Blackwell Publishing Ltd Temporal variation in arthropod sampling effectiveness: the case for using the beat sheet method in cotton Mark R. Wade1*, Brad C.G. Scholz2, Richard J. Lloyd2, Amanda J. Cleary1,2, Bernie A. Franzmann2 & Myron P. Zalucki1 1 Department of Entomology and Zoology, School of Integrative Biology, The University of Queensland, St Lucia, Queensland 4072, Australia, 2Queensland Department of Primary Industries and Fisheries, Toowoomba, Queensland 4350, Australia Accepted: 29 March 2006 Key words: cage, diel variation, Gossypium hirsutum, ground cloth, Nabis kinbergii, phytophagous, predatory, Rhyzobius lophanthae, sampling techniques, suction, visual sampling Abstract Predatory insects and spiders are key elements of integrated pest management (IPM) programmes in agricultural crops such as cotton. Management decisions in IPM programmes should to be based on a reliable and efficient method for counting both predators and pests. Knowledge of the temporal constraints that influence sampling is required because arthropod abundance estimates are likely to vary over a growing season and within a day. Few studies have adequately quantified this effect using the beat sheet, a potentially important sampling method. We compared the commonly used methods of suction and visual sampling to the beat sheet, with reference to an absolute cage clamp method for determining the abundance of various arthropod taxa over 5 weeks. There were significantly more entomophagous arthropods recorded using the beat sheet and cage clamp methods than by using suction or visual sampling, and these differences were more pronounced as the plants grew. In a second trial, relative estimates of entomophagous and phytophagous arthropod abundance were made using beat sheet samples collected over a day. Beat sheet estimates of the abundance of only eight of the 43 taxa examined were found to vary significantly over a day. Beat sheet sampling is recommended in further studies of arthropod abundance in cotton, but researchers and pest management advisors should bear in mind the time of season and time of day effects. Introduction Regular sampling of phytophagous and entomophagous arthropods is necessary to estimate changes in their population size for ecological studies. These estimates, combined with knowledge of their potential impact, can be used to formulate and assess pest management decisions as part of an integrated pest management (IPM) programmes (Binns et al., 2000). Failure to accurately estimate arthropod abundance can lead to inappropriate selection and timing of management tactics, such as an insecticide application. It is undesirable to not spray when required, or to apply sprays when not required. Various *Correspondence: Mark R. Wade, Department of Entomology and Zoology, School of Integrative Biology, The University of Queensland, St Lucia, Queensland 4072, Australia. E-mail: markrwade@yahoo.com.au. methods have been used by researchers, pest management advisors, and growers to sample arthropods, such as the beat bucket, beat sheet, fumigation cage, pitfall trap, sweepnet, suction or D-Vac, visual examination, and whole plant bagging. Choice of method is dependent on several interrelated variables, such as plant type, plant phenology and condition, target species, accuracy, precision, ease of use, speed, and cost. Consideration of these variables has formed the basis for numerous studies aimed at comparing the effectiveness of various methods for sampling arthropods (e.g., Shepard et al., 1974a; Young & Tugwell, 1975; González et al., 1977; Byerly et al., 1978; Wilson & Gutierrez, 1980; Bechinski & Pedigo, 1982; Fleischer & Allen, 1982; Garcia et al., 1982; Nuessly & Sterling, 1984; Browde et al., 1992; Knutson & Wilson, 1999; McLeod, 2000). In general, the ‘best’ method should detect all key arthropods and be suitable for use over the whole growing season. Furthermore, sampling equipment, if any, should ideally be readily © 2006 The Authors Entomologia Experimentalis et Applicata 120: 139–153, 2006 Journal compilation © 2006 The Netherlands Entomological Society 139 140 Wade et al. available, cheap to purchase and maintain, easy to carry, simple to use, and unaffected by user bias. Although no single sampling method has been unanimously identified as the ‘best’, the beat sheet has often ranked highly (e.g., Shepard et al., 1974a; Young & Tugwell, 1975; Studebaker et al., 1991). Also known as a beat cloth, drop cloth, drop sheet, ground cloth, plant shake, shake cloth, or shake sheet, the beat sheet involves beating or shaking a plant, or group of plants, to dislodge the arthropods in the foliage onto a sheet spread on the ground, where they can be quickly counted. The beat sheet is considered fast, inexpensive, easy to use, accurate, and precise compared with visual inspection. However, it is limited to dry conditions, upright plants grown in rows and for sampling arthropods that are easily dislodged, slow moving, and rapidly distinguishable (Shepard et al., 1974a; Young & Tugwell, 1975; Bechinski & Pedigo, 1982; Knutson & Wilson, 1999; Kharboutli & Allen, 2000). The beat sheet is classified as a relative sampling method, as not all arthropods are detected and estimation of the actual population size requires reference to an absolute sampling method (Marston et al., 1979; Studebaker et al., 1991). To use beat sheet sampling in cotton and other agricultural crops requires an appreciation of some of the interrelated abiotic and biotic factors that influence its effectiveness. However, the influences of the time of the growing season and time of day on beat sheet estimates of abundance have received scant attention in the literature. Maintaining accurate estimates of arthropod abundance at various stages of the growing season is important because tolerance to insect damage varies with the plant developmental stage and the seasonal abundance of arthropods is known to vary. Sweep-net, suction, and visual sampling methods are acknowledged as being less effective later in the growing season when the plant canopy is large and hence less of it is sampled (Shepard et al., 1974a; Smith et al., 1976; Byerly et al., 1978; Wilson & Gutierrez, 1980; Garcia et al., 1982; Snodgrass, 1993). However, the sensitivity of the beat sheet to seasonal variation is less certain. There is a notion that beat sheet sampling collects arthropods from the entire plant canopy, regardless of the canopy size and thus the time of the growing season (Shepard et al., 1974a), but this has not been confirmed. In only a small number of studies have the experimental design and analyses adequately permitted consideration of the influence of seasonal variation on beat sheet estimates of arthropod abundance (Shepard et al., 1974a; Adams et al., 1984; Studebaker et al., 1991; Snodgrass, 1993; McLeod, 2000). Furthermore, these have only pertained to a handful of predaceous and phytophagous species. For example, spider (unidentified) abundance in soybean was estimated to be approximately three per sample early in the growing season, regardless of the sampling method used, but 10 weeks later the estimate was five per sample using the beat sheet vs. only two per sample using sweep-net and suction sampling (Shepard et al., 1974a). Determining the ‘ideal’ time of day to conduct beat sheet samples is important because estimates of arthropod populations within a habitat are likely to vary over a day due to changes in their activity and distribution. The diel or diurnal rhythms of arthropods may cause them to move between vegetation types (Dempster, 1957) or vertically in the same vegetation type (Fewkes, 1961; Shepard et al., 1974a). There may be changes over a day in ‘alertness’, which enables them to more readily escape when disturbed by the sampler (or a predator) at certain times, and in the proportion of individuals that are airborne (Southwood et al., 1961). Many studies have recognized the time of day effects involving sampling methods other than the beat sheet (Fewkes, 1961; Dumas et al., 1962, 1964; Benedek et al., 1972; Sevacherian & Stern, 1972; González et al., 1977; Leathwick & Winterbourn, 1984; Braman & Yeargan, 1989; Schotzko & O’Keeffe, 1989; Browde et al., 1992; Rancourt et al., 2000). For example, Fewkes (1961) collected 7.6 times more damsel bugs (Nabis spp.) with a sweep-net in grass at night than by day and González et al. (1977) detected equivalent numbers of damsel bugs and spiders (unidentified), but more big-eyed bugs (Geocoris spp.), pirate bugs [Orius tristicolor (White)], and green lacewings (Chrysopa spp.) in suction samples conducted during the morning (06:00–09:00 hours) than afternoon (17:00 – 20:00 hours). One study that involved a beat sheet found more big-eyed bug nymphs, but not adults, in the morning (09:45–10:25 hours) than afternoon (15:45–16:15 hours) in soybean (Shepard et al., 1974b). They concluded that due to the mechanics of the method, the beat sheet would preferentially dislodge big-eyed bugs when they were in higher plant strata positions during the morning. In contrast, Studebaker et al. (1991) found no differences between beat sheet or sweep-net estimates of soybean looper, Pseudoplusia includens Walker, abundance in soybean at 09:00, 13:00, and 17:00 hours. The effect of the time of day on beat sheet estimates of the abundance of a wider range of taxa remains uncertain. Here we report on the both influences of the time of day and time of growing season on sampling effectiveness in an agricultural crop. In trial one, suction and visual sampling methods were compared with the beat sheet, with reference made to an absolute cage clamp method for determining the abundance of entomophagous arthropods over 5 weeks. In the second trial, relative estimates of entomophagous and phytophagous arthropod abundance were made using beat sheet samples taken at hourly intervals between 06:00 and 18:00 hours. Temporal variation in sampling effectiveness 141 Materials and methods Trial 1: time of season effects The first trial was conducted at the Queensland Department of Primary Industries and Fisheries, Gatton Research Station, Australia. The trial area was 160 m (rows) wide by 150 m long (ca. 2.4 ha) and planted with cotton, Gossypium hirsutum L. (Malvaceae) cv. Siokra V16 in rows 1 m apart on 14 December 2000. Cotton was grown using standard agronomic practices, but without pesticide applications and supplementary irrigation. Sampling was conducted weekly for five consecutive weeks between the vegetative and late-squaring plant growth stages, 34 –60 days after planting (DAP). This period coincided with a rapid increase in plant canopy size and arthropod abundance. Cage, beat sheet, visual, and suction samples were made on 10 randomly selected lengths of rows of cotton plants on each census date to assess densities of entomophagous arthropods, but only eight visual samples were made at 39 DAP due to rainfall (see Table 1 for details of species recorded). Therefore, on each census date the design comprised a completely randomized design. Cage sampling was considered an absolute method, while the remainder were relative methods for assessing abundance. In cage sampling, an A-frame cage was clamped over a 1-m row of cotton, similar to Bechinski & Pedigo (1982). The cage measured 1 m2, and was made from two pieces of aluminium fly screen hinged on one side. A fine polyester gauze bag was fitted to each section of the frame. The cage was held in place over the plants by large clips. The caged plants were hit vigorously with an open hand to knock the arthropods off the plants into the cage. All the arthropods inside the cage were counted and removed to avoid recounting once the cage was opened. Plants were subsequently inspected for arthropods that remained on the plants. In beat sheet sampling, the arthropods in 1 m of row were dislodged from the plants with a stick onto a yellow sheet, from where they were counted. The sheet was 2.5 × 1.5 m and made from yellow woven polyethylene fabric (Canvacon®, Southcorp Industrial Textiles, Clayton, Victoria, Australia), and had two 1.5-m-long wooden dowel rods fixed to each end to prevent the ends from being lifted easily by the wind. The sheet was placed behind the row of cotton plants to be sampled, along the ground in the interrow and up over the adjacent row of cotton, to create a ‘wall’ to deflect or catch flying arthropods (see Deighan et al., 1985). A single 1-m-long wooden dowel rod was used to shake the cotton plants in 1 m of row. Plants were struck 6 –10 times from the base to the top of the plant. The arthropods that remained on the sheet after counting was completed were shaken back off onto the foliage from where they came from. In visual sampling, the entire plant surface was carefully inspected and the numbers of arthropods on each of five consecutive plants counted. In suction sampling, a Stihl BG72 garden blower/vacuum machine was used to draw the arthropods off cotton plants along 20 m of row. Two passes of the machine were made over the top and both sides of the cotton row, i.e., six passes in total. Each collection was emptied into a jar of 70% ethanol and returned to the laboratory for sorting and counting. Visual and suction sampling data were transformed to numbers per metre of row for comparison with the beat sheet and cage methods, based on 7.6 ± 0.4 (n = 10) cotton plants per metre. It was predicted that the relative estimates of arthropod abundance would increase over the five consecutive weeks of sampling, but that relatively fewer of these arthropods would be counted over this period using visual and suction sampling compared with beat sheet and cage sampling. A series of repeated measures ANOVA tests (SAS MIXED procedure, SAS release 8.2, SAS Institute, Cary, NC, USA) was used to determine the effects of sampling method, census date, and sample method × census date on the abundance of each arthropod taxon. The use of repeated measures ANOVA tests rather than separate one-way ANOVA tests on each census date permitted the effects of census date, and critically, the interaction of sampling method and census date to be determined (Everitt, 1995). In the repeated measures tests, sample date was applied as a fixed, repeated effect with a compound symmetry covariance structure, sampling method as a fixed effect, and replicate sample nested in sampling method as a random effect. The data were rank transformed prior to ANOVA tests to improve normality and/or homogeneity of variances (Conover & Iman, 1981). Where the transformation did not correct the violation, often due to a high frequency of zero values in a particular data set, the results of the statistical analyses are not presented, as their interpretation was not valid. Untransformed means are reported in the results. All statistical tests were considered at an overall significance level of α = 0.05. Trial 2: time of day effects The second trial was conducted in a section of unsprayed cotton (cv. NuPearl) of ca. 3.5 ha at ‘Coondarra’, a commercial farm at Jimbour, Queensland. The trial area was 12-row pairs (36 × 200 m), and surrounded by three row-pairs of cotton on each side and at least 200 m of cotton on each end as a non-sampled buffer. The cotton was planted in rows 1 m apart with a single-skip planting configuration on 10 November 2000. Here, the two rows planted out of a possible three were referred to as a 142 Wade et al. Table 1 Details of entomophagous, phytophagous, and non-pest arthropods commonly recorded1 in trials 1 and 2 in unsprayed cotton Order Entomophagous Araneae Coleoptera Hemiptera Hymenoptera Neuroptera Orthoptera Family Species Common name Clubionidae Oxyopidae Salticidae Theridiidae Thomisidae Assorted Carabidae Coccinellidae Cheiracanthium spec. Oxyopes spp. Assorted species Achaearanea veruculata (Urquhart) Assorted species Assorted species Assorted species Coccinella transversalis Fabricius Coelophora inaequalis (Fabricius) Diomus notescens (Blackburn) Harmonia octomaculata (Fabricius) Hippodamia variegata (Goeze) Micraspis frenata (Erichson) Rhyzobius lophanthae (Blaisdell) Dicranolaius bellulus (Guérin-Méneville) Orius spec. Geocoris lubra Kirkaldy Campylomma liebknechti (Girault) Deraeocoris signatus (Distant) Nabis kinbergii Reuter Oechalia schellembergii (Guérin-Méneville) Iridomyrmex spec. Pheidole spec. Assorted species Mallada signata (Schneider) Micromus tasmaniae (Walker) Assorted species Yellow night-stalking sac spider Lynx spider Jumping spider Tangle web spider Crab spider Other spider Ground beetle Transverse ladybeetle Variable ladybeetle Minute two-spotted ladybeetle Maculate ladybeetle White-collared ladybeetle Striped ladybeetle Brown ladybeetle Red and blue beetle Pirate bug Big-eyed bug Apple dimpling bug Brown smudge bug Pacific damsel bug Spined shield bug Meat ant Big-headed ant Wasps Green lacewing Tasman’s brown lacewing Crickets Assorted species Corticaria spec. Aethina concolor (Macleay) Steganopsis melanogaster Thomson – Assorted species Amrasca terraereginae (Paoli) Austroasca viridigrisea (Paoli) Cicadulina bimaculata (Evans) Oliarus lubra Kirkaldy Nysius spp. Oxycarenus luctuosus (Montrouzier) Dysdercus spp. Helicoverpa spp. Assorted species Flea beetles Minute mould beetle Hibiscus flower beetle Bent-wing fly Sciarid fly Aphids Cotton leafhopper Vegetable leafhopper Maize leafhopper Treehopper Rutherglen bug, grey cluster bug Cottonseed bug Cotton stainer, pale cotton stainer Cotton bollworm, native budworm Thrips Melyridae Anthocoridae Lygaeidae Miridae Nabidae Pentatomidae Formicidae Mostly Braconidae Chrysopidae Hemerobiidae Gryllidae Phytophagous and non-pest Coleoptera Chrysomelidae Lathridiidae Nitidulidae Diptera Lauxaniidae Sciaridae Hemiptera Aphididae Cicadellidae Cixiidae Lygaeidae Lepidoptera Thysanoptera Pyrrhocoridae Noctuidae Thripidae 1 More than 10 individuals recorded in each trial. row-pair. Plants were at the 20 –80% open boll growth stages, 124–142 DAP. This period coincided with consistently high arthropod abundance. The diurnal part of the day between 06:00 and 18:00 hours was divided into 12 1-h periods (treatments) to examine variation in beat sheet estimates of arthropod abundance over a day. Sunrise occurred at 06:05 hours (range 06:00–06:09 hours) and sunset at 18:04 hours (17:54–18:15 hours). The study was replicated over four dates on 14, 21, and 28 March and 1 April 2001, but on Temporal variation in sampling effectiveness 143 21 March no sampling was undertaken after 13:00 hours due to rainfall. Arthropods were sampled using only a beat sheet (see trial 1 for details). On each census date, the 12 hourly sampling periods were randomly assigned to a given row-pair of cotton in the study area to create a completely randomized design. At least 10 samples were collected during each period at randomly selected locations along the focal row-pair, and these samples were averaged prior to analyses (see Table 1 for details of entomophagous and phytophagous species recorded). The time taken to complete each sample (rounded to the nearest minute) at different times of the day was recorded as indicator of effort, as it takes longer to count when arthropod abundance is high. A series of repeated measures ANOVA tests (SAS MIXED procedure) was used to determine the effects of sampling period (time of day) on sample duration or the abundance of each arthropod taxon. In the tests, sampling period was applied as a fixed, repeated effect with a compound symmetry covariance structure, and replicate trial was applied as a fixed effect. Arthropod abundance but not sample duration data were rank transformed prior to analyses to improve normality and/or homogeneity of variances in ANOVA tests. Results Trial 1: time of season effects The abundance of 18 of the 24 arthropod taxa varied significantly between the sampling methods. These arthropods were mostly represented by the orders Araneae, Hemiptera, Neuroptera, and all orders combined, and generally more arthropods were recorded using beat sheet and cage clamp than visual or suction sampling methods (Tables 2 and 3, Figure 1). For example, the estimated abundance of the tangle web spider, Achaearanea veruculata (Urquhart), when pooled across all dates was 0.13 for visual, 0.21 for suction, 1.60 for beat sheet, and 2.22 per metre row for cage clamp sampling. Arthropod abundance generally increased over time as the plants grew; the estimated abundance of 18 taxa varied significantly over the sampling period. For example, the abundance of A. veruculata using all methods averaged was 0.17, 0.59, 0.57, 1.48, and 2.41 per metre row at 34, 39, 46, 53, and 60 DAP, respectively (Tables 2 and 3, Figure 1). The contrast between the different sampling methods became more pronounced over time, as indicated by significant method–date interactions for 14 of the 24 taxa. In general, visual and suction sampling was less sensitive to Table 2 Results from repeated measures ANOVA tests (significant values in bold) on abundance estimates of each entomophagous arthropod taxon using different sampling methods in unsprayed cotton on five dates between 34 and 60 days after planting. ‘I’ denotes the immature and ‘A’ the adult lifestage. Significant P-values are highlighted in bold. Analyses were not valid for immature Coccinellidae, adult Coccinella transversalis, and adult Pheidole spec. Order Taxon Lifestage Treatment effect d.f. F-value P-value Araneae A. veruculata I, A Cheiracanthium spec. I, A Oxyopes spp. I, A Assorted Salticidae I, A Assorted Thomisidae I, A Other Araneae I, A All Araneae I, A Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 63.59 20.28 3.51 5.19 8.46 0.79 56.92 8.61 4.23 4.91 2.48 1.96 8.35 2.09 1.84 6.19 11.96 4.16 52.68 19.79 4.27 <0.0001 <0.0001 0.0001 0.0044 <0.0001 0.6631 <0.0001 <0.0001 <0.0001 0.0058 0.0464 0.0323 0.0002 0.0851 0.0468 0.0017 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 144 Wade et al. Table 2 Continued. Order Taxon Lifestage Treatment effect d.f. F-value P-value Coleoptera C. inaequalis A M. frenata A Other Coccinellidae A All Coleoptera I, A D. signatus I D. signatus A G. lubra I G. lubra A N. kinbergii A O. schellembergii I O. schellembergii A Other Hemiptera I, A All Hemiptera I, A Iridomyrmex spec. A All Formicidae A Neuroptera M. tasmaniae A All orders All I, A Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date Method Date Method*date 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 3,36 4,142 12,142 2.62 2.02 1.88 1.73 10.44 1.22 1.28 0.48 0.92 3.34 7.85 1.25 19.56 87.8 6.86 5.59 4.06 1.93 5.48 13.45 2.43 5.43 5.77 1.56 1.92 4.96 1.18 3.24 3.02 1.16 4.64 2.4 1.65 11.46 4.89 4.82 7.89 35.85 4.45 1.77 0.74 0.93 2.02 0.70 0.85 28.43 9.37 6.69 38.36 41.44 3.24 0.0654 0.0951 0.0417 0.1783 <0.0001 0.2752 0.2960 0.7524 0.5319 0.0298 <0.0001 0.2550 <0.0001 <0.0001 <0.0001 0.0030 0.0038 0.0350 0.0033 <0.0001 0.0066 0.0035 0.0002 0.1090 0.1438 0.0009 0.3042 0.0331 0.0198 0.3159 0.0076 0.0526 0.0847 <0.0001 0.0010 <0.0001 0.0004 <0.0001 <0.0001 0.1698 0.5664 0.5236 0.1287 0.5929 0.6000 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.0004 Hemiptera Hymenoptera Temporal variation in sampling effectiveness 145 Table 3 Abundance estimates of each entomophagous taxon per metre row using different sampling methods in unsprayed cotton on five dates between 34 and 60 days after planting. Note that some taxa are not listed on every census date because they were not detected using either sampling method on a particular date. ‘I’ denotes the immature and ‘A’ the adult lifestage. Data are the means ± SE of 10 replicate samples Date Order Taxon Lifestage Suction Visual Beat Cage 34 Araneae A. veruculata Cheiracanthium spec. Salticidae Thomisidae Other Araneae All Araneae C. inaequalis C. transversalis Coccinellidae Other Coccinellidae All Coccinellidae D. signatus G. lubra All Hemiptera Iridomyrmex spec. A. veruculata Cheiracanthium spec. Oxyopes spp. Salticidae Thomisidae Other Araneae All Araneae C. inaequalis C. transversalis All Coccinellidae D. signatus I, A I, A I, A I, A I, A I, A A A I A I, A A A I, A A I, A I, A I, A I, A I, A I, A I, A A A I, A I A A A A I, A A A A I, A I, A I, A I, A I, A I, A I, A A A A I A I, A A I A A 0.17 ± 0.04 0.02 ± 0.01 0.02 ± 0.01 0.01 ± 0.01 0 0.21 ± 0.05 0.01 ± 0.01 0.01 ± 0.01 0.01 ± 0.01 0.01 ± 0.01 0.03 ± 0.01 0.01 ± 0.01 0.02 ± 0.01 0.03 ± 0.01 0.03 ± 0.02 0.14 ± 0.03 0.05 ± 0.01 0.11 ± 0.02 0.01 ± 0.01 0.03 ± 0.01 0 0.33 ± 0.04 0 0.01 ± 0.01 0.01 ± 0.01 0 0.08 ± 0.03 0.05 ± 0.02 0.02 ± 0.01 0.02 ± 0.01 0.16 ± 0.04 0.02 ± 0.01 0 0.04 ± 0.01 0.19 ± 0.04 0.07 ± 0.02 0.12 ± 0.04 0.01 ± 0.01 0 0.10 ± 0.02 0.48 ± 0.07 0.01 ± 0.01 0.01 ± 0.01 0.02 ± 0.01 0.01 ± 0.01 0 0.03 ± 0.01 0.20 ± 0.03 0 0 0.03 ± 0.01 0 0 0 0 0.91 ± 0.41 0.91 ± 0.41 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.19 ± 0.19 0 0.19 ± 0.19 0 0.19 ± 0.19 0.19 ± 0.19 0 0 0.38 ± 0.25 0 0 0 0 0.30 ± 0.30 0.15 ± 0.15 0.15 ± 0.15 0.15 ± 0.15 0.15 ± 0.15 0.91 ± 0.46 0.15 ± 0.15 0 0 0 0 0.15 ± 0.15 0.61 ± 0.34 0 0.30 ± 0.20 0.30 ± 0.20 0.20 ± 0.13 0.20 ± 0.13 0 0 0.10 ± 0.10 0.50 ± 0.22 0 0 0 0 0 0.10 ± 0.10 0 0.10 ± 0.10 1.00 ± 0.47 0.80 ± 0.25 0.50 ± 0.22 0 0.30 ± 0.15 0 0 1.60 ± 0.40 0 0 0 0.10 ± 0.10 0.10 ± 0.10 0.20 ± 0.13 0 0 0.40 ± 0.22 0 0 0 0.70 ± 0.26 0.50 ± 0.22 0 0.30 0 0 1.50 ± 0.43 0 0 0 0.10 ± 0.10 0.10 ± 0.10 0.20 ± 0.13 0.20 ± 0.13 0.10 ± 0.10 0.10 ± 0.10 0.10 ± 0.10 0.30 ± 0.15 0.10 ± 0.10 0 0 0.20 ± 0.13 0.60 ± 0.16 0 0 0 0 0 0 0 0 0.30 ± 0.21 1.30 ± 0.37 0.40 ± 0.31 0.10 ± 0.10 0 0.20 ± 0.13 0.10 ± 0.10 2.10 ± 0.41 0.20 ± 0.13 0 0.20 ± 0.13 0 0.10 ± 0.10 0.20 ± 0.13 0 0 0.30 ± 0.21 0.20 ± 0.13 0.10 ± 0.10 0 1.40 ± 0.60 1.00 ± 0.26 0.20 ± 0.13 0.40 ± 0.16 0.50 ± 0.22 0.30 ± 0.15 3.80 ± 1.02 0 0.10 ± 0.10 0.10 ± 0.10 0 0 0.20 ± 0.13 0.50 ± 0.17 0 0.50 ± 0.27 0.10 ± 0.10 Coleoptera Hemiptera 39 Hymenoptera Araneae Coleoptera Hemiptera Hymenoptera 46 Neuroptera Araneae Coleoptera 46 Hemiptera G. lubra N. kinbergii O. schellembergii All Hemiptera Iridomyrmex spec. Pheidole spec. M. tasmaniae A. veruculata Cheiracanthium spec. Oxyopes spp. Salticidae Thomisidae Other Araneae All Araneae C. inaequalis C. transversalis M. frenata Coccinellidae Other Coccinellidae All Coccinellidae D. signatus G. lubra N. kinbergii 146 Wade et al. Table 3 Continued. Date 53 Order Hymenoptera Araneae Coleoptera Hemiptera Taxon Lifestage Suction Visual Beat Cage O. schellembergii I A I, A I, A A I, A I, A I, A I, A I, A I, A I, A A A A I, A I A I A A I A I, A I, A A A A I, A I, A I, A I, A I, A I, A I, A A A A I A I, A I A I A I A I, A I, A A A A 0.02 ± 0.01 0.01 ± 0.01 0.01 ± 0.01 0.26 ± 0.03 0.03 ± 0.03 0.32 ± 0.03 0.13 ± 0.03 0.15 ± 0.04 0.06 ± 0.02 0.03 ± 0.02 0.14 ± 0.02 0.82 ± 0.06 0.04 ± 0.02 0.01 ± 0.01 0.04 ± 0.01 0.09 ± 0.03 0.11 ± 0.02 0.09 ± 0.02 0.03 ± 0.01 0.09 ± 0.02 0.01 ± 0.01 0.01 ± 0.01 0 0.05 ± 0.02 0.37 ± 0.04 0.01 ± 0.01 0.01 ± 0.01 0.05 ± 0.01 0.22 ± 0.04 0.06 ± 0.02 0.09 ± 0.01 0.02 ± 0.01 0.02 ± 0.01 0.16 ± 0.03 0.57 ± 0.05 0.01 ± 0.01 0.01 ± 0.01 0.01 ± 0.01 0.01 ± 0.01 0.01 ± 0.01 0.03 ± 0.01 0.45 ± 0.08 0.05 ± 0.02 0.07 ± 0.02 0.02 ± 0.01 0 0.03 ± 0.01 0.02 ± 0.01 0.62 ± 0.10 0 0.02 ± 0.02 0.04 ± 0.01 0 0 0 1.22 ± 0.59 0.15 ± 0.15 0 0.61 ± 0.25 0 0 0 0 0.61 ± 0.25 0.15 ± 0.15 0 0.15 ± 0.15 0.30 ± 0.30 0 0.76 ± 0.41 0 0.30 ± 0.20 0 0 0 0 1.06 ± 0.51 0.46 ± 0.46 0 0 0.61 ± 0.34 0.46 ± 0.23 0.15 ± 0.15 0 0 0.15 ± 0.15 1.37 ± 0.70 0 0 0.15 ± 0.15 0.15 ± 0.15 0 0.30 ± 0.20 0.61 ± 0.46 0 0 0 0 0 0 0.61 ± 0.46 0.15 ± 0.15 0 0 0 0 0 0.50 ± 0.17 0.80 ± 0.51 2.10 ± 0.41 1.10 ± 0.35 0 0.20 ± 0.13 0 0.40 ± 0.16 3.80 ± 0.51 0 0 0.50 ± 0.22 0.50 ± 0.22 0.70 ± 0.21 0.10 ± 0.10 0.50 ± 0.22 0.20 ± 0.13 0.20 ± 0.20 0 0 0 1.70 ± 0.45 0.20 ± 0.13 0 0 4.20 ± 0.73 0.90 ± 0.35 0 0.10 ± 0.10 0.20 ± 0.13 0.90 ± 0.28 6.30 ± 0.84 0.10 ± 0.10 0 0.30 ± 0.15 0.60 ± 0.40 0 1.00 ± 0.37 7.20 ± 1.34 0.50 ± 0.27 0.30 ± 0.15 0 0 0 0.10 ± 0.10 8.10 ± 1.43 1.80 ± 1.58 0.10 0.20 ± 0.13 0.80 ± 0.47 0 0 1.90 ± 0.59 0.10 ± 0.10 3.50 ± 0.43 1.50 ± 0.48 0.20 ± 0.13 0.10 ± 0.10 0.20 ± 0.20 0.60 ± 0.22 6.10 ± 0.77 0.30 ± 0.30 0 0.40 ± 0.16 0.70 ± 0.40 0.60 ± 0.22 1.40 ± 0.43 0.50 ± 0.22 0.30 ± 0.15 0.10 ± 0.10 0.10 ± 0.10 0.10 ± 0.10 0 3.10 ± 0.53 0.40 ± 0.27 0 0.10 ± 0.10 4.60 ± 0.69 1.30 ± 0.37 0.20 ± 0.13 0 0.50 ± 0.31 0.90 ± 0.28 7.50 ± 0.96 0 0.10 ± 0.10 0.40 ± 0.22 0.40 ± 0.22 0 0.90 ± 0.28 5.00 ± 1.21 0.70 ± 0.52 0.20 ± 0.13 0.10 ± 0.10 0.10 ± 0.10 0.20 ± 0.13 0 6.30 ± 1.16 0.70 ± 0.70 0.70 ± 0.60 0 Other Hemiptera All Hemiptera Iridomyrmex spec. A. veruculata Cheiracanthium spec. Oxyopes spp. Salticidae Thomisidae Other Araneae All Araneae C. inaequalis C. transversalis M. frenata All Coccinellidae D. signatus G. lubra N. kinbergii O. schellembergii Hymenoptera 60 Neuroptera Araneae Coleoptera Hemiptera Other Hemiptera All Hemiptera Iridomyrmex spec. Pheidole spec. M. tasmaniae A. veruculata Cheiracanthium spec. Oxyopes spp. Salticidae Thomisidae Other Araneae All Araneae C. inaequalis C. transversalis M. frenata Coccinellidae Other Coccinellidae All Coccinellidae D. signatus G. lubra O. schellembergii Hymenoptera Neuroptera Other Hemiptera All Hemiptera Iridomyrmex spec. Pheidole spec. M. tasmaniae Temporal variation in sampling effectiveness 147 Table 4 Results from repeated measures ANOVA tests (significant values in bold) on the relative variation between sampling periods in beat sheet estimates of arthropod abundance in unsprayed cotton. ‘I’ denotes the immature and ‘A’ the adult lifestage. Significant P-values were highlighted in bold. Analyses were not valid for immature and adult Campylomma liebknechti, adult Diomus notescens, adult Gryllidae, adult Micraspis frenata, adult Nabis kinbergii, immature Oxycarenus luctuosus, and immature Orius spec. Order Figure 1 Abundance estimates for all species of entomophagous arthropods using different sampling methods in unsprayed cotton on five dates between 34 and 60 days after planting. Data are the means + SE of 10 replicate samples. The same letters indicate no significance difference at P>0.05 by Fisher’s LSD tests in entomophagous arthropod abundance using different sampling methods over census dates. detecting changes in abundance over time compared with beat sheet and cage clamp techniques. For example, abundance estimates of A. veruculata at 34 DAP were 0, 0.17, 0.2, and 0.3 per metre row for visual, suction, beat sheet and cage clamp, respectively. However, by 60 DAP there were 0.61, 0.22, 4.2, and 4.6 per metre row for these same methods, respectively (Tables 2 and 3, Figure 1). Trial 2: time of day effects The estimated time to complete each beat sheet sample was 4.4 ± 0.1 min (mean ± SE) and did not vary significantly between sampling periods: 4.8 ± 0.1 at 06:00–07:00 hours, 4.9 ± 0.3 at 07:00 –08:00 hours, 4.4 ± 0.4 at 08:00– 09:00 hours, 4.3 ± 0.4 at 09:00–10:00 hours, 4.4 ± 0.5 at 10:00–11:00 hours, 4.5 ± 0.2 at 11:00–12:00 hours, 4.4 ± 0.2 at 12:00–13:00 hours, 4.4 ± 0.5 at 13:00– 14:00 hours, 4.3 ± 0.7 at 14:00–15:00 hours, 4.2 ± 0.4 at 15:00–16:00 hours, 4.2 ± 0.3 at 16:00–17:00 hours, and 4.0 ± 0.4 at 17:00–18:00 hours (F11,28 = 1.25, P = 0.3036). Relative abundance estimates of Rhyzobius lophanthae (Blaisdell), all Coleoptera, Formicidae, all entomophagous arthropods, Steganopis melanogaster Thomson, Sciaridae, adult Cicadellidae, and Oliarus lubra Kirkaldy varied significantly between sampling periods, while the remaining 35 taxa did not (Tables 4 and 5, Figure 2). Discussion Overall, the beat sheet was the best sampling method identified. Significantly more predators were recorded using the beat sheet and cage clamp than by using a suction machine or visual assessment. The disparity between the Taxon Lifestage F11,28 value P-value Entomophagous Araneae All Araneae Coleoptera C. inaequalis C. transversalis D. bellulus H. octomaculata H. variegata R. lophanthae Carabidae Coccinellidae All Coleoptera Hemiptera D. signatus G. lubra I, A A A A A A A A I I, A I I A N. kinbergii I O. schellembergii I Orius spec. A All Hemiptera I, A Hymenoptera Formicidae A Parasitoid wasps A Neuroptera M. signata I A M. tasmaniae I A Orthoptera Gryllidae I All orders All I, A Phytophagous and non-pest Coleoptera A. concolor Chrysomelidae Corticaria spec. Unknown spec. Diptera S. melanogaster Sciaridae Hemiptera Aphididae Cicadellidae A A A A A A I, A I A Dysdercus spp. I A Nysius spp. I A O. lubra A O. luctuosus A Lepidoptera Helicoverpa spp. I Thysanoptera Thripidae I, A All orders All I, A 1.66 1.71 1.35 0.71 1.25 1.28 2.22 1.74 1.55 2.97 1.85 1.12 0.43 1.35 0.63 2.00 0.40 2.25 1.45 1.90 1.27 1.78 1.10 0.77 2.74 0.1351 0.1232 0.2485 0.7171 0.3028 0.2845 0.0437 0.1149 0.1696 0.0097 0.0932 0.3857 0.9271 0.2488 0.7897 0.0683 0.9459 0.0413 0.2055 0.0832 0.2901 0.1061 0.3972 0.6668 0.0155 0.85 1.42 2.10 2.10 3.72 2.24 1.12 1.17 2.83 0.96 0.60 1.01 1.69 2.50 0.75 0.56 0.79 0.80 0.5916 0.2177 0.0550 0.0550 0.0024 0.0417 0.3806 0.3494 0.0127 0.4988 0.8165 0.4624 0.1284 0.0248 0.6845 0.8437 0.6519 0.6361 Taxon Entomophagous Araneae All Araneae Coleoptera C. inaequalis C. transversalis D. bellulus D. notescens H. octomaculata H. variegata M. frenata R. lophanthae Carabidae Coccinellidae All Coleoptera Hemiptera C. liebknechti Lifestage 06:00 –07:00 07:00 –08:00 08:00 –09:00 09:00–10:00 10:00–11:00 11:00–12:00 12:00–13:00 13:00–14:00 14:00–15:00 15:00–16:00 16:00–17:00 17:00–18:00 I, A A A A A A A A A A I I, A I A D. signatus I G. lubra I A N. kinbergii I A O. schellembergii I Orius spec. I A All Hemiptera I, A Hymenoptera Formicidae A Parasitoid wasps A Neuroptera M. signata I A M. tasmaniae I A Orthoptera Gryllidae I A Phytophagous and non-pest Coleoptera A. concolor A Chrysomelidae A Corticaria spec. A Unknown beetle A Diptera S. melanogaster A Sciaridae A 5.15 ± 0.47 0.45 ± 0.21 0.10 ± 0.10 0.05 ± 0.03 0.08 ± 0.08 4.25 ± 2.58 0.30 ± 0.14 0.03 ± 0.03 0.10 ± 0.06 0.08 ± 0.05 2.38 ± 1.45 7.88 ± 2.57 0.08 ± 0.08 0.10 ± 0.04 0.28 ± 0.19 0.25 ± 0.10 0.30 ± 0.11 0.23 ± 0.08 0.15 ± 0.09 0.05 ± 0.03 0 0.08 ± 0.05 1.63 ± 0.21 0.40 ± 0.24 0.73 ± 0.41 0.08 ± 0.08 0.15 ± 0.09 0.13 ± 0.06 0.40 ± 0.23 0.23 ± 0.10 0.10 ± 0.06 5.60 ± 0.44 0.63 ± 0.27 0.18 ± 0.10 0.13 ± 0.06 0.10 ± 0.04 1.53 ± 0.27 0.18 ± 0.08 0.05 ± 0.05 0.13 ± 0.05 0.13 ± 0.05 2.70 ± 0.98 5.78 ± 1.15 0.05 ± 0.05 0.05 ± 0.05 0.23 ± 0.09 0.45 ± 0.13 0.23 ± 0.10 0.45 ± 0.13 0.15 ± 0.05 0.08 ± 0.05 0.03 ± 0.03 0.23 ± 0.13 1.95 ± 0.25 0.35 ± 0.23 0.60 ± 0.25 0.03 ± 0.03 0.05 ± 0.03 0.38 ± 0.23 0.40 ± 0.25 0.15 ± 0.12 0.03 ± 0.03 4.98 ± 0.83 0.43 ± 0.14 0 0.08 ± 0.05 0 1.40 ± 0.49 0.18 ± 0.03 0.05 ± 0.03 0.03 ± 0.03 0 4.35 ± 2.49 6.53 ± 2.25 0.05 ± 0.05 0.03 ± 0.03 0.20 ± 0.11 0.35 ± 0.12 0.28 ± 0.09 0.35 ± 0.06 0.03 ± 0.03 0.15 ± 0.09 0.08 ± 0.05 0.25 ± 0.10 1.75 ± 0.39 1.88 ± 1.09 0.45 ± 0.17 0.03 ± 0.03 0.03 ± 0.03 0.48 ± 0.03 0.20 ± 0.14 0.05 ± 0.03 0.03 ± 0.03 4.80 ± 0.72 0.28 ± 0.08 0.03 ± 0.03 0.10 ± 0.04 0.20 ± 0.14 1.55 ± 0.49 0.13 ± 0.05 0 0.13 ± 0.03 0 4.10 ± 2.78 6.58 ± 3.16 0.13 ± 0.13 0 0.03 ± 0.03 0.45 ± 0.09 0.33 ± 0.10 0.18 ± 0.08 0.05 ± 0.03 0.03 ± 0.03 0.15 ± 0.09 0.25 ± 0.16 1.58 ± 0.31 1.98 ± 0.74 0.35 ± 0.19 0.13 ± 0.05 0 0.38 ± 0.21 0.13 ± 0.09 0.13 ± 0.03 0 5.83 ± 0.98 0.45 ± 0.36 0.05 ± 0.03 0.05 ± 0.05 0 1.33 ± 0.62 0.10 ± 0.04 0.03 ± 0.03 0.05 ± 0.03 0.03 ± 0.03 3.15 ± 2.33 5.23 ± 2.48 0.08 ± 0.08 0.05 ± 0.05 0.28 ± 0.19 0.68 ± 0.28 0.18 ± 0.09 0.18 ± 0.05 0.10 ± 0.00 0.03 ± 0.03 0.08 ± 0.03 0.03 ± 0.03 1.65 ± 0.50 1.63 ± 0.93 0.43 ± 0.18 0.13 ± 0.03 0.03 ± 0.03 0.45 ± 0.18 0.15 ± 0.09 0.20 ± 0.04 0.03 ± 0.03 4.83 ± 0.53 0.33 ± 0.20 0.03 ± 0.03 0.03 ± 0.03 0.03 ± 0.03 1.15 ± 0.29 0.13 ± 0.13 0.10 ± 0.07 0.05 ± 0.03 0 3.78 ± 2.25 5.63 ± 2.32 0.13 ± 0.09 0.05 ± 0.05 0.05 ± 0.05 0.90 ± 0.23 0.30 ± 0.11 0.35 ± 0.20 0.10 ± 0.07 0.15 ± 0.06 0.08 ± 0.03 0.18 ± 0.09 2.28 ± 0.39 1.35 ± 0.70 0.18 ± 0.08 0 0.05 ± 0.03 0.18 ± 0.12 0.08 ± 0.03 0.13 ± 0.03 0.03 ± 0.03 4.78 ± 0.29 0.43 ± 0.22 0.10 ± 0.10 0.05 ± 0.03 0.03 ± 0.03 1.00 ± 0.18 0.08 ± 0.05 0.03 ± 0.03 0.13 ± 0.03 0 1.90 ± 1.05 3.85 ± 1.55 0.05 ± 0.05 0.05 ± 0.05 0.05 ± 0.05 0.68 ± 0.22 0.28 ± 0.05 0.35 ± 0.15 0.18 ± 0.06 0.08 ± 0.05 0.10 ± 0.07 0.25 ± 0.17 2.05 ± 0.41 1.00 ± 0.23 0.38 ± 0.21 0.05 ± 0.03 0.05 ± 0.03 0.18 ± 0.06 0.15 ± 0.06 0.10 ± 0.04 0.05 ± 0.03 4.20 ± 0.47 0.10 ± 0.06 0 0.07 ± 0.03 0 1.23 ± 0.56 0 0 0 0.07 ± 0.07 1.47 ± 0.72 2.93 ± 1.33 0.03 ± 0.03 0 0.10 ± 0.06 0.57 ± 0.22 0.27 ± 0.03 0.17 ± 0.09 0.03 ± 0.03 0.07 ± 0.03 0.07 ± 0.07 0.07 ± 0.03 1.40 ± 0.38 1.33 ± 0.66 0.17 ± 0.03 0 0 0.07 ± 0.07 0.07 ± 0.07 0.10 ± 0.06 0 4.97 ± 0.48 0.17 ± 0.07 0.10 ± 0.06 0.03 ± 0.03 0.10 ± 0.06 1.07 ± 0.62 0.10 ± 0.06 0.03 ± 0.03 0.03 ± 0.03 0.03 ± 0.03 1.37 ± 0.70 3.03 ± 0.79 0.07 ± 0.07 0.03 ± 0.03 0.53 ± 0.15 0.67 ± 0.03 0.33 ± 0.15 0.13 ± 0.07 0 0.13 ± 0.09 0.03 ± 0.03 0 1.93 ± 0.18 2.47 ± 0.50 0.10 ± 0.06 0.07 ± 0.03 0 0.20 ± 0.20 0.07 ± 0.03 0.20 ± 0.06 0.03 ± 0.03 5.57 ± 0.33 0.17 ± 0.07 0.13 ± 0.13 0 0.03 ± 0.03 0.93 ± 0.20 0.17 ± 0.12 0.03 ± 0.03 0 0 1.47 ± 0.70 2.97 ± 0.70 0.03 ± 0.03 0 0.23 ± 0.03 0.63 ± 0.09 0.23 ± 0.09 0.17 ± 0.03 0.03 ± 0.03 0.03 ± 0.03 0.10 ± 0.00 0 1.50 ± 0.23 1.77 ± 0.78 0.17 ± 0.09 0.03 ± 0.03 0.03 ± 0.03 0.07 ± 0.07 0.03 ± 0.03 0.20 ± 0.06 0 5.87 ± 1.42 0.20 ± 0.06 0.03 ± 0.03 0.17 ± 0.17 0 0.90 ± 0.47 0.13 ± 0.09 0 0 0 0.93 ± 0.35 2.40 ± 0.40 0 0.03 ± 0.03 0.17 ± 0.09 0.53 ± 0.07 0.27 ± 0.15 0.27 ± 0.12 0.07 ± 0.03 0.10 ± 0.06 0.07 ± 0.07 0.17 ± 0.17 1.77 ± 0.44 2.07 ± 1.29 0.40 ± 0.21 0 0 0.10 ± 0.06 0 0.17 ± 0.17 0.03 ± 0.03 7.42 ± 0.93 0.25 ± 0.06 0.08 ± 0.08 0.02 ± 0.02 0.05 ± 0.05 1.50 ± 0.37 0.10 ± 0.10 0 0.10 ± 0.05 0.03 ± 0.03 0.80 ± 0.28 2.97 ± 0.41 0.05 ± 0.05 0.03 ± 0.03 0.33 ± 0.27 0.49 ± 0.03 0.16 ± 0.03 0.49 ± 0.06 0.17 ± 0.05 0.04 ± 0.02 0.10 ± 0.06 0.09 ± 0.02 2.01 ± 0.45 1.30 ± 0.77 0.49 ± 0.07 0.05 ± 0.02 0 0.24 ± 0.09 0.08 ± 0.08 0.09 ± 0.05 0 0.33 ± 0.23 0.80 ± 0.51 1.40 ± 0.41 0.03 ± 0.03 0.28 ± 0.15 0.40 ± 0.21 0.55 ± 0.27 1.00 ± 0.77 1.78 ± 0.19 0.03 ± 0.03 0.25 ± 0.06 0.33 ± 0.29 0.43 ± 0.14 0.40 ± 0.16 1.18 ± 0.45 0.15 ± 0.12 0.03 ± 0.03 0.10 ± 0.10 0.18 ± 0.14 0.55 ± 0.29 1.13 ± 0.22 0.20 ± 0.12 0.03 ± 0.03 0.20 ± 0.09 0.15 ± 0.10 0.33 ± 0.13 1.15 ± 0.43 0.15 ± 0.12 0 0.03 ± 0.03 0.43 ± 0.21 0.18 ± 0.14 1.33 ± 0.29 0.15 ± 0.12 0 0 0.40 ± 0.23 0.38 ± 0.14 1.28 ± 0.38 0.33 ± 0.21 0.05 ± 0.03 0.05 ± 0.00 0.17 ± 0.17 0.37 ± 0.09 1.50 ± 0.23 0.43 ± 0.22 0.03 ± 0.03 0.03 ± 0.03 0.53 ± 0.35 0.10 ± 0.06 1.27 ± 0.22 0.17 ± 0.09 0 0.03 ± 0.03 0.03 ± 0.03 0.13 ± 0.03 1.40 ± 0.12 0.30 ± 0.15 0 0.07 ± 0.07 0.57 ± 0.32 0.37 ± 0.22 1.73 ± 0.44 0.50 ± 0.32 0.03 ± 0.03 0.07 ± 0.03 0.20 ± 0.17 0.31 ± 0.13 2.52 ± 0.31 0.71 ± 0.36 0 0.05 ± 0.02 Wade et al. Order 148 Table 5 Relative variation between sampling periods in beat sheet estimates of arthropod abundance in unsprayed cotton. ‘I’ denotes the immature and ‘A’ the adult lifestage. Data are the means ± SE of four replicate trials 1.05 ± 1.02 9.29 ± 2.13 3.33 ± 0.39 0.44 ± 0.22 0.05 ± 0.05 0 0.05 ± 0.02 0.07 ± 0.04 0 1.01 ± 0.30 0.05 ± 0.05 0.62 ± 0.62 2.30 ± 2.30 9.03 ± 2.99 3.27 ± 0.70 2.30 ± 1.33 0.17 ± 0.12 0.13 ± 0.09 0 0.03 ± 0.03 0.03 ± 0.03 0.73 ± 0.24 0.03 ± 0.03 0.33 ± 0.33 3.03 ± 3.03 10.50 ± 3.89 2.87 ± 0.91 1.93 ± 1.74 0.17 ± 0.09 0 0 0.10 ± 0.06 0 1.93 ± 0.88 0.03 ± 0.03 0.17 ± 0.17 1.83 ± 1.83 10.73 ± 3.97 3.17 ± 0.41 0.80 ± 0.47 0.17 ± 0.09 0.33 ± 0.33 0.03 ± 0.03 0 0 1.17 ± 0.59 0.10 ± 0.06 0.10 ± 0.10 6.80 ± 6.75 8.60 ± 2.05 3.57 ± 0.42 0.10 ± 0.06 0.07 ± 0.07 0.03 ± 0.03 0 0.03 ± 0.03 0.03 ± 0.03 1.03 ± 0.43 0.07 ± 0.07 0.03 ± 0.03 3.30 ± 2.94 12.70 ± 5.19 2.93 ± 1.02 2.08 ± 1.72 0.13 ± 0.08 0.08 ± 0.08 0.03 ± 0.03 0.08 ± 0.03 0 1.30 ± 0.53 0 0.03 ± 0.03 5.38 ± 3.12 17.33 ± 4.94 3.53 ± 0.65 1.88 ± 0.93 0.10 ± 0.04 0.05 ± 0.05 0 0 0 0.93 ± 0.28 0.08 ± 0.05 0.15 ± 0.15 2.93 ± 2.25 13.43 ± 5.23 2.60 ± 0.17 1.08 ± 0.30 0.18 ± 0.18 0 0 0.03 ± 0.03 0 0.53 ± 0.15 0.15 ± 0.12 0.13 ± 0.13 3.28 ± 2.18 12.80 ± 4.15 3.73 ± 1.04 0.13 ± 0.06 0.15 ± 0.06 0.10 ± 0.10 0.05 ± 0.05 0.08 ± 0.03 0.05 ± 0.05 0.80 ± 0.23 0.05 ± 0.03 0.35 ± 0.35 2.45 ± 1.70 18.88 ± 9.56 5.20 ± 0.63 1.33 ± 1.29 0.23 ± 0.08 0 0.08 ± 0.03 0.10 ± 0.04 0 0.50 ± 0.08 0.05 ± 0.03 0.25 ± 0.25 1.38 ± 0.80 15.10 ± 6.58 8.43 ± 3.29 0.58 ± 0.51 0.38 ± 0.17 0.05 ± 0.03 0 0.10 ± 0.04 0.35 ± 0.35 0.80 ± 0.27 0.05 ± 0.05 0.20 ± 0.17 1.88 ± 1.68 12.30 ± 5.25 10.45 ± 4.35 0.03 ± 0.03 0.18 ± 0.09 0 0.08 ± 0.05 0.30 ± 0.12 0 1.20 ± 0.64 0.08 ± 0.08 0.23 ± 0.14 I, A I A I A Nysius spp. I A O. lubra A O. luctuosus I A Lepidoptera Helicoverpa spp. I Thysanoptera Thripidae I, A Taxon Aphididae Cicadellidae Cicadellidae Dysdercus spp. Order Hemiptera Table 5 Continued. Lifestage 06:00 –07:00 07:00 –08:00 08:00 –09:00 09:00–10:00 10:00–11:00 11:00–12:00 12:00–13:00 13:00–14:00 14:00–15:00 15:00–16:00 16:00–17:00 17:00–18:00 Temporal variation in sampling effectiveness 149 Figure 2 Relative variation between 1-h sampling periods in beat sheet estimates of the abundance of all species of entomophagous (triangles) or phytophagous (squares) arthropods in unsprayed cotton. Data are the means ± SE of four replicate trials. The same letters indicate no significance difference at P>0.05 by Fisher’s LSD tests in entomophagous arthropod abundance at sampling periods. No significant differences were detected among phytophagous arthropods over a day. sampling methods became pronounced as the growing season progressed. Beat sheet samples were easy to complete by one person, while the cage clamp was cumbersome and required two persons to effectively operate it. Based on factors such as estimates of abundance, precision, and cost, the beat sheet was found to be the most suitable method for sampling the majority of predators and pests in agricultural crops such as cotton, snap bean, sorghum, and soybean in nine of the 18, or 50% of articles reviewed (Shepard et al., 1974a; Young & Tugwell, 1975; Bechinski & Pedigo, 1982; Adams et al., 1984; Studebaker et al., 1991; Michels & Behle, 1992; Snodgrass, 1993; Smith & Stewart, 1999; Kharboutli & Allen, 2000), while the remainder found it less suitable (Rudd & Jenson, 1977; Nuessly & Sterling, 1984; Deighan et al., 1985; Fleischer et al., 1985; Sparks & Boethel, 1987; Kharboutli & Mack, 1993; Knutson & Wilson, 1999; McLeod, 2000; Ludy & Lang, 2004). Several papers have concluded that no single technique provided the ‘best’ estimate for all lifestages of every species (Shepard et al., 1974a; Garcia et al., 1982; Kharboutli & Mack, 1993). For example, Frankliniella occidentalis (Pergande) nymphs were found to be most effectively sampled using berlese funnels and adults using visual sampling (Garcia et al., 1982). Therefore, it is perhaps unrealistic of pest management advisors and growers to seek a single ‘best’ sampling method for all occasions and of researchers to try to meet this expectation. Instead, samplers should adopt 150 Wade et al. two or more complementary sampling methods and it is suggested that one of these should be the beat sheet. There was a critical flaw in the conclusions of many of the papers we reviewed that hinders future meta-analyses of this literature. The flaw relates to a failure to express arthropod abundance on a common basis. For example, Kharboutli & Allen (2000) compared the number of arthropods collected per ‘sample’ from 1.8 m row of cotton using a beat sheet, 3.8 m of row with a sweep-net (10 passes with a 38-cm diameter net), and 12.2 m of row with a suction sampler. Similarly, Shepard et al. (1974a) compared the number of arthropods collected per ‘sample’ from 1.2 m row of soybean with a beat sheet, 15.2 m of row with a sweep-net (20 passes across two rows with a 38-cm diameter net), and 9.1 m of row with a D-Vac suction sampler. Although such comparisons of sampling techniques based on numbers per ‘sample’ are statistically valid, they do not provide biologically valid comparisons of sampling method effectiveness. Nonetheless, several studies have partly (e.g., Fleischer et al., 1985) or fully standardized their data, such as the number collected per metre of row (e.g., Nuessly & Sterling, 1984; Ludy & Lang, 2004; this study). The influences of the time of growing season and time of day are two temporal constraints that may account for a large component of the observed variation between studies of otherwise similar design. We originally thought that the abundance of entomophagous arthropods in cotton crops would decline from mid- to late summer onwards, as reported by Smith et al. (1976), Pyke et al. (1980), and Mensah (1999). This decline was attributed to widespread insecticide use and/or arthropod phenology, as numbers waned even in unsprayed plots. Nonetheless, some studies have recorded the highest abundance of entomophagous arthropods in unsprayed and ‘IPM’ cotton plots late-season in autumn months (Bishop & Blood, 1981; Nuessly & Sterling, 1984; Mansfield et al., 2006; MR Wade, unpubl.). This anomaly could well be explained by lowered sampling efficiency when the season progressed and the plant canopy became larger. Visual or vacuum sampling was used without any correction for this effect in all studies that reported a late-season decline, while those studies that reported a late-season peak in abundance used beat sheet and/or absolute sampling techniques. Indeed, our study provides evidence to support this hypothesis and more generally, highlights the superior effectiveness of beat sheet and cage methods compared with visual and suction methods. Furthermore, several studies have concluded that the effectiveness of sweep-net, suction, and visual sampling methods declined later in the growing season when the plant canopy was large and hence less of it was sampled (Shepard et al., 1974a; Smith et al., 1976; Byerly et al., 1978; Wilson & Gutierrez, 1980; Garcia et al., 1982; Snodgrass, 1993). In any case, changes in the proportion of arthropods collected using various methods are likely to be affected by the absolute number of arthropods present at a given moment. It has been suggested that sampling effectiveness would drop at higher densities because more arthropods escape before being counted (Deighan et al., 1985). In our study, beat sheet but not suction or visual sampling collected a similar number of most arthropods relative to the absolute sampling cage clamp method. An inconsistency in the effectiveness of sampling methods as abundance changes seasonally (i.e., an interaction between sampling method and date) has important implications for using arthropod counts in pest management decision-making (e.g., when using the predator-to-pest ratio, Mensah, 2002); visual and suction assessments will likely grossly underestimate the numbers occurring on plants when arthropod abundance is high, which may coincide with late in the growing season, but should be suitable when abundance is low. This means a more complicated correction factor will need to be applied to data collected because the relationship between the absolute and relative sampling methods is not constant over the growing season (e.g., Fleischer et al., 1985; Snodgrass, 1993). However, if the estimates from relative sampling methods happened to be consistently proportional to absolute sampling, then it would be possible to rely on them season long, but this was only the case for beat sheet sampling. Artificial infestation techniques may help untangle the effects of variation between sampling methods, plant height, or volume and time of season (Young & Tugwell, 1975; Snodgrass, 1993). The influence of time of day is a second temporal constraint that may account for a large component of the observed variation between studies of sampling method effectiveness. A recommendation exists in sampling protocols to suspend sampling after 10:00 or 11:00 hours when air temperature exceeds 25–30 °C in an attempt to minimize the escape of insects that readily take flight or drop to the ground when disturbed (González et al., 1977; Garcia et al., 1982). Restrictions placed on sampling when condensation (dew) is present have been advised (i.e., between approximately 18:00 –08:00 hours) (Sevacherian & Stern, 1972; González et al., 1977; Garcia et al., 1982). However, these recommendations appear to be based on anecdotal evidence. In our study the time required to sample all arthropods at different times of the day was not significantly different, and beat sheet estimates of the abundance of the majority of taxa (35 of the 43, or 81%) were no different over a day. The response was wide-ranging for the remaining eight taxa that did vary significantly over a day. For instance, Formicidae was high after 08:00 hours, R. lophanthae was high between 06:00 and 13:00 hours and Temporal variation in sampling effectiveness 151 17:00 and 18:00 hours, and O. lubra was highest between 06:00 and 07:00 hours. Thus practitioners should indeed be mindful of time of day effects for each taxon studied, but for the majority of taxa there appear to be no effects. This recommendation however, only pertains to beat sheet sampling between 06:00 and 18:00 hours and differences may exist when using alternative sampling methods or when sampling at night. It is likely that sampling methods that bias collection to the upper part of the canopy, such as suction sampling, will be more sensitive to time of day effects (Rancourt et al., 2000). Diel variation in estimates of arthropod abundance has been generally attributed to altered arthropod activity and distribution, however, this linkage has rarely been confirmed. Shepard et al. (1974b) used direct observations of the diel variation in the distribution of big-eyed bugs (Geocoris spp.) within the plant canopy to confirm beat sheet sampling estimates, but no mention was made of altered bug activity or plant structure location that may influence abundance estimates. For example, visual estimates of N. americoferus abundance at different times of the day are associated with bug feeding activity (Braman & Yeargan, 1989). Walking activity was directly correlated with visual abundance estimates of a coccinellid beetle, Coccinella trifasciata L., at different temperatures (sensu lato time of day) (Frazer & Gilbert, 1976). González et al. (1977) predicted that arthropods located under bracts of squares and inside flowers were less likely to be sampled. However, according to Snodgrass (1993), beat sheet and sweep-net estimates of the densities of tarnished plant bug, Lygus lineolaris (Palisot de Beauvois), were not influenced by their distribution (i.e., on leaves, terminals, or inside bracts of squares). Future manipulative studies may utilize an artificial infestation technique to better understand the relationship between time of day, arthropod behaviour, and distribution and beat sheet estimates; this would involve placement of a known number of arthropods at various plant locations followed by beat sheet sampling at different times of the day (e.g., Marston et al., 1979; Snodgrass, 1993). In conclusion, our results provide evidence to support day-long and season-long use of beat sheet sampling for counting numerous arthropods in agricultural crops such as cotton. Certainly, those wanting to use this method should be prepared for potential fatigue, tedium, and exposure to high humidity and ant and mosquito attack (Nuessly & Sterling, 1984; Drees & Rice, 1985; Knutson & Wilson, 1999, personal experience). Our preliminary findings on the effectiveness of the beat sheet have been incorporated into IPM guidelines for the Australian cotton industry (see Deutscher et al., 2005) and have been adopted by many pest management advisors and researchers (e.g., Pearce et al., 2004; Murray et al., 2005; Mansfield et al., 2006). Future research should develop robust correction factors to adjust arthropod count data derived from various methods by repeating this study across numerous fields, over several growing seasons, and with numerous persons involved (to test for observer bias) (see Ruesink & Hayes, 1973; Fleischer et al., 1985; Schotzko & O’Keeffe, 1989). Without this additional research it is not possible to generalize about the effect of temporal variation in arthropod sampling effectiveness. Finally, action thresholds for key predators need to be better refined to facilitate the use of predator information in pest management decisionmaking. A prerequisite to establishment of these thresholds is the development of a reliable and efficient sampling method for key entomophagous and phytophagous arthropods (Sparks & Boethel, 1987), which has been met by this study. Acknowledgements We gratefully acknowledge St. John and Edwina Kent (‘Coondarra’) and Dave Schofield (DPIF Gatton) for maintaining the trial sites, Allan Lisle (UQ Gatton) for statistical advice, and the Australian Cotton Research and Development Corporation for partly funding the research (projects DAQ96C and UQ29C). References Adams GD, Foley DH & Pyke BA (1984) Notes on the pest status of and sampling methods for sap-sucking bugs in cotton. 1984 Australian Cotton Growers Research Conference, 5 – 6 December 1984, pp. 160 –166. Australian Cotton Growers’ Research Association, Toowoomba, Queensland, Australia. Bechinski EJ & Pedigo LP (1982) Evaluation of methods for sampling predatory arthropods in soybeans. Environmental Entomology 11: 756 – 761. Benedek P, Erdelyi C & Fesus I (1972) General aspects of diel vertical movements of some arthropods in flowering lucerne stands and conclusions for the integrated pest control of lucerne. Acta Phytopathologica Academiae Scientiarum Hungarica 7: 235 – 249. Binns MR, Nyrop JP & van der Werf W (2000) Sampling and Monitoring in Crop Protection. CABI Publishing, Wallingford, UK. Bishop AL & Blood PRB (1981) Interactions between natural populations of spiders and pests in cotton and their importance to cotton production in southeastern Queensland. General and Applied Entomology 13: 98 –104. Braman SK & Yeargan KV (1989) Intraplant distribution of three Nabis species (Hemiptera: Nabidae), and impact of N. roseipennis on green cloverworm populations in soybean. Environmental Entomology 18: 240 – 244. Browde JA, Pedigo LP, Degooyer TA, Higley LG, Wintersteen WK & Zeiss MR (1992) Comparison of sampling techniques for 152 Wade et al. grasshoppers (Orthoptera: Acrididae) in soybean. Journal of Economic Entomology 85: 2270 – 2274. Byerly KF, Gutierrez AP, Jones RE & Luck RF (1978) A comparison of sampling methods for some arthropod populations in cotton. Hilgardia 46: 257 – 282. Conover WJ & Iman RL (1981) Rank transformations as a bridge between parametric and nonparametric statistics. American Statistician 35: 124 –129. Deighan J, McPherson RM & Ravlin FW (1985) Comparisons of sweep-net and ground-cloth sampling methods for estimating arthropod densities in different soybean cropping systems. Journal of Economic Entomology 78: 208 – 212. Dempster JP (1957) The population dynamics of the Moroccan locust (Dociostaurus maroccanus Thunberg) in Cyprus. AntiLocust Bulletin 27: 1– 60. Deutscher SA, Wilson LJ & Mensah RK (2005) Integrated Pest Management Guidelines for Cotton Production Systems in Australia, 2nd edn. Australian Cotton Cooperative Research Centre and Cotton Research and Development Corporation, Narrabri, NSW, Australia. Drees BM & Rice ME (1985) The vertical beat sheet: a new device for sampling soybean insects. Journal of Economic Entomology 78: 1507 –1510. Dumas BA, Boyer WP & Whitcomb WH (1962) Effect of time of day on surveys of predaceous insects in field crops. The Florida Entomologist 45: 121–128. Dumas BA, Boyer WP & Whitcomb WH (1964) Effect of various factors on surveys of predaceous insects in soybeans. Journal of the Kansas Entomological Society 37: 192 – 201. Everitt BS (1995) The analysis of repeated-measures – a practical review with examples. Statistician 44: 113 –135. Fewkes DW (1961) Diel vertical movements in some grassland Nabidae (Heteroptera). Entomologist’s Monthly Magazine 97: 128 –130. Fleischer SJ & Allen WA (1982) Field counting efficiency of sweep-net samples of adult potato leafhoppers (Homoptera: Cicadellidae) in alfalfa. Journal of Economic Entomology 75: 837 – 840. Fleischer SJ, Gaylor MJ & Edelson JV (1985) Estimating absolute abundance from relative sampling of Lygus lineolaris (Heteroptera: Miridae) and selected predators in early to mid-season cotton. Environmental Entomology 14: 709 – 717. Frazer BD & Gilbert N (1976) Coccinellids and aphids. Journal of the Entomological Society of British Columbia 73: 33 – 56. Garcia A, González D & Leigh TF (1982) Three methods for sampling arthropod numbers on California cotton. Environmental Entomology 11: 565 – 572. González D, Ramsey DA, Leigh TF, Ekbom BS & van den Bosch R (1977) A comparison of vacuum and whole-plant methods for sampling predaceous arthropods on cotton. Environmental Entomology 6: 750 – 760. Kharboutli MS & Allen CT (2000) Comparison of sampling techniques for tarnished plant bug and predaceous arthropods. Proceedings of the 2000 Cotton Research Meeting and Summaries of Cotton Research in Progress, pp. 167 –171. Arkansas Agricultural Experiment Station, Division of Agriculture, University of Arkansas, Fayetteville, AR, USA. Kharboutli MS & Mack TP (1993) Comparison of three methods for sampling arthropod pests and their natural enemies in peanut fields. Journal of Economic Entomology 86: 1802 –1810. Knutson AE & Wilson WT (1999) The beat bucket: a rapid, reliable method for sampling predatory insects and spiders in cotton. Proceedings of the Beltwide Cotton Conference, pp. 1120 –1126. National Cotton Council, Memphis, TN, USA. Leathwick DM & Winterbourn MJ (1984) Arthropod predation on aphids in a lucerne crop. New Zealand Entomologist 8: 75 – 80. Ludy C & Lang A (2004) How to catch foliage-dwelling spiders (Araneae) in maize fields and their margins: a comparison of two sampling methods. Journal of Applied Entomology 128: 501– 509. Mansfield S, Dillon ML & Whitehouse MEA (2006) Are arthropod communities in cotton really disrupted? An assessment of insecticide regimes and evaluation of the beneficial disruption index. Agriculture, Ecosystems and Environment 113: 326– 335. Marston NL, Dickerson WA, Ponder WW & Booth GD (1979) Calibration ratios for sampling soybean Lepidoptera: effect of larval species, larval size, plant growth stage and individual sampler. Journal of Economic Entomology 72: 110 –114. McLeod PJ (2000) Comparison of insect sampling techniques in snap bean. Journal of Vegetable Crop Production 6: 37 – 43. Mensah RK (1999) Habitat diversity: implications for the conservation and use of predatory insects of Helicoverpa spp. in cotton systems in Australia. International Journal of Pest Management 45: 91–100. Mensah RK (2002) Development of an integrated pest management programme for cotton. Part 1: Establishing and utilizing natural enemies. International Journal of Pest Management 48: 87 – 94. Michels GJ Jr & Behle RW (1992) Evaluation of sampling methods for lady beetles (Coleoptera: Coccinellidae) in grain sorghum. Journal of Economic Entomology 85: 2251– 2257. Murray DAH, Lloyd RJ & Hopkinson JE (2005) Efficacy of new insecticides for management of Helicoverpa spp. (Lepidoptera: Noctuidae) in Australian grain crops. Australian Journal of Entomology 44: 62 – 67. Nuessly GS & Sterling WL (1984) Comparison of D-Vac® and modified drop cloth methods for sampling arthropods in cotton. Southwestern Entomologist 9: 95 –103. Pearce S, Hebron WM, Raven RJ, Zalucki MP & Hassan E (2004) Spider fauna of soybean crops in south-east Queensland and their potential as predators of Helicoverpa spp. (Lepidoptera: Noctuidae). Australian Journal of Entomology 43: 57 – 65. Pyke B, Sterling W & Hartstack A (1980) Beat and shake bucket sampling of cotton terminals for cotton fleahoppers, other pests and predators. Environmental Entomology 9: 572 – 576. Rancourt B, Vincent C & De Oliveira D (2000) Circadian activity of Lygus lineolaris (Hemiptera: Miridae) and effectiveness of sampling techniques in strawberry fields. Journal of Economic Entomology 93: 1160 –1166. Rudd WG & Jenson RL (1977) Sweep net and ground sampling for insects in soybeans. Journal of Economic Entomology 70: 301– 304. Temporal variation in sampling effectiveness 153 Ruesink WG & Hayes DL (1973) Sweepnet sampling for the cereal leaf beetle, Oulema melanopus. Environmental Entomology 2: 161–172. Schotzko DJ & O’Keeffe LE (1989) Comparison of sweep net, D-Vac, and absolute sampling, and diel variation of sweep net sampling estimates in lentils for pea aphid (Homoptera: Aphididae), nabids (Hemiptera: Nabidae), lady beetles (Coleoptera: Coccinellidae), and lacewings (Neuroptera: Chrysopidae). Journal of Economic Entomology 82: 491– 506. Sevacherian V & Stern VM (1972) Sequential sampling plans for Lygus bugs in California cotton fields. Environmental Entomology 1: 704 – 710. Shepard M, Garner GR & Turnipseed SG (1974a) A comparison of three sampling methods for arthropods in soybeans. Environmental Entomology 3: 227 – 232. Shepard M, Garner GR & Turnipseed SG (1974b) Dispersal of Geocoris spp. in soybeans. Journal of the Georgia Entomological Society 9: 120 –126. Smith JD & Stewart SD (1999) Comparison between drop-cloth and suction sampling in cotton during 1998. Proceedings of the Beltwide Cotton Conference, pp. 1212 –1214. National Cotton Council, Memphis, TN, USA. Smith JW, Stadelbacher EA & Gantt CW (1976) A comparison of techniques for sampling beneficial arthropod populations associated with cotton. Environmental Entomology 5: 435– 444. Snodgrass GL (1993) Estimating absolute abundance of nymphs of Lygus lineolaris (Heteroptera: Miridae) in cotton using drop cloth and sweep-net sampling methods. Journal of Economic Entomology 86: 1116 –1123. Southwood TRE, Jepson WF & van Emden HF (1961) Studies on the behaviour of Oscinella frit L. (Diptera) adults of the panicle generation. Entomologia Experimentalis et Applicata 4: 196 – 210. Sparks AN Jr & Boethel DJ (1987) Evaluation of sampling techniques and development of sequential sampling plans for three cornered alfalfa hoppers (Homoptera: Membracidae) on soybeans. Journal of Economic Entomology 80: 369– 375. Studebaker GE, Spurgeon DW & Mueller AJ (1991) Calibration of ground cloth and sweep net sampling methods for larvae of corn earworm and soybean looper (Lepidoptera: Noctuidae) in soybean. Journal of Economic Entomology 84: 1625– 1629. Wilson LT & Gutierrez AP (1980) Within plant distribution of predators on cotton: comments on sampling and predator efficiencies. Hilgardia 48: 3–11. Young SC & Tugwell P (1975) Different methods of sampling for clouded and tarnished plant bugs in Arkansas cotton fields. Arkansas Experimental Station Report Series 219, Fayetteville, AR, USA.