Indirect versus direct effects of grasses on growth of a cactus (Opuntia fragilis): insect herbivory versus competition.

Variation in plant performance between microhabitats is usually attributed to direct mechanisms, such as plant physiological tolerances or competitive interactions. However, indirect mechanisms, such as differences in herbivore pressure mediated by microhabitat differences, could create the same pattern of variation. In this study, we investigated the effect of insect herbivore pressure on the growth of the grassland cactus Opuntia fragilis under different regimes of grassland canopy cover. Our purpose was to establish the extent to which canopy cover plays a direct, competitive role versus an indirect, mediatory role in cactus growth. We manipulated aboveground microhabitat, specifically the cover of adjacent grasses. The three treatments were: (1) open canopy, with grass pinned down away from the cactus; (2) shaded canopy, with a partial mesh cage staked over the cactus; and (3) ambient grass canopy. We measured seasonal plant growth and recorded changes in insect herbivore occurrence and damage in relation to cover. Cactus growth, defined as the change in number of live cladodes, was higher in the open than under either treatment where the plant was more shaded (P<0.05). However, allocation to new growth, measured as the proportion of new segments (cladodes) in a patch, did not differ among cover treatments. Thus, the hypothesis that physiological constraints, or competition for light, limited cactus performance in grass is rejected. Instead, we found that both cladode mortality, caused by the larvae of a cactus moth borer (Melitara dentata), and occurrence of the moth were lower in the open microhabitat than in either shaded microhabitat. Thus, higher net growth in the open, unshaded treatment, rather than representing a release from competition for light with grasses, was better explained as an indirect effect of grass cover on the activity and impact of the cactus moth. These results show that indirect effects can lead to a misinterpretation of experimental data on direct effects. These data also contribute to an improved understanding of mixed results in the biological control of weedy cacti. Clearly, future evaluations of the relative importance of physiology, competition, and insect herbivory in plant performance must be environmentally explicit.