RESEARCH COMPONENTS

Methodology for Montana State University Elk Monitoring Activities

Methodology for the prey base (primarily elk) studies will follow monitoring techniques employed over many years by FWP biologists and technicians. In addition, we will use research techniques developed by the MSU team and utilized for 10 years in the Madison-Firehole (MF) study and applied to the Gallatin Canyon (GC) and East Madison (EM) study sites during the pilot year of data collection during the winter of 2000-01. The primary focus will be on estimating the abundance of each elk population annually, describing patterns of distribution approximately weekly during the intensive winter-spring study periods, and bimonthly during the remainder of the year. In addition, the key vital rates of adult female survival, reproduction, and recruitment, important to understanding elk demography, will also be estimated or indexed annually.

We plan to deploy 40 VHF radio collars on cow elk in the GC during late fall - early winter 2001 and a similar number of collars on the EM site in 2002. Collars will be deployed by helicopter netgunning in the first year. For replacement of small numbers of GPS collars in subsequent years, we will use either netgunning or clover traps. Within the GC site, capture operations will be made in areas outside YNP (i.e. in the Taylor Fork, Teepee and Porcupine drainages). This effort would result in a sample of instrumented elk on each study site that would be comparable to the number of collared elk maintained on the MF study area. We will attempt to maintain a minimum annual instrumented sample of 30 cow elk per site throughout the study period. In the long term, radiotagged elk will provide data on annual rate of mortality, and on causes of death. In the short term, radiotags will provide detailed data on potential changes in the distribution, habitat selection and movements of elk in response to predation risk.

The radiotransmitters will be potted in tan epoxy and attached to brown collars to be as visually inconspicuous as possible, and all collars will be fitted with a drop-off mechanism programmed to release the collar at the end of its life. The radiotransmitters will be equipped with a motion-sensitive "mortality" sensor that allows monitoring for deaths without the need to visually locate each animal. Hence we will be able to quickly and conveniently monitor all instrumented animals on a daily basis for mortality throughout the winter. All mortalities will be investigated as soon as they are detected in order to determine cause of death based on condition of the carcass, presence of wounds and subcutaneous hemorrhaging, blood trails, signs of struggle, and evidence of predators such as tracks and scat. We do not anticipate needing to replace standard VHF collars due to battery depletion as the transmitters deployed will have a minimum life of 10 years.

The location of all instrumented animals will be determined at 3-5 day intervals from November through May using ground-based homing and triangulation techniques aided by hand-held telemetry equipment (White and Garrott 1990). In order to establish summer ranges for animals, we will use aircraft to monitor the spring migration (2 flights in months of May and June) of all instrumented animals during the first year each animal is monitored and conduct a single flight per month for the remainder of the year. In subsequent years will we will locate each instrumented animal monthly during the non-winter period to verify continued fidelity to summer ranges and monitor for survival. If the study continues for the anticipated 5 years, we will have obtained 150-175 animal years of survival and movement information for each study site.

The application of GPS technology for tracking detailed movement and distribution patterns of elk would provide similar advantages as those described for wolf monitoring. Currently, commercial GPS technology is quite expensive, costing $2000-3000 per collar. While this expense can be justified for the deployment of a few collars on each wolf pack, we do not anticipate having adequate resources during the initial phase of this study to purchase and deploy meaningful numbers of GPS collars on elk. Anticipating this limitation, the MSU research team has developed their own GPS units using approximately $300 in components. These inexpensive data collection units are designed to attach to standard VHF collars equipped with preprogrammed break-away links for collar retrieval. We propose to deploy 10 GPS-equipped VHF collars on elk during the 2001-02 field season to test their reliability and effectiveness. If these units perform as expected, we intend to increase the number of elk instrumented on each study site as data needs dictate and resources permit. GPS collars will be programmed to collect one fix every two hours, and will provide extensive, fine-grained data on the movement and distribution of elk in response to the risk of predation. We will deploy these collars initially in the GC site (in areas outside YNP), because the Chief Joseph pack does not use the site continuously, and predation risk will consequently vary on both long and short time scales.

Annual reproductive rates will be estimated by assaying progestogen (P₄) concentrations in fecal samples collected during late gestation (March-April). A study using radioimmunoassay procedures and 153 fecal samples collected from captive and free-ranging elk of known pregnancy status attained almost perfect discrimination (Garrott et al. 1998), hence, the procedure is considered very reliable and has been consistently applied in the MF study since its inception in 1991. Samples will be collected during routine ground monitoring of each elk population by locating feeding and/or bedding areas recently used by cow-calf elk groups and collecting 3-4 fecal pellets from individual pellet groups. Whenever possible, pellets will be collected from those fecal groups where sign (size of bed impression or tracks) suggests the fecal pellets were deposited by an adult animal. In order to assure that all fecal samples collected originated from adult females we will also explore the possibility of using several additional steroid assays to discriminate samples collected from males and young animals All fecal assay work will be performed in Scott Creel’s lab on the MSU campus with the proportion of samples containing progestogen concentrations indicative of pregnancy considered an estimate of annual reproductive rates for each study herd.

Stress hormone (glucocorticoid) levels will be determined from the same set of fecal samples used to determine pregnancy rates. A radioimmunoassay to measure glucocorticoids in elk feces has been fully validated, including ACTH challenge, and procedural and biological validations (Creel et al, in press).  The nutritional status of each elk herd will be assessed each winter by collecting snow-urine samples and assaying to determine allantoin:creatinine ratios. Bob Garrott and a number of his graduate students have developed and extensively tested this urinary metabolite index for elk using both controlled feeding experiments with captive animals as well as comparative studies using free-ranging elk (Garrott et al. 1996). We will attempt to collect a total of 20-25 snow-urine samples from cow-calf groups at two-week intervals throughout the winter-spring field season using protocols similar to those described for collecting fecal samples for determination of pregnancy rates. Samples will be assayed in Creel’s lab and analyzed as recommended by Pils et al. (1999) to estimate nutritional trends for each herd. This approach will permit comparison of the nutritional status of each herd across seasons, among years, and between herds to determine if overwinter nutrition has any affect on wolf prey selection and predation rates.

Recruitment rates will be estimated by monitoring changes in calf-cow ratios from mid-summer through April each year. FWP biologists will be responsible for conducting aerial age-sex composition surveys three times each year for the GC and EM elk herds using techniques similar to those applied when collecting these data during previous years. Attempts will be made to conduct the aerial surveys once during mid to late summer, once in early winter after the end of all hunting seasons, and once in the spring just prior to elk migrating off winter ranges. MSU research crews will also conduct ground-base composition surveys throughout the November to April field seasons to augment aerial survey data. During the winter 2000-01 pilot studies on the EM and GC study sites graduate students divided the study areas into units of appropriate size that could be routinely traveled at fixed intervals (1-3 weeks). During each ground survey attempts will be made to observe and classify all elk groups along the transects and within designated units. Ground-based surveys cannot cover all of the study areas in any one survey so sampling schemes will be developed prior to initiating the 2001-02 field seasons to assure adequate and representative coverage of all areas of each study site. Determination of calf:cow ratios for the MF study site will follow ground-based procedures developed in conjunction with an intensive relocation sampling scheme for instrumented elk. This methodology has been in place since the MF study was initiated and will provide estimates of calf:cow ratios in early October and monthly from December to April each year.

Fall or early winter calf:cow ratios for each elk herd will be compared to historic prewolf data and estimated parturition calf:cow ratios based on annual pregnancy rates determined from fecal steroid assays. These comparisons will provide estimates of calf mortality during the first six months of life and estimates of the contribution of wolf predation to summer-fall calf mortality. Comparisons of post-hunting season calf:cow ratios and April calf:cow ratios will provide estimates of overwinter calf mortality. April calf:cow ratios will be considered an index of annual recruitment and compared to historic prewolf data to estimate the contribution of wolf predation to overwinter calf mortality.

Population estimates will be obtained during mid-winter to spring each year by FWP biologists who have established standardized protocols for counting elk on both the GC and EM study sites. During the pilot field season on the EM site the MSU research team established a protocol for estimating the size of the wintering elk herd based on intensive ground counts that was very similar to the aerial count obtained by FWP personnel. The open grassland habitat and the tendency of the elk to be aggregated in large groups (300-1300) facilitate counting the elk on this study site. The ground-based population estimates on the EM site will be obtained monthly during the November-April field season to augment aerial survey data. Ground-based population estimates cannot be obtained in the GC study site due to a combination of small average group size, dispersed distribution, and extensive forest cover so all population estimates for this site will be obtained by FWP aerial surveys. On the MF study area elk congregate in the meadows that are generally associated with the road system during spring due to early melt-off of snowpack, and again in fall during harem formation and breeding. During both the spring and fall aggregation periods we conduct a minimum of 10 mark-recapture experiments by traversing the entire study area along the road system. These individual experiments are then combined to provide a point estimate, which has considerably smaller confidence intervals than the technique previously employed. This methodology was initially used to estimate size of the MF population during the winter of 1996-97, partially in response to the desire of YNP administrators to reduce research overflights within the Park. The technique has worked well over the past five years and will be continued throughout the duration of this study.