Date of Award


Degree Name

Biological Sciences


College of Science

Type of Degree


Document Type


First Advisor

Frank S. Gilliam

Second Advisor

Dan K. Evans

Third Advisor

Donald C. Tarter

Fourth Advisor

Mary Beth Adams


The herb layer (vascular plants ≤1 m in height) of montane forests responds to a variety of environmental gradient conditions. This study compared herb layer characteristics, soil fertility, and plant-soil nutrient interactions of two young (~20 yr) and two mature (~>70 yr) watersheds of the Fernow Experimental Forest, Parsons, West Virginia. Fifteen 0.04-ha circular plots were established in each of four watersheds [WS4 control, >80 yr; WS13 select cut, >65 yr; WS3 clearcut, ~20 yr; WS7 clearcut/herbicide, ~20 yr] to sample the extremes of watershed elevation and aspect. Cover (%) was measured for all vascular species (≤1 m in height) within 10 1-m2 subplots in each plot. The two subplots with the greatest cover per plot were harvested for biomass. Harvested material was separated by species, oven-dried, and weighed. Regressions of cover and biomass were used to estimate biomass for nonharvested subplots; cover was significantly correlated with biomass (r2=0.62). One 10-cm soil sample was taken from each harvested subplot. Tissue and soils were analyzed for nutrient concentrations. All living woody stems ≥2.5 cm diameter breast height (dbh) were measured for dbh and categorized as overstory (>10 cm dbh) or understory.

Herb layer cover was significantly higher on WS7 (37.5%) due to an abundance of Dryopteris marginalia and Polysticbum acrostichoides on this watershed. Herb layer cover on WS4, WS3, and WS13 was 26.4, 19.3, and 17.9%, respectively. Herb layer biomass was highest on WS7 (18.5 g/m2) followed decreasingly by WS4 (13.3), WS3 (9.7) and WS13 (9.1). Herb layer vegetation of WS4, WS3, and WS13 was dominated by Laportea canadensis, Acer pensylyanicum, and Viola spp. Mean values for species richness (~3.8 species/m2/plot) and diversity (H′=~1.7) were comparable among the watersheds.

Soils of all watersheds were acidic sandy loams of similar fertility. Canonical discriminant analysis (CDA) of soil physical and chemical properties revealed that WS13 soils were distinctly lower in clay and N03-N content compared to the other watersheds. CDA of herb layer tissue nutrients showed that herbaceous species were higher in N, K, and Mg than woody herb layer species. Correlations among herb layer tissue nutrients showed significant relationships between N:P (all watersheds), Mg:Ca (WS3 and WS13), and N to base cations (WS3, WS4, and WS13), suggesting N-limitation in these forest soils. Strong correlations between soil and tissue nutrients were seen only on WS3 and WS4, yet soil nutrient availability appeared to greatly influence herb layer species niche differentiation.

Principal component analysis (PCA) of individual watersheds based on plot soil and canopy data and correlating cover to each PCA axis revealed that herb layer cover was negatively correlated with canopy characteristics (e.g. understory density) on WS4 and WS13 and positively correlated with soil nutrients (e.g. Mg, Mn, Zn) on WS3 and WS7. Relative importance of herb layer vegetation shifted from high herbaceous species, low woody species cover in the younger stands to increased woody species cover in the mature stands. These data suggest that in early forest succession, light availability is relatively high and uniform and herb layer development is nutrient-limited with herbaceous species out-competing woody species for available nutrients. In later stages of succession, as canopy stratification and closure increase, the herb layer becomes limited by light availability and woody species are superior competitors.


Forest ecology - West Virginia.

Biotic communities - West Virginia.

Fernow Experimental Forest (Tucker County, W. Va.)