Near-term spatial reorganization of forest ground-layer vegetation in response to experimental forest gap creation

Background/Question/Methods
We examined the effects of experimental canopy opening treatments designed to mimic the canopy structure of old-growth forests in a second-growth forest in northern Wisconsin. We sampled the response of five guilds of forest herbs the first growing season following treatments. Replicated treatments (n=5) included small (50 m²), medium (200 m²) and large (380 m²) canopy openings, a mechanical treatment that consisted of driving harvesting equipment over permanent plots to simulate the gap treatment without removing the overstory, and a control treatment. Guild classification was based on species phenology and life form. Though dominated by sugar maple (Acer saccharum), site-level variation in overstory composition and structure led us to partition treatments among three blocks according to the importance of co-dominant tree species: Hemlock (Tsuga canadensis), Ash (Fraxinus spp.) and Other (Tilia americana and Carya cordiformis) that correspond to a hypothesized productivity gradient. The difference in percent cover between pre- and post-treatment surveys of: spring ephemerals, early summer and late summer herb species, ferns, and graminoids was examined according to our fully nested split-split plot design, using hierarchical linear modeling and post-hoc comparisons.

Results/Conclusions

Responses varied significantly among scales and guilds. Experimental treatments interacted with overstory composition and structure to affect the productivity response of summer-green herbs and ferns. For early summer herbs, cover increased among all treatments and blocks. Increases in productivity among control plots corresponded to the hypothesized productivity gradient in the control treatment; productivity was marginally greater in the Other block compared to the Ash block (p=0.09) and the Hemlock block (p=0.04). However, for gap-treated plots, increases in productivity were greatest in the Ash and Hemlock blocks followed by the Other blocks (p=0.08). In Ash blocks, productivity differed significantly between Gap and Mechanized treatments (p=0.04) but not between Gap and Control treatments (p=0.12), corresponding to the marginally significant treatment main effects (p=0.07). For ferns, significant interactions among treatments and blocks (p=0.04) resulted from observed increases in fern cover within Gap treatment plots in the Ash block (p<0.05) while all other treatment block combinations do not vary significantly among each other (p>0.05). Significant higher-order interactions in both summer-green herbs and ferns suggest an importance of fine-scale mechanisms. We did not observe any significant main effects or interactions on the change in percent cover of spring ephemerals, late-summer herbs or graminoids. We infer that phenology and life form are important characteristics affecting the response of forest herbs to disturbance.