Ontogeny of cannibalism in rangeland grasshoppers (Orthoptera: Acrididae).

Abstract

The ontogeny of necrophagy and necrophily in rangeland grasshoppers was determined using videotaped observations of attraction to cadavers and determinations of gut contents in second instars through late season adults. Attraction to cadavers began in the third instar and increased through early season adults. Utilization of grasshopper ca davers by other scavengers (primarily ants) lagged about 2 weeks behind grasshopper necrophily. Necrophagy in grasshoppers began as early as the third instar and increased with development, until September. From 0 to 15% of adult grasshoppers were found to have chitin in their gut contents, depending on subfamily. Melanoplinae had the greatest frequency of necrophagy, followed by Oedipodinae, Gomphocerinae and Romaleinae. Based on these findings, the current timing of application of Nosema locustae Canning (a mi crosporidian parasite of rangeland grasshoppers, transmitted by cannibalism) during the third instar appears to be appropriate. Adult rangeland grasshoppers are both necrophilous (i.e., attracted to cadavers, Lockwood, 1989) and necrophagous (i.e., consume cadavers, Gangwere, 1961; Mulkern et al., 1962; Lavigne and Pfadt, 1964; Lockwood, 1989). Although the vast majority of work on the feeding behavior of acridids has addressed phytopha gy, studies of carnivory indicate that necrophagy may play an important role in the behavioral ecology of rangeland grasshoppers. The practical significance of necrophagy lies in the role of this behavior in the biological control of rangeland grasshoppers. The only federally approved, commercially available biological control agent for grasshoppers is the microsporidian, Nosema locustae Canning. Nosema lo custae is standardly applied at a rate of 1 lb of bran (109 spores) per acre. Once the bran bait is consumed, necrophagy is the primary mode of transmission of the pathogen within the grasshopper community (Henry, 1972). Current recom mendations call for application of the pathogen when the majority of a grasshopper assemblage is in the third instar (Henry et al., 1973). This timing strategy is intended to correspond to the period of maximum bran acceptance by nymphs (USDA, 1987). However, since the ontogeny of necrophily and necrophagy are not known, this timing of application may not optimize efficacy. To optimize the rate of infection of a grasshopper community within a season, and perhaps between seasons, TV. locustae must be applied when grasshoppers are at a developmental stage that will transmit the pathogen. That is, the timing of application must take into account not only the rate of bait consumption but the rate of necrophily and necrophagy as well. It may be important to differentiate between necrophily and necrophagy since both behaviors are critical to the trans mission of TV. locustae. Necrophily becomes especially relevant if the pathogen is applied to moderate densities of grasshoppers, as has been suggested (USDA, Accepted for publication 20 March 1989. This content downloaded from 207.46.13.76 on Sat, 10 Sep 2016 05:47:23 UTC All use subject to http://about.jstor.org/terms VOLUME 62, NUMBER 4 535 1987). Under such conditions, it is most important that grasshoppers are at a stage of development in which they are actively attracted to cadavers. The objectives of this study were to determine the ontogeny of necrophily and necrophagy within subfamilies of rangeland grasshoppers. In addition, the poten tial impact and seasonality of competition with other scavengers was assessed. Materials and Methods The study site was located 10 km north of Fort Collins, Colorado, at an elevation of 1800 m. The site was dominated by western wheatgrass, Agropyron smithii, needle-and-thread, Atipa comata and smooth brome, Promus inermis. This area received 10.7 cm of precipitation in the course of the study, primarily in the form of thunderstorms. The average daily temperature during the study was 19?C, with a high of 34?C and a low of 2?C (National Oceanic and Atmospheric Adminis tration, 1987). The research was conducted between May 28 and September 25, 1987. The grasshopper assemblage consisted of Melanoplinae (67%), Gompho cerinae (35%), Oedipodinae (4%) and Romaleinae (1%). To determine the ontogeny of necrophagy, 100 grasshoppers were collected weekly using a 38 cm diameter sweep net and killed and stored in 70% ethanol. The developmental stage and subfamily of each specimen was determined and the foregut and midgut were removed. The gut contents were washed from the tissue using distilled water and then analyzed for chitin (Campbell, 1929). Camp bell's test for chitin was periodically performed on cuticular fragments to protect against false negative tests arising from procedural or interpretive inconsistencies. To ascertain the rate of false positive tests, fifty grasshoppers were individually caged and fed for 5 days on western wheatgrass, smooth brome and dandelion leaves. The gut contents of these grasshoppers were randomly included in field samples using a single blind design (i.e., the technician performing the tests did not know which samples were from plant-fed grasshoppers). Since none of these specimens yielded a positive test, no correction of the data was necessary. The procedure of Lockwood (1989) was modified to determine the ontogeny of necrophily. Three grasshopper cadavers were placed in the field; these grass hoppers had been killed by freezing the previous week and allowed to thaw just prior to their placement. After placement, a 400 cm field around the cadavers was filmed for 1 hr using a Panasonic AG-100 videocamera on a telescoping tripod. Three replications of the treatment and two replications of the control (a site without cadavers) were recorded each week in a random sequence. The number of grasshoppers and other scavengers that were seen in contact with the cadavers (target) and in the field of view, excluding the cadavers (arena) was determined every 2 min. To quantify attraction, a grasshopper present at the arena/target for 1 min was termed a grasshopper-minute; two grasshoppers at a site for 3 min would be scored as 6 grasshopper-minutes, etc. The differences in the proportion of grasshopper gut contents testing positive for chitin as a function of subfamily and developmental stage were analyzed using a chi-squared test. The effects of instar, adult age and sampling date on the rate of attraction of grasshoppers to the arena and target were analyzed by ANO VA and protected least significant difference tests. Adult ages consisted of early season (July), mid season (August) and late season (September) categories. The difference in the rate of attraction to the arena between the control and treatment sites at This content downloaded from 207.46.13.76 on Sat, 10 Sep 2016 05:47:23 UTC All use subject to http://about.jstor.org/terms JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY