European Commission logo
français français
CORDIS - Résultats de la recherche de l’UE

When to reproduce? Genetic, phenotypic and environmental factors explaining the variation in timing of avian reproduction

Final Report Summary - POPBYTIME (When to reproduce? Genetic, phenotypic and environmental factors explaining the variation in timing of avian reproduction)

The timing of breeding is a primary factor in affecting the reproductive success of birds and many breeding performances are linked to the time of the season. For example, the seasonal decline in clutch size is a common pattern among birds and many authors deem it to be a result of plasticity depending on individual quality: earlier breeders are of better condition and can produce larger clutches compared to later breeders in less good condition producing smaller clutches. An alternative but not mutually exclusive hypothesis is that the breeding date pattern is a phenotypic trait genetically determined and that early and late breeding could represent specific strategy aimed at optimising the reproductive performances in accordance with environmental conditions. Accordingly, both these two strategies, breeding early vs. breeding late, can be adaptive in maximising the parents fitness, especially if they breed across an extensive period with changing weather conditions and food availability as in the Mediterranean area. Here, during the late period of the breeding season the meteorological conditions are more stable than the early period, and late breeders can benefit from this situation. Thus, early and late breeding could reflect two adaptive strategies, the first one a risky strategy but that yields, whether conditions are fine, a high reproductive output, while the latter one reflects a less risky strategy with a stable but lower output.

The main aim of this project was to evaluate the fitness consequences on free ranging European kestrels (Falco tinnunculus) of the timing of breeding by investigating the effects of between years variation of weather conditions on the reproductive success of early and late breeders and two ascertain the existence of any genetic differentiation between early and late breeders. It has also been shown in this species, but in a different population, that the sex ratio is of one male vs. one female (1:
1) but with more males produced by early pairs and more females produced by late pairs. Since the species in sexual dimorphic as females are heavier than males, we tested the abovementioned hypothesis by considering that late breeders produce smaller clutches but with higher reproductive values (female biased). To achieve these main goals we studied a wild population of kestrels breeding in nest boxes attached to the pylons of utility lines in Parma during three consecutive years (2009-2011) considering also data coming from other 2 years of research (2006-2007).

Breeding adults were trapped once during the young rearing phase (from mid May to mid June; 6-15 days after hatching) using traps placed on the ground (bal-chatri) or in the nest (sliding door controlled remotely) to gain biological samples (e. g. blood, feathers, parasites). Nestlings hatched from each breeding pair were also sampled. Diet and nest activities (laying date, hatching date, hatching success and fledging success) were recorded by placing time-lapse cameras before the onset of reproduction. Climate conditions were obtained from the official archive of the regional environmental agency (ARPA Emilia Romagna). Data are referred to 134 nests and 514 individuals, 125 adults (66 females and 59 males) and 389 nestlings. Some lab analyses are still in progress (STR analysis of three loci 70 samples each, plasma analysis of physiological markers of condition of 370 birds) and this report describe a part of the main output of the research.
1. The most important finding was that kestrels responded to warmer seasons by anticipating the mean laying date of the population (F (4, 124) = 21. 71, p < 0. 001). Consequently, due to the general covariation between laying date and clutch size, the mean clutch size was higher in warmer seasons (F (4, 124) = 4. 13, p < 0. 01). Early and late breeders responded similarly to temperature variation as both anticipated the laying date in relation to higher temperature (year X timing: F (8, 113) = 0. 69, p = 0. 70). During these favorable years the decline of the clutch size with the progress of the season was not significant as also late breeders produced large clutches. This result give indirect evidence that the decline of the clutch size with the season is related to the season itself, such as, e. g. the increase of the photoperiod registered during the laying date, and not to a direct individual constrain.

2. RAPD analysis revealed a significance divergence between early and late breeders (p = 0. 007) but not between early and intermediate breeders of between intermediate and late breeders. This molecular differentiation was reinforced at phenotypic level by the finding that females (but not males) breeding later were bigger (longer tarsus) than the ones breeding earlier (r = -0. 45, n = 19, p = 0. 05). This is a very interesting discovery that deserves a deeper investigation and we are working on STR and other molecular markers. We have already analyzed locus 347 and 79-4 of 70 birds and we are working on the STR analyses of other 3 loci. The gathered samples will also shed light on the nature of this genetic differentiation that could be related to, e. g., a genetically determined condition or to the existence of two distinct subpopulations (e. g. resident and migratory).

Finally, in order to evaluate the existence of a geographical cline of the genetic differentiation by time considering the genetic structure of populations at different latitudinal locations and characterised by different habits (migratory vs. resident) we have collected 40 DNA samples of nestlings (one per nest) belonging to the population of Groningen (Netherlands) and collected in collaboration with the University of Groningen.
3. From the analysis of the DNA it was also possible to molecularly determine the sex of 279 nestlings showing that the brood sex-ratio was always female-biased (F (4, 72) = 0. 205, p = 0. 935) and that this bias was more marked during adverse seasons due to food scarcity (in turn caused by a extreme rainy season that caused a crush a voles, the main prey item of the studied population). Contrary to what expected, the number of male nestlings increased with the progress of the season, but only in favorable years, while in unfavorable years the seasonal trend of the brood sex ratio was female biased. This evidence is important because clearly it clearly indicates the consequences of climate conditions on the structure of the population.
4. In a five-year study on the Eurasian kestrel Falco tinnunculus we systematically observed the disappearance of eggs during incubation. In 2010 we equipped 11 nests and 20 in 2011 with a time-lapse digital camera (WSCA04, Wingscape, AL, USA) programmed to record photographs at 1-min interval from dusk to sunset. Cameras were powered by 7 A, 12 V lead batteries which were fixed to the roof of the nest-box and replaced every eight days together with the 2 GB memory card from which frames were downloaded. By recording the incubation behavior with these camera we discovered that females were responsible for all the clutch reductions. We therefore analyzed egg losses retrospectively, in relation to the trophic conditions characterising each year to search for a possible functional relationship between food availability and clutch reduction. We expected that eggs removal would have been more frequent (1) during years of adverse trophic conditions, when parents can benefit in reducing the costs of incubation when food is scarce; (2) in larger clutches, because costs of incubation increase with the number of eggs that must be heated; and (3) in females in low body condition, because they could benefit from a lower parental effort during incubation and brooding. In addition, we considered if clutch reduction could benefit the parents in terms of fitness by foreseen (4) a higher fledging success (fledged/hatched nestlings) among pairs showing clutch reduction. We found that clutch reduction varied with year (GLM year, F (4, 72) = 2. 90, p = 0. 03; nest, F (43, 72) = 1. 55, p = 0. 05; clutch size, F (1, 73) = 24. 85, p < 0. 0001), being more frequent during 2009 and 2010 (negative peaks of voles), being more frequent during years of low food availability (2009 and 2010). Clutch reduction was more frequent in larger clutches (r = 0. 37, n = 121, p < 0. 001,), but was not associated with females'body condition trapped during mating (r = 0. 48, n = 9, p = 0. 19), nor during young rearing (r = 0. 04 n = 23, p = 0. 85). Clutch reduction benefitted parents in terms of relative fitness since it was positively associated with fledging success (Final model: GLM year x clutch reduction, F (4, 111) = 2. 21, p = 0. 07; year, F (4, 111) = 6. 56, p < 0. 0001, clutch reduction F (1, 111) = 2. 94, p = 0. 09. The estimate of parameters showed that relative fledging success was enhanced by clutch reduction only in 2009 (estimate = 0. 13±0. 04, p = 0. 001; for all others estimates all p > 0. 29), thus only in year when food scarcity was not predictable. These findings indicate for the first time that incubating females can adjust clutch size after clutch completion by secondary clutch reduction in an adaptive way (see draft paper in Annex I).


The study revealed a clear response of breeding European kestrels to warmer season anticipating the laying date and producing more nestlings during warmer seasons. We found an overall female biased sex-ratio that was more marked in adverse years. The implication of these discoveries should be considered by wildlife conservationists and ecologist and should be further addressed by future studied on population dynamics influences by climate change, and at the end, by timing of reproduction. At present, concerns about the disentanglement between the biotic and abiotic parts of the ecosystem due to climate change and landscape management is a hot topic of the European Union policy and these results can contribute to provide useful information to foresee the effects of the modification of the environment on the structure and status of the population and to know to what extent birds, and raptors in particular, are flexible in reacting to global change. Moreover, the effect of climate change on wildlife is a topic of broad interest to the general public as well, since it relates to the biodiversity of the Planet and it could have a great impact on mankind. Knowing the consequences of global warming at individual and population level may be important for wildlife managers interested in predicting sustainable harvesting plans for wildlife. Indeed, wildlife management plans have to foresee the pressures that can act on the structure of a population, such as the sex-ratio of the juveniles produced during a season or the condition of individuals, which is directly related to their population dynamics and productivity.