ADAPTATIONS AND DISTRIBUTION 



and c)\ Illation ensues. This sequence is untenable lor Rana where the female 

 will not permit the male to retain his clasp unless ovulation is almost 

 complete." 



This outline iits perfectly my statement concerning those forms which 

 breed only ajter rain (i.e., those having the xeric breeding pattern). Con- 

 cerning others, in the same paper I said, "In other species wherein a breed- 

 ing season occurs irrespective of rainfall, factors other than rainfall, of course, 

 start the cycle of breeding activities; but in those (e.g.. Rami berlandieri and 

 B. w. woodhousil) whose breeding season is modihed or extended by the 

 coming of the rain, either some internal factor alone, or rain alone, or the 

 two of them working together at times stimulate the pituitary of the male, 

 the remainder of the cycle progressing as before." Whatever the details may 

 be, the essential point is that what applies to a mesically adapted species like 

 Rana pi pi ens does not necessarily apply to a xerically adapted form like Bujo 

 cognatHs or to a form like the leopard frog of Oklahoma, called by me Rana 

 berlandieri , which behaves differently from either the mesically adapted or 

 the xerically adapted species. I am convinced that Noble and Aronson are 

 substantially correct for the frogs which they studied, but I am also convinced 

 that their results cannot be applied to Salientia in general, particularly to 

 xerically adapted forms. (For further discussion, including detailed evidence, 

 see Bragg, 1937, 1940, 1940a, 1940b, 1940d, 1941, 1941d, 1942, 1942b, 1943, 

 1943a, 1944, 1944a, 1944b, 1945, 1945a; Bragg and Smith, 1942, 1943.) 

 (2) Selection of and Migration to Breeding Sites with Diverse Habitats 



The frogs and toads of Oklahoma may be classified into (a) those which 

 migrate little or none to reach a breeding site, and (b) those which character- 

 istically migrate considerable distances. The first group includes those whose 

 habitat is in or near water (all Rana, and Jcris crepitans). The second, of 

 course, contains the aboreal, fossorial, and, in general, the more nearly terres- 

 trial species (in Hyla, Bufo, Microhyla, Scaphiopus, and Pseudacris). In the 

 first group active selection of breeding sites is rare or nonexistent; in the 

 second, it is common although not universal. 



There is some evidence that so-called selection of breeding sites is, in 

 some forms, no more than the chance result of the marked attraction of the 

 calls of males which happen to arrive first at a given pool. In other species, 

 there seems to be a clear selection of one or a few types of sites among sev- 

 eral types which are equally available. (See Fig. 15.) In each case "selection" 

 may or may not cut across taxonomic lines. The following discussion illus- 

 trates this: 



In an earlier paper (Bragg and Smith, 1942) notes were given on the 

 breeding sites of frogs and toads based upon four years of field studies in 

 Oklahoma and immediately adjoining areas. Observations have continued 

 each year since that time; now a total of about ten seasons gives a fairer basis 

 for generalizations concerning many species inhabiting the area. Breeding 

 congresses and pools inhabited by tadpoles have been examined in every 

 county in the state but the most detailed data are available from the area 

 within a twenty-five mile radius from Norman, Cleveland County. Some 



69 



