The production of Liza ramada (Risso) in Lake Mariul, Egypt 255 



Another possible alternative or supplementary cause of this change might be a defer- 

 ment of maturity of O group fish resulting from the increasingly adverse conditions 

 in the lake. 



With regard to growth there is also the appearance of a change in my samples 

 between 1928 and 1933. For all ages the trend is upwards (Table II). 



The continued diminution in the numbers of O group coming into the catch between 

 1928 and 1933 is difficult to reconcile with the apparent stability of the stock from 

 1930 to 1935, and it may be that this has been achieved by a deferment of maturity 

 causing more of the O group to stay in the lake and away from the spawning shoals 

 until they become I group, and by an increase in the growth rate. The observations 

 however are quite insufficient to make this any more than a reasonable conjecture. 



The Incidence of Mortality 



The decline in the landings and almost certainly of the whole stock of L. ramada 

 in Lake Mariut does not appear to have been directly attributable to changes in the 

 supply of fry, for, as we have seen, four million fry at the commencement of the 

 period is estimated to have produced about a million O group fish and in 1933 only 

 a little over thirty thousand. Indeed we have seen that there is some indication that 

 the strength of the stock in the lake may have influenced the supply of fry. 



Undoubtedly the shrinking habitat of the lake caused by successive reductions in 

 its level has been the cause of the big and successive reductions of numbers indicated 

 by the estimates of O group fish shown in Table 1. We then come to the question at 

 what period in the life of the O group fishes is this reduction in numbers brought about. 



On the observations here available I am inclined to think that the mortality takes 

 place very early in the period. If there were a mortality acting slowly over the whole 

 first year's life, such, for instance, as overcrowding on a ground where the food 

 supply was not adequate for the whole population, there would be a corresponding 

 effect on the growth-rate which has not been observed. In these circumstances I 

 believe that an early mortality, perhaps of a catastrophic kind, occurs very soon after 

 or possibly during the actual introduction. It must be remembered that the little 

 mullet on first introduction has a short gut and is able to eat plankton animals, but 

 that later it develops a long and much convoluted intestine and becomes a vegetable 

 and detritus browser. This change-over takes place at a very early age, and would 

 produce a condition when a food shortage caused by a shrinkage of the available 

 shallow water habitats due to a reduction of the lake-level might be critical, and cause 

 an immediate thinning out of stock sufficient to explain the successive reduction in 

 O group strength. 



If this is so, it would mean that each year the lake accepted a certain number o\' 

 recruits corresponding to the lake level. After this acceptance was over, there appears 

 to have been enough food for the survivors to make normal growth. Indeed, as we 

 see in Table II, there are indications that growth has been rather better as the lake 



level has fallen. 



The mechanism of this sort of mortality is of the greatest interest in the study ol 

 the populations of our food fishes. The parts played by the nature of the organic 

 and inorganic environment, and in particular the role of competition, presents problems 

 for solution by careful observation and experiment. 



