1996. The Journal of Arachnology 24:43–57 EFFECTS OF CULTURAL PRACTICES ON THE SPIDE R (ARANEAE) FAUNA OF LOWBUSH BLUEBERRY FIELD S IN WASHINGTON COUNTY, MAIN E Judith A. Collins', Daniel T. Jennings 2, and H. Y. Forsythe, Jr.' : 'Department of Applied Ecology and Environmental Sciences, University of Maine, Orono, Maine 04469 USA ; and 2Northeastern Forest Experiment Station, 180 Canfield Street , Morgantown, West Virginia 26505 USA ABSTRACT . Spiders of 17 families, 53 genera, and 87 species were captured in pitfall traps (n = 45 traps/year) placed in lowbush blueberry fields in Washington County, Maine, during the summers of 1986 and 1987. Species and numbers of hunting spiders (Lycosidae, Gnaphosidae, Thomisidae) were numeri - cally dominant. Significantly more (ANOVA, G-tests) spiders were captured in 1987 than in 1986 . Sex ratios were highly biased toward males both years . Species richness, diversity, and evenness of trapped spiders varied among three blueberry cultural treatments (mowing, burning, bearing crop). In 1986, rich- ness and diversity were greatest in crop bearing fields, with spiders more evenly distributed in burne d fields. In 1987, species richness, diversity, and evenness were greatest in burned fields . Over all weeks in 1986, there were no significant differences (ANOVA, DMRT) in mean numbers of individuals or specie s captured among treatments . Significant differences in mean catches among treatments were observed o n one of nine sampling dates in 1986. Greater variation was seen in 1987 for both individuals and species ; significant differences in mean catches among treatments were noted on six of 12 sampling dates. Per- centage similarity (PS) of species quantities among treatments was > 60; PS values were greater in 1986 than in 1987 . The blueberry-spider fauna had more species in common (QS) with terrestrial habitats than arboreal habitats in Maine . Lowbush blueberry, Vaccinium angustifol- populations were higher and somewhat mor e ium (Ait.), is a perennial shrub native to north- diverse in wild blueberries (Vaccinium spp.) eastern North America (Vander Kloet 1978) . than in cultivated, highbush blueberries (V. In Maine, lowbush blueberry is fostered and corymbosum L.). However, in Maine the ar- nurtured for berry production ; it comprises a thropod fauna associated with blueberr y major commercial crop. Numerous cultural plants and with blueberry fields has received practices have been developed and imple- scant attention ; most studies concern only pest mented to promote berry production . These insects (Forsythe & Collins 1986, 1987 , practices include : herbicidal control of com- 1988). We know of no previous studies con- peting weeds ; insecticidal control of pestifer- cerning the possible effects of either burnin g ous insects; irrigation of fields during period s or mowing on the araneofauna associated wit h of drought; and artificial pruning of older lowbush blueberry in Maine . plants by either burning or mowing . Recently, Some information is available about the ef- mowing blueberry fields has evolved as an al- fects of burning and mowing on spider pop- ternative management practice to burning . ulations in other ecosystems . Most studies Since early times, burning blueberry field s have shown that spider numbers decline fol- was the method used not only to prune blue- lowing burning (Riechert & Reeder 1972 ; Na- berry plants, but also to suppress competin g gel 1973 ; Dunwiddie 1991) ; however, Aitch- vegetation. However, the long-term effects of ison-Benell (1994) found that numbers of bo g burning, mowing, or other cultural practice s spiders were high two months after fire an d have not been fully evaluated for the blueber- then decreased . Other investigations hav e ry agroecosystem. shown that spider numbers also decline afte r In Arkansas, Johnson et al . (1981) and Hop- mowing (Howell & Pienkowski 1971 ; Nyf- kins & Johnson (1984) reported that spider feler & Breene 1990 ; Dunwiddie 1991) . 43 44 THE JOURNAL OF ARACHNOLOG Y In this paper, we : 1) describe the araneo- supports, was placed over each trap and re- fauna associated with lowbush blueberry mained in place until the traps were serviced . fields in commercial production in Washing- In 1986, traps were deployed on 20 Jun e ton County, Maine; 2) compare the density of and serviced weekly until 22 August for a to- pitfall-trap catches among three blueberry cul- tal of nine trap weeks . In 1987, traps were tural treatments (pruning by mowing = deployed on 15 May and serviced weekly un- "mow", pruning by burning = "burn", an d til 14 August for a total of 12 trap weeks . At bearing crop = "bear"); 3) evaluate the ef- each servicing, traps were removed from th e fects of these three cultural treatments on spi- ground and their contents passed through a der species composition and abundance; and fine mesh strainer. Captured organisms were 4) compare the terricolous spider fauna of placed in small jars with 70% ethanol an d Maines blueberry fields with that of other transported to the laboratory for sorting an d habitats . identification of spiders . Potential sampl e sizes were: n = 45 traps/year ; 5 traps/field X METHODS 3 cultural treatments X 3 replications/treat- Study sites.—Commercial lowbush blue - ment X 9 weeks = 405 in 1986, and X 1 2 berry fields that represented three blueberry weeks = 540 in 1987 . —Only cultural treatments (mow, burn, bear) were Spider identifications . sexually sampled for spiders in Washington County, mature spiders were identified to species; spe- Maine in 1986 and 1987 . Three fields were cies determinations follow the identification keys and descriptions of Kaston (1981) . Ju- selected and monitored for each treatmen t veniles, including penultimate stages, wer e each year. identified to genus . A few specimens (mostly Treatments.—Burn treatments were ap- Linyphiidae) were sent to Dr. C. D. Dondale, plied in November prior to each study year, Ottawa, for species determination or confir- i.e., 1985 or 1986 . Mow treatments were ap- mation . Representative specimens of all iden- plied in April of each study year, i .e., 1986 or tified species will be deposited in the arachni d 1987. Mow treatment areas were flail-mowed collection, U . S. National Museum of Natural by the grower using standard commercia l History, Washington, DC . flail-mowers. Burn treatment areas were sim- Data analyses.—Spider taxa: We used ilarly burned with commercial oil burners by nonparametric procedures (Sokal Rohl f the grower. Hexazinone (herbicide) and fertil- 1981) for statistical comparisons at P = 0.05. izer were applied to mow and burn fields b y The G-test was used to compare spider abun- the grower following standard lowbush blue- dances between years and by foraging strate- berry management practices (University of gy, and to compare sex ratios of trapped male s Maine Cooperative Extension Service 1986) . and females . Null hypotheses were : expected In 1986, phosmet (insecticide) was applied to abundances equal between years ; expected one bear field between 4–11 July ; in 1987, abundances equal between foraging strategie s azinphos-methyl (insecticide) was applied t o (web spinner, hunter) ; and expected propor- one bear field between 10–17 July. Both in- tions (0.50: 0 .50) of spider sexes. secticide treatments were standard applica- Data analyses .—Species richness, diversi- tions to control blueberry maggot, Rhagoletis ty, and evenness : Computations of species mendax Curran . richness, diversity, and evenness were mad e Pitfall traps.—At each study site, we de- using the program of Ludwig Reynolds ployed five pitfall traps along line transects ; (1988), where : species richness, NO = the starting points and orientations of transects number of all species in the sample regardles s were chosen at random and were at least 20 of their abundances (Hill 1973) ; species di- m from field edges and roads . Pitfall trap s versity, H = Shannons index (Shannon & were 0 .5 liter plastic drinking cups (height , Weaver 1949) ; and species evenness, E5 = a 13 .0 cm ; top diameter, 8 .6 cm) . Each cup was measure of how evenly species are distribute d filled to a depth of about 5 cm with ethylen e in a sample. For species comparisons, we in- glycol (antifreeze). A rain cover (12 .7 X 12.7 cluded only those species represented by adult cm) constructed of 0 .6 cm exterior plywood , spiders; hence, our estimates are conservative . and with four 16d nails (length = 8 .9 cm) as For comparisons of species similarities COLLINS ET AL.—SPIDER FAUNA OF LOWBUSH BLUEBERRY 45 among blueberry cultural treatments (Q-mod e hunter families comprised 20 .6% of the total analysis), we used the percent similarity (PS) species (n = 73) trapped that year. index of Bray & Curtis (1957), which take s Numbers of individuals also differed by into account species quantities in sampling foraging strategy each year ; individuals of the units. hunter guild were by far the most commonl y Sorensen's similarity quotient (QS), as de - trapped spiders each year and for both years fined by Price (1975), was used to compare combined (Fig. 2). In 1986, the hunter guil d the terricolous spider fauna we found in blue - consisted chiefly of individuals of Lycosidae berry fields of Maine with that of other similar (72.2% of all individuals), Gnaphosidae and dissimilar habitats in Maine . For these (6.8%), and Thomisidae (8 .8%); collectively, comparisons, we excluded species identifie d the remaining hunter families comprised 6 .7% only to genus in the various studies . Hence, of the total trapped individuals (n = 832). our estimates of spider faunal similarities i n Again, in 1987 the Lycosidae were numeri- Maine may be conservative . cally dominant (77 .8% of all trapped individ- Data analyses .—Treatment effects: Para- uals), followed by the Gnaphosidae (7 .0%) metric procedures were used to evaluate treat- and the Thomisidae (6 .5%). Collectively, the ment effects . Prior to analysis, pitfall-catch remaining hunter families comprised 4 .2% of data were subjected to Hartley's test for ho- the total trapped individuals (n = 1,890).