurface gelatinous sheaths that bind soil particles remain low even after crusts appear to have recovered (Belnap 1990). The finding that the crustal surface rose during the first few years following the cessa- tion of trampling is intriguing. The process by which pinnacled crusts develop is not well understood, but this result suggests that they may develop through accretion rather than erosion. If they were erosional features, the undisturbed strips should have remained con- spicuously higher than the treatment lanes. This was not the case. Given the fragility of these crusts, random trampling by backcountry recreationists is ca- pable of seriously impacting large areas. Very low levels of ongoing use will maintain high levels of disturbance. This shows up most commonly as webs of trails that surround trail junctions, camping areas, and points of inter- est. In arid parks of the southwestern United States it is important to educate visitors about the nature, importance, and fragility of cr\p- togamic crusts. With this knowledge, visitors are more likely to voluntarily minimize tram- pling of crusts and support management actions taken to protect areas of crust. Most visitors neither recognize cryptogams as fragile vegetation nor realize their importance to site stability. It is also important to locate trails, camping areas, and other activity sites away from places with well-developed crust and, where this is not possible, to try to con- fine traffic to one well-developed route. The one positive management implication of this research lies in the finding of rela- tively fast visual reco\ery. Where it is possible to eliminate trampling, crusts can ({uickly reestablish themsehes. In this experiment trampling left an apparently sterile surface of sand that, in reality, was heavily inoculated with crustal organisms. Managers can speed recovery of disturbed areas by inoculating them (St. Clair et al. 1986). Moreover, even though complete recovery may take much more than five years, the rapid elimination of 1990] THAMri,lN(; DiSTUHBANCKOFCHVnXKiAMS 325 the visual evidence oi damage is helplul. This makes it easier for managers to keep visitors oft certain trails and campsites. Acknowledgments Partial support for this research was provided by the Western Region, National Park Service, USDI. This paper profited from the comments, on an earlier draft, of Jayne Belnap, David Chapin, Jeff^ Johansen, and Jeff" Marion. Literature Cited Anderson, D. C, K. T. Harper, and S. R Rushforth. 1982. Recovery of cryptogamic soil crusts from grazing on Utah winter ranges. Journal of Range Management 35: 355-359. Belnap, J. 1990. Microbiotic crusts: their role in past and present ecosystems. Park Science 10(3): 3-4. Brotherson, J. D., andS R. Rushforth. 1983. Influence of cryptogamic crusts on moisture relationships of soils in Navajo National Monument, Arizona. Great Basin Naturalist 43: 73-78. Cole, D. N. 1985. Recreational trampling effects on six habitat types in western Montana. USDA Forest Service Research Paper INT-350. 43 pp. 1988. Disturbance and recovery of trampled mon- tane grassland and forests in Montana. USDA Forest Service Research Paper INT-389. 37 pp. Johansen, J. R., and L. L. St Clair 1986. Cryptogamic soil crusts: recovery from grazing near Camp Floyd State Park, Utah, USA. Clrcal Basin Natu- ralist 46: 632-640. Johansen, J. R., L. L Sr Clair, B L Webh, and C. T. Nebeker. 1984. Recovery patterns of cryptogamic soil crusts in desert rangclands following fire dis- turbance. Bryologist 87: 238-243. Kleiner, E F., and K. T. Harper. 1972. Environment and community organization in grasslands of Canyonlands National Park. Ecology 53: 299-309. Marble, JR.. and K.T Harper. 1989. Effect of timing of grazing on soil-surface cryptogamic communities in a Great Basin low-shrub desert: a preliminary report. Great Basin Naturalist 49: 104-107. Snyder, J M., and L. H. Wullstein 1973. The role of desert cryptogams in nitrogen fixation. American Naturali.st 90: 257-265. St. Clair, L. L , J R Johansen, and B L Webb 1986. Rapid stabilization of fire-disturbed sites using a soil crust slurry: inoculation studies. Reclamation and Revegetation Research 4: 261-269. St Clair, L. L., B L Webb, J R Johansen, and G T Nebeker. 1984. Cryptogamic soil crusts: enhance- ment of seedling establishment in disturbed and undisturbed areas. Reclamation and Revegetation Research 3: 129-136. Warren, P L . K L. Reichhardt. D A Mouat, B T. Brown, and R. R Johnson 1982. Vegetation of Grand Canyon National Park. Technical Report 9, USDI National Park Service, University of Ari- zona, Tucson. 140 pp. Received 1 June 1990 Revised 2 November 1990 Accepted 8 January 1991 Creat liasi.i Naturalist 50(4). 1990, pp. 327 331 EIMERIA SP. (APICOMPLEXA: EIMERIIIME) EUOM WYOMINC; GROUND SQUIRRELS (SPERMOPHILUS KLF.GANS) AND WHITE-TAILED PRAIRIE DOGS {CYNOMYS LEUCURUS) IN WYOMING Lain M. Shiilts' ", H()l)frt S. Sc\illf', Nancy L. Stanton', and (Ivmgv K, Menknis, Jr.'' Abstract. — Six species of the coccidian genus Eimcrui (E. larimcroisis [prevalence - 17%], E. hilcmicllata [12%], E. beechcyi [34%], E. inorainensis [43%], E. cdUospcnnophiH [21%], and E. spcrmophili [5%]) were recovered from WVominu ground squirrefs {Spcrniopliilus ch'^ans ele^,(ins) collected dining 1983, 1984, 1985, and 1986. Infected ground s))liilus clcfi.dns) inlfclcd by Eimeriu sp. by Near. Samples (N ^ 1007) wvvv taken liom two haliitats in Wyomini;;. Animals ma\ be infeeted simultaneously by more than one speeies oi Eiiiicria. 1983 1984 1985 Species Mesic Xeric Mesic Xei ric Mesic Xeric of {iV = 314)' (\ 212 :) (iV - 34) (N 72: ) (N 86) (TV 74) No. No. No. No. No. No. Eimcrid of scjuir reh ol 1 scjuinels of scjuirrel;! of s- 37; 279-290. Hu.ton, D F J. AND J L. Mahrt 1971. Eimeria sper- mophili n.sp. and other Eimeria spp. (Sporozoa, Eiiueridac) irom three species of ,\lberta Sper- mophihis (Rodentia, Sciuridae). C>'anadian Journal ofZoology 49; 699-701. LKVINE, N. D, V IVENS. and F J KRI IDENIER 1957. New species oi Eimeria from Arizona rodents. Journal of Protozoology 4; 80-88. 1990] COCCIDIAN OCCL'HHKNCK IN WVOMINC Scil'HOMOHI'lIS 331 1990. 'Ihc cottidiaii iiarasitcs (I'roto/.oa: Sjioro/oa) of roclcnts. (MU' I'lx-ss, lioca Katon, I'lorida. 228 pp. Mknkk.ns, C E , Jh., and S. H Andf.hson 19(S9. Tt'iiipo- ral-spatial Nariation in the white-tailed prairie do,n deinoiiraplu and life liistories in Wyoniinfi. (Cana- dian Journal ot Zoology 67: 343-349. Pellkrdv, L., and a. Babos. 1953. Zur Kenntni.s der Kokzidien aus Citcllus citellus. Acta Veterinaria. Academiae Scientariurn Hungaricae3: 167-172. Seville, R S , and E. S. Williams 19(S9. Endopara.site.s of the white-tailed prairii- dog, Cynoiiii/s Ivttcunis, at Meeteetse, Park CCountx', WVoniiiig. Proceeding.s of the Hehninthological Soeiet\' of Washington 56: 204-206. Shults, L. M. 1986. Coccidian parasites (Einieriidae) of the Wyoming groinid sciiiirrel, Spcrmophiliis elegans. Unpublished dis.sertation. University of Wyoming, Laramie. 101 pp. Todd, K S . Jr . and D M. Hammond 1968a. Life cycle and host specificity of Eimeria callospennopliili Henry, 1932 from the Uinta ground squirrel Spennophilus anuatus. Journal of Protozoology 15: 1-8. 1968b. Life cycle and host specificity oi Eiiiicria larimcrensis Vettcrling, 1964, from the Uinta ground s(juiri el Spennophilus (irnialiis. Journal of I'rotozoologN 15: 268-275. Todd K S , Jh , 1) M Hammond, and L. C Anderson 1968. Obserxatious on the life cycle oi' Eimeria biUimcUata llenr\, 1932, in the Uinta ground scjuirrel Spennophilus anudtus. Journal ol Proto- zoology 15: 732-740. ToRHEiT. B. E., W. C. Mahquahdt, and a. C Cahey. 1982. A new species o[ EAmeria from the golden- mantled ground scjuirrel, Spennophilus lateralis, in northern (Colorado. Journal ol Protozoology 29: 157-159. Veluvolu, P., andN. D Levine 1984. Eimeria helclin^,ii n.sp. and other coccidia (Apiconi])lexa) of the ground sciuirrel SpcrmopJiilus hehlin^ii. Journal of Protozoology 31: 357-358. Vetperlinc. J M 1964. Coccidia (Eimeria) from the prairie dog, Ctpunntjs ludovicianus, in northern Colorado. Journal of Protozoology 11: 89-91. Zeoers. D a 1984. Spennophilus ele^ans. Mammalian Species No. 214: 1-7. American Society of Mam- malogists. Received 1 August 1990 Revised 8 January 1991 Accepted 24 January 1991 Creat Basin Naturalist 50(4), H)9I), pp. 333-338 EMERGENCE, ATTACK DENSITIES, AND MOST RELATIONSHIPS FOR THE DOUGLAS-FIR BEETLE (DENDROCTONUS PSEUDOTSUGAE HOPKINS) IN NORTHERN COLORADO E. I). Lcssard'andJ. M. Schinicl- AbstraCT. — Douglas-fir beetle-infested Doiiglas-fir trees were partially eaged to deteniiine the emergenee period and beetle production. Beetles began emerging in April, but emergence peaked between 10 and 26 June. In 1987 and 1988 beetle emergence averaged 20 or more per sq. ft. of bark. Annual growth of the infested trees showed a decline prior to the beetle outbreak followed by an increase during the outbreak. The Douglas-fir beetle, Dcndroctonus pseudotsugae Hopkins, is usually an insignifi- cant pest of Douglas-fir (Pseudotsugae men- ziesii [Mirb.] Franco) in the Front Range of Colorado. The beetles life cycle generally lasts one year, although a partial second gen- eration has been noted in other parts of its range. The beetle prefers windthrown trees but will infest standing trees during droughts or high population levels (Wood 1963). Stand- ing trees also become more susceptible to infestation by the Douglas-fir beetle after severe defoliation bv insects (Wright et al. 1984). In 1972 western spruce budworm (Choris- toneura occidentalis Freeman) populations began to increase in Roosevelt National Forest west of Fort Collins. Defoliation was noticeable, moderate, and limited to 3500 acres (1378 ha) in 1974 (Minnemeyer 1974), but by 1976, 54,000 acres (21,260 ha) were moderately to severely defoliated. Parts of Poudre Canyon, the location in this study, were severely defoliated (Cresap 1976). In 1977, the area of severe defoliation more than doubled on the forest (115,840 acres [45,606 ha]). Defoliation continually increased from 1977 until 1983, reaching a maximum of 469,000 acres (184,646 ha) (see Raimo 1983). Defoliation in Poudre Canyon, while noted as early as 1976, was confined to particular por- tions and was not extensive throughout until 1980 (see Linnane 1977, 1981). Thereafter, it was extensive and moderate to severe on most north-facing slopes until 1983-84. Although the acreage of moderately to severely defoliated stands progressively in- creased from 1977 to 1983, egg mass densities peaked in 1980, four years prior to the maxi- mum acreage defoliation, and had declined substantially by 1984 (see Raimo 1983, 1984). By 1985 population levels became endemic, with only light defoliation visible. As the budworm outbreak subsided, Douglas-fir beetle populations began to in- crease. Scattered groups of faded trees were observed in the mid-1970s. Subsequently, beetle-killed Douglas-fir have increased both in numbers and in geographic extent (J. M. Schmid, personal observation). Numerous stands of Douglas-fir on north-facing slopes suffered significant tree mortality. Because tree mortality became significant and our knowledge of the life history and habits of the Douglas-fir beetle in Colorado was deficient, the current infestations pro- vided an opportunity to learn more about the beetle s life history and habits in Douglas-fir stands in Colorado and expanded our knowl- edge of the geographic variation in these aspects of the beetle s biology. Methods To monitor beetle emergence, we attached 1 X 2-ft. (.3 X .6-m) wire screen emergence cages to infested trees in Poudre Canyon west of Fort Collins, Colorado, in late Febru- ary 1987. Two cages were attached at breast height on each of five randomly selected 'Forest Pest Management, R-IO, Anchorage. Alaska 99601. "Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colorado 80.526. 333 334 E. D. LessardandJ. M. Schmid [Volume 50 1986-infested trees at each of three locations: the Narrows, Pingree Park road turnoff, and near BM 6998 (east of Indian Meadows). On April 3 another five trees near Crystal Lakes (northwest of Red Feather Lakes) were caged in the same manner. To monitor emergence in 1988, we at- tached cages, as described above, to five 1987- infested trees on each of four sites: near Crys- tal Lakes (northwest of Red Feather Lakes) and in Poudre Canyon in late September 1987, and also near Camman Springs (south of Poudre Canyon) and near Black Mountain (north of Red Feather Lakes) in late October. Cages were attached at breast height for practical purposes. Furniss (1962) recom- mended sampling for attack and brood densi- ties at the midpoint on the bole because attack densities were twice as great there as at breast height, attack success was greatest in the mid- point zone, and live brood was greater. Fur- niss worked on standing trees that averaged 20 inches in diameter at breast height and ranged in height from 79 to 162 ft.; the sam- pling point on the average tree was thus 40 ft. or more aboveground. In contrast, our trees averaged 15 inches in diameter at breast height with trees only at the Pingree Park road site averaging 20 inches; tree height ranged from 34 to 88 ft. Although Furniss recommended sampling at or near the mid- point of the bole, it should be noted that he felled his trees for sampling and did not ex- tract his samples from standing trees. In addi- tion, the zone of optimum sampling is lower on smaller trees in the southern portion of the trees' range, i.e., southern Utah, than in Idaho, where Furniss did his study. Because our trees were smaller and were not to be felled, the terrain was difficult at some sites, and there was no evidence to suggest that the beetle s emergence pattern varied with height, we attached cages at breast height. During 1987 and 1988, cages were checked at one- to two-week intervals from 1 April to 1 July. After I July, cages were checked at irregular intervals tlnough September. Dur- ing each check period, the number of emerg- ing adults was recorded for each cage, and observations were made on the discoloration of foliage on the infested trees. The density of emerging beetles per sci. ft. (.09 m") of bark siuface was determined by dividing the total number of beetles emerging in each cage by the surface area covered by each cage (ca. 2 sq. ft. [.18 m~]). Beetle num- bers were subjected to one-way analysis of variance to determine if differences among locations were significant (alpha = .05). Beetle numbers at breast height were also tested against their respective tree diameters to determine if beetle production was related to tree diameter (DBH). For each year, tree diameters were grouped into three classes, and beetle numbers among diameter classes were tested for significant differences using analysis of variance (alpha = .05). For 1987 the diameter classes in inches (cm) were: 7.5-9.6 (19-24), 11.2-15.7 (31-61) and 16.2-24.3 (41-62). For 1988 diameter classes were: 9.3-12.5 (24-32), 12.6-13.3 (32-34), and 13.4-18.0 (34-46). Diameter classes dif- fered between 1987 and 1988 because the diameters of the infested trees were different. A one-way ANOVA was used because all diameter classes were not equally present on all locations. Population trend was evaluated by dividing the density of emerging beetles by twice the density of attacks (this assumes a pair of beetles creates each attack). When the ratio of emerging beetle density to attack density exceeds one, the beetle population is increas- ing. When the ratio is less than one, the popu- lation is decreasing. The density of beetle attacks on standing trees was determined by removing 6 x 12- inch (15 X 30-cm) bark samples from or near breast height. Two samples were removed from each 1986-infested tree in late October 1987. Two samples from each 1987-infested tree were removed in late September 1987. The bark samples from 1987-infested trees were also used to determine brood density and stage of development. To determine past growth rates of the 1986- infested trees, we extracted increment cores from the caged trees at breast height. Annual radial growth for each of the last 20 years was measured to .001 inch (.03 mm). Mean annual growth was determined for all trees from each of the four locations. Annual growth during the three preceding five-year periods (1972-76, 1977-81, 1982-86) was analyzed for significant differences in the periodic growth rate using one-way analysis of \'ariance (alpha - .05). Separate one-way analysis of variance was used for each location because 1990] Dou(jLAs-FiH Bi-:i: ill: 335 1000 r- ■^ 750 - (0 O) 3 s c 0) E a> E 3 z 500 - 250 - Jun 1 Jun 22 Jul 6 1000 r- (0 © 0} i3 ;^ 750 (0 iS C3) o Q c E 0) 0} E 3 500 250 0 o o Poudre Canyon O- — -O Red Feather 1988 Apr 1 May 1 Jun 1 Jun 22 Jul 6 Fig. 1. Total number of Douglas-fir beetles emerging from five trees at fom- loeatit)ns in the Arapaho-Koosex-elt National Forest, Colorado, in 1987 and 1988. the variability in site and stand conditions would not yield meaningful results in a more complex statistical design testing for differ- ences among locations and their interactions. Periodic growth for the five-year periods was also used to compute Mahoney's PGR (Mahoney 1978), which is the ratio of the growth for one five-year period to the growth for the previous five years. When analysis of variance indicated signifi- cant differences among the means, Tukey's test was used to determine which means were different (alpha = .05) (Steel and Torrie 1960). Results and Discussion Emergence. — Adults began emerging in mid-April of both years (Fig. 1), continuing to emerge at low rates until early June. Emer- gence peaked between 10 and 26 Jtme in both 336 E. D. LessardandJ. M. Schmid [Volume 50 Table 1. Mean number of Douglas-fir beetles emerging per sq. ft. (.09 m") of bark for several locations in the Arapaho-Roosevelt National Forest. Within the same year, means followed by the same letter are not significantly different (alpha = .05). Nuni ber Numl )er of beetles of trees Cages (x -+- S.D.) Location 198" 1988 per tree 19S7 1988 Narrows 5 2 2 27 ± 14 a Pingree Park Road 5 2 2 40 ± 20 a BM 6998 5 5 2 21 ± 15 a 30 ± 28 ab Crystal Lakes 5 5 2 39 ± 18 a 55 ± 33 a Camman Spring 5 2 8 ± 10 b Black Mountain 5 2 10 ± lib years. Adults rarely emerged after 1 July. In terms of percentage of the emerging popula- tions, 18% of the beetles had emerged by 10 June in 1987 and 4% in 1988; 77% and 92% emerged between 10 and 26 June in 1987 and 1988; 5% and 4% emerged after 26 June, respectively. After 1 July, 2% or less emerged in both years. Wood (1963) noted two principal flight periods for the Douglas-fir beetle in Califor- nia, Oregon, and Utah, depending on the overwintering life stage — one during May- June and another during July- August. In this study we found only one principal flight period. If a second flight period is occurring, we believe the beetles are reemerging adults, not new adults emerging later from the caged hosts. Density' of emerging adults. — ^The num- ber of adults emerging per sq. ft. (.09 m") of bark surface ranged from 6 to 82 in 1987 and 0 to 88 in 1988. Mean numbers per sq. ft. (.09 m") of bark showed significant variation among areas in 1987 and 1988, but Tukey's test did not reveal significant pairwise comparisons in 1987 (Table 1). Although the number of emerging beetles did not significantly correlate with DBH, areas where mean tree diameter was 8.5 inches (22 cm) or less produced the lowest number of beetles. In addition, numbers were influenced by the density of attacks and tree diameter. The population trend ratio was generally >1 when attack densities were <12 per sc}. ft. (.09 m") and tree diameter was >10 inches (25 cm) DBH. When tree diameter was <10 inches (25 cm), the trend ratio was <1. Similarly, when the density of attacks was >14 per scj. ft. (.09 m"), the trend ratio was generally <1. Population trend thus appears to be influenced by competition (McMullen Table 2. Mean number of Douglas-fir beetles emerg- ing per sq. ft. (.09 m~) of bark by diameter class. Within the same \ear, means followed by the same letter are not significantly different (alpha ^ .05). Number Diameter class Niunber of beetles (inches [cm]) of trees (X ± S.D.) 7.5- 9.6(19-24) 11.2- 15.7(28-40) 16.2-24.3(41-62) 9.3 12.6 13.4 12.5(24-,32) 13.3(32-,34) 18.0 (.34-46) 1987 5 7 8 1988 6 7 21 ± 15 a 36 ± 16 a 34 ± 21 a 22 ± 25 a 20 ± 20 a 35 ± 38 a and Atkins 1961) and quantity of food (tree size, not number of trees) as hypothesized by Wright etal. (1984). Larger trees provide ade- ({uate food to produce an increasing popula- tion until the attack density exceeds 12 per sq. ft. (.09 m"). At greater densities, competition causes beetle production to decrease. Smaller trees generally have production rates less than one, even when attack densities are 8—12 per sq. ft. (.09 m~), because smaller trees do not provide sufficient phloem for developing larvae. Beetle densities in this stud\ were about the same as or greater than those found by Fredericks and Jenkins (1988) in Logan Canyon, Utah. Beetle numbers of 21-22 per s(j. ft. (.09 m") in our diameter classes of 7— 13 inches (18-33 cm) were comparable to the beetle numbers at 22-24 per sq. ft. (.09 m") of Fredericks and Jenkins (1988). In trees of comparable diameters (i.e., 22 inches [56 cmj), beetle numbers of 34 per sq. ft. (.09 m") (Table 2) in this study were slightly greater than the 22-24 per sq. ft. (.09 m") of Fredericks and Jenkins (1988). 1990] D()U(;[.AS-FiH Bkkti.k 337 Tablk 3. Mean annual radial growth in .001 inch (.025 cin) for the periods 1972-76, 1977-81, and 1982-86 for the four 1987 locations. Within the sanu' loeation, means followed bv the same letter are not significautlv different (alpha .05). Me an aTiuiial growth (.001 inch) (.V ± S.E.) Location 1972-76 1977-81 1982-86 Narrows Pingree Park Road BM 6998 Crystal Lakes 11 ± 1.0 a (28 ± 3) 38 ± 2.3 a (97 ± 6) 11 ± 0.9a (28 ± 2) 7 ± 0.6ab(18 ± 2) 12 ± 1.4 a (30 ± 4) 30 ± 1.2 b (76 ± 3) 7 ± 0.3 b (18 ± 1) 5 ± 0.3 b (13 ± 1) 19 ± 1.3 b (48 ± 3) 33 ± 1.2ab(84 ±3) 12 ± 1.8 a (30 ± 5) 8 ± 1.0 a (20 ± 3) Att.ack densities. — The nuniher of attack.s per sc^. ft. (.09 m~) of bark surface ranged from 8 to 20 in 1986 and 6 to 14 in 1987. Within each k)cation, attack densities were not signifi- cantly different between aspects. Mean densi- ties ranged from 9 to 15 in 1986 and 8 to 10 in 1987, comparable to the fifth-year attack densities in Oregon of Wright et al. (1984). Because the Colorado outbreak appeared to be in its fifth year, the pattern of attack densi- ties during the outbreak may be the same as in the Oregon outbreak. In contrast, attack densities from our Colorado locations were 62-80% lower than those of the Utah out- break. In the recent outbreak in Utah, attack densities were high and essentially the same throughout the first three years (Fredericks and Jenkins 1988). Apparently, the Utah out- break exhibited a pattern of attacks different from either the Oregon or Colorado out- breaks. Discoloration of infested trees.— In February following the attack, foliage of most infested trees was predominantly green, only the lower two or three whorls of branches having discolored to red-brown. By late April most trees had discolored, the color ranging from yellow-green to reddish. By mid-May most trees were reddish. Trees with extensive woodpecker debarking and foliage discol- oration in February turned reddish first, usu- ally by late April. Those without these charac- teristics discolored later but had turned by May. Foliage usually discolored at different rates in different crown levels, the lower crown fading first. When it was yellow-green, the rest of the crown was green. When the upper crown yellowed, the lower crown was already reddish. From August through Octo- ber, the best external clues for Douglas-fir beetle infestation were cinnamon-colored boring dust and/or clear pitch "streamers." During winter the most notable external char- acteristic was the debarked bole caused by woodpecker activity. These boles are lighter in color and can be discerned from more than 100 feet away. After October, but before the foliage turned red, woodpecker activity was the best characteristic for locating currently infested trees. Annual growth. — Annual radial growth varied significantly among and within loca- tions. Significant variation in growth among locations was expected because of differing site conditions, stand densities, and tree ages. In three of four locations, mean annual growth declined significantly in the 1977-81 period, presumably a result of the budworm outbreak (Table 3). Mean annual radial growth in each location increased during 1982-86. Thus, the increase in Douglas-fir beetle populations coincided with increasing growth of the host. Mean annual growth for the 1977-81 period ranged from .005 inch to .03 inch (.013 to .08 cm) and for 1982-86 from .008 to .033 inch (.02 to .08 cm) (Table 3). The growth rate was greatest on large trees situated in a ravine, a more favorable site. The periodic growth ratio (PGR) exhibited changes similar to the changes in mean annual radial growth. In three of four locations, PGR became <1 when 1977-81 was com- pared against 1972-76. PGR then became >1 when 1982-86 was compared with 1977-81. Growth rates declined for 1977-81 because of the budworm defoliation; thus, the change in PGR for 1977-81 vs. 1972-76 was expected. However, the increase in Douglas-fir beetle populations with >1 PGR for the 1982-86 period was unexpected. Most stands suscepti- ble to the mountain pine beetle (D. pon- dcrosae Hopkins) exhibit PGRs <1, and so a beetle outbreak coinciding with a period of increasing growth is unusual. 338 E. D. LessardandJ. M. Schmid [Volume 50 Literature Cited Cresap. v. L. M. 1976. Western spruce hucKvorm. USD A Forest Service, Rocky Mountain Region, Forest Insect and Disease Management, Lakewood, Col- orado. Biological Evaluation R2-76-11. Fredericks, S. E., and M. J. Jenkins 1988. Douglas-fir beetle (Dendroctontis pseudotsugae Hopkins, Coleoptera: Scolytidae) brood production on Douglas-fir defoliated by western spruce bud- worm {Choristoneura occidentalis Freeman, Lep- idoptera: Tortricidae) in Logan Canyon, Utah. Great Basin Naturalist 48: 348-351. FuRNISS, M. M 1962. Infestation patterns of Douglas-fir beetle in standing and windthrown trees in south- ern Idaho. Journal of Economic Entomology 55; 486-491. LiNNANE, J. P. 1977. Western spruce budworm on na- tional forest lands in Colorado. USDA Forest Ser- vice, Rocky Mountain Region, Forest Insect and Disease Management, Lakewood, Colorado. Bio- logical Evaluation R2-77-15. 1981. Western spruce budworm. Results of an egg mass survey on the Pike, and San Isabel and Ara- pahoe and Roosevelt National Forests in 1980. USDA Forest Service, Forest Pest Management, Rocky Mountain Region, Lakewood, Colorado. Biological Evaluation R2-81-2. Mahoney, R. L. 1978. Lodgepole pine/mountain pine beetle risk classification methods and their appli- cation. Pages 106-113 in Theory and practice of mountain pine beetle management in lodge- pole pine forests. Proceedings of a symposium at Washington State University, Pullman, Washing- ton, 25-27 April 1978. McMullen, L H , AND M D. Atkins. 1961. Intraspecific competition as a factor in the natural control of the Douglas-fir beetle. Forest Science 7: 197-203. MiNNEMEYER. C. D 1974. Western spruce budworm. USDA Forest Service, State and Private Forestry, Region 2, Denver, Colorado. Biological Evalua- tion R2-74-9. Rmmo, B J 1983. Western spruce budworm in the Rocky Mountain Region 1983. USDA Forest Service, Timber, Forest Pest and Cooperative Forestry Management, Rocky Mountain Region, Lake- wood, Colorado. Biological Evaluation R2-83-4. 1984. Western spruce budworm in the Rocky Mountain Region. USDA Forest Service, Timber, Forest Pest and Cooperative Forestry Manage- ment, Rocky Mountain Region, Lakewood, Colo- rado. Biological Evaluation R2-84-5. Steel, R G. D., and J H Torrie 1960. Principles and procedures of statistics with special reference to the biological sciences. McGraw-Hill Book Com- pany Inc., New York. 481 pp. Wood, S L. 1963. A revision of the bark beetle genus Dendroctontis Erichson (Coleoptera: Scolytidae). Great Basin Naturalist 23: 1-117. Wright, L. C, A A Berryman, and B E Wickman. 1984. Abundance ofthe fir engraver, Scohjtiis ven- tralis, and the Douglas-fir beetle, Dendroctontis pseudotsagae, following tree defoliation by the Douglas-fir tussock moth, Orgijia pseudotsiigata. Canadian Entomologist 116: 293-305. Accepted 3 October 1990 Great Basin NatmalisI 50(1). 1990, pp. 339-3-15 ECOLOGICAL REVIEW OF BLACK-TAILED PRAIRIE DOGS AND ASSOCIATED SPECIES IN WESTERN SOUTH DAKOTA Jon C. Sharps' and Daniel VV. Urcsk" AbstraCT. — Blatk-tailt'd prairie dogs {(Upioiiit/s Itidovicitinus) once occupied extensive areas throughout the (Ireat Plains. In recent years massive control programs have been initiated to reduce prairie dog populations, primarily to benefit the livestock grazing industry. Currently in western South Dakota most prairie dogs are found on public lands. Control programs using toxicants for prairie dogs have been found to be economically unfeasil)le when not combined with reductions in li\estock grazing. Control programs also have negatively impacted some nontarget species of birds and small mammals. Livestock grazing is directly related to prairie dog densities. Prairie dog and livestock grazing activities are responsible for keeping plant phenological development in a suppressed vegetative stage with higher nutritional qualities that attract greater herbivore use. Prairie dog colonies create and enhance habitat for many wildlife species; in western South Dakota 134 vertebrate wildlife species have been documented on prairie dog towns. Scientific evidence strongly suggests that prairie dogs are valuable components of the prairie ecosystem. They are responsible for maintaining, creating, and regulating habitat biodiversity through soil and vegetative manipulation for a host of vertebrate and invertebrate species dependent upon prairie dog activity for their survival. Quantified information regarding verte- brate wildlife species living on or closely asso- ciated with black-tailed prairie dog {Cynomys hidovicianus) colonies is lacking or is only alluded to in scientific literature. To promote a better understanding of the complexity of prairie dogs and their habitat requirements and their importance to vertebrate species of wildlife, we conducted a review of scientific literature regarding prairie dog biology, ecol- ogy, and associated biopolitics pertaining to land management practices. Most of the stud- ies and observations reported in this paper were conducted in western South Dakota. Where possible, corroborating studies and lit- erature from other areas are presented and their importance discussed. Historical Background Historically, prairie dogs occupied exten- sive areas on the Great Plains, ranging from Texas to Saskatchewan (Hall 1981) (Fig. 1). Merriam (1902) noted that prairie dogs com- pete with livestock for forage and are system- atically targeted for elimination by livestock producers. The largest areas of land in the United States currently occupied by prairie dogs are federally managed lands (Schenbeck 1982). In South Dakota most black-tailed prairie dogs are found on lands administered by USDA Forest Service, primarily the Buf- falo Gap National Grasslands and Fort Pierre National Grasslands (Schenbeck 1982). Storch (1989) estimated that prairie dogs inhabited 3,000 acres on the South Dakota portion of the Nebraska National Forest in the 1960s. In the mid-1970s prairie dogs inhabited approxi- mately 20,000 acres on the Conata Basin por- tion of the grasslands (Schenbeck 1982); Schenbeck's estimate represents an 87% in- crease over an eight-year period. The live- stock grazing industry claimed estimated losses of up to $10.29 per acre on pasture and rangeland and $30. 00 per acre for hayland on a statewide basis (Dobbs 1984) and objected to the increase in prairie dogs. Economics of Control and Livestock Grazing The South Dakota livestock industry has recommended and instigated widespread wholesale reductions in prairie dog densities on public land, and in 1983 the state legisla- ture listed the prairie dog as a pest and preda- tor (Clarke 1988). Of the 707,000 acres in the Ft. Pierre and Buffalo Gap National Grasslands, 'Wildlife Sy.stems. HC 82 Box 172B, Box Elder, South Dakota .57719. "USDA Forest Service, Rocky Mountain Forest and Range Experiment Station, 501 E. St. Joseph Street, Rapid City, South Dakota 57701. 339 340 J. C. Sharps and D. W. Uresk [Volume 50 Fig. 1. Distribution of black-tailed prairie dog (C(/no- im/s ludovicianus) on the Great Plains (adapted from Hall 1981). approximately 10,000 acres are currently oc- cupied by prairie clogs (Storch 1989). Control of prairie dogs has usually been initiated with- out consideration of the value of forage gained (Collins et al. 1984) or the effect on wildlife species associated with prairie dogs and their habitat (Sharps 1988). An economic analysis of prairie dog control by Collins et al. (1984) found it was not eco- nomically feasible to poison prairie dogs in the Conata Basin using zinc phosphide because the annual control costs exceeded the value of forage gained. Also, based on burrow counts, prairie dog densities were significantly less on areas excluded to cattle than on areas grazed by cattle (Uresk et al. 1982). Herbicide appli- cations to reduce forb production and thus reduce prairie dog densities were also found to be an inefficient control method because prairie dogs changed their diets from forbs to grasses (Fagerstone et al. 1977). It has long been known and extensively reported that cattle grazing will influence and is directly proportional to prairie dog densities (Koford 1958, Knowles 1982, Uresk et al. 1982, Cin- cotta 1985, Snell 1985). Schenbeck (1986) re- ported that habitat suitability for prairie dogs can be reduced by combining rodenticide use with changes in livestock grazing practices. The poison bait effects of zinc phosphide- and stiychnine-treated oats on nontarget birds, small mammals, and other nontarget species were evaluated by Uresk et al. (1988). The effects on nontarget bird species showed varied losses to Horned Larks, depending upon the density of strychnine-treated oats used, with no losses to other avian seed- eaters. No measurable reductions in Horned Larks were found using zinc phosphide- treated oats, although there were indirect impacts on Horned Larks resulting from habi- tat changes. Prairie dog towns provide habitat for many seed-eating and insectivorous birds. Significantly, Apa (1985) reported that 50 species of birds were observed using prairie dog towns during the course of his study. While zinc phosphide may not be detri- mental to Horned Larks and the smaller seed- eating birds, it has been reported to be rela- tively toxic to gallinaceous birds (Record and Swick 1983). Studies by Koford (1958), Smith (1958), Snell and Hlavacheck (1980), and Uresk et al. (1982) indicated that excluding or decreasing cattle grazing increases cool-season grass den- sity (wheatgrass and needlegrass) and reduces prairie dog colony size on mid- and short-grass rangeland. This method of prairie dog control has historically been opposed or rejected by the livestock grazing community. Although heavily grazed rangelands give rise to very slow forage improvement, prairie dogs alone are generally not responsible for range deteri- oration (Uresk 1987). Prairie dog expansion is related to livestock grazing (Uresk et al. 1982, Uresk and Bjugstad 1983). Black-tailed prairie dogs usually disperse during May and June and have been reported to move and become established an average of three miles from their original towns (Garrett and Franklin 1981, Cincotta et al. 1987). They will repopu- late their towns to initial population numbers in three years (Schenbeck 1982, Cincotta et al. 1987). Economically, control of prairie dogs is not feasible except at very low main- tenance levels — below 5% — based on an in- crease of forage for livestock of only 50 pounds per acre, a 4.4% increase (Uresk et al. 1982, Collins et al. 1984, Uresk 1985, 1986). 1990] Ecol()(;i(:ai,Kk\ iKWOi' lii.ACK r\ii,i;i) Phaiiuk Docjs 341 Associ atkdX'khikhhaikSi'kciks Prairie dogs create a biological niche or habitat for many species of wildlife (King 1955, Reading etal. 1989). Agnewetal. (1986) tonnd that bird species diversity and rodent abundance were higher on j^rairie dog towns than on mixed-grass prairie sites. The high diversity of bird species was attributed to het- erogeneous plant cover and species composi- tion (Agnew et al. 1986, Cincotta et al. 1987). In a survey of prairie dog towns extending through portions of Utah, Colorado, and New Mexico, Clark et al. (1982) recorded 107 ver- tebrate species and subspecies of wildlife; more species were associated with larger prairie dog towns than with smaller towns. Sixty-four vertebrate wildlife species were recorded by Campbell and Clark (1981) on 25 white-tailed and 21 black-tailed prairie dog colonies in Wyoming. Reading et al. (1989) listed 163 vertebrate species sighted on black- tailed prairie dog colonies. They suggest that "richness of associated vertebrate species on black-tailed prairie dog colonies increases with colony size and regional colony density. Data pertaining to vertebrate wildlife species associated with black-tailed prairie dog colonies were obtained from an extensive literature re- view, personal field notes (J. C. Sharps, unpub- lished), observations while conducting endan- gered species surveys, or observations inci- dental to other research on prairie dog colonies. In South Dakota, 600 vertebrate wildlife taxa were found statewide. There are 332 species located west of the Missouri River (excluding fish) (Sharps and Benzon 1984). Of western wildlife species, 40% were found to be associated with prairie dog colonies. This 40% represents 134 vertebrate wildlife species (Table 1) associ- ated with prairie dog colonies in western South Dakota: 88 birds, 36 mammals, 6 reptiles, and 4 amphibians (Agnew 1983, Apa 1985, Mac- Cracken et al. 1985, Agnew et al. 1986, Uresk et al. 1986, Deisch et al. 1989). Whitney et al. (1978) reported that approximately 33 bird spe- cies, or 39% of the birds found in South Dakota, are conspicuous on the grasslands. Of those 33 species only 5, or approximately 15%, were not observed or reported on prairie dog colonies. Plant-Soil-Animal Interactions Agnew et al. (1986) and Deisch et al. (1989) found five classes of invertebrates on prairie dog colonies localc'd on the Badlands National l\uk and Buffalo Cap National Grasslands, respectively. The five classes consisted of Insecta (6 orders, 26 families), Arachnida (4 orders, 10 families), Chilopoda, Diplopoda, and Crustacea. Agnew et al. (1988) found that ins(>ctivorous rodent species favor prairie dog colonies; these mammals, by consuming arthropods, may reduce localized arthropod outbreaks. Prairie dog colonies provide habitat diver- sity in the prairie ecosystem by mixing soils and regulating vegetative species diversity (Koford 1958, Bonham and Lerwick 1976, Ag- new et al. 1986, Detling and Whicker 1988, Sieg 1988). This in turn creates interactions and numerous niches, thereby contributing to the food chain for a host of invertebrate and vertebrate wildlife species. Prairie dogs alter soil structure and chemical composition by their burrowing activities, excrement, and addition of plant material, which contribute to vegetation diversity (Gold 1976, Hansen and Gold 1977, O'Meilia et al. 1982, Cincotta 1985, Agnew et al. 1986). Prairie dog activity results in the aeration, pulverization, granula- tion, and transfer of considerable quantities of soil (Buckman and Brady 1971, Sieg 1988). Soils in prairie dog colonies are richer in nitro- gen, phosphorus, and organic matter than soils in adjacent grasslands. Sheets et al. (1971) found prairie dog and cattle feces, grass seeds, stolons, roots, and remains of prairie dogs and mice while excavating 18 prairie dog burrows to retrieve black-footed ferret scats in south central South Dakota. Soil-enrichment activ- ity of the prairie dog is beneficial to the macro- arthropods living in the soil. Forbs and grasses in prairie dog colonies are constantly clipped by prairie dogs and remain in a state of regrowth (O'Meilia et al. 1982, Cincotta 1985). Ingham and Detling (1984) reported that prairie dog colonies support higher popu- lations of nematodes than adjacent areas away from the colonies. They also stated that prairie dog activities suppress plant phenological development, thus maintaining the plants in a vegetative state. Young vegetation, which is higher in nutritional qualities than mature plants, attracts cattle, bison, and pronghorn to prairie dog colonies (Uresk and Bjugstad 1983, Coppock et al. 1983, Knowles 1986, Krueger 1986, Detling and Whicker 1988). 342 J. C. Sharps and D. W. Uresk [Volume 50 Table 1. Vertebrate wildlife species associated with black-tailed prairie dog colonies in western South Dakota. Eastern tiger salamander Amhiistoma ti^rinum ti^rinuin Black-billed Magpie'' Common Raven ' Pica pica Great plains toad Biifo co^natus Corvus corax Western chorus frog Pseudacris triserata American Crow'' C. brachyrhnchos Bullfrog Rana catesl>eiana Northern Mockingbird' Mimus polyglottos Turtles Lizards Plains garter snake Smooth green snake Bullsnake Prairie rattlesnake Emydidae ukn spp. Iguanidae ukn spp. Thamnophis radix Opiieodnjs vcrnalis Pituophis melanoleuciis miji Crotalus viridis viridis Gray Catbird'' American Robin ' Eastern Bluebird' Mountain Bluebird' Water Pipit' Northern Shrike'' Loggerhead Shrike'' Dumetella carolinensis Tardus migratorius Sialia sialis S. currucoides Anthus spinoletta Lanius excubitor L. ludovicianus Great Blue Heron"" Ardea herodias European Starling' Sturnus vulgaris Trumpeter Swan"" Cij^nus buccinator Yellow Warbler'' Dcndroica petechia Canada Goose^ Bra Ufa canadensis Common Yellowthroat ' Geothlypis trichas Mallard' Anas platyrhynclios Yellow-breasted Chat'' Icteria virens Gadwall' A. strepera House Sparrow'' Bobolink ' Passer domesticus Northern Pintail' A. acuta Doliclwnyx oryzivorus Blue-winged Teal" A. disco rs Western Meadowlark' Sturnella neglecta Northern Shoveler'' A. clijpcata Yellow-headed Blackbird Xanthocephalus Canvasback' Aijthija valisineria xanthocephalus Turkey Vulture ' Catharics aura Red-winged Blacktiird'' Agelaius phoeniceus Red-tailed Hawk'' Butcojamaicensis Brewer's Blackbird'' Euphagus cyanocephalus Swainson's Hawk'' B. swainsoni Common Grackle ' Quiscalus quiscula Rough-legged Hawk' B. la^opus Brown-headed Cowbird'' Molothrus ater Ferruginous Hawk' B. recalls Western Tanager ' Piranga ludoviciana Golden Eagle'' Aquila chrysactos Dickcissel'' Spiza americana Bald Eagle'' Northern Harrier'' Haliacctus leucocephahis Common Redpoll'' Ca rduelis fla mmea Circus cyaneus Pine Siskin ' C. pinus Prairie Falcon'' Falco mexicanus American Goldfinch ' C. tristis Merlin' ill F. columharius Rufous-sided Towhee ' Pipilo eryth rophthalmus American Kestrel F. sparverius Lark Bunting'' Calamospiza melanoco rys Sharp-tailed Grouse'' Tynipanuchus phasianeUus Cirasshoiiper Sparrow'' Ammodramus savannarum Ring-necked Pheasant' Pliasianus colcliicus 1 Vesper Sparrow ' Pooecetes gramineus Sora' Porzana Carolina Lark Sparrow' Chondcstes grammacus Killdeer'' Cliaradrius vocifcrus Slate-colored Junco' Junco hyemalis Long-billed Curlew' Numcnius americanus Oregon Junco' J. oreganus Upland Sandpiper'' Bartramina longicauda Chipping Sparrow ' Spizella passe rina Long-billed Dowitcher' Limnodromus scolopaceus White-crowned Sparrow' Zonotrichia leucophrys Wilson's Phalarope' Phalaropus tricolor McCown's Longspur' Calca rius mccoivn ii Ring-billed GulF Lartts delawarensis Chestnut-collared Rock Dove'' Columha livia Longspur' C. ornatus Mourning Dove' Zcnaida niacroura Great-horned Owl ' Bubo lirginianus Shrews Soricidae ukn. spp. Snowy Owl' Nyctca scandiaca Bats W'spertilionidae ukn. spp. Burrowing Owl ' Atlienc cunicularia Eastern cottontail Sylvilagusjloridanus Short-eared Owl'' Asio flamnwus Desert cottontail S. auduboni Common Nighthawk ' Cliordcih's minor White-tailed jackrabbit Lepus fownsendii Belted Kingfisher' Ccrylc alcyon Black-tailed jackrabbit L. californicus Northern Flicker' Colaptcs (luratus Thirteen-lined S))ermo})hilus Red-headed ground scjuirrel tridecendineafus Woodpecker " Mclancrpcs crythroccphalus Black-tailed prairie dog Cynomys ludovicianus Downy Woodpecker' Picoidcs pubcscens Northern pocket gopher Thomonnjs talpoides Eastern Kingbird ' Tyrannus tyrannus Plains pocket gopher Geomys bursa rius Western Kingbird ' T. verficalis Olive-backed Say's Phoebe ' Sayornis saya pocket mouse Perognatlntsfasciatus Horned Lark'"' F.rcmopliila (dpcstris Hispid pocket moust' P. hispidus Violet-green Swallow'' Tachycincta thalassina Ord s kangaroo rat Dipodonujs ordii Northern rough-winged Plains har\ est mouse Rcithrodontonuis montanus Swallow ' Stclgidoptcryx scrripcnnis Western har\ t-st mouse R. inegalotis Barn Swallow ' Hirundo rustica Deer mouse Perotnyscus manicidattis Cliff Swallow'' H. pyrrhonota Northern grasshopper Blue Jay' Cyanocitta crislata mouse Onychomys Icucogaster 1990] Ecological Review OK Black tailkd Piv\ii{ie Docjs 343 Tabi.k 1 fontiiuR'cl. Prairie vole Nonvay rat House mouse Porcupine Racoon Lon^-tailed weasel Black-footed ferret Mink Badtjer Spotted skunk Striped skunk Coyote Red fox Northern swift fox Bobcat Mule deer White-tailed deer Pronghorn Bison Microtus ochro^aster lidtttts ii(>rv('ireet Significance level (control versus treated)" Treatment Pretreatment (19S3) Posttreatment (1984) Adjusted means Immkdiatkimp ACTS September Treated Control 9.3 16.3 ± 0.9 2.7 4.0 13.0 ± 1.2 ± 5.5 -5.3 -3.3 1.9 3.7 -2.0 ± 2.7'' Posttreatment impacts May Treated Control 17.0 20.3 _^_ 3.1 3.0 5.3 7.7 ± 0.9 ± 1.5 -11.7 -12.7 2.3 4.7 1.0 ± 2.1 0.864 June Treated Control 20.7 21.3 ± 4.3 2.2 0.3 2.7 ± 0.3 ± 2.2 -20.3 -18.7 4.5 4.3 -1.7 ± 1.6 0.795 July Treated Control 10.3 11.0 3.0 3.8 0.0 3.0 ± 0.0 ± 2.1 -10.3 -8.0 I 3.0 5.9 -2.3 ± 1.6 0.726 August Treated Control 9.3 16.3 + 0.9 2.7 0.7 0.3 ± 0.7 ± 0.3 -8.7 -16.0 0.3 3.0 7.3 ± 1.1 0.063 'Raiuiomization test used to detect ditlerenees l)etween pairs of adjusted uieans, after signifieaut F-protectiou at a < . 10. T'reatmeut effects were not significant [P 29.5). tfierefore. statistical significance of contrasts was not determined for September. with prebaited strychnine on deer mice from May through July. Impact of zinc phos- phide in August (F = .027) was greater than that of prebaited strychnine. Comparison of treatment effects between the two strych- nine rodenticides indicated that strychnine alone was more effective than prebaited strychnine for lowering densities of deer mice in June (P= .174). Discussion Of the three rodenticide applications used for prairie dog control, only zinc phosphide consistently lowered deer mouse densities. On these sites zinc phosphide was also most effective in reducing prairie dog burrow activ- itv (Apa 1985). Deer mice consume seeds (Baker 1968, Flake 1973, Sieg et al. 1986) and are susceptible to granular rodenticides. After initial rodenticide treatments, long- term changes in deer mouse populations are associated with habitat changes such as in- creased density of vegetation (Uresk 1985) because of lack of clipping by prairie dogs. Deer mice are adapted to live in more open habitat (Baker 1968, Jones et al. 1983, Mac- Cracken et al. 1985, Agnew et al. 1986), and their numbers decrease with increased vege- tation height and canopy cover. Prairie dog burrows were initially devoid of vegetation before rodenticide application; increased plant canopy cover and aboveground biomass occurred with absence of prairie dogs (Klatt 1971, Potter 1980) and contributed to a de- crease in deer mouse densities. Deer mouse densities were variable over the long-term period with the two strychnine treatments, especially when prebaiting was applied. Deer mouse populations generally increased after treatment with the strychnine only. This increase can be attributed to lim- ited control of the black-tailed prairie dogs (Uresk et al. 1986), which provided and main- tained suitable habitat for deer mice (Agnew et al. 1986). Changes in densities of deer mice may also be attributed to seasonal movements of these animals from other areas (Terman 1968) and possible lower predation. An influx of rodents usually occiured in the spring when yearling deer mice established home ranges (MacCracken et al. 1985), and lower densities in August were due to dispersal of young-of- the-year (Falls 1968, Metzgar 1980). Crabtree (1962) and Marsh et al. (1970) found that zinc phosphide produced a 352 M. S. Deischetal. [Volume 50 response-stimulating odor that proved attrac- tive to small mammals, but strychnine did not have an attractive effect on rodents. Based on these findings, discontinuation of zinc phos- phide for prairie dog control is not recom- mended or required, but land management plans should include considerations for possi- h\e nontarget deer mouse losses. We found that use of strychnine alone or prebaited strychnine generally showed a long-term increase in deer mouse densities. Use of these two strychnine treatments for prairie dog control appears to impose the least threat to nontarget deer mice. While this study addressed direct effects of rodenticides (zinc phosphide, prebaited strychnine, and strychnine alone) on deer mouse densities, impacts on other nontarget small mammals could not be evaluated be- cause of the small populations observed. We suspect that granivores, such as Perognathus spp. and Dipodomys spp., found on prairie dog towns in western South Dakota, may also be affected by rodenticides. Further investi- gations are needed to assess nontarget losses of small mammals other than deer mice. Acknowledgments This study was funded under cooperative agreement IAG-57 with the USDA Forest Service, Rocky Mountain Forest and Range Experiment Station, Nebraska National For- est, USDI Fish and Wildlife Service, the South Dakota Cooperative Fish and Wildlife Research Unit, and the National Pesticide Impact Assessment Program (NAPIAP). Thanks are extended to Nebraska National Forest and Badlands National Park for provid- ing study areas. Literature Cited Agnew, W , D W Uresk, and R M Hansen 1986. Flora and fauna associated with prairie dog colonies and adjaci'iit ungrazed mixed-grass prairie in western South Dakota. Journal of Range Management 39: 135-139. Allen, D. L. 1967. The life of prairies and plains. McGraw-Hill, Inc., New York. 2.32 p|). Apa, a. D. 1985. Efficiency of two hlack-tailed prairie dog rodenticides and their impacts on non-target bird species. Unpublished thesis, South Dakota State University, Brookings. 71 pp. Bakek, R H I96S. Habitats and distribution. Pages 98-122 in J. A. King, ed., Biolog) oi Pcroini/sctis (Rodentia). The American Society of Mammalo- gists. Special Publication No. 2. Bell, H. B , and R W Dimmick 1975. Hazards to predators feeding on prairie voles killed with zinc phosphide. Journal of Wildlife Management 9: 816-819. Carmer, S G 1976. Optimal significance levels for appli- cation of the least significant difference in crop performance trials. Crop Science 16; 95-99. Carmer, S G , and M R. Swan.son. 1973. An evaluation of ten pairwise multiple comparison procedures by Monte Carlo methods. Journal of American Statis- tical Association 68: 66-74. Collins, A, R., J P. Worknl\n, and D W Uresk. 1984. An economic analysis of black-tailed prairie dog {Ciinomt/s hidovicianus) control. Journal of Range Management 37: .358-.361 . Crabtree, D. G 1962. Review of current vertebrate pes- ticides. Pages 327-362 in Proceedings, Vertebrate pest control conference, California Vertebrate Pest Control Tech, Sacramento. Deisch, M S 1986. The effects of three rodenticides on nontarget small mammals and invertebrates. Un- published thesis. South Dakota State University, Brookings. 149 pp. Delsch, M S., D, W. Uresk, and R L Linder 1989. Effects of two prairie dog rodenticides on ground- dwelling invertebrates in western South Dakota. Pages 166-170 in A. J. Bjugstad, D. W. Uresk, and R. H. Hamre, eds.. Ninth Great Plains wildlife damage control workshop proceedings. USDA Forest Service General Technical Report RM-171. Fort Collins, Colorado. 181pp. Edgincton, E S 1980. Randomization tests. Marcel Dekker, Inc. , New York. Falls, J, B 1968. Activity. Pages 543-.567 in J. A. King, ed. , Biology ofPeromyscus (Rodentia). The Amer- ican Society of Mammalogists. Special Publication No. 2. Flake, L. D 1973. Food habits of four species of rodents on a short-grass prairie in Colorado. Journal of Mammalogy 54: 636-647. Green, R. H. 1979. Sampling design and statistical meth- ods for environmental biologists. John Wiley and Sons, New York. 257 pp. Hayne, D. W. 1976. Experimental designs and statistical analyses in small mammal population studies. Pages 3-13 in Populations of small mammals un- der natural conditions, Pymatuning Laboratory of Ecology. Special Publication Series Volume 5. Universit\' of Pittsburgh, Hegdal, p. L., and T. a G.viy 1977. Hazards to seed eating birds and other wildlife associated with sur- face strychnine baiting for Richardson s ground squirrels. EPA report under Interagency Agree- ment EPA-IAG-D4-0449. He(;dal, p. L,, T. a. Gati. and E, C Fite 1981. Sec- ondary effects of rodenticides on mammalian predators. Pages 1781-1793 (ii J. A. Chapman and D. Pursley, eds.. The world furbearer conference proceedings. Hilton, II W , W H Robison, and A H Teshima, 1972. Zinc phosphide as a rodenticide for rats in Hawaiian sugarcane. Entomology 1972: 561-571. Jones, J K , Jr , D M ARMsmoNc. R S Hoefmann. and C. Jones. 1983. Manunals of the northern Great Plains. University of Nebraska Press, Lincoln. 1990] Effects OK Kodkmicidkson Mick 353 Ki.Ai'r, 1j Iv 197K A c()iiii)aiis()ii ol tlic ccolo^} ol attivc aiul al)aiKl()necl hlack-tailrd prairie dou; {Cynoniiis lu(li)iici(iniifi) towns. Uiipuhlishc-cl thesis, ("olo- rado State University, Fort C^ollins. KoKOUD, C. B 1958. Prairie dogs, vvhitcfaces, and hhie grama. Wildlife Monographs 3. 78 pp. MacCr^ckkn, J G , D \V Urksk, .\ni) R A II.anskn 1985. K()dent-\egetation relationships in south- eastern Montana. Northwest Seience 4: 272-278. Maksh.R E ,\\' E Howard. AND S U. Palm.ateer 1970. Etteets of odors of rodentieides and adherents on attracti\eness of Oats to ground S(}nirrels. Jomiial ofWildlife Management 34: 821-825. Mktzcar, L H 1980. Dispersion and numbers in Per- oini/scus populations. Ameriean Midland Natural- ist 103: 26-31. O Meilia. M E , F L Knopf, and J C Lewis 1982. Some consequences of competition between prairie dogs and beef cattle. Joinnal of Range Management 35: 580-585. Potter. R L 1980. Secondary successional patterns fol- lowing prairie dog removal on shortgrass range. Unpublished thesis, Colorado State University, Fort Collins. Romesburg, C. 1981. Randomization tests. Resource Evaluation Newsletter. Pages 1-3 in Technical Article 1. USDl Bureau of Land Management, Denver Federal Center, Denver. RuDD. R L , AND R. E Genelly 1956. Pesticides: their use and toxicity in relation to wildlife. California Fish and Game Bulletin No. 7. 208 pp. Schenbeck. G L 1982. Management of black-tailed prairie dogs on the National Grasslands. Pages 207-217 in R. M. Timm and R. J. Johnson, eds.. Fifth Great Plains wildlife damage control work- shop proceedings. University of Nebraska, Lin- coln. Schitoskey, F , Jr. 1975. Primary and secondary hazards of three rodentieides to kit fox. Journal ofWildlife Management 39: 416-418. SiKc, C 1! , D W Urksk, and H M, Hansen 1986. Sea- scmal diets of deer mice on bentonite mine spoils and .sagebrush grasslands in southeastern Mon- tana. Northwest Science 60: 81-89. Taher, R D , AND 1 McT Cowan 1969. Capturing and marking wild animals. Pages 277-317 in R. H. (Jiles, ed.. Wildlife management technicjues. 3rd ed. The Wildlife Society, Washington, D.C. Tacha, T. C , W D Wahdk. and K P Burnham 1982. U.se and interpretation of statistics in wildlife jour- nals. Wildlife Society Bulletin 10: 355-362. Terman, C R 1968. Populations dynamics. Pages 412- 445 in J. A. King, ed.. Biology of Peromyscus (Rodentia). The American Society of Mammalo- gists. Special Publication No. 2. TiETjEN, H. P 1976. Zinc phosphide: its development as a control agent for black-tailed prairie dogs. U.S. Department of International Fish and Wild- life Service. Special Science Report Wildlife No. 195. 14 pp. Uresk, D. W 1985. Effects of controlling black-tailed prairie dogs on plant production. Journal of Range Management 38: 466-468. Uresk, D W, R M King, A D Apa, M. S. Deisch, and R. L Linder 1988. Rodenticidal effects of zinc phosphide and strychnine on nontarget spe- cies. Pages 57-63 in D. W. Uresk, G. L. Schen- beck, and R. Cefkin, eds.. Eighth Great Plains wildlife damage control workshop proceedings. US DA Forest Service General Technical Report RM-154. Uresk, D W , R M King. A. D Apa, and R L Linder. 1986. Efficacy of zinc phosphide and strychnine for black-tailed prairie dog control. Journal of Range Management 39: 298-299. Wood, J E. 1965. Response of rodent populations to controls. Journal of Wildlife Management 29: 425-438. Received 20 December 1990 Accepted 28 January 1991 Cif.il Basin N.ilni.il(st5(M), HWO. pp. 3ri5-3W) ON THE TYPIFICATION OF OXYTROPIS BOREALIS DC. Stanley L. Wclsli' Abstract. — The statu.s oi the name Oxytropis hoiralis DC^. is levii-wcd as it ai)i:)li('s to Noitli Anu'iican plants. A sunmiaiy of the infraspecilic taxa is presented, and several nonienelatinal eoinhinations are proposed: Oxtjtropis horcalis DC. var. Iitidsonica (Greene) Welsh; O. Ijorealis var. siilplutrca (Pors.) Welsh; O. horcalis DC var. viscida (Nutt. ) Welsh. One new ta.xon, Oxytropis horcalis DC. var. ciiist talis Welsh, is deserihed from Utah and Ni'vada, USA. Preparation of a revi.sionary .summary of the genus Oxytropis DC. for the Flora North America Project necessitates that nomencla- tnral changes and new taxa be presented prior to pnbhcation in that project. The principal reason for this paper involves the nomencla- ture of O. horcalis, a name that has figured in various taxonomic treatments of the genus in North America and elsewhere for more than a centurv (Barnebv 1952, Bunge 1874, Gray 1884, jurtsev 1986, Torrey and Gray 1838, Vasil'chenko, Fedchenko, and Shishkin 1948). The American phases of Oxytropis section GloeocepJiala have passed under a series of names centering on Oxytropis viscida Nutt. ex Torr. & Gray (1838). Since the section Gloeoccphala has circumboreal or at least amphiberingian representation, American workers were almost certain that there was an older name in the Old World literature. Indeed, Barneby (1952) in his revision of the North American species of Oxytropis cited two specific epithets older than that of O. viscida. And Boivin (1967), in his attempt at summarizing the Canadian portion of the section, transferred the infraspecific taxa to O. leucantha (Pallas) Pers. An examination of the type of that species demonstrated that it lacked glands typical of members of the sec- tion Gloeocephala; it was indeed a portion of the O. campestris (L.) DC. sensu lato (Welsh 1972). The transfers to that entity, thus, are incorrect and are merely nomenclat- ural baggage that accompanies the genus in perpetuity. Welsh (1967, 1974) and Welsh et al. (1987) essentiallv followed the lead of Barnebv (1952), who chose a wait-and-see attitude with regard to the earliest name for the North American complex. Examination of the types was necessary prior to a final determination of the (}uestion of an earlier name for the North American materials. Bunge (1874) treated two main sections of glandular oxy tropes, Gloeoccphala and Polyadenia. The main diagnostic feature used in segregation of members of these sec- tions is the arrangement of the leaflets — Glococcpliala having opposite, subopposite, or scattered leaflets and Polyadenia having pseudofaciculate leaflets. Since North Ameri- can viscid oxytropes have both leaflet arrange- ments, but mainly opposite, subopposite, or scattered, it is necessary to review the names of Old World representatives of both Gloeoccphala and Polyadenia. The names O. muricata (Pallas) DC. {PJuica muri- cata Pallas, Reise 3: 318. 1776) and O. mi- crophylla (Pallas) DC. (Phaca microphylla Pallas, Reise 3: 744. 1776) were both pub- lished prior to the next available name in Gloeoccphala, i.e., O. horcalis DC. Authen- tic (probable type) specimens of these and others of the Polyadenia were obtained on loan from the herbarium of the Komarov Botanical Institute herbarium (LE). Neither O. muricata nor O. microphylla seems to be within the concept of the glandular phases of O. horcalis with pseudofaciculate leaflets as they occur in North America. Thus, the earliest name available in section Gloeocephala in North America is O. horcalis DC, which is based on a specimen (Fig. 1) deposited in the Prodromus herbarium at Life Science Museum and Department of Botany and Range Science, Brigham Young University, Provo, Utah 84602. 355 356 S. L. Welsh [Volume 50 I ^ = in EH ^ rr rj =" E — u >^-, Fig. 1. Holotype of Oxi/fropts horealis DC. The specimen i.s at G-DC. 1990] TvlMFl(:ATl()N()l'().\VllU)ri.S HOKKALIS 357 Geneva. Tlie specimen is in poor, hut not terrible, condition, essentially what is ex- pected for many historical t\pes. The speci- men appears never to ha\ e been in ,t!;ood con- dition after its collection. The flowers are crumpled as though they had been wet follow- ing collection, or even following mounting. The (juestion of glandularity was left unan- swered in the description by de Candolle in Prodromus (see below). The need to examine portions of the material was critical as to its nomenclatural importance in North American taxonomy. Detailed photos and fragments of the speci- men were sent for examination through the generosity of Dr. A. Charpin, conservateur at Geneve (G-DC.). Of particular importance among the fragments loaned is a black, hairy bud with calyx teeth still connivent. The teeth are clearly glandular verrucose. Other frag- ments include a portion of a flower and part of a floral bract. The bract, a very long structure not unlike those of many Alaskan specimens, is definitely dorsally glandular also. The plant size and nature of other features, though shat- tered, are well within the morphological lim- its of the group as it occurs in North America. Clearly this material belongs to that portion of the Gloeocephala complex treated by Barneby in 1952 as O. viscida var. suhsuccu- lenta. Having priority, the name O. borcalis must replace O. viscida for North American portions of the complex. The author hopes the transfers proposed below are not additional nomenclatural baggage. Oxytropis borealis DC, Prodr. 2: 275. 1825. O. borealis, subacaulis, pilis scaporum stipularuiiKjiie setosis patulis, petioloruni paucis, foliolis elliptico- lanceolatis suhtus glabris superne pilosis scapi folii longitudine, floribus capitatis, biacteis calycis nigro- hispidissinii longitudine. In terra Tschuktschorum ad Sinum Sancti-Laurentii. Stipulae pallidae. (v.s. Comm. acl. Fisch.) (I.e.). Type locali-H'. — "In terra Tschuktscho- rum ad sinum Sancti-Laurentii," collector not stated. Type. — "e sinu S. Laurentii in terra Tschuktschorum (pays des Tchouktchi) sep- tentrionem versus a fretu Beringii. Legumina diversa a leg. ox. montana. m. [Messien] Fischer 1825" G-DC.!. The specimen cited above is the only one bearing the name O. borealis in the Prodromus herbarium, and it is regarded as the holotype (Fig. 1). The species, as it occurs in North America, consists of a series of mainly intergrading vari- eties as indicat(xl below. They difler in com- pactness of inflorescence, size of flowers, length ol floral bracts, and other features that tend to grade individually and collectively into each other. As intergradation occurs, the taxa within the borcale complex match those of infraspecific taxa in other specific com- plexes in this genus. Presented below is a summary of the in- fraspecific taxa as they occur in North Amer- ica. The writer has examined herbariinn ma- terials from all regions of distribution in the continent. Additionally, he has examined the species in the held from the arctic regions of Alaska, Yukon, and Northwest Territories south to its southern limits in Utah and Ne- vada. Variation is huge in the species as a whole and in the infraspecific taxa. The group has received several interpretations in the past and will undoubtedly be interpreted dif- ferently in the future. Oxytropis borealis DC. var. borealis Distribution; N.W.T. and Alaska; Chukotsk. Oxytropis tiralcnsis (3 suhsucculenta Hook., Fl. Bor.- Anier. 1; 146. 1831. Oxytropis viscida vAr. suhsuc- culenta (Hook.) Barneby, Proe. Calif. Acad. IV, 27; 246. 1952. Type; "Arctic seashore, to the east of the Mackenzie River," Dr. Richardson s.n.; holo- type K. Oxytropis borcalis p Hook. & Arnott. Bot. Beechev Bot. 122. 18.32. Oxytropis campestris var. verrucosa Ledebour, Fl. Ross. 1; .591. 1842. Type; "in terra Tschuktschorum ad sinum Sancti-Laurentii," the collector not stated. The relatively few leaflets, ample flowers, and condensed, copiously hirsute inflores- cence in combination allow this entity to be rather readily identified. It consists, at least in part, of what has passed under the name of O. glutinosa Pors., who excluded the type of "subsucculenta" from consideration in treat- ment of the genus in "Vascular Plants of Conti- nental Northwest Territories, Canada" (Por- sild and Cody 1980). Included within the concept of var. borealis is the O. uralensis 3 subsucculenta Hook., the basis of O. viscida var. subsucculenta (Hook.) Barneby. Oxytropis borealis var. hudsonica (Greene) Welsh, comb. nov. Aragallus huilsonicus Greene, Proc. Biol. See. Wash. 18; 17. 1905. Oxytropis viscida var. hudsonica (Greene) Barneby, Proc. Calif Acad. IV, 27: 245. 358 S L. Welsh [Volume 50 1952. O. viscida ssp. hticlsonica (Greene) Love & Love, Taxon 3L 347. 1982. O. Icucuntha var. imd- sonica (Greene) Boivin, Naturaliste Canad. 94: 76. 1967. Tvpe; Whale River, HutLson Bay; A. P. Low 14272, 24 June 1896; holotype NDG!.' Oxijtropis leucantha var. hiidsonica f. ^alactantha Boivin, Naturaliste Canad. 94: 76. 1967. T\pe: Canada: Franklin District, Melville Peninsula, Repulse Bay, along Nauja River, 27 July 1950, P. F. Bruggeman 52; holotype DAO!. Oxijtropis leucantlia var. leuchippiana Boivin, Natural- iste Canad. 94: 76. 1967. Type: Yukon: White- horse, airport area, steep slope, flowers varying in color from yellow to piuple, abundant, Gillette & Calder318i; lectotype here selected DAO!. This is the phase of the species that occurs in North America mainly east of the Yukon, but with some representation in that province, where it is transitional with both var. viscida and var. siilphurea. The main diagnostic feature involves the short calyx teeth. Oxijtropis borealis var. sulphurea (Pors.) Welsh, comb. nov. O. viscidula ssp. sulphtirea Pors., Bull. Nat. Mas. Canad. 121: 247. 1951. Type: Yukon, Rose-Lapie Pass, shalv cliffs bv waterfall E of Lapie Lake, mile 105 [Canol Road], Pors. & Breitung 10198, 19 July 1944; holotype CAN; isotypes ISC!, S!. Oxijtropis sheldoncnsis Pors., Bull. Nat. Mus. Canad. 121: 246. 1951. Type: Mount Sheldon, on rocky granite ledges at or near the simimit, opposite mile 122 [Canol Road], Pors. & Breitung 11750, 11 August 1944; holotype CAN!; isotypes ISC!, US!. Oxijtropis verriictdosa Pors., Bull. Nat. Mus. Canad. 121: 246, 1951. Type: Yukon: Rose-Lapie Pass, rocky ledges on dry slope W of mile 116 [Canol Road], Pors. 10072, 1944; holotype CAN!; i.sotype S!. These are the pallid-flowered plants of the Yukon and Alaska. In their most typical condi- tion the racemes are compactly and uniformly small flowered. They vary from that norm to elongate racemes with small to large flowers. The bracts are mainly small, but in some they are very long and conspicuous in the inflores- cence. On the one side the plants seem to grade with var. hiidsonica and on the other with both var. viscida and var. borealis. Oxytropis borealis var. viscida (Nutt.) Welsh, comb. nov. O.xytrupis viscida Nutt., e.xTorr. & Gray, Flora N. Amer. 1: 341. 1838. Arag.allus viscidus (Nutt.) Greene, Pittonia 3: 211. 1897. A.stnifialus viscidus (Nutt.) Tidestrom, Proc. Biol. Soc. Wash. 50: 19. 1937. O. campcstris var. viscida (Nutt.) S. Watson, U.S. Geol. Expl. 40th Parallel, Bot. 5: 55. 1871. Spiesia viscida (Nutt.) Kuntze, Rev. Gen. 206. 1891. O. leucantha var. viscida (Nutt.) Boivin, Naturaliste Canad. 94: 77. 1967. Type; Rocky Mountains, near the sources of the Oregon [SW Wvoming], Nuttall s.n. 1834; syntypes NY!, PH. ArafiaUtts viscididiis Rvdb., Mem. N.Y. Bot. Card. 1: 253. 1900. O. Li.s'df/«/rt (Rydh. ) Tidestrom, Contr. U.S. Nat. Herb. 25: 332. 1925. Type; Montana, Melrose, Silver Bow County, Rydberg 2716; holo- type NY! (type specified by Barneby 1952). Ara^alliis viscidula var. depressus Rydb., Mem. N.Y. Bot. Card. 1:523. 1900. Oxytropis leucantha var. depressa (Rydb.) Boivin, Naturaliste Canad. 94: 77. 1967. Type: Haystack Mt., Stillwater County, Montana, Tweedy 120; holotype NY!. Oxijtropis gaspcnsis Fern. & Kelsey, Rhodora 30: 123. 1928. Astrag.alus fiaspcnsis {Fern. & Kelsev) Tide- strom, Proc. Biol. Soc. Wash. 50: 19. 1937. O. leucantha var. gaspensis (Fern. & Kelsev) Boivin, Naturaliste Canad. 94: 76. 1967. Type: Quebec, Mont St. Pierre, Gaspe County, Fernald & Smith 25874, 14 August 1933; holotvpe GH; isotvpes CAS!, NY!. Oxijtropis ixodes Butters & Abbe, Rhodora 45: 2, tab. 745, figs. 1-6. 1943. O. leucantha var. ixodes (Butters & Abbe) Boivin, Naturaliste Canad. 94: 76. 1967. Type: Minnesota, South Fowl Lake, Cook County, Butters, Abbe, & Burns 611, 27 June 1940; holotvpe MINN; isotvpes GH, NY!, PH!, US!. Oxytropis leucantha var. niagnifica Boivin, Naturaliste Canad. 94: 77. 1967. Type: Alberta, Macloed, High River, 27 June 1902, J. Fletcher s. n.; holo- type DAO!. Oxytropis ixodes var. ecaiidata Butters & Abbe, Rhodora 45: 4. 1943. Type: Ontario, Thunder Bay District, Butters, Abbe, & Burns 682; holotype MINN. Distribution. — Alaska, Yukon, N.W.T., Quebec, British Columbia, Alberta, Minne- sota, Oregon, Idaho, Wyoming, Nevada, Utah, Colorado, and California. This variety includes almost as much diver- sity as the species as a whole. The numerous subunits are held together by tenuous charac- teristics that are diflicult to define or place in a key. Variation is often great in subpopulations from adjacent hillsides or even on a single gravel bar, especially in the arctic. One is reminded of the conditions of morphological variation occurring in the boreal O. nigres- cens var. nigrescens, as regarded by this au- thor. Unless one is willing to support a ta.\on- omy wherein the purported ta.\a are largely sympatric and consist of morphological sub- imits whose genetic continuitx' is (luestion- able, made up of a series of similar plants held together by that similarity and not by genetic linkage, there seems to be no reasonable way to segregate the morphological variants as taxa. The rather large number of synonyms, often at specific or \ arietal levels, reflects the attempts at segregation. 1990] TypikicationokOxyihoims Bokkalis 359 Oxytropis horcdlis \ ar. (iiistntlis Welsh, \ar. \\o\ . Siniilis (). horcali \ar. viscida (Nutt.) Welsh seel in llorihus pallidis et iiifloreseentia viilgo fbliis saepe subae(jiialis distinguitiir. Caespitose, aeaiilescent, 6-19 cm tall; pu- hescence basifixed; stipules glabrous to glan- dular or sparingly so; leaves 4-15.5 cm long; leaflets 15-33, 1.5-20 mm long, 1-5 mm wide, oblong to lanceolate or elliptic, spar- ingly pilose to glabrate or glabrous on both sides, sometimes also glandular; scapes 2-16.5 cm long, spreading-hairy; racemes 2- to ll-flowered, the flowers spreading-ascend- ing, the axis 1-3 cm long in fruit; bracts glabrous dorsally, glandular; flowers 11-19 mm long, whitish or rarely suffiised with pink; calyx 5-11 mm long, the shortly cylindric tube 4-7 mm long, the teeth 1.5-3.5 mm long, triangular-subulate, commonly glandu- lar; pods erect, sessile, ovoid to subcylindric, 8-16 mm long, 4-6 mm thick, glandular. Distribution. — Utah and Nevada, USA. Type. — Utah: Sevier Co., open hillside, E of Hogan Pass, along Utah Hwy 72, at 8300 ft. elevation. Flowers white. T25S, R4E, 23 July 1967, S. L. Welsh, D. Isely, & G. Moore 6452; holotype BRY!, isotype ISC!, NY! (a total of 17 duplicates distributed earlier as O. viscida Nutt.). Other collections: Utah: Emery Co., 10 km due W of Ferron, 2 June 1977, E. Neese & S. White 3022; do, E end of Bald Ridge, T16S, R8E, SIO, 11 July 1979, R. Foster. Sanpete Co., 20 km up Ferron Canyon, T19S, R5E, S36, 9 June 1977, S. Clark & K. Taylor 2473; do, Ferron Mt., T20S, R5E, S33, 11 July 1989, M. A. Franklin 6794. Sevier Co., Aspen Spring, Salina Canyon, 18 June 1943, W. P. Cottam 9191; do, 1 km SE of Mt. Hilgard, 25 August 1965, R. Stevens 110; do. Desert View, ca 1.5 km S of Hogan Pass, ca 23 km N of Fremont, T25S, R4W, 10 May 1969, S. L. Welsh, D. Atwood, L. Higgins 8971; do, 21 km due SSW of Fre- mont Jet., T26S, R4E, S4, 8 July 1977, S. L. Welsh 15359; do, head of Clear Creek below Hilgard Mt., T24S, R4E, S26, 30 June 1977, S. Clark 2662; do. Clear Creek ca 3 km SE of Clear Creek Guard Station, T24S, R4E, 10 June 1981, D. Atwood 7947; do, milepost 18 on Utah Hwy 72, T25S, R4E, S22, 31 Mav 1986, R. Kass, E. Neese, B. Neelv 2345; do, milepost 18 on Utah Hwy 72, T26S, R4E, S4, 31 May 1986, E. Neese, B. Neely, R. Kass 17521; do, ca 13 km N of Fish Lake, T24S, R3E, S33, 25 July 1987, B. Franklin & J. & ]. Chandler 4999. Wavne Co., Elk Horn Ciuard Station, T27S, R4E, S15, 17 June 1977, S. Welsh 14982; do, Paradise Valley, T25S, R4E, 24 July 1978, D. Atwood 6922; do, Elkhorn C:ampgr()und, T24S, R4E, S15, 16 June 1986, J. M. Porter 3918; do, on the slopes overk)oking Deep Creek, T27N, R4E, S25, 17 June 1986, J. M. Porter 3863. Nevada: Elko Co., Rubv Mountains, S of Harrison Pass, T28N, R57E, ca S25, 7 August 1967, J. L. Gentry & G. Davidse 1823. Nye Co., Tocjuima Range, Pine Creek drainage, TUN, R45E, 24 July 1964, J. L. Reveal 657; do, Toquima Mts. ca 110 km S of Austin, TUN, R45E, ca S28, 15 July 1973, A. Cronquist 11048; do, Toquima Range, Mt. Jefferson, head of South Fork Pine Creek, TUN, R45E, S29 & S32, 18 July 1978, K. R. Genz 8246; do, north side of Timber Mountain, Grant Range, T6N, R57E, 27 June 1979, M. J. Williams & A. Tiehm 79-109-4. This southern phase of O. borealis, though mainly montane in distribution, occurs mostly on xeric sites in sagebrush, black sagebrush, grass, ponderosa pine, and aspen parkland communities, often on exposed ridges or out- crops. Main substrate types are of igneous origin, either granitic- or basaltic-derived soils, but limestone also serves as a substrate. Elevational range varies from 2135 to 3355 m. The differences cited in the diagnosis are not absolute, as is usual for infraspecific and even specific taxa in this genus. Flower color is typically white or ochroleucous, but some are occasionally tinged with pink; and some that appear to be white when fresh fade slightly lavender on drying. Inflorescences tend to be only slightly longer than the leaves, or even slightly shorter, but some have inflo- rescences much surpassing what appear to be juvenile leaves with tiny leaflets. The herbage is often conspicuously glandular, with sand grains and plant fragments adhering. The stip- ules are occasionally quite glandless, how- ever. In spite of the variation in morphology, these plants appear to represent a xeric south- ern phase related to the typically more mesic var. viscida. That variety, shorn of var. aus- tralis, is not much less polymorphic. There are individual plants, and possibly even sub- populations, within var. viscida that simulate 360 S. L.Welsh [Volume 50 var. australis. Plants from the Wallowa Moun- tains of northeastern Oregon are almost as variable as var. viscida as a whole. Literature Cited Barneby, R. C. 1952. A revision of the North American species oiOxytropis DC. Proceedings of the CaH- fornia Academy of Science IV, 27: 177-.312. BoiviN, B. 1967. Etudes sur les Oxtjtropis DC. II. Natu- rahste Canadienne 94: 73-78. BUNGE, A. 1874. Species Generis Oxtjtropis, DC. Mem- oirs of the Academy of Science, St. Petersburg VII, 22: 1-166. Gray, A. 1884. A revision of the Nortli American species of Oxtjtropis DC. Proceedings of the American Academy of Science 20: 1-7. JURTSEV, B. A 1986. Oxtjtropis DC. Arctic flora S.S.S.R. 9:61-146, 178-182. PORSILD. A E., AND W. J. CoDY. 1980. Oxijtropis DC. Pages 438-442 in Vascular plants of the continen- tal Northwest Territories, Canada. National Mu- seum of Natural Sciences, National Museums of Canada. 667 pp. ToRREY. J , AND A. Gray. 1838. Flora of North America. Vol. 1. G. C. CarviU & Co., New York. Vasil'chenko, I. T., B. A Fedchenko. and B. K. Shishkin. 1948. Oxtjtropis DC. In Flora U.S.S.R. 13: 1-229. Welsh, S L 1967. Legumes of Alaska II: Oxtjtropis DC. Iowa State Journal of Science 41: 277-303. 1972. On the tvpification oi Oxtjtropis leticantha (Pallas) Pers. Taxon 21: 155-157. 1974. Oxtjtropis, DC. Pages 275-282 in Ander- son's flora of Alaska and adjacent parts of Canada. Brigham Yoimg University Press, Provo, LItah. 724 pp. 1987. Oxt/tropis DC. Pages 396-398 in S. L. Welsh, N. D. Atwood, S. Goodrich, and L. C. Higgins, A Utah flora. Great Basin Naturalist Memoir 9: 1-984. Received 5 Febnianj 1991 Accepted 25 Fehruanj 1991 Great Basin Naturalist 50(4), 199(), pp. 3(il-.365 REPRODUCTION OF THREE SPECIES OF POCKET MICE {PEROGNArHUS)lN THE BONNEVILLE BASIN, UTAH Kfiim-tli L. ClraiiuT " and Joseph A. {,'ha|)iiian' Abstract. — Data on reproduction of three species ol pocket mice (Pero^mitlms) occurring; in northern Utah are siMiimarized. Pcn)d in the lai)oratory (Hayden et al. 1966). Our results suggest that poeket miee in northern l^tah generally breed only in the spring although they may produce more than one litter per year. Long-tailed pocket mice and little pocket mice usually have six young per litter, while Great Basin poeket mice usu- ally produce about five young per litter. These data are consistent with previous hterature with the exception of our litter estimates for little pocket mice. Even given our relatively small sample sizes, the large discrepancy (two young per litter) between our field data and previous lab estimates (Hayden et al. 1966) suggests that caution be exercised in extrapo- lating from the lab to the field. This could be particularly misleading when drawing infer- ences from large literature reviews of diverse data sets (e.g., Jones 1985). Acknowledgments This study was funded in part by U.S. Air Force Department of Defense Contract No. F42650-84-C3559 through the Utah State University Foundation. The Department of Fisheries and Wildlife and the Ecology Cen- ter of Utah State University provided vehicles and other logistical and technical support. The Utah Department of Wildlife Resources granted the necessary collection permits. This paper is part of a dissertation completed by the first author in partial fulfillment of the requirements for a doctoral degree in Wildlife Ecology at Utah State University. Literature Cited Brown, J H , and G. A. Lieberman. 1973. Resource uti- lization and coexistence of seed-eating desert ro- dents in sand dune habitats. Ecology 54: 788-797. Brown, L N 1964. Reproduction of the brush mouse and white-footed mouse in the central United States. American Midland Naturalist 72: 226-240. Brown, L. N., and C H Conawav 1964. Persistence of corpora lutea at the end of the breeding season in Peromysctis. Journal of Mammalogy 45: 260-26.5. Cramer, K L. 1988. Reproduction and life history pat- terns oiPeromyscit.s manictilatiis and Pcro^ncithus spp. in the northern Bonneville Basin, Utah. Doc- toral dissertation, Utah State Uuiversitw Logan, I'tah. 101 pp. DliKK, K L 19.57. lieproduclion in i'cro^iuilluix. Journal of Mammalogy 38: 207-210. French, N R , B (; Ma/a, and A P Aschwanden. 1974. A i^opulation study of irradiated desert rodents. Ecological Monographs 44: 45-72. IIai.i,, E R. 1946. The mammals of Nevada. University of ('alifornia Press, Berkeley. Hayden, P , J J Gambino, and R C Lindber(; 1966. Laboratory breeding of the little pocket mouse, Pcri)'VSv-, ♦' ^ '■■-'.V -, '". ■:^-' i,' "•' .-.- >^*.^ "^V--- '■^;^- l*-^:'^ , '^^V^iT 0 10 m from any other vegetation, it is hkely the fungus derived its requisite carbohydrates solely from these seedlings, as most ectomycorrhizal fungi are assumed to rely on the carbohydrates ob- tained through the infection of an autotrophic host for completion of their life cycles and subsequent fruiting body production (Marx and Bryan 1975b). Thus, the apparent lack of ectomycorrhizal formation on the juniper, rose, and peashrub may indicate the develop- ment of a parasitic, or perhaps ectendomycor- rhizal, relationship between P. tinctorius and these hosts in the Leviathan Mine, although there are no reports of this fungus forming either of these relationships with any of its previously identified host species. Acknowledgments This paper contains results of the Nevada Agricultural Experiment Station Research Project 612 funded by the Mclntire-Stennis Cooperative Forestry Research Program. The author is indebted to P. M. Murphy of the Division of Forestry, Nevada Department of Conservation and Natural Resources; A. T. Leiser of the Department of Environmental Horticulture, University of California, Davis; and D. C. Prusso of the Department of Biol- ogy, University of Nevada, Reno, for their invaluable assistance. Literature Cited BuTFERFiELD, R. I.. AND P. T TuELLER. 19S0. Revegeta- tion potential of acid mine wastes in northeastern California. Reclamation Review 3: 21-31. COKER, W. C. AND J. N. Couch. 1928. The Gastero- mycetes of the eastern United States and Canada. University of North Carolina Press, Chapel Hill. 201 pp. HarleyJ. L.andS. E SMrni, 1983. Mycorrhizal symbio- sis. Academic Press, New York. 483 pp. HiLE, N., AND J F Hennen 1969. In vitro culture of Pisolithus tinctorius mvcelium. Mycologia 61: 195-198. Lampky, J R., AND J H. Peterson. 1963. Pisolithus tincto- rius associated with pines in Missouri. Mycologia 55: 675-678. Lampky, S. A , and J R Lami'KV 1973. Pisolithus in cen- tral Florida. Mycologia 65: 1210-1212. Marx, D. II. 1975. Mycorrhi/.ac and estahlishmcnt oi trees on strip-mined land. Oliio journal ol Science 75: 288-297. 1977. Tree host range and world distribution of the ectomycorrhi/al hmgus PisoUtlius tinctorius. Canadian Journal of Microbiology 23: 217-223. Marx, D H , and J. D. Artman. 1979. Pisolithus tincto- rius ectomycorrhizae improve survival and growth of pine seedlings on acid coal spoils in Kentucky and Virginia. Reclamation Review 2: 23-31. Marx. D. H., and W C Bryan. 1975a. Growth and ecto- mycorrhizal development of loblolly pine seed- lings in fumigated soil infested with the fungal svmbiont Pisolithus tinctorius. Forest Science 21: 245-254. 1975b. The significance of mycorrhizae to forest trees. Pages 107-117 in B. Bernier and C. H. Winget, eds.. Proceedings of the fourth North American forest soils conference, August 1973, Laval University, Quebec. International Scholarly Books Services, Portland, Oregon. 67.5 pp. Marx. D H . W C Bryan, and C. E Cordell. 1976. Growth and ectomycorrhizal development of pine seedlings in nursery soils infested with the fungal svmbiont Pisolithus tinctorius. Forest Science 22: 91-100. Marx, D H , C E Cordell, D S Kenney. J G Mexal, J. D. Artman, J W. Riffle, and R. J. Molina. 1984. Commercial vegetative inoculum of Piso- lithus tinctorius and inoculation technicjues for development of ectomycorrhizae on bare-root tree seedlings. Forest Science Monograph 25. 101 pp. Marx, D. H., C E Cordell, S B. Maul, and J. L. RuEHLE. 1989a. Ectomycorrhizal development on pine by PisoUtlius tinctorius in bare-root and con- tainer seedling nurseries. I. Efficacy of various vegetative inoculum formulations. New Forests 3: 45-56. 1989b. Ectomycorrhizal development on pine by PisoUtlius tinctorius in bare-root and container seedling nurseries. II. Efficacy of various vegeta- tive and spore inocula. New Forests 3: 57-66. Medve, R. J., F. M Hoffman, andT. W. Gaither. 1977. The effects of mycorrhizal-forming amendments on the revegetation of bituminous stripmine spoils. Bulletin of the Torrey Botanical Club 104: 218-225. Schramm. J. R. 1966. Plant colonization studies on black wastes from anthracite mining in Penns\lvania. Transactions of the .American Philosophical Soci- ety 56: 1-194. Tuc:ker, J. M , W. P. Cottam, and R Drobnick. 1961. Studies in the Quercus undulata complex. II. The contribution of Quercus turhinclla. .\merican Journal of Botany 48: .329-.3.39. Walker. R F 1989, Pisolithus tinctorius. a Gastero- mycete, associated with Jeffrey and Sierra lodge- pole pine on acid mine spoils in the Sierra Nevada. Great Basin Naturalist 49: 111-112. 1990. Formation of Pisolithus tinctorius ecto- mycorrhizae on California white fir in an eastern Sierra Nevada minesoil. Great Basin Naturalist 50; 85-87. Walker R. F., D C. West, S B M<:Lau(;iilin, and C C. Amundsen. 1989. Growth, xylem pressure potential, and nutrient absorption of loblolly pine on a reclaimed surface mine as allected b>' an induced Pisolithus tinctorius injection. Forest Science 35: 569-.581. Received 10 Noveniher 1990 Accepted '2H}aniianj 1991 Ciri-at Basin Naturalist 50(4), 1990, pp. ,371-372 NATURAL HYBRID BETWEEN THE GREAT PLAINS TOAD (B(7FOCOCNAr(:/S)ANDTHE RED-SPOITEDTOAD (/^l^^C; PC/NC7ArL7S) FROM CENTRAL ARIZONA Brian K. Sullivan' Hybridization among toads of the genus Bufo is well known (Sullivan 1986). In the southwestern United States hybridization has been documented within both the ameii- canus and punctatus species groups (Fergu- son and Lowe 1969, Sullivan 1986). However, natural hybridization between members of more distantly related species groups is rela- tively rare. Natural hybrids between Bufo punctatus and both B. borcas (Feder 1978) and B. woodJiousii (McCoy et al. 1967) have been described. Documentation of hybridiza- tion is important because it provides informa- tion about the genetic relatedness of taxa, as well as potential insights into proximate as- pects of species recognition and reproductive behavior. Herein I report on a natural hybrid between B. cognatus (Great Plains toad) and B. punctatus (red-spotted toad), members of separate species groups within the "thin- skulled lineage of North American toads. The hybrid male was collected in a rain- formed pool on the evening of 13 August 1990 at the Cave Buttes Recreation Area along ' Cave Creek, 12 km southwest Jt^ Ga\^ Creek, Maricopa County, Arizona. Approximately 50 mm of rain fell from 11 to 13 August; on each evening, a number of anurans called along a narrow (3-m wide), flowing stream and a large, shallow pool (40-m diameter) created by an earthen dike across the stream channel. Male B. alvarius, B. cognatus, and Scaphio- pus couchii called from the pool, while male B. punctatus were restricted to the channel. The hybrid was calling among the male B. cognatus at the large pool. I recorded a series of its advertisement calls using a Marantz PMD 430 cassette recorder and Sennheiser ME-80 microphone, and I mea sured its cloacal temperature with a Weber Quick Recorder thermometer. Five calls were analyzed with a DATA Precision 6000 Waveform Analyzer (see Sullivan 1989 for details), and mean values were calculated for each call variable. Data are reported as the mean ± standard deviation. The mean pulse rate of the advertisement call of the hybrid was 45 p/s, and the mean duration was 7.8 s, at a cloacal temperature of 24 C. The corresponding values for 13 B. punctatus recorded on the same night were 55 ± 2.60 p/s and 6.4 ± 1.23 s (cloacal tem- peratures = 25 ± .34 C); the values for 8 B. cognatus were 24 ± 1.66 p/s and 18 ± 7.58 s (cloacal temperatures = 25 ± .82 C). The dominant frecjuency of the hybrid's ad- vertisement call was 2.109 kHz, lower than both B. punctatus (2.538 ± .111 kHz) and B. cognatus (2.700 ± .207 kHz). Hence, the advertisement call of the hybrid, although more similar to that of B. punctatus, was in- termediate in pulse rate and duration and dra- matically lower than either parental species in dominant frequency. However, the vocal sac was darkly pigmented and sausage-shaped when inflated (Fig. 1), the condition typical of B. cognatus. The hybrid was intermediate in size (63 mm snout-vent length) relative to B. punctatus (54 ± 2.59 mm) and B. cognatus (75 ± 4.83 mm). The oval parotoid glands, enlarged cra- nial crests, and boss of the hybrid were also intermediate to B. cognatus and B. punc- tatus. Following the methodology of Fergu- son and Lowe (1969), I determined four ratios (parotoid length/parotoid width, svl/parotoid width, parotoid length/eyelid length, tibia/ parotoid length) for the two parental species 'Life Sciences Program. Bo.\ .37100, Arizona State University West, Phoenix, Arizona H.50H9-710U. 371 372 Notes [Volume 50 Fig. 1. (a) Bufo cognatus, (b) hybrid, and (c) Bufo punctatus from Cave Biittes Recreation Area, Maricopa, County, Arizona. and the hybrid. All of the ratios calculated for the hybrid were between the mean values and exclusive of the 95% confidence intervals for the two parental species. Unfortunately, the hybrid escaped after these observations were completed. Docu- mentation of a natural hybrid between mem- bers of these two distinct species groups is noteworthy. Although B. cognatus and B. punctatus typically breed in dissimilar habitats, the present observations reveal that they may interact if they breed sympatrically, and that they can produce hybrid offspring. Additional work will be recjuired to determine the evolutionary importance, if any, of such interactions. Acknowledgments I thank Ken Johnson of the Maricopa County Flood Control office for providing as- sistance in the early stages of my work. This investigation was supported by a Faculty Grant in Aid award and a Summer Research Grant from Arizona State University. Literature Cited Feder. J H 1979. Natural hybridization and genetic di- vergence between the toads Bufo horeas and Bufo punctatus. Evolution 33: 1089-1097. Ferguson, J H., and C. H Lowe. 1969. Evohitionary relationships in the Bufo punctatus group. Ameri- can Midland Naturalist 81: 435-466. McCoyC J, H M Smith,.\ndJ A TiHEN. 1967. Natural hybrid toads, Bufo punctatus x Bufo woodhousei, from Colorado. Southwestern Naturalist 12: 45-.54. Sullivan, B. K 1986. Hybridization between the toads Bufo microscaphus and Bufo woodhousei in Ari- zona: morphological \'ariation. Journal of Her- petology20: 11-21. 1989. Interpopulational variation in vocalizations of Bufo wooclhousii. Journal of Herpetolog>' 23: 368-373. Received 18 December 1990 Accepted 28 January 1991 Great Basin Naturalist 50(4), 199(), pp. 373-37 NEW VARIETY OF OXYTROPIS CAMPESTRIS (FABACEAE) FROM THE COLUMBIA BASIN, WASHINGTON Elaiiu' Joyal' In 1984 I found an Oxytropis in central Washington that I was unable to identify. Col- lection was made and sent for determination to Rupert Barneby, who puzzled over it for some time before concluding that it lacked a published name. What follows is a description of that taxon. This is a rare taxon, presently known from a single population on an isolated mountain. Habitat and ecological notes are included, therefore, to facilitate understand- ing of the taxon's conservation status. Oxytropis campestris (L.) DC. var. wanapum ]oyd\, var. nov. Fig. 1 O. cat7ipestri(L.) DC. var. gracili{A. Nels.) Barneby affinis, plantis dense sericeo pilosis, robustis, foliolis 20-25, corollis lavandulis, carinis maculatis, diflfert. Caespitose perennial, acaulescent, 17-30 cm tall; herbage silvery, densely silky-pilose to villous; stipules membranous, pilose to densely pilose, the blades free for half their length, (5) 6.5-9 (16) mm long, margins ciliate to densely ciliate; leaves (11) 14-18 (22) cm long, with (13) 19-26 (32) linear to narrowly oblong leaflets, (8) 15-25 (33) mm long, scat- tered, sub-opposite; scapes erect to spread- ing, (10) 17-21 (30) dm long, pubescence spreading-appressed; racemes in part exceed- ing the leaves, (5) 6-12 (17) flowered, con- gested in flower, (4) 6-8 (12) cm long in fruit; calyx sericeous-pilose, 7-9 mm long, greater than half the length of the corolla, with a few dark hairs, the tube 5—7 mm long, the teeth (1) 2-3 mm long, linear-lanceolate; corolla pale lavender with darker penciling, keel maculate, drying blue; banner obovate, 14-16 (23) mm long; wings 13-15 (19) mm long; keel (10) 11-14 (17) mm long; pod sessile to short-stipitate, erect, 1-celled with the su- ture not or only slightly intruded, the wall membranous-leathery, 10-20 mm long, beak about 6 mm long. TYPE: United States; Washington, Grant County, Saddle Mountain, above Lower Crab Creek' and E of Beverly, T15N, R24E, S2, Nl/2, elev. ca 550 m, NNE aspect at crest of ridge, in sandy (volcanic ash) soils above steep basalt talus, 25 May 1987 (flower and early fruit), Joyal 1264 (Holotype: US; Isotypes: BRY, CAN, CAS, ISC, K, MO, MONTU, NY, OSC, S, UBC, WS, WTU). PARATiTE: United States: Washington, Grant County, Saddle Mountain, above Lower Crab Creek, T15N, R24E, S2, elev. ca 550 m, NNE aspect at crest of ridge, in sandy soil, 15 May 1984 (flower), Joyal 467 (BLM— Spokane, NY, OSC). There are currently at least 10 varieties of O. campestris recognized in North America (Barneby 1952, Elisens and Packer 1980, Welsh, personal communication). Characters used to distinguish the infraspecific taxa are: length of leaves and of scapes, number of leaflets, numbers of flowers per raceme, length and density of flowers in raceme, color of corolla, habitat, and distribution. There are notable differences in these characters in variety wanapum when compared with other varieties of O. campestris. The three varieties that occur in eastern Washington with var. wanapum are compared below; a key also is provided to separate these four varieties (see Barneby 1952, Elisens and Packer 1980, Welsh, personal communication, for compari- son with other O. campestris varieties). On the average, plants of var. wanapum are more robust and have a greater number of leaflets. The length of the leaves (16 cm) averages Department of Botany, Arizona State University, Tempe, Arizona 85287. 373 374 Notes [Volume 50 Fiji. I. Oxytropis campestris vAr. wanapimi. IIal)it. Flower (har 1 tin). C()mi)()sitc draw iiiti from Joval 467, 1264, and photos of the Satlclk' Mountain population. 1990] Notes 375 greater than those of var. columhiana (St. John) Harnehy, cusickii (CIreenni.) Barnel)y, and ^'/7/r///.v (A. Nels.) Barnel)y (11, 6, and 11 cm, respectively); leallet len,ti;th (20 mm) like- wise averages greater than tliose of the other three varieties, (14, 8, and 12 nun); scape length (20 cm) is also greater than the other three (17, 7, and 16 cm); the mean nnmber of leaflets (22) is greater than the first two vari- eties (each = 15) and within the range of the third variety (17); the average number of flow- ers per raceme (8.5) is within the range of three related taxa (8-12), with columhiana and gracilis occasionally having as many as 30 flowers/raceme; keel length (12 mm) is similar for all four taxa, with gracilis showing slightly larger dimensions; the pale lavender flower color, while not unique in the group, is un- known among northwest members of O. campestris. Some of these differences might lie explained as a phenotypic response of a primarily montane taxon to a desert environ- ment. Its desert habitat sets this taxon apart from its close relatives in nearby moimtains; precipitation is about half that of the moun- tains (20 cm vs 40 cm/yr), the climate is warmer, the vegetation is shrub-steppe rather than forested, the geologic substrate is sedi- mentary and volcanic rather than intrusive with some volcanic rocks, and it lies south of the glaciated portions of the Okanogan High- lands and Cascade Range. Key to clo.sely related varieties of Oxijtropis campestris in the Paeific Northwest (after Hitehcoek and Croncjuist 1973) 1 Corolla white with maculate keel; leaflets 12-17 (23); in wet gravel along the Columbia River in Washington (historically) and near Flathead Lake, Montana var. cohnnhiaiiii r Corolla other than white with maculate keel; leaflets often more than 17 2 Stipules glabrous or glabrate; scapes rarely greater than 15 cm; leaflets seldom greater than 17; range of var. gracilis, but not above 2000 m elevation and not west of the Cascades var. cusickii 2' Stipules very hairy; scapes mostly greater than 15 cm; leaflets generally greater than 17 3 Corolla ochraleucous or white, keel rarely mac- ulate; leaflets 15-20; plants averaging smaller than the next, scapes averaging 16 cm; usually montane plants (in prairies east of the Rocky Mountains); western Washington to Alberta and South Dakota, south in Rocky Mountains to Col- orado var. gracilis 3' C^orolla pale lavender with darker penciling, maculate keel, drying blue; leaflets 20-25; plants larger than the preceding, scapes averag- ing 20 cm; desert plants; at low elevation in the (Columbia Basin of central Washington var. wanapum Elisens and Packer (1980) most recently treated the O. campestris complex in north- western North America. They introduced new cytological information for several of the taxa in this difficult complex; on the basis of these data they reelevated several taxa, in- cluding the eastern Washington var. colum- hiana, to full species status. While accepting their findings, I do not see that it necessarily follows that taxa such as O. campestris var. "columhiana^ should be given specific status based on Elisens and Packer's new data. More importantly, Barneby (personal communica- tion) and I agree that it is preferable to treat the undescribed taxon in a conservative fash- ion and place it at what we consider the appro- priate rank as a variety of O. campestris, near var. gracilis. It may well be that future studies (Welsh, personal communication) in the O. campestris complex will result in this entity being raised to a higher rank. However, until that work is completed, varietal status under O. campestris seems more appropriate. Oxijtropis campestris var. wanapum is presently known only from Saddle Mountain in the Columbia Basin of central Washington (Fig. 2). Saddle Mountain is an isolated east- west trending ridge formed from a partly faulted anticline that stretches approximately 50 km, being cut by the Columbia River at Beverly. Several ranges to the southwest con- ceivably may contain habitat suitable for var. wanapum. Whereas the north slope is steep basalt talus, the south slope is gentle and sandy and dominated by Artemisia tridentata. The Oxijtropis grows in a narrow band of deep sand, derived from volcanic ash, slightly be- low the crest of the north-facing ridge. The community is very open, as is typical of many sandy habitats. It is dominated by Chryso- thamnus nauseosus. Salvia dorrii, Monar- della odoratissinm, Agropyron spicatum, and Bromus tectorum. Other species present in- clude Achillea millefolium, Arenaria frank- linii. Astragalus caricinus, A. purshii, Cas- tilleja cf. tJiompsonii, Chaenactis douglasii, Comandra umhellata, Crepis modocensis, Cryptantha pterocarya, Erigeron linearis, 376 Notes [Volume 50 oo 0«yl ropis campeslfis * vat Columbiana • »ar cusickii O var gracilis o vat wanapum 50 km , Fig. 2. Distribution of Oxt/tropis campestris varieties in eastern Washington and adjacent Oregon. Data points are of representative specimens from MO, OSC, US and WTU. Elevations in the Coknnliia Basin physiographic province average less than 500 m (unshaded); those in the Blue Mountain, North and South Cascade, and Okanogan Highland physiographic provinces average greater than 500 ni (shaded). Eriog,onum tnicrofhecum, E. ovaUfoIiiim, EriophyUiim latiatiim, Galium nutltiflorum, Gilia sinuata, Hackelia arida, Lupinus sp., Penstemon richardsonii, and Poa sp. (taxon- omy follows Hitchcock and Cronquist 1973). No other Oxytropis spp. were noted in the immediate vicinity. Plants of this taxon are frequent (several hundred individuals) in this restricted area, and there is a good size-class distribution of individuals. I observed seedlings, which I presumed to be from the previous year, small vegetative individuals, and flowering plants. The largest plants had many flowering stems (as many as 48 stems per plant observed) and covered areas up to 0.5 meter across. This Oxytropis flowers profusely. My first collec- tion of the taxon was made at peak flowering, in the middle of May 1984, an average season with respect to temperature and precipita- tion. During my second visit in late May 1987, an early and dry spring, I found the plants mostly past flower and well into fruit. The flowers of this Oxytr()})is are held at a 45-degree angle from the rachis, or higher, becoming erect in fruit. The only floral visitors I observed were several iridescent blue-green metallic-leafcutter bees {Osmia Integra Cresson, Hymenoptera: Megachili- dae), working Oxytropis flowers on the upper slope. The pods have a short pubescence and redden as they mature. Seed set appeared to be good, but predation of seed pods was high. Some pods had their sides chewed out in a pattern typical of departing larvae with not a single seed remaining within; more often the upper one-half or one-third of the pod had been eaten away entirely, along with all de- veloping seeds. No larvae were observed, but several small weevils collected from the pods were identified as species ofTychius (Coleopr tera: Tychinae). Oxytropis campestris var. wanapum occurs on land that is a "checkerboard" of Bureau of Land Management (BLM) and private lands. The primary land use is grazing; some recre- ational vehicle use occurs on the mountain as does natural gas exploration. The area in which the Oxytropis grows is isolated by low rimrock from the bulk of the grazing activity to the south. The BLM s Spokane District is treating the ta.xon as a "sensitive species. The varietal epithet honors the Wanapum tribe, who originally called Saddle Mountain and the desert surrounding it home. The Wanapum, except for one small community on the south side of the mountain, have mostly disappeared from the landscape. Acknowledgments The original discovery of this taxon was made while I was employed by the Spokane District Office of the Bureau of Land Manage- ment, U.S. Department of Interior. Gary Par- sons identified insects; Richard Rust provided the specific epithet for Osmia; Kay Thorne did the illustration; Kenton Chambers of Oregon State University provided work space and herbarium support on a regular basis during my western tenure; the Smithsonian Institu- tion staff has generously allowed me use of their herbariinn; and Rupert Barneby and Stan Welsh provided valuable comments on the manuscript. I am especially indebted to Rupert Barneby, for whom I first thought of collecting this taxon. It was he who later con- firmed why I was imable to put a name on it, and who encoinaged me to write this paper. 1990] NoTKS 377 Literature Cited "nrf '""S'lKii"'^!;^'""'"" '""'^"" ^'"™'' oi Botany 5H: io2l)-ioJl. r 1 Ni .1 * line hc(k:k C L and A. Cronquist. 1973. Flora of BAKNEBVRC. 1952. ArevisionoftheNorth AnKM.can '"'^ '\,^^, i,^^,,.,. N,,thwest. Un.Versity of Washington species oiOxytropis DC. Proceedings of the Cah- fornia Academy ofScience IV, 27: 177-312. mss, :,cattic. EusENs, W J , AND J G PACKER. 1980. A Contribution to Reccwccn) January 1991 the taxononnofthe0.v,yf,.>,;<.vc«,»,>r.^ri.v complex Accepted 2S January 1991 Great Hasiii Naturalist 5()( IK 1990. pp. :i79-3,S() EFFECT OF BACKPACK RADIO TKANSMITTEH ATIACHMENT ON CHUKAR MATING Bartfl T. Slaii^li , Jt'iran T. l-'IiTuln-.s', Ja\ A. liohcr.son", and N. I'aiil loliiiston' Results of a previous study (Slaugh et al. 1989) indicate that backpack radio transmitter attachment is more compatible with Chukars {Alectoris chukar) than is a poncho apparatus. It appears, though, that backpacks, especially the antenna angle, could inhibit C'hukar mat- ing. The objective of this study was to deter- mine the effects, if any, of backpacks and an- tenna position on mating and fertility. Materials and Methods Chukars were housed (as pairs or trios) in 45-cm-high x 75-cm-wide x 90-cm-long wire cages. Six pairs had no radio transmitters attached (group I). In group II each of six cages contained one male and one female without radios plus one female with a simu- lated backpack radio with the antenna angled downward along the tail (Fig. 1). In group III each of six cages contained one male and one female without radios plus one female with a simulated backpack radio with the antenna angled upward (Fig. 1). The purpose was to determine if the males would prefer to mate with the females without radios and exclude the females with radios. Eggs were collected from females for one week prior to exposure to males to ascertain fertility status. Females were exposed to males for four days and then separated and caged individually to facilitate individual fertility observations. Eggs were collected for one week, incubated for one week, and then opened to determine fertility. Results and Discussion Females in all groups produced fertile eggs, indicating that males did not exclude radio- Fig. 1. Backpack attachment of simnkited radio trans- mitters with antenna angled downward (left) and upward (right). attached females from their mating. The radios and antennae did not impair mating even when antennae were angled upward. Males were observed to either straddle the antenna or grasp it with a foot and bend it downward while mating. These results in- dicate no mating problems with captive Chukars fitted with radio transmitters. Their behavior, however, could possibly differ in the wild. This study did not include any field obser- vations of mating or fertility. The only prob- lem observed with released Chukars carrying backpacks was that, with the antenna angled upward, some birds experienced difficulty in flying as a result of a wing coming in contact with the antenna. Attachment one week prior to release (to allow time to become accus- tomed to radios) did not affect flight ability or survival. Department of Botany and Range Science, Brigham Young University. Provo, Utah 84602. ^Utah Divi.sion of Wildlife Resources, 1596 West North Temple, Salt Lake City, Utah 841 16. Department of Animal Science, Brigham Young University. Provo, Utah 84602. 379 380 Notes [Volume 50 Acknowledgments Literature Cited This research was funded by the Utah Divi- Slaugh, B T , J T Flinders, J A Roberson, M R. f\x7-iJi-f B^.^,,.-^^c Olson.andN P.Johnston 1989. Radio transmit- Sion of Wlldhfe Resources. ^^^ attachment for Chukars. Great Basin Natural- ist 49; 632-636. Received 15 July 1990 Accepted 15 January 1991 Great Basin Naliiialist 50(11. 19W), pp. 381-383 FOOD CACHING AND HANDLING BY MARTEN Steijlien K. Ilcm) , Martin (J. Ilajjliai'l", and Lt-onard F. Huggiero' Various studies provide evidenee of food caching by marten {Maiies americana). Marten have been seen uncovering or retriev- ing food items (Murie 1961, Simon 1980, Buskirk 1983), but whether these items were initially cached by marten was unknown. Hawbecker (1945) and Thompson (1986) doc- umented food concealment by marten, but neither reported subsequent recovery of prey. Due to lack of evidence, Stordeur (1986) concluded that caching of food is un- common in marten. Prey caching has impor- tant implications for foraging frequency and energetics of marten. Study Area and Methods The primary objective of our research was to quantify changes in marten home range characteristics and habitat use following the fragmentation of a subalpine coniferous for- est. An ancillary research objective was to describe the characteristics of marten resting sites. Our study area was in the Medicine Bow National Forest, 18 km south of Encamp- ment, Wyoming. The area was characterized by stands of lodgepole pine (Pinits contorta), Engelmann spruce (Picea engelmunnii), and subalpine fir (Abies lasiocarpa). Small mead- ows and rock outcrops were interspersed throughout the area. Elevations at the obser- vation sites ranged from 2935 to 3387 m. Most observations were made during field efforts to locate resting sites of radio-collared marten. Resting sites were defined as loca- tions in which a marten remained stationary and inactive for at least 0.5 h. The radio-signal strength was monitored for 0.4-1.5 h from a distance of at least 70 m. After the signal indicated inactivity, the potential resting site was quietly approached on foot to avoid alert ing the marten or causing it to flee. Precau- tions were made to minimize the observer's influence in order to maximize observations of natural behavior. These precautions in- cluded reduction of receiver volume, conceal- ment of the observer, and removal of shoes if necessary. Observations Case L— 9 June 1987, 1445 h. Adult male marten M3 was seen carrying the hind half of a snowshoe hare (Lepus americanus) for about 20 m near a known resting site. He cached the hare under a leaning stump and then foraged within 400 m for about 0.5 h before returning to the hare and carrying it away. Case 2.-8 July 1987, 1015 h. M3 was seen foraging in and around a rock outcrop. After 5 min the marten emerged from the rocks, grasping a juvenile yellow-bellied marmot (Marmota flaviventris) by the neck. He im- mediately carried the prey approximately 550 m, deposited it in a rock crevice in a road fill, and then left the site. There were six marten scats at the entrance of this den, indi- cating prior use by this or other marten. Case 3.— 1 September 1987, 1900 h. M3 was found resting in a bushy-tailed wood rat {Neotoma cincrea) nest in a rock outcrop. He growled a few times and then ran away, carry- ing an unidentified mammal. Case 4.-9 September 1987, 1730 h. M24 was found resting in a rock outcrop. As the observer approached, the marten peered from a crevice before disappearing back into the rocks. After a few seconds he emerged and fled, carrying a chipmunk (Tamias spp.). Case 5.-22 September 1987, 1300 h. M3 was seen feeding on a freshly killed Blue Grouse (Dendragapus obscurus) next Rocky Mountain Forest and Range Experiment Station, 222 .South 22nd St. , Laramie, VVvoming 82070. "Pacific Northwest Forest and Range Experiment Station. .3625 93rd Ave., S. W,, Olympia, Washington 98502. 381 382 Notes [Volume 50 to a large log. The marten carried the grouse about 40 m and cached it in branches of a recently felled pine, then retreated to a nearby resting site under a different log. Two Blue Grouse feathers, one Northern Flicker (Colaptes auratus) feather, and two Gray Jay {Perisoreus canadensis) feathers were found at the entrance to the resting site. After 25 min the marten ran from the den, took the grouse, and headed downslope. Case 6.-27 July 198S, 1605 h. The observ- er heard F28 killing a juvenile Blue Grouse in an alder (Alnus tenuifolia ) bog. A few minutes later she was observed eating the grouse in- side a hollow log approximately 25 m from the kill site. During this time an adult grouse was heard giving the brood-gathering call. When one of the young responded with a call, the marten left the dead grouse in the log and stalked the live young. It located a young grouse in a tree and made an unsuccesshil attempt to catch it. Upon returning to the original prey, the marten saw the observer and left without the cached grouse. The prey had been removed from the log by 1400 h the following day. Case 7.— 31 January 1989, 1240 h. M35 found a piece of beaver meat (trap bait) at our field camp. He carried the meat 15 m away, climbed up a tree, and moved out onto a limb heavily laden with snow. He dug a hole in the snow, placed the meat in it, and then covered the meat with snow before descending the tree. He continued to move about the camp area, searching for additional food. Case 8.-26 July 1989, 0710 h. A hidden observer witnessed M35 cache a red squirrel (Tamiasciuriis hudsonicus) under a shelter at the field camp. After caching, the marten im- mediately left. Approximately 12 h later (1930 h), M35 was observed retrieving the squirrel. Case 9.— 3 August 1989, 1050 h. Red scjuir- rels were heard scolding F37 at their midden. At 1 1 15 h the marten was seen at the base of a snag 150 m from the midden. She had a s(}uir- rel forearm in her mouth as she ascended into the broken top of the snag. Within moments she descended without the forearm and leit the area. Inside the snag was found the scjuir- rel forearm and the hind one-third of a scpiir- rel. Within an hour the marten was out of telemetry range. Discussion For our purposes "caching" is defined as the act of concealing food for later consump- tion. Marten meet the criteria for "cachers" (Macdonald 1976); viz., they are solitary hunters with fixed home ranges, and they are not large enough to protect their prey from larger scavengers. Our observations show that marten will cache large prey items, and cases 1, 5, and 8 are rare documentation that the same individual that made the cache had sub- sequently returned to it. In addition, these observations show that sometimes the cache site also serves as a resting site or den. We have also documented hunting behavior that is generally associated with surplus killing. Small rodents are consumed quickly by marten, and they are not necessarily removed from the kill site (Pulliainen 1981b). How- ever, it has been reported that marten readily carry larger prey from 9 m to several hundreds of meters awav from kill sites (Murie 1961, Hargis 1981, Pulliainen 1981b, Raine 1981, Spencer and Zielinski 1983). Our observa- tions (cases 1, 2, 5, 7, 8, 9) demonstrate that marten cache food at varying distances from kill sites, especially if the prey is too large for one meal. We suggest that the removal of prey from capture sites may provide security for marten. The noise of the pursuit and kill (e.g. , cases 6 and 9) and the distress calls of the prey could alert competitors or predators to the location of a marten and its kill. This is consis- tent with observations made by Simon (1980), who found that marten typically consume food in secluded cover. Cases 2, 3, and 4 are similar to the find- ings of Pulliainen (1981a), Raine (1981), and Spencer (1981), who also observed that marten sometimes carry prey items to their resting sites. The selection of a resting site mav depend upon the proximitv to the kill site (Marshall 1951, Buskirk 1984) and the amount of protection afforded. When a marten uses a specific rest site on consecutive days (e.g., Steventon 1979), it may be because of cached food. In Finland, Pulliainen (1981b) found sur- plus killing of pre\ 1)> European pine marten (Af. martes). Our observations (cases 1, 6, 7, 8, 9) suggest that marten participate in surplus killing. Animals that we observed resinned an ajiparent foraging actixity after caching. 1990] NOTKS 383 Marten meet the criteria tor species prone toward smplns killing, suggested by Oksanen et al. (1985), because they are small members of a predator guild in a cool, dry environment (at least throughout portions of their range). However, there is no evidence suggesting that marten are involved in siuplus killing or hoarding to the same extent as other muste- lids (e.g., Johnsen 1969 [as cited in Oksanen 1983] reported a stoats [Mustela erminea] single cache of 153 lemmings and a shrew). Important knowledge of marten ecology would be gained if researchers could devise a way to examine the interior of resting sites to determine if food caches vary seasonally. Acknowledgments We thank S. W. Buskirk, R. T. Reynolds, and W. D. Spencer for comments on an earlier draft, and K. O. Christensen and L. J. Kelly for manuscript preparation. Literature Cited Buskirk, S. W 19S3. The ecology of marten in southcen- tral Alaska. Unpublished dissertation, University of Alaska, Fairbanks. 131 pp. 1984. Seasonal use of resting sites by marten in southcentral Alaska. Journal of Wildlife Manage- ment 48: 950-953. Hargis, C. D. 1981. Winter habitat utilization and food habits of pine martens in Yosemite National Park. Unpublished thesis. University of California, Berkeley. 57 pp. Hawbecker, a. C. 1945. Activity of the Sierra pine marten. Journal of Mammalogy 26: 435. Macdonald, D. W. 1976. Food caching by red foxes and some other carnivores. Journal of Tierpsvchologie 42: 170-185. Marshall, W. H. 1951. Pine marten as a forest product. Journal of Forestry 49: 899-905. MuRlE, A. 1961. Some food habits of the marten. Journal of Mammalogy 42: 516-521. Oksane.N.T 19S.3. I'rcN caching in tlic hunting strategy of small mustelids. Acta Zoologica Fennica 174: 197-199. Oksanen, T, L Oksanen. and S D Fretwell. 1985. Surplus killing in the hunting strategy of small predators. American Naturalist 126: 328-346. PULLIAINEN. E. 1981a. Winter habitat selection, home range, and movements oi the pine marten {Marte.s martes) in a Finnish Lapland forest. Pages 1068- 1087 in J. A, (IhajMnan and D. Pursley, eds.. Worldwide lurbearer conference proceedings, Frostburg, Maryland. 1981b. Food and feeding habits of the pine marten in a Finnish Lapland forest in winter. Pages 580-598 in J. A. Chapman and D. Pursley, eds., Worldwide furbearer conference proceedings, Frostburg, Maryland. Raine, R. M 1981. Winter food habits, responses to snow cover and movements of fisher (Mat-tes pennanti) and marten (Martes americana) in southeastern Manitoba. Unpublished thesis. University of Manitoba, Winnipeg. 145 pp. Simon, T. L. 1980, An ecological study of the marten in the Tahoe National Forest, California. Unpublished thesis, California State University, Sacramento. 187 pp. Spencer. W. D. 1981. Pine marten habitat preferences at Sagehen Creek, California. Unpublished thesis. University of California, Berkeley. 121 pp. Spencer, W. D., and W J Zielinski 1983. Predatory behavior of pine martens. Journal of Mammalogv 64: 715-717. Steventon, J. D 1979, Influence of timber harvesting upon winter habitat use by marten. Unpublished thesis. University of Maine, Orono. 25 pp. Stordeur, L a 1986, Marten in British Columbia with implications for forest management. WHR-25. Research Branch, B.C. Ministry of Forests and Land, Victoria, B.C. Thompson, 1. D. 1986. Diet choice, hunting behaviour, activity patterns, and ecological energetics of marten in natural and logged areas. Lhipublished dissertation. Queen's University of Kingston, Ontario. 179 pp. Received 23 May 1990 Revised 7 February 1991 Accepted 12 February 1991 (;r.'at Basin Naturalist 50(4), 19W). p, 385 HOLOCENE PREDATION OF THE UINTA GROUND SQUIRREL BY A BADGER Michael E. Nelson' In 19S5 J. H. Mudsen, Jr., then the state paleontologist of Utah, collected several fossil bones at an elevation of 1524 m in Morgan Connty, Utah (Utah Antiquities locality 42Mo029v). The specimens were recovered from a large burrow intruded into shoreline sands deposited by Pleistocene Lake Bonne- ville. All fossils were found in a single pocket that probably represents the distal end of a burrow of the North American badger, Taxidea taxus. The specimens consist of (1) numerous post-cranial elements of a juve- nile badger and (2) several bones, including a right dentary, of the Uinta ground squirrel, Spermophilus armatiis. Many of the ground squirrel bones are crushed or broken, a condi- tion also noted by Long and Killingley (1983) in their study of badger prey. Taxidea taxus is virtually an exclusive carni- vore and does not eat significant amounts of plant material (Ewer 1973). Rodents are the most common prey, but the animals are not adverse to eating a variety of other vertebrates and arthropods (Long and Killingley 1983). Messick and Hornocker (1981) noted that Townsend ground squirrels, Spermophilus townsendi, are the most important prey spe- cies of badgers in southwestern Idaho. The animals either burrow after the active squir- rels, catch them hibernating in their burrows, or opportunistically wait at a burrow entrance (Ralph 1961). Badgers also will eat carrion and sometimes make food caches (Snead and Hendrickson 1942). Postdeath disturbance of the bones proba- bly accounts for missing elements of both Taxidea and Spermophilus. All preserved bones show extensive gnawing by small rodents; mice and other rodents commonly occupy badger burrows after the structures are deserted (Choate 1989, personal com- mvmication). Part of the badger pelvis was sacrificed for a radiocarbon date completed by Tandem Accelerator Mass Spectrometry at the Labo- ratory of Isotope Geochemistrv, Universitv of Arizona. The date of 2790 ± 74 yr. B.P. (AA-2514) suggests that Holocene diets of Utah badgers were similar to their extant counterparts. The remains of the ground squirrel may represent the last meal of the badger. All specimens are accessioned into the Sternberg Memorial Museum at Fort Hays State University (FHSM VP-10648 [ground squirrel] and FHSM VP-10649 [badger]). I thank Dr. Dave Gillette (Utah Antiquities), James H. Madsen (DINOLAB), and John Lund (FHSU). Literature Cited Balph, D. F. 1961. Underground concealment as a method of predation. Journal of Mammalogy 42: 42.3-424. Ewer, R. F. 1973. The carnivores. Cornell University Press, Ithaca, New York. 494 pp. Long, C. A., and C. A. Killingley. 1983. The badgers of the world. Charles C. Thomas, Publisher, Spring- field, Illinois. 404 pp. Messick, J. P., and M G. Hornocker. 1981. Ecology of the badger in southwestern Idaho. Wildlife Mono- graphs No. 3. .53 pp. Snead E., and G. O. Henderson. 1942. Food habits of the badger in Iowa. Journal of Mammalogy 45: 380-391. Received 16 January 1991 Accepted 28 January 1991 Department of Earth Sciences, Fort Hays State University, Hays, Kansas 67601-4099. 385 Cif.it H.1SJ11 Naturalist SOitl, 1990, pp. 387-.W9 PATTERNS OF MICROHABITAT USE BY SOREX MONTICOLUS IN SUMMER MaikC. 15clk'-, Clvdi-L. PrilclK-tt', and 1 1. Diiaiic Sinith' Sorex nionticolus is found tioin Alaska to Mexico in a variety of montane and boreal habitats (Hennings and Hoffmaini 1977). In previous characterizations of niicrohabitat used by this species, few measures of physi- cal or vegetative structure were significantly correlated with captures of S. monticolus. Typically, only some measure of near-ground cover (or related variables) is significantly associated with abundance. Sorex monticolus favors habitats with dense ground cover but seems to have few other niicrohabitat recjuire- ments (Hawes 1977, Terrv 1981, Gunther et al. 1983, Reichel 1986, Doyle 1989). In most montane areas the annual cycle of snow accumulation and melting, followed by herbaceous growth and decay, causes large-scale changes in the near-ground envi- ronment. During summer rapid herbaceous growth greatly increases the area covered by dense, near-ground vegetation. Previous studies of niicrohabitat use by S. monticolus have not addressed temporal changes in habi- tat use relative to this change in available cover (Terry 1981, Doyle 1989). During summer 1986, in conjunction with a study of niicrohabitat use by rodents in a mon- tane area, we recorded 104 captures of shrews in Sherman live traps. These shrews all ap- peared similar, and 17 specimens, retained for positive identification, subsequently were identified as S. monticolus. Given the possi- bility that some of the shrews captured may have been another species, we used a bino- mial probability to calculate the proportion of the 104 captures that could be regarded as S. monticolus; at a .05 level of confidence at least 85% of shrews captured were S. monticolus. Based on this, we feel confident that the ma- jority, if not all, of the shrews captured were S. monticolus. In this paper we examine teiii poral patterns of niicrohabitat use by these shrews during summer in relation to changes in niicrohabitat. StudyArea AND Methods The study site (111°37'N, 40°26'W) is on the east slope of Mount Timpanogos at an elevation of about 2400 m in Utah County, Utah. The habitat includes stands of aspen (Populus trcmuloides) and Douglas fir {Pseu- (lotsu^a )}wnziesii) interspersed with herba- ceous meadows and shrub-dominated ridges (principally snowberry, Symphoricarpos al- bus). Three trap grids were located in sepa- rate areas considered similar in overall habitat structure. Each grid covered 1 ha and con- tained 100 trap stations arranged in 10 rows of 10 each. Two folding Sherman traps were placed at each station, and stations were 10 m apart. Grids were trapped in a rotating fashion (see Belk et al. 1988 for details). Trapping began in early June, immediately after snow- iiielt, and continued until mid-September, resulting in 13,800 trap nights. Nineteen habitat variables were measured at each trap site characterizing live woody structure (trees and shrubs), dead woody structiu-e (fallen logs), and herbaceous cover and height (see Belk et al. 1988 for details). Five variables were correlated with shrew captures at the .10 level of significance during at least one month. These variables — percent canopy cover, average overstory tree size, average understory tree size, density of fallen logs, and number of woody species — were analyzed with principal-components analysis (SAS Institute, Inc. 1985). Two com- ponents had eigenvalues greater than one, but shrews exhibited little variation of habitat use on the second component (all means near Department of Zoology. Brigham Young University, Provo, Utah 84602. 'Present address: Savannah River Ecology Laborators . Drawer E. Aiken, South Carolina 29802. 387 388 Notes [Volume 50 0)5 "o 5-5 0.2 ■::; 0.0 LU E ci)z « .r ' S., and Deborah A. Hall, note by, 289 Anderson, Gar>- A., Jack E. Williams, Mark A. Stern, and Alan \'. Munhall, article by, 243 Apa, Anthony D., Daniel W. Uresk, and Raymond L. Linder, article by, 107 Asquith, Adam, article by, 135 Bats in Spotted Owl pellets in southern Arizona, 197 Beard, James M., and Harold M. Tyus, article by, 33 Belk, Mark C, Clyde L. Pritchett, and H. Duane Smith, note by, 387 Berna, Howard, J., article by, 161 Bibliography of Nevada and Utah vegetation description, 209 Birds of a shadscale (Atriplex confeiiifolia ) habitat in east central Ne\ada, 289 Black-tailed prairie dog populations one year after treat- ment with rodenticides, 107 Blackburn, W'ilbert H., and M. Karl Wood, article b>-, 41 Bone chewing by Rocky Mountain bighorn sheep, 89 Bourgeron, P. S., L. D. Engelking, J. S. Tuhy, and J. D. Brotherson, article by, 209 Britton, Carlton M., Guy R. McPherson, and Forrest A. Sneva, article by, 115 Brotherson, J. D., P. S. Bourgeron, L. D. Engelking, and J. S. Tuhy, article by. 209 Buchholz, Todd D., William C. McComb, and James R. Sedell, article by, 273 California Gull populations nesting at Great Salt Lake, Utah, 299 Chapman, Joseph A., and Kenneth L. Cramer, article by, 361 Chapman, Joseph A., Kenneth L. Cramer, and A. Lee Foote, note b\', 283 Cole, David N., article by, 321 Conservation status of threatened fishes in Warner Basin, Oregon, 243 Cox, George W. , article by, 21 Cramer, Kenneth L. , and Joseph A. Chapman, article bv, 361 Cramer, Kenneth L. , A. Lee Foote, and Joseph A. Chap- man, note by, 283 Dam-site selection by bea\ers in an eastern Oregon basin, 273 Deisch, Michele S., Daniel W. Uresk, and Ra\niond L. Linder, article by, 347 Distribution of limber pine dwarf mistletoe in Nevada, 91 Doescher, Paul S., Richard F. Miller, Jianguo Wang, and JeffRose, article by, 9 Duncan, Russell B., and Ronnie Sidner, note by, 197 Ecological review of black-tailed prairie dogs and associ- ated species in western South Dakota, 339 Ectomycorrhizal formation b\' Pisolithtis finctorius on Qucrcufi gembclii x Quercus turhincUa h\brid in an acidic Sierra Nevada minesoil, 367 Edminster, Carleton B., Robert L. Mathiasen, and Frank G. Hawksworth, articles b\', 67, 173 Effect of backpack radio transmitter attachment on Chukar mating, 379 Effects of burning and clipping on fi\ e bunchgrasses in eastern Oregon, 115 Effects of dwarf mistletoe on growth and mortality of Douglas-fir in the Southwest, 173 Effects of nitrogen availability on growth and photosyn- thesis oi Ai'temisia triclcntata ssp. wijomingensis, 9 Effects of prairie dog rodenticides on deer mice in west- ern South Dakota, 347 Eimcria sp. (Apicomplexa: Eimeriidae) from Wyoming ground squirrels, Spermophilus elegans, and white- tailed prairie dogs, Ct/nomiis leiicurus, in Wyoming, 327 Elliott, Charles L. , and Richard Guetig, article by, 63 Emergence, attack densities, and host relationships for the Douglas-fir beetle {Dendroctomis pscudotsiigae Hopkins) in northern Colorado, 333 Engelking, L. D., P. S. Bourgeron, J. S. Tuhy, and J. D. Brotherson, article by, 209 Esox liicius (Esocidae) and Stizostedion vitreum (Per- cidae) in the Green Ri\ er basin, Colorado and Utah, 33 Evans, Howard E., note 1)\, 193 Flinders, Jerran T., Bartel T. Slaugh, Ja\' A. Roberson, and N. Paul Johnston, note by, 379 Foliage biomass and cover relationships between tree- and shrub-dominated communities in pin\on-juniper woodlands, 121 Food caching and handling by marten, 381 Foote, A. Lee, Kenneth L. Cramer, and Joseph A. Chap- man, note by, 283 Forage quality of rillscale {Atrii)lcx sucklcyi) grown on amended bentonite mine spoil, 57 Form and dispersion of Mima mounds in relation to slope steepness and aspect on the Columbia Plateau, 21 Formation of Pisolitlitis tinctarius ectomycorrhizae on C^alifornia white hr in an eastern Sierra Ne\ada mine- soil, 85 Fres(iuez, P. R. , article b\ , 167 Fungi associated with soils collected beneath and be- tween ]>in\'on and juniper canopies in New Mexico, 167 Cialatowitsch, S. M., article b\. 181 392 1990] Index 393 George, Sarali 15., aTui Mark A. Forts, note 1)\, 93 Glenn, James L., Hieliard G. .Stiaiulit, and Jack \V. .Sites, Jr., article h\, 1 Gnetig, Hicliartl, and (^liarlcs L. Klliott, article hv, 63 Hall, i)el)()rali A., and Penny S. Amy, note by, 289 Hawksworth, Frank t;., Robert L. Mathiasen, and (^arle- ton B. Edniinster, articles b\', 67, 17.3 Hawksworth, Frank C;., and Kobert L. Mathiasen, note by, 91 Henry, Stephen E., Martin G. Hai)hael, and Li'onard F. Ruggiero, note b\', 381 Holocene predation ol the L'inta gronnd scjnirrel bv a badger, 385 Home range and acti\ it\ pattiMiis of black-tailed jaekrab- bits, 249 Humpback chub (Gihi cyplia) in the Yampa and Green rivers. Dinosaur National Moniuuent, with observa- tions on ronndtail chub ((.. rohttsta) and other sym- patric fishes, 257 Infection of young Douglas-firs by dwarf mistletoe in the Southwest, 67 Influence of soil frost on infiltration of shrub coppice dune and dune interspace soils in southeastern Nevada, 41 Jehl, Joseph R., Jr., Don S. Paul, and Pamela K. Yochem, note by, 299 Johnston, Paul, Bartel T. Slaugh, Jerran T. Flinders, and Jay A. Roberson, note by, 379 Joyal, Elaine, note by, 373 Karp, Catherine A., and Harold M. Tyus, article by, 257 Keating, K. A., note by, 89 Lessard, E. D., and J. M. Schmid, article by, 333 Linder, Raymond L., Anthony D. Apa, and Daniel W. Uresk, article by, 107 Linder, Raymond L., Michele S. Deisch, and Daniel W. Uresk, article by, 347 Longitudinal development of macroinvertebrate commu- nities below oligotrophic lake outlets, 303 Management of endangered Sonoran topminnowat Bylas Springs, Arizona; description, critique, and recom- mendations, 265 Marsh, Paul C., and VV. L. Minckley, article by, 265 Mathiasen, Robert L., Garleton B. Edminster, and Frank G. Hawksworth, articles by, 67, 173 Mathiasen, Robert L., and Frank G. Hawksworth, note by, 91 Mayfly growth and population density in constant and variable temperature regimes, 97 McComb, William G., James R. Sedell, and Todd D. Buchholz, article by, 273 McPherson, Guy R., Garlton M. Britton, and Forrest A. Sneva, article by, 115 Medica, Philip A., note by, 83 Medin, Dean E., note by, 295 Menkens, George E., Jr., Larry M. Shults, Robert S. Seville, and Nancy L. Stanton, article by, 327 Microbiology and water chemistry of two natural springs impacted by grazing in south central Nevada, 298 Miller, Richard F., Paul S. Doescher, Jianguo Wang, and Jeff Rose, article by, 9 Minckley, W. L., and Paul G. Marsh, article by, 265 Minshall, G. Wayne, and Ghristopher T. Robinson, arti- cle by, 303 Munhall, Alan V., Jack E. Williams, Mark A. Stern, and Gary A. Anderson, article by, 243 Natural hybrid between the Great Plains toad (Btifo cognatus) and the red-spotted toad (Bufo piinctatus) from central Arizona, 371 Nelson, Michael E., note by, .385 New distribution records of spider wasps (Ilymenoptera, Pompilidae) from the Rocky Mountain states, 193 New Mexico grass types and a selected bibliograpln of New Mexico grass taxonomy, 7.3 New variety o\'()xiitr()f)i.s cain))c.stris (Fabaceae) from the (Columbia Ikisin, Washington, 373 Notevvorthv mannual distribution records for the Nevada Test Site, 83 Observations on the dwarf shrew (Sorcx luiiius) in north- ern Arizona, 161 On the typification oiOxiitropis horcalis D(I, .355 Oxyfr(>f)is horcalis \ ar. australis, 359 ()x{itr(>))is horcalis var. huclsonica ((ireeiie) Welsh, 357 Oxtjtropis horcaUs \ar. sttlchurea (Pors.) Welsh, .358 Oxytropis horcalis var. viscidi (Nutt.) Welsh, 358 Oxytropis campestris var. wanapum, 373 Patten, Duncan T. , and Juliet G. Stromberg, article by, 47 Patterns of microhabitat use bv Sorcx monticoltis in sum- mer, 387 Paul, Don S., Joseph R. Jehl, Jr., and Pamela K. Yochem, note by, 299 Pollination experiments in the Mimulus cardinalis- M. Icwisii complex, 155 Ports, Mark A., and Sarah B. George, note by, 93 Pritchett, Glyde L., Mark G. Belk, and H. Duane Smith, note by, .387 Rader, Russell B., and James V. Ward, article by, 97 Raphael, Martin G., Stephen E. Henry, and Leonard F. Ruggiero, note by, 381 Reproduction of three species of pocket mice {Per- ognathus) in the Bonneville Basin, Utah, .361 Roberson, Jay A., Bartel T. Slaugh, Jerran T. Flinders, and N. Paul Johnston, note by, 379 Robinson, Ghristopher T., and G. Wavne Minshall, arti- cle by, 303 Rose, Jeff, Paul S. Doescher, Richard F. Miller, and Jianguo Wang, article by, 9 Ruggiero, Leonard F., Stephen E. Henry, and Martin G. Raphael, note by, 381 Schmid, J. M., and E. D. Lessard, article by, 333 Sedell, James R., William G. McGomb, and Todd D. Buchholz, article by, 273 Seed production and seedling establishment of a South- west riparian tree, Arizona walnut (Juglans major), 47 Seville, Robert S., Larry M. Shults, Nancy L. Stanton, and George E. Menkens, Jr., article by, 327 Sharps, Jon G., and Daniel W. Uresk, article by, .339 Shults, Larry M., Robert S. Seville, Nancy L. Stanton, and George E. Menkens, Jr., article by, 327 Sidner, Ronnie, and Russell B. Duncan, note by, 197 Sites, Jack W., Jr., James L. Glenn, and Richard C. Straight, article by, 1 Slaugh, Bartel T. , Jerran T. Flinders, Jay A. Roberson, and N. Paul Johnston, note by, 379 Slauson, William L., Gharles W. Welden, and Richard T. Ward, article by, 313 Small mammal records from Dolphin Island, the Great Salt Lake, and other localities in the Bonneville Basin, Utah, 283 Smith, David R., note by, 287 Smith, Graham W., article by, 249 Smith, H. Duane, MarkG. Belk, and Clyde L. Pritchett, note by, 387 Sneva, Forrest A., Carlton M. Britton, and Guy R. McPherson, article bv, 115 394 Index [Volume 50 Sorex prcblei in the northern Great Basin, 93 Spatial pattern and interference in pinon-juniper wood- lands of northwest Colorado, 313 Sprouting and seedling establishment in plains silver sagebrush {Ai-femisia cana Pursh. ssp. cana), 201 Stanton, Nancy L., Larry M. Shults, Robert S. Seville, and George E. Menkens, Jr., article by, 327 Stern, Mark A., Jack E. Williams, Alan V. Munhall, and Gary A. Anderson, article by, 243 Straight, Richard C., James L. Glenn, and Jack W. Sites, Jr., article by, 1 Stromberg, Juliet C., and Duncan T. Patten, article by, 47 Sullivan, Brian K., note l)y, 371 Summer food habits of coyotes in Idaho's River of No Return Wilderness Area, 63 Tausch, R. J., and P. T. Tueller, article by, 121 Taxonomy and variation of the Lopidea nigridia complex of western North America (Heteroptera; Miridae, Or- thotylinae), 135 Trampling disturbance and recovery of cryptogamic soil crusts in Grand Canyon National Park, 321 Tueller, P. T., and R. J. Tausch, article by, 121 Tuhy, J. S., P. S. Bourgeron, L. D. Engelking, and J. D. Brotherson, article by, 209 Two pronghorn antelope found locked together during the rut in west central Utah, 287 Tyus, Harold M., and James M. Beard, article by, 33 Tyus, Harold M., and Catherine A. Karp, article by, 257 Uresk, Daniel W., Anthon> D. Apa, and Raymond L. Linder, article by, 107 Uresk, Daniel W., Michele S. Deisch, and Raymond L. Linder, article by, 347 Uresk, Daniel W. , and Jon C. Sharps, article by, 339 Using the original land survey notes to reconstruct pre- settlement landscapes in the American West, 181 Vickery, Robert K., Jr., article by, 155 Voorhees, Marguerite E., article by, 57 Walker, R. F. , notes by, 85, 367 Walton, T. P., C. L. Wambolt, and R. S. White, article by, 201 Wambolt, C. L., T. P. Walton, and R. S. White, article by, 201 Wang, Jianguo, Paul S. Doescher, Richard F. Miller, and Jeff Rose, article by, 9 Ward, James V., and Russell B. Rader, article by, 97 Ward, Richard T., Charles W. Weklen, and William L. Slauson, article by, 313 Weklen, Charles W., William L. Slauson, and Richard T. Ward, article b\ , 313 Welsh, Stanley L., article by, 355 White, R. S., C. L. Wamliolt, and T. P. Walton, article by, 201 Williams, Jack E., Mark A. Stern, Alan V. Munhall, and Gary A. Anderson, article b>', 243 Wood, M. Karl, and Wilbert H. Blackburn, article by, 41 Yochem, Pamela K. , Don S. Paul, and Joseph R. Jehl, Jr. , note bv, 299 75 TABLE OF CONTENTS \Olimu' 50 No. 1 - March 1990 Articles A plasma protein marker for population genetic studies of the desert tortoise (Xerohates agassizi) James L. Glenn, Richard C. Straight, and Jack W. Sites, Jr. 1 Effects of nitrogen availability on growth and photosynthesis of Artemisia tridentata ssp. wijomingensis Paul S. Doescher, Richard F. Miller, Jianguo Wang, and Jeff" Rose 9 Form and dispersion of Mima mounds in relation to slope steepness and aspect on the Columbia Plateau George W. Cox 21 Esox hicius (Esocidae) and Stizostcdion vitrciim (Percidae) in the Green River basin, Colorado and Utah Harold M. Tyus and James M. Beard 33 Influence of soil frost on infiltration of shrub coppice dune and dune interspace soils in southeastern Nevada Wilbert H. Blackburn and M. Karl Wood 41 Seed production and seedling establishment of a Southwest riparian tree, Arizona walnut (Jufilans major) Juliet C. Strombergand Duncan T. Patten 47 Forage quality of rillscale (Atriplcx siicklcyi) grown on amended bentonite mine spoil Marguerite E. Voorhees 57 Summer food habits of coyotes in Idaho s River of No Return Wilderness Area Charles L. Elliott and Richard Guetig 63 Infection of young Douglas-firs by dwarf mistletoe in the Southwest Robert L. Mathiasen, Carleton B. Edminster, and Frank G. Hawksworth 67 New Mexico grass types and a selected bibliography of New Mexico grass taxonomy Kelly W. Allred 73 Notes Noteworthy mammal distribution records for the Nevada Test Site Philip A. Medica 83 Formation o[ Pisolithus tinctoriits ectomycorrhizae on California white fir in an eastern Sierra Nevada minesoil R. F. Walker 85 Bone chewing by Rocky Mountain bighorn sheep K. A. Keating 89 Distribution of limber pine dwarf mistletoe in Nevada Robert L. Mathiasen and Frank G. Hawksworth 91 Sorex prehlei in the Northern Great Basin Mark A. Ports and Sarah B. George 93 No. 2 -June 1990 Articles Mayfly growth and population density in constant and variable temperature regimes Russell B. Rader and James V. Ward 97 Black-tailed prairie dog populations one year after treatment with rodenticides Anthony D. Apa, Daniel W. Uresk, and Raymond L. Linder 107 Effects of burning and clipping on five bunchgrasses in eastern Oregon Carlton M. Britton, Guy R. McPherson, and Forrest A. Sneva 115 Foliage biomass and cover relationships between tree- and shrub-dominated communi- ties in pinyon-juniper woodlands R. J. Tausch and P. T. Tueller 121 Taxonomy and variation of the Lopidea nigridia complex of western North America (Heteroptera: Miridae, Orthotylinae) Adam Asquith 135 Pollination experiments in the Mimulus cardinalis-M . lewisii complex Robert K. Vickery, Jr. 155 Observations on the dwarf shrew {Sorex nanus) in northern Arizona Howard J. Berna 161 Fungi associated with soils collected beneath and between pinyon and juniper canopies in New Mexico P. R. Fresquez 167 EflFects of dwarf mistletoe on growth and mortality of Douglas-fir in the Southwest Robert L. Mathiasen, Frank G. Hawksworth, and Carleton B. Edminster 173 Using the original land survey notes to reconstruct presettlement landscapes in the American West S. M. Galatowitsch 181 Notes New distribution records of spider wasps (Hymenoptera, Pompilidae) from the Rocky Mountain states Howard E. Evans 193 Bats in Spotted Owl pellets in southern Arizona Russell B. Duncan and Ronnie Sidner 197 No. 3 -October 1990 Articles Sprouting and seedling establishment in plains silver sagebrush {Ai-temisia cana Pursh. ssp. cana) C. L. Wambolt, T. P. Walton, and R. S. White 201 Bibliography of Nevada and Utah vegetation description P. S. Bourgeron, L. D. Engelking, J. S. Tuhy, and J. D. Brotherson 209 Conservation status of threatened fishes in Warner Basin, Oregon Jack E. Williams, Mark A. Stern, Alan V. Munhall, and Gary A. Anderson 243 Home range and activity patterns of black-tailed jackrabbits Graham W. Smith 249 Humpback chub {Gila cijpha) in the Yampa and Green rivers, Dinosaur National Monu- ment, with observations on roundtail chub (G. rohusta) and other sympatric fishes Catherine A. Karp and Harold M. Tyus 257 Management of endangered Sonoran topminnow at Bylas Springs, Arizona: description, criticjue, and recommendations Paul C. Marsh and W. L. Minckley 265 Dam-site selection by beavers in an eastern Oregon basin William C. McComb, James R. Sedell, and Todd D. Buchholz 273 Notes Small mammal records from Dolphin Island, the Great Salt Lake, and other localities in the Bonneville Basin, Utah Kenneth L. Cramer, A. Lee Foote, and Joseph A. (Chapman 283 Two pronghorn antelope found loeked together duriufj; llie rut in west central Utah .... David H. Smith 287 Microbiology and water chemistry of two natural sprini!;s impacted by gra/ing in south central Nevada Deborah A. Hall and Penny S. Auiy 289 Birds of a shadscale {Atn})h'X confcrt [folia) habitat in east central Nevada Dean E. Mediu 295 California Gull populations nesting at Great Salt Lake, Utah Don S. Paul, Joseph K. Jehl, Jr., and Pam(