Off the top of my head, I don't know if anyone has explicitly tested this effect, but it would not surprise me at all if fish could learn to avoid a net.  There is a rich literature on fish learning, in particular, their ability to learn and avoid 'risky' stimuli (i.e. predators, dangerous habitats, risky foraging opportunities).  I suggest taking a look at Culum Brown's edited volume 'Fish Cognition and Behaviour, 2nd edition.  Might be a good place to start.

Yes, there are observed such behavior, especially if the catch is made at the same place.

Every animal has the learning capacity for its survival, still we are getting some catch means more fishes might have avoided the fishing net. We can also expect the instinct power from the parents to escape methods.

I've published on gear efficiency, but not learned learned avoidance per se.

Hey Paul, what kind of nets are they? Hoop nets? 

Hi Dan,  small fyke nets with a single wing - very effective for capturing small bodied fish.

We estimated effectiveness of catchability of traps for chum salmon fry by tagging and recapture technique. It was found that catchability coefficient is equal to 0.14.

We also estimate the effectiveness of smolt traps for catching several anadromous species at various ages using mark/recapture experiments 2-5 times per week. The efficiency of capture can vary by species, age at smolting, lunar phase, stream flows, and stream clarity. One way to test learned behavior is to capture fish with a different trap or net, mark those fish, and look at recapture rates  when captured with the net in question vs fish that were captured with the net in question, marked, and released and subsequently recaptured. Another way might be to install video cameras (if feasible) to see if fish are actively avoiding the nets. In your  case, I would suspect that fish are learning to avoid capture.

The literature on capture-mark-recapture designs has frequent discussion of capture heterogeneity (some individuals more risk-prone or trap-happy, others more risk-averse) and I think some discussion on "trap avoidance."  These analyses typically require use of individually recognizable marks, but some analyses may be possible for small-bodied fish given a series of different batch marks.   I think some small experiments such as Michael Sparkman suggested would be the more reliable way to estimate the magnitude of learned net avoidance, though I would add that you consider conducting some recapture samples with alternate gear as well. We often use fyke nets for the marking sample and electrofishing for the recapture sample to overcome any learned trap avoidance.

I agree with Charles Anderson, that you use alternate gear to test avoidance. Good luck!

Thanks Everyone - Enjoy the new year festivities

My paper below (PDF available if you e-mail me) dealt with before vs. after sampling in the same microhabitats. We found that the second sampling had lower total-fish abundances regardless of whether seining or electrofishing wwas done first:

Vadas, R.L. Jr., and D.J. Orth. 1997. Species associations and habitat use of

stream fishes: the effects of unaggregated-data analysis. Journal of

Freshwater Ecology 12: 27-37 (cf. http://www.tandfonline.com/doi/abs/ 10.1080/02705060.1997.9663506).

-Bob Vadas, Jr. ([email protected])

In many applied situations it seems likely that fish may avoid nets based on probable acoustic detection of the noise of water flowing through the nets.

Certainly the noise of the net deployment vessel has been shown to influence net catches though that is not a direct effect of the net on the fish.

Hi. since 2009 I have used individual encounter histories of pre-spawning Chinook salmon caught and marked with beach seine at Burman RIver in Nootka Sound to  address transient and trap avoidance behaviours.. I used U-CARE (Choquet et al. 2009)  to detect survival (transients) and capture P (trap avoidance) heterogeneity and correct both with time-since marking , or age2*t models.