Wiley and British Ecological Society are collaborating with JSTOR to digitize, preserve and extend access to Journal of Animal Ecology. http://www.jstor.org Wiley British Ecological Society Environmental and Demographic Correlates of Intraspecific Nest Parasitism in Lesser Snow Geese Chen caerulescens caerulescens Author(s): D. B. Lank, E. G. Cooch, R. F. Rockwell and F. Cooke Source: Journal of Animal Ecology, Vol. 58, No. 1 (Feb., 1989), pp. 29-44 Published by: British Ecological Society Stable URL: http://www.jstor.org/stable/4984 Accessed: 07-10-2015 00:27 UTC 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 [email protected]. This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTC All use subject to JSTOR Terms and Conditions
Transcript
Wiley and British Ecological Society are collaborating with JSTOR to digitize, preserve and extend access to Journal of Animal Ecology.
http://www.jstor.org
WileyBritish Ecological Society
Environmental and Demographic Correlates of Intraspecific Nest Parasitism in Lesser Snow Geese Chen caerulescens caerulescens Author(s): D. B. Lank, E. G. Cooch, R. F. Rockwell and F. Cooke Source: Journal of Animal Ecology, Vol. 58, No. 1 (Feb., 1989), pp. 29-44Published by: British Ecological SocietyStable URL: http://www.jstor.org/stable/4984Accessed: 07-10-2015 00:27 UTC
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 [email protected].
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
ENVIRONMENTAL AND DEMOGRAPHIC CORRELATES OF INTRASPECIFIC NEST PARASITISM IN LESSER SNOW GEESE CHEN CAERULESCENS CAERULESCENS
BY D. B. LANK, E. G. COOCH, R. F. ROCKWELL* AND F. COOKE
Department of Biology, Queen's University, Kingston, Ontario, Canada K7L 3N6
SUMMARY
(1) Intraspecific nest parasitism was the source of 5-3% + 0.02 (S.D.) of the goslings hatched annually at the La Perouse Bay snow goose colony near Churchill, Manitoba, between 1969 and 1986. Yearly values ranged between 1-8 and 9-3%.
(2) The rate of parasitism was low in years when nest-sites were readily available at the start of laying, and high when fewer sites were clear of snow or water. High parasitism rates did not occur in years of high nest failure during laying. This suggests that more parasitism occurs as a response to a lack of suitable nest sites than occurs as a consequence of nest failure.
(3) Nest parasitism was higher at times when a higher than average proportion of young birds did not attempt to nest.
(4) Nest parasitism had no measurable effect on the hatching success of host eggs. On average, the hatching success of parasitic eggs was lower than that of host eggs, primarily due to poor timing of laying by parasites relative to the onset of incubation.
(5) Due to physiological constraints on total clutch size, nest parasitism in arctic nesting geese is largely a 'salvage' female reproductive tactic, used by females that have a low likelihood of nesting successfully. This contrasts with the situation in temperate-nesting ducks and many other species of birds, where nest parasitism has the potential to be part of a mixed female reproductive strategy that may produce higher reproductive success than that obtained by laying the optimal number of eggs in a female's own nest.
INTRODUCTION
Intraspecific nest parasitism, the laying of eggs in nests attended by conspecifics, has been documented in many families of birds (Yom-Tov 1980), and is especially common among waterfowl species (Weller 1959). We know little of the conditions affecting the balance between parasitic and non-parasitic laying in a population, however, or how the probability of parasitic laying varies among individuals. Parasitic and non-parasitic laying may be maintained as reproductive tactics by processes analogous to those maintaining variation in male reproductive tactics (Dunbar 1982; Austad 1984). First, nesting and parasitism might be equivalent-payoff, genetically differentiated strategies, maintained by such mechanisms as negative-frequency dependent or disruptive selection (Gadgil 1972; Slatkin 1979). Second, equivalent-payoff tactics could be maintained
* Present address: Department of Biology, City College of New York, New York, New York 10031 U.S.A. and Department of Ornithology, American Museum of Natural History, New York, New York 10024 U.S.A.
29
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
through developmental or seasonal phenotypic plasticity in the absence of genetic differences (Austad 1984; Gross 1984). Third, nest parasitism might be one tactic in a 'mixed reproductive strategy' (sensu Trivers 1972), in which females enhance their reproductive success by both nesting themselves and parasitizing the nests of others. This enables females to obtain additional or lower cost parental care than they could provide for their own brood, and under certain conditions, mixed strategists might benefit slightly by spreading the risk of total reproductive failure (Rubenstein 1982). Males of many avian species pursue the analogous mixed strategy of pair-bonding plus extra-pair copulation (McKinney, Cheng & Bruggers 1984; Birkhead, Atkin & Moller 1987). Finally, whether a female lays an egg in her own or another's nest may be conditional on the availability of reproductive resources to her (Dawkins 1980; Eberhardt 1982). For females with fewer or lower quality resources, parasitic laying may be more profitable than attempting to nest with little chance of successfully rearing young.
Direct observations have documented mixed strategies in a few species (Brown 1984; Gibbons 1986; Moller 1987), and conditional parasitism by lower-status females in others (Emlen & Wrege 1986), but there are no examples of equivalent-payoff parasitic and non- parasitic strategies of egg-laying (Yom-Tov 1980; Andersson 1984; Emlen & Wrege 1986). In this paper, we draw inferences about the nature of intraspecific nest parasitism in a colony of lesser snow geese, Chen caerulescens caerulescens (Linnaeus 1758). We describe temporal variation in the occurrence of parasitism in relation to environmental and demographic variables, and examine the effect of parasitism on aspects of nesting success. This approach highlights the mechanisms responsible for parasitism in our
population and enables evaluation of their reproductive significance. Finally, we examine our results in light of current information on the distribution and dynamics of nest
parasitism in geese, waterfowl, and birds in general.
METHODS
Data were gathered at the La Perouse Bay snow goose colony (58?24'N, 94?24'W), 40 km east of Churchill, Manitoba, near the southern edge of the species' breeding range, from 1969 to 1986. General field methods are described in Finney & Cooke (1978), Rockwell, Findlay & Cooke (1983), and Cooke & Rockwell (1988).
Measures of nest parasitism We measured parasitism in several ways, each of which had strengths and limitations
for use in different analyses.
Abnormally large clutch size For comparison with parasitism indices computed in other studies, we determined the
proportion of clutches thought to be too large to have been laid by a single female (seven or more eggs, Ankney & Macinnes 1978, but see Rockwell, Findlay & Cooke 1987).
Unusual laying sequences To obtain information on parasitism at the egg-laying stage, we tabulated unusual
laying events at nests that were checked daily from the laying of the first egg until 3 days after the onset of incubation, and the eggs were numbered. Cases where more than one egg appeared in a nest per day, new eggs were found outside the nest, or eggs were laid after a gap of 2 or more days at the end of laying were taken as evidence of parasitic events. These
30
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
criteria underestimated parasitism, since eggs laid in normal sequence would not be detected. To compute an index of parasitism frequency, the number of parasitic events detected at each nest was divided by the number of opportunities for detection. The number of opportunities was the clutch size minus 2 (since detection of parasitism by the above criteria was unlikely for the first two eggs, and parasitism could not be detected after the last egg), minus additional detection opportunities lost due to any missed daily visits. Nest sequences were available from 235-521 nests per year for 12 years (1973-86; insufficient data from 1979).
Annual parasitism rates based on plumage genetics Snow geese have two colour morphs: blue and white. Basic plumage colour is
controlled by a single Mendelian locus, with the blue allele (B) incompletely dominant to white (b) (Cooke & Cooch 1968). Incomplete dominance enables identification of probable homozygous individuals (Rattray & Cooke 1984). Goslings are also dimorphic, but heterozygotes are not identifiable. This natural genetic marker allowed identification of colour 'mismatched' goslings that could not be produced by homozygous pairs of breeders. White goslings hatching in nests of homozygous dominant (BB x BB) pairs could not be offspring of either member of the pair (barring spontaneous mutation), and may thus be attributed to nest parasitism. Blue goslings hatching from nests of homozygous recessive (bb x bb) white pairs may be the result of parasitism, but such goslings also could arise through extra-pair fertilizations of females by sperm carrying a B allele. By comparing observed and expected parental exclusion rates calculated from pairs of homozygous dominants and homozygous recessives, we estimated that extra-pair fertilization accounted for 30% of such goslings produced (1973-86) (Lank et al. 1989). When estimating annual parasitism rates, we attributed the remaining 70% of blue goslings in nests of white pairs to parasitism. Data were insufficient to allow use of year- specific corrections for extra-pair fertilization rate in these estimates.
For comparisons among years, we calculated an annual parasitism rate, our best estimate of the parasitic proportion of goslings hatching each year. Data on the colour of nest attendants and the number and colour of hatched goslings were recorded from large samples (> 500) of nests each year. These provided our most accurate and longest-term information on annual variation in nest parasitism (1969-86). Annual parasitism rate (F) was computed as the sample-size (n) weighted average of the parasitism rate estimated from nests of presumed homozygous dominant pairs (FBB), and that estimated from nests of homozygous recessives (Fbb):
F= (FBB* nBB + Fbb* nbb)/(nBB + nbb)
The parasitism rate for BBxBB pairs was estimated as: FBB=(1-X)/(1--p), where x = the proportion of blue goslings in the sample nests and p = the proportion of all blue goslings produced at the colony for the year in question. The division by 1 -p corrects for undetected blue parasitic goslings (Cooke & Mirsky 1972). Although the data are not contaminated by goslings produced by extra-pair copulation, the sample sizes from nests of BB x BB pairs were barely adequate to produce accurate estimates of parasitism (x = 297 + 144 (S.D.), ranging from 59 to 545 goslings per year). Thus, FBB alone was not a useful measure of parasitism rate. Since about 70% of breeding adults at La Perouse Bay were white, and mating was largely assortative with respect to colour (Cooch & Beardmore 1959; Cooke & Davies 1983), pairs of homozygous recessive geese made up the bulk of the nesting sample (x = 2814 + 1142 (S.D.), ranging from 919 to 5029 goslings
D. B. LANK et al. 31
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
per year). For bb x bb nests, we computed Fbb, analogous to FBB, as: Fbb = 07x/p (Cooke & Mirsky 1972), where the 0-7 corrected for goslings attributed to extra-pair fertilizations, and dividing by p corrected for undetected parasitic white goslings. Calculation of both rates assumed that parasitic females produced the same colour ratio of goslings as the colony as a whole, and that the rate of parasitism in BB x BB and bb x bb nests were similar, both of which appear to be the case (Lank et al. 1989). FBB and Fbb were significantly correlated (r= 0-63, n = 18, P= 0-005).
Number of nests affected For comparing relative parasitism frequency within seasons, we used the proportion of
BB x BB and bb x bb nests where mismatched goslings hatched. Finally, we estimated the number of nests affected by parasitism by correcting the proportion of homozygous pair- type nests where mismatched goslings hatched by detection rates based on gosling colour ratios, the clutch-size specific proportion of multiple parasitic events detected per nest, and the fraction of detections attributed to extra-pair copulations.
Female geese could incubate unrelated eggs for several reasons other than nest parasitism. Two females might commence laying in the same nest, and one subsequently evict the other. Pairs might take over an abandoned nest that already contains eggs, or actively displace earlier residents. Our sequence measures included the first phenomenon, and our colour measures included all of these events.
Nesting chronology We looked for changes in relative parasitism frequency among nests hatching eggs that
were initiated in the early, middle and late parts of each season (1973-86). The mean initiation date and two adjacent days comprised the middle period; the early period preceded and the late periods followed these dates. This procedure produced relatively even sample sizes among the periods. The initiation date of each nest was estimated by subtracting a year and clutch-size specific correction factor from its hatch date. Correction factors were derived from clutch-size specific multinomial regressions of each year's sample of nests with known initiation and hatch dates. A simpler method (Hamann 1983) was used for 1979, where an insufficient number of known initiation dates precluded the regression approach.
We examined correlations of annual parasitism rates with the mean, variance and skew of initiation dates.
Potential environmental correlates
We correlated measures of precipitation and temperature with annual parasitism rates. Meteorological data were recorded by Environment Canada at the airports at Winnipeg and Churchill, Manitoba. Data from Winnipeg were used because conditions on migration correlate with aspects of breeding at the colony (Davies & Cooke 1983). We tested data from a set of time periods to search for critical windows that might affect parasitism rates. Meteorological variables computed were: mean temperatures and total precipitation for Winnipeg and Churchill in April, and for Churchill for the following time periods: all of May, 1-15 May, 15 days prior to and including mean initiation, the week of mean initiation, 9-5 days prior to mean initiation, 5 days prior to and including mean initiation, 4 days including and after mean initiation, 5-10 days after mean initiation, and from 17 days prior through 3 days after mean initiation. Precipitation data were highly non-normal with positive skew, and were log square-root transformed prior
32
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
to analysis. In addition to mean values, we computed a series of 'deviation variables', which were the absolute value of each year's deviation from the mean of annual means. Regressions using these deviations tested whether response variables changed in unusual years, regardless of the direction of the deviation from normal. Finally, we also calculated an accumulated snow variable, which consisted of the snow depth on the ground at the end of April plus additional snowfalls from 1 May to the mean nest initiation date.
We ranked years in terms of the rate that nest-sites became available once laying had
begun. Persistent snow cover or flooding associated with rapid snow melt or ice jams were the major causes of nest-site limitation. During the middle years of the study (1973-80) the rankings were derived from the proportions of dry ground tabulated along transects
prior to and during the laying period (Finney 1975). Annual rankings were taken from Finney (1975), Abraham (1980), and unpublished annual reports and personal observa- tions since 1980. We used a 1-5 scale ranging from highly restricted to total availability of nest-sites.
Potential demographic correlates
The number of breeding pairs (colony size) was estimated with Jolly-Seber capture- recapture methods, using data on females only, from annual post-breeding banding drives (Healey 1985; T. A. Armstrong, personal communication). We also calculated an index of the proportion of surviving young females (2- and 3-year-old birds) that bred at the colony each year (Appendix) and correlated this index with parasitism rate.
Measures of clutch size and reproductive success
We examined the correlation of annual parasitism rate with mean clutch size, an adjusted mean clutch size corrected for annual parasitism rate (see below), the proportion of six-egg nests, and the proportion of seven-or-more-egg nests. The large clutch sizes were used as possible indices of parasitism rate (see above).
We examined the correlation of annual parasitism rate with three measures of colony- wide reproductive success that might plausibly affect or be affected by nest parasitism. Preincubation nest failure, calculated only from nests visited daily, was the proportion of these nests that failed prior to clutch completion. Incubation failure was the proportion of nests reaching incubation that failed to hatch one or more eggs. Failure at either stage could be due to nest predation or abandonment. Hatching success, calculated only for nests hatching one or more young, was the proportion of eggs laid that hatched and left the nests as goslings.
Statistical methods
We examined a variety of potential annual correlates of nest parasitism rate. Except where noted, we have used all of the years available for each variable. SAS (SAS Institute 1985) was used for data manipulation and to calculate most of the statistics presented. Multidimensional contingency analyses (MDCA) were performed with the P4F subrou- tine of BMDP (Dixon 1985). The BIOM programs (Rohlf 1986) were used for Jack- knifing and for calculation of principal axes in correlation analyses. Pearson's correlation coefficient (r) was reported from most correlations. The squared correlation coefficient (r2) between observed and predicted values was reported as a summary statistic in regression analyses. Spearman's correlation coefficient on ranks (rs) was used for correlations with habitat availability, an ordinal-level variable, and with other variables not normally distributed. Suspected outliers in univariate distributions were tested using
D. B. LANK et al. 33
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
FIG. 1. Annual intraspecific nest parasitism rate at the La Perouse Bay snow goose colony, 1969-86.
Dixon's test (Sokal & Rohlf 1981). Suspected outliers in bivariate correlations were tested against the distribution of Jack-knifed z-transformed correlation coefficients using Dixon's test (F. J. Rohlf, personal communication).
RESULTS
Correlates of annual parasitism rates
The estimated annual parasitism rate averaged 5-30 ? 002 (S.D.) between 1969 and 1986, ranging from 2-4 to 9-9% (Fig. 1). Parasitism affected 22% of hatching nests, ranging from 12 to 31%. The La Perouse Bay snow goose colony grew from about 2000 to 9000 pairs over the course of the study, but parasitism rate did not change systematically over the years or correlate with colony size (regression of parasitism rate on year: r2=0.03, n= 18, N.S.; correlation with estimated number of nesting pairs: r=-0 12, n= 16, N.S.).
Parasitism rates were higher in years when nest-sites became available more slowly after the start of laying (Fig. 2, Table 1). The nest-site availability index was negatively correlated with the total accumulated snow through mean laying date, total precipitation throughout April to mid-May, and with the deviation in temperature just after mean laying (Table 1). The temperature variable was not correlated with either precipitation variable. Thus, nest-site availability was lower when more snow accumulated prior to nesting of the geese, and when unusual temperatures occurred during the latter half of laying. Unusually cold temperatures would maintain snow cover, while unusually warm temperatures in conjunction with accumulated snow, or rain, would result in restriction of nest sites due to flooding.
None of the twenty-five basic meteorological variables examined correlated signifi- cantly with parasitism rate. Parasitism rate correlated with two of the deviation variables: positively with deviations from mean precipitation in the 2 weeks prior to mean initiation date (r = 0-71, n = 18, P = 00009), and negatively with deviation in precipitation over the 5 days following mean inition ( = 6 n = 1 P = reation (r=65, 18, P=0004). Precipitation during the
34
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
FIG. 2. Parasitism rates as a function of nest-site availability, 1970-86. Numbers next to each point indicate the year.
TABLE 1. Spearman correlations of nest-site availability with annual parasitism rate, the proportion of nests with white attendants that hatched blue goslings within early, middle, and late nest initiation periods, annual rates of preincubation nest failure, accumulated snow as of the annual mean laying date, total precipitation 1 April-15
May, and mean temperature during 5 days after mean laying date
All years Excluding 1977
Variable rs n P rs n P Annual parasitism rate -0-51 17 0-03 -0-75 16 <0-001 Early -0-14 14 0-64 -0-40 13 0.18 Middle -0-21 14 0-47 -0-45 13 0-13 Late -0-56 14 0-03 -0-79 13 0.001 Preincubation nest failure 0-60 13 0-03 0-52 12 0.09 Accumulated snow -0-56 17 0-02 -0-47 16 0-07 1 April-15 May precipitation -0-58 17 0-02 -0-55 16 0-02 Temperature post-mean laying -0-65 17 0-004 -0-60 16 0-02
former period was almost entirely snow. The two precipitation variables were themselves
negatively correlated (r= -0-61, n=18, P=0-007). Postlaying precipitation was not
significant in a multiple regression model when entered after pre-laying precipitation, but both terms were significant (P< 0-01) when postlaying precipitation was entered first. This suggests that postlaying precipitation was significant only because of its correlation with prelaying precipitation.
There was no systematic difference in parasitism in years with early or late mean nest initiation dates. We expected the skew in laying dates to reflect relative nest-site
availability at the start of laying, but correlations with measures of laying skew or variance were not significant. Skew measures themselves also did not correlate with nest- site availability. Parasitism rate was not significantly correlated with mean clutch size, nor with clutch size adjusted for laying date. It was significantly correlated with the annual
35
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
TABLE 2. Seasonal variation in the proportion of nests with two white attendants containing blue goslings. Backdated nest initiation dates were used to assign nests to time periods (see Methods). Contingency analysis tested for heterogeneity among periods each year (d.f. = 2), and the Mantel-Haenszel X2 statistic (MH) tested for a seasonal trend (d.f. = 1). Weighted mean percentages are given at the bottom, and the G value over all years is for the parasitism x period term of a full-model MDCA. There
proportion of large clutches (> 6 eggs, r= 065, n= 14, P = 001). The lack of relationship with total clutch size is unexpected, since parasitism should increase individual clutch sizes. This effect would be masked if clutch sizes were lower for other reasons when parasitism is higher. To test this, we adjusted annual clutch sizes by subtracting the proportion of eggs attributed to successful parasitism from the annual mean. Parasitism rate was still not significantly correlated with adjusted clutch sizes, but the trend was negative as predicted (r= -0-50, n= 14, P =0 07).
One year was a clear exception to the general negative relationship between parasitism rate and habitat availability (Fig. 2). Due to a lack of snowfall, all potential nest-sites were exposed when geese arrived at the colony in 1977, yet the parasitism rate was the highest ever recorded. The correlation between parasitism rate and the index of nest-site availability is much stronger without this year (rs= -0-75, n= 16, P<0 001 vs. r= - 0-51; Table 1), which is a statistical outlier from the bivariate distribution of other years (Dixon's test, r22 = 076, d.f. = 17, P < 0-005). 1977 has been identified previously as a drought year at migratory staging areas (Davies & Cooke 1983), where geese store nutrients for use in egg production and incubation (Ankney & Macinnes 1978). The estimated proportion of 2- and 3-year-old females that nested (Appendix) in 1977 was far lower than the estimates for any other year (0-23 in 1977 versus the mean of 0-66 + 0-072 (S.D.) for 1973-84, excluding 1977; 1977 was a statistical outlier from the proportions in other years: Dixon's test, r2 =0.62, d.f.= 12, P<002). Exclusion of 1977 improved correlations between habitat availability and several measures of parasitism, but had no such effect on the correlations between habitat availability and meteorological variables (Table 1). We conclude that unlike other years, the high parasitism rate in 1977 was not largely due to restricted nest-site availability. It may have been associated with the failure of a large number of young birds to nest successfully.
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
TABLE 3. Correlations and partial correlations of the relationships between annual rates of intraspecific nest parasitism (F), preincubation nest failure, and habitat availability. The data set (n= 12) included 1973-86 except for 1977, which was an exceptional point (see text) and 1979, when no estimate of preincubation nest failure
was available
Simple Partial correlation correlation
Fx preincubation nest failure -0-64 -0-44 Fx nest-site availability -0-82 -0-75 Preincubation nest failure x nest-site availability 0.51 -0-03
37
Variation within years One or more mismatched goslings were found more frequently in nests initiated earlier
in the season, averaging 7-3, 5-8, and 4-8% for early, middle, and late parts of the nest initiation distribution (Table 2). Differences between periods were significant in only 4 of 14 years, but there was a significant overall 'parasitism x period' term, controlling for
years (MDCA: G = 32-31, d.f. = 2, P < 00001, no significant three-way interaction). The
early period had the highest rate in 9 out of 14 years, while the late period was lowest in 9
years. While these results show that early nests have a greater risk of parasitism, the
pattern does not necessarily mean that parasitic laying was more frequent earlier in the season. The timing of laying by parasites relative to host incubation is poor (Lank et al.
1989), as evidenced by significant numbers of unhatched eggs (see below). Nests begun early are exposed longer to potential parasites than nests begun later. Eggs laid during the middle or late part of the season could be laid into nests initiated early in the season, for
example. This analysis does not demonstrate differences in the timing of laying by nesting and parasitic females.
We compared nest-site availability with the proportion of nests hatching mismatched
goslings initiated during the early, middle, and late periods of the laying curve. All three correlations were negative. However, the only significant relationship was in the late
period (Table 1). Later-laying females, which include a higher proportion of young breeders (Finney & Cooke 1978; Hamann & Cooke 1987), may have been more sensitive to environmental conditions.
The mean age of attendant females found at nests that hatched tended to be higher in
periods when nest parasitism was higher (14 years x 3 periods=42 points, r=0-29, P = 007; to eliminate the problem of using older samples of known age birds each year, this analysis included only birds 2--5 years of age). When year and period were controlled for, there was a significant effect of age on parasitism (two-way ANCOVA, no significant interactions among year, period and mean-age terms, F of age term=4 61, d.f.=2, P=0-04). This indicates that parasitism was higher at times when fewer young birds initiated nests that survived to hatching.
Parasitism and annual reproductive success
We expected parasitism rates to be higher when rates of preincubation nest failure were
higher. Parasitism might increase the probability of preincubation failure, or birds that
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
lost their nests might lay the remainder of their clutch parasitically. To our surprise, parasitism was higher in years when preincubation nest failure was lower, with the prominent exception of 1977, when both rates were high (all years: r= -0-15, n= 13, N.S.; without 1977: rs= -0-64, n= 12, P=0-03; 1977 is an outlier from the bivariate distribution of values in other years: Dixon's test, r2i =0'73, d.f. = 13, P<0-005). The negative relationship (for years other than 1977) suggests that, in general, nest parasitism was not a major cause of preincubation failure, and that parasitism following nest failure did not account for a large fraction of all nest parasitism.
Preincubation failure also correlated, positively, with nest-site availability (Tables 1, 3). When parasitism rate was controlled for, however, the partial correlation coefficient was reduced to nil (Table 3), suggesting that the apparent relationship was a spurious consequence of the correlation between parasitism rate and nest failure. In contrast, the relationships between parasitism rate and the two other variables were much less affected when controlling for third variables. Since habitat availability is a rank-ordered variable, we cannot report meaningful significance values for these partial correlations. Nonethe- less, this analysis suggests that habitat availability influences parasitism rates, and that parasitism rate is associated with variation in nest failure rates, but habitat availability is not directly associated with nest failure rates.
Parasitism rate was uncorrelated with the failure rate of nests during incubation (r= 0-13, n = 14, N.S.). This implies that nest loss due to parasitism is a small contributor to nest failure at La Perouse Bay. However, parasitism rate was strongly negatively correlated with annual hatching success (r= -067, n= 13, P=0-013). To determine whether the poorer annual hatch was attributable to poor timing or placement of presumed parasitic eggs, we used the sequence irregularity index from nests checked daily. The index was strongly negatively correlated with annual hatching success (r= -0-79, n = 13, P= 0-001), and also with the hatching success of the sequenced nest sample only (r = - 077, n = 13, P = 0 002). When we removed parasitic eggs laid too late to hatch and those laid outside of nests from the calculation of hatching success, however, the relationship was much weaker, and not statistically significant (known egg sequence nests: r= -0-36, n= 13, N.S.). Thus, poorly timed or placed parasitic eggs have a lower hatching success than non-parasitic eggs, and we cannot detect a correlation between the level of parasitism and the hatching success of the remaining eggs.
DISCUSSION
Intraspecific nest parasitism accounted for 53 % of the young hatching annually at the La Perouse Bay snow goose colony, varying from 1 8 to 9 3% between 1969 and 1986 (Fig. 1). Since 22% of the nests were parasitized in an average year, nest parasitism will affect most nesting females during their reproductive lifetimes (a median of three seasons, from the year of first breeding; Cooke & Rockwell 1988). The rate of parasitism increased in years when nest-site availability was restricted by snow or water (Fig. 2, Table 1; Finney 1975). Nest parasitism in snow geese appears to be largely a conditional reproductive tactic used by females that might have nested had nest sites been readily available.
In 1 year out of 18, the parasitism rate was high despite readily available nest-sites (Fig. 2). Other unusual aspects of this year included a severe drought at migratory staging areas (Davies & Cooke 1983), an unusually low proportion of young females nesting successfully, unusually small clutches, and a high rate of nest failure. We infer that despite
38
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
nesting opportunities, nest parasitism became the primary breeding tactic for many females due to their relatively poor physiological condition.
Parasitic laying is an inefficient method of obtaining parental care for eggs. Parasites frequently lay outside the nest cup, and eggs must then be rolled in by the attendant female (Lank et al. 1989). Although adopted parasitic eggs and host eggs laid synchronously had similar hatching success, eggs left outside of the nest and those laid after hosts had begun incubation hatched at lower rates. Despite the strong selective advantage of appropriate placement and timing, parasites often fail to obtain care for their eggs. Unless nest parasites obtain fecundity or survival advantages that offset their lower hatching success per egg, parasitism is generally a poorer reproductive tactic than nesting. Despite this, for individuals facing a high probability of nest failure, and perhaps for those laying small or partial clutches, the net reproductive benefit from parasitism may be higher than that from attempting to nest. The negative relationship between rates of parasitism and preincubation nest failure (Table 3) suggests that a component of the population regularly switches between two low-return reproductive tactics. Younger females appear to be especially likely to face this choice (Lank et al. 1989).
Nest parasitism following partial clutch loss or abandonment logically should increase total parasitism rate. Cooch (1958), working at Boas River on Southampton Island (63?42'N, 85?45'W), described individual females being unable to lay in prepared nests due to morning flooding of their nest-sites, and speculated that females in this situation might lay their eggs parasitically. Our findings show that parasitism following nest failure contributes relatively little to annual variation in parasitism rate.
The adoption of parasitic tactics highlights the importance of breeding synchrony and coloniality as constraints on the reproductive tactics of female snow geese. At La Perouse Bay, all geese could obtain nest-sites by waiting or moving away from the colony. With regard to time, lower gosling survivorship and recruitment from later broods selects against waiting to nest (Findlay & Cooke 1982; Cooke, Findlay & Rockwell 1984), and earlier parasitic laying provides a higher probability of synchronizing eggs with those of the nesting female. Factors selecting for colonial breeding in this species are unknown. At more northerly locations, where the bulk of snow geese breed, however, the option of moving to unoccupied habitat is precluded in years when all snow-free areas are utilized (Syroechkovksiy 1979; Bousfield & Syroechkovskiy 1985). Colonial breeding not only increases opportunities for parasitism (Alexander 1975; Brown 1984; Moller 1987), it also constrains parasites from searching for nest sites in other locations.
At the snow goose colony on Wrangel Island, north of Siberia, nest parasitism is of greater magnitude, and even more variable annually, than at La Perouse Bay (Syroechkovskiy 1979; Bousfield & Syroechkovskiy 1985). Snow-free colony area ranged between 125 and 2600 ha among 13 years, the percentage of adult pairs at the colony that nested ranged from 4 to 100%, and the average clutch size of incubated nests ranged from 3-7 when much area was clear to 5-9 when habitat was restricted. In an extremely restricted year, at least 82% of nests were parasitized, 20-30% of all eggs laid were parasitic, and an additional 20-30% of the eggs were laid outside nests, often in piles of 25-30 or more 'dumped' eggs. Long, thin eggs, attributed to first-time breeders, were especially common in such piles. Large dump nests have also been reported in certain years at other arctic locations (Barry 1956; Prevett, Lieff & Macinnes 1972) but are extremely rare at La Perouse Bay.
We failed to find higher rates of parasitism when birds bred later in the season, as has been suggested from observations of frequent nest parasitism in one or a few late years
39
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
(Prevett, Lieff& Macinnes 1972; F. G. Cooch, personal communication). Wrangel Island snow geese also show no such relationship (date of the beginning of peak laying versus proportion of nests larger than 6: r= -010, n=10, N.S., Ye. V. Syroechkovskiy, personal communication). The hypothesized relationship between late breeding and parasitism may reflect a slow rate of nest-site emergence in certain years when snow melt was late.
Strategies of nest parasitism among birds
Pure parasite strategies Intraspecific nest parasites that achieve equivalent reproductive success to the most
productive nesting females have not yet been found in birds. Annual reproductive equivalence for parasites would require either perfect transfer efficiency of eggs to host nests or increased egg production by parasites to offset losses. Equal lifetime reproductive success might be achieved with a lower annual success by parasites, however, if they avoided any increased risk of mortality associated with nesting or brood rearing. In species examined thus far, however, pure parasites would be at a clear annual disadvantage (Yom-Tov 1980; Andersson 1984; Emlen & Wrege 1986). Snow geese may also fit this pattern, as loss of parasitic eggs is extensive. A purely parasitic female might be able to transfer resources normally used during incubation into egg production (Ankney & Macinnes 1978), but it is not yet clear whether this would be sufficient to compensate for egg-losses. While we cannot exclude pure parasites as a source for the baseline level of annual parasitism, most parasitism is conditional on nest site availability and female body condition.
Mixed strategies Parasitism as part of a mixed strategy to enhance nesting has been documented in wood
ducks (Clawson, Hartman & Fredrickson 1979; Heusmann, Bellville & Burrell 1980), two species of colonial-nesting swallows (Brown 1984; Moller 1987), moorhens Gallinula chloropus L. (Gibbons 1986), and is suspected for redheads Aythya americana (Eyton 1838) and numerous other species of waterfowl (Weller 1959). Mixed strategies are more likely to occur in species where the number of offspring that may be reared by a female (or pair) is not limited at the egg-laying stage. Moorhen females may lay parasitically prior to nesting, despite a lower hatching success for parasitic eggs, primarily because of poor timing of parasitism with regard to host incubation (Gibbons 1986). Gibbons speculates that parasitism is adopted as a second best laying tactic until a female's mate achieves sufficient body condition to sustain incubation. If so, nest parasitism may be the best option available to females while they wait for their mates to gain weight, just as female snow geese may parasitize while they wait for the snow to melt.
In altricial species, where clutch size may be adjusted to match the demands of post- hatching care, egg production may be relatively cheap. Among colonial swallows, females pursuing mixed strategies obtain substantially more parental care for their young by laying 'extra' eggs in nearby nests (Brown 1984; Moller 1987). However, the higher host cost of rearing young in such species creates a stronger selective advantage for defence against parasitism than in precocial species (Yom-Tov 1980; but see Andersson 1984 for a different view). In white-fronted bee-eaters, where parasitism appears limited to a salvage strategy (Emlen & Wrege 1986), clutch sizes are small, rearing young is difficult, and anti- parasite behaviour such as nest guarding is well developed.
A mixed strategy is unlikely in snow geese because females can adequately incubate
40
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
more eggs than they can afford to lay, as shown by the success of parasitic eggs laid coincident with those of the host. Arctic-nesting geese store and transport nutrients to the breeding grounds to support both egg production and incubation (Ryder 1970), which
places a clear physiological limit on clutch size. Since parasitic laying involves a substantial risk of egg loss, selection should favour nesting females laying in their own nests.
Salvage tactics Parasitism as a low-return salvage tactic has been suggested for nest-site limited species
(e.g. wood ducks Aix sponsa L. (Jones & Leopold 1967), and other cavity-nesting waterfowl; starlings Sturnus vulgaris L., see Andersson 1984), and for white-fronted bee- eaters Merops bullockoides L. (Emlen & Wrege 1986), where breeding opportunities are limited by ecological and social dynamics associated with cooperative breeding.
A relationship between salvage parasitism and constrained breeding opportunities seems clear, as is a relationship with coloniality. However, snow geese lack several other
predicted attributes of parasitic species. Large clutch sizes, long laying seasons and a lack of breeding synchrony have all been suggested as ecological factors favouring parasitism (Yom-Tov 1980; Andersson 1984; Gibbons 186). In snow geese, clutch size limitation
prevents geese from relaying a full clutch following partial clutch loss, which may favour parasitism as a laying tactic for the remaining eggs. The short, synchronous laying season favours laying parasitically rather than waiting for nest sites to become available.
Intraspecific nest parasitism in waterfowl
Nest parasitism in many species of north temperate zone ducks provides a revealing contrast to arctic geese. Females in both groups make a substantial energetic investment in egg production and nesting, and the young in both groups are precocial. While the net reproductive effect of parasitism on hosts has not yet been measured for any duck or goose (Amat 1987), potential costs to the hosts at the laying and incubation stage (Andersson & Eriksson 1982; Pienkowski & Evans 1982; Amat 1987; but see Rohwer 1985) may be totally offset by the advantages of larger family size during the brood stage (Raveling 1970; Nudds 1980; Eadie & Lumsden 1985; Gregoire 1985). In both cases, therefore, selection for defence against parasitism may be relatively weak.
Many species of ducks have more potential than geese for pursuing a mixed strategy. While factors limiting clutch size in ducks are less clear than those in geese (Rohwer 1984), many species obtain nutrients for egg production from their immediate environment and regularly produce second clutches. The most interesting cases are those of cavity species such as wood ducks, where nest-site limitation may be a factor. The nature of nest parasitism in wood ducks is controversial and complex (cf. Andersson 1984; Semel & Sherman 1986). On one hand, local populations often increase when nest boxes are provided (e.g. McLaughlin & Grice 1952; Jones & Leopold 1967; Haramis & Thompson 1985), suggesting a shortage of natural sites. On the other hand, parasitism persists, even at high levels, despite unoccupied nest boxes (e.g. Semel & Sherman 1986). These apparently contradictory observations are resolved if wood duck populations contain both mixed and salvage strategy parasites. Providing additional nest sites would facilitate nesting, but parasitism would nonetheless persist among mixed strategists and females who, without sufficient reserves to incubate nests, adopt a salvage strategy. Direct observations suggest that this is the case (Clawson, Hartman & Fredrickson 1979).
41
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
This analysis was possible because a great many persons have collected and collated data since 1969. To all of them, our thanks. Ye. V. Syroechkovskiy provided information on Wrangel Island snow geese, T. A. Armstrong provided estimates of colony size for recent years; we thank him and F. C. Rohwer for discussions contrasting nest parasitism in snow geese with that in ducks. F. J. Rohlf provided statistical advice on Jack-knifing and outlier detection in correlation analyses. The manuscript benefited from the comments of an anonymous adaptationist reviewer. Generous discretionary computer grant funds from Computer and Communications Services of Queen's University facilitated our analyses. Snow goose research at La Perouse Bay has been supported by the Natural Sciences and Engineering Research Council of Canada, the Canadian Wildlife Service, the Manitoba Department of Renewable Resources, the Wildlife Management Institute, the Mississippi and Central Flyway Councils, and the City College of New York.
REFERENCES
Abraham, K. F. (1980). Breeding site selection of lesser snow geese. Unpublished Ph.D. thesis, Queen's University, Kingston, Ontario.
Alexander, R. D. (1975). The evolution of social behaviour. Annual Review of Ecology and Systematics, 5, 325- 383.
Amat, J. A. (1987). Is nest parasitism among ducks advantageous to the host? American Naturalist, 130, 454- 457.
Andersson, M. (1984). Brood parasitism within species. Producers and Scroungers (Ed. by C. J. Barnard), pp. 195-228. Chapman & Hall, New York.
Andersson, M. & Eriksson, M. O. G. (1982). Nest parasitism in goldeneyes Bucephala clangula: some evolutionary aspects. American Naturalist, 120, 1-16.
Ankney, C. D. & Macinnes, C. D. (1978). Nutrient reserves and reproductive performance of female lesser snow geese. Auk, 95, 459-471.
Austad, S. N. (1984). A classification of alternative reproductive behaviors and methods for field testing ESS models. American Zoologist, 24, 309-319.
Barry, T. W. (1956). Observations of a nesting colony of American brant. Auk, 72, 193-202. Birkhead, T. R., Atkin, S. D. & M0ller, A. P. (1987). Copulation behaviour of birds. Behaviour, 101, 101-138. Bousfield, M. A. & Syroechkovskiy, Ye. V. (1985). A review of Soviet research on the lesser snow geese on
Wrangel Island, U.S.S.R. Wildfowl, 36, 13-20. Brown, C. R. (1984). Laying eggs in a neighbour's nest: benefits and cost of colonial nesting in swallows. Science,
224, 518-519. Brownie, C., Anderson, D. R., Burnham, K. P. & Robson, D. S. (1985). Statistical Inference from Band Recovery
Data-A Handbook, 2nd edn. United States Fish and Wildlife Service Research Publication 15. Clawson, E. L., Hartman, G. W. & Fredrickson, L. H. (1979). Dump nesting in a Missouri wood duck
population. Journal of Wildlife Management, 43, 347-355. Cooch, F. G. (1958). The breeding biology and management of blue goose Chen caerulescens. Unpublished Ph.D.
thesis, Cornell University, Ithaca, New York. Cooch, F. G. & Beardmore, J. A. (1959). Assortative mating and reciprocal differences in the blue-snow complex.
Nature, 183, 1833-1834. Cooke, F. & Cooch, F. G. (1968). The genetics of polymorphism in the snow goose Anser caerulescens
caerulescens. Evolution, 22, 289-300. Cooke, F. & Davies, J. C. (1983). Assortative mating, mate choice and reproductive fitness in snow geese. Mate
Choice (Ed. by P. Bateson), pp. 279-295. Cambridge University Press, Cambridge. Cooke, F., Findlay, C. S. & Rockwell, R. F. (1984). Recruitment and the timing of reproduction in lesser snow
geese (Chen caerulescens caerulescens). Auk, 101, 451-458. Cooke, F. & Mirsky, P. J. (1972). A genetic analysis of lesser snow geese families. Auk, 89, 863-871. Cooke, F. & Rockwell, R. F. (1988). Reproductive success in a lesser snow goose population. Reproductive
Success (Ed. by T. Clutton-Brock), pp. 237-250. University of Chicago Press, Chicago. Davies, J. C. & Cooke, F. (1983). Annual nesting productivity in snow geese: prairie droughts and arctic springs.
Journal of Wildlife Management, 47, 291-296. Dawkins, R. (1980). Good strategy or evolutionary stable strategy? Sociobiology: Beyond Nature/Nurture (Ed.
by G. W. Barlow & J. Silverberg), pp. 331-367. American Association for the Advancement of Science, Washington D.C.
42
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
Dixon, W. J. (1985). BMDP Statistical Software. University of California Press, Berkeley. Dunbar, R. I. M. (1982). Intraspecific variations in mating strategy. Perspectives in Ethology, Vol. 5 (Ed. by P. P.
G. Bateson & P. Klopfer), pp. 385-431. Plenum Press, New York. Eadie, J. McA. & Lumsden, H. G. (1985). Is nest parasitism always deleterious to goldeneyes? American
Naturalist, 126, 859-866. Eberhardt, W. G. (1982). Beetle horn dimorphism: making the best of a bad lot. American Naturalist, 119,420-
426. Emlen, S. T. & Wrege, P. H. (1986). Forced copulations and intra-specific parasitism: two costs of social living in
the white-fronted bee-eater. Ethology, 71, 2-29. Eyton (1838). A monograph of the Anatidae. London. Findlay, C. S. & Cooke F. (1982). Synchrony in the lesser snow goose (Anser caerulescens caerulescens). II. The
adaptive value of reproductive synchrony. Evolution, 37, 724-734. Finney, G. H. (1975). Reproductive strategies of the lesser snow goose, Anser caerulescens caerulescens.
Unpublished Ph.D. thesis, Queen's University, Kingston, Ontario. Finney, G. & Cooke, F. (1978). Reproductive habits in the snow goose: the influence of female age. Condor, 80,
147-158. Gadgil, M. (1972). Male dimorphism as a consequence of sexual selection. American Naturalist, 106, 574-580. Gibbons, D. W. (1986). Brood parasitism and cooperative nesting in the moorhen, Gallinula chloropus.
Behavioral Ecology and Sociobiology, 19, 221-232. Gregoire, P. E. J. (1985). Behavior offamily and other social groups in wintering and migrating lesser snow geese.
Unpublished. M.Sc. thesis, University of Western Ontario, London, Ontario. Gross, M. R. (1984). Sunfish, salmon, and the evolution of alternative reproductive strategies and tactics in
fishes. Fish reproduction: Strategies and Tactics (Ed. by G. Potts & R. Wootton), pp. 55-75. Academic Press, London.
Hamann, J. (1983). Intra-seasonal clutch size reduction in lesser snow geese. Unpublished M.Sc. thesis, Queen's University, Kingston, Ontario.
Hamann, J. & Cooke, F. (1987). Age effects on clutch size and laying dates of individual female lesser snow geese Anser caerulescens. Ibis, 129, 527-532.
Heramis, G. M. & Thompson, D. Q. (1985). Density-production characteristics of box-nesting wood ducks in a northern greentree impoundment. Journal of Wildlife Management, 49, 429-436.
Healey, R. F. (1985). Estimating population parameters of a nesting snow goose colony using the Jolly-Seber method. Unpublished M.Sc. thesis, Queen's University, Kingston, Ontario.
Heusmann, M. W., Bellville, R. & Burrell, R. G. (1980). Further observations on dump nesting by wood ducks. Journal of Wildlife Management, 44, 908-915.
Jones, R. E. & Leopold, A. S. (1967). Nesting interference in a dense population of wood ducks. Journal of Wildlife Management, 31, 221-227.
Lank, D. B., Mineau, P., Rockwell, R. F. & Cooke, F. (1989). Intraspecific nest parasitism and extra-pair copulation in lesser snow geese. Animal Behaviour, (in press).
McKinney, F., Cheng, K. M. & Bruggers, D. J. (1984). Sperm competition in apparently monogamous birds. Sperm Competition and the Evolution of Animal Mating Systems (Ed. by R. L. Smith), pp. 523-545. Academic Press, New York.
McLaughlin, C. L. & Grice, D. (1952). The effectiveness of large-scale erection of wood duck boxes as a management procedure. Transactions of the North American Wildlife Conference, 17, 242-259.
M0ller, A. P. (1987). Intraspecific nest parasitism and anti-parasite behaviour in swallows, Hirundo rustica. Animal Behaviour, 35, 247-254.
Nudds, T. D. (1980). Canvasback tolerance of redhead parasitism: an observation and an hypothesis. Wilson Bulletin, 92, 414.
Pienkowski, M. W. & Evans, P. R. (1982). Clutch parasitism and nesting interference between shelducks at Aberlady Bay. Wildfowl, 33, 159-163.
Prevett, J. P., Lieff, B. C. & Macinnes, C. D. (1972). Nest parasitism at McConnell River, N.W.T. Canadian Field-Naturalist, 86, 369-372.
Rattray, B. & Cooke, F. (1984). Genetic modeling: an analysis of a colour polymorphism in the snow goose (Anser caerulescens). Zoological Journal of the Linnaean Society of London, 80, 437-445.
Raveling, D. G. (1970). Dominance relationships and agonistic behavior of Canada geese in winter. Behaviour, 37, 291-319.
Richards, M. H. (1986). A demographic analysis of lesser snow goose band recoveries. Unpublished M.Sc. thesis, Queen's University, Kingston, Ontario.
Rockwell, R. F., Findlay, C. S. & Cooke, F. (1983). Life history studies of the lesser snow goose (Anser caerulescens caerulescens). I. The influence of time and age on fecundity. Oecologia, 56, 318-322.
Rockwell, R. F., Findlay, C. S. & Cooke, F. (1987) Is there an optimal clutch size in snow geese? American Naturalist, 130, 839-863.
Rohlf, F. J. (1986). BIOM A package of statistical software to accompany the text Biometry. Applied Biostatistics, Inc., Seatauket, New York.
Rohwer, F. C. (1984). Patterns of egg-laying in prairie ducks. Auk, 101, 603-605.
D. B. LANK et al. 43
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
Rohwer, F. C. (1985). The adaptive significance of clutch size in prairie ducks. Auk, 102, 354-361. Rubenstein, D. I. (1982). Risk, uncertainty and evolutionary strategies. Current Problems in Sociobiology (Ed. by
King's College Sociobiology Group), pp. 91-111. Cambridge University Press, Cambridge. Ryder, J. P. (1970). A possible factor in the evolution of clutch size in Ross' goose. Wilson Bulletin, 82, 5-13. SAS Institute (1985). SAS version 5 Edition. SAS Institute Inc., Cary, North Carolina. Semel, B. & Sherman, P. W. (1986). Dynamics of nest parasitism in wood ducks. Auk, 103, 813-816. Slatkin, M. (1979). The evolutionary response to frequency- and density-dependent interactions. American
Naturalist, 114, 384-398. Sokal, R. R. & Rohlf, F. J. (1981). Biometry. W. H. Freeman, San Francisco. Sulzbach, D. & Cooke F. (1978). Elements of nonrandomness in mass-captured samples of snow geese. Journal of
Wildlife Management, 42, 437-441. Syroechkovskiy, Ye. V. (1979). The laying of eggs by white geese into strange nests. Zoologicheskiy Zhurnal, 58,
1033-1041 (in Russian). Trivers, R. L. (1972). Parental investment and sexual selection. Sexual Selection and the Descent of Man (Ed. by
B. Campbell), pp. 136-179. Aldine Publishing Company, Chicago. Weller, M. W. (1959). Parasitic egg laying in the redhead (Aythya americana) and other North American
anatidae. Ecological Monographs, 29, 333-365. Yom-Tov, Y. (1980). Intraspecific nest parasitism in birds. Biological Reviews, 55, 93-108.
(Received 10 December 1987)
APPENDIX
Calculation of the proportion of surviving young (2- and 3-year-oldfemales) that attempt to breed
A large-scale banding operation during the post-breeding moult provided marked samples of each year's gosling cohort, and band recoveries of these birds enabled estimation of year-specific survival rates for goslings and older birds (Brownie et al. 1985; Richards 1986). The number of potential 2-year-old breeding females among the banded sample each year (N2) was estimated as:
N2= {(Ng-2*Sg-2) +Ny- l}*Sa-
where Ng-2 and sg_2 are the number and first year survival rate of goslings banded 2 years previously, Ny_ 1 is the number of yearlings banded the previous year, and Sa- is the adult (including yearlings) survival rate the previous year. The number of potential 3-year-old females (N3) was calculated similarly, with the inclusion of an additional year-specific survival probability at the end. Together, N2 and N3 estimated the potential maximum number of young breeding females. To estimate the number actually breeding, we tabulated 2- and 3-year-olds recaptured in each year's banding drives. Recaptures provide a larger and less age-biased sample than that obtained through direct observations at nests (Sulzbach & Cooke 1978; Healey 1985). We divided the recapture total by the proportion of the colony believed to have been handled in banding drives to produce a final estimate of young breeders present. This was divided by N2 + N3 to produce the final index.
44
This content downloaded from 142.58.26.229 on Wed, 07 Oct 2015 00:27:41 UTCAll use subject to JSTOR Terms and Conditions
