Radial Dispersion of Neighbors and the Small-Scale Competitive Impact of Two Annual Grasses on a Native Perennial Grass Jeffrey S. Fehmi, 1,3 Kevin J. Rice, 2 and Emilio A. Laca 2 Abstract In California’s Mediterranean type grasslands, native perennial grasses such as Nassella pulchra are surrounded by introduced annual species and these annuals are thought to have displaced natives through much of their range. Amongst other invaders, two grasses Lolium multiflorum and Bromus hordeaceus, commonly dominate portions of the grassland with potential for N. pulchra restoration. We hypothesized that competitor species differences and small- scale gaps (150 cm 2 ) could be important determinants of N. pulchra survival and performance on these sites. Lolium multiflorum and B. hordeaceus were planted in 20 cm diameter circular plots at a constant rate of 1 seed per cm 2 surrounding newly transplanted N. pulchra plants. Nassella pulchra showed no significant effect of the species of competitor or from the distribution of the competitors. Both interspersion of patches of bare ground and separation of competitors into patches did not increase N. pulchra pre- dawn water potential, basal area change, number of seeds produced, or average weight of seeds. The presence of L. multiflorum was associated with a decrease in N. pulchra survival compared with plots with only B. hordeaceus. Plants with increases in basal area of less than 0.75cm 2 during the growing season had 74% mortality compared with no mortality in plants with more growth. However, initial N. pulchra plant size was not a good predictor of mortality. Limiting competition from annuals may increase survival of N. pulchra plantings, but 60% of the plants survived for at least 1 year, despite being transplanted into soil containing substantial annual grass seed. Key words: additive design, asymmetric competition, neighbor effect, plant interference, radial dispersion, spatial pattern, spatially mediated interaction. Introduction Where they remain, individual native perennial grasses in California’s Mediterranean type grasslands are surrounded by introduced annual species that have spread widely since the mid-1800s (Bartolome et al. 1986; Heady et al. 1991). Nassella pulchra (A. Hitchc.) Barkworth (purple needle- grass), the most widespread native bunchgrass, probably declined from dominance due to heavy livestock use (Heady et al. 1991) and competition with exotic species (Bartolome & Gemmill 1981; Dyer & Rice 1999; Brown & Rice 2000). Yet, removal or reduction of livestock use has not resulted in widespread conversion to native-dominated communities (Heady et al. 1991). The scarce distribution of native perennials combined with the recent rise in desir- ability of natives generally has made areas with substantial cover of grasses, like N. pulchra, the target of conservation or the goal of restoration efforts. Although a good competitor after establishment, N. pulchra seedlings generally do not survive well in the normal conditions found on California grasslands. In one study (Dyer et al. 1996), only 0.01% of N. pulchra seedlings survived through the end of three growing seasons. One alternative is the establishment of plants through preger- minated plants (‘‘plugs’’) grown in a greenhouse for several months before planting. Nassella pulchra established through plugs should have an increased chance for survival because of the potentially unimpeded access to deep soil water in areas dominated by shallow-rooted annuals (Holmes & Rice 1996; Dyer & Rice 1999), although some of the invasive forbs such as Centaurea solstitialis (yellow starthistle) can also tap deep soil moisture (Roche et al. 1994). The limited research on plug transplants has shown them to survive well in cleared or weeded sites, but mortality was high in untreated annual grasslands (Dyer & Rice 1997). It is unknown how plug transplants interact with the normal gaps found in California’s grasslands. Small gaps and patchy species distributions are common in many grasslands and rangelands. Heady (1958) reported the occurrence of numerous 6.5 cm 2 (1 in 2 ) bare quadrats in California grasslands. Bartolome et al. (1980) reported foliar covers of 2070% on grasslands distributed across California, and Pitt and Heady (1978) reported similar covers. Our recent work (Fehmi et al. 2001) demonstrated that the resources of small grassland gaps can be used by surrounding annual plants and thus aboveground biomass 1 U.S. Army ERDC-CERL Ecological Processes Branch, 2902 Newmark Drive, Champaign, IL 61826, U.S.A. 2 Department of Agronomy and Range Science, University of California, One Shields Avenue, Davis, CA 95616-8515, U.S.A. 3 Address correspondence to J. S. Fehmi, email jeffrey.s.fehmi@erdc.usace.army.mil Ó 2004 Society for Ecological Restoration International MARCH 2004 Restoration Ecology Vol. 12 No. 1, pp. 6369 63