Content uploaded by Andrew Laurie
Author content
All content in this area was uploaded by Andrew Laurie on Jan 15, 2018
Content may be subject to copyright.
361
EFFECTS OF
THE
1982-83 EL NIRO-SOUTHERN OSCILLATION EVENT ON MARINE
IGUANA
(AMBLYRHYNCHUS CRISTATUS
BELL, 1825) POPULATIONS ON
GALAPAGOS
W.
ANDREW LAURIE
Max-Planck-Institut fur Verhaltensphysiologie,
8
131, Seewiesen, West Germany, and Large
Animal Research Group, Department of Zoology, University of Cambridge, Cambridge CB2
3EJ,
United Kingdom.
ABSTRACT
Laurie.
W.
A..
1989. Effects of the 1982-83 El Niiio-Southern Oscillation event
on
marine iguana
(Amblhhynchus Gristatu Bell, 1825) populations on Galapagos.
The effects of the 1982-83 El Niiio-Southern Oscillation event (ENSO) on marine iguanas
(Amblvrhvnchus cristatus) were observed during a long term study of marine iguana population
dynamics on Santa Fe, Galapagos, begun in 1981. The 1982-83 ENSO was the most severe ever
recorded: there were record sea-surface temperatures, sea-levels and rainfall, and a major change
in marine algal flora resulted in disappearance of most of the iguanas' preferred food species and
colonization of the intertidal zone
by
the brown alga Giffordia mitchelliae, a species not previously
recorded in Galapagos. This led to widespread starvation, with about 60% of the Santa Fe
population dying between March and August 1983, and similar mortality on other islands of the
archipelago. Adult males and juveniles suffered the highest mortality, with 1982 hatchlings being
almost completely exterminated. Body condition and growth rates were greatly depressed during
ENSO,
with adult growth ceasing almost entirely, but both increased again rapidly after the
population crash and reached levels higher than before ENSO. There was almost no breeding in
the post ENSO season (1983-84) but since then frequency of breeding, age at first breeding, and
clutch size have all increased above pre-ENS0 levels. It is suggested that the increases in rates
of
growth and reproduction are due to a reduction in competition for food after the return of normal
feeding conditions at greatly reduced population density.
Y
1 INTRODUCTION
A long term study of marine iguana (AmblYrhvnchus Gristatus) population dynamics in
Galapagos, begun in 1981, provided an opportunity
to
study the effects
on
the iguanas
of
the
1982-83 El Niiio-Southem Oscillation (ENSO) event (Philander, 1983), the most severe
on
record
(Quinn et al., 1978, 1987; Glynn, 1988). ENSO events are characterized by a massive advection
of warm, low salinity, nutrient poor surface water to the south in the eastern tropical Pacific,
mainly along the coasts
of
Ecuador and Peru (Houvenaghel, 1978, 1984; Rasmusson, 1984;
Hansen,
this
volume). The biological productivity of the euphotic zone declines rapidly leading to
reduced survival and reproduction of animals at higher trophic levels (Barber and Chavez, 1983,
1986; Trillmich and Limberger, 1985; Arntz, 1986; Barber and Kogelschatz, this volume),
although the increased rainfall leads to increased reproduction in land-based ecosystems, for
example, in Darwin's finches on Galapagos (Gibbs and Grant, 1987; Grant and Grant, 1987) and
in the floras of the "lomas" formations in the Atacama and Peruvian deserts (see Dillon and
Rundel, this volume).
362
The marine iguana is endemic to Galapagos and is widely distributed over the archipelago with
highest concentrations on the western islands (Laurie, 1983a). The iguanas feed on fleshy
macrophytic marine algae, either diving for them
or
grazing on exposed intertidal
rocks
at low tide
(see Trillmich and Trillmich, 1986). They are sexually dimorphic, with adult males typically
weighing
70%
more than adult females. Adult male body weight varies from a maximum of 12.3
kg on southern Isabela to about 1.2 kg on Genovesa (Laurie, in prep.). Males defend mating
territories during the breeding season, and females lay one to six eggs about one month after
copulation. The eggs take three months to incubate in nests dug
30-80
cm deep in sand
or
volcanic
ash. The time of the breeding season varies between islands (Fig. l), being earliest (nesting in
January) on Santa Fe and latest (nesting in March-April)
on
southern Isabela and Espaiiola (Laurie,
in prep.).
During 1983 unusually high mortality
of
marine iguanas was observed in populations on all the
islands of the archipelago, except Wenman and Culpepper, which were not visited (Laurie,
1983b). A major change in marine algal flora was observed during the same period and
abnomially high rainfall, sea-surface temperatures
(SST)
and sea-levels associated with the
Fig. 1. Map of the Galapagos Islands, showing main study site, Miedo.
1982-83
El
Niiio-Southern Oscillation event
(ENSO)
were recorded from November 1982 until
July 1983. The mean monthly SST anomaly reached +4.3OC in June 1983 (Fig. 2), and the
pattern
of
SST
fluctuations was similar to those on the coast of Peru (Chavez, 1987). The
363
tradewinds failed almost completely, and
on
Santa Cruz, where the mean annual rainfall between
1965 and 1981 was 374 mm, 3,264
mm
of rain fell between November 1982 and July 1983
(Robalino, 1985). There was an increase
in
sea level over the same period that varied between 20
and 45 cm (Wyrtki, 1985). ENS0 events occur frequently but are poorly predictable and vary
I
I I I
1965 1970 1975 1980 1985
30
-
V
0
W
n
3
$
25
a:
W
a
I
W
c-
W
u
-
a
k
20
3
ul
W
ul
a
15
1965 1970 1975 1980 1985
YEAR
Fig. 2. Monthly mean sea-surface temperature
(SST)
anomalies (above) and monthly mean sea-
surface temperature (below) observed
on
the shore
in
Academy Bay, 1965-8 (courtesy Charles
Darwin Research Station). Broken curve denotes long-term (22 years) annual mean
SST.
greatly in intensity, extent of influence and duration.
Quinn
et al. (1987) have documented 24
events of near moderate to very strong intensity since 1900 and
50
between
1800
and 1987, a
mean frequency
of
one event
in
3.8 years. The average interval between strong
or
very strong
364
ENSO
events, with mean monthly sea surface temperature anomalies of 3.0
to
5.0OC, is 12.3
years (Quinn et al., 1978,1987) but the 1982-83 ENSO was exceptional, and there is evidence that
the last event of comparable magnitude occurred in 1877-78 (Kiladis and Diaz, 1984).
2
STUDY
AREA
The main study area was at Miedo, on the south coast of Santa Fe (Fig. 1) and consisted of 2
km of low, rocky coastline with extensive intertidal flats and abundant marine algae. There is an
old, uplifted beach deposit 300 m inland at the base of an escarpment, and most of the marine
iguanas in the study area nested there. A number of other sites were chosen on other islands for
comparative observations during regular visits over the study period. The climate of Galapagos is
biseasonal: January to May is the hot season, with the only substantial rainfall, and June to
December is cool, and frequently overcast, with persistent, very light drizzle (Colinvaux, 1984).
3 METHODS
3.1 Census tec hnicues
In order to collect comparative data on population densities and composition on Santa Fe and
other islands binoculars were used to count animals and classify them according
to
sex and size.
Iguanas were divided into 11 different size classes, based
on
snout-vent length and, with practice,
animals could
be
classified accurately
to
size class without being captured
or
otherwise disturbed.
Differences between the sexes in body size, head width and nuchal crest development were
sufficient to determine the sex
of
most adults without capture. Prominent hemipenes were visible
in some younger males when held in the hand but even after capture sex determination was
generally possible only for the older animals. Counts
of
iguanas along the same stretch of
coastline produced different results in terms of both numbers and population composition
according to the time of day and state of the tide. Experiments showed that the censuses in the late
afternoon gave the highest counts, and that consistent estimates of population size and composition
can be made by a simple mark, release and count method (Laurie, 1982). An annual census was
made each April on Santa Fe using this method: the animals released after weighing and measuring
constituted the marked population.
3.2 CaDturine and marking
ieuanas
A total of 3,833 iguanas was marked over the study period: 3,482
on
Santa Fe and 351 on
Caamaiio, a small islet off Santa Cruz (Fig. 1). A further 1,440 captures were made
on
other
islands but the animals were weighed, measured and released without marking and may not all be
different individuals. Fifty-two recaptures were made of animals marked earlier by C. Rohrbach
and
N.
Rauch on Caamaiio and Punta Nuiiez.
Adult iguanas were caught
by
hand on the shore in the early morning when still torpid,
or
later
with the help of a running noose on a bamboo pole. Hatchlings were caught on emergence at the
nesting area in an enclosure fenced
by
45 cm high plastic sheeting. An intensive effort was made
to recapture
all
marked animals each year between February and May for measuring and weighing,
365
and many adult males and females were recaptured each October
or
November too, at the start of
the breeding season. Permanent marking was achieved by hot branding with wire brands heated in
a
portable gas burner. Coloured glass beads, threaded on nylon line through the nuchal crest,
provided a second method of marking that was less permanent but allowed identification of
individuals at greater distances.
For
very rapid identification of adults during limited periods of
detailed observation, particularly over the breeding season, animals were painted with white,
orange or yellow paint on both flanks with numbers
5
cm high.
A
small patch of orange
or
yellow
paint on the neck
or
base of the tail was used to distinguish animals caught at different sites. There
was no evidence that the numbers
or
the small patches of colour affected the iguanas' behaviour.
3.3 Observations of reproductive behaviour
Intensive observations were made during each breeding season from 1981-82
to
1985-86.
Observations were made from suitable vantage points above colonies on the coast and the nesting
area, during continuous (0700-1730) daytime watches that spanned and were maintained
throughout each breeding season. These involved two
or
three observers who worked in shifts
every day for eight weeks each season. Checks were also made during the night, with some
prolonged nocturnal observations at the nesting area. The proportion of females that nested each
year was estimated from a combination of direct observations and changes in weight of females
caught at the beginning and the end of each nesting season (Laurie, in prep.).
3.4.
Measurements
of
ivuana
were captured. They included snout to vent length (SVL), tail length
(TL),
maximum head width
(HW) (at the point on the living animal with maximum width across the quadrates, just in front
of
the tympanum) and length of longest nuchal crest spine
(SL).
A
600
mm
rigid stainless steel rule
and Vernier calipers (Mitutoyo) were used for these measurements. Spring balances (Pesola) were
used to determine body weight
(WT)
to the nearest
1
g up
to
100
g,
to the nearest
10
g up
to
1,000
g, to the nearest
50
g up to 2,500 g and to the nearest 100 g above 2,500
g.
Two people are needed
to
measure an iguana accurately. There appeared
to
be
more potential
for error in the way the iguanas were held for measuring, than in the actual measuring, particularly
for
SVL.
So,
as
I
was present throughout the study,
I
always held the iguanas and a number of
different people did the measuring. Accuracy checks were carried
out
within each season using
repeat measurements of
100
animals within an hour of
fiist
measuring. The standard deviations of
measurements were: SVL, 2.2%, TL, 0.8%, HW, 2.5%,
SL,
5.0%
and WT, 0.5%; with only
small differences between size classes. The length measurements of adult males were slightly less
consistent (2.25% sd for SVL) than for smaller animals (hatchlings: 2.13%), due to small
variations in the extent to which the animals were stretched for measurement.
The
major error
appeared to be in
SVL
measurement
so
I
tested for consistency between seasons and assistants by
comparing the ratio of
SVL
to total length (SVL
+
TL)
for each size class in each year, first
discarding all measurements of animals that had lost part of their tails. If SVL and TL are assumed
Standard linear measurements were made
to
the nearest
1
mm
on each occasion that animals
366
to have independent errors and the variances are small, the standard deviation of 3.04% in
SW(SVL
+
TL) corresponds to
a
standard deviation of 2.15% in
SVL
measurement between
seasons; i.e. the same as within seasons.
3.5 Growth
actual date of capture, usually not more than one month. Differences in growth rates between
years and between
size,
sex and age classes were tested by an&jsis of variance and the t-test for
the difference between means, making full use of the different types of data involved: paired
increments (for the same individuals in
two
years), unpaired increments (for the means of different
individuals of the same size class at the beginning of the years in question) and data for hatchlings
that can be separated into each year's cohort.
Growth rates were calculated as annual rates, and corrected for differences between years in the
3.6 Survival rm
annual recapture of marked individuals, recovery of corpses (marked and unmarked) and
the
annual censuses. The best data are those on recapture of marked individuals. Estimation of
survival was based on the method of Pollock (1981) and reported by Laurie and Brown (in press
a).
Three
sources of data were available for estimating survival rates in each sex and age class:
4
RESULTS
4.1 Morta litv
.
awed with Em
Only ten corpses were recorded from the coastline of the study area between April 1981
and
October 1982 during eight months on Santa Fe, but between November 1982 and July 1983 more
than 800 corpses were recovered during six months
on
the island. Most corpses were washed
away by the sea
so
the figures indicate an enonnous mortality in a population estimated to number
less than 8,000 individuals in June 1982. The data for recaptures of marked individuals were used
to estimate annual mortality rates (April to April) for each sex and size class and cohort (Laurie and
Brown, in press a). During the 1982-83 ENSO the relative rates for each group were checked
using the data from recovery of corpses and the annual census, which, although less accurate, gave
very consistent results.
Animals began to die as a result of ENSO as early as November 1982,
so
the annual mortality
rates shown in Table
1
do
not accurately indicate the size of the effect of ENSO. Further analysis,
using November recaptures, showed that mortality over the period November 1982 to November
1983 rose from a pre-ENS0 level of 8-15% in adult males, 2-4% in adult females and 46% in
hatchlings to 58%
in
adult males,
47%
in adult females and 84% in hatchlings. These rates have
since returned towards pre-ENS0 levels (Table 1) but adults of both sexes have lower survival
than before
ENSO
and the 1985 hatchlings had
a
survival rate in the first year similar
to
hatchling
survival rate in 1982 and 1983 and considerably lower than in 1981 (Laurie and Brown, in press
a). Fig.
3
shows the percentage of animals that survived to the end of each year according to sex
(in adults) and cohort (year of emergence).
367
TABLE
1
Estimated percentage annual mortality (April to April) for each adult sex class and cohort.
COHORTS
Adultmales Adultfemales
1980 1981 1982 1983 1985
N=
464 453 140 643 404 422 722
1981-82 14.7 3.8 22.9 35.7
1982-83 7.5 1.5 11.1 47.3 60.6
1983-84 57.7 47.1 77.7 72.9 83.7 55.6
1984-85 5.3 13.6 17.5 7.1 13.1
1985-86 27.7 16.9 13.1 21.5 3.1 60.4
19 86- 87 (7.2)
-
(14.6)
-
(14.6) (10.9)
1986-87
figures
are
over-estimates (see Laurie and Brown, in press a).
The abnormal mortality started in December
1982
and continued
until
August
1983,
with the
highest mortality occurring between April and July
1983
(Laurie,
1983b).
Adult iguanas weighed
a mean of
54.2%
(s.e.
0.9%.
n
=
42)
of their normal weight at death, and were extremely
emaciated, with
no
fat reserves and considerable reduction of musculature, particularly at the base
of the tail (Cooper and Laurie,
1987).
They were very weak and in
the
last few days before death
could hardly move. Their stomachs generally contained very little: the mean weight of the contents
of
89
adults' stomachs was
17
g (s.e.
3
g, range
0-220
g) compared with a mean of
196
g (s.e.
22
g, range
95-228
g) for
6
adults' stomachs collected during normal conditions. Apart from
algae, stomach contents included small stones, pieces of crab
IGraDsus
m,
iguana skin,
iguana and sea-lion (7alophus dornianu) faeces, sea-lion hairs, and earth. These other items
were also observed being picked
up
on
land by animals apparently too weak to venture into the
water
or
the intertidal zone.
The algae present in the stomach and the intestines consisted mainly of Giffordia mitchelliap
and were largely undigested. In marked contrast
to
the normal semi-liquid state of algae in marine
iguana intestines, Giffordia was relatively
dry
and
fibrous
and remained
so
in the faeces, which
are
normally liquid and amorphous containing few recognizable parts of algae. Gross and
histopathological examination of iguanas that died during ENSO and comparison with others that
died under normal conditions indicated that the former died of starvation (Cooper and Laurie,
1987).
4.2
Changes in aleal
flora
Spermothamnium spp. had almost entirely disappeared by March
1983
and was replaced by the
brown alga Giffordia mitchelliac, which dominated the intertidal and splash zones (Laurie,
1983b).
Giffordia mitchelliae has not been recorded before in Galapagos but may have been present in
small quantities. It tolerates a wider range of temperatures than the red algae and was thus able to
The normal red algal turf, consisting of Polvsiphonium, aelidium, Centrw and
'"1
a
I
1 1
I
-7
1981 1982 1983 1984 1985 1986
YEAR
Oi-
1981 1982 1983 1984 1985 1986
YEAR
Fig.
3.
The percentage
of
animals that survived to the end
of
each year, adults according
to
sex
(34
and juveniles according to cohort, or year
of
emergence
(3b).
369
colonize the sites that the red algae had occupied previously (J. Price, pers. comm.).
organic matter digestibility of the Gjffordia was 21% compared with a mean
of
78% for the
preferred red algal species and 64% for the green algae
spp.) (Laurie and Uryu, in prep.). Brown algae generally contain more cellulose than red and
green algae (Black, 1955; Paterson, 1984) and thus would be expected to be more difficult to
digest.
The sea level and sea-surface temperatures in Galapagos had returned to the normal range for
the time of year by September 1983, and the dense mat of
Giffordia
algae had begun
to
disappear
by early November and was almost completely gone by December, being slowly replaced by red
algal turf,
immediate: there was
no
more than normal mortality after August 1983 and the surviving adults
had recovered to near their 1981 condition by November 1983 (Laurie, 1987).
In
vitro digestibilities of algae with sheep rumen fluid (Tillev and Terry, 1963) showed that the
Chaetomorpha and Enteromorpha
and Chaetomorpha spp. The response of the marine iguanas was almost
4.3 Growth ram
Mean annual relative increase in snout-vent length decreased in adults from 6.8% in 1981-82 to
0.5%
in 1982-83, increased slightly
to
1.2% in 1983-84, most of the growth being after August
1983, and then returned to pre-ENS0 levels of 6.3% in 1984-85. Juvenile growth rates recovered
much faster after the ENSO, 1983 hatchlings grew faster than 1981 hatchlings in their first year.
Figure 4 shows the growth rates of males and females separately; the mean annual increase in
SVL
(April to April) is plotted against
SVL
at the beginning of the year for 29 one cm size classes. The
data include all records of individuals captured at the beginning and the end of the year under
consideration; younger animals that did not reach 225
mm
SVL during the study were not sexed,
these juveniles of unknown sex are included in both curves. For some individuals there
are
data
for only one year; for others up to six years. Points
on
the
graphs are the means of between 8 and
178 individuals' growth rates.
There was a significant decrease in growth rates from 1981-82 to 1982-83 in all size classes
(1 1.2
>
t
>
3.4;
p
c
0.01), followed
in
1983-84 by increased growth rates in hatchlings (16.2
>
t
>
3.6; p
<
0.001), but continued low growth rates in adults (t
=
4.73; p
c
0.001 for animals of more
than 22 cm
SVL).
Adults of both sexes hardly grew at all between April 1982 and April 1983
(mean of
0.5%
increase in
SVL).
Growth in the year April 1983 to April 1984 was concentrated in
the last part of the year after the recovery of the algal flora. The 1984-85 growth rates exceeded the
1982-83 and 1983-84 rates in
all
size classes (3.0
<
t
<
13.2; p
<
0.01). For juveniles the 1984-85
growth rates were considerably higher, more than double in some size classes, than the pre-ENS0
growth rates of 1981-82 (t
=
18.7; p
c
0.001 for animals of less than 22 cm
SVL).
However,
adult growth rates in 1985-86 were similar to those
of
198 1-82, although juvenile growth rates
were still considerably higher than in 1981-82. Juvenile growth rates
in
1986-87 were lower than
in 1985-86 (4.3
c
t
<
6.2; p
c
0.01) and appear
to
be decreasing towards pre-ENS0 levels.
Figure
5
shows the growth curves for each hatchling cohort from 1980-1986 (there were
extremely few hatchlings in 1984 and none was measured
on
emergence). The highest first year
growth was recorded for 1985 hatchlings and had decreased again for the 1986 cohort. The rapid
.z
0
R
P
W
F.
9
A
0
n
=h)
DO
-4
2
D
nh)
v-
Q
v
MEAN GROWTH RATE CM
SVL
PER YEAR
0
A
h)
W
P
cn
0)
n
rn
I
b
Q
4
I-
rn
v,
U
MEAN GROWTH RATE CM
SVL
PER YEAR
0
A
h)
W
P
VI
a,
-
--
I
I'
-1
6
w
4
0
370
Fig.
4.
Mean annual increment in
SVL
plotted against
SVL
at beginning
of
the year
for
males
(4a)
and females
(4b),
1981-82
to
1986-87.
37
1
0
C
v)
1981 82 83 84 85 86 87
Year
Fig. 5. Growth curves of hatchlings of 1980, 1981,1982, 1983, 1985 and 1986 cohorts.
growth rates of the 1983 hatchlings resulted in that cohort having a greater mean size at 2 years
of
age than the 1981 hatchlings had at
3
years of age (Tables 2 and
3).
There was also a considerable
overlap in size in 1985 between 1982 and 1983 hatchlings; growth rates increased again after the
slowed growth during ENSO, showing a plasticity of response to environmental conditions.
Figure 6 shows the mean predicted growth in SVL for both sexes, based on the 1981-82 and
the 1985-86 data on annual increments in SVL
for
each one cm size class (Fig.
4).
It shows a clear
shift to the left,
so
that both males and females could be expected to reach the mean size of 1981
breeding animals about two years earlier than in 1981.
4.4
Condition
The relationship between log SVL and log WT was examined in order to find a suitable index
of condition applicable across sex and size classes. Simple regression and principal component
analyses both showed that
WT/SvL3
varied little over sex and size classes at the time of
measurement (Wong, 1985). Figure 7 shows the changes in mean condition of adult iguanas
(SVL
>
225 mm) on Santa Fe over the seven years 1981-1987. Results for other islands, with the
exception of the small islet CaamaAo, are similar (Table
4).
The clear trough in 1983, when many
animals lost almost 50% of their body weight before either dying
or
recovering, was followed by a
sharp rise in condition to well above the pre-ENS0 level, and then a return to that level. There was
a marked decrease in condition of animals well before the 1982-83 ENSO and before the
first
negative sea surface temperature anomaly of the event
(-1.OOC)
was recorded in Galapagos by
A.
372
TABLE
2
Age specific snout-vent lengths of hatchlings of different cohorts,
1981-1987.
Cohort Emergence
1
yr
2P 3P 4P 5P 6P
1981
SVL
s.e.
n
1982
SVL
s.e.
n
1983
SVL
s.e.
n
1984-1985
SVL
s.e.
n
1986
SVL
s.e.
n
a
106.7
(0.2)
643
a
107.0
(0.2)
404
ab
107.5
(0.2)
422
a
107.2
(0.1)
743
b
108.2
109
(0.4)
a
136.1
a
170.4
a
197.7
a
247.2 a288.5 297.0
(0.4)
(0.8)
(1.6) (2.0) (3.7) (5.4)
282 181 41 37 22 22
a
137.4
a
167.9
b
217.9
b
261.2
a
286.7
(0.6) (3.1) (2.5) (3.0) (5.2)
140 14 19 12 7
b
146.0
b
204.1
c
252.4
c
273.1
(0.8)
(1.2) (1.6) (2.3)
137 118 94 75
Almost no reproduction took place in
1983-84
c
162.0
c
199.4
(0.7) (1.1)
222 156
d
149.0
(0.7)
75
Within years, cohorts with different letters (a,b,c, etc.) have significantly different snout-vent
lengths (p
<
0.001).
Matson (pers. comm.,
1983)
in
May
1982.
There is an inverse correlation
(rs
=
-0.6,
p
<
0.05)
between the condition index (Fig.
7)
and the mean monthly sea surface temperature anomaly
between
1981
and
1987
(Fig.
8).
4.5
Effects on reproduction
in the
1983-84
season. On Santa Fe, territorial defence was less intense than normal, with only
25%
of the normal number of territorial males and fewer extended fights (Laurie,
1984),
but the
main difference from previous years was in the reactions of the females, who consistently avoided
the males' approaches. Not a single copulation was seen during one month of observation,
although during the same period of observation in each of the other four years of the study between
55
and
70
copulations were recorded.
In the
1981-82
and
1982-83
seasons the males finished mating by early January, but in the
1983-84
season temtorial defence continued until early March, with the difference that the territory
holders fed more frequently and lost significantly less weight than in a normal year.
The most obvious effect of ENS0 on reproduction was the almost complete failure of breeding
373
TABLE
3
Age specific weights of hatchlings of different cohorts,
1981-1987.
Cohort Emergence
1
yr
2yr
3yr 4yr 5yr 6yr
1981
WT(g)
s.e.
n
1982
WT
s.e.
n
1983
WT
s.e.
n
1984-1985
WT
s.e.
n
1986
WT
s.e.
n
a
58.8
(0.3)
643
a
59.4
(0.4)
404
b
56.5
422
(0.4)
c
62.9
(0.3)
743
c
63.5
(0.7)
109
a
121.1 a203.4 a436.8
a
851.9
a
1,288.6 1,413.6
(1.3) (3.9) (12.7) (21.3) (44.7) (81.6)
282 181 41 37 22 22
b
105.5
b
251.4
b
584.7
a
922.5
a
1,271.4
(2.1) (12.8) (19.9) (37.0) (102.9)
140 14 19 12 7
c
166.8 c468.4
c
826.4
b
1,101.2
(2.8) (8.1) (15.7) (26.0)
137 118 94 75
Almost no reproduction took place in
1983-84
d
212.7
d
423.5
(2.8) (6.8)
222 156
e
183.7
(4.4)
75
Within years, cohorts with different letters
(a,b,c,
etc.) have significantly different weights
@<O.OOl).
In later years there were many changes in reproductive rate compared with pre-ENS0 years,
described in full by Laurie (in prep.). Table
5
summarizes the main differences between pre- and
post-ENS0 years. One of the consequences of the increased post-ENS0 growth rates was that
females started breeding in their third year in
1985
compared with in their fifth year in previous
years. Furthermore, before ENSO the mean interval between breeding was approximately
2.5
years, but afterwards most females bred annually. Only
5%
of females bred in both
1981
and
1982
and
in
each of these two years about
40%
of females bred (Laurie, in prep.).
In both
1984
and
1985
the proportion of females that bred was about
87%,
more than double
than before ENSO, and the higher proportion of nesters was accompanied by an increase in mean
clutch size from
two
to three (Laurie, in prep.). On Isabela, although the sample size is very small,
mean clutch size increased from
3.0-10.0
eggs (n
=
2)
in
1982-83
to
5.3M.5
(n
=
3)
in
1984-85.
5
DISCUSSION
The
1982-83
ENSO event led to a
60-70%
reduction in marine iguana numbers on Galapagos
within six months and greatly affected the sex and age structue of the populations, with some
374
n
E
s
Males
1985-86
data
Fern
a
I
es
__---------
---
------------
1111111l111111
1
2
3
4
5
6
7
8
9
1011121314
Age
/
Years
Fig.
6.
Mean predicted growth in
SVL
for males and females based
on
1981-82
and
1985-86
data.
TABLE
4
Mean index of condition, measured as
WT/SVL3
(kg m-3) for each island at each visit.
Island
n
1981 1982 1983 1984 1985 1986 1987
YEAR
Santa Cruz
235 52.5 52.0 43.8 55.2 54.3 51.9
Caam~o
233 52.9 51.5 50.1 53.9 55.0 44.3
Plaza Sur
48 55.5 48.8 41.7 54.5 54.7
Santiago
35 55.7 48.9 42.5 56.9
Genovesa
78 56.1 45.4 62.1
Marchena
60 55.3 41.3 64.7
Pinta
80 54.9 41.7 60.5
Isabela
251 58.8 57.1 43.1 54.0 58.8 54.2
Femandina
415 57.7 49.6 39.5 53.6 58.7 55.3
Floreana
31 54.9 55.0
Santa Fe
4,168 55.0 47.1 37.2 55.8 57.0 54.1 54.2
SeymourNorte
21
57.5 60.3
Each
figure
is the mean of the condition index of three classes of iguanas: adult males, adult
females and juveniles.
sd
in each case
is
between
1.5
and
13.5
with overall
sd
=
10.8.
E
s
p a
A
o
1
a
78 54.2 39.9 53.5
classes being almost completely eliminated and hardly any breeding taking place during the
1983-
84
post-ENS0 breeding season. Since the population crash there have been large increases in rates
of growth, survival and reproduction, with growth rates and fecundity exceeding pre-ENS0
levels, but survival rates still lower. Male mortality was greater than female mortality under the
375
70-
60-
X
a
3
c
I
50-
.-
.w
.-
m
c
0
0
40-
'C
Fig. 7. Condition measured as WT/SVL3
(WT
in
kg,
SVL in m) for adult iguanas
on
Santa Fe
between April 1981 and February 1987.
conditions of low food availability, as has been found in several species of mammals and birds, for
example, in pinnipeds (Limberger, this volume) and deer
(Cervus
elaDhus)
(Flook, 1970), and this
is analysed and discussed by Laurie and Brown
(in
press a, b).
The evidence from post-mortem examination suggests that iguanas died slowly of starvation
after their preferred food species of red algae were replaced by the brown alga Giffordia ~tchellias
(Cooper and Laurie, 1987; Laurie and Uryu, in prep.). The iguanas showed typical signs of
starvation and gut impaction, and death by ingesting toxins was considered unlikely. W. Fenical
(pers. comm., 1983) has found toxins in
Bifurcana.
'
Laurencia
and
Ochta
spp.,
all
of which are
avoided by marine iguanas, but not in Giffordia spp.
5.1
Indirect comaetition
for
food
Marine iguanas
are
limited by the tides in the time they can spend feeding.
On
islands where
iguanas are larger, more sub-tidal feeding takes place (Trillmich and Trillmich, 1986), but juveniles
and many adult females are restricted to intertidal feeding. Whenever the sea is rough, animals go
without food, sometimes for several consecutive days.
Food
type, abundance and ease of grazing
81 1982 1983 1984 1985 1986 1907
YEAR
Fig.
8.
Mean monthly sea-surface temperature anomaly at Academy Bay, Santa
Cruz
between
1981
and
1987.
TABLE
5
Estimated proportion of females that bred in each year.
Year
N
%
Females
90%
confidence
nested limits
1981-82
1982-83
1983-84
1984-85
1985-86
99
157
202
206
228
41.1
40.4
1.0
87.9
86.1
34.2-52.2
30.9-58.0
78.2- 100
75.0-97.2
varies considerably between islands and between sites
on
the same island, and probably accounts
for
a large amount of the size differences between populations (Laurie,
1983a).
On Santa
Fe
in the
ENS0 year, growth rates and survival were higher on some parts of the coastline, where grazing
was easier and the lower intertidal zone approachable, than
on
steeply shelving unsheltered
377
neighbouring areas (Laurie, 1987; Laurie and Brown, in press b).
reproduction after the 1982-83 ENSO is reduction in competition for food after the return of the
normal food species and the disappearance of the colonizing Giffordia sp. Marine iguanas do not
fight at feeding sites and there is generally abundant food in the grazing areas. However,
individuals are limited in their feeding times by the state of the tide, the strength of the swell and
the temperature of the water (Carpenter, 1966; White, 1973) and there is an advantage in
occupying feeding sites easy of access, nearest to the shore or particularly well sheltered from the
waves, which can knock iguanas off exposed sites.
So
there is indirect competition for food in
that individuals have to feed on inferior feeding sites if the best ones are already occupied, and they
may have to spend relatively more of their foraging time in reaching the feeding site and clinging
on as waves pass over them, losing more heat in the process and further reducing the time available
for feeding.
the lower, more exposed sites have a lusher growth of algae. There appears
to
be an optimum
level at which to feed for each state of the sea: low enough to avoid the heavily cropped sites yet
high enough
to
avoid spending too much time and energy searching and holding on in heavy
surf.
On Santa Fe there is relatively little sub-tidal feeding (Trillmich and Trillmich, 1986), but similar
effects could result in indirect feeding competition in sub-tidal feeders also. By 1985 the algae had
returned to the normal species composition, and the fewer iguanas grazing could take more food in
a shorter time than before the 1982-83
ENSO.
Population size has returned
to
near the pre-ENS0
level by 1989, although compared
to
1981 the population composition is heavily skewed towards
juveniles
(T.
Dellinger, pers. comm., 1989).
The most likely explanation for the increase in rates of juvenile survival, growth and
The inshore and sheltered feeding sites are often grazed down very short like a lawn, whereas
5.2 Growth rates and aee at first reproduction
after the 1982-83
ENSO,
leading to a two year decrease in
the
age of first breeding. This
flexibility in growth rates, typical of reptiles (Dunham, 1978; Vogel, 1984) and the year of first
breeding being related to size rather than age, means that there is relatively fast recovery of
population size and structure after periods of high mortality.
As
the population density increased
again, particularly in juveniles, growth rates have begun to decrease.
The growth curves of iguanas of both sexes were shifted about two years forward immediately
5.3 Dominant cohorts
The almost complete loss
of
the 1982 cohort, and heavy losses to the 1980 and 1981 cohorts,
and the failure to breed in 1984 have resulted in the 1983 cohort dominating the recruitment from
the five years 1980-1984. This has important implications for investigation of whether individual
reproductive strategies have adaptive significance, or whether there is
so
much uncertainty in the
environment, both in terms
of
feeding conditions (Laurie, 1987), predation (Laurie, in prep.) and
distribution of females between mating territories (Laurie, in prep.), that any variation is a direct
response to environmental conditions (Howe, 1978; Price, 1985).
378
5.4 costs
0
f breeding
It appears that marine iguanas
are
limited in growth rates and frequency of reproduction by
food availability. Adult females did not recover breeding condition in time to breed in consecutive
years, and adult males that mated with high numbers of females in one year often spent the next
year as a non-territorial male and returned to their territories after two years (Laurie, 1984).
Skipping opportunities for reproduction has been recorded for a large number of reptiles and
amphibians (Ballinger, 1977; Bull and Shine, 1979).
5.5 Other effects oft he 1982-83 ENSO in Galapapm
Grant, 1987), but seabirds and sea mammals all suffered high mortality, failure of breeding
or
loss
of whole cohorts (Trillmich and Limberger, 1985; Gibbs et al., 1987; also see contributions by
Duffy, Limberger, and Smith, this volume.). The 1982-83 event was exceptional (Cane, 1983;
Glynn, 1988) but nevertheless important in that repeated Occurrences within a relatively short time
would be expected to lower the average population size of affected species, and could be critical for
species with small, isolated populations with a slow recovery rate such as some fishes and corals
(Glynn, 1988). The marine iguanas have shown a rapid recovery, but the fur seals
are
considerably slower, and more vulnerable to repeated occurrences of strong ENSO events
(F.
Trillmich, pers. comm.; Limberger, this volume).
Land animals thrived during the ENSO (Gibbs et al, 1984; Gibbs and Grant, 1987; Grant and
5.6
Pole of
EP
but other, weaker events have not been shown to affect the iguanas in the same way. No
widespread mortality of marine iguanas was recorded by scientists or others during previous
ENSO events, but there is evidence that some events led to a decline in condition and possibly,
therefore, reduced reproduction (G.K. Trillmich, pers. comm.). Certainly ENSO events must be
considered as potential regulators of population density
in
marine iguanas. There was another,
weaker ENSO event in early 1987 that led
to
a reduction in the standing crop of red algae and the
reappearance, at a few places high
on
the shoreline, of Giffo&
'
in February, the first
record since 1983. However,
no
increase in mortality has been recorded,
so
the event was
presumably
tea
brief
to
affect the marine iguana population.
The 1982-83 ENSO event has had profound effects
on
the population size and composition,
6
ACKNOWLEDGMENTS
I
am
very grateful to the Galapagos National Park Service (GNPS) for permission to work in
Galapagos and
to
the Charles Darwin Research Station (CDRS) for logistic support. M.
Cifuentes,
F.
Cepeda and
H.
Ochoa of the GNPS and F. Koster and
G.
Reck of the CDRS and
their staff were always ready to help. Don Ramos and the crews of many boats were invaluable in
supplying me with water and food
on
Santa
Fe.
I thank
J.
Marshall,
H.
Uryu,
R. and M. Zavala,
D. Watling, the late D. Villalba, M. Eckstein, C. Fairhurst,
T.
Woollard, M. Wells,
D.
Hams, B.
Iglesias, A. Dik, E. Cruz,
G.
Molina, C. Cevallos,
A.
Balmford,
F.
Trillmich and
T.
Dellinger for
help in the field, and D. Brown for help with analysis. The research was funded by The
379
Leverhulme Trust, The Royal Society and the Max-Planck Gesellschaft, and the data analysis was
carried out at the Large Animal Research Group of the Department of Zoology, University of
Cambridge. This is contribution number 504 of the Charles Darwin Foundation.
7 REFERENCES
Arntz, W.
E.,
1986. The two faces of El Niiio. Meeresforschung, 31: 1-46.
Ballinger,
R.
E., 1977. Reproductive strategies: food availability as a source of proximal variation
Barber,
R.
T. and Chavez, F. P., 1983. Biological consequences of El Niiio. Science, 222: 1,203-
Barber, R. T. and Chavez, F. P., 1986. Ocean variability in relation to living resources during the
Bell, T., 1825.
A
new genus of Iguanidae. Zool.
J.,
2: 204-208.
Black, W. A. P., 1955. Seaweed and their constituents in foods for man and animal.
J.
SOC.
Chem.
&
Ind., 74: 1,640-1,645.
Bull,
J. J.
and Shine, R., 1979. Iteroparous animals that skip opportunities for reproduction. Am.
Nat., 114: 296-303.
Cane, M. A., 1983. Oceanographic events during El Niiio. Science, 222: 1,189-1.194.
Carpenter, C. C., 1966. The marine iguana of the Galapagos Islands, its behaviour and ecology.
Proc. Calif. Acad. Sci., 4th Ser., 34(6): 329-376.
Chavez, F. P., 1987. The annual cycle of SST along the coast of Peru. Trop. Ocean-Atmos.
Newslet., 37: 4-6.
Colinvaux, P. A., 1984. The Galapagos climate: present and past. In:
R.
Peny (Editor), Key
Environments
-
Galapagos, Pergamon Press, Oxford, pp. 55-69.
Cooper,
J.
E. and Laurie, W. A., 1987. Investigation of deaths in marine iguanas on Galapagos.
J.
Comp. Pathol., 978: 129- 136.
Dunham, A. E., 1978. Food availability as a proximate factor influencing individual growth rates
in the iguanid lizard Scelopom merriami. Ecology, 59: 770-778.
Flook, D. R., 1970. A study of sex differential in the survival of wapiti. Can. Wildl. Serv. Rep.,
Ser. No. 11, Ottawa, 71 pp.
Gibbs, H. L. and Grant, P.R., 1987. Ecological consequences of an exceptionally strong El Niiio
event on Darwin's finches. Ecology, 68: 1,735-1,746.
Gibbs,
H.
L.,
Grant,
D.
R. and Weiland,
J.,
1984. Breeding of Darwin's finches at an unusually
early age in an El Niiio year. Auk, 101: 872-874.
Gibbs, H. L., Latta,
S.
C.
and Gibbs,
J.
P., 1987. Effects of the 1982-83 El Niiio event on blue
footed and masked booby populations on Isla Daphne Major, Galapagos. The Condor, 89:
Glynn, P. W., 1988. El Niiio-Southern Oscillation 1982-83: nearshore population, community
and ecosystem responses. Annu. Rev. Ecol. Syst., 19: 309-345.
Grant,
P.
R. and Grant B.
R.,
1987. The extraordinary El Niiio event of 1982-83: effects on
Darwin's finches on Isla Genovesa, Galapagos. Oikos, 49: 55-66.
Houvenaghel, G. T., 1978. Oceanographic conditions in the Galapagos Archipelago and their
relationships with life on the islands. In:
R.
Boje and M. Tomczak (Editors), Upwelling
Ecosystems. Springer Verlag, Berlin, pp. 18 1-200.
Houvenaghel, G. T., 1984. Oceanographic setting of the Galapagos Islands. In:
R.
Perry
(Editor), Key Environments
-
Galapagos. Pergamon Press, Oxford, pp. 43-54.
Howe,
H.
F.,
1978. Initial investment, clutch size and brood reduction in the Common Grackle
Kiladis, G. and Diaz,
H.
F.,
1984. A comparison
of
the 1982-83 and 1877-78 ENS0 events.
Laurie,
A.,
1982. Marine iguanas
-
where have all their babies gone? Noticias de Galapagos, 35:
Laurie, W.
A.,
1983a. Marine iguanas in Galapagos. Oryx, 17: 18-25.
Laurie, W. A., 1983b. An ill wind for iguanas. New Sci., 100: 108.
Laurie, W. A., 1984. Marine iguanas: the aftermath of El Niiio. Noticias de Galapagos,
40:
9-
11.
Laurie, W. A., 1987. Marine iguanas and the Galapagos marine reserve. Oceanus, 30: 54-60.
in
a lizard. Ecology, 59: 628-635.
1,210.
1982-83 El Niiio. Nature, 319: 279-285.
440-442.
(Ouiscalus-L.).
Ecology, 59: 1,109-1,122.
Trop. Ocean-Atmos. Newslet., 25: 7-8.
17-19.
380
Laurie,
W.
A. and Brown, D., In press a. Changes in annual survival rates of marine iguanas
(AmblvrhynchyS
-)
on Galapagos, and the effects of size, sex, age and fecundity in a
population crash.
J.
Anim. Ecol.
Laurie,
W.
A.
and Brown, D.,
In
press b. The factors affecting survival of marine iguanas
(Amblvrhvnchus on Galapagos.
J.
Anim. Ecol.
Paterson, I.
W.,
1984. The foraging strategy of the seaweed eating sheep of North Ronaldsay,
Orkney. Ph. D. Dissertation, University of Cambridge.
Philander,
S.
G.
H.,
1983. El Niiio Southern Oscillation phenomena. Nature, 302: 295-301.
Pollock,
K.
H.,
198 1. Capture-dependent models allowing for age-dependent survival and capture
rates. Biomemcs, 37: 521-530.
Price, T., 1985. Reproductive responses to varying
food
supply in a population
of
Darwin's
finches: clutch size, growth rates and hatching synchrony. Oecologia,
66:
41 1-416.
Quinn,
W.
H., Neal,
V.
T. and Antunez de Mayolo,
S.
E., 1987. El Nifio Occurrences over the
past four and a half centuries.
J.
Geophys. Res., 92: 14,449-14,461.
Quinn,
W.
H.,
Zopf, D.
O.,
Short,
K.
S.
and
Kuo
Yang, R. T.
W.,
1978. Historical trends and
statistics of the Southern Oscillation El Niiio, and Indonesihn droughts. Fish. Bull.,
7(3):
663-
678.
Rasmusson,
E.
M.,
1984. El Niiio: the ocean/atmosphere connection. Oceanus, 27: 5-12.
Robalino,
M.,
1985. Registros meteorologicos de la Estacion Cientifica Charles Darwin para
1982-83. In: G. Robinson and E.
M.
del Pino (Editors), El Niiio en las Islas Galapagos: El
Evento de 1982-83. Charles Darwin Foundation for the Galapagos Islands, Quito, pp. 83-90.
Tilley,
J.
M.
A. and Terry, R. A., 1963.
A
two stage technique for the in vitro digestion of forage
crops.
J.
Brit. Grassl. SOC., 18: 104-1 11.
Trillmich,
F.
and Limberger, D., 1985. Drastic effects of El Niiio on Galapagos pinnipeds.
Oecologia, 67: 19-22.
Trillmich,
G.
K.
and Trillmich,
F.,
1986. Foraging strategies of the marine iguana,
Amblvrhvnchus gristatus. Behav. Ecol. Sociobiol., 18: 259-266.
Vogel, P., 1984. Seasonal hatchling recruitment and juvenile growth of the lizard Anolis
lineatopus. Copeia, 1984: 747-757.
White,
F.
N.,
1973. Temperature and the Galapagos marine iguana
-
insights into reptilian
thermoregulation. Comp. Biochem. Physiol.,
45A:
503-513.
Wong,
K.
T., 1985. An investigation
of
growth and survival rates of marine iguanas on the
Galapagos. Diploma in Statistics Dissertation, University of Cambridge.
Wyrtki,
K.,
1985. Pacific-wide sea level fluctuations during the 1982-83 El Niiio. In: G.
Robinson, and E.
M.
del Pino (Editors), El Niiio en las Islas Galapagos: El Evento de 1982-
83. Charles Darwin Foundation for the Galapagos Islands, Quito, pp. 29-48.
