. I
( 7 Population and reproductive trends of Nene Branta sandvicensis in Hawaii Volcanoes National Park,1989-1999
E.H. Rave1, A. Cooper1,2, D. Hu3, R. Swift4,5 & K. Misajon4
1Department of Biology, Bemidji State University, Bemidji, MN 56601 USA. Email: [email protected] 2Biology Department, Minnesota State University Moorhead, Moorhead, MN 56563, USA. 3Pacific West Regional Office Honolulu, P.O. Box 52, Hawaii National Park, HI 96718, USA. 'Hawaii Volcanoes National Park, P.O. Box 52, Hawaii National Park, HI 96718, USA. 5Pacific Cooperative Studies Unit, University of Hawaii, P.O. Box 52, Hawaii National Park, HI 96718, USA.
The Hawaiian Goose or Nene Branta sandvicensis is an endangered species endemic to Hawaii. Hawaii Volcanoes National Park contains about 41% of Nene on the island of Hawaii. Population numbers, reproductive success and causes of known adult mortality of Nene in the park were summarised from 1989 to 1999 in order to determine population and reproductive trends. Reproductive data were also pooled across years to determine overall reproductive success at several sites within the park. Numbers in the park generally increased throughout ( the decade, from 153 birds in 1990/1991 to 203 in 1997/1998. Annual hatching success averaged 55%, whereas annual fledgling production averaged 11 birds. Fledgling production varied among sites, from Oat Mauna Loa and Kipuka Kahalii to 32 at Pen 7. The most common cause of known adult mortality was road kill. Population numbers increased throughout the decade probably, in part, because of intensive predator control. However, recruitment remained low because of starvation, dehydration and predation on goslings. Nene at certain sites within the park had high reproductive success, perhaps because those sites contained more nutritious food for adults and goslings or had supplemental food and water provided to the birds. Managers need to continue to control predators and improve or restore habitat to maximise Nene population numbers.
Key Words: Branta sandvicensis, Hawaiian Goose, Nene, population trends, reproductive success
© Wildfowl & Wetlands Trust Wildfowl [2005) 55: 7-16 I
8 Reproductive trends of Nene
The Hawaiian Goose or Nene Branta throughout the 1990s, population sandvicensis is an endangered species estimates, reproductive success and endemic to Hawaii. Nene were causes of known adult mortality of perilously close to extinction in 1950, Nene were summarised from 1989 when only 17 wild birds were known to to 1999. It is hoped that any trends exist, all on the island of Hawaii !Elder & observed will assist park personnel in Woodside 19581. To help increase Nene managing Nene better in HVNP. numbers, the state of Hawaii initiated a captive colony in 1949, with the goal Methods of re-establishing self-sustaining wild populations. Beginning in 1960, captive This study was conducted in HVNP, a birds were released at several sites 930-km2 park located on the island of on Hawaii, including Hawaii Volcanoes Hawaii (Figure 11. From 1989 to 1999, National Park !HVNPI !Kear & Berger yearly numbers of Nene in HVNP were 19801. Captive birds have continued to determined. Because Nene breed on be released at HVNP throughout the a declining day length !Kear & Berger past 45 years. Nene currently found 19801, and thus begin to breed in in HVNP descend from these captive October, yearsweredefinedas 1 October released birds, as well as from the to 30 September. During each year, last remaining wild Nene found on the all ringed wild and captive-released island in the 1950s. birds were counted. !Wild birds were Today, about 41% of the island's caught and ringed opportunistically, ( Nene population resides in and around and captive birds were ringed prior to HVNP (U.S. Fish & Wildlife Service their release.I Unringed birds were 20041. Management practices such as counted by their affiliation with their predator control, habitat restoration ringed mates, as Nene usually pair for and reducing the birds' contact life !Black et al. 19961. Unringed birds with humans have concentrated on without ringed mates were counted enhancing reproductive success on the one day each year when the and survival of wild Nene. In order greatest number of unringed birds to determine whether management was seen. These measures ensured practices are working, Nene numbers that a high proportion of the population in HVNP have been surveyed. Surveys was counted. Captive birds located in conducted in the 1970s showed a HVNP were also counted because they substantial decrease in Nene numbers contributed offspring to the park's Nene in HVNP, from 300 birds in 1975 to 50 in numbers. All ringed birds counted in a 1979 !W.S. Devick, Hawaii Department given year were also included in the of Land and Natural Resources, estimate for the following year, even unpublished data!. However, numbers if they were not sighted the following increased throughout the 1980s, from year. However, if two years elapsed 75 birds in 1981 to 179 birds in 1988 without sighting a specific bird, it was !Hoshide et al. 19901. considered dead and was removed To determine numbers and from that year's estimate. Additionally, reproductive trends of Nene in HVNP if a bird was known to have died during ( ;----. � I •
� z -n a, u::i-· z -.....·C Island of Hawaii (I) 0 -, �. � "'"" C1) "U • -,a, r 0 -;,:;-n - a, :::t. I 0 a, ::, ,· :E 0 , -a, , :-· ,.....- (f)(I) ::, "" / ;:::;.: - (I) ..... '<'�-.,,' - -,' (/1 0 .,, -, (/1 -...::.. ,, (I) ..... --- J ,.-' .,,"/r 3' ':l , -cc.. /' -. nro '< ,. t ... ,,- ,, \.,-...... r\ ..... (/1 I ::, -· ,;' CraterRim \ \ ro Park Boundary , (/1 '°(/1 D ... , /',,, Roads 9�- Road ) of Chain §"· :E ., 1, lO ;:::;.: a-aters ,, Road a, ::, NeneAreas 5''..''\.. , 8 I ::, 3· "-" I 7 ,,, 4 ,-,� Cl. a, 1 Mauna Loa C" ::, j-l 6 '-�---:;=,, 2 Kau Desert ., cl Cl. .:f-,/' ,, 0 a, 3 Kilauea /2 ,...... ,, Cl. Cl. / ::;·'iii·n 4 Ainahou ,' lO (I) 5 Ainahou Big Pen -,a, ;;o (I) � CD / ..... "tJ-, 6 Pen? / � 0 0 C. 7 KipukaNene ::,-I C ,,,, a, !:l. 8 KipukaKahalii -· :.:· � :E CD 0 a, 9 UppperChain of - -, -, · CD 5 0 5 10 Kilometers n· ::, Craters Road < C. a, 0 "' 10 Coast �n e. a, C z + ::, CD (/1 0 ::, (I) CD Cl. � -0 10 Reproductive trends of Nene ( a given year, it was not counted in that enclosures that were built to enhance year's estimate. reproductive success of Nene nesting Similar to the methods of Hoshide et within them. Supplemental food and al. (19901, the number of breeding pairs water are provided to Nene inside the and the number of nesting attempts of pens, and a fence and a perimeter wild and captive-released Nene were trapline protect nests and birds determined each year in HVNP by from introduced predators such as counting [11 number of known nests, Indian Mongoose Herpestes javanicus, (21 birds associated with goslings or rats Rattus rattus, R. exulans and R. newly-fledged young, and (31 females norvegicus, Feral Cat Felis catus and that were gravid [based on abdominal Feral Pig Sus scrofa. Although Ainahou profile, Zillich & Black 20021 or had Big Pen (constructed in 1994) is still brood patches. Yearly egg production in use, Pen 7 was built at a time when and hatching success were determined little was known about appropriate by finding nests and counting the habitat for Nene. Consequently, it was number of eggs laid and the number of constructed in sub-optimal habitat and eggs that hatched. Because nests were used by nesting Nene only through the often difficult to find, egg and hatching 1996/1997 breeding season. data from known nests were a subset Additionally, causes of known adult of total nesting attempts. Fledgling mortality were summarised across all production was determined by years of the study. following the fate of all known goslings, regardless of whether the nests from Results ( which they came were found. Reproductive data across all years Numbers of Nene generally increased were pooled to determine overall throughout the study, from 153 reproductive success for each of 10 birds in 1990/1991 to 203 birds in sites within and adjacent to HVNP. The 1997/1998. However, numbers declined chosen sites reflected nesting and approximately 10% during the last year brooding locations historically used by of the study (Table 1). Nene. Sites were [11 Mauna Loa (2,012 Reproductive success fluctuated ml. (21 Kau Desert [726-914 ml. [3) greatly throughout the study. The Kilauea (1,220 ml. (41 Ainahou [726- percentage of the population that 838 ml. (5) Ainahou Big Pen [838 ml. bred each year varied from 26% to [6) Pen 7 [729 ml. (7) Kipuka Nene [900 46% (Table 11. with an overall mean of ml. [8) Kipuka Kahalii [853-945 ml. (9) 35%. The number of nesting attempts Upper Chain of Craters Road (31-1,128 varied from 22 to 48, with some pairs ml, which includes all non-coastal attempting more than one nest per areas along the road, and [101 Coast year [Table 1). Of those nests that were [<31 ml. which includes the coastal found, the mean clutch size was 2.7 region of Chain of Craters Road and all eggs. Yearly hatching success of known other regions along the coast [Figure nests varied from 44% to 77% (Table 11. Ainahou Big Pen [5 ha) and Pen 7 11; of these hatchlings, only 30%, on (0.5 ha) are both fenced, open-topped average, fledged [range = 0-50%). ( /""\. � I -
� a. ,, er a.1 -u z -u G) rn < ::J --t IDc lDO
12 Reproductive trends of Nene
Yearly fledgling production varied from birds were counted at Ainahou (W.S. 0 to 18 birds (Table 11, with an annual Devick, Hawaii Department of Land mean of just 11. Of those fledglings that and Natural Resources, unpublished were ringed and whose fate could be data!. By 1988, Nene numbered 179 followed, 42% eventually attempted birds, an increase of 258%; however, to breed. Thus, overall recruitment of much of this increase resulted from a Nene into the population was generally high number of captive birds that were low throughout the study. released during the 1980s (Black et Some sites in HVNP were used more al. 1991; Hoshide et al. 19901. Captive frequently than others by Nene during releases declined to fewer than 60 the reproductive season. Based on the birds throughout the 1990s. Over number of breeding pairs and nesting 80% of these releases occurred in attempts, the most frequently used the first five years of the decade (U.S. sites were Kipuka Nene, Kau Desert, Fish & Wildlife Service 20041, a time Kilauea and Ainahou (Table 21. Although when population numbers remained Kipuka Nene had more breeding pairs relatively stable. Although these birds and nesting attempts than any other may have contributed to some of the site, only 16 birds fledged from the site observed population increase during throughout the 10 -year study period the last half of this study, an important (Table 21. The highest number of birds contributing factor was probably fledged from Pen 7 (321. even though intensive predator control. In 1995, this site had less than half the number a large-scale predator control effort of nesting attempts as Kipuka Nene using trapping and the anticoagulant (Table 21 and was used only through diphacinone began. Thereafter, the 1996/1997 breeding season. numbers of Nene increased until Forty-nine known adult Nene are 1998. In 1998/1999, however, numbers known to have died during this study. declined by 10%, perhaps due to the Of those, 30 birds (61 %) were killed by drought that affected HVNP during the vehicles, seven (14%1 were depreciated last two years of this study. Although (two by Feral Cats, one by an Indian the population decline was within an Mongoose, one by either a feral cat or expected range of yearly variation for dog, and three by unknown predators!. Nene (U.S. Fish & Wildlife Service six (12%1 were struck by golf balls, one 20041, it is nevertheless problematic (2%1 was caught in a small mammal in an endangered population with such trap, one (2%1 died after ingesting low recruitment rates. fertiliser, and four (8%1 died of unknown The mean percent of the population causes. that bred each year was only 35%. This low number may be caused by Discussion low food availability, which results in insufficient body reserves for egg Numbers of Nene in HVNP generally production by females (Banko 19921. increased throughout the 10 years of Hu (19981 concluded that a larger the study. Indeed, numbers have been proportion of females must breed increasing since 1979, when only 50 each year to establish self-sustaining I .
0. f"I Cl" GI -u z G) rn -0 :::r --1 <1> c OlCl Q) Q) Ill Breeding Nesting Eggs Eggs Goslings -, ,.... CT ;:: 3 �� a b A° n ,... 11) er - :::r C1) ::, 11) lCl3· OJ� Site pairs attempts laida hatched fledged ru°'(/)0: Q)I�N • � � ;::!: QJ Mauna Loa 4 4 8 0 0 :E: Q) z o ; [ � -Q) C. ::, C � Cl) - · ; 0 lCl c d -· lCl ::E 0.. [3l [Ol lC 3 .... 0 O" � � 3· � Kau Desert 61 64 41 25 28 --0 l/l (0 ... ron::rOJ co ..- ., ::Tu, i:: [161 [61 I --0 -· l/l ....I ::, � in" g- n lC 0 ci a. � Kilauea 49 50 68 37 10 --0 l/l - �,_ C QJ a. n :::JOJ ,_ru [28 ::,- 0.. -, 1 [54] �;:::!:• (0 z 11) • C. (0 0.. 47 47 40 23 17 ::, a; Ainahou (0 --, 3 0 � 0 [161 [571 C. 3 OJ 3 .,O" -:,,. Ainahou Big Pen 9 11 28 8 2 -., (0 ::, ::,; s 0 0 (0 ::E0 0 ::E [1 OJ [29] 3 !?: ::, ::E ::, ::, 32 34 84 51 32 (0 lC ::, lCl::, CD ::, Pen 7 Q) 11) 0 (/l (/l (/l - n -o [261 161 I Q) (/l =: !" :::r ::, l/l ., lCl Kipuka Nene 74 75 75 41 16 l/l =II: (/l .....-· - 0 (0 ::, [31] [55] 11) lCl -::,- Kipuka Kahalii 9 9 6 4 0 :E: � lCl CD .....-· l/l..... (/l -c [2] :::r -· ::,- 0 [67] -· ::, OJ -c ::, lC ;:; C Upper Chain of Craters Road 4 4 3 3 1 ::,- ii, I a, CD - 11 I [1001 Q) ..... 0.. 0 -... ::, :E: ::0 =II: Coast 5 5 13 9 7 '°
( Reproductive trends of Nene 15 could have been due to frequent Ainahou Big Pen was built in 1994, and human disturbance, as there had been perhaps Nene will use it more in the a public campground located within future after females that fledge from Kipuka Nene. To reduce this potential the pen attain breeding age [females problem, the campground was closed generally breed close to their natal permanently in 1999. sites, Stone et al. 1983a]. The number of goslings that fledged While high gosling mortality is a from Kau Desert was relatively high serious problem and one that needs (281. probably because this area to be rectified before the population contains a large pasture where goslings can increase [Hu 19981, the influence frequently forage. Indeed, Banko (1988] of adult mortality on population size found . that gosling survivorship is in HYNP is poorly understood. Of the positively correlated with the nearness known causes of adult mortality, road of nests to grass pastures. Additionally, kill was the most common, a result Kau Desert has a relatively low density also reported by Hoshide et al. [19901. of mongooses !Hoshide et al. 1990; Vehicles killed 30 adult Nene in the Stone et al. 1983bl, further contributing 1990s, a large number for a small to the survivorship of goslings. population, especially if some of those Kilauea had a high number of killed are breeding. Indeed, one female nesting attempts [50]; however, gosling killed by a vehicle had an egg in her mortality was high, with only 10 birds reproductive tract IHYNP, unpublished fledged from the area. Though Nene data]. The number of Nene hit by cars are not discouraged from nesting here, could probably be reduced if more broods are sometimes translocated park visitors obeyed the posted speed to more suitable habitat, where there limits, noted the "Nene Crossing" signs is better forage and less disturbance posted at various locations in HYNP, from people. Unfortunately, long and did not feed Nene, which attracts distance translocations may not always them to roads and parking areas. be advisable because they can destroy Continued management is pair bonds and disrupt family groups necessary to increase population IHYNP, unpublished data]. numbers and reproductive success Many nesting attempts occurred of Nene in HYNP. Predator control at Ainahou [471. However, few nesting efforts should be improved so that attempts were made inside Ainahou fewer eggs, goslings and adults Big Pen [11 I. which is located within the are depreciated. To increase food Ainahou area, in spite of supplemental availability for goslings and breeding food and water supplied to Nene and females, habitat restoration efforts perimeter traplines protecting them should be expanded. Lastly, continued from predators. The success of Pen public education is necessary to reduce 7 [32 fledglings during the decade] harmful human-Nene interactions. indicates that Ainahou Big Pen should With these management strategies in be successful, as it is bigger and place, perhaps the survival of Nene situated in better nesting and foraging within HYNP will be ensured. habitat [Woog & Black 2001] than Pen 7. 16 Reproductive trends of Nene
Acknowledgements Hoshide, H.M., Price, A.J. & Katahira, L. 1990. A progress report on Nene Branta sandvicensis in Hawaii Volcanoes National Park from We thank Bemidji State University, the 1974-89. Wildfowl 41 : 152-155. Bemidji State University Foundation Hu, D. 1998. Causes of Endangerment and and Hawaii Volcanoes National Park Extinction in Hawaiian Birds. Unpublished for financial support. We also thank 0. PhD thesis, University of California, Davis, USA. Rave and two anonymous reviewers for Kear, J. & Berger, A.J. 1980. The Hawaiian Goose: helpful comments on the manuscript. an Experiment in Conservation. T. & A.O. Lastly, we thank the numerous Poyser Ltd., Calton, England. volunteers who helped collect Nene Rave, E.H. 1995. Genetic analyses of wild monitoring data. populations of Hawaiian Geese using DNA fingerprinting. Condor 97: 82-90. Stone, C.P., Walker, R.L., Scott, J.M. & Banko, References P.C. 1983a. Hawaiian Goose research and management: where do we go from here? Banko, P.C. 1988. Breeding Biology and 'Elepaio 44: 11-15. Conservation of the Nene, Hawaiian Goose Stone, C.P., Hoshide, H.M. & Banko, P.C. 1983b. Nesochen sandvicensis. Unpublished PhD Productivity, mortality, and movements of thesis, University of Washington, USA. Nene in the Ka'u Desert, Hawaii Volcanoes Banko, P.C. 1992. Constraints on productivity National Park, 1981-1982. Pacific Science 38: of wild Nene or Hawaiian Geese Branta 301-311. sandvicensis. Wildfowl 43: 99-106. U.S. Fish & Wildlife Service 2004. Draft Revised Black, J.M. & Banko, P.C. 1994. Is the Hawaiian Recovery Plan for the Nene or Hawaiian Goose Goose Branta sandvicensis saved from Branta sandvicensis. U.S. Fish & Wildlife extinction? In: Creative Conservation: Service, Portland, OR; 148 + xi pp. ( Interactive Management of Wild and Captive Woog, F. 1999. Ecology and Behaviour of Animals (eds. P.J.S. Olney, G.M. Mace & Reintroduced Hawaiian Geese. Unpublished A.T.C. Feistner) , pp. 394-410. Chapman & PhD thesis, University of Hannover, Hall, New York. Germany. Black, J.M., Duvall, F., Hoshide, H., Medeiros, Woog, F. 2002. Distribution and timing of J., Hodges, C.N., Santos, N. & Te lfer, T. nesting in Hawaiian Geese in relation to 1991. The current status of the Hawaiian f_o od phenology in scrublands. Wildfowl 53: Goose Branta sandvicensis and its recovery 79-107. programme. Wildfowl 42: 149-154. Woog, F. & Black, J.M. 2001. Foraging behavior Black, J.M., Prop, J., Hunter, J.M., Woog, F., and temporal use of grassla nds by Nene: Marshall, A.P. & Bowler, J.M. 1994. Foraging implications for management. Studies in behaviour and energetics of the Hawaiian Avian Biology 22: 319-328. Goose Branta sandvicensis. Wildfowl 45: 65- Zillich, U. & Black, J. 2002. Body mass and 109. abdominal profile index in captive Hawaiian Black, J.M., Choudhury, 5. & Owen, M. 1996. Geese. Wildfowl 53: 67-77. Do Barnacle Geese benefit from lifelong monogamy? In: Partnerships in Birds: the Study of Monogamy, (ed. J.M. Black). pp. 91- 117. Oxford University Press, New York. Elder, W.H. & Woodside, D.H. 1958. Biology and management of the Hawaiian Goose. Transactions of the North American Wildlife Conference 23: 198-215. (