Question 1182685
It is a relatively large sample size, and if one believes in the rule of 30 regardless, a z-test would be used. However, I prefer using a t-test since the sd used is from the sample. The difference is <1% with a sample size this large.  While the population weight is skewed somewhat, it is not likely to such a degree that a random sample of 187 could not be construed to be normally distributed. 
t(alpha/2)=1.97
z(alpha/2)=1.96
If you've not studied the rule of 30 for sample size, use t.
If you were taught the rule of 30 means you can use z, use it.
But notice the small difference between the two with this sample size.