SEARCH

SEARCH BY CITATION

Keywords:

  • behaviour/social evolution;
  • conservation biology;
  • ecological genetics;
  • empirical;
  • fish;
  • population genetics

Abstract

  1. Top of page
  2. Abstract
  3. References

Studying the movement of individuals in the wild has always been a challenge in ecology. However, estimating such movement is essential in life sciences as it is the base-line for evaluating connectivity, a major component in developing management and conservation plans. Furthermore, movement, or migration, is an essential parameter in population genetics, as it directly affects genetic differentiation. The development of highly variable markers has allowed genetic discrimination between individuals within populations and at larger scales, and the availability of high-throughput technologies means that many samples and hence many individuals can be screened. These advances mean that we can now use genetic identification for tracking individuals, and hence follow both survival and reproductive output through the life cycle. The paper by Morrissey & Ferguson (2011, this issue) is a demonstration of this new capability, as authors were able to infer the movement of salmonid fish initially captured as juveniles, and later as reproductively mature adults.

Understanding patterns of dispersal (through the life cycle) is a major goal in 21st century marine ecology (Botsford et al. 2009). Patterns of dispersal determine the rates of exchange, or connectivity, among populations. Population connectivity, in turn, has major consequences for all aspects of an organism’s biology, from individual behaviour to metapopulation dynamics, and from evolution within metapopulations to the origin and extinction of species. Further, understanding patterns of larval dispersal is critical for the design of effective networks of marine reserves—vital tools in development of sustainable fisheries (Sale et al. 2005).

Up to recently, capture-mark-recapture protocols in natural populations were the only way to follow individuals in real time. More recently, telemetry, pop-up satellite tags and other pit-tags and recorder tags systems have revolutionized the concept of following individuals in real time. Unfortunately, while satellite tracking offers an ultimate tool for following individuals, there are also limitations to this system. The first limitation is the time: any such system requires energy, but batteries do not last for years and years, limiting the survey to some portion of the lifespan. The other major limitation is water, as satellite radio cannot penetrate water and is useless in rivers, seas and oceans. Although some acoustic systems are currently in development, they suffer the same limitations of energy use. Finally, the last problem is the size of such equipment: as while there is no problem to adapt such equipment to lions in the jungle or whales or turtles in the ocean, it remains impossible for small fish and even more so for their larvae. Overall, such real-time tagging approaches are very good in the perspective of understanding the behavioural ecology of individual movement, but they appeared not adapted where the goal is to understand population-level patterns and processes.

Individual-based genetic identification has been mostly used in the context of parentage analysis. Parentage analysis is a practical form of assignment test, which involves identifying the parents of specific individuals (Manel et al. 2005). Patterns of parentage play a central role in the study of diverse ecological and evolutionary topics (reviewed in Jones et al. 2010) and can be particularly useful for detecting ecological and evolutionary patterns in systems with high levels of gene flow (Christie 2009). This approach has become particularly useful for disentangling patterns of dispersal in many biological systems. The first applications of parentage analysis to estimate dispersal were conducted with plants, with the aim of obtaining estimates of pollen immigration in wild populations (Ellstrand & Marshall 1985). Since then it has been used to address questions of dispersal in a wide variety of organisms including rodents (Waser et al. 2006; Nutt 2008), insects (Tentelier et al. 2008) and coral reef fish (Jones et al. 2005; Planes et al. 2009; Saenz-Agudelo et al. 2009).

Stream-dwelling fish populations, and especially salmonids, have long provided a major source of information on animal movement (Rodriguez 2002). A common finding is of intrapopulation heterogeneity in movement, where a substantial portion of any given population is highly sedentary. The brook charr (Salvelinus fontinalis) appears to generally display this pattern (Hutchings & Gerber 2002; Rodriguez 2002), but so far (and as in other species), data are only available from adult or large-bodied individuals. The apparent contradiction between these results illustrates the limitations of both methods for obtaining accurate estimation of movement patterns (Wilson et al. 2004).

In an effort to better understand the movement throughout the entire life cycle of stream-dwelling brook charr (Fig. 1), Morrissey & Ferguson (2011) conducted a 4-year capture-recapture study in a freshwater river over multiple intervals and long portions of the life cycle. The first step was to sample 835 mature individuals in 2002 and then sample their offspring cohort twice a year except at the very end, where the last two samples were collected only in the (fall) spawning seasons. The major challenge was first to determine the movement of juveniles and between juvenile and mature stages. The second challenge was to genetically identify the individuals after each sampling effort to determine their movement since their last capture. By combining population genetics, parentage analysis and statistical models to estimate individual movements, the authors showed that substantial gene flow occurred throughout lower and upper portions of the freshwater river. The gene flow was biased in the downstream direction with a maximum likelihood of migration from the upper to the lower population of 8.37 (95% CI: 7.98–8.78) against 6.83 (95% CI: 6.51–7.16) from the lower to the upper population.

image

Figure 1.  Photograph of a brook charr, Salvelinus fontinalis, from freshwater river (Photograph credit to Jennie Knopp).

Download figure to PowerPoint

This result was explained by the observation of juvenile movement from their redd sites by assigning each alevin to a couple of parents (or only one parent) and comparing their recapture position with the redd/parental initial position. The interesting point of this result is that through assignment and parentage analysis, the authors were able to infer movement at very early life stages, which would be impossible with physical capture-mark-recapture markers (tags, dye etc.). It should be noted as well that a lot of genetic homogeneity is explained by the characterization of the distribution of alevin movements.

When analysing the total movement of the individuals captured as juveniles and again as adults, it appears that a single distribution model best explains net movement, refuting the hypothesis of intrapopulation heterogeneity supported by previous analysis conducted on adult or large-bodied individuals (Hutchings & Gerber 2002; Rodriguez 2002). In addition, studying the movement of the brook charr by partitioning the sampling throughout the life cycle revealed that movements occurred during the earlier life stages before they became more sedentary adults, hence resolving the discordance between the previously reported direct and indirect estimates of migration (Wilson et al. 2004).

This study highlights the insights that can be gained when by combining different analysis methods to study individual movement within a population. It also brings a new methodological approach to capture-mark-recapture studies that should decrease movement biases and handling time of the fish, as well as minimizing behavioural modifications in response to physical tagging components on the body of the fish. Morrissey and Ferguson’s work also underlines the importance of sampling scheme choices and of parentage analysis when estimating population dynamics.

The main limitation of such study is the difficulty of obtaining sufficient microsatellite markers to perform accurate parentage analysis. This consideration can be an issue when working on non-model species, although acquiring subsequent polymorphic markers is now more convenient with the arrival of next generation sequencing. A further limitation is that the sampling method described in Morrissey and Ferguson is dependent on the comportment and life cycle of the studied species. One must also take into account that the environment in which the species occurs: in this study, the authors had the advantage of studying an easily accessible one-dimensional environment.

The prospective uses of this approach are very broad – they can lead to an improved knowledge of individual movements in species that can be difficult to observe or capture, or to assess movements of animals with long-distance migration, such as Atlantic eels, leatherback turtles or tiger sharks. Furthermore, with enough genetic information and the capacity to achieve a sampling of candidate parents, parentage analysis can be applied to large populations with high migration, making it possible to uncover patterns of dispersal in scenarios where indirect methods fail (Waples & Gaggiotti 2006). Such individual-based genetic identification also opens a new window in population genetics. Recent works on kin selection (Buston et al. 2009) and on individual selective mortality are among the first to use these new approaches, and many others are coming as the genetic markers make these protocols possible. Because obtaining large numbers of highly variable genetic markers is no longer time-consuming or unaffordable, individual-based genetic analysis will become an ultimate tool in many approaches dealing with connectivity and local selection. Further limitations will now come from the completeness of sampling of candidate parents that remains hard to achieve in the sea.

References

  1. Top of page
  2. Abstract
  3. References