Worldwide, shorebird populations are declining. Our objectives were to examine abundance trends of shorebirds regularly wintering at Tomales Bay, Marin County, California, accounting for the local effects of rainfall, raptors, and the restoration of part of the bay to tidal wetlands. From November 1989 to February 2019, we conducted 177 comprehensive winter shorebird surveys of Tomales Bay; we averaged 5.7 ± 0.9 (mean ± SD) winter surveys per year. In 30 yr, we counted 1,215,821 shorebirds of 31 species. We used generalized linear models and multi-model inference to evaluate trends in shorebird abundance while accounting for local sources of variation. We conducted separate analyses for 14 species seen in at least 20 of the 30 yr of monitoring and for all shorebird species combined. During the study, the abundance of all species combined declined 66% (52% in the North Bay and 81% in the South Bay) with the most rapid decline in the first 10 yr of monitoring. Of 13 species for which year was in the top model, 10 species decreased in abundance and 3 species increased. Dunlin and Western Sandpiper accounted for the greatest losses in total numbers. The best-supported models to estimate trends in shorebirds included predictors for year and North Bay vs. South Bay. Of the local variables we considered, rainfall was included in 10 of the 15 best-supported models (including all species combined), negatively affecting the numbers of all species except Willets. The wetland restoration project was included in 5 top models, with a short-term positive impact. Raptor abundance was included in 3 top models with mixed results. Our results show that effective conservation and management of local shorebird populations must be linked with regional/global efforts if we are to reverse negative shorebird trends.

中文翻译：

---

加利福尼亚温带河口越冬鸻鸟数量下降：30 年展望

在世界范围内，鸻鹬种群正在减少。我们的目标是检查经常在加利福尼亚州马林县 Tomales 湾越冬的鸻鹬的丰度趋势，考虑降雨、猛禽的局部影响以及将海湾部分恢复为潮汐湿地的影响。从 1989 年 11 月到 2019 年 2 月，我们对托马莱斯湾进行了 177 次冬季鸻鸟综合调查；我们平均每年进行 5.7 ± 0.9（平均值 ± SD）冬季调查。在 30 年中，我们统计了 31 个物种的 1,215,821 只鸻鹬。我们使用广义线性模型和多模型推断来评估鸻鹬丰度的趋势，同时考虑当地的变异来源。我们对 30 年监测中的至少 20 年中看到的 14 种物种以及所有鸻鹬物种进行了单独分析。在学习期间，所有物种的丰度总和下降了 66%（北湾 52%，南湾 81%），在监测的前 10 年下降最快。在最高模型年份的 13 个物种中，10 个物种减少，3 个物种增加。邓林和西鹬在总数中的损失最大。估计鸻鹬趋势的最佳支持模型包括年份和北湾与南湾的预测因子。在我们考虑的局部变量中，降雨量包含在 15 个最佳支持模型中的 10 个中（包括所有物种的组合），对除 Willets 之外的所有物种的数量产生负面影响。湿地修复项目入选5个顶级模式，短期产生积极影响。Raptor 丰度包含在 3 个顶级模型中，结果好坏参半。