Accessions A-68, L-11, L-19, L-28, and L-38 of Lehmann lovegrass (Eragrostis lehmanniana Nees); P-15608 Cochise lovegrass (E. lehmanniana Nees X E. trichophora Coss & Dur.); A-84 and Catalina boer lovegrass (E. curvula var. conferta Nees); Palar Wilman lovegrass (E. superba Peyr.) and P-15630 blue panicgrass (Panicum antidotale Retz.) were seeded at a study site near San Simon, Ariz., in spring 1970 and 1971. Seedbeds were prepared by root plowing and furrow pitting immediately before planting. Growing season precipitation was 136 mm in 1970 and 218 mm in 1971. Mean accession densities in the fall after the initial growing seasons were 18 plants/M2 for both the 1970 and the 1971 plantings. Between fall 1971 and 1972 mean accession densities declined 44% and forage production was unchanged on the 1970 plantings. Accession densities declined 22% and forage production increased 250% on the 1971 plantings. Between fall 1972 and 1982 the majority of seeded plants died and forage production declined 90% on the 1970 plantings. Accession densities declined 78% and forage production declined 84% on the 1971 plantings. Southeastern Arizona and southwestern New Mexico rangelands were overutilized and deteriorated rapidly between 1880 and 1900. Griffith (1901) documented the deterioration and corresponding livestock losses. Cooperative studies to restore these rangelands were initiated in the early 1900's by the Division of Agrostology (USDA) and State Experiment Stations at Tucson, Ariz., and Las Cruces, N. Mex. Blount (1892), Griffith (1907), Keefer (1899), and Thornber (1905) seeded native and introduced grasses on irrigated and rangeland sites and evaluated emergence and survival. Teff [Eragrostis abyssinica (Jacq.) Link.] emerged on irrigated and nonirrigated sites, but long-term survival occurred only at irrigated sites. Native grass either failed to emerge or to survive at southwestern revegetation sites between 1910 and 1934 (Barnes et al. 1958, Cassady 1938, Glendening 1937, and Hendricks 1936). Numerous grass, forbs, and shrub species were introduced after 1930 (Cox et al. 1982). These introduced species were screened for germination, drouth tolerance, and seed production potential at Soil Conservation Service Plant Materials Centers, and a few promising grasses were released for rangeland plantings. Among these were A-68 Lehmann lovegrass and A-84 boer lovegrass; both were introductions from southern Africa. Lovegrass species and newly developed accessions were sown in summer (Bridges 1941 and Herbel et al. 1973) and fall (Bridges 1941) at desert sites in southern New Mexico. A-68 Lehmann and A-84 boer lovegrasses emerged in moist summers, and A-68 emerged in wet winters at lower elevations. Jordan (1970) conducted studies for 9 years to determine the best combinations of mechanical brush control, seedbed preparation, The authors are range scientist, USDA, Agr. Res. Serv., Arid Land Ecosystems Improvement, 2000 East Allen Road, Tucson, Ariz. 85719; and professor, range management, School of Renewable Natural Resources, University of Arizona, Tucson 85719. This paper is published with approval of the Director, University of Arizona College of Agriculture, Agricultural Experiment Station, as Paper No. 3705. The paper reports on work supported by the U.S. Department of the Interior, Bureau of Land Management, and is a cooperative investigation of Agr. Res. Serv., USDA, and the Arizona Agr. Exp. Sta., University of Arizona. Manuscript received January 20, 1983. and time of seeding for emergence and survival of forage grasses at 3 sites in southeastern Arizona. A-68 Lehmann lovegrass emergence and survival was optimized when root plowing and pitting were used to control brush and prepare the seedbed in spring, and when seed were sown immediately after a mechanical treatment. Comparative seedling trials were conducted to select adapted lovegrass and blue panicgrass accessions at a study site near San Simon, Ariz., in 1970 and 1971. The purpose of this paper is to quantitatively document and compare initial and long-term plant densities and forage production for these seeding trials. Study Site and Methods The study site is located 25 km southwest of San Simon, Ariz., near the Arizona-New Mexico State Line in southeastern Arizona. Average annual precipitation is 280 mm, and 30 to 40% occurs in winter. Winter months (November to March) are cold, dry, and windy, and typical of the Chihuahuan Desert (Mabry et al. 1977). Winter precipitation is either evaporated or transpired by shrubs, and apparently not used by seeded grasses (Jordan 1970). Effective summer precipitation falls in late July through October and varied from 98 to 230 mm at San Simon between 1972 and 1982 (National Oceanic and Atmospheric Administration, Annual Climatological Data Summaries 1972-1982). Mean summer precipitation was 165 mm over the 10 years. Figure 1 shows the departure of annual summer precipitation from the 10-year average. Average annual air temperature is 17?C and the frost-free period is 220 days. Soils are deep, well drained, and formed in old alluvium from mixed sources. Soils are classified as Eba gravelly sandy loam, mixed, thermic Typic Haplargids (Vogt 1980). Native perennial grass forage production was 2.5 to 5.0 g/ m2 on