Evaluating detection and monitoring tools for incipient and relictual non-native ungulate populations

Abstract

Hawai‘i Volcanoes National Park (HAVO) encompasses 1,308 km2 on Hawai‘i Island. The park harbors endemic plants and animals which are threatened by a variety of invasive species. Introduced ungulates have caused sharp declines of numerous endemic species and have converted ecosystems to novel grazing systems in many cases. Local ranchers and the Territorial Government of Hawai‘i had long conducted regional ungulate control even prior to the establishment of HAVO in 1916. In 1995 the park’s hunting team began a new hunt database that allowed managers to review hunt effort and effectiveness in each management unit. Target species included feral pigs (Sus scrofa), European mouflon sheep (Ovis gmelini musimon), feral goats (Capra hircus) and wild cattle (Bos taurus). Hunters removed 1,204 feral pigs from HAVO over a 19-year period (1996‒2014). A variety of methods were employed, but trapping, snaring and ground hunts with dogs accounted for the most kills. Trapping yielded the most animals per unit effort. Hunters and volunteers removed 6,657 mouflon from HAVO; 6,601 of those were from the 468 km2 Kahuku Unit. Aerial hunts yielded the most animals followed by ground hunt methods. Hunters completed eradications of goats in several management units over an 18- year period (1997‒2014) when they removed the last 239 known individuals in HAVO primarily with aerial hunts. There have also been seven cattle and five feral dogs (Canis familiaris) removed from HAVO. Establishing benchmarks and monitoring the success of on-the-ground ungulate removal efforts can improve the efficiency of protecting and restoring native forest for high-priority watersheds and native wildlife. We tested a variety of methods to detect small populations of ungulates within HAVO and the Hō‘ili Wai study area in the high-priority watershed of Ka‘ū Forest Reserve on Hawai‘i Island. We conducted ground surveys, aerial surveys and continuous camera trap monitoring in both fence-enclosed units and unenclosed units where populations of introduced mouflon and feral pigs threatened sensitive native plants and forest bird habitats. Beginning in June 2014, twenty infrared camera traps were positioned in areas occupied by ungulates. The cameras were active for at most 198 days, and then half of the cameras were baited with oats and salt blocks for 126 days. There were a total of 1,496 observations of mouflon captured on camera, totaling 2,592 individuals: 1,020 ewes, 900 rams, 276 lambs, and 396 sheep of unknown sex. There were no detections of the illegally introduced axis deer (Axis axis). There were 11 observations of feral pigs and 109 observations of other animals (birds, rats, and other small mammals), including one detection of the federally endangered Hawaiian hawk (Buteo solitarius). Mouflon detection rates did not increase near baited cameras until three months after the initial baiting. Ground-based surveys for ungulate presence were conducted along six transects in Kahuku in October 2014. Evidence of ungulates were detected in 27.5% of plots surveyed within an unenclosed unit, while an enclosed unit had sign in only 3.6% of plots surveyed. An aerial survey by helicopter was conducted in October 2014. A total of 378 mouflon were detected during the survey: 192 in the Kahuku Paddocks, 186 in the Kahuku East unit and no mouflon were detected in the actively controlled Mauka unit. Two baseline ungulate surveys have been completed at the Hō‘ili Wai study area in the highpriority watershed of Ka‘ū Forest Reserve adjacent to Kahuku prior to the completion of an exclusionary ungulate fence. Ground-based surveys were conducted on four transects within a 4.99 km2 area on 5 August and 5–6 November 2014. In August, 20.71% of 565 plots surveyed 2 had fresh or intermediate ungulate sign. In November, 17.41% of 557 plots surveyed had fresh or intermediate ungulate sign. These surveys represent baseline levels of ungulate activity prior to management; therefore comparative inferences can be made about ungulate distribution and relative abundance, but inferences about absolute abundance cannot be made until all ungulates have been removed from the enclosed area. Additional ground-based surveys will be conducted when the fenced area has been fully enclosed, and until ungulate removals have been completed.