No data available.
Please log in to see this content.
You have no subscription access to this content.
The full text of this article is not currently available.
Examination of the Appropriate Inference Procedure in a Model Structure for Harvest-Based Estimation of Sika Deer Abundance
Rent:
Rent this article for
This article is currently unavailable for purchase.
Abstract
Abstract.
To obtain proper estimates of wildlife abundance by harvest-based models (HBMs), an understanding of the model structure and data properties is required. Otherwise, there may be a risk of failure to obtaining adequate estimates. In this study, we estimated the abundance of sika deer using several spatially fine-scale HBMs with different structures and aimed to clarify the effects of the model structure and data quality on estimates. We used monitoring data collected by the Gifu Prefectural Government and other data collected by the authors. Four HBMs were constructed according to the combinations of the model structure (considering overdispersion in the observation models) and data (with or without additional observation data), and their parameters were estimated. The results showed that among the four HBMs, reasonable deer abundance was estimated by two HBMs in which overdispersion was considered in the observation models of the less precision data only. As the parameters failed to converge in the other two HBMs in which overdispersion was considered in all observation models, the abundance would be overestimated. Thus, our results confirmed that understanding the model structure and data properties was essential for obtaining proper estimates of wildlife abundance from currently available data with HBM.
Abstract in Japanese
(要旨).ニホンジカ個体数推定のためのHarvest-based modelsにおける適切なモデル設計の検討.Harvest-based models(HBMs)を用いて野生動物の適切な個体数推定値を得るためには,モデルの構造とデータの特性を理解することが必要である.これらに対する理解が不十分な場合,適切な推定値を得られないリスクが大きくなる.本研究では,狩猟メッシュを単位とした空間解像度の高いHBMsを複数構築してニホンジカ個体数の推定を試み,モデル構造とデータの質が個体数推定値に及ぼす影響を明らかにすることを目指した.データとして,岐阜県が収集したモニタリングデータと,筆者らが収集した観測データを用いた.モデル構造(観測モデルにおける過分散の考慮の有無)とデータ(追加観測データの有無)の組み合わせにより,4つのHBMsを構築し個体数推定を試みた.その結果,4つのモデルのうち,精度の低いデータに対する観測モデルのみに過分散を設定した2つのモデルでは妥当なニホンジカ個体数が推定された.一方,すべての観測モデルで過分散を考慮した他の2つのモデルではパラメータは収束せず,また個体数推定値は過大であった.本研究の結果から,ある時点で利用可能なデータからHBMsを用いて野生動物の個体数を適切に推定するためには,モデル構造とデータの特性に対する理解が不可欠であることが確認された.
Received 20 10月 2021
Accepted 07 10月 2022
Acknowledgments:
We thank the Environmental Policy Planning Division, Department of Environmental Affairs and Citizen Support, Gifu Prefecture, for providing most of the data. This work was supported by JSPS KAKENHI grant numbers 21H02247. We also appreciate the two anonymous reviewers' helpful comments on the models and the discussions.
Full text loading...
/content/article/1343-4152/48020/91