Historic, archived document

Do not assume content reflects current
scientific knowledge, policies, or practices.

y 3
ATLANTA BRANCK “YF
214 GLENN BUILDIl
STLANTA GEORGIA _4

BPEAN FO

It is my prid

CREeY

DOAVIUE 5
PUERTO RIGCC

BUBBLY Tuo

OCTOBER 1942

This journal serves as a medium of interchange of knowledge
among those interested in forestry in the islands and countries in

or near the Caribbean Sea.

Invitations to cooperate in this project

have been sent to forestry and agricultural officials in the follow-

ing places:

Bahama Islands
Barbados
Brazil

British Guiana

British Honduras

Canal Zone
Colombia
Costa Rica
Cuba
Curacao
Dominica

Dominican Republic

French Guiana
Grenada
Guadeloupe
Guatemala

Haiti

Honduras
Jamaica

Leeward Islands
Martinique
Mexico
Montserrat
Nicaragua
Panama.

St. Lucia

St. Vincent
Salvador

Surinam

Trinidad & Tobago
Venezuela

Virgin Islands

The journal is presented quarterly, in January, April, July,
and October. Material for publication should be submitted at least
two months before publication date and be addressed to the Director,
Tropical Forest Experiment Station, Rio Piedras, Puerto Rico.

Articles may be submitted in the contributor's own language
and preferably should be accompanied by # short summary of the paper.
Authors’ names should be typed or printed clearly and the title or
position of the author sent with the paper.

CONTENTS

Forestry and forest resources in Mexico . © + o o .e
H. Arthur Meyer, Pennsylvania State College

Roofing shingles in Jamaica - . . 2 © o © © © o «» e
L. V. Burns, Jamaica

ihe pine forests of Haiti oo o o 0 © ¢ 6 0 6 0 © ©
L. R. Holdridge, Haiti

Creosote penetration in tabonuco wood as affected
by preliminary boiling treatments in organic
SOUMOMGSE ee oetol ion moMcEhoulo Por chic choo jo oil. 6 © .6:. 0.6

David Reid, Duke University School of Forestry

Catalogue des cryptogames vasculaires des Antilles
frangaises ° ° ° °o ° ° S ° ° ° ° ° ° ° ° ° ° ° °
H. Stehlé, Martinique

°

16

23

35

FORESTRY AND FOREST RESOURCES IN MEXICO

H. Arthur Meyer
The Pennsylvania State College

Not until the beginning of this century did influential citizens of
Mexico and representatives of the Government take active steps to protect
the Mexican forests from unnecessary and ruthless destruction. It is true
that long before many far-sighted men had warned the people of the conse-
quences of forest devastation, and often large forest fires drew the at-
tention of the public to this important problem. But such warnings were
not heeded or followed by protective measures. While the Mexican Congress
concerned itself from time to time with various forest problems the legis-=
lators were unwilling to enact any drastic legislation which would favor
the Mexican forest. The governments of the various states were likewise
unable to promote laws to conserve the forests of the mountainous regions
of the country because such laws would naturally conflict with the immediate
interests of the local communities and the large landowners. Local cutting
restrictions and other decrees which were actually in force under the
Spanish rule, that is to say before 1810, were discarded after the liber-
ation, but not replaced by new ordinances. The National Government did not
dare to restrict by federal legislation the sovereignty of the states.
Under these unfortunate circumstances the destruction of the forests by the
indigenous population, in progress for centuries, was allowed to continue
and even to increase as the lumbermen invaded the virgin Sierras of the
North and the forested areas in the vicinity of the larger centers of
population.

It is an established fact that most of the denuded and badly eroded
mountains and hills so familiar to the visitor of the many famous archae-
ological sites throughout the country were once covered with forests. The
gradual destruction of the Mexican soil by erosion, which took place after
the destruction of the original forests, is paralleled on the same scale
only in the countries around the Mediterranean basin and in China.

Towards the end of the nineteenth century the forest conditions of
Mexico became a subject of deep concern to many patriotic citizens of Mexico.
Under the leadership of Ing. Miguel A. de Quevedo a movement was started .to
protect the Mexican forests from further destruction and devastation. Mem-
bers of a Central Committee began their work in the vicinity of Mexico City,
Veracruz, and other centers of population. A forest school and a large
forest nursery were established in Coyoacan near Mexico City. In 1909 a
Bureau of Forestry was established within the Department of Agriculture.
All this work was interrupted by the Mexican Revolution, but it was resumed
again soon after-. An excellent forest law was enacted in 1926. Under the
presidency of General Lazaro Cardenas the federal forest service, together
with wildlife and fisheries, was administered under an autonomous department.
In 1938 it again became a bureau within the Department of Agriculture.

= le October 1942

Area, Distribution and Types of Forests

Although there is considerable known about type and composition of a
great number of Mexican forests, it is still rather difficult to give a
complete picture of the existing forest types and their distribution. The
varied topography and the great differences in elevation result in a corres-
ponding difference in climatic conditions and hence types of vegetation.
Tropical forests cover the coastal regions on the gulf of Mexico and a narrow
strip along the Pacific coast. Embedded between the western mountain range
of the Sierra Madre and the mountains in the East near the Gulf coast Lies
the vast tableiand of Mexico with an altitude of from 5000 to 8000 feet dbove
sea-level. <A sub-tropical vegetation covers the foot hills of the main moun-
tain ranges, which unite in the southern part of the country. Coniferous
forests of great commercial importance cover the mountains of the Sierras.
Large areas in the west, central, and northern parts of the tableland are
extremely dry, supporting only a scarce vegetation of mesquite, cactus, and
AZAVE «

Botanically, the existing trees and shrubs of Mexico are rather well
known. They are described by Paul C. Standley in five volumes, published by
the Smithsonian Institute in Washington. Yet the practicing forester may
still find himself in great difficulty when trying to determine the scientific
name of an important forest tree found in remote regions of the country.

There is no doubt but that a great deal of botanical research remains to be
done, especially in tropical forests, before all the existing timber trees
of Mexico are adequately described.

At the present time less is known of the existing forest types. In
the files of the Mexican Forest Service there are a great many valuable re-
ports on specific forest areas cf the country. Many of these reports have
also been published in official bulletins. However, up to the present time
no one has compiled and made this information availiable in a comprehensive
study covering every region of the country. An attempt to outline and to
map the major forest regions or "zones" of Mexico was made in 1937 by José
Garcia Martinez who, at that time, was Chief of the Statistical Office of
the Department of Forestry, Game and Fish. The forest zones established by
Martinez are chiefly based on the topographic features of the country. The
resulting map, which is reproduced in Figure 1, gives quite a satisfactory
pisture of the distribution of the major forest zones since the forest type
is intimately correlated with altitude.

It would taks too much space to give a description of each of these
major zones and of the prevailing climates and subsequent types of vegetation.
Some pertinent date. have been assembled in Table 1, showing the total exten-
sion of various zones of altitude and the estimated forested area within each
of these zones.

According to this table only 28,471,000 hectaresl/ or 14.5 per cent
of the total land area of Mexico is covered by forest vegetation. The esti-
mates cf the percentage of forested areas within each zone of altitude

J

l/ One hectare = 2.47 acres.

»
Q

Caribbean Forester = Vol. 4, No. L

Mexico.1/

Fig. 1.—Forest Zones of

alder,

3
7

gany, ceiba, chicle

palme.
cedar (Cedrele) and many other tropical species.

(3) Oak, copal (Elephrium), pinon pine, senna

(2) Mahogany, mountain maho

(1) Mangove,

arbutus.:

(4) Agave,

mesquite.

(5) Pine, white cedar,

2

cactus,

2

eletore

cypress,

ay Reproduction of map published in Boletin del Departa-

mento Foresteal y de Caza y Pesca, No. 8, 1937.

October 1942

probably refer merely to the areas actually covered by forests and do not in-
clude arid lands sustaining only a poor vegetation which might be classified
as forest land as contrasted with agricultural or tillable land. It has been
estimated that of the total land area of Mexico which is 197 million hectares
only 20 million hectares are agricultural land, 30 million hectares sterile
land, the balance representing a type of land which could be classified as
forest and range land, of which only a small portion or roughly 30 million
hectares is covered by high forests. A large part of the remainder is suit-
able for restoration and reforestation.

Table 1.—Forested area of Mexico in different zones
of altitude.

Altitude above Total land area Area covered with forest

sea-level. miJ.lion Per vegetation
meters hectares cent million hectares Per cent

Total 196.9 100.0 28.5 14.5
0-500 64.7 $229 IU GI) 8.6
500-1000 32.7 16.6 4.3 Jie
1000-1500 40.6 20.6 2.6 1.3
1500-2000 $8.8 19.7 Diode eG
2000-2500 18.9 SAS Wok 0.6
2500-3000 nie 0.6 0.4 0.2
and over

With a population of 17 million the forest area per head amounts to
1.54 hectares, which is rather low when compared with other Central and South
American countries, but about the same as the per capita area of the commer-
cial forest lands of the United States, and about six to eight times larger
than that of Germany and France.

Forest Production

Statistics on forest preduction are still scarce and by no means all
inclusive. The letest officiel statistics on wood and charcoal production
are found in the September-October issue of the Boletin Forestal y de Caza,
Volume 2, Number @. According te this publication the output of wood pro-
ducts in the years 1937-1938 was as shown in Table 2.

The increase in the production of saw timber is partly expiained by
the recent construction of highways leading through heavily forested regions.
The decline in hewn timber is due to the heavier tax imposed on such timber,
the purpose of the tax being to discourage this kind of products, and to the
reduced demand for railroad ties which resulted from the increased use of
creosoted ties The actual exploitation is probably larger than the amounts
recorded in official statistics; still it represents only a small average

Vol. 4, No. lL

He
i]

Caribbean Forester =

exploitation per hectare. Ten million cubic meters per year would only
amount to 0.35 cubic meters per hectare, while the productive capacity of

the estimated forest area of 28 million acres may be estimated to be at least
3 cubic meters per acre. These few figures give an idea of the future po-
tentialities of forestry in Mexico.

Table 2.—Wood production in Mexico 19353-1938.

1933 1934 1935 1936 1937 1938

Saw timber in 1000 cu: m- 369 1419 585 600 648 477
Hewn timber in 1000 cu. nm. 698 1817 387 569 SL7 399
Logs in 1000 cu. m- 136 380 199 61 66 250
Fuelwood in 1000 cu- nm. 1447 2140 514 567 791 pip Er By
Charcoal in 1000 tons 170 188 167 156 L57 160

With the large number of forest types in Mexico it is not surprising
that the country is producing a great variety of secondary forest products.
Some of these products if produced in greater quantities would represent a
basis for @ number of important industries. At the present time the pro-
duction of gum or chicle, used for the manufacture of chewing gum, is perhaps
the most important one. This gum is produced by the chicczapote, Achras
sapota. It comes from tropical forests, the principal producers being the
State of Campeche and the Territory of Quintana Roo. The annual production
for the entire country according to official statistics is as folilows:

1935 - 2,069,718 kgs.
1936 - 2,760,851 kes.
1937 - 2,577,898 kes.

Other secondary products are dyewoods, tanning materials, rubber, naval
stores, coconut and palm oils.

Accomplishments and Future Problems of Mexican Forestry

in recent years the Mexican government became fully aware of the pres-
Sing problems concerning the maintenance and conservation of the forests. The
executive branch of the government, entrusted with the conservation of these
forests, has undertaken an extensive campaign in all parts of the country to
impress upon the average citizen the national importance of forest conser-
vation. One of the outstanding features of this campaign is the establishment
of forest nurseries by schools throughout the country- Over 1500 such nurser-
ies were in existence in 1938- A popular magazine "Proteccién a la Naturaleza"
has been published and freely distributed during recent years.

A large number of beautiful natural forested areas have been set aside
as national parks. Up to 1938 twenty-five such parks had been definitely

Se Gie October 1942

established and many more were being contemplated. The areas of these parks
vary from several hundred to seventy thousand hectares. A great many tourists
visit these parks each year. Most visitors to Mexico probably nave seen the
beautiful fir forest of the "Desierto de los Leones" near Mexico City, or the
virgin oak forest in the National Park Alexander von Humboldt, located near
the famous city of Tasco which also has been declared a national monument.

In accordance with the provisions made in the forest law, a large
number of forests or partly forested areas have been declared protective
forest zones. These areas have been established as a protection against .
erosion in steep mountainous sountry, as @ protection for watersheds, and
for artificial systems of irrigation. Protective zones also have been de-
clared in the neighborhood of cities and other centers of population. Any
exploitation of these forests is subject to the strictest regulation. The
size of these protective zones varies greatly, ranging from about 20,000 to
200,000 hectares.

Much progress has been made in forest protection, especially against
forest fires. More than 10,000 farmers have been organized in “defense
cooperatives" throughout the country.

The enforcement of forest regulation in practice has been very effec-
tiveiy tied up with the collection of the forest yield tax. Cutting and
shipping permits must be secured before any large quantity of wood may be
cut and sold. The application of differential tariffs has greatly increased
the radius of forest operations; the tax on wood cut near the centers of
population is higher than for products coming from remote regions. In ad-
dition high quality products are taxed higher than defective or dead wood,

he removal of which is 4 definite advantage to the forest. By extending
the current wood operations over a larger territory, the cutting in any par-
ticular region becomes Jess intensive. The forest tax, therefore, is an
effective fastor in helping to control forest exploitation.

The distribution of the large land holdings and “haciendas" has in
many ways created a new situation in regard to the administration and manage-

mens of the Mexican forest; The creation of communal forests and of Ejidales
forests has been greatly promoted in recent years.

It is the opinion of leading Mexican foresters that communal forests
and forest cooperatives offer the best possibilities for sustained yield
management and conservative forest practices, as contrasted to large scale
exploitations which have been, and to a certein extent are still, carried
out by large lumber ccmpanies.

The intensification and extension of forest activities throughout the
country was only possible through a substantial increase in the personnel of
the forest service, It was not too difficult a problem to train, in a
relatively short time, a large number of forest rangers, and this has actually
been done at the ranger school in Tlalpam, near Mexico City. These rangers
are being placed at stations and forest delegations throughout the country.
They represent a police force which sees to it that the regulations of the
forest law are observed In the enforcement of these regulations the depart-
ment of forestry is assisted by the regular army.

\

Caribbean Forester = 6 = Vol. 4, No. Ll

Amore difficult task is the education of technically trained forest
engineers. Today young men are receiving technical training in forestry at
the National Schocl of Agriculture in Chapingo near Mexico City. The forest
service seems to have in mind also a plan to send promising young men to
foreign schools where they can improve and widen their technical knowledge.
Most of the well trained foresters of the country are at the present time
holding important positions in the federal forest service.

The extensive organization of the Mexican forest service based on
progressive forest legislation, together with the patriotic spirit of the
Mexican forester, promises the most successful continuation of the forestry
program,

Summary

Mexisce’s forest history is not unlike that of many other nations in
which exploitation greatly reduced the forest resource before the application
of effective remedies started. Toward the end of the nineteenth century,
Ing. Miguel A. de Quevsde started & protection movement. A forest school and
a large Porest nursery were established near Mexico City. In 1909 a Bureau
of Forestry was set up within the Department of Agriculture. An excellent
forest law was enacted in 1926.

There is a great variation in climate and soils throughout the length
of Mexico, resulting in a corresponding wide variety of forest types. Tro-
picai forests cover the coastal regions on the Gulf of Mexico and a narrow
strip along the Pacific coast. SEI SST vegetation covers the foothill
and important coniferous forests cover the Sierras. A vast majority of the
tree species have been identified but little is known concerning the charac-
ter and extent of distinct forest types.

There are at present 28 million hectares of forested land and an ad-
pack million hectares of non-forested forest land. Present saw-timber

ion is Low snough so that if properly conducted the present forest

be adequate for perpetual production.

>

eres. ala

A sesondary product of Mexice
gum> The preduction, which was more
centered in Campeche and Quintana Roo.

"5 je, used in chewing
t i

h egrams in 1937, is

2

Mexico has a sate set aside a large number of parks and protection
forest yiel
1

forests. A flexible for gd tex provides government control of the

3
type and iocaticn of ferest exploitation. Communal forests and forestry
education are being encouraged.

Resumen

A principios de este sigle 91 go

; bierno mexicano empezé a tomar accién
contra la destruccion innecesaria del bo

»sque. Se inicié un movimiento de

= 7 = October 1942

proteccién forestal bajo la direccién del ingeniero Sr. Miguel A. de Quevedo.
Se establecio en Coyoacdén cerca de Ciudad de México una escuela forestal y un
vivero grande de arboles. En el 1909 se cred una Direccién de Bosques que
dependia de la Secretaria de Agriculture. Bajo la presidencia del General
Lazaro Cardenas se constituyS el Departamento Forestal y de Caza y Pesca que
tenia cardcter auténomo, pero en el 1938 volvié éste a depender de la Secre-
taria de Agricultura.

Apesar de lo mucho que se sabe del tipo y composicién de gran nimero
de los bosques mexicanos, todavia es dificil dar una idea completa de los
tipos forestales alli existentes y de su distribucién. Las variaciones topo-
gréficas y les grandes diferencias en slevacion dan como resultado una dife- ~
rencia. correspondiente en condiciones clim&ticas y por lo tanto en el tipo de
vegetacion. El litoral del golfo de México y una faja estrecha a lo largo
del litoral Pacifico estén cubiertos de bosques tropicales. Las bases de las
cordilleras principales que se unen a1 sur del pais estén cubiertas de vege-
tacion subtropical. Las montafias de las sierras estén pobladas de bosques
de coniferos de gran importancia comercial, Grandes areas en el oeste, centro
y parte norte de la meseta central son extremadamente secas y tienen por lo
tanto solamente una escasa vegetacioén de mezquite, cacto y agave.

Sdlo e1 14.5 por ciento del a&rea total de México o sea 28,471,000
hectéreas estan cubiertas de bosques. El area por persona viene siendo de
1.54 hectéreas.

Los principales productos forestales son madera de aserrio, madera
lefia y carbén. Ademaés un producto secundario importante es el chicle’
ue se obtiene en el estado de Campeche y el territorio de Quintana Roo.

P y

El Bon 27110 mexicano ha emprendido recientemente una campafia extensiva
para recalcarle al pueblo la importancia de los bosques. También se han es-
tableside grandes parques nacionales,. Las diferencias en los impuestos sobre
aprovechamiento forestal (v.g. que sean mis altos cuanto més cerca al mercado
se efectis 21 sorte) se usan para favorecer determinadas prdcticas mds con-
venientes y racionales.

S
6

saron catedras de selvicultura, dasometria, ets. para establecer
el grado de Ingeniero de Montes y también se crearon Escuelas de Guarderia y
en el presente se considera un plan pare que los ingenieros de montes reali-
cen estudios avanzados en el extranjero.

Le ox ganizacion y trabajo extenso del Servicio Forestal Mexicano, ba-
sado en una legislecién forestal progresiva y el espiritu patridtico de los

ingenieros d@ montes mexicanos promete la més prospera continuidad del pro-

grama forestal de aquella nacién.

Caribbean Forester = 8 = Vol. 4, No. 1

ROOFING SHINGLES IN Jamarcal/

L. V. Burns
Forest Supervisor
Jamaica

The use of shingles as a roofing material in Jamaica has been a prac-
tice from the earliest days of the colony. The practice appears to have been
introduced by the English colonists who settled in the island towards the end
of the seventeenth century. Prior to this the Spaniards had employed tiles
almost exclusively for their better dwellings. Most of their towns were
built on the coastal plains within easy distance of the lower limestone hills,
and the soil was suitable for the manufacture of tiles, to which they were
accustomed. After 1655, however, the English colonists spread through the
island and began to establish temselves further inland. The island abounded
in valuable timbers many of which were suitable for shingles, and their master
craftsmen were set to work. In the next 50 years, most of the magnificent
colonial mansions were erected. Today only massive ruins remain but these
pay a silent tribute to the skill of the early builders who so influenced the
school of colonial architecture, and introduced the art of fashioning hand-
riven shingles.

Advantages of Shingles

Coolness.-—In tropical sountries, it is necessary to select a roofing
material which will always remain relatively cool. This is perhaps the best
feature about shingles, as buildings roofed with shingles maintain an almost
constant temperature sometimes many degrees iower than that outside.

Relative durability.—There are many examples on the sugar estates
throughout the island where cedar shingles are still in good condition after
nearly a century of constant use- This is common on the lowland houses but
even in the hills, where conditions are very humid, shingles will last for 50
years or more. Ali these early shingles were split and dressed entirely by
hand and are credited with being much more durable than sawn material.

Local materials —Timbers suitable for shingles occur throughout the
island mainly over 1.000 ft. abowe sea-level. Consequently there is not much
distribution of the finished product from one point to another. Generally,
shingles are utilized essentially for rural construction as in the towns
imported shingles, galvenized iron, and tiles are the most popular roofing
materials. However. in rural districts iow transport costs and ready replace-
ments when necessary are of great assistance, particularly to the peasantry who
are slowly changing from the use of thatch and grass to shingles, and who are
usually required to provide their own transport - by head or donkey.

af Acknowledgement is made to C. Swabey, the Conservator of Forests in
Jamaica for valuable assistance in the preparation of this paper.

= 9c October 1942

Suitability to environmment.—Unfortunately, seasoning is not a very |
popular or widely practised operation in Jamaica. Shingles are usually manu-
factured only when they are required and are often used within a few weeks.
Therefore, shingles produced in any district are generally best suited for
use in the same district rather than in another locality where humidity and
temperature, etc. may be quite different.

Disadvantages of Shingles

Fire hazard,—-Fire is the bugbear of all wooden buildings. Shingles
have become less popular as a result of the exaggerated advertising of mer-
chants who are anxious to sell less inflammable roofing material. However,
with ordinary elementary precautions, shingles are as safe and suitable as
any other form cf roofing.

Water col TECULnE ~In the rural areas where people depend on rainfall
for their water supply, roof catchment is very important. Water collected
off new roofs which are made of some aromatic wood like cedar will definitely
be tainted for a fewweeks. As the water does not become injurious, and as
the faint aroma may be removed by boiling it will be seen that this is not a
serious disadvantage, In any event, tainted water is usually the result of
uncovered tanks, birds perching on roofs and probably also to freshly painted
roofs.

Appearance.--The rough appearance of an unfinished split shingle consti-
tutes a serious drawback and is one of the chief reasons for the limited use
of shingles in urban construstion.

Wood Used for Shingles

Most straight-grained timbers are used for shingles but other desirable
qualities are light weight, durability, and minimum tendency to warp and split
under varying conditions of wind, temperature, and moisture. In the tropics
where a heavy downpour often occurs during a blazing sun, special regard must
be paid to all qualities in selecting material for use in the better buildings.
The following timbers are those chiefly used for shingles:2

Cedrela odorata L. (Meliaceae) West Indian cedar.—This tree furnishes
what is undoubtedly the best shingle wood in the island. It is a large tree
and the timber is light (36 lbs. fou £4.) durable, fissile and is available in

iarge quantities. It occurs throughout the island up to 4,000 ft. Very
closely allied to C. mexicana,

Terminalia latifolia Sw. (Gombretaceae or Terminaliaceae) Broadleaf.—
Next in importance to cedar. This is a tall tree reaching 80 to 100 ft. in
height with a girth of 8 to 10 ft. The trunk is usually very straight with
horizontal branches arranged in whorls every 15 to 20 ft. The timber is
straight-grained, light (41 lbs. cuft-), splits easily and is often used for
shingles in the limestone districts where it occurs.

We The information here presented was obtained from "The Principal Timbers
of Jamaica” by C. Swabey, B. Sc.

Caribbean Forester - 10-2 Vol, 4, No. 1

Calophylium brasiliense Camb. var. antillanum Britton (Guttiferae or
Clusiaceae) Santa Maria.—-A lofty straight tree, 100 to 120 ft. in height with
a girth of 8 to 10 ft. The wood is of medium weight (53 lbs.) fissiie, and
often used for shingles. The shingles are not durable and usually require
some form of preservative treatment. A widely distributed forest tree in
damp woodlands up to 3,000 feet.

Nectandra spp. and Ocotea spp. (Lauraceae) Sweetwocods.—The term sweet-
wood is rather loosely applied to several trees of different identity. The
timber is usually soft, light, (approx. 35 lbs./cu.ft.) and fissile. It is
not durable but is often used for shingles by the peasantry.

Mosquitoxylum jamaicensis Kr. & Urb. (Anacardiaceae) Mosquito wood. —
A tall tree common in the cockpit region and often used for shingles in the
surrounding districts. The shingies are very durable but because of the cross
grain they do not fit nor do they lie evenly.

Peltophorum brasiliense (Sw.) Urb. (Leguminosae or Caesalpiniaceae)
Braziletto.—-A small tree common on the dry lowlands with 4 first class, heavy
(67 lbs./cu.ft-.), durable, hard wood occasionally used for shingles. Never
used where cedar, etc. are obtainable but occurs under xerophytic conditions
where usually there are few other suitable timbers.

Samanea_saman (Jacq.) Merril (Leguminosae or Mimosaceae) Guango. (Syn.
Pithecolobium saman Benth).—A large tree common on the southern lowlands of
the island. The timber is good but the tree is rarely used except as shade
and for fodder in pastures. When mature it is cross-grained, and difficult to
work but fairly light (44 lbs./cu. ft.) and strong. The branches are often
quite large and are sometimes used for making shingles on estates.

Macrocatalpa longissima Britton (Bignoniaceae) Yokewood. (Syn. Catalpa
longissima Sims).—A large tree to 80 ft. in height. It yields very good
timber which has recently been tried in the production of sawn shingles. The
tree is of limited distributicn, being found on the dry coastal plains of the
South and Southeast and along sandy river benks.

Hibiscus elatus Sw. (Malvaceae) Blue mahoe.—A fast growing, straight-
stemmed tree- The wood is hard, tough, durable, and fairly light (47 lbs./cu.
ft.). This timber is popular for cabinet work and is now almost unobtainable.
It has been used cccasionally to make first class shingles but is definitely
of very limited use for this purpose, owing to its rarity.

In addition to the above recognized shingle timbers, there are many
others which are used by the poorer classes for splitting rough shakes to be
used on temporary "bush huts". Among these may be mentioned:

Cecropia peltata L. (Moraceae) Trumpet

Anons. montane. Macf. (Anonacese) Wild sour sop or mountain sour sop
Clethra tinifolie Sw. (Ericaceae) Soapwoed

Crescentia acuminate (Bignoniaceae) Calabash

Eugenia jambos L. (Jambose jambos) (Myrtaceae) Roseapple
Calyptranthes spp. and Eugenia spp. (Myrtaceae) Rodwood

= li = October 1942

Spondias mombin L. (Anacardiaceae) Hog Plum
Zanthoxylum martinicense {Lam,}) DC. {Rutaceae) Prickly yellow
Unknown Black fig

Method of Preparation by Hand

The method usually employed is extremely wasteful. Under the Crown
Lands Forest Produce Rules all green standing trees cut with the object of

yielding shingles are assessed for royalty purposes on their girth measure-
ment. This does not, however, apply to remnants, as the people are encouraged
to work on these.

Usually trees are felled with a generous portion remaining on the stump.
All work is done with axes. After the branches have been removed, the trunk
is cut into sections 2 ft. 6 in. in length. These may be cross-cut but axes
are still very generally used as they are cheaper, easier to obtain, easier to
sharpen, can be used by one man, and probably more than any other reason, the
men are familiar with this versatile tool which can be such a formidable
weapon of offence or defence. Cross-cut saws are now being stocked by the
village shops and their advantages are slowly being recognized. After cutting,
the short barrel-like blocks are stood on end, and while one man holds the
"froe" in place, his partner drives it tangentially through the block with a
series of blows. The "froe" is usually native made and is an ordinary cutting
implement about 12 in. long and 3 to 4 in. wide. It has a handle about 3 ft.
in length providing great leverage to assist in the process of getting off the
undressed shingle. No guide lines are drawn on the end of the block and the
shingle splitter depends on his amazingly accurate judgement to get off the
rough shakes which will dress down to half~an-inch in thickness at the larger
end. The minimum shingle width is 5 in. but usually the timber is discarded
as 6& result of cross grains when the now polygonal block is about 8 to 9 in.
across. As immature trees of small girth are often cut it is obvious that
the majority of the timber is wasted.

The method of "shaping" or “dressing” shingles is very primitive and
full use is made of the powerful prehensile toes with which most of the
peasantry are equipped. The operator sits on a low wooden horse, the main
section of which consists of a hardwood scantling approximately 4 ft. long
and 4 in. square. The end of this scantling is raised about 12 to 15 inches
above the ground and usually is supported by an ordinary wooden cross. The
end Seas the cperator and away from him, rests on the ground. About 9 in.

from this end, a rectangular hole about 3 in. by 2 in. is made through the
beam fitting freely in this hole is a hardwood pin so designed that it pro-
trudes for 3 to 4 inches at its lower end but is prevented from falling through

the hole by @ head which is extended on the side towards the operator and so
rests on the top of the beam, A nail is driven into the head of the wooden
pin so that the sharp point protrudes for approximately half an inch like a
solitary upper tooth in the "jaw" which is formed by the main beam and the
head of the pin. When a shingle is pushed into this jaw, the operator grasps
the lower end of the pin between his toes and pushes forward. This lowers the
upper end of the pin, closes the jaw and forces the nail into the shingle
which is thus held firmly in place. A drawing knife is then used to smooth
off the shingle on its upper side, after which the pressure is redieased and

Caribbean Forester - 12 = Vol. 4, No. l

the shingle is turned to present the other side.

The native shingle splitters are remarkably dextrous. A pair will
split and dress up to 400 in one day when the blocks are already cut. The
quantity depends on the kind of timber, cedar being easiest to work.

At this stage the shingle-splitter leaves his shingles. The next step
is known as "jointing" and is done by the carpenters who are laying the shin-
gles on the roof. Jointing consists of planing the sides of the shingles so
that they are parallel and the shingles wiil fit closely. This is a selective
operation and only about 20 per cent of the shingles will require attention.

Costs

Production.—Cost of manufacture varies considerably but is chiefly
influenced by locality and type of timber used for the shingles. Splitting
and dressing can usually be done for 40 to 50 shillings per M., while joint-
ing will cost an additional 8 to 10 shillings per HM. :

In some districts where the shingies must be carried for some distance
often over very rocky ground, the cost of "heading" must be considered. This
usually works out at approximately 6 to 9 pence per 100 shingles per mile.
The amount carried depends on the terrain but a good carrier wiil usually
make bundles consisting of about 100 tc 150 shingles.

Sale.—Shingles are offered for sale in two grades, No. 1 being 22 in.
by 5 by 1/2 in. and over, while all shingles measuring less than this are
classed as No. 2. In actual practice, with the sale of native split shingles,
no such sharp distinction exists although sawn shingles are rigidly graded.
Under the Crown Lands Forest Produce Rules the principal timbers are grouped
in 3 classes. Of the shingle woods, cedar and mahoe are in Class 1, Braziletto
in Class 2 and all others fail into Ciass 3: No. 1 cedar shingies, all heart,
will fetch about 6 to 7 pounds per M., wnile broadleaf, santa maria, sweetwood,
and mosquite wood shingles cf the same grade are sold for 3 to 4 pounds per M.
Sawn shingles are usually seld for approximately 20 per cent less than split
Shingles cf the same grade and species. Sawn yokewood shingles were recently
put on the market at 5 pounds per M., but did not compete very well against
the best sawn cedar which was sold at the same price.

Local Shingle Mills
Within the past 5 years many local shingle mills have been erected.
Power is almost always derived from an ancient, erratic, and usually ineffi-
cient motor-car and is conveyed by slack, wildly slapping beits to the saw
table, There are however, a few mills where a small but powerful diesel
engine runs 41] the units necessary-

One of the best small mills in the island uses a 25 horse power Ruston

Hornsby engine, and by adequate belts and shafting runs a breakdown log saw,
5 saw tables, one 12 in. jointer and a small lighting plant.

= 13 = October 1942

The logs are purchased already cut to length and are collected by truck.
On arrival at the yard, the green logs are sawn into slabs 5 in. thick. These
slabs are then removed to the shingle tables where the actual shingles are
turned out. These tables are constructed of good solid heavy hardwood in order
to reduce vibration. A circular saw is operated so that 5 in. protrudes above
the table top and the slab is drawn at an angle to the saw so that the necessary
taper is given to the shingle.

It is important that the teeth be kept sharp and properly set or the
shingles will have raised and torn grain. When this type of shingle is laid,
water usually collects on and is absorbed through the ragged surface, thus re-
sulting in early decay. Very few mills pay sufficient attention to these points
and as a result the quality and quantity are not as satisfactory as they might
be-

Summary

The use of shingles in Jamaica as a roofing material has been 4 practice
ever since the first English colonists settled on the island near the end of
the seventeenth century.

Shingle roofs are for several reasons very desirable. They are cool,
keeping inside temperature below that outside. They are durable, some old
roofs in Jamaica being still serviceable after nearly a century. They can be
built of local materials, thus obviating the necessity of high transportation
costs. On the other hand they are inflemmable and have a rough appearance.

Cedrela odorata and Terminalia latifolia are the most important shingle
woods in Jamaica. Cedrela odorata and Hibiscus elatus produce Class 1 shingles;
Peltophorum brasiliense produces Class 2; and all other species produce Class 5.

The best shingles are made by primitive hand methods. Tangential pieces
are splits out of blocks of wood with a froe. They are then dressed with a draw
knife while lying on a small timber before the seated worker. The jointing or
trimming of the edges is done by the carpenters when laying the shingles. A
pair of laborers can split and dress 400 shingles per day, or at a cost of 40
to 50 shillings per thousand.

Split shingles are sold in two grades, the first being 22 x 5x 1/2
inches or over and the second grade those measuring less than this. First
grade split shingles of Cedrela odorata will bring about 6 to 7 pounds per
thousand -

Recently several miljs have been erected for the manufacture of sawn
shingles. When efficiently run these can produce shingles cheaper than by
hand, @ circumstance required for their existence, because of the 20 per cent
premium paid for split over sawn shingles. When these mills do not properly
maintain their saws torn grain results. This increases the absorption of
water by the shingles and accelerates decay.

Caribbean Forester - 14 = Voll. 45° Noi:

Resumen

El uso del tejamani para techar ha sido una prdctica empleada en
Jamaica desde los tiempos en que los primeros colonizadores ingleses se es-
tablecieron en la isla a fines del siglo 17.

Los techos de tejemani son deseables por varias razones: son frescos,
pues conservan una temperature en el interior mis baja que la del exterior;
son duraderos, ya que algunos techos antiguos son servibles atin después de
casi un siglo. y porque pueden construirse con materiales locales obviando
asi la necesidad de pagar costos altos de transporte. Pero por el contrario
son inflamables y tienen una apariencia ordimria.

Las principales maderas usadas en Jamaica para tejamani son: Cedrela
odorata, Terminalia latifolia, Calophyllum brasiliense var. antillanun,
Nectandra spp-, Ocotea spp-, Mosquitoxylum jamaicense, Peltophorum brasiliense,
Samanea saman., Macrocatalpa longissima, e Hibiscus elatus. Las primeras dos
mencionadas son las més importantes. Cedrele odorata e Hibiscus elatus pro-
ducen tejamanies de primera clase; Peltophorum brasiliense de segunda clase;

y las demis de tercera clase.

El mejor tejamani se hace por métodos manuales primitivos. Se cortan
listones tangenciales de trozos de madera, luego se desbastan con una cuchilla
especial sobre un madero que el trabajador coloca delante de su asiento. El
ribete de los bordes lo hacen los carpinteros al techar.

Dos trabajadores pueden cortar y desbastar 400 tejamanies al dia a un
costo de 40 a 50 chelines por miller.

El tejamani se vende de dos tamafios, el primero es de 22 x 5x 1/2

pulgada o mas, y el segundo, aquél que mide menos. $1 tejamani de primer
tamafio pesa de 6a7 libras por millar.

Recisntemente se han construido mclinos para la fabricacién de teja-
mani aserrado. Debide a que el tejamani hecho a mano vale 20 por ciento mis,
es una circunstancia indispensable para le existencia de los molinos el que
funcionen eficientemente produciendo tejamani mais barato. Si las sierras en

Linos no se conservan en buen estado se producen granos torcidos en la
madera. Esto aumenta la absorcion de agua y acelera el deterioro.

= 5) = October 1942

THE PINE FORESTS OF HAITI

L. R. Holdridge
Manager, Forestry Division
Société Haitiano-Americaine de Developpement Agricole
Port-au-Prince, Haiti

The pine genus belongs essentially to the temperate zone, but since it
hes invaded the warmer zones of the world in such places as the West Indies,
Central America, and the East Indies, no consideration of tropical forestry
can be quite complete without mention of this important group of trees.
Furthermore, pine is of special importance to a country such: as Haiti, where
due to the pressure of a dense agricultural population, the greater portion
of its timber resources have been removed already from the richer soils which
previously supported excellent stands of hardwoods.

The one species of pine native to Haiti is Pinus occidentalis Sw., a
tree scarcely heard of in northern climes but one which is destined to play
an important role in forest management on the island of Hispaniola and in
eastern Cuba. Two papers!/ which appeared previously in this journal have
reported on the taxonomic characteristics and the range within the Dominican
Republic of this species,

In Haiti, the range of the species includes most of the country,
although in most sections it occurs only as scattered specimens. In the
southern ranges of mountains, good stands of pine are found in large blocks.
At the eastern end of the Morne la Sells range, in the section called Mornes
des Commissaires is located a stand estimated to cover approximately 10,000
acres. This is contiguous with the pine forest across the border in the
Bahoruce Mountains of the Dominican Republic. The Haitian tract is located
on a plateau at an elevation of about 5,000 feet above sea-level, although
scattered patches of pine to the north and south grow down to 2,000 feet.
This is the only forest in Haiti which is traversed by a public road and when
"le forét de pins" is mentioned in the capitol, it is almost invariably with
reference to this unit.

Only a fewmiles to the west of this area la Selle peak, the highest
point of land in the Republic, attains an elevation of over 9,000 feet. It
is covered to the very peak with a pine forest which has an area probably
greater than 25,000 acres. West of La Selle on the main plateau and on the
various ridges extending therefrom scattered individuals and patches are
encountered to slightly beyond Furcy and Kenscoff.

if Chardén, Carlos BE. Los Pinares de la Repiblica Dominicana. Caribbean
Forester, 2(3): 120-130. April 1941.

Carabia, J. P. Contribuciones al Estudio de la Flora Cubana, Gymnospermae.
Caribbean Forester 2(2); 94-97. January 1941.

Caribbean Forester feycws Vol. 4, Nowa

Traveling still farther westward on the southwestern peninsula various,
at present inaccessible, scattered patches in the Morne la Hotte range are
found, but only near Pic Macayo are solid stands of pine found to be of any
appreciable extent. Here, they ascend to 7,000 fest on the peak and occupy
the sides and crests of precipitous ridges, which in combination with the very
deep, intervening valleys make up a region which is very difficult to traverse
and hence little known.

North of the Cul de Sac plain, which runs eastward from Port-au-Prince,
many different mountain ranges occur and on all of these so far observed scat-
tered trees of this pine have been seen. Only in the Montagnes du Nord, which
come across the border between Mont Organisé and Banica and extend northwest-
ward, are there extensive areas of pine. Although this area has been invaded

4

to a considerable extent by agriculturalists so that the forest is very broken,

there is nevertheless an area of about 100,000 acres on which pine is the
dominant feature of the landscape and thus it comprises an important forest
section. In addition to this, the extensive central plateau should be men-
tioned since its soil is very unproductive for agriculture and there remain a
great many scattered pines, making it possible to develcp this in the future
as a solid block of pine which would probably be more profitable to the
country than the present farming and grazing uses.

As is te be expected, the site conditions in the various parts of the
range are extremely variable and attest to the great adaptability of the
species. Within the Republic of Haiti alone, pine ranges from an elevation
of about 300 feet above sea-level on the northern coastal plain at Acul
Samedi toe over 9,000 feet in the Morne la Selle section. Soils are derived
from many parent rocks, and scil types are not the limiting factor for the
distribution cf the species as supposed by some previous observers. The
accompanying vegetation varies greatiy so that the species actually trans-
cends several vegetation types. It is really too early for me to attempt to
explain the esological status of the species but so far the only logical
explanation for the occurrence of the pine seems to be tied up very closely
with fire. I would call it an igneous type and believe that it would not
have been able to persist in solid stands without the occasional occurrence
of a conflagration. At the same time, this would explain the scattered
patches of hardwocds which are found inside the blocks of pine. The only
obvious difference between the two such adjacent sites is apparently that of
greater moisture in the hardwood stands.

The tres itseif is a good commercial timber tree. Although the major-
ity of the trees are small due to the fact that the stands have had no atten-
tion and have been at the mercy of unrestricted cutting and burning during
the last century, there still exist many magnificent specimens. I have
measured trees 48 inches d.b.h. in the Morne la Hotte range of the southwest
peninsula and have found trees up to 44 inches in the more accessible Mornes
des Commissaires forest. Specimens 125 feet or more in height are common,
some having a clear bole for 80 feet. The species grows straight and we
have had no difficulty in filling orders for masts 60 feet in length. Not
much is known about the lumber but from general appearance it would seem to
be similar to slash pine, P. caribaea, in quality. Recently, in cooperation
with Duke University and the Tropical Forest Experiment Station, timbers were

= if = October 1942

shipped to North Carolina for timber tests which will throw light on this
question. At any rate, local timber purchasers prefer the native lumber to
that imported from the States. In addition to its lumber, the species is a
naval stores pine which makes it a dual purpose tree.

Naturally, im a country such as Haiti which has been densely populated
for centuries since the early dsys when it was one of the most productive of
colonies, it is very difficult to even estimate what the effects of man have
been on these forest areas. One might suppose that their inaccessibility in
the highest mountains had been an important factor in their persistence, and
yet we have found at a distance of over 50 miles from Port-au-Prince on a
recently constructed road remnants of masonry construction which are none
other than the establishments of the French commissaires for which this
section is named. These were going concerns before 1702 and being located
between the present Morne la Selle and Mornes des Commissaires units must
have exerted some influence on the closely adjacent pine forests. At any
rate, we can safely assume that the pine is a hardy species very resistant
to fire after its sapling stage and also a good reproducer. The dense repro-
duction in protected corners, such as in small triangles at trail intersec-
tions affords excellent evidence of the latter.

With a species of such potential value and possessing excellent sil-
vicultural characteristics, it was a surprise to the author to find in 1939
that Haiti, a steady importer of pine lumber from the States, was doing
nothing with her own pine forests, and several government contacts were made
to discuss the possibility of development. Also, shortly before my visit,
Mr. C. Swabey, Gonservator of Forests in Jamaica, had visited the same areas
and prepared 9 favorable report for the government. Mr. Atherton Lee, then
Agricultural Advisor to the Government of Haiti, assisted in the fostering
of the idea and in the winter of 1939-1940, the progressive group of workers
of the Service Technique at Damien started work at Mornes des Commissaires.

Ths oeginning was crude because it was extremely difficult to get sup-
port and a great number of psople in the capitol sincerely believed that the
Jumber would prove te be of panes quality and not worth transporting to
town. Thanks to the aggressiveness of M. Georges Heraux, Chief of the
Service Technique, they area on with the project even though all of the
first lumber wes pit-sawed and not very uniform due to the inexperience of
the labor avai LLable in that section. These first boards which were brought
to market did at least serve the purpose of destroying the skepticism about
the qualities of the lumber and mili. This was set up early in 1941 and
continued to cperete under the direction of the Service Technique until
October with oscasional official trips being made by the writer from the
Tropical Forest Experiment Stetion in Puerto Rico in an advisory capacity.

This was the picture previous to October of 1941, when the Mornes des
Commissaires, Mcrne le Selle, and Cerca la Source (northern unit) forests
were turned over to the newly organized Societé Haitieno-Americaine de
Developpement Agricole and the writer came to Haiti on an assignment as
Manager of the Forestry Division of the new sorporation, The Société was
organized primarily for the development of rubber and other agricultural
crops but Mr. T. A. Fennel, General Manager of the Corporation and previously

Caribbean Forester - 18 - Vol. 4, Noo 1

Agricultural Advisor to the Republic following Mr. Lee, was personally inter-
ested in the forestry project and sufficiently satisfied with the prospects

to give it the benefit of the capital resources of a large enterprise. Fortu-
nately, the Service Technique had pushed the idea of exploiting the pine
forests beyond the troublesome starting point and had made rapid strides in
such items as fire-control, even though this work was largely confined to the
Mornes des Commissaires Unit.

I think the most significant fact about this whole project is that
Latin American leaders can grasp the idea of profitable forestry and set up
their own projects. Naturally, anyone entering a new field is up against
many technical and organizational difficulties and where it would be impos-
sible for the individual smaller countries to finance and await the training
in technical forestry of their own students in order to make a start, they
could probably help in a ees way to contribute to the maintenance of a staff
of experts at some centrally located point such as the Tropical Forest Experi-
ment Station and thus obtain the maximum benefits of technical assistance for
@ minimum of expenditure. On the other hand, the Station should be organized
so that eventually its own staff may contain several workers from the various
countries whose forestry departments would be serviced by a truly internation
al experiment station.

f

But, to return to the pine forests of Haiti, profiting by the opportu-
nity to obtain sufficient capital for the necessary expansion of the work,
the Forestry Division has made considerable progress in the first few months,
although the organizational work remaining is so great that comparatively
speaking this is just a stars. Lands for the operation of forestry were
secured in a iease by the Haitian Government of approximately 150,000 acres
of pine forest on public lands to the Corperation in exchange for stock in
the Societé The exact boundaries of this land have not yet been defined and
it is a part of the contract that the Government perform this cperation. From
the Division standpeint, this seems edvisable in taking away from iandowners
who may dispute actual forest boundaries any feeling that the Division is
solely responsible for the settlement of the boundary, which in the event of
dissatisfaction could readily be transposed into such destructive measures
as incendiarism-

Since the establishment of the Forestry Division happened to be timed
during difficult shipping periods, it was primarily essential that production
of lumber to meet the needs cf the country be given preference over all other
considerations - This is being done as far as possible although the same
oceanic transportation problems coupled with the scarcity of certain materi-
@ls work against the rapid instellation of new plants. fFians are to cut and
market cne million board feet during the first twelve months and push this up
to three millions during the second year, the latter production equivalent to
the averége importations from the exteri or during the past few years:;

Transportation problems within the Republics constitute the most serious
ebstacle te be oversome and this difficulty will become increasingly more
acute with the growing scarcity of rubber tires. As a long time solution to
the problem a primary road system is being constructed and fire-breaks which
are suitable for log transportation in good weather now are laid out in such
@ manner that they can later be improved to constitute the secondary road

=- 19 = October 1942

system of the forests. For the solution of the immediate problem of trans-
portation, which can be assumed to last anywhere from one to five years, we
are making plans to drop back to more simple means of transportation such as
ox-carts and water. Thus, locations of new mills and sites selected for re-=
Location of existing mills are on the sea-coast or at lower elevations where
hauls to them may be made by oxen or river driving and where the transporta-
tion to market of the finished products may be accomplished by ox-cart or
small coastal schooner.

The pressure for production coupled with the minimum of trained per-
sonnel makes it necessary for the time being to neglect the finer points of
forest management and put the maximum emount of attention on the proper lo-
cation and operation of the cutting and milling. This has entailed a large
amount of preliminary reconnaissance on horse-back, on foot, and in small
river boats. And, in spite of the need for rushing this type of work, it
must be essentially accurate in order that present developments may fit
inte future management plans, which will be developed after the accumulation
of sufficient factual data, with the minimum of changes possible. At the
same time, mapping work has been started and although far behind present
needs will be availiable for some areas within a relatively short time. As
fast as such maps become available, preliminary developmental plans which
exist now only in our minds or in the form of sketchy notes, will be written
down in more permanent and quantitative records.

Training of personnel is another tremendous job which must be carried
out on all phases of the work and with all classes of employees. This is too
important a phase to be neglected, and since the early trained workers are
expected to pass their learning on to @ progressively increasing number,
special pains are being taken to see that these first men obtain the correct
training at the start. The forest administration system is being built
around the femiliar District Ranger. Local college trained men are selected
for these positions and because no forestry trained men are available this
type of practical, educated workers must carry the load untii forestry gradu-
ates are available in the future. j

Fire-sontre) is an important factor in the satisfactory stocking of
the forest since reproduction dees not become fire-hardy until fifteen or
twenty years of age. Due to the nature of the topography and the fact that
the major portion of the fires are caused by peasants, emphasis of the fire
control program has been placed on fire prevention rather than on organi-
zation for fighting fires. Agricultural police are located throughout the
forest areas, who although under the direct supervision of the Forestry Di-
vision are government agents and thus have the power to arrest and take to
local courts any one who is found burning without permission within or in
the vicinity of the forests. This system is working very satisfactorily and
where forest developments are under way and. providing work to the local resi-
dents the pines have already attained a high place in the respect of the
native so that there should be less and less trouble with fires as the various
units reach a higher stage in development.

Presumably, control burning will be necessary in the future in order to
exclude hardwood trees and brush from the pine stands. Botanical collections

Caribbean Forester = 20 = Vol. 4, No. lL

as @ basis for an understanding of the ecology of the areas and plots to

be established in the near future will provide us with the factual basis

as to whether or how this will have to be carried out. The forests will

be managed on the uniform silvicultural system which can easily include
controlled burning. This essentially will consist of burning at the time

of main fellings followed by from fifteen to twenty-five years of strict
protection so that a new crop of seedlings becomes established and attains
the size of large saplings with the thickened bark which gives them immunity
to fire damage.

Division headquarters are being constructed at Mornes des Commissaires,
about one mile above sea-level where the climate is invigorating and very
healthful. At this place the temperature drops down to the freezing point
just about once every year. In the vicinity, are areas of good fertile soil
where northern vegetables thrive and thus, with the tropical vegetables and
fruits which are brought in from the lowlands, providing a broad range of
diet. All these factors are considered important in maintaining the adminis-
trative staff in good working condition.

As one side line to the management of the pine forests, this same rich
soil of the high plateau is being tested for its suitability for the produc-
tion of Cinchona. Itis still far too early to know what the ultimate chances
for success with this crop will be but it is hoped that success may be ob-
tained as this is not only a very much needed crop at the present time but
also is one which would very well complement the forestry activities in this
region.

As a final point, it might be of interest to mention the financial pos-
sibilities of the pine forests. So far the lumber operations are easily able
to pay their ecwn way. Forest developments such as the road building and
construction ef Division headquarters are tco heavy a load to be borne by
present timber exploitation and are therefore capitalized but there are excel-
lent possibilities of cancelling out all of these developmental costs within
a relatively few years at which time it should be possible to proceed with the
developmen; of other forest types in the Republics.

Summary

The native pine of Haiti represents one of her most important forest
resources. This tree with a wide range in the Republic is a valuable timber
and naval stores species which fortunately possesses good silvicultural
characteristics. Initial development was started by the Service Technique
in 1939 but since October 1941, 150,000 acres have been turned over to the
Société Haitiano-Americaine de Developpement Agricole, which has carried on
a program primarily. designed to provide for the timber needs of the country
but also constituting a long range development plan with construction of
roads, fire protection, mapping and the necessary investigations for setting
up proper management. The first few months of operation indicate satisfactory
financial possibilities.

= 21 = Octeber 1942

Resumen

El pino oriundo de Haitf, Pinus occidentalis, constituye uno de los
recursos forestales més importantes de ese pais. Este a&rbol, que estdé exten-
samente distribuido en la republica suministra madera muy util y trementina,
y afortunadamente, posee caracteristicas silvicolas favorables.

Es dificil encontrar especimenes desarrollados debido a las cortas sin
distincion a que fué sometido en el pasado. Sin embargo en algunos sitios hay
arboles de 44 pulgadas de d.a.p. y 125 pies de alto. Son bien formados y por
consiguiente producen madera de grano recto que tiene buena demanda. Consi-
derando estas circunstancias es un hecho sorprendente que en época tan recien-
te como 1939 no se estuviese explotando sistemaéticamente.

En 1941, 150,000 acres de tierras forestales fueron puestas bajo la
administracion de la recién organizada Société Haitiano-Americaine de
Developpement Agricole que est& llevando a cabo un programa destinado no sdlo
& proveer al pais de la madera que necesite sino que incluye un plan de gran
alcance para construccién de carreteras, proteccién contra incendios, delinea-
miento de mapas y las investigaciones necesarias para el manejo apropiado de.
este recurso. Los problemas mis importantes con que se confronta la Adminis-
tracioén son aquellos propios de cualquier orgeanizacién similar en un pais
donds la Dasonomia esta en sus comienzos, Hoy existe adem&s el problema de
la transportacion debido a la escasez de neum&iticos y también la dificultad
de adquirir equipo y maquinarias necesarias. Ha sido necesario hacer estu-
dios y reconocimientos en gran escala répidamente aunque con la necesaria
exactitud, La labor de los primeros meses ya indica que cuando la inversién
inicial se distribuya sobre las mejoras permanentes, se demostraré un bene-
ficio o ganancia en el negocio.

o0o

Errata

On page 75 of Vol. 3, No. 2, paragraph 2, “Calophyllum globulifera"
should have been "Calophyllum brasiliense™.

The headings of the tables on pages 93 and 94 of Vol. 3, No. 3 should
read "Girth class in feet".

In the table of Non-American woods on page 172, Vol. 3, No. 4, Jarrah
should have been classified as an Australian, rather than an Indian wood.

Caribbean Forester = 22 = Vol. 4, No. 1

CREOSOTE PENETRATION IN TABONUCO WOOD AS AFFECTED BY
PRELIMINARY BOILING TREATMENTS IN ORGANIC SOLVENTS

David Reidl/
Duke University School sof Forestry

Most Puerto Rican woods are not destined to become important on a
large commercial scale but rather will be useful only for local consumption.
This is explained by the fact that approximately 80 per cent of the timber
resources on the island have been exploited in order that land could be made
available for agricultural use.2/ In time some of the land now in agricul-
ture may be available for reforestation. Specific information as to potential
uses of the woods and proper management practices must be determined in order
that species of useful qualities may be produced.

This paper deals specifically with the problems in the preservative
treatment of tabonuce, Dacsryodes excelsa Vahl, one of the more important
Puerto Rican species. A review of the literature indicates that the preser-
vative treatment of tropical timbers has received little or no attention.
Therefore, as a part of a cooperative program of research dealing with the
physical-mechanical properties of important Puerto Rican timbers being
carried on by the Duke University School of Forestry and the Tropical Forest
Experiment Station, a study was initiated with e two-fold purpose: (1) to
ascertain correct treating schedules for tabonuco, and (2) to present some
of the problems arising from attempts to force preservatives into woods of
this kind.

Selected tabonuco trees in Puerto Rico were felled and bucked into
four-foot lengths. The bolts were then sawn into test-specimen flitches
according to recommendations of the American Society for Testing Materials.
The flitches, after coding, were wired into bolt form and the ends painted
with tar to prevent drying and checking during shipment. Upon receipt of
the material at Duke University, the flitches were re-sawn and milled into
standard test specimens used for determining strength data. The surplus
stock, largely heartwood, was cut and milled into 3 inches by 23 inches
material and allowed to air dry for five months before being used in this
study -

if A thesis submitted in partiel fulfillment of the requirements for the
degree of Master of Forestry in the School of Forestry of Duke University.

The writer wishes to acknowledge his indebtedness to Dr. E. S. Harrar for
directing the experimental work. Appreciation is expressed to Professor

A. E. Wackerman for his many valuable suggestions concerning the preservative
treatment, and to Professor F. X. Schumacher for his advice on the statistical
analysis.

af Gill, Tom. Tropical Forests of the Caribbean. Baltimore, Maryland, 1931.

= 25 = October 1942

Preliminary Study

The original proposal was to make a comparative study of oil-preserva-
tive retentions in tabonuco wood, using the Bethell and Rueping Processes.
Preliminary experiments of preservative treatment were conducted by the Bethell
Process, Rueping Process, and a modification of the Bethell Process using creo-
sote as the preservative.3/ These methods of oil impregnation failed to ac-
complish any penetration in the heartwood although the sapwood, when present,
was completely penetrated. Failure to obtain penetration of the heartwood by
creosote prompted the use of chromated-zinec chloride, a water soluble salt.
Negative results were obtained also with this salt using the Bethell Process
and a modification of the Bethell Process.

Hunt and Garratt4/ assert that three main factors affect penetration;
viz; the anatomy of the wood, the preparation of the timber, and the treating
procedure. Since the preparation of the timber and the treating procedure
were controlied in this study, it was thought that some anatomical feature of
the heartwood was prohibiting penetration.

Microscopic examination of the wood showed that the fiber walls are
exceedingly thick and possess very minute pits. Since these gross anatomical
features were found to be characteristic of both heartwood and sapwood, they
were not considered as factors prohibiting penetration because, as previously
indicated, preliminery experiments resulted in excellent sapwood penetration
but no heartwood penetration. It is quite evident that when the wood changes
from sapwood to heartwood the fiber walls are infiltrated with materials that
block the passage of preservatives. Practically no organic material is
conteined within the intercellular spaces except in the vessels which consti-
tute only a smell. percentage of the total wood volume. Therefore, the infil-
trated material in the fiber walls of the heartwood, as will be indicated
later, apparently affects preservative penetration much more than the included
material within the vessels and other cell cavities.

It was concluded, therefore, from the results of the preliminary pre-
servative treatments and the microscopic examination of the wood, that it
would be necessary to subject the wood to a pre-treatment before impregnation
with &@ preservative was attempted.

Accordingly, this paper deals with the effects of certain boiling
treatments on the penetration of a creosote preservative in tabonuco wood.

Experimental Technique

Preparation of material

ecimens 1-1/4 in. by 1-1/4 in. by 5 in. were cut and milled from the
3S ine by 3 in- by 23 in. stock and placed in a drying oven maintained at 100

3/ Anonymous. Manual of Recommended Practice. American Wood Preservers
Association. 1941

4/ Hunt, George M. and George A. Garrat. Wood Preservation. McGraw-Hill
Book Co., Inc., New York and London. 1938.

Caribbean Forester = pon Vol. 4, Now 1

ee

|
|

a

degrees Centigrade. After the specimens were oven-dry, 60 of the clear heart-
wood pieces were chosen for this study. These were numbered consecutively
with steel dies on the ends and sides, and placed in a dessicator in order to
maintain the oven-dry condition until ready for use in subsequent treatments.

Boiling treatment

Four organic solvents were chosen for this phase of the study: viz:
Distilled water, alcohol, kerosene, and xylene. A set of three oven-dry
specimens was taken from the dessicator, their respective numbers and weights
recorded, and then placed in a flask and flooded with 600 cc. of one of the
solvents.

Boiling the specimens was accomplished by using a reflux condenser
system as shown in Figure 1. The six liter flask was heated by an electric
hot plate in order to control temperature during the boiling process since
the four solvents used possessed very different boiling points.

The boiling treatment lasted 10 hours after which time the specimens
were removed from the flask, weighed, and then placed in a closed container
to prevent drying. This procedure was followed with each of the solvents ©
chosen, using a new set of three oven-dry specimens for each boiling treat-
ment.

Preservative treatment

Preservative treatment was undertaken in an experimental wood preser-
vation plant. Figure 2 shows details of the plant construction. This plant
is designed to withstand working pressures of 200 pounds per square inch.
The vacuum system can effect and maintain vacuums as high as 28 inches of
mercury. Heating the preservative is accomplished by means of steam heating
coils located in the storage tanks, charging tank, and treating cylinder.
Live steam may be forced into the treating cylinder in order to give timber
steaming treatments preliminary to introduction of the preservative. Gauges,
valves, and thermometers are connected to the system which permit continuous
observation and control of temperatures, vacuums, and pressures for the
duration of the treating procedure.

After the four sets of three specimens had received a boiling treat-
ment in their respective solvents, they were encased between two steel end-
plates along with three oven-dry specimens which served as controls. The
end=-plates were used to minimize end-penetration of creosote into the wood;
this study being concerned only with side penetration. These 15 specimens
were placed in the treating cylinder and subjected to a steaming period of
50 minutes at a steam pressure of 25 pounds per square inch. This was fol-
lowed by & vacuum period of 25 to 28 inches of mercury for 30 minutes. Final-
ly, without breaking the vacuum, creosote at a temperature of 190 degrees
Fahrenheit was introduced into the cylinder and held under a pressure of 200
pounds per square inch for 90 minutes. The properties of the creosote used
are presented in Table 1. The records of the preservative treating schedule
used are given in Tables 2 to 5.

= 25 = October 1942

Figure 1.—Reflux Condenser System Used for the
Boiling Treatment of Wood Specimens.

. Figure 2.—Side View of the Treating Cylinder.

Table 1.—Properties of Creosote Used in Preserving Tabonuco.

Standard Distillation
of FractionsL
UpecorLorc.

Specific Gravity | Water Content

1.103 1.5 3.3% 12.6%

a/ Percentage by weight. —

Measurement of preservative penetration

After the preservative treatment, a transverse sample wafer, one-
guarter inch in thickness, was sawn from each specimen. These samples were
representative of side penetration since they were taken from the middle of
the specimens. Arranging them in columns according to their respective
boiling treatments, they were photographed in order to record depth of creo-
sote penetration. The negative was placed in a photographic enlarger and
the cross-section images, showing the outline of the specimens and region
of preservative penetration, are shown in Figures 3 to 6.

= 27 = October 1942

Table 2.—Percentage of Cross-Sectional Area of Specimens Penetrated
by Creosote Preservative - First Replication.

Boiling Treatment

Alcohol Control Kerosene Water Ay lene
% Tp % % I,

36.6 55.0 93.7 79.0 72.6
ST0e 42.4 98.0 71.9 60.3
41.5 3505 96.8 85.3 45.1

Figure 3.—Transverse Samples of First Replication
Showing Regions of Preservative Penetration.

- Pre-treated in Alcohol.

- Oven-dry Control.

Pre-treated in Kerosene.
Pre-treated in Distilled Water.
- Pre-treated in Xylene.

a SR QS
0

Caribbean Forester - 28 = Vol. 4; No. i

Table 3.—Percentage of Cross-Sectional Area of Specimens Penetrated
by Creosote Preservative = Second Replication.

Boiling Treatment

Alcohol Control Kerosene Water Xylene
Ve jos qe To Ui
Alol 45.1 44.6 3250 29.6
38-3 55.1 42.3 27.9 30.4
$5.0 48.6 46-6 RATE 33.8

Figure 4.—Transverse Samples of Second Replication
Showing Regions of Preservative Penetration.

A
C
Kae
W
x

Pre-treated in Alcohol.

- Oven-dry Control.

Pre-treated in Kerosene.

- Pre-treated in Distilled Water.
Pre-treated in Xylene.

= 29 = October 1942

Table 4.—Percentage of Cross-Sectional Area cf Specimens Penetrated
by Creosote Preservative = Third Replication. |

Boiling Treatment

Alcohol Control Kerosene Water Aylene
fo fo lo () VA
40.5 45.7 71.2 87.5 45.9 :

S108 42.0 63.7 29.4 3109

:

Figure 5.—Transverse Samples of Third Replication
Showing Regions of Preservative Penetration.

A - Pre-treated in Alcohol.

C - Oven-dry Control.

Pre-treated in Kerosens.

W - Pre-treated in Distilled Water.
X - Pre-treated in Xylene.

ae
j

Caribbean Forester - 30 = Vol. 4, No. 1

Table 5.—Percentage of Cross-Sectional Area of

Specimens Penetrated

by Creosote Preservative = Fourth Replication.

Boiling Treatment

Alcohol Control Kerosene - Water Xylene
% ae: % wee Sp
40.5 43.5 53.6 37.9 68.8
2701 38.6 78.0 59.2 51.1
26.2 45,1 97.5 26.8 51.8

Figure 6.—Transverse Samples of Fourth Replication

Showing

Me aRaQPr
1

Regions of Preservative Penetration.

Pre-treated in Alcohol.
Oven-dry Control.
Pre-treated in Kerosene.

Pre-treated in Distilled Water.

Pre-treated in Xylene.

= $1 s

October 1942

The total area and the penetrated area cf each specimen was determined
from the tracings by using @ planimeter. The percentage of creosote penetra-
tion was established for each of the fifteen specimens by dividing the value
of the penetrated area by the value of the total area.

Compilation of data

Four replications of 15 specimens each supplied the data as shown in
Tables 2 to 5. Each datum expresses the per cent of cross-sectional area of
the specimen penetrated by the creosote. Figures 3, 4, 5, and 6 show the
region of penetration for each specimen in the four replications.

Analysis of variance

An analysis of varianceS/ was computed from the data in Tables 2 to 5.
This was used to determine whether or not the boiling treatments had a signi-
ficant effect on preservative penetration for the treatment given.

The 60 observations in Tables 2 to 5 were summed according to replica-
tion and specific boiling treatment which produced 20 values composed of three
observations each. The analysis of variance is shown in Table 6.

Table 6.--Analysis of Variance Between Replications and Treatments

Degrees Sums
Source of of Mean
Freedom Squares Squares
Replication 3 12646 .224 4215.408
Treatment 4 24836.458 6209.458
Error 12 17691..646 1474.471
Total 18) 95176.328

The minimum significant difference at the one per cent leve16/ is
27.65. Therefore, any boiling treatment, to have a significant effect on
preservative penetration at the one per sent level, must have an average
penetration per cent as great or greater than the average per cent of the
oven-dry control, plus 27.65. Therefore the kerosene boiling treatment, with
an average penetration per cent of 70.5, is the only treatment having a sig-

nificant effect on preservative penetration,
The Standard Error for the average penetration per cent of each boiling

treatment was found to be £ 6:4 per cent.

5/ Fisher, R. A. Statistical Methods for Research Workers. Sixth Edition-
Revised and enlarged. London, England. 19384
6/ Probability less than 1 in 100 that obserwed difference is due to chance.

Caribbean Forester = 32 = Vol. 4, No. 1

Observations and Conclusions

From the statistical analysis of the data on percentage of creosote
penetration, the kerosene boiling treatment was the only one of the four
showing a@ significant effect at the one per cent level on preservative
penetration when compared to the oven-dry control.

The average penetration percentage of the 15 specimens in each treat-
ment gave the following values: kerosene, 70.3 per cent; distilled water,
50.4 per cent; xylene, 47,5 per cent; and alcohol, 35.2 per cent; the oven-dry
control being 42.5 per cent. The standard error of each of these averages is
+ 6.4 per cent. The alcohol boiling treatment indicated a retarding effect
on preservative penetration as compared to the oven-dry control. Specimens
that received the distilled water and xylene treatment showed 4 greater average
penetration percentage than the control but not a significant difference, such
as was the case with the kerosene boiling treatment.

The photographs of the cross-sections of the specimens clearly show
that the side penetration of the preservative into the specimens was very
irregular. This may be attributed largely to the effect of interlocked grain
which is an anatomical characteristic of tabonuco wood.

Even though the kercesene treatment is the only one showing a significant
effect on penetration, it would be impractical to apply on @ commercial basis.
Incising of pre-framed timber is a possibility that merits consideration since
some of the refractory woods growing in the United States are treated commer-
cially after such preparation (Hunt and Garratt).

There are specific problems brought out by this study that are deserv-
ing of further consideration. A chemicel analysis of the solvents, after the
boiling treatment would show what materials, if any, had been extracted from
the wood. Also it would be desirable to know the effect of the boiling treat-
ment on the cell walls and the mechanical properties of the wood.

It would also be interesting to conduct a series of service tests in
Puerto Rican soils on the treated wood. Data of this sort could probably be
used to indicate how much penetration of @ preservative is required to render
the wood safe from fungal and insect attack under local conditions.

Summary

}

s

sa Vahl, is very refractory
and cannot be treated successfully by standard preserving methods. ©

2. The presence of organic material infiltrated in the cell walls was
assumed to be the factor prohibiting penetration rather than gross
anatomical features.

5. Pre-treatment by boiling samples of

eartwood in organic solvents
to extract infiltrated materials, befo

h
efore subjecting them to

= $3 = October 1942

preservative treatment showed an average percentage effect on
creosote penetration for the fifteen specimens in each treat-
ment as follows:

Kerosene 70.3 ~ 6.4%
Distilled Water 50.4 = 6.4%
Xylene 47.5 = 6.4%
Alcohol 35.2 ~ 6.4%
Oven-dry Control 42.5 = 6.4%

4. Results in this study show that only the preliminary boiling
treatment with kerosene had significant effect at the one per
cent level on preservative penetration.

Resumen

Recientemente la Escuela de Técnicos Forestales de la Universidad de
Duke llevé a cabo un estudio sobre la penetracién de preservativos en la ma-
dera del tabonuco, Dacryodes excelsa, una de las especies de arboles més im-
portantes de Puerto Rico.

Se encontré que los preservativos penetraban con facilidad en la albura
mientras que el duramen, que contiene substancias inhibidoras, requiere un
tratamiento preliminar para remover o modificar dichas substancias para que
la penetracién sea satisfactoria.

Se trato una ebullicién preliminar en cuatro solventes: alcohol, kero-
sina, agua destilada y xileno. Se hirvieron especimenes en cada uno de estos
liquidos por espacio de 10 horas. Después se encajaron entre dos laminas de
acero para impedir que la penetracién se efectuase por las extremidades y
luego se sometid al tratamiento con preservativo.

El tratamiento con preservativo comenzé aplicando vapor por un periodo
de 30 minutos a una presién de 25 libras por pulgada cuadrada. A éste siguid
un periodo en un vacio de 25 a 28 pulgadas de mercurio por 350 minutos. Final-
mente, sin destruir el vacio se introdujo creosota a una temperatura de 190
grados Fahrenheit en el cilindro y se le sometié entonces a una presién de
200 libras por pulgada cuadrada por espacio de 90 minutos.

Después de la preservacién se tomé una muestra transversal que se corté
del centro de cada pedazo. El por ciento de drea penetrada con respecto al
area total del corte se toméd como medida indicativa de la efectividad de la
penetracién.

El unico tratamiento que tuvo efecto significativo positivo sobre la
penetracién desde el punto de vista estadistico fué la kerosina. De los demé&s
el agua destilada y el xileno tuvieron efecto positivo en la penetracién
mientras que el alcohol tuvo efecto negativo.

Caribbean Forester = $4 = Vol. 4, Noo. 1

CATALOGUE DES CRYPTOGAMES VASCULAIRES DES ANTILLES FRANCAISES

H. Stehlé
Ingénieur Agricole et d*Agronomie Coloniale
Martinique

Le présent catalogue comporte l1'énumération systématique des Ptérido-
phytes ou Cryptogames vasculaires des Antilles frangaises récoltées princi-
palement par L'Herminier, Mazé et Duss @ la Guadeloupe, par Duss et par Hahn
& la Martinique, ainsi que par l’auteur entre 1934 et 1942, avec l’aide de
Madame M. Stehié et du R. P. Lo. Quentin, collecteur infatigable et botaniste
avisé, dans l’une et l'autre de ces iles. A un petit nombre d’espéces prés,
dont la difficulté de rassembler la bibliographie compléte en milieu tropical
isolé est la raison, la totalité des plantes apparteneant aux Fougéres,
Lycopodes et Selaginelles des Antilles frangaises, figure dans le présent
travail.

L'ordre suivi et la nomenclature adoptée sont ceux de Wm. R. Maxon,
Curator of Plants, Smithsonian Institution, de Washington (Studies of tropical
American ferns et Pteridophyta of Porto Rico and the Virgin Islands, 1926),
spécialiste le plus remarquable des Ptéridophytes d‘Amérique et des Antilles.
Il a bien voulu étudier nos récoltes au cours de ces 8 derniéres années et
reviser nos déterminations, ce pourquoi nous lui adressons ici l’expression
de nos plus vifs remerciments.

La derniére publication relative aux Ftéridophytes des Antilles
francgaises est celle du R. P. Duss (Division, Nomenclature et Habitat des
Fougéres et Lycopodes, Lens-le-Saunier, 1903), dans laquelle de nombreuses
especes sont désignées par des binémes aujourd'hui caducs et que les botanis-
tes les plus récents ont classés en synonymie.

Nous nous sommes référés dans cette étude 4 1l'Histoire des Fougéres et
des Lycopodiacées des Antilles (Paris, 164 p., 34 tab., 1866), par A. Lo A.
Fée, publication magistrale et aux monographies les plus récentes des ptérido-
graphes des Antilles voisines, en particulier & Urban: Symbolee Antillanae,
Flora Portoricensis, Vol. IV, 1903-1911, Pteridophytae p. 5-70 (1903) et
p- 648-651 (1911), Pteridophyta domingensia, Vol. IX, 1925, p. 273-397 et &
Domin: Pteridophyta of the Island of Dominica (Prague, 1929).

Dans notre Essai d‘Ecologie et de Géographie Botanique de la Guadeloupe,
tome I, Basse Terre, 1935, ainsi que dans l'’Esquisse des Associations végételes
de la Martinique (in Bull. Agr. Mart. 6, 193-264, 1937), nous avons cité des
Fougeres et végéteaux alliés récoltés dans les différents secteurs, mais
d‘'apres leur répartition et leurs exigences édapho-climatiques. La publication
suivante a surtout pour objet de donner une idée de la richesse en ptérido-
phytes des Antilles francaises, de faire connaftre, sous leurs dénominations
les plus récentes et les plus exactes, la plupart des espéces de notre flore
et les raretés que nous y avons récoltées au cours de ces derniéres années,
surtout dans les secteurs de la chaine centrale des deux fles qui avaient été

=- $5 = October 1942

moins explorés. Les numéros indiqués isi sont ceux de nos collections et
nous les faisons précéder du nom du collecteur. Enfin, les lettres G. et M.
correspondent respectivement @ la localisation en Guadeloupe et en Martinique.

Nos collections antérieures & 1938 sont déposées au Muséum d'Histoire
Naturelle de Paris et celles récoltées ultérieurement figurent dans l'herbier
de H. et M. Stehlé, actuellement & la Martinique. Tous les doubles des
Ptéridophytes ont été adressés & 1'Herbarium de la Smithsonian Institution &
Washington.

L'énumération faite ci-aprés des Fougéres et Alliées des Antilles
francgaises fait ressortir un total de 56 genres et 313 espéces en écartant
les espéces douteuses, les variétés et les formes. Ces plantes sont réparties
en 10 familles, chacune représentée par un petit nombre d'especes & l'exception
de la famille des Polypodiacées qui a elle seule groupe presque les trois-
quarts des espéces énumérées: 225 réparties dans 44 genres. C'est aussi la
plus représentée dans le monde ou on lui connait environ 6500 espéces. Ces
chiffres montrent la richesse ptéridophytique des Antilles frangaises parti-
culiérement élevée puisque sur un total d’environ 2500 végétaux vasculaires
(Phanérophytes et Ptéridophytes), le groupe des Fougeres et plantes affines
est représenté dans le rapport de 1 pour 7, 8 plantes, alors que Fée, dans
son introduction @ l°Histoire des Fougeres et des Lycopodiacées des Antilles
(p. IX) en 1866, indiquait pour les Tropiques et l1'Equateur que les Fougéres
"forment environ le neuviéme de la végétation totale".

Tableau Comparatif de la Richesse Pteridophytigque
des Antilles Frangaises

Antilles frangaises | Puerto Rico
(H. Stehlé) (W. R. Maxon)

Dominique
(K. Domin)

Familles

Ophioglossacese 2 2 3 =
Marattiaceae L fs) 4
Osmundaceae = = 1
Ceratopteridaceae = = i =
Schizaeaceae = = 6 =
Glacheniaceae 1 6 3 2
Cyatheaceae 3 9 12 5
Polypodiaceae (y compris
Dicksoniaceae) 44 225 205 140
Hymenophyllaceae 2 39 24 16
Salviniaceae = = 1 =
Marsileaceae = = 2 =
Lycopodiaceae 1 17 13 10
Psilotaceae 1 2 1 L
Selaginellaceae 1 9 9 4
Total 56 513 285 180

Caribbean Forester = $6 = Vol. 4, No. lL

Il nous parait intéressant d‘indiquer en-un tableau comparatif les
familles représentées dans notre flore, le nombre des genres et celui des
especes par famille et de comparer ce dernier nombre @ celui de la Dominique,
fle anglaise voisine située entre la Martinique et la Guadeloupe et a celui
de Porto Rico, tile la plus petite et la plus proche des Grandes Antilles.
Dans ces deux tiles voisines la connaissance des Fougéres a été trés poussée -
grace aux remarquables monographies de Maxon (1926) et de Domin (1929).

En ce qui concerne Puerto Rico et les Antilles frangaises, les chiffres
indiqués ci-dessus serrent certainement de trés pres la réalité alors que pour
la Dominique ce chiffre, bien que sensiblement inférieur @ celui des Antilles
francaises, doit & notre avis, @tre bien plus élevé. Dans Domin, il y a un
nombre plus élevé d'espéces que celles consignées, d’aprés notre décompte dans
ce tableau, mais nous n‘y avons pas fait figurer naturellement celles qui,
dans ce travail intitulé "The Pteridophyta of the Island of Dominica", sont
citées par l’auteur lui-méme comme des endémiques de Trinidad, de la Jamsique
ou méme de Colombie, celles qu'il énonce comme existant aux Antilles
francaises et pouvoir 6tre trouvées un jour @ la Dominique ob on ne les a
pas collectées encore et enfin les douteuses les variétés et formes pour
pouvoir comparer avec Puerto Rico et les Antilles frangaises dans des
conditions égales. W. H. Hodge au début de ses Notes on Dominican Ferns
(in Amer. Fern Journ. Vol. 31, No. 3-4, 1941) estime que ce grand groupe de
plantes, dominant en foréts humides dans la Dominique montagneuse est re=
présenté par environ 400 espéces. Ce chiffre nous parait tres élevé car si
l’on ajoute aux 180 especes dominiquaises de Domin les 23 trouvées en outre
de celles connues pour cette ile, par Hodge au cours de ces 4 dernitres
années d'investigation nous arrivons seulement & un total de 203 espéces.

Il semble donc que la Dominique ait été insuffisamment explorée mais que
liabondance des Fougéres qu’elle héberge demeure notabliement moins élevée
que celle de Puerto Rico, de la Martinique et de la Guadeloupe. Une visite
malheureusement trop courte de la Dominique en 1935 en nous rendant de Gua-
deloupe en Martinique, nous a semblé & premiére vue correspondre avec les
résultats du précédent tableau comparatif.

Les chiffres indiqués ici ne sont pas intangibles: ce sont des relevés
mis en lumiére d‘apres les récoltes effectuées et les publications les plus
récentes; la végétation des iles évolue, des découvertes peuvent étre
réalisées, les conceptions varient également. Notre intention en rédigeant
ce catalogue a été surtout de faire un recensement tenu & jour des Ptérido-
phytes des Antilles frangaises, de démontrer leur richesse floristique, de
donner les noms logiques et les binémes rationnels des especes avec la
réference des ouvrages dans lesquels leur publication et description ont été
données afin de combler la lacune qui sur ces points, existait dans la Flore
Cryptogamique de Duss de 1903, seul traveil antérieur sur les Fougéres des
Antilles frangaises.

- $7 = October 1942

4.
5.0

Ordre I. OPHIOGLOSSALES
Famille 1. OPHIOGLOSSACEAE
1. OPHIOGLOSSUM (Tourn.) Le

Ophioglossum retisulatum L. Sp. Pl. 1063. (1753). G.: S.n. 476, n. 927 et
M.S. n. 3850 et no. 4106.

2. CHEIROGLOSSA Presl.

Cheiroglossa palmata (L.) Presi, Abh. Béhm. Ges. Viss. V. 4:317 (1845), G.
Duss n. 43508.

Ordre II. MARATTIALES

Famille 1. MARATTIACEAR

1. DANARA J. Eo Smith

Danaea nodosa (L.) J. E. Smith, Mém. Acad. Scienc. Turin 52420. (1793) G.

Mazé n. 64 et Duss n. 4317, n. 4519 et M. Duss
no 1690.

Danaea elliptica J. E. Smith, in Rees, Cycl. II. n. 2 (1808), G. Duss
n. 4316, n,. 4336 et M. (ex Maxon).

Danaea alata J. E. Smith, in Mém. Acad. Science. Turin 5:420. (1790), G.
Duss n. 4318 et M. Duss n. 1687.

Danaea stenophylla Kunze, G.S. n. 534, n. 12235 et M.S. n. 3447.

Danaea Mazeana Underw., G.S. n. 1811.

Ordre III. FILICALES
Famille 1. SCHIZABACEAE
1. ACTINOSTACHYS Wall.
etinostachys ponnula (Sw.) Hook. Gen. Fil. pl. III, A. (1842). G. Duss
n. 4348.
Actinostachys Germaini Fée, Hist. Foug. et Lycop. Ant. 29:123 (1866) G.
Duss no. 4349.
2. ANEMIA Sw.

Anemia hirta (L.) Sw. Syn. Fil. 155 (1806), G. Duss n. 4306 et M. Duss
no 1689 et n. 4579.

Anemia adiantifolia (L.) Sw. Syn. Fil. 157. (1806). G.S. n. 89, no 192,

n. 540, n. 840, n. 1842, n. 2016, n. 2534,
n. 2704 et M.S. no. 4159.

Caribbean Forester = 38 = Vol. 4, Noo 1

Famille 2. GLEICHENIACEAE
1. DICRANOPTERIS Bernh.

Dicranopteris flexuosa (Schrad.) Underw. Bull. Torrey Bot. Club 34:254.
(1907). M. (ex Maxon).

Dicranopteris pectinata (Willd.) Underw. Bull. Torrey Bot. Club 34:260
(1907), G» Duss n. 4310, n. 4311 et M. Duss
n. 1694, n. 1696.

Dicranopteris bifida (Willd. ) enou N. Amer. Fl. 16:60 (1909), G.S. n.
94, n. 653, no 1441 et M.S. n. 3419.

4. Dicranopteris Bancroftii (Hook,) Underw. Bull. Torrey Bot. Club 34 (1907),

5e

60

G.S. no 2424.
Dicranopteris furcata (L.) Underw. Bull. Torrey Bot. Club 34:257 (1907),
GaSaune 2415.
Dicranopteris farinosa (Kaulf.) Underw. Bull. Torrey Bot. Club 34:254
(1907), G. Duss no. 4312 et M. Duss n. 1693 b.

Rapportée par Duss & Gleichenia subtrisperma
(Fée) Duss, synonyme.

Famille 3: CYATHEACEAE
1. CYATHEA

Cyathea arborea (L.) J. E. Smith. Mém. Acad. Turin 5;417 (1793). G. S.
no $32 et M.S. no. 5297, n. 3309 et n. 3329.

Cyathea serra Willd. Spec. Plant. 5:1,491 (1810), G. Duss n. 43522, ne
4546, n. 4591 et M. Duss n. 1604. Espéce tres
affine de C. arborea (L.) J. E. Smith.

Cyathea tenera (J.Smith) Griseb. Fb. Br. W.I.I. 704 (1864), M. (ex Urban,
Symb. Ant. 9:289 (1925).

2. HEMITELIA R. Br.

- Henitelia grandifolia (Willd.) Spreng. Syst. 4. 1. 125, (1827). G.S. n.

96, eae SSs ne MO2 sm. TO ne eed ne 7.74!

3 : ‘ et M. Duss n. 4605.
Hemitelia horrida Duss, non R. Br., G. Duss n. 4155 et M. Duss n. 1605 et

no 1610, se rapporte & H. grandifolia. Il en
est de méme de H. obtusa Duss, G. n. 4153 et M.
Duss n. 1607 et n. 1608, non Kaulf., limité a
Grenade et Saint Vincent (ex Maxon).

Hemitelia muricata (Willd.) Fée, Gen. Fil. 350 (1850-52). G.S. n. 9, ne
$21, no 5326, n. 529, n. 527, n. 827 et M.S. n.
3414 et n. 3426. Cyathea Tussacii Desv. (Syn.
C. Imrayanea Hook. et C. Beryi L'Herm.) n'est
autre que cette espece. Alsophila nitida Kuntze
basé sur des spécimens de la Martinique s‘y
rapporte également. Elle est maintenue par Domin
Pterid. Dominica, 66 (1929), comme esp&ce valable,
sous le binéme C. Imrayana Hook., qu’il considére
comme trés différent de C. Tussacii Desv.

>

- 39 = October 1942

Alsophile aspera R.Br. ex Duss, Foug. et Lycop.
Ant. fr. 20 (1903), G. Duss n. 4157, n. 4323 et
M. Duss n. 4602, se confond aussi avec H. muricata
(Cf. Maxon, Stud. trop. ferns 5, in Contr. U. S.
Nat. Herb. 419 (1914).

7, Hemitelia Kehautiana (Presl,) Kunze, Bot. Zeit. 2:298 (1844). G. ex Maxon
et M.S. no 3281, n. $294, n. 3404 et n. 3422.

8. Hemitelia insignis (Fée) Christ. Index Fil. 349 (1905), G. Duss n. 4449,
ne parait pas distinct de H. grandifolia (Willd.)
Spreng. Cette espece a été basée sur 1'échantillon
de Willdenow n. 20167 récolté en Martinique.

3. ALSOPHILA R. Br.

9. Alsophila sp. M.S. n. 3316. Spécimen stérile mais se référant au genre

Aisophila str. sensu tel que le congoit Wm. R.
Maxon.

Famille 4. POLYPODIACHAE
1. BLAPHOGLOSSUM Schott.

1. Elaphoglossum piloselloides (Presi) Moore, Ind. Fil. 13 (1857). G. Duss n.
4137 et M. Duss n. 1615 et n. 4576.
2. Elaphoglossum apodum (Kaulf.) Schott, Gen. Fil. pl. 14 (1834). G. Duss n.
4321 et M. Duss n. 1621 et n. 4577.
3. Blaphoglossum undulatum (Willd.) Moore, Ind. Fil. 16. (1857), G. (ex Chris)
et M. Duss n. 1616 et n. 4122.
4. Elaphoglossum perelegans (Fée) Moore, Ind. Fil. 16. (1857), G (ex Fée et
_ Christ).
5. Blaphoglossum erinaceum (Fée} Moore, Ind. Fil. 9 (1857), G. L'Herminier
n. 7 (1862), S. n. 694 et M.S. no. 4160. Espece
affine de BE. scolopendrifolium (Raddi) J. Smith,
& laquelle divers auteurs l'’ont rapportée.
6. Elaphoglossum cuspidatum (Willd.) Moore, Ind. Fil. 16. (1857, G. Duss
ro 4589.
7. Elaphoglossum tectum (H. et B.) Moore, Ind. Fil. 15 (1857). G. (ex Christ).
8. Elaphoglossum decoratum (Kunze) Moore, Ind. Fil. 8 (i857). G. Duss n. 4152.
9, Elaphoglossum Herminieri (Bory et Fée) Moore, Ind. Fil. 16 (1857), G. Duss
n. 4136, n. 4137 et M. Duss n. 1626.
10. Elaphoglossum pteropus C. Christ. Ind. 5 et 314 (1905), G. (ex Urban) et
M. (ex Domin).
ll. Elaphoglossum flaccidum (Fée) Moore, Ind. Fil. 356 (1862), G. Duss n. 4143
et M. Duss no. 4149.
12. Elaphoglossum petiolatum (Sw.) Urban, Symb. Ant. Fl. Port. 4:61 (1903) et
93373 (1925). Ge Duss n. 4130 et M. Duss n. 1622.
13. Elaphoglossum glabellum J. Smith, Lond. Journ. Bot. 1:197. (1842), G. S.
n. 342 et M. (ex Maxon).
14, Elaphoglossum rigidum (Aubl.) Urban Symb. Ant. 9:374 (1925). G. et M. Les
variétés de EB. latifolium Sw. et cette esptce
citée par Duss (loc. cit. p. 27) ainsi que

Caribbean Forester = 40 = Vol. 4, No. lL

15.

16.

~

be

19.

20.

gic

22.

23.

24.

25.
26.

E. longifolium (Sw.) J. Smith pour les Antilles
frangaises doivent etre rapportées & cette
espéce.

Elaphoglossum Dussii Underw. in Maxon, Pterid. Porto Rico in Sc. Surv.
P. Rs 6:598. (1925)enG. S. n. 1061, n. 1448,
n. 1466, no. 1808, n. 1828, n. 2418 et M.S. n.
2384. Confondu souvent avec E. petiolatum
(Sw.) Urban.

Elaphoglossum simplex (Sw.) Schott, Gen. Fil. (1835) in Obs. sub Bolbiti,
t. 14. G (ex Fée et Hook.), Duss n. 44101, n.
4414 et M. Duss n. 1624, n. 1628 et n. 4687.
La var. martinicense (Desv.) Urban, Symb. Ant.
4:60 (1903), G. et M. (ex Urban) n'est autre
que E. glabelium J. Smith.-

- Elaphoglossum Underwoodianum Maxon, Pterid. Porto Rico 6:397. (1926), G.

et M. est trés voisin de E. simplex (Sw.) Schott.

Elaphoglossum Plumieri (Fée) Moore, Ind. Fil. 13 (1857), G.S. n. 2405 et
M. Duss n. 1614. Espece affine de E. vililosum
(Sw.) J. Smith et souvent confondue avec elle.

Elaphoglossum Feei (Bory) Moore, loc. cit. (1857), G.S. n. 652, n. 2406,
note lao ne vor Db, No L782, meet.
n. 2406 et M. Duss n. 4135.

Elaphoglossum Boryanum (Fée) Moore, Ind. Fil. 7. (1857) G. Duss n. 4151
et M. Duss n. 1613.

Elaphoglossum scolopendrifolium (Raddi) J. Smith. in Curt. Bot. Mag. 72:
17 (1846), G. (ex Urban). Il s‘agit peut-étre
d*une confusion avec E. erinaceum (Fée) Moore.

Elaphoglossum lingua (Raddi) Bracken. in Wilkes Explor. Exped. 16:74
(1654), G. Mazé n. 314, S. n. 1063 et n. 1199.
Syn-: E. brevipes (Kunze) Moore et E. scandens
(Bory et Fée) Moore.

z2- HYMENODIUM Fée

Hymenddium crinitum (L.) Fée, Mém. Foug. 2:90 (1845}. G. L'Herminier, 1862,
A. Fée n. 23, Duss n. 4148 et M.S. n. 2284,
Stl SIS) CIF daly Seite

3. RHIPIDOPTERIS Schott

Rhipidopteris peltata (Sw.) Schott, Gen. Fil. pl. 14. (1834), forme type:
M.S. no. 2098, n. $274, n. 33354, n. 3398 et
no 5441. Var. flabellata nov. comb. Syn.+s
R. flabellata Fée, Acrostichum peltatum Sw.
var. flabellatum (H.B.K.) Duss, G. Duss n. 4147.

4, ACROSTICHUM L.
Acrostichum aureum L. Sp. Pl. 1069. (1753), G. Duss n. 4146 et M.S. n. 3310.

Acrostichum danaefolium Langsd. et Fisch. Icon- Fil. 1:5, pl. 1. (1810) G.
L°Herminier n. 25 et M. (ex Maxon).

= 4] = October 1942

5. ANETIUM Splitg.

27. Anetium citrifolium (L.) Splitg. Tidjsch. Nat. Gesch. 7:395. (1840). G.
Duss n. 4227 et M. Duss 1505.

6. VITTARIA J. Smith

28. Vittarie lineata (L.) J.E. Smith, Mém. Acad. Turin 5:413, pl. 9. (1793),
Go.So. no 1194 et M. Duss n. 1553.
29. Vittaria filifolia Fée, Mém. Foug. 3:20, pl. 3. (1851-52). G. Duss n. 4266.

7. ANANTHACORUS Underw. et Maxon

30. Ananthacorus angustifolius (Sw.) Underw. et Maxon, Contr. U. S. Nat. Herb.
10:487. (1908). G. Duss n. 4225 et M.S. n. 2274.

8. PTEROPSIS Desv.

3l. Pteropsis martinicensis (Christ) Maxon, Stud. trop. amer. ferns, in Contr.
U.S. Nat. Herb. 3351 (1912), M. Duss n. 250 b.
Endémique rare décrite par Christ, Bot. Jahrb.
Engler, 24:137 (1897) sous le binédme Drymoglossum
martinicense Christ. Duss précise que, par un
malheureux accident, il a perdu tous les spécimens
(Foug. et Lycop. Ant. fr. 93, 1903). Nous avons
recherché véinement la plante jusqu’é& présent sur
les rochers et les arbres de la Montagne du
Vauelin, localité type.

9. HECISTOPTERIS J. Smith

32. Hecistopteris pumila (Spreng.) J. Smith, Lond. Journ. Bot. 1:193, (1842).
' G. (ex Urban).

LO. POLYTAENIUM Desv.

33. Polytaenium Feei (Schaffn.) Maxon, Pterid. Porto Rico, Se. Surv. P. R.
6, 405 (1926), G.S. n. 505 et M.S. n. 3448.
Antrosphytum lanceolatum (L.) Kaulf. Enum. 198.
(1824), non P. lanceolatum Desv., est synonyme
de cette espéce.

34. Polytaenium Dussianum Benedict, in Bull. Torrey Bot. Club, 169. (1911),
G.S. no. 1457 et M.S. n. 2275. C'est le corres=
pondant insulaire de P. brasilianum (Desv.)
Benedict.

ll. PALTONIUM Presi.

35. Paltonium lanceolatum (L.) Presl., Epim. Bot. 156. (1851). G. Husnot n.
293, S. n. 882 et M. Hahn n. 73, Duss no. 1547.

Caribbean Forester - 42 = Vol. 4, No. 1

36.

37.

38.

39.
40.

Aes

48.

49.

12. COCHLIDIUM Kaulf.

Cochlidium seminudum (Willd.) Maxon, Pterid. Porto Rico, Sc. Surv. P. R.,
6, 407 (1926), G.S. n. 357, N. 651, n- 1419 b,
n. 1886 et M.S. n. 3423.

Cochlidium linearifolium (Desv.}) Maxon, Contr. U. S. Nat. Herb., G. S.
n. $41, n. 1419 a et M.S. n. 4167.

13. POLYPODIUM L.

Polypodium duale Maxon, Contr. U. S. Nat. Herb. 16:61. (1912). G.S. n.
324 a, n. 341 an. 356, n. 542, n. 649, n. 650,
n. 657, n. 658, n. 842, n. 891, n. 1195 et M.S.
n. 3397, n. 3407 et n. 3424. Souvent désigné sous
le synonyme de P. serrulatum (Sw.) Mett, binéme
prétant a confusion.
Polypodium trifurcatum L. Sp. Pl. 1084 (1753). G.S. n. 2407 et M.S. n. 3442.
Polypodium taenifolium Jenmann, Bull. Bot. Dept. Jamaica II, 4:114 (1898).
GeS.on. 1222- n: 1207,.n. 1469 b,:m. 447000, n.
1784 a, n.- 2413, n. 2422 et M. Duss n. 1654 a.
Polypodium asplenifolium L. Sp. Pl. 1. ed. 2. 1084. (1753). G.S. n. 246,
n. 327, n. 524, n. 648, n- 1129, n. 1210, n. 1212,
n. 1412, n. 1413, n. 1414, n. 1433, n. 1434, et M.
S. m. 3313, n. 3521 a, n. 3336,-n.. $416, n.-5434,
n. 5445. Syn. Polypodium suspensum L. Spec. 1 ed.
2. 1084 (1753), ainsi que ses formes diverses.
Considéré comme espéce distincte par certains
auteurs.

- Polypodium jubaeforme Kaulf. Flora 6: 364. (1823). G.S. n. 538, n. 656 a,

Mosoes ne L202, n. 1209, n.l4b7- on. L770. -n-
1773. a, n- 1865 et M.S. n. 2100, n. 2108, n. 5291,
n. 3292, n. 3335, n. 3403, n. 3410, n. 3425, et
n. $430.

Polypodium mollissimum Fée, Mém. Foug. 11:47, pl. 12, f.2. (1866). G.
Duss no. 4106 et M. Duss n. 4571.

Polypodium taxifolium L. Sp. 51.1086. (1753). G.S. n. 539, n. 839 et M.S.
Wo Sa ie

Polypodium Plumula Humb. et Bonpl.; Willd. Sp. Pl. 5:178 (1810). G. S.
WM, 525, mu. 835 et M. (ex Fée).

Polypodium pectinatum L. Sp. Pl. 1085. (1753) G.S. n. 346, n. 350 et M.S.
ae oae lt

Polypodium dissimile L. Syst. Nat. ed. 10, 2:1325 (1759), G. Duss n. 4063,
n. 4065 et M. (ex Maxon). Syn. P. sororium (Humb.
et Bonpl.) Willd. (1810), bindéme sous lequel Urban
l'a désignés.

Polypodium piloselloides L. Sp. Pl. 1083. (1753). G.S. n. 348, n. 901,
n. 1430, n. 1917 et M.S. n. 2658, n, 3283, n. 3340
a. Souvent confondue avec P. ciliatum Willd.,
espece affine.

Polypodium polypodioides (L.) Watt, Canad. Nat. I1.13:158. (1867). G.S.
Re 265, 507, no 5OS, mo S29. .n- -1¢88 et M.S.
n. 4168. Syn.: P. incanum Sw. (1788).

= 43 = October 1942

50. Polypodium loriceum L. Sp. Pl. 1086. (1753). G.S. no. 103, n. 1229, n.
1450, no. 2104 et M.S. n. 3327.

51. Polypodium suruchense Hook. Icon. Plant. 1:69. (1837). G. (ex Urban), M.
Duss n. 1652.

52. Polypodium chnoodes Spreng. Neu. Entd. 3:6. (1822). G. Duss n. 4120 et M.
Duss n. 1661.

535. Polypodium angustifolium Sw. Prodr. 130. (1788). G Duss n. 4103.

54. Polypodium astrolepis Liebm. Dansk. Vid. Selsk. Skrift. V. I:185. (1849).
G. S. n. 506, n. L191, n. 1431 et M.S. n. 2252, n. 3514,
n. $340 b. Es péce reppontce par divers paboune Bo Be
lanceolatum et & P. elongatum (Pw). Mett, non Ait. (1789).

55. Polypodium exiguum Hedw. in Mag. Nat. Hist. n. ser. 22458. (1838), G. Duss
n. 4087 et M. Duss n. 1655. Espece affine de P. hetero-
phyllum L.)

56. Polypodium lycopodioides L. Sp. Pl. 1082. (1753). G.S. n. 507, no 1189 et
M. (ex Urban).

57. Polypodium phyllitidis L. Sp. Pl. 1083. (1753). G.S. n. 642, n. 703, no
1449 et M.S. n. 3317.

58. Polypodium latum (Moore) Sodiro, Crypt. Vase. Quit. 371. (1893). G. et M.
(ex Maxon).

59. Polypodium crassifolium L. Sp. Pl. 1083. (1753). G.S. n.. 1794 et M. Duss

no 1671.

60. Polypodium sectifrons Kunze; Mett. Abh. Senckenb. Ges. Frankfurt, 2:99.
Pl. 2, f. 3,4. (1857). G. (ex’Maxon).

61. Polypodium aureum L. Sp. Pl. 1087. (1753). G.S. n. 347, n. 2439 et M.S. n.
4169. Espéce polymorphe et variable dont Duss, Foug.
Ant. fr. 51 (1903) retient var. areolatum (H.B.K.) Hook
et Bak. et var. Mazei (Fourn.) Duss.

62. Polypodium triseriale Sw. loc. cit. G.I. n. 208, no. 210, n. 257, n. 1187,
n. 1228, n. 1437, n. 1877, n. 2420.

63. Polypodium serricula Fée Gen. Fil. 238. (1852) G.S. n. 656, n. 1208, n.
1467 a et b, n. 1468 b, no. 1470 c, no. 1471, n. 1474,

n: 1763, n. 1767 a; n. 1768; nm. 17695 no 177305 no eliie.
n. 1773 b, n. 1781, no. 1885, ns 2413 a et M. Duss n-
1654 et n. 1654 b.

64. Polypodium tenuiculum Fée Gen. Pil. 239 (1852), G.S. n. 655, n. 1204,
no. 1206, n. 1415.

65. Polypodium glaucophyllum Kunze Dic Farnk. 227, 93. (1846). G.S. n. 97, no
592, no 1432, n. 2411.

66. Polypodium brasiliense Poir. in Lamk. Encycl. meth. bot. 5:525 (1808). G.
S. no 95 et M.S. n. 4170. Espéce trés polymorphe avec
de nombreuses formes écologignes intermédiaires.

67. Polypodium Hartii Jenman, Journ. Bot. Brit. and For., 24:272 (1886). G.S.
n. 1125, n. 1469 a, no. 1470 a, n. 1472, n.'1473, n.
1484 b, n. 1793.

68. Polypodium Grisebachii Underw. in C. Christ. Ind. Fil. 53, I, (1906). G.
(ex Grisebach) et M. Duss n. 1655.

69. Polypodium flabelliforme Poir. in Lamk. Encycl. 5:519 (1804). G.S. n. 1203,
n. 2416.

70. Polypodium limbatum (Fée) Maxon loc: cit. G. (ex Maxon).

71. Polypodium pendulum Sw. Syn. Fil. 33 (1806). G.S. n. 1205, n. 1418.

72. Polypodium cultratum Wild. Spec. Pl. 5, 187 (1810). G.S. n. 1416 et M. (ex
Plumier).

Caribbean Forester = 44 = Vol. 4, Noo 1L

73. Polypodium induens Maxon, Bull. Torrey Bot. Club 32, 75 (1905). G.S. no
1468 a. Espéce d'altitude des Grandes Antilles et
d‘'Amérique Centrale, nouvelle pour les Petites
Antilles.

74. Polypodium Knowltoniorum W. H. Hodge, Notes on Dominica Ferns, in Amer. Fern.
Journal, 31, 105, (1941). G. Questel n. 1033.

Polypodium trichomanoides Sw. Prodr. 131 (1788).

Espéce indiquée par Duss Hieronymus et par Urban, Symb. Ant. Fl. Port.
4:53 (1903), pour la Guadeloupe et la Martinique, est une fougtre de répar-
tition géographique apparemment limitée & la Jameiique et au Guatemala.
Pour les Antilles frangaises, il s‘agit du P. taenifolium Jenman et de P.
serricula Fée.

Polypodium laxifrons Liebm. Mex. Bregn. 52 (1849).

Espéce citée par Urban, Symb. Ant. Fl. Port. 4:54 (1903) pour la
Guadeloupe et la Martinique, n’est autre que le P. asplenifolium L. Le
type de P. laxifrons, décrit pour le Mexique et l‘'Amérique Centrale,
correspond exactement avec les échantillons du P. asplenifolium des Peti-
tes Antilles. C'est aussi l’opinion de Wm. R. Maxon, Pterid. Puerto Rico
6:41] (1926).

Polypodium suspensum L. Spec. I, ed. 2:1804 (1753).

Espéces citée par Duss et par Urban, Symb. Ant. Fl. Port. 4:54 (1903)
pour la Guadeloupe et la Martinique, est basée sur la planche 87 de Plumier
représentant une plante martiniquaise de détermination critique, paraissant
Stre une forme de P. asplenifolium L.

Polypodium trichomanes A. St. Hil. Voy. Distr. Diam. I. 378 (1833).

Espece décrite par Christ, in Engl. Bot. Jahrb. 24:127 (1897), comme
P. taxifolium L. var. trichomanes Christ, citée par Urban, Symb. Ant. Fl.
Port. 4:54 (1903) pour la Guadeloupe et la Martinique, ne parait pas
spécifiquement distincte de P. taxifolium.

Polypodium areolatum H. et B. in Willd. Spec. Plant. 5:1, 172 (1810), cité
par Urban, loc. cit. 57, pour le Guadeloupe et la Martinique, considéré par
Hooker et Baker, Syn. Fil. 2,347 (1847) comme var. areolatum de P. aureum L.,
n'est pas spécifiquement distinct de P. aureum L.

(To be concluded in the January 1943 issue.)

Summary

This catalogue contains the systematic enumeration of the pteridophytes
and vascular cryptogams of the French Antilles collected chiefly by L‘Herminier,
Mazé and Duss in Guadeloupe, by Duss and Hahn in Martinique and also by the
author, between 1934-1942, with the assistance of Mrs. Stehlé and R. P. L.

- 45 - October 1942

Quentin, indefatigable collector and well-informed botanist who collected them
in one or the other of these islands. Except for a few species whose complete
bibliography is difficult to obtain in an isolated tropical region, the major=
ity of the plants belonging to the ferns, club mosses, and Selaginella of the
French Antilles appear in this work.

The order and nomenclature adopted here is that used by William R.
Maxon, Curator of Plants, Smithsonian Institution, Washington. (Studies of
tropical American ferns and Pteridophyta of Puerto Rico and the Virgin Islands,
1926), the most noted specialist in the Pteridophytes of America and the
Antilles. He has willingly studied our collections and revised our determi-
nations during the last eight years for which here is a sae my most
sincere acknowledgement.

Reference is made to A. L. A. Fée'’s "History of Ferns and Lycopodia of
the Antilles” (Paris, 164 p., 34 tab., 1866), a superb publication, and also
to the most recent monographs of adjacent islands specially those of Urban:
Symbolae Antillanae, Flora Portoricensis, Vol. IV, 1903-1911; Pteridophytae
p. 5-70 (1903) and p. 648-651 (1911), Pteridophyte dominguensia, Vol. IX,
1925, p. 273-397; and those of Domin: Pteridophyta of the Island of Dominica
(Prague, 1929).

In an essay on the Ecology and Botanic Geography of Guadeloupe, Vol. 1
Basse Terre, 1935 as well as in the Outline of Plant Associations of Martinique
(in Bull. Agr. Mart. 6, 193-264, 1937) the fern and allies collected in the
different sections, according to their distribution and their edaphic and
climatic requirements were mentioned. The object of the present paper is to
present information concerning the abundance of pteridophytes in the French
Antilles; to make known, under their most recent and exact denominations, the
majority of the species of our flora and the rarities which have been collected
during recent years, especially in those sections of the central range of both
islands which have been least explored. The numbers here mentioned are those
of our collections and they are preceded by the name of their respective col-
lector. Last of all, the letters "G" and "M" stand for location at Guadeloupe
and Martinique respectively.

Our collections previous to the year 1938 are found at the Museum of
Natural History in Paris while those collected after 1938 appear in H. and M.
Stehlé's Herbarium in Martinique. Samples of all species have been sent to
the Herbarium of the Smithsonian Institute, Washington.

Resumen

Este cat&logo contiene una enumeracién sistemética de las Pteridofitas
o Criptégamas Vasculares de las Antillas francesas coleccionadas principalmente
por L'Herminier, Mazé y Duss en Guadalupe, por Duss y Hahn en Martinica, asi
como por el autor entre 1934 y 1942 con la ayuda de la Sra. Stehlé y del R. P.

Caribbean Forester = 46 = Vol. 4, No. 1

L. Quentin, botaénico y coleccionista incansable, en una u otra de éstas islas.
Con excepcién de un pequefio numero de especies cuya bibliografia completa es
dificil de conseguir en un medio ambiente tropical aislado, la totalidad de
las plantas que pertenecen a los helechos, Licopodias y Selaginelas de las
Antillas francesas figura en este trabajo.

El orden seguido y la nomenclatura adoptada es aquella de Wm. R. Maxon,
Curador de Plantas del Smithsonian Institution de Washington (Studies of
tropical American ferns and Pteridophyta of Puerto Rico and the Virgin Islands,
1926), especialista de renombre de las Pteriofitas de América y las Antillas.
El ha estudiado con gusto nuestras colecciones en estos ultimos ocho afios y
ha revisado nuestras determinaciones, por lo cual expresamos aqui nuestro mas
vivo agradecimiento.

Hacemos referencia eneste estudio a la Historia de los Helechos y
Licopodidceas de les Antillas (Paris, 164 p., 34 tab., 1868), de A. L. A. Fée,
publicacién mgistral y también a las monografias més recientes sobre las
pteridofitas de las Antillas vecinas, en particular a Urban: Symbolae
Antillanae, Flora Portoricensis, Vol. IV, 1903-1911, Pteridophytae p. 5-70,

(1903) y p. 648-651 (1911), Pteridophyta domingensia, Vol. IX, 1925, p. 273-
397 y a Domin: "Pteridophyta of the Island of Dominica" (Prague, 1929).

En nuestro ensayo sobre ecologia y geografia botdnica de la Guadeloupe,
tomo 1, Basse Terre, 1935, asi como en el Extracto de las asociaciones vege-
tales de la Martinica (en Bol. Agr. Mart. 6, 193-264, 1937) hemos citado los
helechos y plantas afines coleccionadas en los diferentes sectores, pero segun
su distribucién y exigencias edafo-climaticas. El presente articulo tiene por
objeto sobre todo dar una idea de la riqueza en pteridofitas de las Antillas
francesas, de dar a conocer con sus denominaciones mas recientes y més exactas,
la mayoria de las especies de nuestra flora y las rarezas vegetales que hemos
colectado en el curso de estos Ultimos afios, sobre todo en los sectores de la
cadena central de las dos islas que habian sido menos exploradas. Los numeros
que aparecen en este trabajo son los de las especies en nuestras colecciones
y que preceden al nombre de su colector. En fin las letras "G" y "M" corres-
ponden a la localizacién en Guadeloupe y Martinica respectivamente.

Nuestras colecciones anteriores al 1938 estan depositadas en el Museo
de Historia Natural de Paris y aquéilas recolectadas despues figuran en el
Herbario de H. y M. Stehlé, actualmente en la Martinica. Hemos enviado dupli-
cados de las pteridofitas al Herbario del Smithsonian Institution en Washington.

- 47 = October 1942

PROBLEMAS DE SUELO EN LA PRODUCCION DE

COSECHAS TROPICALES:/

Es una creencia popular que todas las tierras donde crecen bosques
virgenes son tremendamente fértiles. Esto es una falacia total. A los ojos
del lego todos los bosques parecen mas o menos iguales—densos y exuberantes
—y se ha creido que ésto prueba que el suelo donde crecen debe ser muy rico.»

Pero aquéllos que conocen el bosque saben que hay diferencias enormes en su
densidad y exuberancia—muchos arboles gigantes, bien desarrollados en algu-

nas areas y sélo aérboles relativamente pequefios en otras. Sin embargo, adn
en areas de bosque mis pobre hay una densidad de vegetacién mucho mayor que
aquella que habita en la mayoria de los suelos de las regiones templadas.

El hecho es que el suelo puede ser bastante fértil bajo condiciones forestales;
es decir, mientras esté cubierto de bosque, pero de poca fertilidad si se le
desmonta y planta con otro cultivo. Bajo condiciones forestales, con la llu-
via abundante y el calor de los himedos trépicos hay una rdpida circulacién
de material nutritivo. Continuamente se estén desprendiendo hojas y ramas;:
caen al suelo, se pudren pronto y dejan libres los materiales nutritivos que
contienen—nitrdégeno, fosfato, potasa, etc.—que de inmediato estén listos
para ser asimilados otra vez por absorcién en las raices de las plantas fores-
tales... Este proceso es muchas veces més répido que en los bosques de la
zona templada. En esta forma una cantidad dada de material nutritivo puede
rendir servicio mucho més a menudo que en la zona templada. La caida copiosa
de la hoja da como resultado una capa permanente de hojarasca putrefacta y de
humus en la superficie del suelo y también lea presencia de una rica zona
hameda en las primeras pulgadas del suelo en si. La circulaciém més rdpida
de alimento del suelo—absorcién por la raiz, traslacién a la hoja, caida y
pudricién de ésta—todo se origina por medio de los procesos quimicos y bio-
légicos que se efectuan en esta capa humifera, y mucha de la aparente ferti-
lidad del suelo forestal depende de estos procesos. Por lo tanto, bajo con-
diciones forestales esta capa humifera es permanente; se provee de material
nutritivo con la caida de la hoja a la par que se va pudriendo o se la van
llevando los micro-organismos e insectos.

Pero cuando se elimina el bosque de este suelo el ciclo forestal cerra=
do se rompe. La parte de desgaste del ciclo o sea la pudricién y oxidacién
del humus continta y mis atin se acelera debido a la exposicion al sol tropical
—pero la parte de provisién del ciclo cesa - no hay caida de hoja que equipa-
re las pérdidas, desaparece el humus = el suelo queda desnudo y sélo resta a
la disposicién del agricultor la fertilidad innata del suelo. Las grandes
diferencias que puedan existir en cuanto a la fertilidad innata se revelan
entonces. Asi se conoce como muchos suelos se ven desprovistos de ciertas
materias nutritivas desde el mismo instante en que se elimind el bosque; es
decir, que la cantidad de material nutritivo suministrado por el ciclo cerrado
no es suficiente para permitir el crecimiento vigoroso de un cultivo a menos
gue se le aflada abono.

V/ Tomado de un articulo publicado en el nimero de "Tropical Agriculture"
de febrero 1942. Autor H. J. Page.

Caribbean Forester = 48 = Vol. 4, No. Ll

CONTENTS

A forest policy for the American tropics .... .
Arthur Bevan, Puerto Rico

The evaluation of forest tree species in Puerto
Rico, as affected by the local forest problem .
Frank H. Wadsworth, Puerto Rico

Roble, a valuable forest tree in Puerto Rico...
Frank H. Wadsworth, Puerto Rico

Comments on the silviculture of Cedrela .....
Le. R. Holdridge, Haiti

Lady-beetles don't behave ..- © « 0 « «© « © « « «
George N. Wolcott, Puerto Rico

Catalogue des cryptogames vasculaires des Antilles
frangaises. (Continuation) ....-.-+-+-«e.
H. Stehlé, Martinique

49

54

59

ali

81

83

A FOREST POLICY FOR THE AMERICAN TROPICS

Arthur Bevan, Director
Tropical Forest Experiment Station

The history of forest exploitation in temperate North America from
the days of "inexhaustibility" to the formation of organized conservation
signifies the natural relationship between man and @ valuable rescurce. The
virgin forests, conteining many valuable species, were exploited in such a
manner that the margin between cost and sales price was as wide as possible.
There was nothing malicious in this manner cf expiciting forests, but as it
happened, cutting methods which brought the greatest immediate profits were
not compatible with what are considerec good siivicultural practices. With
iess than 20 per cent of the original commercial forest area of the United
States still uncut, conservationists, even with the support of Federal and
State Governments, have been unable to balance the growth-drain budget. Only
slightly more than one half of the commercial forest area which has been cut-
over is classified as "fair to satisfactorily restocking".

Why has this happened? Was it the fault cf the lumbermen, the econom-
ic system, or the democratic form of govermment? Each probably contributed
to some extent, but the real blame must be placed upon the shoulders of the
people who throughout the history of the country have, to a large measure,
condoned or permitted the prevaiiing methods of forest exploitation. Only.
when wood shortages become acute will the public demand better silvicultural
practices, thereby stopping destructive cutting of remaining stands, and
bringing devastated lands back into production so that ultimetely 4 sustained
annu&l yield will be reached which will meet the requirements cf the country
for forest products.

For the present, however, the rate cf depletion is being greatly ac-
celerated by the needs of the war. Shortages of strategic materials are being
alleviated by substitutes produced from wood, for when other raw materials are
no longer avaiiable, there is an almost universal tendency to turn to wood.
For this reason, the demand for wood end more wood is growing with each month
of the war. How then, in the face of these growing demands, can our depleted
forests continue to produce our needs in the future during the long period
necessary to bring our cut-over lands into production? Will pressure to
relieve this approaching pinch necessitate increased imports of forest pro-
ducts from the Americar tropics? Such & move is already under way, and it is
expected that Inter-American lumber trade will grow rapidly.

Tropical America is 4 region with a very heterogeneous distribution of
population. Areas of dense population, such as most of the West Indies and
in the vicinity of large cities in Central America are already denuded, but
@ large area of forest remains at least in the countries of the mainiand.
Much of it is remcte from civilization, some is virtually inéccessible. A
certain analogy exists between the present stage of development here and that
in the United States in 1800, as far as the forests are concerned. At that
time a limited amount of forest exploitation in the United States had already

a5. January 1943

taken place and the species and the values of some of the more important
forest products were known. It is possible that we are on the threshold of
another era of exploitation of virgin forests. Are the forests of Central
America now to be slaughtered, or will it be possible to stem the tide of
depletion?

What is the present general concept of forests and their management in
the forested countries cf the Americen tropics? That concept is largely con-
cerned with the exploitation of valuable fancy woods for export and the dis-=
covery of other species which might be suitable for these same markets. A
study of literature, descriptions of forest stands, reports of forest officers,
planting programs in the tropics, clearly indicates that this policy has been
the incentive for nearly all forestry programs. Forest laws in tropical
American countries, where they exist, fortunately are more advanced than were
those of the United States at a corresponding period of development, but pres-
sure of circumstances still largely dictates actual practice.

In the main the tropical hardwood forests, which make up the bulk of
the forest resources of these countries, are highly complex stands of many
species. Species yielding valuable export woods make up only a small part
of the forests, averaging perhaps one tree to the acre although sometimes
occurring as occasionally as one tree to five acres. Cuttings made primarily
for these species result in high-grading of the forests and in many countries
where forest exploitation is "organized" such species are being cut at a
depletion rate; certainly this is true where the timber is readily accessible.
Under such a system logging costs are very high, damage to the remaining stand
is excessive, and regeneration of the valuable species is not planned for.

We do not too often find consideretion of the forest problem of the
tropics of the Western Hemisphere based on the needs of the local inhabitants
and their future welfare. Denuded forest lands are much more common than
most people imagine. Near all heavily populated sections in this region,
this type of land abuse is to be found. Despite this situation, no serious
consideration has yet been given to short rotation quick-growing forest crops
suitable for fuel, stakes, posts, and other materials needed for local cone-
sumption and for the improvement of the standard of living, of the peasant or
peon class which makes up by far the greater portion of the population of
these countries. Import statistics of lumber and wood products even in
forested countries are indicative of the almost total neglect of local require-
ments and of attempts to solve this problem. In most of the regions lumber
imports could be substantially reduced by better utilization of forest products
and proper management of local cut-over forest lands, many of which are idle.

In some countries efforts are being made to maintain the forests through
the planting of deforested lands with the more "valuable" species. Where suc-
cessful these may eventually help to supply the foreign market, but, with the
exception of teak plantations in Trinidad, very little success has been
attained to date. Past plantings in Puerto Rico, may be cited as an example.
Large expenditures have been made on a forest program which has centered
around the planting of deforested lands. The first species tried was Spanish
cedar, Cedrela sp., as this proved easy to raise in the nursery and early
survival was high. However, later these plantations, largely pure in

Caribbean Forester = 50 = Vol. 4, No. 2

composition, wilted and died, and today as far as our knowledge goes not a
single plantation of cedar exists which can really be called successful.
Later, plantings were made cf other fancy-wocd species, such as mahogany,
Swietenia sp.; teak, Tectona crandis: and maga, Montezuma specicsissima.

On good sites some of these plantations are growing well. Mahogany, where
planted under a light overstory, looks good, but on poor sites or on denuded
abandoned and eroded lands, the areas which should first be reforested, plan-
tations fail or are very expensive to establish.

More recently, following such failures, local species have been planted,
but because previous interest was restricted to species producing high value
woods, little is known of the characteristics of these trees. Studies of the
normal sucsessions on such sites and value of tree species showing promise
must be determined before reforestation at reasonable cost is possible. Re-
forestation programs are almost certain to start from and around centers of
greatest need, that is to say, centers of high population density where exten-
Sive areas are denuded. This is the pian being carried out by Mexico, and it
has much of merit in it, not only from the viewpoint of supplying the greatest
need first, but alse for the psychological and educational effect on the
people. However, to secure continued and increased public support the plan-
tations must be successful.

If the vision of high prefits in the distant future from the production
of exportable fancy woods is not to "rule and ruin” the domestic forest economy
of trepical America, due consideration must be given to the development of a
policy which recognizes that domestic supplies ere of first importance. Pra=
visicn for the producticn of “luxury" woods should be secondary to the main-
tenance and management of sufficient areas of forest land under sustained yield
management to supply present and prospective domestic needs. This does not
necesserily mean that the production of forest products for export need be
restricted to areas not producing for domestic needs. On the contrary the
proper management of these lands will restore them ultimately to a more normal
condition where such species can be grown one or two trees to the acre as in
the virgin forest. There is evidence to indicate that the reestablishment of
@ forest cover will permit successful introduction of “luxury” species on sites
where pure plantaticns have proved a failure. Such introduced dominants or
emergents could be carried over as growing stock through several rotations of
the understory or true canopy. The cutting of the latter will supply the
local market and pay the costs of carrying the slower growing fancy woods.

This would develop a system something of the order of “coppice with standards"
practised in Europe with hardwood forests.

A sound forest policy should be based, in the order named, on:

lL. The needs of the locai population.

2. Determination of the properties and uses of the more commen
species for lumber and other wood products for home consumption.

3. The production of "luxury" fancy wood for export.

Such an approach should solve the problem of proper utilization and man-

agement and lead to perpetuation of the high value species rather than their
elimination under present destructive high-grading methods. On permanent

- 51 = January 1943

forest lands, a method of management such as has been suggested is more
compatible with natural processes and the economy of such extended use is
obvious»

Consideration should be given to the fact that areas of what most
foresters would consider brush, composed of so-called inferior species,
close to centers of dense population, have much greater values than forest
containing high value fancy woods which is generally costly to log because
the commoner species are economically inaccessible at present.

The recent cutting of a tract of pomarrosa, Jambosa jambos, in Puerto
Rico close to transportation facilities illustrates the extraordinary values
occasioned by proximity to market and a dense population. This stand was cop-
pice, and about 12 years old. Largest stems were 7 inches d.b-h. This pure
stand, clear cut, yielded posts, poles, stakes, charcoal and fuelwood which
gave @ net return of over $40 per acre per annum. This may be an extreme case
but it shows definitely that where a domestic market exists it is just as im-
portant as exports. It is doubtful that much of our high forest with fancy
woods will show any such return as this coppice.

Undoubtedly there will be a large area of abandoned land in tropical
America following the war as certain large agricultural production programs
now very necessary become less important. Many of these programs, such as
the following one, result in the destruction of forest. "Several hundred
acres a week in Panama and Costa Rica are being cut from the jungle and seeded
to Manila fiber plants as one phase of the world-wide program recommended by
the Combined Raw Materials Board to fill the United Nation's war need for
rope."4/ Thus we place more land under the plantation system so familiar in
the tropics with sugar, cocoa, and bananas, In times of high prices in the
past large areas of forest land have been cleared for these crops which, when
disease or low prices came, were abandoned. Naturally the poorer and most
eroded areas were abandoned first. The sugar boom in Cuba in 1920 is an ex-
ample where large areas of sugar cane were abandoned when the slump came.
Imagine the forest that would exist today on these areas had a policy of re-
forestation been followed. With the coming of peace similar abandonments will
occur when products are again procurable from lands now cut off from the world
markets. Are these areas to become an impenetrable jungle of weeds, vines, and
brush or will provision be made to see that a forest is restored, not only to
ensure the rebuilding and protection of the soil for the future should a need
for these lands arise, but to provide a supply of wood products for local use
and for export? When abandoned, such areas have usually been provided with a
ready means of access and can help to support the economy of the country.

The war has already caused a change in the point of view and is encour-
aging consideration of the home requirements of a country and its citizens.
However, much more should be done and research in particular should be directed
to a solution of these problems which will receive added consideration with
the coming of peace. New territories will be opened up for settlement and the
tropics of the Western Hemisphere are a fertile field for such purposes. We
must be prepared with a new and up-to-date forest policy which will not

if Victory. (Washington, D.C.) 3(38):13. September 22, 1942.

Caribbean Forester - 52 = Vol. 4, No. 2

perpetuate the mistakes of the past. If provision for domestic needs is made
now, first by keeping all true forest land in forest, and then by keeping
them productive to the extent that at least domestic needs are met, production
for export will then assume its proper role in the forest economy of tropical
American countries, that of a surplus crop over and above all local needs.

Resumen

La rapida explotaciGn de los bosques en la zona templada de Norte
América a través del siglo pasado ha reducido grandemente la fuente dispo-
nible de productos forestales. La gran demanda de madera originada por la
presente guerra esta acelerando més aun el compas de esta explotacién.
Todos los indicios son de que se acerca una era de destrucci6én similar en
los bosques tropicales del hemisferio occidental para ayudar a suplir las
naciones de la zona templada donde los bosques estan casi exhaustos.

El concepto general del bosque tropical en el pasado se limitaba a
tomar en cuenta principalmente las maderas preciosas de exportacién que se
consiguen sole aqui y allé en el bosque. De ahi que el corte continuo de
los bosques con el objeto de conseguir estos pocos arbcles dispersos lleva
consigo la deterioracién costosa y desperdiciadora del bosque.

Si el ritmo de esta explotacién ha de acelerarse, el corte debe efec-
tuarse de modo que la produccion se conserve como una fuente de ingresos
valiosa para el puebio de todas las pequefias naciones de la América latina.
Estos cortes deben dar lugar también a la produccién continua de productos
forestales de utilidad doméstica tales como combustible, estacas y postes.
En las tierras despobladas que no son utiles para la agricultura deben sem-
brarse arboles de crecimiento répido especialmente cuando estas tierras se
hallen cerca de los centros de poblacién. Los arboles que producen maderas
preciosas son generalmente de crecimiento lentc y deben sembrarse solo en
aquellos sitios donde la demanda de combustible no es grande ya que a fin de
cuentas el ingreso que producen las especies forestales de crecimiento répido
se asemeja mucho ai que proviene de las maderas preciosas debido al largo
periodo de crecimiento de éstas.

La politica forestal sana, que sirva para asegurar los intereses del
pueblo de las pequefias naciones de la América tropical debe basarse, segun el
orden mencionado, en:

1. Las necesidades de la poblacidon local.

2. La determinacién de las propiedades y usos adecuados de las
especies forestales més corrientes como fuente de madera y
otros productos para uso local.

3. La exportacion de maderas preciosas.

- 53 - January 1943

THE EVALUATION OF FOREST TREE SPECIES IN PUERTO RICO,
AS AFFECTED BY THE LOCAL FOREST PROBLEM

Frank H. Wadsworth
Associate Forester
Tropical Forest Experiment Station

Tropical forests, most of which are composed of a complex mixture of
many species of trees, present many and varied problems to the forester at-
tempting to convert them into highiy productive managed stands. One of the
most important decisions which must be made is in the selection of the tree
species of greatest value: those which should be favored, or if necessary
introduced, in order that they may make up the managed forest. This decision
should be influenced by consideration of not only the properties of the wood,
rate of growth, and the complex relationship between any species and the
physical and biological factors of its environment, but also of the forest
problem of the region which is directly dependent upon past forest land use,
and present and prospective supply and demand for forest products. In Puerto
Rico the present forest problem is almost entirely a result of the manner in
which forests have been exploited in the past.

Past Forest Use

As the population of Puerto Rico has steadily grown, so has the demand
for forest products. The sawtimber requirements for construction lumber and
for attractive furniture woods have long since exceeded the local supply.
Most of the lumber now imported comes from the southern United States, and
for furniture, mahogany and Spanish cedar are imported from the Dominican
Republic and Central America.

With parts of the Sierra palm type as a possible exception, all of the
forests of the island have been cut or culled at least once, most of them
many times. In the past most of the cut-over lands suitable for agriculture
were immediately put under cultivation, having been cleared primarily for
that purpose. The poorer lands were either permitted to grow up immediately
to second growth forest or were farmed until the soil was gone or worn-out,
and then abandoned to return slowly to forest.

Until recently these second-growth forests, together with the remaining
old-growth, have been sufficiently extensive to supply local needs for many
small forest products, including fuel wood, fence posts, poles for house
construction, stakes for vegetabie crops and tobacco, small poles, hoe handles,
wagon parts, and ox-yokes. These are referred to as "secondary" forest pro-
ducts in the United States where they are by-products of the main harvest or
are generally harvested at the time of forest improvement cuttings rather than
at the end of the rotation. However, on an island the sizé of Puerto Rico,
with 1,300,000 rural inhabitants, these products are of primary importance.
They have a low value per unit of volume compared with other forest products,
and therefore it is imperative that Puerto Kico supply her own requirements

Caribbean Forester - 54 = Vol. 4, No. 2

to and obviate importation, the cost of which will approach that of the
products themselves.

The great demand for these products, accompanied by the fact that they
may be taken from small trees, has led to severe cutting in young stands, and
naturally the more suitable tree species have suffered most. Continued forest
deterioration as a result of this practice, and the clearing of nearly every
acre of land that could be farmed even if for only a few years, have steadily
lowered production until at present the market price of these smaller forest
products is rising rapidly.

Thus, Puerto Rico‘s forest problem is not merely that of increasing
production but, fuliy as important, the establishment of a forest cover on
some of the worst lands of the island for their protection.

Silvicultural Measures Required

Considering the circumstances a logical objective of a forest land
management policy is the improvement of existing stands and the reforestation
of all cleared lands unsuited to agriculture.

The available area is sufficient to produce all of the smaller forest
products such as posts and fuel, that are needed. The satisfaction of this
demand is of first importance. Without a great reduction in population,
attended by less demand and reduced intensity of cultivation of marginal
lands, the local forests cannot produce an appreciable portion of the con-
struction lumber, railroad ties and timbers consumed here. However, this is
not intended to convey the impression that lumber and valuable furniture woods
cannot or should not be produced. Our tropical climate gives us an opportu-
nity to produce very attractive furniture woods which are in great demand on
both the domestic and foreign markets. When our forests become fully pro-
ductive the yield of some of them can include these high-priced products, and
part of the least accessible area can be managed primarily for this purpose.

The greater part of the improvement of existing forests consists of
increasing the number of trees per acre of species which will rapidly produce
the desired products. In some of the better forests, trees of suitable
species are already growing and merely need encouragement by protection from
frequent cutting and by liberation from the domination of other less valuable
individuals. However, more frequently, years of indiscriminate cutting have
eliminated all of the better species except a few persistent sprouters. In
such forest underplanting is often needed. A moderately dense tree canopy
should be left until the young planted trees are several feet tail, the re-
moval of the older trees being made in several light cuts similar in effect
to shelterwood removal cuttings which leave sufficient shade to control weeds
and vines.

Many areas of worn-out farm land throughout the island are steep and
rocky and should never have been deforested. The great pressure of the large
agricultural population has resulted in the cultivation of lands which else-
where, because of their slope or low productivity, would have remained perma-
nently in forest. When these lands are abandoned reforestation is very

- 55 = January 1943

necessary to adequately supply the local farmers with forest products and
also to protect the soil.

Regardless of the tree species or the method of reforestation practised,
young trees on deforested lands require protection from the competition offered
by weeds and vines. On the basis of site there are two recommended methods of
weed control. On the best of the lands available it is possible to interplant
the area with food crops for a few years until a canopy sufficiently dense to
dominate other vegetation is formed. This method has the advantage of pro-,
viding payment for the weeding with the value of the crop harvested. However,
most of the lands now available for reforestation are idle because of steepness
of slope or impoverished soil. On these lands intercultivation is inadvisable
and therefore weeding is necessary, the cost of which must be considered an
investment until the trees are harvested or at least until full site protection
is provided.

Silviculturally Desirable Tree Characteristics

To be of value in the solution of Puerto Rico's forest problem, a forest
tree must have some of the following characteristics. It musts (1) produce a
wood satisfactory for the products needed, (2) grow rapidly, (3) be relatively
insusceptible to damage or mortality as a result of epidemic attacks by pests
or diseases, (4) be easily established and able to grow on poor soils, (5) be
able to grow in competition with existing herbaceous vegetation, (6) if under-
planted, be sufficiently tolerant of shade to grow under existing forests
until dominance is attained. Considering the 500+ native and the many intro-
duced species of the island it appears that there should be a wide choice.
Several species have 1 or 2 of these characteristics, but no tree so far used,
either native or exotic, has all of them to the desired degree, and it is
extremely doubtful that such a species will be found. As a result any species
chosen, though it may be very suitable in some respects, will also have
certain disadvantages.

Suitability of the Species Now in Use

For underplanting a tree tolerant of shade must be chosen, for other-
wise the survival and continued growth of the young trees would require that
the canopy be kept sufficiently open to also admit dense herbaceous vegetation.
When in the open, tolerant trees generally grow more slowly than those re-
quiring full light, but in the shade they will grow with virtually no care
except the gradual removal of trees dominating them. One species which has
been used in underplanting is broadleaf mahogany, Swietenia macrophylla, a
tree whose chief advantage is its valuable wood and chief disadvantage is the
long wait before harvesting. Among the relatively shade-tolerant species are
many natives producing useful woods, such as the laurels, (Lauraceae, parti-
cularly Ocotea spp.); guaraguao, Guarea guara Jacq.; algarrobo, Hymenaea
courbaril Lo; 1 Le; capé prieto, Cerdana alliodora R. & P.; maga, Montezuma
speciosissima Sessé & Moc.; maricao, Byrsonima spicata (Cav.) DC.; and jacana,
Lucuma multiflora A. DC.

A number of species because of their rapid growth, have been used here
recently for reforestation of bare lands. Most of these are exotics, such as

Caribbean Forester - 56 = Vol. 4, No. 2

pino, Casuarina equisetifolia Forst.; albizzia, Albizzia procera Willd.;
cassia, Sciacassia siamea Britton; and eucalyptus, Eucalyptus robusta Smith.

All of these species require intercultivation or thorough weeding to make
their rapid growth while young. While these species are able to grow on

poor sites they are all intolerant and unsuited for underplanting in existing
deteriorated stands. When raised in pure dense plantations they form a tem-
porary forest which must be periodically replaced by planting. Cassia in
block plantings has been found to stagnate after 3 or 4 years of growth. 1/

Pino, cassia, and albizzia are well suited for lands where cultivation
or frequent weeding during the period of establishment is possible. However,
as has been stated, on the greater part of tne lands which now are, or will
become, available for reforestation, cultivation is difficult and unprofit-
able, and usually incompatible with soil conservation. To obtain the neces-
sary amount of weeding of young trees of these species on private lands
farmers must be convinced of the value of spending money to make trees grow.
This is no small task in a country where people have always closely associated
trees with machetes. Education is needed and will help, but encouragement by
government subsidy or a government land acquisition program may be necessary.
However, the forester may go a long way toward the solution of this problem
merely by the use of aggressive trees which require a minimum of weeding. On
many private lands it may be best to use an aggressive species even if it
lacks many other desirable characteristics. If desired, other species can
easily be introduced by underplanting as soon as the herbaceous vegetation
has been shaded out. Then the gradual removal of the "pioneer" trees in post,
stake, and fuel cuttings will transfer dominance to the young trees of the new
species.

Few trees are known which are suitable for use in reforestation on the
large aggregate area of land, much of it in small privately-owned tracts,
where cultivation is undesirable and adequate weeding is difficult to obtain.
Three species which can be used successfully under these conditions are maria,
Calophyllum antillanum Britton; pomarrosa, Jambosa jambos (L.) Millsp.; and
roble, Tabebuia pallida Miers. These are not extremely rapid growing but
they are aggressive and will withstand adverse site conditions with a minimum
of weeding. Pomarrosa, slow growing at first, makes phenomenal growth from
coppice. A discussion of roble appears in the following article of this
issue.

Tne trees in the first group named are most suitable for underplanting
in existing deteriorated forests. Those in the second group are suitable for
reforestation with cultivation. The third group contains species suitable
for reforestation of poor sites where it is expected that the plantations will
receive little care. These species and others with similar good character-
istics must be used in their proper place if the objectives of Puerto Rican
forestry are to be fulfilled.

l/ Betts, T. F. The Tiv Plantations - 1939 to 1941. Farm and Forest
II (3), Ibadan, Nigeria.

- 57 = January 1943

Resumen

Una de las decisiones més importantes que hay que tomar al convertir
los bosques tropicales mixtos en rodales altamente productivos utilizando
métodos de selvicultura es la de seleccionar las especies que mejor se adap-
ten a los propésitos del manejo forestal. La seleccién depende casi por con-
pleto del problema forestall que se confronta y que en Puerto Rico proviene
del abuso ejercido en el pasado sobre las tierras forestales y ademés de la
demanda actual y venidera de productos forestales.

El desmonte desmedido que se efectuéd en el pasado para poder suplir
las necesidades de una poblaci6én crecients cred el problema del surgimiento
de bosques deteriorados y tierras abandonadas al deslave y de la escasez de
productos forestales. Asi tenemos que la especie forestal que se precisa
debe producir rapidamente los productos necesarios a la par que crece en lo-
calidades relativamente pobres. A pesar de la produccién local, la mayor
parte de la madera que se usa en Puerto Rico es importada y adem&s muchos de
los productos menores tales como postes, estacas y lefia se necesitan perento-
riamente. Por lo tanto el objetivo primordial del presente es el de proteger
las tierras despobladas que no son utiles para la agricultura con una capa
forestal y la produccién de estos productos forestales menores.

Los bosques deteriorados deben mejorarse levantando plantaciones nue-
vas al amparo de ia sombra del bosque. Las especies que se usen deben tole-
rar la sombra y a la vez crecer rdpidamente. Las tierras despobladas que no
Sirven para la agricultura se dividen en dos clases; las tierras mejores en
que es posible intercalar cultivos agricolas hasta que la sombra lo impida
pues de este modo no habrén gastos adicionales ya que la plantacioén forestal
se beneficia del cuidado que se le prodiga a la cosecha agricola, en segundo
término las tierras més pobres que no pueden cultivarse ni siquiera temporal-
mente debido a la topografia accidentada o al suelo somero y que deben sem-
brarse de 4rboles agresivos que requieren un desyerbo minimo.

Algunas especies que sugerimos y otras que estdén usdndose ya para esos
propoésitos estén incluidas en la lista que sigue. Todas estas especies tie-
nen sus ventajas pero también sus desventajas.

Para Sembrar Bajo Sombra Para Repoblar
Junto con cosechas
agricolas Con menor atencioén
Ocotea spp. Casuarina equisetifolia Tabebuia pallida
Guarea guara Albizzia procera Calophyllum antillanum
Hymenaea_courbaril Sciacassia siamea Jambosa jambos
Cerdana alliodora Eucalyptus robusta

Montezuma speciosissima

Byrsonima spicata
Lucuma multiflora

Caribbean Forester - 58 = Vol. 4, No. 2

ROBLE, A VALUABLE FOREST TREE IN PUERTO RICO

Frank H. Wadsworth
Associate Forester
Tropical Forest Experiment Station

The forest production of Puerto Rico is dependent upon the future man-
agement of not only Government-owned forest lands but, equally important, the
large aggregate area of privately-owned non-agricultural land which has been
deforested and is now unprofitably cultivated, serves as poor pasture, or is
virtually idle. Because of wartime pressure for food production a portion of
this land is being cleared for temporary cropping. Probably, normal submargi-
nality of such lands has been removed because of the emergency, but following
the war much of this land must be abandoned regardless of population increases,
because of impoverishment of the soil.

These lands, and others which are idle because of steepness of slope or
poor soil can best serve the people in forest production, but in order to in-
terest private owners in investing in reforestation and management it is neces-
sary that the crop yield early returns and that establishment and management
costs be at a minimum. One tree species which has been found well adapted for
planting on such lands because of its aggressiveness and ability to withstand
adverse site conditions is roble, Tabebuia pallida Miers (or T. pentaphylla
Hemsl.).

The Tree

Roble, or roble blanco, is a member of the family Bignoniaceae. Con-
trary to the implication in this common name, which is applied to the species
throughout most of its range, it is not closely related to the robles of the
Mediterranean region or the oaks (both groups belong to the genus ercus in
the Fagaceae). Roble is, however, in the same genus as the commercially
important Central American prima vera, or white mahogany, Tabebuia Donnell-
Smithii Rose. The natural range of roble extends from Hispaniola throughout
the Lesser Antilles, and from Mexico well into northern South America.

There are 4 other species of Tabebuia native to Puerto Rico, but roble
blanco may be distinguished from them all by the large leaflets, 7 to 15 cen-
timeters long, which make up its 3- to 5-foliate leaves. Roble is a rather
small tree, few Puerto Rican specimens exceeding 20 inches in diameter and
60 feet in height.

Distribution in Puerto Rico
Roble is found from sea-level to elevations of 3,000 feet. Scattered

trees or young volunteer stands are found in all parts of the island which
receive an annual rainfall above 45 or 50 inches.

- 59 = January 1943

Roble is commonly planted in towns, frequently in the plazas, where
its abundant pink or white trumpet-shaped flowers make it very attractive
during the one to several periods each year in whicu it is in bloom. Another
important use of the tree is in fence rows. As freshly cut posts generally
will sprout if promptly set, roble makes a very "durable" fence post. Robles
are to be seen in fence rows in nearly all parts of the island.

Most of the roble forests are on slopes in the interior. A large
group of nearly pure roble stands is to be found on the north and west slopes
of the Luquillo Mountains at elevations ranging from 500 to 2,000 feet above
sea-level.

Products

As has been stated, roble is not a large tree. It will probably never
be important here for its sawn lumber. The wood is not highly prized like
that of mahogany or cedar. However, the tree grows straight and in the for-
est prunes itself early, producing a good bole (see figure 1).

The wood is very useful. It is white in color, hard, heavy, and strong,
and according to Britton and Wilsoni/ the specific gravity is about 0.8.
petonae/ states that it is easy to work, finishes smoothly, and seasons with-
out difficulty. It is used for general construction, tool handles, fence
posts, ox-carts and other similar uses.

Growth

Rapidity of growth is a primary consideration in the selection of a
tree species for use in plantings on private forest lands where the owners
must be convinced of the value of planting and caring for trees. To be of
any practical significance, growth data must be based upon conditions ina
stand or plantation where the tree is subjected to normal forest environment,
including competition of neighboring trees. Many tropical trees are capable
of phenomenal growth when isolated, or when in an open stand or young plan-
tation. However, if they grow slowly when subjected to the natural compe-
tition for light and soil which exists ina forest sufficiently dense to shade
out herbaceous vegetation, these species are of little value for reforestation.

An indication of the growth of roble trees and forests is found in
data collected in 5 one-tenth acre growth plots established in 1938 by A. K.
Thurmond, then a District Ranger in the Caribbean National Forest. The plots
are in a young nearly pure roble stand in the humid northeastern Luquillo
mountains. The elevation of the plots is 300 to 500 feet above sea-level,
and they are located on a degraded heavy clay soil (Catalina stony clay)
about 50 feet above Rio Espiritu Santo. Annual precipitation averages about

see ce server oe

ay Britton, N. L. and Wilson, Percy. Botany of Porto Rico and the Virgin
Islands, New York Academy of Sciences, 1925.

2/ Record, S. J. Timbers of Tropical America. Yale University Press, New
Haven, 1924.

Caribbean Forester - 60 = Vol. 4, No. 2

Fig. 1.—A 4-inch roble left after a
recent clearing, showing the good form of
forest grown trees. Trees such as this one
produce excellent posts and poles.

=G). =

January 1943

100 inches. The present stand came up when previous cultivation was discon-
tinued. A few older trees nearby are probably the progenitors of the stand.

The plots were established to determine the effects of thinning, inter-
cropping, and removal of wolf trees upon the growth of roble. All trees with-
in the plots were numbered with metal tags and diameters and heights of most
of them were measured. Three plots were established in a mixed stand of roble
and other hardwoods; the other two are in a pure roble stand. The mixed stand
was so open and heterogeneous that the three plots give no conclusive indica-
tion of the results of intercropping and the removal of wolf trees. In all
three plots the openings are rapidly being filled with young robles.

The two pure roble plets in the pure roble stand are more fully stocked
and represent average forest conditions in such a stand. They were established
to study the effects of thinning, but at present they are of value chiefly for
their record of growth. The forests in the two plots differ somewhat in den-
sity (see table 1) but this factor is of insufficient importance to produce a
great difference either in their appearance or in measured growth, so the data
are combined for analysis. Figure 2 shows a general view of the unthinned
plot.

Table 1.—Stand per acre in thinned and control plots in
@® pure roble stand just after thinning and four years later.

Unthinned Thinned
1938 1942 1938

No. Basal No. Basal No. Basal No.
trees | areal/| trees | areal/| trees | areal/| trees

0-0.6 760 0.22 770 0523: |) 410 0.12 480 0.14
12/ 380 2.32 580 3.54... 430 2.62 400 2.44
2 420 9.66 450 10.35 430 9.89 360 8.28
3 950, 11.66. 350) ” 27.74.) 220. “iVelsyoescor spon
4 80 7.16 170 ° 15.22 40 S158" 140 des ss
5 10 1.39 60 8.35 50 6.96 90 12.53
6 30 5.99 10 2.00 40 7.98
7 20 5.42 30 8.13
8 10 3.54 20 7.08
9 10 4.47

10
it 10 6x66
12 10 7.92

Total 1950 54.02 2460 85.03 1580 54.32 1810 59.11

Wf Basal area is the aggregate cross sectional area of all trees at 4.5
feet above the ground (at breast height).
Wp The l-inch d.b.h. class includes trees 0.6 to 1.5 inches, d.b-h., etc.

Caribbean Forester - 62 = Volsau4.) Nosec.

Fig. 2.—A young pure roble stand, showing the tall
narrow crowns of the trees, and the characteristic open
canopy.

The preponderance of trees in the small diameter classes, as shown in
table 1, indicates the youth of the stand. As mortality was evidently negli-
gible, the increase in the number of trees in each plot during the four years
equals the number of saplings entering the smallest diameter classes. It
might be expected that the greater number of seedlings would have come up in
the thinned plot because of the openings created by the thinning, but, as is
indicated by the small basal area, both stands were sufficiently open for
such reproduction. Also, under the thinned stand, areas of dense grass
slowed seedling and sapling growth.

A complete study of increment cannot be based entirely upon diameter
growth, for height growth is equally important, particularly where production
of products, such as posts, is a major objective. Figure 3 shows the curvi-
linear relationship between diameter and height in all the trees in the two
plots. There was surprisingly little individual variation from this curve.
It will be noted that roble in this stand had grown to about one half of its
maximum height by the time that it reached 4 inches d.b-h.

- 63 -. January 1943

Height - Feet

Diameter at Breast Height - Inches

Fig. 3.--Relationship between roble diameter
and height in a young pure stand, El Verde.

Information as to the increment of entire stands is valuable for many
purposes, but fully as important is a knowledge of the growth of individual
trees in relation to their various environments within the forest, for with
such information foresters can control reproduction, both as to quanbatyes and
species, the growth of each individual tree, and tree form.

The growth of a tree in a forest is dependent upon many factors, but
the most easily controlled is light (and root) competition. Analysis of the
growth of trees of different diameters is one method of determining the im-
portance of light and root competition as growth-inhibiting factors. Trees
in the larger diameter classes are generally among the tallest and their
crowns therefore receive an abundance of light, also they generally have well-
developed root systems which give them added advantage.

Caribbean Forester - 64 - Vol. 4, No. 2

The permanent tags on each tree in the plots made possible a study of
the relationship between diameter and tree growth. A summary of the data
shows that, as might be expected, the larger trees made the most rapid diam-
eter growth. While those in the l-inch d.b-h. class in 1938 averaged a growth
of 0.3 inch in diameter for the 4-year period, those larger than 5 inches
d.-boh. grew almost 1 inch (see figure 4). The relationship with height growth
is less marked, for many small trees under openings were found to be growing
well in height. Linear correlation of diameter with diameter growth and height
growth gave correlation coefficients of +0.33 and +0.26 respectively.

ee. ol

Four~year Diameter Growth - Inches

Diameter at Breast Height

Fig. 4.—The relationship between 1938 roble diameter and
four-year diameter growth in a young pure stand, El Verde.

The significance of this finding that tree diameter is related to sub-
sequent growth, is that it indicates which part of the stand is responsible
for most of the increment. However, with this information alone one might
logically cut all small trees in an effort to increase growth (leaving the
rapid-growing large trees). This practice might increase growth to some
extent but it would be undesirable silviculturally because it makes no pro-
vision for replacement of mature trees when they are cut.

- 65 - January 1943

Diameter and growth in @ pure stand of roble are related, chiefly be-
cause diameter and dominance, or growing space, are related. Therefore,
dominance is to be considered a better index of growth than diameter. A clas-
sification of trees by the position of their crowns with respect to nearby
trees (assuming this to be also a good index of relative position of the root
system beneath the soil surface) has been widely used in the study of compe-
tition between trees within the forest. The four following broad classes may
be used for pure stands:

Dominants - Trees with crowns extending above the general level of
the crown cover and receiving full light from above and partly
from the side.

Codominants = Trees with crowns forming the general level of the
crown cover and receiving full light from above, but compara-
tively little from the side.

Intermediates - Trees with crowns extending into the crown cover
formed by dominant and codominant trees; receiving a little
direct light from above, but none from the side.

Suppressed - Trees with crowns entirely below the general level of
the crown cover; receiving no direct light either from above or
from the side.

The use of these classes as a basis for the study. of individual tree
growth permits a much more accurate expression of environment than does
diameter. In 1942 all of the trees in the plots were classified according
to this system. It would have been desirable to have the classification made
at the beginning rather than at the end of the 4 years for use in growth pre-
diction, but because of the brevity of the period it is extremely unlikely
that the number of trees that changed classes was sufficient to affect the
results of the study.

As has been pointed out, generally the larger trees are at the dominant
end of the scale and the smailer ones are in the intermediate and suppressed
classes. In fact, linear correlation between diameter and dominance is much
closer than that between diameter and growth. (See table 2.)

Table 2.—Linear correlation coefficients between size

and dominance of roble trees and rate of diameter and height

growth.
Simple Correlations Multiple Correlations
Variables Dominance Diameter and Dominance
Diameter growth +0.33 +0.49 +0.50
Height growth +0.26 +0.53 +0. 53

Caribbean Forester = 66 = Vol. 4, No. 2

Growth was found to be definitely related to dominance. Linear corre-
lation of dominance with diameter and height growth gave coefficients of
+0.4857 and +0.5318 respectively. These relationships are shown graphically
in figure 5. Proof of the superiority of dominance over diameter as an index
of future growth is found in the comparison of the various correlation coef-
ficients in table 2. The multiple correlation coefficients shown also bring
out the weakness of using diameter alone in estimating the growth of indivi-
dual trees in these young stands. The correlation using both diameter and
dominance is no stronger than that with dominance alone.

Inches

0.6}

0.4}.
a eLehG frowtbh (0)

Four-year Height Growth - Feet

0.24. om———= Diameter growth (e)

Four-year Diameter Growth -

‘Dominant Codominant Intermediate Suppressed

Crown Class

Fig. 5.—The relationship between roble diameter and height
growth and position in the crown canopy in @ young pure stand,
El Verde.

Pests and Diseases

Roble is attacked by several insects but the only-one of general im-
portance here is a Lepidopterous shoot borer. The larvae of this insect at-
tack roble twigs throughout the island and are responsible for crowded pro-
fuse branching in many trees. The slight zig-zag appearance of the bole of
some trees is possibly the result of past attacks of shoot borer in the
leader. Roble is also attacked by scales and a leafhopper, neither of which
cause extensive damage.

- 67 = January 1943

Recently, a young roble was found dying as a result of a canker in the
bole. Only two such trees are known to the writer, and the pathogen has not
been isolated. A witches-broom is commonly seen on robles but does not de-
form the merchantable part of the tree.

Roble can be considered relatively free from attacks of pests and dis-
eases but because of the large pure stands which have recently grown up, an
attack of some new pest might easily assume epidemic proportions. This vul-
nerability, a characteristic of all pure stands, regardless of species, pro-
vides argument for gradual conversion to a mixed forest by underplanting or
favoring naturel reproduction of the better species in the understory.

Tolerance of Poor Sites

The tolerance of roble for some of the poorest sites of the island is
common knowledge. Good stands of roble are frequently found growing on steep
slopes and ridges where the topsoil has been entirely removed by erosion fol-
lowing cultivation, and the soil remaining is a nearly sterile heavy clay.

Aggressiveness

Attention was attracted to roble chiefly because of its role in plant
succession. The outstanding silvical characteristic of the species in Puerto
Rico is its ability to invade non-forested pastures. It is a prolific seeder,
isolated trees sometimes producing several large seed crops within one year.
Along the edge and in the openings in most stands are young trees which have
obviously come up through herbaceous vegetation, particularly where grazing
has kept the grass low. Also there is unmistakeable evidence in existing
young mixed forests that roble was a pioneer species in the development of
the cover, having invaded a grassy or weedy deforested area and gradually
attained dominance. These observations alone are of sufficient significance
to justify intensive study of this tree.

Tolerance of Shade

In many different ways roble shows its intolerance of shade. In dense
stands heavy seed crops may fall to the ground each year, but there is a gap
in the size classes between the small saplings and poles with crowns in the
lower edge of the canopy in which no young trees seem to be able to grow be-~
cause of shade. Presumably sapling mortality is high under these circum=-
stances. In more open stands this gap does not exist. The high mortality of
young trees is shown by a small seedling plot established under a dense roble
stand. The April 1942 seed crop had stocked the 64 square feet within the
plot at the rate of more than 2,500,000 seedlings per acre. Nine months
later two-thirds of them were dead, and judging by the small number of older
seedlings and saplings in the vicinity it is doubtful that 5 per cent will
survive for 12 months.

The intolerance of roble is shown also by its less prominent position
in older, more dense forests. None of the pure roble stands are very old,
few having dominant stems larger than 8 or 10 inches d.beh. The oldest roble
trees found are in mixed forests, and in many of them roble is a minor

Caribbean Forester = 68 = Vol. 4, No. 2

representative. These old specimens are apparently remnants of a former more
open stand containing many robles. Moreover, robdle is obviously not maintain-
ing its numbers in denss stands, regardless of the number of seedlings which
may start.

The Place of Roble in Forest Management

General observations supported by plot records indicate that foresters
will be wise to assign roble primarily the same task on managed forest lands
as that which it accomplishes so well on unmanaged lands: the reforestation
of poor sites. In existing forests it is doubtful that conversion to roble
would be a simple matter. As has been pointed out, roble is intolerant of
shade and conversion would require destruction of the canopy to give light to
the young trees. In an experimental underplanting made in a moderately dense
woodlot the trees are growing well only where they were planted under scat-
tered openings in the canopy.

Roble can be looked upon as merely & nurse crop to be replaced later
with other species, or it may be desirable to perpetuate it as the major
species in the stand. Whichever decision is made, no difficult silvicultural
problems need be faced, provided that conversion to more intolerant species
is not attempted.

Establishment

Roble can be propagated by seed or cuttings. The use of cuttings has
little practical importance in forestry here at present, primarily because on
@ large scale it is considered more difficult than seed propagation. However,
as has already been pointed out, roble fence posts, which are cuttings of a
type, will generally grow if set in moist soil promptly after cutting.

Not to be overlooked in any reforestation program is an already exist-
ent supply of wilding planting stock to be found under most of the roble
stands of the island, The production of these wilidings by a forest is not
entirely reliable as a source of planting material over a period of years, and
seedlings grow more slowly in the forest than in 4 nursery, but by using this
type of planting material the investment required for nursery maintenance is
obviated. Also, wiidings, because of their slow growth need not be trans-
planted immediately when they become of size, but may be lifted as needed and
at the best season of the year. In the existing stands of the island there
is a supply of wilding plants sufficient to plant thousands of acres. In
many pure stands wildings are so abundant that they can be pulled nearly as
rapidly as stock ina nursery. In the two small pilots measured there were
1,770 saplings from 1 to 4.5 feet tall. Of these, probably 50 per cent or
about 4,500 per acre would be suitable for transplanting.

Propagation by seed begins with collection. The greater part of the
seed is borne in the spring, but open-grown trees sometimes flower several
times during each year, so it is possible to procure at least a small amount
of seed at almost any time. The seeds should be collected when the pods have
turned yellow in color one or two days before they open. As the seeds are
winged and light (20,000 to the pound) they are blown from the pods almost

= 69 = January 1943

immediately after the latter have opened. The seed is relatively perishable.

When stored with the natural moisture content at air temperature it loses its

viability rapidly during the first few weeks but a few seeds may remain viable
for 3 or 4 months.

The apparent aggressiveness of the species led to the establishment of
several direct seeding plots on the island at both low and high elevations. In
several plots seeds were sown about one-quarter inch below the soil surface in
the center of 2-foot cleared spots. In several others they were broadcasted,
some of the plots being left with dense herbaceous vegetation, and others
cleaned. All of these plots failed, and led to the following conclusions:

(1) Roble seedlings are very delicate during the first month of growth.

(2) Because of this condition the seeds should not be more than just
covered with soil, particularly if the soil is heavy.

(3) During the first two or three months many seedlings die from the
effects of long exposure to direct sunlight. (Seeds broadcasted
under dense grass had a high early survival.)

(4) Broadcasting is inferior to spot seeding because most of the
broadcasted seeds lie on top of the soil, remaining dry, except
under very favorable weather conditions, until they are no
longer viable. Also, broadcasted seedlings, being scattered at
random, are more difficult to care for than seedlings in rows of
spots.

(5) If for no other reason, the success of direct seeding of roble in
the field is jeopardized by the slow growth of the young seed-
lings during the first year. Even in a nursery the seedlings do
not exceed 3 or 4 inches in height after 3 months. This slow
early growth in the field might greatly prolong the period during
which care is necessary.

Wherever wildings are available it is recommended that they be used for
planting stock. Where the supply is inadequate it will be necessary to grow
nursery stock, since direct seeding has several disadvantages.

In the nursery the seed should be sown at a depth not greater than one-
quarter inch. To yield sturdy stock seeds should be spaced 4 x 2 inches apart.
During the rainy season no watering is necessary. Seedlings raised under 50
per cent lath shade frames were of better color than those completely exposed,
but no effect on growth was noticed, so evidently the shade is desirable but
not entirely necessary. No nursery diseases of the species are known here.

After 5 or 6 months the seedlings are about 18 inches tall and may be
lifted, or they may be held until they are 48 inches tall if necessitated by
weather conditions. With ordinary care in lifting, wrapping of roots, and
planting, mortality can be kept very low. Of 100 wildings planted on a moist
site in the interior, there were 98 living 8 months later. The trees should
be spaced at either 6 x 6 or 8 x 8 feet, preferably the former, in order to
rapidly obtain a cover.

Caribbean Forester - 70 - Vol. 4, Now 2

Data are not yet available on the growth and mortality of the planted
trees during the period of their establishment in the field. However, numer-
ous plots have been established where trees have been planted following two
types of site preparation, complete clearing of herbaceous vegetation, and
clearing of a circle 2 feet in radius about the tree. A series of control
plots without preparation has also been established.

It is recognized that it will be necessary to keep the trees free from
vines, and during their first year, also to occasionally weed them. However,
observations indicate that little additional care should be necessary.

During the next few years none of the trees should be removed, regard-
less of the fact that some may be misshapen, for crown canopy which shades
out the herbaceous vegetation beneath it is the first objective of reforest-
ation.

Silviculturali Control of Established Forests

When the crowns of the trees meet and the canopy starts to kill the
herbaceous vegetation by shading, a@ silvicultural policy should be formulated
to guide the production of the forest, in both quantity and quality. To
properly control composition and increment, such a policy must designate which
species are to be favored and what cutting plan should be followed.

Much can be learned about desirable silvicultural practices by obser=
vation in existing roble forests. Nearly every significant silvical charac-
teristic of roble stands reflects the intolerance of the species for shade.
The trees may grow close together but their crowns are small and therefore the
shade produced is not dense. (See figure 2.) The data here presented show
that even in this relatively light environment the young robles below the
canopy are making negligible growth.

Another important consideration in roble stands is the present and
prospective representation of other species. In all pure stands seen there
is evidence of impending invasion by other species which are more tolerant
of shade and can therefore grow up under the roble and eventually replace it.
Two common invaders are pomarrosa, Jambosa_ jambos (L.) Millsp. and the
laurels (Lauraceae).

If pure roble is to be perpetuated, the stand must be selectively cut
frequently to keep it sufficiently open to permit growth of the reproduction.
All other species should be gradually removed from the stand. However, over-
cutting must be carefully avoided, for it may set the stand back several
years. An example is shown in the thinned plot data presented in table 1. It
was recorded that 55 posts were removed from one plot for comparison with the
other, a control. Unfortunately, details of the cutting policy are lacking,
except that the "best" trees were left. A comparison of the stand table of
the reserve stand with that in the control plot in 1938 indicates that many of
the trees removed were large, for only those below 6 inches d.b-h. were left,
While in the other stand there were several up to 8 inches d.b.h. Table 1
shows that 510 saplings entered the unthinned stand (grew to 4.5 feet in height)
from below during the 4 years, but in the thinned stand the increase was only

= 4h '= January 1943

250. This is because parts of the thinned plot were so open that herbaceous
vegetation invaded and slowed tree reproduction.

Overcutting also sets back the aggregate increment of the dominant
class. In this same thinned plot the 4-year basal area increment was 72.2
per cent as compared to only 57.4 per cent in the unthinned stand. However,
the higher increment per cent of the thinned stand is of little significance
as it was calculated from a smaller base (1938 basal area) and the actual
square foot basal area increment of the two stands was about the same.

The close relationship between the dominance and growth of roble trees,
regardless of their size, must be borne in mind when cuttings are made. Thin-
nings which remove the subordinate trees to increase the growth of the domi-
nants will have little beneficial effect because such cuttings do not materi-
ally change the environment of the already dominant trees. In fact, if the
other extreme was used (i.e. the removal of the dominants, releasing the now
slow-growing subordinates and making them dominant) it would produce acceler-
ated growth in these new dominant trees which would approach that of dominants
in the uncut stand. However, because quality as well as quantity production
is desired neither extreme is satisfactory. Consideration of the potential
value of each tree, both dominants and subordinates, is necessary. If among
the dominant trees there are some of excellent form which are capable of
large value increment they should be left until they produce a 6- to 10-inch
bole for small lumber or large posts, regardless of the number or quality of
the trees beneath them. OQther. dominants will increase only very slewly in
value because they are as large as is needed, or because, due to crook or
other abnormalities, they are suitable only for fuel. These should be cut
to give the growing space to more promising trees, even though they be much
smaller at present. Subordinate trees which show little promise because of
poor form should also be cut. This policy is flexible and can be followed
both in light cuttings in open stands and heavier cuttings in dense stands.

In open stands it may be impossible to remove all undesirable trees in one
cutting without unduly exposing the site to invasion by herbaceous growth.
In such forests only a few of the worst trees should be removed in each cut-~
ting. Likewise in very dense stands it may be desirable to take a few trees
of good form in addition to those of low value in order to open up the stand
sufficiently to provide for reproduction.

If it is considered desirable to establish other species in the stand,
two methods may be used: (1) the encouragement of other species already in-
vading, or (2) underplanting other species.

Conclusions

Thus roble can admirably serve in the reforestation of idle lands, at
present an all-important phase of forest land management in Puerto Rico.
Early growth of the seedlings is slow but planting stock is readily available
in existing woodlands and little care is necessary. When the stand is a few
years old forest conditions are established, and any of a great number of
species may be introduced into the stand if desired, while the gradual har-
vesting of the roble will provide early income. On the other hand, the main-
tenance of a nearly pure roble stand will provide e continuous high yield of

Caribbean Forester - 72 = Vol. 4, No. 2

excellent small poles, house posts, tool handles, and similar products. It
is necessary to prevent the invasion of other species but this can be accom-
plished by occasionally cutting the saplings of other species.

In & country such as Puerto Rico, where the biggest immediate job con-
fronting foresters is that of providing an abundant cheap supply of small wood
products and fuel, and where it is necessary to reforest large areas of pri-
vate lands where little control or care of the trees is possible, aggressive
species such as roble must be used if early success is to be achieved.

summery

One of the greatest problems which must be faced before forest produc-
tion in Puerto Rico can reach the maximum compatible with other land uses is
the reforestation of cleared lands which have been, or will be, virtually
abandoned because of impoverishment or loss of soil. To accomplish this ob-
jective the forester must find various means of making reforestation and for-
est management attractive to private owners. One way is in the use of trees
which can be cheaply established and which will grow rapidly on the type of
land available for forest. A species which has been found well adapted is

roble, Tabebuia pallida Miers (or TI. pentaphylla Hemsl.).

Roble is found throughout tropical America. In Puerto Rico it grows
in most parts of the island, but chiefly in the mountainous interior. The
hard, heavy, strong wood is used in general construction, tool handles, ox-
carts and similar uses.

The growth rate of open-grown trees is not extremely rapid, but in the
forest, growth compares well with other species. The trees grow straight and
have few branches. The larger dominant trees were found to grow about an
inch in diameter and 7 feet in height in four years. The growth of the sub-
ordinate trees is in proportion to the amount of light they receive, with the
suppressed trees making very little growth. While there is a definite cor-
relation between diameter and diameter growth that between dominance (amount
of light received) and diameter growth is much stronger. By classifying a
tree as dominant, codominant, intermediate, or suppressed it is possible to
predict its growth from the data presented.

One of the outstanding characteristics of roble is its tolerance of
adverse site conditions. Good stands are often seen on ridges where the top-
soil has been largely removed by erosion.

The aggressiveness of the species is one of its best characteristics
for reforestation purposes. It competes well with grass and weeds and is
often the first tree to naturally appear on abandoned land. Undisturbed
areas surrounding roble forests frequently contain many saplings which are
a result of the large amounts of wind-blown seeds produced.

Roble is definitely intolerant of shade and forms 4 temporary forest
type. Young pure forests are gradually invaded by other species, as roble

- 73 = January 1943

seedlings do not grow well under the shade of the parent stands. In old stands
the species is generally a minor representative.

If roble is to be used in reforestation it must be planted. Direct
seeding has been unsuccessful because the seedlings are delicate for several
months, and early growth is slow. Propagation in the nursery is not difficult,
but large amounts of planting stock can be obtained from existing forests.

Wild stock 18 to 36 inches in height is plentiful in moist pure roble forests,
and it transplants well during the wet season.

The planted trees need early protection from the competition offered by
herbaceous vegetation. The cutting of weeds in a circle 3 or 4 feet in diam-
eter around each tree is necessary during the first year.

When the stands have grown to the point where a canopy is produced the
poorer specimens should gradually be removed in charcoal or post cuttings -
but always leaving the stand sufficiently dense to prevent invasion by grass.
Seedlings of other species, which will begin to appear in the forest floor
should be encouraged if they are of value. These other species will gradually
replace the roble unless they are cut back and unless the stand is kept suf-
ficiently open for the growth of roble saplings. However the development of a
mixed forest will provide a wider variety of products.

As roble is a small tree from which little lumber is produced, the main
products for which the species is used do not require trees larger than 8 or
10 inches in diameter. Therefore all trees of this size can be considered
mature, but cuttings should always be selective, leaving sufficient stems to
preserve forest conditions.

Thus roble, a species which does not produce a valuable furniture wood
for export, is important in Puerto Kico because of its usefulness in reforest-
ation, because a supply of wild planting stock is already available in exist-
ing stands, because the species is aggressive and able to withstand adverse
site conditions, and because the wood is useful for the many wood products re-
quired by our own huge agricultural population. The species has certain limi-
tations as do all trees when considered for use under the wide variety of
conditions existing in Puerto Rico, but such species, the products of which
are primarily of value for local consumption, must be grown in part of our
forests if they are to be of greatest service to our people.

Resumen

Uno de los problemas més grandes que es preciso allanar para que la
produccién forestal en Puerto Rico alcance un grado m&éximo, compatible con
otros aprovechamientos del terreno es la repoblacién forestal de las tierras
despobladas que han sido o serdn virtualmente abandonadas debido al empobre-
cimiento o deslave del terreno. Para lograr este objetivo el forestal debe
buscar los medios para hacer que esta repoblacién y manejo forestal halague a
los terratenientes. Uno de estos medios es usando arboles que puedan estable-
cerse econémicamente y que crezcan rapidamente en las tierras disponibles para

Caribbean Forester =- 74 = Vol. 4, No. 2

bosques. Se ha determinado que el roble, Tabebuia pallida Miers (o T. penta-
phylla Hemsl.) es una de las especies que se adaptan a estas condiciones.

El roble se encuentra a través de toda la América tropical. En Puerto
Rico crece en casi toda la isla pero principalmente en la regién montafiosa del
interior. La madera es dura, pesada y se usa en construcciones en general,
mangos, carros de bueyes, y demds usos similares.

El indice de crecimiento de los d4rboles de roble dispersos no es extre-
madamente rapido, pero su crecimiento en el bosque si puede muy bien compararse
con el de otras especies. Los arboles crecen rectos y desarrollan pocas ramas.
Se determiné que la mayor parte de los 4rboles dominantes crecen a razén de l
pulgada en didmetro y 7 pies en altura en cuatro afios. El crecimiento de los
&rboles subordinados esté& en proporcién a la cantidad de luz que reciben, sien-
do los &rboles oprimidos los que crecen muy poco. A pesar de que existe una
correlacién definida entre el didmetro y el aumento progresivo en didmetro, la
correlacién que existe entre la dominacién (que depende de la cantidad de luz
recibida) y el crecimiento en didmetro es mucho mayor. Si clasificamos un
arbol en dominante, codominante, intermedio u oprimido podemos predecir su cre-
cimiento de acuerdo con los datos incluidos en el texto.

Una de las caracteristicas més sobresalientes del roble es su habilidad
para tolerar localidades adversas. A menudo observamos rodales buenos aun en
cuchillas que han perdido las capas superiores del suelo por deslave.

La agresividad de la especie constituye una de sus mejores caracteris-
ticas para fines de repoblacién forestal. A menudo es de las primeras en sur-
gir espontaéneamente en terrenos abandonados y compite bien con las malas yerbas.
Debido a la enorme cantidad de semilla que produce y que el viento esparce por
todos lados, las tierras adyacentes a los bosques de roble se ven frecuente-
mente pobladas de numerosos arbolitos.

Se sabe positivamente que el roble no tolera la sombra por lo cual for-
ma un tipo forestal temporero. Los rodales jévenes van siendo invadidos gra-
dualmente por otras especies ya que los arbolitos de roble no pueden desarro-
llarse debidamente bajo la sombra de los 4rboles madres y por lo general en
los bosques mas viejos esta especie abunda menos.

Si el roble ha de usarse en repoblacién forestal debe recurrirse al uso
de arbolitos. La siembra directa no ha dado buenos resultados porque las plan-
titas son muy delicadas durante los primeros meses y su crecimiento inicial es
lento. La propagacion en los viveros no es dificil pero pueden obtenerse gran-
des cantidades de material de siembra de los bosques ya existentes. Arbolitos
silvestres de 18 a 36 pulgadas de alto abundan en la mayoria de los rodales
puros de roble y trasplantan bien durante la época de lluvias.

Durante el primer afio los plantones deben protegerse contra la compe-
tencia de la vegetacién herbdécea, manteniendo un circulo de 3 6 4 pies de
didmetro limpio de yerbas alrededor de cada arbolito.

Cuando el rodal empieza a cerrar los ejemplares més pobres deben cor-

tarse gradualmente para postes o carbon, pero dejando suficiente sombra para

i 75. = January 1943

mantener a raya las malas yerbas. Deben estimularse aquellas especies de va-=

oe que aparezcan bajo el roble. Estas irdén reemplazadndolo gradualmente a
menlos que no se corten con miras de perpetuar el roble y se mantenga el rodal
suficientemente abierto para permitir el crecimiento de éste. Sin embargo un
josque mixto nos proveerd de mayor variedad de productos.

Como el roble es un Arbol que proporciona poca madera, los principales
productos para los cuales se usa no requieren drboles de mds de 8 6 10 pulga-
das de didmetro. Por lo tanto todos los drboles de ese tamafio pueden conside-
varse listos para la corta, que debe ser siempre selectiva pero dejando sufi-
clientes a&rboles para conservar asi las condiciones forestales.

Por lo tanto, el roble, una especie que no suministra madera de valor
o2ra exportar, es importante en Puerto Rico por su utilidad en repoblacién
forestal porque el material de siembra esta disponible en los rodales existen-
i283, porque es agresivo y puede soportar localidades adversas y proporciona
ademéis numerosos productos necesarios a la enorme poblaci6én agricola de la
“sla. La especie, como todos los 4rboles, tiene ciertas limitaciones cuando
su uso se considera bajo la inmensa variedad de condiciones existentes en
Puerto Rico pero las especies cuyos productos son de valor principalmente
para el consumo local, deben cultivarse en nuestros bosques si es que han de
wendir el mayor provecho a nuestro pueblo.

000

LATIN-AMERICAN FOREST RESOURCE SURVEY ORGANIZED

In November 1942 Mr. Bevan, Director of the Tropical Forest Experiment
Station was called to Washington on detail to organize a Latin-American Forest
Resource Survey to be conducted by the U. S. Forest Service. The objective of
the survey is the determination of the character, extent, and availability of
the forest resources of certain Latin-American countries. The preliminary

organization has been completed, and Mr. Bevan is now in the field starting
tWO CYreWS o

In the party are several well known foresters and botanists, including
William R. Barbour, formerly Forest Supervisor of the Caribbean National Forest
-n Puerto Rico; L. R. Holdridge, formerly in charge of the Forest Management
Division of the Tropical Forest Experiment Station and recently Forest Manager
for Société Haitiano-Americaine de Developpement Agricole; Elbert L. Little Jr.,
e dendrologist in the Washington Office of the Forest Service; and José Marrero,
Principal Biological Aide of the Tropical Forest Experiment Station.

Caribbean Forester =- 76 - Vol. 4, Now 2

COMMENTS ON THE SILVICULTURE OF CEDRELA

L. R. Holdridge
Manager, Forestry Division
Société Haitiano-Americaine de Developpement Agricole
Port-au-Prince, Haiti

Mr. Beard’s summary of silvicultural troubles with Spanish cedar in
Trinidad in the Aprii issue was read with extreme interest not only because
the material was well organized but also due to its treatment of a silvi-
cultural subject of such vital interest to all Caribbean foresters. Reading
the article has set my mind working once more on this extremely thought
provoking problem and in my mind’s eye I see again the hundreds of acres of
Spanish cedar graveyards and infirmaries which are scattered through the
West Indies.

Sitting here in the high mountains of Haiti with a heavy jacket on for
warmth may or may not be conducive to the proper mental attitude for consider-
ing a tropical tree, but I would like to make a few notes and theorize a bit
on the cedar question. I recognize only two species in the West Indies,
Cedrela odorata L. and C. mexicana Roem., and although herbarium specimens
might puzzle me, the two can readily be distinguished as trees even at some
distance by the combination cof a greater number of leaflets per leaf and
smoother bark of C. mexicana. These characters must be accorded however, with
the relative size or age of the tree. The two are so similar that silvicul-
tural practices for one should fit the other.

Although I personally recommended discontinuation of planting of
Cedrela in Puerto Rico, I am firmly convinced that the tree can be grown sat-
isfactorily. But to do this, we will have to stop making plantations and
make plantings instead, and discard all the meters and test-tubes for the
time being in favor of some good logic and common sense until the problem is
narrowed down sufficiently that such instruments may be employed with success.
This statement implies no criticism of Trinidad’s efforts as their forestry
work is exceptionally sound and for 99 trees out of 100, their investigations,
as carried out on Cedreia, would have cracked the problem. It just shows that
Spanish cedar is the one tree in a hundred for which some radically new ap-
proach is necessary.

Let us forget for a moment our own respective islands or countries and
look at the whole of the western hemisphere. Surprisingly, we find Cedrela
(consider it one species or several depending on your school of thought),
widely scattered, extending over a range far surpassing most other tropical
American trees. This means that the tree is not very particular about macro-
climatic factors or soil types. We have a wide range of conditions in which
to plant the tree and correspondingly, plenty of space in each island or
country where it should be possible to grow it successfully. But, when we
make a plantation within the limits of this range we are almost certain to
have a complete failure on our hands, a condition cuite perplexing and very
inconsiderate of the foresters. Maybe we had better look more closely at the
tree in its natural habitat.

a a January 1943

Cedrela is a prolific seeder with light, winged seeds readily trans-
ported by the wind over considerably distances. The seeds have a high germi-
native capacity, they germinate rapidly, the tree gets away fast and continues
to grow as rapidly as most of its associates. With a horse of this color, in
every race for the roof following the crash of a forest giant in natural woods,
we ought to have one or more cedars place or be in the running for a long
while. But we don’t find that condition existing when we cruise natural for-
est. We might assume then that Cedrela, within the broad limits of its range,
can only work up enthusiasm for survival and growth when it finds certain con-
ditions together in particular spots. This theory of a happy combination of
certain conditions or factors would seem to explain why only scattered trees
are found in natural forest, why plantations as a whole are failures, and why
research, in running down any one factor at a time, would give insignificant
results. Also, there is a clue indicated for further probing, namely, that
the factors for success which must be determined will be those which are not
widespread and which could occur in different soil types and under different
macroclimatic conditions. The need for a set of special site factors in com-
bination is further borne out by observations in young plantations, where a
tremendous difference in growth of individuals is observed up to the time when
the whole plantation fails.

This all reminds me of something which may seem far-fetched, namely,
black-snakes in southern Connecticut. In walking through the woods at random,
I would suddenly become conscious that right before me was a certain combina-
tion of conditions which I immediately associated in my mind with a black-
snake. There was the flat ledge on a sunny slope surrounded by brush and with
plenty of loose rocks and crevices for a fast snake getaway, which, out of all
the territory through which you could find the snakes, represented the partic-
ular spots where they were located. Isn't that just like Cedrela?

Now, to press the analogy beyond its limit, suppose we took a big
square section of country in which we knew the black-snakes thrived, and then
we tied out hundreds or thousands of snakes in geometrical lines. A few
snakes would be happy and thrive because they happened to be located on the
right spots but the majority, although all right for a while, would become
increasingly morose and weakened until an epizootic or vermin, supposing
there were such, would start spreading through the whole area. We all can
imagine that such an affair would spread so thoroughly that no snake would
be left healthy. But, whether or not such a fantasy has any connection with
trees, we still continue to space out Cedrela evenly across the country.

My own recommendation would be to forget all about plantations of the
tree and try instead to bring to maturity 5 trees per acre in every forest
plantation within the range of Cedrela whose component species would not be
incompatible with it. To produce an average of 5 mature trees, about 25
trees per acre should be planted a day or two before the other species are
filled in, and these 25 seedlings should be planted on the most likely spots
for their success. In subsequent examinations of the plantation, any cedar
developing signs of disease or unthriftiness should be cut down immediately
so as to remove it as a source of infection for the remaining thrifty trees.
At best, the resuits of such a system would be a far cry from the always
hoped for solid crops of Cedrela but, at any rate, one twentieth of a loaf

Caribbean Forester - 78 = Vol. 45 Nos) 2

would be better than none at all.

The only difficult part of such a system seems to be that of selecting
the right spots, but with the pooling of our present knowledge, it really isn't
very difficult to start, and the results of successive years of experience
would give us an exact set of criteria. To begin with, we have to think back
to all the thrifty cedars we have seen and combine that knowledge with the re-
sults of research in order to set up the specifications to be searched for
before planting the trees.

I would look for 4 spot in full sunlight, where the soil is porous or
well aerated, where the tree could, so to speak, keep its toes wet and its ~
ankles dry, and where there is a rich soil or the topography is such that
nutrients can be washed down to the growing tree. I have seen a planted tree
in such a spot that reached 10 inches d.b.h. in 10 years time. Consider for
a moment how well this combination explains the distribution of Cedrela in
natural forest; it is only here and there that one can find such a set of
conditions and yet they could occur over a wide range of forest types, soil
types, geological types, etc. This probably would have been observed long
ago in plantations if it were not for the fact that cedar is a sensitive tree
and the buildup of insects and diseases on the hundreds of unthrifty trees was
too much for the properly located trees to resist. At any rate, it is the
only set of factors which I can connect consistently with thrifty cedars in
all the places where they have been seen. Trinidad foresters have always
pointed out the contrast between their unthrifty cedar plantations and the
healthy individuals which came up in the hedge-rows along the highways. Yet
the plantations would not qualify as to the above conditions whereas the
cedars in the hedges usually bordered the road ditch which gave the proper
moisture relation, if the soil itself were not porous then the proximity to
the bank gave sufficient aeration, ditches or hedges make good catch-alls of
rubbish for tree food whether thrown, blown or swept and, of course, there was
plenty of sunlight. It is interesting to note the frequency of the bank-cedar
combination in areas of heavy clay soil. In such areas in Puerto Rico, good
cedar trees are almost invariably found on a bank of some sort, whether a road-
fill bank, a trail bank, or a stream bank. In the case of the road or trail
fill it is understandable that the soil would be more porous but with stream
or other natural banks, there is probably sufficient movement and instability
of the soil to give a satisfactory aeration equivalent to that of a soil
porous by nature.

The full sunlight will be easy to find but I have mentioned that be-
cause of Hypsipyla grandella. It would seem silviculturally wise to plant
under a shelter-wood to reduce the insect damage, but a glance at the colcr
of the leaves coupled with the knowledge that the tree is deciduous would
indicate the desirability of full sunlight. The combination of good thrifty
trees and only a few per acre should eliminate the shoot-borer.

As for water and nutrients, by simply considering the fast growth of
the tree as well as its thin, freely transpiring leaves, we can understand why
the tree must take up plenty of water and nourishment through its roots. Sand
culture studies of small seedlings might not be able to throw much light on
the quantitative needs of a large tree.

- 79 = January 1943

But why the tree wants to be able to reach down to sufficient water
while in leaf and yet have its roots well aereted is hard to explain. Evi-
dently it is just the “nature of the beast". This, however, would be deduced
from the Trinidad notes. The rotting roots indicate lack of aeration and
this is further seconded by the tree's efforts to balance this by taking in
more air through the bark, viz. the enlarged lenticels. In selecting a
planting spot in a wet or humid type, it would not be necessary to put the
tree on a bank above running water. A normally dry-ditch bank would provide
the desired drainage-aeration relation and during drier periods there would be
considerably more moisture available at the bottom of the ditch than in the
ground above.

Well, anywey, I agree fully with Mr. Whitney's previous article in the
Caribbean Forester, namely, “Isn't Research Fun?". I have thoroughly enjoyed
developing these thoughts on cedar and whether or not the proposed system is
the real solution, it should come much closer to the silvicultural truth than
the previous pure plantation methods. Madame Cedrela is a temperamental lady,
but we should be able to win her heartwood before too long.

oOo

TRINIDAD LANDS ADVISORY COMMITTEE PROVES SUCCESSFUL

In a recent article "The Lands Advisery Committee of Trinidad" in the
Empire Forestry Journal, R. L. Brooks, Trinidad's Conservator of Forests,
describes a method of bettering and strengthening the land policies of various
government agencies through the formation of a committee consisting of their
directors. The success of this committee in improving land policies, as re-
ported by Mr. Brooks, should be of considerable interest to members of govern-
ment agencies in other countries.

The Trinidad committee is made up of representatives of all interests
of land use and development, the Director of Agriculture, the Inspector of
Mines, the Director of Works and Transport, and Sub-Intendant of Crown Lands,
and the Conservator of Forests. As three members are also in the legislative
Council, the committee has a strong influence.

One of the first recommendations of the committee greatly accelerated
progress on the soil and vegetational survey being conducted by the Soil
Department of the Imperial College of Tropicai Agriculture. Another led to
steps to prevent further clearing of steep unstable lands in the Northern
Range. Other recommendations indicate that through such a committee greater
recognition and appreciation of the relative importance of the various forms
of land use, including forest, can be obtained.

Caribbean Forester = 80 = Vol. 4, No. 2

LADY-BEETLES DON'T BEHAVE

George N. Wolcott, Entomologist
Agricultural Experiment Station
Rio Piedras, Puerto Rico

(A paper presented before a Joint Meeting of the Zoology and Forestry
Sections of the Sociedad Americana de Ciencias Agricolas. )

The misbehavior of lady-beetles, in eating what they want to, and not
necessarily what the person responsible for their introduction from abroad
wants them to, once again has been illustrated in Puerto Rico. That big,
glossy-black hemispherical lady-=beetle with a big red spot on each elytron,
Chilocorus cacti L., ostensibly introduced into Puerto Rico to feed on the
scale-insects infesting bamboo, may eat bamboo scales in captivity and under
force of circumstance, but free and able to do its own choosing, finds the
scales oh papaya much more to its liking. Of all the lady-beetles ever intro-
duced into Puerto Rico, just at present it is very much the most in evidence,
having repeatedly been collected in cane fields, where there is nothing in
particular for it to feed upon, and often noted as great aggregations of
larvae or pupae on scale-infested papaya trees. Despite the desirability of
the scales on papaya as food for these beetles, the naked trunk of the papaya
offers no protection from the attack of lizards, and many a promising colony
of lady=beetles has been wiped out by a single lizard living in or on a par-
ticular papaya.

Hardly a fence-post in Puerto Rico but serves a a look-out post for
the big, crested brown lizard, Anolis cristatelus Dumeril & Bibron, and,
from the standpoint of this lizard, the smooth, unbranched trunk of a papaya
is almost as good as a fence-post as a base for foraging for insects. Adult
scale-insects are immobile, as well as being too small to attract the notice
of this lizard, and the crawling young are much too minute; but the lady-
beetles and their larvae which feed on scale-insects are of a suitable size
and move quite fast enough to attract the attention of the lizard, but not
so fast as to escape capture.

Much better from the standpoint of protection against lizards is the
common red-leaved hedging plant, Acalypha wilkesiana, which is also often
heavily infested with the same grey scale, Pseudoparlatoria ostreata
Cockerell, that occurs on papaya. Extensively planted on the grounds of the
Agricultural Experiment Station at Rio Piedras, a close examination of the
bushes will show that, extensively infested with the scale-insects, they are
harboring thousands of lady-beetles feeding on these scale-insects, and by
their bushiness, protect the ledy-beetles from lizard attack. Eventually,
of course, when all the desirable scales have been eaten and are no longer
available, the lady-beetles may go back to eating those on bamboo, but for
the present they are doing very nicely on Acalypha.

- 81 = January 1943

A much earlier parallel instance was the introduction of Cryptolaemus
montrouzieri Mulsant, ostensibly to control the mealy-bugs of sugar cane.
In captivity in Puerto Rico, these beetles fed upon and thrived eating sugar
cane mealybugs, but free and in the open, they were much too clever to seek
out mealybugs snugly hidden under cane leaf-sheaths when an abundance of
naked and unprotected mealybugs and fat green soft scale-insects awaited
them on many another host. Not until an outbreak of Pulvinaria iceryi Guerin
occurred on sugar-cane many years after their introduction were they ever
seen in a cane field, and due to their fondness for this particular kind of
soft green scale they promptly disappeared as soon as this outbreak was
brought under control. Much more abundant in Puerto Rico, however, previous
to the introduction of Cryptolaemus, was another soft green scale, Pulvinaria
psidii Maskell, especially on trees of bucare, Erythrina glauca. These
scales were so abundant on the avenues of bucare trees to the Experiment
Station and to the cemetery of Rio Piedras as to cause @ premature shedding
of the leaves in early winter, and for months the trunks and branches would
be naked and ghostly with the thick whitish festoons of excretions of the
scales. The introduction of Cryptolaemus changed all that, and a few years
later hardly a scale was to be found on this host. Now and then one still
finds it on palo de mufieca, Rauwolfia tetraphyila, but almost invariably
accompanied by Cryptolaemus. Now that these fat green scales have almost
disappeared, Cryptolaemus beetles often feed on exposed mealybugs, but never
yet have been found trying to get under the leaf-sheaths of sugar=cane to
eat the mealybugs ensconced there.

It is fortunate for the survival of all of these introduced lady-
beetles that they can eat some other kinds of scale-insect or mealybug than
the particular one that they prefer, for too great specialization in food has
most unfortunate results. Rodolia (Vedalia) cardinalis Mulsant feeds only
on cottony cushion scale, Pulvinarie iceryi Maskell, and absolutely on not
another insect present in Puerto Rico. It is altogether too effective in
the control of this pest, and for the past year we have been unable to find
the beetle at all, or even the scale in sufficient numbers to support even
& wery modest-sized colony of lady-beetles. Locally the lady-beetle has
actually been almost starved out of existence by its own restricted diet.
Eventually, a flare-up in cottony cusnion scale may be anticipated. If any
Rodolia beetles have survived by that time, of course they will bring this
outbreak of cottony cushion scale under prompt control, but if none survive
the present scarcity of their sole host, the only method of natural control
will be to re-introduce the lady=-beetle.

Caribbean Forester OG Vol. 4, No. 2

90.

91.

CATALOGUE DES CRYPTOGAMES VASCULAIRES DES ANTILLES FRANCAISES

H. Stehlé
Ingénieur Agricole et d‘Agronomie Coloniale
Martinique

(Continued from previous number )

14. ADIANTUM (Tourn.) L.

Adiantum macrophyllum Sw. Prodr. 135 (1788). G. Duss n. 4350 et M. Duss n.
174.

Adiantum phyllitidis J. Smith, in Hook. Lond. Journ. Bot. 1:197. (1842).
G. (ex Christensen).

Adiantum obliguum Willd. Sp. Pi. 5:429 (1810). G. Duss n. 4253 et M. Duss
alo JE(BUGe

Adiantum petiolatum Desv. Ges. Naturf. Tround. Berlin Mag. 5:326 (1811),
G. Duss n. 4256 et M. Duss n. 1712.

Adiantum pulverulentum L. Sp. Pl. 1096 (1753). G. Duss n. 4391 et M. Duss
Me 46l3%

Adiantum villosum L. Syst. Nat. ed. 10, 2:1328 (1759). G. Duss n. 4331 et
M.S. n. 3401.

Adiantum latifolium Lamk. Encycl. 1:43 (1783). G. S. n. 644, n. 1092, n.
1196, n. 1796, n. 1914 et M.S. n. 3308. Espéce rapportée
par Urban, Symb. Ant. Fl. Port. Ric. 4:45 (1903) & A.
denticulatum Sw.

- Adiantum tetraphyllum Humb. et Bonpl.; Willd. Sp. Pl. 5:441 (1810), G. Duss

mn. 4527 et M. S. n. 2101, n. 2655 et n. 3265.
Adiantum cristatum L. Syst. Nat. ed. 10, 2:1328. (1759), G. Duss n. 4351
et M. Duss n. 1720.
Adiantum crenatum Willd. Spec. 5:446 (1810), G. (ex Fée) et M. (ex Plumier).

. Adiantum concinnum Humb. et Bonpl.; Willd. Sp. Pl. 5:451 (1810), G. et M.

(introduit).

Adiantum nigrescens Fée, Gen. Fil. 117 (1850-52), G. (ex Urban) et M. Duss
me 1709.

Adiantum Capillus - Veneris L. Sp. Pl. 1096 (1753), G. et M. S. n. 4164.
(introduit).

Adiantum fragile Sw. Prodr. 135 (1788), G. Duss n. 4250, et M. S. n. 2093
et n. 2095.

Adiantum tenerum Sw. Prodr. 135 (1788), G.S. n. 1495, n. 2015 et M. Duss
n. 1722, forma reducta, G.S. n. 1095.

15. DORYOPTERIS J. Smith

Doryopteris pedata (L.) Fée, Gen. Fil. 133 (1852), G. Duss n. 4188 et M.
Dussencmelooc.

16. HYPOLEPIS Bernh.
Hypolepis repens (L.) Presl. Tent. Pter. 162 (1836), G. L'Herminier 1862,

Duss n. 4235 et M. Hahn n. 61, Duss n. 1708. Forme
monstrueuse, M. Duss n. 4679.

- 85.— January 1943

92.
93.

94.

95.

96.

97.
98.
99.
100.

LOL.

102.

103.

104.

105.

17. CHEILANTHES Sw.

Cheilanthes microphylla Sw. Syn. Fil. 127 (1806), G.S. n. 1798.

Cheilanthes elongata Willd. ex Kunze Enum. 213 (1824), G. (ex Urban).
Considéré par Duss, Foug. et Lycop. Ant. fr. 65 comme
var. elongata Willd. de C. microphylla Sw., G. Duss n.
4132 et n. 4231.

18. ADIANTOPSIS Fée

Adiantopsis radiata (L.) Fée, Gen. Fil. 145 (1852), G.S. n. 512, N. 812,
et M. Duss n. 1706.

19. ANISOSORUS Trev.

Anisosorus hirsutus (L.) Underw. et Maxon, in Maxon, Pterid. Puerto Rico,

Sc. Surv. P. R. 6:429 (1926), G.S. n. 87 et M. Hahn n.
17, Duss n. 1545.

20. HISTIOPTERIS (Ag.) J. Smith

Histiopteris incisa (Thumb.) J. Smith, Hist. Fil. 295 (1875), G. Husnot n.
292, S. n. 646 et n. 2425.

21. PTERIS (Tourn.) L.

Pteris longifolia L. Sp. Pi. 1074 (1753), G. ex Maxon et M. Duss n. 4128.

Pteris grandifolia L. Sp. Pl. 1073. (1753), G.S. n. 1466.

Pteris arborea L. Sp. Pl. I ed. 2, 1073 (1753), M. (ex Urban).

Pteris pungens Willd. Sp. Pl. 5:387. (1810). G. Duss n. 4192 et M. et Duss
n. 1539. Espéce considérée parfois comme une simple
variété de P. biauritum L., var. pungens (Willd. Spec.
Pl. 5:380 (1810), M. (Plumier, Descr. 13:9 et Fil. 13:11.

Pteris quadriaurita Retz. Obs. Bot. 6:38 (1791). G.S. n. 2438 et M. (ex
Maxon). Espéce de synonymie abondante et de relation
confuse; elle a été rapportée par Kuhn, in Engi. Bot.
Jahrb. 24:99 (1897) a P. biaurita L. dans une var.
repandula (Link) Kuhn, adoptée par Urban, in Symb. Ant.
Fl. Port. 4:48 (1903). P. repandula Link., Fil. Hort.
Borol. 56 (1841) est un synonyme de P. quadriaurita Retz.

Pteris biaurita L. Sp. Pl. 1076. (1753), G. Duss n. 4193 et M. Duss n.
4585.

Pteris crassipes Agardh, Recens. Pterid. 59 (1839), M. (ex Maxon). Lito-
brochia Agardhii Fée, Foug. Ant. 25 (1866), de Guadeloupe
et Martinique indiquée par Urban comme species dubia est
probablement synonyme de cette espéce.

Pteris altissima Poir. in Lamk. Encycl. 5:722. (1804), G. et M. (ex Maxon).
Pteris Kunzeana Agardh, Roc. Pter. 62 (1839) citée pour
Guadeloupe par Urban n'est pas spécifiquement distinct
de P. altissima Poir.

Pteris vittata L. Sp. Pi. (1753), GS. n. 104, n. 645 et M.S. n. 4162.

Caribbean Forester - 84 = Vol. 4, No. 2

106.

107.

108.

109.

110.

LTT

112.

113.

119.

120.

Pteris multifida Poir. Encycl. 5:714 (1804), G.S. n. 2616. Espece asiatique
naturalisée, citée également par Jenman, Ferns Br. West Ind.
and Guy. 119 (1909) pour la Martinique.

22. HEMIONITIS L.
Hemionitis palmata L. Sp. Pl. 1077. (1753), G.-S. n. 88 et M.S. n. 4163.
23. PITYROGRAMMA Link.

Pityrogramma sulphurea (Sw.) Maxon, Contr. U. S. Nat. Herb. 17:173 (1913),
G. Duss n. 4376.

Pityrogramma chrysophylla (Sw.) Link, Handb. Gewichs. 3:19 (1833), G.S. n.
248 et M.S. n. 2281. Var. Plumieriana Domin, Gen. Pityr.
Clav. 6 (1928), a laquelle appartient Ceropteris Herminieri
Link. est répandue aux Antilles frangaises.

Pityrogramma calomelanos (L.) Link, Handb. Gewdchs 3:20 (1833), G.S. n. 92,
n. 105, n. 352, n. 2437 et M.S. n. 4165. Var. distans
(Link.) Urban, Symb. Ant.Pterid. doming. 9:339 (1925), G.
(ex Urban).

Pityrogramma tartarea (Cav.) Maxon, Contr. U.S. Nat. Herb. 17:173 (1913),
G. Duss n. 4173 et M. Duss n. 1507.

Pityrogramma Schaffneri (Fée) Weatherby in Contr. Gray Herb. 114:25 (1936),
G, et M. (ex Maxon).

Pityrogramma caribaea Domin, Gen. Pityr. Clav. 6, 7 (1928) existe également
aux Antilles frangaises (ex Domin) comme ® la Dominique:
G. Mazé et M. Sieber n. 237.

24. DIPLAZIUM Sw.

Diplazium plantaginifolium (L.) Urban, Symb. Ant. Fl. Porto R. 4:31 (1903),
G.S. no. 1217, no. 1451 et M. Duss n. 1640.

Diplazium flavescens (Mett.) Christ, Farnkr. 217. (1897), G. (ex Urban).

Diplazium centripetale (Baker) Maxon, Pterid. Puerto Rico, Sc. Surv. P. R.
6:441 (1926), G. ex Fée et M. Duss n. 1646. Espece
homologue et souvent référée a D. coltidifolium Kunze,
localisée au Nord d l’Amérique du Sud.

Diplazium arboreum (Willd.) Presl, Tent. Pterid. 114 (1836), G. Duss n.
4207 et M. Duss n. 4690.

Diplazium striatum (L.) Presi. Tent. Pterid. 114 (1836), G.S. n. 528, no
828, n. 1190, n. 1452, n. 2425 et M. Duss n. L631. Var.
truncatum (Presl.) Hieron. in Hedwigia 59:328 (1917), G.
et M. (ex Urban). D&éprés C. Christ. ne différe pas du type.

Diplazium L’Herminieri Hieron. in Urban, Symb. Ant. Fl. Porter. Add. et
Corr. 4:648 (1911), G. (ex Hieronymus).

Diplazium Shepherdii (Spreng.) Link. Hort. Borol. 2:70 (1833). Cité pour
la Guadeloupe et la Martinique par Urban, Symb. Ant. Fl.
Porter. 4:32 (1903) et Pterid. doming. 9:325 (1925);
considéré par Presl, Grisebach, Hooker et Baker ainsi que
par Duss, G. n. 4218 et M. n. 1645, comme une espece
distincte de D. arboreum (Willd.) Presl, mais par

= 85 = January 1943

Hieronymus et par Maxon, Pterid. Porto Rico, Sc. Surv.
P.R. 6:442 (1926), comme synonyme de cette derniére.

121. Diplazium trinitense Domin, Pterid. Dominica, 155 (1929), G. Mazé n. 420
623 (1881-32), est intermédiaire entre D. arboreum (Willd. )
Presl et D. caracasanum Kunze.

25. ATHYRIUM Roth.

122. Athyrium conchatum Fée, Gen. Fil. 186 (1850-52). Cité par Kuhn pour la
Guadeloupe ex Urban, Symb. Ant. 4, 31, parait 6tre une
des espéces énumérées de Diplazium, probablement D.
l'Herminieri Hieron.

26. HEMIDICTYUM Presi.

123. Hemidictyum marginatum (L.) Presl. Tent. Pterid. III, pl. 3, f. 24 (1836),
G.S. no 93, ne 1776 et MoS. n. 3439.

27. ASPLENIUM L.

124. Asplenium serratum L. Sp. Pl. 1079 (1753), G. Duss n. 4203 et M. Duss n.
1641.

125. Asplenium dentatum L. Sp. Pl. 1080 (1753), G. Duss n. 4223.

126. Asplenium formosum Willd. Sp. Pl. 5:329 (1810), G. Duss n. 4202 et M.S.

n. 2092.

127. Asplenium pumilum Sw. Prodr. 129 (1788), G. Duss n. 4201 et M. Duss n.
1639.

128. Asplenium cirrhatum Rich. ex Willd. Spec. Pl. 5:321 (1810), G. Duss n.
4002.

129. Asplenium Serra Langsd. et Fisch. Icon. Fil. 16, pl. 19 (1810), G.S. n.
90, n. 1440 et n. 1460.

130. Asplenium lactum Sw. Syn. Fil. 79, 271 (1806), G. Duss n. 4199 et M. Duss
n. 1635. Esptce affine de A. resectum Smith, a laquelle
Duss l‘a rapportée.

131. Asplenium obtusifolium L. Sp. Pl. 1080 (1753), G. Duss n. 4205 et M. Duss
n. 1636 b. Les plantes des Antilles frangaises se classent
dans la var. aquaticum basée sur A. aquaticum Kl. et Karst.
de Colombie.

132. Asplenium pseudoerectum Hieron. Hedwigia 60:239 (1918), G.S. ne 1224, no
1439 et M. Duss no. 4601. Les échantillons de Guadeloupe
et de Martinique se rapportant a cette espece ont souvent
été identifies avec A. pteropus Kaulf. Enum. Fil. 170
(1824), espece localisée au continent Sud-Américain, en
particulier par Urban, Symb. Ant. 4, 34 (1903) et 9:328
(1925), Krug et Duss l’avaient considéreé comme la var.
pterotus de 1'A. lunulatum Sw., espece continentale.

133. Asplenium abscissum Willd. Sp. Pi. 5:321 (1810), G.S. n. 1461 a et M. Duss
no 1637 be

134. Asplenium salicifolium L. Sp. Pl. 1080 (1753), G.S. n. 245, n. 325; n. 334,
no. $59, n. 5356, n. 836, n. 1193, n. 1461 b, et M. Duss n.
1648, Hahn n. 21.

135. Asplenium harpeodes Kunze, in Linnaea 18:329 (1844), G. Duss n. 4214 et M.
Duss n. 1647.

Caribbean Forester - 86 = Vol. 4, No. 2

141.
142.

143.

144.

145.

146.

147.

Asplenium auritum Sw. Journ. Bot. Schrad. 1800:52 (1801), G.S. no 349 et
M.S. n. 3267, n. 3300 et n. 3324, var. rigidum (Sw.) Hook.
Spee File o (1860), G.S; n. 889, n. 1227 jet! n-. 2250.

Asplenium fragrans Sw. Prodr. Fil. 130 (i788), G. Duss n. 4382 et M. (ex
Duss).

Asplenium radicans L. Syst. Nat. ed. 10, 2:1323 (1759), G.S. n. 1462.
Espece parfois confondue avec A. flabellatum Kunze in
Linnaea 9, 71 (1835).

Asplenium cristatum Lamk. Encycl. 2:310 (1786), G. Duss n. 4334 et M.S.
n. 2094. Espéce souvent désignée par son synonyme A.
cicutarium Sw. Prodr. 130 (1788).

Asplenium cuneatum Lamk. Encycl. 2:309 (1786)., G. Duss n. 4222 et M. Duss
n. 1649.

Asplenium auriculatum Sw. in Vet. Acad. Handl. Stockh. 38:68 (1817).

Indiqué pour la Guadeloupe et la Martinique par Urban, Symb. Ant.
4:35 (1903), ainsi que A. Samanae Brause, RKepert. Nov. Sp. Fedde, 18:247
(1922), ne paraissent que des variations de A. salicifolium L., espéce
trés polymorphe. Les petites formes de cette derniére espeéce rencontrées
aux Antilles frangaises et & Puerto Rico devront 6tre comparées, d‘aprés
Maxon, Pterid. Puerto Rico, Sc. Surv. P. R. 6, 450, avec A. hastatum Kl.
et A. semicordatum Sw. D'apres Domin, Pterid. Dominica, 164 (1929), cette
comparaison permettrait de conclure que les seuls spécimens de Mazé: n.
246, 317, 734, 781, de 1882,-85, de Guadeloupe représentent distinctement

A. semicordatum Raddi, du Brésil et sont identiques & A. pimpinellifolium

Fée et Schaffn.
28. STRUTHIOPTERIS Scop.

Struthiopteris polypodioides (Sw.) Trev. Atti Inst. Ven. 3. 14:571 (1869),
M.S. n. 2106.

Struthiopteris exaltata (Fée) Broadhurst, Bull. Torrey Bot. Club, 39:264
(1912). G. L'Herminier 1863, Mazé n. 276 et M. Hahn n. 76.

Struthiopteris Plumieri (Desv.) Broadhurst, Bull. Torrey Bot. Club, 39
(1912), G. Mazé n. 861, Duss n. 4169 et M. S. n. 3312.
Espece basée sur Lomaria Flumieri Desv., non Hook et Bak.
et souvent confondue avec L. divergens Kunze. La désig-
nation de Blechum divergens (Kunze) Mett., pour les
Petites antilles, en particulier, in Urb. Symb. Ant. 4:40,
19 (1903) et 9:336 (1925), parait s’appliquer & cette
espece.

Struthiopteris L'Herminieri (Bory) Broadhurst, Bull. Torrey Bot. Club,
39:267 (1912) G. L’Herminier n. 99 ter, Mazé n. 95, n.
100, Sen. 1214 et M. (ex Broadhurst).

Struthiopteris violacea (Fée) Broadhurst, Bull. Torrey Bot. Club, 39:379
(1912), G. L'Herminier 1862, Mazé n. 86 et M. (ex Broad.)

Struthiopteris striata (Sw.) Broadhurst, Bulli. Torrey Bot. Club, 39:375
(1912), G. L'Herminier 1862, S. n. 1116 et M.S. n. 4163.

29. BLECHNUM L.
Blechnum indicum Burn. Fl. Ind. 231 (1768). G. Duss n. 4379 et M. Duss n.
1901. Espece souvent désignée par son synonyme Bl. serru-

Tatum B. Cis Rich. (1792).

a pes January 1943

148. Blechnum occidentale L. Sp. Pl. 1077 (1753), G.S. n. 345, n. 1213, n. 1913,
no 1920, n..2417 et M.S. n. 5265.

149. Blechnum unilaterale Sw. Ges. Naturf. Freund. Berlin Mag. 4:79, pl. 3, f.
I (1810). M. Duss n. 1702.

150. Blechnum fraxineum Willd. Spec. Pl. 5:413 (1810), M. Duss n. 1703. Espece
assez affine de Blechnum occidentale L.; citée par Domin,
Pterid. Dominica, 186 (1929) pour la Guadeloupe (ex Duss)
sous le synonyme de B. longifolium H.B.K., cela par erreur
car Duss, Foug. Ant. fr. 88 (1903) précise bien qu'il ne
l'a pas trouvée dans cette Ile.

50. SALPICHLAENA J. Smith

151. Salpichlaena volubilis (Kaulf.) J. Smith, Journ. Bot. Hook, 4:168 (1841),
G.S. n. 640, n. 1215 et M. Duss n. 1700.

$1. STENOCHLAENA J. Smith
152. Stenochlaena sorbifolia (L.) J. Smith, Journ. Bot. Hook. 4:149 (1841),
G.S. ns 518, ne. 51k9, no 5325 m.. 818,on. elo misSeeens
1447 et M. Duss n. 1611.
52. POLYBOTRYA Humb. et Bonpl.

153. Polybotrya cervina (L.) Kaulf. Enum. Fil. 55 (1824), G. Duss n. 4140 et
MoS. no. 5437.

33. NEUROCALLIS Fée

154. Neurocallis praestantissima (Bory) Fée, Hist. Acrost., 89, t. 52 et Hist.
Foug. et Lyc. Ant. 8 (1866),. G.S..m. 520) u.820i et uM.
Hahn 1868 s.n., Duss n. 1510.

54. LEPTOCHILUS Kaulf.

155. Leptochilus nitianaefolius (Sw.) C. Christ. Bot. Tidsskr. 26:285 (1904) G.
Duss n. 4145 et M. Duss n. 1699. Espéce désignée récemment
par Alston sous le nom de Bolbitis nicotianaefolia (Sw.)
Alston.

156. Leptochilus cladorrhizans (Spreng) Maxon, Pterid. Puerto Rico, Se. Surv.

P. KR., 6:460 (1926), G. et M. (ex Maxon).

157. Leptochilus alienus (Sw.) C. Christ. in Bot. Tidsskr. 26:285 (1904), G. Duss

n. 4144 et M. Duss n. 1705.

35. CYCLOPELTIS J. Smith

158. Cyclopeltis semicordata (Sw.) J. Smith in Curtis, Bot. Mag. 72:36 (1846),
G. Duss n. 4358.

36. POLYSTICHUM Roth.

159. Polystichum rhizophyllum (Sw.) Presl. Tent. Pterid. 82 (1836), G. (ex Maxon).

Caribbean Forester - 88 = - Vol. 4, No. 2

160.

161.

162.

163.

164.

165.

166.

L167.

L168.

169.

170.

ne

172.

173.

Polystichum andiantiforme (Forst.) J. Smith, Hist. Fil. 220 (1875).
Signalé par Urban: Symb. Ant. 4, 25 (1903), pour la
Guadeloupe d'aprés Fée, dans une variéteé.

57. DRYOPTERIS Adams.

Dryopteris opposita (Vahl) Urban, Symb. Ant. Fl. Portor. 4:14 (1903), G.S.
n- 24352 et M.S. n. 3418.

Dryopteris panamensis (Presl.) C. Christ. in Viol. Seisk. Skr. 7:4, 292,
Pig. 19 (1907) est cité par Urban (1925) pour la Guade-
loupe; trés voisin de D. opposita (Vahl) Urban.

Dryopteris sancta (L.) Kuntze, Rev. Gen. Pl. 2:813 (1891), G.S. n. 467,

n. 1446, n. 1762, n- 1862, n. 2065 et M. (ex Maxon).
oasis concinna (Willd.) Kuntze, Rev. Gen. Pl. 2:812 (1891), G. (ex
C= Christ).

Deopeeris consanguinea (Fée) C. Christ. Ind. Fil. G. Duss n. 4126 et M.
Duss n. 1584. Le type de l'Aspidium consanguineum Fée
(1866), base de l’espéce est de Guadeloupe.

Dryopteris Linkiana (Presl) Maxon, Journ. Wash. Acad. Sci. 14:199 (1924),
G.S. n. 1453, n. 2430, n. 2433 et M. Kohaut in Sieber,
Herb. Mart. Suppl. n. 65. Parmi les nombreux synonymes
de cette espéce, il convient de remarquer: Phegopteris
Duchassaingiana Fée (1866) et D. diplazioides (Desv.)
Urban (1903) non Kuntze (1891). .

Dryopteris gracilis (Hedw.) Domin, Pterid. Dominica, 210 (1929), G. Husnot
n. 391. Syn.: D. consimilis (Fée) C. Christ.

Dryopteris Germaniana (Fée) C. Christ. Ind. Fil. 267 (1905), G. Duss n.
4067 et li. Duss n. 1560. Le type de cette espeéce est un
échantiilon de Germain, collaborateur d l*Herminier, de
ja Guadeloupe.

Dryopteris Sprengelii (Kaulf.) Kuntze, Rev. Gen. Pl. 2:813 (1891), G.S.

n. 1444 et M.S. n. 1039, n. 3276, n. 3277, n- 3280 et
n. 3286. Espéce rapportée & Dryopteris Balbisii Urban,
non Polypodium Balbisii Spreng. L'échantillon type est
de la Martinique.

Dryopteris limbata (Sw.) Kuntze, Rev. Gen. 2:813 (1891), G. (ex Domin),

M. (ex Krug).

Dryopteris decussate (L.) Urban, Symb. Ant. 4:19 (1903), G.S. n. 514 et
M.S. n. 3293, no. 3435 et n. 3444.

Dryopteris gongylodes (Schkuhr ) Kuntze, Rev. Gen. Pl. 2:811 (1891), G.S.
n. $25, n. 639, n. 643 et M. Duss n. 1587.

Dryopteris dentata (Forsk) C. Christ. Dansk. Vidensk. Skrift. 8. Naturv.
Afd. 6:24 (1920), G.S. n. 2431, n. 2435, n. 2436 et M.S.
n. 35272, no 5502, n. 3505 et n. 33507.

Espéce souvent désignée par D. parasitica Kuntze (1891),
synonyme invalidé, non Polypodium parasiticum L. (1753).
Récemment dénommée par Alston: D. quadrangularis (Fée)
Alston, in Pterid. St. Kitts, Journ. Bot. 75:253 (1937).
W. H. Hodge, in Notes of Dominica Ferns, Amer. Fern.
Journ. 31:3-4, 121 (1941) a démontré l'invalidité de ce
binome et la grande variabilité de D. dentata (Forsk.)
C. Christ.

- 8S = January 1943

174. Dryopteris patens (Sw.) Kuntze, Rev. Gen. Pl. 2:813 (1891), G.S. n. 100
et M. Duss n. 4583.

175. Dryopteris Sloanei (Bak.) Kuntze, Rev. Gen. Pi. 2:813 (1891), G. Duss
n. 4061.

176. Dryopteris normalis C. Christ. Ark. fdr Bot. 9:31 (1910), M.S. n. 2254.

177. Dryopteris Poiteana (Bory) Urban, Symb. Ant. Fl. Portor. 4:20 (1923), G.
S. n. 209 et M. Buss n. 4118. Var. Rivoirei (Fée) Domin
de cette espece (Goniopteris Rivoirei Fée, 1850), G.
Mazé n. 1065, M.Mme Richard-Rivoire n. 164.

178. Dryopteris subtetragona (Link) Maxon, Pterid. Puerto Rico, Sc. Surv. P. R.
62473 (1926) ,°Go.S:-ner64l: ne 2196.) no) 2529et ul. Sean.
2096 et n. 3270. Crest l’espéce désignée par Urban, Symb.
Ant. Fl. Portor. 4:20 (1903) par D. tetragena (Sw.) Urban
(1903), non Kuntze (1891) et qui est basée sur Polypodium
tetragonum Sw. (1788), non Nephrodium tetragonum Presl

(1825). Var. guadalupensis nov. (Goniopteris guadalupensis
Fée) existe & la Martinique et & la Guadeloupe.

179. Dryopteris domingensis (Spreng.) Maxon, Pterid. Puerto Rico, Sc. Surv. P.R.
§:474 (1926), G. Duss n. 4059 et M. Duss n. 4389. Espece
souvent désignée par son synonyme: D. scolependrioides
C. Christ. (1905), non Kuntze (i891).

180. Dryopteris reptans (Gmel.) C. Christ. Ind. Fil. 288. (1905), G. 1l'Herminier
1862, in herb. Fée n. 128. L’espece de Guadeloupe se
refére & la var. tenera C. Christ. in Vid. Selsk. Skr.
TelO, 219° (1925).

181. Dryopteris guadalupensis (Fée) Kuntze, Kev. Gen. Pl. 2:812. (1891), G. S.
n. 1454 et M.S. n. 3323. Espece désignée aussi sous le
nom de D. nephrodioides Kiotzsch. forma guadalupensis
(Pée) C. Christ., basée sur Nephrodium guadalupense Fée,
Mém. Foug. I1:89, pli. 24, f.3 (1866).

182. Dryopteris hastata (Fée) Urban, Symb. Ant., Fi. Portor. 4:21 (1903), G.
Duss no. 4115 et M. Duss n. 4146.

183. Dryopteris megalodus (Schkuhr) Urban, Symb. Ant., Fl. Portor. 4:21 (1903)
G. Duss n. 4110 et M. Duss n. 1585.

184. Dryopteris leptocladia (Fée) Maxon, Pterid. Puerto Rico, Sc. Surv. P. R.
6, 476 (1926), G. Duss n. 4116 et M. Duss n. 4117.

185. Dryopteris pyramidata (Fée) Maxon, Contr. U.S. Nat. Herb. 10:489 (1908),
G. Duss n. 4115 et M. (ex Maxon).

186. Dryopteris serrata (Cav.) C. Christ. Ind. Fii. 291. (1905), G. Duss n.
1612.

187. Dryopteris reticulata (L.) Urban, Symb. Ant. Fi. Portor. 4:22 (1903), G.S.
n. 511, no “526, no 654, n. 8H, om. 826, nm. LOG. ne 2 O9Ge
no 1876, n. 1919 et M.S. n. 5264 et n. 3315.

188. Dryopteris meridionalis (Poir.) C. Christ. in Vid. Selsk. Skr. 8:46 (1920),
G. (type ex Christ.) et M. (ex Christ.)

189. Dryopteris excelsa (Desv.) C. Christ. Ind. Fil. 264 (1905) et in Vid.
Selsk. Skr. 8:54 (1920), G. et M. (ex Christensen).

190. Dryopteris subincisa (Willd.) Urban, Symb. Ant. Fl. Portor. 4:19 (1903);
G. et M. S. no 5325. L'Alsophila martinicensis Spreng.
Neu. Entd. 3:7 (1822), n’est autre que cette espéce,
ainsi que le Phegopteris martinicensis Fourn. Mex. Pl.
5:90. (1872).

Caribbean Forester = 90 = Vol. 4, Noo 2

191.

192.

193.

194.

195.
196.

197.

198.

199.

200.

201.

202.

203.

204.

205.

206.

Dryopteris macrotheca (Fée) C. Christ. Ind. Suppl. 35 (1913), G. (ex
Urban) .

Dryopteris effusa Ae Urban, Symb. Ant. Fl. Portor. 4:16 (1903), G. et

- (ex Urban).

Dryopteris Pilate (Afz.) C. Christ. var. funesta (Kunze) C. Christ.
Ind. 286 (1905}, G. et M. (ex Urban).

Dryopteris 1l’Herminieri (Kunze) C. Christ. Ind. Fil. 275 (1905). G.S. ne
1220, n- 2109 et H.S. n. 3402, n. 3408 et n. 5420.

Le type de cette fougére est de la Guadeloupe. Elle a
eté rapportée par Enis auteurs et par Urban, Symb.
Ant. 4:13 (1903) & D. trichophora (Fée) Kuntze, &
considérer comme Deere

Dryopteris extensa (Bl.) Kuntze, loc. cit., M.S. n. 3260.

Dryopteris pubescens (L.) Kuntze, Rev. Gen. Pl. 2:813. (1891). Var. sericea
(Mett.} Urban, Symb. Ant. Fl. Portor. 4:15 (1903), M. Duss
n. 4690. Var. muscosa (Vahi) Urban, Symb. Ant. Fl. Portor.
4:16 (1903), G. et M.S. n. 3282.

Dryopteris retundata (Willd.) ©. Christ. Ind. Fil. 289 (1905), GS. n.
1104, n. 1463, n. 1887 et M. Hahn n. 37, S. n. 3303 et
n. $436.

ee glandulosa (Desv.) C. Christ. Mon. Dryopt. 1:171 (1913), non
Kuntze, G.S. n. 530, no. 830, n. 1200, n. 1442, n. 1455
et M.S. no. 2105, n. 3417 et n. 3438.

Dryopteris mollicella Maxon in Proc. Biol. Soc. Wash. 36:49 (1923). G. S.
n. 1443, n. 1857, n. 2427 et n. 2434. Espéce considérée
jusqu'a présent comme endémique de la Dominique (Cft.
Domin, Pterid. Dominica, 199 (1929).

Dryopteris rustica (Fée) C. Christ. Ind. Fil. 290 (1905). G. S. n. 1759,
n. 2426, n. 2428, n. 2434 a, et M. Duss n. 4580. D. domi-
nicensis C. Christ doit étre considéré comme synonyme de
cette espéce (W. H. Hodge) Notes on Dominica Ferns, in
Amer. Fern. Journ. 31, 3-4, 122 (1941).

Dryopteris hydrophila (Fée) C. Christ. Ind. Fil. 271 (1905). G. S. n.
1445 et M. S. n. 2429. Espeéce rare basée sur Phegopteris
hydrophila Fée (1866) de Guadeloupe.

Dryopteris deitoidea (Sw.) Kuntze, Rev. Gen. Pl. 2:812 (1891), G. (ex
Hooker).

Dryopteris asplenioides (Sw.) Kuntze, Rev. Gen. Pi. 2:812 (1891). Var.
subpinnata (Mett.) Urban, Symb. Ant. Fl. Portor. 4:18
(1903), G. et M. Forma exindusiata (Kuhn) Urban, Symb.
Ant. Fl. Portor. 4:18 (1903), G. var. pinnata (Mett.)
Urban, Symb. Ant. Fl. Portor. 4:18 (1903), G. (ex Urban).

58. TECTARIA Cav.

Tectaria plantazinea (Jacq.) Maxon, Contrib. U.S. Nat. Herb. 10:494 (1908),
G.S. n- 85, n. 1218 et M.S. n. 3304.
Tectaria heracleifolia (Willd.) Underw. Bull. Torrey Bot. Club. 33:200
(1906), G. Duss n. 4049 a et M. Duss n. 4683.
Tectaria martinicensis (Spreng.) Copel. Philippine Journ. Sci. Bot. 2:410
1907) G. et M: S:'n.° 2253, n. 3306 et n. 3400.

= Sf 2 January 1943

207.

208.

209.

210.

Pi

212.

213.

214.

215.

216.

BLT.

Tectaria trifoliata (L.) Cav. Descr. Pl. 249 (1802), G.S. n. 86, n. 1060,
no 1197 et M.S. n. 3301. L'’Aspidium Plumierii Presl,
cité par Urban, Symb. Ant. Fl. Portor. 4:23 est
Tectaria Plumierii Copel. Philippine Journ. Sci. Bot.
2:410 (1907), espece considérée par Urban comme
distincte d'Aspidium trifoliatum (L.) Sw. = T. trifo-
liata (L.) Cav., mais placée par Maxon, Pterid. Puerto
Rico, Sc. Surv. P. R. 6:482 (1926), comme synonyme
invalidé de cette derniére.

59. OLEANDRA Cav.

Qleandra articulata (Sw.) Presl. Tent. Pterid. 78 (1836), G. L*Herminier
no 162, Duss n. 4161 et M. Duss n. 4587.

40. NEPHROLEPIS Schott.

Nephrolepis cordifolia (L.) Presl. Tent. Pterid. 79 (1836), M.S. n. 3396.
Espeéce nouvelle pour les Antilles frangaises.
Nephrolepis exaltata (L.) Schott, Gen. Fil. pl. 3 (1834), G. et M. (ex
Urban).
Nephrolepis rivularis (Vahl) Mett.; King, Bot. Jahrb, Engler 24:122 (1897),
GS. no 101, n. $58, n. 1455 et M.S. n. 3298 et n. 33530.
Nephrolepis biserrata (Sw.) Schott, Gen. Fil. pl. 3 (1834), G.S. n. 504,
n. 2059 et M.S. n. 83271, n. 3278 et n. 3851.
Nephrolepis Duffii Hort., GS. n. 344, et M. S. n. 4171.
Cultivée et subspontanée; considérée comme une forme
ornementale de N. cordifolia (L.) Presl, par Bailey,
Cyclop. 2:2132.

41. ODONTOSORIA (Presl) Fée

Odontosoria flexuosa (Spreng.) Maxon, Stud. trop. amer. ferns, in Contr.
Us S. Nat. Mus. 17), 23165, (1915)),) G: Duss n-, 4238vet
M. Sieber n. 23 (type) et Duss n. 4238. Ctest 1!
espece rapportée par Urban, Symb. Ant. 4:29 (1903) a
0. aculeata (L.) Smith, qui est localisée aux Grandes
Antilles et nettement distincte.

42. LINDSAEA Dryand.

Lindsaea montana Fée, Mém. Foug. I1:17, pl. 6, fig. 2 (1866), G.S. n. 521,
n. 821, no IZ19et n. 2419)
Indiqué par Urban, Symb. Ant. 4:31 (1903) sous le nom
de L. quadrangularis Raddi, & réserver pour les fougéres
voisines du continent américain.

Lindsaea lancea (L.) Bedd. Ferns Br. Ind. Suppl. 6 (1876), G. Duss n. 4240,
n. 4241 et M. Duss n. 1707. Syn. Lindsaea trapeziformis
Dry. (1797).

Lindsaea L'Herminieri Fée, Hist. Foug. et Lycop. Ant. 15, t. 6, fig. 1 (1866),
G. L'Herminier 1862, A. Fée n. 43, Duss n. 4237.

Caribbean Forester = 92 = Vol. 4, No. 2

218.

219.

220.

221.

222.

2235.

224.

225.6

43. SACCOLOMA Kaulf.

Saccoloma _domingense (Spreng.) Prantl, Arb. Bot. Gart. Breslau 1:21 (1892),
G.S. n.- 1216 et M. (ex Maxon).

Saccoloma inaequale (Kunze) Mett. Ann. Sci. Nat. 4, Bot. 15:80 (1861), G.S.
n. 1465 Var. Galeottii (Fée} Hieron. in Urban, Symb. Ant.
Pterid. Doming. 9:318 (1925), G. (ex Fée) et M. (ex
Urban).

Saccoloma Imrayana Kunze, Hook. Gen. Fil. t. 58B, nom. nud. (1839), Kunze,
Farrnkr. 1:86 (1842) et Mett. Ann. Sci. Nat. 4 (1861),
G. Duss n. 4176.

44. DENNSTAEDTIA Bernh.

Dennstaedtia ordinata (Kaulf.) Moore, Ind. Fil. 306 (1857), G. Duss n.
4179 et M.S. n. 2102. Espéce désignée par Urban, Symb.
Ant. 4:30 sous le nom de D. dissecta (Sw.) Moore, basé
sur Dicksonia dissecta Sw. appliqué primitivement 4 des
espéces jamaiquaises et synonyme de D. cornuta (Kaulf.)
Moore, du continent Centre et Sud-américain.

Dennstaedtia Plumieri Hook. Spec. Fil. 1, 72 (1844). G. Duss n. 4342 et
M. Duss n. 1680.

Dennstaedtia adiantoides (Humb. et Bonpl.)} Moore, Ind. Fil. 97 (1857), G.
Duss n. 4841.

Dennstaedtia cicutaria (Sw.) Moore, Ind. Fil. 97 (1857), G. Duss n. 4184
et M. Duss n. 1686. Espece souvent confondue avec
D. ordinata (Kaulf.) Moore.

Dennstaedtia incisa (Fée) Kuhn in Limnaea 34:146 (1869), G.S. n. 1464 et
M. Duss n. 1685. Espéce basée sur Dicksonia incisa
(Fée), Hist. Foug. et Llycop., Ant. 14, t. 25, f. I.
(1866) sur une plante de Guadeloupe: L’Herminier 1862,
A. Fée n. 172 et considérée par Baker et par Duss comme
une simple varieté de D. cicutaria (Sw.) Moore.

Famille 5. HYMENOPHYLLACEAE

lL. TRICHOMANES L.

1. Trichomanes sphenoides Kunze, Linnaea 9:102 (1834); Farrnkr. 1:215, pl. 88

(1846) G. Duss n. 4419 et M. Duss n. 4704.

Trichomanes punctatum Poir. in Lamk. Encycl. 8:64 (1808), G.S. n. 659 et M.
Duss n. 4689.

Trichomanes pusillum Sw. Prodr. 136 (1788); Fl. Ind. Occ. 3:1729 (1806), G.
Duss n. 4420 et M. Duss n. 4160, n. 4578 et n. 4609.

Trichomanes Kraussii Hook et Grev. Icon. Fil. 2: pi. 149 (1831). G.S. n.
503 et M. Duss n. 1560.

Trichomanes membranaceum L. Sp. Pl. 1097 (1753), GS. n. 513, n. 515, n.
o22, no 815, n- 815, n. 822 et M.S. n. 3427.

Trichomanes Hookeri Presl, Abh. Bohm. Ges. Wiss. 5:3:108 et Hymen, 16.
(1843), G.S. n. 662 et M. Duss n. 1636. Souvent désigné
sous le binédme de T. muscoides Hook et Grev. (1830), non

- 93 = January 1943

Sw. (1801). La var. angustifrons (Fée) Duss (et
non angustifolium Fée comme |e rapporte par erreur
Domin, Pterid. Dominica, p. 47) est synonyme de

Tf. pusillum Sw.

7. Trichomanes pinnatum Hedw. Fil. Gen. Sp. pl. 4. f. I. (1799), G. Duss n.
4287 et M. Duss n. 1528.

Trichomanes Kaulfussii Hook. et Grev. Icon. t. 10 (1830) cité par Duss,
Foug. Ant. fr. 16 (1903), comme espéce différente de
T. pinnatum Hedw. est synonyme de cette derniére.

8. Trichomanes polypodioides L. Sp. Pl. 1098 (1753), G. Duss n. 4286 et M.
Duss n. 1524. Espéce souvent désignée par son
synonyme I. sinuosum L. Cl. Rich. in Willd. Sp. Pl.
5.1:502 (1810), en particulier par Lamk. Illustr. t.
871, f.I; Griseb. Flor. 654; Hook. et Grev. Icon. t.
13; Hook. et Bak. Syn. Fil. 2 ed. 78; Duss, Foug. et
Lycop. Ant. fr. 16 (1903) et Urban, Symb. Ant. 4:6
(1903).

9. Trichomanes holopterum Kunze, Farrnkr. 1:185, pl. 77, f.2 (1845), G. Duss
n. 4275 et M. (ex Maxon). Rapporté par Duss sous le
synonyme de T. Bancroftii. Le T. L'Herminieri Fée,
Hist. Foug. et Lye. ant. 107, te29urien be iees)west
la var. L'Herminieri (Fée) Domin, Pterid. Dominica, 49.

' (1929) de T. holopterum, G. (ex Fée).
10. Trichomanes accedens Presl, Epim. 14 (1849), G. (ex Van den Bosch), et M.
(ex Fée). Var. procerum (Fée) Domin, Pterid. Dominica
53 (1929), G. L*Herminier.
ll. TIrichomanes crispum L. Sp. Pl. 1097 (1753), G.S. n. 1459 et M. Duss n. 1522.
Forme remotum Fée, Hist. Foug. et Lycop. Ant. 105 (1866),
Gs -L'Herminier n./192. 1867.
12. Trichomanes pellucens Kunze in Linnaea 9:104 (1834), M. (ex Urban). Syn.:
T. fastigiatum Sieber (1843).
13. Trichomanes alatum Sw. Journ. Bot. Schrad. 1800: 2, 97 (1801) et Fl. Ind.
oce. 3:1752., ClS06)> (GuScns (90 in. 3540 n. 5554) momsoor
n. 1775, n. 2051, n. 2067 et M.S. n. $320 et n. 3415.
Domin, Pterid. Dominica, 51 (1929) distingue plusieurs
variétés dont var. ptilodes (V. d.B.) Dom., M. et var.
attenuatum (Hook) Dom. G.
14. Trichomanes crinitum Sw. Prodr. 136 (1788) et Flor. Ind. Occ. 3:1730, G.
Duss no 4293.
15. Trichomanes rigidum Sw. Prodr. 137 (1788), GS. no 1422 a, n. 1779 et M.
Duss 1529.
Trichomanes Krugii Christ. Bot. Jahrb. Engler 24:90 (1897), décrit dans
Duss Foug. Ant. fr. 14 (1903) pour la Martinique,
Duss n. 4677, comme espeéce distincte de TI. rigidum Sw.
n'en est pas spécifiquement distinct.
16. Trichomanes radicans Sw. Journ. Bot. Schrad. 1800:97 (1801), et Fl. Ind.
Occ. 3:1736 (1806), G. Duss n. 4289 et M. Duss n. 1537.
17. Trichomanes hymenophylloides v.d.B. Nederl. Kruidk. Arch. 5:209 (1863).
G.S. n. 1421 d et M.S. n. 2103.
Confondu par Hook. et Grev. Icon. Fil. 2spl. 206 (1831)
par Duss, Foug. Ant. fr. 13 (1903) et par Urban, Symb.
Ant. 4:7 (1903), avec TI. pyxidiforum L. Spec. Pl. I ed.
2, 1098 (1753), espéce distincte.

Caribbean Forester - 94 = Vol. 4, No. 2

18.

19.

20.

21,

220
Zoe

24.

250

26.

otic

28.

29.

30.

51.

32.

Trichomanes elegans L. Cl. Rich. in Act. Sec. Hist. Nat. Paris I:114 (1792),
Gases oso, tic Ges, nN. 1225, n. 1888 b, et M. Duss n-
1530. T. anceps Hook. (1846) est synonyme. Duss a
rapporté ses échantillons & T. Prieurei Kunze, espece
affine continentale.

Trichomanes osmundioides Sw. 'Prodr. (1788) et DC. ex Poir. Encycl. 8:65

1808), G.S. n. 1888 c et M.S. n. 3421.

Trichomanes trigonum (Desv.) Kaulf. Enum. Fil. (1824), G.S. n. 1458, n.
2421 et M.S. n. 3406 et n. 3411. Syn. T. Kaulfussii
Hook. et Grev. (1831).

2- HYMENOPHYLLUM J.E. Smith

Hymenophyllum delicatissimum Fée, Cryp. vasc. Brésil 2:83, t. 105, fig. I
(1872-73), G. (ex Fée).
Hymenophyllum crispum H.B.K. Nov. Gen. et Spec. 1:26 (1815), G. (ex Maxon).
Hymenophyilum macrothecum Fée, Mém. Foug. Ant. II:115, pl. 31, f.2 (1866),
G. Duss n. 4402. Syn. H. vincentinum Baker (1891).
Hymenophyllum Francavillei v.d.B. in Nederl. Kruidk. Arch. 4:411 (1859)
G. Duss n. 4272 et M.S. n. 2248.

Hymenophylium interruptum Kunze in Linnaea 9:107 (1834), G. Duss n. 4270

et M. Duss n. 4374.

Hymenophyllum L'Herminieri Mett. ap. Kuhn, in Linnae 35:392 (1868), G. (ex
Mattennius).

Hymenophyllum polyanthos Sw. Journ. Bot. Schrad. 1800:102 (1801), G.S. n.
$24, n. 1201, n. l42lc, n. 1780a, n- 1786, n. 1790 et
M.S. n. 3273a et n. 5228.
Espece variable, polymorphe et de synonymie abondante
et délicate, pour laquelle H. clavatum Sw. (1800) et
H. polyanthos var. clavatum (Sw.) Hook. (1846) sont des
synonymes. Duss, Foug. et Lycop. Ant. fr. 6. 1903 cite
en outre une var. Blumeanum Spreng. de cette espece
et H. Mazei Fournier et Mazé qui ne sont que des formes
écologiques de H. polyanthos Sw.

Hymenophyllum Kohautianum Presl. Hymenoph. 32 et 56 (1843}, G. et M. (ex
Urban). Espéce trés affine de P. polyanthos Sw. et
désignée aussi comme var. Sieberianum Presl, de H.
decurrens Sieber.

Hymenophyllum lineare Sw. Journ. Bot. Schrad. 1800:100 (1801) et Fl. Ind.
Occ. 3:1749 (1806), G.S. n. 1422 b, n. 1778 et n.
1780 b- H. catherinae Hook. (1867) considéré par Duss,
Foug. et Lycop. Ant. fr. 7 comme espéce distincte n'est
pas specifiquement différent de H. lineare Sw. Var.
Dussii Christ, in Duss, Foug. et Lycop. Ant. fr. 9
(1903), G. Duss n. 4267, type et M. Duss n. 1519 a.

Hymenophyllum lanatum Fée Mém. Foug. Ant. 116, t. 31, fig. 3 (1866), G.
Duss n. 4247 et n. 4364.

Hymenophyllum ciliatum Sw. Journ. Bot. Schrad. 1800:2,100 (1801) et Fl.
Ind. Occ. 3:1755 (1806), G.S. n. 1222 et M. Duss n.
1516.

Hymenophyllum sericeum Sw. in Journ. Bot. Schrad. 1809:2, 99 (1801), G.
Duss n. 4266 et Ms Duss n. 1517.

a January 1943

56.4

37 o

38»

39.0

2»

Hymenophyllum hirsutum (L.) Sw. in Journ. Bot. Schrad. 1800;:2, 99 (1801)
et Fl. Ind. Occ. 3:1746 (1806), G. Duss n. 4267 et
M. Duss n. 1519 a.

Hymenophyllum hirtellum Sw. Syn. Fil. 149, G.S. n. 328, n. 1421 b, n. 1888a,
no 2412 et M. Duss n. 1514.

Hymenophyllum latifrons v.d.B. Hymen. 67 (1859), G. Duss n. 4247, n. 4265,
n. 4361 et M. Duss n. 1519. Considéré par Duss (Foug.
Ant. fr. 10, 1903) comme H. hirsutum Sw. var. latifrons
Hook. et Bak. mais démontre distinct par W. H. Hodge
(1941).

Hymenophyllum gratum Fée, Hist. Foug. et Lycop. Ant. 3118, t.30, fig. I
(1866). ° G.S. n. 2052 et M.S. n. 2097, n. 2107a, n.
2655 et n. 3405.

Hymenophyllum fucoides Sw. Fl. Ind. Occ. 3:1747 (1806), G.S. n. 531, n. 831,
mo-1521 a, no L780 bon. 2207 6b et M. Dossier Sie.

n. 1515, n. 1518 et n. 4603.

Hymenophyllum Plumieri Hook. et Grev. Ic. Fil: (1831) [eo MeS.in ise ouDmet
n. 3446. Espéce parfois confondue a tort avec
H. ciliatum Sw.

Hymenophyllum ectocarpon Fée, Hist. Foug. et Lycop. Ant. 115, t 31, f. I,
G. (ex Fée). Considéré par Duss, Foug. et Lycop. Ant.
fr. 7 (1903) comme synonyme de H. fucoides Sw. auquel
il est étroitement ailié, mais différent de ce dernier
ainsi que l'a montré W. H. Hodge in Notes on Dominica
Ferns, Amer. Fer. Journ. 31, 3-4, 125 (1941).

Ordre IV. SALVIN TALES
Famille 1. MARSILEACEAE

MARSILEA L.

, Marsilea Berteroi A. Br. Monatsb. Preuss. Akad. Wiss. Berlin 1870:747 (1870).

Indiqué avec doute par Urban, Symb. Ant. 4:65 pour la
Martinique d’aprés un échantillon de Plée probablement
de Puerto Rico. Nous avons recherché en vain cette
espéce & la Martinique, & la Guadeloupe et dans les
dépendances au voisinage des lagunes, des marécages

et du littoral, au cours de ces 8 derniéres années.

Ordre Ve LY C10*F0.) T2aiL) Bos
‘Famille 1. LYCOPODIACEAE

Ll. LYCOPODIUM L.

- Lycopodium reflexum Lamk. Encycl. 3:653 (1789), G. Duss, n. 2748 et M. S.

n. 3279. Considéré par Duss, Lycop. 8 (1903) comme
var. reflexum de L. rigidum Sw. et comme synonyme de
L. Sieberianum Spring; espéces nettement distinctes.

Lycopodium rigidum Gmel. Syst. 2:1289 (1791), Cond. in Dill. t.57, Fig. 4,
G. (ex Urban).

Caribbean Forester - 96 = Vol. 4, No. 2

12.

13.

14.

Lycopodium Sieberianum Spring. in Bot. Zeit. I:153 (1838) et Monogr. Lycop.
i250 (1882). G.S.-n. $30, n- L801, n. 2410, n- 2414 et
M.S. n.- 1054, n. 2656 et n. 3289.

Lycopodium setaceum Lamk. Encycl. 3:653 (1789), G.S. n. 516, n. 816, n.
1186 et N.S. n. 2099, n. 2581, n. 2386, n. 3275, n.
3409 et n. 3445. Ciest liespéce rapportée par Urban et
par Duss 4 L. verticillatum L. f., le correspondant de
L. setaceum pour les régions tropicales d‘Asie et
d'Afrique.

Lycopodium Wilsoni Underw. et Lloyd, Bull. Torrey Bot. Club, 33:III (1906),
G. (ex Mexon).

Lycopodium funiforme Cham. ex Bory, in Brogn. Hist. Vég. Foss 2:10, 18, pl.

7, fig. 9 (1837), G. Duss n. 3744 et M. Duss n. 1601.

Syn. L. colubrinum 1'*Herm.

Lycopodium tenuicaule Underw. et Lloyd. Bull. Torrey Bot. Club, 33:113

(1906), G.S. n. 543, n. 843, no 1183, n. 1425, n. 1426,
ne 1427 et n. 1785.

Lycopodium mollicomum Spring in Mart. Flor. Brasil. I, I:113, 1840 et
Monogr. Lycop. I, 42 (1849), M. Duss n. 1589.

Lycopodium linifolium L. Sp. Pl. 1100 (1753), G.S. n. 510, n- 660, n. 810
et M.S. n. 2380 et n. 3284.

Lycopodium guatemalense Maxon in Contr. U.S. Nat. Herb. 17:177, t.9, fig.a
(1913), G. Duss n. 3976.

Lycopodium dichotomum Jacq. Enum. Stirp. Vind. 314 (1762), G.S. n. 339, n-
1184, n. 1185, n. 1226, n. 1438, n. 1977 et M.S. n.
1053, n. $318 et n. 3319 b.

Lycopodium mexicanum Herter in Engl. Jahrb. 43, Beibl. 98:49 (1909), M.

(ex Urban) n’est considéré par Maxon que comme synonyme
de L. dichotomum Jacq.

Lycopodium taxifolium Sw: Prodr. 138 (1783) et Fl. Ind. Occ. 3:1753 (1806),
type G.S. n. 661, n. 1423, n. 1428, n. 1792, n. 1800,
n. 2062 et M.S. n. 2659, n. $295, n. 3319 a, n. 3332,
n. 3432 et n. 3433. Var. passerinoides (H.B.K.) Bak.
Fern. All. 16 (1887). Syn.: L. passerinoides H.B.K.
Nov. Gen. 1:41 (1815), L. Schwendenerii Herter, Bot.
Jahrb. Engler 43: Beibi. 98:50 (1909) et Urostachys
Schwendenerii Herter, Report. Nov, Spec. Fedde 19:162
(1923). G.S. n. 1424 et M. Duss n. 4167.

Lycopodium guadalupeanum Spring, Mem. Acad. Sci. Belg. 15, 6:68 (1842),
errore typogr. sub. L. aqualupiano, G.S. n. 1783, n-~
2409 et M. Duss n. 1588.

Lycopodium meridionale Underw. et Lloyd, Torrey Bot. Club. 33:121 (1906),
G-S. n.- 883 et n. 1866. Espéce primitivement rapportée
par Duss et par Urban a L. carolinianum L. Pl. I, ed.
2, SOL (1755):

Lycopodium curvatum Sw. Journ. Bot. Schrad. 1800, 2:116 (1801.) Indiqué par
Fée et par Urban pour la Guadeloupe ot nous ne l‘avons
pas trouvé. Pout 6tre confondu avec L. cernuum L.,
duquel Grisebach Fl. Brit. 647. (1864} ne fait qu’une;
variété; la plupart des spécimens des Petites Antilles
rapportés a cette espece serait, d‘aprés Maxon, L- tortum
Sieber et le L. curvatum Sw. serait endémique de P. Rico.

- 97 - January 1943

15. Lycopodium cernuum L. Sp. Pl. 1103 (1753); G.S. n. 340, n. 517, n. 817, n.
890, n. 1059, n. 1429, n. 1795 et M.S. n. 35269. Ia Var.
Dussii Christ de cette espece, ex Duss, Lycop. Ant. fr.
6 (1903), G. Duss n. 3639, n. 4105 et M. Duss n. 4166,
n'est autre que L. tortum Sieber.

16. Lycopodium tortum Sieber, ex Underw. et Lloyd, in Bull. Torrey Bot. Club
$3:118 (1906), GS: n. 331, im. 552.0. 3350 ne wosGn ane
1895 et M.S. n. 3285.

17, Lycopodium clavatum L. Sp. Pl. ed. 2, 1564 (1753), forme type: G.S. n. 335
et M.S. n. 3290. Var. montanum nov. G.S. n. 537, n. 544,
ne 837, n. 844, n. 1902, n. 2408 et M.S. n. 3288.

Famille 2. PSILOTACEAE
Ll. PSILOTUM Sw.

1. Psilotum nudum (L.) Griseb. Abh. Ges. Wiss, Gétting. 7:278 (1857), G.S. n.
338, ne 663 et M.S. n. 2251, n. 3266, n- 3268 et n. 5299.
2. Psilotum complanatum Sw. Syn.Fil. 188, 414, t.4, fig.5 (1806) M. (ex Fée).

Famille 3. SELAGINELLACHAE

lL. SELAGINELLA Beauv.

1, Selaginella ovifolia Baker, Journ. Bot. Brit. et Fern. 22:90 (1884), G. et M.
(ex Urban et ex Maxon).

2. Selaginella flabellata (L.) Spring. in Bot. Zeit. 1:198 (1838) et Mon. Lycop.
2:174 (1849); G.S. ne $357, n. 343, n. 355, n. 1L82 et Mas.
n. 3337.

5, Selaginella rotundifolia Spring, in Bull. Acad. Sci. Brux. 10:139 (1843) et
Mon. Lycop. 2:85 (1849), G.S. n. 355, n. 541, n. 841, n.
2064, n. 2644 et M.S. n. 3262 et n. 3399.

4, Selaginella portoricensis A. Br. in Ann. Sc. Nat. 5, Bot. 3:288 (1865), M.
(ex Urban).

5. Selaginella albonitens Spring, in Bull. Acad. Sc. Brux. 10:139 (1843) et Mon.
Lycop. 2:80 (1849), G. et M. (ex Maxon).

6. Selaginella substipitata Spring, Mon. Lycop. 2. 198 (1849), G. Duss n. 4078, n.

4084, n. 4086 et M.S. n. 2250, n. 3287 et n. 3338.
- Selaginella Mortensii Spring, Mon. Lycop. 2 (1849), G. Duss n. 3752 et M.
Duss n. 3969.

8, Selaginella Willdenovii (Desv.) Baker, Bot. Jahrb. Engler 24 (1897), M.S.
n. $261.

9, Selaginella didymostachya (Desv.) Spring. Enum. in Bull. Acad. Brux. 10:144

1843) et Mon. lycop. 2:130 (1849), G. (ex Urban).
Selaginella_serpens (Desv.) Spring. Enum. in Bull. Acad. Brux. 10:228 (1843)
et Mon. Lycop. 2:102 (1849), décrit par Duss (Lycop. Ant.
fr. 14, 1903) pour la Guadeloupe et la Martinique, est en
réalité confinée &@ la Jamaique et a Saint Domingue.
Selaginella porelloides (Lamk.) Spring. Mon. Lycop. 2:97 (1849). Décrit par

Duss (Lycop. Ant. fr. 14, 1903) pour la Guadeloupe et la
Martinique, est synonyme de S. Mayerhoffii Hieron. var.
Nectouocii Hieron, in Hedwigia 58:298 (1917), endémique
d'Haiti.

(For summary see previous issue)

Caribbean Forester = 98 = Vol. 4, No. 2

CONTENTS

A seed storage study of some tropical hardwoods. .cccosccccscscssece 99
José Marrero, Puerto Rico

The manufacture of shingles from local woods in Trinidad
and TODA ZO. sce vcccccccccccccscc verre ce veer cece ce rec vesveseoee 107
R. Smeathers, Trinidad and Tobago

Classification des arbres a latex et a secretions de gommes,
resines et matiéres colorantes aux Antilles Frangaises....... 112
H. Stehlé, Martinique

Future may see mahogany forests in Florida..c.scssssesccsccseese 124d
S. J. Lynch and H. S. Wolfe, Florida

Retention of creosote oil in the wood of Pinus occidentalis
WER TUIZ icicle o ro rolce tice ratchonovoletoistcve 6: cvel ore essere cllowove tanovel oie Gibieloreve 66 evehere iene ero 129
E. S. Harrar and D. G. Reid, Duke University

Forests and forest entomology..cccccccssccvcsvsccsvccccecvscc0es Loe
Luis F. Martorell, Puerto Rico

The importance of race in teak, Tectona grandis L.ocewccescccocvccee LOD
Jo S. Beard, Trinidad and Tobago

The Las Cobanitas campeche plantation..cceccecccccecccvceccccers L4Q
Frank H. Wadsworth, Puerto Rico

HEADQUARTERS OFFICE AND LABORATORY BUILDING COMPLETED

Ralph A. Shull
Senior Administrative Assistant
Tropical Forest Experiment Station

We are pleased to announce the completion of the Headquarters Office and
Laboratory building of the Tropical Forest Experiment Station. The building,
which was designed by W. E. Groben, Consulting Architect for the U. S. Forest
Service, is a two-story, concrete structure, painted white, with Spanish-type
tile roof. It has approximately 10,000 square feet of office and laboratory
space which will not only accommodate the present staff but will provide for
expansion in all phases of forest research for some years to come. The build-
ing is planned so that wings can be added without affecting the present struc-
ture.

When Congress appropriated funds in July 1939 for the organization of
a Research Station within the tropics, to be administered by the Forest Ser-
vice, U. S. Department of Agriculture, it was decided to locate the Station on
an area provided by the University of Puerto Rico within the grounds of their
Agricultural Experiment Station in Rio Piedras.

The Insular Department of Agriculture and Commerce and the University
of Puerto Rico, jointly sponsored a Federal Works Project for the construction
of the building, and it has been the writer’s responsibility to see that the
building, other structural improvements upon the grounds, and the landscaping
were properly and successfully carried to completion. The building and other
facilities are shown in the photographs on the following two pages.

On the first floor are located the offices of the Administrative Assis-
tant's Division, the Statistics Division, and three large laboratories with
appurtenant offices. On the second floor are located the office of the Dires-
tor, the Library, and the offices of the Chiefs of the Divisions of Forest
Management, Forest Economics and Forest Influences, a herbarium room, and a
photographic dark room. A large storage room for supplies and equipment is
provided in a basement.

The Library, occupying the central section, is paneled and shelved with
laurel sabino, Magnolia splendens Urban, and has a total of approximately 500
lineal feet of shelf space with further room for book-stacks as the need arises.

The building, surrounded by five and one-quarter acres of grounds, is
situated on a small hill overlooking the harbor of San Juan to the North and
an extensive vista of the island to the West. The area is entirely surrounded
by a road making it self-contained. The building is located on the highest
point of this hill, and an asphalted access road has been constructed to the
rear forming a circle, in the center of which is a 20-car parking lot.

Other improvements which have been constructed upon the grounds are a
concrete eight-car garage, a lumber drying shed, a nursery shade shed, and a
half-acre experimental nursery.

ae

: y
: “s

oie.

ae

Top: Tropical Forest Experiment Station's new office and laboratory building.

Bottom: A close-up of the entrance feature.

Bottom:

1

Rear view of the building showing entrance road which forms a
circle behind the building, with a 20-car parking area in the center.

Showing gerage on left, lumber drying shed on right, nursery shade
shed immediately behind it, and nursery beds to the right of the shade
shed. The relationship of this area to the headquarters building can
be seen from the cars in the parking area in both pictures.

A SEED STORAGE STUDY OF SOME TROPICAL HARDWOODS

José Marrerol/
Tropical Forest Experiment Station
Rio Piedras, Puerto Rico

One of the first and foremost phases of forestry work has always been
the reforestation of cleared non-agricultural lands, Among the numerous
factors which contribute to the success or failure of forest tree plantings
is the quality of the nursery stock planted. To provide high quality stock
tropical forest nurserymen must have access to an adequate supply of good
seed for scheduled sowings during a period of weeks or months which will pro-
duce a continuous supply of seedlings of the proper size during the entire
planting season.

Obviously when continuous production is required during several weeks
or months seed must be continually collected throughout the sowing period or
during part of the period. Fortunately some tropical tree species produce
more than one seed crop annually, or the crop is produced over a long period
during each year. Fresh seed of these species is frequently available on
the trees when needed.

Other species fruit less frequently, some during a brief period each
year, and some producing seed at irregular intervals and in varying amounts.
With such species it is necessary to collect large amounts of the seed when
available and to store it until needed. The rapid loss of viability of some
seeds in the warm moist tropical climates presents a difficulty which must be
removed if stock production is to be concurrent with the best planting season
or if it is to be possible to send seeds great distances for introduction
elsewhere. The collection of large amounts of perishable seed for sowing at
& later date is clearly a hazardous practice unless successful storage
methods are known.

The Problem in Puerto Rico

In the past it has been the custom in Puerto Rico to import during the
Spring large amounts of Spanish cedar, Cedrela mexicana, from Guadeloupe and
broadleaved mahogany, Swietenia macrophylla and S. candollei, from Panama and
Venezuela respectively, to be sown during the balance of the year. Although
the seed was of good quality and it had been possible to protect it from in-
sects and rodents during storage, the cost of the seed made it imperative
that the best practical storage methods be determined.

Algarrobo, Hymenaea courbaril, seed is produced once a year, consis-
tently, and in good amounts. The seed of this species was known to retain

i This study was designed by Messrs. Philip C. Wakeley and L. R. Holdridge
and was carried out under the supervision of the latter. The Caribbean
National Forest provided nursery beds and the necessary labor. The writer
is responsible for the analysis of the data.

- 99 = April 1943

its viability well in storage during short periods, but conclusive information
on storage beyond six months was lacking. Accordingly a study was made to de-
termine the effect of storage under various circumstances upon the viability
of the seeds of these species. It was recognized that there are many factors
which affect viability during storage, such as length of period, temperature,
humidity, seed moisture content, and sealing of storage containers. The study
was designed to show results after various storage periods up to 2 years, and
to test the relative effect of sealed and unsealed containers and of room and
refrigerated temperatures upon loss of viability. Because of a lack of equip-
ment it was impossible to include a study of moisture relationships.

The Seeds and General Storage Practice

The fruits of Swietenia macrophylla are large, ellipsoid, brown, woody
capsules about 6 inches long, containing about 40 seeds. The seeds are about
3 inches long by 3/4 inch thick, most of the length being made by the long
membranous wing. There are about 900 seeds (including the wings) to the pound.

The fruits of Cedrela mexicana are ellipsoid, dark brown, dehiscent
capsules, 1-1/2 to 2 inches in length. The seeds have the same general out-
line of those of mahogany but are considerably smaller, about 1 inch long.
There are about 14,000 seeds (including the wings} to the pound.

The fruits of Hymenaea courbaril are rough, oblong, dark brown, woody
legumes about 4 inches long. The seeds are large and heavy, oblong to ovoid,
and leguminous and are enclosed by a hard testa. They are reddish brown, and
are found in a mealy pulp within the pod. There are about 120 seeds per pound.

Usually when the pods are received they are laid out in the sun in
trays or screens to dry. The dehiscent capsules of mahogany and cedar are
easily opened when dry but in the case of Hymenaea courbaril the thick walled
capsules must be opened by force. If the seeds are not to be sown immediately
they are stored in tin cans, wooden boxes or sacks. The tin cans are prefer-
able as they protect against losses through rodent damage.

Experimental Procedure

Seeds which were all from the same lot, were counted and grouped for
the various treatments. No attempt was made to modify the natural moisture
content, so all treatments of each species can be considered comparable re-
garding this factor. Seeds to be stored unsealed were placed in small muslin
sacks while those sealed were placed in quart fruit jars with a rubber sealing
ring. Seeds stored at room temperature were placed in a wooden office building
at Rio Piedras where the temperature average is 77° F., minimum 54° F., and
maximum 98° F.; and where the average relative humidity is approximately 80
per cent. The refrigerated temperature was that of a commercial cold storage
plant, ranging between 35° and 40° F.

Eight storage periods were used, ranging from 2 weeks to 2 years. Each
treatment included 100 seeds (4 blocks of 25). On the date that storage was
begun (0 days) 400 seeds were germinated to provide a basis for determination
of subsequent losses in viability.

Caribbean Forester = 100 = Vol. 4, No. 3

Sowings were made in randomized blocks in the nursery. All seeds were
sown in drills, the cedar being covered with sand, and the other species with
earth.

Results

Spanish Cedar

There are two obvious characteristics of the cedar germination data as
shown in Table 1. One is the rapid loss of viability during storage at room
temperature, and the other is the great fluctuation in the germination per-
centages, particularly after the longer periods of refrigerated storage.

Table 1.—Germination by blocks
and storage treatments = Spanish cedar.

Room Temperature Refrigerated

Storage Average Average| Total
Period Percent Percent| Average
No. No. No> No. ° No. No. No. No. °

Open Storage

Odaysl/ 13 15 le 14 54 LOW <5 Che LOMIELO 38 46.0
2 weeks Sune su Ls 42 See ey 55 48.5
1 month 5 a ee 25 De ee 22 23.5
SeronEHemee 2h) 5) $B 2 12 io eee eG 41 26.5
4month 0 0 0 O ) pie GB 9G 31 15.5
6months 0 0 0 0 0 Cie aliee Oe aL 4 2.0
8 months 0 0 0 O 0 es cee 41 20.5
l year Gao. O 0 0 i een, 13 6.5
2 years Om Gp. 05. 0 0 One’ Ole © 1 0.5
Potal 29 36 35 33 53 76 58 59
Sealed Storage
Odays3/ 9 7 9 10 35 Qua fe 41s 4 44 39.5
2weeks 11 18 16 12 57 i vg, wate.) 9 56 56.5
1 month Bee 8. 4 24 Bini a. Gi 20 22.0
2months 5 7 #5 5 22 7 ie a 5 50 36.0
4months O 0 2 0 2 Gee 8m na eG 34 18.0
6 months 0 0 0 0 0 Ghee a) Pie we wu 7 3.5
8months 0 0 OO O 0 Neh Laue ie PS 45 22.5
lL year Gr 1 | 6 0 BM 2. Mog) 1G ll 5.5
2 years oP oO 10 Yo 0 OF) foe aes es 10 5.0
Total 29 39 41 331 66 78 74 59.

if This period, corresponding to a check, involves no treatment.

=1@l = April 1943

The effect of temperature was not pronounced until after 2 months of storage
but after 4 months at room temperature practically all seeds were dead. The
high germination obtained after 8 months of cold storage falls off rapidly
during the following four months.

The fluctuation in the germination percentages, such as the "rise" in
germination between 1 and 2 months and 6 and 8 months of cold storage, is ap-
parently due to variation in the weather conditions in the nursery beds during
the different seasons when sowings were made. This factor places a limitation
on the interpretation of these data, but the sharp drop in germination after
. month of room temperature and 8 months of cold storage can be considered a
reliable indication of what to expect.

Sealing the seed had no statistically significant effect upon loss of
viability, though sealed seeds tended to show a higher germination during some
periods. Sealing is therefore unnecessary unless it is used to protect the
seed from weevils or rodents.

Honduras Mahogany

Mahogany seed was found to be similar to cedar in its response to the
various storage treatments, as shown in Table 2. The fluctuations in germi-
nation percentages after the various periods is less pronounced, indicating
that the germination of this species is less affected by weather. The seed
used had been stored in a sack at room temperature for 6 weeks prior to the
study. While this circumstance probably had little effect upon the treat-
ments, this period of time should be added to the storage periods to get
actual age of the seed.

Table 2.—Germination by blocks and
storege treatments =- Honduras mahogany »

oe ees gerated

Storage Blocks of 25 seeds | Average | Blocks of 25 seeds| Average| Total
Period Percent | 1 | 2 | 3 | 4 | Percent a

No. No. No. No. % No. No. No. No.

Open Storage

Odayslf 10 15 16 13 54 TGS alle ean ealS 55 54.5
2 weeks 145 oe aS 15 55 L293 has as 56 55.5
1 month 15 one Peele 55 LAS sion aly, 62 58.5
2 months 5 4°12 9 30 Qi) FSAI. 8 41 35.5
4 months Oh 0 1 1 2 Li dy PLT 1 teeueeke 56 29.0
6 months On 440 07.430 0 9 Bc eaal 9 37 18.5
8 months ) fe) ) i 1 12) PAB, CAG yan 67 1 129.40
1 year fe) 1 Or yr 2 5 7 8 7 27 14.5
2 years ) fe) fe) fe) a LO) ) ) ) 0) 0.0
Total 44 44) 1,55), 156 93 104 99 95

Caribbean Forester - 102 = Vol. 4, No. 3

Table 2.—Continued

Refrigerated
Blocks of 25 seeds | Average Blocks of 29 seeds of = |Blocks of 25 seeds| Average

Storage
Period

Now ONoe! Now Noe _ No. = No.

Sealed Storage

Oldayal/ 92 U9 24. le 61 ie Aegan dee tt5G 58.5
2 weeks TZ) Ge Peay? Welk: 59 16P 16h) 628 66 62.5
L month 15 Te Ste 1S 50 6p 14) 4 Ls 57 53.5
2 months TK) Dols} 9 15 50 13 14 #19 °« 16 66 58.0
4 months ) L 4 3 8 Ieee 5 ow 81s 58 3300
6 months fo) 0 0 fe) 0 Ly ts CeensLi7 51 25.5
8 months 0 0) i 0 1 16 8 16 9 49 25.0
l year 0 fo) fe) fe) fe) Sif) al, alee 8 42 21.0
2 years 0 0 0 0 @) a 3 L 0 5 209
Total Sano? 158; 60 lll 109 118 108.

if This period, corresponding to a check, involves no treatment.

As with cedar seed temperature was an important influence upon loss of
viability. After 1 month unsealed and 2 months of sealed storage, the viabil-
ity of seeds kept at room temperature fell off rapidly. In cold storage the
loss in viability commenced at about the same time but was much more gradual,
particularly that of the seeds which were in sealed containers.

The use of sealed containers preserved the viability of seeds stored at
room temperature better than unsealed containers after 2 months storage. With
«clad storage the effect is shown to be similar in slowing the loss of viabil-
ity after 8 months. Therefore with mahogany the sealed container can be con-
sidered the better practice.

Al. carrobo

In contrast to the seeds of cedar and mshogany, which lost their viabil-

ity after 1 or 2 months of storage at room temperature and 8 months or a year
in cold storage, algarrobo seeds retained viability well during the first year
regardless of storage conditions. (See table 3.)

It will be noted that viability increased in all treatments until 4
months had passed. Apparently this is partly due to after-ripening. The
possibility of the need for after-ripening of this seed is indicated by the
long period during which the seeds remain in the pods on the tree before
falki ING.

- 103 - April 1943

Table 3.—Germination by blocks
and storage treatments - Algarrobo.

Refrigerated

storage Blocks of 25 seeds Blocks of 25 seeds | Average
Period “| 1 | 2 }'3 .|k4 jbercent | Vai 2in [esis reeinenc 2

No. No. No. No. 0 No o No. No. No. jo Se

Open Storage

Q days h/ 15: 10°: 43% | 44 52 16 43 13) 420 52 92.0
weeks ti i i424 » 200-415 48 Lee Ow, ad 9 45 46.5

, month 14 15 16... 15 59 15 13 15 9 52 5505

4 months 15 12 1a Li 49 12 el 13 14 60 54.5

“ months 22 iM) ANT 13 TL 18 25 17 4 72 71.5

5 months 10.)—s«d14 Lg 1h 54 1, 15 LO 2) 45 49.5

‘ months 18 14 Lee) 13 63 S) PO Ras 9 41 5200
veer 19 el 15 14 69 19 16 13 14 62 65.5
years ff 8 3 9 ot di 16 16 16 59 43.0
Total SL (127 “Lek sils 2b. d42> i217 Loo

Sealed Storage

Odeyst/ 18 16 13 Ul 58 2500 ial tee aed) fe 58.0
weeks 1721s) 86 70 19, a6 » a2Oy ot 68 69.0
month 16 18 20 ai 75 22 lula toe ek | ame 76.0

2months 22 17 17 15 71 257 KLGi 2a y more 480 75.5

4 months 21 21 20 19 81 18:09) LBA Bae ATs ers 77.0

A menths 12 17 15 21 65 13°. kegs Sek ses 64.5

8 months 17 18 13 13 61 1S LG uewal eer menere 66.5
year on sop tile as 44 2a 1 isi) 422) aera wee 63.0

2 years @ Jor 6: 2 2 1g: | aSe 6 20" Tags ese 28.5
Total 133 136 127 131 178 133 169 149

_/ This period, corresponding to a check, invoives no treatment.

Though the seeds store well regardless of temperature or container
cenerally there was a significant difference between the sealed and unsealed
storage, in favor of the former. However, at room temperature this relation-
ship is reversed after 1 year. This may be a result of the high humidity
maintained in the sealed container which could lead to more rapid exhaustion
or the seeds.

The effect of temperature is not marked until after 1 year of storege

when cold stored seeds were clearly better than those kept at room temper-
eture. This difference is most pronounced in the seeds kept sealed.

Caiibbean Forester - 104 = Vol. 4, No. 3

Conclusions

Under the conditions of the study the following conclusions are
justified:

1. Refrigeration is necessary for the safe storage of Spanish cedar
seeds beyond two months. In this way seeds can be stored safely
up to eight months although some germination was obtained up to
2 years.

2. Cedar seeds will keep equally well in either sealed or unsealed
containers.

5. Honduras mahogany seed must also be refrigerated for satisfactory
storage beyond 3 months. Seeds store fairly weil up to 1 year in
cold storage.

4. Honduras mahogany seeds retain their viability longer when stored
in sealed containers.

5. The viability of algarrobo seeds can be satisfactorily preserved
for 1 year at room temperature, but for longer storage, refriger-
ation is required.

6. For algarrobo seeds unsealed storage is to be recommended if
storage is to be for periods in excess of 12 months.

Summary

In order to arrange nursery sowing schedules to provide a continual
production of stock throughout the planting season it is frequently neces-
sary to store seed for varying periods following its collection.

In Puerto Rico where seed has often been imported from other isiands
at considerable expense the development of improved storage practices has
been &4n important phase of nursery research. Seed storage of many species
has been tried, among them Cedrela mexicana, Swietenia macrophylla, and

Hymenaea courbaril.

In a study covering 2 years and testing the relative effects of
storage at room and refrigerated temperatures and sealed and unsealed tem-
peratures it was found that:

1. The viability of seeds of Cedrela mexicana can best be preserved

(satisfactory after 8 months) when kept at about 35° F., but
need not be sealed.

= Moly & April 1943

2. Seeds of Swietenia macrophylla retain their viability longest
(satisfactory after 1 year) also when refrigerated and prefer-
ably kept sealed.

3. The storage of the seeds of Hymenaeca courbaril presents no
problem during the first year but if storage for longer periods
is contemplated, refrigeration is to be recommended.

Resumen

Para poder organizar un programa de siembra en el vivero que provea
una produccién continua de arbolitos durante toda la época de trasplante es
necesario frecuentemente almacenar la semilla por periodos diversos después
de la recoleccién.

En Puerto Rico, donde las semillas se han importado a menudo de otras
islas incurriéndose asi en grandes gastos, el desarrollo y mejoramiento de
las prdcticas de almacenaje ha constituido uma fase importante en las inves-
tigaciones relacionadas con los viveros.

En un estudio que se prolongdé dos afios se probé el efecto relativo
dé almacenar semilla bajo temperatura ordinaria y fria y en envases sellados
y sin sellar. Se encontré lo que sigue:

1. Para un almacenaje efectivo de Cedrela mexicana la refrigeracién
es necesaria para periodos en exceso de 2 meses. A temperaturas
ba jas las semillas se pueden aimacenar convenientemente hasta
los 8 meses aunque se obtuvo alguna germinacién hasta les 2 afios.

2. En el almacenaje de Cedrela no existe diferencia marcada con
respecto @ envases.

3. Para que la semilla de Swietenia macrophylla se conserve satis-
factoriamente por mis de 3 meses debe almacenarse a temperatura
baja. En estas condiciones puede conservarse bien hasta un afic.

4. Las semillas de Swictenia retienen mejor su viabilidad si se
almacenan en envases sellados.

5. La viabilidad de las semillas de Hymensea courbaril se conserva
satisfactoriamente por un aflo bajo temperatures ordinarias pero

por um periodo més largo se requiere refrigeracién.

6. Si se desea almacenar semillas de Hymenaesa por periodos de més
de 12 meses se recomiende el uso de envases sin sellar.

Caribbean Forester = 106 = Vol. 4, No. 3

THE MANUFACTURE OF SHINGLES FROM LOCAL WOODS IN
TRINIDAD AND TOBAGO

R. Smeathers
Assistant Conservator of Forests
Trinidad and Tobago

Introduction

The manufacture and use of shingles from local woods was once common in
Tobago where the two species Cedrela mexicana (cedar) and Qcotea leucoxylon
(duckwood) were used for this purpose, both being naturally durable and suffi-
ciently fissile to be riven into shingles by hand. The standard size was 18
inches by 4 inches with a thickness of 1/2 inch at the stouter end, while in
1933 their cost was around 60 cents per 100 or $3.60 per "square". Lengths of
service for shingles used on roofs and walls of 20 and 30 years respectively
2re reported. From about 1935 their manufacture has been discontinued, prob-
ably on account of the ease and cheapness with which galvanized iron roofing
became obtainable. It is not known to what extent locally made shingles were
used in Trinidad itself, but in the Mayaro district a number of shingle roofs
are still in existence and are in perfect condition after a known life of
from 30 to 40 years. In these cases the shingles were made from Copaifera
officinalis (balsam).

Research Work Carried out by the Forest Department = 19402/

Mainly with a view to finding an outlet for the less valuable and, at
present, worthless species felled in large numbers yearly, both in the for-=
mation of teak plantations and by petroleum companies in the course of oil
exploitation, the Forest Department conducted a small scale experiment in
1940 with the object of finding out whether any of these species were suit-
able for shingle manufacture.

To produce the test shingles the method described in the Malayen
Forester (Volume 6. °1937) was employed as it involves only a small circular
s@w rip-bench and a simply constructed wooden craddle. A standard size of
18 inches by 6 inches tapering from 3/8 inch to 1/16 inch was adopted.

The timbers selected were mostly light, soft and not naturally dura-
ble, the shingles were therefore treated with a 50-50 creosote-diesel oil
mixture using the open-tank hot and cold process. Except for one species
absorption was heavy and penetration complete.

if The experimental work referred to in this section was carried out by
J. C. Cater, Assistant Conservator of Forests, Trinidad.

— LOT = April 1943

The species used and results obtained are summarized in the following

table;

Absorption
of preser-
vative

(approx. )

Timber

Hura crepitans Sandbox 25 20 Cuts fairly easily without
much pick-up when not ab-
solutely green. Seasons
well, and retains its shape
if properly stacked.

Remarks on working

Botanical Name qualities, etc.

Erythrina Anauca 25 40 Cuts fairly easily but
poeppigiana immortelle (about) with some pick-up. Sea-

sons and retains shape
well, but is very brittle.

Hernandia sonora Toporite 20 56 Cuts well with little or
no pick-up, except samples
containing the false black
heart. Seasons and retains
shape very satisfactorily.

Spondias mombin Hogplum 34 - The sample was from a small
tree and was very difficult
to cut. The shingles were
woolly, and warped and col-
lapsed badly on seasoning.
None suitable for use was

obtained.
Bravaisia Jiggerwood 35 14 Cuts with a fine smooth
interrigima finish and seasons very
well. Somewhat hard.
Cordia Lay lay 30 24 Cuts moderately well, but
collococca inclined to pick-up.

Liable to warp and split
during seasoning,

Sapium Mi lkwood 35 28 Cuts well with smooth
aucuparium finish and little pick-
up. Seasons well.
Virola Ca juca 50 - The sample was almost im=
surinamensis possible to cut owing to

woolliness. Only one shin-
gle of poor quality was
obtained and this split
badly during seasoning.

Caribbean Forester - 108 - Vol. 4, No. 3

The foregoing list by no means exhausts the possible species, but re-
presents those of little commercial value at the time, although it is of inter-
sst to note that sandbox has since acquired considerable populerity as boarding
for house construction with a corresponding increase in value. Additional
‘ocal species which might repay investigation are Alchornea glandulosa (honey-
wood), Ficus tobagensis (figuier), Didymopanax morototoni (jereton) and

Simaruba_amara (maruba).

These experiments showed that at least three local species are capable
of yielding a satisfactory roofing shingle as far as machining and seasoning
properties are concerned: namely, sandbox, toporite and milkwood. Unfortu-
nately cost of production using the cradle method was high and while the
nature of the investigation prevented accurate costing it was estimated that
a shingle should be produced for around 4 cents, corresponding to about $16.00
per square. This could not then compete with the price of imported Western
ved cedar shingles, or alternatives such as corrugated iron or roofing felt.
Nor did the potential market seem sufficiently attractive to induce local saw-
miils te undertake production with specialized machinery capable of high out-
put, though it is yet to be discovered whether a machine designed and con-
structed for use with a wood such as Western red cedar will prove suitable
when used on structurally different tropical woods. Thus for a time the mat-
ter fell into abeyance.

Subsequent Investigation by a Petroleum Company ~ 1940

With the rapid rise in price and threatened lack of galvanized iron at ~
the beginning of the present war considerable interest in the manufacture of
shingles - or rather the possibilities thereof - from local woods was evinced
by the various petroleum companies in Trinidad; and one of these, Messrs.

Apex Trinidad Oilfields Ltd. began an investigation early in 1940 especially
with @ view to obtaining comparative costs of shingles and galvanized iron
roofing >

Using the cradle method already referred to sufficient shingles were
cut to cover a smali lean to roof in order that they might be tested under
working conditions. Two species only were used, toporite and sandbox. Of
these, part of the toporite and all the sandbox were treated with a 50-50
creosote-diesel oil mixture by the open-tank hot and cold process. The re=
mainder of the toporite shingles were laid untreated.

As far as costs were concerned this investigation lost much of its
value when force of circumstance compelled the adoption of the slow cradle
method of cutting, but the resulting shingles — however expensively produced—
have given a good indication of their durability, for after almost three
years the treated ones are perfectly sound and in excellent condition. The
untreated toporite shingles lasted well for two and a half years but have
subsequently rotted: they still form a watertight roof but cannot be expected
te stand up much longer.

=- 109 = April 1943

Full Scale Production by a Petroleum Company = 19422/

In January 1942 Messrs. Trinidad Leaseholds Limited commenced the pro-
duction of shingles from timber felled in the course of clearing forest for
well sites, etc. In the beginning a wide variety of timber was supplied to
the shingle plant but many proved to be unsuitable owing to seasoning dif-
ficulties—warping, splitting and cellular collapse—while others were obvi-
ously too hard and heavy. Now toporite is being used to the exclusion of
all others and its supply, therefore, has to be supplemented by selective
fellings outside the normal oil clearings.

Logs are first cut at the sawmill into blocks measuring 18 inches by
6 inches by random widths of about 1 foot: these are then cut at the shingle
plant on a machine, designed and made by the Company, which comprises a 24-
inch diameter circular saw powered by a 3 H.P. electric motor, together with
@ metal cradle running on tracks and designed to give the correct degree of
taper (5/16 inch to 1/16 inch) to the shingles as they are cut. The cradle
is hand fed and by an ingenious rack device the block is advanced to the saw
in such a way that alternate shingles are cut with their thick and thin ends
together; this ensures that the grain of the wood runs parallel to the shin-
gle’s face. .

After cutting, the sningles are soaked for 24 hours in a copper naph-
thenate solution. They are then removed to drain and are finally bundled in
50's. The plant is operated by two men, one cutting, the other dipping and
bundling, and two shifts are worked daily = each shift producing 1000 shin-
gles. By the end of 1942 well over 400 "squares" had been manufactured and
put to use on buildings of all kinds throughout the Company's fields.

Wastage is rather excessive and with the present set-up embodying a
saw of 1/4 inch kerf amounts to 55 per cent by volume of the wood blocks sup-
plied to the shingle plant = exclusive of those rejected on account of splits,
shakes, false heart and other defects. In spite of this the Company states
that it is now producing shingles at a cost of $16.00 per "square", which may
be itemized as follows:

Royalty value of wood $0.36
Transport to mill, conversion into

blocks and transport to shingle plant 9.88
Cutting into shingles and preservative

treatment (labor) 1.36
Preservative treatment (material) 4.40
$16.00

The present cost of galvanized iron is $13.00 per 100 square feet of
roof covered, but it must be remembered that this materiai is now often

We Grateful acknowledgement is made to the General Manager, Trinidad Lease-
holds Ltd. for assistance in the preparation of this section.

Caribbean Forester = 110 = Vol. 4, No. 3

tainable or only so in comparatively small quantities. This venture of
rs. Trinidad Leaseholds in the manufacture cf shirgies is, therefore, not

y @n extremely interesting one from the point of view of local wood utiii-
sation but would appear to be an economically sound one for themseives, at
-24St under wartime circumstances.

Summary

Shingles were onte made by hand from local species in Trinidad and
Tobago and durebility as high as 40 years is reported. In about 1935 their

manufacture was discontinued = probably on account of the ease and cheapness
with which alternative materials were obtainable

2 1940 both the Forest Department and a local petroieum company
experiments in the manufacture of shingles by machine methods; the
anes ae with a view to finding a use for certain less valuable species,
th xr in order to compare the cost of shingles with that of gaivanized
ron roofing. Details of the species tested ere given.

In 1941 another petroleum company started the production of shingles
=o meet their own needs, in part. The method of manufacture is described
and the cost of timber, labor and preservative material given.

Resumen
En Trinided y Tobago el tejamani se fabricaba a mano, de especies lo-
cales y seen informecién al efecto su indice de durabilidad ascendia a tanto
somo 40 afios. Por el 1935 su fabricacién se suspendi6éd debido probablemente a

que podian Bueenanee materiales substitutos con més facilidad y a menos costo.

En el 1940 el Departamento Forestal asi como una compafiia petrolera
locel llevéron & cabo experimentos para fabricar tejamani usando maquinerie
edecueds; el primero con el propésito de encontrarie provecho a ciertas ma-~
deras de menos velor; la segunde, para poder comparar el costo del tejamani

om el del cinc de techar.

Los experimentcs consistian en hervir las muestras previamente en una
mezcla 50-50 de creosota y aceite diesel. Tres de las especies que se podian
sortar y curar bien y que retenian su forma si se apilaben debidamente fueron
hure crepiteans, Hernandia sonora y Sapium @ucuparium. Otras especies que po-
ian us@rse pero con resultados menos satisfactorios son Erythrina poeppigiana .

Bréveisia interrigima y Cordia collecocca. Dos especies gue no dieron resul-
tade fueron Spondias mombin y Virola surinamensis.

0 9

Una prueba posterior hecha por otra compefiia petrolera mostré la posi-
bilidad econGmica de fabricar tejamani de Hernandia sonora por lo menos mien-
tres dure le guerra ya que el cine que se usa para techar es virtualmente
imposible de conseguir.

- lll = April 1943

CLASSIFICATION DES ARBRES A LATEX ET

A SECRETIONS DE GOMMES, RESINES ET MATIERES COLORANTES

AUX ANTILLES FRANCAISES

H. Stehleé
Ingénieur Agricole et d'Agronomie Coloniale
Martinique

Parmi les arbres, spontanés ou cultivés, aux Antilles frangeises, une
catégorie est nettement reconnaissable par le sylviculteur non spécialisé en
botanique ou méme par le profane grace aux secrétions laticiféres, aromatiques:
zommeuses ou résiniféres, oléiféres ou colorantes, émises par certaines de
leurs parties ou par tous leurs organes. On peut en compter 70 environ dans
ses Tles et les classer rationnellement d’aprés la nature physico-chimique de
la matiére secretee ou d’aprés des caracttres morphologiques simples et
extérieurs. Dans cette 6tude, une classification générale de ces végétaux a
été élaborée, tenant compte & la fois de l’aspect extérieur du produit secrété,
de la constitution botanique des feuilles et des fruits ainsi que de la
couleur du bois qui revét une certaine importance en Economie Forestiére. Les
distinctions taxonomiques faisant appel & l'organisation florale ont été
laissées de c8té pour simplifier l'emploi de la clef forgée dans un but d'uti-
lisation pratique et en permettre aisément la généralisation.

Les sucs aromatiques de certaines espéces, comme celles du genre
Zanthoxylum, brulent avec une grande clarté et sont utilisées pour la confection
de torches ou flambeaux-caraibes alors que les résines des gommiers: Dacryodes
excelsa Vahl, Protium attenuatum (Rose) Urban, Elaphrium simaruba (L.) Rose
et Tetragastris balsamifere (Sw.) Kuntze, dégagent une odeur d’encens. La
présence de ces résines balsamiques dans ces bois explique leur faculté d'@tre
imputrescibles dans l*’eau de mer, qualité mise & profit dans la confection des
bateaux de péche légers appelés “gommiers".

Les feuilles du Marila racemosa Sw., possédent des canaux oléiféres et
lies de Pilocarpus racemosus Vahl, une huile essentielle d‘ot l'on extrait
pilocarpine; dans le Pterocarpus officinalis Jacq. et le Maclura tinctoria
) Gaud. des matiéres colorantes rouges sont secrétées par le bois ainsi
gue dans diverses espéces d’Acacia et dans 1*Haematoxylon campechianum L.; le
Myristica fragans Houtt. contient un suc visqueux péle dans 1’écorce du trone.
Des gommes-résines, &cres ou colorées, sont exudées par le Moronobea cocinea
Aubl., le Tovomita plumieri Griseb., le Mammea_ americana L., le Sapium
caribaeum Urb., etc.

Il y aurait lieu encore d'ajouter les végétaux a tannins qui ne sont
pas.traités dans la présente étude. Les exemples sont donc nombreux et la
° ° 2 ° ° ° ° ° 2 2 °
connaissance chimique de la constitution intime du produit secrete permettrait
certainement de prévoir de nouvelles utilisations, de connaitre l'influence de
ces matiéres sur la texture et la valeur du bois et de déduire les relations
qui existent d'une part entre les diverses catégories d'excrétion et d'autre

Caribbean Forester - ll2 = Vol. 4, No. 3

part la place occupée par les arbres qui les produisent, dans la classification
générale des végétaux- Des affinités et des corrélations existent certainement.
Pour les Antilles frangaises, les arbres producteurs de résines aromatiques 4
odeur d'encens et d’élemis sont surtout de la famille des Burséracées (ou
Térébenthacées), les substances améres dans les espéces des Méliacées, les

sucs aromatiques entretenant la combustion et l’éclairement dans les Rutacées,
les gommes aux Mimosacées, les tannins et phlobaphénes aux Combrétacées et
Rhizophoracées, etc.

Le synopsis présenté ci-aprés a pour objet de faciliter la déterminatioa
et la distinction de ces arbres et dans l’appendice il est indiqué la référence
du lieu ot le diagnose a été publiée et les noms créoles des espéces @ la
Guadeloupe et & la Martinique.

I. Arbres & Latex

1. Feuilles par 3 ou 4:
2. Feuilles oblongues ou elliptiques, acuminées... Rauwolfia lamarkii A-D-C.
2. Feuilles lancéolées-oblongues, obtuses... Rauwolfia biauriculata J. Mill.
Feuilles par 2, opposées:
2. Feuilles coblongues-lanc4o0lées, & pointe
ODtuSE>.2ccccccccceeccccccex0x00cccecccvece LAbDErnAemontana citrifolia L.
2. Feuilles ovées-oblongues & orbiculaires,
PEROBTie et IO tetera lee = cle ielo) a) ssaletelciceile cleiciele ele!e civic «o's WOM LOLEODIS procera, Agte
iL. Feuilles alternes:
2. Feuilles lobées:
3. Limbe &@ 8-11 lobes pennéS....cccccccccce+cooe Artocarpus communis Forst.
4. Céte, nervures et pétioles glabres....2ccseccoooe Vars non seminifera
4. Cdte, nervures et petiole pubescents..csccccecoccecce Varo Seminifera
3. Limbe des jeunes fevilles & 3-5 lobes, feuilles
adultes entiGres..ccccccccvcccccccvcccceveces Artocarpus integrifolia L.
2. Feuilles entiéres:
3. Feuilles tipulées. Bois blanc tendres et mous:
4, Feuilles trés grandes, de 15 & 25 cm. de

t-

es ovées, stipules glabres...-...00¢0000 Ficus urbaniana Warb.
es oblongues, stipules 4 poils
GOPESccccccvccccccvexv0000000000000e0ccov0s Ficus crassinervia Desf.
4. Feuilles moyennes, de 8 & 15 cm. de long:
5. Feuilles elliptiques ou oblongues,
arrondies ou obtuses & la base, &
13 nervureS.coscscccsxcsceccscccccecccccecceceee FiCUS krugiana Ward.
5. Feuilles ovées cordées a la base, & 3-5
NEPVULES «co cccccce score cccccccceecccevcvcccvcv00 Ficus laevigata Vahl
4, Feuilles petites, de 34 10 cm. de long....... Ficus omphalovhora Warb.
3. Feuilles non stipulées, bois résistants et
colorés:
4, Bois mi-dur, jaune foncé, de densité moyenne........ Hura crepitans L.
4. Bois dur, gris, de grande densité:
o- Feuille elliptique, coriace; bois gris,
veiné de brun, nuancé de jaune Hippomane mancinelia L.
4. Bois durs, lourds, rouges ou brun-rouges;

=e April 1943

5. Feuilles assez étroites, lancéolées, elliptique

6. Feuilles. a6 nceéolées-oblongues........ Achras sapota L. var. typica
Feuilles & plus long pétioles.cocccccscccccee Varo candollei Pierre
6. Feuilles nettement: Lancéo0l6es o.. o.cisis ordiciaterol o slonovopeistetenel o chaveterene eich hen nee
cececocccccccoocvccrecs POuteria martinicensis (Pierre) comb. nov.
6. Feuilles elliptiques-oblongues..ccccccrcsrc voces s0000000000e000000
sovoecccccesroccvoccevooss ue Pouteria: fabri lis (Pilorre)icombsanodws
6. Feuiiles nettement eiliptiques ;
7. Weuililes glabres ou presque:
8. FPeuilles grandes, de 13-18x6-8, 5 cm. limbe
de 10 & 16 cOtes..cccccrccooee Pouteria semecarpifolia Pierre
8. Feuilles moyennes, de 5-l2x4-6 cm., limbe
de 42 cOtes...-oeec0ceece Manilkara riedleana (Pierre) Dubard
8. Feuilles petites, de 4-6x2=-4 CMcoeccsnccccevrccrvr000c 00000000
ooccceevecccces Pouteria chrysophylloides (Pierre) comb. nov.
7. Feuilles pubescentes & la face inférieure;
8. Poils soyeux argentbés...cse...o Chrysophylium argenteum Jacq.
8. Poils soyeux dorés;
9. Pétiole de 12-15 mn.
Feuille cuspidee au sommet et
acutéee A la bas@...coces0ceo0 Chrysophylium caeruleum Jacq.
9. Pétiole de 15-28 mm.
Feuille moyennement acuminée:

LO. /Base obtuse cu acubee se cocoons cto ote ecoroyekoustotosaretetotene /ouatctckers
Chrysophylium cainito L. var. typicum
10. Base arrondie on atténuees «co-ciaa' cinie > mole oletorcie ectoncretne
sgo0000ss0000007 Varo pomiferum (Tussac) Pierre

SWS 2

i

©oP70007F7TDT LOX D000 000

Ol Base bri

usgquement acuminée..... var. martinicense Pierre
5. Feuilles ass eee, obovees:
a1) :

6. Feuilles See
oc occcccovcccccccczcc0cc00000 CAaLOGarpUM mammosum (A.DC.) Pierre
6. Feuilles obovee es-elliptiques :
7. Limbe de 16-21 om., & pointe

spends ea eet . Pouteria dussiana (Pierre) comb. nov.
7. Limbe de 15-16 cm., & pointe
ObEUSE sco 00000800008 se ausoseesd POUterianmultimlonas (pCa) mnyme

6. Feu ll es tena nen OBDOVECS Sicte dno lotonokolsl oione GuoecuoMahcnovenonobouonoRCconotcaoronOnonomonc
oooccoceccccss Pouteria discolor (Walp. et Duchass.) comb. nov.
6. Feuilles obovées-arrondies..Pouteria hahniana (Pierre) comb nov.

II. Arbres & Excrétions Aromatiques ou Colorées

1. Feuilles simples, opposées ou verticillées, rarement
alternes :
2. Feuilles longues et peu larges:
3. Feuilles elliptiques-lancéo0léess.o.ccccceccco000 Sapium caribaeum Urb.
3. Feuilles elliptiques-oblongues.....ceccccsc0cccc00 Styrax glabrum Sw.
3. Feuilles ovales-lancéolées.-.o000000 Chlerophora tincstoria (L.) Gaud.
3. Feuilles oblongues-lancéolées:

Caribbean Forester = ld = Vol. 4, No. 3

4. Feuilles aliternes, grandes
CamaUxVOLGT PETES cles oe 0c «ca
4. Feuilles distiques, moyennes,

OletPerosielecc eae eo8

3. Feuilles nettement lancéolées

coossoo00n000000 000000000

et longues, sans

ecooogocoeceo cseogosescoces# © 9 80 Mangifera indica ine
possédant des canaux
Marila racemosa Sw.

et acuminées:

4. Arbre cultivé,

sans racines aériennes, écorce 4&

suc visqueux rouge p&le..

sesesecoe00csceoocs 00600

o Myristica fragans Houtt.-

4. Arbre spontané,

a racines adven

tives Be tence

bois & gomme-résine jaune.
2. Feuilles amples et large:
3. Feuilles suborbiculaires,

Og LOBEOS sciocicrs owcts «

3. Feuilles ellipticues-obovées.
ues:

5. Feuilles largement

elliptigq

cordées,

ecoeooeodco

ococo

Moronobea coccinea Aub.

entiéres

«eso Ochroma pyramidale (Cav.) Urb.
Mammea americana L:.

osoe#to0e0e0eg0c Oesco eae 090000

4, Feuilles peeeesar une gomme-résine,
mucronées a 1 IgsdheGnee og se vsdo leon co abeobe Tovomita plumieri Griseb
4. Feuilles secrétant une huile essentielle
par des cryptes, arrondies ou échancrées
Bel SextrOma Ge a ctejdiclclclo sucteicc delocvlloccscaecoe Ptlocarpus racemosus: Vahl
3. Feuilles ovées ou oveles:
4. Feuilles membraneuses, dépourvues de
canaux secréteurs, racines & matiére
Golorante violette.....-..-se..+. Nectandra membranacea (Sw.) Griseb
4. Feuilles rigides et OSES &
gomme résine jaune et dorante..sssssssseceees Rhoedia lateriflora L.
3. Feuilles obovées ou ee
4. Feuilles verticillées & l'extrémité des
YAMAUX, COPIACEES.....2escccecoocvecesoeece AnAcardium occidentale L.
4. Feuiiles opposées, SESE et charnues:
5. Feuilles veinées, semi-amplexicaules........0-- Clusia venosa Jacq.
5. Feuilles & nervation peu apperente,
attenuées & la base
OpirinteoOvenOuwe WianL 1 GUE. 560s cclsjc oc cee ocuesivcs, ClUSaasrOSeER dBCd<

6. Fruit globuleux
1. Feuilles composées, pennées,

2. Feuilles & pennies subdivisées en f
3. Arbrisseaux & épines stipulaires
de 4 cm. de long; fruit piat:

4. Feuilles & 3-8 paires de penne

p@ires de fol LioOlesa. > -

3. Arbrisseaux & épines stipulaires

& 4 cm. de long; fruit de 8 a 10

4. Feuilles @ 2-6 paires de penne
paires de folioles lineaires,

de LONE ecscrieiccsscccocees

e

°

coo °o

=
Ve

ece Clusia plukenetii

aiternes:

olioles:
ce plus

s et 10-30

edamecia ACACIA millotica (bs) Delile
de 0.5 cm.

mm. d*’épaisseur:

s et 10-25

de 3-4, 5 mm.

secccceoe Acacia farnesiana (L.) Willd.

Jo

4. Feuilles & 2-8 paires de pennes et 10-20

paires de folioles

linéaires-oblongues,

ade 4.5=-7 mm. de

Longe « o«

2. Feuilles & folioles simples:

3. Folioles crénulées sur
le plus souvent épineux:

4. Folioles peu nombreuses,

rides cisic ste siet > aoe ACACIA tortuossa, (L,.) Willd.

les bords; arbrisseaux

2-7:

- 115 =

5. Petit arbuste, folioles 2-7, généralement 3,
ovées ou elliptiques-lancéolées, sessiles... Zanthoxylum punctatum Vah!
5. Arbre élevé, folioles 3-5 généralement 5,
lancéolées ou lasangiques, pétiolées..cos+cooecccees Amyris elemifera L.
4. Folioles nombreuses, 5-15:
5. Fruit en folicules connés en leur milieu:

9

6. Folioles oblongues ou elliptiques, non
OVSECSs occoscoscscvccuses soooees o ZAnbhoxylumemnantvini censem@Lam.)) De.
5. Fruit & folicules libres:
6. Foliole elliptique oblongue ou
ovée; folicule reniforme obliquement
orbicula Arocccccccccooseocccavccevvce0ccce Zanthoxyium caribaeum Lam.
6, Foliole ové, lancéolé ou elliptique;

1

folicule sbovolieschuas ans uate eae ce Zanthoxylum flavum Ven
3. Folioles entiéres, arbres le plus souvent inermes:
4. Feuilles paucifoliolées, 1-2 folioles:
5. Foliole unique, ovale ou el! iptique=ovée;
fruit: petit folicule globuleux ou
globuleux-covoide.:>scoooccoscooo Zanthoxyium monophyllum (Lam.) P. Wilson
5. Folioles 2 inégaux, oblongues ou obl.ongues
lancéolées; fruit gousse cblongue, comprimée..-... Hymenaea Gourbaril L.
4. Feuilles plurifoliolées, 2-10 folioles:
o> Fruit gousse oblongue:
6. Folioles 2-4 peires cunéiformes-cvées.... Haematoxylon campechianunm i.
5. Fruit petit folicule globuleux ou ovoide;:
6, FPolioles 3-9, linésires ou HS Dep anes wei oa aera es
FO eee Ne Re biol aleyotornerlen » Zanthoxylum spinifex (Jacq.} DC.
5. Fruit capsule Rporme. dchieeen
Loculicide, par 4 valves:
6. Folioles 1-3 paires membraneuses et
elliptiquessccccoccacerccvcv00evevcvcesveovcveovvcces Guarea plabra Vahl
6. Folicles l-5 paires coriaces et

obloe OM BUCS cs02220e2200e00CCvo7C Cov HD 00000 Guarea perrcttetiana ING JUSS 6

>

5. Fruit drupe de 2-4 loges cu & 3 valves:
6. Felicles 3-7, ovées ou obovees:

7. Drupe cbhlengue, triangulée et

CTRL NES oe eke ere tae ee a Elaphrium simarube (L.) Rose
6. Folicles 5-9, oblongues ou elliptiques
7. Drupe sphérique n’atteignant pas
2 om, et comportant 2-4 Logesc.coccscocccccervcsrcs020000000000060900
ec derscesrsvoseearoeccee does Letragasibris sbausamitorar sw.) mkuntiae
7. beiee de Z2-& cM:
8, Fruit oblong-elliptique, indéhiscent...... Dacryodes exceisa Vah!.
8. Fruit ovoide & déhiscence valvaire
et & 3 MoS Sree yo BELE Go mace no? Protium attenuatum (Rose) Urba
4. Feuilles multifoliées: 10-25 folioles
5. Fruit: capsule & 5 vaives Hemneeencs
septifrage:
6. Folioles 10-16, ovales-elliptiques..cooceccceceoees Cedrela odorata L.
5. Fruit: drupe ovoide ou globuleuse:
6. Fruit globuleux, de 1 cm. & | om.
1/2 de diamétrea:

3

Caribbean Forester = 116 = Vol. 4, No. 3

7. Folicles 7-13, oblongues ou ovées-

oblongues, drupe ovoide..... Picraena antillana (Eggers) comb. nov.
7. Folioles 5-9 paires, lancéolées-

oblongues; drupe bianguleuse...o2..ccccooceccoss Simaruba amara Aubl.

6. Fruit: ovoide de 2.5 8 10 cm. de long:

7. Folioles 9-25; drupe pourpre, de

2-5-3 cMo de Longececocecccocccccececcvccccccce Spondias purpurea L.
7. Folioles 5-7 paires; drupe jaune,

de 10 cm. de lLongesessceccecccccccoeceece Spondias dulcis Forst. f.

Références et Désignations Vernaculaires des Arbres Cités

——————————E———————— ee

APOCYNACEAE

Rauwolfia lamarkii A.DC. Prodr. 8:337. 1844. Bois lait (G. et u.)2/

Rauwoifia biauriculata J. Mill. Linnaea 30.1860. Arbre & lait (G.)
Taberneemontana citrifolia L. Sp.Pi. 210:8. 1753. Bois lait (G. et H.)

ASCLEPIADACEAE

r a) Ra Bree in Aitsrs Hort. Kewsedses 2270. LSie
Arbre & soie, cot rance, bois-lait (G.); Coton-Siam, bois-canon,

bois-pétard (M.).

MORACEAE

Artocarpus communis Forst. Char. Gen. 102. 1776. La var. non seminifera
est l’arbre & pain (G. et M.) et la var. seminifera est le chataigner
du pays (G. et M.).

Artocarpus integrifolia L.f. Suppl. 411. 1781. Jacquier (G. et M.).

Ficus urbaniana Warb. in Urb. Symb. Ant. 3:459. 1903. Figuier grandes
feuilles (G. et H.).

Ficus crassinervia Desf. in Willd. Sp.Pi. 4:1158. 1806. Figuier banian
(G.); figuier maudit (M.).

Ficus krugiena Warb. in Urb. Symb. Ant. 3:487. 1903. Figuier maudit
(G. et M.); figuier marron, figuier &@ agoutis (G.) et cocoyer
rivitre (M.).

Ficus laevigata Vahl, Enum. 2:183. 1805. Figuier maudit (G. et M.);
figuier banian, figuier blanc (G.).

Ficus omphalophora Warb. in Urb. Symd. Ant. 3:466. 1903. Figuier ti-
feuilles (G. et M.), multipliant (G.).

Chlorophora tinctoria (L.) Gaud. Bot. Freyc. Voy. 508. 1826; murier
du pays (G. et M.).

BUPHORBIACEAE

Hura crepitans L. Sp. Pl. 1008. 1753. Sablier (G. et M.).
Hippomane mancinella L. Sp. Pl. 1191. 1753. Mancenillier, maximilier
(Go tet Ms):

yy G - Guadeloups et M - Martinique. lLieux ot le nom vernaculaire cité est
employé.»

- 117 = April 1943

Sapium caribasum Urb. Symb. Ant. 3:308. 1902. Bois de soie (G.); bois
la glue et aralie (M.).

SAPOTACEAE

Achras sapota L. Syst. 10, ed. 2:988. 1759. On distingue aux Antilles
frangaises la forme type: var. typica nov., sapotillier ordinaire
(G. et M.) et la var. candollei Pierre, @ fruit presque ovoide.

Pouteria martinicensis (Pierre) comb. nov. Syn.:; Lucuma martinisensis
Pierre in Urb. Symb. Ant. 5, fase. I, 105. 1904. Pomme pain, pain ~
d'épice (M.). |

Pouteria dussiana (Pierre) comb. nov. Syn.: Lucuma dussiana Pierre in
Urb, Symb. Ant. 5, fasc. I, 105. 1904. Pomme pain, pain d’épice (M.).

Pouteria multiflora (DC.) Byma, Not. Guy. Sapot. in Rec. Trav. Bot.
Neerl. XXXIII, p. 164 (1936) et in Meded. Bot. Mus. Herb. Utr. n. 27,
p. 164 (1936). Syn.: Lucuma multiflora DC. Prodr. 8: 168. 1844. Pomme
pain, pain d’épice (G. et M.).

Pouteria fabrilis (Pierre) comb. nov. Syn.: Oxythece fabrilis Pierre
in Urb. Symb. Ant. 5, fase. I, 161. 1904. Balate blanc (G.).

Pouteria hahniana (Pierre) comb. nov. Syn.: Oxythece hahniana Pierre
in Urb. Symb. Ant. 5, fase. I, 161. 1904. Balata rouge (G.);
palate, balata, palata, barac (M.).

Pouteria semecarpifolia Pierre in Symb. Ant. 5, fasc. I, 108. 1904.
Bois caraibe, bois contrevent, bois créole (M.).

Manilkara riedleana (Pierre) Dubard, Ann. Mus. Col. Marseille, III,
3:14. 1915. Sapotillier marron, sapotillier noir, bois noir (Ga);
balata, balate (Mt. ).

Pouteria chrysophylloides (Pierre) comb. nov. Syn.: Micropholis
chrysophylloides Pierre in Urb. Symb. Ant. 5, fasc. I, 122. 1904.
Caimitier-grand bois, Catiimitier-bois (M.).

Pouteria discolor (Walp. et Duchass,) comb. nov. Syn.: Chrysophy] lum
discolor Walp. et Duchass., ms.script. in herb. Berolin; Micropholis
discolor (Walp. et Duchass.) Pierre in Urb. Symb. Ant. 5, fase. I,
121. 1904. Caimitier ti-feuilles, caimitier grand bois (G.).

Chrysophylium argenteum Jacq. Enum. 15. 1760. Bois la glue, Caimitier
bois, bois rabi, bois de bouis, petit bouis, acomat (G.), bois
bouis, petit bouis (M.).

Chrysophyllium caeruleum Jacq. Sel. amer. pict. 30. 1780. Gros bouis,
caimitier noir (M.).

Chrysophyllum cainito L. Sp. Pl. I ed. I, 192. 1753. A l’état indigéne
et cultivé, existent trois variétés: var. typicum nov. & baies
subsphériques violacées, caimitier (il.), var. pomiferum (Tussac)
Pierre, loc. cit., & baies sphériques a demi vert-rouges et a demi
violacées et la var. martinicense Pierre, @ fruits subglobuleux
blancs: la grosse blanche (M.).

Calocarpum mammosum (L.) Pierre in Urb. Symb. Ant. 5, fase. To3e
1904. Gtest la forme type qui se trouve aux Antilles frangaises:
Sapote, grosse sapote {M.) et zapote ou sapote & créme (G.).

OLEACEAE

Styrax glabrum Sw. Prodr. 13. 1788. Oranger-bois, cypt-orangé, cypre
orange (G.); laurier-caraibe, bois chypre, bois madame (M.).

Caribbean Forester - 118 = Vol. 4, No. 3

ANACARDIACEAE (Terebenthaceae)

Mangifera indica L. Sp. Pl. 200. 1753. Manguier (G. et M.).

Anacardium occidentale L. Sp. Pl. 383. 1753. Pommier d‘acajou (G. et
MM. NS

Spondias purpurea L. Sp. Pl. ed. 2, 613. 1763. Prunier d'Espagne,
prunier du Chili (G. et M.).

Spondias dulcis Forst. f. Prodr. 34. 1786. Pomme-cythere et prune
cythere (G. et M.).

HYPERICACEAE

Marila racemosa Sw. Fl. Ind. Occ. 19. 1797. Bois casse rosse, résolu-
martinique (G.), bois cachiman, cachiman-grand bois (M.).

MYRISTICACEAE

Myristica fragans Houtt. Handleid. 3:333. Muscadier (G. et M.).

GUTTIFERAE

Moronobea coccinea Aubl. Pl. Guyane 2, 789:313. 1775. Palétuvier-
jeune (G.).

Tovomita plumieri Griseb. Fl. Brit. W. I. I. 106. 1864. Mangle-bois,
palétuvier grand bois (M.).

Memmea americana L. Sp. Pi. I, ed. I, 512. 1753. Abricotier, abricot
des Antilles, abricot de Saint-Domingue, mamey, abricot-pays (G.
hae sale

Rheedia lateriflora L. Sp. ed. 2, 1193. 1753. Bois l'onguent, abricotier
baterd (G.), ciroyer, abricotier-batard, abricotier-bois, abricotier
montagne, abricotier bord de mer (M.).

Clusia venosa Jacq. Enum. 34. 1760. Mangle, mangle montagne, mangle
rouge, figuier maudit des hauts, palétuvier montagne (G.).

Clusia rosea Jacq. Enum. 34. 1760. Figuier maudit, figuier marron,
abricot batard, abricotier maudit (G.); aralie zabricot, aralie
grande eauinen aralie (M.).

Clusia pluckenetii Urb. Symb. Ant. 5, 432. 1908. Aralie rose, aralie
grande feuille, grande aralie grise, aralie z'abricot (G.).

BOMBACEAB

Ochroma pyramidale (Cav.) Urb. in Fedde Repert. Beihefte 5, 123. 1920.
Pripri, fromager-mapou, patte de liévre, patte de lapin (G.) et bois
flot (G. et M.)-

RUTACEAE

Pilocarpus racemosus Vahl Eclog. I, 29:10. 1796. Flambeau caraibe (G.);
bois flambeau, flambeau noir (M.).

Zanthoxylum punctatum Vahl; West Bidr.Sante Croix 310. 1793. Lépineux
rouge, bois flambeau, bois d’Inde marron, lépuni (G.); bois flambeau
noir lépineux (M.).

- 119 = April 1943

Zeanthoxylum spinifex (Jacq.) DC. Prodr. I, 728. 1824. Bois chandelle,
bois lépineux blanc, bois @ piano, bois blanc 4 flambeau (G.); bois
flambeaux (M.).

Zanthoxylum martinicense (Lam.) DC. Prodr. I, 726. 1824. Lépiné jaune,
lépuni jaune, lépineux jaune (G. et M.).

Zanthoxylum caribaeum Lam. Encycl. 2, 39. 1786. Lépineux blanc, bois
chandelle blanc (G. et M.).

Zanthoxylum flavum Vahl, Eclog. 3. 48. 1807. Noyer, bois noyer (G.).

Zanthoxylum monophyllum (Lam.) P. Wilson, Bull. Torrey Bot. Club 37. 86.
1910. Lépuni jaune, bois noyer (G. et M.).

Amyris elemifera L. Syst. 10, ed. 2. 1000. 1759. Bois chandelle (G. et

M.), bois pini, bois flambeau, bois chandelle blanc (G.).

LAURACEAE

Nectandra_membranacea (Sw.) Griseb. Fl. Br. W.I. 282. 1860. Bois doux,
laurier (G.); laurier-chypre (M.).

MELIACEAE

Guarea glabra Vahl Eclog. amer. 338. 1794; bois pistolet (G. et M.),
néflier des bois (G.).

Guarea perrottetiana A. Juss. in Mém. Mus. Paris, 19,.2413285. 1830.
Bois pistolet, bois rouge & balles (G.), bois cacao (M.).

Cedrela odorata L. Syst. 10. ed. 2, 940. 1759; Acajou (G. et M.);
acajou amer, acajou senti, acajou 4 meubles, acajou pays, acajou
rouge (G.).

BURSERACEAE

Elaphrium simaruba (L.) Rose North Amer. Flor. 25, 246. 1911. Gommier
rouge (G. et M.), gommier, gommier-barriére (G.). |

Tetragastris balsamifera (Sw.) Kuntze Rev. Gen.Pl. 107. 1891. Gommier |
encens (G.). |

Dacryodes excelsa Vahl Skr. Nat. Selsk. 6. 117. 1810. Gommier blanc,
bois cochon (G. et M.).

Protium attenuatum (Rose) Urban, Symb. Ant. VII, p. 240 (1912); Icica
heptaphylla Duss, Fl. Ph. Ant. fr. p. 182, non Aubl. Bois d‘'encens,
gommier blanc, bois gommier (G. et M.).

SIMARU BACEAE

Picraena antillana (Eggers) comb. nov. Syn.; Picrasma antillana Urb.
Symb. Ant. 5, 378. 1908; Rhus antillana Eggers Fl. Sante Croix and
Virg. Isl. 41. 1879; Aeschrion antillana Small, North Amer. Fl. 25.
3335. 1911. Peste & poux, bois noyer, graines vertes (G.), bois
amer (M.).

Simaruba_amara Aubl. Pl. Guyane 2. 859. 332. 1775. Bois blanc (G.
et M.); acajou blanc (G:).

Caribbean Forester = 120 = Vol. 4, No. $

LEGUMINOSAE

Acacia nilotica (L.) Deli
(G.)3; acacia savane (M.

Acacia farnesiana (L.) Willd. Sp. Pl. 4. 1083. 1803. Acacia jaune,
acacia senti, pompons jaunes (G. et M.).

Acacia tortuosa (L.) Willd. Sp. Pl. 4. 1083. 1806. Acacia piquant,
acacia savane (G. et M.).

Hymenaea courbaril L. Sp. Pl. 1192. 1753. Courbaril (G. et M.).

Haematoxylon campechianum L. Sp. Pl. 384. 1753. Camp€chier (G. et.M.).

e Fl. Aegypt. 79. 1812. Acacia de Cayenne

Summary

Among the spontaneous and cultivated trees of the French Antilles a
category is clearly recognizabls by the forester not specialized in botany
and even by the layman: trees which secrete gummy or resinous, oil or dye-
producing substances by certain or all of their organs. There are approxi-
mately 70 of them in these islands and they can be reasonabiy classified ac-
cording to the physico-chemical nature of the secretion or by certain exterior
simple morphological characters. A general classification of these trees is
presented, bearing in mind the exterior appearance of the secretion, the con-
stitution of leaves and flowers and the color of the wood which has certain
importance in forest economy. The taxonomic distinctions in relation to
floristic organizations have been left out in order to simplify the use of
this key to facilitate practical utilization.

The aromatic latex of certain species, for instance those of the
genus Zanthoxylum, burns with great brightness, and is used to make torches
and candles; while the resins of gum-trees -such as Dacryodes excelsa Vahl,
Protium attenuatum (Rose) Urban, Elaphrium simaruba (L.) Rose, and Tetra-
gastris balsamifere (Sw.) Kuntze, give an incense-like odor. The presence
of these balsamic sssences in the wood is responsible for their durability
when submerged in sea-water, a property required in light fishing boats or
"commiers" .

The leaves of Marila racemosa Sw. possess oil ducts, and those of
Pilocarpus racemosus Vah!, an essential oil from which pilocarpine is ex-
tracted. From the wood of Pterocarpus officinalis Jacq. and Maclura
tinctoria (L.) Gaud. as well as from various species of Acacia and Haema. =
toxylon campechianum L., a red coloring matter is secreted. Myristica _
fragans Houtt. contains a pale, viscous juice in its bark. Bitter or col-
ored gum-resins are exuded by Moronobea cocinea Aubl., Tovomita plumieri
Griseb., Mammea americana L., and Sapium caribaeum Urb.

The key presented does not include trees yielding tannin. There are
many of these. Chemical analysis of these secretions will certainly produce
new uses, will provide information on the influence of these substances on
the texture and value of the wood, and the relations existing between the
different secretions and the place occupied by the trees which secrete them

= l2l = April 1943

in the general classification of plants. Affinities and correlations certainly
exist. In the French Antilles the trees which produce aromatic resins, with
incense or gum elemis scent are mostly from the family Burseraceae (or Tereben-
thaceae); these producing bitter substances belong to species of the Meliaceae;
those yielding aromatic juices which support combustion and burn with brightness
belong to the Rutaceae; the gum-producing, to the Mimosaceae; the tannin and
phlobaphene-yieiding, to the Combretaceae and Rhizophoraceae.

The key serves to facilitate the determination and distinction of these
trees. It is followed by @ list of the references from which the diagnoses were
obtained and also the creole names of the species in Guadeloupe and Martinique.

Resumen

Entre los arboles esponténeos y cultivados de las Antillas francesas
el silvicultor no especializado en botaénica y aun el lego, puede reconocer
claramente una categoria especifica gracias a las secreciones laticiferas
aromiticas; gomoses o resinosas, colorantes o aceitosas, segregadas por al-
gunos o por todos sus érganos. En esas islas pueden hallarse alrededor de
70, los cuales pueden clasificarse razonablemente segun Ja naturaleza fisico-
quimica de la materia segregada y segun ciertos caracteres morfolégicos sim-
ples y externos. Bl texto comprende una clasificacién general de dichos
arboles tomande en cuenta a la vez el aspecto exterior de la secrecién, la
constitucion botaénica de las hojas y frutas asi como el color de la madera
gue reviste cierta importancia en la economia forestal. Las distinciones
taxonémicas relacionadas con la organizacién floristica no han sido incluidas
pare. poder asi simplificar el uso de esta clave forjada con un propésito de
utilizacion practica.

Bl latex aromético de ciertas especies, como por ejemplo, las del géne=-
ro Zanthoxylum, arden con mucho brillo y se usan para hacer antorchas y velas
mientras que las resinas de Dacryodes excelsa Vahl, Protium attenuatum (Rose)
Urban, Elaphrium simaruba (L.) Rose, y Tetragastris balsamifera (Sw.) Kuntze,
emanen un oloy de incienso. A la presencia de estes esencias balsdmicas en
le madera se debe su incorruptibilidad al estar sumergidas en el agua de mar,
Dor lo cual se usan en la confeccién de dos botes de pesca livianos llamados

"rommiers" .

Las hojas de Marila racemosa Sw. poseen canales aceitosos y las de
Pilocarpus racemosus Vahl segregan un aceite del cual se extrae la pilocar-
pina. De la madera de Pterocarpus officinalis Jacq., Maclura tinctoria (L.)
Gaud., Haematoxylon campechianum L., asi como varias especies de Acacia, se-
gregan un tinte nee La corteza de Myristica fragans Houtt. contiene un
jugo p&lide wiscosco. Moronobea cocinea Aubl.,. Tovomita plumieri Griseb.,
Mammea. americana Lo y Sapium Sepium caribseum Urb. , paucan gomorresinas amargas o
coloreadas »

Caribbean Forester = 122 - is Vol. 4, No. 3

En la clave no estén incluidos los arboles que contienen tanino los
cuales son numerosos. El andélisis quimico de sus secreciones podré delinear
nuevos usos, dara un indice informativo de la influencia de estas substancias
sobre la textura y el valor de la madera y ademés las relaciones que existen
entre las diferentes secreciones y el sitio que ocupan los arboles que las
segregan en la clasificaci6én general de las plantas. Sin lugar a duda exis-
ten afinidades y correlacicnes. En las Antillas francesas los arboles que
producen resinas arométicas con olor de incienso o elemi pertenecen en su ma-
yoria a la familia de las burserdéceas (o terebentaceas); los que producen sub-
stancias amargas, a ciertas especies de las melidceas; los que segregan jugos
aromaticos que eayudan en la combustion y arden con brille, pertenecen a las
rutaceas; los que producen goma, a las mimosaéceas; los que producen tanino y
flobafenc, a las combretaceas y rizofordceas.

La clave sirve, por lo tanto, para facilitar la determinacién y la dis-
tincién de estos arboles. Le sigue una lista de las referencias que se usaron
en su elaboracién junto con los nombres criollcs de las especies en Guadalupe
y Martinica.

000

UN_ARBORETO DE ARBOLES NaTIVOS

El Institute Politécnico de San Germén, Puerto Rico, ha reservado un
rodal de 10 cuerdas en un lindero de sus terrenos con el firme propésito de
crear alli un arboreto de a&rboles nativos exclusivamente.

El rodal esté admirablemente eadaptado para tal fin. El Dr. I. Vélez,
profesor de biolcgia y el Sr. L. E. Gregory, uno de los colaboradores de la
Estacion Experimental de Silviculture Tropical encontraron alli setenta y
cinco especies de arboles nativos. Entre los més numerosos se encuentran

Hymenaea _courbaril L., Ocotea leucoxylon (Sw.) Mez, y Bugenia sp.

Entre las dos especies exéticas mejor representadas en el rodal se en-
cuentren el mango, Mangifera indica L. y la pomarrosa, Jambosa Jambos (L.)
Millsp. Estos deberén descuajarse gradualmente y el drea que ocupan seré de-
Gicada al cultive de especies natives.

El arboreto continuar&é siendo un lugar de recreo y esparcimiento come
lo fué en el pasado y ademas servird en el futurc de bosque de experimentacién
para las clases de biologia y otras ciencias afines. Para aumentar su utili-
dad a este respecto el bosque se proveera de los caminos necesarios para hacer
eaccesibles todas sus partes y se le pondré un rétulo met&lico permanente por
lo menos a un ejemplar de cada una de las especies que estén representades en
el arboreto.

La Estaci6n Experimental de Silvicultura Tropical coopera ofreciende su

ayude. en la fase técnica del trabajo y enviando semillas y brinzales de las
especies que faltan.

= 123 - April 1943

FUTURE MAY SBE MAHOGANY FORESTS IN FLORIDAL/

S. J. Lynch
Associate Horticulturist
Florida Sub-Tropical Experiment Station

H. S. Wolfe
Head, Department of Horticulture
College of Agriculture, Univ. of Florida

Whet kind of worid shall we live in after this war is over? How dras-
tic will be the economic changes that will determine the course of life and
world commerce in those days? In an attempt to answer such questions one
point is outstanding. The Americas, and the United States in particular, are
determined to be more self sufficient, more completeiy contained. Hundreds
of experiments are today under way that will, if successful, make Florida a
new source of raw materials that formerly came from across the seven seas.

Production of Rhodesian mahogany, Khaya nyasica Stapf., on thousands
of acres of Little-used South Florida land is among the more commercially sig-
nificant of these prospects. Observations in « forestry block containing 20
Khaya trees, at the Sub-Tropical Experiment Station, Homestead, and presented
ia table 1 indicate a marked superiority in growth of the Khaya over slash
pine, Pinus caribaea Morelet, in the same pict and at the three ages measured.

Table 1.—--Comparative growth data,
Kheya ivorensis and Pinus caribaea.2

Average
Log Volume

Inches Inches Feet Peet Cullen
9 Khaya 5 2.8 6.0 Zoe 3702 4167
g Pine J Q 200 ier L} ue 26. 739
LO Khaya 5 309 6.9 2 od 59.3 5538
LO Pine ff Zio O04 UAT 25.8 1066
di Khaya 10 304 ao Parra 39.9 5754
AA Pine 12 PAS ie o Stor 25.8 - LO9L

Khaya nyasica—The bark of 10 year old tree was 0.25 in. thick at top of log
and 0.38 in. thick at bottom.

Pinus caribaea=-The bark of 10 year old tree was 0.335 in. thick at top of log
and 0.67 in. thick at bottom.

i/ Reprinted from Florida Grower, August 1942.
ef This table has been summarized from the original article.

Caribbean Forester - 124 = Vol. 4, Noo. $

The mahogany trees were almost twice as tall in over-all tree height
and produced about twice as long a clear log length as pines of the same age.
The story, from its beginning, is a fascinating and significant one.

The name mahogany has been applied, properly and improperly, to many
kinds of-woods during the last century. Originally mahogany was obtained
solely from the West Indies and Central America from trees of the genus
Swietenia in the family Meliaceae. This wood gained 4 unique reputation for
its color, hardness, remarkably slight shrinkage, and its power of repelling
the attacks of boring insects.

At present the woods on the timber markets of the world deserving the
name “mahogany” come from Central America, the West Indies, and Tropical West
Africa. The American woods are considered products of the genus Swietenia as
mentioned before; and the West African wood is obtained in the main from two
genera, Khaya and Entandrophragma, closely related in the same family to the
American genus. Trees in these genera are so closely similar in foliage,
flowers and seed, as well as in wood, that an experienced botanist is required
to separate some of them. Lumbermen following the tecnnical data of the
Mahogany Association, Inc., consider the three species iisted below as true
mahogany -

West indian Mahogany - Swietenia mahagoni Jacq., which grows in
the West Indies and the southern tip of Flori

to
fo¥)
A)

Tropical American Mahogany = Swietenia macrophylla King., which
grows from southern Mexico to northern South America.

African Mahogany - Khaya ivorensis A. Chev., is the principal
species exported from Western Africa, but several other
species are included under this type of mahogany.

African mahogany is often solid under the name of its port of shioment
or region of derivation, such as Gambia mahogany or Cape Lopez mahogany, and
is thus often confused in the markets. The color of the wood ranges from
light pink through bright red to red-brown. The wood is hard, works well and
produces veneer of unusual lengths and widths. As African mahogany i
figured, most mahogany veneers are of African origin. It is alse used in
boat building, furniture making, and many types of finish work,

One of the several species considered as an African mahogany, Khaya
nyasica, is natiwe to the evergreen forest on the banks of streams in
Nyasaland and the eastern part of Northern Rhodesia. R. J. Miller, Fores
Office, Ndoia, Northern Rhodesia, states it grows into a large, fine, erect
tree attaining a height of over 100 feet and @ girth of more than 15 feet.
Trees have been reported of 200 feet in height and almost 50 feet in girth
at breast height. "The wood is a rich brown color, is not attacked by borers

= 125 = April 1943

and termites3/, weighs approximately 40 pounds per cubic foot, and works well
except for a tendency to interlocking grain, which, however, results in an
attractive ‘striped! figure." The timber of this particular species, although
of good quality, is practically unimown on our markets. In its native land
the wood is often used for general building purposes but is usually employed
for furniture and decorative work.

Seeds of Khaya nyasica were first secured forty years ago by the Plant
Introduction Service, Bureau of Plant Industry, United States Department of
Agriculture. The earliest introduction was received on January 31, 1902, from
Dr. W. Le Thompson of Mt. Silinda, Southern Rhodesia, and was given the plant
introduction number 8311. Dr. Thompson also sent to the Plant Introduction
Service no fewer than six other lots of seed of this species from Mt. Silinda
between the years 1921 and 1930. The Sub-Tropical Experiment Station received
0 plants of P. I. 85748 received at Washington on February 15, 1930, and 10
plants of P. I. 90449, received on December 5, 1930.

The February 15 introductions were planted in the Sub-Tropical Experi-
ent Station forestry block in August 1932, as were also five of the plants
of the December 5 introduction. The other five trees of the last introduction
were transferred into the forestry block in May 1934, from a windbreak row.
These last five were planted in dynamited holes, using one-half stick per hole.
The other fifteen were planted in raw soil in any available pockets or crevices,
in rows approximately 20 feet by 20 feet. The soil is Rockdale series with
only a moderate amount of topsoil available. The forestry block was in second
growth Pinus caribaea saplings 10 to 20 feet tall at the time the Khayas were
planted among them.

5

Further introductions of seed were made by the Sub-Tropical Experiment
Station from the Conservator of Forests, Ndola, Northern Rhodesia, in January
1936, March 1940, and May 1940. One hundred seventy plants of the 1936 intro-
ductions were planted in the forestry block under different degrees of shade
and at various spacings. The remainder of this introduction and the plants
from the two introductions made in 1940 were distributed to interested cooper-
ators throughout South Florida. In three locations trial forests of fifty
trees each were planted in 1941.

All of the Khaya trees growing in the forestry block are making very
good growth. The oldest trees are as tall as, and in some cases taller than,
the surrounding pine trees. The trees have survived low temperatures with
the loss of only a few leaves. In March 1941 temperatures of 25 degrees F.
were recorded in an open area within one-fourth mile of the forestry block.
They do not show any injury from insects or diseases. In Northern Rhodesia
this Khaya, when planted under open plantation conditions on dry lend, has
grown rapidly but has invariably been attacked within a few years by a shoot-

3/ Ed. Note.—Samples of Khaya ivorensis were recently tested for termite

resistance by Dr. G. N. Wolcott of the Puerto Rico Agricultural Experiment
Station at Rio Piedras. This related species, while possibly differing in
this respect from K. nyasica was found to have low resistance to attack by
Cryptotermes brévis Walker, the common dry-wood termite of the West Indies.

Caribbean Forester - 126 - Vol. 4, No. 3

borer (probably Hypsipyla Spo). This attack so kills back the stems that, in
spite of frequently throwing out new shoots, growth is so completely arrested
that the trees are often killed.

In the summer of 1941 measurements were taken of the 20 Khaya trees
first planted in the forestry block. The ten oldest trees were considered as
11 years from seed, the five from the next introduction as 10 years old and
five from the last Bureau of Plant Industry introduction which were trans-
planted were considered 9 years from seed. The volume of clear saw log was
the object of the measurement. A good number of the Caribbean pines among
which the Khaya were planted were measured also and their age determined by
the use of an increment borer at 4 feet height, with adjustments for seedling
growth. Enough pines were measured so that there were data for as many pines
of the same ages as there were Khayas. The length of the log was measured
from a normal stump-cut to where the heavy branches were forming on the trees.
Girth measurements were made with a steel tape at the top and at the bottom
of the log. If the log was more than 20 feet the girth was taken at 20 feet.
The average of these two was considered the log circumference, from which the
averages diameter was calculated. Total tree height was measured by triangu-
lation to the tops of the tree foliage.

The maximum and minimum values for each measurement varied widely in
each group for both species. As the number of trees measured in each sample
was necessarily small, this variation must be borne in mind in the evaluation
of the data.

The diameter at the bottom of the log of the Khaya trees was almost
twice 2s great at 9 years and more than twice as great at 10 and 11 years of
age as that of pines of a similar age. The greatest difference, however, is
in the comparison of log volumes. At all three ages the Khaya produced a
little more than five times the volume of potential log of the pines. The
bark was calculated into this volume. It can be noted also from table 1 that
the bark on the Khava was not as thick as on the pine. This would tend to
accentuate the difference in actual wocd volume. The general inerease in tree
size and log volume as the trees increase in age is shown for both Khaya and
pines.

The growth made by these young African mahogany trees has been most
spectacular, especially when compared to the native Caribbean pines. How they
will continue to grow and what diseases and insect pests they may eventually
fall heir to remain te be seen in the future. The older trees have not
bloomed to date. In the summer cf 1940 one 3-1/2 year old tree put out a
large panicle of bloom but failed to set seed. In a conversation with Dr.

Wo Le Thompson in 1940, he stated that the trees attained a large size and
were about 20 years of age before they bloomed in Northern Rhodesia.

From their performance to date the Khayas appeas to be the most promis-
ing hardwoods for reforestation in South Florida that have been tested by the
Sub-Tropical Experiment Stetion. As there are many othexs species of African
mahogany belonging to the genus Khaya, to say nothing of the species of the
related genus Entandrophragma, this species can well be named “Rhodesian
mahogany".

= eh = April 1943

Summa ry

.

A recent introduction of Rhodesian mahogany, Khaya nyasica, into
southern Florida by the Sub-Tropical Experiment Station, Homestead, has
shown this species well adapted to the region.

Seeds of Khaya nyasica were brought from northern Rhodesia and were
planted on the Experiment Station property in 1930. Three separate plant-
ings were made which are now considered 9, 10, and 11 years old.

Measurements in these three plantations and in adjacent plantings of
Pinus caribaea indicate that the growth of Khaya is definitely superior to
that of the pine. Diameter growth was about twice as great and average
merchantable and total heights were nearly twice as great for the Khaya.

As this species produces a wood with most of the desirable ‘character-
istics of Central American mahogany, Swietenia sp., it would seem desirable
to plant it on little-used lands in southern Florida.

Resumen

la reciente introduccién de la caoba de Rhodesia, Khaya nyasica, en
el sur del estado de Florida por la Estacion Experimental Subtropical de
Homestead, ha demostrado que la especie se adapta bien a esa region.

En el 1930 se trajeron semillas de Khaya nyasica de la Rhodesia sep-
tentrional las cuales se sembraron en los terrenos de la Estacion Experimen-
tal formando tres plantios distintos que hoy tienen 9, 10 y 11 afios.

Al medir los arboles en estos tres plantios y hacer une comparacién
con plantios de pino adyacentes, se encontré que el crecimiento de la Khaya
es definitivamente mayor que ei del pino. El crecimiento en diaémetro fué
el doble y la altura total asi como Ja altura maderable fueron casi el
doble.

Por lo tanto, debido al hecho que esta especie produce una madera
con casi todas las caracteristicas deseables de la caoba centroamericana,
Swietenia sp., seria conveniente sembrarla en aquellos terrenos de poca
utilizacioén en el sur de Florida.

Eye)
Oo
q

Caribbean Forester = 1 Vol. 4, No. 3

RETENTION OF CREOSOTE OIL IN THE WOOD

OF PINUS OCCIDENTALIS swaRTZl/

_ §E. S. Harrar and D. G. Reid
Duke University School of Forestry

The Duke University School of Forestry in collaboration with the
United States Tropical Forest Experiment Station in Puerto Rico has in-
itiated a pregram of research dealing with the determination of pertinent
physicel-mechanical properties of important tropical American timbers. One
phase of this program is concerned with the behavior of certain of these
woods in the treating cylinder. This paper reports the results of a series
of o11 impregnation experiments using wood of Pinus occidentalis Swartz
from Hata .2/

Selected trees were felied and bucked inte four-foot lengths and then
sawn into flitches in accordance with A.S.7.M.3/ recommendations. After
ceding, the flitches were wired intc belt form and end-coated with tar to
minimize drying and checking during shipment. The flitches, upon their
arrivel. were sawn and dressed into pieces of standard dimensions used in
ascertaining strength data. The surplus heartwood was cut inte 3 in. by 3
in. by 24 in. columns, numbered, painted with a 2 per cent solution of mer-
curic chloride to prevent the activity of blue stain fungi, and then stacked
in open cribs for two months te permit drying.

Using Tippett’s random sample numbers, four sets of four pieces each
were drawn from the air-dried stock for use in the pressure cylinder. The
moisture content (12 to 16 per cent) of each piece was determined with a
Tag-Hepenstel resistance meter. Four expsrimental runs, simulating indus-
trial practices as nearly as possible, were made using No. 1 creoscte oil
with schedules as indicated in Table 1.

Tables. 2 and 3 present the observe physical data for each specimen
treated, together with calculated oil retention per cubic foot of material,

Bod Reprinted from Tropical Wocds No. 71, September 1942.

oa Acknowledgements are due Mr. L. R. Holdridge. Société Haitianc-Americaine
de Develeppement Agricole, Port-au-Prince, Haiti for his kindness in supplying
the timber used in this study.

af American Society for Testing Materials. Standard Methods of Testing Smell
Clear Specimens of Timber, D143, 1927.

Sipe April 1943

Table 1.—Treating schedules for each run.

Preliminary Preliminary
vacuum air pressure Oil pressure

Final vacuum

Pounds Pounds
Process | Run | Time per per Time Inches
in square sque.re in of
hours inch inch | minutes| mercury

Bethell ib L 28 | = = 4 200 5 26
Rueping 2 = = J 150 4 200 30 27
Bethel] 3 i 28 = = é 200 5 27
Rue ping 4 = = at 150 hie 200 30 27

Table 2.—-Specimen data, Betheli rocess o2/

To
Moi. scnre
Content

Ok
Perens ion
LOS o

Caleulated oil
CACEARBCIL ON,
Lpsof/euo £60

Airodry
wai gh
Lbs.

Run | Specimen |

Numbex

L 4 14 5o49 125 3077 30.16
1h 14 Boek okZD 5.55 28.40
17 ia 50e8 125 35 78 30.32
on ak 9.30 okB9 4518 55.04

3 25 \4 5.51 125 4.03 32 0e4
Aly 14 6.08 12d ooOT 28.56
18 12 5.67 o bed 3090 B12
fs) 4 6.0L L295 1.65 13.020

l/ Specimen 3 was encased between steel end-plates to minimize end penetration.

Caribbean Forester - 150 = Vol. 4, Now 3

Table 3.—Specimen data, Rueping process.1/

Calculated oil
retention

Volume of
specimen
cu. ft.

Moisture
Content

2 10 16 6.02 125 81 6.48
20 14 5.61 125 089 7.12

Us 12 5.98 ol25 278 6.34

14 12 4.83 0125 1.25 L0.00

4 25 LZ 5.49 oL25 2 59 4.72
6 13 5.93 225 066 5.28

8 1 6.06 oL25 2 82 6.56

13 12 4.95 ol2Z25 o 78 6.24

ia Specimen 4 was encased between steel end=-plates to minimize end penetration.

Results of this study indicate that the wood of Haitian pine will ac-
cept adequate amounts of oil using standard industrial practices. Specimens
treated by means of the Bethell process absorbed oil to the point of refusal
Suggesting that a shorter pressure pericd could have been used and that
Suitable penetration in materials of larger dimensions may be anticipated.
The average oil retention for the four pieces of Run No. 1 wes 30.48 lbs./cu.
ft. and that for Run No. 3, 26.32 lbs.feu. ft.; retentions far above those
specified as minimum by the american Wood Preserver's Assosiaticn, The aver-
age retention, however, for the two runs in which the Rueping process wes
employed, was only 7.46 and 5.68 lbs./eu. ft. respectively. These are mini-
mum and it is recommended that somewhat enese! pressure periods should be
used than those currently employed.

Resumen

Le Universidad de Duke en cooperacién con la Estecicn Experimenta)
de Selvicultura Tropical efectud un estudio sobre le impreguacién dal pins
de Haiti, Pinus occidentalis, con aceite combustible creosctado usando los
procedimientos Rueping y Bethell.

Los resultados de las diversas pruebas e!
jas tablas del texto indican que la retencidén ds 2
miento Bethell fue 4 6 5 veces mayor que la del ; procedimice

= Align April 1943

FORESTS AND FOREST ENTOMOLOGY

Luis F. Martorell
Agricultural Experiment Station
Rio Piedras, P. R.

(Presented at a joint meeting of the Zoology and Forestry Sections
of the S.A.C.A., Puerto Rico Chapter, Rio Piedras, April 4, 1942.)

The science of forest entomology had its beginnings in Germany, the
first country to develop forestry, during the first part of the nineteenth
century, and has since made great progress in many countries paralleling the
development of scientific forestry. In our small island very little has been
done on this phase of forestry although the protection of forests from destruc-
tion is a basic requirement. Fire, fungi, and insects are generally the most
destructive elements in forests. The first, fortunately is not important in
Puerto Rico, but our climate substitutes another direct source of forest de-
struction: the hurricane. Hurricanes have extensively damaged our forests, yet
no even approximate figures are available to show the extent of the damage and
little is known concerning the less obvious losses due to fungi.

It is the purpose of this paper to indicate some of the kinds of inju-
ries caused by insects to the forests of Puerto Rico. Injury due to insects
starts even before the tree seeds are planted. Many seeds are attacked by
granary pests during storage, and often by other insects while they are still
on the tree. Injury continues at every stage of the development of the forest:
when seeds are planted, as the seedlings. grow, later on when transplanted, and
finelly after a period of years develop into large forest trees. Injury by
insects does not cease when these are cut, the timber sawed, and then used by
men. In all this cycle some insects are in constant relation with the trees,
from the beginning to the end, from the small, tiny seeds attacked during
storage to the finished product.

To illustrate the great damage caused by insects to forest and forest
products, I give some examples:

In a survey made in California in 1931 the losses of merchantable timber
due to bark beetles in that year totaled about 1,250,000,000 board feet, or
nearly $3,000,000 in stumpage value.

The lodgepole pine forests of Idaho, Montana and Wyoming, particularly
those near Yellowstone National Park, have suffered losses during the ten years
ending .in 1932 totaling 7,250,000,000 board feet; more than 36,000,000 trees
having been killed in one National Forest alone.

The Planting Division of our Caribbean National Forest has given me the
following information on losses systained as a result of attack by a single
insect. The cedar shoot borer, Hypsipyla grandella Zeller, has been responsible
for the loss of 835,000 mahogany trees and nearly 1,000,000 cedars from our

Caribbean Forester - 132 = Vol. 4, Now 3

plantations since 1935. With a vaiue of 7-1/2 cents per tree the total loss
is in excess of $137,000. No control is known for the borer, which has pre-
cluded the pianting of cedars in Puerto Ricco.

Of the many urgent entomological problems of our Island, I present a
few:

Practical control measures for insect pests attacking seeds are rarely
practised. Methods of collection and storage of forest tree seed in order to
prevent destruction by insect pests must be developed.

No satisfactory methods for the control of such nursery pests as mole-
crickets, white grubs, curculionid iarvae ("vaquites"), ants, and cutworms
exist.

One of our common scale insects, Asterolecanium pustuians Cockereli,
commonly known as the pustule Scale, has been responsible for the nearly com-
plete eradication of the silver oak, Grevillea robusta Cunn., in Puerto Rico,
and now is attacking the yellow cassia, Sciacassia siamea (Lam.) Britton, a
species of value for firewood. One of cur beautiful cabinet woods, "maga",
Montezuma speciosissima Sessé & Moc., appears to be the next victim. Scale
insects were alsc responsibie for the nearly complete destruction of a stand
of an introduced ash (Fraxinus sp.} at El Guineo iake, in the Toro Negro Unit,
and of Cinchona plantings near Maricac.

The aceitille weevil, Apion martinezi Marshail, is important in that
it prevents successful propagation of aceitillo, Zanthoxylum flavum Vehl,
possibiy our most valuable cabinet wood. Nearly all the seeds are destroyed
by the larvae of the weevil.

The seeds of Neltuma juliflore (Sw.) Raf., our common mesquite are alse
subject to attack by three species of insects found in the pods, two of which
were recently described as new to science.

The termite is always troublesome, not only in our forests and in road-
Side plantings, but primarily attacking forest products. With the increased
importation of conifercus building woods as sitka spruce and southern pine,
and temperate zone hardwoods, all of which ere susceptible to attack by the
"“polilia"™ or dry wood termite, Cryptotermes brevis Walker, this insect has
increased rapidiy during recent years. "Comejenes" ocr moist wood termites are
also at work, attacking the beautiful Puerto Rican roadside trees. A flamboyda
tree, Delonix regia (Bojer) Ref., without a "comején" nest or tunnel is hardly
typical. We have become used to seeing them everywhere.

These are but a few examples. There are many more entomological prob-
lems related to forestry, and more will continually present themselves as
forestry and entomology progress in our American tropics. The warm regions of
the world present & more complicated study than the cold, for here the insects
thrive throughout the year, making control more difficult.

One of the first requisites for forest entomology is a sound basis of
silvics and silviculture. Not until one knows trees is he in a position to

. = Ge = April 1943

apply his entomological knowledge to them. No study of a tree species is
complete without a related study of the insects as well as diseases affecting
it in all stages cf its development.

Now that the practice of forestry is increasing in Puerto Rico and
will in future years become very important in South America, the study of
forest entomology should also be expanded. As in the very beginnings when
the work of the forester and the forest entomolcgist started together, their
relaticnship should be maintained by parallel development and cooperation.

I hope to see forest pathologists as well as entomologists working on
eur forest problems in the near future. Conditions are now abnormal, but
peace will come end with it the thoughts cf men will return to the preservae-
tion of natural resources.

Resumen

En Puerto Rico, los insectos del bosque y su importancia en la econo-
mia engi se han estudiade poco. En el futuro debe prestérsele m&s aten-
cién a esta se de la dasonomia pare prever los problemas relativos a los
insectos antes de que surjan en toda su intensided.

©

3

En los Estados Unidos #1 dafic que causan los insectos 41 bosque
asciende anualmente a millones de doleres. En Puerto Rico, el Hypsipyla
grandella Zeller ha destruido cientcs de miles de arboles de cedrc y caobe.
No se conece ningun método de combate effective pare esa plaga.

Es poco lo que se sabe acerca de los métcdos practices para combatir
los insectos que destruyen las semillas. El Asterclecanium pustulans Cockerell
ha atacado severamente y casi destruido algunas de nuestras especies forestales
mas importantes. Bl Apion -martinezi Marshall destruyea casi todas las semillas
del pee ey Zanthoxyilum flavum Vahi, haciendc casi imposible la propagacion

Los termes, particularments e1, Cryptotermes brevis Walker, atacen tan-
tas de las maderas locales que le mayoria no en utilizarse en la fabrica-
eiém de muebles.

Debido al enfasis ereciente que se le esta dando hoy dia a la praéctica
de la dasoncmia tante aqui come en todcs los paises de la América latina,
deben hacerse los preparatives pertinentes para que las investigaciones ento-
molégicas destinadas a resolver los ewer creados por los insectos sean
efectuadas conjuntamente con las précticas de selvicultura.

Caribbean Forester = 134 = Vol. 4, No. 3

THE IMPORTANCE OF RACE IN TEAK, TECTONA GRANDIS L.

J. S. Beard
Assistant Conservator of Forests
Trinidad and Tobago

Teak is one of the foremost timber trees of the East and is native to
monsoon-type forests (i.e. where there is a seasonal alternance of wet and dry
seasons) in India, Burma, Siam and Indo-China, Centuries ago it was introduced
into Java where it became naturalized, and was widely planted, so that in 1930
Java possessed 720,829 hectares of pure teak forests (1,800,000 acres). In
view of its peculiarly fine qualities, teak has been carried to all the cor-
ners of the tropics. It arrived in the West Indies during the nineteenth
century and old trees may be seen in various botanical gardens. It does not
seem to have been tried as an economic timber crop, however, until 1913, in
which year Mr. C. S. Rogers, Forest Officer, Irinidad, imported a small parcel
of teak seed from Tenasserim in Lower Burma. Two small plantations were es-
tablished in Trinidad which at the time of writing are just 30 years old.
These trees yielded seed, with which trials were extended. Such success re-
sulted that the species was taken up as a major commercial proposition and is
now being planted at the rate of some 400 acres annually. All the seed re-
quired for annual operations is obtained from trees which were raised from
seed collected in the plantations of 1913; in other words, present plantings
are in the second "creole" generation.

The teak thus being grown in Trinidad fall, according to sample plot
measurements, in a quality class slightly above the average of Quality II of
Nilambur, India, according to the Nilambur Yield Tables. They are subject to
no pests or diseases and are of excellent bole form.

In 1935 the late Professor R. S- Troup suggested that an investigation
should be made as to whether Trinidad’s stock was satisfactory or could be
improved by fresh importations of seed from selected strains. There was
reascn to believe that the seed brought from Tenasserim had been ccllected
from low, branchy trees in the paddy fields, though there was no evidence that
these trees were genetically distinct from true forest teak. A request was
therefore sent te the Silviculturist at the Forest Research Institute, Dehra
Dun, India, for selected samples of seed, and a reply was received as follows;

"T have asked the Silviculturists of Burma and Madras each to send
you 4 lbs. teak seed from their best moist teak localities. There
are well defined vegetative differences between the teak from these
two localities though inherent timber and bole form differences
have not yet been systematically sought for or discovered."

In June 1936, 4 lbs. of teak seed were received in Trinidad from Shen-
cottah Division, Travancore State (Southern India) and were scwn. No seed
arrived from Burma. Unfortunately most of the seed sent did not germinate
and for the planting season of 1937 (June) only about 150 plants were avail-
able- A plot containing 0.137 acre was however formed in the 1937 teak coupe

ob April 1943

at the Southern Watershed Reserve by the standard local method of stump-
planting at 6 x 6 feet. The plot is located at an altitude of 200 feet and
receives an annual rainfall of 65 inches. During the dry season from January
to April L/2 to 3 inches of precipitation falls monthly. During the rest of
the year it varies from 4 to 14 inches monthly. The soil is a silty clay
derived from a sedimentary formation. The aspect is southeast; the slope is
moderate; and drainage is good.

The plot is entirely surrounded by teak of the local stock, which will
be referred to as "Burma teak" in distinction to the "Indian teak".

At six years old the plot of Indian teak was thinned and measurements
were taken with the result shown in table 1.

Table l.—=A comparison of the growth of Indian and Burma teak; .

Criteria Burma Teakl/

Standing Crop after Thinning

Number of trees per acre 493 615
Average tree

Girth at 4 ft. 3 in. - inches 14.75 14

Total height - feet — 32 | 42
Average height of dominants - feet 36 47
Form factor 0156 0204
Bark per cent 53 30
Basal area per acre = square feet 46.0 48.9
Volume per acre under bark - cubic feet 250.0 $12.5

Intermediate Yield

Number of trees per acre 456 362
Average tree
Girth at 4 ft. 3 in. - inches 10.75 10.50
Total height = feet SL 38
Basal area per acre - square feet 2501 16.4

Volume per acre under bark - cubic feet 98.6 | 72.6

Total Crop

Basal area per acre - square feet 69.0 74.03
Volume per acre under bark - cubic feet 528.6 585.1

Mean Annual Increment

Basal area per acre = square feet 11,5 12.4
Volume per acre under bark - cubic feet 54.8 64.2

i/ Average at Southern Watershed Reserve.

Caribbean Forester - 136 = Vol. 4,\No. 3

The comparative figures shown were obtained by averaging the figures of
three sample plots in Burma teak on similar soil and situation within a radius
of half a mile from the plot of Indian teak. . These three sample plots agree
fairly closely in quality with each other and also, according to the measure-
ment of a few selected trees felled, with the teak immediately surrounding the
Indian plot. It is unfortunate that the latter is so very small: more accurate
measurements could have been taken if it were of larger area.

A heavier thinning is shown as having been made in the Indian plot.
This is merely in accordance with recent practice in Trinidad, the comparative
figures dating from 1935-57. It does not affect the results under consider-
ation. The Indian trees are markedly inferior in height growth, a whole qual-
ity class lower. In point of over-bark girth they are superior but this is
offset by the much greater thickness of the bark, bark per cent being 53 as
against 30 in the Burma teak. Total volumes and mean annual increment are
inferior in the Indian trees. Form of the Indian trees is very definitely
poor and must be inherent, for the immediately adjoining Burma teak are of
excellent form. Almost all stems are wavy, very few being absolutely straight,
and a fair proportion (12 per cent) had bent over towards the southeast, i.e.
down the slope, and towards the prevailing wind. Burma teak grown locally are
of very straight growth. Table 2 summarizes vegetative differences between
the two varieties.

Table 2.—Vegetative differences between Indian and Burma teak.

Character . Indian Teak Burma Teak

1. Length of internodes:

Young stem up to 7 ft. a oh BM a!
Branches 2-1/2" = 3" 2-1/2" - 4-1/2"
2. Stem Wavy Straight
5. Small branchlets Many Few
4. Leaf surface Shiny and smooth Rough
5. Leaf size
Sun leaves:
Petiole 1-1/2" at
Lamina 9"-10-1/2" x 4-1/2"-6" 20"-24" x 15"-16"
Shade leaves;
Petiole 3/4" Nil
Lamina gN=135" x 6-8" 20"-24" x 12"=-15"
6. Flush leaves Green Deep red-purple
7. Flowers and fruit Not examined Not examined

- 137 - Aprii 1943

An apparent greater density of the Indian teak is to be inferred from
a noticeably greater hardness. It is not yet possible to test relative dura-
bility as hardly any heartwood is as yet being formed.

The marked vegetative differences between the teak of India and Burma
referred to in the Silviculturist’s letter are thus established, and it is
evident that the two belong to markedly different "races". These are not
recognized botanically even as varieties, presumably because there is no vari-
ation in the flowers.

Race in trees has long been recognized in Europe, where trees such as
larch and pine in different localities are known to be of different growth
and form—hereditary differences which persist in a new environment. In Java
where much work has been done on teak, several races of teak are recognized.
Experimental plots have even been laid out to compare them. 1/ Seven strains
in all were comparede from Malabar, Godavari and the Javanese. Indian forms
were all found to be of bad shape with heavy branching, and those from the
north of India showed inferior height growth. Siam and Burma teak races were
found to be the best both in form and height-growth. Each race could be
easily distinguished by vegetative differences such as those noted in Trinidad.

The indications certainly are that Trinidad luckily obtained one of
the best races of teak in the preliminary trials. Had the first importation
of seed been made from Northern India, teak in Trinidad might well have been
judged a failure. Others in tropical America who contemplate trials with
teak should bear this in mind. Best results will be obtained from planting
in Burma or Siam, and seed should be obtained only from a reliable source.
There are a number of old teak trees in the Botanic Gardens in St. Vincent,
which appear to belong to an Indian race. When experimenting with a view to
starting afforestation two years age, the Agricultural Department of St.
Vincent imported planting stock (of Burmese race) from Trinidad instead of
collecting seed from these trees. The young trees show considerable promise
and are probably superior to those which would have been raised from the
trees already growing in St. Vincent.

Summary

Teak has been raised in Trinidad from seed imported both from India
and Burma. There are marked differences between the quality of the two
strains, and vegetative differences which divide them into separate "races".
In Java several distinct races of teak are recognized and trials have shown
those from Burma and Siam to be the best. Those who wish to introduce teak
in tropical America should be careful as to the source of their seed.

BVA See reports in "Tectona", organ of the Forest Service in Netherlands’
India.

Caribbean Forester = 138 = Vol. 4, No. 3

Resumen

El establecimiento de plantaciones de “teca" en Trinidad se inici6é en
el 1913 con la importacién de semilla del sur de Burma. Los arboles crecieron
bien y por lo tanto anualmente se siembran 400 acres de arbolitos obtenidos de
la semilla de los plantios més maduros.

Para determinar si la semilla que se estaba usando era la mejor que se
podia obtener se importé un segundo lote del sur de la India. Un estudio de
los plantios que se lograron con esta semilla demostré que es inferior a la
de Burma ya que el rendimiento es menor y la forma del drbol es deficiente.

Los resultados obtenides muestran que cualquier Departamento Forestal
interesado en la "teca" debe obtener su semilla preferiblemente de Burma y
Siam.

00a

LA CREACION DEL NUEVO BOSQUE EXPERIMENTAL CAMBALACHE

La Estacién Experimental de Silviculture Tropical en representacién del
Servicio Forestal de los Estados Unidos, firmé recientemente un acuerdo con la
Autoridad de Tierras de Puerto Rico para reservar un area de 660 cuerdas con
el propésito de establecer un bosque experimental. Se escogiéd un trecho situa-
do en la region caliza seca cerca de la costa norte entre Barceloneta y Arecibo.

El drea consta de colinas calizas de 200 a 300 pies de altura separadas

por pequefias dreas llanas. La precipitacion pluvial anual es de 50 pulgadas.

El bosque que alli existe ha sido talado ya, pero contiene atm algunos de nues=
tros &rboles de maderas mis preciadas aunque en su mayoria son pequefios. Entre
ellos encontramos a Montezuma speciosissima Sessé & Moc., Coccolobis grandifolia
Jacg., Bucida buceras L., e Hymenaea courbaril L. Un reconocimiento preliminar
indic6é que el bosque comprende un gran numero de especies, algunas de las cuales
se conocen hoy dia séle de nombre. |

Después que se hagan las provisiones neces@rias para la proteccién ade-
cuada y la ordenacién en banda el area ser& dividida en tres compartimientes:
(1) una superficie circular para estudiar el rendimiento, (2) una parte para
estudios de siembra,, crecimiento y silvicultura y por Ultimo (3) un area que
no ha de tocarse y que seraé dedicada a un estudio ecologico.

Los objetivos de las investigaciones en este bosque serén; (1) determi-
nar la utilidad y durabilidad de varias de las maderas que pueden obtenerse de
alli y que son relativamente desconocidas, (2) determinar los métodos silvico-
las que deben emplearse para encauzer las mejores especies, (3) determinar si
tiene valor prdécticc el introducir en este bosque varias especies preciosas, y
(4) determinar el crecimiento de los rodales y de las distintas especies para
tener una base con la cual pueda estimarse el m&éximo rendimiento anual constan-
te y poder preparar asi un plan dasocrdético detallado.

soe April 1943

THE LAS COBANITAS CAMPECHE PLANTATION

Frank H. Wadsworth
Tropical Forest Experiment Station

The chief forest problems of Puerto Rico have resulted from the clean-
ing of forest lands. Consequently efforts to reforest such lands have con-
stituted a major phase of the work done in the government-owned National and
Insular Forests.

One of the first plantings made in 1924, was with campeche, or logwood,
Haematoxylon campechianvum L. Though the early records of this plantation have
been lost as a result of the 1928 hurricane, part of the information was found
te be still available through the Forest Guard, Mr. Oscar Rivera, who was on
hand at the time the planting wes made.

‘The planting site was Las Cobanitas arroyo in the northern portion of
the Gudnica Insuler Forest in the arid scuthwestern part of the island. The
soil is a colluvial stony clay of volcanic origin. The slope averages 10
per cent, and drainage is good. The scil surface is loose and granuler in
texture o

The average armual precipitation at nearby Ensenada is 27 inches and
the temperature averagas approximately 80° F. The original forest was short,
deciduous, and rather cpen as @ result of the low rainfell and the protracted
dry season each year. The dominant species in the climax was evidently tecar,
Bucida buceras L.

The seed was imported from Mexico, presumably to start local production
of dyewood. The stock was raised in the Rio Piedras forest nursery. When
about 15 inches tall it was sent by train to the field. The trees arrived in
November, at the beginning of the dry season. The site, which was covered
with low herbaceous vegetation at that time, was prepared by plowing furrows
for the trees. Approximately 30 acres were planted with 20,000 trees, spsced
8x 8 feet. Early survival was good and no replanting or weeding was done.

The general appearance of the plantation at the present time, 18 years
after planting, is shown in figure 1. No thinning has been done, so it is
possible to determine the survival, which is approximately 75 per cent, giving
@ stand of 510 tress per acre.

The characteristic spreading growth of this species, as shown in
figure 2, makes measurement of growth difficult. Nearly one fourth of the
trees were forked at the ground level, and at the breast height point (4.5
ft.) the average number of stems per tree was 2.8. Probably the measure-
ment which has the greatest meaning is’ diameter above the butt swelling, as
at this point (6 to 12 inches above the soil) most of the trees have but one
stem. However, even at. this point error results because most of the stems

Caribbean Forester - 140 = Vol. 4, No. 3

’

are fluted and therefore present an irregular cross-section. The average
diameter of the trees at this point, as determined with a diameter tape in
a 0.37 acre plot is 7.0 inches, and the basal area is 138.2 square feet per
acre.

Fig. 1.—A general view of the 18 year old plan-
tation showing the spreading habit of the trees.

The growth of this plantation is slower than that in Jamaica .2/ Part
of the difference may result from the fact that this site is more arid than
any on which this species is found in that country. However, another factor.
adversely affecting growth, and one which is obvious, is stagnation. The
spreading habit of this species, so characteristic of many trees on dry sites,
indicates its preference for wider spacing than that of this plantation. The
branches are interlaced, little herbaceous cover is found beneath the trees,
and the basal area is undoubtedly very high for this site. Trees as large as
15 inches in diameter are found in openings.

The density of the canopy has prevented the abundant natural repro-
duction from growing beyond the sapling stage except in scattered openings.
These young saplings, found under many parts of the stand, do not appear
destined to survive unless they are released.

u/ Comparative notes were furnished by C. Swabey, Conservator of Forests,
Jamaica.

- 141 ~- April 1943

Pages

Es

‘"
‘a
<

Fig. 2.—An 18 year old logwood or cam=
peche tree, 8 inches d.b.h., showing the
fluted bole.

At present the stand is of no commercial value for dyewood, and in view
of the uncertain future demand for natural dyes, it is doubtful that the trees
should be appraised for other than their fuel value. For this purpose the
stand may be considered mature and should be cut either partially or complete-
ly, depending upon the objectives of management. The apparent stagnant con-
dition points to the need of a partial cutting to improve growth of the smaller
trees and the saplings. Such a cutting, removing half of the basal area will
tend to perpetuate campeche which is almost the only species reproducing on
the site.

If campeche is to be valued only for fuel either of two other species
should be artificially introduced to increase the yield of the site. Baya-
honda, Neltuma juliflore Sw., has invaded and is growing well on similar sites
nearby. It is a rapid growing coppice and produces excellent fuel and durable

Caribbean Forester - 142 = Vol. 4, No. 3

posts. Dominican mahogany, Swietenia mahagoni Jacq., has already been under-
planted in parts of the stand. This species grows well in the vicinity and
should provide more yield than the campeche.

Measurements of current growth and studies of reproduction and conver-
sion techniques are being planned as a part of the Station's research program
for the near future.

Resumen

En el drido Bosque Insular de Gudénica, en Puerto Rico, se establecié en
1924 una plantacién de campeche, Haematoxylon campechianum L. Los arboles se
plantaron en surcos y no se lies ha desyerbado ni aclarado desde entonces. La
supervivencia es de cerca de 75 por ciento y el promedia de crecimiento en
didmetro en la base es de 7.0 pulgades.

El rodal esté ahora apifiado lo cual parece haber influido adversamente
en el crecimiento. Los a4rboles no son le suficientemente grandes para usarse
en la fabricacién de tintes y como la demanda de madera para extraer tintes
esta decayendoc la plantacién debe valiorarse primordialmente de acuerdo con la
lefia que puede suministrar.

El campeche no parece dar un rendimiento en lefia tan grande como la
bayahonde, Neltuma juliflora Sw., debido a lo cual se tratard de supiantar el
campeche por esta especie en una parte de la plantacién. Come la caoba domi-
nicana, Swietenia mahagoni Jacq., crece bien aqui, otra parte del drea en

cuestion serd substituida experimentalmente por esta especie.

000

gQUE PUEDE LOGRAR LA REFORESTACION?

la reforestacion va més allé del mero heche de sembrar arboles o de
levantar una ccsecha forestall. Significa de por si el mejoramiento de la
tierra. Sirve (1} para mantener el suelo arenose fijo, (2} para evitar las
inundaciones, (3) para proveer con un refugic natural a la vida silvestre,
(4) para proteger las tierras inhabitadas del vienteo y la nieve, y por Glti-
mo (5) para devolver la fertilidad al terreno baldio formando la capa humi-
fera. Ya estos servicios tan titiles de los Arboles se afiaden las bellezas
del bosque y las oportunidades que ofrece para la recreacién saludable.
(Tomado de una conferencis. dictada en el N. Y. Botanical Garden por E. W.
Littlefield, miembro del State Department of Conservation.)

- 143 - April 1943

ests.

FOREST LAND ASSESSMENT POLICY IN JAMAICA

In the Empire Forestry Journal, Vol. 21, No. 2 (1942) appears an article
"The Development of Forest Policy in Jamaica” by C. Swabey, Conservator of For-
Obviously the article was prepared only after considerable study of

Jamaica's forest problem and deserves the attention of all foresters in Latin

America.

As many of the countries of the Caribbean, and particularly the Antilles
are now, or will be, fased with problems similar to those in Jamaica, a portion
of Mr. Swabey’s article dealing with forest land assessment considered of espe-

cial interest, is quoted because of its fundamental nature;

"It is not possible or desirable to lay down hard and fast rules as to
what areas should or should not be in forest; it is, however, possible to in-
dicate certain basic considerations which should govern every assessment;

1.

rae

In view of the high and increasing density of population it is essen-
tial that every acre in the island should be put to its highest use.

Land which is capable of permanent productive agriculture must be re~-
tained or developed as such, provided this can be carried out without
affecting disproportionateily water supplies, fiood control, etc.

In normal times the proximity of the North American continent permits
the importation into Jamaica of supplies of softwood at a reasonable
rate; there is no reason to suppose that there is a likelihood of
this source of supply becoming permanently unavailable. It is there-
fore, believed that there is no justification for maintaining forest
solely for timber production on land intrinsically suitable for agri-
culture. In this the Division is guided by the realization that
probably 25-30 per cent of the total land area of the island is un-
suitable for agrisulture, and it is on these areas that timber pro-
duction should be concentrated.

It therefore follows that forest areas should be confined:

(a) to areas unsuited to any form of permanent agricultural usage,
or to areas yielding higher returns under forest management,
than under agricultural management; and

(b) areas possibly suitable for agriculture, but which must be main-
tained under forest for protective purposes (water supplies,
flood control, etc.).

Finally, it cannot be too strongly stressed that the selection of
forest areas for protective purposes must be influenced by the agri-
cultural practices followed on adjoining lands. Protective forests
can only be of value if adequate soil and water conserving methods
are practised on the agricultural lands in the watershed."

Caribbean Porester - 144 = Vol. 4, No. 3

ae

REORGANIZATION OF FOREST SERVICE ACTIVITIES

IN TROPICAL AMERICA

All activities of the Forest Service of the United States Department
of Agriculture, which deal with research, administration and cooperation in
tropical forestry, have been reorganized and placed under a Director of
Tropical Forestry with headquarters at Rio Piedras, Puerto Rico. This
action brings together under one head the direction of the Tropical Forest
Experiment Station, the protection and management of the Caribbean National
Forest, the services rendered to the Insular Government in the administra-
tion and management of the Insular forests, and the forest resources pro-
jects being handled cooperatively with certain Latin-American countries.
Arthur T. Upson has been appointed by the Secretary of Agriculture to this
newly created federal position of Director of Tropical Forestry, and by the
Commissioner of Agriculture and Commerce of Puerto Rico to the position of
General Superintendent of Insular Forests.

The consolidation of the heretofore separately supervised forestry
activities, besides continuing the present programs of forest research and
administration, will aid the progress of tropical forestry in a number of
ways. The forest research program itself can be more closely correlated
with the needs of the protection and management of forest lands and the
utilization of forest products throughout Puerto Rico, whether those lands
be in private, Insular, or Federal ownership. The Caribbean National For-
est, besides being administered for the social and economic benefit of the
Island's people, can be used as a testing ground for the practical applica-
tion of results of forest research. Similarly the results of such research
will be employed on the Insular forests. The protection, management and
development of the two units making up the Caribbean Forest and the many
units making up the Insular forests, can be coordinated to the extent that
the difference in laws applicable to each will permit. The inclusion of
the general direction of the cooperative forestry projects conducted for
Latin-American countries is a still further step in the correlation and
integration of all elements of tropical forestry. Thus the Tropical
Forestry Unit should become the center for the interchange of knowledge
in the entire forestry field between the United States and its American
neighbors.

A graduate in forestry from the University of Nebraska, Mr. Upson
has had 33 years of experience in research and administrative and indus-
trial forestry. It comprised 21 years in the U. S. Forest Service, ll
years with the National Lumber Manufacturers Association, and one year
as Chief of the Lumber and Lumber Products Division of the War Production
Board. From the latter position he took up his duties as Director of
Tropical Forestry in May 1943.

CONTENTS

How to make wood unpalatable to the West Indian
Cryptotermes brevis Walker.

dry-wood termite.
I. With inorganic compounds ...
George N. Wolcott, Puerto Rico

Encina,- Quercus virginiana Mill. ..
Alberto J. Fors, Cuba

Apuntes sobre la Myrica cerifera L. de Honduras .

Luis Landa Escober, Honduras

La vegetation muscinale des Antilles frangaises

et son interét dans la valorisation sylvicole.

H. Stehlé, Martinique

Pomarrosa, Jambosa jambos (L.) Millsp and its

place in Puerto Rico... °
Frank H. Wadsworth, Piiseto Rico

°

°

°

°o

°o

e

145

158

163

164

183

HOW TO MAKE WOOD UNPALATABLE TO THE WEST INDIAN DRY-WOOD TERMITE,
CRYPTOTERMES BREVIS WALKER. I. WITH INORGANIC COMPOUNDS

George N. Wolcott, Entomologist
Agricultural Experiment Station
Rio Piedras, Puerto Rico

The scarcity of cld wooden houses in Puerto Rico is largely due to their
destruction by hurricanes and termites. Hurricanes are unpredictable and may
not occur for long periods of years, but the termites are present all the time,
and the damage they cause is steady and cumulative. Indeed, one may say that
they are even more certain than death and taxes. Of the numerous species of
termites occurring in Puerto Rico, the wet-wood termite, or "comején", which
constructs nigger-head nests and large carton tunnels from its next to points
of attack on wood, is comparatively little to be feared, for it can readily be
controlled by putting dry Paris green in its main tunnels or on top of the nest.
The various species of subterranean termites, have, until recently, been scarce
or of localized distribution, and methods of control have not been needed. The
West Indian dry-wood termite, Kalotermes (Cryptotermes) brevis Walker, locally
known in Puerto Rico as "“polilla", and generally present throughout the neo-
tropical region, is by far the most destructive and most to be feared because
of the insidious nature of its attack. It constructs no external nest, and no
external tunnels, but lives and feeds entirely within infested wood, giving no
outward indication of the injury it is causing. All the moisture it needs for
any stage of existence can be extracted from dry wood, thus it can live within
the timbers of houses kept dry by a roof, and in furniture, in books and stored
papers, and even in picture frames on the wall. As a commensal with man, it is
no country cousin, but essentially a city or town dweller, and as truly symbi-
otic with man as cats and dogs, cockroaches, bedbugs and silverfish, and, if
anything, even mors difficult to get rid of.

The Spanish have no word differentiating butterflies from moths, but the
minute clothes-moth (and the injury which its larva causes to woclen clothes)
is called "polilla". In the tropics, the little insect that causes a somewhat
similar injury to wood came to be called "polille de madera", and eventuall
simply "polilla". Even this does not begin to show all the changes that may
occur in a language in a new environment and confronted with new conditions to
be described, for as the clothes-moth is recognized (if at all) largely by the
damage caused by its lerva, which is a minute caterpillar, so the dry-wood
termite is recognized largely by the grains of its hard, lozenge-shaped excre-
ment, which is seen in little conical piles on top of infested wood, or loosely
scattered over the floor when dropped from an infested ceiling, and it alse is
called "polilla". Only well-informed people kmow the winged adults of the
polilia, which swarm about lights on humid nights in the late spring, breaking
off each others wings and seeming to be playing "follow the leader", with the
leader invariably trying to be last. By next morning, one recognizes a flight
of polilla of the night before by the piles of gauzy, irridescent wings, their
former owners having been eaten piecemeal by ants during the night, or swallowed
whole by pet house lizards, with a possibly fortunate few avoiding such an

- 145 = JULY 1943

untimely fate by escape under the cover of a book, or into cracks in furniture
or. woodwork. Figures 1 and 2 show a de-alate adult and a soldier.

Radios seem especially designed for becoming most easily infested by dry-
wood termites, for, when in operation, their lights attract the adults, and they
have any number of points of entrance for such small insects, in addition to the
often open and unfinished rear. The thin plywood covering of panels of expensive
cabinet woods merely deludes the purchaser, for the termites eat the cheaper
foundation wood inside, and finally make repeated holes through the plywood
covering to void their excrement, or in search of new worlds of wood to devour.
Pianos are larger and necessarily have even more places where termites can gain
access tc their interior. Pictures on the wall, especially those over a table-
lamp, are similarly very susceptible to attack, for the de-alate adults collect
where the picture back is in contact with the wall, and, if undisturbed, burrow
into the back and frame of the picture, eventually often eating the picture it-
self. Many a framed doctorate "sheepskin" is shown to be nothing but parchment
of vegetable origin when polilla start eating it.

Dust covers on books lift the cover just enough so that de-alate adults
find a secure refuge and prompt protection if the house-holder leaves them
lying about on tables. Daily inspection is the price of safety for books left
exposed during the flight periods of the adults in April, May, and June. News-
papers get carried away or destroyed, and are of temporary value in any cage,
but stacks of undisturbed magazines and correspondence on the living room table
will be found to harbor dozens of termites. The more expensive magazines,
printed on heavily coated paper, are considerably less subject to attack than
the "pulps", which present an almost undiluted food supply of cellulose for
termites to eat.

The problem of furniture is more complex, and some articles escape attack,
even if made of susceptible woods. Not even constant use and being moved about
prevents others from becoming infested, if some ready avenue of infection is
present. The angle of a rocker on the floor presents a perfect opportunity for
termites to burrow inside, and often the rocker becomes infested when the re=-
mainder of the chair is still sound. Bureau and desk drawers fit with just
enough clearance to permit termites to enter, and the sides, backs and bottoms
of drawers are almost invariably made of woods susceptible to termite attack.
Solid wicker, caning or rattaning gives perfect security against disturbance
when termites are starting their tunnels, and also serves to hide a termite-
susceptible piece of wood in an article of furniture in all exposed parts made
of a tropical cabinet hardwood resistant or immune to termite attack. (See
Fig. 3.)

All the common building woods are susceptible to termite attack, and
even if the walls of a house are of concrete and its floors of tile, anyorall
of its wooden members may become infested with termites. Even if the semi-
circular openings where corrugated zinc sheets used for roofing rest on the
plate are carefully closed to prevent the entrance of bats, mice and rats to
the attic, and the ventilators under the eaves screened, this still leaves an
abundance of entrances large enough for winged termite adults to get into the
attic, and, entirely undisturbed, begin their tunnels into roof beams and ceil-
ings of the rooms below. Hurricanes pull such weakly-attached zinc sheets off
termite-infested roof beams with ease, and to the ordinary dangers for anyone

Caribbean Forester - 146 = Vol. 4, No. 4

Fig. 2.—Soldier of Cryptotermes

brevis Walker, six times natural size.

Fig. 1.—De-alate adult
of Cryptotermes brevis Walker,

eight times natural size.

Fig. 3.—Destruction by termites of a susceptible rung,
which, covered with caning, forms part of the seat of a
chair whose other members are made of mahogany.

= inp July 1943

out in such storms adds that of being cut by the flying sheets of metal. Hur-
ricanes also feel out all the weak places in shutters and doors where termites
have been at work, and under the unusual pressure they give way and permit the
entrance of the full fury of the storm. Of course the termites suffer as much
from the hurricane as do the human inhabitants of houses, for they are specif-
ically dry-wood termites, and many die as much from exposure to rain as from
having their tunnels broken open when the wood gives way. Enough of them sur-
vive, however, so that there is no danger of extermination, and reconstruction
after the hurricane furnishes them with fresh supplies of uninfested wood to
attack. Is it surprising that the cost of living in the tropics in anything
better than a thatched shack of palm fronds proves to be absurdly high? .

Hundreds of years of experience have shown, however, that some native
tropical hardwoods are disliked, or unpalatable to dry-wood termites to such
an extent that articles made of these woods will survive from one generation
to the next as family heirlcoms, while those of inferior woods disappear in 4
few years. The well-earned reputation of West Indian mahogany is primarily
because of its durability in the tropics. A mahogany four-poster bedstead,
with canopy, is not only immune to termite attack, but its canopy protects
against termite excrement falling from the ceiling. These hard little pellets
are like so many grains of sand on a smooth sheet, and if one gets in your eye,
proves to be exceptionally irritating and hard to disiodge. Dry-wood termites
can and do eat mahogany (usually the sapwood, it will prove to be upon investi-
gation), but in practice its heartwood is so nearly immune to attack as to
constitute a standard by which one may judge of the value of other woods.
These statements apply only to West Indian mahogany, Swietenia mahagoni; not to
Philippine mahogany, Shorea negrosensis, nor to African mahogany, Khaya ivorensis,
the latter being very susceptible to polilla attack, and the former not much
better. Not one of the temperate zone hardwoods can be compared with real ma-
hogany in resistance to termite attack, although the hardness of the wood has
nothing to do with the ease with which it is eaten by termites. In this respect,
hard maple is no better than willow. . The soft coniferous woods commonly used
for building are also varyingly susceptible to attack, the exceptionally gummy
heartwood of cypress or southern pine being resistant only because of their im-
pregnation by gums. Dipping a sample of a susceptible wood in oil of cedar will
prevent its being eaten by termites, even though the amount of cedar oil natural-
ly present in red cedar heartwood will not prevent termites eating it, and the
white sapwood, so admired for contrast in cedar chests, is very susceptible to
attack. Many other substances besides the gums and oils naturally found in
woods serve to protect them against termite attack, and the number of commercial
remedies available for this purpose on the market is legion. Most of them are
"shotgun" remedies, intended to protect against rot and decay as well as against
termites, and not only against dry-wood termites, but also against wet-wood and
subterranean termites. Most unfortunately, substances and combinations able
effectively to do all this incidentally possess properties that are undesirable,
such as objectionable odor, stains that color the wood, or they leave a surface
that cannot be painted.

Such general purpose wood-preserving substances have their place and
value, but against Cryptotermes brevis, the dry-wood termite of the West Indies,
some of these supposedly essential characteristics are valueless. In all shot-
gun remedies, one requirement is that of deep penetration of the toxic or repel-
lent substance in solution, followed by its precipitation in insoluble form

Caribbean Forester - 148 - Vol. 4, Now 4

ieee Tg oe

ee ea

inside the wood, so that it will not subsequently be leached out when exposed
to moisture. Obviously, this is a needless requirement for any substance to
be used only to repel the attack of dry-wood termites, for the woods normally
subject to attack by them will never be exposed to the weather. Roof-beams,
ceilings, sash, doors, shutters, furniture, pictures, radios, pianos, books,
and magazines: all these normally exist under a roof, and we hope and expect
to keep them dry. With this requirement eliminated, the problem of protection
solely against dry-wood termites is greatiy simplified. Methods of testing
out the varying natural resistance or susceptibility of many woods to the
attack of this termite have already been worked outi,2/ and it is a simple
matter to test common inorganic substances on the basis of the toxic or repel-
lent element.

Termites that naturally live in a picture frame can quite as easily live
in a petri dish, if this is kept in the dark most of the time, and only exposed
to light for a few minutes when observations are being made. The termites ob-
viously do not enjoy exposure to light, but this constitutes a minute fraction
of their existence while being used in the experiments, and in practice has
little effect on their choice of woods, or survival, while it admits of ready
inspection at all times. To correlate with previous tests, samples of wood a
quarter of an inch thick, an inch wide and an inch and a quarter long were
used, the wocd selected being almdécigo, or West Indian birch, Bursera simaruba
(L.}) Sarg. a common tropical soft wood of little commercial value and very
susceptible to termite attack. For the preparation of a large number of these
samples, the writer is indebted to the Tropical Forest Experiment Station at
Rio Piedras.

West Indisn mahogany, Swietenia mahagoni, was used as the basis of com-
parison, and the tests conducted to determine whether alm&cigo when treated
proved to be eaten in preference to mahogany, or the untreated mahogany eaten
in preference to the treated almfécigo. Other woods, such as gummy southern
cypress heartwood, Taxodium distichum; lignum-vitee, Guaiacum officinale; maga,
Montezuma speciosissima; aceitille, Zanthoxylum flavum; cobana negra, Steahlia
monosperma; acacie, Albizzia procera; and swamp-seasoned mangrove, all sub-
stantially or more than equal to mahogany in resistance to termite attack,
filled out the space in the petri dish, so that all pieces fitted together
tightly and did not shift about to injure the termites. It is obvious that
the termites were presented with a real opportunity to make a chcice of what
they would eat: @ normally unpalatable wood or 4& treated very susceptible wood;
very decidedly not the kind of choice presented to the American public as to
the color of their mcdel T Ford. Cryptotermes is not much of & builder, but
it promptly accepts the conditions as it finds them in the petri dish, and uses
most of its excrement to cement some of the samples to each other, or to the
dish. Admittedly conditions are abnormal, but even with only mahogany or less
palatable woods and treated almécigo to eat, some of these treated samples most
cf the time would prove to be more palatable and then the termites would have a
real meal, especially once they had eaten inside the treated crust.

dy, Wolcott, GN. The comparative resistance of woods to the attack of termite,
Cryptotermes brevis Walker. Bull. #33, Insular Experiment Station. Aug. 1924.
7) Wolcott, G. N. A list of woods arranged according to their resistance to
the attack of the "polilla", the dry-wood termite of the West Indies, Crypto-
termes brevis Walker. Caribbean Forester 1 (4): 1-10. July 1940.

See) = July 1943

It should be emphasized that the purpose of these tests was not to dis-
cover the substances toxic to the termites, but merely those which were suffi-
ciently repellent so that the termites would not eat the wood treated with them.
It happens that the elements toxic to termites (thallium and arsenic) are not
repellent to them in small amounts, and before the reaction on the termites
treated with such substances became known, and the sampies could be removed,
hundreds of the termites died, and the experiments in unpalatability were thor-
oughly disorganized. A few of these poisoned termites continued to live for
weeks afterwards (after being removed from the regular tests), too sick to gat
@ little more and die promptly as other members of the coleny had done, but
showing no tendency to throw off the effect of the poison and become normally
active.

The tests were started with a saturated solution of copper sulfate, in
which samples were dipped six seconds, one minute, ten minutes, one hour and
forty minutes, and finally sixteen hours and forty minutes. All of these
samples proved to be sce repellent that the termites would not even approach
them. After discussion with the plant physiclogist, Dr. E. F. Hopkins, a
standardized procedure was adopted. For testing the value of each element, the
chloride was used by preference because of usual availability and solubility in
water, but in many cases, and also as check, the acetate, nitrate and sulfate
might also be used. For elements of which these compounds are not possible,
the salt of sodium or potassium was used, the potassium or sodium affecting
palatability so little that 5 per cent of these alkaline earths as chlorides
were apparently unnoticed by the termites.

For preparing the test samples, a solution was prepared which contained
a 2 per cent solution of the element to be tested, samples dipped, the solution
diluted to 1 per cent of the element and samples dipped, and a final series of
samples dipped in a ©.5 per cent solution of the element. Treatment by dipping
was restricted to three periods: the six seconds dipped samples were kept in
constant motion to dislodge air bubbles and favor even penetration of the
solution, and the ends of the one minute and ten minutes dipped samples were
reversed in the middle of the period. All samples were dried in the sunshine
on & window ledge, and not tested until the following day, by which time they
were thoroughly air-dry. Before treatment, each sample was marked with india
ink as to the kind of wood, length of treatment, per cent of element and kind
ef compound, such &n extensive legend being very useful later in determining
just where and how much the termites had eaten, if the attack was slight, for
they completely disregard the legend of india ink, and eat it or leave it
entirely on the basis of the wood and treatment underneath. Of ccurse this
method of precedure does not give a definite concentration of the chemical in
the wocd, its strength varying from the outside where it is most concentrated

and quickly tapering off to nothing in the interior of the sample. In practice,

however, it appears to work well, with complete consistency as indicated by the
reactions of the termites.

The sample dipped six seconds was placed in a petri dish with other woods
somewhat less attractive to Cryptctermes than untreated alm&écigo; that dipped
one minute was placed with woods considerably less attractive; while only the
ten minutes dipped sample was tested against mahogany and similarly resistant
woeds. Practically, the ten minutes dipped samples vs. mahogany constituted

Caribbean Forester ~ 150 = Vol. 4, No. 4

the real test, but the others were useful checks when the treatment had little
or no value and all treated samples were promptly attacked, as well as when they
were effective and all the treated samples were untouched by the termites. No
set time was required by the tests. If the termites had begun to eat the treat-
ed samples by the first day, and the attack was unquestionable by the second day,
that was ample time for a positive test. On the contrary, the treated samples
that proved to be almost the exact equivalent of mahogany in attractiveness re-
quired a longer time; while somewhat less palatable samples might remain in a
test for a week or longer, and be put back again at a later date, to make sure
of the termites’ attitude towards them. After all eaten samples had been elim-
inated, all those remaining were finally tested against each other, alternating
with mahogany and other less palatable woods, using fresh termites that had

been just taken from the original colony and had previously eaten only sycamore,
Platanus occidentalis. All of the termites in all the tests came from this same
colony that had developed in a box that originally contained two five-gallun oil
Cans «

If the termites ate the 2 per cent concentration, the other samples were
valueless, and further tests with that element could be abandoned as requiring
too large an amount of the chemical to be practical for commercial use. On the
contrary, if the 0.5 per cent solution treated sample was uneaten by the ter-
mites, the solution could be further diluted to 0.2 per cent, 0.1 per cent, and
0.05 per cent, and samples dipped at each concentration for additional tests.
By continuing the tests, it was possible to determine the minimum concentration
at which the treated alm&cigo samples were less palatable than mahogany, and
the next dilution at which the almdcigo was preferred to the mahogany. The
final results appear as a table on the following page—the two concentrations
on opposite sides of the mahogany dividing line.

In the table as given, there appears to be no consistency by our stand-
ards as of what elements the termites are conscious. Of course they are pre-
sented with a problem with which they or none of their ancestors have had any
experience, for it would never occur in nature, yet it seems inexplicable that
ferric iron should be repellent, and many times as much of the chemically
similar cobalt, nickel and manganese (and chromium at less than 2 per cent
strength) be apparently imperceptible to their senses. Why should silver and
tin be imperceptible, and copper and zine repellent? When placed in the peri-
odic table, four repellent elements: zinc, cadmium, barium, and mercury are
in the same group II, but this makes even more inexplicable why the interme-
diate strontium is imperceptible. Indeed, aside from this, the reactions of
the termites appear to be largely or entirely determined by the specificity
of the organism.

In the selection of elements to be tested, those whish are so rare as
to be unavailable in the chemical laboratories of the Experiment Station (Mr.
F. J. Ramirez Silva of the Plant Phisiology Department), the University (Prof.
Osvaldo Ramirez Torres), and the Institute of Tropical Agriculture (Dr. J. A.
Bonnet) were not considered in the preliminary stages of the investigation,
and it is quite possible that the results of tests with these rare elements,
locally unavailable, might considerably change the picture. From the practi-
cal standpoint, however, a sufficiently large number of elements, cheap and
abundant, are sufficiently repellent to the attack of the dry-wood termite as

- 151 = July 1943

The relative value of various elements as repellents to the attack
of the West Indian dry-wood termite, Cryptotermes brevis Walker

West Indian mahogany eaten
in preference to almdcigo
treated ten minutes with:

0.15% Copper (as sulfate)
0.2% Copper (as chloride)

Mahogany
LINE

0.2% Zinc (as sulfate, chloride

and nitrate)
0.2% Mercury (as chloride)
0.3% Barium (as chloride)
0.3% Cadmium (as nitrate)

0.2% Ferric Iron (as chloride)

0.4% Ferric Iron (as sulfate

and nitrate)

0.5% Aluminum (as sulfate)
1% Antimony (as chloride)
1% Lead (as nitrate)

2% Boron (as sodium perborate)

Caribbean Forester

2% Chromium
(as chloride)

2% Fluorine
(as potassium
fluoride)

- 152 =

Almécigo treated ten minutes,
as below, eaten in preference
to West Indian mahogany.

0.1% Copper (as sulfate)

0.1% Copper (as chloride)

0.1% Zine (as sulfate, chloride,
and nitrate)

0.1% Mercury (as chloride)

0.2% Barium (as chloride)

0.2% Cadmium (as nitrate)

0.1% Ferric Iron (as chloride)

0.2% Ferric Iron (as sulfate and

* nitrate)

0.3% ARSENIC (as sodium arsenate)
toxic, not repellent

0.2% Aluminum (as sulfate)

0.5% Antimony (as chloride)

0.5% Lead (as nitrate)

1% Boron (as sodium perborate)

1% Chromium (as chloride)

2% Chromium (as sulfate and as

potassium dichromate)
1% Fluorine (as potassium
fluoride)

2% THALLIUM (as acetate)
toxiz, not repellent
2% Ferrous Iron (as sulfate)
2% Silicon (as sodium metasilicate)
2% Tin (as chloride)
2% Iodine (as potassium iodide)
2% Bromine (as potassium bromide)
2% Manganese (as sulfate)
2% Cobalt (as chloride)
2% Nickel (as chloride)
2% Silver (as nitrate)
2% Strontium (as chloride)
2% Uranium (as acetate)
2% Thorium (as nitrate)
2% Cerium (as sulfate)
2% Calcium (as chloride)
2% Magnesium (as chloride)
5% Sodium (as chloride)
5% Potassium (as chloride)

Vol. 4, No. 4

a ee ae

to furnish a considerable choice as to what will be selected in practice for
treating wood or wood products to be made immune to termite attack. It now
becomes apparent why heavily coated paper, in the manufacture of which various
compounds of aluminum are used, should be more resistant to termite attack, and
the possibility of making paper entirely immune seems to be quite possible if
the amount of this element (or others) is increased somewhat.

Translating the amounts of elements of positive value as shown in the
table, into pounds of compound per gallon of water, gives as the minimum pro-
tective concentration approximately 3/4 ounce of copper sulfate, zinc chloride,
barium chloride, and cadmium nitrate per gallon of water. This is such a small
amount of copper sulfate in solution that its presence is almost imperceptible
because of its color, while the other solutions are colorless. Ome-third of an
ounce of mercuric chloride per gallon of water is completely protective, but
over an ounce is needed of ferric chloride. From the practical standpoint,
mercuric chloride, despite its greater efficiency, is at once ruled out because
the price per pound wholesale is in dollars, while that of the other substances
averages around 5 cents, with barium chloride the cheapest of this group. Alu-
minum sulfate is even cheaper, but aluminum being a very light element requires
twelve to fifteen times as much of its compound to equal in effectiveness the
heavy metals, and its use in repelling termites would be indicated as prefer-
able only on account of its other desirable properties. The otner substances
are either commercially unavailable, or tcoo expensive for use unless for some
other special value. Certainly the cost of the materials for making wood or
wood products repellent to termite attack is negligible by comparison with the
benefits obtained.

The possible combinations of repellent substances is infinite, but no
obvious advantage is anticipated, at least so far as the termites are concerned.
Only a few combinations were tried: copper with chromium, zinc with chromium,
and zinc with copper, and in no case did the combination prove more repellent
than a single chemical, at comparable strength. Because of unavailability,
such chemicals as the dichromates, borates, and fluorides of copper, zinc, and
barium, which contain two repellent elements and should theoretically therefore
be doubly repellent, were not tested.

The time and strength of application as given made the wood sample u-
palatable to the termites, but not repellent to the extent that they would not
rest upon it or walk over it. By somewhat or considerably increasing the con-
centration of the chemical, the wood is made actually repellent to the termites,
but this seems an unnecessary precaution, except possibly in special cases. It
should be emphasized, however, that the reactions of the termites applying to a
minimum amount of an element applied under optimum test conditions might well
be considerably modified to give a somewhat greater safety factor of time or
concentration, or both, for commercial application. Presumably depth of pene-
tration in each kind of wood is proportionate to length of time of application,
and should be long enough so that the wood is thoroughly wetted. This may not
be essential, however, if the cracks and crevices where the termites are most
likely to seek entrance are given special attention. Indeed, flat underside
surfaces may be merely brushed over with a more concentrated solution, which
in the case of copper sulfate on wood not readily penetrated forms a thin
crystalline film, like frost on a windowpane, that is actually repellent to any

= 153 - July 1943

termite happening to alight upon it. Of course, in the manufacture of paper,
wallboard, and lignin plastic, all the wood fibers are finely divided, and pos-
sibly in such cases the minimum amount of chemical can be used with reasonable
safety. In soaking logs for making them ready for cutting into sheets or
panels of plywood, the liquid used for this cooking might well contain a mini-
mum amount of appropriate chemical, or the flitches dipped or sprayed with a
stronger solution before being placed in the press. In the construction of
furniture, the manufacturer should use a considerable margin of safety, either
in the strength of the solution or the length of time for which susceptible
parts are to be soaked (and dried) before assembly in the completed furniture.

Anyone in Puerto Rico, or in the Caribbean area, purchasing furniture
known to be made of a susceptible wood, in whole or in part, can protect it
against termite attack by @ suitable treatment. It is rarely possible to dip
the entire article of furniture in @ liquid, and the results might be disas-
trous for some kinds of plywood. It is quite possible, however, to paint or
spray the underside and hidden parts of new furniture with a strong solution
of & repellent chemical, making an especial effort to get into the cracks and
crevices where termites are most likely to seek shelter. This should not af-
fect the varnish or other finish applied by the manufacturer, for it is not
through such parts that termite attack is at all likely. This is not because
the paint, varnish or other finish applied by the manufacturer is repellent to
the termites, for it is not, but these painted or varnished surfaces will not
be attacked by the termites from the front, where you can see them, but from
the hidden back and inside, and it is these parts which must be treated if the
infestation by termites is to be prevented.

Two methods of procedure are possible in preventing termite infestation
of attics in new houses. When the house is being built, all lumber used in
permanent construction should be painted with a solution of the chemical after
it has been cut to fit, and preferably the ends soaked in a solution of the
chemical for some hours, or overnight, before being nailed into place. All
holes left by pulling out nails must be filled with putty, or better, the
nails driven farther in so that they will not be removed. Each and every
nailhole is an invitation to termites to enter, all the more enticing if all
the remainder of the exposed wooden surface is repellent. The other method
consists in spraying the interior after completion, being especially generous
with the liquid in corners and at the eaves, and in all crevices and nailholes
where the termites are likely to seek shelter. It is effective only as the
person doing the spraying thoroughly understands the habits of termites and
takes suitable precautions to outwit them.

When entire structures, such as houses and army barracks, are constructed
of wood, complete protection with water-soluble chemicals cannot be anticipated
for parts exposed to weather. Furthermore, the wear and tear of constant use
will wear away the thin external layer of repellent wood, and the mopping of
floors gradually dissolve away the repellent chemical, Actually, the surfaces
most subject to wear and tear are those where termites would have least chance
of entering, and in practice, thcrough treatment of all members cut to fit in
the process of construction, or spraying the interior of the completed struc-
ture will at least very considerably reduce the danger of infestation. A
discussion of more effective methods and chemicals to be used in such cases

:
Caribvean Forester ~- 154 =| Vol. 4, No. 4 j

is outside the purpose of this paper, as is also the treatment of houses or
barracks already infested. For practical purposes, they seem a hopeless pro-
position. Replacing the worst infested members with new wood only temporarily
postpones eventual total depreciation. Treatment of the new wood is of course
desirable, but almost invariably some small colonies of termites will remain
in the apparently uninfested wood of the original structure, and after the
repairs have been made, will continue tc eat and spread destruction.

The case for infested furniture is somewhat different. For those vindi-
cative persons (this includes the writer) who are satisfied with nothing less
than the destruction of the termites in infested furniture they already have
had for some time, one can suggest with caution the 0.3 per cent @rsenic solu-
tion, of which somewhat more than a gram of sodium arsenate will make 100 cc
of solution in water. (Thallium is also effective, but the quantity required
is several times as great, making it much more expensive, and it is by no means
so readily available as is arsenic.) First, with a sailmaker's or upholsterer's
needle, one should open up the carefully plugged holes through which the ter-=
mites have voided their excrement, and to each such hole apply one drop of the
solution with the medicine dropper. The major lesions in the wood where the
external crust left by the termites has broken through may also be used for &p-
plying the solution, but most of it should go inte the smaller holes, some of
which will accept considerable amounts, if it is administered gradually. This
renders the wood permanently toxic, not alone to the termites present in it,
but to any others that may attempt to re-colonize the deserted burrows, and
incidentally it is also toxic to domestic animais cther than termites that may
chew the poisoned wocd.

Summéry

The West: Indisn dry-wood termite, Kalotermes (Cryptctermes) brevis
Walker, is possibly the most important species in Puertc Rico because cf its
ability to iive in dry wood within houses, and bec&use it does not construct
@n exposed nest or tunmneis, therefore being very difficult to contre.

The adults are frequently seen flying about lights at night during the
late spring. During this pericd those which are not killed by their many
enemies crawl beneath or behind some article of furniture or pile of papers
or books and start tunneling into the material. Radios and picture frames
sre frequently attacked, as are large pieces of furniture such as bureaus,
desks, and pianos, end aiso parts of houses constructed of susceptible woods.
Gradually the tunneling comsumes the insides of the various wocden articles
until they are mere sheils without strength.

Experience has shown that some wocds sre disliked or are unpalatable
to dry-wood termites. A notable exemple is West Indian mahogany, Swietenia
mahagoni. Other more susceptible woods are sometimes treated with various
solutions to reduce their susceptibility to both termite attack and decay,
but some of these "remedies" are not entirely satisfactory, as the appearance,
or some other character of the wood is adversely affected. As dry-wood ter-
mites attack wood which is not generally exposed to the rain, a satisfactory

IMSS) tS July 1943

chemical, repellent need not be insoluble in water and the treatment may be
shallow in its penetration.

Tests were made with about 40 of the more common inorganic chemical
compounds to determine which elements would most effectively repel termites
when deposited on the wocd in aqueous solution by dipping treatment. Small
samples of the very susceptible wood of Bursera simaruba were dipped in solu-
tions of various strengths for various periods of time and were then made
available to the termites, along with untreated mahogany and other woods of
knowm susceptibility for comparison. It was found that samples dipped for 10
minutes in a solution of approximately one-half ounce of either copper sulfate,
zinc chloride, barium chicride, or cadmium nitrate per gallon of water became
more resistant to termite attack than mahogany.

The practical importance of this finding is obvious. Treatment of new
furniture and other wooden articles by dipping or by applying the chemicals
with a brush previous to painting or varnishing will provide cheap protection.
The added finding that arsenic is toxic but not repellent leads to the sug-
gestion that the introduction of a 0.3 per cent arsenic solution into the
tunnels of infested wood might not only kill the present colony but prevent
reinfestation.

Resumen

Puede decirse que el termes de las Indias Gccidentales, Kalotermes
(Cryptotermes) brevis Walker, insectc que corroe Ja madera, es el). mds dafiino
de Puerto Rico ya que vive en el interior de las casas y las galerias o ttme-
les que construye no se ven superficialmente lo cual hace sumamente dificil
su exterminacién.

Ya avanzada le primavera los termes adultos se ven con frecuencia vo-
lando alrededor de las luces durante la noche. En este periodo los que no son
matados por sus muchos enemigos se deslizan debajo de cualquier mueble, en les
libros y papeles y empiezan a construir sus timeles. Atacan tanto los marcos
de los cuadros y los aparatos de radio asi como los muebles m&s grandes, como
escritorios y pianos y atm aquellas partes de la casa que estén hechas de ma-
deras susceptibies. Gradualmente esos tuneles van corroyende el interior de
los objetos hasta que se desgranan.

Por experiencia sabemos que hay ciertas maderas que no son del agrado
de estos termes. El ejemplo tipico es la caoba de las Indias Cccidentales,
Swietenia mahagoni. Puede reducirse la susceptibilidad de otras maderas tanto
al ataque de los termes como a la podredumbre si se tratan con ciertas disolu-
ciones pero algunos de estos "remedios" no son del todo satisfactorios porque
alteran adversamente el aspecto y otros caracteres de las maderas. Debido al
hecho de que los termes atacan la madera cuando esta seca, lejos de la intem-
perio, no es preciso que el reactivo quimico que se use sea insoluble en agua
y el tratamientc por lo tanto puede ser superficial.

Se hicieron varias pruebas con cerca de 40 de los compuestes inorgéni-
cos mas comunes para determinar qué elementos al ser aplicados en las maderas

Caribbean Forester - 156 = Vol. 4, No. 4

A i I ae

en disolucién acuosa por el procedimiento de sumersi6n eran capaces de repe-
ler més efectivamente a los termes. Pequefias muestras de la madera Bursera
simaruba, una de las m&és susceptibles al ataque de los termes, fueron sumer-
gidas en disoluciones de distintas concentraciones por diversos periodos de
tiempo y se sometieron luego al ataque del insecto junto con otras muestras
de caoba y de maderas de susceptibilidad conocida, que no habian sido tratadas
previamente y que habrian de servir como indice de comparacién. Se encontré
que las muestras que habian sido sumergidas por 10 minutos en una disolucién
de aproximadamente 1 onza de sulfato de cobre, cloruro de cinc, cloruro de
bario o nitrato de bario por cada galén de agua eran més resistentes al ata-
que de los termes que la caoba.

La importancia prdctica de este hallazgo es obvia. Al fabricar objetos
de madera, pueden sumergirse o puede aplicdrseles cualquiera de estos reacti-
vos con una brocha, antes de pintarios o barnizarlos, logrando asi un metodo
de proteccién barato- Como también se encontré que el arsénico aunque no lo
ahuyenta es sumamente téxico al insecto, llegamos a la conclusién que si se
afiade una disolucién de arsénico al 0.3% y se trata la madera infestada no
sélo se exterminaré la colonia que alli exista, sino que se vera libre de fu-
turos ataques.

oOo

PLANTAS QUE SUMINISTRAN TANINO EN EL AFRICA ORIENTAL

En un articulo publicado en el Vol. 39 (3) del Boletin del Imperial
Institute, el Sr. P. J. Greenway, botdnico sistem&ético del East African Agri-
cultural Research Station, Amani (Tanganyika) expone un compendio de la infor-
macién existente sobre las plantas que se usan para la extraccidén de tintes y
tanino.

Este cientifico indica que el tanino se usa poco en el Africa oriental
pero que existen alli dos fuentes comerciales de dicho producto: la corteza de
Acacia decurrens Willd. var. mollis Lindl y la corteze de los mangles Bruginers
ceriops y Rhizophora. La lista de plantas que él somete incluye notas que es-=
pecifican la cantidad y la clase de tanino que se puede obtener de las distin-
tas plantas, la distribucién geogrdéfica y los diversos usos dé dichas plantas.
Muchas de éstas se encuentran comunmente en el Hemisferio Occidental, entre
ellas las que siguen:

Por ciento de tanino y

Especie localizacion en la planta
Avicennia nitida Jacq. Corteza, 12%
Cassia fistula L. Corteza, 18%
Casuarina equisetifolia Corteza, 15-18%
Haematoxylon campechianum L. Madera, 6%
Hymenaea courbaril L. Corteza, 19.2%
Mangifera indica Corteza, 16%
Pithecolebivm dulce Benth - Corteza, 25%
Prosopis juliflora DC. Corteza; raices, 6-7%
Psidium gua java L. Corteza, 30%; hojes, 9-10%
Terminalia catappa L. Corteza, 11-23%: frutas, 6%

- 157 = July 1943

ENCINA, QUERCUS VIRGINIANA MILL.2/

Alberto J. Fors
Ing. Encargado de Viveros Forestales
Habana, Cuba

Notas Generales

La encina es un 4rbol corpulento pero de poca altura. Su tronco es

corto y grueso. Su copa es amplia o irregular y esta formada por gruesas y
tortuosas ramas. Corteza gris, profundamente agrietada. Hojas pequefias,
duras, de color verde claro, opace. Sistema radicular amplio y profundo. En
suelos estériles crece lentamente y adquiere poca altura. En suelos fértiles
crece rdépidamente, pudiendo llegar a la categoria de arbol grande. Sembrada
en las mejores condiciones, alcanza la altura de 1 metro al afio; 1.50 metros
a los 2 afios; 4 metros a los 5 afios; 15 metros y 6 pulgadas de didmetro a los
10 afios. ;

La encina es nativa. Existe naturalmente también en Virginia, Florida
y sur de Texas. En Cuba se le encuentra en la mitad occidental de la provin-
cia de Pinar del Rio y norte de Isia de Pinos, en cuyas localidades, ocupa
los suelos siliceos y silico-arcillosos de reaccién dcida, secos o bien sanea-
dos y de fertilidad escasa. Especie intolerante y dvida de sol, se encuentra
generalmente formando montes homogéneos, o mezclada en proporciones variables
con pinos, Pinus caribaea y P. tropicalis; granadillos, Brya ebenus; y alguna
otra especie de follaje claro. No vive en el monte denso de los suelos cal-
céreos fértiles donde predominan las especies de copa amplia y tupida. Los
mejores ejemplares de encina de cualquier edad son los que se encuentran en
las faldas de las lomas, o en terreno lJlano de fertilidad apreciable, en mez-
cla con el macurije, Matayba sp.; el pino macho, Pinus caribaea; el encinillo,
Pithecolobium obovale; y alguna especie de menos importancia, indicando que
se beneficia con esta asociacién. En las cuchillas estériles, donde la roca
madre ha sido expuesta en muchos lugares por erosion, el encinar es homogéneo
y de poca altura.

La encina es monoica. Comienza a florecer desde marzo. Las flores mas=
culinas en amentos que presenta en pequefios racimos. Las femeninas solitarias.

1/ No empleamos la palabra "“roble" como seria lo correcto, ya que en Cuba
este nombre se aplica a diferentes especies maderables que no pertenecen 4
las fagdceas ni tienen semejanza alguna con nuestra encina. Tales son:

Roble amarillo, roble dulce Citharexylum fruticosum L.
Roble australiano Grevillea robusta Cunn.

Roble blanco, roble de yugo Tabebuia pentaphylla Hemsl.
Roble caiman Ekmanianthes actinophylla Urb.
Roble de olor ' Macrocatalpa punctata Brit.
Roble de guayo Petitia domingensis Jacq.
Roble prieto Ehretia tinifolia Lin.

Caribbean Forester - 158 = Vol. 4, Noo 4

El fruto es una bellota sostenida en una cipula caracteristica. Madura de
octubre a diciembre. La recoleccién se hace en el suelo. Las bellotas son
de forma subcilindrica y tamafio muy variable: 2 a 4 cm. de largo por la 2 de
didmetro. De 250 a 300 bellotas pesan 1 kilo y una medida cibica de 1 litro
puede contener 200. La semilla fresca, recién colectada, comienza a germinar
a los 10 dias y puede alcanzar una germinacidén de 85 por ciento en 20 dias.

La germinacién es hipogea y comienza con un pivote carnoso y profundo
de cuya parte superior brota una pldntula de aspecto delicado y hojas alternas
iguales a las del arbol adulto. Se trasplanta sin riesgo entre marzo y abril.

La encina se reproduce espontdneamente con notable abundancia. Los
a&rboles, plenos de sol, fructifican copiosamente desde edad temprana y la es-
pecie se establece y persiste a pesar de tales factores adversos como la Agri-
cultura, los fuegos de las sabanas y los animales domésticos.

Selvicultura

El encinar puede ser regenerado por métodos naturales. Bajo las con-
diciones cubanas, se sigue el método de selecci6én pura, extrayendo del monte
solamente los drboles que han alcanzado cortabilidad legal (12 pulgadas de
didmetro a la altura del pecho de un hombre normal) y sirven para horcones de
viviendas rurales o traviesas de ferrocarril de primera clase.

Como el encinar es un monte claro, aun en las mejores condiciones, la
regeneracién tiene lugar todos los afios, no siendo necesarias las cortas uni-
formes que se emplean en otros tipos de monte. Las cortas totales, aun en
pequefias dreas, no son recomendables en el caso de la encina. En suelos po-
bres donde el monte tiene poco vuelo o es simplemente un matorral, lo mejor
es dejar el encinar intacto. En la mayoria de los encinares cubanos, la re-
generacién tiene que comenzar con cuidadosas cortas de mejora.

Esto ser& efectivo en los mejores suelos. En estas operaciones se tra-
tara de fomentar la asociacién de la encina con especies subordinadas, como
el macurije, por ejemplo. Con las cortas de mejora se extraen los drboles
viejos dafiados, los copeyes y los bejucos y cualquiera otro &rbol que tienda
@ suprimir a la encina. Conjuntamente se pueden plantar en lugares convenien-
tes, tocones fuertes de encina y pimpollos de pino macho si se quiere proteger
o introducir esta especie. ‘

Desde luego, la regeneracion de los encinares requiere medidas eficien-
tes de proteccion contra los fuegos, frecuentes en estas localidades durante
el periodo seco.

Artificialmente la encina se puede establecer por siembra directa o por
plantacién, empleando en este caso, tocones de plantas de viveros de 15 a 18
meses o mejor aun, de 2 afios y medio.

En la siembra directa conviene obtener una germinacién minima de 5
plaéntulas por metro cuadrado, o sean 50,000 por hectérea. Esto es necesario,
porque la encina no tiene una tendencia al crecimiento en altura, el que es
necesario estimular por medio de la densidad. Asumiendo para la semilla de

= se) 2 July 1943

encina un valor real de 50 por ciento, se necesitardn 364 kilos de bellotas
para sembrar una hectérea (275 bellotas pesan un kilo).

La siembra directa puede ser a voleo, en terreno parcialmente prepa-
rado, bastando alterar la capa superficial del suelo con una o dos labores
de roturacién. Si el drea elegida contiene una cubierta vegetal densa, se
procede a chapear y quemar antes de la preparacién del suelo. Esta siembra
debe hacerse con las primeras semillas de la estacién en octubre si es po-
sible, para aprovechar la humedad del suelo y las probables lluvias de este
MES o

La reforestacién a base de encinas puede hacerse por plantacién, emple-
ando tocones procedentes de viveros. Con la debida anticipacién se establecen
los semilleros. Las bellotas se recogen maduras en octubre y es necesario
transportarlas rdpidamente al lugar de la siembra, evitando envases volumino-
sos que provocan calor y fermentacién. El semillero se establece con prefe-
rencia junto a la plantacién para evitar el transporte de los plantones. Se
prefiere suelo muy suelto y muy profundo. Los suelos pesados son indeseables,
pues dificultan las operaciones de trasplantar.

Se preparan canteros o terraplenes a pleno sol donde se siembran las
bellotas a 5 x 19 cms.; un total de 200 por metro cuadrado de semillero. Con
un valor real de 50 por ciento, germinardn a razén de 100 por metro cuadrado
de cantero. Las bellotas se colocan en pequefios surcos y se entierran o cu-
bren con una pulgada de suelo arenoso. Se mantienen moderadamente humedas
hasta el comienzo de la germinacién que tendraé lugar hacia los 10 dias. No
es necesario hacer ningin trasplante provisional en el vivero. Las plantulas
pueden medir 25 cms. de altura a los 5 meses y de 80 cm. a 1.20, a los 15
meses, con un didmetro maximo de 1.5 cm. en la base, que ya es buen tamafio
para plantar. En los semilleros de 27 a 30 meses las mejores plantas miden
2 metros de altura con un didmetro de 3 cm. Los tocones de esta edad planta-
dos en marzo, parten con extraordinaria rapidez. La plantacion se realiza
entre marzo y abril, con preferencia en marzo, antes de que las encinas se
encuentren en plena actividad vegetativa.

Se trasplanta a raiz limpia empleando una coa grande para extraer las
plantas del semillero. Estas se llevan a un lugar sombreado donde se escogen,
desechando las que han quedado demasiado pequefias y se recortan, dejandolas
reducidas a pequefios tocones de 1-1/2 a2 pulgadas. En estas condiciones se
cuentan y atan en pequefios manojos, se sumergen por sus raices en una suspen-
siodn de tierra arcillosa, se envuelven en paja, se forran de yagua y se remi-
ten a la plantacion.

La plantacién se realiza en suelo que puede haber sido preparado de
antemano o no. Cuando la cubierta vegetativa del drea a repoblar es espesa,
se puede chapear y quemar. No es necesario romper el terreno. Entendemos
que no es conveniente hacerlo en el momento de plantar. La plantacién se

ejecutaré a jan, empleando una barreta de hierro, coa fuerte o jan de madera

dura. Se practica un hoyo estrecho y profundo, donde la simple raiz de la

encina pueda ser insertada cémodamente hasta el cuello de la raiz y apretada
con un segundo golpe de barreta dado hacia un lado. Plantada en estas con-

diciones, la encina se pone desde luego en contacto con la humedad del suelo

Caribbean Forester - 160 = Vol. 4, Noo. 4

—

profundo que es mantenida por capilaridad, pudiendo hacerse la plantacién aun
en tiempo seco.

Se aconseja plantar 10,000 tocones por hectérea como espaciamiento pro-
visional, entendiendo que no resulta una plantacion demasiado densa y que en
suelo arenoso bastante fértil, no necesitard aclareos antes de los 5 afios.

Los Encinares del Pinar del Rio

Las encinas de la provincia de Pinar del Rio se pueden contar billones,
pero el volumen de madera de primera clase disponible, es muy pequefio. Un
examen superficial de nuestros encinares da a conocer con bastante aproxima-
cién la cantidad y calidad de la madera que contienen.

Con motivo de las limitaciones impuestas por la guerra se ha pensado
en establecer en Cuba la fabricacién de toneles, prefiriendo la madera de en-
cina, pero esta no existe en las cantidades y de la calidad que se requiere
en esta industria especializada. Los industriales de este género estableci-
dos en Cuba no deben demorar en ensayar maderas como el roble blanco, el
roble guayo, el mangle prieto y otros, en la fabricacidon de toneles.

Los mejores toneles de fabricacién americana, destinados al trasiego y
envase de vinos y aguardiente, se fabrican exclusivamente de la encina blance,
uercus alba y de los montes en que se encuentra esta especie, solamente los
mejores arboles son apeados. La madera destinada a la fabricacién de toneles

no puede tener defectos de ninguna clase. Considerando nuestros encinares
desde este punto de vista, encontramos que no contienen el volumen de madera
que se necesita para la industria tonelera. La poca madera de nuestros enci-
nares, es una madera dafiada y defectuosa y se emplea casi exclusivamente en
construcciones rurales de la localidad y para traviesas.

El uso de la madera de encina se encuentra limitado también por otro
factor: el valor alimenticio de la bellota, empleada en la ceba de cerdos.
Muchos de los terrenos ocupados por la encina son de propiedad privada y los
propietarios encuentran més utilidad en la recoleccién de la fruta que en la
venta de la madera y por consiguiente no se interesan en extraer los arboles
de cortabilidad legal o de efectuar aclareos.

Los encinares de Pinar del Rio son en su mayor parte homogéneos y ocu-
pan dreas de extension considerable interrumpidas frecuentemente por sabanas
donde la yerbe@ es escasa. En algunas localidades han sido reducidos a peque-
ios bosquetes diseminados.

En todos nuestros encinares, son evidentes aun los efectos de los ci-
clones que visitaron la provincia de Pinar del Rio el afio 1910 y algunos afios
después. Los 4rboles de 20 pulgadas o m&s son muy escasos y estén sumamente
dafiados. Las clases de edad inferiores se encuentran en mejores condiciones
y prometen formar un monte de altc valor maderable si contintan bajo la pro-
teccion de la ley y libres de ciclones por algunas décadas.

A pesar de que la encina es eminentemente gregal, se nota a primera
vista la calidad superior de las que crecen en terrenos de alguna fertilidad,

- 161 - July 1943

en sociedad con especies subordinadas que proporcionan proteccién al suelo.
Este es el tipo de monte que podria ser mejorado por medio de cuidadosos acla-
reos seguidos de regeneracién artificial.

Bajo las condiciones econémicas que han prevalecido en Cuba los encina-
res nunca han recibido tratamiento alguno. El producto de las cortas inter-
medias no paga su ejecucién. Lefia, traviesas y postes podrian obtenerse de
las cortas de mejora, pero en la actualidad no es econémico iniciar estas ope-
raciones, ni conviene confiarlas a los duefios del monte poco interesados en su
(lejano) futuro valor maderable.

La encina en Pinar del Rio persiste tenazmente, sobre todo en los suelos
arenosos, que con frecuencia son dedicados al cultivo del tabaco.

La encina y también el pino macho debian ser bases principales de la
riqueza de Pinar del Rio. Grandes extensiones de suelos arenosos propios para
el sostenimiento de este tipo de monte se encuentran actualmente desocupados,
improductivos, pudiendo estar plantados de pinos y encinas. Dada la facilidad
con que estas especies se pueden criar y trasplantar y el valor relativamente
bajo de sus semillas, que se pueden recoger en cantidades considerables bajo
los arboles, es lamentable que todavia no hayamos intentado el establecimiento
de extensos pinares y encinares artificialmente.

Summa ry

Encina, a short tree with a thick trunk, is native to the western half
of the province of Pinar del Rio and the northern part of the Isle of Pines.
It is confined chiefly to sandy and sandy clay soils and prefers the base of
slopes and fertile plains, though it persists on poor soils despite fire and
grazing. It forms extensive pure stands and also occurs in mixture with
Pinus caribaea, P. tropicalis, Brya_ ebenus, and others.

Though there are extensive encina forests in Pinar del Rio the quanti-
ty of high grade sawtimber is very limited. The production of encina barrels,
a recent proposal for the alleviation of limitations imposed by the war, is
impracticable for this reason. However, much of Cuba's defective encina is
widely used for rural construction and railroad ties. Many owners of encina
forests consider the fruits of more value than the wood, for they are valued
for fattening swine.

Encina may be regenerated naturally by partial cutting. To favor enci-
na in mixed forests the overstory should be gradually removed, first removing
the declining trees and others of low value. Artificial regeneration is suc-
cessful either by direct seeding or planting. The fruits mature between Octo-
ber and December. There are 125 per pound, and germination is 85 per cent
complete after 20 days. Seeds should be sown soon after collection, but if
storage is necessary they should be kept exposed to the air, as they tend to
heat. Direct seeding is most successful at the time when the seeds first ri-
pen in October, as it is then possible to take advantage of the probable rains
during that month. Provision should be made for at least 5 seedlings per
square meter, as close spacing is required to produce straight trees.

Caribbean Forester - 162 = Vol. 4, No. 4

In the nursery the seeds are sown in unshaded beds at a spacing of 5x
10 cms. They are covered with an inch of sandy soil. After 15 months without
transplanting they average 80 cms. in height and are ready to transplant.
Transplanting is most successful during March and April. About 4,000 trees
are planted per acre. The trees are moved with bare roots and are cut back
to 1 to 2 inches in height. The holes are made narrow and deep, just enough
_to permit the insertion of the taproot to the same depth as in the nursery.

To the present economic conditions have prevented improvement of encina
forests, and as a result they are now valuable chiefly for firewood, ties, and
posts. It is regrettable that it has not been possible to build up these
forests on the extensive idle sandy lands in Pinar del Rio, where the species,
together with Pinus caribaea should provide a principal source of income.

000

APUNTES SOBRE LA MYRICA CERIFERA L. DE HONDURAS

Recientemente tuve la oportunidad de estudiar la planta de la cera veze-
tal, Myrica cerifera L.,' que crece sobre las faldas del Picacho, inmediatas a
las Crucitas (lugares préximos a la capitel). Esta planta, llamada ccomimmente
"encinillo" pertenece a la familia de lias Miricdéceas. Se trata de un arbustco
que sobrepasa la estatura humana, bastante ramificado, consistente, de hojas
alternas, simples, menudags, incompletas, coriaéceas y tiene aspecto de encina
tierna a lo cual se debe su nombre vulgar. Prefiere los sitios elevados y
frescos y por eso se manifiesta en los claros sobre la parte ascendente del
camino a San Juancito. También se encuentra en occidente en los pueblos de
Ojojona y Maraite.

El periods fructificativo se inicia desde noviembre y los frutos en es-
tado de madurez son apifiados, esféricos, pequefios, cenicientos y casi sentados.
Deben los frutos el color blanquecino a la capa grasienta que los envuelve, la
que no tiene un espesor uniforme sino estructura de tejido,- semejando pavimen-
to, que por la friccién entre los dedos se desintegra en grasa y carcoma, de-
jando en libertad la semilla y una emanacién aromAtica.

La presencia de esta grasa en los frutos es una adaptaci6n natural que
favorece la propagaci6én de la especie. Dado el hecho de que la planta crece
en parajes elevados y de que fructifica en tiempo fric, es légico suponer que
esa capa grasienta aisladora protege la semilla contra la humedad y las hela-
das. Pasado el periodo de la fructificacién viene el cambio favorable del
tiempo que con el calor de la estacién seca, derrite la proteccién grasienta
de los frutos caidos par&é que con las lluvias venideras germinen las semillas
y evolucionen las plantas.

La cera hoy dia sdélo se utiliza en pequefia escala para la fabricacién
de jabones y velas pero ya una fabrica de la capital traté de industrializar
el producto. En mis experiencias yo pude obtener un rendimiento de dos libras
de la recolecta de tres arbustos. La cera se extrae mediante la ebullicién lo
cual hace flotar la grasa que luego se desnata y recalienta hasta eliminar
todo el agua.—PROF. LUIS LANDA ESCOBER, Tegucigalpa, Honduras.

SERS July 1943

LA VEGETATION MUSCINALE DES ANTILLES FRANCAISES

ET SON INTERET DANS LA VALORISATION SYLVICOLE

H. Stehlé
Lauréat de l'Institut et de la Société Botanique de France
Ingénieur Agricole et d’Agronomie Coloniale
Martinique

Les mousses, d'une maniére générale, n'ont guére été l'objet d'études
écologiques ou phytosociologiques dans les régions tropicales et cette obser-
vation s'applique aussi bien sux Antilles frangaises qu’aux autres Iles Caraibes
et aux Grandes Antilles.

Cependant, l'étude systématique ou descriptive de ces végétaux crypto-
gamiques a donné lieu dans la plupart de ces Iles a de nombreuses publications
intéressantes. Les bryologues ont eu en effet la possibilité de procéder aux
déterminations spécifiques et aux études taxonomiques ou anatomiques les plus
diverses dans les laboratoires botaniques, alors qu'il est plus délicat ad’!
observer avec précision les groupements muscinaux in situ, leur biologie et
leur évolution, d’interpréter leur présence et l’aspect particulier qu'ils
impriment au paysage floristique et de montrer enfin l'intérét qu'ils offrent
dans la valorisation sylvicole d'une région déterminée.

L’étude suivante, appliquée au cadre des Antilles frangaises, n'a pas
la prétention de résoudre complétement les problémes posés, mais elle a pour
but de faire ressortir la richesse et l’aspect de la végétation muscinale de
nos Iles, la composition générale des groupements bryologiques et hépaticolo-
giques qu’elle héberge et de déduire les indications qu’ils peuvent fournir
pour caractériser les associations végétales et leur évolution, les stades de
colonisation sur sols nus ou sur laves récentes et pour orienter le forestier
vers des localisations ou des modes de valorisation plus judicieux.

Une légtre adaptation permettrait certainement son application dans les
Tles Caraibes voisines oh des relevés analogues pourraient mettre en lumiére
les nombreux points communs qui nous paraissent y exister et les différences
qui se manifestent dans leur constitution intime.

La connaissance approfondie de la végétation muscinale des Antilles
frangaises est rendue plus difficile encore pour les deux raisons suivantes:
Ces Iles, comme tout le groupe de l’archipel antillais, sont situées en
Amérique tropicale, région, qui, d’aprés 1!’Index du grand bryologue, le
Géneral Paris, est la plus riche en mousses du monde entier. Le caractere
insulaire, montagneux et humide de ces Petites Antilles renforce encore cette
richesse die & leur situation générale. En effet, la Guadeloupe et la Marti-
nique possédent 4 elles seules un total d’environ 410 mousses se répartissant
en 200 Héepatiques et 210 Muscinées.

Une comparaison analogue & celle que nous avons effectuée pour les
Fougeres et Alliées dans le "Catalogue des Cryptogames Vasculaires des

Caribbean Forester - 164 = Vol. 4, No. 4

Antilles frangaises", publié dans un numéro antérieur du Caribbean Forester,
(Vol. 4, No. 1), permet de chiffrer proportionnellement cette richesse bryophy-
tique. Sur les 2500 végétaux vasculaires qu'on y rencontre, nous avions décompté
313 Ptéridophytes, soit un rapport de 1 fougére pour 7, 8 plantes vasculaires.
Si nous faisons le décompte des végétaux feuillus des Antilles frangaises: Pha-
nérogames, Fougéres et Mousses, le total approximatif s'éléve & 2910 espéces,
sur lesquelles 410 sont des Muscinées, ce qui représente un rapport de 1 mousse
pour 7 plantes feuillues (1:7,09). A cette proportion spécifique élevée, doit
s'ajouter encore le foisonnement quantitatif de certaines espéces et leur large
extension dans les milieux o4 l*humidité et la topographie les favorisent,
expliquant la prépondérance du caractére imprimé par le paysage bryophytique
dans liaspect général de la végétation.

Les études systématiaques relatives aux Muscinées de notre flore dignes
de retenir l'attention sont celles de T. Husnot: Catalogue des Cryptogames
recueillies aux Antilles frangaises en 1868 et essai sur leur distribution
géographique dans ces Iles, Caen 1870, puis dans la Revue bryologique 1875-76.
Pour les Hépatiques, E. Bescherelle a donné en 1893 une "Enumération des
Hépatiques connues jusqu’ici aux Antilles frangaises", & laquelle le R.F. Duss
® ajouté en 1903 une trentaine d'iespéces (nommées par Stephani) dans son
"Enumération des Muscinées des Antilles frangaises: I. Hépatiques, Lons-le-
Saumier". Ces espéces ont fait ensuite l'objet de révision et de compléments
avec d'autres nouveautés par Stephani dans son "Spec. Hepaticarum", in Bull.
Herb. Boissier et dans "Hepaticae novae Dussianae, in Urban, Symb. Ant. Vol.
II, Oct. 1901, puis par A. W. Evans dans diverses publications principalement
in Bull. Torrey Bot. Club, Trans. Conn. Acad. 28 en 1927, Ann. Bot., Bryo-
logist et Annales Bryologici (1937). Pour la Guadeloupe, une étude récente:
"The Hepaticae of Guadeloupe" a été publiée par F. M. Fagan.

Pour les Muscinées (str. sens.), la connaissance des mousses de nos Iles
est basée principalement sur la “Florule bryologique des Antilles frangaises"
de E. Bescherelle, de 1876, in "Ann. Sci. Nat. Paris ser. 6, III, & laquelle
une trentaine d’espéces nouvelles, de méme que pour ies Hépatiques. ont été
ajoutées également par le R. P. Duss, dans son “Enumération des Muscinées" de
1803: II. Mousses. Ces espéces ont été déterminées par Bescherelle lui-méme,
puis par Brotherus, le savant bryographe du “Nattrlischen Pflanzenfemilien"”

d*Engler et Prantl dans la premiére édition (1901-09) et la deuxiéme édition
(X.1924 et KI.1925) duquel sont décrites de nombreuses mousses antilleises.
Quelques especes nouvelles, guadeloupéemnes et martiniaguaises sont ajoutées

en outre dans les "Symbolee Antillanae" d’Urban, Vol. IV, Mai 1903: “Musci
novi Dussiani". Dans la North American Flora“, Williams, XV, 2, 1913, en cite
également.

Au fur et &@ mesure de nos réccltes botaniques, commencées en 1934 aux
Antilles frangaises, nous avons collecté les mousses et hépatiques, plus
specialement au cours de ces quatre derniéres années. De 1939 & 1942 les
doubles de nos récoltes ont été adressées & Dr. Winona Welch de Depauw Univer-
sity, quia fait parvenir les hépatiques 4 Dr. Margaret Fulford de University
of Cincinnati et les Muscinées (str. sens.) & M. Edwin B. Bartram de Phila-
delphie. Ces spécialistes ont procédé & l'étude de nos spécimens et nous
eur exprimons ici l’expression de notre vive reconnaissance, De nombreuses
espéces se sont révélées nouvelles pour nos Iles et deux Muscinées (str. sens.)

Ses = July 1943

nommées par Bartram sont des nouveautés pour la Science: Pilotrichum Stehlei
Bartr. et Meiothecium antillarum Bartr. inéd.

Les principaux collecteurs de Muscinées aux Antilles frangaises, grace
auxquels la flore bryologique a pi tre établie sont: Bertero, Badier, Lessueur,
Wickstroém, Beaupertuis, Perrottet, Duchassaing, L'Herminier, Husnot, Bélanger,
Hahn, Duss et pour ce dernier siécle le R.P. Quentin, d’abord seul, puis avec
Madame H. Stehlé et nous-méme, les botanistes de la Mission Cryptogamique du
Prof. P. Allorge & la Guadeloupe (1936) et plus récemment M. A. d'Questel. Il
ne nous semble pas que leurs récoltes aient fait l'objet de publications bryo-
logiques autres que celles ci-dessus énumérées. Chacun de ces collecteurs a
apporté sa contribution, mais la plus large est certainement celle die au R.P.
Duss, chercheur infatigable, qui a accumulé de nombreuses récoltes d'hépatiques,
de bryées et de sphaignes. Cependant, la publication relative @ "l'énumération
méthodigue des Muscinées des Antilles frangaises: I. Hépatiques et II. Mousses",
réalisée par lui en 1903, sans description, synonymie, dessins, références a!
auteur ni clef dichotomique, est le seul travail que le R. P. Duss ait publié
sur les Cryptogames cellulaires feuilluee. Depuis 40 ans qu'elle a vu le jour,
de nombreux binémes ont changé dans la désignation rationnelle des espéces, des
nouvelles mousses ont été décrites, aussi un nouveau catalogue & jour, ou mieux
une Flore bryologique, descriptive et illustrée, des Antilles frangaises serait
elle indispensable dans l'état actuel de nos connaissences.

Les affinités géographiques des mousses et hépatiques des Antilles
frangaises sont surtout accentuées avec le domaine intertropical de 1'Amérique
du Sud et les Grandes Antilles. Pour les mettre en évidence, une étude criti-
que des especes collectées ou publiées s’imposait en se plagant au double point
de vue taxonomique et géographique. La comperaison des espéces de notre
Catalogue préliminaire des Muscinées et Hépatiques des Antilles francaises, en
cours de mise au point définitive, avec les plus récentes publications des
Antilles voisines fait apparaitre une étroite liaison bryophytique. Ainsi,
pour Porto Rico, la plus petite des Grandes Antilles et le plus proche de nos
Iles, F. M. Pag&én a publié en 1939 (in the Bryclogist, Vol. XLII, Febr., Apr.,
Jun. 1939) une liste: "A preliminary list of the Hepaticae of Puerto Rico
including Vieques and Mona Island", comportant 244 espéces d’Hépatiques. Sur
ce total nous avons pu en reconnaitre 90 qui appartiennent A notre Flore, soit
une proportion commune par rapport & l'ensemble de Porto Rico s'élevant & 37%
et, par rapport aux 200 Hépatiques des Antilles frangaises, une commmeute de
45% avec notre flore. Pour les mousses proprement dites, la comparaison de
notre flore avec 1l'Ile de Cuba, la plus grande Ile de l’archipel des Grandes
Antilles, est rendue aisée par la publication du “Catalogue des Mousses de
Cuba" par le Frére Leén, de la Havane (in Annales de Cryptogamie exotique,
tome VI, fasc. 3-4, Sept. 1933) ot 268 mousses sont citées et quia été com-
plétée en 1939 par I. Thériot (in “Memorias de la Sociedad Cubane de Historia
Natural", Vol. XIII, n. 5, 20 Déc. 1939): Complément au Catalogue des Mousses
de Cuba, €levant le nombre des mousses de cette Ile & un total approximatif de
300 espéces. Sur ce nombre, 75 appartiennent & notre flere muscinale, soit wm
taux de communauté de 25%. Sur les 50% que le frére Leén indique (loc. cit.

p. 4) comme se trouvant aux Antilles, la moitié existe done dans nos deux Iles.
Ces valeurs comparatives sont d’autant plus appréciables que le taux d'endémicité
des hépatiques et muscinées des Iles Antillaises est assez élevé: celui de Cuba
pour les Mousses (str. sens.) est indiqué par le Frére Leén & 20% et le nombre

Caribbean Forester - 166 = Vol. 4, No. 4

des hépatiques spéciales 4 Porto Rico et n'étant pas connues ailleurs, s'éléve,
d'aprés notre calcul sur le travail cité de Pagan, & 9% (22 espéces sur 244).

Les relations géographiques examinées confirment donc pour les Hépati-
ques et les Mousses, les conclusions formulées pour les Phanérogames et les
Pteridophytes relatives & la connexion antérieure des Grandes Antilles par les
Iles Caraibes avec 1’Amérique du Sud dans le vaste domaine floristique de
laquelle nos Iles doivent é&tre inclues.

Du point de vue écologique et stationnel, les mousses présentent des
exigences variées et bien définies. Elles reflétent l'action du milieu aussi
nettement que les végétaux supérieurs, action qui se manifeste par des adapta-
tions, des variations écologiques et des formes de croissance particuliéres.

Aux Antilles frangaises, les milieux biologiques naturels sont extréme-
ment favorables au foisonnement muscinal: terrestre, épilithe, corticicole.ou
épiphylle, car les différents facteurs: lumiére, température, état hygromé-
trique, ainsi que l’eau et le sol, exercent sur les mousses leur action simul-
tanée dans des conditions souvent optimales. L‘influence de chacun de ces
facteurs devrait faire l'objet d’observations et de mesures pour définir
exactement la part qui leur revient séparément dans le complexe observé.

Il n*y a qu’une cinquantaine d‘années que, dans les régions tempérées.
cette action du milieu sur les muscinées a été envisagée et a notre connaissance,
le seul travail d’ensemble publié sur ce sujet est celui du Dr. Adrien Davy de
Virville, de la Faculté des Sciences de Paris: “Action du milieu sur les Mousses",
in Revue Générale de Botanique, 1927. Les résultats qu’il a obtenus par experi-
mentation sur des espéces des régions tempérées d'Europe, exposés dans ce
magnifique travail, concordent pleinement avec les constatations que nous avons
pa faire sous les conditions naturelles de la forét dense intertropicale des
Antilles frangaises. Cette forét, du point de vue physique, est caractérisée
par une atténuation trés accentuée de la lumiére dont les rayons ne passent
que difficilement 4 travers les frondaisons épeisses, par une grande constance
de la température, moyenne en général, par un état hygrométrique toujours
élevé et parfois voisin de la saturation totale, par le calme de 1‘*atmosphtre,
par le suintement de l'eau le Aemg des branches et des feuilles et par 1’
imprégnation permanente de son sol humifére.

Cet ensemble de causes détermine des effets assez comparables & ceux du
milieu cavernicole et dans les sylves les plus sombres de la Guadeloupe et de
la Martinique, les muscinées sciaphiles sont en grande abondance formant de
nombreuses cclonisations. Mais, lorsqu’une clairiére ou une trouée se produit
en forét ou que l’aspect de haute futaie disparait pour céder le pas au taillis
régressif, l°augmentation de la luminosité et le passage de courents diair
éliminent rapidement ces espéces sensibles. Elles appartiennent & la catégorie
des espéces que Jd. Amann, qui a étudie leur biologie a désignées sous le nom
d'apanémophiles. Cette disparition est rapide si les causes qui la provoquent
agissent brusquement et la dessication du tapis muscinal, qui suit celle encore
plus immédiate des épiphytes diverses, se fait alors sans transition. Si, au
contraire, le passage & une plus forte luminosité ou & me atmosphére plus
ventilée, se fait de fagon progressive, des modifications des mousses s'obser-
vent avant la mort du vegetal: le tapis vert foncé devient plus clair puis

=e Goffe July 1943

vert-jaunitre, les feuilles deviennent moins colorées, leur taille diminue et
elles se recroquevillent, se resserrent les unes contre les autres et les dents
s'accentuent. Presque toutes les mousses vertes peuvent présenter en forét de
telles adaptations en milieu défavorable alors que nous n‘'en avons observé
aucune apperemment sur trois mousses blanch&tres ou d’un vert-jaune et pale,

assez fréquentes: Octoblepharum albidum (L.) Hedw., Leucobryum antillanum
Schimp. et L. crispum C. M.

Ces mousses sont au contraire douées d'une grande stabilité dans la na-
ture et des cas semblables ont été signalés expérimentalement par Davy de
Virville au laboratoire en particulier sur Leucobryum glaucum Hpe et Anomodon
viticulosus H. et T., d*Europe, “espéces rebelles & toute modification pour
des raisons d’ailleurs inconnues". Ce comportement négatif nous parait devoir
@tre en liaison avec la nature des membranes celluleaires des tissus et 1!
intensité de coloration plus faible des chloroplastes du limbe.

Dans l'ensemble, les mousses sylvicoles sont des appareils météorolo-
giques enregistreurs, vivants, sensibles et complexes, qui réagissent sur un
ensemble de conditions édaphiques et climatiques 4 un instant donné. Cette
réaction est favorable & leur développement lorsque, ainsi qu’on peut l'observer;
ja lumiére solaire ne les frappe pas directement et exerce son action de fagon
discontinue, l*humidité atmosphérique est trés élevée, l’atmosphere ambiante
non soumise @ des agitations sensibles, le degré thermique moyen, l'eau abondan-
te sans tomber violemment et le sol humide et acide. L’acidité du milieu in-
fluence non seulement le déveleppement des mousses et des hépatiques, mais
encore la germination de leurs spores et la presque totalité des bryophytes de
la Martinique et de la Guadeloupe nous parait se développer en solutions acides.

Le milieu "for@t caraibe" a done suscité la création des conditions
énumérées nécessa@ires au développement et @ l’équilibre biologique des végetaux
qu'elle abrite et parmi lesquels les muscinées sciaphiles sont les plus abon-
dents et les plus sensibles. L’étude de leurs colonisations pourra @tre utili-
sée pour reconneitre une ambiance forestiére, pour orienter la reforestation
vers une évolution progressive artificielle ou vers la culture et pour déduire
des indications biolegiques complétant, s'il y en a, les observations météoro-
logiques et plus précieuses que celles-ci pour le forestier ou l’agriculteur
car la présence et le comportement des groupements muscinaux sont le reflet de
ces actions physiques sur le végétal.

A la différence des régions tempérées oti les mousses sont généralement
terrestres ou aquatiques, le paysage bryologique antillais comporte une variete
abondante de mousses que l'on peut classer en cing groupes écologiques: elles
peuvent en effet @tre plaquées sur le scl, latéritique, ponceux, humifére ou
calcaire; ce sont des terrestres comme en renferment les genres Bryun,
Rhodobryum, Ricciea et Isotachis. Certains vivent sur les rochers secs ou
suintants comme les Hyophila, Fissidens, Dicranella et Marchantia alors que
d'autres recherchent plus particuliérement le bois pourri ou les substances
végetales en voie de décomposition; ce sont les saprophytes des genres Rhizo-
gonium et Conomitrium. Les muscinées qui vivent sur la matitre organique
comprennent deux catégories; celles qui vivent sur les feuilles des arbres ou
des plantes, les épiphylles représentées surtout par les genres Qdontolejeunea,
Diplasiole jeunea, Cyclolejeunea et Meteoriopsis; celles qui vivent sur les

Caribbean Forester - 168 = Vol. 4, No. 4

trones des arbres et des écorces vivantes des branches, les épiphytes cortici-
coles qui sont trés abondantes et variées. Ces épiphytes peuvent appartenir

& trois types distincts: type pendant: Phyllogonium, type appliqué: Radula,
Meiothecium, type dressé: Porotrichum et Pilotrichum. Certaines cependant qui
présentent le port dressé ou appliqué dans un stade juvénile passent au type
pendant par la quite, comme c'est le cas de 1l'Qmphalanthus debilis Lehm. et
Lindenb.

La strate n'est pas 4 la fois une "strate herbacée et muscinale" comme
on la désigne souvent dans les études écologiques dans les régions tempérées;
elle est discontinue et surélevée. La strate muscinale, par l'’épiphytisme
arboricole, s'étend depuis le sol jusqu'au sommet des grands arbres mais de
fagon discontinue aussi bien en élévation que le long du tapis inférieur. C!
est sur ie sol qu'il y a le moins de mousses en forét humide et un substratum
organique ou minéral: trone vivant ou pourri, branches, feuilles, talus ou
rocher, est nécessaire & leur colonisation. Dans l‘eau des riviéres, il ne
s'en trouve pas, sauf sur les rochers émergés ou sur les berges formant cein-
ture au niveau des hautes eaux, comme i'indique le Sematophyllum caespitosum
(Sw.) Mitt.

Sur les coulées boueuses, les dépots meubles et les laves andésitiques
récentes de la Montagne Pelée, les muscinées épilithes et terricoles constituent
la premiére phase de la colonisation végétale depuis les abords méme des cratéres
et des aiguilles volcaniques. Comme piconniers de colonisation dans les
successions végetales qui apparaitront sur de tels sols jusqu’& leur mise en
valeur, les mousses offrent par conséquent un intérét pratique digne de consi-
dération.

L’étude phytosociologique aux Antilles frangaises n'a été entreprise
qu'au cours de ces derniéres années et a porté naturellement d‘abord sur les
phanérophytes qui occupent la place prépondérante dans l’aspect et la consti-
tution des grands groupements végétaux. Déjé, dans l’Introduction & sa "Flore
Phanérogamique des Antilles frangaises" le R. P. Duss, en 1896, donnait quelques
indications de Géographie botanique sur les deux Iles et la distribution
succinte des plantes en zones et en régions. Dans notre "Essai d'Ecclogie et
de Géographie botanique" publié en 1935, pour la Guadeloupe et dans 1’"Esquisse
des associations végétales de la Martinique" in Bulletin Agricole Mart. Vol.
VI, No. 3-4, Déc. 1937, les climax forestiers, les associations et les succes-
sions végétales ont été caractérisées. Dans le cadre de chaque groupement, 1?
aspect du paysage bryophytique y a été indiqué chaque fois, mais aucune étude
spéciale n'a encore été faite de maniére détaillée.

L*évolution du tapis végétal aux Antilles faitapperaitre parfois un réle
important des Muscinées, surtout dans la région moyenne en for@t dense, duns
la région supérieure, en forét de transition et dans le sphagnetum guadelupense
des plateaux et sommets volcaniques. Dans les associations de la végétation
inférieure, au contraire, le caractére xero-héliophile des secteurs qui les
abritent, convient peu au développement muscinal et les bryophytes y sont trés
rares. Parmi les stades évolutifs que présente la végétation de nos Iles,
diverses associations muscinales apparaissent dans lesquelles les groupements
précurseurs de la bonne for@t ou de la culture doivent @tre mis en évidence.

3 e5) July 1943

La présente étude a pour objet de tracer l'esquisse des colonisations
muscinales et de leur écologie en les plagant dans les groupements végétaux
dont l'évolution des phanérophytes et des ptéridophytes a déja& été établie
antérieurement dans les travaux énumérées. Comme les diverses strates d'une
méme association sont étroitement interdépendantes les unes des autres et que
leurs réactions réciproques ou sur le milieu se manifestent dans l'entité
genérale qu'elles constituent, il nous a paru nécessaire et logique de décrire
ces formations bryophytiques dans le cadre défini de la végetation étudiée.

Il sera done envisdgé successivement:;

Forét xérophytique et taillis héliophiles de l’étage inférieur.
Forét mésophytique et cultures.
Forét hygrophytique de l'’étage moyen avec son complexe muscinal.
Peuplements & fougéres arborescentes et & bambous.
Forét sciaphile de transition & Clusia.
Sylve montagnarde rabougrie de l'étage supérieur.
Marécages des plateaux et sommets volcaniques et cuvettes & sphaignes.
Colonisations végétales sur laves récentes aux abords des cratéres
et cOnes volcaniques. Réle des mousses.

Comme nous l'avons fait pour les végétaux vasculaires; Les associations
végétales de la Guadeloupe et leur intér@t dans la valorisation rationnelle,
in Rev. Bot. appl. XII, 186-87, 1937, nous ferons ressortir ici, pour chacun
de ces groupements l’intéret pratique qui s’attache & la présence et 1'évolution
des strates muscinales, les indications qu’elles donnent et méme les espoirs
qu'elles autorisent dans la valorisation sylvicole ou agronomique.

Forét Xérophytique et Taillis Héliophiles de 1’Etage Inférieur .

Les associations xéro-héliophytiques de l'étage inférieur ont été
décrites au double point de vue Géographique et Botanique dans le détail,
pour la Guadeloupe dans notre Essai d’Ecologie et pour la Martinique dans 1!
Esquisse des Associations Végétales, mais aucune muscinée ne figure dans les
relevés indiqués, Leur localisation est cependant étendue dans les deux Iles
ott elles couvrent toute la Grande-Terre & la Guadeloupe et le Sud de la
Martinique. Pour ces régions, d'’assez vaste superficie comparative, ni Husnot,
ni Bescherelle, ni le R.P.A. Duss, n'indiquent des mousses ou des localisations
de bryophytes. Il ya en effet trés peu de muscinées dont le type stationnel
est la forét xérophile, le taillis ou la pelouse qui endérivent. Néanmoins,
une recherche minutieuse & travers les bois et sur les mornes permet de
détecter quelques mousses dont la présence doit @&tre sujette 4 interprétation.

En effet, la lumiére directe et l'air see qui caractérisent cet étage,
sont des conditions de vie néfastes aux muscinées et les mousses héliophiles
qui s'y adaptent n'y trouvent pas leur optimum biologique mais les supportent
alors que les autres y périssent. Cette adaptation est plutdt une tolérance
et nécessite des modifications que W. Watson: Xerophytic adaptations of Bryo-
phytes in relation to habitat (in The new Phytologist t. 13, 1914), a décrites
d'une maniére générale; longueur et incurvation du pédicelle, développement des
papilles, carnosité des tissus, absence de cellules chlorphylliennes. Les
touffes de ces mousses sont plus serrées et disposées en coussinets compacts;

Caribbean Forester = 170 = Vol. 4, Now 4

elles sont d'un vert-jaunaétre ou grisatre et les denticulations des limbes
sont plus accusées. LA ob poussent ces végétaux, les conditions optimales
pour leur croissance ne sont pas réalisées, mais il y existe un minimum da!
humidité ou d‘ombrage périodiques pendant certaines époques, de courte durée
sans doute, mais toujours présentes, expliquant non seulement cette végetation
bryophytique mais encore des possibilités culturales ou sylvicoles demeurées
inapergues.»

Le faciés sableux de la forét xérophile 4 Ceiba caribea-Tabebuia pallida
ne comporte pas de mousses caractéristiques de l'association. Des débuts de
colonisation sans fructification ont pfi seuls, dans certains cas, 6tre obser-
vés sur des troncs d’arbres dont les organes véegétatifs se rapportaient les uns
& des Hépatiques de la famille des Lejeuneacées et les autres & des Muscinées
du genre Calymperes- Parfois, 1'Qctoblepharum albidum y installe ses touffes
cespiteuses mais elles demeurent sporadiques et d’un développement limite.

Le faciés calcaire, abondant en Grande-Terre et réduit & quelocues par-
celles du Sud & le Martinique, posséde une flore bryophytique d’une extréme
pauvreté et les quelques espéces récoltées ne sont que des accidentelles d'
autres secteurs dont le microclimat favorabie: suintement de l'eau, maintenant
un état permanent d‘humidité sur ce sol perméable, abord de ravin ou pente
topographiquement bien abritée du vent desséchant, explique le présence, d’
@illeurs étroitement localisée.

Le faciés volcanique décrit pour la Martinique sous le nom dtassociation
& Fagara (Bull. Agr. Mart. VI, n. 3-4, p. 211, 1937): Fagarea microcarpa-Myrcia
paniculata var. Imrayana et ses dérivés régressifs, possédent au contraire de
petites associations muscinales dépendantes.

Nous pouvons ies classer en trois catégeries en liaison avec le substra-
tum: terrestre, saxicclies et murales.

1. Association muscinale terricole des taillis & Acacias ou Hypophiletum
microcarpae.—Au dessous de la strate herbacée du taillis & acacias et @ cam-
péches de Sainte-Anne, une strate muscinale parfaitement développée et fructi-
fiée, s‘observe sur une grande étendue constituant un véritable peuplement bas
et plaqué sur le sol. Elle est formée uniquement par 1l’Hyophila microcarpa
(Schp.) Broth., (Dét. E.B. Bartram), espéce qui n’avait jamais pourtant, &
notre connaissance, été signalée aux Antilles francaises. Elle se développe
sur une large étendue, entre 10 et 100 métres d’altitude, sur le sol de tuff
volcanique d’origine sous-marine et sur les pierres andésitiques dont l'ensemble
constitue le substratum du Morne Saline de Sainte-Anne, le hauteur boisée
située la plus & 1'Extréme Sud de la Martinique. Ses touffes cespiteuses et
denses, son port dressé et ses feuilles d’un vert gris&tre, font reconnaitre
& premiére vue son caractére xéro-héliophile. Dans ce secteur, l’élevage du
cabri, & demi-sauvage d‘ailleurs, est un élément destructeur de la végétation
ot les mimosées épineuses (Acacia et Haematoxylon) et méme les cactacées
(Qpuntia et Cephalocereus) ont apparu.

x

La régénération naturelle de la forét xérophytique & poiriers: Tabebuia
pallida Miers, qui constitue le climax de cette région; pour réaliser ensuite
sous son couvert des peuplements d’essences plus précieuses, de mahogany du

Sa ie July 1943

pays, Swietenia mahagoni L. ou du Honduras, S. macrophylla King d‘acajous
rouges, Cedrela odorate ou de bois blanc, Simaruba amara Aubl., devra faire

intervenir les localisations de cette mousse. Pendant les périodes de caréme,
des semis de ces espéces devront étre réalisés dans les pentes, les creux et
les ruissellements abrités des vents secs du Sud ot le minimum d‘humidité
nécessaire pour la vie de la mousse subsiste, indiquant manifestement des po-
ssibilités d*habitat pour ces essences forestiéres. Quelques "seedlings" natu-
rels de poiriers ont pi &tre observés en effet parmi les buissons de campéches
qui leur offrent une protection et un ombrage dans le jeune age. La détermi-
nation des stations les plus favorables par le groupement muscinal, le choix
des essences rustiques et de haute progressivité et la meilleure saison pour
les semis, apres parcage des cabris, permettront d’allier toutes les conditions
les meilleures pour la réussite de cette valorisation. Dans ce "Sud déshérité",
sur le sort duquel on se lamente depuis des années et pour lequel tout agri-
culteur, forestier, géographe ou m@ne économiste, a proposé en vain une
solution différente en vue de son amélioration, le seul moyen de contrecarrer
l’oeuvre néfaste de l’érosion et de ia stérilité est, pour nous, l'étude de
l'évolution du tapis végétal qui seule fait connaitre la vocation naturelle
d'une terre.

Un tel aménagement, convenablement réalisé, constituerait le berceau
d’oh pourrait se développer en tache dhuiie une plus large zone boisée rejoig-
nant progressivement un centre voisin et réalisant ainsi un quadrillage de
reboisement de plus en plus ténu grace @ des points d’appuis judicieusement
disposés. Cela permettrait de récupérer l’ambiance perdue de cette sylve mé-
ridionale dont on atteste la présence passée et & laquelle se sont substituées
l’érosion, la sécheresse excessive et la stérilité du sol. Le Sud cependant
était autrefois la région de 1*Ile la plus fertile celle ot les premiers
défricheurs de P. Gournay en 1623 se sont installés pour la culture du tabac
et celle ob les usines @ sucre et distiileries de cannes étaient les plus
nombreuses. L*action de l'homme, par la destruction de la forét, la culture
intensive sans récupération pour le milieu exploité et les méthodes primitives
d'élevage, a déclanché, puis accéléré, cette évolution régressive facilitée
par les conditions naturelles.

Du point de vue agricole, les taillis et savanes hyperxérophiles sont
difficiles 4 valoriser, ainsi que nous l'avons exposé dans l'Esquisse des
Associations végétales de la Martinique, (III. Paratypes de substitution p.
232-242), tant la dégradation a été poussée ioin. Néanmoins, dans 1!installae
tion des jardins eréoles du Sud et dans l’organisation de parcelles mises en
défens, pour la constitution de prés-bois, dans la réalisation d’un élevage
plus rationnel, ]*Hyophiletum microcarpae sera un indicateur précieux.

Un autre groupement muscinal terricole des taillis héliophiles est
celui constitué par 1'’Hymenostomum Breutelii (C.M.) Broth. Cette mousse, qui
ne figure pas dans l'’Enumération Méthodique de Duss, caractérise nettement les
talus xéro-héliophiles des Anses d’Arlet, au Sud de la Martinique. Elle colo-
nise les taches humiféres qui subsistent encore sur les talus dioritiques en
cours de latéritisation et ot les hydrarguilites sont parfois mises 4 nu; elle
appartient au facits & Fagara microcarpa - Myrcia paniculata Kr. et Urb. et
M. leptoclada DC.

Caribbean Forester = 172 = Vol. 4, No. 4

Elle indicue l*existence d'un complexe argilo-humique élevé et des
possibilités culturales sur ces taillis.

Le Bryum crugeri (Hampe) C.M., mousse terrestre, nouvelle pour les
Antilles frangaises, trés polymorphe, variable tant dans ses caracteres mor-
phologiques que dans son habitat, a été récolté par nous le long du ruisseau
s'écoulant du bassin de la prise d'eau des Trois Ilets (Propriété G. Hayot),
sous un peuplement artificiel de mahoganys du Honduras, Swietenia macrophylla
King d’une douzaine d’années et de trés belle venue. Une pépiniére forestiére
a 6té installée 1a.

2. Groupement muscinal practicole xérophytique a Riccia dussiansa.—Dans

une pelouse xérophytique et sous le taillis & Croton du Diamant, au Sud de la
Martinique et dans une prairie graminoide aux Saintes, dépendances proches de
la Guadeloupe, dans 1'Ile de Terre de Haut, & la base du Morne Chameau, nous
avons récolté une hépatique le Riccia dussiana Steph. plaquée sur le sol de
tuff. C'est dans la savane & Bouteloua-Sida (loc. cit. p. 239) ou les cacta-
cées (Opuntia) étaient disséminées, que nous l'avons détectée, & quelques
kilométres du bourg du Diamant.

Leur localisation révéle les emplacements & adopter pour tenter la régé-
nération forestiére par série progressive ou pour procéder en bonne saison
pluvieuse aux semis de bonnes fourragéres (Tricholaena repens, Panicum,

Digitaria) dont la dissémination anémophile permettrait 1l’extension naturelle.

5. Groupement muscinel saxicole: Barbuletum agrariae.—Sur les roches

calcaires et les murs, quelques mousses héliophiles plaquées, tant a la Guade-
loupe qu'é la Martinique, forment de petites colonisations 4 partir d'une trés
faible altitude.

Constituant plus spécialement une association murale, l'on rencontre
entre 0 et 400 m. d’altitude Barbule agraria Hedw. et B. husnoti Schp. Deux
autres constituant une petite association homogéne, sont terrestres ou saxi-
coles et forment parfois de larges pelouses plaqués contre le sol humide,
surtout dans les environs de Saint-Pierre. Ce sont le Barbula hymenostegioides
Broth. et l’Anaectangium breutelianum Br. et Schimp.

Ce Barbuletum est 1l’indication de champs pierreux maix cultivables et
présente un caractere nettement rudéreal.

Les mousses électives de la forét xérophytique sont donc des Muscinées
proprement dites appartenant surtout & la famille des Pottiacées.

Forét Mésophytique et Cultures

La forét mésophytique, de par sa topographie, et ses conditions édapho-
climatiques favorables & l’exploitation forestiére et & la culture, n'est plus
guére représentée aux Antilles frangaises que par des lambeaux disséminés et
corrodés entre 150 et 400 métres d’altitude. Pour la Martinique, "l’Esquisse
des Associations" en 1937 (loc. cit. p. 212-215 et p. 242-247) a fait connaitre
ses conditions naturelles, ses caractéres, l’organisation floristique de ses
strates, ses paratypes de substitution et les successions régressives auxquelles

- 173 = July 1943

a donné lieu cette forét. A la Guadeloupe et dans les Dépendances, elle était
localisée au “domaine intérieur" dont les différents secteurs ont été décrits
dans L'"Essai d'Ecologie" en 1935 (p. 126-191). Li'absence d'épiphytisme sur
écorces et sur feuilles y était mise en lumiére pour établir une différence
marquée entre la physionomie de la forét mésophytique et celle de la forét
dense et humide, située supérieurement 4 celle-ci dans l’étage moyen. Parmi
les Cryptogames épiphytes, une seule mousse du genre Leucobryum était citée
comme caractéristique.

Il y a en effet trés peu de muscinées dans cette forét mésophytique, mais
les reliquats qui en subsistent encore, les peuplements artificiels établis sur
les emplacements en certains points favorables pour constituer des foréts de
protection ou d’exploitation, et les cultures qui s"'y sont installées un peu
partout aux Antilles Frangaises, hébergent cependant quelques groupements corti-
cicoles ou, avec les lichens gris incrustants, des mousses prennent place. Pour
les détecter, les foréts de reboisement du Sud et du Centre de la Martinique
ainsi que les arbres des exploitations d’arboriculture fruitiére et des jardins
d'essais de nos Iles ont été fouillés avec soin.

1, Relicts de la forét mésophytique naturelle.—On peut distinguer trois

petits groupements muscinaux sur les arbres de la for&t mésophytique dont les
relicts, comme ceux qui s'étendent au Nord de la Martinique, entre Céron et
Grand-Riviére, sont suffisamment éloignés des grands bourgs, pour ne pas étre
encore détruits ou, comme ceux situés au Sud de cette Ile, appartiennent
encore au patrimoine domanial de la Colonie et sont l'objet de protection par
le Service forestier. Ce sont;

Octoblepharetum albidi de la forét & Lonchocarpus.
Calymperetum richardii de la forét & Inga.
Pilotrichetum crypheoidei de la forét a Simaruba.

La forét & Andira inermis-Lonchocarpus latifolius, décrite in Bull. Agr.
(loc. cit. p. 214), héberge quelques épiphytes de phanérogames et fougeres et
sur l'écorce des arbres, des lichens gris et verts, plaqués ou inerustants,
des algues de la série des Trenthepohliacées, aux couleurs rouge et brique et
deux mousses vert-pale ou blanch&tre, de la famille des Leucobryacées: Leuco-
bryum antillarum Hampe et surtout; Octoblepharum albidum (L.) Hedw. Ces deux
mousses cespiteuses, en rosettes, forment de petites colonisations parfois
assez larges sur les bois-savonnettes lorsqu'une aspérité du trone permet 1!
accumulation d'un peu d*humus.

Dans le Sud de la Martinique, les bois de Montravail, dans les hauteurs
de Sainte-Luce, vers 230 métres d’altitude, abritent des relicts végétaux
mésophytiques ot les bois blancs, Simaruba amara Aubl. dominent nettement. Sous
cette forét, présentant un bel aspect de haute futaie dans les endroits les
mieux abritées une strate herbacée & orchidées et graminées s'’y développe large-
ment mais, exception faite des mousses banales des abords de cours d'eau, ov
elles sont localisées, aucune muscinée terricole ou saxicole ne se rencontre.
Sur l'’écorce des bois blancs et des autres arbres de cette for8t, 1'Qctoblepharum
est assez abondant mais les colonisations corticicoles plaquées les plus nom-
breuses sont celles du Pilotrichum crypheoides Schp. Elles ne s'observent ce-
pendant que dans les secteurs les plus centraux et les plus humides du peuplement
de Montravail.

Caribbean Forester - 174 = Vol. 4, No. 4

A Rivitre-Pilote, les bois de Préfontaine et la forét de Beaudelle,
situés & 300m. d'altitude, sont hétérogénes mais avec, comme essences domi-
nantes les poix-doux du genre Inga et surtout le pois-doux poilu, Inga ingoides
Willd. ainsi que le bois blanc. Sur les trones des pilus gros arbres, des
thalles de fougéres et des jeunes organes végétatifs de mousses forment de nom-
breuses plaques vertes mais leur identification est difficile. Les deux plus
abondants qui colonisent ces écorces sont le Calymperes richardii C.M. et un
Macromitrium sp.

Ces lambeaux forestiers méridionaux sont séparés du massif central de
la forét humide, trés riche en mousses, par de vastes champs de cannes 4 sucre
qui sont totalement exempts de muscinées et constituent une barriére 4 la
dissémination des mousses, surtout épiphytes. Les groupements muscinaux cor-
ticicoles quiils ont conservés sont l’indice d'une ambiance favorable a la
vocation forestiére. Autour de ces relicts, la dissémination des bonnes
espéces forestitres sera facilitée par des éclaircissages du sous-bois, de 1!
élagage artificiel et l'enlévement du mort-bois; la fabrication du charbon en
forét sera évitée et un nettoyage des herbes coupantes (Scleria) faciliterait
la germination par la chute naturelle des semences sur le sol.

Il est @ noter que ces Muscinées électives de la forét mésophytique,
appartiennent & la famille des Calympéracées et des Leucobryacées, sont des
mousses acrocarpées comme les Pottiacées de la forét xérophile. La présence
en outre d'une Pilotrichacée, mousse pleurocarpée, d’une famille dont tous les
autres représentants caraibes sont localisés en forét dense, dans um des rares
bois du Sud doit é6tre interprétée comme l‘indice d'une reforestation possible,
non seulement par les bois-blancs mais encore, sous leur couvert, par des

mahoganys du pays, Swietenia mahagoni Jacg. & du Honduras, S. macrophylle King,
essences plus sciaphiles.

2. Reconstitution forestiére par les peuplements de mahoganys.—Depuis
une vingtaine d‘années, mais de fagon sporadique, des reboisements par le

mahogany du pays et le mahogany du Honduras ont eté effectués en diverses
régions de l'Ile, aussi bien séches qu’humides ou moyennes, sans distinction
de 1l'’évolution antérieure du tapis végétal. En forét xérophile et en forét
hygrophile, ils n'ont pas donné d‘intéressant résultats. Ceux situés aux
abords des postes forestiers, faisant l’objet de surveillance du garde, et
ceux topographiquement bien situés ont seuls subsisté. Les peuplements de la
forét de la prise d'eau du Marin (12 ha), de Saint-Joseph: Bois de la Savane
Bouliqui, de Sante-Anne: Propriété Desportes & Caritan, sont situés sur des
emplacements d'anciennes foréts mésophytiques dégradées ou détruites et ont
pris un trés beau développement.

Dans les reboisements de Swietenia macrophylla King, de Bouliqui, agés
de 22 ans, entre 350 et 400 metres, une colonisation muscinale forme une strate
discontinue, depuis le sol jusque sur les branches. Il y figure sur le sol
humide, étroitement associées: Macromitrium scoparium Mitt., M. mucronifolium
(H. et G.) Schwaeg. et Calymperes donnellii Aust; constituant le Macromitrietum
scoparii des talus humiféres de la forét & Swietenia. Lorsque le ruissellement
a entrainé les matiéres solubles de la couche superficielle, le pouvoir absor-
vant est réduit et l*humus entrainé, l*hydrargile ferrugineuse apparait et les
éléments de l'association muscinale précitée disparaissent, sauf cependant le

Sus = July 1943

Macromitrium mucronifolium qui subsiste sur les talus en cours de latéritisation
si l’érosion des talus n'a pas été trop forte.

A la base des plus gros troncs de la méme forét du réle colonisateur des
empatements et des contreforts de mahoganys incombe & la var. galipense (C.M.)
Smith du Sematophyllum caespitosum (Hedw.) Mitt., dont L'écologie sylvicole est
toute différente de l'espece type, rivulaire et terricole. Plaquées sur 1!
écorce du trone et des branches il y a lieu de noter la coexistence de deux
mousses épiphytes: Lepidopilum subenerve Brid. et Meiothecium scabriusculum
Besch. Ces dernitres espéces qui figurent également dans les relevés floristi-
ques effectués en forét dense 1a ot son atmosphére est moins saturée, peuvent
étre adoptées comme indicatrices du reboisement en forét humide dégradée par
l’essence précieuse et de haute progressivité que constitue le mahogany du
Honduras, en particulier pour le comblement des trouées ou des clairiéres.
Ailleurs, en général, la saturation et l’ombrage de la forét dense, manifestés
par un épiphytisme muscinal abondant, ne permettent pas aisément, ainsi que le
démontre la venue difficile des peuplements de 1l'Alma et des Deux Choux, le
reboisement par ce mahogany, éssence méso-sciaphile, pour laquelle un éclaire-
ment suffisant est cependant nécessaire.

5. Epiphytisme muscinal des cultures.—L’arboriculture fruitiére fait
l'objet aux Antilles francaises de culture hétérogéne suivant la méthode créole
ot, le verger mixte, contigii aux plantations vivriéres, comporte de nombreuses
especes en mélange. Des mousses et hépatiques colonisent souvent le trone et
les branches de ces: végétaux cultivés ou introduits et il est curieux de
constater qu’en outre de l’altitude qui agit par le microclimat, une adaptation
spécifique parait exister entre la mousse et l’hote qui l'héberge.

D'une fagon générale, partout ob nous avons observé des groupements
muscinaux corticicoles sur les arbres cultivés, les rhizoides de ces muscinées
n'étaient pas fixés directement sur l’écorce de l‘arbre mais sur le lacis
formé par les racines et les tiges enchevetrées de fougéres et surtout de
Polypodium polypodicides L. et de P. astrolepsis Liebm. D’autres épiphytes,
surtout des Orchidées: Jacquinella globosa (Jacq.) Schltr. et Epidendrum sp.
pl. etc...., se fixent également sur le m&me support avec des lichens et les
mousses.»

D'aprés les nombreux relevés effectués, il ressort l'électivité suivante:

Sur manguiers; Macromitrium mucronifolium (Hook. et Grev.) Schweegr.
(Jardin de Tivoli, alt. 300 m.)

Macromitrium apiculatum (Hook.) Brid.
(Fonds Saint Denis, alt. 480 m.)

Celymperes richardii CoM.
(Jardin de Tivoli, alt. 300 m.)

Sur arbre & pain: Macromitcium cirrhosum (Hedw.) Brid.
(Camp de Colson, alt. 500 m.)

Macromitrium apiculatum (Hook.) Brid.
(Morne Rouge, alt. 520m.)

Caribbean Forester - 176 - July 1943 j

Sur orangers: Syrrhopodon martinicensis Broth.
Lepidopilum integrifolium Broth.
(Fonds~Saint-Denis, alt. 480 m.)
Macromitrium perichaetiale Hoox. et Grev.
(Camp de Balata, alt. 450 m.)
Macromitrium cirrhosum (Hedw.) Brid.
(Camp de Colson, ait. 500 m.)

Frullania Kunzei L. et Lindenb.
(Camp de 1'Alma, alt. 500 m.)

Sur pommiers-cythéres: Macromitrium mucronifolium (Hook. et Grev.) Schwaegr.
(Poste forestier de l’Adonis-Balata, alt. 450 m.)

Sur cocotiers; Macromitrium Husnoti Schp.
Macromitrium mucronifolium (Hook. et Grev.) Schwaegr.
(Morne Rouge alt. 375 m. et Camp Colson, alt. 500 m.)

Sur palmiers; Roystonia regia Bailey: Pterogonidium pulchellum (Hook. }
CM.

(Poste forestier de l’Adonis-Balata, alt. 450 m.)

Chrysolidocarpus lutescens L.: Calymperes Richardii C.M.
(Jardin de Tivoli, alt. 320 m.)

Elacis guineensis L.: Octoblepharum albidum (L.) Hedw.
(Jardin de Tivoli, alt. 310 m., également sur Oreodoxa).

Il y a lieu cependant de faire intervenir la proximité de la forét ou des
rivitres qui hébergent des mousses. Ainsi, les abords de la Rivitre Mademe &
Tivoli, étant des réserves &: Taxithelium planum (Brid.) Mitt., Vesiculeria
amphibola (Spr.) Broth. Isopterygium tenerum (Sw.) Mitt. et Sematophyilum
caespitosum (Hedw.) Mitt. sur des talus humiféres et ses racines de bambous,
ces espéces se retrouvent @ la base des manguiers ou orangers voisins et des
divers arbres cultivés. A Saint-Joseph sur l*habitation Deslands, humide et
peu éloignée de la forét humide, ce sont 1’Hookeriopsis guadelupensis (Brid.)
Jaeg. et le Calicostella Belangeriana (Besch.) Jaeg., qui cclonisent la base
des manguiers.

Néanmoins, l'’affinité du genre Macromitrium pour les arbres fruitiers
de culture ressort nettement quels que soient les secteurs et l’altitude en-
visages. Le M. apiculatum préfére les écorces de manguiers, le M. cirrhosum
celles d'arbres & pain, le M. perichaetiale celles d'orangers, le M. mucronife-
dium celles de pommiers-cytheres et le M. Husnoti celles de cceotiers. Ilya
done 1&4 un Macromitrietum corticicole des arbres fruitiers des Antilles fran-
caises. Les lisiéres culturales arbustives ont également leurs muscinéés
électives. Les espéces du genre Macromitrium y sont plus rares, remplacées par
des Hookériacées. Ainsi, au Morne Rouge et au Camp de Colson ot les lisiéres
culturales sont entre 400 et 600 m. d‘altitude on trouve sur les gliricidias:;

Gliricidia sepium (Jacq.) Steud., sur des racines de Polypodium polypodioides L.
Lepidopilum radicale Mitt.

Crossomitrium subepiphyllum Besch.
Macromitrium mucronifolium (H. et G.) Schwaegr.

Sly, July 1943

Ces espéces constituent une association corticicole sur les troncs vivants
de Gliricidia sepium (Jacq.) Steud.

Sur les bois-carrés ou bois-c6éte: Cytharexylum gquadrangulare Desf. on
trouve en Guadeloupe, & 50 métres d'altitude (Vieux Bourg), le Thuidium involvens
(Hedw.) Mitt.

Les pois-doux du genre Inga surtout: I. laurina Willd. souvent utilisés
pour brise-vents ou arbres d'ombrage, tant en Guadeloupe qu'en Martinique, por-
tent des Macromitrium et le Frullania kunzei Lehm. et Lindenb.

Il est & noter que les arbres de lisiére conservés de la forét a la li-
mite des cultures portent des mousses plus xérohéliophiles que celles de la
forét d'ou elles dérivent par régression ou qui sont les moins sciaphiles de
cette forét qui se sont adaptés & plus d'insolation et de chaleur.

Forét Hygrophytique de 1'Etage Moyen

La forét hygrophytique, constituant l'étage moyen aux Antilles frangaises
entre 400 et 800 metres d’altitude, qui regoit une tranche d'eau moyenne de 5 &
6m. par an, est dense, humide, polystrate et sempervirente, conditions favo-
rables au foisonnement muscinal, Elle présente des aspects variés déja décrits:;
des noyaux constituant au centre des relicts préliminaires, de la forét native,
trés belle en certains points, & la Guadeloupe; autour de ce noyau, de la forét
dégradée ot l'aspect de futaie se méle aux taillis et ot parmi les essences
primaires s'intercalent abondamment des espéces régressives, des peuplements
secondaires sur humus ou sur terrains dénudés, enfin des paratypes de substitu-
tion issus de la superposition de l’action de l'homme & celle des éléments na-
turels. Alors que la sylve primitive et la forét primaire dégradée sont hété-
rogénes et pourvues de groupements muscinaux abondants sur les écorces les plus
diverses et au niveau de toutes ses strates, les peuplements purs de bambous et
fougéres arborescentes, qui sont trés homogénes, possédent au contraire des coe
lonisations bryophytiques homogénes et stables ot l'on retrouve d'une fagon
constante les mémes espéces avec la méme biologie. Ils seront étudies séparé-
ment.

A tous les stades de régression, cette forét demeure polystrate et depuis
la strate des "miscinae humifusae" (décrite in Bull. Agric. 1937. Esquisse des
Assoc. Végét. p. 223) jusqu'é la strate supérieure épiphytique au sommet des
branches des plus hauts arbres, les mousses sont représentées avec les coeffi-
cients de présence les plus élevés. Suivant le mode de vie et le substratum,
les principales strates que l'on peut distinguer sont: les muscinées terrestres
ou humicoles, les muscinées épilithes ou saxicoles, les corticicoles: celles
dont le support est constitué par les arbres vivants ou épiphytes et celles
recouvrant les troncs pourris et pourrissants, ou saprophytes; enfin les
épiphylles formant des colonisations plus ou moins étendues sur les feuilles
toujours vertes des arbres.

En forét primaire, l'abondance des muscinées corticicoles et le peu de
mousses terricoles est un fait saillant aux Antilles frangaises (voir Ecologie,
1935, p. 208), ainsi que l’absence de muscinées aquatiques dans les cours d'eau
Gloc. cit. ps 225)i.

a

Caribbean Forester - 178 - Vol. 4, No. 4

|

Lorsque la dégradation a permis la compétition de nombreux arbres plus
petits et d'un sous bois feuillu trés dense, abrité du vent, de la lumiére et
des pluies d'orage, l*’épiphytisme sur feuilles prend une large place et parmi
ces 6piphylles, les mousses dominent avec les hépatiques, des lichens et des
algues (Ecologie p. 215). lLeur phytosociologie est particuliére et elles sont
l‘indice d'une humidité hygrométrique élevée et du calme ambiant de l‘atmosphére
des bois. Les peuplements entitrement substitués possédent surtout des mousses
saprophytes & leur base sur les racines enchévetrées ou fibreuses et & peine au
dessus du niveau du sol, 1& ot la matiére organique s‘accumule.

Pour déterminer avec précision les associations muscinales qui entrent
dans le complexe muscinal de la forét hygrophytique, des comptages minutieux,
la détermination des complexes in situ, dans leur strate et des hdtes qui les
hébergent, sont nécessaires. L'étude de la composition intime des strates
doit d'abord @tre effectuée en faisant ressortir l’action du milieu inerte ‘et
vivant.

1. Strate des muscinées terrestres ou humicoles.—Cette strate est dis-
continue, sporadique et n’existe pas partout en forét ot les brindilles, les
détritus végétaux et de toutes sortes, jonchent la surface du sol et ne per-
mettent pas l‘installation des bryophytes terricoles. Une surélévation du sub-

stratum est pour les mousses de la for&t humide une condition plus favorable.

Cependant, cette, réduction comparative des mousses et hépatiques n'est
pas si poussée qu’elle ne puisse permettre de distinguer des colonisations
terrestres sur les talus forestiers qu‘ils soient volcaniques, latéritiques ou
humiféres.

(a) Colonisation des talus volcaniques: Ponceux, Andesitiques ou

Dioritiques.—Sur les talus ot la roche mére d'origine volcanique a été mise

& nu, puis désagrégée, sur le sol ponceux, dioritique, basaitique ou labradori-
tique, sur les coulées boueuses et partout ot l’humus forestier a été entrainé,
sans que les terrains argiloides se soient constitués, on trouve des colonisa-
tions bryophytiques, formant parfois des tapis étendus, verts et foncés, qui
tranchent sur ce support clair. Certains sont méme de couleur noire comme ile
Dicranella caespitans Besch. Les électives des ponces et les colonisatrices
des coulées boueuses et laves éruptives s’y trouvent mélées, mais ces derniéres,
qui constituent les pionniers du premier stade de la végétation des abords des
cratéres, ne tardent pas & étre éliminées pour demeurer confinées 4 1’étage
supérieur sur les pentes et les sommets des cénes volcaniques et sur les
cailloux et m@me sur les rochers ponceux d’éruption récente. Telles sont par
exemple; Breutelia scoparia (Schwaegr.) Schimp., Dicranella perrottetii (Mont. )
Mitt. et Philonotis uncinata (Schwaegr.) Brid. Ces espéces qui sont des saxi-
coles volcaniques, suivent les rochers qui les hébergent sur les sommets, dans
leur fragmentation, leur décomposition et leur entrainement vers les parties
inférieures.

Les électives des taillis ponceux sont des muscinées proprement dites,
minces et étroites, cespiteuses et dressées en brosses courtes au dessus de la
terre; on n’y rencontre pas d’Hépatiques qui, au contraire, caractérisent les
talus argileux. Ce sont surtout des acrocarpes: Dicranellacées et Fissidenta-
cées. Les principales figurent ci-aprés.

= 179 = July 1943

Dicranella subinclinata Lor. (Syn. D. martinicae Broth.)
Dicranella hilariana (Mont.) Mitt.
Dicraneiia longirostris (Schweegr.) Mitt.
Dicranella caespitans Besch.

Fissidens martinicae Besch.

Fissidens polypodioides (Sw.) Hedw.
Microdus crispulus Besch.

Entosthodon husnoti Schwaegr.

Philonotis glaucescens (Hsch.) Par.
Raphidostegium dicranelloides Schimp.
Ectropothecium globitheca (C.M.) Mitt.

Les conditions bien définies de ce milieu humide mais d'une grande per-
méabilité ot la pluviométrie est réguliére et élevée, & une altitude de 500%
700 métres, nous permettent de désigner cette colonisation homogéne sous le nom
d'association @ Dicranella hilariana-Fissidens martinicae des ponces de la forét
humide. Elle présente son optimum biologique sur la terre des bois des Deux-
Choux, de Colson, de l'Alma, du Calvaire-Gros Morne, de la Médaille, de la
Boutaud, de L’Ajoupa Bouillon et Fonds Saint-Denis, pour la Martinique, aux
Bains-Jaunes, au Matouba, @ la Madeleine pour le Guadeloupe, ot d’ailleurs elle
est moins développée.

Du point de vue forestier ou agricole, elles indiquent un sol léger d'
éruption relativement récente, dont la perméabilité est corrigée par une forte
pluviométrie et qui peut 6tre valorisé par la reforestation.

Il existe une autre colonisation, & petit nombre d'espéces, également
homogéne, sur terres volcaniques, dioritiques ou andésitiques, trés humides et
sur les canaux d’écoulement des eaux dont ces pierres forment les revétements
extérieurs. Elle est constituée par des Bryacées uniquement.

Bryum rubrifolium Schimp.

Bryum crugeri (Hampe) C.M.

Bryum argenteum L.

Splachnobryum obtusum (Brid.) C.M.
Splachnobryum mariei Besch.

Ces espéces, dans les chemins arrosés du Morne Rouge, aux Deux-Choux au
Camp de Colson, en Martinique et & Gourbeyre, au Camp Jacob, @ Saint-Claude en
Guadeloupe constituent le Bryetum des terres volcaniques constamment imbibées
d'eau, aux Antilles frangaises.

(b) Colonisation des talus argileux: Hydrargileux ou Lateritoides.—Les
ponces légéres et tuffs volcaniques, les roches méres décomposées, les dép6ts

rivulaires ou éoliens, facilement solubles, étant entrainés par les eaux, un
processus de latéritisation, plus ou moins poussée et aboutissant parfois
jusqu'a la kaolinisation du sol, s'ensuit. Des argiles diversement colorées

par des oxydes de fer et de mangantse, des hydrargiles et hydrargilites ou des
latéritoides apparaissent aprés décapement de la couche arable et le sol est
dénudé. Avant que toute autre végétatiem prenne place, ce sont les Bryophytes
qui apparaissent d’abord, surtout des Hépatiques contrairement & la colonisation
muscinale des talus ponceux. Ces hépatiques sont fortement adhérentes au sol

‘

Caribbean Forester - 180 = Vol. 4, No. a

contre lequel elles se plaquent, s’étalent largement et s'étendent peu & peu
pour former parfois de larges plaques minces. Elles sont mélées & des mousses
proprement dites, dressées, au milieu desquelles elles forment une t&che plate
et d'un vert plus clair ou constituent méme des petits groupements d'hépatiques
seules.

Ce sont surtout des Marchantiacées du genre Marchantia et des Ptilidia-
cées du genre Isotachis pour les Hépatiques qui caractérisent ces groupements
dans lesquels des muscinées (str. sens.) de diverses familles sont electives
mais dont la présence n’est pas aussi constante que les Hépatiques. Blles
figurent ci-aprés.

Hépatiques: Marchantia chenopoda L. (Syn. M. brasiliensis Lehm. et Lindenb.)
Marchantia papilleta Raddi (Syn. M. domingensis Lehm.et Lindenb.)

Manoclea gottschei Lindenb.
Isotachis erythrorhiza (Lehm. et Lindenb.) Steph.

Isotachis auberti (Schwaegr.) Steph. (Syn. I. serrulata (Sw.) Mitt.
Cephalozia obtusata Steph.

Syzygiella contigua (Gott.) Steph.

Anthoceros dussii Steph.

Anthoceros papulosus Steph.

Cyclolejeunea accedens (Gott.) Evans

Lopholejeunea sagraeana (Mont.) Schiffn.

Muscinées: Calymperes guildingii Hook. et Grév.
Funaria bonplandii (Brid.) Broth.
Dicranella herminieri Besch.
Hyophila mollis Broth.
Hookeria densiretis Broth.
Hookeria densirostris Broth.
Hookeris entodella Besch.
Pogonatum glaucinum Besch.
Bryum mnicides Broth.

Les Hépatiques, surtout: Marchantia chenopoda L. et M. papillata Raddi,
ainsi que Isotachis erythrorhiza (Lehm. et Lindenb.) Steph. et I. auberti
(Schwaegr.) Steph. sont trés communes sur les sols ocracés et humides formant
de larges plaques et sur les talus latéritiques suintants, les Lejeunéacées
et Hookeriacées citées colonisent abondamment. Les savanes argilo-ferrugineu-
ses et les sols forestiers latéritisés du Lamentin, la Médaille, Balata, etc.,
a la Martinique, hébergent ces électives qui constituent l'association &
Marchantia chenopoda-Dicranella herminieri des talus latéritoides des Antilles
frangaises. Elles désignent des sols ot: le complexe argilo-humique est élevé,
trés imperméables et humides, dans lesquels la plus grande partie des matiéres
organiques a été entrainée. Ces sols sont difficilement valorisables.

(c) Colonisation des talus forestiers humiféres.—Sur les talus, au bord
des traces ou chemins forestiers, dans les clairiéres, en sous-bois, sur les li-
sitres ou sur les alluvions déposés par les cours d’eau torrentiels, des colo-
nisations muscinales humicoles se développent lorsque la terre posséde des
brindilles, des débris fibreux de racines ou de branches. Lorsque la matiére
organique a été entrainée, ces groupements disparaissent également pour laisser

= igh = July 1943

place & l'association bryophytique des latéritoides. La compétition des espéces
est plus intense que sur les talus ponceux et on y rencontre aussi bien des
Hépatiques plaquées que des mousses dressées ou méme cespiteuses.

Les relevés effectués dans les divers secteurs de la Guadeloupe et de la
Martinique nous permettent de désigner comme électives les espéces suivantes:

Hépatiques: Riccardia stipatiflora (Steph.) Pagan mst. (Syn. Aneura
stipatiflora Steph.)
Riccardia fucoides (Sw.) Schiffn.
Symphyogyna trivit trivittata Spruce
Odontoschisma denudatum (Mart.) Dum.
Anthoceros cucullatus Steph.
Anthoceros flexivalvis Gottsche et Nees.

Muscinées: Floribundaria usneoides Broth.
Dicranella stenocarpa Besch.
Hyophila guadalupensis Broth.
Philonotis sphaericarpa (Sw.) Brid.
Philonotis subsphaericarpa Broth.
Philonotis minuta Schweaegr.

Pogonatum pleanum Besch.

Pogonatum laxifolium Besch.
Lepidopilum Lapidopilunlpolytricoides (Hedw.) Brid.

Homalia glabella Sw.

Stereophyllum matoubae Besch.

Ectropothesium leucocladium Schimp.

Hookeria hyalina Schump.

Hookeria subsimplex Broth.

Ces terres trés humiféres peuvent étre désignées sous le nom d'asso-
ciation & Riccardia stipatiflora-Floribundaria usneoides. En outre de ces élec-
tives, il y a lieu de citer quelques espéces dont le réle colonisateur ne s'ar-
réte pas & l’humus des talus mais qui, de celui-ci, passe sur l’humus des
écorces d’arbres, en s’élevant de bas en haut le long des brindilles, des raci-
nes tragantes, puis des contreforts inclinés. Ce sont:

Calymperes donnellii Aust.

Macromitrium husnoti Schimp.

Calicostelle belangeriana (Besch.) Jaeg.
Syrrhopodon flavescens C.M. (Syn. S. dussii Broth.)

Le Floribundaria usneoides Broth. posséde la faculté de s’installer sur
les talus non protégés, de se blottir dans les moindres cuvettes que 1’écorchure
des parois fait apparaitre & leur surface, de peupler progressivement ces talus
et par la décomposition continue de ses feuilles usnéiformes, ténues et abon-
dantes, de donner une quantité minima de matiéres organiques, permettant ainsi
la recolonisation des pentes dénudées par d'autres espéces. C’est un pionnier
de colonisation dont le réle édificateur d*humus est a,souligner.

(To be concluded in the October 1943 issue.)

Caribbean Forester - 182 = Vol. 4, No. 4 F

POMARROSA, JAMBOSA JAMBOS (L.) MILLSP.

AND _ITS PLACE IN PUERTO RICO

Frank H. Wadsworth
Tropical Forest Experiment Station

As has been previously pointed out,4/ normally the most critically
scarce forest products in Puerto Rico are fuelwood, charcoal, posts, and other
small products vital to our large rural population. It is true that local lum-
ber production is negligible, but there is no reason to suspect that importa-
tion of this product will be more difficult or more expensive in the post-war
world.

Because of population density only lands clearly non-agricultural in
character are available for forest production. Nearly 90,000 acres of this
class of land are in government ownership under the administration of the
Federal and Insular Forest Services. A much larger aggregate area, however,
and one of greater potential value for this purpose because of its proximity
to consumers, is in privately-owned tracts scattered throughout the island.

As an overwhelming majority of forest land owners here consider these areas

of negative or negligible productivity, basing their evaluation only upon the
return from immediate and complete liquidation, it is the forester's responsi-
bility to provide inducements to improved management practices through the de-
velopment of silvicultural methods yielding a maximum return upon a minimum of
outlay. Imperative to the achievement of this goal is the selection of
rapidly-growing tree species requiring little care.

One tree species which deserves careful consideration because of its
high productivity of fueiwood and posts, is pomarrosa, Jambosa jambos (L.)
Millsp. This species, introduced into the West Indies from the Far East many
years ago (prior to 13212/), became widely distributed through natural repro-
duction and is now found in dense pure stands on abandoned lands throughout
many parts of the island, particularly in the humid Cordillera Central. In
some regions the only existing forests are composed purely of pomarrosa.

The name “pomarrosa" (rose apple) is derived from the consistency and
odor of the fleshy fruit. The tree seldom grows to more than 6 inches d-b-h.,
although 14-inch trees have been seen. Likewise it is short, generally less
than 35 feet tall. The main stem generally branches several times, the first
fork usually being less than 6 feet above the ground. Following cutting a
large number of sprouts are produced, giving the tree the appearance of a
large shrub. Because of the intense competition between these sprouts their
stems are often fairly straight for 190 feet.

aU Wadsworth, Frank H. The evaluation of forest tree species in Puerto Rico,
as affected by the local forest problem. Caribbean Forester, January 1943.

2/ Descourtilz, M. E. Flore pittoresque et médicale des Antilles. Paris,
1829.

- 183 - July 1943

The wood of pomarrosa is light colored, hard, and heavy. It is strong
and is used in the round for various construction purposes. Its chief use is
for fuel, and it makes an excelient charcoal. Because of its weight it is
not highly esteemed for hoe handles. Although it is not outstanding in dura-
bility in the ground it is frequently used for fence and barn posts and vege-
table stakes. It is widely used in the tobacco region for the construction
of drying sheds, and an important part of our annual consumption of approxi-
mately 4,500,000 tobacco stakes3/is supplied by this species.

Productivity

Pomarrosa deserves consideration as a species for farm forests chiefly
because of its high productivity. While seedling trees grow slowly, coppice
growth following clear-cutting of an established stand is phenomenal. Within
@ year a dense growth of sprouts will reach 12 feet in height, dominating all
competing vegetation, and with proper management a rapid growth rate will
continue until a very dense stand is formed. Thus production by coppice is
high, and, as no weeding is necessary during the regeneration period, cost
is low.

Without control of stand density pomarrosa tends to form thickets with
a@ very regular canopy, containing no clearly dominant trees. There are indi-
cations in such stands that a marked reduction in diameter growth takes place
in extreme cases possibly approaching stagnation. In a young stand in the Rio
Piedras woodlot of this Station, it wes found that the growth rate is decreas-
ing even though the basal area is but little more than 50 square feet. Here
some trees are still dominant and have grown as much as 0.7 inches in diameter
during the past year, but the annual diameter growth of the average tree is
only slightly more than 0.2 inches. .

A good indication of pomarrosa growth and the character of the stands
formed can be found in an old pure forest in private ownership near Cidra in
east central Puerto Rico. This forest, growing at an elevation of 1400 feet
and receiving an average annual rainfall of about 90 inches, is on & degraded
heavy clay soil (Cialitos). It covers about 60 acres, chiefly on slopes
bordering a stream. It is evident that the area has been forested for many
years, for there are many old stumps more than 2 feet in diameter still
sprouting. Parts of this forest have been cut periodically to supply posts,
stakes, and fuel for a bakery. As cutting has generally been clear, each
cutting area can be distinguished by the size of the trees.

In 1943 one-tenth~-acre temporary sample plots were established and
measured within this forest to determine the density of two stands of dif-
ferent ages. The desirability of marking permanent plots was recognized
but as no control can be exercised over management practices this was con-
sidered unwise. In one part of the stand clear-cut six years ago the number
of trees per acre taller than 4.5 feet and their diameter distribution were
found to be as shown in Table 1.

3/ Unpublished information supplied by the Tobacco Institute, Rio Piedras.

Caribbean Forester - 184 = Vol. 4, Noe 4

Table 1.—Stand Per Acre in Six-year-old
Pure Pomarrosa Coppice, Cidra

Inches Sq. Ft. Sq. Ft. Sq. Ft.
Below 0.6 2,544 v5) 1,945 0.8 4,489 LS
it 9,641 57.8 1,881 1G S22 69.1
2 3,676 84.6 192 4.4 3,868 89.0
3 te 7Al: 8.7 128 6.5 299 See
4 64 Dail 64 S51
5 21 2-9 21 2.9
Total L6G Ly, 160.7 4,146 2550 20,263 183.7
Per cent 79.5 87.5 20.5 W358 100.0 100.0

Several indications of the density of the stand are to be found in Table
1. While the number of trees per acre may not be surprisingly large when tree
size is also considered, the basal area is high for a stand of this age. A
result of stand density is the comparatively small number of saplings of less
than 0.6 inch d.b-h. This characteristic is at once apparent to the eye, as
the ground is virtually bere of small plants. Also an indication of density
is the subordinate position of the seedlings, which, subjected to the intense
competition of the more rapidly-growing sprouts, are unable to attain a domi-
nént position.

On the basis of the tally within the plot there are 2,159 stumps per
acre, each supporting an average of 7.5 sprouts. One large stump supports 46
stems, the aggregate basal area of which is the equivalent of a single stem ll
inches in diameter. The density of this stand can be compared with high for-
ests by assigning each stump the aggregate basal area of all its sprouts and
converting this into one stem of the corresponding diameter. On this basis
the average "tree" in the coppice alone is 3.7 inches d.b-h., and the average
spacing between "trees" is about 4.5 feet. With the seedlings included, these
figures become 2.3 inches and 2.6 feet respectively. This, coupled with an
average tree height of about 25 feet, is ample evidence of density.

This stand, if cut immediately, would yield at least 12 to 15 cords of
wood per acre worth $30 or $40 on the stump. If the posts were sold separetely
the return would be considerably increased.

Another part of this forest was older, probably having been left 12 to
15 years since the last cutting. Based on measurements in a second small
sample plot, this stand was found to be as indicated in Table 2.

To the eye this stand appears fully as dense as the younger one, yet

the basal area is less. The apparent density is due to the deep shade pro-
duced by profuse branching in the larger crowns of these trees. It appears

- 185 - July 1943

that in the older stands such as this one there is a cubic volume of branchwood
nearly equal to that in the main stems. Here, even more than in the younger
stand, the seediings are insignificant, making up less than 5 per cent of the
basal area. Almost no seedlings reach the canopy level.

Table 2.—-Stand Per Acre in 12- to 15-year-old
Pure Pomarrosa Coppice, Cidra

DeB.H: Seedlings Total
_Class Basal Area
Inches Sq. Ft. Sq. ; Sq. 6
Below 0.6 746 0.3 1,430 0.6 Zeno 0.9
1 een 6.0 816 4.1 fa rtoplts) Lore
2 1,412 $20.09 88 7309.0) 1,500 34.5
3 719 36.4 9 0.5 728 36.9
4 35 28.2 $15 28.2
5 NS 24.4 175 24.4
6 44 8.8 44 8.8
i 18 4.9 18 4.§
Total 4,631 141.5 2,543 Waste 6,974 148.7
Fer cent 66.4 95.2 33-6 4.8 100.0 100.0

The stumps are spaced at an average distance of 8 feet, or 676 per acre.
On the average they support 6.8 stems each. One stump with 40 stems has an
aggregate basal area corresponding to a single stem of 15.6 inches d.b-h. The
average aggregate stump diameter determined by this method is 6.2 inches. With
the stumps considered as individual stems of a size corresponding to the total
basal area of their sprouts, and with the seedlings included, the average d.b.h.
is 3.0 inches and the average spacing is 3.8 feet. The total height of these
older trees is only slightiy more than that of the younger stand, probably
averaging less than 30 feet.

Silvical Characteristics

Some of the more desirable characteristics of pomarrosa have already
been mentioned. Its high productivity as a coppice makes it outstanding for
private forest management in certain parts of the island. Once established,
it can cope with herbaceous competition by rapid growth and immediate domina-
tion of the site following cutting. While the growth rate of seedling trees
is considerably slower than that of coppice, seedlings appear to be equally
able to fight through herbaceous vegetation. In many parts of the isiand
young seedling stands can be found in brush or in dense grass, gradually at-
taining dominance without assistance.

Any species able to form forests on this overpopulated island must be

tolerant of degraded sites considered too poor for agriculture. Pomarrosa has
proven very successful in this respect. Although it is not known what effect

Caribbean Forester - 186 - Vol. 4, Now 4

site has upon its growth rate, even on exhausted steep slopes it flourishes
without indication of chlorosis.

ae Spe , LS st
Fig. 1.—An old pomarrosa stump with 25
sprouts averaging 2 to 3 inches d.b-h.

= ae

A criticism of pomarrosa has been that it forms such dense shade that
on steep slopes no subordinate vegetation grows on the forest floor, and thus
the soil is exposed to sheet erosion. This has been observed where the soil
is of an erodable type, and it takes place even on rather gentle slopes.
Certainly such a forest could never be expected to retain any considerable
quantity of water entering it from higher up a slope. However, it has been
found that frequent light cuttings, which can be made at a profit, will main-
tain a sufficient open canopy to provide adequate ground cover to protect the
soil. Ina quarter-acre stand improvement plot a light opening resulted in
quick invasion by a herbaceous ground cover, largely of Palicourea sp., which
has no detrimental effect upon the stand.

Pomarrosa has been criticised also because of its massive root system
which makes land-clearing very arduous. The importance of this characteristic

- 187 - July 1943

as a “disadvantage" veries with the point of view. If pomarrosa is to be grown
tor but one rotation and then the lend is to be cleared for cultivation, this
is certainly a factor to consider. Under this policy tne rapid coppice growth
is not realized, as the trees are all seedlings, and therefore other species
should be recommended. On the other hand, if the lend supporting the forest is
clearly non-agricultural, and therefore should not return to cultivation, this
Gisadvantage disappears, and in fact might be considered an advantage, es it
discoureses an undesirable practice.

Netural Regeneration

If the most important factor in the development of our volunteer po-
merrosa forests has been the ability of the species to fight through competing
vecetation, certainly next in importance is tne vigor of its seeds. The sceds
ere made up of several sections, one to four of which have embryos which may
germinate, thus providing a safety factor. For this reason fresh seed always
has a germinetion percentage "greater than 100". As the seeds vermineate best
in e shady and moist environment they are well adapted for conditions on which
competing vegetation is already growing. Germination is slow, requiring 10 to
2O days to start, and continuing for as long as 120 days. The endosperm pro-
vides sufficient food for the young plent for several months, and new sprouts
are produced if the first stem becomes broken off any time during this period.
The natural seedlings apparently can withstand shade sufficiently dense to
prevent much growth during the first few years, as meny 24-inch wildings ere
tound with very thick root collars and extensive woody root systems.

The discovery of numerous pomarrosa seeds or seedlings at some distence
from the nearest parent tree is not infrequent end is rather unexpected with a
species having heavy seeds (175=200 to the pound). Verious speculations have
been made to explain this, particularly where the scedlings come up in a com-
pact group of 30 or more within ae square foot. Although it is reasonable to
suppose that bats or rodents may carry the seeds or fruits, this has not been
sezn by the writer.

Artificial Regeneration

The main seed production period here is in the spring, from April to
June. The seeds fall gradually throughout this period, end therefore collec~
tion of ea large quantity at any given time necessitates coverage of a large
area, collecting a few from the ground beneath each tree. The fresh seeds
have e 50 per cent moisture content and do not keep well, After unscaled
storage for a month at room temperature one sample had a viability of only 18
per cent. Another lot, stored unsealed at 35-40° F., retained 50 per cent
viability for 3 months.

During 1942 a plot experiment was laid out on two sites to test various
metnods of artificial regeneration of this species, using methods which are
not difficult or costly. The results indicate that, taking advantage of the
characteristics which contribute to pomarrosa's abuncant natural reproduction,
the artificial regeneration of this species will not prove difficult. One
site used was in the Rio Abajo Insular Forest in the moist limestone region
of the north coast. The elevation above sea-level is 500 feet and the annual

Caribbean Forester - 188 = Vol. 4, No. 4 i

precipitation is 80 inches. The other site was in the Toro Negro Unit of the
Caribbean National Forest in the Cordillera Central, at 2,700 feet elevation,
receiving an annual rainfall in excess of 100 inches. Two ground preparation
treatments were tested, (1) complete clearing and burning prior to sowing or
planting, and (2) cleaning only the area within 1 foot of the seeds or seed-
lings. A set of untreated control plots was also included. Methods of estab-
lishment included broadcasting, spot seeding, and planting. Each treatment

was tested in two randomized 20 x 49-foot plots at each site. The seed spot
and planting spacings were 4 x 4, permitting 50 per plot. Four seeds were sown
per spot. The broadcasting was at the rate of 6, 8, and 10 seeds per square
foot in the burned, weeded, and untreated plots respectively. In the broadcast-=
weeded plots the seeds were sown in cleared strips two feet wide.

All phases of the study have not been completed as yet, but some of the
results are already clear. During the first year the seedlings have grown
slowly, the tallest being about 18 inches. The nursery stock was not ready
for transplanting until it was 5 months old. It was slow to recover from the
shock, but now after 6 months it is starting to grow. The data are shown in
Table 3.

Table 3.—Survival of Pomarrosa Sowing and Plantingl/

Cordillera Central

Treatment Block II Average
Spy f Sr Y, LZ 12
Broadcasting
Burned 0 0 0 hat sk 6
Weeded 6) 6) 0 ae 5 14
Untreated 5 @) 2 52 43 48
Seed Spots
Burned 118 90 164 79 45 62
Weeded 100 101 sere) 90 75 82
Untreated 119 118 114 78 104 91
Planting
Burned 46 52 49 96 80 88
Weeded 58 26 42 70 94 82
Untreated 40 38 39 98 94 86
General Average 53 47 66 60

a7 Broadcasting and seed spot survival after 12 months, planting sur-
vival after 6 months.

of As frequently more than one seedling was SRoaueeal: from an individual
seed, it was impossible to accurately determine which had germinated.
Therefore, each plant was counted, making these percentages higher than
the true germination percentage.

- 189 - July 1943

It will be noticed that the relationship between the results at the
two sites is not consistent throughout all treatments. This difference re-
sulted from shallow sowing in the seed spots in the Cordillera. These seeds
were soon exposed by hard rains and many died as a result of insolation. It
is probable that, barring this error, the survival would have been higher in
all treatments at high elevation.

The results of broadcasting are not yet final, as with no care subse-
quent to the sowing, the seedlings must still attain dominance over the herba-
ceous vegetation, which is now dense in all plots. There are no signs of
appreciable losses, so the results to date measure the effect of the various
environments upon germination, rather than upon ultimate survival. The data
from the broadcasted plots corroborate laboratory tests which show that the
seeds require a very moist environment during the long germination period.

The contrast between the germination in the treated and untreated plots is
sharp, and is largely due to the effects of shade in the untreated plots.

The losses in seeds sown in direct sunlight were obvious after only a few
days. The poor survival in the untreated plots in the limestone region is
probably partly due to higher temperature and lower rainfall but also to the
fact that there the grass was shorter as a result of previous grazing, permit-
ting direct exposure of many of the seeds to the sun. In these plots it was
noticed that nearly all of the seedlings appeared from beneath the denser
clumps of grass where the seeds had been well shaded. If future losses are
not large, broadcasting without soil preparation may well prove to be the most
practical method, as it is by far the cheapest, and, using the sowing density
of this study, only a one per cent survival would be necessary to obtain a

4x 4 spacing. Probably much less seed is actually necessary.

The seed spots are considered satisfactory so far at both sites despite
the fact that, as has already been mentioned, the survival at the limestone
site is significantly superior due to shallow sowing in the Cordillera. At
least 90 per cent of the spots in every treatment have at least one seedling.
No significant difference was found between ground preparation treatments and
none should develop, as no more care is planned. Observations will continue
in order to record the effects of competition upon survival and growth.

As it is but 6 months since the planting was done, the data indicate
only actual planting loss. This loss has proved to be significantly higher
in the limestone region, again bringing out the preference of this species
for cool moist sites. No relationship between ground preparation treatments
was anticipated, but should the trees appear unable to cope with herbaceous
competition at any time in the future, some of the plots will be weeded for
comparison.

With observation of the abundant wildings beneath existing pomarrosa
forests came the thought that they might constitute an important source of
planting stock. One hundred wildings 18 to 24 inches tall were pulled by
hand at the aforementioned site in the Rio Abajo Insular Forest and in the
Luquillo Unit of the Caribbean National Forest in the eastern mountains at
about 1000 feet elevation (100 inches annual precipitation). These were re-
planted at their respective locations on sites covered with a dense stand of

Caribbean Forester - 190 = Vol. 4, No. 4

grass and herbaceous vegetation 2 to 4 feet tall. After 6 months without care
survival is more than 90 per cent on each site, and growth is starting after a
somewhat lengthy period of readjustment. With this success in mind it would
seem that if the final data from the study now in progress should show that
planting is to be preferred to broadcasting or spot seeding, wilding stock
should be used to the fullest possible extent before inaugurating a nursery
program.

Management

Judging from observations in existing pomarrosa stands, management
should be a simple matter. The stand described at Cidra has been "managed"
on a clear-cutting system very successfully for at least 4 rotations and the
stumps are still vigorous. The chief fault to be found with present practice
in this stand is that the soil is kept bare beneath the overdense canopy,
permitting sheet erosion and thus site deterioration. In a few small areas
in this forest where partial cuttings have been made the greater ground cover
density is obvious.

Fig, 2.—A cutover area within an old pomarrosa coppice
forest, showing the large stumps and the dense stand in the
background. The lack of herbaceous vegetation and sapling
growth indicates the dense shade which prevailed prior to
cutting.

- 191 - July 1943

It is difficult to make any definite statement es to optimum stand den-
sity, es thinning experiments have not been made. However, it is probable tnet
cubic foot yield is greatest in stends containing less then 100 squere feet of
besal area. In the rapidly-growing coppice at Cidra limitation to tais fizure
would require (or permit) thinning at 2 year intervals foilowing the second or
tnird yeer, and would probably admit a protective herodaceous ground cover.
Thinning should be from above, as there is little reason to encourage the domi-
nants to grow to large size, certainly not to more tnan 6 or 8 inches d&b.h.,
as the stems, being straight for only a relative snort length, do not rapidly
increase in velue witn additional growth. The most crooked stems which will
serve oniy tor fuel should be taken first, regardless of position in the canopy
or potential growth rate, and then, if necessary, the dominants. working toward
@ uniform final crop of posts and poles. It is probable that on a rotation of
10 to 15 years this policy will provide cubic foot volume in thinnings equal to
thet of the final cut and will greatly increase productivity over that in dense
unthinned stands.

One practice which is common here and snould not be permitted in well-
managed torests is that of cutting nigh stumps (See Fig. 2). In the Cidra
forest four rotations were visible, the cuttings having been made progressive-
ly higher until some of the last poles were cut at 8 feet above the ground,
resulting in waste end the exposure of dead stumpwood to termites and fungi.

It is said that pomarrosa has become a serious weed in some parts of the
fmerican tropics, as in the Pinus occidentalis forests of the Dominican Repub-
lic. There the dense understory of pomarrose prevents reproduction of the
intolerant pine, The only logical treatment under such circumstances seems to
be poisoning with sodium arsenite, although this method is expensive end, to
tne writer's knowledge, has not been tried. Attempts at removal by frequent
weedings will probably be equally costly and less certain of success because
of persistent sprouting and the shade tolerance otf the species. The introduc-
tion of pomarrosa into an area which is to sooner or later be converted into a
high torcst may well present serious problems in the future, particularly if
conversion to a shade intolerant species is anticipated. For this reason its
use is recommended only in regions where a permenent fuel and post market is
assured.

While the products of pomarrosa sre badly needed now as a result of
forest depletion, it is by no means recommended, on the basis of present know=
ledge, that every forest acre be plented to this species. Pomarrosa forests
can be of greetest service in wood-starved intensive farming regions such as
the sugar and tobacco producing areas. Here only a small percentase of the
land aree is non-agricultural, in many places the only locations available for
trees being fence rows. However, on areas witn steep slope or rocky soil,
many of which are now used only for poor pasture, well-managed pomarrosa ftorcsts
can increase income while affording permanent protection for the soil.

Ceribbeen Forester - 192 = Vol. 4, Neo

4

Summary

The discovery of forest tree species wnich grow rapidly and require
little care is a prerequisite to improved management of forests in private
ownership in Puerto Rico. Observations and studies indicete that one species
deserving consideration is pomarrosa, Jambosa jambos (L.) Millsp. Pomarrosa
does not grow to a large diameter, nor very tall, and the wood is not used for
lumber. However, when the high demand for posts and fuel is considered, this
species can be considered of importance because of its suitability for these
products. Sample plot measurements in en old coppice stand indicate that
Sprout growth is phenomenal, producing a very dense stand within a few years.
The volume in a 6 year old coppice stand was estimated at lz to 15 cords per
acres

Since its introduction into Puerto Rico many years ago pomarrosa has
formed forests on abandoned lends throughout the moist parts of the island.
Tnis is largely due to the ability of its seedlings to fight through herbaceous
vegetation and then to form such a dense shade that its dominance is assured.
Also of importence is the fact that it grows well on some of our worst soils.

Netural regeneration of pomarrosa is abundant, and seedlings are often
found at some distence from the nearest tree, despite the fact that the seeds
are heevy. As 3 or 4 seedlings may come from each seed the “germination per-
centage" is high, A partially completed study of artificial regeneration,
indicetes thet broadcasting and direct seeding will prove successful, thus ob-
viating nursery expense. The study also brought out the importance of snade
and moisture during the germination period, pointing to the desirability of
leaving undisturbed the existing vegetetion on the planting site. If planting
is preferred, the large available supply of wilding stock in and near existing
stands should be used wherever possible. In the nursery the trees grow slow-
ly, requiring about 6 months before lifting.

Management appears to be simple. Clear cutting produces a dense cop-
pice growth which will rapidly dominate and protect the cutover area. Pomarro-
se tends to form overstocked stands and may stagnate if not occasionally
thinned, possibly as often as every two years. Frequent thinning, while pro-
viding an income, will also admit a herbaceous ground cover to protect the
soil on slopes. Thinning should remove deformed trees first and then dominants,
working toward a uniform final crop of posts or poles of 4 to 8 inches in diam-
eter- Such a policy snould not require a coppice rotation longer than 15 years.

Fomarrosa can be a weed under certain circumstances and should not be
introduced into areas as a nurse crop for other species, as it will be difficult
to eliminate, because of its shade tolerance. However, where a post and fuel
market is assured, well-managed pomarrosa stands can provide frequent income
which will probably exceed that of pasture on the poorer lands, and can perma-
nently protect the soil.

- 193 = July 1943

Resumen

Un requisito previo para lograr el mejor aprovechamiento de los bosques
privados de Fuerto Rico es el descubrimiento de especies forestales de creci-
miento raépido y que requieran poca atenciédn. Las observeciones y estudios
efectuados indican que la pomarrosa, Jambosa_jambos (L.) Millsp. es una de las
especies que merece consideracién especial, La pomarrosa no crece mucho ni en
altura ni en diametro y su madera no se usa en construcciones. Sin embargo
cuando se toma en cuenta la gran demanda de postes y lefia, la importancia de
esta especie es obvia debido a lo bien que suple estas necesidades, Medidas
efectuadas en cuarteles de prueba sefialados en tallares de pomerrosa indican
que el crecimiento de los renuevos es fenomenal, produciendo rodales muy den-
sos en pocos efios. En un tallar de 6 afios se obtuvo un volumen de 12 a 15
cuerdas de madera por acreo

Desde que se introdujo en Puerto Rico hace muchos afios, la pomarrosea
ha formado bosques en tierras abendonadas en todas las regiones htmedas de la
isla. Esto se debe esencialmente a la habilidad que tienen sus brinzales de
luchaxr con la vegetacion herbécea y formar después una sombra tan densa, que
asegura su dominacion., También 6s importante el hecho de que crece bien en al-
gunos de nuestros peores suelos.

La regeneracién natural de la pomarrosa es abundante y los brinzales se
encuentran a menudo a cierta distancia del arbol m&s cercano atm a pesar de
que la semilla es pesada. Como de cada semilla nacen 3 o 4 brinzaies, la capa-~
cidad germinativa es muy alta. Un estudio en vias de terminacion indica que
la siembra al welo y la siembra directa son satisfactorias obviendo asi los
gestos de vivero. Bl estudio sefialé tembién la importancia de la sombra y de
la humedad durante el perfiodo de la germinacion lo cual demuestra que no debe
eliminarse la vegetacién existente en el sitio elegide pera la siembra, Si
se prefiere el trasplante deben usarse los brinzeles silvestres que crecen en
abundancia en y cerca de los rodales ya establecidos., En el vivero los arbo-
litos crecen despacio y necesitan cerca de 6 meses para poderse trasplantar,

Bl aprovechamiento parece ser simple. Después de la corta total, el
denso brote de renuevos domina rapidamente y protege el area talada. La po-
marrosa tiene la tendencia de former rodales espesos y se estanca si no se
aclara de vez en cuando, posiblemente cada dos afios, Bl clareo frecuente,
ademas de ser uma fuente de ingreso, protege el suelo en las pendientes ya
que permite también la formacion de una superficie herbacea, eben removerse
primero los arboles deformes y luego les dominantes para asi producir final»
mente una cosecha unifforme de postes co espeques de 4 a 8 pulgadas de diémetro.
Siguiendo esta politica las rotaciones deben hacerse de no mas de 15 afios.

Bajo ciertas circunstancias, la pomerrosa puede tornarse arbusto y por
lo tanto no debe utilizarse como tutor de otras especies pues serd& dificil de
eliminar debido a su tolerancia, No obstente, en aquellos sitios donde el
mercado de postes y lefia esta asegufedo, los rodales de pomarrosa convenien-=
temente manejados pueden proveer un ingreso frecuente que probablemente exce=
da al yue rinde el pastoreo en los terrenos pobres y puede ademas proteger el
suelo permanentemente 5

Caribbean Forester 2» 194 = Vol. 4, Noo

:
f
:
e

INDEX TO
VOLUMES 3 AND 4 OF THE CARIBBEAN FORESTER
OCTOBER 1941 TO JULY 1943

American tropics, a forest policy for the IV,
Antilles Frangaises, Catalogue des Cryptogames vasculaires des IVeToD
Antilles Frangaises, Classification des arbres & latex et &

secretions de gommes, resines et matiéres colorantes aux LV,
Antilles Frangaises, et son intérét dans la valorisation

sylvicole, La vegetation muscinale des LV 5
Antilles, Montane vegetation in the Leas
Barbour, William R. Lie
Beard, J. S. Daisies G15 SAO saves
Bevan, Arthur LV.5
British Honduras, Forest associations of ia
Brooks, R. L. Diy, 1253
Burns, L. V. IV,
Campeche plantation, The Las Cobanitas EViS
Carabia, J. P. LEEW lOr
Cedar in Trinidad, Summary of silvicultural experience with Lae
Cedrela, Comments on the silviculture of ING
Celebracién del Dia del Arbol Teielas
Classification des arbres & latex et a secretions de gommes,

resines et matiéres colorantes aux Antilles Frangaises. Ves
Classification des essences forestiéres de la Martinique d’aprés

leur utilisation. Dies
Conservacién de los recursos naturales, la; El Problema, sus

diversas fases y la importancia relativa de éstas.(Continuacién) AEIEIC
Creosote oil in the wood of Pinus occidentalis Swartz, Retention of LV,
Creosote penetration in tabonuco wood as affected by preliminary

boiling treatments in organic solvents. Ve

- 195 -

49

83

i2

164

137

114

LZ

Croton eleuteria and Croton cascarilla, The question of HeGib yaa)

Croton en Cuba, El género III, 114
Cryptotermes brevis Walker, How to make wood unpalatable to the

the West Indian dry-wood termite. I. With inorganic compounds. IV, 145
Cryptogames vasculaires des Antilles frangaises, Catalogue des IV, 35, 83
Cuba, Bl género Croton en Iii, 114
DeLeon, Donald Tip ac
Dia del Arbol, Celebracién de Iii, 89
Dominica, A synopsis of the palms of TII,, 163
Dominican Republic, Notes on some forest insects found in

Pinus occidentalis Swartz near Jarabacoa TDD e482
Encina, Quercus virginiana Mill. Vie eESS
Evaluation of forest tree species in Puerto Rico, as affected

by the local forest problem, The Iv, 54
Florida, Future may see mahogany forests in IV, 124
Forests and forest entomology Iv, 132
Forest associations of British Honduras III, 164

Forest insects found in Pinus occidentalis Swartz near

Jarabacoa, Dominican Republic, Notes on some Til, 42
Forest policy for the American Tropics, 4 Iv, 49
Forest policy of Trinidad and Tobago, The Ti leeled
Forest types of Tropical America TLE peusi7,
Forestry and forest resources in Mexico LV cone!

Formation and management of mahogany plantations at Silk

Grass Forest Reserve, The Lie no
Fors, Alberto Jd. Iv; 158
Gonzalez Vale, Manuel A. Deda 5 vet
Grades of broadleaved mahogany planting stock, Study of PD se ane
Haiti, The pine forestsof Iv, 26
Harrar, B. 8. IV, 129

- 196 =

Hodge, W. H.
Holdridge, L. R.
Honduras, Apuntes sobre la Myrica cerifera de

How to make wood unpalatable to the West Indian dry-wood termite,
Cryptotermes brevis Walker. I. With inorganic compounds

Importance of race in teak, Tectona grandis L., The

Tia

Iv, 6s

Iv,

Ly.

Introduction of a beneficial insect into Puerto Rico, The accidental III,

Islander looks at the mainland, An

Jamaica, Roofing shingles in

Jambosa _jambos (L.) Millsp., Pomarrosa, and its place in P, R.
Lady beetles don't behave

Landa Escober, Luis

Lynch, S. d.

Maga, A seed storage study of

Mahogany forests in Florida, Future may see

Mahogany plantations at Silk Grass Forest Reserve, The formation
and management of

Mahogany planting stock, Study of grades of broadleaved

Manufacture of shingles from local woods in Trinidad and
Tobago, The

Marrero, José Lik tg

7

Martinez Oramas. J.

Martinique, Classification des essences forestiéres d’aprés
leur utilisation

Martinique, Plan d’aménagement et d’exploitation rationelle
de la forét Martiniquaise

Martorell, L. F.
Mexico, Forestry and forest resources in

Meyer, H. Arthur

- 197 -

pa

103

Lid

163

145

IG?

89, 173; IV, 99

158

29

52

Montane vegetation in the Antilles TTI

Myrica cerifera de Honduras, Apuntes sobre la IV,
Nelson-Smith, J. H. JE
Palms of Dominica, A synopsis of the Lek
Pierce, John H. AbIOIL
Pine forests of Haiti, The IV;

Pinus occidentalis Swartz, near Jarabacoa, Dominican Republic,
Notes on some forest insects found in ACE

Pinus occidentalis Swartz, Retention of creosote oil in the

wood of IV,
Plan d'aménagement et d'exploitation rationelle de la forét

Martiniquaise sCAUIE
Plant new to the Western Hemisphere, A BuO
Planting with tar-paper pots on difficult sites in Puerto Rico PEG JL
Pomarrosa, Jambosa_jambos (L.) Millsp., and its place in P. R. IV,

Puerto Rico, Pomarrosa, Jambosa_jambos (L.) Millsp., and its
place in IV

Puerto Rico, The evaluation of forest tree species, as affected
by the local forest problem IV,

Puerto Rico, The accidental introduction of a beneficial insect into III,

Puerto Rico, Planting with tar-paper pots on difficult sites in AGAE AL.
Puerto Rico, Robie, a valuable forest tree in IV,
Quercus virginiana Mill., Encina IV,
Reid, David WW, cop
Reproductive cycles in plants AEDS 7
Research fun, Isn't Es
Retention of creosote oil in the wood of Pinus occidentalis Swartz IQ)
Roble, a valuable forest tree in Puerto Rico ive
Roofing shingles in Jamaica | TV),
Seed storage study of maga, A ibIG os

- 198 =

103

158

158
129

173

Seed storage study of some tropical hardwoods, A
Seifriz, William

Silk Grass Forest Reserve, The formation and management of
mahogany plantations at

Silviculture of Cedrela, Comments on the
Silvicultural experience with cedar in Trinidad, Summary of

Smeathers, R.

Tih

Iv,

99

11

75

77

91

107

164

39

25

Stehlé, H. TE a 29e oes, IV, 35), So, bier los
Stevenson, N. S. ia
Swabey, C. dei aiys
Tabonuco wood, Creosote penetration in, as affected by preliminary

boiling treatments in organic solvents IV,
Teak, Notes on pure plantation in Trinidad Pie
Teak, Tectona grandis L., The importance of race in IV,
Trinidad, Summary of silvicultural experience with Cedar in Tis
Trinidad and Tobago, The forest policy of III
Trinidad and Tobago, The manufacture of shingles from local

woods in IV,
Trinidad, Notes on pure teak plantation in rE
Tropical America, Forest types of DME
Wadsworth, Frank H. TV, 94, 59, 1407
Vegetation muscinale des Antilles Frangaises et son intérét

dans la valorisation sylvicole, La Vis
West Indian dry-wood termite, Cryptotermes brevis Walker, How to

make wood unpalatable to the. I. With inorganic compounds IV,
Whitney, Willis R. sia
Wolcott, George N. Pie DS; LV, ois
Wolte, H. S» Vie

- 199 -

124

uta e b
Orc eTy Mean Ga

Dike) oh