MATCHING THE OUTCOME OF SMALL-SCALE DENSITY MANIPULATION EXPERIMENTS WITH LARGER SCALE PATTERNS AN EXAMPLE OF BIVALVE ADULT JUVENILE INTERACTIONS/

Generalising or scaling up from small-scale experiments to lar er area s is an important challenge for both ecology and conservation biology. This study describes a technique that attempts to meet this challenge by combining spatial mapping with small-scale process experiments. Sp ecifically, we evaluate the density effects of large individuals (> 15 mm shell length) of a tellinid bivalve (Macomona liliana Iredale) on macrofauna in 0.25 m(2) experimental plots within the natural density variation of large Macomona over a 12.5 ha site. By mapping the spatia l distribution of large Macomona before conducting the experiment, we were able to identify homogeneous areas with different background dens ities of large Macomona and embed 22 experimental locations within the natural density-scape. Within each location, four experimental densit ies were added to plots from which all large macrofauna (>4 mm) had be en previously removed. Macrofauna were sampled 22 days after the start of the experiment and significant negative treatment effects of high densities of large Macomona were identified by ANOVA for juvenile biva lves Macomona (<4 mm), Austrovenus stutchburyi (Gray) (<4 mm), the iso pod Exosphaeroma falcatum Tattersall and the total number of individua ls. Generalised linear models were then used to include the effect of background density variation of large Macomona in the analysis. Only A ustrovenus (<4 mm) demonstrated a significant interaction between the background and experimental densities of large Macomona. This resulted from background densities of large Macomona having a significant effe ct on Austrovenus (<4 mm) in the two lowest density treatments only. S ignificant effects were detected only because we had planned the study to cover the various background densities of Macomona. The effect of experimental and background density variation of large Macomona on Mac omona (<4 mm), Exospheroma, nemerteans and the total number of individ uals were similar in direction and strength. Except for nemerteans, al l relationships were negative, with low densities of macrofauna associ ated with high experimental and background densities of large Macomona . This implies that large-scale extrinsic factors (e.g., elevation, ex posure to wave disturbance) are not the only features influencing the distribution of Macomona at the scale of the study site; intrinsic pro cesses operating on smaller seals are also important. This scale-depen dent response would not have been uncovered, had we not conducted a la rger-scale survey in concert with the smaller-scale experiment. (C) 19 97 Elsevier Science B.V.