Page:Condor17(5).djvu/12

182 THE CONDOR Vol. XVII older birds. To get their approximate average length of life, it is then only neces- sary to make a sufficiently large count of young and old birds, so that the propor- tion may be approximately correct for the whole Herring Gull population. Dur- ing the winter of 1913-14 this was attempted. Second-winter Herring Gulls are in an intermediate plumage easy to identify, and, as the length of life before count of the group one is studying has nothing to do with the working of the formula, during the season 1914-15 an attempt was made to determine the propor- tion of second-winter birds to those of that age or over, from which the length of life of second-year birds could be computed. The Herring Gulls observed in New York Harbor are a mixed lot represent- ing all ages and plumages. In the fall the first-winter birds are uniformly dark, the second-winter birds much paler, with white rumps, and all others have grey mantles and white tails. The tendency of birds of an age to flock together, some- times marked on adjacent ocean shores, is not noticeable here. In the season of 1913-14, 58 counts of the proportion of first-year birds to the total were made. Out of a total of 7794, 1294 were in their first year. Divid- ing, we get 6.0 for the proportion of total to young. That is, first year birds will live to an average age of 6.0 years. In the season of 1914-15, 46 counts of the proportion of second-year Herring Gulls to the total (excluding first-winter birds) were made. Out of a total of 3267, 215 were in their second year. Dividing we get 15.2 for the proportion of total to young. That is, second-year birds will live on the average 15.2 years, or to an age of 16.2 years. It is at once apparent that the mortality between first and second winters is very high, amounting to about 66 percent. It will be noted that the same year's brood was dealt with as second-winter in 1914-15 which was dealt with as first winter in 1913-14. The personnel of the flocks. is constantly changing, as often on one day one sees twice as many as on the day previous, and a bird with peculiar plumage, perhaps some rare species, has often been looked for, but never seen a second, time. In estimating proportions the writer has ,made actual count each day of birds of different plumages, so far as possible (it is not entirely possible) avoid- ing repetition in a count, and added together the figures obtained on all the days when count was made. Doubtless many individuals were counted on more than one day, and the figures obtained must be too high for the actual number of indi- viduals observed, but this should not affect the proportions. The plumages vary somewhat, but at this season there is seldom hesitation in placing any individual in one of the three categories (first-winter, second-winter, adult). Through the winter young birds become gradually lighter and after about March I there is more change. There seems to be much variation in the persistence of the dark plumage in spring, and the percentage of puzzling plumages increases, so that during the second season, 1914-15, no attempt at count after March I was made. Chance determinations of plumages were omitted, as of a few birds flashing white in a distant flock which might contain undeterminable young birds,- and thus one probability of error in proportions has been avoided. CI-IANCES OF ERROR The formula followed is true only of fixed populations; if the I?erring Gull population is increasing notably the number of young birds will be larger than normal and the determined average age too low. The same is true if the 1913 brood of young was for any reason an abnormally large one.