Robustness of adult female survival maintains a high-density sika deer (Cervus nippon) population following the initial irruption M. Ueno A,H , H. Iijima B , K. Takeshita C , H. Takahashi D , T. Yoshida E , H. Uehara E , H. Igota E , Y. Matsuura F , T. Ikeda C , M. Azumaya G and K. Kaji C A Eastern Field Station, Institute of Environmental Sciences, Hokkaido Research Organisation, Kushiro, Hokkaido 085-8588, Japan. B Laboratory of Wildlife Biology, Department of Forest Science, Forestry and Forest Products Research Institute, Tsukuba, Ibaraki 305-8687, Japan. C Laboratory of Wildlife Management, Department of Agriculture, Tokyo University of Agriculture and Technology, Fuchu, Tokyo 183-8509, Japan. D Kansai Research Center, Forestry and Forest Products Research Institute, Kyoto, Kyoto 612-0855, Japan. E Department of Environmental and Symbiotic Science, Rakuno Gakuen University, Ebetsu, Hokkaido 069-8501, Japan. F Hokkaido Research Center, Forestry and Forest Products Research Institute, Sapporo, Hokkaido 062-8516, Japan. G The Yezo Deer Association, Sapporo, Hokkaido 064-0803, Japan. H Corresponding author. Email: jmayueno@gmail.com Abstract Context. Irruption of large herbivore populations is characterised by three distinct phases: (1) an exponential increase in population to a peak abundance; (2) a population crash; and (3) a second increase to another population peak, typically lower than the first peak of abundance. However, there has been little study of age- and sex-specific factors that affect the post-initial irruption interactions with food sources. Aims. We aimed to investigate annual survival rates of sika deer (Cervus nippon Temminck, 1838) in the sequent irruption of a population on Nakanoshima Island, Lake Toya, Japan, with a chronically high density during the period 2002–12. Methods. Survival monitoring data were obtained for 219 individuals (93 males and 126 females) using radio-collars. Annual survival was quantified, and related factors, i.e. deer abundance and winter severity, were determined by model selection using Akaike information criterion values. Key Results. The results showed that annual survival rates across sexes and age classes (fawn, yearling, prime-aged, old) decreased with increasing population density, snow depth and winter precipitation. Winter severity had a greater effect on adult survival than density regulation. Nevertheless, female adult survival was maintained at a high level, with a mean of 0.84 (95% CI: 0.80–0.88). Key conclusions. Robust survival rates for adult females might contribute to the maintenance of a high-density sika deer population in the post-initial irruption. Implications. We suggest that in the absence of predation and hunting, sika deer population is not able to self-regulate to the density level that avoids an irreversible impact on plants. Additional keywords: population density, population dynamics, population growth, radio-telemetry, survival, wildlife management. Received 26 July 2017, accepted 8 February 2018, published online 1 May 2018 Introduction Irruptive growth of a herbivore population can occur when the controls maintaining a very low density are removed, or when a population is introduced into a new habitat that is abundant in resources and free of predation and hunting (Caughley 1970; McCullough 1979). An irruption is characterised by three distinct phases (Caughley 1970; McShea et al. 1997): (1) an exponential increase in population up to a peak abundance; (2) a population crash; and (3) a second increase to another peak of population size typically lower than the first peak of abundance (Kaeuffer et al. 2010). The third stage assumes an equilibrium between plants and animals (Caughley 1979). However, it is unclear CSIRO PUBLISHING Wildlife Research, 2018, 45, 143–154 https://doi.org/10.1071/WR17103 Journal compilation Ó CSIRO 2018 www.publish.csiro.au/journals/wr