A terrestrial vacuum sampler for macroinvertebrates

Macroinvertebrates (hereafter invertebrates) are a vital component in the diets of upland game birds (Kimmel and Samuel 1984, Healy 1985, Landers and Mueller 1986), providing a rich source of protein and calcium (Stiven 1961, Reichle et al. 1969, Pattee and Beasom 1981, Hurst and Poe 1985), which are needed for rapid bone and tissue growth in poults and egg production in hens. In the past, to better manage for upland game bird brood range, wildlife biologists have sampled fields and the forest floor to determine which habitats harbor abundant invertebrates (Blackburn et al. 1975, Martin and McGinnes 1975, Owen 1976, Healy 1985, Jackson et al. 1987, Knox 1994, HoWield and Dimmick 1995, Peoples et al. 1996). These efforts have been particularly intensive in the eastern and southeastern United Stated where biologists have managed for wild turkeys (Meleagds gallopavo)T ruffed grouse (Bonasa umbelius), and cbobdbhite quail (ColW.4s virginianus). #’ .’ A number of methods havebeen used ingathering data on invertebrates. Use of the “Americaia” sweep” net (Beall 1935) is common, although it has many Emitations: First, abundance of many invertebrates, particularly those dwelling along the forest floor (e.g., snails, spiders@llipedes, centipedes) and those able to firmly grasp *vegetation, is underestimated when bsing a sweep net (Whittaker 1952, Hughes 1955). The;position of .an invertebrate on a plant poses another problem when sampling with a sweep net (DeLong 1932). Variations in temperature and wind velo.city cause invertebrates to, be found either higher on vegetation or lower near the ground surface (Romney ~945,$h.tghes 19557. Thus, sweepmg must be conducted at different heights during different conditions for comparable results. Additionally, ex. , : : ‘. ‘: i treme variation is encountered when 22 collectors sample with unequal intensity (Lockwood 1924, Whittaker 1952) and when sampling vegetation within certain habitats (e.g., old field with thick, waist-high grass and forbs vs. ankle-high clover patch), which prevents equal sweep intensity (Gray and Treloar 1933). Although the sweep-net method provides qualitative data important in determining predominant and least prevalent species and indications of distribution (Whittaker 1952), accurate estimates of true density are not possible, because there is no measure of the area sampled. This precludes the comparison of different studies and sites, since few (if any) researchers use the same sized sweep net, the same number of sweeps per unit sampled, or the same sweep intensity. In addition, it is impossible to capture all, or even a definite proportion of, the invertebrates within a given area using a sweep net (DeLong 1932). These limitations also are present b+hen sampling invertebrate populations using pitfall traps (Gist and Crossley 1975, Mader et al. 1990, Oliver and Beattie 1996) unless impassable borders (i.e., an enclosure) are constructed (Gist and Crossley 1973), and even then capture of flying invertebrates is limited. Another commonly used device for sampling invertebrates is the D-Vat, named after E. J. Dietrich who described a gasoline-powered suction device in 1959 and a modified version in 1961 (Dietrich et al. 1959, Dietrich 1961). The D-Vat is avacuum sampler which consists of a metal backpack frame with a 3-hp gasoline motor attached. The motor turns a fan that creates a vacuum through a 2O-cm (8inch)-diameter hose with a large plastic nozzle. An advantage to using a vacuum sampler is that density estimates can be ’ _