Distribution of plants in a California serpentine grassland: are rocky
hummocks spatial refuges for native species?
Wendy K. Gram
, Elizabeth T. Borer
, Kathryn L. Cottingham
, Eric W. Seabloom
Virginia L. Boucher
, Lloyd Goldwasser
, Fiorenza Micheli
, Bruce E. Kendall
Rebecca S. Burton
National Center for Ecological Analysis and Synthesis, Santa Barbara, CA 93101;
Current address: Sam
Noble Oklahoma Museum of Natural History, University of Oklahoma, Norman, OK 73072;
Ecology, Evolution, and Marine Biology, University of California, Santa Barbara, Santa Barbara, CA 93106;
Current address: Department of Biological Sciences, Dartmouth College, Hanover, NH 03755;
address: Environmental Science and Policy, University of California, Davis, Davis, CA 95616;
Science Institute, University of California, Santa Barbara, Santa Barbara, CA 93106;
Southwest Fisheries Science Center, National Marine Fisheries Service, Tiburon, CA 94920;
Hopkins Marine Station, Stanford University, Paciﬁc Grove, CA 93950;
Current address: Donald Bren School
of Environmental Science and Management, University of California, Santa Barbara, Santa Barbara, CA
Current address: Department of Biology, Alverno College, Milwaukee, WI 53234;
correspondence (tel.: 405-325-8562; fax: 405-325-7699; e-mail: firstname.lastname@example.org)
Received 23 July 2002; accepted in revised form 25 April 2003
Key words: Exotic plant species, Hummock, Invaded communities, Spatial refuges
Invasions by non-native taxa can have severe consequences for native species. In the heavily invaded serpentine
grasslands of central California, many native species appear to be restricted to isolated outcrops of shallow ser-
pentine soil, or “hummocks,” although the extent to which these hummocks function as refuges for native veg-
etation has never been quantiﬁed. We tested whether native plant species were restricted to hummocks within a
serpentine grassland at the University of California Sedgwick Reserve near Santa Barbara, California by sam-
pling species along hummock-grassland gradients. We also examined the inﬂuence of soil parameters, hummock
area, proximity to other hummocks, and spatial location on species composition across 16 hummocks at this site.
Both the hummocks and the surrounding grassland had high Mg, low Ca, and low Ca to Mg ratios typical of
serpentine systems. Hummocks appeared to be more stressful environments because of their shallower soils,
lower cation exchange capacity, and greater percent sand. Of the 27 most common plant species sampled along
hummock-grassland transects, we identiﬁed 8 hummock specialists, 7 edge specialists, 8 matrix specialists, and
4 generalists. Importantly, both the hummock and matrix specialist groups included native species. Plant com-
munity composition was correlated with spatial positioning of the hummocks and with soil Ca, Na, K, and N.
The number of species increased and community composition changed with increasing hummock area. Species
composition was most similar among hummocks in close proximity to each other, and decreased with increasing
distance between hummocks. Our results suggest that the community structure of serpentine grasslands is spa-
tially complex and an effective management or restoration plan must address this complexity.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
Plant Ecology 172: 159–171, 2004.