I do not know how the yields will be for your specified tissues, but I've used QIAGEN's RNeasy kit to extract RNA from rat hypothalamus. The yields for ~100mg of the rat brain matter was ~500ng/ul or 15ug total cellular RNA. You can obtain similar or better yields with Phenol-chloroform extraction followed by isopropanol precipitation.

after nanodrop your samples, dilute them to equal amount and run a gel to see if there are clearly three bonds. 

if you need more accurate data, run the isolate on agilent bioanalyzer.

a realtime PCR for 16s gene after transcription can also tell you how much the RNA was there, by using known concentration gene for standard curve.

QIAGEN protocol wiill give you all details about extraction and control RNA quality 

I've used the RNeasy Mini Kit and Mini Kit plus (QIAGEN) and these kits work fine for the fish tissues you mentioned. Also, a cheaper useful protocol is using TriReagent (Sigma).

After that, check the concentration with Nanodrop and quality with an agarose gel stained with ethidium bromide.

Good luck!

The Qiagen RNeasy plus kit works well with fish tissues but it's expensive, so you could consider using other commercial kits. Tri Reagent has the advantage of giving you increased RNA yields, but you'll have to clean up the RNA using columns from the RNeasy kit (or other commercial kit). If using the RNeasy kit, don't forget to use 50% ethanol for liver tissue instead of 70% ethanol (as indicated in the protocol) to add to the cleared lysate.

As for DNasing your RNA, I would recommend using the TURBO DNA-free kit from Life Technologies. Ideally you would extract your RNA, DNase, and then clean up with a column (from a kit) before quantifying on a Nanodrop.

And is there a good working reference gene or it's primers set for detecting fish species (to cover most possible number of abundant fish species) in food samples using routine PCR?

As it happens for mammalian tissues, efficiency and quality of RNA extraction may be tissue-specific. General-purpose RNA extraction protocols should be suitable; for instance, Trizol classic extraction protocol always works well (with or without implementation with spin-columns).

If fish are small-sized (for example, zebrafish, or juvenile stages, embryos and larvae), RNA extraction can be limited by availability of poor tissue amount... in this case, also micro-kits are common and available.

For downstream RT-PCR analysis, you should keep in mind that more than a few teleost (and non) fish species, as low vertebrates, undergo a lot of gene duplication events. Don't forget it when checking for primer pairs :))

Thank you everybody because of your friendly comments! will update my result :) 

Please find the link.

http://www.digital-biology.co.jp/manufactured/products/ms-100/pdf/report_ms-100_06.pdf

Munir

Thank your kind Mohammad :)