Venue and outcome in ecological experiments: manipulations of larval anurans

The choice of venue is believed to be a critical decision for ecologists co nducting experiments, yet the existence of tradeoffs in design and the impa ct of venue on outcome have not been quantitatively evaluated. To these end s we reviewed a set of 227 comparisons from 52 studies designed to quantify the effect of density on the growth performance of larval anurans. We pred icted that both design and outcome would be related to experimental venue ( laboratory, mesocosm, and field). We found that, in contrast to our predict ions, laboratory experiments did not include more factors or yield more pre cise estimates of responses compared with mesocosm and field manipulations. In partial support of our prediction, we found that laboratory experiments did utilize somewhat larger numbers of replicates and included more levels per factor compared to the alternative venues. In addition we found that, as predicted, raising tadpoles at higher densities of conspecifics, or in t he presence of heterospecifics tended to decrease growth. This effect, cons istent with competition, was strongly dependent on experimental venue. In p articular. we found that interspecific effects were much stronger in mesoco sm versus field manipulations. This result is particularly relevant because conceptions of amphibian assemblages have included interspecific competiti on as integral to natural patterns of amphibian distribution. We conclude t hat hypothesized tradeoffs among experimental venues have not acted in a pa rticularly strong way in shaping the design of tadpole density manipulation s. However, venue may mediate strong effects on the outcome of such experim ents. Quantification of such methodology based biases will help ecologists more effectively match the goals of their experiments with decisions regard ing venue.