EXTERNAL FACTORS ON THE SPAWN DATE I47 



at low rainfalls. For example, in the Thames basin, there is no run-off 

 when the rainfall over a period is less than 52 per cent of normal. In 

 many places on the diagrams it can be seen that the effects of other 

 factors, such as temperature, are only shown when the rainfall is high. 

 The exception is in the sunshine diagrams, and this has been explained 

 as being due to the antagonistic effects of sunshine, which causes 

 evaporation and so reduces run-off, but has another effect as well in 

 the opposite direction. When the temperature is so high that evapora- 

 tion can be expected to be abnormally high for the season, spawning is 

 delayed. The mineral composition of the pond water is also im- 

 portant. Frogs prefer ponds with high potassium, and those wliich, 

 from a high level of phosphate in January, fall to a low one at spawning 

 time. This pattern is well known to be characteristic of ponds that 

 have a rapid growth of algae. 



To suppose that rainfall is acting directly on the frogs is surely most 

 improbable. Many of them are already hibernating under water at the 

 time when these factors are operating, and could not be affected by 

 evaporation. Moreover, they may leave these aquatic liibernating 

 places for other ponds in wliich they spawn, so that they could not be 

 under the influence of potassium or phosphate in the ponds, for they 

 are not even there. The same applies to light. Some frogs may be in 

 the light when they are hibernating, but most are not. Yet sunsliine in 

 a month when they are still hibernating affects the spawn date. Now, 

 it caimot be argued that some frogs are affected by these factors, and 

 that some are not, for the spawning season in one pond is so short that 

 all the frogs in it must have been under the influence of the same factors. 

 If they were not, there should be two widely different groups arriving 

 at quite different dates, separated perhaps by months. This does not 

 happen. If, on the other hand, ponds differed so that some contained 

 frogs in the winter, but some did not, and that this affected the date of 

 spawning in them, then ponds should fall into two groups : early and 

 late. The charts. Figs. 29-32, show no sign of this but Fig. 41 is more 

 convincing. In this diagram are plotted all the available results from 

 South-East England, as departures from the mean date for the year in 

 that district, so as to ehminate the variation from year to year. If there 

 were two groups of ponds, tliis graph should have shown two peaks. 

 It did not, so that we must suppose that spawning in a district is a 

 continuous process. The shape of this diagram has some statistical 

 interest, and is discussed in greater detail in Chapter 10. 



