There will often be a small discontinuity between the number of births (and deaths) in the first year compared to subsequent years. Since Spectrum calculates mid-year populations (July 1 to June 30) we determine the number of women of reproductive age during this period as the average between women last July 1 and this July 1 (WRA = [females15-49(t-1) + females15-49(t)]/2
With this formula you cannot calculate births (or deaths) in the base year since you don’t know how many women there were in the previous year. The professional demographers always want us to leave the first year blank to show that you cannot calculate that value. Even though that is technically correct we don’t do that because it is not helpful to programs. So we estimate births in the first year using the base year population. Since the previous year’s population was probably somewhat smaller this may tend to over-estimate births by some small amount. These births are not used in any subsequent calculations, they are just for display. So the slight drop you see from the first to the second year is just due to this approximation.