Introduction

The most pressing environmental issues of our times are understanding the ecological effects of ongoing climate change and predicting the ensuing implications for maintaining biodiversity (Lovejoy & Hannah, 2019). Foundational to these issues are an understanding of species range limits, which are the geographic limits to species’ spatial distributions. Projection inaccuracy of how species ranges will shift with climate change is alarming (Urban, 2019), and it is unclear whether our inability to project species range shifts comes from our poor knowledge of which climatic variables are important for which species or from the many other factors that can influence range shifts (e.g. microhabitat variation, dispersal limitation, species interactions, and many others). Even explaining the mechanisms underlying observed shifts is difficult, and it is clear that responses are highly species-specific (Freeman et al., 2018; Rumpf et al., 2018).
To address this issue, ecologists are urgently seeking to understand species niche limits in order to project species distribution shifts with climate change. These are commonly predicted with correlative Species Distribution Models (SDMs; Elith et al., 2010; Pacifici et al., 2015), which correlate species occurrences to current macro-climate in order to predict the relative probability of occurrence in space and time (Wiens et al., 2009). Such models have become an integral part of biogeographical research and progress has been made to address some of their shortcomings (Franklin, 2023). While these models implicitly assume that they include sufficient niche variables to accurately describe a species distribution, few models actually test what niche components are necessary to quantify population growth rate and shape species distributions (Pulliam, 2000).
A main driver of population growth rate is the fine-scale microhabitat that mediates the success of individual plants through abiotic (e.g., canopy cover, above-ground temperature, soil moisture) and biotic (e.g., soil biota, soil nutrients) conditions (Kephart & Paladino, 1997; Zurbriggen et al., 2013; Oldfather & Ackerly, 2019; Tanner et al., 2021; Sanczuk et al., 2023; Allsup et al., 2023; Kemppinen et al., 2023). However, studies that quantify population response to microhabitat variation along species ranges are rare, impeding an understanding of which aspects of fine-scale microhabitat are important in shaping species broad-scale distributions (but see Tourville et al., 2022). Despite this, recent work has still shown that incorporating microhabitat into SDMs can provide substantially improved predictions (Maclean & Early, 2023).
How microhabitat and microclimate themselves vary along large elevational, and therefore macroclimatic, gradients that characterize species ranges is poorly understood, and research is increasingly finding that microclimate is often decoupled from macroclimate (Scherrer & Körner, 2010; Ford et al., 2013; Lembrechts, 2023). It is, however, now widely accepted that microclimate differs from macroclimate due to factors such as forest canopy cover (De Frenne et al., 2019; Haesen et al., 2021) and topography (Lawson et al., 2014). This can modify species response to habitat use (Lawson et al., 2014) and create climate change refugia (Dobrowski, 2011; Pradhan et al., 2023) that mediate species response to climate change (Morelli et al., 2020). Very few studies have been able to document how variable plant-relevant environmental drivers of performance are across macroscale gradients, although substantial progress is underway in including microclimate as part of biogeographical research (Kemppinen et al., 2023).
Testing how sensitive species’ life stages are to variations in microhabitat can address many of the sources of bias that distribution models face and accounting for demographic processes can greatly improve our understanding of range dynamics (Fig. 1; Normand et al., 2014; Copenhaver-Parry et al., 2020). For plant species, how early life stages (i.e. recruitment and seedling survival; henceforth establishment) respond to microhabitat may be the key to determining species’ ability to expand their ranges (Kroiss & HilleRisLambers, 2015), as establishment is essential for a range expansion to occur. Understanding how early life stages respond to variations in microhabitat is thus critical to understanding microhabitat suitability for range shifts and population persistence, especially since life stages can respond differently to environmental conditions (Goodwin & Brown, 2023). Yet, few studies test how microhabitat influences establishment, and important microhabitat variables, such as soil moisture and temperature, are often left out of such work (Schurr et al., 2012).
Species show remarkable variability in their responses to climatic aspects of microhabitat (Bullied et al., 2012) and aspects of microhabitat not directly influenced by climate (Selmants et al., 2016; Castro et al., 2022) . There are no general trends, except for that diverse components of microhabitat are important for establishment (Table S1). With regards to climatic aspects of microhabitat, soil moisture and temperature, plant-height temperature, and snow duration can affect establishment in a myriad of ways (Szeicz & Macdonald, 1995; Thompson & Naeem, 1996; Yates et al., 1996; Graae et al., 2009; Santana et al., 2010; Rodríguez-García et al., 2011; Bullied et al., 2012; Moyes et al. 2013; Caldeira et al., 2014; Mondoni et al., 2015; Renard et al., 2016; Kueppers et al., 2017; Andrus et al., 2018; Elliott & Petruccelli, 2018; Lett & Dorrepaal, 2018; Dolezal et al., 2021; Ósvaldsson et al., 2022)
How early life stages respond to microhabitat variables not directly related to climate is also variable by species, with linear, unimodal, or no responses to canopy cover (Lloret et al., 2005; Käber et al., 2021), soil fungus and bacterial content (Rigg et al., 2016; van der Heijden et al., 2016; Tobias et al., 2017; Xi et al., 2018), soil carbon and nitrogen (Monaco et al., 2003; Pérez-Fernández et al., 2006; Li et al., 2011; Pröll et al., 2011; Bateman et al., 2017; Kołodziejek et al., 2017; Zhong et al., 2019), and water holding capacity (Moser et al., 2017; Smithers, 2017; James et al., 2019; Khurana & Singh, 2000). However, studies examining responses to microhabitat beyond range edges are rare and can show inconclusive effects of microhabitat (Lee-Yaw et al., 2016), highlighting the need to measure responses to microhabitat along multiple species’ ranges.
To comprehensively understand how microhabitat, and other factors, influence species distributions, transplant experiments beyond species’ ranges have been suggested as a promising approach (Lee-Yaw et al., 2016; Morris & Ehrlén, 2015). Transplant experiments show either congruence with SDM predictions (Sanczuk et al., 2022) or markedly different responses of populations than predicted by SDMs (Greiser et al., 2020), highlighting the importance of these experiments. Such experiments are particularly valuable at the leading edge of a species range (i.e. edge expanding with climate change), where novel species interactions are likely to be found (Thuiller et al., 2008). Transplant experiments often find that species are establishment, not dispersal, limited, and this can be due to unfavorable microsite conditions (Clark et al., 2007; Davis & Gedalof, 2018). How microhabitat suitability is distributed beyond species current leading edge therefore likely determines species’ ability to shift their ranges (Tourville et al., 2022), yet few studies quantify microhabitat along and above species distributions. We utilized the large elevational macroclimatic gradients of the West and East sides of the Washington Cascade Range, USA for a seed transplant experiment encompassing common grasses, forbs, shrubs, and trees to ask:
  1. How does microhabitat variation within sites compare to among-site macroclimate variation, and does microhabitat covary expectedly along elevational gradients?
  2. Does microhabitat explain establishment of common species in our system, and if so, are responses consistent among species?
  3. If microhabitat variables explain establishment, does the distribution of microhabitats along macroclimatic gradients suggest range shifts will be facilitated, constrained, or unaffected by the availability of suitable microhabitat?