Managing Tropical Rice Pests Through Conservation of Generalist Natural Enemies and Alternative Prey

Abstract

The cultivation of tropical Asian rice, which may have originated 9000 yr ago, represents an agricultural ecosystem of unrivaled ecological complexity. We undertook a study of the community ecology of irrigated tropical rice fields on Java, Indonesia, as a supporting study for the Indonesian National Integrated Pest Management Programme, whose purpose is to train farmers to be better agronomists and to employ the principles of integrated pest management (IPM). Two of our study objectives, reported on here, were (1) to explore whether there exist general and consistent patterns of arthropod community dynamics related to natural or intrinsic levels of biological control, and (2) to understand how the existing levels of biological control are affected by insecticide use, as well as by large—scale habitat factors relating to differing patterns for vegetational landscapes, planting times, and the length of dry fallow periods. We performed a series of observational studies and two experimental studies. Abundant and well—distributed populations of generalist predators can be found in most early—season tropical rice fields. We took samples from plants and water surface using a vacuum—suction device, and from the subsurface using a dip net. Our results show that high populations of generalist predators are likely to be supported, in the early season, by feeding on abundant populations of detritus—feeding and plankton—feeding insects, whose populations consistently peak and decline in the first third of the season. We hypothesize that since this abundance of alternative prey gives the predator populations a "head start" on later—developing pest populations, this process should strongly suppress pest populations and generally lend stability to rice ecosystems by decoupling predator populations from a strict dependence on herbivore populations. We experimentally tested our hypothesis of trophic linkages among organic matter, detritivores and plankton feeders, and generalist predators and showed that by increasing organic matter in test plots we could boost populations of detritivores and plankton—feeders, and in turn significantly boost the abundance of generalist predators. These results hold for populations found on the plant, on the water surface, and below the water surface. We also demonstrated the link between early—season natural enemy populations and later—season pest populations by experimentally reducing early—season predator populations with insecticide applications, causing pest populations to resurge later in the season. Overall, these results demonstrate the existence of a mechanism in tropical irrigated rice systems that supports high levels of natural biological control. This mechanism depends on season—long successional processes and interactions among a wide array of species, many of which have hitherto been ignored as important elements in a rice ecosystem. Our results support a management strategy that promotes the conservation of existing natural biological control through a major reduction in insecticide use, and the corresponding increase in habitat heterogeneity.