in size of the nesting population or its distribution 

 among major habitat types. Migration and 

 chronology of nesting were 1 week earlier than in 

 1!)62. An index to annual fluctuations in the nest- 

 ing populal ion on the refuge was based in roosting- 

 fliglit counts at the mouths of tributaries. 



Factors affecting waterfowl foods in rice 

 fields. — Ecological factors that influence the loss 

 of waterfowl foods associated with rice culture 

 were studied during the winter of 1962-63 by the 

 Louisiana Unit. Emphasis was on the rate and 

 magnitude of seed loss, the causes of loss, and the 

 effect on viability and nutritive content of seeds 

 exposed on the ground and under water. Under- 

 water storage resulted in less loss in all species ex- 

 cept domestic rice, which showed an almost total 

 loss shortly after flooding. Red rice, signal grass, 

 jungle rice, barnyard grass, and browntop millet 

 showed little deterioration during 120 days of sub- 

 mergence; signal grass, browntop millet, and 

 smartweed showed low viability after 120 days of 

 inundation, and birds and small mammals caused 

 major losses of foods when the seeds were exposed 

 on the ground. 



For the rice grower interested in waterfowl 

 management, the following recommendations can 

 be made: (1) Delay flooding of rice stubble until 

 after the first frost to prevent rice seed deteriora- 

 tion and germination, and (2) flood fallow fields 

 from October 1 through March 31 to provide rest- 

 ing and feeding areas for waterfowl. 



Fulvous tree duck in Louisiana. — In the 1963 

 breeding season, fulvous tree ducks returned to 

 the fresh marshes of north-central Cameron 

 Parish in Louisiana in late March and early April. 

 Nesting had begun by late May, and the last 

 broods were flying by November 1. Early in No- 

 vember practically the entire tree duck population 

 in the State concentrated on Lacassine National 

 Wildlife Refuge and shortly thereafter departed 

 for the wintering grounds. Aerial censuses of 

 tree ducks on spring and fall concentration areas 

 showed a decline in the State population. 



Tree duck crop depredations were serious in 

 1963 in only a few fields, and were limited to 

 water-planted rice seed. Primary losses of breed- 

 ing water fowl are apparently largely confined 

 to deliberate actions of farmers to protect their 

 crops. 



The investigation suggested the following 

 recommendations: (1) make an annual appraisal 

 of the tree duck population; (2) inform rice 



farmers of nonlethal control methods; (3) feed 

 ducks on Lacassine National Wildlife Refuge in 

 an effort to hold them on the refuge during periods 

 when serious crop depredations occur; (4) clarify 

 and publicize the legal status of tree ducks in an 

 effort to reduce killing in summer; and (5) obtain 

 further data necessary to the proper management 

 of the species. 



Watt rfowl re/nesting mid homing. — A long-term 

 waterfowl renesting and homing study by the 

 Maine Unit was completed at the end of its eighth 

 year. Results in 1963 continued to indicate that 

 half or more of the breeding populations on the 

 study areas consist of females that had previously 

 nested here, the majority of surviving hens return- 

 ing to the same small islands in succeeding years. 

 Many important relations with respect, to cover 

 used, laying dates, and clutch sizes have been in- 

 dicated in the data gathered, and hold important 

 implications for waterfowl management in the 

 Northeast. The report is now being readied for 

 publication. 



Canada goose sex ratios. — In a project recently 

 completed by the Missouri Unit, under the direc- 

 tion of Dr. William H. Elder of the University 

 of Missouri, it was found that accurate age ratios 

 of Canada geese could not be deduced by compar- 

 ing the number of old with the number of young 

 geese trapped at baited banding stations without 

 taking into account the length of time the bait was 

 available to the birds. The proportion of adults 

 in the catches increased with the time the bait 

 was exposed. There were no significant differ- 

 ences in the ease of trapping male and female birds. 

 This means that sampling birds captured in band- 

 ing traps at random sites is not a satisfactory 

 means of determining the success of the previous 

 nesting season. 



Wood ducks prefer metal boxes. — In a 2-year 

 study of wood duck nesting in southeastern Mis- 

 souri, wood and metal boxes were dispersed and 

 available in almost equal numbers. Metal boxes 

 apparently were preferred throughout the nesting 

 period. Despite the ducks' preference for metal 

 boxes, losses to predators were slightly higher in 

 them. Metal boxes also were often hotter than 

 wooden boxes. Daytime temperatures in the 

 wooden boxes closely paralleled the ambient tem- 

 perature, but daytime temperatures in metal boxes 

 usually exceeded ambient temperatures by a few 

 degrees. 



22 



