Environmental,parental and adaptive variation in egg size of Tengmalm's owls under fluctuating food conditions

We studied egg size variation of Tengmalm's owls in western Finland during 1981–1990. The owls fed on voles whose population fluctuated in a predictable manner: low (1981, 1984, 1987, 1990), increase (1982, 1985, 1988) and peak (1983, 1986, 1986) phases of the cycle occurred every third year. Eggs were largest in the increase phase of the vole cycle, even though that voles were more abundant and egg-laying started earlier in the peak phase than in the increase phase. This suggests that owls invest mostly in egg size when vole abundance increases along with survival chances of offspring. Territory quality and female age had no effects on egg size, but egg size decreased with laying data in the increase phase of the vole cycle. Egg size was significantly positively related to the male age in the increase phase, but the opposite relationship was significant in the peak phase of the vole cycle. The partners of adult males also decreased their egg volume from the increase to the peak phase, whereas the partners of yearling males produced their largest eggs in the peak phase of the vole cycle. This suggests the importance of experience in prevailing food fluctuations. Possibly male Tengmalm's owls can adjust the intensity of courtship feeding not only in relation to the food abundance on their territories at the time of egg laying, but also to the survival prospects of their offspring. Phenotypic plasticity seems to play a substantial role, as the egg size repeatabilities of individual females and partners of individual males were low. Obviously, under cyclic food conditions, predictability and inter-generational trade-offs are important to life history traits.     

Heritability, plasticity and canalization of Ural owl egg size in a cyclic environment

Avian egg size is highly variable on the population level, but is considered inflexible on the individual level. On the basis of 2969 measurements of individual eggs collected during 1981-2005, we analysed heritability, plasticity and selection on egg size in the Ural owl, a long-lived bird that preys on voles. Vole abundance varied in a 3-year cycle, creating varying food supply across the cycle's phases. Ural owl egg size is heritable (h(2) = 60%). Ural owls lay larger eggs in improved food conditions. On the basis of repeated breeding records of 59 females that bred in all vole cycle phases, we show that intra-individual adjustment (plasticity) explained 22.4% of the variation in egg size across phases. Egg size was under stabilizing selection. Extremely small and extremely large eggs had reduced hatchability, and individuals who laid either large or small eggs had lower lifetime fledgling production than the ones laying intermediately sized eggs. Our findings illustrate how maternal investment in egg size can both be heritable and highly responsive to variable environmental conditions, and suggest that variation in the investment in egg size across individuals is canalized.     

Egg size and clutch size in the reproductive investment of American Kestrels

We studied causes and consequences of egg-size variation among clutches of American kestrels ( Falco sparverius ). Egg. from 275 clutches were measured 1990 to 1992. To test the hypothesis that the size of eggs was contrained by food availability in the pre-laying period, we censused small mammal populations in the three years and performed a food supplementation experiment in 1990 and 1991. Kestrels did not advance the date they laid their first egg but did lay significantly larger eggs in response to extra food. The size of eggs was correlated with small mammal abundance on the territoty, and females in good body condition tended to lay large eggs. Body size did not affect egg size, and there were no relationship between agg size and laying date except in 1900, the poorest food year. Clutches with a large mean egg volume had better hatching success than clutches containing small eggs. We argue that there is a phonetypic component to egg size in kestrels, and that kestrels use egg size to fine-tune reproductive investiment to available resources.     

Variable but predictable prey availability affects predator breeding success: natural versus experimental evidence

Food supply is a major source of variation in breeding success for predators, and to what extent individuals are able to cope with temporal variability in food availability remains an outstanding question in life-history studies. We confronted the natural variation in clutch size and breeding success with results from a food supplementation experiment during egg formation, conducted over several contrasted years of natural food supply in an avian specialist predator, the Montagu's harrier Circus pygargus . This raptor mainly preys on common vole Microtus arvalis a cyclic microtine under temperate latitudes. Vole abundance together with timing of breeding accounted for most of the variance in clutch size and number of fledglings. Results from empirical and experimental data were overall in agreement. Fed pairs consistently increased clutch size compared with controls in all experimental years, whereas no effect of food supplementation on egg volume was detected. Supplemented pairs, however, did not fledge significantly more chicks than controls. The costs entailed by the increase in clutch size appear nevertheless to be limited compared with previous studies. Food supply seemed therefore to display sufficient predictability throughout a breeding season to afford individuals the opportunity to adjust their breeding effort to an optimal number of offspring, in agreement with Lack's anticipation hypothesis.     

Laying date, clutch size and egg size of the Mallard Anas platyrhynchos and Tufted Duck Aythya fuligula

The effect of spring temperature on first egg date, laying period and last clutch date was studied in the Mallard and Tufted Duck. Seasonal clutch size and egg size trends were also examined. In years When Mallard laid early, Tufted Duck also did so. The first Mallard clutch was started earlier and laying period was longer in years with a high mean February temperature. The last clutch was started later in years with a high mean June temperature. In Tufted Duck the laying period increased and the last clutch was started later in years with high mean April temperatures.
Mean hatching date of Mallard clutches was later in years when the date of peak chironomid emergence was late.
In both species, clutch size declined through the season. Egg volume was not related to clutch size in either species, but egg volume in the Tufted Duck declined through the season. No difference in Tufted Duck egg size existed between sites, but the significance of egg size on duckling survival is discussed. Genetic factors related to individual consistency in egg size in Mallard may have obscured egg volume trends during the season because of renesting.     

Annual and seasonal reproductive trends in the Lesser Spotted Woodpecker Dendrocopos minor

Annual and seasonal variation in reproductive timing and performance were studied in a population of the Lesser Spotted Woodpecker Dendrocopos minor over 10 years in southern Sweden. The median laying date of the first egg varied by up to 17 days between years, being generally larger than the variation of laying dates within years. Neither clutch size, brood size in successful nests, fledging success in successful nests nor mean nestling weight differed significantly between years. There was no trend for mean clutch size to vary between early and late years. In spite of a more than threefold variation in population size, no reproductive variable demonstrated an apparent density-dependence. Within the season, clutch size declined steeply with increasing clutch initiation date, whereas fledging success and nesting success did not, leading to a trend in brood size almost identical to the trend in clutch size. The survival prospects of fledged young declined with increasing clutch initiation date, and it is argued that the clutch size laid is a strategic adjustment to laying date. Out of 124 breeding attempts, 34% did not produce fledged young. In 9% of the breeding attempts, pairs laid no eggs. At least 20% of the breeding attempts failed after egg-laying. The most common cause of breeding failure was loss of the breeding partner followed by nest abandonment (40% of the failures). Only 16–28% of the failures were due to predation on the nest. Most complete failures, and also partial losses from nests, occurred at the early breeding stages. It is argued that the early nestling phase may be a critical stage, which the woodpeckers adjust to coincide with the seasonal food peak, explaining the strikingly late breeding season compared with other non-migrant species.     

Seasonal variation in reproductive measures of tropical Roseate Terns Sterna dougallih previously undescribed breeding patterns in a seabird

Seasonal variation in egg-laying, egg size, hatching success, hatchling mass, fledging success and chick growth of Roseate Terms Sterna dougallii breeding on Aride Island (Seychelles), Indian Ocean, were studied in 1997 and 1998. I investigated to what extent two patterns, common in a range of species, were followed by tropical Roseate Terns: (a) seasonal decrease in clutch size, egg size and breeding success and (b) an increase in breeding success with increasing egg weight. In 1997 (a poor year), the earliest nesting birds laid significantly smaller eggs, and chicks were lighter at hatching than those of peak nesting birds. The mean clutch size, of 1.04 eggs, showed no seasonal variation and no 'b'-eggs hatched. In 1998 (a good year) the earliest nesting birds laid eggs of similar size and their chicks were of similar weight to those of peak nesting birds. Mean clutch size, of 1.25 eggs, increased significantly through the season and about 60% of the 'b'-eggs hatched. In 1997, hatching success was 57% whereas in 1998 it was 80%. In both years, breeding success declined significantly through the season. The fact that the earliest breeding birds laid smaller eggs in a poor year and smaller clutches in a good year is in marked contrast to a range of other species, and to temperate-nesting Roseate Terns. Egg volume explained about half of the variance in hatchling mass in both years, but only 15% of the variation in linear growth rate. Hatching date was the only variable with a significant effect on fledging success. Roseate Terns on Aride seemed to sacrifice egg size and clutch size for earliness of laying. Presumably it is a strategy of older birds to lay as early as possible and may be regarded as a response of tropical Roseate Terns to breeding under relatively poor, and seasonally declining, food conditions.     

Breeding performance of kestrels at Eskdalemuir, south Scotland

The performance of 144 pairs of European kestrels breeding in upland young conifer plantations was studied from 1976 to 1979. Laying started between 15 April and 1 June, and was 14 days later, on average, in the poor vole year of 1977 than in the other years. Some 38% of pairs failed to fledge young, mainly because they deserted clutches during incubation. The proportion of pairs that failed was positively correlated with total rainfall in May, but apparently not with estimated spring vole numbers. There was a marked seasonal decline in both clutch size and the probability of breeding successfully, so that early laying pairs fledged 2.7 times as many young per attempt as did late pairs. Within years, there was considerable variation in the laying date between pairs. On the available data, environmental factors, such as the habitat or vole numbers around the nest, appeared to have less influence on this variation than did the age of the pair members, their pairing date or their previous breeding experience.     

The effects of habitat quality and female size on the productivity of the lesser spotted eagle Aquila pomarina in the light of the alternative prey hypothesis

Habitat quality is an important but insufficiently understood concept in ecology and conservation biology, due to geographic and temporal variation as well as interaction with individual quality. In 1994–2002, we studied the Estonian population of the lesser spotted eagle Aquila pomarina in order to (1) explore the relative contributions of habitat and female size in reproductive success; (2) check for a switch to alternative prey in vole‐poor years and the relevant variation in annual habitat quality as confirmed in the common buzzard Buteo buteo in the same area. We measured five landscape variables, the number of neighbouring conspecifics and the relative size of the female according to large moulted feathers in 77 nesting territories, and related this to the eagles’ productivity in vole‐rich and vole‐poor years. Nesting lesser spotted eagles benefited from heterogeneous landscapes and suffered from the neighbourhood of conspecifics. There was no evidence that different‐sized females used different habitats. In general, female size was positively related to productivity in vole‐poor but not vole‐rich years, but in the presence of competitors, large size appeared to be disadvantageous. The mean annual productivity of the eagle was well correlated with that of the buzzard, both having peaks after every three years. In contrast to the buzzard, the share of voles in the eagle's diet and its habitat quality did not differ significantly between good and poor years. We concluded that despite a superficial ecological similarity to the buzzard, the lesser spotted eagle did not behave as predicted by the alternative prey hypothesis, but the study confirmed that annual variation in prey utilization and relative habitat quality are parts of the same functional response. Non‐switching to alternative prey may be related to a historical foraging strategy, used by the eagles before they spread to agricultural landscapes, since the current effects of body size strongly suggested food shortage in vole‐poor years.     

Body size and determinants of laying date variation in the Snow Petrel Pagodroma nivea

We studied several determinants of laying date variation and the relationship between laying date and reproductive success in the Snow Petrel Pagodroma nivea . The effects of female body size and condition, year, individual laying period, colony size, mate fidelity, previous reproductive success, and duration of the pre-laying exodus on laying date, were investigated during a 3-year study. The average laying date was 4 December. The laying period was compressed into 10–16 days and was very constant from year to year, both for the population as a whole and for individual females. The laying period of individual females varied from year to year on average by less than one day. Body size explained 24% of the variation in laying dates, with large females laying their egg later than small ones. Laying in large colonies occurred c. 2 days later than in small colonies, mainly because a higher proportion of large females bred in large colonies. There was no effect of mate fidelity, age, body condition and previous reproductive success on laying date, but the duration of the pre-laying exodus was strongly correlated with laying date. Smaller females had shorter pre-laying exodus (c. 17.7 days) than larger ones (c. 20.4 days). During the three years of the study reproductive success either increased, decreased or did not vary with laying date. Although body size is not maintained by selection on laying date alone, the genetic body size component of this species suggests that balancing selection on body size may act through laying date.     

Survival of Herring Gull Larus argentatus chicks: an experimental analysis of the need for early breeding

In 1985 and 1990, we manipulated the hatching date of Herring Gulls Larus argentatus by exchanging complete clutches between pairs which had started egg laying on different dates in the season. The aim was to investigate whether the observed seasonal decline in Hedging success can be explained by laying date. In both years, fledging success of advanced pairs followed the natural seasonal trend, supporting the date hypothesis. However, in both years, Hedging success of delayed pairs remained as high as that of control pairs with the same original hatching date as the foster parents, supporting the parental quality hypothesis. These data suggest that the delayed pairs may have compensated for less favourable environmental conditions. Although conspecific predation was the main direct cause of chick mortality, food shortage probably indirectly determined the mortality rate. A model of fledging success in relation to the timing of egg laying predicts that when food is abundant the optimal hatching date coincides with the mean hatching date of the colony (synchronization advantage). When food is less abundant, early breeders have the highest fledging success.     

Reproduction of the least weasel in captivity: basic observations and the influence of food availability

The least weaselMustela nivalis nivalis Linnaeus, 1766 is unique among carnivores because of its small body size and capacity for fast reproduction. It has been suggested to be the main agent for maintaining cyclic fluctuations in northern vole populations, largely based on its high reproductive potential and dependence on small rodents. This study describes basic reproductive data on the least weasel obtained during a captive breeding program. In the experiments, food availability was manipulated at the onset of breeding. Altogether, 65 litters were born during the 5-year study. The mean litter size was 5.1 (in 53 litters of known litter size), the most common litter size being six. The sex ratio of weaned young was not biased from 1∶1. The median date of birth was June 4. Food manipulations did not affect mating frequency suggesting that the observation of failure in breeding in years of low abundance of small mammals is due to mortality of embryos or young before weaning, but not due to avoidance of breeding at low food availability. In this respect, weasels differ from other main vole predators, owls and raptors, which often skip breeding if small mammals are scarce.     

Fluctuating food supply affects the clutch size of Tengmalm's owl independent of laying date

Summary In western Finland, yearly median laying dates of Tengmalm's owls varied from 14 March to 27 April during 1973–1989 and were negatively correlated with the winter densities of voles. Yearly mean clutch sizes varied from 4.0 to 6.7 and were more closely related to the spring than to the winter densities of voles. The yearly mean clutch size decreased with yearly median laying date. The 3-year vole population cycle is typical of the study area. The start of egg-laying was earliest in the peak phase of the cycle (median laying date 22 March), when vole numbers are high during egg-laying, but decline rapidly to low numbers in the next autumn or winter. In the increase phase (1 April) vole abundances are moderate at the time of laying, but increase to a peak in the next autumn or winter. In the low phase (15 April) voles are scarce in spring and in the preceding winter, starting to increase in late summer. Clutch size and female body mass were independent of laying date in the low phase, decreased slowly but significantly in the increase phase, and declined abruptly in the peak phase. These trends also held when the effects of territory quality, female age and male age were ruled out. When comparing the same laying periods, clutch sizes were significantly larger in the increase than in other phases of the cycle, but there was no difference between the peak and low phases. Supplementary feeding prior to and during egg-laying increased clutch size independent of laying date. These results agreed with the income model (the rate of energy supply during laying determines clutch size). Tengmalm's owls invest most in a clutch in the increase phase, as the reproductive value of eggs is largest because of high survival of yearlings. A high reproductive effort may be adaptive during this phase, because the availability of voles is predictable during the laying period.     

Allocation of offspring size and sex by female black ratsnakes

How females allocate resources to each offspring and how they allocate the sex of their offspring are two powerful potential avenues by which mothers can affect offspring fitness. Previous research has focussed extensively on mean offspring size, with much less attention given to variance in offspring size. Here we focussed on variation in offspring size in black ratsnakes, Elaphe obsoleta . We collected and hatched 105 clutches (1283 eggs) over 9 years. We predicted that females should lay larger eggs, or more variable eggs, when the environment is less predictable. We also predicted that females laying early or laying larger eggs should produce mostly sons because adult males are larger than adult female ratsnakes. The largest hatchling was more than twice the length and almost four times the mass of the smallest hatchling. Variation in offspring size was itself highly variable, with CVs in offspring mass among clutches ranging from 1% to 25%. With one exception, the variables we expected should influence variation in offspring size had little effect. We found that clutch size increased with maternal size and that egg size decreased with clutch size, but we found no evidence that variance in egg size among clutches increased as the season progressed or that females increased the mean size of their offspring the later in the season they laid their eggs. Females in better condition after they finish laying their eggs did produce larger eggs. There was no relationship between within-clutch variation in egg size and laying date or mean egg size. Finally, sex ratio did not vary with mean egg size or hatching date. Given evidence that offspring size in snakes affects survival, selection should reduce variation in offspring size unless that variance enhances maternal fitness and yet we found little support for hypothesized advantages of varying offspring size.     

Breeding success and chronology of Wood Storks Mycteria americana in northern and central Florida, U.S.A.

The breeding success and chronology of Wood Storks Mycteria americana were studied at eight colonies in northern and central Florida during 1981–1985. Mean ± s.d. clutch size for all colony-years was 3.07 ± 0.56 (n = 2694 nests), with three-egg clutches (72%) most frequent. Mean clutch size among all colonies and years ranged from 2.73 ± 0.55 to 3.41 ± 0.61. Many colonies exhibited significant negative trends in clutch size with, hatching date because of a proportional decrease in four-egg clutches later in the season. Mean colony clutch size was not correlated with nest numbers, nesting density or mean hatching date within most years. Mean ± s.d. number of fledglings for all colonies and years was 1.29 ± 1.16 fledglings per nest (n = 2812 nests). Mean annual fledging rates in colonies ranged from 0 (colony failed) to 2.66 fledglings per nest. Most breeding failure occurred prior to egg hatching, and the second highest mortality occurred between hatching and 2 weeks of age. Four-egg clutches fledged more storks than three-egg clutches, which in turn were more successful than two-egg clutches. However, all clutch sizes showed similar fledgling per egg rates. The seasonal decline in productivity was associated proportionally with smaller clutch sizes later in the breeding season. An increase in mean hatching date was correlated with an increase in latitude. There was greater within-year breeding synchrony among colonies than interyear breeding synchrony within each colony. Breeding synchrony was not correlated with mean hatching date, latitude, longitude, nest numbers or nesting density.     

INTERACTIVE EFFECTS OF OFFSPRING SIZE AND TIMING OF REPRODUCTION ON OFFSPRING REPRODUCTION: EXPERIMENTAL,MATERNAL, AND QUANTITATIVE GENETIC ASPECTS

We demonstrate that egg size in side-blotched lizards is heritable (parent-offspring regressions) and thus will respond to natural selection. Because our estimate of heritability is derived from free-ranging lizards, it is useful for predicting evolutionary response to selection in wild populations. Moreover, our estimate for the heritability of egg size is not likely to be confounded by nongenetic maternal effects that might arise from egg size per se because we estimate a significant parent-offspring correlation for egg size in the face of dramatic experimental manipulation of yolk volume of the egg. Furthermore, we also demonstrate a significant correlation between egg size of the female parent and clutch size of her offspring. Because this correlation is not related to experimentally induced maternal effects, we suggest that it is indicative of a genetic correlation between egg size and clutch size. We synthesize our results from genetic analyses of the trade-off between egg size and clutch size with previously published experiments that document the mechanistic basis of this trade-off. Experimental manipulation of yolk volume has no effect on offspring reproductive traits such as egg size, clutch size, size at maturity, or oviposition date. However, egg size was related to offspring survival during adult phases of the life history. We partitioned survival of offspring during the adult phase of the life history into (1) survival of offspring from winter emergence to the production of the first clutch (i.e., the vitellogenic phase of the first clutch), and (2) survival of the offspring from the production of the first clutch to the end of the reproductive season. Offspring from the first clutch of the reproductive season in the previous year had higher survival during vitellogenesis of their first clutch if these offspring came from small eggs. We did not observe selection during these prelaying phases of adulthood for offspring from later clutches. However, we did find that later clutch offspring from large eggs had the highest survival over the first season of reproduction. The differences in selection on adult survival arising from maternal effects would reinforce previously documented selection that favors the production of small offspring early in the season and large offspring later in the season—a seasonal shift in maternal provisioning. We also report on a significant parent-offspring correlation in lay date and thus significant heritable variation in lay date. We can rule out the possibility of yolk volume as a confounding maternal effect—experimental manipulation of yolk volume has no effect on lay date of offspring. However, we cannot distinguish between genetic effects (i.e., heritable) and nongenetic maternal effects acting on lay date that arise from the maternal trait lay date per se (or other unidentified maternal traits). Nevertheless, we demonstrate how the timing of female reproduction (e.g., date of oviposition and date of hatching) affect reproductive attributes of offspring. Notably, we find that date of hatching has effects on body size at maturity and fecundity of offspring from later clutches. We did not detect comparable effects of lay date on offspring from the first clutch.     

Geographical variation in egg size of the Great Tit Parus major: a new perspective

A recent study on geographical variation in egg size of Great Tits Parus major concluded that: (1) mean egg size tended to increase with increasing latitude; and (2) mean egg size was positively correlated with mean clutch size. Including new data on both egg and clutch size, we reanalysed the relationships between egg size, clutch size and latitude, and investigated the possible effects of habitat type, female body size and egg shape on these relationships. We found that (1) egg volume showed minimum values around 51°N, increasing both north and southwards; (2) female body size increased linearly with increasing latitude; (3) female body size was positively correlated with egg breadth, but not with egg length or egg volume; (4) the sphericity index of the eggs (breadth to length ratio) was largest at medium latitudes, and eggs were more elongated towards the north and the south; (5) the relationship between clutch size and latitude was curvilinear, with the largest clutch sizes at intermediate latitudes; (6) egg size was not correlated with clutch size when the complete latitudinal range was considered, but egg size was negatively correlated with clutch size between 40 and 51°N; and (7) egg size did not differ among habitat types. We suggest that female body size (which probably limits egg breadth), and the pressure for producing large eggs (which in turn increases the reproductive success) are the main determinants of geographical variation in egg size and shape. Populations of small-bodied Great Tits seem to escape from the limits of their size, producing relatively elongated eggs, so that from a certain latitude southwards, egg volume does not decrease in spite of a decrease in female body size. Moreover, the negative relationship between egg and clutch size at low latitudes suggests that energetic trade-offs may also contribute to determine egg size in the south.     

Field Vole Microtus agrestis abundance and Hen Harrier Circus cyaneus diet and breeding in Scotland

In many parts of the global range, voles form an important part of the diet of Hen Harriers Circus cyaneus , and breeding numbers are correlated with the abundance of these small mammals. In Scotland, however, little information is available on harrier diet in the spring and our understanding of causes of variation in harrier breeding density is complicated by human interference. In this paper we explore the relationship between Field Vole Microtus agrestis abundance and harrier spring diet, density and productivity in southern Scotland. Over three years, voles occurred on average in 67% of pellets, and 79% in years of high and intermediate vole abundance. From 1992, the number of breeding harriers increased following protection from illegal persecution. After accounting for this trend, harrier numbers correlated strongly with vole abundance. Harrier clutch size was also correlated with vole abundance. Although fledging success tended to be greater in years of vole increase than in years of vole decline, fledging success was not significantly correlated with the relative abundance of voles, or with the abundance of Meadow Pipits or Red Grouse chicks.     

Life-history variation within a three-spined stickleback population in the Camargue

Among individuals of female three-spined sticklebacks Gasterosteus aculeatus from a population in the Camargue, southern France, studied in 12 successive years, adult L _T ranged from 31–64 mm, clutch size ranged from 33–660 eggs, and mean egg diameter per clutch ranged from 1.15–1.67 mm. Because the population was strictly annual, inter-annual variation corresponded to variation among generations having experienced different environmental conditions. Body mass varied significantly among years, suggesting an effect of varying environmental conditions. Gonad mass and clutch size increased with body mass, but mean egg diameter was not correlated to body mass. Body mass-adjusted gonad mass, interpreted as reproductive effort per clutch, did not vary significantly among years, suggesting that this trait was not influenced by environmental conditions. Body mass-adjusted clutch size and egg size varied significantly among years. Inter-annual variation in body length at breeding, clutch size and egg size was of the same order of magnitude as inter-population variation reported by other authors for this species. During the breeding season, reproductive effort and clutch size tended to increase. Egg size tended to decrease during the breeding season but this seasonal pattern varied among years. Observed life-history variation is discussed both in terms of its evolutionary significance and methodological implications in the study of life-history variation among populations.     

Population dynamics in a cyclic environment: consequences of cyclic food abundance on tawny owl reproduction and survival

1. Understanding which factors regulate population dynamics may help us to understand how a population would respond to environmental change, and why some populations are declining.
2. In southern Finland, vole abundance shows a three-phased cycle of low, increase and decrease phases, but these have been fading out in recent years. During five such cycles (1981–1995), all tawny owls Strix aluco were censused in a 250-km² study area, and their reproduction and survival were monitored.
3. Males and females showed similar dynamics, but experienced breeders recruited more offspring and had higher survival than first breeders. Offspring recruitment, but not survival of breeding individuals varied in accordance with vole abundance.
4. The population's numerical response to prey abundance was primarily due to first-breeding individuals entering the population in the increase phase when immigration was the highest. First-breeding birds were younger, but experienced breeders were older in more favourable vole years.
5. A stage-specific matrix population model integrating survival and fecundity showed that, despite obvious variation in fecundity between vole cycle phases, this variation had limited importance for overall tawny owl population dynamics, but that the survival of experienced breeders during the low phase is most important for population growth.
6. Model and data agreed that the vole cycle drives the dynamics of this avian predator by limiting the recruitment of new breeders during the low phase. Population dynamics hence differ not only from the classic example of the species in a more temperate region in the UK where the number of territories is stable across years, but also from the dynamics of other avian vole predators in Fennoscandia where the recurring crash in vole abundance drastically lowers adult survival thereby creating vacancies.