7987 - Laserfiche WebLink

Home Browse Search

Sheldon The Riverine Archipelago A higher order of faunal complexity is imposed on river systems by major watershed divides, the largest lowland rivers, and saltwater barriers. Some of these barriers have been breached by low sea levels during the Pleis- tocene and stream capture so rivers are not only insular but archipelagic. Interdrainage differentiation of fish faunas is shown in Figure 5. Faunal dissimilarity was calculated from data in Hocutt & Wiley (1986) and Stauffer et al. (1982) using Preston's (1962) resemblance equation. Preston's index z ranges from 0 (complete identity) to 1.0 (no species in common). Preston (1962) proposed that z = .27 implies "isolates in equilibrium" under his hypothesis of the canonical lognormal distribution of species abun- dances. f e . I t.........:.67 /-27- .24/ ............... .15'p"C .2~1 190 '\. h cio~.66 0 .31\i d .2o\i U 22y>V --ll-~ .2~~02 . 48 ,. k 0_0 t . 1'.16 .36 \ .Is/.m .66 o......-:..n 36 q · .27 ~._ .24/ S. r .53 .~p Figure 5. Faunal dissimilarity (Preston's z) of nearby drainages in the Mississippi (0) and Gulf Coast-Atlantic (e) regions. a Big Sandy b. Guyan- dotte c Kanawha (excluding New) d New e Monon- gahela f. Susquehanna g. Potomac h. James i. Roanoke j Tar k. Neuse l Cape Fear m. Pee Dee n Santee 0. Savannah p. Altamaha q. Apalachicola r. Mobile.\: Pearl t Tennessee u. White v. lower Cum- berland UJ. Cumberland above falls. Conservation of Stre8IlI Fishes 153 Neighboring rivers on the Atlantic coast and within the Mississippi system are differentiated to a degree that, especially in the coastal rivers, approaches that of "iso- lates in equilibrium." Gulf coast drainages are more strongly differentiated. (This region is also one in which intraspecific genetic similarities between drainages are low [Bermingham & Avise 1986].) The greatest dissim- ilarities are between drainages separated by the coastal- Mississippi divide (e.g., the Potomac and the Mononga- hela, the Mobile and the Tennessee). Kanawha Falls isolates the New River from the lower Kanawha to a comparable degree. The upper and lower Cumberland are less differentiated. (In this calculation I assumed that several species presently found in tributaries below the falls are upper Cumberland species transferred by head- ward erosion of the falls [Starnes & Etnier 1986]. Alter- natively, these species may be ecologically restricted to the Cumberland Plateau physiographic region.) A single trans-Mississippi comparison (the Tennessee and the White) shows the effect of isolation of the Ozark and Appalachian uplands by lowland and large river habitats. Faunal distinctions among rivers are being eroded by introductions from other continents (Courtenay & Stauffer 1984) and transfers within North America (Moyle et al. 1986). Stocking of game and forage fishes, aquarium releases, and bait-bucket introductions have all contributed to the breakdown of regional unique- ness. Proportionally, the effects of introductions are espe- cially noticeable in the depauperate faunas of western drainages. For example, the fauna of the Clark Fork in western Montana (Fig. 1) has been trebled by introduc- tions (Brown 1971), to the clear detriment of some native fishes such as the cutthroat trout Salmo clarki (Allendorf & Leary, this issue). Introductions are less obvious in the diverse faunas of the southeast. The New River, in which 43% of the fishes are nonnative (Hocutt et al. 1986), is a regional extreme, and Jenkins's (1987) careful analysis indicates from 12 to 32% of the species in 11 other Virginia drain- ages were introduced Most of these introductions oc- curred in the last 30 years and adverse effects on the native fishes are not apparent. However, diverse faunas are not immune to impacts from introduced species, and Lemley (1985) has documented reductions of na- tive fishes in North Carolina streams by introduced green sunfish, Lepomis cyanellus. Wholesale invasions become possible when drainages are linked by navigational or hydroelectrical projects. Catastrophic changes in the Great Lakes (Smith 1968) were triggered by invading species (lamprey, alewife) entering through navigation canals. Balon et al. (1986) predict reciprocal invasions when the Rhine-Danube ship canal is completed. In the United States, two (Mo- bile, Tennessee) of the richest (Fig. 1) and most distinc- Conservation Biology Volume 2, No. 2,June 1988