Could you please specify what elements of the dynamics you want to know? If you want to know the population size at a given point of time, I think the best way is to apply tagging. You can also analyse catches of the species, if these are exloited commercially and good stats are available. Third, sample the fish in the lake and study the size and age structure of the stock. Well, there is a lot of literature which might help you in your studies, I'll attach some. Best wishes, Piotr.
Article Effects of size-selective mortality on estimates of the grow...
Article Growth rate of zander (Stizostedion Lucioperca L., 1758) in ...
may be it is good to study seasonal dinamics of population growth and appaerance of invasive fish species wich could decrease other fish species. if the water is polluted you can try to study fish interaction (coaction) with other organisms, because most probably the fishes are infected with different type of parasites.
Depending on your desired outcome of the study, there a a variety of different ways of of assessing population dynamics and age structure. A very good reference that helped me through my work is the book "Analysis and interpretation of freshwater fisheries data" published by AFS. I attached the link, and it is also on sale right now for 20$ plus shipping.
I just want to turn your attention to the potential of using active acoustic methods combined with net sampling. This not only give you abundance estimates but will also picture the biological environment of your target species. A reference volume is the symposium proceedings. Go to Doi: 10.1016/j.fishres.2011.11.013 and from there to the full volume if of interest.
Tagging is a good idea, even a great one, but in a large lake system it can be very hard to impossible to get robust estimates of parameters like survival or abundance. A key design element would be finding "pinch points" where fish can be sampled and tagged where they concentrate, such as spawning areas.
Article Improving Inferences from Fisheries Capture-Recapture Studie...
Divide your lake system into different sampling regions based on the chances of fish aggregations. Use rigorous gear based sampling to collect the major fisheries resources in the lake system. Analyse the length-frequency distribution of fishes on spatio-temporal platform. Note the site wise (sampling) CPUE of various fisheries resources in the Lake. This is suggested because of the difficulties in tagging and recovery. You should sample for at least two years continuously to get a clear picture of the fish population.
All of you may be right, but there're lakes and lakes. And there're different water climates and, hence, species, with different behaviour patterns (say, schooling or solitary), and different prey-predator relations. I, for eample, did some work on L.Victoria, Kenya & co., world's second largest, not counting the Caspian Sea, where introduction-invasion of Nile perch had changed the whole structure of fish population, and since kept changing (with the help of fishing). Also worked on the relatively small L.Kinneret, Israel, where more or less regular fluctuations affect population size of some fish, while its level changes, and hence size of inshore spawning areas affect that of most fish. Consequently, I'd advise Yahya to choose assessment method according to the local climatic situation, the fishes, lake area and depth, and other ecological conditions, including the character of fishing on the lake. Happy 2015.
I've omissed one more factor: the reliability of the collected data. Where exact numbers are not available and the error may be within a wide range (sometime such as over +- 50%), feeding them into even the most precise models, would result in misleading figures. The fact is that no model can produce results better than the data it's fed with.