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 areas 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. Specifically, 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 spatial distribution of large Macomona before conducting the experiment, we were able to identify homogeneous areas with different background densities of large Macomona and embed 22 experimental locations within the natural density-scape. Within each location, four experimental densities were added to plots from which all large macrofauna (>4 mm) had been 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 bivalves Macomona (<4 mm), Austrovenus stutchburyi (Gray) (<4 mm), the isopod Exosphaeroma falcatum Tattersall and the total number of individuals. Generalised linear models were then used to include the effect of background density variation of large Macomona in the analysis. Only Austrovenus (<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 effect on Austrovenus (<4 mm) in the two lowest density treatments only. Significant 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 Macomona (<4 mm), Exospheroma, nemerteans and the total number of individuals were similar in direction and strength. Except for nemerteans, all relationships were negative, with low densities of macrofauna associated with high experimental and background densities of large Macomona. This implies that large-scale extrinsic factors (e.g., elevation, exposure to wave disturbance) are not the only features influencing the distribution of Macomona at the scale of the study site; intrinsic processes operating on smaller seals are also important. This scale-dependent response would not have been uncovered, had we not conducted a larger-scale survey in concert with the smaller-scale experiment. (C) 1997 Elsevier Science B.V.