The construction of an Ecopath-type ecosystem model relies on the truism that, for any producer, (e.g., a given fish stock,) and time period, (e.g., a year or season)
Production = fisheries catch + predation mortality + other mortality + biomass accumulation + loss to adjacent systems .
In addition the groups in the system are linked through predators consuming prey. Such consumption can be described by
Consumption = Production + non-assimilated food + respiration.
This implication of the two relationships is that the model is mass-balanced, i.e. mass is "conserved", or accounted for. Call it as we may, this principle provides an extraordinary rigid framework - formalized using a system of linear equations - through which the biomasses of different consumer groups within an ecosystem can be estimated, along with the trophic fluxes among them.
Important here is that the information required to complete an Ecopath model is of the very type routinely collected by fisheries scientists, (e.g., catches,) or estimated in the context of single-species studies (biomasses, mortality rates, diet composition, etc.), along with biological information on the non-harvested components of the ecosystems -- as usually studied by marine biologists.
Thus, at least these two disciplines are brought together every time an Ecopath model is constructed, and a large amount of scattered information is standardized, and rendered mutually compatible that would otherwise have languished in scattered journals, reports and filing cabinets. The "EcoRanger" approach in Ecopath makes this process statistically rigorous.
So far, over 50 models of marine and estuarine ecosystems have been published, authored by over 60 scientists. Moreover, these models have not only summarized a large amount of data for the systems they represent, but jointly, they have allowed for powerful generalizations to emerge. Ecopath has also been used extensively for training and education.
One of these is that the transfer efficiency of biomass between trophic levels in aquatic ecosystems, though highly variable, tends to have a mean value of 10%, as long suspected, but never before demonstrated conclusively (see Pauly and Christensen 1995).
Another generalization is that the fraction of fish production that is consumed by other fishes, is even within strongly exploited systems, much larger than the catches, hence invalidating the guesses which led to Gulland's equation (see Christensen, 1996, and Pauly, D. 1996. One hundred million tonnes of fish, and fisheries research. Fish. Res. 25(1):25-38).
Yet another generalization is that the primary production required to sustain the world fisheries is far higher than previously anticipated, itself suggesting broad limits to the carrying capacity of the world's oceans (Christensen and Pauly 1995).
Earlier, we had presented new approaches for rigorous estimation of trophic levels and their variance, Christensen and Pauly 1992), for calculating the aggregate flows into and out of each trophic level of a given ecosystem (Ulanowicz 1995), and thus to estimate TE values by trophic levels and ecosystem type. This approach now permits the re-evaluation and the testing of alternative hypotheses concerning empirical relationships established earlier, and based on indirect, and often rather inaccurate methods of unknown precision.
For example Ryther (1969) when estimating world fisheries potential guessed all the TE values he used, stating "Slobodkin [1961] concludes that an ecological efficiency of about 10% is possible, and Schaeffer (sic) [1965] feels that the figure may be as high as 20 percent. Here, therefore, I assign efficiencies of 10, 15, and 20 percent, respectively, to the oceanic, the coastal, and the upwelling provinces, though it is quite possible that the actual values are considerably lower."
When we look back at those attempts it is tempting to quote Parson et al. (1979), who wrote in a now classic textbook with regards to similar relationship that "by definition, these relationships are advanced as being the most acceptable at the time of writing the text but it is to be expected that researchers will improve or disprove many of the processes discussed in the light of future scientific advancement. Such is the nature of science." And now is the time.