Behavioral Ecology Vol. 11 No. 5: 495-501
© 2000 International Society for Behavioral Ecology
Is egg-damaging behavior by great spotted cuckoos an accident or an adaptation?
Departamento de Biología Animal y Ecología, Facultad de Ciencias, Universidad de Granada, Spain
Address correspondence to M. Soler, Departamento de Biología Animal y Ecología, Facultad de Ciencias, Universidad de Granada, E-18071-Granada, Spain. E-mail: msoler{at}goliat.ugr.es .
Received 25 June 1999; revised 13 January 2000; accepted 13 January 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Adult great spotted cuckoos Clamator glandarius damage the eggs of their magpie Pica pica host without removing them from the nest or eating them but by producing the death of the embryo. Observations as well as experiments were used to test several predictions of two different possibilities: great spotted cuckoo egg-damaging behavior is a parasitic tactic resulting from a direct selection process (the adaptation hypothesis), or egg damage is caused by thick-shelled cuckoo eggs which evolved to avoid breakage during rapid laying (the nonadaptation hypothesis). Previously, we provided experimental evidence that egg damage increased the breeding success of cuckoos when they laid late during the laying sequence of the magpie. However, when they laid early, egg-damaging behavior did not increase cuckoo breeding success, contrary to the adaptation hypothesis. In an experimental study, when we simulated laying behavior by the great spotted cuckoo, we found that (1) the number of damaged magpie eggs was significantly lower than in natural parasitism, and (2) whereas in the experimental manipulations the number of damaged eggs did not depend on the number of magpie eggs, in natural parasitism, the number of damaged eggs increased with clutch size of the magpie. These results support the predictions of the adaptation hypothesis, implying that egg damage is not an incidental consequence of rapid egg laying, but an adaptation.
Key words: adaptation, brood parasitism, Clamator glandarius, egg-damaging behavior, great spotted cuckoo, magpie, Pica pica.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The females of many avian brood parasites (species that lay their eggs in the nest of other birds) remove or damage at least one host egg shortly before they lay their own egg (review in Soler et al., 1997
). The great spotted cuckoo Clamator glandarius
is an obligate brood parasite, which, in contrast to most other cuckoos,
parasitizes large birds, mainly members of the corvid family, and, in Europe,
especially the magpie Pica pica, with the carrion crow Corvus
corone as its secondary host (Cramp,
1985). Great spotted cuckoo eggs usually hatch before the magpies'
eggs, but, unlike other cuckoos, the chick does not eject the host's eggs or
young, although many magpie chicks frequently starve in competition for food
with the larger cuckoo chick (Cramp,
1985; Soler, 1990;
Soler and Soler, 1991). The
egg-laying behavior of the great spotted cuckoo has been discussed elsewhere
(Soler, 1990;
Soler et al., 1997).
Generally, cuckoo females lay from the rim of the nest
(Arias de Reyna et al., 1982)
and frequently damage the eggs of their hosts by either pecking, crushing, or
cracking them, always without removing or eating them
(Soler, 1990;
Soler et al., 1997). These
three types of egg damage result in desiccation and death of the embryos in
all cases.
Magpies frequently remove the most damaged eggs. Soler
(1990) found that removed eggs
had larger cracks than those not removed, and later, in an experimental study,
we (Soler et al. 1999) found
that host response varied significantly according to the type of egg damage.
Pecked eggs were removed more frequently than crushed ones, whereas cracked
eggs were never removed. However, the egg damage that most readily causes egg
removal is albumen leaking from the egg
(Soler et al., 1999). In 37.8%
of parasitized magpie nests (n = 360), there were no damaged eggs
(Soler et al., 1997). However,
in some cases this could have been due to magpies removing the most damaged
eggs. In fact, host clutches in parasitized nests were significantly smaller
than those in unparasitized nests (Soler,
1990), and clutches in parasitized nests without damaged eggs were
also significantly smaller than those in parasitized nests with damaged eggs
(Soler et al., 1997).
Egg-damaging behavior by great spotted cuckoos can be explained by two
different hypotheses: (1) the "reduction of nestling competition"
hypothesis states that egg-destruction behavior increases the survival of the
parasite chick by reducing the number of competing host chicks in the nest,
and (2) the "enhancement of hatching success" hypothesis states
that egg-damaging behavior may increase the likelihood of latelaid cuckoo
egg(s) hatching by destroying eggs that would otherwise hatch earlier than the
cuckoo egg(s) and result in terminated incubation for the cuckoo egg(s).
Recently, in an experimental study, we demonstrated that the damage is
inflicted by the parasite, and we found that egg damage increases the breeding
success of the great spotted cuckoo, both by reducing the number of competing
host chicks in the nest and by increasing the likelihood that late-laid cuckoo
eggs hatch (Soler et al.,
1997).
Brood parasites in general lay eggs with shells thicker than those of their
hosts (Brooker and Brooker,
1991; Lack, 1968;
Picman, 1997; Spaw and Rohwer,
1987). In Clamator species, eggs are particularly strong (thicker and
rounder) compared with the eggs of the hosts
(Brooker and Brooker, 1991;
Gaston, 1976). Pecked eggs bear
the only damage directly inflicted by the female great spotted cuckoo; crushed
eggs are considered the consequence of impact from the dropping parasite egg,
and cracked eggs are the consequence of host eggs being jostled against one
another as the parasite quickly leaves the nest after laying its egg
(Soler, 1990,
Soler et al., 1997). Thus, it
is not clear whether the crushing or cracking of host eggs during the laying
of the parasitic egg is an adaptation because this egg-damaging behavior could
be the consequence of hurried laying, which incidentally may have led to more
frequent destruction of host eggs, but this was not the selective force
driving the evolution in the first place
(Soler et al., 1997).
In this study we attempted to answer a basic question: is egg damage really a parasitic tactic resulting from a direct selection process (the adaptation hypothesis)? There is one alternative possibility; namely, that egg damage is incidental, caused by thick-shelled cuckoo eggs which evolved to avoid breakage during rapid laying and which are not driven by natural selection (the nonadaptation hypothesis). Parasitic cowbirds apparently have developed thick shells to resist puncture by hosts (Spaw and Rohwer 1987). However, in great spotted cuckoos thick-shelled eggs have not evolved for the same reason because this species parasitizes large birds that do not need to puncture the parasitic egg because they can easily grasp it.
The aim of this study was to test different predictions of these two
hypotheses. First, we present tests of a prediction that fail to support the
adaptation hypothesis: (A1) a positive association should exist between the
number of damaged eggs and cuckoo reproductive success. Second, we present the
results of tests of two predictions that support the nonadaptation hypothesis.
(B1) If egg damage is an incidental consequence of parasitic laying, in
multiparasitized nests, we predict that as more parasitic eggs are laid per
nest, the number of damaged host eggs will also increase. However, if egg
destruction is an adaptive behavior, then in multiparasitized nests (three or
more cuckoo eggs laid by at least two females), destruction of cuckoo eggs
laid previously (the "cuckoo competition hypothesis") will be
expected (a tendency that has been found in the European cuckoo Cuculus
canorus; Davies and Brooke,
1988). (B2) The adaptation hypothesis predicts that host-egg
damage should be greater with the larger carrion crow than with magpies
because larger nestlings pose a greater competitive threat to the cuckoo
nestling. In contrast, the nonadaptation hypothesis predicts less egg damage
with larger hosts because bigger eggs are more resistant to breakage
(Rahn and Paganelli,
1989).
Third, we present the results of an experiment in which we reproduced great spotted cuckoo damage behavior (i.e., laying from the rim of the nest and jostling the host eggs against one another with its feet when the parasite quickly leaves the nest after laying the parasitic egg). This damage experiment would allow us to determine the extent of damage one might expect from (1) the act of laying and (2) the contact of harder- and softer-shelled eggs during manipulation by the parasite, which subsequently would allow us to compare the number of damaged eggs in these experimental manipulations with those provoked by the great spotted cuckoo in naturally parasitized nests. The adaptation hypothesis predicts (C1) that, if the cuckoo purposefully damages host eggs, natural parasitism should damage more magpie eggs than our experimental manipulations reproducing the great spotted cuckoo laying behavior, and (C2) that in naturally parasitized nests the number of damaged eggs should increase as the clutch size of the host increases because as more eggs are laid by the host, the competitive threat to the cuckoo chick is greater; however, in our experimental manipulations, the clutch size of the host does not affect the number of damaged eggs. The nonadaptation hypothesis bears the opposite predictions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Field work
Observational data were collected during the breeding seasons of 1990-1995, and the damage experiment was conducted during 1997 in Hoya de Guadix, southern Spain (37°18' N, 3°11' W), a high plateau (1000 m) with sparse vegetation, cultivated cereals (especially barley), and many groves of almond trees (Prunus dulcis) and holm oaks (Quercus rotundifolia). Magpies nested in higher density in almond trees but also in oaks, whereas carrion crows showed the opposite tendency (Soler, 1990
). During the same
period we also found 21 parasitized carrion crow nests which were used to test
prediction B2.
During 1990-1994, we found 692 magpie nests, 404 of which (58.4%) were
parasitized by the great spotted cuckoo. Magpie nests were located during
usually weekly visits to our study plots from the start of the nest-building
period. The nests were inspected at least once a week. We recorded the
percentage of magpie nests parasitized in each plot, considering a nest
parasitized if it contained at least one cuckoo egg. Although we performed
several field experiments, we have only included data from nonexperimental
nests. We determined the number of damaged eggs by carefully examining all
eggs in each nest. During 1997 we monitored 96 magpie nests [40 unparasitized
and 56 (58.3%) parasitized by the great spotted cuckoo], and we recorded the
number and type of damaged magpie eggs in all parasitized nests. After a
careful inspection of damaged eggs, they were classified as pecked, cracked,
or crushed. In 11 nests that did not contain cuckoo egg, we found damaged
eggs. These nests have not been included in the analyses because presumably
they were parasitized by the great spotted cuckoo, but the parasitic eggs had
been ejected by the magpies (Soler et al.,
1997).
Experimental procedures
The egg-damage experiments were carried out during the 1997 breeding
season. We reproduced the two types of damage caused indirectly by the great
spotted cuckoo (crushing and cracking) in two different experiments (see
below). In both experiments, we used a real magpie nest that was collected
after being deserted by the magpies. The magpie eggs were taken from their
nests and put in the experimental magpie nest, where all manipulations were
made. Sometimes more than one test (between 1 and 4) were made using the same
magpie clutch (the eggs that remained without any damage), considering that
the number of magpie eggs does not affect the number of damaged eggs (see
below). In each test, we recorded the number of magpie eggs and, after careful
inspection, the number and type of damaged eggs. Also, by observing carefully,
we determined which eggs in the clutch actually received the blow.
Experiment 1
In experiment 1 we determined the extent of damage expected from the act of
laying from the rim of the nest (eggdropping experiment), which in theory
crushes host eggs (Soler,
1990). After introducing the magpie clutch into the experimental
nest, a strip of cardboard was situated on the top of the magpie nest, and an
unincubated cuckoo (or magpie) egg was dropped from the strip of cardboard
(i.e., from the rim of the nest).
Three kinds of tests were made: (1) a real cuckoo egg was dropped onto a magpie clutch, (2) a real magpie egg was dropped onto a magpie clutch (control), and (3) a real cuckoo egg was dropped onto one or more cuckoo eggs. In all these cases, we recorded number of eggs in the nest, number and type of damaged eggs, and also which egg was damaged, the one dropped or the ones receiving the blow.
Experiment 2
The aim of experiment 2 was to determine the effect of jostling host eggs
against one another and / or against the parasitic eggs (egg-jostling
experiment) as the parasite supposedly does with its feet when quickly leaving
the nest, thereby cracking eggs (Soler,
1990). This second damage experiment would allow us to determine
the extent of damage one might expect from the contact of harder-and
softer-shelled eggs during manipulation by the parasite.
The experimental manipulation consisted (after introducing the magpie clutch into the experimental nest) of jostling the eggs with a pen both when one or more cuckoo eggs were together with the magpie clutch (experimental), and when only magpie eggs were included in the experimental nest (control). During preliminary tests, we increased the strength with which we jostled the eggs because no damage occurred. We jostled the eggs strongly to damage some eggs. The strength was such that some of the eggs moved nearly to the rim of the nest. This means that this movement could not be caused incidentally by the cuckoo when it leaves the nest; rather, a movement of this magnitude must be intentional by the cuckoo. Thus, by jostling the eggs as strongly as we did, and considering the results as incidental damage, we have obtained highly conservative results.
Comparison between egg damage in naturally parasitized nests and
experimental manipulations
We compared the number of damaged eggs as a consequence of simulated cuckoo
egg-laying behavior (that is, incidental egg damage; the results of our
experimental manipulations) with those inflicted by the great spotted cuckoo
in naturally parasitized nests. For this comparison we used naturally
parasitized nests with only one cuckoo egg, and only experimental nests where
both types of experimental manipulations (egg dropping and egg ostling) were
carried out. We have quantified incidental egg damage by adding eggs crushed
and cracked during the two types of experimental manipulations and quantified
the damage caused by great spotted cuckoos by adding number of eggs crushed,
cracked, and pecked. Considering that we have experimentally demonstrated that
most damaged eggs removed by magpies are pecked
(Soler et al., 1999), the
number of pecked eggs was calculated by adding to the number in the nest the
difference between the clutch size of the magpie and the mean clutch size of
the magpie in unparasitized nests.
Statistical procedures
We followed the statistical methods of Sokal and Rohlf
(1989), and when nonparametric
statistics were needed, we used the methods described by Siegel
(1988). Residual analyses were
used to control for the effects of the number of cuckoo eggs, the number of
magpie eggs, and the number of damaged magpie eggs because these three
variables can affect the number of egg damaged in the nests and the cuckoo
breeding success. The number of damaged eggs (0, 1, 2, or more rarely 3) was
normally distributed (Kolmogorov-Smirnov test, p >.1), but this
was because of the low power of this test with small sample sizes. So, when
analyzing experimental results we have used non-parametric statistic. With our
sample sizes, Spearman correlations and Mann-Whitney U tests should
be quite powerful. Values given are means ± SEs. All tests are two
tailed.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Tests of predictions that fail to support the adaptation hypothesis
Prediction A 1
Controlling the effect of number of cuckoo eggs and number of remaining magpie eggs in the nest in a multiple regression, we found that the relationship between the residuals of the number of damaged magpie eggs and variables of cuckoo reproductive success were not significant (partial correlation coefficients: hatching success, R =.05, p =.48, n = 185; fledging success, R =.11, p =.18, n = 162; breeding success, R =.09, p =.24, n = 184). On the other hand, smaller cuckoo nestlings suffered a higher rate of postfledgling mortality compared to larger ones (Soler et al., 1994
).
Therefore, chick mass can be considered a good indicator of cuckoo
reproductive success. However, the relationship between the number of damaged
eggs and the mass of the cuckoo chicks after controlling for the number of
magpie and cuckoo chicks was not significant (partial correlation coefficient
R =.12, p =.37, n = 62). Therefore, there is no
direct link between number of damaged eggs and cuckoo reproductive success in
naturally parasitized nests.
Evidence supporting the nonadaptation hypothesis
The nonadaptation hypothesis states that egg damage is incidental and is
caused by the thick-shelled cuckoo eggs which evolved to avoid breakage during
rapid laying. We tested two predictions that involve both the adaptation and
the nonadaptation hypotheses:
Prediction B 1
If egg damage is nonadaptive, in multiparasitized nests, we predict that as
more parasitic eggs are laid per nest, the number of damaged host eggs will
also increase. As magpies remove damaged eggs, we can test the above
prediction by analyzing the relationship between the number of parasite eggs
and the number of undamaged host eggs per nest. The prediction was fulfilled
because the number of undamaged host eggs decreased with the number of
parasite eggs (rs = -.27, p <.0001, n
= 353). However, this is a nonexclusive test, and the number of damaged host
eggs could also increase for adaptive reasons as more parasitic eggs are laid
in a nest.
With respect to the alternative possibility (cuckoo competition hypothesis), damaged parasitic eggs in multiparasitized nests (three or more cuckoo eggs laid by at least two females) were found only sporadically [in 3 of 48 (6.3%) multiparasitized magpie nests], and in two of these three cases there were five or more cuckoo eggs laid by three or more different females in the nests with broken cuckoo eggs. This scarcity of damaged parasitic eggs suggests that the cuckoo competition hypothesis is not supported.
Prediction B2
The adaptation hypothesis predicts that host egg damage should be greater
for a larger corvid host (carrion crow) than for magpies because larger
nestlings pose a greater competitive threat to cuckoo nestlings. On the other
hand, the nonadaptation hypothesis predicts less egg damage for larger hosts
because they have stronger eggs. The percentage of damaged eggs was
significantly lower in carrion crow nests (16.4 ± 6.7, n = 21
nests) than in magpie nests (27.8 ± 1.6, n = 355; Mann-Whitney
U test, z = 2.16, p =.03). This is especially strong support
for the nonadaptation hypothesis considering that reproductive success of the
great spotted cuckoo reached low values when parasitizing the carrion crow
(hatching success = 82.2% ± 8.4, n = 15; fledging success =
62.5% ± 12.5, n = 14 and breeding success = 50.6% ±
12.5, n = 15) than when parasitizing the magpie (hatching success =
80.8% ± 2.3, n = 213 ; fledging success = 92.2% ± 1.3,
n = 195 and breeding success = 73.2% ± 2.4, n = 213;
Mann-Whitney U test, z = 0.26, p =.79; z = 2.89, p
=.004 and z = 1.87, p =.06, respectively), which imply that selection
pressures for egg destruction are higher when parasitizing carrion crows than
when parasitizing magpies.
Experimental egg-damage manipulations: effect of the act of
laying
When we dropped one great spotted cuckoo egg onto the magpie clutch to
simulate cuckoo laying behavior (laying from the rim of the nest), in most
tests only one magpie egg was damaged (85.1%), and 88.5% of the damaged eggs
were crushed (Table 1). The
albumen leaked out of the egg in only one of the damaged magpie eggs (1.9%),
and the cuckoo egg was never damaged; the damaged eggs were invariably those
that received the blow (Table
1). There was no significant relationship between the number of
magpie eggs used in the test and the number of damaged eggs
(rs = -.13, p =.37, n = 47).
|
When one fresh magpie egg was dropped onto the magpie clutch in control
tests, the results were similar to those found when the egg dropped was that
of a cuckoo (Table 1).
Similarly, the number of magpie eggs did not significantly affect the number
of damaged eggs (rs = -.14, p =.41, n =
37). The number of damaged magpie eggs
(Table 1; multiple loglinear
model, 
2 = 5.34, df = 2, p =.07), the type of damage
(Fisher's Exact test, p =.75) and the percentage of eggs with albumen
leaking out (Fisher's Exact test, p =.56) were not significantly
different for the two groups of tests. However, the dropped egg was more
frequently damaged when it was a magpie egg (51.5%) than when it was a cuckoo
egg (0%; Table 1; Fisher's
Exact test, p <.00001).
When a cuckoo egg was dropped onto one or more cuckoo eggs, only in one out of 12 tests (8.3%) there was any egg damaged (the one receiving the blow).
Experimental egg-damage manipulations: effect of the contact of hard
and softer eggs
When we jostled the eggs with a pen to reproduce the supposed behavior of
the parasite when quickly leaving the host nest, large differences were found,
depending on the content of the nest (Table
2). When there were only magpie eggs in the nest, only a single
egg was damaged in one test. However, when there was one or more cuckoo eggs
with the magpie clutch, significantly more magpie eggs became damaged
(Table 2; multiple log-linear
model, 2 = 16.7, df = 2, p =.002). This jostling
manipulation provoked a similar percentage of crushed and cracked eggs
(Table 2), and similar to that
of the egg-dropping manipulation, only sporadically produced egg damage with
albumen leaking out of the egg (Table
2).
|
The number of magpie eggs did not significantly affect the number of damaged host eggs (rs = -.15, p =.39, n = 35). However, as the number of cuckoo eggs in the nest increased (between 0 and 3), the number of magpie eggs that became damaged also increased (rs =.60, p =.001, n = 35).
Egg damage during natural parasitism
During 1997, the same year when the experiment was carried out, damaged
magpie eggs were found in 37 (66.1%) parasitized nests, and 35.7% of the nests
contained one damaged magpie egg. However, we also found nests with two
(21.4%), three (3.6%), and four (5.4%) damaged eggs. We recorded 62 damaged
magpie eggs that had been pecked (8.1% of the damaged eggs), crushed (48.4%),
or cracked (43.5%). Pecked, crushed, or cracked eggs were found in 10.8%,
67.6%, and 37.8%, respectively, of the parasitized nests with damaged eggs
(n = 37). In 19 parasitized nests (33.9%), no damaged eggs were
found.
As the number of cuckoo eggs per nest increased, the number of magpie eggs decreased (rs = -.28, p =.002, n = 96), and the number of damaged magpie eggs increased (rs =.50, p <.0001, n = 96).
Comparison between egg damage in naturally parasitized nests and
experimental manipulations
Crushed eggs were more common in the experimental manipulations (1.18
± 0.48, n = 28) than in naturally parasitized nests with only
one cuckoo egg (0.60 ± 0.65, n = 25; Mann-Whitney U
test, U = 182.5, p =.003). However, the average number of
cracked eggs per nest was similar in naturally parasitized nests (with only
one cuckoo egg) (0.44 ± 0.82, n = 25) and in the experimental
manipulations (0.39 ± 0.63, n = 28; Mann-Whitney U
test, U = 343, p =.9).
In agreement with the first prediction of the adaptation hypothesis (C1), we found that natural parasitism by great spotted cuckoo damaged more magpie eggs (2.3 ± 1.4, n = 25) than did our experimental manipulations simulating what could be incidental damage during laying (1.6 ± 0.63, n = 28; Mann-Whitney U test, U = 239, p <.05).
In agreement with the second prediction of the adaptation hypothesis (C2), in our experimental manipulations the clutch size of the host did not significantly affect the number of damaged magpie eggs (egg-dropping experiment: rs = -.13, p =.37, n = 47; egg-jostling experiment: rs = -.15, p =.39, n = 35, and rs =.006, p =.74, n = 36, respectively, for tests with and without cuckoo eggs in the nest), but in naturally parasitized nests with only one cuckoo egg, the number of damaged host eggs increased as clutch size of the host increased (rs =.41, p =.031, n = 25).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Adaptation is an important concept in evolutionary theory. Adaptations are always the consequence of natural selection. Beneficial consequences that are independent of natural selection are not adaptations, and all known adaptations are in principle explicable by natural selection (Ridley, 1993
). However,
adaptation has frequently been overestimated; it has sometimes been regarded
as the sole influence on evolution, and evolutionary biologists have tended to
view natural selection as absolutely dominant on evolutionary mechanisms,
assuming that it explains almost any phenotypic difference; however, this
assumption is not accurate (Ridley,
1993; Travisano et al.,
1995).
Experimental egg-damage manipulations
Crushed magpie eggs have been considered the consequence of host eggs being
struck by the parasite's egg (Soler,
1990; Soler et al.,
1997), as the cuckoo lays from the rim of the host nest
(Arias de Reyna et al., 1982).
In agreement with this idea, we found that when a cuckoo egg is dropped on a
magpie clutch, usually one magpie egg is crushed, although sometimes one
magpie egg may be cracked (Table
1). Cracked eggs have been considered the result of the host eggs
being struck by the feet of the parasite when quickly leaving the nest
(Soler, 1990;
Soler et al., 1997). However,
in our egg-jostling experiment, we found that this egg manipulation provokes
egg cracking as well as egg crushing. Thus, we can conclude that egg dropping
usually crushes eggs and may crack some, whereas egg jostling both cracks and
crushes eggs.
In the experimental manipulations, crushed eggs were more common than in
naturally parasitized nests with only one cuckoo egg. Given that a cuckoo egg
dropped onto a magpie clutch almost always damages at least one magpie egg
(Table 1), that egg jostling
also results in some crushed eggs (Table
1), and that crushed eggs are only rarely removed from the nest by
magpies (Soler et al., 1999),
we can conclude that cuckoo females lay from the rim of the nest in
approximately 50% of the cases (there were magpie eggs crushed in 13 out of 25
magpie nests parasitized with one cuckoo egg).
In the egg-dropping experiment, no cuckoo egg was ever damaged, but when a magpie egg was dropped onto the magpie clutch (control tests), the one dropped was damaged frequently (Table 1). In the egg-jostling experiment, when there was no cuckoo egg in the magpie clutch, usually no magpie eggs were damaged (Table 1), even when the eggs were jostled very strongly (see Methods). This was an expected result because magpie eggs should be adapted to resist natural jostling against each other by the wind. Egg-jostling manipulations caused egg damage when there was at least one cuckoo egg in the magpie clutch, and thus this damage must be the consequence of contact between harder- and softer-shelled eggs; in fact, as more cuckoo eggs were present, more magpie eggs were damaged.
All these results support the idea that the great spotted cuckoo, like
other avian brood parasites, lays eggs with shells thicker than those of their
hosts (Brooker and Brooker,
1991; Lack, 1968;
Spaw and Rohwer, 1987). According to Brooker and Brooker
(1991), in Clamator
species, eggs are thicker and rounder compared with host eggs. This strong
shell protects the great spotted cuckoo egg from damage when it collides with
the hosts eggs, but it also damages more magpie eggs (egg-jostling
experiment). However, this strong shell could also protect the parasitic egg
from damage if the nest is multiply parasitized
(Brooker and Brooker, 1991). In
the egg-jostling experiment, when more than one cuckoo egg was introduced into
the magpie clutch, no cuckoo eggs were ever damaged, and in the egg-dropping
experiment, when one cuckoo egg was dropped on another cuckoo egg, only in 1
test out of 12 was a cuckoo egg damaged
(Table 1). Thus, the strong
shell of the great spotted cuckoo egg, besides causing substantial egg damage
to the magpie clutch, protects the parasitic egg from damage both when it
collides with the host eggs and when other cuckoo eggs are present in multiply
parasitized nests.
The strong shell of the cuckoo egg is responsible for part of the egg
damage caused by the cuckoo. However, it is unclear whether the damage to host
eggs is an adaptation. If cuckoo eggs became more thick shelled as a way of
destroying host eggs, it would be an adaptation, but perhaps cuckoos started
to lay thick-shelled eggs because they were in a hurry when laying, and
females with thick-shelled eggs less often suffered breakage of their own
eggs. This incidentally led to more frequent destruction of host eggs, but
this was not the selective force driving the evolution in the first place. If
this was the case, we would be dealing with an exaptation rather than an
adaptation (Gould and Vrba,
1982; but see Skelton,
1992, for problems with terminology).
The basis for the exaptation scenario described above (rapid laying) is a
characteristic frequently reported in brood-parasitic females: parasitic eggs
are laid quickly, usually in only a few seconds (see review in
Sealy et al., 1995). This
rapid laying has frequently been suggested to be an adaptation for brood
parasitism, as recently shown by Sealy et al.
(1995). Rapid laying is
adaptive because it allows parasitic females to avoid detection by hosts
(Davies and Brooke, 1988;
Scott, 1991;
Sealy et al., 1995) and
thereby provides parasitic females with some important advantages by reducing
the chance of (1) host attacks and possible injury
(Briskie and Sealy, 1987;
Sealy et al., 1995;
Soler, 1990), (2) the
possibility of attracting other parasite females or predators
(Sealy et al., 1995;
Wyllie, 1981); and (3) a
careful nest examination by the host which would increase the likelihood of
parasitic-egg rejection (Davies and Brooke,
1988; Moksnes et al.,
1993; Rothstein,
1975; Sealy et al.,
1995; but see Soler et al.,
1999).
Evaluation of the adaptation and nonadaptation hypotheses
A previous experimental study (Soler et
al., 1997) demonstrated that egg-damaging behavior in the great
spotted cuckoo is advantageous to the chicks of the parasite when the magpie
has laid numerous eggs before parasitism. A second experiment revealed that
increasing reproductive success of the cuckoo by egg damage is the consequence
of both reduction of the number of competing host chicks in the nest and
increased probability that late-laid cuckoo eggs will hatch
(Soler et al., 1997). This
information alone provides strong support for the idea that egg destruction is
an adaptive strategy which increases the reproductive success of the great
spotted cuckoo (Reeve and Sherman,
1993). However, great spotted cuckoos also destroyed magpie eggs
when laying early during the breeding cycle of the magpie. Contrary to the
prediction of the reduction of nestling competition hypothesis, the breeding
success of the great spotted cuckoo was high and similar both in nests
parasitized naturally (with host-egg destruction) and experimentally (by us,
without host-egg destruction) at the beginning of the laying sequence of the
magpie (Soler et al.,
1997).
We have some information that suggests that egg-damage behavior could not
be maintained by natural selection: the result of the above mentioned
experiment, which did not support the prediction of the reduction of nestling
competition hypothesis, the prediction A1 (this study), which also failed in
this respect, and two other predictions (A2 and A3) that can be tested
considering previously published results
(Soler et al 1997).
Under prediction A2, egg destruction could not be considered an incidental
behavior if the frequency of pecked eggs is high compared with crushed and
cracked eggs (crushed and cracked eggs could be an incidental consequence of a
parasite's visit, but pecked eggs could not). However, pecked eggs accounted
only for 14.1% of the damaged eggs (n = 241) and were found in 18.5%
of the parasitized nests (n = 157). Crushed and cracked eggs,
respectively, represented 55.2% and 30.7% of the damaged eggs (n =
241) appearing in 60.5% and 31.2% of the nests with damaged eggs (n =
157) (Soler et al., 1997).
Thus, the frequency of the different types of egg destruction does not support
the adaptation hypothesis.
Under prediction A3, if egg damage was an adaptive behavior, as adult
cuckoos also destroy magpie eggs when laying early in the laying sequence of
the magpie (Soler et al.,
1997), then egg damage during early laying should also increase
the breeding success of the cuckoo. However, this was not the case when the
experiment was made simulating laying of a parasite egg during the breeding
cycle of the host because the reproductive success of the cuckoo was similar
in naturally and experimentally parasitized nests
(Soler et al., 1997). Even the
opposite tendency to that predicted was found in most cases, but differences
were far from significant (Soler et al.,
1997). Thus, the results do not support the above prediction, and
egg-damaging behavior of the great spotted cuckoo does not increase the
breeding success of the cuckoo when the egg of the parasite is laid early
during the breeding cycle of the host (one to three host eggs already present
in the nest). This possibility that egg-damage behavior could not be
maintained by natural selection was confirmed by two predictions which
distinguished between the adaptation and nonadaptation hypotheses (B1 and B2,
see Results).
Adaptation is demonstrated by observed conformity to a priori design
specifications (the a priori method of recognizing adaptations), and this is
what we have done by testing the previously mentioned predictions. This method
consists of searching for the reason why a character is favored by natural
selection (Maynard Smith, 1993;
Williams, 1992). This method
of studying adaptation works well if we are studying adaptation, but the
problem is that, for many characters, it is not obvious how (or whether) they
are adaptive (Ridley, 1993),
and not all differences between animals can be explained as adaptive
(Maynard Smith, 1993). If the
character under study is an adaptation, then it must exist because of natural
selection, and it is correct to persist in looking for the reason it is
favored. However, if the character is not favored by natural selection, the
method breaks down (Maynard Smith,
1993; Ridley,
1993). Another problem of this method of studying adaptation is
that, frequently, it is assumed that animals will always be perfectly adapted
and that every detail of a tactic makes adaptive sense
(Ridley, 1993). However,
adaptations may be imperfect because natural selection cannot operate as fast
as the environment of a species changes or because the mutations that would
enable perfect adaptation have not arisen
(Ridley, 1993).
Perhaps these problems of the a priori method of studying adaptations are responsible for the fact that our results in the above-mentioned tests have not supported the predictions. However, these tests are based on comparisons between individuals, and the results are not conclusive. Rather, an experimental approach is needed to draw stronger conclusions. In our experimental manipulations, we found that when we simulated the laying behavior of the great spotted cuckoo the number of damaged magpie eggs was significantly lower than in natural parasitism. In addition, although the number of eggs damaged did not depend on the number of magpie eggs in the experimental manipulations, the number of damaged eggs increased as magpie clutch size increased in natural parasitism. These results support the predictions of the adaptation hypothesis (C1 and C2), implying that egg damage is not an incidental consequence of rapid egg laying, but an adaptation.
Over the history of evolutionary biology, some traits have been described
as nonadaptive; however, in all cases after the appropriate analysis it was
concluded that the characteristic under study was controlled by natural
selection and thus adaptive (Ridley,
1993). This is what has happened in our study concerning great
spotted cuckoo egg-damaging behavior: although some tests of predictions did
not support the adaptation hypothesis, the behavior turned out to be adaptive
after a detailed experimental study.
|
ACKNOWLEDGEMENTS |
|---|
We are most grateful to Tim Birkhead, Michael Brooker, Juan P. Martínez, Juan Moreno, Pascual Rivas, Spencer Sealy, Amotz Zahavi, and two anonymous referees and specially Anders Pape Møller for valuable comments on different versions of the manuscript. We are particularly grateful to Juan J. Soler for his help with the field work, for assistance with various aspects of this study, and for a constructive review of the manuscript. Financial support was given by the Commission of the European Communities (SC1*-CT92-0772) and DGI-CYT (PS90-0227 research project).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Arias de Reyna L, Recuerda P, Corvillo M, Aguilar I, 1982. Reproducción del críalo (Clamator glandarius) en Sierra Morena Central. Doñana Acta Vert 9: 177-193.
Briskie JV, Sealy SG, 1987. Responses of least flycatchers to experimental inter- and intraspecific brood parasitism. Condor 89: 899-901.
Brooker MG, Brooker LC, 1991. Eggshell strength in cuckoos and cowbirds. Ibis 133: 406-413.
Cramp S, 1985. The birds of the Western Paleartic, vol. IV. Oxford: Oxford University Press.
Davies NB, Brooke M de L, 1988. Cuckoos versus reed warblers: adaptations and counter-adaptations. Anim Behav 36: 262-284.
Gaston AJ, 1976. Brood parasitism by the pied crested cuckoo Clamator jacobinus. J Anim Ecol 45: 331-348.
Gould SJ, Vrba E, 1982. Exaptation—a missing term in the science of form. Paleobiology 8: 4-15.[Abstract]
Lack D, 1968. Ecological adaptations for breeding in birds. London: Methuen.
Maynard Smith J, 1993. The theory of evolution. Cambridge: Cambridge University Press.
Moksnes A, Røskaft E, Korsnes L, 1993.
Rejection of cuckoo (Cuculus canorus) eggs by meadow pipits
(Anthus pratensis). Behav Ecol
4: 120-127.
Picman J, 1997. Are cowbird eggs unusually strong from the inside? Auk 114: 66-73.
Rahn H, Paganelli CV, 1989. Shell mass, thickness and density of avian eggs derived from the tables of Schönwetter. J Ornithol 130: 59-68.
Reeve HK, Sherman PW, 1993. Adaptation and the goals of evolutionary research. Q Rev Biol 68: 1-32.
Ridley M, 1993. Evolution. Boston: Blackwell Scientific.
Rothstein SI, 1975. An experimental and teleonomic investigation of avian brood parasitism. Condor 77: 250-271.
Scott DM, 1991. The timing of day of egg laying by the brown-headed cowbird and other icterines. Can J Zool 69: 2093-2099.
Sealy SG, Neudorf DL, Hill DP, 1995. Rapid laying by brown-headed cowbirds Molothrus aler and other parasitic birds. Ibis 137: 76-84.
Siegel S, 1988. Estadística no paramétrica aplicada a las ciencias de la conducta. Mexico: Trillas.
Skelton PW, 1992. Adaptation. In: Paleobiology: a synthesis (Briggs DEG, Crowther PR, eds). Oxford: Blackwell Scientific; 139-146.
Sokal RR, Rohlf FJ, 1989. Biometry. San Francisco: W.H. Freeman.
Soler M, 1990. Relationships between the great spotted cuckoo Clamator glandarius and its magpie hosts in a recently colonized area. Ornis Scand 21: 212-223.
Soler M, Palomino JJ, Martinez JG, Soler JJ, 1994. Activity, survival, independence and migration of fledgling great spotted cuckoos. Condor 96: 802-805.
Soler M, Soler JJ, 1991. Growth and development of great spotted cuckoos and their magpie host. Condor 93: 49-54.
Soler M, Soler JJ, Martinez JG, 1997. Great spotted cuckoos improve their reproductive success by damaging magpie host eggs. Anim Behav 54: 1227-1233.[ISI][Medline]
Soler M, Soler JJ, Pérez-Contreras T, 1999. The cost of host egg damage caused by a brood parasite: experiments on great spotted cuckoos (Clamator glandarius) and magpies (Pica pica). Behav Ecol Sociobiol 46: 381-386.
Spaw CD, Rower S, 1987. A comparative study of eggshell thickness in cowbird and other passerines. Condor 89: 307-318.
Travisano M, Mongold JA, Bennett AF, Lenski RE, 1995.
Experimental tests of the roles of adaptation, chance, and history in
evolution. Science 267:
87-90.
Williams GC, 1992. Natural selection: domains, levels, and challenges. Oxford: Oxford University Press.
Wyllie I, 1981. The cuckoo. London: B.T. Batsford Ltd.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||