Three long-term (26 years) data sets were assembled from the literature and from unpublished records to test the following hypotheses concerning mallard recruitment rates: (1) they vary inversely with the size of breeding and wintering populations, and (2) they vary directly with wetland conditions on the breeding and wintering grounds. The ratio of numbers of immature:adult mallards harvested in the Mississippi Flyway between 1959 and 1984 and the ratio of numbers of broods: breeding pairs of mallards from ground surveys on the Caron Study Area and the Redvers Study Area in southern Saskatchewan between 1950 and 1977 were used to index recruitment rates for mallards breeding in the Prairie-Parkland Area of Canada. Age ratios of Mississippi Flyway mallards for 1959-84 varied inversely with the size of breeding (P = 0.001) and wintering (P 50% of MF mallards winter in the MAV. Crissey (1969) demonstrated a positive relati nship between annual production of mallards on a continental basis and number of July ponds in the PPA. The U.S. Fish and Wildlife Service predicted annual production (i.e, total number of young present in the Aug population) of mallards using spring breeding population size, number of July ponds in the PPA, and an index of brood production in the PPA (Martin et al. 1979). These 3 variables are now used by the U.S. Fish and Wildlife Service to ' Deceased, 17 July 1986. This content downloaded from 157.55.39.215 on Tue, 30 Aug 2016 04:28:16 UTC All use subject to http://about.jstor.org/terms 142 MALLARD RECRUITMENT * Kaminski and Gluesing J. Wildl. Manage. 51(1):1987 predict mallard age ratios (young: adult) in August (J. D. Nichols, pers. commun.). Additionally, Heitmeyer and Fredrickson (1981) demonstrated a positive correlation between mallard age ratios in the MF harvest and precipitation during the preceding winter in the MAV, which implied to the authors that mallard recruitment rates may be influenced by wetland conditions on the wintering grounds. To acquire further insight into factors possibly influencing mallard recruitment rates, we assembled 3 long-term data sets from the literature and from unpublished records to test the following hypotheses concerning mallard recruitment rates: (1) they vary inversely with the size of breeding and wintering populations, and (2) they vary directly with wetland conditions in the PPA during the breeding season and the MAV during the previous winter. We thank R. J. Blohm, S. M. Carney, E. M. Martin, K. J. Reinecke, and D. L. Trauger of the U.S. Fish and Wildl. Serv., and Ducks Unlimited Canada (DUC) for providing data used in this study, and D. H. Arner, R. J. Blohm, M. E. Heitmeyer, E. P. Hill, G. A. Hurst, W. G. Leitch, J. W. Nelson, J. D. Nichols, T. D. Nudds, K. J. Reinecke, and M. E. Wise for their suggestions on improving the manuscript. J. A. Dubovsky, L. E. Miranda, and D. E. Steffen assisted with data analyses, and M. E. Wise typed the manuscript (Miss. Agric. For. Exp. Stn. Publ. 6,415).