Challenges to introducing and managing disturbance regimes for Holocarpha macradenia, an endangered annual grassland forb

Summary

Introducing rare plants to new sites to requires knowledge of habitat requirements and plant demography. In this study an experiment was conducted at three coastal prairies in southwest USA to test the effects of clipping frequency and litter accumulation on seed germination, seedling survival, reproduction, and seedling recruitment of introduced populations of the endangered Santa Cruz tarplant Holocarpha macradenia. The species is listed as endangered by the State of California and as threatened under the U.S. Endangered Species Act.

Study sites: The research was undertaken at three species-rich coastal prairie grasslands (Elkhorn, Swanton and the University of California, Santa Cruz campus (UCSC)) within the historical range of H.macradenia near the town of Santa Cruz, California, southwest USA. All sites had been cattle-grazed prior to the study, contained coastal prairie plants normally associated with H.macradenia, had <10% slopes facing south to southeast, and had sandy loam soils. The sites were chosen as they appeared to be suitable for, but lacked, H.macradenia. Site selection also depended on permission to establish experimental populations of an endangered species.

During the 6 study years, annual rainfall ranged from 29 to 84 cm. Seasonal rainfall was close to or below average in all years, and well below it in 2000–2001.

At the beginning of the study, non-native annual grasses Bromus spp., Italian rye-grass Lolium multiflorum and squirrel-tail fescue Vulpia bromoides dominated all sites. Exotic forb cover was 47% at Swanton, 29% at UCSC and 15 % at Elkhorn. Native grasses (Danthonia californica and Nassella pulchra) and forbs (Eschscholzia californica and Madia sativa) constituted 2–11% cover.

Study species: H.macradenia (Asteraceae: Madiinae) is an annual that flowers from July to October. The basal rosette produce a stem which reaches a height of 0.3–0.8 m. Plants produce flower heads that have two types of achenes (seeds) with different morphologies and germination requirements. Seeds from ray flowers have a thicker seed coat and long-term dormancy. Seeds from disk flowers are lighter and do not appear to maintain viability in the soil after 1 year. Most seeds fall within 45 cm of the plant.

Experimental design: At each site a 52 × 52 m cattle exclosure was installed in autumn 1998, experiments commenced in January 1999. Within each exclosure 30, 7 × 7 m plots (a 1-m mown buffer separating the plots) with three replicates of 10 treatments were allocated.

A motorized rotary trimmer was used to clip vegetation to approx. 5 cm at three frequencies: (1) 2 clippings/year, spring (March) and autumn (September); (2) 3 clippings/year, every other month through the growing season (January, March, May); and (3) 6 clippings/year, monthly through the growing season (January–June). This provided a gradient of disturbance frequencies and mimicked common management regimes.

Each was subjected to secondary treatments to test disturbances related to cattle grazing: reduced live vegetation cover, litter removal, soil disturbance, and also a "no disturbance" treatment (neither clipped nor grazed).

Three additional 7 × 7 m plots were set up adjacent to the exclosure that were grazed by cattle. The seedlings in grazed plots at Swanton were heavily affected by feral pigs; therefore data from these are not presented. At Elkhorn cattle grazed the plots at a stocking rate of 6 animals/ha for approx. 4 days at 45- to 60-day intervals December–June each year. At UCSC, 3 cattle/ha grazed the site during March and May.

Germination, seedling survival, reproduction and recruitment: Only disk seeds (locally collected) were used as ray seeds do not germinate until after at least 1 year in the seed bank. In early October 1999 and 2000, just before onset of winter rains, 36 seeds were placed 10 cm apart in a 6 × 6 seed grid on the soil surface in each plot; 36 seeds were used in the litter accumulation, litter removal, and no disturbance treatments at Elkhorn to test whether litter accumulation affects germination.

In 2003, 36 seeds were placed in petri dishes on moistened filter paper in the laboratory and germination was recorded over a 2-week period; five dishes were in ambient light and five were covered to provide dark conditions.

In January 1999 and 2000 (all sites) and January 2001 (Elkhorn only), 25 seedlings were grown in a greenhouse from seed sown in late October the prior year. Seedlings had approximately four secondary leaves, a similar stage to the natural populations, when planted out.

Each year the number of small, medium and large flowers on each plant was recorded. For seedlings planted in January 1999 the production of disk and ray seeds produced per plot was estimated.

In January to May 2000, 2001 and 2002 the number of seedlings was monitored and seedlings recruiting and the number surviving to reproduction was calculated. In 2003 and 2004 the number of plants that reproduced at the end of the growing season was monitored.

Seed bank: In autumn 2003, 10, 2.5 (diameter) × 2 cm (deep) soil cores were taken in each of six plots at Elkhorn (with the highest seed set in previous years) to determine whether seeds remained in the seed bank after 2 years of minimal seed inputs.

Germination: Germination was much higher in 2000 than in 1999 (29 vs. 12%), although precipitation patterns were fairly similar in both years. Germination was highest at Elkhorn and lowest at Swanton in both years (1999: Elkhorn 16%, Swanton 6%, UCSC 13%; 2000: Elkhorn 46%, Swanton 9%, UCSC 34%).

In 2003 germination (Elkhorn only) was 35% in the field compared with 63% in light conditions and 64% in dark conditions in the laboratory. Clipping and grazing enhanced seed germination in all clipping years (1999, 2000, 2003) but litter removal had no effect.

Seedling survival: Survival of planted seedlings was much higher in 1999 (51%) than in 2000 (10%) or 2001 (11% [Elkhorn only]) and was significantly higher at Elkhorn. Litter-removal treatments did not affect seedling survival in any year.

Survival patterns differed across clipping treatments in the 3 years. In 1999 survival at Swanton was highest in the 3/year and 6/year treatments and lowest in the no-disturbance treatment. Survival at Elkhorn showed a similar trend, whereas survival was equal in all treatments at UCSC. In 2000 survival was lowest in 2/year clipped plots at Elkhorn and Swanton. Only one plant survived to reproduction at UCSC in 2000. Survival at Elkhorn in 2001 was low in all treatments.

Flower production: In general, total flower production was higher and proportion of small flowers was lower at Elkhorn and in clipped plots at all sites. In 2000 and 2001, survival was too low in many plots to make any meaningful comparisons.

Seed production: Seed production per plot was much higher at Elkhorn than at UCSC or Swanton. Seed production was generally higher in more frequently clipped plots and intermediate in grazed plots.

Seed bank: Of the six seed bank samples, four contained 1,019–1,218 ray seeds/m², one 406 ray seeds/m², and one no seeds.

Recruitment and survival of recruits: The number of seedlings recruiting from 1999 plants was highest at Elkhorn in 2000–2002 (where seed production had been greatest). Here there were no treatment effects on recruitment in 2000, but in 2001 and 2002 it was higher in more frequently clipped plots compared with no-disturbance and grazed plots. At Swanton and UCSC the number of seedlings recruiting was low in all years. At Swanton recruitment was higher in more frequently clipped plots in 2000 only. At UCSC recruitment was higher in grazed plots in 2000, primarily due to a large number of recruits in one plot, with no significant treatment effects thereafter. The number of recruits per plot was correlated with the number of seeds set.

Few recruiting seedlings survived to reproduction at Swanton or UCSC (<10 at each site in 2000 and 2001, 0 in 2002). At Elkhorn 2.2 (2002) to 3.8 (2001) seedlings per plot survived with no significant treatment effects. Survival of recruits from 1999 seedlings at Elkhorn (2000: 9%; 2001: 12%; 2002, 11%) was much lower than for out-planted plants. In 2003 only three plants survived, one at Swanton and two at Elkhorn. In 2004 none survived to reproduction.

Conclusions: Clipping favoured H.macradenia, mostly by enhancing germination and flower production, but surprisingly litter removal did not enhance germination success. Large numbers seedling recruits were recorded at only one of the sites. Although recruitment was higher in clipped plots for 2–3 years, by 4–5 years post-introduction very few seedlings survived to reproduce in any treatment. The lack of successful establishment was attributed to a combination of poor habitat quality, small population size and lack of a seed bank. In terms of other plants, clipping favoured exotic annual forbs and reduced exotic annual grass cover, but had no effect on native perennial grasses.


Note: If using or referring to this published study, please read and quote the original paper, this can be viewed at:

http://www.blackwell-synergy.com/doi/abs/10.1111/j.1523-1739.2006.00416.x