I posted a reply to an earlier question from you on this subject. See http://r.789695.n4.nabble.com/Errors-in-SIAR-td4029804.html. In it I note that there are problems with the function from siar (not SIAR) you are using, but that this may not be "your" problem, that the function calls for matrices (you were using data frames), and that you are unlikely to get further help on this until you post your data (or data that resemble yours).
It's not that people don't want to help you, but you have to give them something to work with (see the "famous" footer of this message). One of the demos in the siar package mostly works, the other one does not. It's possible that there is a minor "glitch" somewhere, which could easily be fixed, so that given data in the correct format you get a result.
Why don't you dput() a subset of your data, so that anyone who is interested in helping you can have a go? If your data set is called myData, and is stored as a data frame, then do something like the following and copy the result of dput() into your next email. Of course, if your data set has many rows then you want to adjust the "by" argument (increase it). Twenty to thirty rows should be sufficient.