Yokomizo Hiroyuki, Yamashita Jin, Iwasa Yoh
Department of Biology, Faculty of Sciences, Kyushu University, Fukuoka, 812-8581, Japan.
J Theor Biol. 2003 Jan 21;220(2):215-31. doi: 10.1006/jtbi.2003.3157.
We study the optimal conservation effort for a population in a fluctuating environment. The survivorship of a population is affected by unpredictable environmental fluctuation (noise) and can be improved by conservation effort accompanied by a cost. The optimal effort level is the one that minimizes the total cost, defined as the weighted sum of the population extinction risk and the economic cost of conservation effort. The optimal effort depends on the variance and the probability distribution of the noise, the relative importance of the population's survival vs. the economic cost, the effectiveness of conservation effort, and the time scope over which we optimize. The analysis of dynamic programming illustrates that the choice of extinction risk function greatly affects the optimal effort level. The conservation effort level that is the best solution of a multiple-year optimization may be higher than that for the corresponding single-year optimization, if the population is relatively safe. However, the conservation level for the multiple-year optimization becomes lower than for the single-year optimization if the population is endangered. In a similar manner, the optimal conservation effort level for the problem with a short time scope is either higher or lower than that for the problem with a long time scope, depending on the extinction risk of the population. Next, for each parameter of the model, we define five different sensitivities of extinction probability or of the total cost. We then study the mean increase in the total cost caused by the uncertainty of parameters. To achieve the best conservation result, we need to invest the limited research effort to the parameter with the largest effect to the optimal effort level, rather than to those with large impacts on the extinction probability or on the total cost. The recommended policy should depend critically on the choice of the criterion to optimize, which shows the importance of theoretical study of the relationship in performing proper decision making in conservation practice.
我们研究了波动环境中种群的最优保护策略。种群的存活率受到不可预测的环境波动（噪声）的影响，而保护策略虽伴有成本但可提高存活率。最优策略水平是使总成本最小化的水平，总成本定义为种群灭绝风险与保护策略经济成本的加权和。最优策略取决于噪声的方差和概率分布、种群生存与经济成本的相对重要性、保护策略的有效性以及我们进行优化的时间范围。动态规划分析表明，灭绝风险函数的选择对最优策略水平有很大影响。如果种群相对安全，多年优化的最佳解决方案中的保护策略水平可能高于相应单年优化的水平。然而，如果种群濒危，多年优化的保护水平会低于单年优化的水平。同样，时间范围短的问题的最优保护策略水平高于或低于时间范围长的问题的最优保护策略水平，这取决于种群的灭绝风险。接下来，对于模型的每个参数，我们定义了灭绝概率或总成本的五种不同敏感性。然后我们研究参数不确定性导致的总成本平均增加。为了实现最佳保护效果，我们需要将有限的研究精力投入到对最优策略水平影响最大的参数上，而不是投入到对灭绝概率或总成本影响大的参数上。推荐的策略应严格取决于优化标准的选择，这表明在保护实践中进行正确决策时研究这种关系的理论重要性。
