7861 - Laserfiche WebLink

Home Browse Search

658 COPEIA, 1989, NO.3 DC . o . . . ') ")0 o 0 0 o 0 1.0 . : . . . .1. . 1.0 2.0 N U a... o. .. 0 . .0 1.0 . 0 , . . . o. .. 0 . .. 5 to c o o 08 1.0 . ...,-. o 1.0 . ..- . .~. . . '0 .... o 0 . : ,., I .. . . . . .- -.: . ... o o 0 , 0 . 0 0 0 o 0 o ........ 2.0 3.0 PC, Fig. 3. Plot of PC, vs PC. for qualitative characters collected from fishes ofthe genus Gila from the Yam- pa River, Colorado: A) Plot of seven characters, scored 1-5; B) Plot of seven characters, scored 1-3; C) Plot of five characters, scored 1-5. Symbols: open circles, specimens identified as G. cypha; and closed circles, specimens identified as G. robusla (fewer than the orig- inal number of specimens may be represented due to superimposition of individuals in vector space). For further information, see text. 1.1 titative characters, but that under specific con< be relatively confident i characters for field iden Of most importance to manager is the rapid an of species, not the mea fication is accomplishe< The marked success in segregating these tax inherent information ( type. By this we mean 1 are not single charactel ones), but in fact are I overall variation withir separation of species (e.g., skull, mouth/sno so forth). Although a' cially like the nuchal numerous separate I themselves may not 1: methods of quantifical et aI., 1979). Also, iT comprise these region informative when eval separately. This, in fa titative characters we separate taxa. Qualita separation because I marized discriminatin separation was maint, in scale of the data I (although remaining) lation features were I examine use of these tures in future studie~ Does this mean th cannot be discrimim morphometric data? variety of morphom from preserved speci' separation (Smith et 1985). We demonstl of this study, in wh collected data for a ters in a brief time I conditions, twO forn arated on the basis ters used (Fig. 4). metric analysis of th of Gila will require unambiguous quan plished of those criti marized by our qu; has already been , C\I I . :. ..,. e. e. -1- c:....~. :O-t. 0 .. 0 .~ I CD o 0 o . . .. .. o 0 ... o 09 0.1 0.' 0.' H3 A Fig. 4. Plot of sheared PC. vs sheared PC, for eight qualitative characters computed for cyprinid fishes of the genus Gila from the Yampa River, Col- orado (symbols as in Fig. 3). Specimens identified by qualitative features as G. c)'pha stratify somewhat within the cloud of G. robusta, but the two species cannot be un- equivocally discriminated. Counts of fin rays were excluded from PCA to avoid combination of continuous and discon- tinuous variables. Dorsal rays for G. robusta var- ied from 7-9 and averaged 8.98 :t 0.17 (:tSE) (n = 305), and those for G. c)'pha were 9-10, average 9.03 :t 0.18 (n = 30). Anal rays were 8-10, average 9.12 :t 0.35, and 8-11, average 9.43 :t 0.63, respectively. o DISCUSSION Results of this project may be discussed at a variety of levels: recognition of organismal dif- ferences by humans and how it relates to the trained eye of the systematist; types of phenetic characters used in analysis and their resolving power; phenotypic differentiation in Gila and the intuitive processes involved with grouping these fishes; future directions with regard to Gila morphometrics; and the critical nature of population studies involving Gila from the Col- orado River. Our results demonstrate that technicians with no special training in systematics successfully discriminated a variety of qualitative characters that separate wild-caught Gila into distinct groups. In this instance, two such groups from the Yampa River corresponded to G. cypha and G. Tobusla. We do not propose on the basis of this preliminary work that qualitative charac- ters will prove universally applicable in identi- fication ofthese difficult taxa, but do argue that new approaches must be fostered to aid in solv- ing long-standing problems. Qualitative rank- ings are but one of many such approaches. Our intent is not to polarize qualitative versus quan- C 4.0