Trends in the demography of long-tailed bats, Eglinton Valley, Fiordland
Background
This measure relates to indicator 1.4.2 – Security of threatened and at risk taxa.
Long-tailed bats were common throughout Aotearoa New Zealand in the 1800s but have been declining in many regions since the early 1900s as a result of forest clearance, land development and predation. Consequently, they are now listed as Nationally Critical under the New Zealand Threat Classification System. DOC has been monitoring bats in the Eglinton Valley, Fiordland, since 1994 to detect changes in their survival and abundance in response to predator management.
Long-tailed bats at Eglinton have high survival due to predator management.
What did we measure?
DOC has been monitoring the annual survival of long-tailed bats in two social groups in the Eglinton Valley for over ten years (see Figure 1 - Maps). This site is prone to predator irruptions driven by periodic mast seeding of beech forest.
DOC staff catch long-tailed bats using harp traps and mark them with identifying bands before their release. The banded individuals that are recaptured each year are then used to estimate survival and the size of each social group. If a marked animal was not caught one year but is caught in later years it is included in the minimum number alive (MNA) for the earlier year. Results are modelled in relation to beech mast seeding and predator management using multi-state mark-recapture models.
What did we find?
- Survival of long-tailed bats in both groups has been variable since 2010 (Figure 2), but overall, numbers of adult females have increased (Figure 3).
- Before 2010, beech mast seeding was associated with reduced survival, but this was not evident in the 2010, 2012 or 2015 mast events, suggesting that pest management in the Eglinton Valley has reduced the impacts of predator irruptions.
- However, there was a significant drop in survival for the Walker Creek social group after the 2016/17 beech mast. Causes of this are unclear, but it is possible that the 2016 aerial 1080 operation did not successfully reduce rat numbers for sufficient time. Alternative causes include some members moving to a different area, or loss of foraging habitat from land clearance south of the National Park boundary.
- Because rat numbers were high, additional predator management with bait stations was used in 2017/18 and 2018/2019, and bat numbers continue to increase.
- Management of the 2019/2020 mast with aerial 1080 (an operation in September 2019 and another in May/June 2020) resulted in an increase in numbers at Walker and a small decrease at Mackay.
Figure 1: Location of the two long-tailed bat social groups that are monitored by DOC in the Eglinton Valley, Fiordland.
Figure 2: Survival of adult and juvenile female long-tailed bats at Walker Creek and Mackay Creek calculated using RMark. Coloured bars indicate the beech mast and management response in the preceding season. Values are means ± 95% confidence intervals.
Figure 3: Abundances of adult female long-tailed bats at Walker Creek and Mackay Creek estimated as the minimum number alive using recapture rates. Coloured bars indicate the beech mast and management response in the preceding season. Values are means ± 95% confidence intervals.
Data quality
This measure is classified as a case study and complies with the data quality guidelines used in New Zealand’s Environmental Reporting framework.
These survival estimates are considered robust because capture histories have been collected for many bats over a long period of time and recapture probabilities are consistently high. The general trend is similar to that modelled using the index of MNA and the recapture rates.
Glossary of terms
95% confidence interval is the range of values that have a 95% likelihood of containing the true value.
Mast seeding is the synchronous production of large quantities of seeds within a population of plants at irregular intervals. This occurs in a number of New Zealand forest tree and tussock grass species.
Minimum number alive (MNA) is the number of individuals caught in a capture session plus those any previously marked that were not caught but were caught in subsequent capture sessions (Krebs, 1966). Here, we model MNA taking into account the estimated recapture probability in each year.
RMark is an interface to the software package MARK developed by Laake (2013). MARK was developed by Gary C. White to derive parameter estimates from animals that are marked and then re-encountered at a later time.
Additional resources
Krebs, C.J., 1966. Demographic changes in fluctuating populations of Microtus californicus. Ecological monographs 36, 239–273.
Laake, J.L., 2013. RMark: An r interface for analysis of capture-recapture data with MARK (AFSC Processed Rep. No. 2013-01). Alaska Fisheries Science Centre, NOAA, US Department of Commerce., Seattle, WA.
McGlone, M.S., McNutt, K., Richardson, S.J., Bellingham, P.J., Wright, E.F., 2020. Biodiversity monitoring, ecological integrity, and the design of the New Zealand biodiversity assessment framework. New Zealand Journal of Ecology 44, 3411.
O’Donnell, C.F., Pryde, M.A., van Dam-Bates, P., Elliott, G.P., 2017. Controlling invasive predators enhances the long-term survival of endangered New Zealand long-tailed bats (Chalinolobus tuberculatus): Implications for conservation of bats on oceanic islands. Biological Conservation 214, 156–167.
Walker, S., Kemp, J.R., Elliott, G.P., Mosen, C.C., Innes, J.G., 2019. Spatial patterns and drivers of invasive rodent dynamics in New Zealand forests. Biological Invasions 21, 1627–1642.