03 Vial Bolitoglossa
03 Vial Bolitoglossa
03 Vial Bolitoglossa
James L. Vial'
INTRODUCTION
BACKGROUND AND SCOPE OF WORK
lvfodern ecological concepts have becn established and refined mostly frcm
data obtained in north temperate zone studies. Despite the fact that about ane
third of the land mas'S of the earth and ncarly two-thirds of its biota oecu! in the
tropic zone, little js known about basic biological processes or ecological relation
.ships lUlder the influence oE the tropical sun. Tbe tropical environment is distin.
guishable from that of extratropical latitudes on the
temperature, heat budget, and
productivity;
basis of
combination
photoperiodicity,
of
factors
which
27'
The tropies are usually defined as the warm latitudes lying between
23'
N and S of the equator. Because of the sun's relation to the earth this latit
udinal region receives maximum solar radiation. If the earth's surface were oE uni
form composition and physiography, it might be expected that a
characteristic
..
13
REVISTA
14
DE BIOlOGIA
TROPICAL
in teuestrial and fresh water situations by the 24 e mean anoual isctherms) ('or
reded to sea leve!. Tropical marine habitats occu.r in area:s whecc tlle mean sur
tace temperature in the coldest month is greater than 210 C. The tropical regien
size and elevation of the included land areasl and effects of oceanie currents.
Within this regioo the daylight period is neve! less than la hours in
24,
nor mOre than 14. Farther north and outh the photopericd may range frcm zero
to 24 houes depending upon the SCaSon. Solar [adiation in the tropc belt aver
ages
heat budget basis the tropics receive the greatest prcportion oE solar ee;:gy and
the energy i5 evenly distributed over the year.
Lack of seasonality is .sometirnes regarded as a tropical character, bl..lt actnally
the seasons are obscured because diel extremes oE temperature and humidity show
greater range of variation than annual climatic cydes. Seasonal changes in the tea
pics are expressed principally by differences in ra,nfall distribution.
In a general way there is a correspondence {)f biota with the deliiting
thermal boundaries, although broad areas of transition may be present. High mon
tane and alpine regions dissect much of the tropies. \'J'ithin the 24' C (sea level)
isotherm these regioas are tropical in that they are exposed to the same light periad
kity and energy magnitudes as the lowlands, although their dimatc parameters
may at first sight corrcspond to those of extratropical localities.
Extratropical plethodontid salamanders have been One of the most widdy
studied groups o amphibians. The great diversity within the family and its broad
distribution ofEer excellent opportunities for studies of comparative ecology, as
have been dcmonstrated by DUNN
ORGAN (56).
(22),
GROBMAN
(32),
HAlRSTON
(33), and
found in the tropics, no intensive study oE a..'1y tropical representative has been
previously attempted. Probab!y few otIler species of salamanders are so suitcd for
sustained field investigations as is Eo/ilog/ouo slIhpalmofo (Boulenger). AI
though practicalIy nothjng was known of its life history, the species' remarkable
abundance on th! Cerro d::.: la :Muerte of Costa Rica indicated an almost ideal
situation for continuous study throllgheut its vertical range of distribution, acces
sible by the oter-American Highway (Fig.
1).
The objectives of the present study are fivefold: (1) to determine the
'
local distribution, habitat selection, and life history of B. stlbpolmafa; (2) to
evaluate the significance of populational char:-...."ters as expressed by the ecology
of Ihe species;
feasibility of
organisms; and
applying
ecologie
and
demographic
(4) to determine
methods
to
tropica!
ecology.
Preliminary field work was carrie out in June oE 1961. Fulltime study was
IN
COSTA RICA
15
were traveled in Costa Rica in direct relation to the study. Information was gatber
ed on ovee
4,000
spedmens of
B. subpalmala.
More than
3,000
weee preserved
and catalogued. Almost 1,000 otbee, were maeked and studied in the field througb
300
cedures.
METIlODS
To evaluate local
feom September
e,tablished at
facing slopes of the Ceero de la Mueete in Costa Rica. Because of the po;sibility
of theft oe tampering, alJ stations were placed on local fincas. As a eesult of this
p,ecaution, the st.tion at Finca Chvez was
intervalo The Iocations are ShOVlll in Fig.
could not be obtained for
2.
37
( 1 20
ft).
1961.
stations
At San
use
oE
TayIor
gauges were especially constructed and ,Iso calibeated by this agency. Readings of
peecipitation
",ere
25 ft.
to the ntarest meter. Because of rapid fluctuations in atOlo
at
<l
an
of the Cerro de la Mueete weee not available during the peeiod of field study; how
ever, quadrangles for the llorthern paet of the eegion have been published sub
sequently by the Instituto Geogcifico
Americao Geodetic Survey. Elevations and place names given by these maps
included in Fig.
2.
The maximum
50
aee
m.
taken.
15
cm, +
150
cm),
ed at the start and. completion of every study peeiod of more than one-half houe
at
the saIne
site.
REVISTA
16
DE BIOLOGIA TROPICAL
(7,000 ft.) on both Ihe NW aod S slopes of Ihe Cerro. This area was estimated
to offee optimal results after seveeal peelirninary teials. Colleeting proeedures within
quadrats were consistent; all surEace litter was cleared and removed, but no dig
ging was done. Habitat, size, and sex were noted for a11 specimens. Arcas were
Dot resampled except for immigration studies. Efforts wcrc made t collect from
similar environmental situations; i.e, topography, slope, and vegetation, in each
site.
At other times areas of various dimensions were sampled te supplement
observations en relative densities, activity cycles, and local distribution. Near1y
(2,286 m. on the NW
30 marked and rc!eased individuals
305
ro
(1)000 ft.)
vertical intervals. Three were located cn the !\TW slope and hvo on the soutb
een incline. Population IV, at
a total of
individuals eaptured and marked duriog the year. Populatioo 111, located at
46
2,926
(9,600 ft. ) , iocluded 302 marked anima!s; 407 salamanders weee mark
3,200 m ( 10,500 ft. ) and was the
highest study site established. 00 the southern slope, at ao elevation of 2,926 m,
ed in Population Ir near the summit al
Populations 1 and lA were set up ; the latter aflcr destmction oE much of the
habita! oi N' 1 by eoad building. Subsequently, N< I was studied ooly in respeet
to individuals previously marked. These two "mples included
respeetively.
The marking period of Population JI was from August,
1961, through
1962. Numbers III and IV were studied from Septemher, 1961, to Aug
ust, 1962. The site of Populatioo 1 was destroyed in November, 1961, and at that
time ao alternate site, N lA, was established aod studied until August, 1962. At
Ihe experimental site (Ix) marked specimeos were eeleased in August, 1961, and
July of
the area was examined at two month interval:>o After October o that yea!, no
animals were recovered.
At eaeh study site a stake geid was set up, umally at iotervals of about
15 m (50 ft.) providing fixed references foc azimuths and measurements. To tbe
extent that it eould be aeeomplished with rninimum distuebance of Ihe habitat, all
surfaee eover was eemoved and replaeed. Objects that could not be moved without
permanent di'Sturbance or destru.ction were Ieft in place. Every avaiIabIe specimeo
within the aeea was marked by toe dipping, measueed for snout-yeot length, ' and
sexed. Notes were taken on breeding condition, and other pertinent characters.
SUBPALMATA
IN COSTA RICA
17
Tho animals \Vere then returaed to the spot at which they "'ere taken. Although
marking of new indivduals was pcrfocmed ooly within a dcfined area, the sur
rounding 30 m (100 ft.) zone was exarnined in the same manner for marked
salamanders. Al! recaptured animals were studied as above. In these sites, 966
individuals were marked and studied during the year.
.
11" study aceas varied Erom 652 sq. m (7,020 sq. Et.) to 1,104 sq. m
( 1 1,875 sq. Et.) . The size oE eaeh area was determined pcimarily by the number
oE animals that eould be processed in a single day, thus avoiding any nocturnal
pop"lation shiEts.
Measurements oE distances \Vere made with a 30 m ( 100 Et.) Lufkin steel
tape. Azimuth readings to the neacest degcee were taken with a Uned States Army
Lensatic compass. Cace was practiced to avoid any magnctic attraction by ex
traneOlls objects.
Nocturnal activity of B. sflbpa11lJflla was studied on six diffcrent occasions
on the Cerro de la Muerte, between 2,560 m (8,400 Et.) and the summit. Prc
dusk to p osLdawn observations were made at five localities which induded the
macked pop ulations. In eaeh study the site was checked, using a battecy head
lamp, at two hour intervals throughout the night. A consistent path of examino
ation of the area was followed and eql.1al time allotted te each visito Following the
method oE HAIRSTON (33), a salamandec was considered 100% active iE it had
completely emerged from cover and wa"S found in the opeo, or 50% active if but
partially exposed. To avoid negative responses by active animals, none were touch
ed oc distu,rbed. On the night of Mal' 12, 1962, at 2,926 m (9,600 Et.) on the
southern slope, all active specimens \Vere collected at each vjst. Aftec sunrise the
area was thoroughly examined and all a"ailabIe salamanders wece collected.
At the beginning and cnd of each visit, relative humidit:y and tempcratures
o thc microhabitat, soil and air wece recorded.
Eaeh month, samples oE B. JIIbpalmala were candomly taken at altitudinal
intervals oE about 150 m ( 500 Et. ) ; quadcat sites and macked pop"lations wece
avided. AH sp ecimens were noted as t standard length. number and size of test
icular Jobes Oc number and sizes oE ova, as wc:ll as the concltion of fccondary sex
characters. From each sample paraffin 'Sections were made from individuals select
ed at 5 mm size intervals to determine the relationship of size and sxual develop
mento Gonads from 250 salamanders were Eixed in Carnoy solution, stained with
hematoxylin-eosin and sectioned at 8 to 10 micra.
Eggs and attending adults obsecved in the Eield wece examincd at intervals
and records taken on egg sizes and stages oE development. Thirtcen egg clutches
with the .ttending aduJts ",ere brought to the laboratocy and maintained at con
ditions approximating those of the nameal habitat in arder to permit study oE
brooding behavior and development. Information was obtained Erom a total oE 31
clutches .nd attending aduJts.
To observe courtship and breeding behavior, males and females werc kep t
together in cap tivty. The animals were rep laced at bi-weekly intervals to ensure
optimal physiologicaJ conclitions. Twenty-four experimental animals were sub-
REVISTA DE
18
BrOLQGIA TROP.TCAL
B. mbpalmata as well as
Canned soil samples were taken in the field during most pertods of sludy
01 the
marked populations, quadrats, and activity cycles ; classified as te texture and color;
are
105'
C for
48
hOills. Soil
tnoisturc
value,
the groove. dorsal side clown, and a rnillimeter rule placcd gently. but firmly. ayer
the ventenl surface. Reading were taen to the nearest
0.5 mm.
Repeated measurC
0. 1
mm before prescr\'
0.2
mm.
ACKNOWLEDGMENTS
aton Professor of Ecology. in fue years 1961 through 1964. During this memor
able period, my work was enthusiastieally supported by Des. Rafael Lueas Rodr
guez C. aod John DeAbate. Professor Elliott Coeo, Chief of the Servicio Meteoro
lgico de Costa Rica, made available his original data on tropical dimates arrd as
sisted
in
the coHection of wcather records, as did Sr. Vicente Contero. Lic Pedro
technical
al1 miccotechnical laoocs. Se. Jaime Villa R. aidcd with several of the drawings.
Ing. Federico Gutirrez B. , Director of the Instituto Geogrfico de Co_ta Hjea
provided mueh of the topograpbie information.
am
Robert M. Chew and John L. Mohr offered numeCOL1S suggestions in their critica!
reviews of the dissertation manuscripts, from which this paper has beco extractcd
and revised.
The help and eneouragement of my wife, Lynda W. Vial, eontributed gre:1t
Iy to all phases of the work. 1 also wish to express my gratitude to many individ
uals who, by their anonymous assistanee, in sorne way cantributed to tbe study.
Financial support was granted by the American Philosophical Society (Peo
rose Fund Grant N.
3029), the
4705)
The School
and
of Graduate
rN
COSTA
RICA
19
dioe in which the southwestern (Pacific) ,ide is more abrupt thao the north
day OCcur in limited areas. Clayey soils, usually yellawgray, are common bct
weeo 2,440 aod 2,745 m. At higher elevatioos a graybrown gravelly loam is typic.
al, a lthough in sheltered areas aod depressioos d eposits of black silty 10am oecur;
the result of localized decomposition of vegetatioo. WEBER (78) report, pH
values of 3.9 to 4.0 for mil samples from Jaboncillo Peak.
CLIMATE AND WEATHER
20
Recorels for precipitation during the first j'ear of the study period are pre
sented in Table L The lowest annual precipitation, 2,117 mm, was observed at El
EmpaLne (2,134 m ) . On two occasions the rain ",lUge was tampered with at
Finca San Cristbal (Hermanos Camacho) ; however, the ,nnual total of 2,429
mm recorded for this locality is reasonably accurate. At Finca Olvez (2,706 m)
3,142 mm fell, the mos! recorded. Although no region of the Cerro receives less
than 2,000 mm of rainfall per year, there is a distinct dry period from January
tbrough March. Following the wettest months, Mal' and June, a short dry period,
or Veranillo de San Juan, usually occurs in late July and eady August.
Temperature record s obtained from six weather sta.tions are given as mean
monthly mnima and maxima in TabIe 2. The lowest mean monthly mnimum
was O' C, obtained for January and February at Finca Chve. (2,706 m)_ The
highest mean monthl)' minimum of lO' C at the El EmpaLne station (2,134' m)
occurred from Mal' through November, and at La Cueva del Tigre (2,438 m) on
the southern slope in May and JuIy. Average mean minima and maxima for the
complete 12 montbs period ranged from 3.3 to 9.2' C aod from 18.1 to 24.
The ovecriding influence oE local conwtions is easily seeo.
l' e.
4.
At lO' N Lal. tbe period of daylight at the solstice extremes is from 11.
to 12.5 hours. The sun's zenith lies to the north for four months of the year.
Notwithstanding this slight seasonal variation in photoperiod, south_facing mootane
slopes cxhibit markedly higher maximuffi temperatures because of more extended
direct exposure te solar insolation.
Probably the mo,t effective use of temperaturc records obtaioed from the
weathcr stations can be made by application of tbe In!an anoual biotic temperaturc
of
HOLDRlDGE (39, 42) . His concept of biotemperature assumes that plants exhibit
ages are utiJized in the computation. Spccific values are derived from the formula
B = ::: nlean monthly temperatures > Oe.
"
as
Biotemperatures
are
treated
by
currents develop and by mid_morniog the sky is deosely overcast. Dudng the dry
periods, cloud developrneot mal' be delayed to mid-aftcrnoon and raio, if aoy, is
sligh!. For almost teo months of the year visibility becomes redueed to
few meter>
IN
RICA
COSTA
21
TABLE
Precipi/a/oll h TiJ011lhly 101ttJ al six wetlther s/aJiollJ 011 Ihe Cerro de la MI/erle. ReconiJ
were auumulaJed
al
2,134 ro
Month
El Empalme
2,438 ro
Mar.
Apr.
Mar
June
July
Aug.
Sept.
Oct.
No,'.
Doc.
3,048 m
2.706 m
Chve,
San Crjstb1
L1. Georgina
148
25
27
159
40
394
312
340
366
392
525
251
32
18
O
170
416
382
207
421
330
402
342
J6
3,142
2,SS
45
2
2
69
245
604?
259
203
233+
343
270
14
68
10
6
S
210
419
218
200
261
248
271
121
Jan.
Feb.
1962.
2,743
ro
L1 Siberia
2.438 ro
La Cueva
382
394
ISO
314
5 3!
341
333
1 16
2
7
3
167
03
482
196
283
325
266
417
147
2,609
2,798
20
5
3
157
Totals
mm.ppt./yr.
2, 117
TABLE
Temperstlure . records 01 mO,lIhJy
Cerro
la
the
vl1lion! ami loca/iou! are shown ;'1 Fig. 2. BiolemperaJures tll'e discuued i11 texto
LMWF.w::::: Low" MonJane Wet Pore!t MRF= Montane Rain PoreJt.
Elevation and Location
2,134 m
Montb
El Empalme
mino
max.
June
10
19
20
22
21
21
20
July
10
20
Aug.
10
lO
10
10
9
9.2
20
20
2Q
19
18
20.0
Jan.
Feb.
Mar.
Apr.
May
Sept.
Oct.
Nov.
Dcc
Ave.
Biotcmp.
7
7
8
9
lO
14.6
2,438 m
2,706 m
San Cristbal
mino
m3X. mino
6
6
7
7
8
9
16
19
21
18
19
17
5
7
9
S
7
7.2
Chvez
18
17
19
lB
17
IR
1fU
12.6
3,048
ro
u Gcorgina
2,743
m:tx. mino
max. mino
O
O
18
6
(,
I r.
19
20
22
21
20
20
19
21
19
19
18
19
20
4
1"
3
3.3
16
17
J S.5
2
3
3
(,
2
4
4
4
-1
3
3,4
18
21
19
18
18
18
19.5
10.9
1 1 .4
MRF
MRF
ro
La Siberia
flUX. mino
4
4
20
2;
22
21
IR
2,438
20
7
6
7
6
4
5.:
21
22
20
20
21
20.R
ro
1-1. Cueva
max.
26
25
2
24
22
23
23
24
8
S
9
10
9
10
9
S
9
9
9
S.8
29
23
22
23
24.1
13.3
1 6. 2
lllWF
LMWF
Vegetation
Formation
LMWF
lMWF
22
soon afte! nooo) foliowed by heavy but sporadic raios. In late afternoon sorne therm
a! stability is reached and the douds fragmento Visibility alternates from zerO to
severa! hundrcd meters as winds move the c10uds over rnountain ridges. Darkness
is followcd by spotty cloud cove! but during the night there is gradual clearing
of the sky, a reduction of humidity and decrel!e in temperature.
VEGETATlON
The area ineludes two maj or fllmatic vegetation formations: the Tropical
Lower Montanc Wet Forest and tbe Tropical Montane Rain Forest (HOLDRIDCE,
40,
41 ) .
3,100 m,
2,500
m and
clmax community. the pramo) resulting from the invasion of the burned over
surnmit regian by elements Df
It is
Dct) however, a
2.
(40 ) ,
and 'Species
STANDLEY (68 ) ,
heights of over
30
BI/ddJeia alpina, Hesperomeles obovatll, Escal1011ia p0aJatlft lV1l1nama pinnal, e{"a n/ata, Didymoprmax Pillieri and Podo
cm'pus montanlfS are characteeistic teees species. The large oak, Q/lerms eOstari
censis, forms dense, extensive staods dominating the physiogoorny of much of this
formation. Above 3,000 m Perl1ett)'d coriacea, Vaccinium cOl1sangrtil1Cllil1, Hper
icum strietllm and H. Jilenoide are sorne of the more common shruhs which, with
the conspicuous dwarf bamboo, Chwqllea sllbtessellata, represent a mixture of
In the Montane Rain Forest,
VIAL:
IN
COSTA
RrCA
23
in
Costa Rica and from the region .oE Carop Boquete jn Panam.. In the lowcr limits
of its altitudinal range, approximately 1,500 m, this salamanclcr is encountcred
sporadically and is common only above
2,1>5
tropical !ower montane and montane regions in a geographic range bisccted into
northern and southern parts by tbe Meseta Central of Costa Rica.
On the Cerro de la Muerte
aboye 2,300
ro
with elevation. On the southern slope it is a11 but absent below 2,600 m.
By compariog information from precipitatian and temperaturcs, an obvious
correlation of popuIation densHies with dimatic conclitions is seen. It has been
generally accepted that moisture, as expressed in the amount and distribution of
rainfall exerts the greatest influence on the distribution of Qrganisrns in tropical
enviroments
(LA RUE
49; RICHARDS,
01 second
ary importan ce. Thi5 is only tClle of regioos with fairly uniform surface featl1ces.
Many areas of low elevations, (i.e., below
ture and refuge requirements of
istic of certain coastal regions of the country, yet lowland tropical wet forest areas
are not occupied by this species. Apparently temperatme is the dominant phr-iml
nfluence in restricting the distribution of these salamanders. No limiting bio
logical factor can yet be suggested. The species has not been observed to occupy
habitat, on tbe Cerro de la Muerte at biotemperatures higher than 1 5 .2' C. 1 do
not imply that microclimate of habitats is measurable by records from weather
stations, but cerwnIy microdimates are influenced. in a general way by gross di
matic regimens
(GEIGER, 27).
HAlllTAT
SELECTION
common retreat. The Cerro is subject to surface drying dlUing winter, but sal
amander populations necd migrate only a few centimetrs below the ground
slJ;,[face in crevices and root-fractured crannies te avoid desCcation.
2,700
m, the seJected
2,400. m
and
retains a high. moisture level and a stable, relatively low, temperaturc is a habitat
most frequently utilized. OccasionaUy saJamanders may be
encountered under
rocks and plant debris, but only when these are deeply imbedded in the soil or
possess a weU-developed lichen-moss margino Where local strearo bank topography
oc persistent raios rnodify local moisture condtions, habitats that would otber
wise be unfavorable can be occupied temporarily.
24
Below 2,400 m B.
mbpalmata
Crevices in moist, dayey soil on well-drained slopes were the most successful col.
lecting sites. Such places affee an excellent balance of excess water run-off and
high retention
abIe moisture and temperature conditions are present throughout the year.
HABITAT UTlLlZATION
mbpalmata
of the "aerial swamp" (DUNN, 24; PICADO, 5 7 ) . During its nocturnal wander
ings, the species has beeo observed emerging from bark crevices in teees and 10gs
and dimbing on the branches of shrubs and mosscovered trees to heights of two
meters above ground leve!. Most frequently, howevec, it js seen moving slowly
over the ground or rocks in the near vicinity of em'cr.
In daylight hours the animals retreat to dark depl'essions ooder rocks and
debris; lie concealed under dense decumbent mossy mats or deep within decay.
ing Iogs; or worm their way through dcep fractures and creviees in the exposed
rock mantle. In view of its ubiql.1itol.1s use of the habitat area, B.
slIbpalmala must
LIFE HISTORY
REPRODUCTIVE CYCLE OF MALES
The testes of B.
mbpalmata
B.
subpdlmata
and ORGAN
DE
(56) have re
Aside froID gross dissection of the aboye sample, inforrnation was obtained
from microscope slides of testes from 165 males. The pattera of maturation oE
sperm ceUs and lobe formation has been treated in detail by BURGER ( 1 2 ) ,
DesmogllathllJ
and
temperate plethodontid
the same basic pattern,
DesmogllaJhlts
and
Plehodon,
ceU cord gives rise to the first testicular lobe which indudes numerOus sm.U
lobules. In the earliest stages of formation the lobules appear to contain spermat
ogonia n. Since no weUformed lobules (i.e., with a complete margin oE septal
tissue) contain cells in earHee mitotic phases, the formation oE the septa and late
spermatogonia seems to take place at about the same lime. By the time the septal
boundary is completed and the lobule weU defined, some germ cells are approach.
VIAL:
JN COSTA RICA
25
26
aod pass into the oviduct during oviposition. Sorne indication that immature ova
are retained and may subsequently mature is offered by five brooding females
that contained numerous small to medium (0.6 to 2.0 mm dia!l1(ter) 'whitish ova
in the ovaries.
The oviducts of mature individuals are enlarged, thickcned, highly con
volu.ted tubes which occupy much of the abdominal cavity. In immature specimens
they are thin aod 'straight dorsolateral tubules.
Fernales
enlarged ova were faund at aH times oE the year, but most oE the marked adults
did TIot become gravid duri11g the periad of study. Two marked females remalo
ed gravid over a periad of eight months without any intervening oviposition. Un
fortunately, no information is available on marked gravid anima1s that subsequent1y were Doted as "spent". The female element of the population is capable
oE
stlbpalmata occurred
in an
adlllt female which had beeo subjectcd to a subcutaoeous gular implant of a single
anterior hypophyseal lobe of fresh
011
C.
SlIbpalm"/(/
VIAL:
i7
in mounds or slopes of granitic talus. These too, typically, have a dense cover of
low vegetation and always possess ao interstitial matrix oE clayey loam or grave!.
On a single occasion an egg dutch was observed
Within
a decomposcd talus slope three clutches with the attending adults were found
unde!
the
same rock fragment, each clutch nearly in contad with thc others. The
stages of development were welI advaneed and essentialIy equal, indieating that
they had alI been deposited at nearly the same time.
The egg mass is usualIy I.id along the sides or m.rgins of the site. The
combination DE clutch positiori and soil texture provides adequate drainage of
surplus water and yet offers a high degree of moisture retenton. Under p[otective
cover and at observed depths of from 3 to 22
cm,
existo Ambient soilair interface temperatures recorded at 1 5 nest sites range from
9.8 to
16.0'C.
(mean 12.8' ) .
EGGS AND EGG CLUTCHES
STEBBINS
(70) for
to those descr.ibed
by
EnlaJ;na eschlcholtzi.
Twa gelatinous membranes surround the opaque yolk, which almost com
pletely occupies the cavjty formd by the inner membrane. The inner membrane
is not visible in newly deposited cggs; however, aEter immersion in water fo!
ap
(b) surrounds the ovum (d) and is only slightly separated from it (e) . Of Ihe
two gelatinous layers the inner eonforms to the spberieal sbape of the yolk mass
and is essentially uniEorm in thickness. Because oE variable thickoess in the outer
membrane
Ca )
the gg assumes
a.xis.
mm
(average 5.0
0.12
4.7
to 5 . 3
masses
month
durjng th"?
study period exeept for April .nd May. Eaeh c1utch was attended by a11 .cluit.
Numerous attempts were rnade to observe a complete sequence of development
in the field without 'uceess. Although disturbance of the nest .nd eggs was kept
REVISTA
28
DE
BIOLOGIA TROPICAL
to a minimum and observatiollS made as infrequently as but once every two weeks,
the brooding adult 'invariably abandoned the cluteh,
The attending parent usuaHy was coiled tightly about the egg clutch with
the forelimbs and throat region resting on the upper surface of the mass, This
embrace was usually sustained for three oc four days without notable change. Dur
ing the period of brooding the adult seldom left the eggs and was almos! always
in contad with them. At intervals of severai days the adults were observed to rotate
the cluteh with the forelimbs and tail, but changing its loeation littIe or not at aH.
Evidence tcm both laboratory and field studies indicates tbat persistent,
but not constant, attendance DE the eggs by an adult is a requisite foc their cont
EMBRYOGENESIS
No deavage is visible in ncwly deposited eggs. The earliest stage recogniz
:lble without sacrifice of the egg is that of the aeural groove. Optic vesicles and leas
appear soon after convergence of the neural folds and undergo rapid enlarge
ment, becoming the most coaspicuous fciture of the post_aeural embryo. Anterior
limb buds, mouth parts and gill filaments, arising as Ol1tgrowths from the guIar
fold region, appear at approximately the same time; the posterior limb buds
being apparent somewhat latero By the time these hind buds are present the em
bryo is very elongate, extending over about three-fourths of the yolk circumfer
enee. The giHs grow more rapidly at this stage, elongating and trifurcating from
a single basal stalk:. Pigmentation becomes more obvious as melaaophores develop
.long the mid-er.nial region and to a lesser degree along the lateral body margins
to the tail tipo Melanophores are at first absent from the vertebral region and limb
buds. Occipital and internasal fontanelles are still conspicuous at this stage. Draw
ings of sorne majar fearnres of embryogenesis are presented in Figs.
19.
12
through
(In eaeh of these figures the horizontal line represents a seaJe of one mm.)
As the limb buds elongate articulations ad digits become distinguish
able. The gills beeome somewhat redueed in overaH size althol1gh Ihe distal brandl
as the costal folds develop. These appear as separate ridges overlying the yolk filass
(Fig.
18).
Pig
mentation rapidly expands over the head and body region, costal ridges and hind
limbs; it is almost absent from the forelimbs. N.solabi,1 grooves are 0150 distinct.
29
At least two weeks later, the head takes on a definitive sh.pe, the guIar
fold becomes conspicuous and the feet weIl formed. Although variable, a reduc
tion io gill filaments may oca" as tbe basal portions are gradually resorbed witb
a concomitant fusion of the costal ridge,. At this time a ,hort yoIk stalk extends
from tbe post-pectoral region. Just before hatching tbe yoIk mass is exteenally
visible as a slight ventral extrusion. Pigmentation at this time is weU developed with
recognizable melanophores, iridophores and erythrophores. The young salamander
oeeupies the entire central cavity of tbe egg and lies tightly reflexed in a bead-to
t.il position (Fig. 19) .
1 have not been able to foUow tbe entire sequenee of development in
relation te time. On the basis of comparative obervations of the periods involved
between variolls and oveelapping stages in different du.tches, it seems probable
that four to five months are required for intraoval development. This estimate
is similar to that given for Ensati,la esChIcholtzi by STEBBINS (70). Moreover, the
meroblastie eleav.ge patteen of the large, neaely isolecithal yolk and tbe prolonged
intraoval period in absence cE a free larval stage are consistent with his observa
tions of that speeies.
HATCHING
30
GROWTH
87
290
reeaptmes of
136
marked
1963,
26
mouths
eacHer. Measurements at time cE capture and last recaptuce for these animals are
54
mm were
six to ten
intervals of from ane to five months) . These examples occllrrcd throughollt the
year and contribute to an
in
which pcriods
oE
negligible size inerease, iE any, alternate with those oE more rapid growth Al
though zero values in a way vouch for the obst'rved slownes oE growth in B. sub
pttlmata they are cf 1ittle use in the determintion oE annual rates, since a11 ani
mals oE mOre than ten months of record demonstmted a size inerease. Thus, estim
ates oE annual growth rates mut exclude those
Information is available for
75
87
port a nu11 hypothesis that no differences exist among the three groups at the
99%
eonfidenee leve!.
VIAL:
ECOLOGY
Fig. 2 2 gives
OF
BOLITOGLOSSA
31
omitting specimens for which there was no measurable gro'\vth. Althollgh a margin
As i.s typical
CHARACTERS
32
base oi the tail. NOBLE (53) has attributed cOl1rtship stiml1lation as the role of
these "hedooic" gland, in male plethodontids. Their function in courtship has
been inferred by ORGAN (54, SS) for two species oi Plelbodol1, and by Snssms
(70) for Emalina eJChfcholtzi.
In my series of B. f1Ibpal",ala, females attain a maximum standard length
of 69.0 mm, 4.0 mm gre.ter than that of the largest maleo The snout is more
rou.."'1.ded than in males and nasolabial grooves are jnconspicuous. Membranous
folds lino the margin of the cloaca. Near-tcrminal gr.vid females comprise the
largest size class of alI 'pecimens observed.
Swollen cloacal lips are characteristic oi both sexes when in breeding
condition. This manifestation, however, is indicative only of tbe actual breed
iog period of the individual aod its absence cannot be used as a general criterion
for reproductive jmmaturity. It does serve as a useful chlractcr in determining
the dur.tion of the breeding cyele in the population. Throughollt the period of
field studYI males and females were observed in this conclitioo.
POPULATION DYNAMICS
AcrIVI1Y
After prelimin.ry investigation to determine the extent of nocturnality in
B. mbpalmala., studies on activity cydes were initiated one hour before dark and
continued until after sunris:e. At least two sites at different elevations were visit
ed every two hours during the night. A total of 13 site studies were made. T"
investigate the possible influence of seasonal changes upon activity, thcse studies
were conducted in October, December, March, May and July of 1961 and 1962.
B. slIbpalmala is completely nocturnal. Diurnal .ctivity has becn induced or
stimulated to a degree by rainfaU in some ,pecies of plethodontids sueh as Desmog11t1thus oc/)o,-op/)aeIlI ct1ro!inellsis (HAlRSTONJ 43), but precipitation has not beco
observed to iofluenee activity periods of salamanders on the Cerro befare the
onset of darlroess. Deep within the forest, or in some densely vegetated quebra
das, it is possible that lighting is so reduced eveo during daylight hours tn.t
some diurnal activity may be stimulated by increased atmospheric moisture. On
one oceasion, March 17, 1962, an exposed juvenile was collected on a talus slope
near the summit between 12 :30 and 1 :00 PM. There h.d been no rainfaIl during
this day. At the time of capture the ground surf.ce temperature registered 12.6"
C; doud cover limited visibility to 20 meter, and the relative humidity was 100%.
It ,hould be noted, however, that sueh conditions are typical on the Cerro, espe
ciaUy during late morning or eady afternooo.
Activity periods may extend throughout !he night, beginning just after
dusk and terminating with the first discernible light of dawn. Varioble emo,
gence times were evident withn the same samplc arcas 00 different nights and
at different sample areas on the same night. Time of initial emergence ranged
from 6 : 1 5 PM to 1 1 :30 PM (GMT-6) . At this latitude ( lO" N) there is only
IN COSTA
RiCA
3"
REVISTA
DE BIOLOGIA TROPICAL
35
with the total observed salamanders i n a marked population within a ,,'eek o f th.
night study gives activity levels from 19.8 to 24.6%. If observed activity level,
.re compared with predicted numbers of individual, in the stu.dy areas computed
froIn capture-recapture data foc months during which activity tudies were made,
the proportion of aetivity is reduced to a range of 0.002 to 0.050% of the total
population.
At the maximum less th.n a quarter of the population is active in any
one periodo No correlation in the intensity oE activity with scasonal dimatic re
gimens has be,n demonstrated, although local weather eonditions do influenee
B.
mbpalmata
is a high carrying capacity sustained by the available habita!:>. Diurnal retreats are
practically always available to the animals within a distance that can be traversed
during a nocturnal activity periodo Spatial utilization of the total habitat is three
dimensional. Potential retreats are not only immedjately bencath surfacc cover,
but also indude subterranean levels to dopths of probably a meter or more in talus
deposits. During period, of activity salamanders primarily utilize the grotlnd
surface, but also have been ob,erved dimbing on rocks and logs or in ,hmbbery.
rocky soil
to a depth of 30
ano
On anothcr occasion
fou:rteen adults were taken from within a rotting lag approximately 2 metcrs in
length and 30 cm in diameter.
distance, statisticaJ analysis of the relationship of these two factors can be made.
Measured distances cE separate moves, rather than total distance covered oc net
distance from a given site, were plotted io one meter units against the actual
TABLE 3
Movements 01 B. subpalmata expressed in momhly inlervals. The nl/ll hypolheIis dJsumes
Ihal disJances moved are rando11l with respecl lo lime.
al
he
Monthly
Inten'aJ
po
Freq.
2.87
2 . 13
3.05
3.90
2.21
4.97
122
39
14
16
20
17
3
4
5
6
(*)
indica/es rejection
Chl
Poisson
NWl!bus
E
in mcrs
Squarc
269.64
99.69
31.82
103.2' "
31.29
61.88
121.'7
38.73
13.98
1 5 .80
19.99
17.G1
'rABLE 4
Comp(11'alive averages 01 dislances moved by males, lema/es and juveniles. The nI/JI hypo.
thesir JaSSl/mes 110 significanl difference hetwem any 01 Ihe three grol/ps. Chi.sqllare
f,:zlues are accepred al Ihe
Males
Fem:l.les
5.4;
'.43
3.71
3.71
Juveniles
2.8l
2.8'
Chi-squ:l.re
0.1637
0.4 1 1 9
0.0566
TABLB 5
Comparalive averages 01 1111mher of 1CParau r;:OfICS of males, lemales Imd jllvenilcs. The "NU
h)'polhesis aI!IImes 110 ignificanl differencer he/lucen any 01 the three grollps. Chi-qu:Jre
tia/l/u are accepled at Ihe 95% confidence leve/ al one degrce 01 freedom.
Males
Females
1.19
1.19
U3
Juveniles
1.13
1.13
hi.square
0.OZ09
0.0007
0.0299
VIAL:
37
period oE time in months between two captures. OE t."c total moves observed,
228 whicb cover periods from ooe to six months, were comparcd separately with
a Poisson distribution te obtaio ao estimation of the nature of randomness in
movement. Table 3 includes pertinent valucs derived from thcse calculations.
The value of .po js the average distance in meters lnoved foc all animals
ducing that time intervalo Chi-square computations are derived by use oE squared
deviations from the Poisson number foc each distance intervalo In al! time inter
vals the distances moved are rejeeted as being random on the basis oE Chi-square
values at N-2 degrees oE Ereedom Eor the samples, eaeh greatly exeeeding the
95 % eonEidenee level. If the distance of movcments inereased randomly with
time, it would indicate that individuals do not oecup}' any deEined area. The
distances are not randollJ., howevec, jrnplying that a home range is utilized.
30 mm standard
Field cecords on movcments of unsexed ju'veniles
Jength) ineluded 15 animals. OE a total oE 17 movcs, tbe extremes were 0.10
and 13.2 m. In adult males the Jimits oE 75 observed moves of 63 specimeos
ranged from 0.10 to 50.39 m and for in moves oE 116 Ecmales between 0.05
and 35.13 ffi. To evaluate this information distances have been assigned jnter
vals of ene meter at frequencies of occucrence for juveniles, males and fcmales
(Fig. 26). This proeedure disregards any time interval.
Among juveniles the average distance oE aU moves is 2.85 0.84 m,
with an average of 1.13 moves per individual. Adult males average a distance
of 5.43 0.97 m with 1.19 moves per speeimeo. The magnitude of eaeh move
by females averages somewhat Jess, 3.71 + 0.42 m, than for males but the aver
age number of moves per individual is slightly greater, 1.53.
Averages of the distanee of movements (the null hypothesis being that
tbe difference in magnitudes are not significant between juveniles, males and
females) are treated in Table 4. In eaeh oE these eomparisons the null hypo
thesis is substantiated at the 95 % level Eor One degree oE freedom.
On the assumption that no signifieance exists in differenees of the num_
ber of moves per individual jn juveniles, females and males, Chi.square evalua
tion gain substantiates the nuU hypothesis at the 95. % probability level with one
degree oE freedom (sce TabJe 5 ) .
The relationsrup of distances moved to sizes of tbe saJamanders can now
be examined. In Fig. 27 the average of observed distanees oE moves for eacb
two mm size elass is plotted for juveniles, males and fcmales. Within ten mm
sizo intervals, males of 30 to 39 mm standard length are represented by 12 re
eords; those between 40 .nd 49 mm by 57; from 50 to 59 mm by five and
those ahove 60 mm by one. Females of 30 to 39 mm ineluded 42 observations;
those from 40 to 49 mm ineluded 60; from 50 to 59 mm, 62; and above 60 mm,
13. Intervals Eor which there were few records demonstrated tbe higher aver
ages. The pattero of these results doc> not suggest any eorrelation between size of
the animal and average magnitude of movement.
The extent of mobility in size clases of B. Ilbpalmala is presented in Fig.
28. PJotted at two mm intervals oE standard length, again for all tlu:ce groups, are
the .verae number oE recorded moves. Here also, the vaJues generaUy reflect
REVISTA DE BJOLOGIA
38
TROPiCAL
and
FITCH (25 ) .
BURT (13), for example, has stated that "occasional sallies out
side !he area, perhaps exploratory in nature, ,hould not be considered as in part
of the home range". In his work on Rana p"tiosa in Yellowstone Park,
TURNER
(74), treated cydic migrations and rapid population shifts as part of !he "act
ivity range" but beyond !he normal limits of what he referred to as a "fixed
horne rangel>. Such seasonal wanderings are well known in aH terrestrial vert
ebrate groups. It rnust be recognized that distinctions between rapid shifts Oc
temporarj' excursions are characterized more in magnitude of time rather than
distance (omitting, of course, long.range migrations) .
CUNNINGHAM (17)
has
pointed out that the essential diHerence between temporary extended sorties and
the possession of two home ranges is !hat of time.
As here conceived, the home range of B. sllbpalmata indudes that
a,ea
within which the individual persistently carries out its normal role in the com
munity. As no breeding migrations oc aggregations occur in this species, the only
activity which js not construed to be within a horne range is movement that may
be associated with dispersion of the population. In the absence of any known
territoriality, and considering that there is seldom a limitation on habitat space
oc a rustance between sites of protective cover in excess Di the average distanee
of rnoves, the horne range is viewed as being essential1y ci [(ular.
1 have adopted in part the methods for determination of home range
(1)
an individual, and
(2)
The average home range area is computed as the area of a circle (". r')
290 recaptures for which measllr.
abl. position changes ( > 0.05 m) were available. Radii extremes (0.05 lO
using straight line distances (radii) recorded in
VIA[.,
IN
COSTA RICA
39
50.39 m) are therefore equal to tbose obtained for separate moves; area extremes
0.008 to 7,977.00 sq. m. lbe average horne range radius is 3.74
1.19 m and tbe average area is 43.6 4.44 sq. m. As no statistically significant
Animals recapturcd more than once at the same site cfEer no real measure
of the extent of home range, although this kind of information is valuable in
indicating the degree of utilization of a horne site.
STEBBINS
(69) observed 39 of 527 marked Ensal;na that remained under the same cover
from one to five months during four years of study. In his work on Batrachoreps
altenltall/S, HENDRICKSON (35) introduced relatively large cover boards whieh
permitted considerable position changes beneath them. It is possible tbat his value,
59% of 133 marked animals which did not change cover in 24 months, may have
beco influenced by tbis factor and would thereore be conservative. Of 375 re
captures in Ealrachoseps pacificlH, CUNNlNGHAM (17) reports that 92% moved
to other cover.
Evidence of tbe sedentary nature of Ba/itoglossa sllbpalmala is based upon
406 (5.4%) recaptures in which tbe animals had not ehanged position
more than 0.05 m (without regarding tbe size of cover) in periods up to ten
22
of
distance moved
by an
individual (for whieh there were two or more recaptures) was sligbtly less tban
21
m.
HOMINCJ> BEHAVIOR
lbe
were placed
salamanders
at
different
directions
.nd
distances
(twQ
REVISTA DE
40
BrOLOGIA TROPICAL
to twelve m) rcm the points of original capture, and these sites were checked
at six weekly intervals following displacement. In only oae instance was a dis
placed female found in its original site, frcm which it had been moved three
m twe Wteks before. This female was again rmoved and placed seven m away
in the same direction but neve! was observed to return. Although no other dis
placed BoJitogloss. returned, several had been placed nearer to the capture point.
These few records can no more thn suggest that hcming behavior is
poorly developed in B.
s1lbpalmata.
attendance, as well as the ease with which adults abandon them, offecs support
ng inferential evidence.
AGGREGATlONS
Below evels of 2,500 m, population densities are low and seldom are
three ar more salamanders found in clase proximity within the same refugiW1l.
In mid-ranges and upward to the sununit area, as densities inerease with mOre
favorable envuonmental conclitions, the occurrence of aggrgations becomes more
frequent. Above 3,000 m the finding of .ix to fourteen animals under a rock
not more than 1 5 cm in diameter is COmIDon.
No indication of sexual behavior has becn observed in the composition
of aggregations. The groups were composed of all males, all females or varying
numbers of both sexes with equal frequency. By following the movements of
marked individuals, the composition of aggregations was
observed to
change
markedly. On only three of 290 (1.03%) occasions were the same two animals
together more than once in the same refugium.
During the dry seaSOn on the Cerro, the ground surface becomes caked
and salamanders are found ooly under deep seated debris and rocks where mois_
ture has been retained. Aggregations are intensified during this time. For example,
between September, 1961 and February, 1962, only 17 spedmens had been found
at the site of Population IV. In March, however, 15 new individuals were marked,
sities of salamander populations is notable only for the limited number of studi",
that deal with this zspect of population ecology and the variety of techniqu<5
cmployed.
RICA
41
42
TABLE 6
Relalive densiJies 01 popuJaliol!J of B. subpalmata
(/s
the mwzbcl 01 illdillidllrZ1s per Jq. m. [Alz a.Ilerk (*) repreents smnpies in ulhiob no
anima/ fuere i'Jreunt itl he
Elevations
S!oPt!
in Meters
Ma:ch
NW
NW
2438
2591
2743
2896
3048
3200
3048
2896
2743
2591
.0(45
<.0(153*
.1291
.01A5
.0430
.1"05
<.0053'
.OE45
.0484
< .0053"
<.0537
NW
NW
NW
Summit
S
S
S
S
Monthly
Averages
June
Sept.
.1129
.0753
.0645
.150S
.2582
.1559
.;333
.1291
.0484
.1022
.0053
.0 1 6 1
.0376
.0713
.2044
.1882
.1 183
.0699
.0806
.1075
<.0053'
.0161
.0645
.0753
.1237
.1452
.2797
.0053
.0430
.02 1 5
< .0470
<.0282
.0739
.0914
.1573
.1492
<.1841
.0672
.0551
<.0591
.1432
.0903
< .0779
< .0930
Dec.
Ave.
TABLE 7
Comparaf,'e demities 01 lome p/ethodontid salamanders,
Stated
Species
Crud. Deruity
Crude DeDsity
lnd./Sq. Meter
. Souece
Test and
Plethodon
.0499
Bingman (72)
84 per
2,000 sq. ft.
.4516
Hendrickson ( 3 5 )
cinereus
Balra.hoseps
attenuatUJ
BatrachoupJ
allenualJU
Aneides
lugubriJ
Ensatina
euhJ&holtzi
Bo/itogloJJa
Jllbpalmata
3,870 - 8,160
.95582.01 5 5
Anderson ( 2 )
A051.4989
Anderson ( 2 )
.1482.1729
Stebbins (70)
per acre
1,640 - 2,020
per acre
600 700
per acre
306 - 3,683
pe! acre
.0756-.9097
VIA[.,
43
were observed in all areas where the species was found, although usuaUy of lesser
magnitudes. Similar responses to annual distribution of rainfall have been report
ed for the plethodontid genera Batrachoseps, Aneides and Ensatina
(ANDERSON,
fuU year and aIso, these sites were at essentially the same elevation as
ing upon tite extent to which their horne range areas overlap the boundaries.
Individuals that have a major portion o their home range interseeted by this
margin wiU be eneountered more frequent1y than those with but a smaU part
within the defined area. In cases where peripheral animals may be present it
should be reeognized that the study site is really inclusive of a larger area than
the measured limits, and without sorne adjustment the computed density values
will be unrealistieally high. To compensate for variations in the extent to whieh
home range overlap oecurs, a margin about the defined study zone equal to the
home range radius should be added and the total area eonsidered in relation
to densities. This procedure has been used by
ing area sizes in sq. m were obtained (eorreeted values in parentheses) : Popu_
lation lI,
a (n + 1)
(r + 1), in
mark
(3,200 m elevation),
0.4795 -+- 0.0556 individuals per sq. m.
Population lII, at 2,926 m, averaged 0.3269 -+- 0.0502; and Population IV, at
2,560 m, 0.0389 -+- 0.0068.
ative density study. Density is greatest in Population II
ro
'\Vere from
44
REVISTA
DE BIOLOGIA TROPICAl.
Population IU and IV deosities raoged from 0.1 170 to 0.6672 and 0.0205 to
0.0516, resp:ctively. Tbe absence of correlaticn in time of Qccurrence or elevaticn
is product of local climatic influence.
In view of the nature of the defincd stu:dy arcas, avaHablc data 011 densites
of extratropical plethadontids appe.r to be consisteot over-estimates. For p,upoees
of comparisoo, crude den,ities of B. subpalmala cao be computed 00 fue hasis
of the actual study area, rather fuao ao adjusted value ohtaioed by .dditioo of
the borne raoge radius_ Uoco",:cted deosities of from 0.0757 to 0.9100 salaman
ders per sq. m represeot the extremes of ohserved annual averages ohtained during
my studies_ Table 7 summarizes fue comparative data avdahle for four other
species aod equates fue numhers of iodividuals io sq. m w'ts. ANDERSON (2)
studied populatioos of Aneides /lIgubris and Batrachoseps allenuatus 00 the
islands of SlIl Francisco Bay" California. The average density of four estimates
he obtaioed for Aneides was 1,892 per .cre. His remarkable estimate of 66,530
Batrachoseps per acre was admittedly ' excessive, which he attrihuted to a local
population shift; however, he considered an estimate of 8,160 salarnanders per
acre to be reliable.
If the observed activity levels of B. subpalma/a are in any way indicative
of low energy demands hy these montane ectotberms, then fue factors limiting
fue numher of individuaIs would appear to he density-independent; most prob
ably, spatial requiremeots and available microhabitats preseot wifuin the distribu_
tion range.
AGE CLASSES
IN COSTA
IUCA
45
between Ihree ana four years; two Iobes represenling five lo six years; and Ihree
lobes
indicative of seven lO eighl years. In Ihe five species ineluded' in his study
all males matured by Ihe fourlh year afler Ihey had been deposiled as eggs.
Fig. 32 ineludes a sample of 140 B. JlIbpalmala males laken during
June of 1962. The two mm inlerval was chosen wilh Ihe purpose of avoiding
any bias in favor of Ihe proposed growlh curve (Fig. 2 3 ) . Singlelobed males
in this srunple demonslraled a modal group al Ihe 41
mm
midpoint. In double
lobed males the mode occurred al 47 mm, wilh a second peak at 5 5 mm, wbile
Ihe mode of Ihe Ihreelobed specimens was at 55 mm. These modes may be co
incider.taI actifacts, buJ: they obviously conform lo the estimated 2.9 mm per year
growlh incremenls derived from marked and recaptured animals and correspond
almost exactly to intervals at the ninth, eleventh and thirteenth years, without
then the averages represenled by Ihe growth curve are indirectly sub,tantiated-.
SEXUAL MATURITY
of. maluro sperm in Ihe testes and Wolffian ducts, aod in fomales by Ihe ap
( > 2.5
mm dia.) . In
both sexes maturity is attained over a fairly wide range of age and size.
ual attaining malurity. Above this size the proportion of mature males generally
increases to the 42 mm size clas, aboye which all specimens examined were pro
dU<:ng sperm. A size difference of 16 mm exists Ix:twcen Ihe onset and completion
taneous with their first occuuence in the testes although in 9.7% oE tbe tissue
male malurity, bul POPE and POPE (60) considered Ihese slructures to be sut>
jeet
me
46
44
maturity level of the population is shown as the median bem'een the largest size
dass at which no individuals were" mature, and that at and aboye whidl a11 spec:i
mens examioed were reproductively capable. The 50% Jevels are represented by
the intersection oE the slanted Ene connecting the extremes of these size c1asses witb
the 50
% level be
tween initiation of maturity and the 100 % level is at the midpoint of the 46.0
to 47.9 mm interval, or 47 mm, but for the range of size lit which aviposition
is known to occur the midpoint is 49 mm.
By osing the previoos estimates foc growth and age elass.s, acquisition
of sexual maturity in B. S1Ibpalmala populations can be translated into
time.
Within this sebeme the earliest maturation of males would occur at four years
and gradually inaease through the ninth year, by whicb time all males were
capable of reproducing. Between these extremes, a time raoge oE five years, the.
populatioo median (50 % level) for maturity in males is predicted at the sixth
year.
Sexual m.turity in females is consideiably delayed as 'compared to. males
The earliest oviposition was a{ter the ninth year, an age at which aH males were
marure, and increased until about the fourteenth year when al! appeared to have
laid eggs at least once Oc were gravid. Maiden females wcre present in the thir
tecn year age category. Tbe time fange of extremes represented hece, four years,
is slightly less than tbat of males and the 50% level of matur.tion is pcedicted
at the twelfth year. The average time factor for the 6 mm differential observed
between the occurrence of females with mature oviducts and the smallest size
'
47
substantiat Ihis
estimate and suggest Ihat a two year period is probably the minimum required
.
between successive ovipositions of egg clutehes.
The variatian in ages at maturity in both sexes af B. stlbpalmal seems
reasanable in view af the estimated '18 year potential life span. Retarded maturity
in females may be a reflection af the time required far tbe relalively high energy
demands imposed by elabaratian of oviducal cells ond by tbat required for starage
of nutrienls in the ova during formalian. It is likely, too, Ihat Ihe introduction
af mature females occurring at a time when aU males of
mature wauld enhance the probability of sexual
encounters
and
reproductive
success.
SEX RATIOS
4S
represe1!led by he 'l1(rrTter/y quadrats. The null hypother i a one-to-oue sex ratio'. Tbe data
wcre IZccepted al Ihe 95 % ollfidence leve! al one def!.ree 01 frcedom.
NI' Males
Month
March
N9 FemaJes
Total
Males
Chi
Total
Square
39
58
97
.4020
1.8788
140
124
264
.530 3
0.4853
September
87
92
179
.4860
0.0698
December
70
71
141
.4964
0.0355
336
345
681
.4933
0.0595
June
TOTAL
dll11ni ( DU MAS,
ar ehanging seasonaUy in
ple/hodon
mogl1athus, ORGAN (56) found Ihree of the included five species to demonstrate
significant difierences, eaeh in favor of Ihe maleo
,
Numerous factors have been considered as operative to a degree result
nathush gravid females may be more secretive, hence a sampling error is intro
VISTA
48
DE B10LOG1.'\
TROP,leAL
subPa/mata. Females with egg dutches demonstrate ao escape respoose not mark
edly different from other individuals. Microhabitats utilized for oestiog offer
optimal habitat conditioos and are commonly frequented by nooreproducing
members of the population. Studies 00 activity cydes and movemeots have of
fered no evidenee to indieate that diel o, s<aSonal patteros are io any way dif
ierent between the sexes. Although fluctuations in population deosities are con
siderable, the proportioo of males to females does not significantly ehange.
The
and the probable continuanee of ieeding during this time would eliminate these
iactors as beiog of any importancc in
differential
survival.
ext,eme seasonal changes demonstrated during the d,y periad on the Cerro de
la Muerte cannot be equated with the impact of protracted extratropical winters.
Briefly stated, in Boli/og/ossa SllbpaJlIIa/a the sexes do not dernonstrate differ
ences in activity, habitat oc reproductive behavicr as have beco
reported far
WILLS
TURNER (74)
Rana
a"d
pretios". If
any signifieant differential io sex ratios does exist at various stages of the life
REPRODUCTIVE POTENTLU
Adult females became recognizably gr.vid between two observations that
varied in time from h:vo to seven months. The longest observed period during
which an individual was eontinuously gravid was eight months. These two periads
together comprise a minimum of ten months from the time of egg en,largement
,
VIAL,
49
The ' most simplified me!hod for estimating mortality rates is to deter
mine the proportion of macked animals present one year after capture. O 34
.individuals at Pop,ation II onl)' three were found after this interval. If all of
tbe original sample were equally at risk, the ruortality rate vlOuld be 0.91 pcr
year (P = 0.09 ) . Although fuU year records from other populations wcre more
limited in numbcrs, approx'm.tely tbe same rates were obtained.
To proceed witb any further examination of survival characters of
B. subpalllla'. several premises must first be aceepted : ( 1 ) tho estimated age
class re!ationships are v.lid; (2) reproductive and death rates are equal; (3) no
diffecenees in mortality are exhibited by the ,.xes; aud (4) no unbalaoced
migration patterns result in reduction or dilution DE the population.
ORGAN (56) has attempted tbe determination of survivorship rates in
fiye spccies oE DeS1llog1athlls utilizing the character oE mllltilobed testes and
interpolating !he age dasses from HUMPHREY'S (43) estimates. 00 the basis o
seasonal reproductive patterns in D. fIlSCII! the latter worker was able to assign
an absolute time interval oE two years between the successive formation al
testicular Iobs after altainment of maturity.
'
50
where
mOl
tn12Ro
equals the oumber oE marked saJamanders recaptured at the secood
COSTA
RICA
51
sampling period; m02, Ihe marked animals from the first sarople not recapturcd
until Ihe third; "'12, animals marked in Ihe fiest sample and recaptured "at the
second and third plus those marked at Ihe second sample and recaptured at
the third; '2 equals the total marked specimens taken in the Ihird samp!e; ul,
Ihe unmarked animals taken in Ihe second sample; and RO equa!s Ihe total num
bor of Ihe' fi[St sample, Substitutng values, P = (8) ( 5 ) + (28) (2) = 0.21
per year and is tr.ated as a constant for all age groups.
(6) (74)
CAUSES OP MORTALITY
REVISTA DE BIblOGIA
'!ROPICAL
Y/AL:
IN
COSTA RICA
'3
lnitial Wt.
jo Gms
Desiccated Wc.
in Gms
1.58
1.67
1.86
2.13
2.95
3.23
CUNNINGHAM
0.32
0.29
0.33
0.42
0.63
0.62
Water Conteot as
% oE Total
79.7
82.6
82.2
80.3
78.6
80.8
ly constant proportion of the total weight. 1 find no basis for his conclusion Ihat
,maller salamanders contain significantly more body water tban larger ones.
Five experimeots were conducted to establish the critical level of water
loss in B. !Nbpalmata. All
tation to cemove excess meisture from the cabinet and then allowed te stabilize
at room temperature. The insulated eabine! walls permitted air temperatures to
move excess body surEace water. Care was taken to avoid any c10acal discharge.
Weights were recorded immediateJy prior to experimentation and at intervals
of about
Six control specimeos were rnaintained in eovered moist chambers and subjected
to otherwise identical conditions.
No behavior attributable to stress was observed in animals Ihat had ost
up to
32.3 % of their initial body weight, but definite avoidance reactions wcre
ob..rved when the weight loss exceeded this value. None of the spedmens sur
vived water 10ss of greater Ihan.
given by STEBBINS
54
separately tested B. Jttbp"lmata weighing from 1.30 to 4.71 gms the loss rate
per boue was found t be an almost consistent nverse proportion to jnitial weight.
The results are presented in Table 10. Rates of loss are from 0.8 to 5.2 % per
houe, averaging 2 . 5 % .
Certain behavioc responses under stress of water 1055 were noted. In two
instances adults autotomized the tail, but the actual water loss level at which this
occurred was not observed. Coiling of individuals is aIso common once the stress
threshold is rOlched. Animals subjected to dehyration in the same container per
sistently congregated in tight, ball-like masses, which effectively reduced the
total exposed body surface area. Results from one experiment indicative of the
effect of such behavior is given in Table 11. Al! six animals ",rvived .the 18
hour test. No isolated specimens ever survived tbis exposure periodo
Tbe smaller animals appear to gain somewhat mQre advantage from this
cooperative behavior than larger enes. Those with initial weights of less than
two grams had the hourly rate of loss reduced by at least 0.90/0 as compared to
rates given in Table lO, and the average hourly cate foc the two experimental
groups was lowered from 2.5 to 1.1%, a difEerence of 1.4%. Similar behavior
has been opserved during aestivatioo of BatraclJoJepJ pafhu! (CUNNlNGHAM,
17) and in experimental studies on An,id" IlIgtlbriJ by ROSENTHAL (63 ) .
TABLE
10
dem01tS/I(/ting aggrega/ion
h/lmidiJicJ be/ween .50 and .58 % .
bcbavior.
Experimenti.1/
Initial
Wt.
Desiccated
Wt.
Time
% Lost
% Lrut/Hr.
1.30
1.65
1.72
1.74
1.96
2.02
2.33
2.41
4.71
0.88
1.24
1.27
1 .38
1.64
1.88
2.16
2.13
4.36
6.25
6.25
6.25
5.25
8.00
8.00
8.00
8.00
8.00
32.3
24.8
26.2
20.7
16.;
6.9
7.3
11.6
7.4
5.2
4.0
4.2
3.9
2.0
0.8
0.9
1.4
0.9
TABLE
11
Wt.
1.87
1.99
2.07
2.46
2.79
4.54
Desiccated
Wt.
1.50
1.69
1.71
1.97
2.01
3.64
% Water Lost
% Lost/Hr.
of Initial We.
16.0
IS.1
17.4
19.9
27.9
19.8
1.1
0.8
1.0
1.1
1.6
1.1
VIAL:
Reh)'dratiolt 01 de/;ydl'aua
55
12
111 21 /0 240 c.
Initial
Wt.
Dehydeated
Wt.
Envieonment
3
4
1.99
2.46
2.79
1.87
1.69
1.97
2.01
2.07
1.71
.l4
3.64
Submerged
Submerged
Submerged
Moist
Chamber
Moist
Chamber
Moist
Chambtc
N'
1
2
1.50
Deviation
Init. Wt.
Recov.
2.30
2.88
2.67
1.83
+0.;1
+0.42
-0.12
-0.04
115.6
117.1
95.7
97.9
1.96
-O.ll
94.7
4.37
-0.17
96.2
Wt. at
3 Hes.
1 5 . 1 and 30.30/0
1.50 to 3.64 gms
were rehydrated. Three of these were submerged in water and the others were
placed in covered ten cm culture dishes lined with saturated filter papero The
vessels \Vere again maintained at temperatures of between 2 1 and 24 C. Per
tincnt measurements on water recovery are presented in Tabie 12.
The three submerged specimens averaged 109.4% of initiaJ weight after
three hours, while those in the moist chamber regained an average of 96.2 % .
Although the difference of the two averages i s considerable (13.2% ) , there
seems to be little ecological advantage gained by submergence as compared to
saturated surfaces in rates of rehydration. A11 animals were we11 above the
moisture stress Emit of 32.3% within three hours and had essentia11y regained
their initial weight. Althol1gh two submerged specimens weighed more than
their nitial weight at this time, it is significant to note that N<:I 3, in recovering
95.7%, was exceeded by two of the moist chamber specimens at 96.2 and 97.9 % .
At the end of five hours a11 specimens equaled o r exceeded theie initial weight.
Comparison of dehydration-rehydration rates of these same specimens
cannot be made as aggregation behavior demonstrated during the former test
resulted in modification of the. weight loss rates. Even if only the moist chamber
animals are considered. however, recovery rates are much greater than dehydra
tion at the same temperatures, averaging 96.2 % for the total test peeiod or 32.1 %
per hour. This is six times gre.ter than the most rapid hou.rly 1055 rate (5.2%)
observed in separately dehydrated B. sI/bpalmata and twelve times that of tbe
average hourly 1055 rate of 2 . 5 % . COHEN ( 16), using the same animals foe
dehydration and rehydration experiments, found AI/eides ugl/briJ to lose water
1.25 times faster tban it gained, and Bmatil/a .eschscho/t to lose 4.66 times
more rapidly. It may be that his differences are due in part to the f.ct that de
hydration was carried out at 20' C and rehydration at 13' C.
Althougb the Cerro de la Muerte is subjected to intense roinfal! during
56
mueh of Ihe year, B. subpaJmata generaliy avoids drainage eourses and inun
dated areas. On only six occasions wcre individuals seen in reeently flooded micro
habitats. Experimeots were conducted lo determioe the length of time these ani.
mals can survive when submerged.
TABLE
13
s..,
Length
(mm)
37.5
39.5
40.0
42.0
42.5
42.0
45.0
40.0
51.5
52.0
S
o'
Wt.
(gros )
0.78
1 01
0.92
.
LIS
1.51
1.26
1.64
1.67
2.31
1.99
cr
(hes)
3.50
3 .75
7.7 5
6.50
5.25
PL
(hes)
6.75
10.25
10.75
7 .75
7.75
10.25
17.50
< 1 1 .00
< 11.00
< 1 1.00
Difference
cr PL
(hes)
3.25
6.50
3.00
1 .25
2.50
Pint jar, eoolaining 300 mI of well water were allowed lO stabilize for
24 hours at 22 to 23' e before experimenlation. The waler was nol .eraled duro
iog the experiments. Animals were maintained at room temperatures in cantain
ers partly filled wilh moss from Ihe habital sile and tesled wilhin a week of cap
Iure. Table 13 ineludes the observalions on leo experimental animals. The 'conlrol
group agaia compdsed six specimens maintained in covered moist chambers
duriog experimenlalion. Lo% of righling ability, (capture threshold) eilher
volunlarily or in response to strong pressure applied to the hind limbs,
represents the limit beyond which escape is impossible. The physiological
lethal point was taken as Ibe cessation of capillary flow in Ihe guiar
regioo. These crileria are essenlially Ihose used by STEBBINS (70). Time
intervals are given to the nearest quarter hOll! of determination, Loss
of righling ability occurred between 3.50 and 7.75 hours, with an
average of 5.35. A much broader range of loleranee was demonslrated in physio.
logical lelbal poinls (6.75 lO 17.50 hours; .average, 10.14). Differences in
periods belween CT and PL averaged 3.30 hours (range 1.25 lo 6.50 hour,) .
No behavioral . or physiological responses were observed in Ihe conlrol group. My
information is inadequale lo justify more Ihan a general inlerprelalion of Ihese
toleranees. It is evident, however, thal at eighl hours B. mbpaJmata has suffered
irreversible lethal damage. SUBBINS (70) found an inverse corcel.lion between
si2:e and tolerance to submergence in EnJatina eschscholtzi, which is not apparent
in Ihese experiments. The maximum lime for loss of righting ability .( equal,
"ecological limil" of STEBBINS) in three geographic races of Enwina was 95
V/
ECOLOGY
51
hours, while the greatest physiological tolerance (in unaerated water) was 98
hours.
The factors contributng to variations in habitat moisture have been treated
in othec sections. Twenty-ix sOll samples were taken fram quadrats ducing the
quarterly collections. The top centimeter of soi! was removed from a nearly level
site that was dear of debris Or vegetation. Inundated or super.saturated sites were
avoided. Laboratory melhods used for determining lhe amount of moisture as pero
centage of dry weigbt were presented earlier. In several imtances !he val"es
given here do not represent the actual moisture content of !he microhabitats;
however, most approximate it dosely.
The percentage of soil moisture and relative (available) densities of
B. subPalmata are given in Fig. 34. The influence of moisture upon the near
surface presence of salamanders is oovious. Below 30 % lhere is almost complete
retreat to subsurface levels. In four cases where salamanders are shown as occur
ring at moisture levels below 20% (indicated by "L") the animals were no!
oear the ground surface, but limited to and aggregated in localized sites of
moisture ' under deeply imbedded racks or decomposing 10gs. Thus, moisture
contents of the soil samples given for !hese densities are not truly representative
of !he microhabitat. Above 30% a wide range oE moisture conditions are tolerated
and !he optimum appears between 55 and 7 5 % . Because of !he influence it exerts
On evaporation cates, temperature undoubtedIy plays ao important role in mois
ture requirements and contributes to the magnitude of this range. Below DI aboye
certain limits, however, available moisture is a critica! factor in itself and canoot
be ameliorated by temperature. The low surface density at 83 % suggests avoid
ance oE excessively wet conditions in the environment and, although an isolated
record in measured soi! moisture values, this condusion is supported by field
observations. ROSENTHAL (63) ha, experimentally determined that Aneidss
I"gubris avoids excessively wet as weJI as very dry soi! surface conditions at
temperatures ranging from 2 to 29' C and !hat soil moisture is the dominant
factor influencing habitat selection.
THERMAL REQUlREMENTS
REVISTA DE
58
BIOLOGIA TROPICAL
specimens in the 37' e room were dead. When placed in a plastie pan eontaining
(CI'M) ,
BRATT_
STROM, 1 1 , for a general review) . It has bec11 eonduded that the approximation
of body and ambient temperatures is a result oi the necessity for these terrestrial
ectotherms to rnaintain body moisture by clase contact with the wet substratum
(BOGERT, 9; BRATT,'TROM, 1 1 ; eOHEN, 16) . .
Ove, 600 temperatures taken of microhabitats ocrupied by B. 'flbpa/mala
have ranged from 2.8
on
the Cerro
and
lO
although
tolerant
of
lower
thermal
levels,
at
which
8.8
thernial leveI,
IN
COSTA RICA
59
salamanders have
been
found. From over 555 records of 27 plethodontid species and subspecies he derives
an average of 14.8 C (range . 2.0 to 26.2 C), which substantiates the eoncIu
sion of BOCERT
(9) and
TIIE BIOLOGY OF
The family Plethodontidae i:s found in both the eastem and western hcmi
spheres, but in the Old WorId only two presumably reliet specios of the genus
Hydromdntes oerur locally ln the mountains and caves of southern Europe.
The family is di,tributcd from southern Alaska to northwestern Mexieo
in a narraw belt along the western coast oE North America. ]n the -east its range
continues frcm Nova Scotia to the northem half of Florida and broadly extends
westward to eover most of the area botween Texas and the Great Lakes region.
Three isolatcd inland forffis, one from Idaho and two from New Mexieo, are the
only plethodontids occurring in the subhumid continental region.
Ahsent from north central Mexieo, plethodontids reappear in tropical
Jowland areas of tbe east and west coasts and along the southern margins of the
Mexiean lateau. In the tropies the family oceurs throughout Central America
and mo genera, Boltogloua and Oedipina range into narthern South America,
the former attaining jts southernmost limits in Bolivja and along the Madira
and Amazon Rivers in Brazi!. A summary of the New World distribution of the
family is presented in Table 14.
Only three temperate zone species inhabit high montane regions. In New
variety of habitats
77). Of
two subfamilies and sixteen oc seventeen genera occurring in North Amerin and
Europc, aquatic, semiaquatic and terrctrial habits are demonstrated in differeot
60
TABLE
14
New WorlJ.
-o
e
o
<
Vl
""
:J ::l
GENUS
Dnmogl1(1/hIlJ
Leurogmtlhiu
Pheogn.:lIh;'J
Aneides
BdJrachoJpJ
Etl!atina
Hemidactyliu1n
PIe/hadan
SJereochi/lIJ
EUf)'cea
G)"hzophiluJ
HttideOlrilolt
'"
e"
o
e
-o
"
-;
e Vl
o
:t
,
(.')
""C
""
:;
.
o
<;
>
2
v
O-
v
o
""C
v
'"
"
Pseudo/riJon
Typh% molge
Typh'olrilo
Ji
", U <
Bo/ilogloJIa
ChiroplerOJrilol1
OediPina
X
X
Par-/limo/ge
PJetldoeurycea
linea/rilo11
(1)
X
X
X
X
.
"
o ..c v
MancuiuJ
Thorius
.!!
HydromalJl(JJ
H:Jplog/ona
:t
TABLE
61
Il
Subfamily
Stceam
GenUJ
Pond
Stream
Bank
Talus, Rock
\X'oodl3.nd
'Rubble
Arboreal
X
X
Desmognathinae
Desmognalhlls
/:.ellfOgllaJhlll X
Phaeogllalhl/s
X
X
X
X
Plethodontinae
Gyrinophi/IIJ
PIcudo/riton
Slereochi/IiJ
Euryeea
AfancllluJ
Typhlolritoll
Haid60lriloll
Typhlomo/gc
Hemidacty/irm.
Pie/hadan
Aneides
ElIsltna
Hydronul/lJ(!
Ba/rllcholcPS
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
TABLE
16
Genus
Bolitog/oua
Chiroptero/riton
Oedipina
Parr!imo/ge
Psclldoeuryua
Linea/rilon
Tborill1
Surface
Burrowing
Saxicolous
X
X
X
X
n....'elliog
X
X
X
Ji:
Arbores.l
JI.
X
X
62
REVISTA DE
BIOLOGIA TROPICAL
species known Erom lowland lo subalpine regions. Nine genera indude species
with primarily aquatie or slream-dwelling adults (Table 15). The most extreme
examples of aquatic liEe are two little known monotypic gellera found in caves
and artesian wells, the apparently paedomorphie HaideotritOn and Typblomolge.
Several species of E1Irycea also attain se.'lffial maturity without metamorphosis.
Typhlotriton is not known to depart from its underground aquatic habitat; while
Leurognathus cornmonly occupies slream beds, seIdom emerging beyond exposed
rocks. Stereochi/uI requ.ires slow-moving streams Oc ponds in low e1evations.
GyrinoPhilus and Pseudo/riton mark a transition between aquatic and terrestrial
habitats. Adults of these two genera are slream an'd slream bank dwelless.
The widest habital ulilization is exhibited within the genera Desmo;
"athus and Euryrea. Both inelude species of aquatic, stream bank or woodland
habitats. ManCl/lru is somewhat more terrestrial that most species of Etlr)'ct&1.
Hemidactyli"m is known from both slream banks and woodlands, while ill
Plethodon, WIquestionably a terrestrial genus, only one species, P. dUn/Ji, is a
Slream bank formo Batrarhoseps, Ensatina and Phaeogna/h", are woodland in.
habitants with sorne propensity for bureowing. A special type oE terrestrial habitat,
the rocky slope, is utilized by sorne species oE Al1eides, Hydromantes and Plelho
diJ1; genera most typically from woodlands. It appears that the axicolous habit
has been derived from a more general tereestrial ad.ptation rather than through
direct transition from an aquatic 'e.xistence, as the ooly genera to inelude rack
dwelling species are those with terrestcial or wcodland reprcsentatives.
Modern tropical plethodontids occupy a considerable range of habital
types, but no known tropical species can be ecologically equ.ted with the aquatic
or slream bank dwellecs of temperate latitudes. Jt seems probable that since direct
tereestrial developrnent has emerged as a broad adaptation there has been little
selective advantage in returning to water. Diversity has resulted fcom cefinements,
oc enensions) of utilization oE the tecrestrial environment for which direct embcyQ10gieaJ development has been a pread'ptation.
Most tropic salamanders are of a general terrestrial habit and are ground
wanderers living under cover provided by litter. Nonetheless, within the seven
recognized genera (omitting the enigmatic genus HaptogloSIa), species manifest
varying degrees of specialization. A surnmary oE the more distinctive habits i,
given in Table 16. The restricted habits known for sorne genera, e.g. ParvimoJge,
Tborius and ul/eatrilon are undoubtedly a result of limited collecting and in
adequate knowledge of tropical salamander ecology. Oedipi1la and Lil/eatriton
are elongate worm-like forrns) well adapted to a 'Subterranean existence. Talus
slopes and rock rubble are especiaJly Erequented by Cbil'opterotrit<>n ",osalJeri
.
and Bolitoglossa sooyarll",.
A unique community exists throughout much of Ihe tropic zone which
has provided amphibians with Eurther ecological opportunity. The "aerial swamp"
is a spectacular development o an exlresneIy complex epiphyte cornmunity Iypical
oE hurnid tropical lowland and montane forests. Many thousands of epiphytic
plants may conceal the major branches o a gigantie supporting trce. In species
composition thc "swamp" shows great variety, but the predominaot plants are
.
63
bromeliads or aroids having scoop_shaped leaves with bases that serve as reser
voirs of rain water. The 'vertical cxtent of the "aerial swamp" usually depends
upon the complexity of the forest and may extend between about 10 to 30 m in
storied jungles. Most reports of arboreal habits in tropical salamanders indude
reference to bromeliad dwellers. Certain species, Chiroptero/ri/<nl picadoi, Chiro/,'
fero/riton natalit and Bolit(Jglossa alvaradoi, are known only from this habitat.
The greatcst generic diversity in habitat utilization is .hown by ehirop
tero/ri/M and B(JIi/ogloSJa. Within the latter genus no species exceeds the valence
of Boli/(Jgl(Jssa stlbpalma/a, which retains the unspecialized terrestrial features
of the family and exhibits a variety of habits (Table 16) .
The extent to which the historical ond present distributioll of salamandcrs
has been influenced by temperature and available moisture cannot be overestim
ated. Because oE the intimate ecological relationships of the two factors it is desir
able to consider them together. It is fortunate that an instrument sueh as the
HOI.DRIDGE (39) model is avoilable by whieh this relationship can be empiric
ally defined.
Consi.tent with my previous use of this .y.tem in treating of the local
distribution of BolilogloSJa stlbpalmaJa, recent ecogeographic patterns of pletho
dontids and possibly aU salamanders, can be measl"ed in terms of biotemperature
(discussed earlier) and the potential evapotraas?iration rateo Poteatial evapo
transpiration (PET) is the amount of water lost by evaporation and transpiration,
presumng water lo be contiauously available. PET = (B) (K), where B is !he
mean annual biotemperature and K is a constant value of 58.93. To obtain !he
annual rate of PET, potential evapotranspiration is divided by the mean annual
precipitation (HOLDRlDGE, 41, 42) .
Supeeimposed on the base diagram of the Holdridge elassification of
Oimatic Vegetation Formations peesented in Fig. 36, is the general ecogeo
graphic distribution of plethodontid salamande. The latitudinal regions oc
cupied by the family are Tropical, Low Subtropicll, Warm Tempeeate aad Cool
Temperate. Equated with these regions aee Teopical, Subtropical, Lowee Mont.
ane and Montane a1titudinal belts. Each latitudinal eegion, defined by biotempee.
ature lirnits, ineludes the vegetation foemations indicated by eaeh of the hexagons
withia !hat latitude regioa and all of !hose ot lower biotempeeatuces. Foc example,
a total of 38 Tropical latitudina! formatioos ace defined (counting seven eaeh
foc Warm Tempeeate and Low Subteopical, whieh are sepacated by a critica!
temperatuce line of 17' C) . The seven latitudinal region, indude a total of 130
fQemations. Four eegioos and 30 formations aee occupied by plethodontids, whieh
are restricted to !he Humid, Peehmid and Superhumid Humidity Provinces
iadicated at the base of Fig. 36. It would be a m,take to assume that a neat
and dear-cut disteibutional picture sbaped from !he peecise mold of Holdridge
hexagons OCCUrs ia plethodontids. Salamanderi and many o!her animals exhibit
behavioral and physiological adaptatioas providing a means of escape from
direct expou[e to the physical environment. Fer example, hibernation, aestivation
and migration are meaos of evading unfavorable extremes and permit !he babit
ation of afeas otherwise ecologically inaccesible.
64
dry
Province are marginal habitats which may be invaded by way of more humid
environments. Aa excellent example of the occupation of marginal conditioas is
demonstrated by sorne salamanders of the warm temperate regions in the western
United States, sueh as Batra.hoseps a/telma/m, Aneides /lIgllbris and Ematina
esehseholtzi, that inh.bit the eak forest community of California. Although actual
Iy a part of the W.rm Temperate Dry Forest formation in the Subhumid Hu
midity Province, this cornmunity is occupied by these three species in localized
areas where topography and drainage compensate for climatically inadequate
moisture.
Few other salamandees inhabit the altitudes at which B. S1Ibpalmala is
commonly found. Pselldoellryeea ,ex is abundant in the pine forest belt of Guat
emala in the Tropical Montane altitudinal belt (the cool temperate dimate of
SCHMIDT, .65 ) .
In North America,
"pha/IIS are high altitude relicts occurring in different localities at mOre than
2,700
m, but neither of these spedes occllpies the subalpine belt. The upper therm
. HOLDRIDGE (39, 42 ) ,
about
4,000 m on the heat equator. Latitudinal increases wOtlld redce this limit
(3,820 m ) , in
the southern Talamancas of Costa Rica, offers an opportunity to test this idea.
Salamanders are not capable of prolonged exposure to subhumid conditions.
As a rule, moisture and temperature operate together in limiting the di,tribution
cf terrestrial species. There are, however, certajn conditjons in which the rn'o
appear to operate independently. la thermal gions characteristically ocmpied by
a species, moisture exerts a determining influencc. For example. A'tleides JI/g".
bri,;. oi the warm temperate region in California exhibits a clese association with
soil mosture and amount of anoual predpitation, which ROSENTHAL (63) con
siders to be more important than temperature.
The humid provinces are subjected to thermal lcvels which may limit
salamander distribution. B. subpaJma/a is es,entially limited to .Lower Montane
aod Montane thermal levels within the extensive Subtropipl bdt of compar
able humidity.
is evidence of adaptation to thermal levels within the humid regions. Most trapic-
' V/ ECOLOOY"OF
BOl.ITOGLOSSA SUBPAiMTA
IN
COSTA RICA
: 65
al 'pecies are upland forms, reflecting Iheir origins from tbe eool temperate Old
Northern. herpetofau.na (SAVAGE, 64) .. Certain genera, most notably BolitogloJ!a
and Oedipina" have ,been sllec;;sful in oeeupying the moist lowlands. ,
Thore are distinct differenees in the types of effeets of
environmental
temperatures upon B. subpalmalo. Mueh of the work done on the eritieal thermal
maxima (CTM) of extratropical salamanders has becn reviewed by HUTCHlN'
SON
(44) .
tbe value,of
CTM
CTM
for B. S1/bpalmaJa lies somewhere near 37' C, yet the species cinnot tolerate
prolonged exposuse to 27' C. Even mere important is the therma! level lirniting
species distribution, which more aerurately refleets the nature of adaptatiou. The
wide range of voluntacy temperanlCes recorded in the microhabitats (21' C)
in9,icates a broad ecological tolerance in adults, but there are subtle, yet critica!
aspects to the pieture. Miaoh.bitats may have temperatures ab'ove 20' C at times,
but B. sl/bpalmata does not ,ecm to be eapable of sustained brooding behavior
when subjected to prolonged exposure to sueb thermal levels. Moreover, the Ime
ecologieal tbermal maximum is probably near 16- C, above whieh eggs have not
bated at 26' C. It should be noted, however, that the altitudinal distribution 'lf
Therefore the tropical form has lo,,'er thermal limits than those tolerated by
can
with only liltle overIap. In aU but ne geilus' the life cyeles are notably eonsistent
and
can
the eggs are anchored in water and an .quatie larva possessing functional gill,
Type II ineludes forms whoSe eggs are deposited on land and have larvae that
migrate to water. After metamorphosis the adult emerges to oceupy a terrestrial
habitat.
are deposited 00. larid and undergo direet development. The embryo hatehes as a
replica of the adult and loses the allantoie gills irrunediately upon 'hatehing or
this group all of the patterns just described are representcd. D. '1tiadrarnd
ct/aIHI
is an aqu.atic form 'in a11 phases of its lite cyde. Terrestrial eggs and adult stages
a.e present jn D. fusa/s, but the larvae entee water upoo hatching and .do not
TABLE
17
Rcproducti....e Types
JI
1
DeJ:J1ogl1aJh/lJ
LeurogJlatbllJ
Stereo('hiluJ
EUf)'aa
Gyrillophi/IIS
T)'phlwilon
Pseudo/riton
ManI.,
Total 8
Typhlomolge
Hadto/ton
(1)
III
DesmugRa/hu!
HetnidaclyliNtJI
(1)
(1)
( 10)
(3)
( 1)
DelmognalhlJS
Aneiae!
B.tlraehoJeps
Ensatina
Plelhodon
Hydf'omftlltes
(1 )
(1)
(2)
( 1)
(l)
(3)
( 1)
(17)
(l)
( 1)
Tot:l..l 2
(26)
(J)
(1)
TABLE
Total 6
( 2)
PhaeognalhllJ
(32)
(J)
18
Winter
Spring
M A M
Swnmer
J A
X
X
Foil
O N
Oesmognatruoae
DeImognalhllJ
1.411rogn4J"1I1
PhaeogndlbllI
Plethodontinae
Gyrinophi/lIs
Pstlldotrlolt
Slereochi/JlJ
EUf'yua
Ma}](JlluJ
TyphJo/ritol
Hahiolriloll
Typhlof1Jolge
Hcmidatlylu11l
PIe/hadan
AneideJ
Ensalina
Hyaromames
Batra(hoseps
X
X
X
X
X
X
X
X
X
X
X
XI
X
X
VIA
IN
COSTA
RICA
67
tic larva. Adults of Mancullls qTtndridigilatlls aee terrestria! and return to swampy
areas to deposit eggs in water.
Complete escape from Ihe necessity of reproductive migration to .quatic
habitats has been attained by severa! extrateopica! geneea induding one species of
Desmognathus. p/,aeqgnatblls h,thrichti is not yet known in al! phases of its tife
cyde, but the adult is a fos>oria! woodland foem, which strongly suggest complete
ly terresteial adaptations. P/etbodOI/ indudes seventeen species,
al! of which
have terrestrial life cydes. E"satina, Eatrachoeps and Aneides also have the feee
living laeval stage completely suppeessed.
DUNN (22, p.
348 ) . NOBLE is probably eefercing to the same species when he states it .. eetains
the eggs in the oviduct and gives bielh to ful!y teansfoemed young" (53, p. 59 ) .
DUNN
(22), who repoeted this phenomenon in Bolitoglossa ( = OediP1iS) adspersa.
STEBDINS and HENDRlCKSON ( 7 1 ) have observed this 'pecies in the field and
question Posada' s observations. BRAME and WAKE (10) have ruade a com
successful members of the famity. Of the extratropical geneea, tbe genlls Plelho
dOIl has diffeeentiated ioto moee 'pecies and occupies a gecater geogeaphic aeea
than any other. What special advantage may be gained by a salamandee occupy
ing a terrestrial habitat is difficult to define peecisely but on the basis of compar
ative suevivorship eates in Desmognatbus,
may be the escape from higher mortality factors present in aquatic habitats, part
icularIy dueing embryogenesis and earIy larval life_
6S
P/e/bodo" all begin reproductioll in early summer and continue well into the fall.
In California, the eggs of Hydromantes are de osited only during the fal! (GOR
MAN, 3 1 ) . Only three genera have been observed to indude winter months in
the period of egg deposition. Batrachoseps is reported by various. workers to ncst
from October through March. Stereocbillls marginatlls is known to brood its
eggs io" January aod March; the latter month may rcprcsent lile termination of
its cycle (SCHWARTZ and ETHERIDGE, 66). Eggs of Matlw/us quadridigita/tls
have
.
been found during December, nuary and February.
The periods of reproduction reported for plethodontid genera are given
REED; 26) . It seems reasonable to predict that the reproductive period could be
prolonged if environmental imits were arneliorated. Except for isolated collect
iog records, information on temporal reproductive patterns in tropical salamanders
other than Bo/itog/ossa sllbpa/mata is !acking, but in view of the adaptive range
VIA'"
69
evidenced by temperate zone genera, it is not at all surprising tIlat this species
is acydic.
KLOPFER (48), citing data from Skutch (1954, el seq.) makes particular
note that breeding cydes in related species of tropical birds occupying the same
habitats are staggered, thus reducing the intensity of competition for food aad
nest sites. Obviously, alternation of reproductive periods pennits most cffective
interspecific use of any environment, tropical oc not, as shown by extratropical
plethodontids at both the generic and specios level. Jt is not yet known if
cydic and staggered reproduction OCCllrs in tropical salamanders.
Klopfer', statement, though valid, tend, te ohsCllre the intraspecific ad.
vantages of acydic reproduction. By e"tending the period of use of availablo
nesting sites, the degree to which reproductive pote:ntal ar population densitites
are limited by the number of sites is reduced as compared to poptilatians having
strongly phased reproductive cydes. Further advantage is gained in th.t most
unfavolabI physical changes in the environment are of short-tCIID duration and
wouJd not affect an acyclic population to the same degree as if encroaching at
a period when a synchronized reproductive phase is in progress.
Formation of testicular lobes in plethodontid salamanders ' has becn ato
tributed by HUMPHREY (43) to the following. cenditions: slow forward move.
ment of the spermatogenic wave; delayed regeneration oE emptied lobules; and
a time interval following which the long postponed regeneration brings into
existence a new lobeo 1 agree with his condusions in spite of the fact that they
are based upon the premise: "the urodelo male it will be recalled completes but
c.ne sexual cycle annually". There may he a gre.t deal of difference in relative
cates oE tissue reactivat.ion in DesmognathuI IIIS(1I1 and B. JubpaJmattJJ but a slow
spermatogenic wave and deIayed regeneration of the secondary germinal cord
dces not inhibit continuous sperm production.
That the environment is not a universal causative agent Di lobed testes io
salamanders is established by the absence oE lobing in the widely rustributed
temperate zone Plethodon gllltinoSt/s. The slow movement of tbe spermatogenic
wave may be continuous .nd tIle period oE time between degeneration of spent
lobuJes and reactivation of this spent ti>sue may be relatively short, permitting,
regeneration oE part of the lobe before the remaining portion has been completely
exhousted DE sperm cells. Testicular lobing is known for several members of
the subEamiJies Desmognathinae and PJethodontinae and probably hec.me genet
ically fixed in thejr early history. This would explain most simply the occur '
rence oE lobed testes in both extratropical ond tropical Eorms. The alternative is
lo conceive of this choracter as having originated separately and independently
in several other tropical species closely related to B. Sllbpalmala, such as B. ads
persa, B. cel'roensis and B. sooyorll1n, which, on the best morphological and geo. .
graphical evidence, are tropical upland forms DE the Central American complex
oE tbe OJd Northern herpetoE.una (SAVAGE, 64). On the basis af observations on
B. sllbpalmata, reduction in the number oE testicular labes . in tropical species
may be a, prevalent cha:racter. The continuous regeneration oE germ cord cells
could contribute ta .n increase in size and consequently a reduced number of Jabes' .
70
(WILLS, 8 1 ) and greater nutritive requirements for yolk storage. Even with su;
tained feeding during brooding, and the absence of .ny prolonged dormant period,
an approximate!y two year egg laying interval is demonstrated. In view of the
apparent absence of environmeotal nfluences it seems likely that the te.mporal
patterns oE maturation exhibited in B . fflbpaltl1flta
originated in the coo1 temper
.
ate antecedents of the species and have proved non-adaptive in the tropies. The
lack of any necessity foc extended hibemation oc aestivatioo) and ao essentially
continuons avaiJability of food, has enabled the population to proloog breeding
activities throughout the year. Continuous reproduction in' anurans has been re_
ported in nine ' species from the Argentinian Chaco by CE ( 14) ; in seven
speO'" from Singapore by' BERRY (7) ; Bufo. mel"nosettlJ from Java by CHURCH
VIAL..
71
DUNN (23) and NOBLE (52) have effectively argued for the aquatic
origin of the Plethodontidae and NOBLE (53) has described brooding behavior
in sorne oi the more primitive contemporary forms. lt seems paradoxical to
consider tbat antibiotic secretions are likely to have originated in conditions where
they are subject to rapid dilution and removal in flowing water. Neither GORDON
(28) nor HrGHTON and SAVAGE (38) found positive evidence of antibiosis.
DANrEL and SlMPSON (18) repnrted negative results in antibiosis experiments
with three species of $alamanders, although none were plethodontids or terres
trial egg brooders. Thre is little doubt that the brooding adult plays an active
role in surviv.1 o the <ggs of
function may
8. s1lbpallla
ldl Ihe observed manipulalon and rotalon of the cluleh is analagous.
The importance of the brooding habit may Jie in a dual or multipurpose relation
ship, one of whieh may be the necessity of frequent kinetic a"istance.
The nature of the brooding instinct appeaes highly variable in pletho
dontids. Females of five species of Desmog,ahtls were never observed to leave
their eggs, even for feeding (ORGAN, 56) ; GoRDON (28)
repnrts a similar
situatiofl foc Aneides aene/lJ. Intermittent wanderings have been oted among
brooding females of A1Ieides hardii (JOHNSTON and SCHAD, 46), and EIIsatill.
72
oE plethodon gluln,oJUS grew betWeen 6 and, 21 mm per year, the smallest baving
the most rapid growth, He also noted that hatcblfngs of 12 to 15 mm almost
double their length 'in the first eight months and may exeeed 30 mm within
the year, In northern populations, P. glutinos/Js grows more slowly, but attains
a larger size than tbose in Florida (HIGHTON, 37), Ba/rachoseps allenllalllJ
demonstrates a slower growth cate aboye 35 mm than in smaller individuals
(HENDRICKSON, 3 5 ) , Ens<tlil1a hatehes at about .20 mm standard length and
may grow to 35 mm by its seeond spring at approximately 18 months of age
(STEBBINS, 70). Bo/iloglossa subpalmala, emerging at 10 mm, and requiring
an estimated two years to double this length, is not greatly diHerent from the
latter' species in its early proportionate growth.
Maturity is attamed at a fairly uniform age in temperate zone pletho
dontids. Ple/bodon g!rlfinostls is usually , capable of breeding at the third year
in Florida and at five years in the oorthern part of its range (HIGHTON, 3 7 ) ;
Ensa/ina usually in the third, although some may b e delayed until the fourth
year (STEBBINS, 70). Aecordiog to POPE and POPE (61 ), Plelbodon ollacbila.
matures, in the Eourth season, anci P. yonah/osee probably requires the same lime
(POPE, 59) . The western species, Bmrdchoseps allenllaltls, is estimated to require
two to fom years before reacbing reproductive capacity (HENDRICKSON, 35 ) .
In some species of Desmognmhlls the males matme io the foutth year, while
females are delayed until the fifth (ORCAN, 56). The smalle,t male B. sllbpa/
mata pcoducing sperm is estimated to be four years oi age and egg laying in
fomales probably OCCUrs o earJier than the tenth year. With variations in growth
ando
. .the absence of seasonality, maturity may occur earlier.
Maximum age estimates of other plethodontids are almost hall tbat
coinputed for the species under stooy, HENDRlCICSON (35) gives a maximum
age of teo years in Ba/rachoseps a/te111ltttTlS aod STEBBINS (70), eight and one
half years for Enstttina. 1 , can ' oHer no .dequate explan.tion for the marked
retitrdation in growth and maturity in' B. I1fbpaI11lata. One possible' factor ' may be
low activity levels of the populations whieh could be reflected in redueed growth
rates and retarded maturity. Unfortunately, there is , no quantit.tive data on
activity levels for any other spe<ies of ,alamander. , HIGHTON (37) has suggested
that the reduced growth rate of .-Ple/hodon g/Tllinosffs in the northero .United
St.tes i related to the decreased. activity imposed by hiliero,:tin.
It is generally recognized that tropical -biotic communlties inelude a
greater number of genera with fewer individuals of eaeh member species than
those of temperate regions. Nonetheless,"the densiries of B. ItIbpalmata on the
Cerro de la Muerte are rarely exeeeded by .ny 'species oe plethodontid in the
temperate zone. Several features of the ecology oC this tropical salamander', elu
cidated by the present study may help oto explain the observed popultiol mag
nitudes : ( 1 ) high produetivity ' levels in the tropiCal environment, ' (2) broad.
ecologieal use of the habitat, (3) the possibility of relarively low energy d.mands
by, the 'pecies, (4) acyclic reproduction, and (5) lack of territorial requirements.
The estimated deosities of the ' Cerro populatioos markedly exeeed tbose 'of all
other terrestrial plethodontids except 'those observed by ANDERSON (2Yin the '
VIAL.. ECLGY
OF BOUTOGLOSSA SUBPALMATA
IN
COSTA
RIU
73
Itlgtlbris. Classieally simplified island ecosystems which may indude large num
bers of individuals of few species are not homologollS to lhe complex tropical
habitat of the Cerro de
animals; and the small proportion of the total which are active during any
given pedod. The values of P = 0.09 and 0.21 derived by the simple propor
tional method and the LESUE and CHITTY
are probably unrealistically low, being based upon the assumption that an animal
ca.ptured oo ane Occasion is equally at risk of capture on the next. As this ii;
but lhe first attempt to apply the latter procedure to salarnander populations, its
reliability under such conditions has yet to be tested.
cantly higher results. Based upon single capture data, age specifie survivorship
rates of
greater ' Ehe ecologieal mortality, the greater the reproductive potential of ' lhe
'pecies... " (ALLEE el al.,
eggs pe! clutch than most plethodontid species known, aCGording to comparative
data given by ]OHNSTON and SHAD
population replacement and shor! life expectancy. Tho estimated individual long
evity of adults is unusually high. These factors would indieate lhat eggs and
young have but little chanee for suival. The absence of any homing behavior
or strong fidelity on the part of btooding adults, as compared to olher species,
infers a higher mortality in eggs than possibly any other phase of lhe life cycle.
As no informatian is available on the survival of eggs, the estimafed reproduetive
potential is definitely exeessive. If; as assumed; the population is neither inereas
ing nOr decreasing, the balance would be best indicated by a lower realized re'
prdductive performance and an annual survivorship rate of between the calculated
minimum and maximum
(P
be made witn his computed life table vaIuesJ as there are SigM
nificant features in the life cycle of desmognathines which contribute to a dif
ferential survival in the sexes. These faetors, specifically .considered earlier, are not
characteristic of B. subpalmata. Certain of Organ's data offef a means by which the
REVISTA
74
DE BIOLOGIA TROPICAl
less equated. By omitting brooding females aod assumiog that males and 000brooding females of Deslllogl1alhlJI represeot the cooditions obtaioiog io B. sub_
palma/a, a reasonable compari"", can be made. ORGAN (56) has determined ao
nual survivorship cates for to age periods, "eari}''' and "later" life, bascd upon
testicular lobes .od year cla.rs<. Although the age peciods, maturit"f and ye.,
classes are not the sme as in Desmog1'lathtts, the lobe formatioJl periad, two years.
approximates lhat of B. subpalmeta and the values are derived by the same pro
cedures (Table 19). It should be carefuUy noted that "Iater agc" of Desmog
""'",s is estimated to be youngor than the "early age" of B. !ubpaJmata (seveo
to eight and nine to ten yeacs, respectively). Therefore the value of 0.65 for
"early age" B. sl/bpalmat is evidence cf greater survivorship than cf any speces
of DeS1nog11atblls at "later agc" and is consistent with Organ's conc1usion that
survivorship vaLu.es and degree of terrestrialism in salamanders are correlated.
TABLE
19
Mean I11mu/ s:w);vorship 'fates in age periods o/ some plethodo1Uid sa/amauders. See Jext
for explanaJion o/ age ategories.
$pecies
Early Age
Later Age
D.
(jllttdramam/altt.r
0043
0.38
D.
m01JIi(ola
0.53
D.
0.51
IUJ(uJ
D.
o(hrophaeuJ caro/hJensir
0.62
0.74
0.91
0.6
D.
wrighti
B.
sllbpaltnaJa
0040
0.7
0.29
0.29
1"5
S U M M A R Y
An intensive field shldy was conducted on the montane tropical salamand
ero BolitogloSJ.1 wbpalmata (Boulenger) in the Cordillera de Talamanca of Costa
Rica. Information was obtained on over
samples and from six marked populations. Measurements oE the physical environ
ment and laboratory experiments supplemented the field data. The results and
conclusions of the study are as follows:
2.
in adult males. The lobular formation of testes is similar to that which has been
described for Desmognathus. Females prabably require about two years to com
plete a cycle oE oviposition. although at any
are
population
in
reproductive
stage.
time
sorne
Convoluted
individuals
oviducts
are
in
the
retained
3.
sites. The se1ected sites permit avoidance oE both excessively wet or dry conditions
and have a stability of temperatures between about
to be the optimal range (the upper level may be ecologically critical) . Usually the
size oE a nesting site restricts its occupatlon to but one cIutch; however, up to
three egg masses with their attending adults have been found in a single "nur
sery".
4.
not situated in a grape.Jike cluster at the time of laying. the ovipositing female
may actively gather them together. The clutch averages
to any object. The eggs are .bout 5 mm in di.meter and cohere by the outer
membranes. Per.sistent, but not constant, attendance oE an adult is required fo!
sustained development. Both males and females have been abserved to brood
the clutches.
5.
After the estimated four to five month period required for develop
6.
'TIle spedes is active throughout much of the night, howcvcrl the p;1t
76
terns of activity are variable and influenced by temperature and, more frequently,
by humidity. Maximum activity levels appear at temperaturcs aboye 6 C and
relative humidities of more than 80 % , and are somewhat reduced at times of
actual precipitation. Probably between 1 0 and 2 5 % of the near-surface population
is active during any given night.
7. Vagility in B. subptilmata is limited and individuals utilize a definite
home range. No statistically significant differences were found in the movements
of males, females and juveniles either in distan ce or in frequencies. The average
home range area for all of thlOse groups is 44 sq. m. There is no evidence of any
homing behavior in displaced individuals.
8 . Aggregations appear to be a response to unfavorable environmental
conditions and are not related to reproductive behavior. A voidance of extremely
dry or wet conditions results in congregation at the most favorable habitat sites.
9. Densities of B. subpalmata on the Cerro attain striking proportiom
and are loca11y controlled by density-indepcndent factors. Population densities
are correlated with elevation ; the greatest densities are found near the summit and
in this area may exceed 9,000 animals per hectare (3,690 per acre) . Both local
and seasonal changes in moisture influence the occurrence of animals at the sur
face and during the dry periods they retreat to 'Sub-surface cover.
1 0. Age class estimates of populations, admittedly speculative, are base!
on the measured hatchling size and calcub.ted growth rates. The maximum age
attained is about 18 years. Sexual maturity is apparently attained by males in the
sixth year although it is not until the twelfth year that mst females have deposited
eggs. Males probably not only mature more rapidly, but reach senility earlier than
females. The expected one-to-one sex ratio occurs at a11 seasons and throughoul
a11 age groups except the olde$t, in which females notably exceed males.
1 1 . Observations of numbers of egg clutches and spent females are used
for determining the estimated annual partal potential reproductive capacity of
58 %
1 2 . Survivorship and mortalty estimates, computed by different math
ematical procedures, are variable to a degree that casts serious doubts upon the
applicati6n of most of these methods to salamander populations. The fluctuation
of population densities, extensive use of the habitat, and low activity levels
contribute to problems in using recapture records. The consistent absence of im
mahue animals in the samples presents almost insurmountablc problems in using
single capture data.
1 3 . The best available annual survivorship 'estimates are between 0. 29
and 0.65, decreasing with age. It has not been feasible to construct life tables or
survivorship curves. The potentially high reproductive capacity and low observed
mortality factors acting upon adults suggest a high juvenile and egg clutch mortal
ity. A high mortality rate of eggs is further indicated by the absence of strong
attendance fidelity on the part of adults and lack of homing behavior in B. sub
palmata.
1 4.
77
trpicl piethodontids. The total body moisture averages abollt 80% of body
weight. No behavioral stress is noted io aoimals losiog up to 32.3% of ioitial
body weight, althougb none survived a loss of more than 41.2 % . Rates of water
loss averaged 2.5 % per hOllr, but are somewhat higher io smaller animals. Loss
15.
merged eonclitioos ranged from 3.50 to 7.75 hOlles (average 5.35), and ees
sation of eapillary eirculation was noted at from 6.75 to 17.50 houes (average
10.14). The average differenee between these periods was 3.30 hOlles.
16.
habitat areas. Salarnanders are virtually absent at less thao 30 %, the optimum
range being between 5 5 aod 7 5 % .
17.
between 4 and
results after less than two hours exposure. Temperatures of 27' e are letbal after
19.
(a differenee of 2 1') is greater than tbat' reported Eer any other pletbodontid
species.
20.
than those in the tropies. None o the latter are known to be aquatie. Probably
there ha:s been little selectve advantage in returning to water afte! the evolution
o direet development. B. wbpalmata is not exeeeded in its habitat utilization
21 .
22.
23.
78
plmala maximizes use of available nesting sites and permits <vasion of shortterro
VIAIo
79
80
usan para determinar una capacidad potencial r,eproductiva parcial anual calcu
lada en
58%.
12, Los
dOs matemticos, son tan variables que hacen dudar seriamente la validez de 'a
aplicacin de muchos de estos mtodos a poblacioneS de salamandras
La fluc
tuacin de la densidad de las poblaciones. el uso muy e.,:tenso del habitat y los
de recaptura.
y. los factores de mortalidad en adultos sugieren que debe haber una elevada
mortalidad juvenil y de las nidadas. Adems. la falta de una fidelidad fuerte en
'i
a las de
80%
2.5%
este ltimo ambiente no ofrece ninguna ventaja ecolgica, o muy poca a lo sumo.
15.
horas.
16.
3.50
17. Las
55
30%. las
75%.
'a
37' C
de unos
VIAL.
19.
RICA
23,8'
81
C (una diferencia de
21'
2,8 a
20.
de habitats que los de los trpcos. No se conoce ninguna especie tropical acu
tica. Probablemente ha habido poca ventaja selectiva en el regreso al agua despus
de la evolucin del desarrollo directo. Ninguna especie de la familia sobrepasa
a B. mbpnlmata en utilizacin del babitat.
21.
nivel de
22.
mbpttl1l1dta.
23. '
24.
festacin de 105 lmites impuestos por las cap.cidades metablicas del tejido
germinativo primordial, y no tiene correlacin con la periodicidad del ambiente.
25.
Se sugiere
que,' de
ADDENDUM
Since June
7, 1965,
2 1 : 228-236)
of an
82
LITERATURE CITED
1.
Ce.,
Saunders
8,7 pp.
2.
Ecology and
c\'olution
3.
in
tbc . San
30: 3593S5.
M. D. D.
BACHMANN,
1964.
P. K.
ANDER..(QN,
.
1960.
cil1erell!.
161 pp,
4.
1952.
29:
120-127.
S.
B.......'>:NlKO\', A. G.
1949.
the
on
USSR.
BANNIKOV, A.
6.
1950.
Notes
1:-os
Angeles.
d.
rett, Unho, Mich. Mus. 2001. Edited $ond prepared for duplication by F. B.
7.
BE:R.RY, P. Y.
1964.
193.
BOGERT, C.
19n.
A,":mrJ
YS5 pp.
M.
1963.
xiv
Rehtivc abund':lOce, habitats and ooonal thermaJ Ie\'ls of sorne Vi rgi nia
salrunanders.
10.
J.
BrSHOP, S. C.
8.
9.
19'7.
&:
D. B. WAKE
Js A'1tgelcf Co.
MI/f.
.Coro/r.
Sci.
1 1.
BRATTSTROM, B.
1963.
H.
gy,
44: 2,8-254.
12.
BURGER, J.
19 H.
W.
60: 4)94870
VIAL'
13.
14.
16.
83
Bun, W. H_
1943,
as
CEI, J. M.
1949.
15.
CHURCH, . G.
1960. Annua! and lunar periodicity in the sexual cyclc oE the ]avanese toad,
Bufo rnelanortictu Schneider. ZooJogica, 45: 181-1$8,
N. W.
Cornparative cates oE dehydration
.sa.lamaoders. Eco1ogy, 33: . 462479
COHEN,
1952,
:lnd
hyclration
in sorne
California
17,
CU!'lNINGH.AM, J, D.
1960.
18.
DANIEL, ]. c..
1954.
19.
D.WIS, D. E.
&:
F. B. GOLLEY
10:
16.
xiii +
DUMAS,
1956.
22.
SIMPSON
DICE, L. R.
1952.
21.
R. W.
1963.
JO.
$:
P. C.
.The ecological relations of sympatry io Plelhod()1; dJumi aod Plelhodo"
r.ehicuJSlm. &010g1, 37: 484-495.
DUNN, E. R.
1926.
versar] PubJ.
23.
DUNN, E. R.
1928.
24.
DUNN, E. R.
1937.
25.
26.
FITCH, H. S,
1958.
Home raoges, territories and season::l.J movements (lE vertehrates o the
natural history reservation. Univ. K,m,r. PuM. Mus. Nttl. Hist., 1 1 : 63-326.
GNAEDINGERJ l. M.,
194ft
&.
C. A. REED,
GEIGER, R.
Tbe climaJe llear be gl'olJ11d. Harvard Univ. Press, Cambridge. 1:82 pp.
Translated by M. N. Stewart. el al.
GoRDON, R. E.
19:52.
A contribution to the life history aod ecology oE the plethodontid sala
mander AneideJ aenetu (Cope aod Packard) . Am. MM/. Nat., 47: 666-701.
1950.
28.
29.
GoRDON, R. E.
1953.
A popuIation oE Holbrook's salamander, Etlf'j'cea lotlgicauda guuole;1ft;t
(Holbrook) . Tu/on, Stud. Zoo/., 1 : 55-60.
30.
GoRDON, R. E.
1961.
31.
!!-colog),
42:
200-202.
GORMAN, J.
Reproduction in pk1:hodontid salamanders of tbe genw Hydrom.-J.i116J. Htt1'
1956.
pCJoJogica, 1 2 : 249-259.
32.
GnoBMAN, A. B.
1944.
33.
34.
HAIRSTON, N. G.
19-19.
The local distribution and ecotogy oE the plethoclontid salamanders of the
southera Appalachians. Ecol. Monographs, 19: 47-73.
HEATWOLE, H. F.
Rates oE desiccation and dehydration oE eggs in
1961.
35.
terrestrial , s.1Jamander,
HENDRlCKSON, J. R.
1954.
Ecology and systematics oE saIamnnders oE the senus Batrachose!,s. Uni1?
CoIij. Pub/o Zoo/., 4: 1-46.
36.
HIGHTON, R.
1956.
The life history oE the slimy salamander, Plethodon glflliltoSNS, in Flori
da. Copeia, 1956: 75-93.
37.
HIGHTON, R.
Geographic variation in the life history oE the slimy salamander. Cop6ia,
1962.
1962:
38.
597-613.
1961.
39.
HOLDRIDGE, L. R,
1947.
Determination oE
data.
HOLDRIDGE, L. R.
la vegetacin de Costa Rica. In Trejos, E.W. and .A. Archer AlZas Es
1953:
85
41.
HOLDRIDGE, 1 . R.
1959. Simple method for determining potential evapotrampiratioD from temp.
erature dat. .kel1cel 130 (3375): 572.
42.
HOLDJUDGE., L. R.
1964. lije Zone EfO/ogy: P':Ofi. EJ. Trop. Science Center: San Jos, Costa Ric .
124 pp.
43.
HUMPHREY, R. R.
1922. The multip e testes in urodeles. Biol. BIIII., 43: 4567.
44.
HUTClDNSON, V. H.
1961. Critica! thermal maxima in salama.oders. Physiol. Zool., S4: 92125.
4.
46.
Univ. Kans.
PMhl.
47.
KINGSaUR-Y, B. P.
1902. The spermatogenesis of Desmogll:1lhUI fusca. Am. /. Anal., 1 : 991 5.
48.
KLOPFER, P. H.
1962. Behat.ioral aspul! of ecQ/ogy. PrenticeHall, Ine.) New Jersey. 175 pp.
49.
50.
"
of
J.
The ID3ximum
51.
mearu
Iik'e1ihood equatons
38: 269292.
LLoYD. J. ].
1963. Tectonic history of the South CentralAmerican oro8en. (In: Childs. O.E.
and B.W. Beehe, Bad:,brme of Ihe Americas) Am. /trlor. Petr.
2: 88-100.
Geol.,
Mnn.,
52.
NOBLB, G. K.
The plethodontid salamanders; !lome aspects oE their evolutioll. Am. MUJ.
1927.
Novil4leS, 249: 126.
3.
NOBLE, G. K.
1931. Th. hiolo" 01 Ihe amphihia. McGraw-Hill Book Co., New York. 577 pp.
54.
O'GAN, J.
19'8.
55.
A.
Courtship
and
1958: 251-259.
spermatophore
oi
P/elhodon
ordon;
metra/fi.
Copeia,
OIlGAN, J. A.
1960. The courtsb..ip and spermatophore o the salamander P/sthodol1 gll/linolllJ.
Coptia, 1960: H-40.
86
16.
ORGAN, J. A.
1961. Studies on thc 1o,:,;11 distribution, life histOfy, and population dynamics of
the salamander genus Desmogna/hus in Virginia. Eco/. Monogmphs, 3 1 : 189220.
l7.
PIc,",o, C.
1913. Les brolneliaccs epiphytes considres comme une milieu biologique. Bull.
Sci. Francf Be/g., 5 : 2 1 5-360.
58.
PIERSOL, W. H.
1910. The habits and larval state of PIe/hadan ciucrcus er)'/h071otuS. Trmu. Ca
nado 1ml.} 8: 469-494.
19.
POPE, C. H.
1950. A statistical and ecological. study of the salam::;.nder
osee, Bull. Chicago Acad. Sci., 9: 79-106.
60.
Pon:, C. H.,
1949.
&
62.
)'{)lu//;.
SAltAH H. POPE
Ple/hodan
Ple/hadon glll'
!lO
63.
ROSENTHAL, G.
1957. The role oi moisture and temperature in the local distribution of the
plethodontid salamander AneideI /ugubris. Univ. Calil. Publ. Zool., '4:
371-420.
64.
SAVAGE, J. M.
1960. Evoluticin oE a peninsular herpetofauna. Syst. 2001., 9: 184-212.
,
6'.
ScHMIDT K . P.
1936. Guatemalan salamanders of the genus Oedipll!. Chhago Fie"! Mm., Zool.
Ser., 20: 131-166.
66.
61.
SMITH,
1906.
B. G.
Preliminary
report
on
the
embryology of
Cryptobranchus aJleghenie7uiJ.
..
:
pLY, P. C.
STA
1
1937-38. Flora of Costa Rica. Pield Alus. Nat. HisJ., BOl. S"., 18 (1-4 ) : 1-1616.
VIAL,
69.
STEBBINS, R. C.
1951.
70.
87
pp.
STEBBINS, R. C.
1954.
71.
&: 1. R. Hn:nRICKSoN
Field studieS oE amphibians iD Colombi:t, South Americ:!..
STEBBINS, R. c.,
1959.
Univ.
Ca/il.
72.
1948.
THORSON, T.
1955.
B.
The rebtionship oE water economy to terrestrialism in amphibians. EC'o/
og;, 36: 100116.
74.
F. B.
Population structure and dynamics oE the western spotted frog, Rana p.
pre/ioa Baird and Girard, in Yellowstone Park, Wyo. Ero/. Monographs, ,O:
TURNER,
1960.
251 278.
7'5.
TURNER, F. B.
1960.
Tests of undomness in raptwes oE Rfl11d p. prelios4. E(% gy, 4 1 : H7239.
76.
TURNER,
1960.
F. B.
Size :md dispersioo of a louisiana population oE the cricket frog,
gryll'J, Ec%
77.
W'YL,
1957.
80.
B.
Comparative osteolob}' and evoJuuon oE the lungless salamanders, family
PJethodontidae. UnpubJ. Ph. D. Dissertation, Univ. So. <:alif. 443 pp.
WEBER, H.
1959.
79.
AuiJ
258268.
WAKE, D.
1964.
78.
gy, 4 1 :
WEYl,
ld61.
REVISTA
88
81.
DE B101OO1A TROPICAL
WJ.LLS, J. A.
1956.
82.
ZWEIFJ!L, R. G.
19S7. Studies cn critical tbermal maxima of salamanders. Ec% gy, 38: 64-69.
Fig.
1.
V/AL:
COSTA R I CA
89
'"
o
Fg.
2.
Elevations
designated by
Geo
"x"
WS
Weather StatioD,
;;
e
'"
'"
;:
...
IN
COSTA RICA
91
1.
O'
;,
,.
-.
o.
!
!
.:l
;f;o
<.
O.
-o
o:
.,
__
;"
92
Fig.
3.
maximum
and
mnimum
showing
temperatures
4.
maximum
and
mnimum
temperatures
5.
1962.
la
Muerte.
Opeo cicdes
the
represent records
"
,
"
..
"
z u
w"
o
'"
te
"
<
,r ,
--,
.'
I 'o' ,
,--....
_
_
--,
,
, ," '"
"
,.
"
"
3
"
.,
'"
"
'"'
'"
"
,
z"
w"
g 1 4
:;: 12
ffi
11 )
" ..-\.../\
.....T..."' , "
,-I
,
'
,
,
,
,
'
',
'"
,
"
,'
"
"
..... ,,/
'"
"i
"
, ,'
, ,
'.
roo
'e
: '
r --
",
, ,-'
oo .
"
.,
"
,.'
,
_._ I
V
,
,
"
,_o',----
A
'
."
'-
,
1 '-__,
,o,
'....
'.'
"
, I
," ,
"
"
MONTHS
..
,
30
"
,
30
,
1
,
"
,
//'\
!-' '0-
7/
.-
\\
,.,--f--,
O -,
,
,
,
"
,
"
93
94
Fig.
6.
VIAL:
RICA
9S
96
REVISTA
Fig.
7.
DE
BIQLOGIA TROPICAL
S.
General aspect of
:lO
LQmarhl loxemis,
/irioides)
terrestrial bromeliad
(Pllya dasy
97
)g
:'i
fig.
9,
upper left
contain
the
testis, viewed
oc
smaller spermatocytes H .
In the
lower
"
'"
'"
-
g
8
-
>
g
:!!
'"
COSTA RICA
99
100
Fig. 1 1.
extraovular sp:!ce;. d
=
=
palma/a.
The
enveloping
geJ::tinous
membranes
are
omitted.
Fig. 1 3
Fig. 14.
ID
Figure 13.
V/AL:
101
o
b
12
11
102
Fig. 1 5 .
formation
in
B. Jubpa/mala,
dorsal
Fig. 17.
Fig 18.
.
developed,
1S
Pig
not shown.
.. )
._-_._
-_.--,/
19
103
Fig. 20.
recaptuJ"cd in Octobel",
196,.
B. Jllbpulll/(lI1
inc!uded.
Males
are
standard
leosth.
Triaogles repn:sent
Fig. 22.
Average
annu!
growth
rates
21.
Fig. 2 3 .
loCro
juvl'niles;
cirdc reprcsents
the
"
'"
..
>
;;l
...
70
60
I
1C>
50
..J
:...:.:.:::.:
l.':',:. .
. . . .,
.:: ::.:.
...
..
' .
.
:: :
;
;
;
,
.
;
;
,
;
,
:;
,,
.
"
;
,
;;
:
:;
;
:
.
:
;:.;:,
,:,;::
. . .'" ,. " " ,.'. , .:. :.',:,:, :
.:::::::::::::::::::.:::...... .
." ",
_, ,'j_' ,
., ....., ,
,.1.'"
o...
o.
.....v.
!::.'::....
O
o:
..
o
%
7
"'''",,
'"
0 40
2
..
'"
1'"
'O
:;
.. . . . .
i
2
o
:;
o
1A$oNI
1961
J. ! J. '.
f1
O
o
o
o
o
o
o
00
IN1T1Al
srAHOAIID
o o
o
l(NGtH
30
0 0
OO
.t
to '
..
STANOARD
... .
.-
INllIAl
O . .
.",
'
'O
L!I;TH
'"
..
11
GROWTH CURVE
of
g
;
>
8. slJbpolmolo
l':
'"
,,-
o
" 30<
O
<
:;; 20-
;:
"
Ollllnld Ran91
/./------
// E$llmalld Ran".
,oJ
11
i<:
0X5
w 40_
"
0 0
00
//:r::d Range
1:-
"
00
,6
.. ! ! 'l. 01'. I'1l. '. '. l , b1
"O
7'-
20
"
. . . .".
!)
10
,
..
12
AGE IN YEARS
'
13 14
15
16 n
18
l' W
13
o
o-
Fig. 24.
during
that periodo
Soil.air
interface
.fffb/IiIIJ1/II"
1 2 1 3 , 1962. The
Fig.
26.
... i l ,.<;
m
'"
...
<:
;;
8
;;
...
VIAL:
=J O
O
e
=
e
o
,
,
,
;
,
,
,
,
,
,
,
I
I
,
n
n
N
2
A,l.IQIWnH 3AI1.'3H
,
,
,
,
,
"
'
!(
'1
J,
,
,
,
,
\.
A,lIAI.1:>tt %
I
j
,
,
I
\..
,
,
,/ \
I
!
I
,
w O
o
U
3 n .llf 3d"3l.
mI (I
,
,
,
,
o
N
o
N
/'
',
$a/lO JO JaqwlIN
3 n H 3 d .. 3l.
o
N
107
:o
"'
d :!
o
o
,
n
,IOIWnH
3AI.L""'3
A l. I AID %
'"
Fig.
27.
Pi!:. 2fL
Average
'
distanccs moved in rel:tion to standard length
size
fOf males.
27 foc explanation
of symbols.
Pig.
29.
Fig. 30.
00
tht Cerro de
square
ciedes
00
111l'tel' for
tH:h
monlhly
represtot compulcd
capture data.
Open.
circ\t:s
denole
ohscrvalion.
dc-nsiti(:s
the
text).
00
>
i'l
!'!:
'"
basis
of re
ro
compulations
"P"
;;1
B!a(K
in
Roman numenls
populalions.
g:;:
equals
"
Vl.1L:
IN
COSTA ruCA
109
!
O
'"
N
>
Z
12
O
'0
'"
O
Z
z
O
::
<t
-'
'"
OC
0 ;0
. ",
"'
'"
>
O
"
S3AOW :10
"'
L>
Z
<t
>"'
Y3SnN
o
"
:;
N
6
;;
.::
;;
3nJ.:"'1:l3dW3.l0!8
fig. ; l.
composition of robl
:samples from
Fig.
32.
tl
'"
lO
:;:
tobes.
[h
occltrrence
nun
si%c
of one, twu
....
VIAL-
111
N
M
'"
'"
..
I
....
"
z
w
-'
el
a:
..
'"
'"
"
.c
o
"
.c
o
-'
.,
.c
o
-'
ro
"
"'
o
N
-,
-- ,
-o
-,
-.
-,
-,
-,
-o
-,
"
<
-,
-,
-.
-,
_o ,
-.
"
-,
-.
-,
-,
-,
-r '
,
S1 ",nOlfllONI
.0
-*
ro
-o
-o
"
- ,
S38wnN
el
'"
..
....
(f)
-'
'"
o
"
-r '
,
'"
"'
'"
'"
o
Z
10
'"
112
Fig. 33.
Fig. 34.
Fig. 35.
"
re
VIAL: EC010GY
OF BOUTOG10SSA SUBPALMATA
100
75
-'
..
....
O
.... 50
dF
u.
O
10
20
30
r,
,
IN
COSTA
113
I
I
I
I
I
I
J ,
/'H
,.,-9../
I
I
I
\
!'
GRAVIO OR BROOOING
I
I
FEMALES
33
)I
40
50
60
STANDARD LENGTH
"
RICA
70
60
"'
<t
:>
O
"
40
...
O 'O
"
W
'"
" 'O
:>
Z
,
,
,o
34
O
30
'O
60
S O I L MOSTURE
60
tOO
'O
U
.
W
"
:>
...
<t
"
w
o.
"
W
...
'O
...
Z
w
;;
"
"
35
O
'O
BOOY TEMPERA'TURE Oc
"
'O
""
>
Fig. }6.
(39. 4 2 ) .
The
salamanders are
.,
g;;:
VIAL:
M.olt
J
<
>
1
:: a ;' ! H;! : : H &a
' ' ' " ' ' 1 1 " ' ' ''
"
"
grod.
l O9r ClMll
AltfWt1I 8o/IIIMf11lur. it
' ,1 '
'
7"
"
J
<
;:
f,
'
;:
...
"
i
:1
J
<
.......,
...... .... " " "
"
"-
"
<
11
,
,
,
,
,
,
,
,
,
-
115