Mark S. Udevitz

Application of Airborne Thermal Imagery to Surveys of Pacific Walrus

Abstract We conducted tests of airborne thermal imagery of Pacific walrus to determine if this technology can be used to detect walrus groups on sea ice and estimate the number of walruses present in each group. In April 2002 we collected thermal imagery of 37 walrus groups in the Bering Sea at spatial resolutions ranging from 1–4 m. We also collected high-resolution digital aerial photographs of the same groups. Walruses were considerably warmer than the background environment of ice, snow, and seawater and were easily detected in thermal imagery. We found a significant linear relation between walrus group size and the amount of heat measured by the thermal sensor at all 4 spatial resolutions tested. This relation can be used in a double-sampling framework to estimate total walrus numbers from a thermal survey of a sample of units within an area and photographs from a subsample of the thermally detected groups. Previous methods used in visual aerial surveys of Pacific walrus have sampled only a small percentage of available habitat, resulting in population estimates with low precision. Results of this study indicate that an aerial survey using a thermal sensor can cover as much as 4 times the area per hour of flight time with greater reliability than visual observation.

Evaluation of aerial survey methods for Dall's sheep

Abstract Most Dalls sheep (Ovis dalli dalli) population-monitoring efforts use intensive aerial surveys with no attempt to estimate variance or adjust for potential sightability bias. We used radiocollared sheep to assess factors that could affect sightability of Dalls sheep in standard fixed-wing and helicopter surveys and to evaluate feasibility of methods that might account for sightability bias. Work was conducted in conjunction with annual aerial surveys of Dalls sheep in the western Baird Mountains, Alaska, USA, in 2000–2003. Overall sightability was relatively high compared with other aerial wildlife surveys, with 88% of the available, marked sheep detected in our fixed-wing surveys. Total counts from helicopter surveys were not consistently larger than counts from fixed-wing surveys of the same units, and detection probabilities did not differ for the 2 aircraft types. Our results suggest that total counts from helicopter surveys cannot be used to obtain reliable estimates of detection probabilities for fixed-wing surveys. Groups containing radiocollared sheep often changed in size and composition before they could be observed by a second crew in units that were double-surveyed. Double-observer methods that require determination of which groups were detected by each observer will be infeasible unless survey procedures can be modified so that groups remain more stable between observations. Mean group sizes increased during our study period, and our logistic regression sightability model indicated that detection probabilities increased with group size. Mark–resight estimates of annual population sizes were similar to sightability-model estimates, and confidence intervals overlapped broadly. We recommend the sightability-model approach as the most effective and feasible of the alternatives we considered for monitoring Dalls sheep populations.

Potential population-level effects of increased haulout-related mortality of Pacific walrus calves

Availability of summer sea ice has been decreasing in the Chukchi Sea during recent decades, and increasing numbers of Pacific walruses have begun using coastal haulouts in late summer during years when sea ice retreats beyond the continental shelf. Calves and yearlings are particularly susceptible to being crushed during disturbance events that cause the herd to panic and stampede at these large haulouts, but the potential population-level effects of this mortality are unknown. We used recent harvest data, along with previous assumptions about demographic parameters for this population, to estimate female population size and structure in 2009 and project these numbers forward using a range of assumptions about future harvests and haulout-related mortality that might result from increased use of coastal haulouts during late summer. We found that if demographic parameters were held constant, the levels of harvest that occurred during 1990–2008 would have allowed the population to grow during that period. Our projections indicate, however, that an increase in haulout-related mortality affecting only calves has a greater effect on the population than an equivalent increase in harvest-related mortality distributed among all age classes. Therefore, disturbance-related mortality of calves at coastal haulouts may have relatively important population consequences.

Estimating survival rates with time series of standing age-structure data.

It has long been recognized that age-structure data contain useful information for assessing the status and dynamics of wildlife populations. For example, age-specific survival rates can be estimated with just a single sample from the age distribution of a stable, stationary population. For a population that is not stable, age-specific survival rates can be estimated using techniques such as inverse methods that combine time series of age-structure data with other demographic data. However, estimation of survival rates using these methods typically requires numerical optimization, a relatively long time series of data, and smoothing or other constraints to provide useful estimates. We developed general models for possibly unstable populations that combine time series of age-structure data with other demographic data to provide explicit maximum likelihood estimators of age-specific survival rates with as few as two years of data. As an example, we applied these methods to estimate survival rates for female bison (Bison bison) in Yellowstone National Park, USA. This approach provides a simple tool for monitoring survival rates based on age-structure data.

Energy demands for maintenance, growth, pregnancy, and lactation of female Pacific walruses ( Odobenus rosmarus divergens )

Decreases in sea ice have altered habitat use and activity patterns of female Pacific walruses Odobenus rosmarus divergens and could affect their energetic demands, reproductive success, and population status. However, a lack of physiological data from walruses has hampered efforts to develop the bioenergetics models required for fully understanding potential population-level impacts. We analyzed long-term longitudinal data sets of caloric consumption and body mass from nine female Pacific walruses housed at six aquaria using a hierarchical Bayesian approach to quantify relative energetic demands for maintenance, growth, pregnancy, and lactation. By examining body mass fluctuations in response to food consumption, the model explicitly uncoupled caloric demand from caloric intake. This is important for pinnipeds because they sequester and deplete large quantities of lipids throughout their lifetimes. Model outputs were scaled to account for activity levels typical of free-ranging Pacific walruses, averaging 83% of the time active in water and 17% of the time hauled-out resting. Estimated caloric requirements ranged from 26,900 kcal d−1 for 2-yr-olds to 93,370 kcal d−1 for simultaneously lactating and pregnant walruses. Daily consumption requirements were higher for pregnancy than lactation, reflecting energetic demands of increasing body size and lipid deposition during pregnancy. Although walruses forage during lactation, fat sequestered during pregnancy sustained 27% of caloric requirements during the first month of lactation, suggesting that walruses use a mixed strategy of capital and income breeding. Ultimately, this model will aid in our understanding of the energetic and population consequences of sea ice loss.

Adaptation of a stage-projection model for species with multiple year reproductive cycles

Abstract We apply stage projection matrices to multiple age and reproductive categories, deriving the stable age—reproductive category distribution for populations with multi-year reproductive cycles and identifying conditions for the existence of a dominant latent root. Given the stable distribution of age and reproductive category and the number of female young for each reproductive category, the age specific fecundity rates can be calculated. For species with multi-year reproductive cycles, the fecundity rate associated with the stable age distribution is often not a smooth function of age, but fluctuates as a larger or smaller fraction of females becomes available to breed. We suggest that fecundity rates for species with multi-year reproductive cycles be defined in terms of litter size and probabilities of conception and successful gestation rather than average fecundity

Estimating Age Ratios and Size of Pacific Walrus Herds on Coastal Haulouts using Video Imaging

During Arctic summers, sea ice provides resting habitat for Pacific walruses as it drifts over foraging areas in the eastern Chukchi Sea. Climate-driven reductions in sea ice have recently created ice-free conditions in the Chukchi Sea by late summer causing walruses to rest at coastal haulouts along the Chukotka and Alaska coasts, which provides an opportunity to study walruses at relatively accessible locations. Walrus age can be determined from the ratio of tusk length to snout dimensions. We evaluated use of images obtained from a gyro-stabilized video system mounted on a helicopter flying at high altitudes (to avoid disturbance) to classify the sex and age of walruses hauled out on Alaska beaches in 2010–2011. We were able to classify 95% of randomly selected individuals to either an 8- or 3-category age class, and we found measurement-based age classifications were more repeatable than visual classifications when using images presenting the correct head profile. Herd density at coastal haulouts averaged 0.88 walruses/m2 (std. err. = 0.02), herd size ranged from 8,300 to 19,400 (CV 0.03–0.06) and we documented ∼30,000 animals along ∼1 km of beach in 2011. Within the herds, dependent walruses (0–2 yr-olds) tended to be located closer to water, and this tendency became more pronounced as the herd spent more time on the beach. Therefore, unbiased estimation of herd age-ratios will require a sampling design that allows for spatial and temporal structuring. In addition, randomly sampling walruses available at the edge of the herd for other purposes (e.g., tagging, biopsying) will not sample walruses with an age structure representative of the herd. Sea ice losses are projected to continue, and population age structure data collected with aerial videography at coastal haulouts may provide demographic information vital to ongoing efforts to understand effects of climate change on this species.

An improved procedure for detection and enumeration of walrus signatures in airborne thermal imagery

In recent years, application of remote sensing to marine mammal surveys has been a promising area of investigation for wildlife managers and researchers. In April 2006, the United States and Russia conducted an aerial survey of Pacific walrus (Odobenus rosmarus divergens) using thermal infrared sensors to detect groups of animals resting on pack ice in the Bering Sea. The goal of this survey was to estimate the size of the Pacific walrus population. An initial analysis of the U.S. data using previouslyestablished methods resulted in lower detectability of walrus groups in the imagery and higher variability in calibration models than was expected based on pilot studies. This paper describes an improved procedure for detection and enumeration of walrus groups in airborne thermal imagery. Thermal images were first subdivided into smaller 200 � 200 pixel ‘‘tiles.’’ We calculated three statistics to represent characteristics of walrus signatures from the temperature histogram for each tile. Tiles that exhibited one or more of these characteristics were examined further to determine if walrus signatures were present. We used cluster analysis on tiles that contained walrus signatures to determine which pixels belonged to each group. We then calculated a thermal index value for each walrus group in the imagery and used generalized linear models to estimate detection functions (the probability of a group having a positive index value) and calibration functions (the size of a group as a function of its index value) based on counts from matched digital aerial photographs. The new method described here improved our ability to detect walrus groups at both 2 m and 4 m spatial resolution. In addition, the resulting calibration models have lower variance than the original method. We anticipate that the use of this new procedure will greatly improve the quality of the population estimate derived from these data. This procedure may also have broader applicability to thermal infrared surveys of other wildlife species.

Walrus haul-out and in water activity levels relative to sea ice availability in the Chukchi Sea

An animals energetic costs are dependent on the amount of time it allocates to various behavioral activities. For Arctic pinnipeds, the time allocated to active and resting behaviors could change with future reductions in sea ice cover and longer periods of open water. The Pacific walrus (Odobenus rosmarus divergens) is a large Arctic pinniped that rests on sea ice or land between foraging trips to feed on the seafloor. We used behavioral data collected from radiotagged walruses in the Chukchi Sea (2008–2014) in a Bayesian generalized linear mixed effects model to estimate the probability a walrus was in water foraging, in water not foraging, or hauled out, as a function of environmental covariates. The probability of a walrus being in water increased with wind speed and decreased with air temperature, and the probability a walrus was foraging, given it was in water, increased with available benthic macrofaunal biomass. The probability of each behavior was also related to the nature and availability of haul-out substrates. The amount of time walruses spent in water foraging and hauled out was greatest when only sea ice was available, which typically occurs when walruses occupy feeding areas during summer and early autumn. This situation may be most energy efficient for walruses because it allows the highest proportion of in water energy expenditure to be allocated to foraging. Conversely, the amount of time walruses spent in water foraging and hauled out was lowest when only land was available, which typically occurs in late autumn, in years when walruses were constrained to land haul-outs because sea ice was absent over the continental shelf.

Evaluation of a method using survey counts and tag data to estimate the number of Pacific walruses (Odobenus rosmarus divergens) using a coastal haulout in northwestern Alaska

Increased periods of sparse sea ice over the continental shelf of the Chukchi Sea in late summer have reduced offshore haulout habitat for Pacific walruses (Odobenus rosmarus divergens) and increased opportunities for human activities in the region. Knowing how many walruses could be affected by human activities would be useful to conservation decisions. Currently, there are no adequate estimates of walrus abundance in the northeastern Chukchi Sea during summer–early autumn. Estimating abundance in autumn might be possible from coastal surveys of hauled out walruses during periods when offshore sea ice is unavailable to walruses. We evaluated methods to estimate the size of the walrus population that was using a haulout on the coast of northwestern Alaska in autumn by using aerial photography to count the number of hauled out walruses (herd size) and data from 37 tagged walruses to estimate availability (proportion of population hauled out). We used two methods to estimate availability, direct proportions of hauled out tagged walruses and smoothed proportions using local polynomial regression. Point estimates of herd size (4200–38,000 walruses) and total population size (76,000–287,000 walruses) ranged widely among days and between the two methods of estimating availability. Estimates of population size were influenced most by variation in estimates of availability. Coastal surveys might be improved most by counting walruses when the greatest numbers are hauled out, thereby reducing the influence of availability on population size estimates. The chance of collecting data during peak haulout periods would be improved by conducting multiple surveys.