| We studied the effects of landscape composition on population abundances of main natural enemies as well as their biocontrol services in northern China using the 3S (RS, GIS, GPS) technology as well as landscape and population eclology and DNA methodology. The results are listed as follows:Principal component analysis (PCA) and an adjusted Akaike’s Information Criterion analysis (AICc) of the data from 41 study sites indicated that landscape composition and diversity affected the abundance of predators on cotton, but the effect varied among different enemy taxa. The models of PC3 at 1 km, PC1+PC3 at 2 km and PC3 at 0.5 km had the lowest AICc values which were the best fit models contained 22.3%,43% and 29.1% of Akaike weight and explained 5.5%,30.3% and 9.1% of the variability in abundance of ladybeetle, lacewings and Orius spp. across sampling sites, respectively. PC3 was weakly positive with ladybeetles at 1 km scale, which was significantly negative with lacewings at 2km whereas significantly positive with Orius spp. abundance at 0.5 km radius. The model of PC3 at 1 km had the lowest AICc value which was the best fit model contained 22.8% of Akaike weight and explained 5.1% of the variability in abundance of all predators in cotton field. However, no powerful competing model can best predict the relationship between landscape variables and abundance of spiders. These results indicated landscape with a high proportion of some small crops (such as peanut, soybean, sweet potato, vegetable crops and fruit trees) surrounding cotton field supported high abundance of ladybeetles, lacewings and Orius spp. In addition, the abundance of lacewings were positively correlated with the proportion of non-crop habitats and urban habitats but was negatively correlated with the main crops (e.g., cotton and maize), whereas the proportion of non-crop habitats and urban habitats had a negative correlation with ladybeetle abundance. The abundance of spiders was not significantly negative correlated with non-crop habitats and urban habitats as well as a weakly positive correlation with the proportion of cotton and maize.Landscape diversity enhanced the aphid parasitoids abundance and their parasitism on cotton at large spatial scales (1-2 km radius) across all 35 study sites. The model of PC3 at 0.5 km and PC1 at 2 km had the lowest AICc values which were the best fit models contained 45.1% and 23.1% of Akaike weight and explained 53.1% and 24.3% variation of mummy abundance and parasitism, respectively. The mummy abundance was significantly positive correlated with PC3 at 0.5 km, whereas the parasitism was significantly positive correlated with PC1 at 2 km. The proportion of mummy ranged from 0-53.5% with a mean value of 12.5%, which was supported with a higher proportion of other host crops than cotton and maize (0.5-2 km) and urban habitats (1-2 km), and was negative correlated with the proportion of cotton and maize.More than 30,000 sentinel eggs of Helicoverpa armigera were exposed in cotton fields for 48 h at 23 sites. The parasitism rates ranged from 0-38.8% on cotton The model of PC2+PC3 at 2 km had the lowest AICc value which was the best fit model contained 28.4% of Akaike weight and explained 45.4% of variation in H. armigera parasitism by Trichogramma chilonis in cotton field, howerer, parasitism rate was not significantly negative correlated with PC2 but was significantly positive correlated with PC3 at this scale. The results showed high proportions of urban and non-crop habitat surrounding cotton field were beneficial for the parasitism on cotton.In recent years, cotton planting adjacent maize field have been popular in the northern China, especially the proportion of maize planting increased rapidly, which may affect the population occurrence of Helicoverpa armigera and its parasitoids. So we also deployed more than 50,000 sentinel eggs of H. armigera in maize fields in 33 sites. The parasitism rates ranged from 0-25.8% on maize, which were significantly negative correlated with the proportion of the total area of cotton and maize and the whole arable area whereas significantly positive correlated with landscape diversity index (D) and other small host crops across all these four scales, it was also positive correlated with the proportion of non-crop habitats at 0.5 and 1 km radius scales (VIP>0.8). PLS (Partial Least Squares) analysis indicated landscape diversity and high proportion of small host crops greatly enhanced the parasitism of T. chilonis on the eggs in maize field from 0.5 to 2 km radius scales. Parasitism rate was significantly positive correlated with the proportion of non-crop habitat at 0.5-1 km radii but negative correlated with the proportion of main crop (such as cotton and maize) and arable areas. In the food shortage period, the small host crops and non-crop habitats can provide pollen, nectar or other alternative preys and serve as shelter or temporary habitats to support the parasitoids, and then favoring their biocontrol services on suppressing the pest.This study provides a primary understanding of the relationship between landscape structure and the community of natural enemies and their ecosystem services for insect pest management in cotton in changing small farming systems, especaly the decreasing of some small crops planting following the rapid urbanization and agricultural intensification. It will be useful for further developing conservation biological control strategy in the specific agricultural ecosystems in China. |
PDF Full Text Request