The relationship between predator density, community composition, and field predation of Colorado potato beetle eggs

Transcription

1 Biological Control 31 (2004) The relationship between predator density, community composition, and field predation of Colorado potato beetle eggs Gary C. Chang *, William E. Snyder Department of Entomology, Washington State University, Pullman, WA , USA Received 26 May 2004; accepted 22 July 2004 Available online 25 August 2004 Abstract Conservation biological control programs seek to increase natural enemy densities through the adoption of more benign farming practices, under the assumption that higher predator densities will lead to more effective pest suppression. However, predator predator interference may lead to diminishing returns in improved pest control as predator densities increase. We examined the relationship between predator density and predation rates on Colorado potato beetle eggs in production potato fields. These potato fields naturally spanned a 10-fold range in predator density, due to differences in management practices. Periodically through the growing season we simultaneously measured predator densities and subjected sentinel eggs masses to predation, allowing us to correlate predator density and egg predation for each field on each sample date. Egg predation rates were significantly positively correlated with total predator densities, a correlation that was not improved when predator densities were scaled to reflect differences in feeding rates on potato beetle eggs of the constituent predator taxa. There was no correlation between per-capita egg predation rates and predator density, and so no evidence that predator interference increased with increasing predator density. We divided predators into six dominant taxa dwarf spiders, crab spiders, minute pirate bugs, big-eyed bugs, damsel bugs, and Lygus bugs (together constituting 93% of all predators collected), and a seventh group, other predators that included all other, less common, taxa and examined correlations between all predator combinations and egg predation rates. The highest correlation was between combined densities of the six most common predator taxa, excluding only the other predators grouping. This suggests that predators may be largely equivalent in their impact on Colorado potato beetle eggs, and that field scouts might be able to ignore uncommon predator taxa when sampling for natural enemies. Ó 2004 Elsevier Inc. All rights reserved. Keywords: Predator guild; Organic agriculture; Generalist predators; Potato 1. Introduction Generalist predators of insects are often expected to be important for controlling pests in agricultural systems (Altieri, 1995). The intuitive prediction is that increasing the density of generalist predators in a crop should subsequently increase predation on pests (Chang and Kareiva, 1999; Walsh and Riley, 1868). On the other hand, increases in the abundance of certain predators can actually decrease predation on a target pest due to negative * Corresponding author. Fax: address: gchang@wsu.edu (G.C. Chang). predator predator interactions such as intraguild predation (Rosenheim et al., 1993). Despite the possibility of intraguild predation, generalist predators contribute to the biological control of pests in crops such as rice (Settle et al., 1996), soybeans (Carter and Rypstra, 1995), grains (Lang, 2003), alfalfa (Snyder and Ives, 2003), and vegetable gardens (Snyder and Wise, 2001). The Colorado potato beetle ( CPB, Leptinotarsa decemlineata Say) damages potato (Solanum tuberosum L.) crops by defoliating plants (Walsh, 1866), substantially reducing potato yields if beetle populations are not kept in check (Hare, 1980). Historically, biological control has played a limited role in CPB management /$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi: /j.biocontrol

2 454 G.C. Chang, W.E. Snyder / Biological Control 31 (2004) (Casagrande, 1987; Hare, 1990). Scientists have identified many species of natural enemies of the CPB (Casagrande, 1987; Hough-Goldstein et al., 1993; Shimer, 1869), but by themselves indigenous natural enemies are generally not capable of suppressing potato beetles below damaging densities (Hare, 1990; Hilbeck et al., 1997; Tamaki, 1981). However, the ability of the CPB to evolve resistance to numerous insecticides (Miyo et al., 1999; Noronha et al., 2001) is one reason why research on alternative control methods is a continuing priority. Combining biological control with other compatible control methods may be a useful tactic (Hare, 1990; Hilbeck et al., 1998). In Washington State potato crops, CPB have historically been controlled using applications of broad-spectrum insecticides (Ruffle and Miller, 2002). However, in response to changing federal regulations (Epstein et al., 2000) and a growing appreciation of the value of natural enemy conservation (Koss, 2003), local potato growers have begun to adopt the use of selective insecticides (Koss, 2003). For example, some potato farmers have been applying a variety of relatively new pest control compounds such as pymetrozine and spinosad (Koss, 2003) that are less toxic to beneficial organisms than are many broad-spectrum insecticides (Sechser et al., 2002; Williams et al., 2003). In addition, organic potato production is growing in the state (Koss et al., 2004). Both the adoption of selective pesticides within conventional agriculture and organic pest management lead to increased densities of many CPB predators (Koss and Snyder, in review). Farming practices that conserve and enhance the assemblage of predatory insects in potatoes may increase the predation rate on potato pests (Hough-Goldstein et al., 1993). In the work we report on here, we sought to answer three questions. First, which predators have the highest per capita feeding rates on CPB eggs? This first question was addressed by measuring predation rates on CPB eggs in the laboratory. Second, what is the relationship between total predator densities and rates of CPB egg predation in the field? Here, we took advantage of the natural differences in predator abundance among production potato fields that differed in their pesticide inputs, and measured the relationship between predator density and CPB egg predation. Third, can we identify predator taxa that have a particularly strong correlation with predation on CPB in the field? Again using our field data, we examined whether such predators could be identified. This information might allow field scouts to concentrate on sampling just the most important taxa when factoring biological control into pesticide application decisions. 2. Materials and methods The community of natural enemies in Washington potato fields is numerically dominated by a diverse group of generalist predators (Koss and Snyder, in review; Tamaki and Weeks, 1972). In potato foliage, the most abundant predators are predatory bugs (Hemiptera) in the genera Nabis, Geocoris, and Orius, spiders (Araneae) in the families Linyphiidae and Thomisidae, and coccinellid beetles (Coleoptera: Coccinellidae) (Koss, 2003; Tamaki and Weeks, 1972). While not typically considered predators, Lygus spp. (Hemiptera: Lygaeidae) bugs are abundant in local potato fields and will opportunistically feed on soft-bodied potato herbivores (Koss, 2003). Other predators more rarely collected in the foliage include members of several beetle families (Coleoptera: Anthicidae, Carabidae), lacewings (Neuroptera: Chrysopidae, Hemerobiidae), mantids (Mantodea: Mantidae), predatory flies (Diptera: Syrphidae, Dolichopodidae), and less-common predaceous bugs (Hemiptera: Reduviidae) (Koss, 2003). The CPB eggs used in our laboratory and field predation assays were the progeny of female beetles collected from production potato fields in the vicinity of Othello, WA, USA, during the summer of Adult beetles were maintained in a greenhouse at 22 C and 16:8 L:D, and fed live potato foliage. Eggs were harvested daily and placed in a refrigerator (4 C) to retard egg development. Clutch size was standardized (at 20 eggs for laboratory feeding trials, or at eggs for field predation measures, see below) by carefully cutting away excess eggs using dissecting scissors Predator feeding rates in the laboratory To compare the appetites of different predators, we measured the number of CPB eggs consumed by predators in simple laboratory microcosms consisting of 9-cm diameter Petri dishes provisioned with a moistened cotton wick. In each bioassay, we placed an individual predator in a Petri dish with 20 CPB eggs attached to a small (ca. 1 cm 2 ) piece of potato leaf. We checked each dish for the number of intact CPB eggs every 24 h until the final check at 72 h. Our trials also included Petri dishes containing 20 CPB eggs but no predator, which served as controls. We conducted the feeding bioassays in an environmental control chamber ( Intellus, Percival Scientific, Perry, Iowa) at 22 C and 16:8 L:D. We tested field-collected individuals of the most abundant predator taxa as revealed in our field sampling; predators were held at 22 C and 16:8 L:D in the laboratory for 16 h (overnight) before being used in the laboratory assays, provided with water but not food Field measurements of predator densities and egg predation rates We conducted our fieldwork in seven potato fields on commercial farms managed by cooperating growers. Fields were between 24 and 48 ha, irrigated, located in

3 G.C. Chang, W.E. Snyder / Biological Control 31 (2004) the Columbia Basin in the vicinity of the city of Othello, WA, and were either managed conventionally or certified organic (Table 1). Fields 1 6 were managed by one grower, while field 7 was managed by a second grower. Vacuum samples were collected using a D-Vac suction sampler (Model B&S, D-Vac, Ventura, CA). On each sampling date, we took D-vac samples from groups of 10 plants in five different locations within each field (total N = 50 plants sampled per field per sample date). The D-vac collecting tube was held over each plant, and vigorously shaken, for 15 s. We sampled all seven fields between June (grouped together into sampling period 1) and July 9 14 (grouped into sampling period 2). On August 1 (sampling period 3), our sampling was limited to only five fields, because we were unable to access fields 3 and 5 on that date. D-vac samples were placed in a cooler (5 C) and returned to the laboratory, where they were held in a freezer ( 20 C) until sorting. All predatory insects and spiders were sorted and counted from these samples. We used a sentinel egg method (e.g., Hilbeck and Kennedy, 1996) to measure the predation rate in each potato field. On each sample date, we attached 10 groups of CPB eggs each, with egg clutches on a ca. 1-cm 2 piece of potato foliage where they had been deposited by the female beetle, to potato leaves using a small drop of ElmerÕs glue (Borden, Columbus Ohio, USA). The glue was entirely covered by the potato foliage and does not alter predator behavior (Koss et al., 2004). Concurrently we also stapled five groups of CPB eggs that were enclosed within a cloth sleeve (52 52 mesh lumite screening; Bioquip, Gardena, CA, USA) so that predators could not reach them; these eggs served as controls for mortality not due to predation. We left the eggs for 2 days, then recovered them and recorded the number of eggs that were still viable by returning eggs to the laboratory and allowing them to complete development to hatching in a growth chamber (22 C and 16:8 L:D). We did this because feeding by some predators, for example those with piercing-sucking mouthparts, does not always leave obvious signs that predation has occurred (Koss, 2003). For each field we then calculated the predation rate by subtracting the average proportion of viable sentinel eggs from the proportion of viable control eggs, in that field on that sample date Observations of predation events in the field We opportunistically recorded predation events that we saw while setting and recollecting the CPB eggs used to measure predation rates. We also inspected naturally occurring CPB egg masses on four dates in August to determine what, if any, predators were consuming them. On August 7, 63 CPB egg masses were monitored. Each was inspected for associated predators three times (once every 2 h) from 18:00 to 01:00 h. On August 12 between 13:00 and 15:00 h, 29 egg masses were located and examined once for predators. On August 14, 17 egg masses were located and checked for predators once beginning at 11:00 h, and a second time beginning at 13:00 h. On August 15, 14 egg masses were inspected for predators three times (every 2 h) beginning at 08:00 h Statistics All analyses were conducted using SYSTAT statistical software (SPSS, Chicago, Illinois, USA). For the laboratory predation rate measurements, we compared CPB egg consumption rates among the different predator species using ANOVA followed by TukeyÕs HSD post hoc test. The field estimates of predation rate were correlated to the total abundance of predators in the corresponding vacuum sample, in that field and on that date. To attempt to identify predator taxa that correlated particularly strongly with CPB predation, we performed correlations between field estimates of predation rate and subsets of the predator assemblage. With these correlations, we sought only to determine whether predators could be found that would indicate high predation on CPB and thus be useful to field scouts, and not necessarily which predators were causing the most CPB mortality. All possible combinations of the seven Table 1 Summary of the characteristics of the potato fields sampled in 2003 Field, location Management Preparation Insecticides used before and during our sampling period 1, central Adams Co. Conventional Fumigated Pymetrozine and indoxacarb (June 9), pymetrozine (July 28) 2, southern Adams Co. Conventional Fumigated Carbaryl (July 14) 3, central Adams Co. Conventional Cover cropped Methamidophos and esfenvalerate (June 9) 4, southern Adams Co. Conventional Cover cropped Carbaryl (July 14) 5, southern Adams Co. Organic Cover cropped Garlic extract (July 3, July 16), cottonseed oil, clove oil, and garlic extract (July 24) 6, southern Adams Co. Organic Cover cropped Garlic extract (July 3, July 16), cottonseed oil, clove oil, and garlic extract (July 24) 7, northern Franklin Co. Organic Cover cropped None Field preparation for conventional potatoes may include either fumigation or growing mustard during the preceding winter to manage soil-borne pests (McGuire, 2003; Riga et al., 2004). Fumigation is not an option for organic land, so three types of potato fields are found in central Washington: (1) conventional-fumigated, (2) conventional-mustard, and (3) organic.

4 456 G.C. Chang, W.E. Snyder / Biological Control 31 (2004) most abundant predator groups were correlated with predation rate. Thus, in addition to the original correlation between all predators and predation rate, 126 additional correlations were calculated (7 subsets of 1 taxon, 21 subsets of 2 taxa, 35 subsets of 3 taxa, 35 subsets of 4 taxa, 21 subsets of 5 taxa, and 7 subsets of 6 taxa). However, conducting so many correlations using the same dataset inflates the probability of finding a statistically significant correlation by chance alone; therefore, we applied the Bonferroni approximation to adjust for multiple comparisons (Jones, 1984). We also attempted to account for the different appetites of various predators by calculating a correlation between predation rate and the field density of particular predator taxa weighted by the number of CPB eggs that each consumed in the laboratory bioassay. Finally, we calculated per capita predation rates dividing predation rate by predator density, and then calculated the correlation between predator density and per capita predation. We arcsine transformed all predation rate data before analysis. We considered measurements of predation rate from the same field but during different sampling periods as independent in this analysis because events such as insecticide sprays that occurred between samples caused substantial changes in the abundance of natural enemies (see below). big-eyed bugs, ground beetles (Bembidion spp.), dwarf spiders, and minute pirate bugs (Fig. 1) Correlations between predator densities and egg predation rates in the field The most abundant predators in our field samples were dwarf spiders, minute pirate bugs, big-eyed bugs, damsel bugs, and crab spiders (Fig. 2). Bugs in the genus Lygus were abundant. While Lygus are typically herbivorous, some were observed feeding on CPB eggs in the field (see below). Relatively rare predators such as ground beetles, lady beetles, green lacewings, brown lacewings, rove beetles, and assassin bugs were also 3. Results 3.1. Predator feeding rates in the laboratory Eight taxaõs predation rates on CPB eggs were quantified in the laboratory bioassay. Assassin bugs (Sinea sp.) had the highest predation rate, followed by ladybird beetles (Hippodamia convergens), damsel bugs, Lygus, Fig. 1. Per capita predation on CPB eggs by selected predators in the laboratory. Sample sizes are given in parentheses following each treatment. Bars represent means ± their standard errors; letters indicate significant differences using TukeyÕs post hoc test. Fig. 2. Composition of the arthropod predator complex in different potato fields during the summer of The top panel graphs data from D-vac samples taken in late June, the middle panel shows data from mid-july, and the bottom panel shows data from early August.

5 G.C. Chang, W.E. Snyder / Biological Control 31 (2004) Fig. 3. Predation rate on sentinel CPB eggs related to density of all predators collected in D-vac samples. Each point represents data from one field during one sampling period. Missing labels on the x-axes indicate fields lacking complete data (both predator densities and predation rate, see Materials and methods) for the correlation analysis (see Materials and methods). found in the samples. We refer to the collective group of relatively rare predators as others. The total densities of predators in a field varied nearly 100-fold, from 1.8 per 10 plants sampled on July 12 in field 3, to per 10 plants sampled in field 7 on July 30 (Fig. 2). Although several of the predatorsõ densities were highly correlated with each other (Table 2), the relative abundances of various predator taxa differed among fields and changed during the summer (Fig. 2). For example, the percentage of predators that were dwarf spiders from one field at one date ranged from 23.2% (in field 6, sampling period 1) to 83.8% (in field 2, sampling period 3). The average survival of caged CPB eggs was 93%. The range of predation rates on sentinel CPB eggs (mortality in excess of that observed on caged CPB eggs) was 0 26%, with an average of 7%. The predation rate in a field was significantly positively correlated with the total density of predators in D-vac samples taken from that field (correlation coefficient = 0.633, P = 0.008; Fig. 3). All possible correlations between predation rate and various subsets of the predators were calculated (Fig. 4). Predation rate was more strongly correlated with certain combinations of predators than with total predators. The highest correlation coefficient between predation rate and a set of predators was 0.668, which occurred with three groups: 1. total predators excluding others, 2. total predators excluding others and bigeyed bugs, and 3. total predators excluding big-eyed bugs and minute pirate bugs. Several other predator combinations had correlations with predation rate that were nearly as high as However, only combinations that included at least six of the seven predator taxa were statistically significantly correlated with predation rate when the Bonferroni correction was applied to adjust for multiple comparisons (Fig. 4). Five of the abundant taxa in the field (dwarf spiders, big-eyed bugs, damsel bugs, minute pirate bugs, and Lygus) were among the predators whose feeding rates on CPB eggs were quantified in the laboratory. Weighting the field densities of that group of five predators by their laboratory consumption of CPB eggs did not improve correlations with predation rate on sentinel eggs. The correlation coefficient weighted by laboratory consumption for dwarf spiders, big-eyed bugs, damsel bugs, minute pirate bugs, and Lygus was 0.635, which was less than the unweighted correlation coefficient for that group (0.663). The values for per capita predation rate in the field ranged from 0 to eggs removed per predator. Fifteen of the 16 values were between 0 and The highest value was from field 3 during the second sampling period. Per capita predation was not significantly correlated with predator density (correlation coefficient = 0.223, P > 0.2). Removal of the point with the highest per capita predation rate changed the sign of the correlation coefficient, but the correlation was still not statistically significant (correlation coefficient = 0.282, P > 0.2) Observations of predation events in the field Predators seen attacking CPB eggs in the field over the summer were lady beetles (twice), damsel bugs (twice), Lygus (twice), stinkbugs (once), assassin bugs (once), and minute pirate bugs (once). Stinkbugs were also observed attacking CPB larvae (four times), as were assassin bugs (twice). However, some of the larger predators were also observed eating other smaller predators. Assassin bugs were recorded feeding on a brown lacewing and on a dwarf spider, while a damsel bug was seen feeding on a Lygus. Table 2 Correlation matrix for predator densities in D-vac samples Dwarf spiders Minute pirate bugs Damsel bugs Big-eyed bugs Crab spiders Lygus Others Dwarf spiders 1 Minute pirate bugs Damsel bugs Big-eyed bugs Crab spiders Lygus Others

6 458 G.C. Chang, W.E. Snyder / Biological Control 31 (2004) Fig. 4. Correlations between predation rate and every possible combination of seven predator taxa. Panels display predation rate correlated with (A) individual predator taxa, (B) combinations of two taxa, (C) combinations of three predator taxa, (D) combinations of four predator taxa, (E) combinations of five predator taxa, and (F) combinations of six predator taxa. On each panel, the horizontal line at correlation coefficient = indicates the critical value for P with n = 16(Zar, 1974). In (A) and (F), the horizontal line at correlation coefficient = indicates the critical value after applying the Bonferroni correction for seven comparisons. In (B) and (E), the horizontal line at correlation coefficient = indicates the critical value after applying the Bonferroni correction for 21 comparisons. In (C) and (D), the horizontal line at correlation coefficient = indicates the critical value after applying the Bonferroni correction for 35 comparisons. Abbreviations used: L, Lygus; O, Orius; N, Nabis; D, dwarf spiders; T, others ; C, crab spiders; and G, Geocoris.

7 G.C. Chang, W.E. Snyder / Biological Control 31 (2004) Discussion We found that predation on CPB eggs was higher when fields contained more predators. However, even at the highest observed predator density, 74% of sentinel CPB eggs were intact after 2 days in the field (Fig. 1). Colorado potato beetle eggs can hatch in as few as 4 days (Logan et al., 1985). Thus, even under the observed conditions with the most predators, it seems reasonable to expect that a substantial proportion of eggs would survive to become larvae and begin feeding on plants. Indeed, the organic fields that we monitored were all eventually sprayed with natural insecticides to control CPB (Table 1; field 7 was sprayed with spinosad on August 7, after our final measurements of predation rate). Hilbeck et al. (1997) also concluded that biological control in potatoes grown in North Carolina would not, by itself, adequately suppress CPB. It remains possible that better CPB suppression could be obtained if total predator density in potato fields were to increase, for example via conservation biological control (Barbosa, 1998), because predation on CPB eggs increased through the entire range of observed predator densities and did not appear to reach a plateau at the higher densities. Despite substantial quantitative work, setting economic injury levels for CPB remains subjective (Nault and Kennedy, 1998). Our results may help growers who are interested in monitoring predator densities and incorporating predation rates in their decisions of whether or not to spray insecticides. Several of the predators that fed on CPB eggs have highly correlated population densities, and so our data do not reveal a single most important predator of CPB eggs. Indeed, correlations of single predator species with field measures of predation on CPB eggs are not as high as the correlation between total predator density and predation. Therefore, when field scouts sample predators to estimate the potential biological control of CPB, several predator species should be counted. However, in our dataset, it is possible to exclude certain groups of predators from the total and obtain a correlation with predation rate that is as high (or higher) as that for all predators. For example, because a high correlation can still be obtained when the disparate group of relatively rare predators ( others ) is excluded, field scouts might be able to concentrate on learning how to identify, and in the field on counting, only the most abundant predators rather than the rarest taxa. However, different environmental conditions and sampling methods may alter which taxa are common and which are rare, and potentially produce different relationships between predator densities and predation rate. Predator density is frequently higher on organic farms than on conventional farms (Drinkwater et al., 1995; OÕSullivan and Gormally, 2002), and individual predators can be larger and more fecund on organic farms (Bommarco, 1998). Thus, biological control may be particularly important to organic farmers (McSorley, 2002). Biological control compensates for the loss of insecticides in the case of Californian tomatoes, where pest damage on organic farms is not significantly different from pest damage on conventional farms (Letourneau and Goldstein, 2001). The action of predators also contributes to lower aphid abundance on organic versus conventional spring barley in Sweden (Östman et al., 2001). However, for many crops the extent that biological control can reduce insecticide usage remains unclear (Gurr and Wratten, 2000). Even when biological control cannot eliminate the use of insecticides, commercial agriculture could integrate knowledge of the impact of predators on crop pests into management decisions (Musser et al., 2004). No particular predator stood out from our field data as having a particularly strong positive correlation with predation rate, despite significant differences in laboratory consumption of CPB eggs by different predator species. This suggests that in the field, the various predators are functionally similar (Harris, 1995) in terms of predation on CPB eggs. Interestingly, Tamaki (1981) suggested treating the predators attacking the green peach aphid in potatoes as a predator complex, wherein different predator species are considered equally important. Factors that could effectively render predators functionally similar for CPB control include intraguild predation and alternative prey. For example, assassin bugs are a relatively rare, generalist predator found only when total predator density was relatively high. Assassin bugs had the highest predation rate on CPB eggs in the laboratory, but also consumed other predators in the field. Thus, higher per capita predation by assassin bugs might be counteracted by their interfering with other, more abundant predators, in an intraguild predation scenario reminiscent of that in cotton found by Rosenheim et al. (1993). However, intraguild predation may not be as strong in potatoes as in cotton, because we did not find a relationship between per capita predation rate and total predator density. Thus, the situation in potatoes may be more similar to systems where assemblages of predators have a net additive effect on prey suppression, despite some predator predator interference (Lang, 2003; Snyder and Ives, 2003). Similarly, the presence of multiple herbivores and detritivores may influence predation on CPB eggs by generalist predators. The lady beetle H. convergens was another species with a high per capita predation rate on CPB eggs in the laboratory bioassay. However, lady beetles prefer aphids to CPB eggs when offered a choice (Snyder and Clevenger, 2004). Thus, in the short term, alternative prey could dilute the impact of a generalist predator on the population density of a particular focal prey species (Ayres et al., 2001; Snyder et al., 2004). Östman (2004), working on organic farms in Sweden, found that

8 460 G.C. Chang, W.E. Snyder / Biological Control 31 (2004) the predation rate on aphids was positively correlated to both predator density and alternative prey density, suggesting that the positive contribution of alternative prey density to predator density more than offset their dilution of the predatorsõ functional response. We found a positive correlation between the total density of an assemblage of generalist predators and predation on sentinel CPB eggs. Because the densities of most of the predators were positively correlated with each other, it is difficult to identify a single, particularly important predator species. Together, six numerically dominant taxa were highly correlated with CPB egg predation, so field scouts might be able to ignore less common predator taxa when sampling to estimate predation on CPB. Because per capita predation does not have a significant relationship to predator density, we suggest that what interference occurs between predators has little effect on their net impact on CPB eggs. Furthermore, the relationship between predator density and predation on CPB eggs is positive over the range of observed predator densities and does not appear to reach a plateau. Therefore, conservation biological control methods that can increase predator densities beyond what we observed may further suppress CPB in potato fields. Acknowledgments We thank Stacy Kniveton (Johnson Agriprises), Brad Bailie (Lenwood Farms), and Alec McErlich (Small Planet Foods) for giving us spray records and access to their fields. We also appreciate Nick Lee and Rebecca Darling, who ably assisted with the fieldwork. Garrett Clevenger maintained laboratory colonies of CPB. Marc Evans discussed our statistical options, and Cory Straub and two anonymous reviewers made valuable suggestions that improved the manuscript. References Altieri, M.A., Agroecology: The Science of Sustainable Agriculture. Westview Press, Boulder, CO, USA. Ayres, B.D., Ayres, M.P., Abrahamson, M.D., Teale, S.A., Resource partitioning and overlap in three sympatric species of Ips bark beetles (Coleoptera: Scolytidae). Oecologia 128, Barbosa, P., Conservation Biological Control. Academic Press, San Diego, CA, USA. Bommarco, R., Reproduction and energy reserves of a predatory carabid beetle relative to agroecosystem complexity. Ecol. Appl. 8, Carter, P.E., Rypstra, A.L., Top-down effects in soybean agroecosystems: spider density affects herbivore damage. Oikos 72, Casagrande, R.A., The Colorado potato beetle: 125 years of mismanagement. Bull. Entomol. Soc. Am. 33, Chang, G.C., Kareiva, P., The case for indigenous generalists in biological control. In: Hawkins, B.A., Cornell, H.V. (Eds.), Theoretical Approaches to Biological Control. Cambridge University Press, New York, pp Drinkwater, L.E., Letourneau, D.K., Workneh, F., van Bruggen, A.H.C., Shennan, C., Fundamental differences between conventional and organic tomato agroecosystems in California. Ecol. Appl. 5, Epstein, D.L., Zack, R.S., Brunner, J.F., Gut, L., Brown, J.J., Effects of broad-spectrum insecticides on epigeal arthropod diversity in Pacific Northwest apple orchards. Environ. Entomol. 29, Gurr, G., Wratten, S., Biological Control: Measures of Success. Kluwer Academic Publishers, Norwell, MA, USA. Hare, J.D., Impact of defoliation by Colorado potato beetle on potato yields. J. Econ. Entomol. 73, Hare, J.D., Ecology and management of the Colorado potato beetle. Annu. Rev. Entomol. 35, Harris, P.M., Are autecologically similar species also functionally similar? A test in pond communities. Ecology 76, Hilbeck, A., Kennedy, G.G., Predators feeding on the Colorado potato beetle in insecticide-free plots and insecticide-treated commercial potato fields in eastern North Carolina. Biol. Control 6, Hilbeck, A., Eckel, C., Kennedy, G.G., Predation on Colorado potato beetle eggs by generalist predators in research and commercial potato plantings. Biol. Control 8, Hilbeck, A., Eckel, C., Kennedy, G.G., Impact of Bacillus thuringiensis insecticides on population dynamics and egg predation of the Colorado potato beetle in North Carolina potato plantings. BioControl 43, Hough-Goldstein, J.A., Heimpel, G.E., Bechmann, H.E., Mason, C.E., Arthropod natural enemies of the Colorado potato beetle. Crop Prot. 12, Jones, D., Use, misuse, and role of multiple-comparison procedures in ecological and agricultural entomology. Environ. Entomol. 13, Koss, A.M., Integrating chemical and biological control in Washington State potato fields. M.S. Thesis. Washington State University, Pullman, WA, USA. Koss, A.M., Snyder, W.E., in review. Alternative prey disrupt biocontrol by a guild of generalist predators. Ecol. Appl. Koss, A.M., Chang, G.C., Snyder, W.E., Predation of green peach aphids by generalist predators in the presence of alternative, Colorado potato beetle egg prey. Biol. Control, in press. Lang, A., Intraguild interference and biocontrol effects of generalist predators in a winter wheat field. Oecologia 134, Letourneau, D.K., Goldstein, B., Pest damage and arthropod community structure in organic vs. conventional tomato production in California. J. Appl. Ecol. 38, Logan, P.A., Casagrande, R.A., Faubert, H.H., Drummond, F.A., Temperature-dependent development and feeding of immature Colorado potato beetles, Leptinotarsa decemlineata (Say) (Coleoptera: Chrysomelidae). Environ. Entomol. 14, McGuire, A., Green manuring with mustard. Agrichemical and Environmental News 206: June. Available from: < McSorley, R., Nematode and insect management in transitional agricultural systems. HortTechnology 12, Miyo, T., Keil, C.B.O., Hough-Goldstein, J.A., Oguma, Y., Inheritance of resistance to esfenvalerate in Colorado potato beetles (Coleoptera: Chrysomelidae). J. Econ. Entomol. 92, Musser, F.R., Nyrop, J.P., Shelton, A.M., Survey of predators and sampling method comparison in sweet corn. J. Econ. Entomol. 97, Nault, B.A., Kennedy, G.G., Limitations of using regression and mean separation analyses for describing the response of crop yield

9 G.C. Chang, W.E. Snyder / Biological Control 31 (2004) to defoliation: a case study of the Colorado potato beetle (Coleoptera: Chrysomelidae) on potato. J. Econ. Entomol. 91, Noronha, C., Duke, G.M., Chinn, J.M., Goettel, M.S., Differential susceptibility to insecticides by Leptinotarsa decemlineata (Coleoptera: Chrysomelidae) populations from western Canada. Phytoprotection 82, Östman, Ö., The relative effects of natural enemy abundance and alternative prey abundance on aphid predation rate. Biol. Control 30, Östman, Ö., Ekbom, B., Bengtsson, J., Landscape heterogeneity and farming practice influence biological control. Basic Appl. Ecol. 2, OÕSullivan, C.M., Gormally, M.J., A comparison of ground beetle (Carabidae: Coleoptera) communities in an organic and conventional potato crop. Biol. Agric. Hort. 20, Riga, E., Mojtahedi, H., Ingham, R.E., McGuire, A.M., Green manure amendments and management of root knot nematodoes on potato in the Pacific Northwest of U.S.A. Nematology, in press. Rosenheim, J.A., Wilhoit, L.R., Armer, C.A., Influence of intraguild predation among generalist insect predators on the suppression of an herbivore population. Oecologia 96, Ruffle, R., Miller, J., Digging for Alternatives: An Analysis of Potato Pest Management Research at Two Northwest Land Grant Universities. Northwest Coalition for Alternatives to Pesticides, Eugene, OR, USA. Sechser, B., Reber, B., Bourgeois, F., Pymetrozine: selectivity spectrum to beneficial arthropods and fitness for integrated pest management. Anz. Schadlingskd. 75, Settle, W.H., Ariawan, H., Astuti, E.T., Cahyana, W., Hakim, A.L., Hindayana, D., Lestari, A.S., Pajarningsih, Sartanto, Managing tropical rice pests through conservation of generalist natural enemies and alternative prey. Ecology 77, Shimer, H., Insects injurious to the potato. Am. Nat. 3, Snyder, W.E., Clevenger, G.M., Negative dietary effects of Colorado potato beetle eggs for native and introduced ladybird beetles. Biol. Control, in press. Snyder, W.E., Ives, A.R., Interactions between specialist and generalist natural enemies: parasitoids, predators, and pea aphid biocontrol. Ecology 84, Snyder, W.E., Wise, D.H., Contrasting trophic cascades generated by a community of generalist predators. Ecology 82, Snyder, W.E., Chang, G.C., Prasad, R.P., Biodiversity and successful conservation biological control: is there a relationship? In: Barbosa, P., Castellanos, I., (Eds.), Ecology of Predator Prey Interactions. Oxford University Press, Oxford, UK, in press. Tamaki, G., Biological control of potato pests. In: Lashomb, J.H., Casagrande, R. (Eds.), Advances in Potato Pest Management. Hutchinson Ross, Stroudsburg, PA, pp Tamaki, G., Weeks, R.E., Efficiency of three predators, Geocoris bullatus, Nabis americoferus and Coccinella transversogutata, used alone or in combination against three insect prey species, Myzus persicae, Ceramica picta, and Manastra configurata, in a greenhouse study. Environ. Entomol. 1, Walsh, B.D., The new potato bug. Practical Entomol. 1, Walsh, B.D., Riley, C.V., Potato bugs. Am. Entomol. 1, Williams, T., Valle, J., Vinuela, E., Is the naturally derived insecticide Spinosad (R) compatible with insect natural enemies? Biocontrol Sci. Technol. 13, Zar, J.H., Biostatistical Analysis. Prentice-Hall, Englewood Cliffs, NJ, USA.