Repeated measures or not?

Hi All,
I sampled crickets weekly over 5 weeks from the same population, and measured their body size. Half of the crickets were taken from predator's nests (sampled within a day of being caught), and half were sampled directly from their natural habitat. My research question is: are predators taking larger crickets than the average cricket size in the population? I could simply have used a t-test, HOWEVER, over the 5 sampling periods, the crickets grew in size. And since I didn't sample the same number of crickets in each sampling period, I can't lump all samples into one t-test. I have asked other stats experts, and they either say to do 5 t-tests (one for every sampling period), and use Bonferroni correction, OR do a repeated measures ANOVA. Which is correct?
To sum up:
IV: Caught or not caught by predator
DV: body size
Samples are not matched
Single population is sampled from on 5 occasions
Overall body size increases over the 5 occasions

Any insight or links would help. Thanks!