Most restoration projects involving invasive plant eradication tend to focus on plant removal with little consideration given to how these invasives change soil microbial communities. However, soil microorganisms can determine invasibility of habitats and, in turn, be altered by invasives once established, potentially inhibiting native plant establishment. We studied soil microbial communities in coastal dunes with varying invasion intensity and different restoration approaches (herbicide, mechanical excavation) at Point Reyes National Seashore. Overall, we found evidence of a strong link between bacterial and fungal soil communities and the presence of invasives and restoration approach. Heavily invaded sites were characterized by a lower abundance of putatively identified nitrifiers, fermentative bacteria, fungal parasites, and fungal dung saprotrophs and a higher abundance of cellulolytic bacteria and a class of arbuscular mycorrhizal fungi (Archaeosporomycetes). Changes in soil microbiota did not fully dissipate following removal of invasives using herbicide, with exception of reductions in cellulolytic bacteria and Archaeosporomycetes abundance. Mechanical restoration effectively removed both invasives and soil legacy effects by inverting or "flipping" rhizome-contaminated surface soils with soils from below and may have inadvertently induced other adverse effects on soils that impeded reestablishment of native dune plants. Land managers should consider additional measures to counteract lingering legacy effects and/or focus restoration efforts in areas where legacy effects are less pronounced.

We asked whether the reproductive capacity of the endangered Cordylanthus maritimus ssp. maritimus (salt marsh bird's-beak) at Sweetwater Marsh (San Diego Bay, California) was sufficient to provide a self-sustaining population, as required by a mitigation agreement under the Endangered Species Act. After seeds were sown for three consecutive years (1990)(1991)(1992) to an upper intertidal marsh where the historical population was last seen in 1987, the number of plants censused rose from ഠ5000 in 1992 to ഠ14 000 in 1995.The transplanted population initiated fewer seed capsules per flower (mean 0.2-0.4 capsules in 1992-1994) than the donor population (up to 0.9 capsules in 1991). Pollen supplementation (hand-pollination) treatments increased the number of capsules by 89% in 1992, and by 52% in 1993. Abundance of pollinators appeared to be less important than the relative abundance of particular genera (i.e., Anthidium and Bombus). However, in areas where pollen supply was sufficient, nitrogen availability seemed to limit capsule set.Plant growth rates of the transplanted population compared favorably with those of the donor population (mean height 12-13 cm). At Sweetwater Marsh, larger plants produced more flowers, and aboveground dry mass of plants correlated strongly with inorganic nitrogen in the soil, which was Ͼ71% sand. Nitrogen fertilization treatments increased plant size, foliar nitrogen content, and mean seed mass.Even though the population produced Ͼ1.5 ϫ 10 6 seeds from 1992 to 1994, it is not clear that C. m. maritimus will be able to sustain itself at Sweetwater Marsh. The number of plants in 1994 was Ͻ3% of the estimated seed production in 1992-1994. Opportunities for seed germination and seedling establishment may be more limiting than reproductive capacity. The most difficult challenge to habitat managers may be maintaining ecological relationships that are dependent on external factors, i.e., pollinators, nutrient inflows, and canopy disturbances (from debris or mammals) that create the necessary openings for seedling recruitment.

This study evaluated the genetic consequences of a reintroduction of the endangered annual plant Cordylanthus maritimus ssp. maritimus to Sweetwater Marsh (San Diego County, California). A survey of 21 enzyme loci in natural populations revealed that genetic diversity is very low and is primarily found as rare alleles at a few loci, making this species especially susceptible to the loss of alleles and heterozygosity through genetic drift. The reintroduction was performed in 1991 and 1992 by sowing seeds (collected from Tijuana Estuary) in numerous small patches of suitable habitat. For this study, leaf tissue was collected from all plants in all patches during flowering in 1995 and surveyed for genotype at the three enzyme loci that are polymorphic at Tijuana Estuary. Rare alleles were absent in 27 out of 30 patches for Pgm‐1, in 17 out of 30 patches for Pgm‐2, and in 10 out of 11 patches for Mdh‐1. In all, half of the patches lacked any rare allele. Rare alleles tended to occur in patches with few individuals. Overall rare allele frequency was lower than in the colonies from which seeds were collected at two of the three loci, and heterozygosity was reduced. The Sweetwater Marsh population is at risk of losing most of its genetic variation at enzyme loci through the extinction of patches with few individuals. Future reintroduction attempts should attempt to create contiguous sets of patches or to periodically reseed existing patches to reduce the loss of genetic variation.

No abstract

Invasive plants create refuge habitat for predators that results in increased levels of seed predation on native plants. We test the effects of a large-scale removal of the invasive plant Ammophila arenaria on the predispersal seed predation rate experienced by an endangered plant in a coastal dune ecosystem. Restoration resulted in a large and lasting reduction in seed predation in this system, and reducing predation could be an explicit goal in future restoration projects. The generalized linear mixed statistical model with a BACI (before-after-control-impact) design used here is a useful, flexible model that can be applied to analyze other large-scale restoration activities. Implications for Practice• This is the first documentation that large-scale removal of invasive plants through habitat restoration reduces the threat of seed predation via refuge-mediated apparent competition for native plants. • Reducing refuge-mediated predation pressure should be a goal incorporated into large-scale invasive plant removal programs.• Generalized linear mixed model analysis is a useful tool for analyzing before-after-control-impact designs in restoration projects.