I think we'd expect that even the most generalised predator will take proportions of prey that are non-linearly density-dependent, as microhabitat use and search-image vary over time and different prey differ in value and catchability. It would be useful (not easy) to get detailed time-series data on encounters and captures to see if frequency of prey types varies quasi-continuously or switches at discrete times.

Thank you very much indeed for your contribution, John. Our work was based on spraints, to determine the seasonality of the diet of the otter, so we should change the methodology to actually monitor the behaviour of the otter (which could be very interesting for future studies). However, taking into account what you´ve just explained, I think there are very few possibilities to monitor this behaviour.

1) Install micro-cameras and monitor the otters in the nature. I think this is very difficult, as this mustelid is a dawn and crepuscular predator, so light could not be enough to see what the otter is actually doing.

2) Studying otter behaviour in "controlled" conditions would be easier, but I believe that the results would include lots of uncontrolled variables (stress, for example).

3) Otherways, the "classic methodology" (collect spraints) would only give us information about the density (and number, if possible) of prey present in the faeces. We could, in addition, start some electrical fishing activity (with all the legal permissions required) to see whether in each season the density of crabs (Pacifastacus leniusculus) and fish (Cyprinidae), is different, and then try to establish a relationship between the density of these prey and their proportion on the spraints.

What do you think? Thank you very much in advance.

Do any spraints contain remains of multiple prey individuals or species? I was once able to identify six vertebrate prey species from bones in a single Canis dingo scat (diverse small prey and carrion diet) but can understand why this sort of thing might be rare in aquatic mustelids. However, any cases of multiple prey per spraint could be used to construct a model time series.

I did think of something like your option (2) using a small captive colony, adding prey types at different concentrations and analysing outcomes. Excellent infra-red and low-light cameras are now available so I wouldn't rule out possibility of (1) either. Good luck!

Surely they do! In fact, identifying species from bones or cuticle is not very difficult. We found some spraints with multiple prey individuals (up to 4 or 5 small Eurasian minnow (Phoxinus phoxinus) in a remain, from pharyngeal archs) and multiple species per spraint too. The problem came when trying to identify the number of crayfish present in each spraint, as their remains are usually quite deteriorated, making the decision on the number of individuals very difficult. Moreover, otters don´t eat the biggest and hardest parts of the cuticle of the crayfish (for example, the first pair of pereiopods), which, unfortunately, would be key when counting specimens...

I will surely take into account your advices, and bring them up to my colleague. Thank you very much!

It's also worth noting that many generalist predators do not always focus on the most abundant prey types but instead exhibit a degree of preference or 'individual specialisation'. This has been observed in sea otters (Newsome et al. 2009, Estes et al. 2003) which are generalist at the population level, but at the individual level otters specialise on certain prey types. A large enough sample of faeces or spraints might give you some information on this (assuming you can assign spraints to a specific otter/territory), ie does resource use correlate with resourec availabilty in some individuals /territories but not others? Ideally If you could obtain hair or whisker samples (I appreciate this is very difficult) you could use stable isotope analysis to investigate this (we have done this with badgers and found some interesting results). I also agree with the earlier comments about cameras, we use Bushnell camera traps fo filming badgers at night (ie in complete darkness) and we have got some good feeding onservations, assuming the badger is close enough to the camera!

Although the Eurasian Otter is a highly specialized species you are right to assume some opportunistic behavior resulting in this change of diet. Usually otters prefer fish of a suitable size and mobility, crabs may be “low value food” which is taken in times when fish is rare or hard to be caught. As you write, a cluster of several factors will influence the animals choice: individual preference, prey availability and costs for catch effort. Physiological needs (lactating females), and geographical conditions will also interfere with these… There is some pretty work on this published by J. Ruiz-Olmo (I guess you may know these already, but I regard them as helpful):

• Observations on the predation behavior of the otter Lutra lutra in NE Spain. Acta Theriologica, 40, 175-180 (1998)
• Capture and consumption of prey of the otter (Lutra lutra) im Mediterranean freshwater habitats of the Iberian peninsula. Galemys, 10, 209-226 (1998b)

Methodological approaches to get further information about this prey consumption pattern are already mentioned by John and Andrew; the informations gained by analyses of prey remains in spraints are clearly limited  (see  in Kruuk, 2006, Otters: ecology behaviour and conservation, Oxford Press ISBN 0-19-856586-0; p. 99 ff) so I agree with Andrew:  it would be interesting to confirm these results by observation of behavior ( no easy attempt in nocturnal activity ) and by stable isotope analyzes to detect the seasonal changes.  

Thank you Ernst! I know those papers, but thank you for the recommendation. The only one I don´t have is Kruuk, 2006, as it is closed for us (even in the university).

Well, after some discussion with some colleagues, we have just started a "pilot experiment/project", with two camera-traps (phototrap) situated in two places where we had seen otter previously (and where some faeces had been deposited). 

Our aim is to see whether we can manage to film otters eating their prey, and try to identify the prey. In the literature I have seen that otters usually consume their prey calmly outside the river (when the prey is not small), so I hope that with this project we can film some otters and, in the future, have more cameras in the river.

With this experiment, we haven´t solved the problem with the differential abundance of the different prey (crab, fish, amphibians) throughout the year (consequently having some bias in the data). But, I think that this is all we can do with our budget (none...), and it can be a good approach to guess the dietary habits of the otter. In the future, as most of you stated, we will think about some stable isotope analisis (not present in our lab, so difficult to achieve at the moment), or some controlled conditions with special cameras.

Thank you all for your contributions!! If you have any other opinion, just tell me, I appreciate all the help given.

Diego

Following on from Andrew Robertson's comment in relation to 'specialisation', while it has been recognised that Eurasian otters are generalists/ opportunists at population level, and that individuals are capable of being specialised in the predation of particular species, there may also be case for specialisation amongst sub-populations inhabiting certain river levels/ stream orders.

While not currently published, my undergraduate research clearly showed a generalist otter population in the main river of the study catchment, and a (salmonid) specialist otter population in one of its tributaries. The study was conducted during the salmonid breeding season, therefore the salmonids were abundant in the system, however, the generalist population showed no significant switch to salmonids in comparison with the specialist population. Therefore, such prey selection may occur at stream order basis (main river/ tributaries).

Are your two camera-traps set up on the same river as it may influence your findings?

The problem is to identify those subpopulations as you state. Past studies in this stream show that, at least, we have 3 individuals in 10 km of the stream (which is a tributary of Ebro river, so it´s pretty small to have different sub-populations of the otter I presume). But it is very difficult to assure the number of individuals, as they move a lot and this movement is very influenced by the seasonality and the abundance of prey.

My undergraduate research also showed a generalist population. And yes, my two camera-traps are set up in the same stream, with a distance of 6-7 km. The first images obtained show, at least, two different individuals.

Thank you for your answer!

Well, I think that not always the most abundant prey is the most suitable one. In general, I think otters (as most of predators) try to balance the hunting effort with the energy intake. Thus, sometimes they prey on suboptimal preys (eg. crayfish) just because they are easier to catch tan preferred preys (fishes).