Habitat fragmentation is a leading cause of biodiversity loss, but relatively little is known about its long-term evolutionary consequences. I addressed this problem using neighbouring wild populations of burying beetles (Nicrophorus spp.), insects which fight to secure carrion upon which they breed. My work focused on the Nicrophorus populations inhabiting a cluster of seven ancient woods west of Cambridge. Whilst the criteria used to identify ancient woods are increasingly controversial, I combined new and existing approaches to show there is strong evidence that these woods have been physically separated for at least a thousand years.

I carried out three years of fieldwork in each of the seven woods and found that four different species of burying beetle compete over carrion. The carrion niche is partitioned among burying beetles by size: but despite overlapping spatially and temporally with its larger competitors, the smallest species N. vespilloides is the most abundant in each wood. Notwithstanding their similarity in age, ecology and geographic location, I found that the community structure and density of burying beetles varied between the seven woods. I argue that these differences likely cause differences in the intensity of competition between the woods.

If this is true, then selection could then favour different levels of investment in competition-related phenotypes between the populations. I tested this hypothesis by establishing seven laboratory populations of N. vespilloides, each derived from a different woodland population. I then measured relevant phenotypes from each population in common garden experiments. I found that the contests between burying beetles over carcasses are fought primarily with their mandibles but that there is also a behavioural component of individual fighting ability. Females from woods with more diverse community structures were found to be better fighters.

A second set of experiments revealed between-population variation in the behaviour of larvae and adults when breeding on a carcass. In woods that had more diverse community structures, larvae arrived at the carcass sooner after oviposition than did larvae from woods with less diverse community structures. I also found that larval arrival time at the carcass causes a sudden drop in the motivation of potential usurpers to take over a carcass. Together, these results suggest that a quicker larval arrival time might provide a mechanism for evading intense interspecific competition for carrion. Finally, I found that males from woods which support a high density of burying beetles had a shorter duration of parental care, presumably because males departed earlier to secure additional off-carcass mating opportunities.

After working with collaborators to obtain pooled whole genome sequencing for the populations, I measured the extent of between population genetic differentiation. I found that the extent of genetic differentiation correlates strongly with the extent of dissimilarity in their community structures. These results suggest that the populations are undergoing divergent adaptation despite gene flow. Together with recent work on other species, the emerging picture is that even populations living in closely-connected habitat fragments can be genetically distinct from one another, at least at some loci. The spatial grain of biodiversity may be far finer than previously believed.