97th ESA Annual Meeting (August 5 -- 10, 2012)

PS 71-67 - Characterization and comparison of forest and steppe ecotones on north- and south-facing slopes in northern Mongolia

Thursday, August 9, 2012
Exhibit Hall, Oregon Convention Center
Julia A. Perzley, Ecology & Evolution, Rutgers University, New Brunswick, NJ, Daniel Brickley, Department of Plant Biology, Michigan State University, East Lansing, MI and Peter S. Petraitis, Department of Biology, University of Pennsylvania, Philadelphia, PA
Background/Question/Methods

To characterize the ecotones between taiga forest and steppe grassland in the Dalbay Valley along the east coast of Lake Hövsgöl in northern Mongolia, we collected data along 24 transects that ran roughly north-south near the peak of the south-facing slope and 25 transects on the north-facing slope. Throughout this region, steppe grassland occurs on the south-facing slope while taiga forest consisting almost exclusively of Siberian larch (Larix sibirica) occurs at the peak of the slope and continues down the north-facing slope into the next valley before switching back to steppe or riparian vegetation. This pattern reoccurs in the other valleys in this region. A better understanding of the switch from steppe to taiga is also currently relevant because this region is undergoing environmental changes due to global climate change that may affect the position and character of these ecotones. Each transect began in the grassland and ended in the forest. Breast height diameter, distance to nearest neighbor, and breast height diameter of nearest neighbor of trees within a meter (north-facing slope) or a half-meter (south-facing slope) of the transect line were all measured. Comparative analyses were of density measurements based on distance to nearest neighbor. Data from the two slopes were analyzed using STRUCCHANGE (Zeileis et al. 2002) to test for the “breakpoint” in a linear regression model along each transect for each metric.

Results/Conclusions

Distance to nearest neighbor was the most consistent measure for a structural breakpoint on both slopes. On the south-facing slope, the average breakpoint was at tree number 3.9 from the beginning of the transect in the 15 transects that returned a result. On the north-facing slope, the average breakpoint was at tree number 8.4 of the 21 transects that gave a result. These results can categorize parts of the slopes into ecotone and forest. On both slopes there is a difference in tree density between the ecotone and the forest. However there are structural differences between the two slopes. The ecotone is wider and more variable on the more forested north-facing slope.

References

Achim Zeileis, Friedrich Leisch, Kurt Hornik and Christian Kleiber (2002). strucchange: An R Package for Testing for Structural Change in Linear Regression Models. Journal of Statistical Software, 7(2), 1-38. URLhttp://www.jstatsoft.org/v07/i02/