No abstract

No abstract

Climate change and forest disturbances are threatening the ability of forested mountain watersheds to provide the clean, reliable, and abundant fresh water necessary to support aquatic ecosystems and a growing human population. Here, we used 76 years of water yield, climate, and field plot vegetation measurements in six unmanaged, reference watersheds in the southern Appalachian Mountains of North Carolina, USA to determine whether water yield has changed over time, and to examine and attribute the causal mechanisms of change. We found that annual water yield increased in some watersheds from 1938 to the mid-1970s by as much as 55%, but this was followed by decreases up to 22% by 2013. Changes in forest evapotranspiration were consistent with, but opposite in direction to the changes in water yield, with decreases in evapotranspiration up to 31% by the mid-1970s followed by increases up to 29% until 2013. Vegetation survey data showed commensurate reductions in forest basal area until the mid-1970s and increases since that time accompanied by a shift in dominance from xerophytic oak and hickory species to several mesophytic species (i.e., mesophication) that use relatively more water. These changes in forest structure and species composition may have decreased water yield by as much as 18% in a given year since the mid-1970s after accounting for climate. Our results suggest that changes in climate and forest structure and species composition in unmanaged forests brought about by disturbance and natural community dynamics over time can result in large changes in water supply.

Abstract. In contrast to stable inland systems, coastal landscape positions are dynamic, changing as shorelines migrate and storms alter topography. We define landscape position by distance to ocean shoreline and elevation above sea level, two metrics that integrate a suite of environmental and biotic factors. As shoreline and elevation change, suitability of a geo-referenced position for a given plant species may also change. The objectives of our study were to use two methods for measuring landscape position (GPS and hyperspectral/light detection and ranging or LIDAR) to develop habitat polygons, compare habitat polygons for five species representing several adaptive strategies, and illustrate change in landscape position due to migrating shoreline for a Virginia, USA barrier island. Habitat polygons for each species were distinct, represented several growth forms or functional groups, and were indicative of tolerances to biotic and abiotic stresses. The habitat polygon for Cakile edentula (annual forb) was relatively small, indicating narrow habitat requirements for the strand environment. Cirsium horridulum (biennial forb), with succulent shoots and roots, occurred on dunes where water is most limiting. For the dunebuilding grass, Ammophila breviligulata, as distance from shoreline increased, minimum elevation also increased. Two woody species occurred across the entire island; however, Morella cerifera (N-fixing shrub), was limited to mesic swales whereas Juniperus virginiana (evergreen tree), with the largest habitat polygon, occurred on both dunes and swales. For a geo-referenced point on the north end of Hog Island, distance to shoreline increased from the shoreline to 1100 m inland over 139 years. In contrast, the geo-referenced point on the eroding portion of the island decreased from 1700 m to 120 m from the ocean shoreline over the same time period. Where sea level rise and storms are expected to alter shorelines and island topography, generation of habitat polygons from hyperspectral and LIDAR imagery provide rapid assessment of potential effects on species distribution patterns at local and regional scales. Habitat polygons have broad applicability beyond coastal systems and may contribute to a rapid assessment or identification of vulnerability for species as climate patterns shift through time.

Infestation of eastern hemlock (Tsuga canadensis (L.) Carr.) with hemlock woolly adelgid (HWA, Adelges tsugae) has caused widespread mortality of this key canopy species throughout much of the southern Appalachian Mountains in the past decade. Because eastern hemlock is heavily concentrated in riparian habitats, maintains a dense canopy, and has an evergreen leaf habit, its loss is expected to have a major impact on forest processes, including transpiration (E(t)). Our goal was to estimate changes in stand-level E(t) since HWA infestation, and predict future effects of forest regeneration on forest E(t) in declining eastern hemlock stands where hemlock represented 50-60% of forest basal area. We used a combination of community surveys, sap flux measurements, and empirical models relating sap flux-scaled leaf-level transpiration (E(L)) to climate to estimate the change in E(t) after hemlock mortality and forecast how forest E(t) will change in the future in response to eastern hemlock loss. From 2004 to 2011, eastern hemlock mortality reduced annual forest E(t) by 22% and reduced winter E(t) by 74%. As hemlock mortality increased, growth of deciduous tree species--especially sweet birch (Betula lenta L.), red maple (Acer rubrum L.), yellow poplar (Liriodendron tulipifera L.), and the evergreen understory shrub rosebay rhododendron (Rhododendron maximum L.)--also increased, and these species will probably dominate post-hemlock riparian forests. All of these species have higher daytime E(L) rates than hemlock, and replacement of hemlock with species that have less conservative transpiration rates will result in rapid recovery of annual stand E(t). Further, we predict that annual stand E(t) will eventually surpass E(t) levels observed before hemlock was infested with HWA. This long-term increase in forest E(t) may eventually reduce stream discharge, especially during the growing season. However, the dominance of deciduous species in the canopy will result in a permanent reduction in winter E(t) and possible increase in winter stream discharge. The effects of hemlock die-off and replacement with deciduous species will have a significant impact on the hydrologic flux of forest transpiration, especially in winter. These results highlight the impact that invasive species can have on landscape-level ecosystem fluxes.