Because high numbers of mites were found drowned in the felt after exposure to spinetoram in the preliminary trial, an additional study was conducted beginning at the same time as the May 4, 2010 field trial by confining mites to the leaf disc and not allowing them to suicide in the moist felt. This was done using Munger cell bio-assays . Briefly, Munger cells were constructed using a 3-layer Plexiglas “sandwich”; the middle cell layer was drilled with 3.2-cm diameter bit to provide a circular test arena . The upper and lower parts of the Plexiglas sandwich were solid and between the lower base and test arena a piece of filter paper was placed under the leaf to allow moisture exchange and to extend the life of the leaf during the bio-assay. Airflow through the test arena was provided through two holes drilled through the center cell layer directly opposite one another, with fine-mesh screening melted onto the interior of the test arena to prevent mite escape. The Plexiglas sandwich was held together with four binder clips positioned such that the passive airflow was not obstructed. The lid of the Munger cell had a 0.5 cm hole that could be plugged and unplugged with a small cork. The mites were transferred into the Munger cell via this hole and the cork remained in place at all times except when probing a mite to evaluate mortality. The control leaves were sprayed with water only. The Munger cells were placed in the same environmental chamber as the open-faced Tupperware containers described above at 24°C, 50% RH and long day light conditions .

The bio-assay was conducted 1, 3, 10, 14, 21 and 28 d post pesticide application and mortality readings were taken daily for 5 d after each bio-assay was set up. Two of the pesticides registered for avocado thrips management,square planter pots abamectin and spinetoram, exhibit translaminar activity. To determine if photodecomposition might affect the impact of these chemicals on Euseius hibisci, twice as many control, abamectin and spinetoram leaves were included in the May 4, 2010 field trial and were bio-assayed on each date. Half of these field-weathered leaves were randomly selected for exposure to intense ultraviolet light for 120 min after they were picked and before the bio-assays were conducted. These leaves were placed perpendicularly into 1-dram vials filled with DI water to prevent desiccation and placed in a hood with a 15-watt UV bulb with a 250-320 nm range . After 120 min UV exposure, the leaves were removed and leaf discs were set up in trays as described above with a minimum of 20 mature female E. hibisci mites. Data for each treatment were analyzed by day by calculating: % of mites found trapped in the felt and % dead mites, excluding those trapped in the felt . To determine how/if E. hibisci females would respond if provided both spinetoram treated and free spaces on a leaf, an additional study was run in which half of each bio-assay leaf was treated with spinetoram. The leaves were selected, flagged and half of each leaf was randomly selected for the field trial on August 23, 2010; the left or right side of the leaf beyond the midrib was randomly selected for treatment on both the axial and abaxial surface, alternating which side was treated.

Depending on which side of the leaf was treated, a paper towel cut into approximately 12 x 16 cm rectangles was covered with clear plastic wrap and was paper clipped to the opposite leaf side protecting it from the pesticide spray. Prior to treatment, the leaves were held such that the leaf mid-rib was parallel to the ground with the covered portion above the bare side such that when the pesticide was sprayed on the leaf, run-off fell to the ground and did not contact or accumulate on the covered side of the leaf. The paper towel and plastic wrap covering remained on the leaves for one hour after treatment and were then carefully removed so as to not tear the leaf or drag any remaining wetness across the untreated side of the leaf. The control leaves were covered in exactly the same fashion, but were treated with water. The half-leaf trial bio-assay was conducted on 1, 3, 7, 10 and 14 d post pesticide application. As described previously, the leaves used in the bio-assays were selected, transported to the laboratory, and the whole leaf was placed abaxial side up on wet white felt covered sponge. The whole leaf was rimmed with strips of wet felt and at least 20 mature female E. hibisci were hand transferred to each leaf. To account for possible positioning bias, the leaves were treated on the left or right side and the mites were deposited either on the treated or untreated sides so that all combinations were accounted for, with three replicate leaves for each combination. We wanted to determine if the mites could detect the spinetoram on the leaf surface; therefore on each of the bio-assay days, the bio-assay was checked every 20 min for the first hour then 3, 5, and 10 h post setup, and once every 24 hours for five days. Mites were scored as being dead, alive or stuck in the felt and which side of the leaf they were found was recorded on at each of the observation time intervals. Data were analyzed by ANOVA with repeated measures using SAS 9.2 with the following factors: treatment , bio-assay day , treatment side , initial mite placement location and observation time . Bio-assays were conducted, 1, 3, 7, 14, 21, 28, 49, 70 90, 111 and 132 d post pesticide application. Mite mortality and mite repellency were recorded separately because it was unknown if the mites that were found dead in the wet felt were dead due to pesticide exposure or from drowning alone. Data were recorded for all bio-assay days for 5 days post bio-assay setup and data from the day 4 count were selected because there was little further activity post 5 days and day 4 counts best represented the data overall. High levels of mite repellency were seen only with the spinetoram treatment and >20% of the mites in the spinetoram treatment were found drowned in the wet felt surrounding the leaf discs through 14 days post-treatment. Mite mortality was the highest in the fenpropathrin treatment and this pesticide showed the most persistent impact . Mite mortality with the spinetoram treatment appeared to increase and then decrease but this was an artifact of mortality being calculated by excluding mites that drowned. Only moderate mite mortality was observed with the abamectin and sabadilla treatments and this dropped to below 10% on 14 and 21 days post-treatment, respectively . Data from the day 4 count are described . Relatively few mites were repelled on abamectin treated leaves and UV exposure showed little impact . After treated leaf exposure to intense UV light for 120 min, high levels of mite repellency with the spinetoram treatment dropped in persistence from 10 d to 3 d. Mite mortality after treated leaf exposure to UV light was reduced to control levels on abamectin treated leaves by the time of the day 1 bio-assay . Because mites trapped in the felt were excluded, percent mite mortality on spinetoram versus spinetoram-UV treated leaves must be interpreted carefully in Fig. 4-7. Based on data from Fig. 4-3 , Fig. 4-8 depicts a more accurate assessment of the results when mites were placed on UV treated spinetoram leaves because it combines the mortality and repellency . There was no effect of the leaf side treated or leaf side the mites were placed on for both treatments and therefore, data were pooled and the new response variable for each treatment became ‘mites deposited on treated side’ versus ‘mites deposited on untreated side’. There was no significant variation observed between the three leaf replicates on any bio-assay date and thus,growing blackberries in containers replicate data were pooled. Based upon repeated measures analysis, each response variable showed significant impacts by day , time and location . With spinetoram treatment, there were more mites alive on the untreated side versus the treated side of the leaf on days 1, 3, 7 and 10 . However, on day 14, there was no difference between the numbers of mites on the treated versus the untreated side of the leaf for any of the observation times. The mites in the control treatment were distributed similarly across both sides of the leaf on all days and times . With spinetoram treatment, mite mortality was different from the control on both the treated and untreated sides of the leaves. On the spinetoram-treated side of the leaf, mite mortality by day was significantly different from the control treatment for days one through 10 but on day 14 mite mortality was no longer different . The same pattern was observed for the untreated side of the spinetoram leaves, i.e. mites were dying at higher levels versus control leaves on both the untreated and treated sides of the leaf. In the control treatment, mite mortality was not different between the water treated and untreated sides of the leaf . Mite repellency on the spinetoram treated and untreated sides of the leaf were different than seen on control leaves . On the spinetoram treated side, mite repellency by day was different from the control for days one and three as well as day 7 and day 10 but not for day 14 . Mite repellency on the spinetoram treated side of the leaf by day for each level of time was not different from the control for times 20 min through 10 hours , but was different for each observation interval from 24 hrs through 5 days . On the spinetoram untreated side of the leaf, mite repellency by day was different from the control for days 1, 3, and 7 as well as day 10 but not for day 14 . Mite repellency on the spinetoram treated side of the leaf by day for each level of time was not different from the control for times 20 min through 10 hours , but was different for observation intervals of 24 hrs through 5 days . In our assessment of the four pesticides currently recommended for avocado thrips management , we found that all four products had some negative effects on E. hibisci. Mite exposure to abamectin resulted in relatively high mortality within the first two weeks of the bio-assay and dropped sharply; presumably as translaminar movement of the material took place and the ultraviolet light rays broke down surface residues . Spinetoram was the only material to which the mites exhibited strong repellency and when mites were bio-assayed in the Munger cells with spinetoram, mortality was high and consistent with the pattern observed with the repellency over the first two weeks of the bio-assays. Mites exposed to sabadilla, a chemical commonly thought to have little non-target effect , showed higher mortality and longer persistence than expected. However, this could be due to the mites feeding on the pesticide-laced sugar on the leaf surface, as sabadilla is formulated with sugar . Data clearly showed that exposing the treated, field-weathered leaves to UV light increased the survival of the mites on both abamectin and spinetoram-treated leaves. Mite mortality to the UV-treated abamectin leaves was no different than with control leaves on day one of the bio-assay , indicating that surface residual activity had been eliminated. With spinetoram treatment, mite repellency and mortality were reduced from 14 d to 10 d. The chief differences in spinetoram from its analog spinosad are: 1) the addition of the 3’-O-ethyl group, which improves potency by altering nicotinic function in the insect nervous system and 2) hydrogenation of the 5,6 double bond, which improves photo stability of the molecule and thereby increases residual control . Our data show that these modifications increased the longevity of the material on the leaf surface but with intense UV exposure, that activity was broken down to some degree. The bio-assays evaluating mite detection of spinetoram on the leaf surface clearly showed more mites alive on the untreated side of the leaf than the treated side, indicating that the mites were able to detect the material and move away from it.