A capture-recapture model for exploring multi-species synchrony in survival

1. Although recent decades have seen much development of statistical methods to estimate demo-

graphical parameters such as reproduction, and survival and migration probabilities, the focus is

usually the estimation of parameters for individual species. This is despite the fact that several spe-

cies may live in close proximity, sometimes competing for the same resources. There is therefore a

great need for new methods that enable a better integration of demographical data, e.g. the study of

synchrony between sympatric species, which are subject to common environmental stochasticity

and potentially similar biotic interactions.

2. We propose a mark–recapture statistical model that uses random effect terms for studying

synchrony in a demographical parameter at a multi-species level, adapting a framework initially

developed to study multi-site synchrony to this novel situation. The model allows us to divide

between-year variance in a demographical parameter into a ‘synchronous’ component, common to

all species considered, and species-specific ‘asynchronous’ components, as well as to estimate the

proportion of each component accounted for by environmental covariates.

3. We demonstrate the method with data from three colonially breeding auk species that share

resources during the breeding season at the Isle of May, Scotland. Mark-resight information has

been collected since 1984 for Atlantic puffins Fratercula arctica, common guillemots Uria aalge and

razorbills Alca torda marked as breeding adults. We explore the relationship between synchrony in

the species’ survival and two environmental covariates.

4. Most of the between-year variation was synchronous to the three species, and the same environ-

mental covariates acted simultaneously as synchronising and desynchronising agents of adult sur-

vival, possibly through different indirect causation paths.

5. Synthesis and applications. The model proposed allows the investigation of multi-species

synchrony and asynchrony in adult survival, as well as the role of environmental covariates in

generating them. It provides insight into whether sympatric species respond similarly or differently

to changes in their environment, and helps to disentangle the sources of these differences. The

estimated indices of synchrony ? asynchrony can facilitate the generation of further hypotheses

about similarities ? differences in these species’ ecology, such as the potential overlap of wintering

areas. The method is readily applicable to other species, ecosystems and demographical parameters.