Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Unified neutral theory of biodiversity wikipedia , lookup
Habitat conservation wikipedia , lookup
Biodiversity action plan wikipedia , lookup
Introduced species wikipedia , lookup
Island restoration wikipedia , lookup
Theoretical ecology wikipedia , lookup
Latitudinal gradients in species diversity wikipedia , lookup
Generalist Predators (Coleoptera: Carabidae, Staphylinidae) Associated with Millipede Populations in Sweet Potato and Carrot Fields and Implications for Millipede Management Author(s): Adam J. Brunke, Christine A. Bahlai, Mark K. Sears, and Rebecca H. Hallett Source: Environmental Entomology, 38(4):1106-1116. 2009. Published By: Entomological Society of America DOI: 10.1603/022.038.0418 URL: http://www.bioone.org/doi/full/10.1603/022.038.0418 BioOne (www.bioone.org) is an electronic aggregator of bioscience research content, and the online home to over 160 journals and books published by not-for-profit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder. BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. BIOLOGICAL CONTROLÑPARASITOIDS AND PREDATORS Generalist Predators (Coleoptera: Carabidae, Staphylinidae) Associated With Millipede Populations in Sweet Potato and Carrot Fields and Implications for Millipede Management ADAM J. BRUNKE, CHRISTINE A. BAHLAI, MARK K. SEARS, AND REBECCA H. HALLETT1 Department of Environmental Biology, University of Guelph, Guelph, Ontario N1G 2W1, Canada Environ. Entomol. 38(4): 1106Ð1116 (2009) ABSTRACT The predatory beetle assemblage of Ontario carrot and sweet potato Þelds was described and assessed to identify species of interest to the control of the emerging pest millipede Cylindroiulus caeruleocinctus (Wood) (Diplopoda: Julidae). Pterostichus melanarius (Coleoptera: Carabidae) was identiÞed as a dominant species, and seven other carabid species [Pterostichus melanarius (Illiger), Harpalus pensylvanicus (DeGeer), Ophonus puncticeps (Stephens), H. erraticus Say, Bembidion quadrimaculatum oppositum Say, Poecilus chalcites (Say), Scarites subterraneus Fabricius, and Pterostichus permundus (Say)] were identiÞed as common species on the basis of activity density. Common species became more abundant as the growing season progressed. In laboratory bioassays, P. melanarius preyed on millipedes regardless of prey size, whereas H. erraticus never selected millipedes as prey. A signiÞcant positive spatiotemporal relationship was found between P. melanarius and C. caeruleocinctus in sweet potato Þelds. P. melanarius was found to be a natural enemy of C. caeruleocinctus, and other common carabid species warrant future study. The role of Staphylinidae in millipede control could not be elucidated, likely because of low trapping efÞciency. Tachinus corticinus Gravenhorst, an introduced staphylinid from Europe, was newly recorded in Ontario, extending its North American range considerably westward from the province of Quebec. The results of this study are an important foundational step toward developing a successful integrated pest management strategy for controlling millipede damage in crops. KEY WORDS Cylindroiulus caeruleocinctus, Pterostichus melanarius, integrated pest management Julid millipedes (Diplopoda: Julidae) are most typically known for their beneÞcial role as key decomposers of organic matter, contributing to soil health (Coleman et al. 2004). Recently in Ontario, millipedes have been implicated as the causal agents of root damage to corn, Zea mays L., and several root crops including ginseng, Panax ginseng C. A. Mey., sweet potato, Ipomoea batatas L. Lam., and carrot, Daucus carota L. (Allen and Filotas 2008a). Millipedes in the family Odontopygidae are well known in African countries as agricultural pests of sweet potato (Ebregt et al. 2005). Carrot and sweet potato yield loss caused by suspected millipede herbivory was recently addressed in Ontario, Canada (Allen and Filotas 2008a,b; Sears et al., unpublished data). That study identiÞed the common European millipede Cylindroiulus caeruleocinctus (Wood) (family Julidae) as the most abundant species in these agroecosystems and showed that it is capable of causing damage to carrot and sweet potatoes. The study concluded that this species has potential to be a pest of these vegetables. The natural predator complex within the carrot agroecosystem in northeastern North America is rel1 Corresponding author, e-mail: [email protected]. atively well known (Baines et al. 1990, Boivin 1999). The sweet potato agroecosystem in North America is less studied, and research has been limited to studies conducted in the southeastern United States (Chalfant et al. 1990). Sweet potato has been increasing in acreage and importance in Ontario (OMAFRA 2008) and the northeastern United States (USSPC 2008) in recent years, and several pests have been recognized (OMAFRA 2008). Information about the composition of the arthropod community inhabiting the sweet potato agroecosystem in northeastern North America, including natural enemies of pest species, is critical to proper pest management. There are currently no registered pesticides or integrated pest management (IPM) strategies in place for millipede control in Ontario, and thus, there is a need to understand millipede ecology, especially their natural enemies. Cylindroiulus caeruleocinctus has a well-armored body and possesses defensive benzoquinones (Jacobson 1966); consequently, millipedes are widely known as unpalatable to many predators (Baker 1985). North American predators that specialize on millipedes are not known to occur in eastern agricultural Þelds, although they occur elsewhere as the western carabid genus Promecognathus Chaudoir 0046-225X/09/1106Ð1116$04.00/0 䉷 2009 Entomological Society of America August 2009 BRUNKE ET AL.: GENERALIST PREDATORS AND MILLIPEDE MANAGEMENT (Arnett and Thomas 2001) and as the rare forest beetle genus Phengodes Illiger (Arnett et al. 2002). Therefore, it is the generalist guild of predators that is of interest to millipede management and the focus of this study. Generalist predatory beetles from the families Carabidae (ground beetles) and Staphylinidae (rove beetles) occur in many agroecosystems (Sunderland 2002) and include species that are known to accept millipede prey in laboratory settings (Snider 1984, Baker 1985). Baker (1985) worked with Australian carabids not known from Ontario, whereas Snider (1984) studied species typical of a Michigan forest. Both of these studies concerned close relatives of species known from agricultural Þelds. We hypothesized that one or more generalist predatory beetle species have the potential to be important to the management of millipede (C. caeruleocinctus) populations through natural predation in carrot and sweet potato Þelds. The objectives of this study were therefore to (1) examine the biodiversity of predatory beetle species and determine the common members of the generalist predatory beetle guild in carrot and sweet potato Þelds; (2) determine the temporal distributions of these common species; (3) evaluate the ability of select common beetles to consume millipedes as prey; and (4) assess spatial and temporal associations between identiÞed millipede predators and C. caeruleocinctus. Materials and Methods Species Richness and Temporal Distributions of Predatory Beetles. Three carrot (C, G, and P) and three sweet potato Þelds (CSP, GSP, and LSP) in southern Ontario were sampled every other week, over 7 d, in the summer of 2007. Site G was located near Glencoe (42⬚44⬘47⬙ N, 81⬚42⬘32⬙ W), and sites C and P were located near Chatham (42⬚24⬘43⬙ N, 82⬚11⬘6⬙ W). Site LSP was near West Lorne (42⬚36⬘13⬙ N, 81⬚36⬘24⬙ W), and sites CSP and GSP were near Blenheim (42⬚20⬘8⬙ N, 81⬚59⬘50⬙ W). The year previously, carrot sites C and G were planted to corn, and site P was planted to tomatoes. Sweet potato sites were planted in 2006 to tobacco (CSP), soybean (GSP), and winter wheat (LSP). All sites were irrigated as needed except for carrot site G, which was without a water source. Soil types were determined by hand texture analysis to be very sandy loam for carrot sites P and C and sandy loam for carrot site G. Soil types at the sweet potato sites were sandy loam for sites LSP and CSP and loam for site GSP. Dry organic matter contents of soils for each site were as follows: G, 2.9 ⫾ 0.17%; C, 5.2 ⫾ 0.96%; P, 26.8 ⫾ 1.45%; LSP, 3.5 ⫾ 0.1%; CSP, 3.17 ⫾ 0.07%; and GSP, 2.37 ⫾ 0.09. Of the sites sampled, insecticides were applied only at site G (broad-spectrum insecticide applied twice in June). Carrot sites G and P were sampled from 23 May to 4 September and site C from 29 May to 4 September. All sweet potato sites were sampled from 20 June to 31 August. Sampling was conducted in three transects per Þeld, with each transect in carrot Þelds corre- 1107 sponding to rows that were 1, 15, and 30 m from a Þeld edge and in sweet potato Þelds in rows at 1, 10, and 50 m from the edge. Ten sampling units each 5 m apart were placed in each transect. Each sampling unit consisted of a central pitfall trap with a corn bait trap placed in the row 1 m away from the pitfall trap and a potato bait trap placed 1 m away in the opposite direction. Bait traps consisted of a mesh bag (holes ⬇1 cm diameter) constructed of a 15-cm-long cylinder of clear plastic netting (MasterNet, Mississauga, Ontario, Canada), containing either a quarter of a potato or 62.5 ml of corn, and sealed at both ends. Bait traps were placed under the soil surface. Plastic 237-ml polypropylene specimen containers of 8 cm diameter (VWR International, Mississauga, Ontario, Canada) were used as pitfall traps. Containers were quarter-Þlled with propylene glycol, placed in the soil with the top lip level with the soil surface, and protected from weather with a plastic canopy. Canopies were constructed from the container lids held above the pitfall trap with wire supports. To minimize impacts on natural populations, all traps were only in place for 1 wk every other week. Bait traps were used to capture millipedes; beetles found in these traps were ignored because beetles often escaped during bait collection. Pitfall traps were placed to capture both millipedes and beetles. Additional pitfall traps were placed in the edge habitat of each Þeld but were used only to enhance the assemblage descriptions for each site (see Appendix 1). Contents of traps were washed with ethanol and sieved. Millipedes, carabids, and staphylinids were counted and identiÞed to species with some staphylinid exceptions: species of Neohypnus Coiffait and Saiz cannot be reliably distinguished without dissection and so were treated together; most species of Aleocharinae encountered were small (3 mm), difÞcult or impossible to separate, and were therefore not considered in this study. The taxonomy of Arnett and Thomas (2001) was adopted for this study. The number of individuals of each beetle species collected from a pitfall trap was used to determine their activity density (AD), an approximation of relative abundance. AD is a product of both true density and the activity level of the individuals caught and is therefore most useful for identifying common assemblage members (see Luff 2002). Millipedes caught within a given sampling unit were combined into a single AD value. As aleocharine rove beetles were excluded, total AD for predatory beetles was determined for all non-aleocharine beetles collected in this study. Species captured were ranked in decreasing order by proportion of the total individuals caught. “Common” beetle species were identiÞed by adding species in ranked order until 95% or more of the total catch was achieved (all Þelds combined) (Luff 2002). Each common speciesÕ AD over the sampling period was compiled into a table for nonstatistical comparisons with each other and C. caeruleocinctus. Assessment of Predatory Ability. Two common carabid beetle species, Pterostichus melanarius (Illiger) and Harpalus erraticus Say, were chosen for 1108 ENVIRONMENTAL ENTOMOLOGY laboratory experiments based on availability of live specimens. P. melanarius was collected at site P and H. erraticus was collected at site LSP, with dry pitfall traps. Traps had large reßective, aluminum pie-plate canopies to reduce light stress and desiccation. Beetles were collected from 31 August to 21 September placed into individual containers with soil from their respective Þelds and transported to the University of Guelph. On the same day of collection, beetles and soil were transferred to individual square plastic 946-ml containers (Ziploc; SC Johnson, Brantford, Canada) that were ventilated by addition of a 3 by 9-cm mesh strip in the lid. Beetles were provided with a 7.5 by 8-cm waxed cardboard refuge and a vial cap as a standing water dish. Beetles were kept in a controlled environment room with a 15Ð23⬚C temperature cycle and 16:8 h L:D photoperiod to simulate natural conditions. Containers were misted daily with water. Commercially bred mealworms, Tenebrio molitor L., and standing water were provided to beetles ad libitum. Millipedes (C. caeruleocinctus) were hand-collected from site LSP on 31 August and 21 September and placed into several opaque plastic, lidded containers with soil, leaf litter, and bakerÕs yeast. Containers were misted daily. A readily available and easily cultured prey item was sought for comparison that lacked both the physical and chemical defenses of millipedes. Cabbage looper, Trichoplusia ni Hübner (Lepidoptera: Noctuidae), was selected for these reasons and because the larvae approximated the size and shape of C. caeruleocinctus. Several cultures of secondinstar cabbage loopers were obtained from colonies at the Southern Crop Protection and Food Research Centre, Agriculture & Agri-Food Canada, London, Ontario, Canada, and maintained on a modiÞed wheat germÐ based diet under the same temperature and lighting conditions as carabid beetles. Cabbage loopers were raised to the desired size range and used in feeding experiments. Experiment 1: Palatability and Size Effects. A choice experiment was conducted using millipedes of two different size categories to examine palatability of millipedes to both carabid species and the effect of prey size on predation. Millipedes were grouped into “small” (mean length: 14.0 ⫾ 0.65 mm; mean width: 1.1 ⫾ 0.04 mm) or “large” (mean length: 24.9 ⫾ 0.60 mm; mean width: 1.92 ⫾ 0.08 mm) size categories, which differed signiÞcantly from each other in length (StudentÕs t-test: n ⫽ 10, t ⫽ 12.28, df ⫽ 18, P ⬍ 0.001) and width (StudentÕs t-test: n ⫽ 10, t ⫽ 9.07, df ⫽ 14, P ⬍ 0.001). Beetles were starved for 2 d before each experiment. One large and one small millipede were randomly selected and randomly assigned to either the far left or right corners of the choice arena. Arenas were rectangular plastic containers (17 ⫻ 11 ⫻ 10 cm; Ziploc; SC Johnson) lined with soil. A staging vial was attached at one end of the arena to hold beetles before introduction into the arena. Trials were conducted under red lighting to minimize the effects of light on the beetles, as they are nocturnal (Luff 2002). Trials were conducted until 10 responsive replicates were Vol. 38, no. 4 acquired (i.e., a prey item was chosen), for each species. In each replicate, an individual beetle was selected without replacement, placed into the staging vial and allowed to move into the arena at will. The length of time from the beetle entering the arena until a choice was made was recorded. A choice was deÞned as sustained grasping of the prey item, and replicates were recorded as “nonresponsive” if no choice was made within 15 min. Beetle choices were analyzed for deviation from randomness (1:1 choice ratio) with a 2 test, in Proc FREQ (SAS 9.1). Time until choice was compared using a least signiÞcant difference test (Proc GLM; SAS 9.1) between prey types and between beetles, with shorter times interpreted as stronger responses to prey items. Statistical signiÞcance was taken at ␣ ⫽ 0.05. Experiment 2: Effects of Physical and Chemical Defense on Predation. A choice experiment using the same methodology as above was conducted with millipedes and cabbage loopers and replicated 20 times for each carabid species. Both beetle species were included in this experiment, regardless of their previous acceptance or rejection of millipedes, to ensure that the experimental setup did not adversely affect a beetleÕs receptivity to prey. Because size was found to have no effect (see Results), millipedes and loopers of similar size were selected by visual approximation, rather than by measurement. Data were analyzed as above. Spatiotemporal Associations Between Millipedes and Predatory Beetles. Beetle species found to consume millipedes in laboratory assays were selected for a study of their spatiotemporal co-occurrence with millipede populations. Linear regression analyses (Proc GLM; SAS 9.1) were conducted between beetle AD and millipede AD by crop and by individual Þeld. Beetle and millipede ADs from each sampling unit for each sampling period were used as data points in these analyses. Statistical signiÞcance was taken at ␣ ⫽ 0.05. Results Species Richness of Generalist Predatory Beetles. Excluding aleocharines, 7,380 individual beetles (7,018 collected from Þeld and 362 collected from edge habitat) were identiÞed to species. Approximately 12 aleocharine “morphospecies” (690 individuals) were found to inhabit carrot and sweet potato Þelds. The number of aleocharine morphospecies in carrot and sweet potato Þelds was similar. Fifty carabid and 25 staphylinid species were recorded and identiÞed in total. The identiÞed carabid and staphylinid species collected from pitfall traps are summarized in Appendix 1. Higher carabid species richness was found in carrot (24 ⫾ 1) than sweet potato (20 ⫾ 2) Þelds. Tachinus corticinus Gravenhorst (Staphylinidae) is newly recorded from Ontario. A species native to Europe, it was known previously in North America from Quebec, New Brunswick, Nova Scotia, Prince Edward Island, and Vermont (Campbell and Davies1991; Byers et al. 2000; Klimaszewski and Majka 2008), and these southwestern Ontario speci- August 2009 BRUNKE ET AL.: GENERALIST PREDATORS AND MILLIPEDE MANAGEMENT 1109 Table 1. Average activity densities of the millipede C. caeruleocinctus and the “common” species of carabid beetles collected in southern Ontario carrot fields Site Species G Cylindroiulus caeruleocinctus Pterostichus melanarius Harpalus pensylvanicus Ophonus puncticeps Harpalus erraticus Bembidion quadrimaculatum oppositum Pterostichus permundus Cylindroiulus caeruleocinctus Pterostichus melanarius Harpalus pensylvanicus Ophonus puncticeps Harpalus erraticus Bembidion quadrimaculatum oppositum Poecilus chalcites Scarites subterraneus Pterostichus permundus Cylindroiulus caeruleocinctus Pterostichus melanarius Harpalus pensylvanicus Ophonus puncticeps Harpalus erraticus Bembidion quadrimaculatum oppositum Poecilus chalcites Pterostichus permundus C P Mean activity density by weeka 23 May 29 May 12 Jun. 26 Jun. 10 Jul. 24 Jul. 7 Aug. 21 Aug. 4 Sep. 4.27 0 0 0.03 0 0.63 0 Ñ Ñ Ñ Ñ Ñ Ñ Ñ Ñ Ñ 15.83 0 0 0 0 4.93 1.17 0 1.48 0 0 0 0 0 0 4.40 0 0 0 0 0.23 0 0 0 2.30 0 0 0 0 0.52 0.30 0 0 0 0 0 0 0 0 5.30 0.27 0 0 0 0.20 0.07 0 0 3.31 9.45 0.03 0 0 0.52 2.31 0 0.07 0 0 0.03 0 0.10 0 0.40 0.40 0 0 0 0.50 0.33 0.03 0 0.37 49.59 0 0 0 0.93 0.85 0.04 0 0.03 0.31 0 0 0.03 0 0.23 0.07 3.40 0.03 0 0.33 0.17 0.03 0 1.03 34.76 0.28 0.03 0 0.48 0.24 0.38 Ñ Ñ Ñ Ñ Ñ Ñ Ñ 0.41 1.69 0.48 0 0 0.03 0 0 0.07 0.80 12.63 0.07 0.07 0 0.07 0.23 0 0 0 1.07 0.10 1.53 0 0 0 0 2.59 0.21 0.14 0.03 0.10 0 0 2.63 38.80 2.93 9.67 0.03 0.07 0.1 0.10 0 0 0.17 0.13 1.70 0 0 0.55 0 0.82 0.29 0.54 0 0.04 0 0 6.14 3.24 2.86 0.31 0.14 0 0 0 0 0.10 0.07 0.07 0.77 0 0.03 0.30 0.50 0.30 0.03 0.50 0 0.03 0.03 0 2.21 1.34 1.93 0.10 0.24 0 0.03 0 a Mean activity density for C. caeruleocinctus is the mean for all sampling units (i.e., bait and pitfall trap units) in the Þeld. Mean activity density for carabids is the mean for all pitfall traps in the Þeld. Dates represent the end of the sampling week. Ñ, sampling did not occur on this date at this site. mens (two male, two female) are the westernmost known in North America. Vouchers of all species and aleocharine morphospecies were deposited in the University of Guelph Insect Collection. Common Predatory Beetles and Their Temporal Distributions. Eight carabid species comprised 95% of the total individuals captured (total AD) and were therefore identiÞed as dominant, P. melanarius (65.5%), or common, Harpalus pensylvanicus (DeGeer) (8.8%), Ophonus puncticeps (Stephens) (4.7%), H. erraticus (4.7%), Bembidion quadrimaculatum oppositum Say (4.5%), Poecilus chalcites (Say) (2.4%), Scarites subterraneus Fabricius (2.1%), and Pterostichus permundus (Say) (1.2%). Species not included as common each comprised ⬍1% total AD. No identiÞed staphylinid species were found to be common members of the predatory beetle assemblage of carrot and sweet potato Þelds (but see Discussion), although some of the excluded aleocharine morphospecies were quite abundant. The temporal distributions of the above eight species and C. caeruleocinctus are described in Tables 1 and 2. At least one common species co-occurred with millipedes at any given time in both crop types, except for 29 May and 12 June at site G (Table 1). No millipedes or predatory beetles were found on 12 June at site G (Table 1). In carrot Þelds, relatively few common species were present in early summer with a greater number present later in the growing season (Table 1). This trend was also observed in sweet potato Þelds but was less pronounced (Table 2). Pterostichus melanarius, H. pensylvanicus, B. q. oppositum, and P. permundus were widespread species present in all Þelds of both crop types. H. erraticus was only absent from one Þeld (GSP), whereas S. subterraneus was more typical of sweet potato Þelds (absent from G and P), and O. puncticeps was more typical of carrot Þelds (absent from CSP and LSP). P. chalcites was present in one sweet potato and two carrot Þelds. Assessment of Predatory Ability Experiment. 1: Palatability and Size Effects. Millipedes were consumed by P. melanarius, but no size preference was observed (2 ⫽ 1.6, n ⫽ 10, df ⫽ 1, P ⫽ 0.21). Time for P. melanarius to make a choice ranged from 22 to 580 s but did not differ between prey size categories (F ⫽ 3.86, n ⫽ 10, df ⫽ 1, P ⫽ 0.08). Twelve replicates of P. melanarius were classiÞed as nonresponsive and excluded from analyses; all nonresponsive beetles were observed to circle the perimeter of the assay arena in an agitated manner without regard to prey items. No millipedes were preyed on by H. erraticus. Although two individuals of H. erraticus were observed grasping millipedes in their mandibles after trials had ended, they promptly released the millipedes and appeared to wipe their mandibles and maxillary palpi together and against the substrate. This behavior was not observed in P. melanarius. Experiment 2: Effects of Physical and Chemical Defense on Predation. When given the choice between a millipede and a cabbage looper, P. melanarius chose loopers (87.5%) signiÞcantly more often than millipedes (12.5%; 2 ⫽ 9.0, n ⫽ 16, df ⫽ 1, P ⫽ 0.003). Four replicates were classiÞed as “nonresponders.” H. erraticus always chose loopers. The two beetle species 1110 ENVIRONMENTAL ENTOMOLOGY Vol. 38, no. 4 Table 2. Average activity densities of C. caeruleocinctus and the dominant species of carabid beetles collected in southern Ontario sweet potato fields Site Species LSP Cylindroiulus caeruleocinctus Pterostichus melanarius Harpalus pensylvanicus Harpalus erraticus Bembidion quadrimaculatum oppositum Scarites subterraneus Pterostichus permundus Cylindroiulus caeruleocinctus Pterostichus melanarius Harpalus pensylvanicus Ophonus puncticeps Bembidion quadrimaculatum oppositum Poecilus chalcites Scarites subterraneus Pterostichus permundus Cylindroiulus caeruleocinctus Pterostichus melanarius Harpalus pensylvanicus Bembidion quadrimaculatum oppositum Scarites subterraneus Pterostichus permundus GSP CSP Mean activity density by weeka 20 Jun. 3 Jul. 17 Jul. 31 Jul. 14 Aug. 28 Aug. 35.77 0.97 0 0 0.07 0 0 2.87 0.43 0 0 0.03 0 0.47 0.73 1.60 0.07 0.03 0.60 3.30 0 6.23 0.37 0 0 0 0 0 2.03 0.10 0.03 0.03 0.03 0 0.33 1.03 0.63 0.07 0.07 0.23 0.20 0 6.53 0.40 0.50 0 0 0 0 0.18 0.07 0 0 0 0 0.29 0.29 0.10 0 0 0 0.17 0 18.73 0.23 0.73 0.20 0.03 0 0.03 3.20 0.07 0 0 0 0 0 0 2.27 0 0 0.07 0.07 0 80.47 2.66 0.47 2.57 0.03 0 0 14.10 0.07 0.03 0 0.03 0.03 0 0.03 1.47 0.07 0 0.07 0.03 0 69.47 2.70 0.23 2.10 0 0.03 0.03 14.13 0.43 0 0 0.07 0 0.03 0.07 6.53 0.50 0.03 0.07 0.13 0.07 a Mean activity density for C. caeruleocinctus is the mean for all sampling units (i.e., bait and pitfall trap units) in the Þeld. Mean activity density for carabids is the mean for all pitfall traps in the Þeld. Dates represent the end of the sampling week. did not differ in the time taken to choose a prey item (F ⫽ 2.18, n ⫽ 27, df ⫽ 1, P ⫽ 0.15). Time until choice did not differ in P. melanarius between those that chose loopers and those that chose millipedes (F ⫽ 0.0, n ⫽ 16, df ⫽ 1, P ⫽ 0.98). Spatiotemporal Associations Between Millipedes and Predatory Beetles. Pterostichus melanarius and C. caeruleocinctus were not found to be spatiotemporally associated in southern Ontario carrot Þelds (R2 ⫽ 0.0061, P ⫽ 0.07; Fig. 1A). However, a signiÞcant positive spatiotemporal relationship was found between P. melanarius and C. caeruleocinctus in southern Ontario sweet potato Þelds (R2 ⫽ 0.2593, P ⬍ 0.001; Fig. 1B). Spatiotemporal associations were found in each of the individual sweet potato Þelds; LSP (R2 ⫽ 0.1791, P ⬍ 0.001; Fig. 2A), GSP (R2 ⫽ 0.079, P ⬍ 0.001; Fig. 2B), and CSP (R2 ⫽ 0.3209, P ⬍ 0.001; Fig. 2C). Discussion Predaceous Beetle Assemblages in Ontario Carrot and Sweet Potato Fields. Our carabid assemblage descriptions are derived from one season of data and as such cannot be considered complete in both composition and proportional abundance because of the strong annual variability in agroecosystems (Scott and Anderson 2003). However, while abundances will vary, the status of some species as common is reasonably stable and predictable from year to year because most of the assemblage variability is caused by uncommon species (Scott and Anderson 2003). Both carrot and sweet potato Þelds were found to be less species rich in Carabidae (24 ⫾ 1 and 20 ⫾ 2 species, respectively) compared with the average value for agroecosystems in North America (30 ⫾ 2 species) (Luff 2002). Low carabid richness relative to this av- erage has been reported in carrot agroecosystems in the Netherlands and Quebec and was attributed to lack of early season crop cover, preventing long-term moist microhabitats (Booij and Noorlander 1992; Boivin and Hance 2002). This explanation Þts with our result that a greater number of common carabid species were found later in the season than earlier. However, species richness in this study may also have been limited by the low rainfall experienced in southern Ontario in 2007, as evidenced by declines in arthropod activity density at some sites in June and July. Between-row cultivation in both crops tends to decrease as the plant canopies close, and this early season disturbance may have played a role in determining richness and abundance patterns (Thorbek and Bilde 2004). The broad- spectrum insecticide applications made at site G may have accounted for the absence of millipedes and predatory beetles in our traps on 12 June, because nontarget effects of sprays are often substantial (Huusela-Veistola 1996). Differences in edge habitat were not quantitatively measured but may also account for some of the differences observed between Þelds (Dauber et al. 2005). Low species richness of Staphylinidae was an artifact of two characteristics of this study. First, pitfall traps do not adequately sample staphylinid assemblages and can underestimate their abundance (Lang 2000). Halsall and Wratten (1988) found a pitfall trap bias against Tachyporus Gravenhorst, a rove beetle genus that was found in most Þelds surveyed (Appendix 1). Alternative trapping methods, such as suction sampling (Holland et al. 2004) or coverboards (Davalos and Blossey 2006) are needed to sample staphylinids and carabids more accurately. Second, aleocharine species were excluded from analyses in this study, and this subfamily is the most speciose of the rove beetles August 2009 BRUNKE ET AL.: GENERALIST PREDATORS AND MILLIPEDE MANAGEMENT 1111 100 90 A 80 70 60 50 y = 0.4506x + 7.7139 R² = 0.0061 p = 0.07 P. melanarius Activity Density 40 30 20 10 0 0 5 10 15 20 25 30 35 12 B 10 y = 0.0195x + 0.2245 R² = 0.2593 p < 0.0001 8 6 4 2 0 0 50 100 150 200 250 300 350 400 C. caeruleocinctus Activity Density Fig. 1. Activity density of P. melanarius in relation to the activity density of C. caeruleocinctus in southern Ontario carrot Þelds (A) and sweet potato Þelds (B). Activity density of P. melanarius was measured using pitfall traps, whereas that of C. caeruleocinctus was measured using corn and potato bait traps as well as pitfall traps. (Arnett and Thomas 2001). Aleocharine morphospecies found in Þelds may consist of many true species. This subfamily was the dominant group of Staphylinidae in this study, consistent with results in New Zealand from the only other study to evaluate the rove beetle assemblage of carrots in detail (Sivasubramaniam et al. 1997). It is unlikely that this group of small beetles (most 3 mm) preys on adult or subadult C. caeruleocinctus, but the role of aleocharines in millipede egg predation could be signiÞcant as some species are known to be important in dipteran egg mortality (Andersen et al. 1983). Common Species in Carrot and Sweet Potato Fields. The eight species identiÞed as common in this study were all carabids and are typical of agroecosystems and of northeastern North American synanthropic habitats in general (Boivin and Hance 2002, Cardenas and Buddle 2008). Indeed, four of the eight common species were found in every Þeld. P. melanarius is frequently the dominant carabid in synanthropic ecosystems in both North America and in its native Europe (Cardenas and Buddle 2008). Predatory Ability of Two Common Carabids. Pterostichus melanarius consumed millipedes in situations with and without the option of alternative prey. This is the Þrst known record of P. melanarius consuming millipede prey. This species does not seem to be limited by the physical or chemical defenses of millipedes. P. melanarius was also observed to consume millipedes in the Þeld, and it is possible that millipedes are accepted more readily in nature than in our experiments. Although soft-bodied prey (cabbage loopers) were chosen more often than millipedes, this response is expected in a generalist predator, where the tendency to minimize handling effort is well known (Sunderland 2002). Native Pterostichus species have been found to readily accept millipedes in the families Julidae and Polydesmidae (Snider 1984). Both larval and adult stages of those Pterostichus spp. survived on a diet solely of julid prey, and viable offspring were produced (Snider 1984). It seems likely that P. melanarius can do the same, particularly because C. caeruleocinctus is also a native European species. Snider (1984) showed that many Pterostichus spp. can consume large numbers of millipedes per day. P. melanarius was never observed to consume large quantities of C. caeruleocinctus, but predation in our studies was less than that observed by Snider (1984), likely because of less favorable conditions, especially lower humidity, in captivity. 1112 ENVIRONMENTAL ENTOMOLOGY Vol. 38, no. 4 12 A 10 y = 0.0162x + 0.6374 R² = 0.1791 p < 0.0001 8 6 4 2 0 0 50 100 150 200 250 300 350 400 P. melanarius Activity Density 4.5 4 B 3.5 3 2.5 y = 0.0138x + 0.1119 R² = 0.079 p = 0.0001 2 1.5 1 0.5 0 0 10 20 30 40 50 60 70 80 90 3.5 3 C 2.5 y = 0.0402x + 0.0324 R² = 0.3209 p < 0.0001 2 1.5 1 0.5 0 0 10 20 30 40 50 60 C. caeruleocinctus Activity Density Fig. 2. Activity density of P. melanarius in relation to the activity density of C. caeruleocinctus in southern Ontario sweet potato Þelds: (A) site LSP, (B) site GSP, and (C) site CSP. Activity density of P. melanarius was measured using pitfall traps, whereas that of C. caeruleocinctus was measured using corn and potato bait traps as well as pitfall traps. Cylindroiulus caeruleocinctus was unpalatable to H. erraticus in laboratory feeding experiments, and it seems that this species is not a millipede predator. Cabbage loopers were eaten readily, and both H. erraticus and P. melanarius took the same time to choose a prey item, eliminating the possibility that laboratory conditions prevented feeding behavior in H. erraticus. It is unlikely that H. erraticus was deterred by the physical defenses of C. caeruleocinctus, because it belongs to a genus of primarily omnivorous beetles that have mandibles useful for the cracking of seeds (Ingerson-Mahar 2002). Given that individuals attempt- August 2009 BRUNKE ET AL.: GENERALIST PREDATORS AND MILLIPEDE MANAGEMENT ing to feed on millipedes promptly rejected them and exhibited wiping behavior, it seems likely that H. erraticus was deterred by the defensive secretions of C. caeruleocinctus. Spatiotemporal Co-occurrence of P. melanarius With C. caeruleocinctus. A signiÞcant relationship between the ADs of P. melanarius and C. caeruleocinctus indicates that that these two species indeed co-exist both spatially and temporally in sweet potato Þelds. The amount of variation explained by millipede activity density was relatively low (r2 ⫽ 0.259), but this is to be expected with generalist predators that are typically not tightly linked to any one prey species. An improved linear model might be achieved by using a trapping method such as fenced pitfalls, which provide a greater measure of true carabid density and do not overestimate densities of large species as unfenced pitfalls do (Mommertz et al. 1996). No relationship was found between the ADs of these species in carrot Þelds, which may be because of a greater abundance of alternative prey or a lower abundance of millipedes in carrot Þelds relative to P. melanarius. In sweet potato Þelds, millipede ADs were consistently much greater than that of their co-occurring common carabids, whereas this was not seen in the carrot Þelds studied (Tables 1 and 2). At high densities, C. caeruleocinctus may become an important prey species for P. melanarius, and this may be mediated by the tendency of C. caeruleocinctus to form large, local aggregations in carrot and sweet potato Þelds (Sears et al., unpublished data). Other Natural Enemies of C. caeruleocinctus. The predatory potential of the other six common predatory beetles remains to be evaluated directly. B. q. oppositum is almost certainly too small to prey on any millipede life stages, except eggs and Þrst instars. This species is known to eat the eggs of pest Diptera (Sunderland 2002) and may also feed on C. caeruleocinctus eggs. H. pensylvanicus and O. puncticeps are omnivorous, consuming a variety of prey from seeds to arthropods (Arnett and Thomas 2001; Honek et al. 2007). Snider (1984) tested the palatability of millipedes to one H. pensylvanicus individual, which refused this prey. P. chalcites is known as a primary generalist insectivore (Sunderland 2002), but its status as a millipede predator is unknown. The genus Poecilus is closely related to Pterostichus, sometimes placed within it (Downie and Arnett 1996), and may share some of its biology. P. permundus overlaps the size range of P. melanarius (unpublished data) and is likely to follow the trend of its congeners in consuming subadult and adult millipedes. S. subterraneus is known to consume a wide range of large arthropod prey (Sunderland 2002), and Baker (1985) reported millipede predation in the congener S. cyclops. The generalist assemblage was examined for species that were common, accepted millipede prey, and coexisted with millipedes. P. melanarius exhibited all of these characteristics, and it is concluded that this species is a natural predator of C. caeruleocinctus in Ontario sweet potato and carrot Þelds and may play an important role in millipede population regulation. 1113 This is the Þrst study to examine the predatory beetle assemblage of these crops in Ontario, and the Þrst study to address the natural control of C. caeruleocinctus in North American vegetable crops. The effect of millipede density on millipede mortality by natural predation is unknown and may be an important aspect of future IPM strategies for C. caeruleocinctus. Acknowledgments We thank L. OÕKeefe and T. Marowa for their role in the intensive Þeldwork of this project, and A. Verhallen (Ontario Ministry of Agriculture, Food and Rural Affairs [OMAFRA]) for soil textural analyses. Drs. J. Allen and M. Filotas (OMAFRA) provided logistical support and background knowledge of carrot and sweet potato production in Ontario. We thank the participating vegetable growers who volunteered Þelds for study and L. Verdon (AAFC) for provision of cabbage loopers. This project was supported by the Canada-Ontario Research and Development Program and the Fresh Vegetable Growers of Ontario through funding to M.K.S. and R.H.H. and by a Natural Sciences and Engineering Research Council Undergraduate Summer Research Assistantship to A.J.B. References Cited Allen, J., and M. Filotas. 2008a. Millipedes attacking sweet potatoes and carrots in OntarioÑthe latest info. Part 1: tracking millipede populations in the Þeld. (http://www. omafra.gov.on.ca/english/crops/hort/news/hortmatt/ 2008/02hrt08a1.htm). Allen, J., and M. Filotas. 2008b. Millipedes attacking sweet potatoes and carrots in OntarioÑthe latest info. Part 2: identifying pest millipedes and the damage they do. (http:// www.omafra.gov.on.ca/english/crops/hort/news/hortmatt/ 2008/03hrt08a1.htm). Andersen, A., A. G. Hansen, N. Rydland, and G. Øyre. 1983. Carabidae and Staphylinidae (Coleoptera) as predators of eggs of the turnip root ßy Delia floralis Fallén (Diptera, Anthomyiidae) in cage experiments. Z. Angew. Entomol. 95: 499 Ð506. Arnett, R. H., Jr., and M. C. Thomas. 2001. American beetlesÑvolume I. CRC Press, New York. Arnett, R. H., Jr., R. H. Arnett, Sr., and M. C. Thomas. 2002. American beetlesÑvol. II. CRC Press, New York. Baines, D., R. Steward, and G. Boivin. 1990. Consumption of carrot weevil (Coleoptera: Curculionidae) by Þve species of carabids (Coleoptera: Carabidae) abundant in carrot Þelds in southwestern Quebec. Environ. Entomol. 19: 1146 Ð1149. Baker, G. H. 1985. Predators of Ommatoiulus moreletii (Lucas) (Diplopoda: Iluidae) in Portugal and Australia. J. Austral. Entomol. Soc. 24: 247Ð252. Boivin, G. 1999. Integrated management for carrot weevil. Integr. Pest Manag. Rev. 4: 21Ð37. Boivin, G., and T. Hance. 2002. Ground beetle assemblages in cultivated organic soil and adjacent habitats: temporal dynamics of microspatial changes. Pedobiologica 47: 193Ð 202. Booij, C.J.H., and J. Noorlander. 1992. Farming systems and insect predators. Agric. Ecosyst. Environ. 40: 125Ð135. Byers, R. A., G. M. Barker, R. L. Davidson, E. R. Hoebeke, and M. A. Sanderson. 2000. Richness and abundance of Carabidae and Staphylinidae (Coleoptera), in Northeastern dairy pastures under intensive grazing. Great Lakes Entomol. 33: 81Ð106. 1114 ENVIRONMENTAL ENTOMOLOGY Campbell, J. M., and A. Davies. 1991. Family Staphylinidae. Rove beetles, pp. 86 Ð124. In Y. Bousquet (ed.), Checklist of beetles of Canada and Alaska. Publication 1861. E. Agriculture Canada, Ottawa, Canada. Cardenas, A. M., and C. M. Buddle. 2008. Introduced and native ground beetle assemblages (Coleoptera: Carabidae) along a successional gradient in an urban landscape. (http:// www.springerlink.com/index/a76v56557m2g7254.pdf). Chalfant, R. B., R. K. Jansson, D. R. Seal, and J. M. Schalk. 1990. Ecology and management of sweet potato insects. Annu. Rev. Entomol. 35: 157Ð180. Coleman, D. C., D. A. Crossley, Jr., and P. F. Hendrix. 2004. Fundamentals of soil ecology. Elsevier Academic, New York. Dauber, J., T. Purtauf, A. Allspach, J. Frisch, K. Voigtlander, and V. Wolters. 2005. Local vs. landscape controls on diversity: a test using surface-dwelling soil macroinvertebrates of differing mobility. Global Ecol. Biogeogr. 14: 213Ð221. Davalos, A., and B. Blossey. 2006. Assessing capture techniques for ground beetles. Can. Entomol. 138: 118 Ð122. Downie, N. M., and R. H Arnett, Jr. 1996. The beetles of Northeastern North America. Volume 1: introduction; suborders Archostemata, Adephaga, and Polyphaga, thru superfamily Cantharoidea. Sandhall Crane Press, Gainesville, FL. Ebregt, E., P. C. Struik, B. Odongo, and P. E Abidin. 2005. Pest damage in sweet potato, groundnut and maize in north-eastern Uganda with special reference to damage by millipedes (Diplopoda). Neth. J. Agr. Sci. 53: 49 Ð 69. Halsall, N. B., and S. D. Wratten. 1988. The efÞciency of pitfall trapping for polyphagous predatory Carabidae. Ecol. Entomol. 13: 293Ð299. Holland, J. M., L. Winder, C. Woolley, C. J. Alexander, and J. N. Perry. 2004. The spatial dynamics of crop and ground active predatory arthropods and their aphid prey in winter wheat. Bull. Entomol. Res. 94: 419 Ð 431. Honek, A., Z. Martinkova, P. Saska, and S. Pekar. 2007. Size and taxonomic constraints determine the seed preferences of Carabidae (Coleoptera). Basic Appl. Ecol. 8: 343Ð353. Huusela-Veistola, E. 1996. Effects of pesticide use and cultivation techniques on ground beetles (Col., Carabidae) in cereal Þelds. Ann. Zool. Fenn. 33: 197Ð205. Ingerson-Mahar, J. 2002. Relating Diet and morphology in adult carabid beetles, pp. 111Ð137. In J. M. Holland (ed.), Vol. 38, no. 4 The agroecology of carabid beetles. Intercept, Hampshire, United Kingdom. Jacobson, M. 1966. Chemical insects, attractants and repellents. Annu. Rev. Entomol. 11: 403Ð 422. Klimaszewski, J., and C. G. Majka. 2008. Introduced Staphylinidae (Coleoptera) in the maritime provinces of Canada. Can. Entomol. 140: 48 Ð72. Lang, A. 2000. The pitfalls of pitfalls: a comparison of pitfall trap catches and absolute density estimates of epigeal invertebrate predators in arable land. J. Pest Sci. 73: 99 Ð 106. Luff, M. L. 2002. Carabid assemblage organization and species composition. In J. M. Holland (ed.), The agroecology of carabid beetles, pp. 41Ð 81. Intercept, Hampshire, United Kingdom. Mommertz, S., C. Schauer, N. Kosters, A. Lang, and J. Filser. 1996. A comparison of D-Vac suction, fenced and unfenced pitfall trap sampling of epigeal arthropods in agroecosystems. Ann. Zool. Fenn. 33: 117Ð124. [OMAFRA] Ontario Ministry of Agriculture, Food and Rural Affairs. 2008. Current minor use registrations, priorities and projects for sweet potato. (http://www.omafra.gov. on.ca/english/crops/minoruse/minoruse_swpotato.htm). Scott, W. A., and R. Anderson. 2003. Temporal and spatial variation in carabid assemblages from the United Kingdom Environmental Change Network. Biol. Conserv. 110: 197Ð210. Sivasubramaniam, W., S. D. Wratten, and J. Klimaszewski. 1997. Species composition, abundance, and activity of predatory arthropods in carrot Þelds, Canterbury, New Zealand. New Zeal. J. Zool. 24: 205Ð212. Snider, R. M. 1984. Diplopoda as food for Coleoptera: laboratory experiments. Pedobiologia 26: 197Ð204. Sunderland, K. D. 2002. Invertebrate pest control by carabids, pp. 165Ð215. In J. M. Holland (ed.), The agroecology of carabid beetles. Intercept, Hampshire, United Kingdom. Thorbek, P., and T. Bilde. 2004. Reduced numbers of generalist arthropod predators after crop management. J. Appl. Ecol. 41: 526 Ð538. [USSPC] U.S. Sweet Potato Council. 2008. 2007 Sweet potato statistical yearbook. (http://sweetpotatousa.org/ yearbook.htm). Received 7 August 2008; accepted 18 April 2009. August 2009 BRUNKE ET AL.: GENERALIST PREDATORS AND MILLIPEDE MANAGEMENT 1115 Appendix 1. Predatory beetle species collected in pitfall traps placed in field (F) and edge (E) habitats of carrot and sweet potato fields, in southern Ontario Carrot Species list Carabidae Cicindelinae: Cicindelini Cicindela Harpalinae: Chlaeniini Chlaenius Cyclosomini Tetragonoderus Harpalini Acupalpus Anisodactylus Bradycellus Dicheirotrichus Harpalus Ophonus Stenolophus Lebiini Microlestes Licinini Diplocheila Platynini Agonum Calathus Pterostichini Poecilus Pterostichus Cyclotrachelus Zabrini Amara Scaritinae: Clivinini Clivina Scaritini Scarites Trechinae: Bembidiini Bembidion Elaphropus Trechini Trechus G Sweet potato C P LSP GSP CSP F E F E F E F E F E F E punctulata repanda X X / / X X / / X / X / X / X / X / X / X / / / emarginatusa nemoralis pusillus / / / / X / / / / / / / X / X / / / / X / / X / / / / / / / / X / / / X fasciatus X / X / / / / / / / / / partiarius rusticusa santaecrucis rupestrisa tantillus cognatus affinis caliginosusa compar eraticus fulvilabris indianus longicollisa pennsylvanicus somnulentrus puncticeps comma ochropezus X X X / / / / / / / X / / / / / / / / / / / / / X / / / / / / / / / / / / / / / / / X X / / / / / / / X X X X / X / X X / X / / / X / / / X / X / / X X / X X X X X X / X X / / X / X X / X / X / X / / / X / X X / / / X / X X X / / X / X / / / X / X / / X / / X X / / / X X / X / X / / X X / / / X / / / / X / / X / / / / X / X X / / / / / / / / / / / X / / X / / / X / / / X / / X / / / / / / / / / X / / / / brevilobus X / X X X / / / X / / / obtusa / / / X / / / / / / / / cupripennea muelleri nutans placidum opacula / / / / / / / / / X / / / X / / / / / / / X / X / / / / / / X X X / / / / / / / / / / / / / / / / / / X X / / / / / / / chalcites lucublandus melanarius permundus sodalis / X X X / / / X X X X / X X / / / X / / X / X X / / / X X / / X X X X / X X X X X X X X X / / / / X / / X X / / / X / / avida cupreolata familiaris musculus rubricaa / X X X / / X / / X / / X / / / / / X / / / / / / / / / / / X X X X / / X / / / / X / X / / / / / / / / / X / / / / / / impressefrons / / X / X / / / / / X / subterraneus / / X / / / X / X X X X affinea nitidum obtusum quadrimaculatum oppositum anceps xanthopusa X X / / / / / / X / / / / / X / / / / / X / / / / / / / / / / / / / / / X X / / / / X X / / / / X X X / / / X X / / / / X X / X / / X X / X / / quadristriatus / / X / X X / / / / X / Continued on following page 1116 Appendix 1. ENVIRONMENTAL ENTOMOLOGY Vol. 38, no. 4 Continued Carrot Species list Staphylinidae Paederinae: Paederini Lobrathium Scopaeus Oxytelinae: Oxytelini Anotylus Thinobiini Carpelimus Staphylininae: Staphylinini Belonuchus Erichsonius Neobisnius Philonthus Tasgius Xantholinini Neohypnus Stenistoderus Steninae Stenus Tachyporinae: Mycetoporini Bryoporus Mycetoporus Tachyporini Tachinus Tachyporus G Sweet potato C P LSP GSP CSP F E F E F E F E F E F E collare dimidiatuma quadriceps / / / / / / X X / / / / X / X / / / X / / / / / / / / / / / / / / / / / brevicepsa insignitus / / / / / / / / / / / / X X / / / / / / / / / / obesusa / / X / / / / / / / / / rufipennis nanus sobrinus carbonarius caucasicus concinnusa janusa ater / X X / X / X / X / / / / / / / X / / / / / / / / / / X / / / / / / / / / / / / / / / / / / / X / / / / / / / / / / / X / / / / X / / / / / / / / / / / / / / / X / / X / X / X / / / / / / / X spp.b rubripennisa X / / / X X / / X / / / X / / / / / / / X / / / pubescensa stygicusa / / / / / / / / / X / / / / X / / / / / / / / / rufescens consors / / / / / X / / X X / / X X / / / X X / / / / / corticinusc dispara maculocollis nitidulus / / / X X / / / / / / X / / / / / X X / / / / / X / X / / / / / / / / / / / / / / / / X / / / / Edge traps were excluded from the main study because of variable edge habitat and asynchrony of the two trap collection schedules. a Species represented by one specimen. b Species included obscurus and melanops. c Newly recorded for Ontario. X, present; /, absent.