BIOLOGICAL AND MICROBIAL CONTROL Plant-Related Factors Influence the Effectiveness of Neoseiulus fallacis (Acari: Phytoseiidae), a Biological Control Agent of Spider Mites on Landscape Ornamental Plants 1 2 3 P. D. PRATT, R. ROSETTA, AND B. A. CROFT USDA/ARS and University of Florida, 3205 College Avenue, Fort Lauderdale, FL 33314 J. Econ. Entomol. 95(6): 1135Ð1141 (2002) ABSTRACT The predatory mite Neoseiulus fallacis (Garman) was evaluated as a biological control agent of herbivorous mites on outdoor-grown ornamental landscape plants. To elucidate factors that may affect predator efÞciency, replicated tests were conducted on 30 ornamental plant cultivars that varied in relationship to their generalized morphology (e.g., conifers, shade trees, evergreen shrubs, deciduous shrubs, and herbaceous perennials), production method (potted or Þeld grown), canopy density, and the prey species present on each. Plant morphological grouping and foliar density appeared to be the most inßuential factors in predicting successful biological control. Among plant morphological groups, N. fallacis was most effective on shrubs and herbaceous perennials and less effective on conifers and shade trees. N. fallacis was equally effective at controlling spider mites on containerized (potted) and Þeld grown plants, and there was no difference in control of mites on plants with Tetranychus spp. versus those with Oligonychus or Schizotetranychus spp. Moderate to unsuc- cessful control of spider mites by N. fallacis occurred mostly on tall, vertical plants with sparse canopies. Acceptable spider mite control occurred in four large-scale releases of N. fallacis into production plantings of Abies procera, Thuja occidentalis ÔEmeraldÕ, Malus rootstock, and Viburnum plicatum ÔNewportÕ. These data suggest that N. fallacis can be an effective biological control agent of multiple spider mite species in a range of low-growing and selected higher growing ornamental plants. KEY WORDS Phytoseiidae, Tetranychidae, predatory mites, nursery PREDATORY MITES IN the family Phytoseiidae are impor- et al. 1998, Skirvin and De Courcy Williams 1999). tant biological control agents of plant-inhabiting pest Plant morphology, including trichomes, acarodomatia, mites, particularly those in the Tetranychidae (Helle and nectaries, may directly affect leaf residency time and Sabelis 1985, McMurtry and Croft 1997). Inocu- and ultimately biological control success (Walter and lative releases of these predatory mites into agroeco- OÕDowd 1992, Pemberton 1993, English-Loeb et al. systems have resulted in local suppression of spider 1999). In addition, phytoseiids are vulnerable to low mite populations and a decreased reliance on miti- humidity, which may be mitigated by foliar density cides (Croft 1990). Releases of the biological control within plant canopies (Croft et al. 1993). Responses to agent Neoseiulus fallacis (Garman), for instance, can plant incompatibility may include emigration (dis- result in suppression of pest mites on hops, pepper- persal) from the plant, thereby releasing spider mite mint, and strawberry in the PaciÞc Northwest of North populations from predation and resulting in excessive America (Strong and Croft 1995, Morris et al. 1996, plant damage (Grevstad and Klepetka 1992, Kareiva Croft and Coop 1998). Recently, N. fallacis was se- and Sahakian 1990). McMurtry and Croft (1997) clas- lected for evaluation as a control agent of multiple siÞed N. fallacis as a type II selective predator of spider mite species in the diverse cropping system of tetranychids and suggested that this group may be outdoor-grown ornamental landscape plants in this more affected by plant features than more specialized region (Pratt 1999, Pratt and Croft 2000a). predators of spider mites (i.e., Phytoseiulus spp.). Unfortunately, little is known concerning the com- Knowledge about the compatibility of N. fallacis with patibility of N. fallacis with plants grown in ornamental a range of plant species from many generalized mor- systems. Studies have demonstrated that phytoseiids phological groups (e.g., conifer, evergreen shrub, de- and their prey are sensitive to plant architecture (Pratt ciduous shrub, shade tree, herbaceous perennial) pro- duced in ornamental nursery systems is integral to the 1 E-mail: [email protected]. development of biological control tactics for these 2 Oregon State University, North Willamette Research and Exten- sion Center, 15210 NE Miley Road, Aurora, OR 97002Ð9543. commodities (DeBach et al. 1976, Stiling 1993). 3 Oregon State University, Department of Entomology, 2046 Cord- The objective of this study was to assess the ability ley Hall, Corvallis, OR 97331. of N. fallacis to suppress spider mites on a range of 1136 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 95, no. 6 plants produced in outdoor ornamental nurseries of and (3) statistically signiÞcant reductions of pest mite the PaciÞc Northwest. To do this, we Þrst identiÞed population densities when compared with controls: plants susceptible to spider mites from among the 1 ϭ unacceptable levels of spider mite control result- morphological groups grown in the region. Initially in ing in pronounced esthetic damage and unmarketable small-scale tests, we inoculated N. fallacis into selected plants (Ͼ3 spider mites/leaf), 2 ϭ acceptable spider spider mite-infested plant canopies and quantiÞed the mite control with only slight esthetic damage resulting level of pest suppression. We then selected a subset of in marketable plants (1Ð3 spider mites/leaf), 3 ϭ ac- these plants and performed large-scale releases of N. ceptable spider mite control with no esthetic damage fallacis into production level plantings. (Ͻ1 spider mite/leaf) and, 4 ϭ 3, except reduction of spider mites to levels that were undetectable by our sampling method. In addition, we classiÞed the foliar Materials and Methods density of each plant canopy into one of three cate- Predator Source and Release. Stocks of N. fallacis for gories based on visual inspection: 1 ϭ dense, 2 ϭ laboratory cultures were initially collected from agri- moderate or 3 ϭ sparse (Table 1). cultural crops in the Willamette Valley, OR (Hadam et Large-Scale Tests. N. fallacis was effective at sup- al. 1986). These cultures have been maintained for Ͼ6 pressing spider mites on many plants in the small-scale yr with periodic additions from Þeld-collected mites. experiments. However, extrapolation of data gathered In all experiments, predators came from a rearing from small-scale tests to production-level plantings facility at Oregon State University: N. fallacis were remained uncertain because of greater variability ex- produced on lima beans (Phaseolus lunatus L.) in- perienced in production-level plantings with regard to fested with Tetranychus urticae Koch under green- predator release distributions, prey densities between house conditions of 26:21 (Ϯ5)ЊC day:night, 75% plants, plant cultural conditions, and other factors (Ϯ10) RH, and a photoperiod of 16:8 L:D (Strong and relating to production and environmental conditions. Croft 1995). Before release of predators, spider mites Therefore, in 1998, we undertook large-scale inocu- had nearly been eliminated on leaves. Predators were lative releases of N. fallacis into production level plant- released by placing bean leaves containing known ings of Abies procera, Thuja occidentalis ÔEmeraldÕ, quantities of N. fallacis individuals directly into the Malus MM.106 EMLA rootstock, and Viburnum plica- canopy of selected plants that were infested with spi- tum ÔNewportÕ. Estimation of mite densities and pred- der mites. ator identiÞcation were done as described earlier. Small-Scale Tests. We selected 30 ornamental plants Abies procera. The study site consisted of a 0.5-ha from among the generalized morphological groups of planting of Ϸ2-m tall (8 yr old) A. procera trees located conifers, shade trees, evergreen shrubs, deciduous near Independence, OR (44.86 N and 123.2 W). Trees shrubs, and herbaceous perennials to quantify the were planted on a 4% NW slope and spaced Ϸ2m ability of N. fallacis to suppress spider mites and re- apart. Populations of the spruce spider mite, Oligony- duce damage symptoms within their respective plant chus ununguis (Jacobi), were monitored weekly by canopies. Experiments were conducted in northern sampling the branches from 40 randomly selected Willamette Valley ornamental production facilities trees within the study site. Three randomly selected under typical cultural practices (Pratt and Croft 10-cm terminal branches were removed from each 2000a). In all tests, Ն10 individual spider mite-infested tree at 0.5, 1, and 1.5 m above the soil surface. Samples plants were randomly assigned one of two treatments: were placed in an ice cooler, transported to the lab- release of N. fallacis, as described, or no release (con- oratory, and reviewed under a 40 ϫ stereomicroscope. trol). A minimum of Þve replicate plants per treatment Data from stratiÞed samples were averaged within a were used for each test. tree and ultimately among trees. On 8 May, spider mite To estimate mite densities over time, Þve randomly densities reached 2.75 (Ϯ1.02) mites per branch, and selected leaves or three branches (Table 1) were re- 8,000 adult female N. fallacis were evenly distributed moved, without replacement, from each replicate for on trees within the study site. several weeks before and a minimum of 8 wk after the Thuja occidentalis [Emerald]. The study site was introduction of predators. Individual leaves or near Woodburn, OR (45.14ЊN and 122.5ЊW) and con- branches were placed into sealable plastic bags and sisted of 500 Þeld-planted T. occidentalis ÔEmeraldÕ held in a cooler chest during transport to the labora- trees. Plants were 1.25 (Ϯ0.11) m tall, spaced Ϸ1m, tory; mite densities were counted under a 40 ϫ mi- and irrigated with overhead sprinklers as needed. croscope within 24 h. All predators found on sampled Spruce spider mites were monitored on 25 randomly branches were mounted on glass slides in HoyerÕs selected trees by removing two terminal branches of medium, heated for 1 wk on a slide-warmer at 40ЊC, 4 cm in length located at 0.5 and 1 m above the soil and identiÞed according to morphological character- surface.