I would suggest a Van-Veen grab, but maybe the sampling surface is too small. It works really well for macrofauna, but I guess you are working with megafauna.

Try to use a standard camera with a light. (see paper).

We sample for grass shrimp monthly with a quadrant that has vertically extended mesh sides that are taller than the water column. However, all of our sites are accessible from the creek bank and we are sampling on foot.

Well, I think the main issue that could address which is the best option for your sampling is which Decapods species are the target. Active species usually are sampling by traps considering the bait's attraction as a relation of the area or in case some low tide crabs, direct counting by time and line of observation is used. Species that stay at the bore the most of the time (specially at the schedule of sampling) could be sampling by core tube, quadrate and other incorporating the depth of the sediment as a factor. In any case important consider transect lines and time as part of the information.

Maybe this study can help you further:

Pranav J. Pandya. and Kauresh D. Vachhrajani (2010). Spatial distribution and substratum preference of the brachyuran crab, macrophthalmus depressus (decapoda, ocypodidae) along the lower estuarine mudflat Ofmahi River, Gujarat, India.

Published in Crustaceana 83 (9): 1055-1067

I've used a cast-net or throw-net (circular, with weighted edges and a rope attached to the centre) to sample shallow water small fishes in the tropics; it might work for your situation. Best wishes Julian

In cae very difficult to catch the animals, use semipermanent plot. Put the plot and waiting for in enough distance you can see. When the animal come or go out from their hole count it (using naked aye, binocular, or camera).

Micha, try these quantitative methods depending on your general or specific purpose: (1) a barrier net staked into the mud in the form of an L or V during high water and enclosing a specific and calculable area, fauna can be collected at the vertex during low water (especially good for studying faunal ingression). (2) A small mud sledge (see design by Pullen et al. (1966). Limnol Oceanogr 13:200-202) will be good for small hyperbenthic shrimps like mysids and sergestids, and settled penaeid postlarvae. Put the net down onto the mud while on a small boat. Pay out a fixed length of tow rope, say 30m, while slowly moving the boat away. Then pull in the net when the boat stops. Hope these work for you. best, vc

Can you use a V-shaped ground net that traps crabs with the outgoing ebb tide? However, this would not give you real densities/abundances...

Alternatively, seign-netting would be even easier (at least where the water is shallow) and you could sample a defined area...

Martin, yes in Malaysia, this type of gear is used by fisherman to fish on the mudflat. The wings span from 100 to 400m in length on either side, depending on the configuration. While fish tend to be trapped a the vertex, prawns and crabs will also be collected on the ground along the wings. Of course, some types of crabs just stay put in the mud. best, vc

If water level allows for wading or using a small boat respectively, you could try a drop trap (e.g. as invented by Phil et.al.). You can find the method and the citation of the original construction in. Polte et. al. 2005a, Mar. Biol -for the boat drop trap, and Polte et. al. 2005b, ECSS for the one used by wading.

Why not try side scanning sonar? See e.g.

Foote, KG; Hanlon, RT; Iampietro, PJ; et al. 2006. JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA Volume: 119 Issue: 2 Pages: 844-856

I you need to normalize your samples by area and perhaps volume or weight of sediment. You must use the same cores diameter and buries them at the same depth. Also you need to sift through your samples individually in diameter sieves with meshes that can retain their organisms of interest. And you need a lot of patience to triage them into a microscope.

Another thing to consider is the portion of the population you are targeting (the whole population, just adults, juveniles or settlers). Collecting juveniles and adults normally have very different efficiencies depending on the sampler. Nets are efficient for adults, but early juveniles are best sampled with some sort of suction sampler over a quadrat, particularly if there is vegetation. "refuge traps" are also efficient, but very species specific.