Distinguishing Causes of Intraspecific Synchrony in Population Dynamics

Abstract

We examined the degree and cause of intraspecific synchrony in population dynamics between 29-42 populations for each of seven mammalian species. Regions containing multiple populations with similar dynamics were identified using cluster analysis. Two explanations for the observed synchrony were evaluated, dispersal and the Moran effect, a spatially correlated density independent perturbation, such as weather, which occurs across populations. Populations synchronized by dispersal are expected to exhibit a negative relationship between synchrony and distance, while populations synchronized by a Moran effect must have a similar density-dependent structure (DDS). To distinguish between these hypotheses we used autoregressive modeling to determine the DDS of each population, and cross-correlation to estimate the degree of synchrony between populations. Dispersal alone was assumed to be responsible for synchrony in populations with a significant negative synchrony-distance (S-D) relationship and heterogeneous DDS. We assumed that a Moran effect ould have produced synchrony in populations with homogeneous DDS and non-significant S-D relationship. It was not possible to assign synchrony to a single factor in regions with a homogeneous DDS and a significant S-D relationship, and we excluded the possibility of a Moran effect in regions with heterogeneous DDS and a non-significant S-D relationship. Using these criteria, we identified dispersal as synchronizing populations within one region each of ermine, lynx, mink and red fox and in two regions of muskrat. A Moran effect may have synchronized one region of ermine, fisher, lynx, and mink, although we were unable to identify a causative factor. One lynx and one mink region showed characteristics of both dispersal and Moran-based synchrony, and 9 regions of synchrony could not be assigned to either factor. These results show that by examining the DDS and the S-D relationship we were able to determine the factors most likely responsible for synchronized population dynamics in 10 of 21 cases. Possible cases of Moran-based synchrony do not appear to be common, occurring in only 18.2% of these regions. It should also be noted that just as correlation does not imply causation, presence of homogeneous DDS does not indicate a Moran effect, only its possible occurrence.