Changes in Tree Species Abundance in a Neotropical Forest: Impact of Climate Change
Author(s): Richard Condit, Stephen P. Hubbell and Robin B. Foster
Reviewed work(s):
Source: Journal of Tropical Ecology, Vol. 12, No. 2 (Mar., 1996), pp. 231-256
Published by: Cambridge University Press
Stable URL: http://www.jstor.org/stable/2560095 .
Accessed: 10/12/2012 15:29
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
 .
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
 .
Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to Journal of
Tropical Ecology.
http://www.jstor.org 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
Jou1rnal of Tropical Ecology (1996) 12:231-256. With 2 figures 
Copyright (C 1996 Cambridge University Press 
Changes in tree species abundance in a 
Neotropical forest: impact of climate change 
RICHARD CONDIT*, STEPHEN P. HUBBELL*t and ROBIN B. 
FOSTER* + 
* Smithsonian Tropical Research Institute, Unit 0948, APO AA 34002-0948, USA; or 
Apartado 2072, Balboa, Ancon, Panama' 
t Department of Ecology, Evolution and Behavior, Princeton University, Princeton, New 
Jersey 08544, USA 
+ Department of Botany, Field AMuseum ofNatural History, Chicago, Illinois 60605, USA 
ABSTRACT. The abundance of all tree and shrub species has been monitored for eight years in 
a 50 ha census plot in tropical moist forest in central Panama. Here we examine population trends 
of the 219 most numerous species in the plot, assessing the impact of a long-term drying trend. 
Population change was calculated as the mean rate of increase (or decrease) over eight years, 
considering either all stems :10 mm diameter at breast height (dbh) or just stems :100 mm dbh. 
For stems B 10 mm, 40% of the species had mean growth rates <1 % per year (either increasing 
or decreasing) and 12%l had changes B5/5% per year. For stems 100 mm, the figures were 38% 
anid 8%. 
Species that specialize on the slopes of the plot, a moist microhabitat relative to the plateau, 
suffered significanitly more declines in abundance than species that did not prefer slopes (stems 
10 mm dbh). This pattern was due entirely to species of small stature: 91 % of treelets and shrubs 
that were slope-specialists declined in abundance, but just 19% of non-slope treelets and shrubs 
declined. Among larger trees, slope and non-slope species fared equally. For stems I100 mm dbh, 
the slope effect vanished because there were few shrubs and treelets with stems I100 mm dbh. 
Another edaphic guild of species, those occurring preferentially in a small swamp in the centre of 
the plot, were no mor-e likely to decline in abundance than non-swamp species, regardless of growth 
form. Species that preferentially colonize canopy gaps in the plot were slightly more likely to 
decrease in abundance than non-colonizing species (only for stems B10 mm dbh, not B100 mm). 
Despite this overall trend, however, several colonizing species had the most rapidly increasing 
populations in the plot. 
The impact of a 25-year drying trend and an associated increase in the severity of the 4-month 
dry season is having an obvious impact on the BCI forest. At least 16 species of shrubs and treelets 
with affinities for moist microhabitats are headed for extinctioin in the plot. Presumably, these 
species invaded the forest during a wetter period prior to 1966. A severe drought of 1983 that 
caused ulnusually high tr-ee mortality contributed to this trend, anid may also have been responsible 
for sharp increases in abundance of a few gap-colonizers because it temporarily opened the forest 
canopy. The BCI forest is remarkably sensitive to a subtle climatic shift, yet we do not know 
whether this is typical for tropical folrests because no other lalge-scale censuses exist fol comparisonl. 
KEY WORDS: demography. population dynamics, tropical populations 
INTRODUCTION 
The myth that tropical climates provide a stable environment for tropical forest 
organisms has long been buried. We know that annual shifts in moisture avail- 
ability can stress vegetation and limit the distribution of many species 
231 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
232 RICHARD CONDIT ET AL. 
(Hartshorn 1992, Wright 1992). Now, thanks to long-term weather records kept 
by the Panama Canal Commission, we also know that the vegetation of Barro 
Colorado Island (BCI) must face supra-annual shifts in moisture availability 
(Windsor 1990, Windsor et al. 1990). Total precipitation at BCI underwent an 
abrupt decline around 1965, averaging 2740 mm prior to 1965 and 2430 mm 
since, paralleling a worldwide reduction in rainfall in the northern tropics 
(Bradley et al. 1987, Diaz et al. 1989). Moreover, associated with a strong El 
Nino event, the annual dry season of 1983 was unusually long and severe, 
causing elevated tree mortality (Condit et al. 1992b, Condit et al. 1995, Leigh 
et al. 1990). The 1983 drought and the long-term drop in rainfall are 
undoubtedly part of the same phenomenon, as the frequency of dry seasons (15 
December to 15 April) receiving less than 100 mm of rain increased from once 
every 6.2 years prior to 1965 to once every 3.5 years since (Windsor 1990). 
How does the composition of a tropical forest change when precipitation 
patterns shift? Unusual droughts and variation in rainfall are recognized more 
and more as important in tropical forests (Foster 1982a, Hartshorn 1992, Leigh 
et al. 1990, Woods 1989), but we know little about how populations of individual 
species change as a result. Certainly we know that past climate changes have 
led to shifts in species' distributions (Bush & Colinvaux 1990, Bush et al. 1990, 
Hamilton & Taylor 1991, Sukumar et al. 1993), and in temperate forests, 
detailed descriptions of range shifts that accompany past climate changes are 
so well documented (Davis 1981, Delcourt & Delcourt, 1987) that precise pre- 
dictions on the impact of future climate scenarios can be made (Botkin & Nisbet 
1992, Dale & Franklin 1989, Franklin et al. 1992, Overpeck et al. 1990, Pastor & 
Post 1988, Shugart & Smith 1992, Solomon 1986, Urban et al. 1993). The 
species-specific information behind these predictions is not available for most 
tropical forests. Only for the Luquillo forest in Puerto Rico has a species-specific 
model been used to predict the impact of climate change; O'Brien et al. (1992) 
assessed the potential impact of increasing hurricane frequency. 
To gather information on many individual species of tropical trees in one 
community, we established a large-scale and long-term population survey of 
forest at Barro Colorado Island in Panama. In order to include substantive 
information on populations of many species, a large plot was mapped: 50 ha of 
forest, with all stems above 10 mm in diameter included (Condit 1995, Hub- 
bell & Foster 1983). This dataset provides detailed information on change in 
forest composition and its relation to climate change. Similar large plots in 
natural forest are now being censused in India, Malaysia, Thailand, Sri Lanka, 
Puerto Rico, Ecuador, Cameroon and Zaire (Condit 1995, Manokaran et al. 
1992, Sukumar et al. 1992, Zimmerman et al. 1994), so we will soon be able to 
make a worldwide assessment on the lability of the species composition of trop- 
ical forests. 
Here we provide population estimates for 313 species of tropical trees found 
in the 50 ha plot in Panama between 1982 and 1990. We address specific hypo- 
theses about how the community is changing, in particular, how it might be 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
Climate variation and tropical forest change 233 
affected by long-term reduction in rainfall (Condit et al. 1992b, Hubbell & 
Foster 1990a, 1992). First, we consider species whose distributions within the 
plot are associated with moist microhabitats: a seasonal swamp and the moder- 
ately sloping terrain that drops off from the plateau in the centre of the plot 
(Becker et al. 1988, Hubbell & Foster 1983, 1986a). These areas remain wet 
throughout most dry seasons because a basalt cap below the plateau accumu- 
lates water during the wet season and drains slowly into the swamp and slopes 
throughout the dry season. Our prediction is that species associated with the 
moist microsites should be especially sensitive to the overall drying trend and 
will have suffered disproportionate losses in population. 
In addition, we consider population changes of species that preferentially 
colonize light gaps within the forest. There are two different predictions about 
colonizing species. First, Hubbell & Foster (1990a, 1992) suggested that the 
plot is undergoing a slow loss of weedy species, because the region just north 
of the plot (plus 2 ha within the plot) was cleared of forest about 90 years ago, 
and has since regrown. Colonizing species probably gained abundance within 
the old forest because of their large populations just outside, and are now 
declining. If this is the case, we should be able to detect disporportionate popu- 
lation declines among colonizing species. The second prediction on colonizers 
is just the opposite, and is based on the observation that the drought opened 
the forest canopy briefly during 1983 (Becker & Smith 1990). With more light 
reaching the ground, colonizing species should increase in abundance. Popula- 
tion changes of species preferring gaps can tell us which of the potentially 
opposing forces is more important. 
MATERIALS AND METHODS 
Study site 
Barro Colorado (BCI) is a 1500 ha island that was a hilltop until the Panama 
Canal was finished in 1914. The island is part of the Barro Colorado Nature 
Monument and has been operated as a research reserve since 1923. It is entirely 
forested, most in old-growth forest with no signs of human disturbance for over 
500 years: 48 ha of the 50 ha plot are in old-growth, with 2 ha in an area cleared 
until about 1900 as part of a French settlement. Temperatures are uniform 
year-round at BCI, but rainfall is seasonal, with almost none falling between 
mid-December and mid-April. Details on climate, flora and fauna can be found 
in Croat (1978) and Leigh et al. (1982). 
Census 
A 50 ha plot on the top of the island was fully censused in 1981-1983, 1985 
and 1990 (Condit et al. 1992a,b, 1993a,b, Hubbell & Foster 1983, 1986a,b, 
1987, 1990a,b, 1992); we refer to the first census, which lasted two years, as 
the 1982 census. All free-standing, woody stems ? 10 mm diameter at breast 
height (dbh) were identified, tagged and mapped. The diameter of each stem 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
234 RICHARD CONDIT ET AL. 
was measured at breast height (1.3 m) unless there were irregularities in the 
trunk there, in which case the measurement was taken at the nearest point 
downward where the stem was cylindrical. Dbhs of buttressed trees were taken 
above the buttresses. There were about 242,000 living stems in each census 
(Hubbell & Foster 1990a), and 305,875 stems over all three censuses; 28 have 
not been identified to species. A total of 313 species have been identified: 304, 
306 and 303 in successive censuses. (Three new species have been added since 
Condit et al. 1992b, all rare plants that had been misidentified as more common 
species.) Included in the list of 313 is a single tree that appeared to be a hybrid 
between Apeiba membranacea and A. tibourbou, and two distinct varieties of Swartzia 
simplex (Croat 1978). 
Analyses 
Species included. Abundances for all 313 species are reported. Species' names 
match those from Croat (1978) and D'Arcy (1987), except for species which 
were discovered, or whose names have been changed, since. An Appendix lists 
all cases where names do not match those found in Croat (1978) or D'Arcy 
(1987), and allows any species listed here to be located in those floras or in our 
previous publications on the 50 ha plot. Authorities for all species can be found 
via these references. 
Tests of hypotheses about changes in abundance included only those species 
that had at least 20 individuals 10 mm dbh in at least one of the censuses. 
We used this cutoff because large percentage changes in abundance of very rare 
species could be caused by minor, chance events. Four species of Bactris palms 
were also eliminated from analyses because we changed methods for counting 
individuals of these species. This left 219 species for analyses of all stems 
?10 mm dbh. Analyses were then repeated for changes in the number of indi- 
viduals >100 mm dbh, including the 136 species that had at least 20 stems in 
at least one census. We included an analysis with this larger cutoff because 
many other studies of tropical forest use the 100 mm limit (Phillips & Gentry 
1994, Phillips et al. 1994). 
Species characteristics. We analysed changes in abundance as they correlated with 
three species characteristics - growth form, moisture preference and tendency 
to recruit into light gaps. Species were divided into four growth forms - large 
trees (?20 m tall), mid-sized trees (10-20 m), treelets (4-10 m) and shrubs (1- 
4 m) - based on the maximum height attained at BCI (Hubbell & Foster 
1986a). Moisture regime was defined using the slopes in the 50 ha plot, which 
have higher soil moisture content during the dry season than the plateau above 
them (Becker et al. 1988), and the swamp, which is flooded throughout the wet 
season and remains moist in the dry season (Hubbell & Foster 1986a). 
Many species have distributions clearly demarcated by the slopes and the 
swamp (Hubbell & Foster 1986a), and we calculated the density of all species 
in the different habitats (unpublished data). We used the ratio of density on 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
Climate variation and tropical forest change 235 
the slopes (all 20 m X 20 m quadrats inclined >7Q) to density on the lower part 
of the plateau (quadrats with slope <7? and elevation <152 m, excluding the 
swamp) as an index of 'slope-specialization', and the ratio of density in the 
swamp (all 20 m X 20 m quadrats holding standing water through most of the 
wet season) to density on the lower plateau as an index for 'swamp- 
specialization'. We considered 'slope-specialists' and 'swamp-specialists' species 
with ratios > 1.5; this cutoff was chosen because chi-squared tests showed that 
nearly all higher ratios were significantly different from 1.0 (P<0.01), while 
most below did not (unpublished data). This index was preferable to a definition 
based on statistical significance, because the latter is sensitive to sample size. 
Finally, as a 'colonizing index' for each species, we used the fraction of recruits 
in light gaps given in Welden et al. (1991): Hubbell & Foster (1986b) used a 
similar but not identical 'index of heliophily'. Colonizers were defined as those 
species with an index ?30; again, this corresponds roughly with a statistically 
significant preference for recruiting in gaps (Welden et al. 1991) but does not 
depend on sample size. Most colonizers are probably 'pioneers' as defined by 
Swaine & Whitmore (1988), but they emphasized seed germination character- 
istics, which we do not consider here. Species for which information was lacking 
were omitted from all analyses requiring that information. The slope and swamp 
indices were calculated for all but 27 of the very rare species, but the colonizing 
index was available for only the 156 species listed in Welden et al. (1991). 
Statistical tests. In order to determine whether certain groups of species suffered 
disproportionate losses, the number of species that increased or decreased in 
abundance between 1982 and 1990 was tallied as a function of the four categor- 
ical variables. For statistical tests, a standard ANOVA was not possible because 
the design was unbalanced, with many empty cells. Instead, chi-squared tests 
were used on each of the variables: for example, a 2 x 2 contingency table for 
slope-specialization category and for population change provided a chi-squared 
statistic with one degree of freedom. To determine effects of each variable separ- 
ately, we proceeded as follows. Swamp effect was assessed by contingency tables 
for swamp and non-swamp species; since swamp status was not associated with 
colonizing nor slope status, the latter two categories were simply ignored when 
swamp effect was tested. But slope and colonizing variables were associated - 
there were fewer slope-colonizer species than expected by chance - so we segreg- 
ated species simultaneously by both categories. All tests were carried out on 
the four growth forms separately. 
For each species, we calculated the annualized rate of population change (r) 
using a standard model of exponential population growth: 
In Nt -In No r = 
t 
where Nt and No are population sizes at time t and time 0 and In means the 
natural logarithm. The time interval, t, for each species was defined as the 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
236 RICHARD CONDIT ET AL. 
arithmetic mean time elapsed between censuses for individuals of that species 
(based on the census data of each 20 m x 20 m quadrat in the plot). 
Earlier publications 
Hubbell & Foster (1990a, 1992) described population changes based on the 
1982-1985 interval, and Condit et al. (1992b) updated this with 1990 data, but 
this is the first presentation on abundance for all 313 species. Discrepancies 
between the numbers reported here and those from earlier reports are slight 
and are due solely to corrections of old errors. Since this is an on-going process, 
future reports might give figures slightly different from those reported here. 
Access to data 
We hope that the table of abundances for 313 species provided here will be 
useful for many future studies, and we will provide computer versions of the 
table to anyone interested. Please send us a diskette and indicate preferred 
formats. 
RESULTS 
Changes in abundance for stems 10 mm dbh 
Of the 219 more common species considered here, 105 had increases in stem 
number between 1982 and 1990, 108 had decreases and six did not change. For 
all 313 species in the plot, 136 increased, 154 decreased and 23 did not change. 
As previously noted (Condit et al. 1992b, Hubbell & Foster 1990a, 1992), rare 
species - in this case the 94 having fewer than 20 stems in all censuses - 
suffered proportionally more declines than common species. Table 1 gives the 
abundance in all three censuses for all 313 species. 
Many populations did not change by much (Figure 1). Of 219 species, 88 
(40%) had population changes <1% per year between 1982 and 1990. But 
some species had dramatic changes in abundance: 27 species (12%) changed 
at rates more than 5% per year, 18 declining and nine increasing (Table 2). 
Some common species underwent substantial declines. Poulsenia armata fell from 
3430 to 2126 stems, and Acalypha diversifolia from 1568 to 827. The most rapid 
rate of decline was Piper aequale, which had 219 stems in 1982 and 83 in 1990 
(Table 2). On the other hand, the population of Palicourea guianensis rose from 
377 to 1475 stems, while the much less common Psychotria gracifora had the 
biggest rate of increase, from 10 to 44 stems over eight years (Table 2). The 
mean rate of change for the 219 species was -0.29% per year, while the mean 
rate of absolute change was 2.25% (the mean of the absolute values of rates of 
change). 
Changes in abundance for stems ?100 mm dbh 
The range of population change among larger stems was no different han 
for smaller (Figure 1B). Of the 136 species considered, 66 increased in abund- 
ance from 1982 to 1990, 61 decreased and seven stayed the same. Fifty-one 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
Climate variation and tropical forest change 237 
A) 30 
E colonizer 
25 _ slope 
c slope-colonizer 
rt other 
(1) 20- 
1 5 Ssoecooie 
0~ 
6 
10- 
5 
-0.12 -0.09 -0.06 -0.03 0.00 0.03 0.06 0.09 0.12 
rate of change 
B) 20- 
ED colonizer 
U slope 
15 ' slope-colonizer 
(no [i other 
10- 
0 
6 
5- 
-0.12 -0.09 -0.06 -0.03 -0.00 0.03 0.06' 0.09 
rate of change 
Figur-e 1. Distribution of population growth r-ates. The 'colonizer' category inicludes all colonizing species 
that were not slope-specialists, and the 'slope' category includes slope species that were non-colonizing; 
'slope-colonizers' are the four specializing in both areas. 'Other' inicludes all species that were neither slope 
nor colonizing plus all species that wer-e missing information on one or bothi categor-ies. (A) Rate of change 
of populations of stemns 310 mm dbh, including 219 species (see text). (B) Rate of change of populations of 
stems 3100 mm dbh, including 136 species (see text). 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
4- - , C 3
.t * o^ C) C) C) "Zi Cf C) ND C, -,I cn C) C) L(- 00 C snNssss o C) C) ?) C) C) O ) "i C) C, C1 (Z . . . . . . - : ct CD CD CD C DC ) 
= = D Q O~~~~Cs 
m 
m 
t 0 mn 
o ,,. , 
4 X.YO 
Qt c :t . -0 
, b. CD C1 cn -4 Q0 "i t- 54 cn t- C ZL qC n CD CD C14C 00 
0 
CD CD CD CD CD 00) Ln CD 
u c CD C1 c n C1 cn c + Ln cn O O 
Ct ,^ t z, A 
v -C ct o n 00 I, C, I, 0 In n CD o CD CD 00 Q0 cn C -oo CD CD C n Ln >oLn CD oCo 
bN -,C! X-- V:Z co L(-) C?) C) C14 Lc) - Cq) LCq L) C4 C>l 
-. _ ._ 
bo M 
C , t u 
_ = = ~ t S S Zco c n C 4 c e 
-C! r. 4 '4<%X4S>> < 
H~~~~~~~~C co mmmmmmmmmmmm t 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
"Zil : C) C) C) -11, C14 -Izt 00 C) 00 1 C:) 1 - Q0 Lr) cn cn C) U- -,Zt : : - C,4 Cf cn C) C) C) C) - C) Lr) Cq - C:) CD D CD - cn -'t 00 -,I, -,Zt - (=) - -,I, - C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) CD CD CD CD CD CD CD CD 6 6 6 C) 6 6 6 6 6 6 6 6 CD CD CD CD 6- CD CD CD CD CD CD CD 
Cn - Cn Cn (:o (:O - - Cq C4 t- C4 Cq CD C4 Cn -,Zt C) C) C) C,4 - C) C) C) Cn t- C) Ln oo Cq Lr) r- C14 (.0 C14 cn cn co C14 C'4 (.0 co Cq Cq 
C) C) cn C) C) - C,4 (.0 cn C) t- Cq CY) -,t Cq C) 00 Cq CD CD - C) L(-) "Zil cn C) C) C) C) - C) - C) C) C) 00 Cq CD CD u- Ln C14 -11, C14 -11, cn co C14 C14 Lr) -1t m - Cq Cq -1t 
C) C) CO C) cn - L(-) C) -,Zt C,4 C) C,4 t- C) 00 c"I Cq Cq cn CD C) - C) C,4 L(-) C) C) C) C) C,4 Co C,4 C) CD U-) CD CD CD CD oo Lr) C14 C14 co cn Cq Lr) Cq "IZII cn - Cq Cq -Izt 
C) t- C,4 t- C) t- C) cn C) C,4 C) Lr) C) 00 00 00 CY) cn U- U- C) CO (.0 C) (.0 -,Zt C,4 C,4 U- oo D CD Cq CD C) cn C) - U- C) C) cn C) C) C) cn - C,4 C) Cq Cq D - Cq cn U- 00 -'t 00 C) L(-) L(-) C) C,4 Lr) U- cn U- C) -,t CD cn CD U 00 CD CD C) C) C) C) C) C) C) C) C) C) C) C) CD CD CD CD CD CD CD CD C) C) C) C) C) - C,4 (=) (=) (=) (=) CD CD CD CD CD CD CD CD 6 6 6 6 6 C) C) 6 C) 6 6 C) CD CD CD CD C5 CD 6 CD CD CD CD CD CD CD CD CD CD CD C5 CD 6 C5 CD CD 
CD U-) - cn CD CD - Q0 U-) C,4 U-) 00 o) 00 Ln t- D C) cn cn C) t- o) CD U-) cn - cn o) o) C,4 00 cn - C'4 C'4 C14 1-n C14 C) Ln 00 Cq - -,I, C,4 CD -,I, Q0 -11, U-) - 00 C,4 00 CD co t- CD -'t 00 D - - Cq u- C) oo U-) "Zil C14 o) -11, - "Zil Q0 C,4 CD D C) CD Cq cn - - - - 
Cq CD Cq t- - C,4 Q0 Q0 cn 11 C,4 CD - t- 00 C,4 t- Q0 U-) t- Cq CD Cq 00 t- CD -,Zt Q0 C,4 cn 00 - - 00 00 U-) cn - Q0 C) C,4 t- t- Cq - cn U-) C,4 cn (.0 cn C14 "Zil cn t- 00 C) -It 00 Lr) D D cn (.0 co u- (.0 -11, C14 co -11, - -It Cq Cq cn 
(.0 C4 Ln - C) (.0 C14 Co U-) C14 t- - Ln Cq Cq t- C) Cq Lr) C,4 cn (=) CO cn CO - Lr) C'4 Lr) 00 Cli - u- D -'t cn (.0 C,4 -,I, (.0 C,4 - (.0 u- Co C) cn C) D cli u- 00 -It - C) C) cn co (.0 Lr) "Zil C14 u- Lr) cn cn t- D Lr) Cq cn 
(1) (1) ct ct ct ct ct (L) (L) (L) (L) (id ct 
U U U U U ct ct t t t t U U (L) (L) (L) (L) ct 
tttuttutt V(L) 4-- 4-1 4-- 't ct ct ct U Ct 
U U U U U U (L) (L) (L) ct (L) (L) (L) (L) (L) (L) ct (L) U U ct ct ct U (L) (L) (L) U (L) ct ct U U U (L) (1) ct U .- ct ct ct Ct C) Ct It It (R) cut ct ct ct ct ct U 4-1 4-1 4-- 4-1 41 ct ct ct (L) (L) (L) U U U U U U 4-1 4-1 4-1 0 0 4-1 4- U U C) t6ttu _"tZZZZ (.)(.) -- t t bi) tf) tf) It ct ct ct ia C) C) C) C) C) ct ct ct ct ct ct ct ti) bi) E s 0_, (.) (.) (.) (.) C) C) ct ct ct 
ct 
Ct Ct Ct U Z C's 't :z C) C) 't ct ct ct ct C) C) U ct (L) - - , V U ct 0 0 - C) C) C) 't (3 Vc)) (/,t) (2) 
IQ 
10 tz -- (Z) rz 
rz - -Q, :,) -Z -Z 
-G (Z' " 
tz (Z' Io 10 110 
--G 
lZq Qq Qq lZq Iq Iq 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
C) Lr) C"l Lr) C) (.0 (.,::) C"l cn - L(-) C) - cn u- CD CD -'t CD C) cn C) C) C) C) Lr) U-) C) c) Cj CJ 
-'t Ln CD cn CD C) Ln C) cn C"l cn C) C:) - C:) Lr) C'I cn ce) CD CD CD CD CD CD CD CD CD C) C-4 CD CD 6 C) CD CD CD CD CD CD CD CD CD CD CD 00 
CD CD Q0 CD - C14 -zll C14 C14 CD D CD Cn D Cn C) C',j CD t- Cn cn CO C) Ln C) u -) C) C) C) - 00 Cn D C4 Cn 00 - L(-) Ln Lr) cn Co C14 C14 C) 
C) 
c) (.C) cn -n C'I cn C,4 Q0 C) CD CY) Ln 00 CD Cq C) t- cn (.0 oo C) (.0 - -zl, Lr) (.C) oo cn m Cq Cj -,Zt c"I u- c"I -,I, -,I, C) cn C"l C14 C) CD CY) C) 
M 
C,4 CO U-) CO C) (.0 U- C) M (.0 CII Cll t- CO C) 00 - 00 C 00 C) CO M C) C) c"I (.0 m C) C) C) V oo M C) C14 ",I, t- C14 ",I, cn c"I u- C14 C"l C) C) Cj 
L<) C4 M M Ln (.0 Lf-) Lf-) D Cn D C) C) U- cn C) C) C) CO Ln (Z') cn V D Cq D - Cn (=) (=) (=) L(-) (=) - - (=) C) C4 -11 Cn Lr) CD - C,4 C) C) C,4 4- C) C) C) C) C) C) C) CD CD CD CD CD CD CD CD C) C) C) C) C) C) C) C) C) C) C) C) CD CD CD CD CD CD Cq CD Cq CD CD 6 6 6 6 6 6 6 6 6 6 6 6 CD CD CD CD CD CD 6 C) 6 6 6 6 6 6 6 6 6 6 CD CD CD 00 
o') 
CD - (.0 (.0 cn (.0 u- CD cn o) cn U-) Cq C) C) D Cl - t- C) CO Lf-) (.0 C) "ZI, "ZI, m C) Ln (:7) D r- Ln t- C14 - CO C) M t- M C) C14 Co C) C) CO M C) cn C"l C,4 (.0 -11, 
C"l 
7 
P-i 
Ln 00 C) -,Jl 00 11 - "Zil C,4 (.0 - c"I (.0 Co Lf-) C) u- C) CO -,Zt C) 00 CD CD 00 Ln D m C,4 co CO co -11, - -,Zt -,Zt Lf-) Ln D CJ - m C) CD -zll Ln c"I C) L(-) 
c"I CO CO C) C) C) C) C"l c"I Or - 00 Lr) CJ - D C"l C"l C"l C,4 Lr) 
M C"l 
14 -Izt Lf-) CIII Lf-) CO C) Ln 00 D m C14 m C) u- "Zil m C) CO C) C) (.0 C) c"I C) C) QD 00 D -,t Cq - C) C) co L(-) L(-) 
(.0 
- Lr) cn -,Zt Lr) t- Cq 00 Cj C) CD D CD 00 (.0 - Lr) Lr) C) cn C) C) Cf m 00 C) Ln C) -'t D D C14 cn cn 
ct ct 
(L) ct ct (1) u u 
u 
ct U ct ct ct ct ct ct ct Ct ct 
ct V (1) (1) Ct (1) C) - C) C) >1 >1 U U U V ct ct (-) ct ct U ct (L) V 
U U U U U U U U U U U U V V U bf) 
4- 4- 4- 4- Ct CS Ct Ct Ct Ct Ct Ct Ct CS Ct ct i4 i4 i4 i4 i4 i4 i4 i4 -ct C-t bi) bi) bf) o o o o o o o o o o o ct t o v z ct v v , z z C-t >, -4 W 4 4 WI' "WI 
r. . . . . . . . . . . . . 
rz rz 
tz - ,, g 
- .- 9 'ZS 
'4 (Z) 
Zs cn rz tz tz -q4, -q4, g rz 
(z 10 u tz tzi ri In In In 11) 11) 11) 11) 11) 11) 
4z), ;Z ;Z ;: ;Z zs ;Z ;Z ;Z ;Z ;Z 
10 'ZO lo ct 
t Z4 --q 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
C) - (.0 cn - c"I c"I 00 (.0 CY) c"I - C) C) L(-) U 00 CD - Cj cn - 00 U-) 00 C) cn 00 L(-) 00 u- 00 "t - u- : u- - C) C) CN CD CN CD CD Lr) CD cn cn c"I C) c"I C) Q0 c"I cn c"I - 00 Q0 cn C) D CD CN D CN - CN - CD (.,::) (.,::) (.,::) (.,::) (.,::) (.,::) (.,::) (.,::) (.,::) (.,::) (.,::) CD CD CD CD (.,::) (.,::) (.,::) (.,::) (.,::) (.,::) (.,::) CD CD CD CD CD CD CD 6 6 CD CD CD CD CD CD CD CD CD CD CD CD CD CD 
QO Cn t- -It (Z:) Ln (:7) L(-) CD D D CD Clj CD .C (.0 C"I C) cn t- CO C) L(-) -,t c"I C) C) U- CN Ln Ln CD Ln CD r- CO u- cn CO L(-) C"l -It 1-- - C) "t c"I -zll Lr) C) -t cn j cn C"l (.0 C"l C"l QC) - C"l 
C) cn U-) C) C) cn C"i CD CD cn D D D C) Co -,t t- cn c"I C) C- -,t cn C) Ln r- CD D r- CN u- CD C) C) C) "t C) (.0 C) (.0 C) oo cn 00 cn U-) U-) Ln C) cn cn cn - cn C"l C"l C"l Lr) Cl 
C) C) U-) U-) CO CO u- - C) - C"l U-) (.0 C) 00 D CD cn -t C) 00 CN - -,t C) Ln CO C) u- c"I -,t C) cn D C) Ln -zll C) C) C) r- CO c"I u- cn -It (.0 -It - C) cn (.0 00 C"i cn - D - CN CN cn - cn CN (.0 C"l C"l U-) - C"l 
00 (.0 -'t u- r- D cn cn Cj L(-) D C'll (=) - C'll C'll t- C'll Co Cn Cn 00 (=) CD D C) CD 00 U- (.0 Ln C) C) C) C) C) C- cn C) - CD CD Cj CD CD CD -t CD C"l C) c"I C) - - C) cn c"I - - - c"I - C) c"I C) CO C"l C) CD CD CD CD CD CD C) C) C) C) C) C) C) C) C) C) C) ( CD CD CD CD CD CD CD CD C) C) C) C) C) C) C) C) C) C) C) C) CD CD CD CD CD CD CD CD 6 6 6 6 6 6 6 6 6 6 6 6 CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD I I I I I I I I I I I I I I I I I I I I 
Clf 00 00 C14 C'-j O') 00 1- C14 CD Q0 00 00 00 U- Ln U- - C) CD C"I (=) C"I - - - t (=) - C"I C- Ln (=) CO C"l "t CO C) CN D Cn Ln D u- u- U-) (.0 C"l C) (.0 C) CO "t "Zil CN - C) C) cn r- D -zll - 00 C"l cn (.0 c"I c"I -,t CY) - CY) CY) cn -11 cn - CY) cn -t cn 00 - 00 - CY) Lr) (=) L(-) cn C"l t- cn C) C"l cn 
Ln CD 
-11 
C) CN C) C) c"I - c"I Cli C) C) u- "t C) "t C) C) C) CD 00 CD CN - CN CO C"l C) "t -,t C"l cn C) t- cn Ln C"j C) C"j Ln 00 C"j C) - C) (.0 cn (.0 L(-) C) L(-) - (.0 (.0 - - C) - -t t- Ln CJ C) - D C) C) Co cn Co C) - C) Ln QC) c"I c"I oo cli u- cn C) cn 
CO t- U-) C'I - "t C) D t- Ln 00 CN C) "t t- c"I t- C) CO CO cn C'I CO r- Ln C) - Cj Cj Cj Ln "t CO C) C) cn U-) CO U-) CO Ln 00 00 t- -zll C) 00 CJ Ln C"l t- C) C) - L(-) C) - - (=) - Cn Co (:7) C',l t J - Ln C) Clj - Lr) u- Ln U-) cn CO 00 Ln 00 C4 -It (:7) C"l 1-- cn C) cn C"l -11 C) 
cn 
ct ct ct ct v v ct u v u u v (L) (L) u u ct ctctct Uctct (L) (L) utuu ct (-) ct ct t t t t t t t t t t t t t ct (L) U ct ct U U V U U Ct ct ct v (L) V v u V U (L) C's ct (-) ct ct ct ct C's ct U Ct ct Ct 't ct ct It ct ct U Ct Ct C) C) C) C) C) C) C) C) C) C) C) C) C) C) U Ct Ct V V u u v v ct Ct v 't It It (L) q) Z:; ct u u u U U U U C) C) C) u 0 0 0 Ct ct ct t t U) U) Ct 4- 0 U) U) U) > > C) u u u It U C) Ct C) U U bi) 1) bl) bl) c-t (-) ct ct bO bo bo bi) bl) bl) bl) bl) bi) bi) bi)C'S cn ct 
,I" 
zs -tz z 
ZS 
-q-4, 
q4, 
zs 
(Z' (Z) 
Io Io 
10 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
C) cn C) 00 -It cn C) C) C) 00 LC) CD 00 CJ - c"I c"I Lr) (.C) c) c"I c"I C) CD Cj Cj CD - CD C) CN CD - 00 - C) cn . . . 00 C-4 (.C) (=) -,t 4-1 C) CD CD CD CD CD C) C) C) C) C) C) C) C) C) C) (Z) (Z) CD CD ct C-4 CD CD CD 6 6 6 6 6 00 o') 
C) C) C) C) Ln QO C) D C"j D CD CD CD '11 t- cn cn (.C) (,o J Cn C\j u- CN Ln C\f C\f t- Ln Cn 
Ln CO t- C) C\f cn C) Ln Ln C) C) m C\f (Z) cn "I, CD CD CD C'I M (Z) (Z) C\f in CO Qo C\j co cr) C) C C) C\f cr) C) m C) 
A\ 
r-N Ln U- C) cn t- C) C) C) cn C) CD "t Ln Cli u- CD CD C) C) C\f M cn C\f C) C) cr) "t cn 0) u- C'I C) oo 00 co C\j C\f C\f C\j - C\f cr) (.0 Ln Ln 
cr) C) C) Ln C\f (.0 "t C) 00 CD C'I Ln Ln CD cr) - cr) C\f m C\f C) C'I 00 - cn - C) "t CD CD C'I o - C\f cr) "'t - C) C) C\f "t (=) Ln CD CD CD CD CD CD C) C) C) C) C) C) C) C) C) C) C) C) CD CD CD CD CD CD CD CD (z) (z) (z) (z) (z) P4 C'I CD CD CD CD CD CD CD CD CD CD (Z) (z) (z) (z) (z) (z) CD CD 00 m 
m m (Z) t- Ln (Z) 110 C\f (Z) cn cli 00 0) C'I (Z) u- 00 CD C\f 00 (o cn M - (,o (Z) 00 t- C\j (Z) C'I 0) Cj M 00 00 cli C) u- (.0 C) C\f co cr) C) cr) cr) C\f - L(-) C\j CO o cn 0) cli o C) t- m co C\f cr) Ln m cr) (.0 cn cn - C'I u- oo C'I 
Ln (.0 Ln C) C\f C) "t Ln C\f (.0 C\f CO - M Ln "t (z) C\f M "t M cr) m m t- "zi, cr) C'I o oo I'll - cn CD cli (.0 -:fl C\f co (7) t- cr) C\f m - "- C\j cn cn 00 ,I, - u- co cr) C<-) (.0 C\f u- cr) cr) - C\f M t- cr) 
A\ 
cli M C'I CO CO C\f (.0 (.0 "t CO C\f C\f (.0 M (z) CO (Z) 00 M u- CD m C\j Ln (,o Ln m C\j 4-1 co I'll - CO M V- Lr) cli 00 u- cr) cr) cr) Ln "t "'t C\f "t - CO "t - (.0 cn 0) C) CO cr) Ln cr) m cr) - cr) C\f C\f t- 
u u u u u u u u 
t t t t t t t t v 
4--l 4--l 4-1 4-1 +-I +-I +-I +-I ct ct u u Ct U u ct ct u v 0 ct V ct .Ct v v t V 
"" v v v v v ct u v v v v Ct U u t t t t t t t ct ct u 0 0 0 0 0 0 u v v v v >1 u u biD 4- 4- 4- 4- 4- 4-1 4-1 4-1 U U U U U U bjD t U U U U 't ct U 0 (4-, 4-1 (n (n (n (n (n (n cn cn ct ct ct ct 
0 
ct C) u i. ct ct ct ct ct ct ct t ct -I--, u 
i- 
biD 't u ct - ;z ;z bP :Z ;Z ;Z ;Z ;Z U ;z ;z ct 0 ;z ;z ct ct -0 0 ;z t t t t 
C5 O 
tz In cr) 
, L-Z  . Z 
-r,- Z In Z2, 
u 
- - - - - - 
v 
ct 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
C) C) C) C) t- (.0 - : : : : : : : C) C) CN C;) CI) CN C) o 0) CD C) Ln l Ln Ln CN Ln C) : : : : : : : C) C) C) C) "t C) . . . . . . . . C) C) - CN OC) CN C) - ce) C) CN - C,,[ CD - CD cn CD . . . . . . . CD CD CD CD C) C) C) C) C) C) C) C) C) C) C) CD CD CD CD CD CD CD CD CD C) C) C) 6 6 6 6 6 CD CD CD CD CD CD CD CD CD CD CD 
CD CD CN M M C\f CD CD CD CD CD CD CD M - CD "t Ln "t Ln C\f C\f M I'D Ln t- CD C'I 00 "t CD CD CD CD CD CD CD CD M CN cn CD Ln I'D 
00 - ce) - ce) 
cn CD CD CN cn 00 M CD CD CD CD CD CD CD M - CD M M Ln t- CN CN C) - M - C) cn M -,-), C) (Z) (Z) (Z) (Z) (Z) (Z) cn L(-) - t- Cn (=) (=) L(-) 
co - ce) - ce) 
"t C) C) CN ce) CO CO C) C) C) - C) C) CD 0) - CD 00 C'I M ce) - CN u- ce) C;) C) Ln C) "t "t C) C) CD CD CD CD CD cn u- Ln C\j - ce) t- Qo Ln m C\f ce) 
CO C) "t Ln (.0 CN - C) cn M - M M (.0 CN CN (Z) (Z) "t t- cn C;) C) "t C;) C) t- in cn (.0 CN CD L(- Ln u- ce) C) - (.0 CN C) C) C) (.0 CN (.0 M C\f Ln Cli C) CD CD CD oo CD C) Ln C) C) C) C) C) (=) - - C-1 CN CO "t C) 00 o CD CD CD CD - C) C) C) C) C) - C) C) - C) C) C) C) CD CD CD CD CD CD CD CD CD C) C) C) C) C) C) C) C) C) C) C) - CD CD CD CD CD CD CD CD CD 6 6 6 6 6 6 6 6 6 6 6 C) CD CD CD CD CD CD CD C=; CD CD CD CD CD CD CD CD CD CD CD CD CD I I I I I I I I I I I I I I I I I I 
C'I I'D Ln ce) ce) t- "t ce) CD -, t- ce) ce) 00 0) - C'I o 0) - cn t- I'D cn cn I'D C\f C\f CD M cn - 00 u- I'D Ln CD C'I (.0 L(-) t- CO t- CO (.0 u- Ln (.0 -t o Ln cli u- cli o "It u- CN CO - CN cl-I C\j "t co C) CD 
CN - cn ce) 
CN (.0 C) cn cn co cn t- 00 cn C) cn Ln cn u- M - CN CO t- - (.0 Cn Cn CO CO Ln Ln Cli Ln Ln "t (.0 - (.0 CO t- Ln CO (.0 (D "t Ln cli 00 00 C) C) - -,-), CN (.0 L(-) CN CN (.0 co co C;) cn CN CN 
CN --I, CN t- Co Co Co M M t- CN M C) M C) (Z) - CN Q0 CN Ln t- Cn CN (.0 Cn Ln CN M CO (Z) Cn 00 Ln L(-) CN t- (.0 (.0 CO (.0 - C) Cli CO Lr) r- cli u- cn L(-) D Ln - - (.0 CN ce) CO ce) Ln Cli - CN CN -t Q0 ce) C;) u- Q0 ce) ce) CN CN 
ct 
V ct ct ctctc u 
o o o 
0 0 0 0 u u u t t t 0 u V V V V V V V ct t-0 
t t t t t t t V - - r- r- t t t V V V - U U U U t t t t t t t t U V ;z V V V V V V V u u V V V V V V V u u u u u u u ul-l- u u U-E ct ct ct ct u V u u u u u t t U u u t U u u u u u u u ct -0 -0 .5 Ct Ct -0 Ct ct ct ct . ct . ct . ct Ct 0 0 0 4-1 - ct bjD bi) bjD -0 -0 :Z bi) bi) bi) bi) -0 -0 -0 -0 -0 -0 -0 bP >1 ;z := ;z ;z ;z :z ;z 0 C) 
;Z 
;Z 
C75 C/) C/) C/) 
:Z 10 
Q5Z5 z 
141 
12 
<Z5 
IZT, -Z:- - ;Z 
-!ZZ -Z -Z 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
CD CD CD OC) C') o C o (.0 OC) 
. . . . . . CD C,,[ CD OC) - (=) ce) (=) - ce) Ln CN ct I (= CD CD CD CD (Z) (Z) (Z) (Z) (Z) - (= (= C'I (Z) CD CD CD CD CD CD CD CD CD CD CD CD CD CD c (z) (z) (z) (z) (z) (z) 
00 
CD CN CD CD CD CD CD CD 00 m 00 CD CD o cn o CD - C'I 0) Ln 00 CD Q0 CD CD ce) CD CD CD t- CD u- CD cf oo (:;) - ce) (.0 CO ce) Ce) Ce) L(-) C-1 
(.0 C"f C"f cn cli 
CN CN Ce) 1- CO Cn CD Cli Lr) Ln D Ln CO Ce) (.0 Cn 00 CD C) Cn cli U-) - CN CO Ln CN CN Ln CN (=) C\f cli cn cli 
C"I CN 
(.0 Ce) U- CO Cn Cn D -tl Ln --I' t- CO "t U-) U-) - CO 00 cn cli l CO t- CN Ln CN C) CN CN cr) 
QO U-) C\j U-) CO o Ln 00 Ln Ln CO Ln CO U- (.0 U-) (.0 Ln CN Cli Cli C') 00 Cli CD CD CD CN CN (.0 - C') C-1 U-) Cli Ln C) CD CD 4-1 
ct 
CN 
00 
m 
CD CD m "o C,-) Ln 00 C,-) o CD ce) o CD o o C'I Cl-I I'D Ln CD t- CN C"f Ln C,,[ CD Ln oo C'I 0) CD "t c7l) (:7,j C') - U- CO (D C) (.0 U- (:;) (.0 t- C-) U-) CN Ln (.0 Ln 
C\f C\f 
C14 Co (:7-, Ln U-) C') 00 CD --tl Ln cj oo j U- Lr) U-) C') Ln C') Ln C) C') co U- o o U-) CC) Cli o C'-) CN C') Ln ce) 
Co Ln C-1 -tl M U- Ln CO -711 Ln Cli CD U-) CN C') C-1 CO Cn Ln co o 00 u- (:;) Ln C-4 u- (=) "t cn C') CN (=) (.0 (:;) (.0 U-) C - - Ln CN CN (.0 co cn cl-I C\f CN ce) CN 
ct ct ct ct ct 
U U ctct(tu u u U U (-) U U U 
'-" (a.) Ct Ct Ct U 0 0 U U U U U 0 ct ct ct ct ct ct r- r- ct -0 ct ct ct ct ct ct U U U U U U ct 0 U U C-) Ct ct ct ct ct U ct ct >, -;z 7; (t ct t ct ct ct (t (t U 0 0 ct (Ut 5 5 5 5 t Ct ct ct - ct 3 
t C) :FD '- 0 t 0 bi) b'f) -0 ct b.0 bi) 0 0 ct ;z ;z ;z ;z ;z ;z U U 0 0 0 0 S. 
p t4 -V4 
ct 
tz 
:,z S 
-2:1 -Izz -z 
U 
6 
Q -- -- - 
<Z, 
ct 
Z, 
cn 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
OC) u- CN OC) C) t- CN CN C 00 o C\j (:D : C) C) C) Ln - -,I - : (=) (,o U- ,J (=) 00 CD C-) (=) U- C\j CD - - r- - Ln . CD C') C'I cn C) - C) C) C) C) C) C) C) C) C) CD CD CD CD CD CD CD CD CD C) C) C) C) C) C) CD CD CD CD CD CD CD C c; c; CD CD CD CD CD 
ce) CD 00 ce) 00 CD Ln 00 "o ce) "o ce) - 'n 00 Ln C'I CD cn C'I Ln 00 t- CD t- ce) 00 C\f CD CD oo CD - CD o Ln - cli cn cli U- CD oo cli Ln - - oo 00 ce) Ln o CD Cli Ln o cn cn cn u- - QD 
CN CO CO Ln CN CN C) C CO C'i C'n U- - Lr) U- 00 Cn Ln CN CO (,o CO CO (=) ce) Cn CD _ 00 CD 00 - ce) CN t- C) CO C-4 -i-I CN C) CN "t Ln "t 00 cli o C) CN rn ce) C - - Ln 
(.0 C) - cn c C) U- 00 cn CD 00 00 cli CN -t (Z) cn (.0 CN (.0 Ln (.0 U- (Z) cn - cn ce) (,.D - ce) C) "t ce) CO C CO cl-I U-) Cli u- C\f cli Ln o cn C) CN U- C) CO CN CN ce) C) Q0 
CN 
(.0 CO 7t' C) (.0 - CO Ln U- (=) (.0 (:D U-) (.0 - CO "t Cn -7i D 7i C) C'i CN U- "t CO - ce) C) C) C) CD CD CD CD CD CD CD CD C) CN ce) C) ce) Ln C) C) Ln - C) o C) Cj CD C C C (=) ce) C\j (=) (=) C) C) C) C) C) C) (Z) CD CD CD CD CD CD CD CD C) C) C) C) C) C) - C) C) C) C) CD CD CD CD CD CD C) C) C) C) C) C) 6 6 6 6 6 6 CD CD CD CD CD CD CD CD 6 6 6 6 6 6 6 6 6 6 6 6 CD CD CD CD CD CD CD CD CD CD CD CD CD I I I I I I I I I I I I I I I I I I I I 
LI-) Ce) CO - U-) CN Ce) (:7'j CO - Ln - U- U- Cli - 00 - 00 Cli C) Ln Ln U- U- (,o Cn (:D C) Cn C) C "t C) C) U- CD Lr) - CN (:D (=) t- Ln (:D CO CN C CO CN C C - CN CO 00 - 00 Ln Cn 00 CN ce) CN CO ce) -1- C) oo o C) Ln cli CC) ce) CD CN C) CN - cn "t ce) CN - 
00 C) 00 (Z) C) 00 (Z) (Z) - C) C) U- 00 (.0 ce) C -tl (.0 "'t Cli cr) cr) Lr) Ln (,o ce) C) U- CD C (.0 - C CN 00 CN t- t- C) CN 00 - cn (.0 C C C14 C) C cl-I t- (.0 "t CN - OC) t- Ln Ln u- CN - o Ln ce) - OC) cn cc) C OC) (.0 (.0 Ln - ce) ce) - CN CN CN C'I C'I Cli ce) cn cli 
OC) (.0 (.0 t- Ce) Ce) r-n CN CN - Ce) C-f V- D Cli Cli Cli V- V- CN ce) (,o CO (=) Ln ce) Lr,) t- - ce) C) t- t- (Z) CD oo u- - CO CN - CN (.0 C U-) CN Cn (.0 CN - Ln V- Ln CD CC) - 00 CN C C) u- - ce) CN C CO (.0 CN Ln C) cn cn ce) 00 - CN CN CN CN CN ce) CN 
ct ct ct U U ct ct ct ct ct ct ct ct t t t U U U U ct ct ct ct U U U U U U U U U v U cn U U ct U v U ct U U U U U U 0 t t U U ct U U U U U U U U U U ct -1 4-1 U 4--l 4-1 Ct U U U U U (t U 
U 
U 
ct 
ct 
U U U U U 0 U >, ct U 
U 4-1 U U U U U U ct 
U U U 4-1 4--, C) U U o 
57' 
0 0 :Z U 4-1 Q- 4 
ct 
-z 
;z-- 2z "-;Z" 
-zS -Z.) -4:" - - -Q -, 2, ZZ5 <Z5 
; E ,Z- L- N N N N 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
246 RICHARD CONDIT ET AL. 
Table 2. Species whose population of stems l10 mm changed at a rate B5/5% per year. The four columns 
of species characteristics are the same as those given in Table 1 (growtth form, coloniizing, swamp and slope 
status). Species are ordered fiom those with the fastest shriniking populations to the fastest growinig; the line 
separ-ates growinig from shl-inikinig populations. 
Population Rate of population chanige 
Species G C W S 1982 1990 1982-1985 1985-1990 1982-1990 
Pilper aequiale S n 1n S 219 83 -0.0984 -0.1223 -0.1131 
Pilper culebr-anu;in S ... W S 120 53 -0.2036 -0.0391 -0.0992 
Cizamaedorea tepejilote S ... II S 32 16 -0.1128 -0.0600 -0.0803 
C(estr-umi egalop/zyllhmn S ... W S 309 157 -0.0769 -0.0783 -0.0777 
Hanmpea ppendicidata M ... II n 76 40 -0.1363 -0.0388 -0.0760 
Acalyp/ha dive-sifolia S C W n 1568 827 -0.0742 -0.0717 -0.0727 
Acaivplya niacrostachya U C n n 80 45 -0.0626 -0.0762 -0.0714 
C'oostegia bracteata S nI W S 391 209 -0.0931 -0.0563 -0.0711 
Pip)er cordutlatuni S n 1n n 3149 1777 0.0545 -0.1407 -0.0693 
T?npiWia occiden2talis T ... n n 153 85 -0.0937 -0.0539 -0.0690 
Piper arboreiurn U ... II S 107 60 -0.0798 -0.0624 -0.0690 
S'en2na da-ienisis S S II n 205 116 -0.1171 -0.0328 -0.0657 
Solaniuiin hzayesii M C II S 125 77 -0.1121 -0.0273 -0.0582 
Pouilseniia ar-mnata T n 11 S 3430 2126 -0.0703 -0.0441 -0.0545 
Piper perlasense S n 1n S 110 68 0.0162 -0.1029 -0.0529 
Eryt/hrina costaricana U ... n S 289 185 -0.0622 -0.0464 -0.0525 
Tremiza rinicrazthla M ... n n 32 21 -0.1167 -0.0172 -0.0519 
Olmtiedia aspe-a U n n S 442 279 -0.0434 -0.0564 -0.0510 
Chirysopzvllhrnz cainito T C W n 70 109 0.0399 0.0579 0.0510 
Chlrysopjlyllnzm argedtezum T C n n 423 683 0.0366 0.0681 0.0560 
SY)onidias nionobin T C V W n 63 101 0.0207 0.0785 0.0575 
Crotonz billbergianus U C W n 620 1012 -0.0005 0.0944 0.0590 
Mliconzia rgentea M C W 11 531 902 0.0764 0.0542 0.0626 
Citp)aia refescens T n W 1n 55 96 0.0775 0.0578 0.0654 
Anomna spragzei M C II nl 55 143 0.0680 0.1328 0.1082 
Palicoitrea gufianzenisis S C w n 377 1475 0.1861 0.1533 0.1654 
Psychotria graciliflora S ... W n 10 44 0.1231 0.2176 0.1853 
(38%) changed by <1% per year, and 11 (8%) changed by more than 5% per 
year (eight of the latter were declines and three increases). The fastest rate of 
change in the larger size class was Inga acuminata, which increased from 11 to 
20 stems; a more common species increasing nearly as rapidly was Xjlopia 
macranth/a, whose population rose from 79 to 128 stems. The greatest decline 
was in Solanum hIzaesii, which had 40 stems in 1982 but just 13 in 1990. A more 
abundant species, Pterocarpus rohrii, fell from 136 to 83 stems. The mean rate of 
population change among 136 species in the large size class was -0.32%, and 
the mean rate of absolute clhange was 1.93%/. These are not significantly differ- 
ent from rates for stems 10 mm (t-test). 
Within-species consistency inpopulation chacnge 
Rates of change during 1982-1985 and 1985-1990 were consistent within 
species (Figure 2A; the correlation is highly significant: r2=0.266, P<0.0001 
for stems ?10 mm dbh; r2=0.246, P<0.0001 for stems ?100 mm). There were 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
Climate variation and tropical forest change 247 
.,.... ... , . ..... t....,... . 1. I . 
A) 0.20 
0.15. 
0.10, 
0.0 
-0.050 
-O .1 0 ,,, .. 
-o . . 1 5 - I . I . . . . I - I . . . ,. . . . . . - 
-0.20 -0.1 5 -0.10 -0.05 -0.00 0.05 0.10 0.15 0.20 
rate 1 982-85 
B) 0.10 
0.05- * l 
-0.1 1 
-E. .1 * . 
-0.1E0 -0.05 o.o O.05 o.1 o 
rate 010 mm 
Figure 2. Scatter plots of population change, with one point plotted foi- each species. Th-e dotted lines aile 
diagonal regressions, slhown only to indicate tlhe dii-ection of trends. (A) Population change for steims :,B IO mm 
dbh, plotting 1985-1990 rate vs 1982-1985; 219 species included. (B) Population change over 1982-1990, 
plotting rate foi- stems :,B1I00 imm dbh vs rate for stems :,B1I0 mim dbh; 136 species included. 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
248 RICHARD CONDIT ET AL. 
Table 3. Numbers of species with increasing and decreasing populations, by colonizing status, slope status 
and growth form. 'Slope' mean-s lope-specialists that were not colonizers, 'colonizer' means colonizing species 
that were not slope-specialists and 'neither' means species that were neither slope-specialists n-or colonizers 
(unlike Figure 1, this latter category does not include species for which information- was missing). The four 
slope-colonizer species ar-e not included. Asterisks between- the increasin-g and decreasing columns indicate 
a significant difference in the fi-action of species increasing between a given group and the 'neither group' 
(** indicating P<0.01 and * P<0.05). The final column gives the mean rate of population change for each 
species group. 
Number of populations 
Mean rate 
Growth form Status Increasing Decreasing No chan-ge of change 
Lar-ge trees Slope 5 6 0 -0.004 ?0.030 
Neither 15 7 0 0.005 ?0.017 
Colonizer 8 8 0 0.003 ?0.032 
Mid-sized trees Slope 5 3 0 0.008 ?0.012 
Neither 17 11 0 0.001 ?0.020 
Colonizer- 4 4 0 0.018 ?0.049 
Understorey trees Slope 1 ** 6 0 -0.025 ?0.016 
Neither 19 2 0 0.016 ?0.016 
Colonizer 3 2 0 -0.002 ?0.049 
Shrubs Slope 0 * 4 0 -0.062 ?0.043 
Neither 10 5 1 -0.004 ?0.024 
Colonizer 1 3 0 0.003 ?0.011 
All species Slope 11 ** 19 0 -0.013 ?0.033 
Neither 61 25 1 0.005 ?0.020 
Colonizer 16 * 17 0 0.006 ?0.050 
cases, though, where populations increased dramatically prior to 1985 then 
declined afterwards, or vice versa. For example, Piper cordulaturm increased from 
3149 to 3713 stems, then declined to 1777 (Table 1). Rate of change was also 
fairly consistent between the two size classes (r2 = 0.265, P < 0.000 1 for the 1982- 
1990 rate), but again there were exceptions (Figure 2B). For example, Trichilia 
tuberculata, the most abundant large tree in the plot, suffered a rather consider- 
able decline among stems 100 mm dbh, but its population 10 mm dbh 
increased (Table 1). 
Population change as a function of species characteristics 
Slope status. Slope status was clearly associated with a species' probability of 
declining in abundance. Excluding all colonizing species (in order to separate 
the effect of that variable) there were 30 slope-specialists, 19 declining in abund- 
ance and 11 increasing. Of 87 non-slope species, 25 declined and 61 increased 
(Table 3). Thus, 29% of non-slope but 63% of slope-species declined in abund- 
ance (X2 = 1 1. 1, df= 1, P < 0.0 1). Sample sizes were augmented if colonizers were 
included, and the pattern remained: 63% of 52 slope species but 42% of 165 
non-slope species declined in abundance. 
The pattern did not hold for all growth forms, though, in fact the distinction 
between slope and non-slope species was due entirely to species of smaller 
stature - shrubs and treelets. Of this group, 91% of slope-species and only 19% 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
Climate variation and tropical forest change 249 
of non-slope species declined in abundance, with the difference statistically 
significant in both growth forms (Table 3). But in large and mid-sized trees, 
there was no such distinction, with 47% of slope and 36% of non-slope species 
declining in abundance and no significant difference in either growth form 
(Table 3). 
Slope-specialists suffered the most impressive population declines. Of the 18 
species declining more than 5% per year, 11 were slope-specialists (Figure IA, 
Table 2). The four most rapid declines were slope-specialist shrubs (Table 2). 
Conversely, none of the nine species increasing more than 5% per year were 
slope-specialists (Table 2). 
The poor performance of slope-specialists was not evident when considering 
stems above 100 mm dbh, because there were few treelets and shrubs included 
in this size class. For large and mid-sized trees, slope-specialists performed no 
worse than non-slope: six of 17 slope species decreased in abundance, whereas 
of non-slope species, 20 of 40 declined. Among treelets and shrubs, slope- 
specialists did suffer more declines than non-slope species (two out of three vs 
four out of 13) but the sample was far too small to evaluate statistically. 
Colonizing species. Species designated as gap-colonizers performed somewhat 
worse than non-colonizing species, although not as poorly as slope species. 
Excluding slope-specialists (to isolate the effect of the colonization variable), 
52% of 33 colonizing species and 29% of 87 non-colonizers declined in abund- 
ance (x2 = 5.3, P < 0.05, Table 3). Each of the four growth forms showed a 
comparable pattern, with colonizers doing slightly worse than non-colonizers, 
but none was statistically significant by itself (Table 3). 
Despite the fact that colonizers on average performed poorly, they were over- 
represented among rapidly increasing populations. Seven of the nine species 
increasing faster than 5% per year were colonizers, and only one was not (the 
other had an unknown colonization index). In contrast, of the 18 species 
decreasing faster than 5% per year, four were colonizers and six were not. 
When considering abundance changes in stems ? 100 mm dbh, there was no 
indication that colonizers performed differently than non-colonizers. Excluding 
slope-specialists, 10 of 25 colonizing species decreased in abundance, while 24 
of 53 non-colonizers declined. The four growth forms did not differ. 
Szwamp status. Swamp status was unrelated to population change. Considering 
stems 10 mm dbh, 26 of 51 swamp species declined in abundance (51/%), 
whereas 78 of 167 non-swamp species declined (47%). For stems 100 mm 
dbh, 10 of 26 swamp species (38%/) and 51 of 110 non-swamp (46%/) declined. 
Neither difference, nor any for individual growth forms, reached statistical signi- 
ficance. Like colonizing species, though, swamp-specialists were over- 
represented among rapid increasers: seven of nine species increasing ?5/5% per 
year were swamp species, but only four of 18 species declining by ?5/5% per 
year were. 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
250 RICHARD CONDIT ET AL. 
DISCUSSION 
Nearly all shrubs and treelets that occur preferentially on the slopes of the 50 ha 
plot declined in abundance. WVe know the slopes around the side of the plateau 
are a wetter microhabitat during the dry season (Becker et al. 1988), and we 
assume that species more abundant there are less able to tolerate drought stress. 
We can support this assumption by casual observations on species distributions: 
some of the familiar slope-specialists at BCI - Poulsenia armata, Olmedia aspera, 
Egythrina costaricana nd Acalypha diversifolia - are common along permanent stre- 
ams in forests near BCI (there are no permanent streams on BCI). Further 
casual support comes from the genus Piper, which is particularly abundant in 
wet forests; five of its eight species in the 50 ha plot species are slope-specialists 
(all eight declined in abundance). 
It seems certain that this group of shrubs and treelets that cannot tolerate 
long drought invaded the plateau forest at BCI during the wetter periods prior 
to 1966 but is now being eliminated by the increased severity of the dry season. 
It is possible that the extreme dry season of 1983 is solely responsible; alternat- 
ively, it may be a continuing problem caused by recurring severe dry seasons. 
We cannot distinguish between the two alternatives now, but future censuses 
will. If the only problem for drought-intolerant species was 1983, then popula- 
tions should level off and perhaps even climb by 1995 or 2000, when the plot 
will be censused anew. We are certain, however, that dry season length and 
severity is the crucial edaphic variable affecting population success and limiting 
species' ranges at BCI (Wright 1992, Wright & van Schaik 1994). Reduction 
of rainfall during the wet season is probably inconsequential (at least for trees) 
since water is never limiting then. 
Why have moisture-demanding trees of larger stature not suffered population 
declines as consistently as shrubs and treelets? We anticipated that they would, 
largely because of two prominent canopy trees and strong slope-specialists that 
suffered severe declines in abundance: Poulsenia armata and Ocotea whitei. But 
other slope-specialists in the canopy, such as Calophyllum longifolium, have 
healthy populations. We suggest the following hypothesis to account for this 
division and the general decline of shrubs and treelets. As adults, some trees, 
like Calophyllum, have longer roots than others, like Poulsenial, long enough to 
reach water below the slopes during the dry season, but not from the plateau 
(which is higher and thus further from the water table, see Wright & van Schaik 
1994); Poulsenia thus suffered high mortality at all sizes during the 1983 El Nifio 
drought, whereas Calophyllzum did not. But both species have drought-sensitive 
seedlings, and are thus largely restricted to the wetter areas within the plot. 
Likewise, there are shrubs and treelets with drought-sensitive seedlings that are 
restricted to moist regions, but nearly all have short root systems as adults 
(Becker & Castillo 1990, Wright 1992) and suffer from long dry seasons at BCI. 
(Some shrubs have other drought-adaptations and are widespread in the plot.) 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
Climate variation and tropical forest change 251 
During moderate dry seasons, there is presumably enough water near the sur- 
face of the slopes for these drought-intolerant plants, and this allowed their 
spread into the plot prior to 1966, when dry seasons were less severe. The 
swamp may remain wet even during the most severe dry seasons, so that 
swamp-specialists can persist despite the drying trend. 
Are the slope-specialists becoming extinct on BCI, or will they persist in 
locally wet sites? It appears not - all shrubs and treelets that declined through- 
out the plot declined on both the slopes and the plateau; in fact, most species 
had similar rates of change in both regions. Thus, it seems that there is a group 
of drought-intolerant species headed for extinction at BCI, at least 16 treelets 
and shrubs, and perhaps as many as 30-40 including the large, drought- 
sensitive trees like Poulsenia and Ocotea whitei. Howe (1990) has already predicted 
that Virola surinamensis, a large, slope-specialist tree, will go extinct on BCI due 
to its inability to tolerate long dry seasons. He based his conclusion on seedling 
survival data, not having seen the 1982-1990 population data which bears out 
his prediction - a decline from 300 to 239 stems. 
Species which preferentially invade light gaps in the 50 ha plot - what we 
called colonizing species - had an almost bimodal distribution of population 
change. A few species had very rapid increases, but the rest did worse than 
average. Condit et al. (1992b) and Hubbell & Foster (1990a) have stated two 
different hypotheses about factors affecting the populations of gap-colonizers. 
One is that the more open canopy caused by drought-induced mortality in 1983 
(Becker & Smith 1990) created more recruitment opportunities for species that 
demand light gaps, leading to a population burst in the 10 mm size class by 
1985 or 1990. The other hypothesis is that the 50 ha plot is undergoing succes- 
sion because the area just north of the plot (plus 2 ha within the plot) was 
cleared around the turn of the century but has matured since. Ruderal species 
abundant in the near-by farmland maintained high sink populations within the 
old forest because of the large number of seeds entering, but these populations 
are now declining. Both factors may in fact be at work at the same time. There 
are some colonizing species which can obviously maintain high populations 
within gaps of old forest (all seven of the rapidly rising species would be 
examples), but other species such as Apeiba tibourbou and Scheffiera morototoni 
which are abundant only in large clearings may not persist in the old forest 
and are now in decline (indeed, the latter dropped out of the plot between 1982 
and 1985). 
An obvious concern with these conclusions is the method for identifying 
edaphic preferences of individual species. Some associations with topographic 
regions may be due to factors having nothing to do with moisture or light 
preference. Artefactual correlations weaken our power to detect effects of mois- 
ture preference, but the trends we did detect should be robust with respect to 
this error and ought to appear even stronger if species with accidental associ- 
ations were segregated. We eventually hope to get physiological information on 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
252 RICHARD CONDIT ET AL. 
some species to define independently those that are drought-tolerant (Mulkey 
et al. 1994, Wright & van Schaik 1994). 
The mean rate of absolute population change in the plot was 2.25% per year, 
a 20% increase or 16% decrease after eight years. About 10% of the populations 
in the plot are changing >5% per year (a 49% increase or 33% decrease over 
eight years). These seem like substantial rates for trees, which ought to have 
rather lethargic population trajectories due to their long life spans and slow 
growth (Condit et al. 1992b, Hubbell & Foster 1990a). Are such changes typical 
for tropical forests, or is BCI unusual because of the climatic shift taking place? 
In a study of 50 ha of dry forest in India, several tree and shrub species under- 
went severe declines in abundance - in just three years - due to elephant 
herbivory (R. Sukumar, unpublished data). Other studies in the tropics have 
been on much smaller areas with irregular censuses, and are very difficult to 
compare. For example, Manokaran & Kochummen (1987) documented some 
abrupt declines and increases in a 34-year record of tropical forest in Malaysia: 
Shorea parvifolia declined from 26 to 16 individuals in 16 years, and Daccyodes 
puberula from 14 to four in 34 years, both consistent declines of about 5% per 
year, but both are based on small samples. The empirical issue of stability in 
tree populations and community composition of tropical forests must be 
resolved by more large datasets, and large-scale plots are now under way at 11 
sites in Africa, Asia and America (Condit 1995). Results from these will settle 
the matter. 
These plots will offer a baseline for assessing the impact of global climate 
change on tropical forests. Long-term changes in precipitation can have tre- 
mendous effects on forests (Foster 1982a,b, Hartshorn 1992), and if Phillips & 
Gentry (1994) are correct, CO-fertilization may be changing forest-wide 
dynamics. We see perhaps 10% of the species at BCI headed for extinction 
because of a 25-year decline in precipitation. Understanding and even anticipat- 
ing climatic effects on tropical forests will be crucial for their long-range 
conservation. 
ACKNOWLEDGEMENTS 
The Smithsonian Tropical Research Institute in Panama provided generous 
logistical and financial support for the censuses, and we thank Ira Rubinoff for 
his long-term support of the project. We also thank the field workers who 
contributed to the censuses on BCI, more than 100 people from 10 countries, 
and R. Perez and S. Loo de Lao for their persistent work maintaining the plot 
and its database. The project has been supported by grants from the National 
Science Foundation, the Smithsonian Scholarly Studies Program, the Smithson- 
ian Tropical Research Institute, the John D. and Catherine T. MacArthur 
Foundation, the World WVildlife Fund, the Earthwatch Center for Field Studies, 
the Geraldine R. Doge Foundation and the Alton Jones Foundation. This article 
is a scientific contribution from the Center for Tropical Forest Science, which 
is supported by the John D. and Catherine T. MacArthur Foundation. 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
Climate variation and tropical forest change 253 
LITERATURE CITED 
BECKER, P. & CASTILLO, A. 1990. Root architecture of shrubs and saplings in the understorey of a 
tr-opical moist forest in lowlan-d Pan-ama. Biotropica 22:242-249. 
BECKER, P., RABENOLD, P. E., IDOL, J. R. & SMITH, A. P. 1988. Water potential gradienlts for 
gaps and slopes in a Pan-am--an-ian- tr-opical moist for-est's dry season. Journzal of Tropical Ecologv 4:173- 
184. 
BECKER, P. & SMITH, A. P. 1990. Spatial autocor-relation of solar radiation in a tropical moist for-est 
un-derstorey. Agricultural anid Forest Meteorology 52:373-379. 
BOTKIN, D. B. & NISBET, R. A. 1992. Projecting the effects of climate change on biological diver-sity 
in forests. Pp. 277-293 in Peters, R. L. & Lovejoy, T. E. (eds). Global zewarnfing anzd biological diversitr. 
Yale Un-iversity Press, New Haven. 
BRADLEY, R. S., DIAZ, H. F., EISCHEID, J. K., JONES, P. D., KELLY, P. M. & GOODESS, C. 
M. 1987. Precipitation- fluctuation-s over northern-i hemisphere lan-d areas since the imiid-19th century. 
Scienice 237:171-175. 
BUSH, M. B. & COLINVAUX, P. A. 1990. A pollen record of a complete glacial cycle fromll lowlanld 
Panam--a. Jour-nal of V7egetationi Sc ence 1:105-118. 
BUSH, M. B., COLINVAUX, P. A., WIEMANN, M. C., PIPERNO, D. R. & LIU, K. 1990. Late 
pleistocen-e temper-ature d pression and vegetation- chan-ge in Ecuadorian Amazon-ia. Quiaternary Research 
34:330-345. 
CONDIT, R. 1995. Research in lar-ge, lon-g-term tropical for-est plots. Trenids in Ecologi' anid Ez'olhtion 
10:18-22. 
CONDIT, R., HUBBELL, S. P. & FOSTER, R. B. 1992a. Recruitmen-t n-ear coinspecific adults and the 
mainten-ance of tree an-d shlub diversity in a neotropical forest. Anmerican Natur1alist 140:261-286. 
CONDIT, R., HUBBELL, S. P. & FOSTER, R. B. 1992b. Stability an-d clhanlge of a ineotropical imioist 
forest over a decade. Bioscience 42:822-828. 
CONDIT, R., HUBBELL, S. P. & FOSTER, R. B. 1993a. Iden-tifying fast-growing n-ative tr-ees from-- 
the n-eotropics usinlg data fiomn a lar-ge, perman-ent cen-sus plot. Forest Ecology, and Managenment 62:123- 
143. 
CONDIT, R., HUBBELL, S. P. & FOSTER, R. B. 1993b. Mortality an-d growth of a comlimer-cial 
h-ar-dwood, 'El Cativo', Prioria copaifera, in Pan-ama. Forest Ecology anid Managenment 62:107-122. 
CONDIT, R., HUBBELL, S. P. & FOSTER, R. B. 1995. Mortality rates of 205 neotropical tree species 
an-d the responses to a severe drought. Ecological Monographs 65:419-439. 
CROAT, T. R. 1978. Flora of Barro C'olorado Island. Stan-ford University Press, Stan-ford, Californ-ia. 
DALE, V. H. & FRANKLIN, J. F. 1989. Poten-tial effects of climate change on stand developmen-t in 
the Pacific Nor-thwest. Caniadiani Journ-lal of Forestiy Research 19:1581-1590. 
D'ARCY, W. G. 1987. Flora of Panama. Part I: Introduzctioni a id checklist. Missouri Botanical Garden, St 
Louis, Missouri, USA. 328 + xxx pp. 
DAVIS, M. B. 1981. Quater-nary history an-d the stability of forest comiimun-ities. Pp 132-153 in West, 
D. C., Shugart, H. H. & Botkin, D. B. (eds). Forest sutccession: concepts and applications. Sprinlger-Verlag, 
New York. 
DELCOURT, P. A. & DELCOURT, H. R. 1987. Long term11 forest diynamnics of the tenmperate Zone: a case 
stuldy of late quzaternarv forests in eastern1 Nor-th Amer-ica. Sprin-ger-Verlag, New York. 
DIAZ, H. F., BRADLEY, R. S. & EISCHEID, J. K. 1989. Precipitationl fluctuations over global land 
areas since the late 1800's. Journ-lal of Geophysical Researlch 94:1195-1210. 
FOSTER, R. B. 1982a. Faminie on Barro Colorado Island. Pp. 201-212 in Leigh, E. G., Jr-, Ran-d, S. 
A. & Win-dsor, D. M. (eds). The ecology, of a tropical forest: seasonial r4ylthmns and lonig-term1l chaniges. 
Smithsonian Inistitution- Press, Washington, DC. 
FOSTER, R. B. 1982b. The seasonal hl-ytlhm of fruitfall on- Barro Color-ado Island. Pp. 151-172 in 
Leigh, E. G., Jr, Ran-d, S. A. & Win-dsor, D. M. (eds). The ecology of a tropical forest: seasonial rhythnis 
anid long-trenm chaniges. Smithsonian- Institution- Press, Washinigton, DC. 
FRANKLIN, J. F., SWANSON, F. J., HARMON, M. E., PERRY D. A., SPIES, T. A., DALE, V. 
H., McKEE, A., FERRELL, W. K., MEANS, J. E., GREGORY, S. V., LATTIN, J. D., 
SCHOWALTER, T. D. & LARSEN, D. 1992. Effects of global climatic chanlge on for-ests in 
northwesterni- Nor-th America. Pp. 244-257 in Peters, R. L. & Lovejoy, T. E. (eds). Global warnming 
and biological dizersity. Yale Un-iversity Pr-ess, New Haven. 
HAMILTON, A. C. & TAYLOR, D. 1991. History of climate and forests in tropical Africa dur-inlg the 
last 8 million years. C'limiiate C'hanige 19:65-78. 
HARTSHORN, G. S. 1992. Possible effects of global warmin-g on- the biological diversity in tropical 
for-ests. Pp. 137-146 in Peters, R. L. & Lovejoy, T. E. (eds). Global zwarming anid biological diversity). 
Yale University Press, New Haven. 
HOWE, H. F. 1990. Survival and growth of juven-ile liirola suiKiarensis in Pan-ama: effects of herbivory 
and canopy closure. Journal of Tropical Ecology 6:259-280. 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
254 RICHARD CONDIT ET AL. 
HUBBELL, S. P. & FOSTER, R. B. 1983. Diversity of canopy trees in a neotropical forest and 
implications for conservation. Pp. 25-41 in Sutton, S. L., Whitmore, T. C. & Chadwick, A. C. (eds). 
Tropical rain forest. ecology and nianagemnenit. Blackwell Scien-tific Publications, Oxford. 
HUBBELL, S. P. & FOSTER, R. B. 1986a. Commonni-ess and rarity in a neotr-opical for-est: implications 
for tropical tree conser-vation. Pp. 205-231 in Soul6, M. (ed.). Conservation biology: the scienice of scarcity 
and diversity. Sinauer Associates, Inc., Sunderland, Massachusetts. 
HUBBELL, S. P. & FOSTER, R. B. 1986b. Can-opy gaps and the dynamics of a n-eotropical for-est. Pp. 
77-96 in Crawley M. J. (ed.). Plant ecology. Blackwell Scientific Publications, Oxford. 
HUBBELL, S. P. & FOSTER, R. B. 1987. The spatial con-text of regeneration in a neotropical forest. 
Pp. 395-412 in Crawley, M., Edwards, P. J. & Gray, A. (eds). Colonization, succession, anid stabilitv. 
Blackwell Scientific Publications, Oxford, England. 
HUBBELL, S. P. & FOSTER, R. B. 1990a. Structure, dynamics, and equilibrium status of old-gr-owth 
forest on Barro Colorado Island. Pp. 522-541 in Gentry, A. (ed.). Fowr Neotropical rain forests. Yale 
University Press, New Haven. 
HUBBELL, S. P. & FOSTER, R. B. 1990b. The fate of juvenile trees in a n-eotropical forest: imllplications 
for the natural inaintenance of tropical tree diversity. Pp. 325-349 in Hadley, M. & Bawa, K. S. 
(eds). Reproduzctive ecology of tropical forest planits. Man and the Biosphere Series, Vol. 7. UNESCO/ 
IUBS, Paris, and Parthenon- Publishinig, Carn-forth, UK. 
HUBBELL, S. P. & FOSTER, R. B. 1992. Short-term population dynamiics of a neotropical forest: why 
ecological research matters to tropical conservation- and managemen-t. Oikos 63:48-61. 
LEIGH, E. G., Jr, RAND, S. A. & WINDSOR, D. M. (eds). 1982. The ecology of a tropicalforest: seasonal 
rhlythms anid long-term changes. Smithsonian Institution Pr-ess, Washington, DC. 
LEIGH, E. G., Jr, WINDSOR, D. M., RAND, S. A. & FOSTER, R. B. 1990. The imnpact of the 'El 
Nifio' drought of 1982-1983 on a Panamanian semideciduous forest. Pp. 473-486 in Glynn, P. W. 
(ed.). Global ecological conisequenices of the 1982-1983 El Nifio-Souzthern Oscillation. Elsevier Press. 
MANOKARAN, N. & KOCHUMMEN, K. M. 1987. Recr-uitmen-t, growth and inortality of tree species 
in a lowland dipterocarp forest in Penin-sular Malaysia. Joutrnial of Tr-opical Ecology 3:315-330. 
MANOKARAN, N., LAFRANKIE, J. V., KOCHUMMEN, K. M., QUAH, E. S., KLAHN, J., 
ASHTON, P. S. & HUBBELL, S. P. 1992. Stand table and distribittion of species in the 50-ha research plot 
at Pasoh Forest Reserve. Forest Research Inistitute of Malaysia, Research Data No. 1. 454 pp. Kepong, 
Malaysia. 
MULKEY, S. S., SMITH, A. P., WRIGHT, S. J., MACHADO, J. L. & DUDLEY, R. 1994. Contrasting 
leaf phen-otypes control season-al variation in water loss in a tropical forest shl-ub. Proceedings of the 
Nationial Academty of Sciences 89:9084-9088. 
O'BRIEN, S. T., HAYDEN, B. P. & SHUGART, H. H. 1992. Global climatic chanlge, hurricanles, and 
a tropical for-est. C'limatic Change 22:175-190. 
OVERPECK, J. T., RIND, D. & GOLDBERG, R. 1990. Climate-induced changes in forest disturban-ce 
and vegetation. Nature 343:51-53. 
PASTOR, J. & POST, W. M. 1988. Response of n-orthern- for-ests to CO2-induced climate chan-ge. 
Nature 344:55-58. 
PHILLIPS, 0. L. & GENTRY, A. H. 1994. Increasing turnover througlh time in tropical forests. Science 
263:954-958. 
PHILLIPS, 0. L., HALL, P., GENTRY, A. H., SAWYER, S. A. & VASQUEZ, R. 1994. Dynamics an-d 
species richlness of tropical rain forests. Proceedinigs of the Nationial Acadenp' of Science 91:2805-2809. 
SHUGART, H. H. & SMITH, T. M. 1992. Usin-g computer models to project ecosystem response, habitat 
change, an-d wildlife diver-sity. Pp. 147-157 in Peters, R. L. & Lovejoy, T. E. (eds). Global zwarming anid 
biological diversity. Yale Un-iversity Press, New Haven. 
SOLOMON, A. M. 1986. Transien-t response of forests to CO2-induced climnate chan-ge: simulation 
inodelin-g experimnents in eastern- North Aim-erica. Oecologia 68:567-579. 
SUKUMAR, R., DATTARAJA, H. S., SURESH, H. S., RADHAKRISHNAN, J., VASUDEVA, R., 
NIRMALA S. & JOSHI, N. V. 1992. Long-term monitoring of vegetation in a tropical deciduous forest 
in Mudumalai, southern Inidia. Cuarrent Science 62:608-616. 
SUKUMAR, R., RAMESH, R., PANT, R. K. & RAJAGOPALAN, G. 1993. A 613C recor-d of late 
Quatern-ary climate chan-ge from tropical peats in southern India. Natutre 364:703-706. 
SWAINE, M. D. & WHITMORE, T. C. 1988. Onl the defin-ition ofecological species groups in tr-opical 
rain forests. Vegetatio 75:81-86. 
URBAN, D. L., HARMON, M. E. & HALPERN, C. B. 1993. Potential r-espon-se ofPacific northlvestern 
forests to climatic change, effects of stand an-d in-itial composition. C'limatic C'hiange 23:247-266. 
WELDEN, C. W., HEWETT, S. W., HUBBELL, S. P. & FOSTER, R. B. 1991. Survival, growth, and 
recruitment of saplings in can-opy gaps an-d forest understorey on Barro Colorado Island, Panama. Ecology 
72:35-50. 
WINDSOR, D. M. 1990. Climate anzd nmoistutre rvaiability in a tropicalforest: long-termn recordsfrom Barro C'olorado 
Island, Pananmd. Smithsonian Contribution- to the Earth Scien-ces, No. 29. Smithsonian Institution- Press, 
Washington, DC. 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
Climate variation and tropical forest change 255 
WINDSOR, D. M., RAND, A. S. & RAND, W. M. 1990. Caracteristicas de la precipitaci6n en1 la isla de 
Barro Colorado. Pp. 53-71 in Leigh, E. G., Rand, A. S. & Windsor, D. M. (eds). Ecologia de uzn bosquze 
tropical. ciclos estacionalesy cambios a largo plazo. Smithsonian Tropical Research Institute, Balboa, Republica 
de Panama. 
WOODS, P. 1989. Effects of logging, drought, and fire on structure and composition- of tropical forests in 
Sabah, Malaysia. Biotropica 21:290-298. 
WRIGHT, S. J. 1992. Seasonal drought, soil fertility and the species diversity of tropical forest plan-t 
communities. Trenids in Ecology and Evolutionz 7:260-263. 
WRIGHT, S. J. & VAN SCHAIK, C. P. 1994. Light and the phenology of tropical trees. American Naturaalist 
143:192-199. 
ZIMMERMAN, J. K., EVERHAM, E. M., III, WAIDE, R. B., LODGE, D. J., TAYLOR, C. M. & 
BROKAW, N. V. L. 1994. Responses of tree species to hurricane winds in subtropical wet forest in 
Puerto Rico: implications for tropical tree life histories. Jowrnal of Ecology 82:911-922. 
Accepted 18 March 1995 
APPENDIX 
The 48 species names that have been changed since the 50 ha plot was initiated 
in 1981 or which do not appear in Croat (1978). Thle current name is the one 
appearing in Table 1. The eight species listed as sp. nov. were newly discovered 
in the 50 ha plot and remain undescribed. 
Current name Namne in D'Arcy (1987) Name in Croat (1978) 
Appunia seibertii Appunia seibertii not appearing 
Ardisia giuianiensis Ardisia gutianenisis not appearing 
Brosimnum guiiienise Brosimurm gutinenise not appearing 
C'zarnaedorea tepejilote Chamnaedorea tepejilote Charnaedorea wenlandiania 
Chanigutava schlippii Psidiuim aniglohotndutrenisis Psidiurin anglohondutrenisis 
Chrnsochlanys eclipes Tovomnitopsis nicaragutensis Tovonmitopsis niicaraguenisis 
C'hIysopkyllurm argenteurin Cynodendron panamense Cynodendron panamense 
Ervthroxylum rnacropIhyllurn Erthroxylrni rnacrophylluzni not appearing 
Garciniia inter?nedia Garcinia intermlledia Rheedia edutlis 
Garcinia mnadrunio Garcinia madrunlo Rheedia acuminiata 
Gutarea granidifolia Gutarea grandifolia Guarea muizltijflora 
Gutarea sp. nov. not appearing not appearing 
Heisteria acurminata Heisteria acurininiata Heisteria longipes 
Hyeronirma alcheornoides Hyeronimla l xifora Hveroniima laxiftora 
Inga acuminata not appearing not appearing 
Lonichocarpus latifolia Lonchocarpus latifolia Lonichocarputs pentaplzvllus 
Lopimia dasypetala Lopimnia dasypetala Pavonia dasypetala 
Maclura tinctoria Clzlorophora tinictoria n-ot appearing 
Malrnea sp. nov. not appearing Crematosperma sp. 
AvIyrosperrnurn frutescens AMiyrospernmum ftutescenis not appearing 
Nectandra piuipurzea Nectandra puiputrea Nectanidra piupur1escenis 
Nectandra sp. nov. 1 not appearing not appearing 
Nectanldra sp. nov. 3 not appearing not appearing 
Ocotea puberilda Ocotea puberilda Ocotea pryamidata 
Ocotea whitei Ocotea zvhitei Ocotea skuztchii 
Oenoca?puts mapoura Oenioca?pus nmapoura Oeniocarputs panarnainus 
Ormosia amazonica Ormosia amnazonica not appearing 
Osniosia croatii Ormosia coccinea Ormosia cocciniea 
Phoebe cinnamomifolia Phoebe cininamonJifolia Phoebe mnexicana 
Pochota quzinata Bombacopsis quzinata Bonmbacopsis quziinata 
Pochota sessilis Bomnbacopsis sessilis Bombacopsis sessilis 
Pourounia bicolor Pourouzma gutianensis Pourouznia giuianienisis 
Poutteria reticulata Pouteria unilocutlaris Pouteria wnilocidaris 
Protiurm sp. nov. not appearing not appearing 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions
256 RICHARD CONDIT ET AL. 
APPENDIX 1. (cont.) 
Current name Name in D'Arcy (1987) Namne in Croat (1978) 
Pscchotria gracflor-a Psvchotria graciflora not appearing 
Pterocacpzus belizensis Pterocarpus belizensis not appearing 
Sapiumn auicuparium Sapizum cauzdatzum both (now considered syn-onyms) 
Sapiuim sp. nov. not appearing not appearing 
Schefflera morototoni Didvmopanax morototonii Didvinopaniax morototoni 
Senina dariensis Seinna dar-iensis Cassia fiuticosa 
Socratea exoirhiza Socratea exorrhiia Socratea dursissima 
Solan1u1m1 steyeniarkii Solanium argenteurn Solanurim argeniteuini 
Terminalia oblonga Terminialia obloniga Termiiiialia chiriquzensis 
Trichilia pallida Trichilia pallida Trichilia monitana 
Trichilia tuberculata Trichilia tutberculata Trichilia cipo 
Trichospermnzn galeottii Trichospe-rmum galeottii Trichospermumii t exicamuiiin 
Urera baccifera Urera baccifera not appering 
Virola sp. nov. not appearing not appearing 
This content downloaded by the authorized user from 192.168.52.67 on Mon, 10 Dec 2012 15:29:57 PM
All use subject to JSTOR Terms and Conditions