Experimental designs for the study of allelopathy

The primary aim of this paper is to discuss the methodological approaches that may best develop studies of allelopathy in the future. Laboratory studies on the functions of isolated chemicals, no matter how mechanistically detailed, cannot demonstrate the significance of allelopathy in communities. Evidence for allelopathy in natural plant communities should include information of concentrations and release rates such as demonstrated in field soils for (±)-catechin and Centaurea maculosaLam. Community-relevant evidence for allelopathy should include some manipulation of exudates such as performed in many experiments with activated carbon and gel filtration columns. Realistic evidence for allelopathy should include separation of resource effects from chemical effects; such as demonstrated by experiments with activated carbon additions, density-dependent responses to additions of competitors and chemicals, and resource addition treatments. Community-relevant evidence should link laboratory effects to field patterns and experiments; such as the links between the inhibitory effects of roots of Larrea tridentataCov., the highly spatially segregated root systems and regular above-ground spacing of this species, strong spatial disassociation of L. tridentata with other species, and removal experiments indicating that segregation of L. tridentata root systems via allelopathy may feed back to sequestering of resource use. Studies of allelopathy should consider chemically initiated shifts in microbial populations, and the effects of organic and inorganic soil components on the function of exudates; which has been done in a number of studies. Finally, studies of allelopathy should include large-scale manipulation of chemical effects; such as performed in field experiments in boreal forests in Sweden with Empetrum hermaphroditumHagerup and Scots pine (Pinus sylvestrisL.). Demonstrating the occurrence and importance of chemically mediated interactions among plants is not trivial. If even a small portion of the thousands of chemicals produced by different plant species have effects on their neighbours, then species-specific interactions, natural selection, community integration, and community coevolution may be quite different than predicted by conceptual models based solely on resource competition. Appropriate methodology is crucial for integrating chemically mediated interactions into ecological theory.