700 



ECOLOGY AND EVOLUTION 



York state, all confined to the species of 

 the white oak group, and most of them to 

 a single species of oak. 



Predation has strongly influenced the 

 evolution of both the exploiter and the ex- 

 ploited. Worthington (1940) postulates an 

 astonishingly rapid adaptive radiation of 

 fishes in Lakes Victoria, Kioga, Edward, 

 and George in Africa (p. 611), and points 

 out that L.':ikes Albert and Rudolf have no 

 such evolution of endemic forms, though 

 similar ecologic conditions have been es- 

 tablished for at least as long a period. 

 Lakes Victoria and Kioga have fifty-eight 

 endemic cichlids, Lakes Edward and 

 George have eighteen, Lake Albert has 

 two, and Lake Rudolf three. 



Lakes Albert and Rudolf have large ac- 

 tive predators, the Nile perch (Lates) and 

 the tiger fish (Hydrocyon) , while in Lakes 

 Victoria and Edward, the only large preda- 

 tors that survived the arid period, or that 

 regained access subsequently, are the 

 rather inactive lungfishes and certain cat- 

 fishes. In other lakes with the same poten- 

 tialities the predators prevented a parallel 

 rapid evolution. Lake Nyasa survived the 

 Pleistocene arid period because of its great 

 depth. Species of the genus Barilitis are 

 less active predators than Lates and Hydro- 

 cyon, which are absent from Lake Nyasa. 

 Nearly twice as many species of endemic 

 cichlid fishes (171) have evolved in Lake 

 Nyasa, as compared with Lake Tanganyika, 

 where there are eighty-nine endemic cich- 

 lid species together wdth Hydrocyon and 

 two species of Lates. The great evolution 

 of cichlids in Lake Tanganyika as com- 

 pared with that of Lake Victoria may be 

 explained by the much greater age and 

 depth of Lake Tanganyika (Hesse, Allee, 

 and Schmidt, 1937). 



Cause (1934a) experimented upon lab- 

 oratory controlled predator-prey relations 

 (p. 371). 



"The destruction of one species by another 

 has been studied with Paramecium caudatum 

 being devoured by another infusoria, Didinium 

 nasuttim. Experiments showed that this bio- 

 logical system presents no oscillations in the 

 numbers of individuals peculiar to itself, and 

 that in spite of abundant food for Paramecium 

 the latter are completely destroyed by predators 

 which perish in their turn later on. However, 

 oscillations appear if we admit a controlled 

 and simultaneous immigration of predators and 

 prey into the microcosm (Fig. 239). There- 



fore, it is not the interaction itself, as would 

 be expected from the mathematical theory 

 developed by Lotka ( 1920 ) and by Volterra 

 (1926), but the constant interference from 

 without that leads to the oscillation in numbers. 

 ... In our experiments an anlaysis was made 

 of the role of cover or refuge for the prey in the 

 processes of the struggle for existence. This 

 showed that when the number of individuals 

 becomes reduced, and the conditions in the 

 microcosm complicated, instead of the 'deter- 

 ministic' processes subject to differential equa- 

 tions we are confronted vdth 'probabilities of 

 change' in one direction or another." 



DeBach and Smith (1941) discuss "in- 

 herent" oscillations in host-parasite systems. 

 Experiments on populations of housefly 

 puparia and one of their parasitic species, 

 Mormoniella vitripennis, reacting through 

 seven generations, followed the theoretical 

 predictions of Nicholson and Bailey (1935) 

 closely. DeBach and Smith think that the 

 periodicity is inherent in the predator- 

 prey relation, but they introduce conditions 

 that Gause considers external, so that prob- 

 ably no basic difference in principle is 

 involved in the two treatments (pp. 384, 

 705, Fig. 239). 



Ullyett (1936) studied host selection by 

 the chalcid, Microplectron fuscipennis. He 

 showed that the insect was able to distin- 

 guish to some extent between parasitized 

 and unparasitized hosts and also to choose 

 proper hosts in contrast with hosts that 

 could not serve for the completion of the 

 parasite life cycle. Local concentrations of 

 parasites indicated intensive searching for 

 hosts in a local region. Lloyd (1938) con- 

 cludes from his studies of host selection bv 

 the egg-parasitic chalcid, Ooencyrtus kti 

 vanae, that eggs suitable for the offspring 

 are generally chosen, and that the fraction 

 of the eggs found rises as the host den- 

 sity diminishes. Salt (1936) studied the 

 effect of host density on parasite den- 

 sity under experimental conditions (see p. 

 383). He used the chalcid egg parasite, 

 Trichogramma evanescens, with its host, 

 the eggs of the moth, Sitotroga cerealella. 

 Five females, capable of depositing 108 

 eggs, produced 84.4 progeny per 100 

 available hosts, while fifty females, capable 

 of depositing 1080 eggs, produced only 

 29.8 progeny per 100 available hosts be- 

 cause of the competition for food when 

 several parasite e?s[s are laid in the same 

 host egg. Of these 29.8 progeny, 12.8 



