Spatial components of plasticity in tit phenology: responses, constraints and amelioration

There are three potential ways in which organisms can respond to changing environments: (1) they may disperse, or migrate, (2) they may evolve so that they adapt to the new environment, or (3) they may produce different phenotypes - in other words display phenotypic plasticity - as the environment changes. A contemporary example relates to understanding responses of populations to climate change. Work to date suggests that, despite the three mechanisms being non-exclusive, population responses to climate change usually involve phenotypic plasticity. Hence, understanding the evolutionary forces acting on plasticity is of central importance in our understanding of viability in the face of climate change.

Understanding of phenotypic plasticity and its role in adaptation to changing environments is hampered by the fact that most studies simply correlate an average phenotype for the population with a single value for the environment, most often at the level of an entire year. This only makes sense if the environmental cues to which organisms respond are very large-scale cues, varying little from the perspective of individuals within populations. However, we know that many organisms experience only a limited part of the environment, and that the environment may vary over quite small spatial scales. Despite this, we don't understand how animals balance these small- and large-scale cues. The central aim of this research is thus to determine how the spatial scale of the environment is important in understanding the evolution of phenotypic plasticity.

Our model system involves reproductive behaviour in small woodland birds - great tits, breeding in Wytham Woods near Oxford - which are under strong natural selection to time their reproduction to coincide with peaks in abundance of moth caterpillars (e.g. the winter moth) that are adapted to feed on newly emerged leaves of deciduous trees. At the population level there is a good match between the timing of birds' breeding and the peak of caterpillar abundance, but there is tremendous variation within each year in the timing of these events over quite a small spatial scale. Furthermore, we have evidence that, despite a common temperature trend, different parts of the population are responding at different rates. Hence, the population level summary statistics disguise several important levels of variation.

We will use long term data on breeding behaviour and fitness, together with detailed environmental data to analyse the spatial scales at which variation in bird reproductive timing can best be explained, and to test hypotheses about the influence of scale on fitness and population dynamics. We will then supplement these data with new data collected across a regular grid of locations to determine phenology of bud-burst and caterpillar abundance, and hence characterise the extent to which birds are able to match the timing of events in their environment at different scales. Because we expect multiple scales to be important, we can make the prediction that the optimal phenotype is a balance between small- and large-scale plasticity, and hence that adaptation will not be perfect at either scale in isolation. Because the environment is patchy, we can further predict that adjusting to small- as well as large-scale cues will lead to some patches having higher productivity than others; hence the spatial scale of plasticity will lead to within population variation in population dynamics. Collecting environmental data on the ground is very time-consuming, and only limited areas can be covered; therefore we will test the extent to which satellite images can be used to estimate phenology at scales that are relevant to organisms in nature. Finally, we will carry out experimental tests of whether mis-matches in phenology between birds and the environment, which have been implicated in population declines in some species, are alleviated by being in more varied environments.

The primary justification for the work proposed here is to advance our fundamental understanding of the evolution of phenotypic plasticity in variable environments. However, because one of the environmental axes along which we will assess plasticity is temperature, and spring temperature has shown an upward trend over the past three decades, and is one of the key drivers of phenotypic change in response to global warming, the research will also be of relevance to understanding how populations adapt to climate change. The unique insights from this work will be an understanding of how the potential for adaptation via phenotypic plasticity to large-scale drivers may be compromised by simultaneous selection for adaptation to small-scale environmental effects. Consequently, the work has the potential to inform the next generation of theoretical work that incorporates phenotypic plasticity, and both spatial and temporal variation in the environment assessments of the risk of climate change to individual populations. Nevertheless, the theoretical framework needed for this work is not yet developed, and its development is outside the scope of this application.

However, there are three distinct areas where this research can be expected to have impact.

First, the development of approaches for remote sensing vegetation phenology at scales relevant to forest-living birds and insects would be an important advance in our ability to assess habitat quality and make management decisions with relatively less ground-level data. This research would thus be of interest to conservation-focussed organisations (at national level these include the BTO, RSPB, Wildlife Trusts, Woodland Trust).

Second, the development of a detailed understanding of the effects of phenological mismatch on reproductive fitness, and how this depends on habitat variability would be very valuable information for understanding population viability in species where phenological mismatch is implicated in low rates of reproduction, and this research would thus also be of interest to conservation-focussed organisations (at national level these include the BTO, RSPB, Wildlife Trusts, Woodland Trust).

Finally, the research will contribute to the public understanding of science. Research into responses of animals to climate change creates tremendous interest among the public, particularly when the research can be related to familiar organisms, such as common garden birds. Further, because the elements of the tri-trophic bird/oak/caterpillar system are visually accessible, they have been popular subjects for broadcast media. In addition, the link between remote-sensing and breeding phenology of birds ("predicting behaviour from space") is also likely to be of considerable interest to the general public. Therefore, we expect that the research in this proposal is likely to be of disproportionate interest to the public.