Introduction

The timing of diverse species' life history events (i.e. phenology) has been dramatically shifting in response to anthropogenic climate change (Thackeray et al., 2008; Boutin & Lane, 2014; Carter et al., 2018; Piao et al., 2019). For instance, highbush blueberry (Vaccinium corymbosum) and yellow wood sorrel (Oxalis europaea) advanced their flowering time in spring by 21 and 32 d respectively over the last 150 yr in Concord, Massachusetts, USA (Miller-Rushing & Primack, 2008) and butterflies in central California have advanced their first flight by > 3 wk over the last three decades (Forister & Shapiro, 2003). Such phenological shifts can have substantial ecological impacts, as they can influence species fitness and abundance (Willis et al., 2008; Springate & Kover, 2013), facilitate biological invasions (Wolkovich et al., 2013), and alter biophysical processes (Richardson et al., 2013; Estiarte & Peñuelas, 2015).

Climate change-induced phenological shifts can also disrupt species interactions (Edwards & Richardson, 2004; Donnelly et al., 2011; Thackeray et al., 2016) because species can vary greatly in their phenological responses to changes in climate (Morin et al., 2009; Primack et al., 2009; Cole & Sheldon, 2017). Phenological mismatches among mutualistic taxa are likely to cause reduced fecundity or increased mortality of those involved, and may have cascading effects throughout the ecosystem (Kudo & Ida, 2013; Kudo & Cooper, 2019; Visser & Gienapp, 2019). In particular, the disruption of synchrony in plant–pollinator systems could negatively impact plants through pollen limitation (Rafferty & Ives, 2011; Kudo & Ida, 2013), and pollinators through a reduction of floral resources (CaraDonna et al., 2018; Schenk et al., 2018). This can lead to collapses of mutualisms (Warren II & Bradford, 2014), reductions in animal-pollinated crop yields (Bartomeus et al., 2013), and local extinction (Revilla et al., 2015). It has been estimated that climate change-induced phenological shifts may cause a reduction of the floral resources available to 17–50% of all pollinator species, resulting in up to half of the historical pollinator activity period falling at times when no food plants are flowering (Memmott et al., 2007).

Further complicating this issue is the fact that phenological sensitivity to climate varies extensively within species across their ranges (Høye et al., 2013; Fitchett et al., 2014; Park et al., 2019; Song et al., 2020; Love & Mazer, 2021; Pearson et al., 2021). For example, the timing of leaf out, flowering, and fruiting have been found to be more sensitive to temperature in warmer regions across many plant species in temperate ecosystems (Zhang et al., 2015; Park et al., 2019). Thus, with climate change, intraspecific interactions and gene flow can be altered (Fox, 2003; Rivest et al., 2021; Park et al., 2022) and phenological synchrony between interacting species may shift heterogeneously across the landscape, differing among populations or morphotypes (Park et al., 2022). Knowledge of both intra- and interspecific variation in phenological responses is therefore critical to assessing the ecological impacts of climate change. However, less is known about the phenological responses of insect pollinators such as bees (Bartomeus et al., 2011), and most studies to date examining climate-driven phenological mismatches have not considered the consequences of intraspecific variability (Charmantier et al., 2008). This may be due in part to the general lack of phenological observation experiments that track multiple interacting species simultaneously across wide geographic scales. Indeed, many phenological studies on plant–pollinator interactions are limited to relatively small geographic areas (Forrest & Thomson, 2011; Kudo, 2014; Pyke et al., 2016; Olliff-Yang & Mesler, 2018), and it is possible that the trends observed in such studies may not apply across species' ranges.

Here we investigate how intraspecific variation in phenological sensitivity to climate can alter ecological interactions simultaneously within and among species using over 120 yr of natural history collections and citizen science data. We focus on a unique system, comprising Claytonia virginica (Portulacaceae; Virginia spring beauty) and its specialist pollinator Andrena erigeniae (Andrenidae; spring beauty miner bee). Claytonia virginica is an understory spring ephemeral that has distinct floral color morphs ranging from white to pink, which often occur in sympatry (Frey, 2004). It is not certain how these color morphs are maintained, but pollinator preference and divergent selection by pathogens and herbivores have been suggested to play a role (Frey, 2004). Variation in flowering phenology can also allow different color morphs to coexist (Tarasjev, 1997) and even small differences can potentially lead to divergence and reinforcement among phenotypes through assortative mating (Hopkins, 2013). Comparatively little is known about how the phenology of C. virginica and A. erigeniae varies across color morphs and populations (Schemske et al., 1978), though some degree of reproductive isolation and phenological differences has been suggested to occur among races of C. virginica (Lewis, 1976; Lewis & Suda, 1976; Doyle, 1981, 1983). Most studies of plant–pollinator mismatch focus on generalist pollinators, and analogous studies of specialist species that may be more vulnerable are rare, particularly for oligolectic bee species such as A. erigeniae (Bartomeus et al., 2011; Maglianesi et al., 2020).

Along these lines, we test the following hypotheses: flowering phenology varies among white and pink C. virginica color morphs, contributing to their maintenance in sympatry; phenological sensitivity to climate varies across the ranges of C. virginica and A. erigeniae and among color morphs of C. virginica; and climate change will result in larger temporal gaps among species and color morphs.

中文翻译：

---

物候敏感性的种内和种间变异的生态学意义

介绍

不同物种的生命史事件（即物候）的时间已经随着人为气候变化发生了巨大变化（Thackeray等人，  2008 年；Boutin 和 Lane，  2014 年；Carter等人，  2018 年；Piao等人，  2019 年） ). 例如，美国马萨诸塞州康科德的高丛蓝莓 ( Vaccinium corymbosum ) 和黄酢浆草 ( Oxalis europaea ) 的春季开花时间在过去 150 年中分别提前了 21 天和 32 天（Miller-Rushing & Primack，  2008 年）) 和加利福尼亚中部的蝴蝶在过去 30 年中将它们的首次飞行提前了 > 3 周 (Forister & Shapiro,  2003 )。这种物候变化会产生重大的生态影响，因为它们会影响物种的适应性和丰度（Willis等人，  2008 年；Springate 和 Kover，  2013 年），促进生物入侵（Wolkovich等人，  2013 年），并改变生物物理过程（Richardson等人，2013 年；Estiarte 和 Peñuelas，  2015 年）。

气候变化引起的物候变化也会破坏物种相互作用（Edwards & Richardson，  2004 年；Donnelly等人，  2011 年；Thackeray等人，  2016 年），因为物种对气候变化的物候反应可能有很大差异（Morin等人，2016 年） 。 ,  2009 年；普里马克等人，  2009 年；科尔和谢尔顿，  2017 年）。共生类群之间的物候不匹配可能会导致相关物种的繁殖力降低或死亡率增加，并可能在整个生态系统中产生级联效应（Kudo & Ida,  2013；Kudo & Cooper,  2019; Visser 和 Gienapp，  2019 年）。特别是，植物-传粉系统同步性的破坏可能通过花粉限制对植物产生负面影响（Rafferty & Ives，  2011 年；Kudo 和 Ida，  2013 年），并通过减少花卉资源对传粉者产生负面影响（CaraDonna等人，  2018 年；Schenk等人，2018 年）。这可能导致互利共生崩溃（Warren II 和 Bradford，  2014 年）、动物授粉作物产量下降（Bartomeus等人，  2013 年）和局部灭绝（Revilla等人，  2015 年）). 据估计，气候变化引起的物候变化可能导致 17-50% 的传粉媒介物种的可用花卉资源减少，导致历史传粉媒介活动期的一半在没有食用植物开花的时候下降(Memmott et al .,  2007 )。

使这个问题进一步复杂化的事实是，物候对气候的敏感性在其分布范围内的物种内差异很大（Høye等人，  2013 年；Fitchett等人，  2014 年；Park等人，  2019 年；Song等人，  2020 年；Love & Mazer，  2021 年；Pearson等人，  2021 年）。例如，已发现在温带生态系统中许多植物物种的较温暖地区，落叶、开花和结果的时间对温度更为敏感（Zhang等人，  2015 年；Park等人， 2019 年）。因此，随着气候变化，种内相互作用和基因流可能会发生改变（Fox，  2003 年；Rivest等人，  2021 年；Park等人，  2022 年），相互作用的物种之间的物候同步性可能会在整个景观中发生异质性变化，在种群或形态类型 (Park et al .,  2022 )。因此，了解物候反应的种内和种间变异对于评估气候变化的生态影响至关重要。然而，人们对蜜蜂等传粉昆虫的物候反应知之甚少（Bartomeus等人，2011 年）), 迄今为止，大多数研究气候驱动的物候失配的研究都没有考虑种内变异的后果 (Charmantier et al .,  2008 )。这可能部分是由于普遍缺乏在广泛的地理范围内同时跟踪多个相互作用的物种的物候观察实验。事实上，许多关于植物与传粉者相互作用的物候学研究仅限于相对较小的地理区域（Forrest & Thomson，  2011 年；Kudo，  2014 年；Pyke等人，2016 年；Olliff-Yang & Mesler，  2018 年），并且有可能在此类研究中观察到的趋势可能不适用于所有物种的分布范围。

在这里，我们使用超过 120 年的自然历史收藏和公民科学数据，研究物候对气候的敏感性的种内变异如何同时改变物种内部和物种之间的生态相互作用。我们专注于一个独特的系统，包括Claytonia virginica（马齿苋科；弗吉尼亚春季美女）及其专业传粉者Andrena erigeniae（Andrenidae；春季美女矿工蜂）。Claytonia virginica是一种林下春季短命植物，具有从白色到粉红色的独特花色变形，通常出现在同种植物中（Frey，  2004 年）). 不确定这些颜色变形是如何维持的，但传粉者偏好和病原体和食草动物的不同选择已被认为发挥了作用 (Frey,  2004 )。开花物候的变化也可以让不同的颜色变形共存 (Tarasjev,  1997 )，即使是很小的差异也可能通过选型交配导致表型之间的分化和强化 (Hopkins,  2013 )。相对而言，关于C. virginica和A. erigeniae的物候如何随着颜色变体和种群的变化而知之甚少（Schemske等人，  1978 年）), 尽管有人提出在C. virginica的种族之间会发生某种程度的生殖隔离和物候差异(Lewis,  1976 ; Lewis & Suda,  1976 ; Doyle,  1981 , 1983 )。大多数关于植物-传粉者不匹配的研究都集中在通才传粉者上，而对可能更脆弱的专业物种的类似研究很少见，特别是对于寡聚蜂物种，如A. erigeniae（Bartomeus等人，  2011 年；Maglianesi等人，  2020 年） ).

沿着这些思路，我们检验了以下假设：开花物候在白色和粉红色的C. virginica颜色变体中有所不同，有助于它们在同源地的维持；物候对气候的敏感性在C. virginica和A. erigeniae的范围内以及在 C. virginica的颜色变体之间变化；气候变化将导致物种和颜色变形之间更大的时间间隔。