>From: Andrew Strak <abstrak@auracom.com>
>To: phowgate@rsc.co.uk; Shuckmans@aol.com; seafood@ucdavis.edu
>Subject: RE:salmon oil's
>Date: 05 November 1998 19:29
>Do we know how much impact any particular sampling plan would have on
those
>variations? Sometimes variations are calculated on composite lots of fish
>coming from different areas, season etc. And moreover, if the sample is
>small, the standard error will be high again. I think, that using the
'>avarage oil content' in salmon context without any clear references to
the
>environment and sampling methods does not make any practical sense for
>processing purposes such as smoking. It may however, for some rough cut
>nutritional estimates.
The lipid contents in the references I cited were measured on individual
fish. The SD's then are the fish-fish variation within the sample. The
samples were large, hundreds of fish, so the precision of the estimates of
means and variances would be very high. The point is that there are
systematic differences in mean lipid contents between sources - between
farms, wild v. farmed, between feed compositions - but these are small
compared with the variance within sources. It appears from these published
data that a salmon processor will have to expect a range in lipid content
of getting on for 10 percentage units within a batch from a particular
source, say between 6% and 16%. I can assure you that it does have a
practical effect on the smoked salmon product. For good quality products
the raw material should have a lipid content below about 10%. Above this
the product shows oil on the surface and it has an oily mouthfeel. Some
multiples - supermarkets - specify a maximum limit for oil content in the
raw material used for manufacture of smoked salmon for their stores, but
the spec. is not all that easy to meet because of the high variance. There
is a better chance of getting low lipid contents if the raw material comes
from farms that do not use high energy feeds.
The original question asked if there was a difference in lipid content
between farmed and wild salmon. The answer is Yes, there probably is, with
the content in wild being a little lower than in farmed, though I did not
come across enough data on lipid content of wild to make a good estimate of
the difference, but the difference is small compared with the within-source
variance.
Peter Howgate