Habitat edges decrease plant reproductive output in fragmented landscapes

Abstract

1 INTRODUCTION

Habitat loss is a major threat to biodiversity (Dirzo et al., 2014; Newbold et al., 2015; Tilman et al., 2017). Although negative effects of habitat loss on biodiversity are clear, there is more debate about the effects of habitat fragmentation, which is often confounded with habitat loss (Fahrig, 2017; Fahrig et al., 2019; Fletcher et al., 2018; Haddad et al., 2015). To resolve this debate, examining mechanisms of biodiversity change, such as demographic processes within species, may clarify biodiversity trends in fragmented landscapes (Fletcher et al., 2023; Pardini et al., 2017). Population demography determines species persistence, particularly for small populations, and cumulative responses of multiple species may lead to community-level changes in biodiversity (Paniw et al., 2023; Schmidt et al., 2022). Past fragmentation research on demography has primarily focused on the processes of immigration and emigration (Honnay et al., 2005; Jacquemyn et al., 2002). However, other demographic processes, such as reproductive success, may also be impacted by fragmentation (Aguilar et al., 2019). Given that reproduction is a component of population growth (Koons et al., 2017), fragmentation effects on reproductive output may have important consequences for population persistence.

Because fragmentation results in several spatial patterns that arise at multiple spatial scales (Fletcher et al., 2023), experiments that are able to separate out the effects of these spatial patterns are valuable. For example, as a given amount of habitat is broken apart, the number of habitat patches increases at the landscape scale, which decreases habitat structural connectivity at the among-patch scale (Fletcher et al., 2023). At the same time, fragmenting habitat also creates more edge habitat, increasing the edge-to-area ratio at the patch and landscape scale and decreasing the average distance to an edge at the within-patch scale (Fletcher et al., 2023). These multiple components of fragmentation may each influence plant reproductive output (i.e. seed production), through impacts on pollination, growth, seed predation, or herbivory (Brudvig et al., 2015). However, despite broad recognition that effects of habitat loss and fragmentation are often confounded (Ewers & Didham, 2005; Fahrig, 2003; Valente et al., 2023), disentangling their effects remains challenging. Previous research on plant reproductive output has typically focused on patch size to test fragmentation effects (Bruna & Kress, 2002; Portela et al., 2021; Tomimatsu & Ohara, 2010), confounding multiple components of fragmentation with habitat loss. Experiments designed to separate the effects of multiple components of fragmentation from habitat loss will clarify the mechanisms of population demography change in fragmented areas, as we do here using an experimentally fragmented system.

Habitat fragmentation creates disconnected populations in isolated patches, which may reduce reproductive output for plants through disruption of pollen movement (Betts et al., 2019). Pollination is a key process for the vast majority of plant species' reproductive success (Friedman & Barrett, 2009; Ollerton et al., 2011), meaning that disruptions to pollination under landscape change can have negative consequences for plant reproductive output. Spatial isolation of populations by fragmentation may reduce pollen movement (Hadley & Betts, 2012), subsequently reducing gene flow and leading to a higher probability of inbreeding (Aguilar et al., 2019; Rosas et al., 2011). Both wind-pollination and insect-pollination may be decreased by fragmentation but through different mechanisms. Pollination for species dependent on plant–pollinator mutualisms is directly tied to fragmentation effects on their pollinators, with pollen movement corresponding to pollinator response (Kormann et al., 2016). Connectivity between patches facilitates movement for pollinators (Tewksbury et al., 2002), increasing pollen movement for insect-pollinated species (Townsend & Levey, 2005). However, for wind-pollinated species, abiotic conditions created by fragmentation such as increased edge and isolation may be the limiting cause of pollination through changing wind dynamics (Aguilar et al., 2019; Damschen et al., 2014). Structural connectivity of open habitats increases wind movement between patches, especially when aligned with predominant winds (Damschen et al., 2014), which may facilitate the movement of pollen between discrete populations (Provan et al., 2008). However, because of variation in species responses to fragmentation (Ewers & Didham, 2005; Fischer & Lindenmayer, 2007), more work is needed to understand whether patterns of pollination are consistent among pollination modes, as well as to disentangle the impacts of multiple fragmentation components on pollination that may confound fragmentation effects (Brudvig et al., 2015; Heinken & Weber, 2013; Newman et al., 2013).

Although pollen movement is often considered in the context of fragmentation, fragmentation may also affect plant reproductive output through population-level shifts in flowering and phenology. Edge habitat often hosts unique microclimate conditions, changing abiotic conditions, such as temperature, moisture, and light availability (Tuff et al., 2016). Because plant growth and flowering are highly determined by abiotic conditions, these abiotic changes could impact plant flowering and seed production (Galloway & Burgess, 2012; Müller et al., 2021; Suzán-Azpiri et al., 2017). Additionally, plant fitness can be affected indirectly through edge effects on insect visitors. Pollinators and insect herbivores may be affected by abiotic edge conditions, further impacting seed set and plant growth (Andrieu et al., 2018; Levey et al., 2016; Ren et al., 2023). As demographic structure (e.g. proportion of flowering individuals) and reproductive output can contribute to population growth (Caughlin et al., 2019), edge effects on plant flowering and seed production may impact plant population dynamics (Bruna & Kress, 2002; Suzán-Azpiri et al., 2017).

Plant population growth is determined by several demographic rates, including fecundity, establishment, survival, and growth (Sibly & Hone, 2002), which all may be affected by habitat fragmentation (Bruna & Oli, 2005; Honnay et al., 2005). However, the relative importance of these demographic rates for population dynamics may vary depending on the species' life history, local abiotic environment, and biotic interactions, among other factors (Crone, 2001; de Kroon et al., 1986). As such, seed production may be highly important for population growth and persistence if a species is seed limited, but less important if habitat conditions constrain survival or growth instead (Clark et al., 2007). Within our experimental system of longleaf pine savanna habitat, previous work has found that for two long-lived perennial species, seed production was the most important demographic parameter for predicting population growth (Caughlin et al., 2019). However, for an early-successional species, microsite conditions and seed predation were more significant than seed abundance (Orrock et al., 2006), highlighting the variability of demographic driver significance, even within a system. As a whole, although the relative importance of seed production for plant population persistence may vary, measuring reproductive output provides insight into how one component of demography may be impacted by landscape alterations (Bruna & Kress, 2002; Caughlin et al., 2019; Suzán-Azpiri et al., 2017).

Here, we test how fragmentation affects plant reproductive output, looking at fragmentation effects on plant flowering, pollination, and seed production. We worked in a large-scale, replicated fragmentation experiment designed to manipulate three aspects of fragmentation: among-patch connectivity, patch-scale edge-to-area ratio, and within-patch distance from an edge. We experimentally planted three wind-pollinated and two insect-pollinated plant species to ask (1) Do connectivity, edge-to-area ratio, and distance from an edge affect the likelihood of a plant flowering and flower abundance? (2) If a plant flowers, do connectivity, edge-to-area ratio, and distance from an edge affect pollination rate and seed production? We expected a reduction in plant reproductive output (seed production) near habitat edges and in unconnected patches. Specifically, in our system with open-habitat patches and forested matrix, we expected abiotic effects of canopy shading from the edge to decrease flowering, decreasing plant reproductive output near edges. Additionally, we expected pollination to be reduced in unconnected patches due to a disruption of pollen movement for both wind-pollinated and insect-pollinated species.