Multidimensional plasticity of phenology: Assessing the effects of population density on plastic responses of breeding time to temperature

Phenotypic plasticity is the key adaptive mechanism behind annual adjustment of breeding time in response to temperature. In nature, organisms are not only subjected to variation in temperature but encounter multiple fluctuating environmental factors that affect phenotypic expression, including conspecific density, which affects individual performances through resource competition. We examined the interactive effects of temperature and conspecific density at two spatial scales (territory and patch level) on breeding time and success utilizing data obtained from long-term monitoring of a wild great tit (Parus major) population in a fragmented woodland. As expected, we detected earlier breeding in response to warmer spring temperatures. We report earlier laying at low territory level density (i.e. a larger available area per breeding pair), but no evidence of density effects at patch level (breeding pairs per hectare). Birds experiencing low territory level density throughout their life bred on average earlier, and this response was also seen at the within-individual level (earlier laying in years when individuals experienced a lower density than average). We found no context-dependence of plastic responses to warmer springs as we detected no interactive effects between density and temperature. In terms of breeding success, earlier laying decreased the risk of brood failure and increased the number of fledglings. The number of fledglings was higher at low territory level density, while higher patch level density increased the probability of brood failure. Altogether, these results indicate that density-related responses were likely mediated by food competition rather than by increased numbers of low-quality birds or increased occupation of poorer territories at higher densities. This study highlights the importance of examining parameters at different spatial scales, along with the study of individual responses to multivariate cues for a comprehensive understanding of the variations in phenological plasticity.

Methods
This study uses data from a long-term great and blue tit monitoring program in Belgium from 1994-2016. The “Boshoek” field site (51°08 N – 4°31 E) encompasses 12 patches of fragmented woodlands (1 to 15 ha) in an approximately 10 km2 area near Antwerp. For the purpose of this study, data from the four largest oak patches— ZZ, (11.1 ha with 96 boxes); KB, (10.1 ha with 90 boxes); LO, (6.2 ha with 62 boxes) and HN (2.8 ha with 23 boxes)—were used due to their comparable habitat characteristics (Matthysen 2002). In all sites approximately 9 boxes were available per hectare, whereby two thirds of boxes had a 32mm opening (allowing both blue and great tits) and one third had a 26mm opening (allowing blue tits only). These patches were characterised predominantly by mature stands of common oak(Quercus robur) , undergrowth of hazel (Coryllus avellana) and black cherry (Prunus serotina) , and had distinct boundaries well-defined by the sharp transition from closed-canopy forest to open areas such as fields or meadows, residential areas or small farms (Matthysen 2002). 

Nestboxes were visited approximately weekly starting before egg laying. 80–90% of all parents were captured on the nest when they were feeding 8-day-old nestlings. Nestlings were ringed and their body mass measured to the nearest 0.1g. All nestboxes were checked at least twice each winter for roosting birds. In some years, additional captures were done with mistnests in autumn and/or winter. All chicks were ringed at day 14 or 15 after hatching. Each year nearly all (> 85%) of the parents were trapped when feeding young of ca. 8 days old.

Phenological parameters

First-egg dates were estimated from the first observation of a partially laid clutch, assuming one egg being laid per day. All laying dates are expressed as days since March 1.