Climate change is expected to increase woody vegetation abundance in the Arctic, yet the magnitude, spatial pattern and pathways of change remain uncertain. We compared historical orthophotos photos (1952 and 1979) with high‐resolution satellite imagery (2015) to examine six decades of change in abundance of white spruce Picea glauca and tall shrubs (Salix spp., Alnus spp.) near the Agashashok River in northwest Alaska. We established ~3000 random points within our ~5500 ha study area for classification into nine cover types. To examine physiographic controls on tree abundance, we fit multinomial log‐linear models with predictors derived from a digital elevation model and with arctic tundra, alpine tundra and ‘tree’ as levels of a categorical response variable. Between 1952 and 2015, points classified as arctic and alpine tundra decreased by 31% and 15%, respectively. Meanwhile, tall shrubs increased by 86%, trees mixed with tall shrubs increased by 385% and forest increased by 84%. Tundra with tall shrubs rarely transitioned to forest. The best multinomial model explained 71% of variation in cover and included elevation, slope and an interaction between slope and ‘northness’. Treeline was defined as the elevation where the probability of tree presence equaled that of tundra. Mean treeline elevation in 2015 was 202 m, corresponding with a June–August mean air temperature > 11°C, which is > 4°C warmer than the 6–7°C isotherm associated with global treeline elevations. Our results show dramatic increases in the abundance of trees and tall shrubs, question the universality of air temperature as a predictor of treeline elevation and suggest two mutually exclusive pathways of vegetation change, because tundra that gained tall shrubs rarely transitioned to forest. Conversion of tundra to tall shrubs and forest has important and potentially contrasting implications for carbon cycling, surface energy exchange and wildlife habitat in the Arctic.

中文翻译：

---

阿拉斯加西部布鲁克斯山脉的树木和高大灌木侵入苔原的路径

预计气候变化将增加北极的木本植被丰度，但变化的幅度、空间格局和途径仍不确定。我们将历史正射影像照片（1952 年和 1979 年）与高分辨率卫星图像（2015 年）进行比较，以检查西北部 Agashashok 河附近的白云杉 Picea glauca 和高灌木（Salix spp., Alnus spp.）丰度的 60 年变化阿拉斯加州。我们在约 5500 公顷的研究区域内建立了约 3000 个随机点，用于分类为九种覆盖类型。为了检查对树木丰度的自然地貌控制，我们将多项对数线性模型与来自数字高程模型的预测变量以及北极苔原、高山苔原和“树”作为分类响应变量的水平进行拟合。1952 年至 2015 年间，归类为北极和高山苔原的点分别减少了 31% 和 15%。同时，高灌木增加了86%，与高灌木混合的树木增加了385%，森林增加了84%。具有高灌木的苔原很少过渡到森林。最好的多项式模型解释了 71% 的覆盖变化，包括高程、坡度以及坡度和“北度”之间的相互作用。树木线被定义为树木存在概率等于苔原概率的海拔。2015 年的平均林线海拔为 202 m，对应于 6 月至 8 月的平均气温 > 11°C，比与全球林线海拔相关的 6-7°C 等温线高 4°C。我们的结果显示树木和高大灌木的数量急剧增加，质疑气温作为树木线海拔预测指标的普遍性，并提出了两种相互排斥的植被变化途径，因为获得高大灌木的苔原很少转变为森林。将苔原转变为高大的灌木和森林，对北极的碳循环、地表能量交换和野生动物栖息地具有重要的潜在影响。