Bird Conservation International (2005) 15:73–88. BirdLife International 2005 doi:10.1017/S0959270905000067 Printed in the United Kingdom Dynamics of extinction: population decline in the colonially nesting Tricolored Blackbird Agelaius tricolor LIZETTE F. COOK and CATHERINE A. TOFT Summary Tricolored Blackbird Agelaius tricolor is a rapidly declining species largely endemic to California and forms larger breeding colonies than any other extant North American landbird following the extinction of Passenger Pigeon Ectopistes migratorius. We present information on its distri- bution, breeding habitat and changes in global population size using data collected since the 1930s. We also present data on reproductive success at 103 colonies between 1992 and 2003. While possibly once the most abundant bird throughout much of its range, it declined by over 50% between the 1930s and early 1990s, and by a further c. 56% between 1994 and 2000. The global population is now smaller than the historic size of some individual breeding colonies. Reproductive success was significantly higher in upland non-native vegetation (primarily Hima- layan blackberry Rubus discolor) than in native emergent cattail Typha spp. and bulrush Scirpus spp. marshes, its likely predominant historic breeding habitat. Contemporary losses of import- ant upland nesting substrate, combined with low reproductive success in native habitats and complete breeding failure in harvested agricultural fields, are the most likely causes of recent declines. Recovery of this species presents possible conflicts in conservation policy because successful reproduction now largely depends on invasive non-native plants and the willingness of farmers to delay harvest or to lose portions of their crops. Introduction Colonially nesting birds are especially vulnerable to extinction. Three of the half- dozen or so modern extinctions among bird species in North America north of Mexico were colonial or highly social breeders: Passenger Pigeon Ectopistes migratorius; Carolina Parakeet Conuropsis carolinensis and Great Auk Pinguinus impennis. Because a small number of colonies may include a relatively large proportion of the population, human activities can have catastrophic effects on colonial birds either directly by the taking of adults or offspring or indirectly through habitat loss. Where breeding is socially facilitated, reduced populations may ultimately be driven to extinction through Allee effects (inverse density dependence defined as a positive relationship between population density and survival and reproduction; Allee 1931, Courchamp and Clutton-Brock 1999, Stephens and Sutherland 1999). Passenger Pigeon, once the most abundant bird in North America, may have ultimately suc- cumbed to extinction following widespread hunting and habitat loss because it could not survive at low population densities (Blockstein 2002, Bucher 1992, Stephens and Sutherland 1999, Wilcove 1999). L. F. Cook and C. A. Toft 74 Tricolored Blackbird Agelaius tricolor form the largest breeding colonies of any North America landbird, a distinction held by Passenger Pigeon prior to its extinction in 1914. The majority of the population can still breed in colonies of tens of thou- sands, but the number of such colonies is now small. Colonial breeding in this species, which includes highly synchronous nesting behaviour (Orians 1961, Collier 1968), is an adaptation that is likely to confer protection from predators through predator saturation and mutual defence (Wiklund and Andersson 1994, Picman et al. 2002). Tricolored Blackbirds also forage communally throughout the breeding season and are likewise social during all other times of the year. These characteristics suggest that both reproduction and survival could be inversely density-dependent. Largely endemic to California (Neff 1937, Orians 1961, DeHaven et al. 1975a), and once among the most abundant bird species throughout most of its range (Baird et al. 1874, Grinnell 1898, Neff 1937), the global population of Tricolored Blackbird declined by over 50% during the 40 years following its first co-ordinated surveys in the early 1930s (Neff 1937, DeHaven et al. 1975a). In the mid-1800s, one observer described how wintering flocks could “darken the sky for some distance by their masses” (Heermann 1859), a reference reminiscent of Passenger Pigeon (Wilcove 1999). Censuses in the early 1930s revealed colonies with as many as 300,000 breeding adults and a total estimated population of over 700,000 in mostly the northern portion of the species’ range (Neff 1937). These data, combined with other information on local populations in the southern portions of its range, (Collier 1968, DeHaven et al. 1975a), suggest numbers historically may well have exceeded 1 million. Multiple colonies of more than 100,000 adults were reported as recently as the 1960s (Orians 1961, Payne 1969), but a decade later the estimated population was reduced by over half of that found in the 1930s (DeHaven et al. 1975a). Tricolored Blackbird is currently classified in California as a Species of Special Concern and federally as a Migratory Bird of Management Concern, categories which identify reduced populations but do not include the legal protections afforded species listed as threatened. Much of the breeding habitat of Tricolored Blackbird today consists of vegetation that differs from that of its original habitats. Of those colonies observed during the 1930s, c. 97% of breeding occurred in the vast deepwater emergent marshes of cattail Typha spp. and bulrush Scirpus spp. throughout California’s Central Valley (Neff 1937). The preponderance of upland nesting that is found today was not reported during this time and the vast majority of upland substrates used now consist of non- native plant species that would not have been present in the Californian landscape prior to the arrival of Europeans. Nesting over water apparently affords protection from predators in many marsh-nesting birds and is a primary criterion for nest- site selection in congeners of the Tricolored Blackbird (Red-winged Blackbird A. phoeniceus and Yellow-winged Blackbird A. thilius, Yellow-headed Blackbird Xanthocephalus xanthocephalus), and other passerines (Picman et al. 1993, 2002, Picman and Isabelle 1995, Hansson et al. 2000, Massoni and Reboreda 2001). Although the historic range of Tricolored Blackbird has changed little since the 1930s, approximately half of all nesting is now in upland habitats. This apparent shift from wetland to upland is surely due to the loss of 96% of California wetlands over the last 150 years from 1,500,000 ha before European settlement (Kreissman 1991). Here we document and evaluate the population decline of Tricolored Blackbird and explore possible causes using data on this species that were collected intermittently Population decline of Tricolored Blackbird 75 Table 1. Proportions of colonies and individuals of Tricolored Blackbirds by nesting substrate in the 1930s, 1970s, 1994 and 2000. Nesting substrate 1932–1934 1968–1972 1994 2000 %%% %% % %% colonies birds colonies birds colonies birds colonies birds Emergent 94.892.769.7 a 47.425.759.654.0 marsh Himalayan 1.30.116.1 a 31.420.820.211.5 blackberry Silage 0.00.00.0 a 5.140.25.816.7 Other flooded 1.30.25.8 a 3.82.93.84.1 plants Other upland 2.67.09.0 a 12.210.410.613.6 plants Total flooded 96.193.073.0 a 51.328.663.558.1 plants Total upland 3.97.027.1 a 48.771.436.541.9 plants Total native 96.793.075.5 a 60.933.365.454.5 plants Total non-native 3.37.024.5 a 39.166.734.645.5 plants Data from 1932–1934 are from Neff (1937), Sacramento Valley and northern San Joaquin Valley. Data from 1968–1972 are from DeHaven et al. (1975a), statewide. Data from 1994 and 2000 are from Hamilton et al. (1995), Hamilton (2000) and Cook (unpubl. data). When nesting substrate vegetation was mixed, the predominant vegetation was used to categorize the nesting substrate. Percentage of colonies and birds are for all colonies located throughout the breeding season and may represent colonies and birds counted more than once (see text). aData not available. over the past 70 years. We hypothesized that if nest-site selection was predator- driven, then reproductive success should be higher in emergent marshes than in upland substrate as has been shown for other marsh-nesting birds. If true, then scarcity of available nest-sites in emergent marshes, and increased use of upland substrates, could explain the continuing decline of the Tricolored Blackbird popula- tion. Because wetland environments are among the most highly threatened world- wide, our results could have implications for management of other marsh-nesting birds. We also searched for evidence that Tricolored Blackbird could be subject to inverse density dependence, and, therefore, under threat of imminent extinction, by exploring the similarity of its circumstances to those surrounding the extinction of another colonially nesting bird, Passenger Pigeon. Study area and methods Study area Over 90% of the Tricolored Blackbird population has historically nested in California’s Central Valley, a basin 64 km wide and 644 km long running north– south, and this continues to be the case (Neff 1937, Orians 1961, DeHaven et al. 1975a, Heitmeyer et al. 1988, Beedy and Hamilton 1999). Most individuals and colonies are found in the southern portion, specifically the San Joaquin Valley, during L. F. Cook and C. A. Toft 76 the first half of the breeding season (DeHaven et al. 1975b, Hamilton 1998). Later, most birds disperse and begin appearing in its northern portion, the Sacramento Valley, for additional nesting attempts (Hamilton 1998). Large numbers of birds also bred historically in southern California (Baird et al. 1874, Grinnell 1898, Collier 1968). Today this region contains a much reduced popu- lation and one greatly smaller than that of the Central Valley. Birds have been observed nesting in other portions of California, at elevations as high as 1,200 m (DeHaven et al. 1975a), and locally in Baja California (Wilbur 1987, Howell and Webb 1995), Oregon (Neff 1933, 1937), and possibly Nevada and Washington (Beedy and Hamilton 1999).