Showing papers in "Journal of the Kansas Entomological Society in 1995"

Diversity, habitat preferences, and seasonality of Kansas carrion beetles (Coleoptera: Silphidae).

Abstract: Pitfall trapping, blacklighting, and examination of institutional collections produced records for 13 species of Silphidae in Kansas: Necrodes surinamensis (Fabricius), Necrophila americana (Linnaeus), Nicrophorus americanus Olivier, Ni. carolinus (Lin naeus), Ni. marginatus Fabricius, Ni. mexicanus Matthews, Ni. orbicollis (Say), Ni. pus tulatus Herschel, Ni. tomentosus Weber, Oiceoptoma inaequale (Fabricius), O. novebora cense (Forster), Thanatophilus lapponicus (Herbst), and T. truncatus (Say). No current populations of the federally endangered silphid, Ni. americanus or Ni. mexicanus were documented in Kansas, and records for both species are more than 50 years old. Data based on 2007 specimens resulting from 1709 pitfall trapnights in 23 Kansas counties are standardized and used in an assessment of habitat preferences and seasonality among the encountered taxa. Four species (M. carolinus, Ni. marginatus, T. lapponicus, and T. trun catus) are nearly restricted to open prairies with sandy soil, 2 species (Ne. americana and O. noveboracense) are dominant in woodlands, and 3 species (Ni. orbicollis, Ni. pustulatus, and O. inaequale) occur in both wooded and open habitats. Necrodes surinamensis and Ni. pustulatus have bimodal peaks of activity in Kansas. Adults of O. inaequale and O. noveboracense were not captured in Kansas after mid-summer. Necrophila americana, Ni. marginatus, Ni. orbicollis, and Ni. tomentosus occur in Kansas from spring to late summer. The purpose of this study is to reveal current patterns of diversity, distribution, habitat preferences, and seasonality for Kansas carrion beetles (Cole?ptera: Sil phidae). Although Peck and Kaulbars (1987) summarized these attributes for all species in the conterminous United States, this study provides more detailed information about the habitats and seasonality of Kansas taxa. Carrion beetles have been widely studied for several reasons. First, most have a well-defined ecological role as primary scavengers of carrion and thus appeal to ecologists. Second, members of the subfamily Nicrophorinae (Nicrophorus spp.) stimulate interest from behavioral and evolutionary biologists due to their parental care (summarized in Anderson and Peck, 1985), a behavior quite uncommon among beetles. Third, the presence within this family of a federally endangered insect, the American burying beetle, Nicrophorus americanus Olivier, has drawn attention from biologists, conservationists, and the news media. Populations of this species have been intensively studied on Block Island, Rhode Island (Kozol et al., 1988), and in Oklahoma (Creighton et al., 1993), while populations in Arkansas (USFWS, 1991) and Nebraska (Ratcliffe and Jameson, 1992) were more recently re-discovered and are currently being investigated. Another reason con tributing to interest in Silphidae concerns the attractiveness of many species that have contrasting aposematic coloration (noted in Anderson and Peck, 1985). With abundant resources, coexistence of species with similar niche requirements can occur. As resources begin to limit populations, however, alternative niche 1 This paper published through a grant from the C. P. Alexander Fund of the Kansas Entomological Society. Accepted for publication 10 November 1995. This content downloaded from 157.55.39.59 on Sun, 16 Oct 2016 04:39:33 UTC All use subject to http://about.jstor.org/terms

The insect fauna of spring habitats in semiarid rangelands in central Oregon

Abstract: In semiarid rangelands, a variety of permanently wet habitats can be classified within the general category of springs. They may grade imperceptibly to either typical streams or to wetlands; habitats differ in water transit time and nature of aquatic and riparian vegetation. Emergence traps were used to collect aquatic insects from helocrene habitats in the Bridge Creek watershed in central Oregon. Water chemistry analyses showed high conductivity, pH, and cation concentration values at all sites. Nematocerous Diptera typically comprised greater than 90% of the aquatic insects in emergence traps. Other aquatic orders were uncommon and low in species richness and abundance. Plecoptera were common at only one site, while the damselfly, Argia vivida occurred at all sample sites. Terrestrial insects associated with emergent vegetation also were present in the sam ples. Many of the insect taxa that were collected do not occur in larger streams and thus insects of springs provide a significant contribution to the diversity of aquatic fauna in semiarid landscapes. In semiarid landscapes, such as those dominating the Great Basin and much of the Columbia River Basin, water is a critical resource for the biota as well as for human endeavors. Land-use issues relating to utilization of aquatic resources have been exacerbated by drought conditions in the West for at least the past six years. Overgrazing by domestic livestock, particularly when they have access to the stream and riparian areas, has detrimental effects on habitat quality as well as on associated biota (Kauffman and Krueger, 1984; Skovlin, 1984). Irrigation poses another threat to aquatic resources, particularly in time of drought. These issues may be quite volatile when fishery resources are jeopardized. Clearly, de cisions regarding use of aquatic resources will influence biodiversity in semiarid lands. However, such decisions are often based on information that is limited in scope due to the tendency of research and public interest to focus on a limited, albeit visible, portion of the ecosystem. Such is the case in semiarid landscapes where the large lotie habitats, especially those supporting salmonid fish, have received the most attention from aquatic scientists and environmentalists. In the headwaters of perennial streams there is a variety of aquatic habitats that can be classified under the broad category of springs. Little is known about the faunal composition or ecological role of these habitats (though see Cushing and Gaines, 1989). In this paper we report on the habitat characteristics and aquatic insect com position of a range of springs that occur in the Bridge Creek drainage in central Oregon. We also discuss the contribution of springs to the biodiversity in these landscapes and the role of isolated aquatic habitats in semiarid rangelands. Faunal surveys, such as we present, are meant to stimulate interest in these habitats and to provide a basis for further ecological studies. 1 This is contribution No. 10,561 from the Oregon Agricultural Experiment Station. This content downloaded from 207.46.13.158 on Wed, 11 May 2016 05:51:57 UTC All use subject to http://about.jstor.org/terms

Distribution and host plants of the blueberry maggot fly, Rhagoletis mendax (Diptera: Tephritidae) in southeastern North America

Abstract: To better determine the geographic range, pattern of host plant usage and phenology of Rhagoletis mendax, the blueberry maggot fly, an extensive set of rearings from fruit of Vaccinium and Gaylussacia (Ericaceae) species was made, primarily from natural populations of these host plants. Collecting focused on the relatively unstudied southeastern United States, although fruit collections as far north as Michigan and as far west as Texas were made. The known range of the blueberry maggot fly has been extended west to the Ozark Plateau, and the fly is now known to be much more widely distributed in southeastern North America than previously thought. Although R. mendax infests many species of Vaccinium and Gaylussacia in the northernmost part of its range, in most of its range it infests almost exclusively V. stamineum (deerberry). The relatively inflexible dia pause characteristics of R. mendax may contribute to this restriction in host plant usage. The blueberry maggot fly, Rhagoletis mendax Curran, is a serious native pest of blueberries {Vaccinium L. spp.), primarily in northeastern North America (Lathrop and Nickels, 1932; Bush, 1966). The economic importance of R. mendax notwithstanding, its geographic range and use of native host plants remain poorly known, especially in the southern part of its range. Bush (1966) lists only 12 localities of R. mendax in the southern United States, and specific host plant information is lacking for many of the specimens. Moreover, most of the Florida specimens mapped in Bush (1966) are actually not R. mendax, but are instead a new species of the R. pomonella species group that is restricted to V. arboreum Marshall. This species is discussed elsewhere (Payne and Berlocher, 1995). Better information about R. mendax is desirable for both practical and basic reasons. Commercial blueberry production is increasing in the southern United States. In 1986, there were 2059 hectares of improved cultivars of rabbiteye blueberries, V. ashei Reade, in production in 10 southeastern and mid-south states (Payne et al., 1988). In 1989, there were about 14,000 hectares of highbush blueberries, V. corymbosum L., in the U.S.A. with the largest production areas in New Jersey, Michigan, and North Carolina (Hancock and Draper, 1989). While R. mendax infestation of commercial rabbiteye blueberries in the South is rare, it may be on the rise (Payne et al., 1993). On the basic side, complete host plant information is needed to understand the population genetics of host race for mation, an area of study to which Rhagoletis species have made substantial con tributions (Feder et al., 1988; McPheron et al., 1988). Here we present new geographic range and host plant data on R. mendax, focusing on Vaccinium and Gaylussacia Humboldt, Bonpland, and Kunth (huck leberry) species in the southeastern North America. However, some collections 1 USDA-ARS, Southeastern Fruit and Tree Nut Research Laboratory, 111 Dunbar Road, Byron, Georgia 31008. 2 Department of Entomology, 320 Morrill Hall, University of Illinois, 505 S. Goodwin Ave., Urbana, Illinois 61801. Accepted for publication 8 August 1994. This content downloaded from 207.46.13.168 on Sat, 09 Apr 2016 06:33:39 UTC All use subject to http://about.jstor.org/terms 134 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY were also made in other poorly sampled areas, such as the Ozark Plateau and the midwestern United States. Materials and Methods Collections were made on seven automobile trips by S.H.B. during summer and fall 1989-1991, and on numerous trips by J.A.P. during 1985-1991 (Payne et al., 1989). All flies were reared from infested fruit. Most fruit collections were effectively random samples from all Vaccinium and Gaylussacia species at a site that had ripe or ripening fruit; only a few of the samples were seen to be obviously infested in the field. Ripening fruit was considered to be fruit beginning to show color. Fallen but non-decayed fruit occurring in collectible numbers was harvested and pooled with fruit picked form plants. Number of fruit collected was approx imated by weighing (Berlocher and Enquist, 1993). The emphasis was on wild host plants growing under natural conditions, with multispecies sites being especially sought out. However, some collections from domesticated cultivars of highbush blueberry, V. corymbosum, and rabbiteye blueberry, V. ashei, were included to obtain phenological information. The do mesticated plants were sampled from both cultivated and formerly cultivated, abandoned plantings. In addition, some undomesticated species and interspecies hybrids growing under cultivation were collected. Fly samples were used for multiple purposes. Larvae were frozen for enzyme electrophoresis (Berlocher, 1995), DNA analysis, and immature morphological studies, and adults were reared for phenological and morphological analyses (studies to be described elsewhere). This multiple use of material affected iden tification of material. At some sites with few fruit and low infestation rates, all material was saved as frozen larvae; at such sites, identification was based on electrophoretic characters (Berlocher et al., 1993). At most sites, however, adults were reared as described by Berlocher and Enquist (1993) for R. pomonella. Adult flies were identified following Bush (1966). Herbarium samples and plant photographs were taken at many of the sites. Herbarium samples are deposited in the Herbarium of the University of Illinois at Urbana-Champaign. The primary reference for plant identification was Vander Kloet's (1988) monograph of Vaccinium and Luby's et al. (1991) chapter on blueberries and cranberries. Regional floras were used as supplemental references for Vaccinium, and to identify Gaylussacia species. Insect vouchers (adult flies) for sites yielding sufficient adults have been deposited in the collection of the Illinois Natural History Survey. Collection sites are described in Table 1. Sites are named using nearest prin cipality or other landmark, but are not necessarily exactly at that point; for ex ample, the two Florida state park collections in Table 1 were actually from road sides outside the parks. A variety of cultivars of V. corymbosum and interspecific hybrids involving this species, collected separately, were pooled together for the N.C. State Sandhills Research Station site in Table 1, as all were uninfested. Additional details for sites included in the electrophoretic study are in Berlocher (1995); for other sites, contact the authors. Statistical testing of independence of whether a sample was infested or not, and date of collection, was carried out using contingency analysis. Instead of the familiar contingency x2> a Monte Carlo program (Macintosh) that simulates an This content downloaded from 207.46.13.168 on Sat, 09 Apr 2016 06:33:39 UTC All use subject to http://about.jstor.org/terms VOLUME 68, NUMBER 2 135 Table 1. Locality, date, and host plant species data for Rhagoletis mendax in eastern North America. Infestation rate is the number of larvae or pupae divided by the number of fruit collected. D indicates domesticated plants originally under cultivation, regardless of current cultivation condition, and C indicates wild plants under cultivation (see text). State/ Site (County) Date collected Host species No. fruit Infest, coll. rate (%)

Habitat separation among chironomidae (Diptera) in Big Springs.

Abstract: Habitat separation among Chironomidae was determined for five habitats in Big Springs, which is located in the high plains of western Kansas. Larvae, pupae, pupal exuviae and adults were qualitatively collected from each habitat on seven dates over a two-year period. The five habitats investigated were: spring source, spring run, a large pool, splash zones, and saturated soils/small seeps. Sixty-six species occurred in the five habitats, with Chironomini being most species rich (31 species), followed by Orthocladiinae (22), Tanypodinae (9) and Tanytarsini (4). The highest species richness was in the pool habitat (45 species), which was dominated by Chironomini. Orthocladiinae dominated all other habitats, with the spring run having second highest species richness (20 species), followed by the spring source and splash zones (each with 13 species) and the saturated soils and seeps (11 species). Forty-eight of the species occurred in only one habitat and only three taxa, Tanytarsus spp., Corynoneura spp. and Thienemanniella spp., were found in all five habitats. Jaccard's Coefficient showed that the composition of the splash zone and spring source, and the splash zone and spring run were most similar. However, an analysis of species composition using the Simple Matching Coefficient indicated that the splash zone and spring source were most similar, and the splash zone and saturated soils and seeps as second most similar. Both indices, however, indicated that the spring source and the large pool had the least similar composition. These results suggest that patterns of longitudinal zonation in the composition of chironomids, and perhaps other macroinvertebrates, along spring and spring run gradients are strongly influenced by physical variations in the micro habitats that occur as the groundwater discharge merges into a well defined stream with alternating pool-rifle habitats.

Seasonal occurrence of a fungal pathogen, Neozygites adjarica (Entomophthorales: Neozygitaceae), infecting banks grass mites, Oligonychus pratensis, and twospotted spider mites, Tetranychus urticae (Acari: Tetranychidae), in field corn.

Abstract: The seasonal occurrence of epizootics caused by the fungus Neozygites ad jarica (Tsintsadze & Vartapetov) were studied in populations of Banks grass mites, Oli gonychus pratensis (Banks), and twospotted spider mites, Tetranychus urticae Koch, on field corn, Zea mays L. The first evidence of fungal infection was 5 August, but the major epizootics occurred 14 August to 8 September. The fungal epizootics followed periods of 8-10 h per day of ambient relative humidity above 80%. N. adjarica epizootics occurred in the four test fields, but it was not clear whether or not the epizootic was the major agent of spider mite mortality. Predatory arthropods, miticide applications and corn maturity also appeared to be important spider mite mortality factors. N. adjarica epizootics were erratic and occurred too late in the season to prevent spider mite damage in corn. Awareness of conditions likely to foster fungal epizootics in spider mites will help us understand spider mite population dynamics. Fungal epizootics are sometimes observed in spider mite populations (Acari: Tetranychidae) on field corn, Zea mays L., in the western Great Plains of North America. Banks grass mites, Oligonychus pratensis (Banks), and twospotted spider mites, Tetranychus urticae Koch, are important pests of field corn in this region and frequently cause economic damage to field corn (Mock et al., 1981 ; Sloderbeck et al., 1988). Very little information is available on the pathogen causing these epizootics, or on the role of these epizootics in the population dynamics of spider mites in corn. Dick et al., (1992) described the morphology of the fungal pathogen that attacks spider mites of corn in this region. They identified the pathogen as Neozygites adjarica (Tsintsadze & Vartapetov). Other species o? Neozygites (=Entomoph thora, =Triplosporium) have been reported to infect the Banks grass mite in Texas (Pickett and Gilstrap, 1986) and the twospotted spider mite in other areas of North America (Carrier and Canerday, 1968; Carner, 1976; Brandenburg and Kennedy, 1981; Smitley et al., 1986a, b; Klubertanz et al., 1991). Initiation of Neozygites spp. infections during the cropping season is thought to be associated with high mite populations and favorable weather conditions, specifically, extended periods of high relative humidity (above 90%) (Wilding, 1981; Brandenburg and Kennedy, 1982; Smitley, 1986a). The fungus is thought to overwinter either as thick-walled resting spores (Nemoto and Aoki, 1975; Carner, 1976; Bitton et al., 1979) or as hyphal bodies in mummified mites (Ken neth et al., 1972). However, Brandenburg and Kennedy (1981) demonstrated 1 1716 Pinecrest Av., Garden City, KS 67846. 2 Corresponding author. Accepted for publication 18 June 1995. This content downloaded from 207.46.13.71 on Fri, 21 Oct 2016 04:54:40 UTC All use subject to http://about.jstor.org/terms 426 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY transmission of N.floridana Weiser & Muma inoculum from year to year in active overwintering populations of twospotted spider mite. Kenneth et al., (1972) did not observe resting spores of N. floridana in twospotted spider mites, but dem onstrated that hyphal bodies in mummified mites could sporulate up to 11 months after mite death. They suggested that the fungus overwintered as hyphal bodies in mummified mites, hidden in protected microhabitats. The field ecology of N. adjarica in Banks grass mites has not been reported. The objectives of this study were to determine the seasonal occurrence of N. adjarica epizootics in populations of Banks grass mite and twospotted spider mite in field corn and to examine some of the environmental factors associated with epizootics of this pathogen. Materials and Methods Commercial corn fields in southwest Kansas were surveyed in May and June 1985 to locate fields with early-season spider mite infestations. Four of these fields are included in this report. The fields were furrow irrigated and subject to normal farming practices, including miticide applications. The Finney County west field was treated 23 July with propargite at 0.76 kg Al/ha and 14 August with carbofuran at 0.34 kg Al/ha. The Stanton County field was treated 13 July with propargite at 0.76 kg Al/ha and 4 August with carbofuran at 0.34 kg Al/ha plus methidathion 0.45 kg Al/ha. The Haskell County field was treated 24 June with propargite at 0.76 kg Al/ha and 11 July with methidathion at 0.91 kg Al/ha. The Finney County east field was treated 15 August with dimethoate at 0.23 kg Al/ha. Relative humidity (RH) and rainfall were recorded at 1-h intervals by an au tomated weather station at the Southwest Research-Extension Center, Garden City, Kansas. The number of hours with ambient RH above 80% was equivalent to the number of hours within-canopy RH above 90% (Dick, 1987). The Finney County west, the Stanton County, the Haskell County and the Finney County east fields were 24, 66, 31 and 8 km from the weather station at Garden City, respectively. In each field, two study sites (one in Finney County East), ca. 2 x 20 m, were identified at least 7 m away from field margins. Weekly arthropod sampling started in early July. Four corn plants were collected systematically (every ninth plant along the row) within each study site. The first two plants were examined im mediately for larger arthropods associated with spider mites, including lady beetles and Onus spp., and specimens were collected for species determination. Plants three and four were prepared for processing with a leaf brushing machine. All green leaves on one side of each plant (two half-plants) were removed, cut into lengths to fit 3.8-1 plastic freezer bags, and transported to the laboratory in an ice chest. Leaf samples were processed immediately or refrigerated at 4-6?C for no more than 24 hours before processing. In the laboratory, the leaf sections were removed from the plastic bag, air-dried briefly to eliminate condensation, and processed with a modified (one brush removed) leaf brushing machine (Henderson and McBurnie, 1943; Dick, 1987). The glass collection disks were coated with a thin layer of Annul 535? (DeSoto, Inc.) solution (10% by volume) in 99% ethanol. The collection disks were centered over a 200-section equal area disk (Morgan et al., 1955) and examined with a dissecting microscope. Numbers of small arthropods, including live spider mites, This content downloaded from 207.46.13.71 on Fri, 21 Oct 2016 04:54:40 UTC All use subject to http://about.jstor.org/terms VOLUME 68, NUMBER 4 427 dead spider mites, predator mites, and thrips, were recorded using disk areas of 1/10 (low mite numbers) or 1/40 (high mite numbers). Numbers of arthropods were converted to numbers per whole plant for presentation. Subsamples of 25 50 live and 25-50 dead spider mites were removed systematically, mounted in Hoyer's mounting medium, and examined microscopically for mite species iden tification and N. adjarica infection. Spider mites were considered infected if they contained hyphal bodies, resting spores or had capilliconidia attached directly to the integument. While attached capilliconidia do not automatically result in in fection we considered them more detectable than hyphal bodies and, therefore, a better indicator of infection levels. The early stages of hyphal bodies were difficult to detect. Capilliconidea appear to be the primary agent of inoculation for this group of pathogens (Carner and Canderday, 1968). Voucher specimens of Banks grass mite cadavers infected with N. adjarica are preserved at the Herbarium ARSEF, USDA-ARS-PPRU, U.S. Plant, Soil, and Nutrition Laboratory, Ithaca, New York (% R. A. Humber).

Distribution of chironomids (Diptera: Chironomidae) and ceratopogonids (Diptera: Ceratopogonidae) along a Colorado thermal spring effluent.

Abstract: Chironomidae and Ceratopogonidae populations were studied in a thermal gradient at Poncha Hot Spring (T48N, R8E, S15) in South Central Colorado, U.S.A. Chemical analyses revealed moderate spring ion concentrations that ranged from trace amounts of nitrates to 210 mg/liter sulfate. Four new chironomid species in the genera Cricotopus, Polypedilum, Tanytarsus, and Rheocricotopus were found. Paratendipes ther mophilus Townes, Cricotopus sp., Micropsectra sp., and Larsia sp. were collected from the thermal spring effluent with P. thermophilus being the most abundant chironomid. The presence of P. thermophilus represents a new state record for Colorado. Dasyhelea cincta Coquillett and Palpomyia sp. were also found in the hot spring effluent with D. cincta being the most abundant ceratopogonid. The relationships between chironomid and ceratopo gonid abundances and biomasses to temperature are described. Chironomid abundance and biomass decreased in a curvilinear manner as temperature increased along the spring effluent's thermal gradient, whereas, there was no significant relationship between cera topogonid abundance and biomass and temperature. The ceratopogonid distribution pat terns suggest they were thermophilic. Thermal springs are unique environments that often have high water temper atures, high ion concentrations and pH values less than 4 or greater than 8 (Cas tenholz, 1969). These extreme conditions not only delimit macroinvertebrate survival, but also determine macroinvertebrate phenology, physiology, and ecol ogy. With temperatures exceeding 50?C it seems improbable that life can exist, but a variety of species have adapted to thermal confines. Thermophilic fungi have been found in waters of 62?C and ostracods and insects have been recorded from waters exceeding 50?C (Brock, 1978). Bacteria have been observed in 80?C waters and algal/bacterial mats have been extensively studied in water temper atures greater than 60?C (Kullberg, 1968; Castenholz, 1969; Brock, 1978). Early studies by Brues (1924, 1927, 1928, 1932) and Tuxen (1944) recorded macroin vertebrates from thermal spring regimes in the Western United States and Iceland, respectively, More recently Winterbourn (1968, 1969, 1970) studied macroin vertebrate populations from some New Zealand hot springs. Although little is known about thermal environments and insects of thermal springs, there is a growing number of studies on hot springs. According to Pritchard (1991) Diptera is the most abundant and diverse insect order found in thermal springs. The thermal ecology of Ephydridae (Diptera) has often been studied because ephydrids (brine flies) are the most abundant and cosmopolitan hot spring insects (Wirth and Mathis, 1979). Brock et al. (1969) observed the feeding rela tionship between two ephydrid genera and thermophilic microorganisms. Collins (1975b, 1977) and Collins et al. (1976) studied the unique relationship between 1 Current address: Kansas Biological Survey, 2401 Constant Ave., Lawrence, Kansas 66045. 2 Current address: 3550 N. Winslow Drive, Tucson, Arizona 85715. This content downloaded from 207.46.13.142 on Fri, 10 Jun 2016 06:02:19 UTC All use subject to http://about.jstor.org/terms 78 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY brine flies and algal/bacterial mats in Yellowstone National Park, USA, and Barn by (1987) described physiological responses of brine flies to extreme saline con ditions in a California hot spring. Pritchard (1991) listed Chironomidae as common thermal spring insects. Chi ronomids usually have been included only in hot spring taxa lists, with little emphasis placed on their behavior, ecology, or life history. Lamberti and Resh (1985) described distribution, abundance, and biomass of chironomids, among other macroinvertebrates, along a natural thermal stream effluent in northern California. In recent years chironomids have been used as environmental indi cators for thermal pollution from power plants and waste treatment outflow. Coler and Kondratieff (1989) studied the relationship between thermal outflow from a nuclear power plant and chironomid density. Pritchard (1991) did not list ceratopogonids from thermal springs. This is not surprising as there is little literature that lists them from the thermal environment. According to Waugh and Wirth (1976) some species of the genus Dasyhelea can be found in hot springs and Sheppe (1975) found Dasyhelea in some Zambian hot springs. Lamberti and Resh (1983) found Palpomyia sp. in artificial thermal streams they constructed to separate the effects of chemicals and temperature on stream benthos. Since there are no broadly accepted criteria for temperatures that define a spring as thermal, we have chosen to follow Tuxen's (1944) definition. He classified thermal springs as either absolutely hot springs, with temperatures of 40?C and greater or relatively hot springs, with temperatures between the annual mean maximum air temperature in the spring locality and 40?C. Often the spring source is too hot to sustain life so most studies have been of the spring outflow which exhibits a thermal gradient possessing distinctive flora and fauna. We studied the ceratopogonids and chironomids from a hot spring effluent located in south central Colorado, USA, including ceratopogonid and chironomid distribution along the effluent's thermal gradient as well as the relationships be tween temperature and biomass, and temperature and abundance. The questions we posed were: do chironomid and ceratopogonid biomass and abundance in crease as temperature decreases along the thermal gradient?

Population fluctuations within a spring community.

Abstract: Springs are known for uniformity of their abiotic conditions, yet they have annual fluctuations in primary and secondary production. A spring in the Arbuckle Moun tains of Oklahoma was sampled for 17 months to document population fluctuations and to determine the possible community level interactions. There was a distinct annual vari ation in the amount of algae, macrophytes and dead leaves within the spring. There was also an annual variation in the abundance of crayfish Orconectes virilis, amphipods Hyalella azteca, and snails Physella virgata. The amphipod fluctuation is hypothesized to reflect the fluctuation in food. Crayfish numbers are a result of annual reproduction. The snail fluc tuation is hypothesized to be a combination of fluctuations of food and pr?dation by crayfish. Springs are known for uniformity of their abiotic conditions such as water chemistry, temperature, and discharge (Odum, 1957; Glazier and Gooch, 1987; I versen, 1988). This uniform environment often has annual variation in primary and secondary production (Odum, 1957; Teal, 1957; Tilley, 1968; Wilhm, 1970; Ward and Dufford, 1979; Iversen, 1988). If environmental conditions are uniform, then population fluctuations are presumably the result of biotic interactions, such as life history traits and community interactions. For many aquatic insects in springs, population fluctuations have been attributed to life history characteristics (Thorup, 1963; Ward and Dufford, 1979; Williams and Hogg, 1988). Most aquatic insects spend part of their life-cycle out of the water where conditions can cause them to synchronize reproduction. Alternatively, constant temperatures in springs allow some species to reproduce asynchronously year round (Thorup, 1963; Wil liams and Hogg, 1988). Animals that are aquatic throughout their life and are not subject to fluctuations in abiotic conditions (e.g., molluscs and crustaceans) also fluctuate in number (Tilley, 1968; Stern and Stern, 1969; Wilhm, 1970; Williams and Hogg, 1988; Gooch and Glazier, 1991). These population fluctu ations are expected to be due to community interactions of food and pr?dation. This study was conducted to look at the relationships between the population fluctuations of the different aquatic species to determine possible causal com munity interactions. Because springs are well bounded with reduced abiotic vari ation and species diversity (Glazier and Gooch, 1987; Danks and Williams, 1991; Glazier, 1991; Gooch and Glazier, 1991), the number of possible confounding interactions are reduced.

Conspecific and Heterospecific Interactions of Male Rhagoletis Flies (Diptera: Tephritidae) on a Shared Host

Abstract: The western cherry fruit fly, Rhagoletis indifferens Curran, is a frequent, chronic pest of sour cherry (Prunus cerasus L.) in western North America, whereas the apple maggot fly, R. pomonella (Walsh), uses this host rarely. We compared male terri toriality and other aspects of host use by these species on sour cherry in the laboratory. When flies were released singly onto fruits, residence times of both sexes of R. indifferens were longer than those o?R. pomonella, and females of R. indifferens attempted to oviposit much more frequently than did R. pomonella females. During conspecific encounters, males of R. indifferens interacted for longer periods of time than did R. pomonella males, were more active per unit time, and were more likely to engage in aggressive behavior (e.g., "pouncing"). One fly was actively driven off the fruit in 17% of trials involving R. indifferens males vs. 4% of trials involving R. pomonella males. In head-to-head encounters between species, R. indifferens males were more likely to be the sole fly remaining on a fruit. Fly size (as determined by maximum head width) did not appear to influence the outcome of conspecific encounters between R. indifferens males. Both phenological and behavioral differences between R. indifferens and R. pomonella appear to account for their different degrees of association with sour cherry in Utah. In some insects, males attempt to mate by intercepting females on or around resources used for oviposition (Baker, 1983; Thornhill and Alcock, 1983). Male fruit flies in the genus Rhagoletis establish short-term territories on host fruits, which serve as the primary mating site (e.g., Prokopy and Bush, 1973; Smith and Prokopy, 1980). Aggression between conspecific or heterospecific males may be common when population density is high (Papaj, 1994), and several types of agonistic behavior have been observed both in the field and laboratory (Boyce, 1934; Biggs, 1972). Here we describe male territoriality in two Rhagoletis species that differ in their degree of association with a common host. The western cherry fruit fly, R. indifferens Curran, is indigenous to northwestern North America, where it has infested cultivated cherries (Prunus spp.) for more than 75 years (Bush, 1966). This species was first reported in Utah in 1980, and rapidly became an abundant, chronic pest in all cherry-growing regions (Davis and Jones, 1986). The apple maggot fly, R. pomonella (Walsh), is commonly associated with native black hawthorn, Crataegus douglasii Lindley, in Utah and Colorado (Davis and Jones, 1986; Kroening et al., 1989). It was confirmed to attack sour cherry, Prunus cerasus L., in one region of Utah in 1983 (Jorgensen et al., 1986), probably following dispersal from nearby hawthorn stands (Mc Pheron, 1990). These localized populations of R. pomonella soon dwindled or disappeared from sour-cherry orchards, however, while densities of R. indifferens increased (Allred and Jorgensen, 1993). Elsewhere in North America, the apple maggot attacks cherries only sporadically (Shervis et al., 1970, and references therein), and never within the geographic range of R. indifferens. This study Accepted for publication 19 November 1994. This content downloaded from 207.46.13.124 on Wed, 22 Jun 2016 05:05:27 UTC All use subject to http://about.jstor.org/terms VOLUME 68, NUMBER 2 207 examines differences in post-alightment behavior of R. indifferens and R. po monella, and focuses on territoriality in males. Aggression between males has been characterized separately for the two species (Biggs, 1972; AliNiazee, 1974); we compared conspecific and heterospecific interactions under identical laboratory conditions. Materials and Methods Flies used in the experiments were obtained from naturally infested fruits at two nearby sites in Cache Co., Utah. Larvae of R. indifferens were collected from an abandoned sour cherry orchard in Providence; R. pomonella larvae were col lected from black hawthorn fruits in Wellsville. Fruits were placed on screens over moist vermiculite, which provided a medium for pupation as larvae left the fruit. Puparia were chilled at 6-8?C for 7-10 months, and then incubated at 25?C and 16L:8D photoperiod. Emerging adults were placed in 30 x 30 x 30 cm cages of plexiglas and organdy cloth, and were provided water and Bio-Serv Adult Apple Maggot Diet #9148 (Bio-Serv, Inc., Frenchtown, NJ; after Boush et al., 1969). We tested flies of both species 15 days after adult emergence. Dissection and behavioral observations confirmed that flies were sexually mature by this time. In one experiment, we also tested 3-day-old R. indifferens flies, which were sexually immature. Test flies had no prior exposure to host fruit; such experience is known to alter male behavior in R. pomonella (Prokopy et al., 1989). Trials were con ducted under ambient laboratory temperatures of 21-24?C. The first experiment compared the residence time of flies released singly onto sour cherries (cv. Montmorency). Unsprayed, fresh-picked fruits were washed in distilled water and dried. A single fruit was suspended with wire from the top of an empty cage that was similar to the one used for emerging adults. A fly was randomly chosen from a colony cage and transferred to the test cage in a small vial. Each fly was allowed to walk or hop onto the test fruit from the mouth of the vial. We recorded the elapsed time before a fly left the fruit, up to a maximum residence time of 120 sec. This trial duration was chosen because preliminary observations suggested that flies stayed on the fruits for either very short periods (<30 sec) or for well over 120 sec. No fruit was used for more than one fly. A second experiment was used to measure the tendency to oviposit on sour cherry. Females of R. indifferens and R. pomonella were alternately released onto fruits, and we recorded whether females attempted to lay an egg within 180 sec. We videotaped interactions between pairs of males on fruit. Individual fruits were suspended from the top of a cage as before, and a randomly chosen male was allowed to walk onto the fruit. After a few seconds, a second male was introduced onto the fruit. Interactions between males were recorded with a video camera that was connected to a time-lapse vid?ocassette recorder. Light was provided by a 22-W, circular, fluorescent tube 25 cm above the cage. Videotapes were later analyzed at slow motion, which enabled us to quantify the frequency of behaviors of short duration (e.g., wing-waving). We classified fly activity into six interactive behaviors (i.e., behaviors that appeared to be directed at the other fly) and one non-interactive behavior (Table 1). The behavior labeled as "boxing" here is equivalent to "pawing" in Biggs (1972); "pouncing" is equivalent to "charging headlong" and "head-on collision" in Biggs (1972) and AliNiazee (1974), respectively. The outcomes of male-male This content downloaded from 207.46.13.124 on Wed, 22 Jun 2016 05:05:27 UTC All use subject to http://about.jstor.org/terms 208 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY Table 1. Behaviors exhibited by pairs of male Rhagoletis flies in laboratory trials.

New species of parasitic wasps attacking cereal aphids in the Pacific Northwest (Hymenoptera: Braconidae: Aphidiinae).

Abstract: Two new parasitoid species, Monoctonus washingtonensis sp. n. and Praon yakimanum sp. n. (Hymenoptera: Braconidae: Aphidiinae), reared from cereal aphids [Diuraphis noxia (Kurdjumov) and Rhopalosiphum padi (L.)] in Washington State are described and illustrated. The appearance of the Russian wheat aphid, Diuraphis noxia (Kurdjumov), and its negative impact on small grain crops, has stimulated research on the whole of the aphid parasitoid fauna in the Pacific Northwest. Recognizing the spectrum of parasitoid species attacking aphids is fundamental to understanding and as sessing their potential as natural control agents in agricultural, forest, rangeland, and urban settings. Two new species of parasitic wasps of cereal aphids in Washington State are named and described, and the key features illustrated. Description of New Species Monoctonus washingtonensis Pike & Stary, sp. n. diagnosis: This species can be distinguished from other Nearctic congeners (see Stary, 1974; Stary and Smith, 1976; Marsh 1979) by a 14-segmented (rarely 13) poorly pubescent antenna (female), the shape of the antescutal depression of the pronotum, the areolation of the propodeum, the shape and direction of the radial vein in relation to the pterostigma, and the presence of distinguishable fused radial and medial cells in the forewing. etymology: The name of the new species is derived from its known distri bution, Washington State. description: Female: Head. Eyes sparsely haired. Malar space V6 to V5 eye length. Tentoriocular line % to V3 intertentorial line. Face with sparse hairs over surface and with single rows of hairs along eye margins. Clypeus with sparse, long hairs. Maxillary palpus 4-segmented, labial palpus 2-segmented. Antenna 14 segmented (rarely 13), reaching to middle of metasoma; flagellum somewhat thick ened in apical half, flagellomere 1 (= F^ (Fig. lb) slightly less than 4 times as long as wide, with sparse, long semierect hairs subequal in length to diameter of segment, without rhinaria. F2 (Fig. lc) slightly wider than Fj (15:14), 3 times as long as wide, with long semierect hairs equal in length to diameter of segment, with 0-1 rhinarium. Middle flagellomere, F7 (Fig. If), a little wider than F2 (17: 15), 2.5 times as long as wide, with sparse, long semierect hairs subequal to diameter of segment, with 2 rhinaria. Preapical flagellomere, F10 (Fig. lg), 2.5 times as long as wide, equal to F7 in width. 1 Washington State University, Irrigated Agriculture Research & Extension Center, Rt 2, Box 29 53A, Prosser, Washington 99350-9687, USA. 2 Institute of Entomology, Academy of Sciences of the Czech Republic, Branisovsk? 31, 370 05 Cesk? Bud?jovice, Czech Republic. Accepted for publication 14 April 1995. This content downloaded from 40.77.167.104 on Fri, 07 Oct 2016 05:58:39 UTC All use subject to http://about.jstor.org/terms VOLUME 68, NUMBER 4 409 Mesosoma. Antescutal depression of pronotum (Fig. lj) wide, obtusely tri angular, almost reaching anterior margin of pronotum. Mesonotum (Fig. la) with notauli distinct in ascendent part, effaced on disc where they are each traced with a row of long sparse hairs, otherwise almost hairless except for a few hairs close to prescutellar groove. Propodeum (Fig. 1 d, e) areolate; lower lateral carinae always distinct, well-developed and definable; central, narrow areola variable in shape and structure; upper carinae from near complete to irregular rugosities with a few rings inside indicated areola. Forewing (Fig. lh). Pterostigma short and narrow, 6-7 times as long as broad. Metacarpus short, equal to V3 pterostigma length. Radial and medial cells confluent and well-defined. Apical half of fused intermedial and medial vein and lower half of interradial cross vein colorless. Radial abscissa 1 (Ra,) straight to slightly arcuate, almost perpendicular to pterostigma. Raj shorter than Ra2, but longer than greatest width of pterostigma. Interradial cross vein shorter than Ra1? but equal to greatest width of pterostigma. Posterior marginal hairs long, 3 times as long as surface hairs. Legs. Hind femur with long, sparse, semierect hairs, equal to 2/3 middle diameter of femur. Hind tibiae with long, sparse, semierect hairs, slightly greater in length than middle diameter of tibia. Metasoma. First metasomal tergite (Fig. Ik) twice as long as greatest width; surface coarsely rugose, with sparse long hairs; spiracles dorso-laterally prominent; interspiracular distance slightly greater than distance from spiracles to basal at tachment. Genitalia (Fig. li). Ovipositor sheath ploughshare-shaped; hairs semierect on anterior margin, uniform in length and equal to Vi widest diameter of sheath; hairs on posterior apex subequal to widest diameter of sheath. Coloration. Head brown. Mouthparts yellowish, apex of mandible brown. Scape, pedicel and base of F{ yellowish, remaining portion of F{ yellowish to brownish, remainder of antenna brown. Mesosoma brown; prothorax and pro podeum sometimes lighter. Wings subhyaline, venation light brown to brown. Tegulae light brownish. Metasoma brown, except tergite 1 yellow brown to light brown. Ovipositor sheaths brownish. Length of body about 1.7 mm. Male: Similar to female with some exceptions: antenna 17-segmented (rarely 16); coloration somewhat darker, mouthparts and legs light brownish, antenna brown, scape brownish, pedicel and narrow base of F, yellowish. Length of body about 1.6 mm. holotype: Female, Washington State University (cereal field), Prosser, Wash ington, 8 May 1992, D. Allison and K. S. Pike. Host: Rhopalosiphum padi (L.) on wheat. Deposited in USNM, Washington D.C., Type No. 105226. paratypes: Females, 9; males, 6; additional paratypes dissected and mounted on slides, same data as holotype, except deposited in the Washington State Uni versity Collection, James Museum, Pullman, Washington. Other material from R. padi (L.) and Diuraphis noxia (Kurdjumov) in the Washington State University Collection at Prosser, Washington and collection of P. Stary, Cesk? Bud?jovice, Czech Republic. remarks: The current Nearctic Monoctonus include not only indigenous spe cies, but also reported introduced European species (Mackauer, 1962; Stary and This content downloaded from 40.77.167.104 on Fri, 07 Oct 2016 05:58:39 UTC All use subject to http://about.jstor.org/terms 410 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY

Burrow construction by the Japanese carpenter bee, Xylocopa appendiculata circumvolans Smith, for overwintering (Hymenoptera: Anthophoridae).

Abstract: Fifteen nests of Xylocopa appendiculata circumvolans were collected to in vestigate the overwintering bionomics. Nine of these were brood nests, which contained up to 7 overwintering bees. Generally, more males than females were found in these nests. The number of brood cells per nest was always higher than that of overwintering bees in the nest. The remaining 6 nests, however, were distinguished by a short and narrow tunnel, and contained only a young female. These observations suggest that some females disperse from their natal nests and excavate new burrows for overwintering, while other females and most males remain in the natal nests to form hibernating assemblages. Carpenter bees belonging to the genus Xylocopa are conspicuous insects, most occurring in the tropics and the subtropics and fewer in the temperate regions (Hurd and Moure, 1963). Bionomic studies which have been undertaken during the active period of the bees (e.g., life cycle, nesting and interactions between conspecific bees in relation to mating and sociality), have greatly advanced our knowledge of the bees during the last 30 years (Marshall and Alcock, 1981; Sakagami and Maeta, 1986; Velthuis, 1987; Gerling et al., 1989). However, information on the bionomics during the inactive winter period remains insufficient. Little attention has been paid to hibernating imagines that over winter jointly in nests. In this note I report the overwintering habit of Xylocopa appendiculata cir cumvolans. materials and methods: A total of 15 nests of X. appendiculata circumvolans were collected on January 15, 1989, on the campus of Kobe University (14 nests, nest nos. 1-9 and 11-15) and on December 23, 1988, in Kobe Municipal Arboretum (1 nest, no. 10), Kobe, western Honshu, Japan. These nests were constructed in dead branches of cherry and other ornamental trees. Some of these nests dropped from the tree and were collected from the ground. Each nest was cut open in the laboratory soon after collection, and the number of overwintering adults, their sexes and conditions were recorded. The number of progeny in each nest was estimated by the number of barrel-shaped concavities left inside the burrow. results: The structure of nest nos. 1-9 agreed well with that of the brood nest reported in previous papers (Ikushima, 1934; Kojima, 1979). Details of nest parameters are presented in Table 1. All nests had a single round or slightly elliptical hole for the entrance, which led to a burrow extending along the grain of the wood in both directions. A nest entrance was always found underside a branch. In addition, the inner surface of the nest burrow was usually clean and smooth but the traces of brood cells were distinct. The length of the nest burrows ranged from 12 to 29 cm. A total of 22 overwintering bees (153379$), all of which had unworn wings, were collected from these 9 nests, 0 to 7 bees (mean 2.4 ? 2.1 SD, n = 9) per nest. More males than females were usually found in each nest. The number of brood cells per nest ranged from 5 to 13 (mean 8.7 ? 3.0 SD, n = 9). Burrow structure of nest no. 10 and of five of the nests (nos. 11-15) was quite similar but distinct from burrows of the brood nests (Fig. 1, Table 2). Their fundamental features were; (1) the nest consisted of a single round to short elliptical hole for the entrance (11 to 14 mm in diameter) which led to a short burrow (less than 10 cm) extending along the grain of the wood in either direction. The burrow was slightly narrower than that of the brood nests. The inner surface of the burrow was rough, and (2) the burrow contained only a young female with unworn wings. The time when these burrows were constructed was unknown. In two occasions, however, the author was able to observe that it Accepted for publication 19 September 1993. This content downloaded from 157.55.39.17 on Thu, 01 Sep 2016 06:27:48 UTC All use subject to http://about.jstor.org/terms VOLUME 68, NUMBER 1 117 Table 1. Brood nest size and contents of Xylocopa appendiculata circumvolans in winter. Entrance diameters Nest contents Burrow length -Estimated no. of Nest no. Vert./horiz. (mm) (cm) Female Male Total cells constructed 1 13/? 29.1 0 1 1 13 2 12/14 22.9 2 3 5 9 3 11/12 19.2 0 2 2 6 4 13/? 28.3 2 5 7 12 5 11/10 13.8 0 115 6 13/? 25.9 0 2 2 11 7 12/? 16.1 2 13 6 8 13/? 12.3 10 15 9 13/14 27.2 0 0 0 11 were female bees constructing such burrows. The first was observed on July 27, 1989, on the campus of Kobe University and the second was on August 9, 1991, in Kobe Municipal Arboretum. Judging from these results, it may be concluded that this type of burrow is constructed for overwintering. discussion: In all brood nests studied, the number of brood cells per nest was more than that of progeny overwintering in a particular nest (Table 1). This phenomenon can be explained in two ways. (1) A significant immature and juvenile mortality may be caused by parasitism, predation and other factors. (2) Part of the progeny may disperse from their natal nests before overwintering. Although no data are available on the survival rate of immature and juvenile stages in X. appendiculata cir cumvolans, it would be unusual for offspring mortality before winter to be as high as 71.8% (56/78; see Table 1). Watmough (1983) reported that in South African carpenter bees, 17% of mortality was due to parasites and 8.5% by predation. Also it is well known that female carpenter bees show some subsocial behaviors to probably enhance the survival rate of progeny, e.g., guarding by blocking the nest entrance with the metasoma during the laying and post-laying periods and trophallactic feeding

Estimating densities of grasshopper (Orthoptera: Acrididae) assemblages: an extension of the binomial sampling technique.

Abstract: The effective range of the binomial sampling technique for determining densities of rangeland grasshopper assemblages from the proportion of samples containing grasshoppers was extended by relating the proportion of 0.1 m2 samples containing at least 2, 3, 4, or 5 grasshoppers (tally thresholds, t) to grasshopper density. This was done by using the Poisson probability density function to calculate the probabilities of obtaining 0, 1, 2, 3, and 4 grasshoppers per 0.1 m2. These probabilities were then summed and subtracted from 1.0 to yield the theoretical probabilities corresponding to >2, 3, 4, or 5 grasshoppers per sample (binomial sampling models). Data collected throughout central and eastern Wyoming were used to evaluate the fit of these sampling models. Using tally thresholds of 2, 3, 4, and 5, and a maximum-allowable standard error to density ratio of 0.15, grasshopper densities could be reliably estimated up to 38, 63, 84, and 104 grass hoppers per m2 (32, 53, 70, and 87 per yd2). Legg et al. (1993) developed a method whereby the average density of rangeland grasshoppers could be predicted from the proportion of 0.1 m2 visual samples that contained at least one grasshopper (binomial sampling). Here, we use the term "prediction" in the statistical sense of determining new, grasshopper densities with binomial sampling, at a given point in time, and not in the sense of deter mining grasshopper densities at some future point in time (forecasting). This method provided good resolution for predicting grasshopper densities up to 16 per m2, above which decreasing resolution occurred. United States Department of Agriculture, Animal Plant and Health Inspection Service, Plant Protection and Quarantine personnel have since expressed interest in using this method for pre dicting grasshopper densities within the range of 30-60 per m2 (25-50 per yd2) (Larsen, pers. comm.). This expanded range is supported by the economic thresh olds developed by Davis et al. (1992) which suggest that treatments are rarely justified at grasshopper densities less than 27 per m2 (23 per yd2), and some control methods are not cost effective until densities exceed 48 per m2 (40 per yd2). To accommodate these higher thresholds, the binomial sampling method was ex tended by using different tally thresholds (t). A tally threshold is the minimum number of grasshoppers necessary to consider a sample to be "infested". Tally thresholds of 2, 3, 4, and 5 grasshoppers per 0.1 m2 were used. Materials and Methods Data collected throughout central and eastern Wyoming by Lockwood and DeBrey (1990), and analyzed by Legg et al. (1993), were re-evaluated to determine how many of 40 0.1 m2 visual samples, at each of 175 locations, contained at least 2, 3, 4, or 5 grasshoppers. The frequencies of visual samples that met these criteria were then determined. Frequencies were divided by the number of samples Accepted for publication 3 October 1994. This content downloaded from 157.55.39.104 on Sat, 18 Jun 2016 07:21:10 UTC All use subject to http://about.jstor.org/terms VOLUME 68, NUMBER 2 179 AVERAGE GRASSHOPPERS PER 0.1 m2 Fig. 1. Plot of fitted (curves) and observed proportions of samples containing at least t grasshoppers on corresponding grasshopper densities. Observed proportions were calculated for / = 2 ( ), 3 (+), 4 (*), and 5 ( ) grasshoppers per 0.1 m2 sample. Data collected throughout central and eastern Wyoming, 1990. taken per location (40) to estimate proportions. Grasshopper densities were also estimated by counting all grasshoppers per sample, summing the counts over samples, and dividing that sum by the number of samples per location. Theoretical Poisson probabilities were then calculated with the following equation: P(r) = (Xr/r\)e~* [1] where P(r) is the theoretical sample-model probability of obtaining r grasshoppers in any one sample, X is the estimated grasshopper density per sample unit, and e is the base of the Napierian logarithm (Snedecor and Cochran, 1967). The following theoretical probabilities were then calculated: P(0\ P(l), P(2), P(3), and P(4). These were summed and taken from 1.0 with the following equations: P2 = 1 [P(0) + P(l)] [2] P3 = 1 [P(0) + P(l) + P(2)] [3] P4 = 1 [P(0) + P(l) + P(2) + i>(3)] [4] P5 = 1 [P(0) + P(l) + P(2) + P(3) + P(4)] [5] where P2_5 are the theoretical probabilities of obtaining at least 2, 3, 4, or 5 grasshoppers per 0.1m2 of visualized area. Multiplying these sample-model prob This content downloaded from 157.55.39.104 on Sat, 18 Jun 2016 07:21:10 UTC All use subject to http://about.jstor.org/terms 180 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY abilities by the number of visual samples taken at a location provided expected Poisson frequencies. A chi-square goodness-of-fit test was conducted to determine how well the expected frequencies fit the observed sample frequencies (alpha = 0.05) (Snedecor and Cochran, 1967). Results and Discussion Plots of Poisson sample model probabilities and observed proportions of sam ples containing at least t grasshoppers are shown in Fig. 1. The chi-square good ness-of-fit test indicated that the sampling models for t = 2 and 4 grasshoppers fit the data well (chi-square = 102.8 and 71.8, 159 df, P ? 1.0). However, chi square statistics for the t = 3 and 5 sampling models were highly significant (384.8 and 358.8, and 159 df, P 3 model, and a single spurious frequency at a low density (5.3 per m2) accounted for 80% of the chi-square value for the >5 model. Without these data points, the chi-square goodness-of-fit sta tistics were 64.7 and 71.8 for the t = 3 and 5 models, respectively (157 and 158 df, P ? 1.0). Given these results, it appears that the extended binomial sampling models adequately described the data. The use of binomial sampling models is normally restricted to a pre-determined range of proportions (e.g., Gerrard and Chiang, 1970; Legg et al., 1992). This range is determined by calculating the ratio of the standard error of prediction to density, and excluding those proportions associated with ratios that are greater than a given value. The relation between ratios and density is approximately "u" shaped, with ratios being high when densities are low or high, and reaching a minimum somewhere in between (Tippett, 1932). Fisher (1971) noted that stan dard errors relative to Poisson densities predicted from the proportion of samples containing no individuals, were minimized when P(0) ? 0.2 and x = 1.6. Gerrard and Chiang (1970) computed standard errors as percentages of densities, and they suggested that densities of Diabrotica spp. eggs in soil could be predicted from a percentage of samples containing at least t eggs that fell within the 50-80% range. This method was based on a standard error to density ratio being 2). Calculated ratios were divided by the square root of sample size (40) (Tippett, 1932) to determine the approximate lower and upper ranges of reliability based on values of 0.15 (Gerrard and Chiang, 1970) and 0.25 (Southwood, 1978, p. 7) (Table 1). To our knowledge, these values were subjec tively chosen by Gerrard and Chiang (1970) and Southwood (1978). Note that the range of reliability for t = 2 at the ratio of 0.15 is just greater than that determined by Gerrard and Chiang, and all ranges increased with increasing t (Table 1). We have not empirically examined the practicality of these sampling models, This content downloaded from 157.55.39.104 on Sat, 18 Jun 2016 07:21:10 UTC All use subject to http://about.jstor.org/terms VOLUME 68, NUMBER 2 181 Table 1. Grasshopper densities per 0.1 m2 (/x), and associated proportions of samples containing at least / grasshoppers (P), which satisfied the condition that the ratio of the standard error to density (ax/fl) be 0.15 or 0.25 for a sample size of 40.