First year survival estimate

[Eurystomus]

I'm using a PIM with three age classes for survival (first age class = year 1, second age class = years 2-5, third age class = rest of years) and two classes for recapture (first class = years 1-3, second = rest). My first year of data is 1983. My observed recaptures are (for the pertinent part of my data):

R(i); m(i,j); r(i)
j=2, 3, 4, 5, 6, 7, 8, 9, j=10
3977; 2, 5, 1, 19, 2, 3, 1, 2, 4; 39
3556; 119, 63, 43, 31, 24, 21, 9, 20; 330
4133; 34, 16, 19, 15, 11, 6, 7; 108
3574; 19, 8, 8, 1, 4, 6; 46

My problem is that for 1983 (the first entry, R(i)=3977) the survival and recapture rates calculated with the PIM I'm using turn out to be 0.999998 and 0.0005 respectively. This seems to be inaccurate, and so when I dropped the first year completely (not the first encounter history for each bird, just all the 1983 data) the next year, 1984, has survival and recapture rates of 0.1805693 and 0.1853282. This is much more accurate. Another year similar to 1983 in observed recaptures is 1986 (R(i)=3574) but I get reasonable parameter estimates for that year. It's only the first year that is giving me trouble. Can anyone point me towards why the first year (1983) seems to be providing incorrect parameter estimates?

Thanks,
Diana