Predicting Aquatic Monsters

A scientific look at the unknown large animals of river, lake and sea.

Introduction

Speculation as to the nature of large unknown aquatic animals has generally occurred in the absence of quantitative data and relied almost solely on eyewitness testimonial. This need not be the case. I (Paxton 1998, abstract here) estimated the number of unknown large open water marine animals awaiting discovery by science based on an assumption that the scientific description rate for unknown large aquatic animals from 1830 could be extrapolated into the future. If this is true then the cumulative species description rate can be modelled as a rectangular hyperbola (Figure Two) and an estimate of the number of large unknown open water marine animals could be made.

There were a number of problems with the assumptions of my analysis. I used species description rather than species discovery as my criteria of recognition by science. There can be a substantial gap between the two although description "more truly reflects recognition by science of a new species than mere discovery." (Paxton's 1998 words!).

More fundamentally, there were all the standard logical objections to induction as a method. Can the description of rate of the last 170 years be usefully extrapolated into the future? How realistic is the use of a rectangular hyperbola? And we really need some error bars on the estimates.

Then there were the statistical problems with the approach. The cumulative species description curve is made of non-independent data points in time. This meant that confidence intervals could not be calculated for the estimated constants. The constants themselves were biased, as the method assumed constancy of "effort in searching for unknown animals" from 1830 to 1995. There has almost certainly been a change in effort. However this overall effort has probably increased with ever more reports of strange carcasses stranded on beaches, more marine biological research and large scale pelagic netting. Thus the estimates of the numbers of new species awaiting discovery are too liberal. The method also assumes that existing methods of species identification and discovery will continue. The results would be invalid if there were a proliferation of cryptic species discovered "within" known ones by say molecular genetic techniques.

How many?

If the assumptions hold (big if) then based on my database of 1998, there are some 47 large (> 2 m long) open water animal species awaiting description by science. The rate of new species descriptions appears to be about one every 5.3 years.

This figure now seems rather high and some unpublished work suggests the figure could be much lower.

Main page

References

Author

______________________

Last modified: 29^th July 2009.