Modifying orchard planting sites conserves entomopathogenic nematodes,reduces weevil herbivory and increases citrus tree growth,survival and fruit yield

In Florida, a root weevil pest of citrus, Diaprepes abbreviatus, is more damaging and attains higher population density in some orchards on fine textured, poorly drained “flatwoods” soils than in those on the deep, coarse sandy soils of the central ridge. Previous research revealed that sentinel weevil larvae were killed by indigenous entomopathogenic nematodes (EPNs) at significantly higher rates in an orchard on the central ridge, compared to one in the flatwoods. We hypothesized that filling tree planting holes in a flatwoods orchard with sandy soil from the central ridge would provide a more suitable habitat for EPNs, thereby reducing weevil numbers and root herbivory. Fifty trees were planted in oversized planting holes filled with coarse sand and 50 trees were planted in native soil in a split plot design where whole plots were species of introduced EPNs and split plots were soil type. Each of Steinernema diaprepesi, Steinernema riobrave, Heterorhabditis indica, Heterorhabditis zealandica, or no EPNs were introduced into the rhizospheres in 10 plots of each soil type. During four years, EPN numbers in soil samples and the relative abundance of seven species of nematophagous fungi associated with nematodes were measured three times using real-time PCR. The efficacy of EPNs against sentinel weevil larvae was also measured three times by burying caged weevils in situ. EPN species richness (P = 0.001) and diversity (P = 0.01) were always higher in sand than native soil. Soil type had no effect on numbers of EPNs in samples, but EPNs were detected more frequently (P = 0.01) in plots of sandy soil than native soil in 2011. Two nematophagous fungi species, Paecilomyces lilacinus and Catenaria sp. were significantly more abundant in nematode samples from sandy soil on all three sampling dates. Efficacy of EPNs against weevil larvae was greater in sandy soil inoculated with S. diaprepesi (P = 0.03) in June 2010 and in all treatments in sandy soil in May 2011 (P = 0.03). Sixty-eight percent more adult weevils (P = 0.01) were trapped emerging from native soil during two years than from sandy soil. By May 2011, the cumulative number of weevils emerging from each plot was inversely related (P = 0.01) to the numbers of EPNs detected in plots and to EPN efficacy against sentinels. Three trees in sandy soil died as a result of root herbivory compared to 21 trees in native soil. Surviving trees in sandy soil had trunk diameters that were 60% larger (P = 0.001) and produced 85% more fruit (P = 0.001) than those in native soil. Although it is not possible to characterize all of the mechanisms by which the two soil treatments affected weevils and trees, substitution of sand for native soil was an effective means of conserving EPNs and shows promise as a cultural practice to manage D. abbreviatus in flatwoods citrus orchards with a history of weevil damage to trees.