Is Sardine a candidate for science based managment measures?

Welcome back after a long hiatus. With due apologies for not keeping this forum as active as intended may I moot this topic for discussion at right earnest, which hopefully absolve me of a part of the blemish? Indian Oil Sardine, OS, has been one marine resource which has been moving and shaking the Indian scenario for sometime now. Whilst it leads the chunk of pelagic resources, it has one of the most perfected selective fisheries of India revolving around it. While estimated landings of this resource off Indian waters, spearheaded by southwest coast, have been declining at a double digit rate per annum since 2012, when they hit a collective peak, efforts are on at various levels to reach the route to its root cause as it has adversely affected a particular thriving fishery to such an extent that Kerala, a very vibrant marine fishing province has lost its preeminence in the hustings amongst Indian states.
Now the robust issue to be addressed to is whether OS, a small Clupeid with a fast rate of growth and preponderance of herbal feeding and relatively shorter lifespan, would be the zebra-fish for scientific management? I for one believe that adding the adjective 'scientific' to management in this context is blasphemous in itself. But still can this resource, which has historically shown significant spikes with heteroskedastic amplitudes, be amenable to a comprehensive scientific assessment and submissive to the predictions and managerial measures emerging thereof ? Hope this thread invokes bridled enthusiasm and unleashes an array of plausibilities in subsequent quips. 

Modeling fish populations: Are we yet to reach to the bottom of the methodological shoal?

Frequentists are often at loggerheads with Bayesians. One believes in frame relay, that is one scenario leading the other as per universal norms, whereas the other does laser show, keeping everything except the sources of change flexible. Though there is a wealth of work in stock assessment of fisheries using the relatively fresh Bayesian methods, I have a serious poser as regards the priori Dn assumptions. Usually robustness of Bayesian inferences are often directly regressed upon the non-informative priors. Can we really have this comfort when we employ Bayesian techniques for Fisheries research? In my opinion opting for non informative priors for stock defining parameters like carrying capacity etc would be too much of a risk to start with. Please post your opinion...

Getting a frame of fishing populace of a country

Census operations have been in existence from time immemorial. Probably General Census which are mostly decennial as in India is one grand exercise which is based on simple conceptualisation. Reach out to every citizen of the country within a specified time interval. The catch points are "individual" and "time period". All other information fall in place smoothly alongside these two pivots. The classic census exercise involves collection of socio-cultural, economic and infrastructural details of persons/ groups of co-existing individuals or families. The greatest strength in this exercise is that there is no ambiguity on who is to be approached. But if the complete enumeration is to cover a certain portion of the population of citizens of a country the spadework becomes huge. As mentioned in the other post, "whom to be considered and contacted to represent the fishing population of a country?" is the first question to be answered.

Another important facet of these exercises is the spotting of the target population. In this context we may have to evolve a strategy targeting the individual as well as his family. Socio- cultural information necessarily happens to be a collation of data on a subject and his/her social grouping (household/ family). A portion of the information is devoted towards assessing towards the combined status of the group (above poverty line/ below poverty line etc). Still within the family individuals will have more than one façade to show. That is the family, at least one member of which is a fisher, can have income from other vocations also. It may be that the individual has taken to fishing as a matter of compulsion than tradition. He or she may not also be the head of the family. Who knows he or she may be the worst off in the family. Information can be collected about them also. But in the hindsight what these figures have to do with the prime objective? If the prime objective is to simply get the number of persons involved in the core activity in the whole country then the mere inclusion of such individual as a count may be with the additional information on his gender and community will fit the bill very cleanly. But if we have to provide information on the individual's economic and social condition vis-a-vis government's welfare dossier, then the whole family has to be projected alongside the individual. Probably there may be scenarios where poor individual from a rich family (?) should not be getting a particular benefit. Another interesting aspect is to trace the economic progression of traditional fisher folks. A simple glance at there infrastructure (including crafts) and the educational status of their family members, this can be easily assessed. This is more interesting if the Census exercise is to be done at regular intervals. But in all these cases, the vexing is issue is how to zero in on them? Whether to focus only their habitats or to focus the crafts in operation is a major poser. As the information has to be first hand the individual has to be contacted or at least his family. Probably the listed fishermen of previous such exercises would be a very good starting point. This issue is of very high importance and sensitivity in countries where access to seas is open, registration of vessels/ crafts is yet to be complete and licensing to fishing is still having the birth pangs.

Sampling Fish Populations

The following piece is an abridged version of the queries flagged by my teacher Dr M. Srinath:

(i) How does one go about with the random selection leading to a sample while studying a particular resource especially when the frame is impossible to construct?

(ii)Are we justified in terming fish samples taken from time to time as random samples?

(iii)Can we consider catches from various gears as independent random samples?

The answers would be serving better if they satisfy both a statistician and a biologist.

Marine Biodiversity valuation

As any other natural resource marine resources too have immense value and utility even on encashable terms. In a way the valuation and budgeting of the marine resources is very much a priority for any avid policy watcher. But unlike its superterranean counterparts the marine ecosystem has lot of things waiting to be unshrouded. Practically none of the pivotal factors worth its utility in budgeting them can be assessed in the mathematical sense. Unlike forests and other animal resources marine resources have unique set of dynamics at every part of their biology. At times quite a few species have shown extreme sense of resilience especially against adversity so much so that post threat certain species have over dominated unusually. Further most of these resources are in a state of flux. So this topic needs careful intellectual assuaging before being unleashed in zones hitherto bereft of such studies.

Stock and Statistics

It has always been fascinating to mix two seemingly immiscible concepts with a firm objective in mind. Statistical ogling of fish biological dynamics happens to be the best suited case under this category. One of the immediate moot points in this line is the stock based inference arrived through quasi-linear or non-linear modeling. Statistics being inductive in nature is bound to blow up the results to some universe and "stock" fits the bill perfectly. But there are non-statistical issues on the very definition of an unit stock. The wikipedia throws up a dozen such definitions and points to the most acceptable one which unfortunately happens to be the latest. (http://en.wikipedia.org/wiki/Fish_stock) this has the potential danger of the query- "what happened to previous inferences". In the sampling concept a sample and population are uniquely, indisputably defined and the methodology sticks on that perfect framework. But it seems to be not the case in Stock assessment. Further stocks by definition vary with the part of the globe in which the study is being conducted. Another issue is the collation of various stock based inferences- are the additive? or some interaction is to be accounted for? Are the pelagic and demersal assessments even footed in terms of setting?

Who is a "Fisherman"

The term fisherman has now been perceived in more ways than what the lexicon would interpret. From the innocuous poor rod wielding person to a multimillionaire entrepreneur all in a way come under the gamut of the meaning conveyed by the term "fisher". The arguments targeting this issue always end up inconclusive. But they always throw up defining assumptions and purposes which brings some finiteness to the whole exercise. It would always be helpful if we could initiate healthy debate on who all should be labelled as a fisherman if the information seeker happens to be a state/ government which has welfare motive as its primary focus.