Hi Jack

Good to hear from you.

The following papers are relevant to your question:

Hayes JW, Goodwin E, Shearer KA, Hay J, Kelly L 2016. Can WUA correctly predict the flow requirements of drift-feeding trout? —Comparison of a hydraulic-habitat model and a drift-net rate of energy intake model. Transactions of the American Fisheries Society 145: 589–609.

Weber N, Bouwes N, Jordan CE 2014. Estimation of salmonid habitat growth potential through measurements of invertebrate food abundance and temperature. Canadian Journal of Fisheries and Aquatic Sciences 71: 1158-1170.

The Hayes et al. paper focusses drift sampling to estimate a drift concentration x flow relationship and to cover spatial variation in drift concentration within a reach - all geared to support modelling trout NREI (and prediction fish numbers) as a function of flow.

The Weber et al. paper is aimed more at estimating spatial variation in invertebrate drift concentration and  benthic invertebrate densities among reaches and streams to support predictions of trout growth and consumption. And related papers from that team are aiming to put all of this together with NREI models to predict spatial variation in NREI/fish abundance.

e,g,

Mchugh PA, Saunders WC, Bouwes N, Wall CE, Bangen S, Wheaton JM, Nahorniak M, Ruzycki JR, Tattam IA, Jordan CE 2017. Linking models across scales to assess the viability and restoration potential of a threatened population of steelhead (Oncorhynchus mykiss) in the Middle Fork John Day River, Oregon, USA. Ecological Modelling 355: 24-38.

Wall CE, Bouwes N, Wheaton JM, Saunders CW, Bennett SN 2016. Net rate of energy intake predicts reach-level steelhead (Oncorhynchus mykiss) densities in diverse basins from a large monitoring program. Canadian Journal of Fisheries and Aquatic Sciences 73 (7): 1081-1091.

So your drift sampling protocol will depend on the aim of your study - i.e., whether it is to isolate the effects of flow on drift and then on fish or is more concerned with explaining spatial and/or temporal variation in drift and fish abundance.

If you want to isolate the effects of flow on drift (and fish abundance) then you might want to minimise the influence of behavioural drift on your drift estimates - by sampling just daytime drift and avoid dawn/dusk peaks (as did Hayes et al. 2016).

If you are more interested in estimating/understanding spatial/temporal variation in fish abundance then you will want to include dawn/dusk behaviour drift peaks in your drift sampling (as did Weber et al. 2014 - they sampled 24h drift).

If you want to deploy drift samplers in deep/fast water or just want to reduce the field input of having to record water velocities regularly at the net mouths (to estimated volume sampled) then invest in making samplers that have an internal current meter (i.e. plankton tow net type arrangement - as used by Hayes et al. 2016). But you'll also want to consider whether you stack multiple nets in the water column to cover the vertical distribution of drift (more necessary for predictions of actual fish abundance than for predicting relative fish abundance with NREI models... Again dirt sampling design depends on your study questions).

If you want to spend less on drift samplers and happy to accept greater field effort (ok if you have student slaves) then deploy samplers that don't have internal current meters. But be sure to take sufficient water velocity measurements to give you confidence in your estimates of volume sampled. Some researchers take this task too lightly. A great advantage of samplers with internal current meters is that they continuously account for clogging - i.e. because they record the flow actually filtered by the sampler. Coping with clogging with samplers that require manual velocity measurements is a pain. Nets need to be regularly cleaned and velocities recorded before and after each cleaning and redeployment. And theses samplers are restricted to wadeable locations where manual velocity measurements can be made. 

You'll need to put some thought into number and spatial distribution of samplers. Ultimately this is strongly influenced by the resources you have available.  

Good luck

John

Thank you John for the very complete and thoughtful answer. Both your 2016 and McHugh et al feature prominently in the proposals. I was also looking at Weber et als paper comparing drift to benthos - very interesting. I wlll try to get the flow meters - GO is what I used in my prior sampling life.

What about the size of the nets - mouth dimensions, mesh size, length of the net, and even material? Again, it's been a while since I've done this so I want to come in with the best available knowledge.

We have one PhD starting, hopefully in September, than a PostDoc proposed  -the later working on a project with the three largest hydropower producers in Sweden on restoration and Eflows, and even dam removal. One of the biologists, Marco Blixt, was in Nelson a while back, visiting with Ian Jowett. It would be very nice if I get the chance for a visit there in the coming years.

Please keep me updated if you change any of the modelling software. I also plan to get started with that directly this fall, so we are prepared when the first results come in.

Best regards,

Jack