I want to determine if there is a difference in the amount of total antibodies in the serum of different bovines, I was thinking of doing it with Elisa sandwich. Has anyone tried it? What initial dilution should I try?
Regarding your question, I would also propose to follow the providers' instructions included in or offered for the kit. This strongly depends on your experimental settings (sample size, quantity, species, the experimentator etc.).
Are you after IgG only or do you want to check for the others, too?
If you know the sensitivity range of your test and have some expectations about how much antibodies are present in your samples, it's easy to calculate. Any hints in the instructions of your test kit?
Otherwise, you'll have to run a series of serum dilutions (1:10,20,50,100,200,500,1k, possibly more) in order to get into a range where your test has optimal accuracy and can cover the whole concentration range of interest - or what is present in your samples.