Effect Of Polycultural Manipulation On The Structure And Dynamics Of Arthropod Communities In Irrigated Rice-Based Ecosystems

The loss of biodiversity in rice ecosystems is one of the main causes of rice insect pest out-breaks. Methods of biodiversity enhancement of rice ecosystems have therefore been promotedunder integrated pest management and other sustainable management practices. Non-rice habi-tats, especially non-crop habitats surrounding rice fields, were shown to be important to the nat-ural enemies of insect pests. However, few attempts have been made to evaluate the ecologicalrole of more profitable crops that could be used to diversify rice ecosystems and aid in naturalenemy conservation and insect pest suppression. In2008and2009, soybean and corn intercropswere examined for their potential to diversify rice ecosystems and enhance insect pest manage-ment. The composition, structure, temporal dynamics and movements of arthropod guilds withinarthropod communities were evaluated using sampling methodologies that included vacuumsampling and directional traps (pitfall, sticky and malaise). Arthropod communities were dividedinto13guilds: ants, detritivores, herbivores (including chewers, miners and sap-suckers), parasi-toids, hyperparasitoids, predators (including predatory insects and spiders), tourists and naturalenemies.The species richness of each arthropod guild in the rice fields was slightly higher in diculturesystems than in monoculture systems. When compared to the species richness of arthropodguilds in rice fields in rice monoculture systems (RMon), a significant improvement in speciesrichness was recorded in2009among herbivores, miners and spiders in rice fields in rice-soybeansystems (RS). However, the diversification of the rice ecosystems had minimal effects on thetemporal dynamics of species richness and the relative abundance of each arthropod guild.In contrast, rice fields in rice-corn system (RC) showed no significant change in the abun-dance of arthropod guilds except for sap-suckers, predators and natural enemies, all of whichshowed significantly higher abundance in RC than in RMon in2009. There was no significant dif-ference in the abundance of arthropod guilds between RS and RMon except that the abundancesof detritivores, herbivores and miners were significantly higher in RS than in RMon in2009. Poly-cultural manipulation did not significantly lower the abundance of herbivores, nor did it signifi-cantly increase the abundance of natural enemies.In2009, the value of the ratio of arthropod species richness (S) to the abundance of arthro-pods (N), S/N, was significantly higher in RC than in RMon and RS, and there was no significantdifference in the value of S/N between RMon and RS. In2008and2009, there were no significantdifferences in the value of the ratio of species richness of natural enemies (Sn) to species richness of herbivores (Sp), Sn/Sp, in rice fields under different cropping systems. However, the value of theratio of the abundance of natural enemies to the abundance of herbivores, Nn/Np, was signifi-cantly higher in RC than in RMon in2008, although there were no significant differences in thevalue of Nn/Npbetween RS and RMon in either2008or2009.The species abundance patterns for all arthropods as indicated by the species rank/relativeabundance curves was significantly different in rice fields under different cropping systems; thedistribution of the species abundance was more consistent in diculture than in rice monoculture.However, the cumulative species abundance curves showed that there were no significant dif-ferences in the distribution of species abundance for the arthropod communities or guilds in ricefields under different cropping systems.The Jaccard similarity coefficients for the arthropod communities and guilds in the rice fieldswere not significantly different among the different cropping systems, with mean coefficient val-ues falling between0.46and0.73. Similarly, no significant differences in the Morisita-Horn simi-larity coefficients were found for the arthropod communities and guilds in the rice fields underdifferent cropping systems, except for the total arthropods and tourists in2009. The mean valuesof the Morisita-Horn similarity coefficients for the arthropod communities and guilds fell be-tween0.53and0.98; the similarity values for tourists were less than0.23between RS and RMonand between RS and RC.Species richness based on the movement of arthropod guilds was lower, but not significant,in diculture systems than in rice monoculture systems. In2008, the species richness of predatoryinsects sampled by pitfall traps was significantly lower in diculture than in rice monoculture, andthe species richness of spiders was significantly lower in rice-corn systems than in rice monocul-ture systems. The species richness of hyperparasitoids and parasitoids sampled by sticky trapswas significantly lower in diculture than in rice monoculture, as was the species richness of min-ers sampled by malaise traps in rice-corn systems relative to rice monoculture.The abundance of emigrant predators and parasitoids was not significantly increased in di-culture than in rice monoculture. The abundance of immigrant natural enemies was not signifi-cantly higher in diculture than in rice monoculture with the exception of the abundance of immi-grant parasitoids, which was significantly higher in rice-soybean systems than in rice monoculturesystems. The abundance of emigrant and immigrant sap-suckers, rice leafhoppers and Ery-throneura subrufa was significantly higher in rice-soybean systems than in rice-corn systems andrice monoculture systems. However, the number of rice leafhoppers (predominantly Cicadulafasciifrons), was not significantly different under different cropping systems. Compared to RMon, RS had lower rice planthopper abundance in2008, but higher abun-dance in2009, though neither value was significantly different. The abundance of rice planthop-pers was significantly reduced in RC than in RMon and RS, where RC had26%47%fewerplanthoppers than RMon. However, neither parasitoid nor predator numbers were affected bypolycultural manipulation. There were no significant differences in the directional movements ofplanthoppers or natural enemies between crop subplots in the different cropping systems. Ingeneral, the movement of planthoppers was very limited.Overall, those rice fields with corn intercrops significantly improved the S/N and Nn/Npval-ues in2008and2009, respectively; rice fields with corn intercrops had a lower abundance ofplanthoppers, which contributed to lower sap-sucker abundance in RC than in RMon in2009. Thisstrategy could therefore be useful as part of an ecologically based pest management strategy forthe sustainable production of rice.