Hi,
This is easy to explain. The two variances represent different pool.

Say a trait has a variance of 100%. In the first step you regress the variance explained by fixed factors e.g. age and sex. Some of this variance is genetic and some of it is environmental but together it explains 61% of the trait variance. The residual 39% of the variance goes into the additive genetic modeling. 
Out of that variance 52% is being explained by the additive genetic factors. Therefore, the total variance explained by the additive genetics is .39*.52 or about 20% and remaining 19% are enviromental. 

So together this will add:

61% fixed factor variance + 20% additive genetic + 19% enviromental factors =100