This research has provided a long overdue test of a hypothesis, increased our quantitative knowledge of carrion community structure, provided insight into the influences of Nearctic, large vertebrates on biological diversity, and provided faunistic information on one of the least known and more ecologically important and diverse taxa within Yellowstone National Park. Clearly, the null hypothesis of no carcass effect has been shown to be false. Given this first step, two avenues of future research are more firmly rooted in fact. The first, which has attracted and continues to attract the most attention from researchers is the controlled, quantitative documentation below the community level--incorporating individual species' life histories to provide a predictive understanding of carrion ecology, such as is common in forensic entomology. By using techniques as employed for this study, researchers can quantitatively identify those species that are carrion associated and measure the strength of those associations.

The second avenue, which currently receives less attention, is the focus on the larger patterns of biogeography and biodiversity effects of carrion ecology at the ecosystem level. Through the analysis of large scale patterns of diverse communities via long-range studies, information valuable to ecosystem management can be obtained. In recent years, there has been a shift from traditional anthrocentric management of natural resources to whole-ecosystem management practices (Glick et al. 1991). This shift incorporates the acknowledgment that all species in an ecosystem are deserving of attention, not merely the larger charismatic or commercially important taxa (loc. cit.). Land managers should consider that the removal of large herds of ungulates from an area likely affects myriad species. Due to the large number of species that are associated with ungulate carcasses, ungulates might be keystone species for these communities. Small parks and preserves that cannot support herds of ungulates, in areas where they were once common, may have experienced numerous, unrecorded, local population extinctions of invertebrates as a result. A total food-web analysis might show that insectivorous vertebrate species are dependent to some degree on the invertebrates of these carrion communities and that they too could suffer from a loss of plentiful carrion.

I hope that this research represents an important step towards understanding carrion-beetle community ecology in Yellowstone National Park and may, at the very least, reduce our ignorance of how large vertebrates contribute to biological diversity within YNP and other Nearctic ecosystems.

To obtain a thorough understanding of the importance of large carcasses to the Yellowstone Ecosystem a considerable amount of research remains to be done. The current project could be expanded and continued to achieve two goals: (1) an almost complete survey of the beetle communities in the Lamar Valley and (2) a thorough faunistic component. The data presented herein indicate that the fauna was not sampled to completion, i.e. species remain uncollected. This is a common problem in insect community ecology due to the great diversity of species. However, within the Lamar Valley it may be possible to obtain a thorough sample. With the use of different trap types and many more trap units, it may be possible to rapidly assess the entire fauna, if the trapping effort is great enough. The knowledge of all species associated with the Lamar Valley region and their consequent association (or lack thereof) with large carcass will provide an understanding of nature that is rarely achieved by the study of the community ecology of diverse insect taxa. With enough trapping effort it should be possible to ascertain the difference between those species whose reproductive success is associated with elements of the Lamar Valley and those that are merely passing through the area--knowledge that is currently lacking.

In addition to an increase in trapping effort, a corresponding and equivalent increase in faunistic research could yield data of importance to management. The identification of every species by trained taxonomists would create the groundwork from which a standardized faunistic report could be produced. This report would incorporate all known information of each species, such as distribution, rarity, trophic level, whether the species is a North American native or was introduced, etc. and the potential exists for the later creation of a means to identify these species. Either through the use of computerized expert-systems technology or by the creation of a traditional published key, a means to identify the local beetle fauna, usable by any personnel, could be of great importance to management. Such a thorough identification of all the species would discover rare, unusual or undescribed species, such as the aberrant Borboropora quadriceps listed within the findings of this study.

However, many habitats of the Yellowstone ecosystem are not found within the Lamar Valley. A more general sampling, using a similar methodology, conducted in each of the representative habitat types and elevation gradient zones of the park would give a more complete understanding of the ecological importance of large carcasses. There exists a relict population of a rare species of silphid, Thanatophilus coloradensis, on the Beartooth Plateau. It may be possible that this carrion feeder will appear in samples taken in alpine habitats.

An important consideration is the costs in effort and money input into such a project compared to the corresponding information obtained. My thesis research indicates that using large carcasses can be very beneficial in this regard. The use of these carcasses essentially doubles the number of species captured per unit of trapping effort. It should be noted that the lab work involved in sorting the samples may be disproportionately greater for carcass samples due to the abundance of the carrion-associated individuals. In addition to the use of carcasses, costs per species documented can be reduced by sampling qualitatively rather than quantitatively. Once the Lamar Valley fauna is known quantitatively, the remainder of the park may best be sampled without counting every individual beetle, but rather by simply recording each new species encountered in the samples. By combining the carcass and qualitative sampling methods, with enough trapping effort, the beetle fauna of the entire park may be sampled in a way that maximizes the information obtained while minimizing the effort input.

Although the ecological diversity of beetles is great enough that they are one of the best choices as an insect order surrogate for the entire ecosystem, information on other groups could be incorporated. One simple study to obtain information on the entire arthropod fauna would be to take random samples prior to the sorting of the beetles and simply count and identify all the arthropods within each sample. By using many independent traps in a variety of habitats it may be possible to discover a reliable relationship between the beetle diversity and the total arthropod diversity within the park. If this method was expanded and refined, it may be possible to identify for each habitat type, trap type, etc. a corresponding ratio of beetles to arthropods. From these data it would then be relatively straightforward to make estimates of the diversity of the entire fauna of a habitat by rapidly and inexpensively documenting the beetle fauna. For example, if the ratio of beetles to arthropods in four pit-fall traps next to carcasses during June in Subalpine fir/Oregon Grape habitat averaged 20% and 200 beetle species were captured then one could estimate a total arthropod fauna of 1000 species for that habitat at that time. These ratios could periodically be verified if a monitoring program was in place, by again counting all the arthropods within samples. It is currently unknown how constant these ratios might be between trap types, habitat types and over time.

I would also recommend the use of multiple bait types for a variety of traps. A carefully planned dung-baiting program would discover another aspect of the relationships between large vertebrates and the diversity of species with which they are associated. With dung, as opposed to carrion, it has been documented that there are many more specialist species that exploit the dung of only one species or type of vertebrate (e.g. there are beetles species that specialize on deer dung as opposed to cattle or bison dung).

The use of baits in general insect faunistics has many advantages, particularly so if used in concert with unbaited traps. The principal advantage is, as already mentioned the increase in species obtained per unit effort invested. However, in addition, there remains the advantage that the species shown to be associated with that bait type logically should have some degree of their reproductive success correlated with that resource type. This conclusion results from the argument that if individuals are trapped more frequently in traps near the bait then they must have actively moved towards the bait and they must have some means of sensing where the bait is located. A mechanism to detect the location of the bait and the fact that the individuals expend energy to locate the bait all point towards the conclusion that they must obtain some advantage in terms of fitness from doing so. Although from the data obtained one can not determine how much of a species' reproductive success is correlated with the bait type, at least it would be documented that there exists some relationship. Because so little is known about the life-history and ecological relationships of most insect species, the little ecological information obtained from baited trapping would represent the first ecological information known. In fact, there are entire families of insects that are equally poorly known, such as the Scraptiidae. The little information presented herein on a species of Scraptiid associated with carcasses was entirely unexpected and may represent an important step forward in understanding the biology of that family.

Once the baseline faunal information is known, the next most obvious step would be the undertaking of a total food-web analysis. Such a project would clearly be complex but would be useful to identify those species that use carcasses as a reproductive resource, those that are predaceous exclusively at carcasses, those that are opportunistic, etc. The identification of all the associated species from mites to birds and the documentation of their behaviors would be difficult, but would greatly increase our understanding of the ecological importance of large vertebrates.