Distribution and Abundance of Ground-layer Bryophytes in Mature Forests of Western Washington

Shelley A. Evans, Donald McKenzie, and Charles B. Halpern College of Forest Resources Box 352100 University of Washington Seattle, WA  98195-2100 saevans@u.washington.edu

Introduction

Bryophytes are among the most diverse members of the forest flora in the Pacific Northwest, yet many aspects of their distribution and ecology remain poorly understood.  In this study we use pre-treatment data from four of the DEMO experimental blocks to explore the factors that shape the patterns of abundance, diversity, and distribution of ground-layer bryophytes in coniferous forests of western Washington. Our work contributes to a small, but growing literature on the abiotic and biotic controls on bryophyte composition in this region. Most previous work has compared bryophyte communities in young and old forests or in managed and remnant old-growth forests.  Our studies explore distributional patterns within mature forests (65-170 yr old) across a diversity of physical environments and forest types.  We address three primary questions: 

• Do the composition and patterns of diversity of ground-layer bryophyte communities differ among mature forests of the region?
• Can we predict the cover and richness of bryophyte communities from local patterns of vegetation structure and environment?
• Which substrates contribute most to the abundance and diversity of bryophyte species; do individual species show an affinity for particular substrates?

• We observed a total of 78 bryophyte taxa (56 moss and 22 liverwort species).
• Of the full set of taxa, 23 were common to all blocks, but 22 were unique to a single block.
• Representation of liverworts varied markedly among blocks (19-36% of the flora).
• Detrended correspondence analysis revealed substantial separation among blocks in ordination space with Capitol Forest plots completely isolated from those of the Cascade Range blocks (Fig. 1). 

• Patterns of richness and distribution of species abundance were fairly similar among blocks (Fig. 2).

Predictive models of bryophyte richness, cover, and composition

• Multiple regression models explained little of the plot-level variation in bryophyte richness or total cover (r² < 0.23).  Richness was positively correlated with log cover and richness of herb species, but negatively correlated with slope. 
• Similarly, canonical correspondence analysis (CCA) indicated that little of the variation in species composition could be explained by stand structural or environmental variables.

Role of substrates

• Although fine litter was the most abundant substrate (75-92% cover), other substrates (e.g., decayed logs and the bases of trees) contributed substantially to bryophyte frequency and richness (Fig. 3).

Figure 3. Mean number of bryophyte taxa per plot supported by each substrate (circles; right axis) relative to mean cover of each substrate (gray bars; left axis).

• Indicator species analysis (ISPAN) revealed many significant associations with substrates (32 of 130 tests): 
• Species were most often associated with decayed logs, and occasionally with fine litter, rocks, or the bases of trees.
• 7 species showed significant associations with the same substrate in more than one block, but 5 showed significant associations with different substrates (“switching” primarily between decayed logs and tree bases).

Significance and Management Implications

• Mature forests in southwestern Washington can support distinctly different bryophyte assemblages.  Many factors contribute to these differences (e.g., biogeography, physical environment, stand structure, disturbance history, dispersal limitation), but it is difficult to tease apart their relative importance.
• Given the “unpredictable” nature of species composition, site-specific surveys must be performed to understand the consequences of forest management for particular species or groups of species. 
• At a local scale (plots within forest stands) it is difficult to predict bryophyte cover or richness from other vegetation or environmental characteristics.  Had we considered a broader range of stand ages or structures, predictive models may have been stronger.
• Our analyses of substrate relationships confirm the results of previous studies in temporal and boreal forests:  downed wood (particularly well-decayed logs) and the bases of trees can contribute significantly to the diversity of ground-layer bryophytes.  However, we also observed that species “preferences” for substrates may vary from location to location.