Successional overlap mimics intermediate disturbance in the diversity of a riparian woodland

Background/Question/Methods

That intermediate disturbance can promote ecological diversity has been revisited numerous times – sometimes yielding support, others opposition. Riparian zones are frequently structured by spatially and temporally varying hydrologic processes (especially destructive flooding), often yielding a disturbance gradient along which vegetational diversity may follow an intermediate disturbance pattern – i.e. with colonizer-dominated high-disturbance and competitor-dominated low-disturbance patches less diverse. In contrast, some riparian corridors (including 6^th-order Zoar Valley Canyon in western NY) are instead structured largely by patterns of landform establishment and stabilization, after which vegetational succession may proceed relatively undisturbed. Interestingly, Zoar Valley’s riparian forest diversity is highest at intermediate stand ages (~40 – 150y) and lower in younger (<30y) and older stands (>200y) – much as might be expected if flood disturbance was in fact driving the system. To assess successional and/or disturbance factors in vegetational diversity here, stand composition, age, and structure have been quantified on ~45 ha of floodplains and terraces within Zoar Valley’s riparian zone. Flood, windthrow, and canopy mortality disturbances have also been catalogued, directly over the past 15 years, and through examination of aerial images (1929 – 2007), core-based stand ages, and downed wood volume and orientation.

Results/Conclusions

Woodland diversity in Zoar Valley (H’, 0.35 – 1.84) followed a unimodal distribution across stand age (second order, R² = 0.659). On upper terraces (114 – 301y) diversity was negatively correlated with shade tolerance (Pearson r = -0.706), and declined concurrently with increasing dominance by late-seral sugar maple, beech, and hemlock. Conversely, dominance by pioneer cottonwood, sycamore, willow, etc. reduced diversity on landforms <30y old. Flood disturbance largely ceased by the beginning of stem exclusion at ~40y, but later, as understory reinitiation commenced at ~90y, treefall disturbance emerged. Deadwood volume remained modest until ~150y but increased in multi-aged stands >200y, as did >60-cm diameter gapmakers (to 145m³/ha and 11 logs/ha). Downed wood was sometimes oriented with prevailing winds (Kolmogorov-Smirnoff goodness-of-fit), which may indicate exogenous windthrow disturbance. Thus, the most disturbed seral stages (early = flooding; later = canopy mortality/windthrow) displayed reduced vegetational diversity. In contrast, the very least disturbed stages (stem exclusion through understory reinitiation [~40 – 150y]) displayed maximal diversity. High diversity here appeared to reflect not a mosaic of differentially disturbed patches, but instead an overlap of successional stages – senescing pioneers, emerging late-successional competitors, and a dynamic assemblage of mid-seral components.