Field assessment of effects of control strategies against the olive fruit fly (Bactroceraoleae (Rossi)) on predatory arthropods: comparison of different methods of data analysis

Abstract: Numbers of predators captured in a one-year field trial in a Spanish olive grove thatreceived different treatments to control the olive fly (Bactrocera oleae (Rossi)) were analysed withtwo different statistical analysis methods. The treatments were kaolin (Suround WP), trichlorfon baitspray as a positive control (Trichlorfon + protein hydrolysate (Nulure®) and unsprayed control. Twoapplications of kaolin and one bait spray were done, and fourteen samplings were carried out using abeating method.Two methods of analysis were used: Analysis of variance (ANOVA) and Principle Response Curveanalysis (PRC). One-way ANOVA was carried out on total numbers of predators captured at eachsampling date and two-way ANOVA was carried out on numbers of specimens of the most abundanttaxa, with “treatment” and “sampling date” as factors. Changes in abundance of the different taxa ofpredators were also investigated using the multivariate method Principle Response Curves (PRC)analysis.Results/conclusions based on one-way ANOVA and PRC analysis were similar, as both showed asignificant decrease of the predatory arthropod community in the kaolin treated trees. PRC shows theeffect of kaolin on the community of predators in an easier and quicker way, but only allowscomparisons between two treatments. Also, PRC gives a global result for the whole season while byusing ANOVA it is possible to pinpoint dates in which the effect of treatment is significant.Both PRC and two-way ANOVA identified the coccinellid Scymnus mediterraneus Iablokoff-Khnzorian and the spider family Philodromidae as the taxa the most affected by kaolin. For other,less numerous predatory groups it seems that in general PRC analysis was able to detect effects ofthe kaolin treatment which were not revealed by two-way ANOVA analysis.