Scale of Predator-Prey Interactions


Spatial structure of bivalves in a sandflat: Scale and generating processes

Journal of Experimental Marine Biology and Ecology
Vol. 216, Issue. 1-2, pp.99-128, 1997

P. Legendre b, S. F. Thrush a, V. J. Cummings a, P. K. Dayton c, J. Grant d, J. E. Hewitt a, A. H. Hines e, B. H. McArdle f, R. D. Pridmore a, D. C. Schneider g, S. J. Turner a, R. B. Whitlatch h and M. R. Wilkinson a

a National Institute of Water and Atmospheric Research, PO Box 11-115 Hamilton New Zealand
Département de sciences biologiques, Université de Montréal, C.P. 6128, succ. Centre-ville Montréal, Québec H3C 3J7 Canada
c Scripps Institution of Oceanography, UC-SD La Jolla, CA 92093-0201 USA
d Department of Oceanography, Dalhousie University Halifax, Nova Scotia B3H 4J1 Canada
e Smithsonian Environmental Research Center, P.O. Box 28 Edgewater, MD 21037 USA
f Biostatistics Unit, School of Biological Sciences, University of Auckland, Private Bag Auckland New Zealand
g Ocean Sciences Centre, Memorial University, St. John's Newfoundland A1C 5S7 Canada
h Department of Marine Science, University of Connecticut, Avery Point Groton, CT 06340-6097 USA


A survey was conducted during the summer of 1994 within a fairly homogeneous 12.5 ha area of sandflat off Wiroa Island, in Manukau Harbour, New Zealand, to identify factors controlling the spatial distributions of the two dominant bivalves, Macomona liliana Iredale and Austrovenus stutchburyi (Gray), and to look for evidence of adult-juvenile interactions within and between species. Most of the large-scale spatial structure detected in the bivalve count variables (two species, several size classes of each) was explained by the physical and biological variables. The results of principal component analysis and spatial regression modelling suggest that different factors are controlling the spatial distributions of adults and juveniles. Larger size classes of both species displayed significant spatial structure, with physical variables explaining some but not all of this variation. Smaller organisms were less strongly spatially structured, with virtually all of the structure explained by physical variables. The physical variables important in the regression models differed among size classes of a species and between species. Extreme size classes (largest and smallest) were best explained by the models; physical variables explained from 10% to about 70% of the variation across the study site. Significant residual spatial variability was detected in the larger bivalves at the scale of the study site. The unexplained variability (20 to 90%) found in the models is likely to correspond to phenomena operating at smaller scales. Finally, we found no support for adult-juvenile interactions at the scale of our study site, given our sampling scale, after controlling for the effects of the available physical variables. This is in contrast to significant adult-juvenile interactions found in smaller-scale surveys and in field experiments. Our perception of adult-juvenile interactions thus depends on the scale of study.

Author Keywords: Adult-juvenile interaction; Autocorrelograms; Austrovenus stutchburyi ; Bivalves; Macomona liliana ; Spatial modelling; Spatial structure