Aerial Ungulate Survey (2015) Moose and White-tailed Deer in Whitemud-Hotchkiss Rivers Wildlife Management Unit (527)

Transcription

1 Aerial Ungulate Survey (2015) Moose and White-tailed Deer in Whitemud-Hotchkiss Rivers Wildlife Management Unit (527) Hanna Neufeld and Luke Vander Vennen Alberta Environment and Parks Lower Peace Region Suggested Citation: Neufeld, H. and Vander Vennen, L Aerial Ungulate Survey (2015), Moose in WMU 527. Alberta Environment and Parks, Government of Alberta. Lower Peace Region, Peace River, Alberta.

2 EXECUTIVE SUMMARY An aerial ungulate survey was conducted in Wildlife Management Unit (WMU) 527 (Whitemud- Hotchkiss Rivers) from January 20-23, The survey was completed by Alberta Environment and Parks and Alberta Environmental Monitoring, Evaluation and Reporting Agency (AEMERA) staff using distance sampling methods to estimate a population density for the WMU. In total, 97 transects were surveyed totalling 891 kilometers. The encounter rate for moose was moose/km 2 (90% CI ), the estimated density was moose/km 2 (90% C.I ) and the estimated moose population across the WMU was 1922 (90% CI ). Of the 144 moose observed; 18% were bulls, 61% cows, and 21% calves. The encounter rate for white-tailed deer was deer/km 2 (90% C.I ). The estimated density for white-tailed deer was deer/km 2 (90% C.I ) and the estimated white-tailed deer population for the WMU was 1584 (90% C.I ). Key words: Alberta, aerial survey, ungulates, population estimates, density estimates, age/sex ratios, moose, deer, distance sampling.

3 1 INTRODUCTION 1.1 Background Every winter Alberta Environment and Parks (AEP; formerly Environment and Sustainable Resource Development) and Alberta Environmental Monitoring, Evaluation and Reporting Agency (AEMERA) complete between two and five aerial ungulate surveys on different Wildlife Management Units (WMU) within the specified Oil Sands Monitoring (OSM) region (Figure 1). During the winter of 2015, three WMU s were surveyed; 503, 519 and 527. Prior to January 2015, WMU 527 was most recently surveyed for ungulates in 2009 with only one other documented survey occurring before that in This unit s survey focused on moose and white-tailed deer populations. Historically, an adequate number of mule deer had been observed during the survey to complete a population analysis on the species. However, mule deer observations were too low to provide an accurate estimate of their current status. Current and past population data can be seen in Tables 1 to 3. Table 1 Historical and current data for Moose population estimates in WMU 527 MOOSE Year Pop. Estimate* Ratio to 100 Cows Density Bulls Calves sq. km (15.6%) (16.7%) (--) *Confidence limit of survey in brackets Table 2 Historical and current data for White-tailed Deer population estimates in WMU 527 White-tailed Deer Year Pop. Estimate* Ratio to 100 Cows Density Bulls Calves sq. km (31.4%) (22.2%) (--) *Confidence limit of survey in brackets Table 3 Historical and current data for Mule Deer population estimates in WMU 527 Mule Deer Year Pop. Estimate* Ratio to 100 Cows Density Bulls Calves sq. km 2015 N/A N/A (17.5%) (--) *Confidence limit of survey in brackets

4 Figure 1 WMU s that are included in the OSM study area are outlined in black

5 1.2 Survey objectives The objective of this survey was to provide an update on moose and deer populations in WMU 527. WMU 527 is located in the Upper Peace region of the Alberta and falls within the OSM boundaries. WMU s within the OSM region are targeted to be surveyed at least once every 5 years, with 527 most recently being surveyed in The purpose of the OSM initiative is to improve population estimates and the accuracy of ungulate trends by increasing survey frequency as oil and gas expansion continues in the area. Additionally, aerial ungulate surveys are completed to provide both decision makers and the public with an understanding of ungulate population size and composition (i.e. gender and age). Area Wildlife Biologists rely on this data considerably when determining appropriate hunting regulations for upcoming seasons. 2 METHODS 2.1 Study area WMU 527 is bordered by the 25th base line (northern boundary of Township 96) to the north, the Peace River to the east, Highway 684 and Highway 35 to the southwest, and bordered on the west by the TransCanada Peace Pipeline from its intersection with the Whitemud River to its intersection with the 25th base line. This WMU is relatively large at 6,767 km 2 and dominated by settlement and agriculture, with only 37% in the Green Zone. There are two main communities (Manning and Peace River), several smaller communities (including Dixonville, North Star, Notikewin and Deadwood), acreage developments near Peace River, and farms throughout this WMU. This WMU is dissected by tributaries to the Peace River, including the Whitemud, Notikewin, Buchanan, Hotchkiss and Meikle rivers. These river valley complexes provide escape and wintering habitat for all ungulate species, while the agricultural areas provide feeding areas (Moyles, D., 2009).The dominant tree species in WMU 527 are white and black spruce, balsam poplar, aspen and birch. The west side of highways 35 in the southern section of the WMU overlaps slightly with the Chinchaga woodland caribou range.

6 Figure 2 Map of the Whitemud-Hotchkiss Rivers Wildlife Management Unit No Survey methods Two Bell Jet Ranger helicopters were used for this survey, each with 3 experienced observers and a pilot. Transects were flown at approximately 300 feet above ground level at 80 knots. Both machines were flight followed using the Peace River Fire Centre where staff their aircraft location every 30 minutes. Crew positioning (front left, back left or back right) and crew manifest were also reported to the flight followers daily. Weather conditions including temperature, precipitation, cloud cover and wind were documented prior to beginning the survey each day. This survey was conducted using distance sampling methods. Transects between 2 and 10 km were flown while observers scanned the landscape for animals. Transects were generated by constructing a grid of rectangles (w=1.2km, l=10km) oriented north-south. A random sample of rectangles was selected and the centreline of selected rectangles were used as transects. Transects

7 measuring less than 2 km in length were not included. Out of 907 possible transects, 97 were randomly selected and sampled. The rear observers were responsible for observing the area beyond 50 meters on their respective sides of the aircraft. The front left observer restricted his vision to include the area directly below the aircraft and 50 meters on either side. When a group of animals were detected on transect, it was documented by taking a waypoint immediately after it was observed (line waypoint). The helicopter remained on transect until it became perpendicular with the group. It then left transect to collect a second waypoint (group waypoint) over the location the group was originally observed at. The composition of the group was documented at this time, which included age class, gender, antler class and activity of animal when observed (standing, bedded, moving). When groups were comprised of more than one animal, the group waypoint was taken at the center point between animals. Any animals within 30 meters of the group waypoint would be considered a part of the same group. In addition to animal observations, covariates such as crown closure, light intensity, snow cover and terrain were documented. 2.3 Analysis Techniques Survey data were analyzed using the program Distance 6.0 (Release 2.0; Thomas et al., 2010). Through this program, the data were fit to five different candidate models (half-normal + cosine, halfnormal + hermite, uniform + cosine, uniform + polynomial and hazard + cosine), with the half normal + cosine model being the default. All models were evaluated, however, if models other than the default did not improve the fit or precision of the detection function, or change the estimate, the default model was selected. The fit of the data to a particular model was assessed using various goodness of fit tests and model selection was based on Akaike s Information Criterion. Data were truncated or binned to improve the model fit if necessary (Buckland et al., 2001, Burnham and Anderson, 2002). When appropriate, we used size-biased regression to estimate cluster. We considered the effect of including stratification in the density estimation by estimating strata specific encounter rates. The detection function and cluster size were estimated at the regional level. 3 RESULTS A total of 97 transects (figure 5) were flown on January 20 and 23, 2015, totalling 891 km, and 23.3 flying hours. The survey was delayed on January 21 and 22 due to high temperatures and winds. In total, 144 moose were observed and successfully classified from 105 independent groups; 88 (61%) of these were cows, 30 (21%) were calves and 26 (18%) were bulls. In addition to moose, 90 white- tailed deer were observed in 27 separate groups; 61 were classified to be 26 (42%) does, 23 (38%) fawns and 12 (20%) bucks. Mule deer were also observed and classified, however too few were documented to run an effective analysis on the species (70 total, 54 classified). The estimated density for moose resulted in moose/km 2 (90% C.I ) and the estimated population across the WMU was 1922 (90% CI ). The estimated density for white-tailed deer was deer/km 2 (90% C.I )

8 and the estimated population for the WMU was 1584 (90% C.I ). The default model using the half normal key and cosine adjustment terms was chosen for both species based on model fit and simplicity (Figures 3 and 4). In both cases, the model fit improved when data were truncated at a distance that excluded outliers; moose data were truncated at 700 meters and white-tailed data at 400 meters. Stratifying for boreal and agriculture areas was considered impractical without larger sample sizes of observed groups in the boreal regions. Figure 3: Distance probability vs. perpendicular distance in meters for Moose in WMU 527 using the best fit model (half normal with cosine adjustment terms, truncated at 700 meters)

9 Figure 4: Distance probability vs. perpendicular distance in meters for White-tailed Deer in WMU 527 using the best fit model (half normal with cosine adjustment terms, truncated at 400 meters) Tables 4 & 5: Parameter estimate results of the five candidate models for both species sampled. Confidence limits are based on a 90% confidence interval. Model names reflect the detection function model keys (hn = half normal, uni = uniform, haz = hazard), adjustment terms (cos = cosine, poly = simple polynomial, herm = hermite) and associated levels (T = truncated) for each model. The asterisk (*) denotes the model chosen for analysis. Table 4: White-tailed deer density and population estimates Model Density Population CV Delta k Estimate LCL UCL Estimate LCL UCL AIC (/km2) default *Hn(cos)T Hn(herm)T Uni(cos)T Haz(cos)T Uni(poly)T

10 Table 5: Moose density and population estimates Model Density Population CV Delta k Estimate LCL UCL Estimate LCL UCL AIC (/km2) default *Hn(cos)T Hn(herm)T Uni(cos)T Haz(cos)T Uni(poly)T Figure 5: Ninety-seven transects were flown in WMU 527 January 20 th and 23 rd, 2015.

11 Figure 6: Moose and white-tailed deer (WTDE) population estimates for WMU s surrounding 527. Table 6: A summary of the most recent Moose and White-tailed Deer data available for the WMU s surrounding 527. WMU Year Moose WTDE Population Estimate Density Estimate (moose/km 2 ) Population Estimate Density Estimate (deer/km 2 ) (17.9%) (19%) (11.8%) (19.2%) (13.6%) (16.5%) (39.1%) (15.6%) (31.4%) 0.23

12 4 DISCUSSION AND MANAGEMENT RECOMMENDATIONS Moose and deer population estimates from this survey were lower than the previous survey, conducted in A series of consecutive mild winters in the mid 2000 s precipitated a large increase in ungulate abundance in WMU 527 and surrounding areas, and the higher population estimate in 2009 was likely partially attributable to that population increase. Since 2009, combinations of factors have likely contributed to the population decline. Winter conditions have not been as mild, therefore acting as a more restrictive population limitation factor than in the mid 2000 s. The winter of 2013 was particularly severe for ungulate populations, as cold weather and deep snow depths decreased deer and moose survival and subsequent recruitment of young into the population the following year. In conjunction with changes in winter conditions, areas of important moose and deer habitat have become increasingly more accessible to hunters, leading to increased hunting pressure. When taken in conjunction with ongoing land clearing activities on private land, since the previous 2009 survey ungulate populations have faced more typical winter conditions, been subject to increasing hunter access, and undergone loss of habitat patches within the white zone of the WMU. Wolf populations are now well established in WMU 527 and have impacts on moose populations in both forested and agricultural areas. These factors are likely all influential in causing a population decrease since the time of very high ungulate densities in the 2000 s. These decreases in moose abundance align with available information from hunters, landowners, and recreationalists. Wildlife managers have adjusted harvest allotments in accordance with the population declines to allow for continued harvest of this population. So while lower abundances of moose were not unexpected, the magnitude of the change was greater than anticipated. Tag allotments for the 2015 moose hunting season will therefore be decreased from 2014 in order to maintain a healthy and sustainable population, while still providing hunting opportunities for Albertans. 5 ACKNOWLEDGEMENTS Surveys were conducted by the following AEP, AEMERA and Justice and Solicitor General Employees: Dave Moyles, Luke Vander-Vennen, Hanna Neufeld, Scott Donker, Andrew Baron, Erik Graf Von Platen and Agnieszka Sztaba. The pilots for this survey were Darvin Mossing with Black Swan Helicopters and Phil Clay with Highland Helicopters. Nikki McKenzie from the Peace River Fire Centre provided flight following. Finally we would like to thank Hannah McKenzie for all her insight into the finer points of Distance. 6 LITERATURE CITED

13 Buckland, S. T., Anderson, D. R., Burnham, K.P., Laake, J.L Distance Sampling: Estimating Abundance of Biological Populations. Environment and Sustainable Resource Development Aerial Ungulate Survey Protocol Manual. Natural Regions Committee Natural Regions and Subregions of Alberta. Compiled by D.J. Downing and W.W. Pettapiece. Government of Alberta. Pub. No. T/852. Moyles, D., Pages In: N. Webb and R. Anderson, editors. Delegated aerial ungulate surveys, 2008/2009 survey season. Data Report, D , produced by the Alberta Conservation Association, Rocky Mountain House, Alberta, Canada. Thomas, L., S.T. Buckland, E.A. Rexstad, J.L Laake, S. Strindberg, S.L Hedley, J. R, B. Bishop, T. A. Marques, and K.P Burnham Distance software: design and analysis of distance sampling surveys for estimating population size. Journal of Applied Ecology 47: DOI: /j x 7 APPENDICES 7.1 Logistics Summary WMU 527 (Whitemud-Hotchkiss Rivers) Dates of survey January 20 th and 23 rd, 2015 Observers Dave Moyles, Scott Donker, Hanna Neufeld, Agnieszka Sztaba, Luke Vander Vennen, Andrew Baron, Erik Graf Von Platen Aircraft 206 Bell Jet Ranger (Highland Helicopters and Black Swan Helicopters) Pilot(s) Phil Clay and Darvin Mossing Cost and time breakdown Total hours flown: 23.3 Total cost: $28, Design Distance sampling design with 907 North-South transects generated 1.2km apart. 7.2 Raw Data Moose Stratum Stratum Area (km 2 ) Transect ID Transect Length (km) Perpendicular Distance Moose Group Size WMU WMU WMU

14 WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU

15 WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU

16 WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU

17 WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU Total Survey Area (km 2 ) Total transect length surveyed (km) Total number of Moose groups observed Total number of moose observed White-tailed Deer Stratum name Stratum Area (km 2 ) Transect ID Transect Length (km) WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU Perpendicular distance WTDE Group Size

18 WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU

19 WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU

20 WMU WMU WMU WMU Total Survey Area (km 2 ) Total transect length surveyed (km) Total number of WTDE groups observed Total number of white-tailed deer observed Mule Deer Stratum Stratum Area (km 2 ) Transect ID Transect Length (km) WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU Perpendicular Distance Mule Deer Group Size

21 WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU

22 WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU WMU

23 Total Survey Area (km 2 ) Total transect length surveyed (km) Total number of MUDE groups observed Total number of MUDE observed 7.3 Summary of Distance Results for Moose Density Estimate Results Effort : # samples : 97 Width : # observations: 104 Model Selection Results Model 1 Half-normal key, k(y) = Exp(-y**2/(2*A(1)**2)) Point Standard Percent Coef. 90% Percent Parameter Estimate Error of Variation Confidence Interval DS E E(S) E D E N Measurement Units Density: Numbers/Sq. kilometers ESW: meters Component Percentages of Var(D) Detection probability : 18.6 Encounter rate : 75.6 Cluster size : Estimation Summary Encounter Rates Estimate %CV df 90% Confidence Interval n

24 k L n/l E Left Width Detection Probability Estimate %CV df 90% Confidence Interval Half-normal/Cosine m LnL AIC AICc BIC Chi-p f(0) E E E-02 p ESW Expected Cluster Size Estimate %CV df 90% Confidence Interval Average cluster size Half-normal/Cosine r E-01 r-p E(S) Density and Abundance Estimate %CV df 90% Confidence Interval Half-normal/Cosine DS D N Goodness of fit tests Kolmogorov-Smirnov test D_n = p =

25 Cramer-von Mises family tests W-sq (uniform weighting) = < p <= Relevant critical values: W-sq crit(alpha=0.600) = W-sq crit(alpha=0.500) = C-sq (cosine weighting) = < p <= Relevant critical values: C-sq crit(alpha=0.400) = C-sq crit(alpha=0.300) = Chi-square Goodness of Fit test Total Chi-square value = Degrees of Freedom = 4.00 Probability of a greater chi-square value, P = Qq-plot

26 7.4 Summary of Distance Results for White-tailed Deer Density Estimate Results Effort : # samples : 97 Width : # observations: 26 Model Selection Model 1 Half-normal key, k(y) = Exp(-y**2/(2*A(1)**2)) Point Standard Percent Coef. 90% Percent Parameter Estimate Error of Variation Confidence Interval DS E E E E(S) D E N Measurement Units Density: Numbers/Sq. kilometers ESW: meters Component Percentages of Var(D) Detection probability : 26.3 Encounter rate : 60.9 Cluster size : Goodness of fit tests Kolmogorov-Smirnov test D_n = p = Cramer-von Mises family tests W-sq (uniform weighting) = < p <= Relevant critical values: W-sq crit(alpha=0.150) =

27 W-sq crit(alpha=0.100) = C-sq (cosine weighting) = < p <= Relevant critical values: C-sq crit(alpha=0.100) = C-sq crit(alpha=0.050) = Chi-square Goodness of Fit test Total Chi-square value = Degrees of Freedom = 1.00 Probability of a greater chi-square value, P = Estimation Summary Encounter rates Estimate %CV df 90% Confidence Interval n k L n/l E E E-01 Left Width Detection probability Estimate %CV df 90% Confidence Interval Half-normal/Cosine m LnL AIC AICc BIC Chi-p E-01 f(0) E E E-02 p ESW Expected Cluster Size Estimate %CV df 90% Confidence Interval

28 Average cluster size Half-normal/Cosine r r-p E(S) Density and Abundance Estimate %CV df 90% Confidence Interval Half-normal/Cosine DS E E D N Qq-plot