Data from: Lincoln estimates of mallard (Anas platyrhynchos) abundance in North America

Estimates of range-wide abundance, harvest, and harvest rate are fundamental for sound inferences about the role of exploitation in the dynamics of free-ranging wildlife populations, but reliability of existing survey methods for abundance estimation is rarely assessed using alternative approaches. North American mallard populations have been surveyed each spring since 1955 using internationally coordinated aerial surveys, but population size can also be estimated with Lincoln's method using banding and harvest data. We estimated late summer population size of adult and juvenile male and female mallards in western, midcontinent, and eastern North America using Lincoln's method of dividing (i) total estimated harvest, H, by estimated harvest rate, h, calculated as (ii) direct band recovery rate, f, divided by the (iii) band reporting rate, p. Our goal was to compare estimates based on Lincoln's method with traditional estimates based on aerial surveys. Lincoln estimates of adult males and females alive in the period June–September were 4.0 (range: 2.5–5.9), 1.8 (range: 0.6–3.0), and 1.8 (range: 1.3–2.7) times larger than respective aerial survey estimates for the western, midcontinent, and eastern mallard populations, and the two population estimates were only modestly correlated with each other (western: r = 0.70, 1993–2011; midcontinent: r = 0.54, 1961–2011; eastern: r = 0.50, 1993–2011). Higher Lincoln estimates are predictable given that the geographic scope of inference from Lincoln estimates is the entire population range, whereas sampling frames for aerial surveys are incomplete. Although each estimation method has a number of important potential biases, our review suggests that underestimation of total population size by aerial surveys is the most likely explanation. In addition to providing measures of total abundance, Lincoln's method provides estimates of fecundity and population sex ratio and could be used in integrated population models to provide greater insights about population dynamics and management of North American mallards and most other harvested species.对物种全分布区的种群丰度、狩猎收获量及狩猎捕获率的估算，是严谨推断开发利用对野生自由栖息种群动态影响的基础，但现有种群丰度估算调查方法的可靠性，极少通过替代方法进行验证。自1955年以来，北美绿头鸭（mallard）种群每年春季都会通过国际协作的航空调查开展监测，不过种群规模也可利用鸟类环志与狩猎收获数据，通过林肯法（Lincoln's method）进行估算。本研究采用林肯法，对北美西部、中部大陆及东部地区的成年与幼年、雄性与雌性绿头鸭的夏末种群规模进行估算：该方法将（i）总估算狩猎收获量H除以估算捕获率h，而捕获率h由（ii）直接环志回收率f与（iii）环志报告率p的比值计算得到。本研究的核心目标是将林肯法得到的种群估算结果与传统航空调查得到的估算结果进行对比。在6月至9月期间存活的成年雄性与雌性绿头鸭的林肯法估算结果，分别为对应区域航空调查估算值的4.0倍（区间：2.5~5.9）、1.8倍（区间：0.6~3.0）与1.8倍（区间：1.3~2.7），对应北美西部、中部大陆及东部的绿头鸭种群；且两种方法得到的种群估算结果仅存在中等程度的相关性（西部种群：1993-2011年，r=0.70；中部大陆种群：1961-2011年，r=0.54；东部种群：1993-2011年，r=0.50）。林肯法估算值偏高具有其合理性：林肯法的推断地理范围为种群的全分布区，而航空调查的抽样框架并不完整。尽管两种估算方法均存在若干重要的潜在偏倚，但本研究的分析表明，航空调查对总种群规模的低估是导致这一差异的最合理解释。除了可获取总种群丰度的估算值外，林肯法还能估算繁殖力与种群性别比例，且可应用于集成种群模型（Integrated Population Models），以进一步深入解析北美绿头鸭及多数其他狩猎物种的种群动态与管理策略。