A program to trap and remove shiny cowbirds (Molothrus bonariensis) was conducted during two successive passerine nesting seasons at Roosevelt Roads Naval Station in eastern Puerto Rico. It sought to improve existing trapping techniques and to determine the effect cowbird removal has on the reproductive success of the endangered yellow-shouldered blackbird (Agelaius xanthomus). Decoy traps of two basic designs were used to capture 2449 cowbirds in 1162 trap-days (average 2.1/trap-day) in June-September 1985 and 850 cowbirds in 1571 trap-days (average 0.5/trap-day) in March-August 1986. The lower capture rate in 1986 suggests that cowbirds removed in 1985 were not being replaced during the non-trapping period. Trapping data from yellow-shouldered nesting areas in mangrove swamps indicated that cowbird capture rates were significantly higher (P=0.02) for large (14.2-14.8 m3) traps than for smaller (4.2 m3) ones. The effect of cowbird removal on the nesting success of the yellow-shouldered blackbird could not be directly determined because only one nest could be found. Cowbird removal, however, greatly reduced parasitism rates of another parasitized species, the yellow warbler (Dendroica petechia).

A 2.2-ha cedar and mixed deciduous tree woodlot in Crawford, Mississippi, harboring 330,000 blackbirds (Icterinae) and European starlings (Sturnus vulgaris) was aerially sprayed by helicopter the night of 1 March 1989 with 1050 I. of a CPT-avicide formulation at a rate of 49 kg CPT/ha. Most mortality occurred within 36 h of treatment. Mortality in the roost was 3% of the pretreatment population. No reliable technique to estimate out-of-roost mortality was identified. Pilot misapplication and probable CPT volatilization of the spray formulation contributed to the low mortality. Only 1 of 37 radiotagged blackbirds using the roost the night of the spray was verified as a CPT-related mortality. Pre- and post-treatment numbers of hawks and owls in and near the roost did not differ (P > 0.05). Dead nontarget animals found in and near the roost included 10 cardinals (Richmondena cardinalis), 1 robin (Turdus migratorius), and 1 least shrew (Cryptotis parva). Additional research is needed to develop methodologies for evaluating the efficacy of slow-acting toxicants.