ALCES VOL. 44, 2008 YOUNG AND BOERTJE – RECOVERY OF BULL:COW RATIOS 65 RECOVERY OF LOW BULL:COW RATIOS OF MOOSE IN INTERIOR ALASKA Donald D. Young Jr. and Rodney D. Boertje Alaska Department of Fish and Game, 1300 College Road, Fairbanks, AK 99701-1599, USA, e-mail: don.young@alaska.gov ABSTRACT: During 1996–1999, hunters killed an estimated 24–30% of the pre-hunt bull moose (Alces alces) in Game Management Unit 20A. As a result, the 1999 post-hunt bull:cow ratios declined to 24:100, well below the management objective of 30:100. During 2000 and 2001 we shortened the hunting season from 25 to 20 days to reduce the harvest of bull moose, but kill rates of bulls remained high (23–27%) and ratios remained unacceptably low (22–26 bulls:100 cows). Subsequently, to recover bull:cow ratios to 30:100, hunters were restricted unit-wide to taking bulls with 1) spike-fork antlers, 2) antlers ≥50 inches wide, or 3) ≥3 brow tines on ≥1 antler. These restrictions were in place from 2002- 2007, but results occurred rapidly. After only 2 years of antler restrictions, hunters killed an average of 36% fewer bulls compared with the previous 2-year average harvest rate ( x = 715 during 2000–2001 and 455 during 2002–2003). Comparing these same 2-year periods, average kill rates of bulls declined from 25% to 12% of the pre-hunt bull population, average number of hunters declined 24% (1,568 to 1,187), and the average hunter success rate declined from 34% to 29%. Bull:cow ratios increased from 26:100 to 32:100 after 2 years of antler restrictions. With an additional 2 years (2004–2005) of antler restrictions and high harvest of cow moose, bull:cow ratios reached 38:100. Modeling indicated that the bull:cow ratio would have stabilized at 33:100 without the high harvest of cows. The recovery of bull:cow ratios to our objective of 30:100 with 2 years of antler restrictions allowed 1) bull seasons to be lengthened from 20 to 25 days beginning in 2004 and, 2) a limited number of drawing permits for any bull during 2006–2007. Elsewhere, similar selective harvest strategies should also allow recovery of bull:cow ratios, unless the total kill rate of bulls is higher than estimated here. ALCES VOL. 44: 65-71 (2008) Key words: Alaska, Alces alces, antler restrictions, bull:cow ratios, Game Management Unit 20A, recovery, selective harvest strategy. Historically, low bull:cow ratios (10:100) of moose (Alces alces) resulted where hunter access was good, and hunting was restricted largely to any bull (Rausch et al. 1974). Low bull:cow ratios did not occur in remote inac- cessible areas (60–80:100, Gardner 2002), so we infer that humans, not wolves (Canis lupus) and bears (Ursus americanus and Ursus arctos), caused the skewed ratios by selec- tively killing bull rather than cow moose. To recover low bull:cow ratios, selective harvest strategies (SHS) were first implemented in British Columbia in 1980 (Child 1983, Child and Aitken 1989). These SHS were based on limiting hunters to particular bulls with regula- tions on antler architecture. In Alaska, SHS were first initiated on the Kenai Peninsula in 1987 (Schwartz et al. 1992). Selective har- vest strategies spread rapidly to other Alaska roadside areas during 1988–1993 because of low bull:cow ratios. This is the first paper that describes suc- cessful recovery of bull:cow ratios in interior Alaska. High kill rates of bull moose by hunters in Game Management Unit 20A (Unit 20A) during 1996–1999 resulted in post-hunt bull:cow ratios declining below the manage- ment objective of 30:100. During 2000–2001 RECOVERY OF BULL:COW RATIOS – YOUNG AND BOERTJE ALCES VOL. 44, 2008 66 we shortened the hunting season to reduce the harvest of bull moose, but bull:cow ratios remained low. In 2002, hunters were restricted unit-wide to taking bulls with 1) spike-fork antlers, 2) antlers ≥50 inches wide, or 3) ≥3 brow tines on ≥1 antler. We report the effects of these antler-based SHS on 1) total kill of bull moose by hunters, 2) kill rate of pre-hunt bull moose numbers, 3) hunter participation, 4) hunter success rate, 5) antler size of harvested bulls, and 6) bull:cow ratios. STUDY AREA Our study area encompassed Unit 20A immediately south of Fairbanks and across the Tanana River. The study area is in interior Alaska and is centered on 64°10′N latitude and 147°45′W longitude. The study area en- compasses 17,601 km2, but only 13,044 km2 contains topography and vegetation used char- acteristically by moose. Gasaway et al. (1983), Boertje et al. (1996), and Keech et al. (2000) described the physiography, habitat, climate, and factors limiting moose during 1963–1997. Young and Boertje (2004) described hunter access, moose seasons, and bag limits from the 1960s through the early 2000s, moose popula- tion status from 1997-2003, and the use of calf hunts to increase yield. Young et al. (2006) described impediments and achievements of managing Unit 20A moose for elevated yield. Moose in Unit 20A (1997–2005) exhibited the lowest nutritional status documented for non- insular, wild moose in North America (Boertje et al. 2007). Boertje et al. (2009) described how predation and reproduction affected the harvest of moose during 1996–2007. METHODS Estimating Harvest and Hunter Statistics We monitored reported moose harvest, hunter participation, hunter success rates, and antler characteristics of the harvest us- ing a mandatory harvest report card system (Schwartz et al. 1992). To estimate the number of moose killed by hunters, we multiplied the reported harvest by 1.34 (47/35) based on the reported harvest of 35 radio-collared moose and an additional 12 radio-collared moose that died from unreported harvest (6) and wound- ing loss (6) in Unit 20A during 1996–2006 (Boertje et al. 2009). We focused this study on the kill of bull moose; bulls included all males ≥15 months of age. We derived the pre-hunt bull kill rate based on pre-hunt bull numbers (Gasaway et al. 1992). We estimated the pre- hunt bull numbers each year by adding the total estimated kill of bulls by hunters, which occurred in September, to the aerial estimates of November bull numbers. We derived hunter success rate using the reported harvest. We primarily compared results from the 2 years prior to SHS (2000 and 2001) versus the first 2 years of SHS (2002–2003) because those 4 years had 20-day hunting seasons, whereas all other years (1996–1999 and 2004–2005) had 25-day seasons. We used Student’s t-tests to compare total kill, kill rate, hunter participation, hunter success rate among years, and antler size in the harvest. We assumed significance at α = 0.05. Estimating Moose Population Character- istics We flew moose population estimation and composition surveys (1999–2005) each November, except in 2002 due to poor survey conditions. To calculate moose numbers, we used spatial statistics (Kellie and DeLong 2006, Ver Hoef 2008) and a sightability cor- rection factor of 1.21 (Boertje et al. 2009). To estimate the finite annual population growth rate (λ), we fitted population estimates dur- ing 1996–2004 with a trendline through parametric empirical Bayes estimates (Ver Hoef 1996:1048). We used the Student’s t- test to compare mean bull:cow ratios prior to and after the initiation of SHS. We assumed significance at α = 0.05. ALCES VOL. 44, 2008 YOUNG AND BOERTJE – RECOVERY OF BULL:COW RATIOS 67 Modeling to Assess Effect of Cow Moose Hunts on Bull:Cow Ratios We used a deterministic model using Microsoft® Office Excel 2003® software to compare bull:cow ratios with and without harvests of cow moose. Cows included all females ≥15 months of age. Input variables in- cluded annual estimates of the number of bulls and cows killed by hunters, annual survival rates of bulls and cows, and calves:100 cows observed during November surveys. Also, during the initial year we input the number of bulls, cows, and calves. We backdated the model to the year 1991 and varied survival rates until output values were similar to estimates obtained from annual surveys. Output values included annual post-hunt moose numbers, bull:cow ratios, and λ. In the initial model we used empirical estimates of 98, 161, 698, and 794 cow moose killed by hunters during 2002–2005. We then ran the same model using a simulated adult kill of 10 cows (i.e., estimated illegal kill) during 2002–2005. The result was paired bull:cow ratios with and without harvests of cows during the 4 years of unit-wide SHS. RESULTS Harvest Statistics The mean estimated kill of bull moose by hunters declined 36% after SHS were initi- ated (Table 1). Specifically, the mean kill of 455 bulls taken during 2002–2003 was lower (P = 0.003, df = 2) than the mean kill of 715 bulls taken during the 2 years prior to SHS (2000–2001). Mean kill of bulls by hunters was not lower during 2004–2005 ( x = 577, P = 0.079, df = 2) when the SHS season was lengthened from 20 to 25 days. The mean estimated kill rate of 12% of the pre-hunt bull population during 2002–2003 was lower (P = 0.029, df = 2) than the 2-year mean 25% kill rate observed prior to SHS (Table 1). Kill rates remained lower ( x = 14%, P = 0.026, df = 2) after the SHS season was lengthened from 20 to 25 days. Hunter Participation and Success The mean reported number of hunters declined (24%, P = 0.002, df = 2) after SHS were initiated (Table 1). Mean reported hunter success rates also declined (P = 0.034, df = 2) from 34% to 29% after SHS were initi- ated. Success rates remained lower (x = 25%, P = 0.004, df = 2) after the SHS season was lengthened from 20 to 25 days. Antler Size of Harvested Bulls We observed a shift away from 30- to 40-inch antlered bulls and toward ≥50-inch antlered bulls in the harvest after SHS were implemented (Table 1). The mean propor- tion of 30- to 40-inch antlered bulls declined (P = 0.019, df = 2) in the harvest from 30% to 6%. Conversely, the mean proportion of ≥50-inch antlered bulls increased (P = 0.023, df = 2) from 26% to 49% in the har- vest. These trends continued into 2004–2005 when the SHS season was extended 5 days (30- to 40-inch antlered bulls: x = 7%, P = 0.019, df = 2; ≥50-inch antlered bulls: x = 52%, P = 0.009, df = 2). We observed no change in the proportions of <30-inch ( x = 22% vs. 23%, P = 0.607, df = 2) or 40- to 50-inch antlered bulls (x = 21% vs. 19%, P = 0.417, df = 2) in the harvest. Bull:Cow Ratio of the Population The mean bull:cow ratio increased (P = 0.006, df = 4) by 11 bulls:100 cows after SHS were implemented when comparing data from 1999–2001 (x = 24) versus 2003–2005 (x = 35, Table 1). However, SHS alone did not increase the bull:cow ratio because high harvests of cows during 2004–2005 also con- tributed to the increase in the bull:cow ratio (Fig. 1). Modeling indicated that without the harvest of cows, the bull:100 cow ratio would have increased to only 33 in 2005 rather than the 38 observed. Regardless, when comparing the observed bull:cow ratios pre-SHS (x =24) versus the simulated bull:cow ratios without the harvests of cows (x = 33, RECOVERY OF BULL:COW RATIOS – YOUNG AND BOERTJE ALCES VOL. 44, 2008 68 R ep or te d ha rv es t r at e (% ) b y an tle r w id th (i nc he s) Y ea r Se as on le ng th / B ag lim ita N o. b ul ls : 10 0 co w s 90 % C I Pr e- hu nt b ul l po pu la tio nb R ep or te d ha rv es t To ta l k ill o f bu lls c Pr e- hu nt k ill ra te (% ) o f bu lls R ep or te d no . of h un te rs R ep or te d su cc es s ra te d (% ) <3 0 30 –4 0 40 –5 0 50 + U nk 19 96 25 d ay /A ny bu ll 39 3, 28 9 60 4 80 9 24 .6 1, 63 6 36 .9 23 29 21 25 3 19 97 25 d ay /A ny bu ll 33 3, 45 4 62 0 83 1 24 1, 59 5 38 .9 22 28 25 22 4 19 98 25 d ay /A ny bu ll 31 2, 99 5 60 8 81 5 27 .2 1, 65 2 36 .8 17 28 23 27 5 19 99 25 d ay /A ny bu ll 24 17 –3 1 3, 01 0 66 9 89 7 29 .8 1, 57 4 42 .5 19 24 24 30 3 20 00 20 d ay /A ny bu ll 22 15 –2 9 2, 66 8 53 4 71 5 26 .8 1, 58 4 33 .7 23 27 22 26 2 20 01 20 d ay /A ny bu ll 26 18 –3 4 3, 06 9 53 4 71 5 23 .3 1, 55 1 34 .4 21 33 20 26 1 20 02 20 d ay S- F/ 50 –e –e 3, 41 4f 35 0 46 9 13 .7 1, 18 5 29 .5 25 5 17 52 1 20 03 20 d ay S- F/ 50 32 25 –3 9 3, 97 5 32 8 44 0 11 .1 1, 18 9 27 .6 21 7 21 45 6 20 04 25 d ay S- F/ 50 35 27 –4 3 3, 90 1 40 0 53 6 13 .7 1, 63 8 24 .4 16 7 25 49 3 20 05 25 d ay S- F/ 50 38 33 –4 2 4, 25 2 46 1 61 8 14 .5 1, 81 6 25 .4 11 6 27 54 3 a “ A ny b ul l” b ag li m it oc cu rr ed in 1 4, 32 5 km 2 ( 81 % ) o f t he 1 7, 60 1 km 2 i n G am e M an ag em en t U ni t 2 0A . W e de si gn at ed u ni tw id e se le ct iv e ha rv es t s tr at e- gi es a s “S -F /5 0” , w hi ch in cl ud ed re st ri ct in g ha rv es t t o bu lls w ith 1 ) s pi ke -f or k an tle rs , 2 ) a nt le rs ≥ 50 in ch es w id e, o r 3 ) ≥ 3 br ow ti ne s on ≥ 1 an tle r. b P re -h un t b ul l p op ul at io n eq ua le d nu m be r o f b ul ls e st im at ed fr om N ov em be r s ur ve ys p lu s es tim at ed to ta l k ill o f b ul ls b y hu nt er s. c T ot al k ill o f b ul ls e qu al ed re po rt ed h ar ve st o f b ul ls x 1 .3 4, b as ed o n a ra di o- co lla re d sa m pl e of m oo se k ill ed b y hu nt er s (4 7/ 35 ). d R ep or te d su cc es s ra te e qu al ed re po rt ed h ar ve st d iv id ed b y re po rt ed n um be r o f h un te rs . e N o da ta . f D at a on p os t- hu nt b ul l n um be rs w er e in te rp ol at ed fr om a dj ac en t y ea rs . Ta bl e 1. C ha ra ct er is tic s of th e bu ll m oo se (≥ 15 m on th s of a ge ) p op ul at io n, re po rt ed h ar ve st a nd k ill o f b ul ls b y hu nt er s, a nd h un tin g re gu la tio ns fo r b ul ls in G am e M an ag em en t U ni t 2 0A , I nt er io r A la sk a, 1 99 6– 20 05 . ALCES VOL. 44, 2008 YOUNG AND BOERTJE – RECOVERY OF BULL:COW RATIOS 69 Fig. 1), bull:cow ratios remained different (P = 0.002, df = 4). DISCUSSION Using antler-based SHS, bull:cow ratios increased from about 26 bulls:100 cows to above our objective of 30 bulls:100 cows in 2 years, 2002–2003. We documented that SHS reduced kill and pre-hunt kill rates of bulls, hunter participation, and hunter success rates. After the initiation of SHS, bull:cow ratios increased to a 3-year average of 35 bulls:100 cows, but the increase above 33 bulls:100 cows apparently resulted from liberal harvests of cows during the last 2 years (Fig. 1). When comparing 3-year averages pre- and post-SHS, bull:cow ratios increased from 24 to 35 with ratios as high as 38. With SHS on the Kenai Peninsula, average bull:cow ratios increased from 16 to 25 with ratios as high as 29 (Schwartz et al. 1992). Schwartz et al. (1992) reported a 27% decline in the mean harvest of bull moose when comparing 5-year periods pre- (x = 636) and post-SHS (x = 466); whereas, we observed a 36% decline in the mean harvest of bulls when comparing 2-year periods pre- (2000–2001) and post-SHS (2002–2003). After SHS were implemented on the Kenai Peninsula, the distribution of the harvest of bulls shifted toward yearlings and away from 30- to 40-inch antler class with little change in the proportion of bulls ≥50-inches (Schwartz et al. 1992). We also observed a shift away from bulls in the 30- to 40-inch antler class. However, we also observed a shift toward bulls with ≥50-inch antlers, presumably be- cause our higher bull:cow ratio had resulted in a larger proportion of large antlered bulls in the population. Unlike the Kenai study, we observed little change in the proportion of ≤30-inch bulls (i.e., yearlings). We speculate that hunters had difficulty identifying yearling bulls because of poor antler development in Unit 20A. For example, during a flight prior to the hunt in 2007, we observed 6 (22%) of 27 known-age, radio-collared yearlings with antlers consisting of small spikes (2–8 cm). Poor antler development among yearlings was likely related to the low nutritional status of this population (Boertje et al. 2007). The rapid recovery of bull:cow ratios allowed us to elevate the harvest of bulls after 2 years. Recovery of bull:cow ratios to 30:100 was a secondary management objec- tive during this study. The primary objective was to improve nutritional status of moose by encouraging greater harvest (Boertje et al. 2007). Therefore, during 2004–2007, when the post-hunt bull:cow ratios exceeded our management objective of 30 bulls:100 cows, we encouraged greater harvest of bulls using 2 methods. First, 5 days were added to the hunting season with SHS during 2004–2005. Subsequently, we issued a limited number of drawing permits for any bull during 2006–2007 (Schwartz et al. 1992). We recognize that increasing harvest of cows, not bulls, was the most effective method to decrease population size, but hunts targeting cows were difficult to implement and far more controversial than hunts targeting bulls (Young and Boertje 2004, Young et al. 2006). We know that low bull:cow ratios result from hunters favoring bulls versus cows, based on data from favored roadside hunt- ing areas (10 bulls:100 cows, Rausch et al. 1974). In contrast, in remote, lightly hunted areas, we observed 60–80 bulls:100 cows. 29.5 33.0 29.3 32.3 33.3 33.5 36.1 38.0 20 25 30 35 40 2002 2003 2004 2005 B ul ls :1 00 C ow s With Cow Harvests Without Cow Harvests Fig. 1. Estimated post-hunt bull:cow ratios with and without cow moose harvests, Game Management Unit 20A, Interior Alaska, 2002–2005. RECOVERY OF BULL:COW RATIOS – YOUNG AND BOERTJE ALCES VOL. 44, 2008 70 We demonstrate that SHS were a key factor allowing bull:cow ratios to increase from 26:100 to 32:100 in 2 years. Bull:cow ratios should increase elsewhere where similar SHS are implemented, unless kill rates of bulls are higher than estimated here. ACKNOWLEDGEMENTS Many individuals contributed to the work reported in this manuscript and we thank them. Those deserving special thanks include skilled pilots, generous volunteers, and dedi- cated Alaska Department of Fish and Game (ADF&G) wildlife biologists and technicians that flew many arduous hours conducting moose survey flights. We particularly wish to thank ADF&G Publications Technician Laura McCarthy for her timely and much appreciated technical assistance. This study was funded by ADF&G and Federal Aid in Wildlife Restoration. REFERENCES Boertje, r. D., M. A. Keech, D. D. Young, K. A. Kellie, and c. t. SeAton. 2009. Managing for elevated yield of moose in Interior Alaska. Journal of Wildlife Management 73(3): in press. _____, K. A. Kellie, c. t. SeAton, M. A. Keech, D. D. Young, B. W. DAle, l. g. ADAMS, and A. r. ADerMAn. 2007. Rank- ing Alaska moose nutrition: signals to begin liberal antlerless harvests. Journal of Wildlife Management 71: 1494–1506. _____, P. VAlKenBurg, and M. McnAY. 1996. Increases in moose, caribou, and wolves following wolf control in Alaska. Journal of Wildlife Management 60: 474–489. chilD, K. n. 1983. Selective harvest of moose in Omineca: some preliminary results. Alces 19: 162–177. _____, and D. A. AitKen. 1989. Selective harvests, hunters and moose in central British Columbia. Alces 25: 81–97. gArDner, c. l. 2002. Unit 20E moose. Pages 406–429 in C. Healy, editor. Moose management report of survey and inven- tory activities 1 July 1999–30 June 2001. Project 1.0. Juneau. Alaska, USA. (ac- cessed February 2008). gASAWAY, W. c., r. D. Boertje, D. V. grAn- gAArD, D. g. KelleYhouSe, r. o. Ste- PhenSon, and D. g. lArSen. 1992. The role of predation in limiting moose at low densities in Alaska and Yukon and implications for conservation. Wildlife Monographs 120: 1–59. _____, r. o StePhenSon, j. l. DAViS, P. e. K. ShePherD, and o. e. BurriS. 1983. Interrelationships of wolves, prey, and man in Interior Alaska. Wildlife Mono- graphs 84: 1-50. Keech, M. A., r. t. BoWYer, j. M. Ver hoef, r. D. Boertje, B. W. DAle, and t. r. StePhenSon. 2000. Life-history conse- quences of maternal condition in Alaskan moose. Journal of Wildlife Management 64: 450–462. Kellie, K. A., and r. A. Delong. 2006. Geo- spatial survey operations manual. Alaska Department of Fish and Game, Fairbanks, Alaska, USA. rAuSch, R. A., R. J. SoMerVille, and R. H. BiShoP. 1974. Moose management in Alaska. Naturaliste Canadien 101: 705–721. SchWArtz, c. c., K. j. hunDertMArK, and t. h. SPrAKer. 1992. An evaluation of selective bull moose harvest on the Kenai Peninsula. Alces 28: 1–13. Ver hoef, j. M. 1996. Parametric em- pirical Bayes methods for ecological applications. Ecological Applications 6: 1047–1055. _____. 2008. Spatial methods for plot- based sampling of wildlife populations. Environmental and Ecological Statistics 15: 3–13. Young, D. D., jr., and r. D. Boertje. 2004. Initial use of moose calf hunts to increase ALCES VOL. 44, 2008 YOUNG AND BOERTJE – RECOVERY OF BULL:COW RATIOS 71 yield, Alaska. Alces 40: 1–6. _____, _____, c. t. SeAton, and K. A. Kellie. 2006. Intensive management of moose at high density: impediments, achieve- ments, and recommendations. Alces 42: 41–48.