Scaling Natal Dispersal Distances: Confounding Factors

Audrey L. Mayer, Kenneth Petren, Alicia Shelton, Michael J. Cramer, Brian Keane, Jeffrey Markert, Ben Heath, Eric Maurer, J. Andrew Roberts, and Brandon Tonnis

• Responses to this Article
• Literature Cited

Published: April 1, 2002

Sutherland et al. (2000) posit several ecological and morphological characteristics of species that may be correlated with natal dispersal distance, and establish empirical models that could be used to predict the probability of dispersal limits. Their methodology raises several interesting hypothetical relationships and offers insight into areas in which more data on dispersal are clearly needed. However, we perceive three important issues that should be addressed before using this analysis for biological conservation. Below is an outline of our suggestions for improving upon the authors' examination of natal dispersal relationships and their use for conservation purposes.

First, the ecological and morphological characteristics analyzed may not be the best suited for conservation purposes. We would recommend examining baseline metabolic rate as well as body mass (regardless of allometric rescaling) or food guild, especially if dispersal is influenced by the distribution of food resources in the landscape. Allometric rescaling of birds and mammals may not be sufficient on its own to adjust for differences in metabolism that may drive dispersal through the landscape. Whereas cougars and shrews are both mammalian carnivores, the extremely high metabolic rate of shrews may mean that they have a greater dispersal distance per kilogram than do cougars. Likewise, reptilian carnivores have much lower metabolic demands per mass, so that one might expect shorter mean dispersal distances for them. Consequently, metabolism may be a better characteristic to use when comparing dispersal across taxa, because it encompasses both body mass and diet. In addition, it may be more appropriate to use juvenile body mass instead of adult body mass for those species in which dispersal occurs mainly among juveniles and in which juvenile mass at age of dispersal differs considerably from that of adults.

Second, the landscape matrix may influence natal dispersal to a greater degree than the ecological and morphological characteristics examined. The authors mention the importance of landscape context in their last paragraph, but we believe that this issue should be thoroughly incorporated into dispersal analyses. The importance of the permeability of the landscape matrix in which habitat fragments are located cannot be understated, although admittedly there are a great number of species and matrix habitat types for which this information is unavailable. However, dispersal data gathered from field studies inherently include information about the matrices in which the fragments are located, and these effects should somehow be teased apart from dispersal capability. Controlling for the permeability of the landscape matrix and the degree to which the matrix aids or hinders dispersal may have substantially affected the results of this study.

Third, the abundance or rarity of a species may influence dispersal distance. Although allometric or metabolic scaling may be a good general predictor of natal dispersal, the possibility of numerous specific exceptions to this rule must be seriously examined. For example, the authors did not distinguish rare from common species in their data pool, and hence make the assumption that rarity and natal dispersal are not related. Indeed, there is probably a great deal of variability within taxonomic groups, and subdividing these groups by other characteristics, such as rarity, in addition to taxonomic affiliation may better illustrate relationships between ecological and morphological characteristics and dispersal. As rare species are more likely to be threatened with extinction, these species are more often the focus of concerted conservation efforts. Examining the relationship between rarity, body size, metabolism, taxonomic affinity, and dispersal would greatly enhance the utility of this research to conservation efforts.

In their Fig. 4, Sutherland et al. (2000) did demonstrate a distinct relationship in birds and mammals between median and maximum dispersal distance. It occurs to us that this relationship, if verified through further data collection and analysis, might be used to help predict the loss of species richness as the distance between habitat fragments increases, with appropriate modifications for the landscape matrix. For example, as the distance between habitat fragments (represented as a vertical line on the graph in Fig. 4) increases from –1 (log₁₀ km) median dispersal distance to -0.5 (log ₁₀ km), the reserve network might lose an additional four mammal and four bird species. This species-distance model may be a corollary to the species-area curve that, if used properly, could help conservationists design networks of biological reserves or prevent the destruction of specific natural areas. More useful for conservation efforts would be the identification of the species that fall at the low end of the dispersal relationship; are they more likely to be rare, or more intolerant of a highly modified landscape matrix?

The results in Sutherland et al. (2000) and our recommendations here must be used with caution by policy makers. The variability inherent in ecological systems makes all predictions based on generalized models dangerous when they are overextended. A standard methodology for determining natal dispersal probability and the use of these data for conservation purposes may resolve several of these issues. Through standardization, the process of dispersal could be more reasonably compared between taxonomic and ecological groups, and generalization may then be appropriate. This methodology would also assign priorities to the different types of data that could be gathered in a specific time period. Because shorter-term studies would incorporate less temporal variability in abiotic and biotic conditions, these studies would have associated with them a larger margin of error. In this way, dispersal estimates would be associated with a degree of uncertainty and could be used conservatively.

Responses to this article are invited. If accepted for publication, your response will be hyperlinked to the article. To submit a comment, follow this link. To read comments already accepted, follow this link.

Sutherland, G. D., A. S. Harestad, K. Price, and K. Lertzman. 2000. Scaling of natal dispersal distances in terrestrial birds and mammals. Conservation Ecology 4(1): 16. [online] URL: http://www.consecol.org/Journal/vol4/iss1/art16