Functional connectivity as a possible indicator of desertification in degraded grasslands
Abstract
Desertification of semi-arid grasslands impacts air quality, climate, biodiversity, and soil fertility. Desertification processes such as wind erosion lead to declining soil resources and sometimes local climate change. Desertification is irreversible; however, identifying when and where ecological changes are irreversible is problematic, requiring observations of a new ecological state, favoring the continued process of wind erosion and continued depletion of soil resources. Scientists hypothesize that an indicator of irreversibility in desertification might be "connected pathways". The connected pathway hypothesis requires that vegetation structure has changed from a grass to a shrub dominated system with increasing number and size of bare soil gaps. These bare soil gaps are functionally connected through the action of wind; therefore, functional connectivity of a landscape is related to the length and size of pathways through vegetation. This study used a combination of field measurements (total horizontal flux (Qtot) and vegetation structure) and landscape modeling to examine the difference in functional connectivity between grassland locations that were either degraded or relatively intact. At our field site, the degradation process was initiated by groundwater pumping, which adversely affects groundwater dependent grasses, providing a useful link to management seeking to limit the effects or extent of desertification. To analyze the functional connectivity of these locations in Owens Valley, we used circuit theory, a novel graph-based approach, which integrates all possible pathways to determine a "resistance distance" between any two points. Circuit theory uses current and resistance to represent movement of wind and the effect of vegetation and soil roughness on wind. Circuit theory was implemented using the open source software package, Circuitscape. To estimate landscape resistance, we performed a supervised classification on 1m aerial photographs. For each landscape class (shrub, grass, and soil), we applied a standardized resistance value which decreased downwind of vegetation. We validated the resistance layer by using Qtot data collected from 13 plots with BSNE catchers (i.e., plots with higher Qtot should be associated with higher connectivity). Next, we developed a focal region grid containing source and ground regions for current at these plots; these regions represent wind directions (N, S, E, W). We ran Circuitscape on data from different locations with similar vegetation structure but differing histories of groundwater withdrawal. From the model output, we found that degraded landscapes have greater connectivity than landscapes that are not degraded. We also found that locations with fewer larger and elongated bare soil gaps had greater connectivity than locations with numerous small bare soil gaps. Our results support the idea that functional connectivity is a possible indicator of desertification and that managing for reduced connectivity might limit the effects and extent of desertification.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFM.B51A0383V
- Keywords:
-
- 0439 BIOGEOSCIENCES / Ecosystems;
- structure and dynamics;
- 1632 GLOBAL CHANGE / Land cover change;
- 1809 HYDROLOGY / Desertification