Multiple disturbances, multiple legacies: Fire, canopy gaps and deer jointly change the forest seed bank

Abstract

1 INTRODUCTION

Forests around the world have experienced substantial changes in their historic disturbance regimes with altered land use, climate change or forced removal of Indigenous peoples (Bowman et al., 2011; Gilliam, 2016; Götmark, 2013; Kelly et al., 2023). Many mesic temperate forests in North America, Europe and Asia have become more even aged, undergone severe fire suppression and experienced increased ungulate browsing (Carpio et al., 2021; Frelich, 2002; Hai et al., 2023; McDowell et al., 2020; Pascual-Rico et al., 2021). These altered conditions create adverse environments for the plant species that coevolved and depend on historic disturbance patterns, such as globally dispersed oak (Quercus) species (Carrero et al., 2020; Tinner et al., 2005). Changes in plant composition with altered disturbance regimes have led managers to restore or manipulate disturbance to support biodiversity and ecosystem function (Long, 2009; Stanturf et al., 2014). However, our understanding of how the reintroduction of multiple historic disturbances influences biodiversity is nascent and represents a key knowledge gap in our long-term management and restoration of temperate forest systems.

Mesic North American forests are expansive ecosystems that have experienced dramatic alterations in their disturbance regimes over the last century (Abrams, 2005; Hanberry & Nowacki, 2016; Vander Yacht et al., 2020; Webster et al., 2018). This scenario is particularly acute in Appalachian hardwood forests, which have lost oak (Quercus spp.) tree regeneration and are transitioning to wetter, maple-dominated (Acer spp.) systems (Nowacki & Abrams, 2008; Pile Knapp et al., 2024). This transition from oak to maple forests was initiated by the forced removal of Indigenous peoples and their use of cultural burning as a management tool (Abrams et al., 2021; Pile Knapp et al., 2024; Poulos, 2015). This was followed by mass deforestation and slash wildfires in the late 19th and early 20th century (Lafon et al., 2017). Negative perceptions of these wildfires led to a century of state-sanctioned fire exclusion and suppression that favoured maple growth and wetter understories (Alexander et al., 2021; Arthur et al., 2021). As a result, Appalachian forests became dominated by even-aged stands with few mid-sized and large (>15-m diameter; >175 m²) canopy gaps and infrequent low-intensity fires (Clebsch & Busing, 1989; Nowacki & Abrams, 2008; Raymond et al., 2009). In regions of Appalachia, the fire return interval is now over 10,000 years, as opposed to the historic 1-to-2-decade fire return interval under Indigenous stewardship and with lightning-ignited fires (Lafon et al., 2017).

Concurrently, white-tailed deer (Odocoileus virginianus) populations have increased dramatically above historical baselines (above 4 to 8 deer/km²) in most of eastern North America, driving ecological change depending on their population density, similar to the effects of overabundant cervid populations in many other areas in Europe and Asia (Côté et al., 2004; Iijima et al., 2023; Reed et al., 2022; Valente et al., 2020). To reverse the long-tailed effects of historic management and sustain oak-dominated plant communities, forest managers are reintroducing disturbances like prescribed burns, canopy gap creation through tree harvesting and lowering deer densities through hunting or fencing off vulnerable areas (Nuttle et al., 2013; Raymond et al., 2009).

Reintroducing multiple disturbances can be a powerful tool in efforts to restore and direct change within ecological communities (Abrams et al., 1985; Batllori et al., 2019; Reed et al., 2023; Sasaki et al., 2015; Yantes et al., 2023). For instance, combined low-intensity fire and canopy gap creation can lead to greater oak growth in both North America and Europe, while these disturbances alone are less effective (Brose et al., 2013; Hutchinson et al., 2024; Izbicki et al., 2020; Petersson et al., 2020). In this example, the surviving oak trees represent a post-disturbance legacy, which is broadly characterized as the adaptations, individuals and biomass that remain on the landscape following a disturbance (Cuddington, 2011; Franklin et al., 2000). Disturbance legacies can be material (e.g. wood and nutrient pools) and informational (e.g. species' adaptive responses and genetic material), although the categories are not mutually exclusive (Johnstone et al., 2016). Each disturbance that occurs in a given area modifies the legacy community of the previous disturbance, and in certain cases, the disturbance combination and timing may lead to unique communities depending on how the disturbances in question interact (Anoszko et al., 2022). Thus, in eastern US forests and in temperate forests around the world, the disturbance legacies of combined low-intensity fire, canopy gap creation and ungulate browsing may have a particularly influential role in determining how forests reorganize and develop into the future when compared to the legacies of these disturbances individually (Cuddington, 2011; Seidl et al., 2014; Turner & Seidl, 2023).

To this end, the soil seed bank represents an important, but understudied, entity that may be strongly influenced by the reintroduced disturbances and may influence future disturbance regimes (Archibold, 1979; Ferrandis et al., 1996; Morgan & Neuenschwander, 1988; Pakeman & Small, 2005; Sousa, 1984). Seed banking is a reproductive adaptation that allows plants to persist belowground as dormant seeds, wherein the soil serves as a buffer from aboveground disturbances (Baskin & Baskin, 2022; Thompson, 1987). The forest seed bank has been shown to be a reservoir of biodiversity in temperate forests around the world, holding many herbaceous and woody early successional species (Grubb et al., 2013; Plue et al., 2010; Yang et al., 2021). The seed bank is also a latent source of genetic diversity (Levin, 1990; McCauley, 2014), making the seed bank both a material and information legacy.

Germinated plants that survive a disturbance eventually mature and release seeds, reestablishing the seed banking process that allows for plant communities to reorganize with future disturbance, thereby setting another legacy depending on the seeds that are returned to the soil (Baltzer et al., 2021; Falińska, 1999; Grubb, 1988; Hyatt & Casper, 2000; Seidl & Turner, 2022). Hypothetically, more disturbance will lead to a seed bank that is more similar to aboveground vegetation, as the herbaceous layer is homogenized and a few ruderal species survive and reproduce (Ma et al., 2021). These changes in the seed bank with disturbance can have long-lasting ecological ramifications. For example, rampant timber harvesting and slash wildfires in the United States during the late 19th and early 20th centuries likely allowed the shrub Rubus to spread and saturate forest seed banks with its long-lived seeds, creating a century-old legacy of heavy Rubus regeneration following overstory disturbance throughout the eastern United States (Dunn et al., 1982; Peterson & Carson, 1996). Rubus can then survive as a recalcitrant understory for decades (Donoso & Nyland, 2006; Kern et al., 2012).

Prescribed burns, canopy gap creation and deer browsing each provide a unique and important opportunity for new vegetation to grow from the seed bank and for the seed bank to change (Gioria et al., 2022; Muscolo et al., 2014; Schuler, 2010). Prescribed fires clear plant material, catalysing seed germination with increased light, heat, smoke and nutrients (Keeley & Fotheringham, 2000; Ooi, 2012; Pausas et al., 2022). In fire-prone ecosystems throughout the world, Pausas and Lamont (2022) found that ≈42% of seed-banking species are adapted to germinate with heat or smoke. Canopy gaps increase understory resources like light, soil temperature and soil moisture, which are critical for seeds to germinate (Dalling & Brown, 2009; Pakeman & Small, 2005). Both fire and canopy gaps result in a temporary depletion of seeds in the seed bank as plants germinate, but over time, newly established vegetation will grow, reproduce and replenish the seed bank (Auld & Denham, 2006; Shinoda & Akasaka, 2020). This replenishment process may be disrupted by ungulate herbivores, as chronic over-browsing can constrain seed set, reduce plant abundance and lower long-term understory plant diversity by shifting composition to browse tolerant species (Brody & Irwin, 2012; Pendergast et al., 2016). These direct consumptive effects may indirectly reduce the abundance and diversity of seed banking species (Beauchamp et al., 2013; Tamura, 2019). However, in regions where deer populations are low and similar to historic estimates, deer browsing has been shown to increase understory diversity by reducing the abundance of otherwise competitive ruderal species, which could then lead to a more diverse seed bank (Royo et al., 2010; Yacucci et al., 2024).

Despite the increasing prevalence and co-occurrence of experimental tests of reintroduced disturbances in the eastern United States and in temperate systems more broadly (Kleinman et al., 2019; Thom & Seidl, 2016), our understanding of how individual and combined low-intensity fires, canopy gaps and ungulate herbivores change long-term forest seed banks is minimal. This highlights a significant gap in our understanding of post-disturbance legacies, as seed banks are critical for maintaining forest biodiversity in light of disturbance. Therefore, the primary question guiding our research is: Do multiple reintroduced disturbances cause more substantial long-term changes in the seed bank than each respective individual disturbance? To test this question we used a unique, multi-disturbance forest experiment that factorially manipulated low-intensity fire via controlled burning, canopy gap creation via girdling and herbicide injection and deer density via fenced exclosures. Thirteen years after the experiment's initiation, we sampled the seed bank in each disturbance combination treatment and tested how seed composition varied by disturbance treatment and in comparison to extant vegetation at multiple time points.

We expected low-intensity fire to be the predominant driver of increased seed density and diversity, as the Appalachian ecosystem has historically experienced frequent low-intensity burns and many plant species are likely favoured by fire (H1). Similarly, we expected canopy gaps to lead to a modest increase in seed bank density and diversity, mirroring the increased aboveground plant diversity with canopy gaps noted by Royo et al. (2010) (H2). We hypothesized that fire combined with canopy gaps would cause the greatest increases in seed bank density and diversity, leading to concomitant changes in seed community composition (H3). Based on studies showing negative impacts of deer herbivory on aboveground plant growth and reproduction, we expected deer to have a negative effect on seed bank density and diversity, particularly when combined with fire (H4). Lastly, when comparing the seed bank to extant vegetation, we expected the seed bank community to be most similar to extant vegetation in highly disturbed plots, as many seed banking species are favoured by disturbance and may have been able to saturate the seed bank (H5).