Spatial variation in population density across the geographical range in helminth parasites of yellow perch Perca flavescens
Robert Poulin and Terry A. Dick R.
Poulin (robert.poulin@stonebow.otago.ac.nz), Dept of Zoology, Univ. of Otago, PO Box 56, Dunedin 9054, New Zealand. Б T. A. Dick, Dept of Zoology, Univ. of Manitoba, Winnipeg, MB, Canada R3T 2N
species Ecography 30: 629Б636, 2007 doi: 10.1111/j.2007.0906-7590.05139.x © 2007 The Authors. Journal Compilation © 2007 Ecography Subject Editor: Carsten Rahbek. Accepted 2 July 2007 .
The abundance of a species is not constant across its geographical range; it has often been assumed to decrease from the centre of a species’ range toward its margins. The central assumption of this ‘‘favourable centre’’ model is tested for the first time with parasites, using different species of helminth parasites exploiting fish as definitive hosts. Data on prevalence (percentage of hosts that are infected) and abundance (mean no. parasites per host) were compiled for 8 helminth species occurring in 23 populations of yellow perch Perca flavescens, from continental North America. For each parasite species, correlations were computed between latitude and both local prevalence and abundance values. In addition, the relationships between the relative prevalence or abundance in one locality and the distance between that locality and the one where the maximum value was reported, were assessed separately for each species to determine whether abundance tends to decrease away from the presumed centre of the range, where it peaks. For both the cestode Proteocephalus pearsei and the acanthocephalan Leptorhynchoides thecatus, there was a positive relationship between prevalence or abundance and the latitude of the sampled population. There was also a significant negative relationship between relative prevalence and the distance from the locality showing the maximum value in P. pearsei, but no such pattern was observed for the other 7 parasite species. Since this single significant decrease in prevalence with increasing distance from the peak value may be confounded by a latitudinal gradient, it appears that the distribution of abundance in parasites of perch does not follow the favourable centre model. This means that the environmental variables affecting the density of parasites (host availability, abiotic conditions) do not show pronounced spatial autocorrelation, with nearby sites not necessarily providing more similar conditions for the growth of parasite populations than distant sites.
The abundance of any given plant or animal species is not constant across its entire geographical range; typically, individuals of one species occur at high densities in some areas, but are only sparsely dis- persed in other parts of their range (Rapoport 1982, Hengeveld 1990, Sagarin et al. 2006). Several empirical studies indicate that this spatial variation can follow a regular pattern: the density of individuals may decrease from the centre of a species’ range toward its margins. This is true for at least some species of plants, insects and vertebrates (Whittaker and Niering 1965, Whittaker 1967, Hengeveld and Haeck 1982, Brown 1984, Root 1988, Telleria and Santos 1993). There are, however, many exceptions to this simple regular pattern (Sagarin and Gaines 2002, Gaston 2003, McGeoch and Price 2004, Murphy et al. 2006), and several alternative scenarios are possible.
In simple terms, a decrease in density from the centre toward the margins can be explained by a relationship between environmental conditions and location within the geographic range of a species. Conditions for survival and reproduction may be most favourable at the centre of the range, and become progressively worse as one proceeds toward the margins (Hengeveld 1990). This does not hold in many cases (Sagarin and Somero 2006), and remains mostly an untested assumption. It would require that local populations respond to local conditions, and that the density achieved locally reflects the extent to which local environments meet the niche requirements of a species Brown 1984, 1995). This niche-based explanation also assumes that the environmental variables affecting the density of a species tend to be autocorrelated over space, such that sites in close proximity have more similar environmental conditions than distant sites (Brown 1984, 1995). Since environmental conditions are likely to change more-or-less continuously across the land- scape, the ‘‘favourable centre’’ model predicts a uni- modal distribution of density in space, peaking near the centre of the range and dropping off along a gradient toward the margins. In contrast, if environ- mental factors, especially those that are important niche parameters, do not change gradually but instead irregularly over space, or periodically over time, then the density of a species is likely to show multimodal variation across its geographical range (Brown 1995). Spatial variation in density in this ‘‘local oasis’’ model would resemble an irregular series of peaks and valleys. With data on density of a species from many sites within its range, there is a simple empirical test to distinguish between these two models. Under the ‘‘favourable centre’’ scenario, the site of maximum observed density should be at, or very near, the centre of the species’ distributional range; thus, when local density is plotted against distance from the site with maximum density, one would expect a significant negative relationship for the ‘‘favourable centre’’ model. In contrast, no relationship at all is expected between local density and distance from the site of maximum observed density for the ‘‘local oasis’’ model.
In the present study, we test the central assumption of spatial autocorrelation underpinning the ‘‘favourable centre’’ model, with data on populations of parasites in freshwater fish hosts. For parasites, local density in a host population is best measured as either prevalence (the proportion of hosts that are infected) or intensity of infection (the mean number of parasites per infected host), or, even better, their product which is referred to as abundance (the mean number of parasites per host, including uninfected ones). The prevalence, intensity and/or abundance of parasites of a given species vary among different populations of their hosts, but within species-specific bounds (Arneberg et al. 1997, Poulin 2006, Krasnov et al. 2006). However, the geographic structure of this variation has never been examined. In the case of parasites of freshwater fishes, the parasites’ distributions will be just as discontinuous as those of their hosts, which only occur in lakes and rivers; on a large geographical scale, however, patterns across the whole distributional range should still be detectable, if they exist. The local abundance of a parasite species will not only depend on the local availability of its fish hosts, but also on the local availability of intermediate hosts necessary for the completion of the parasite’s life cycle, and on the abiotic conditions to which the parasite’s infective stages are exposed. These variables will need to be tightly autocorrelated over space for the ‘‘favourable centre’’ model to apply to parasites (Brown 1984, 1995).
Here, the relationship between both local preva- lence, intensity and abundance of infection and the distance from the site where these values peak, is quantified for several helminth parasites of yellow perch Perca flavesvcens, a widespread North American fresh- water fish. The present study provides the first analysis of the spatial distribution of abundance across the geographical range of parasite species, and points toward local factors as important determinants of population abundance within different parasite species.
ليست هناك تعليقات:
إرسال تعليق