TITLE: Plant-plant interactions change during succession on nurse logs in a northern temperate rainforest
Carrie L. Woods1, Katy Maleta, and Kimmy Ortmann
Biology Department, University of Puget Sound, 1500 N. Warner Ave., Tacoma, WA 98416
Correspondence
Carrie L. Woods, Biology Department, University of Puget Sound, Tacoma, WA, USA. Email: cwoods@pugetsound.edu, ORCID id 0000-0002-2156-3078
ABSTRACT
1. Plant-plant interactions can vary depending on the severity of the environment. Positive interactions, such as facilitation, are often important in early life stages of plants whereas negative interactions, such as competition, predominate in later stages of growth. Thus, through succession, plant-plant interactions often change from facilitative to competitive. In northern temperate rainforests, gap dynamics result in tree falls that facilitate tree regeneration (nurse logs) as well as bryophyte succession. While the importance of nurse logs for tree seedlings is known, how the interactions of bryophyte communities and tree seedlings vary through succession of the log remains unclear.
2. We surveyed tree seedlings, bryophyte community composition, bryophyte depth, and percent canopy cover in 166 plots on nurse logs and the forest floor in the Hoh rainforest in Washington. We examined the relationship of bryophyte community composition, bryophyte depth, percent canopy cover, and tree seedling density on nurse logs of three decay classes and the forest floor.
3. Tree seedling density was highest on young logs with early-colonizing bryophyte species, such as Rhizomnium glabrescens , and lowest on decayed logs with Hylocomium splendens a long-lived moss that can reach depths of >20 cm. As a result, bryophyte depth increased with nurse log decay stage and was negatively associated with tree seedling density. Tree seedling density was 4.6x higher on nurse logs than on the forest floor, which was likely due to competitive exclusion by H. splendens . Nurse logs had 17 species of bryophytes while the forest floor had only six, indicating that nurse logs contribute to maintaining bryophyte diversity.
4. Nurse logs are, therefore, essential for forest dynamics as they enable both tree seedlings and smaller bryophyte species to avoid competition with the dominant forest floor bryophyte, H. splendens . Given that H. splendens has a global distribution and is often dominant in forested systems across the northern hemisphere, it is likely a widespread driver of plant community structure.
Synthesis: Our findings indicate that plant-plant interactions shift with succession on nurse logs from facilitative to competitive and, thus, influence forest community structure and dynamics.
KEY WORDS: determinants of plant community structure; moss; nurse log; Olympic rainforest; plant-plant interactions; Stress Gradient Hypothesis; structural diversity; temperate rainforest
INTRODUCTION
Plant-plant interactions are complex and can vary from facilitative to competitive depending on the severity of the external environment (Bertness & Callaway, 1994; Brooker & Callaghan, 1998; Callaway & Walker, 1997). Most positive plant interactions have been found to occur in severe environments (e.g., arid, salt marshes, polar tundra), and happen through the amelioration of those stressful conditions (Brooker & Callaghan, 1998; Callaway et al., 2002). Facilitation, in particular, appears to be important in colonization and early community development as environments in early succession are typically severe (Callaway & Walker, 1997; Clements, 1916; Connell & Slatyer, 1977; Egler, 1954; Gómez-Aparicio, 2009). As the community develops, the amelioration of disturbance and stress by the physical presence of plants (e.g., stabilizing slopes, herbivory protection, microclimate), results in competitive interactions becoming more important over time (Brooker & Callaghan, 1998; Egler, 1954; Ricklefs, 1977; Walker & Chapin, 1987; but see Maestre, Valladares, & Reynolds, 2005). While vascular plants have largely been the focus of plant-plant interaction research, bryophytes can have facilitative or competitive interactions with vascular plants depending on multiple abiotic and biotic factors (Doxford, Ooi, & Freckleton, 2013; Gornall, Woodin, Jónsdóttir, & van der Wal, 2011; Gough, 2006; Sedia & Ehrenfeld, 2003; Soudzilovskaia et al., 2011; Staunch, Redlecki, Wooten, Sleeper, & Titus, 2012). In locations where bryophytes dominate the ground cover, such as in the arctic and boreal forest, bryophytes can structure the composition of vascular plant communities (Gavini, Suárez, Ezcurra, & Aizen, 2019; Gornall et al., 2011; Gough, 2006). However, how various stages of bryophyte succession influence vascular plants remains largely unexplored.
In old-growth forests, forest regeneration is often triggered by disturbances, such as tree falls, that create gaps into which early colonizing species proliferate (i.e. , small-scale gap dynamics, McCarthy, 2001; Ricklefs, 1977; Yamamoto, 2000). These gaps often result in a change in light levels that facilitate regeneration of early-colonizing shade-intolerant plants if the gaps are large, and shade-tolerant climax species if the gaps are small (< 200 m2); colonization of small gaps by the canopy dominants can perpetuate the current canopy species composition (Denslow, Schultz, Vitousek, & Strain, 1990; McCarthy, 2001; Runkle, 1981). In Appalachian forests in Tennessee, for example, shade-intolerant trees were only able to establish in gaps created by multiple tree falls whereas small gaps were colonized by shade-tolerant trees (Barden, 1981). In contrast, in subalpine fir forests of coastal British Columbia, Pacific silver fir (Abies amabilis ) preferentially colonized all gaps regardless of gap size (Lertzman, 1992). However, western hemlock (Tsuga heterophylla ) was more dominant on stumps in gaps (Lertzman, 1992), suggesting that substrate changes during gap formation could also influence forest dynamics. In temperate coniferous forests of the Pacific Northwest, gap regeneration after tree falls can be quite slow (>25 yr, Spies, Franklin, & Klopsch, 1990), which results in large areas occupied by canopy gaps (13.1%, Spies et al., 1990). The trees that fall to create these gaps become essential sites of forest regeneration (i.e.nurse logs, Franklin et al., 2002) that cover more area in these forests (up to 25%) than in other forests (<4%) (Harmon et al., 1986). Similar to the subalpine fir forests of British Columbia, these nurse logs provide safe germination sites for late succession tree species’, such as shade-tolerant Tsuga heterophylla (Christie & Armesto, 2003; Harmon & Franklin, 1989). Nurse logs are thought to facilitate seedlings through several means, such as by becoming a physical barrier between plants and terrestrial fungal pathogens, limiting competition between seedlings, herbaceous species, and bryophytes on the forest floor, and providing nutrients for potential seedling growth (Franklin, Shugart, & Harmon, 1987; Graham & Cromack Jr., 1982; Harmon, 1986; Harmon & Franklin, 1989). They may also be key to maintaining bryophyte diversity.
Nurse logs are sites for bryophyte colonization and succession. In coniferous forests in Colorado, bryophytes were found to colonize logs after lichen establishment (McCullough, 1948), and in coniferous forests of the Pacific Northwest, epiphytic plants from standing trees were replaced by bryophytes more commonly found on the forest floor (Sharpe, 1956). In the Hoh Rainforest in Washington, bryophyte cover on fallen logs reached approximately 90% in 11-19 y, and succession followed dominance by Dicranum spp. and Hypnum circinale in early succession, Rhizomnium spp. in mid succession, and feather mossesKindbergia oregano and Hylocomium splendens in late succession (Harmon, 1989b). Given that late successional bryophytes are tall and dense enough to prevent tree seedling establishment (Harmon, 1986), they may also out-compete early- and mid-successional bryophyte species. Thus, nurse logs with moderate decay may be a refuge for early- and mid-successional terrestrial bryophytes that have difficulty establishing on the forest floor either due to insufficient light or competition with late successional bryophyte species. Thus, gap dynamics may also be important for the maintenance of bryophyte diversity.
These changes in bryophyte communities may have facilitative or competitive interactions with tree seedlings. In the Hoh rainforest, seedling density was highest when logs were dominated by the mid successional bryophytes Rhizomnium spp. and lowest when logs were dominated by late succession bryophytes (Harmon, 1989b). Harmon & Franklin (1989) argue through their controlled experiments that the interaction of mosses with seedlings changes from facilitative to competitive once the moss layer exceeds 5 cm in depth. Previous research on bryophyte-vascular plant interactions found that thicker mosses can hinder vascular plants by reducing temperatures and nutrient availability (Gornall et al., 2011; Pearce, Woodin, & Van Der Wal, 2003). However, they could also hinder vascular plant seedlings by reducing light levels if the bryophytes are thick and dense enough. In coniferous forests, removal of forest floor mosses showed positive effects on tree seedlings (Wardle, Lagerström, & Nilsson, 2008; Zackrisson, Nilsson, Dahlberg, Jäderlund, & Jaderlund, 1997). Thus, the importance of nurse logs for tree regeneration may be mitigated by bryophyte communities.
Here we build on the findings of previous studies (Harmon, 1986, 1989b; Harmon & Franklin, 1989) by diving deeper into the role of bryophyte species on tree seedling density. Over the course of three years, we conducted separate studies in northern temperate rainforests on the Olympic peninsula in Washington State (Hoh rainforest) to examine the importance of nurse logs for tree seedlings focusing on the effects of the nurse log bryophyte community, how that could change with nurse log decay class, and how it may differ from the forest floor. We tested the following predictions: 1) tree seedling density would be higher on nurse logs than on the forest floor, as shown previously (Christie & Armesto, 2003; Harmon & Franklin, 1989); 2) bryophyte community composition and the depth of bryophyte mats on nurse logs would influence tree seedling densities (Fukasawa & Ando, 2018; Harmon & Franklin, 1989); and 3) the decay class of the nurse log would influence both bryophyte communities and tree seedling densities. Because bryophytes vary in their growth patterns, we also examined if particular bryophyte species were more abundant on the forest floor or nurse logs and were associated with high and low tree seedling density as found previously (Harmon, 1989b). We focused on the dominant feather moss, Hylocomium splendens , given its abundance in these forests and its impact on tree seedling growth in other studies (Fukasawa & Ando, 2018; Harmon & Franklin, 1989). We tested our hypothesis that the effect of H. splendens on seeding growth was due to competition for light by measuring light under and beside this moss. If bryophyte communities on nurse logs change predictably through succession and influence tree seedlings differently at different successional stages, it would further limit the degree of safe regeneration sites in these forests for trees, and highlight the essential role of nurse logs and plant-plant interactions in forest dynamics and diversity maintenance.
MATERIALS AND METHODS