Using phytolith assemblages to reconstruct the history of a degraded desert grassland in Big Bend National Park

Background/Question/Methods

Written historical accounts indicate the northern region of Big Bend National Park once supported extensive arid and semi-arid grasslands.  Decades of overgrazing together with periodic drought, however, has transformed this landscape into an advanced desertified state.  As part of an effort by the national park to restore native grasslands to this region, evidence of the pre-disturbance plant assemblage within the affected area was sought through phytolithic reconstruction of paleovegetation.  The study site is located at the foot of the Santiago Mountains in the northeastern corner of the park near Dog Canyon.  The plant community at this location is currently characterized by a banded vegetation system commonly associated with arid environments.  Here, dense parallel bands of grasses and shrubs are distributed within a matrix of bare soil and scattered shrubs.  Soil cores (1.5 m depth) were taken from three scattered points within areas of bare soil.  Core samples were extracted at 15 cm intervals to create a record of historical change with time.  Phytoliths were extracted from each sample using a standard ashing method, mounted to microscopes slides, then counted and classified to morphotype.  Raw counts within each sample were transformed to proportions and averaged for each depth.

Results/Conclusions

Seven different morphotypes were found.  These include six morphotypes commonly associated with grass (festucoid, saddle, rondel, pyramidal, panicoid opaque, bulliform) and one morphotype typically associated with shrubs (spherical).  Substantial variation in morphotype representation was found with increasing soil depth indicative of a plant assemblage in considerable flux over time.  A clear pattern of directional change in community composition is absent.  Notably, however, several pairs of morphotypes show a strong negative correlation in abundance with changes in soil depth.  That is, an increasing abundance of one type is coupled with decreasing abundance of a second.  Moreover, these associations appear to show an oscillating trend over the full soil profile analyzed.  This is indicative of periodic or regular changes in plant community composition over time.  Such findings could be a product of the extant plant community.  Evidence suggests that the bands in some banded vegetation systems do migrate slowly across the landscape.  Our evidence of oscillating changes in plant composition might be a historical footprint of such movement in our system.  Alternatively, they could reflect a larger-scale pattern of change in these systems.  Further research is needed to identify morphotypes with known species and to establish dates for the time scales analyzed.