Use of Eretmocerus eremicus and Amblyseius swirskii to suppress Bemisia tabaci on potted poinsettias in protected culture: Data set

The purpose of this study was to determine whether using two natural enemies (a predator, Amblyseius swirskii, and parasitic wasp, Eretmocerus eremicus) in combination would be better at suppressing the main pest of poinsettias, the sweetpotato whitefly (Bemisia tabaci), compared to each natural enemy alone. To test whether the combination of the two natural enemies is superior at suppressing the whiteflies compared to each natural enemy alone, we used both a substitutive and additive design (see Sih et al. 1998. Emergent impacts of multiple predators on prey) to test for both intraspecific and interspecific interactions between our natural enemies on whitefly suppression.

Experimental design:
Experimental units consisted of 12 whitefly-infested potted poinsettias (subsamples) within a large cage (4 ft x 5 ft x 3 ft; w x l x h). Each trial consisted a total of 24 cages split between two adjacent greenhouses (block_greenhouse 8 and 9). Cages were assigned to one of six treatments (4 treatment replications):
1. Control. No natural enemies were released. Only whiteflies at the start of the trial.
2. Eremicus. Twenty-five E. eremicus pupae released into the cage weekly.
3. Swirskii. Two and a half milliliters of A. swirskii bulk (in carrier material, ~123 mites) spread across the 12 poinsettias at weeks 0 and 4.
4. Eremicus + Swirskii. Combination of treatments 2 and 3 above.
5. Eremicusx2. Fifty E. eremicus pupae released into the cage weekly.
6. Swirskiix2. 5 milliliters of A. swirskii bulk (in carrier material, ~246 mites) pread across the 12 poinsettias at weeks 0 and 4.

At the start of the trial (week 0), total number of whitefly nymphs, pupae, exuviae, and adults were counted for every other plant within each cage prior to releasing natural enemies. Number of whitefly nymphs, pupae, exuviae, adults, A. swirskii mites, and E. eremicus adults were subsequently counted on the same plants for individual leaves (1 to 10 = lower to upper canopy) for weeks 2, 4, 6 and 8. Due to difference in subsample level (whole plant vs leaf), data is best analyzed if summed to the plant level first, prior to analysis. In our analysis (ob1_analysesv3), we averaged all plants to the cage level prior to analysis to avoid zero-inflation. Entire trial was replicated a total of three times (spring "A", fall "B", and summer). Spring and fall trials are in the same file ("wfcounts"), since initial and environmental conditions were very similar. Summer trial data is found in the "summertrial.csv" file and was ended prematurely (week 6) due to very high whitefly numbers and rapid plant health decline in all treatments. Endeavor (Pymetrozine) insecticide was applied to all treatments on week four in the summer trial, with no apparent suppression to whitefly populations.

Hobo data loggers were placed inside cages in the middle of the greenhouse for each trial and recorded temperature (F) and relative humidity in 30-minute intervals.

本研究旨在探究联合使用两种天敌（捕食性螨类安德里亚绥螨*Amblyseius swirskii*与寄生蜂类烟粉虱寄生蜂*Eretmocerus eremicus*），相较于单独使用每种天敌，是否能更有效地抑制一品红的主要害虫——烟粉虱（*Bemisia tabaci*）。为验证两种天敌联合使用对烟粉虱的控害效果优于单一天敌，本研究采用替代设计与添加设计（参见Sih等人1998年发表的《多种捕食者对猎物的涌现性影响》），以此探究两种天敌在烟粉虱控害过程中的种内与种间相互作用。

实验设计：
实验单元为大型笼养环境（4英尺×5英尺×3英尺，宽×长×高）内的12株感染烟粉虱的盆栽一品红（子样本）。每个试验共设置24个笼养单元，分配至相邻的两个温室（区块温室8与9）中。笼养单元被随机分配至6种处理（每组4次重复）：
1. 对照组：不释放任何天敌，仅保留试验初始阶段的烟粉虱种群。
2. 寄生蜂组：每周向笼内释放25头*E. eremicus*蛹。
3. 捕食螨组：在第0周与第4周，将2.5毫升携带载体的*A. swirskii*种群基质（约123头螨）均匀撒布于12株一品红上。
4. 联合处理组：同时采用上述处理2与处理3的方案。
5. 加倍寄生蜂组：每周向笼内释放50头*E. eremicus*蛹。
6. 加倍捕食螨组：在第0周与第4周，将5毫升携带载体的*A. swirskii*种群基质（约246头螨）均匀撒布于12株一品红上。

试验初始阶段（第0周），在释放天敌前，对每个笼内每隔一株的植株，统计烟粉虱若虫、蛹、蜕壳与成虫的总数量。随后在第2、4、6、8周，针对相同植株的不同叶片（1至10号对应冠层从下到上），分别统计烟粉虱若虫、蛹、蜕壳、成虫数量，以及*A. swirskii*螨类与*E. eremicus*成虫的种群数量。由于子样本的观测尺度存在差异（整株与单叶），建议先将数据汇总至植株水平后再进行分析。在本研究的分析（ob1_analysesv3）中，为避免零膨胀问题，我们先将所有植株数据平均至笼养单元水平后再开展分析。本试验共重复3次：春季试验“A”、秋季试验“B”与夏季试验。春季与秋季试验数据存储于同一文件"wfcounts"中，因二者初始条件与环境因子高度相似。夏季试验数据存储于"summer trial.csv"文件，该试验因所有处理组的烟粉虱种群数量过高、植株健康快速衰退而提前于第6周终止。夏季试验在第4周向所有处理组喷施了Endeavor（吡蚜酮，Pymetrozine）杀虫剂，但未对烟粉虱种群产生明显的控害效果。

每个试验的温室笼内中部均放置了Hobo数据记录仪，以30分钟为间隔记录温度（华氏度）与相对湿度。