A REAPPRAISAL OF MATING CALL DIFFERENTIATION IN HYLA CADAVERINA (= HYLA CALIFORNIAE) AND HYLA REGILLA.

Abstract

Ball and Jameson (1966) in their of reproductive isolation between the broadly sympatric western North American tree frogs, Hyla cadaverina (as H. californiae; see Duellman, 1968) and H. regilla, paid particular attention to differences in mating call structure, which they presumed to be the most important of the possible premating isolating mechanisms. They described the call of H. cadaverina as strictly monophasic and lacking the compact trills (= pulse modulation) present in the usually pulsed, diphasic calls of H. regilla. But instead of concentrating on these marked differences, they selected three other call components for detailed multivariate statistical analysis: mid-point of the dominant frequency, call duration, and call repetition rate. Snyder and Jameson (1965) made a detailed of geographic variation in the mating call structure of H. regilla throughout its extensive western range, and also omitted any consideration of pulse repetition rate. Although they did not consider the presence of H. cadaverina as a possible selective influencing dominant frequency in the southern populations of H. regilla, they showed that H. regilla females could effectively discriminate between the mating calls of the two species and were attracted only to calls of conspecific males. Other monophasic signals, presumably associated with territoriality, were also described. Following a complex statistical analysis involving several samples from singlespecies (allopatric) and two-species (sympatric) breeding assemblages (all of which are from within the area of geographic overlap; see range maps in Stebbins, 1966), Ball and Jameson (1966) arrived at the following conclusions. 1) Since there is overlap in the values of all of the call components measured, one call character appears to be independently operative as the primary isolating factor (page 544). Evidently differences in the total call complex of each species is (sic) significant enough to result in interspecies isolation with no one parameter being the major selective force (page 550). 2) Little evidence of the reinforcement of morphological or call parameter differences between Hyla regilla and Hyla californiae was demonstrated in this study (page 545). Both of these statements, but more especially the first, are widely at variance with what has been reported for other sympatric hylid frogs. Blair (1958), Crenshaw and Blair (19591), Littlejohn (1959, 1965, 1969), Michaud (1964) and Ralin (1968) have drawn attention to the striking differences in pulse repetition rates in mating calls of closely related sympatric species of hylid frogs. In general, there is no overlap in ranges of variation, and means differ by a of two or more (Littlejohn, 1969). Results of two-choice female discrimination tests between sympatric hylids in which the mating calls differ principally in pulse repetition rates (Littlejohn, 1960; Littlejohn, Fouquette and Johnson, 1960; Littlejohn and LoftusHills, 1968; Michaud, 1962) showed that sexually active females could differentiate between the calls and were only attracted to those of the homospecific male. LoftusHills and Littlejohn (1971), using syn-