INTRODUCTION

There is clear evidence that pollinators, especially native bees, are in decline due to a variety of factors including introduction of pathogens and pests, climate change, loss of genetic diversity, and land use change (Potts et al., 2010). Some authors argue that among these factors land use change is the single greatest threat to the sustainability of native bee populations (Brown & Paxton, 2009), and corresponding negative effects on native bees and pollination services may be driven by fragmentation (Cane et al., 2006), loss of landscape-level heterogeneity (Kremen et al., 2007), and degradation of habitats (Thapa-Magar et al., 2020). Consequently, there is a need to understand how localized habitat factors drive changes in bee assemblages, and whether habitat factors can be manipulated to enhance the conservation of pollinators while also meeting ecosystem management goals.

Throughout most temperate regions of the northern hemisphere coniferous forests are a predominant cover type, and a variety of recent studies show that forests are an important reservoir for native bee biodiversity (reviewed by Hanula et al., 2016). Bee assemblages in temperate conifer forests are responsive to disturbances including fire (Burkle et al., 2019), insect outbreaks (Davis et al., 2020; Foote et al., 2020), and forest management (Rivers et al., 2018). One apparent link among abiotic, biotic, and anthropic disturbances is that they each typically involve either destruction or removal of live canopy, which has cascading effects on physical (Prévost & Raymond, 2012) and ecological (Kneeshaw & Bergeron, 1998) properties of forest stands and can impact plants and animals inhabiting the understory, including bees. For example, reduced canopy density or cover may result in recruitment of forbs (Anderson et al., 1969) and a resultant increase in bee foraging habitat (McCabe et al., 2019; Walters & Stiles, 1996). Alternatively, removal of the canopy may alter thermal conditions with consequences for bee foraging behaviours and activity periods (Aleixo et al., 2017). Hence, it is useful to understand how bee assemblages and bee functional traits vary along gradients of canopy cover in conifer forests, as this can promote management efforts that target conserving both biodiversity and functional variation by maximizing cover conditions associated with high bee species richness (Odanaka et al., 2020).

Here, our objective is to characterize the bee fauna in a southwestern mixed pine-juniper ecosystem across a gradient of canopy cover in southcentral Colorado. Across much of the western United States, pinyon pine (Pinus edulis Engelm.) – juniper (primarily Juniperus monosperma (Engelm.) Sarg. and J. osteosperma (Torr.) Little) ecosystems are found in arid habitats; typical stands have low canopy heights (<10 m) and may range in structure from woodland to shrubland to savanna. Although this ecosystem type occupies a vast geographic area in the western United States (~40 M ha, Romme et al., 2009) there remains very little known about the biodiversity of native bees in these systems, with only a few previously published surveys (Carril et al., 2018; Nyoka, 2010; Smith et al., 2015) and none from Colorado. However, arid regions of the southwestern United States are major hotspots of bee biodiversity (Minckley & Radke, 2021), and native bees may be among the fauna most threatened by regional fragmentation due to rapid growth of the wildland-urban interface (Cane et al., 2006). Accordingly, we: (1) describe the bee community in a previously unstudied pine-juniper ecosystem; (2) evaluate the relative effects of canopy openness on bee α- and β-diversity, and (3) compare bee functional traits across a gradient of canopy openness and relative to predominant overstory cover (no canopy, pinyon pine-dominant, or ponderosa pine-dominant). Our results provide new insights into how the structural characteristics of arid pine-juniper woodlands impact bee biodiversity and functional variation, with potential consequences for vegetation management practices.

中文翻译：

---

树冠覆盖和季节性与松树-杜松混合林地本地蜜蜂组合的变化有关

介绍

有明确的证据表明，由于多种因素，包括病原体和害虫的引入、气候变化、遗传多样性的丧失和土地利用的变化，授粉媒介，尤其是本地蜜蜂正在减少（Potts 等，  2010 年）。一些作者认为，在这些因素中，土地利用变化是对本地蜜蜂种群可持续性的最大威胁（Brown & Paxton，  2009），而对本地蜜蜂和授粉服务的相应负面影响可能是由碎片化驱动的（Cane 等人。 ,  2006 年）、景观水平异质性的丧失（Kremen 等人，  2007 年）和栖息地退化（Thapa-Magar 等人，  2020 年））。因此，有必要了解局部栖息地因素如何推动蜜蜂群落的变化，以及是否可以操纵栖息地因素以加强对传粉媒介的保护，同时满足生态系统管理目标。

在北半球的大部分温带地区，针叶林是主要的覆盖类型，最近的各种研究表明，森林是本地蜜蜂生物多样性的重要水库（Hanula 等人综述，  2016 年）。温带针叶林中的蜜蜂群落会对火灾（Burkle 等人，  2019 年）、昆虫爆发（Davis 等人，  2020 年；Foote 等人，  2020 年）和森林管理（Rivers 等人，  2018 年）等干扰做出反应. 非生物、生物和人为干扰之间的一个明显联系是，它们通常都涉及破坏或移除活的树冠，这对物理有级联效应（Prévost & Raymond，  2012) 和生态 (Kneeshaw & Bergeron,  1998 ) 特性，可以影响栖息在林下林下的植物和动物，包括蜜蜂。例如，树冠密度或覆盖率的降低可能会导致蜂群的增加（Anderson 等人，  1969 年），从而导致蜜蜂觅食栖息地的增加（McCabe 等人，  2019 年；Walters & Stiles，  1996 年）。或者，去除树冠可能会改变热条件，从而影响蜜蜂的觅食行为和活动期（Aleixo 等，  2017）。因此，了解蜜蜂组合和蜜蜂功能特征如何随着针叶林冠层覆盖梯度的变化是有用的，因为这可以通过最大化与高蜜蜂物种丰富度相关的覆盖条件来促进旨在保护生物多样性和功能变化的管理工作（Odanaka等人，  2020 年）。

在这里，我们的目标是在科罗拉多州中南部的一个树冠覆盖梯度的西南松-杜松混合生态系统中描述蜜蜂动物群的特征。在美国西部的大部分地区，在干旱的栖息地发现了松树 ( Pinus edulis Engelm.) – 杜松（主要是Juniperus monosperma (Engelm.) Sarg. 和J.osteosperma (Torr.) Little）生态系统；典型的林分树冠高度较低（<10 m），其结构范围从林地到灌木丛再到稀树草原。尽管这种生态系统类型在美国西部占据了广阔的地理区域（~40 M ha, Romme et al.,  2009）对于这些系统中本地蜜蜂的生物多样性知之甚少，只有少数先前发表的调查（Carril 等人，  2018 年；Nyoka，  2010 年；Smith 等人，  2015 年），而科罗拉多州也没有。然而，美国西南部的干旱地区是蜜蜂生物多样性的主要热点（Minckley & Radke,  2021），由于荒地-城市界面的快速增长，本地蜜蜂可能是受区域分裂威胁最大的动物群之一（Cane 等人） .,  2006）。因此，我们：（1）描述了以前未研究过的松柏生态系统中的蜜蜂群落；(2) 评估冠层开放度对蜜蜂 α 和 β 多样性的相对影响，以及 (3) 比较不同冠层开放度和相对于主要覆盖层（无冠层、松树松树或黄松）的蜜蜂功能特征松树为主）。我们的研究结果为干旱松柏林地的结构特征如何影响蜜蜂生物多样性和功能变异以及对植被管理实践的潜在影响提供了新的见解。