Northern Prairie Wildlife Research Center
|Figure 2. Diagram of the Arkansas darter distribution in the study area in south-central Kansas. Circles () represent historic collection localities. Closed circles () are sites where the Arkansas darter was collected during the 1994 through 1997 surveys. Open triangles () mark 37 sites sampled during the 1994 through 1997 studies where the Arkansas darter was absent beyond the southern periphery of the its known range. Some circles and triangles represent more than one sample location.|
We suggest that the 16 sites where both adult and young Arkansas darters were abundant (more than 50 individuals at each site) represent core populations. These larger populations were present throughout most of the study area. Individuals from core populations might periodically disperse to occupy depopulated or marginal habitats. Arkansas darters also might emigrate to adjacent core populations, thereby maintaining the larger metapopulation. Given the localized nature of the small spring-fed habitat preferred by the Arkansas darter, we suggest that there are additional core populations that have not been documented.
There are exceptions to all aspects of the description of the spring-fed habitat typically occupied by the Arkansas darter (Moss 1981), which suggest that the species is tolerant of at least short-term variations. During our survey, individuals collected from generally warmer water (greater than 29°C) were predominantly young fish, whereas mixed populations of adult and young darters were captured at groundwater-fed sites with cooler water temperatures (17 to 28°C). We hypothesize that Arkansas darters in the warmer habitat had emigrated from core populations within the stream, perhaps being forced out by population pressures (Taber et al. 1986) or scouring stream flows (Miller 1984).
The Arkansas darter also has been collected in the main stems of the Arkansas and Cimarron rivers (Cross 1967, Matthews and McDaniel 1981, Pigg et al. 1985, Ernsting and Eberle 1989). Its presence in these larger streams might be due, in part, to reduced flows in these rivers that create a habitat similar to the spring-run tributary typically occupied by the Arkansas darter (Cross et al. 1985). Cross et al. (1985) speculated that the Arkansas darter uses rivers as avenues of dispersal from one system of small tributary streams to another. Consistent with this hypothesis, Eberle and Ernsting (1989) reported the presence of four female Arkansas darters in the Arkansas River in central Kansas 25 km upstream from the mouth of Rattlesnake Creek, the nearest known source. This possible 25-km journey by the darters would be in addition to their movement downstream in Rattlesnake Creek, and might have included passage through the salt-marsh area of Quivira National Wildlife Refuge. However, the long distances from one tributary to another along the Arkansas River on the High Plains of western Kansas and eastern Colorado suggest that rivers might be effective dispersal barriers of suboptimal habitat (Miller 1984), especially with the falling water tables and declines in surface flow caused by groundwater mining, diversions of surface water, construction of impoundments, and other factors. Although tributaries with surface water are rare in this portion of the High Plains, Cross et al. (1985) suggested that groundwater seepage formerly maintained pools in low swales connected to the Arkansas River by shallow runs, similar to the "brook" at the type locality for the Arkansas darter (Gilbert 1885). These habitats could have facilitated the dispersal of this species throughout the High Plains portion of the Arkansas River basin. The use of rivers and their larger tributaries by the Arkansas darter for dispersal in south-central Kansas suggested by Cross et al. (1985) seems plausible to us given the relatively high density of small spring-fed streams in this area and the potential vagility of the species (Ernsting and Eberle 1989).
The future of the Arkansas darter is precarious in the extreme western portion of our study area, west of Crooked Creek (Fig. 1). In this region, groundwater is mined extensively from the underlying Ogallala aquifer (part of the High Plains Aquifer system), primarily for crop irrigation. Of the 420,872 ha in the two counties in this area, 24% (100,366 ha) were available for irrigation in 1984 (Kansas State Board of Agriculture [1985?]). Water rights in this region have been over-appropriated, and Groundwater Management District #3 follows a policy of controlled depletions (Kansas Water Office 1994). As the water table continues to fall in this area, it causes some formerly perennial streams to become ephemeral, and it eliminates the groundwater seepage that maintains the summer temperature of the surface water at a level appropriate for the Arkansas darter (generally 25°C or less). The largest populations of the Arkansas darter in this portion of our study area occurred in the main stem of the Cimarron River upstream from Crooked Creek, rather than small spring-fed creeks.
Arkansas darter populations in portions of the Ninnescah River Basin (Fig. 1) also might be at risk due to habitat loss as a result of groundwater withdrawals. This northern basin in our study area overlies the Great Bend Prairie Aquifer, an eastern extension of the High Plains Aquifer. From 1976 through 1984, the amount of land available for irrigation in four counties located in this basin more than doubled, from 29,988 ha to 64,388 ha (Kansas State Board of Agriculture [1977?], [1985?]). Of the 926,213 ha in these four counties, 7% (64,388 ha) was available for irrigation in 1984 (Kansas State Board of Agriculture [1985?]). Groundwater Management District #5, which includes the Ninnescah River Basin, has a policy of safe yield, in which recharge is intended to be approximately equal to withdrawals (Kansas Water Office 1995).
The south-central portion of our study area lies within the rough topography of the physiographic province known as the Red Hills. The topography and the limited groundwater supply with a high mineral content preclude extensive groundwater mining for irrigation in this region, which lies south and east of the more heavily irrigated areas overlying the High Plains Aquifer. Of the 966,359 ha in the four principal counties within this basin, only 0.8% (7,568 ha) was available for irrigation during 1984 (Kansas State Board of Agriculture [1985?]). The actual number of hectares and the relative proportion of hectares available for irrigation in the Red Hills are substantially lower than those in the other two parts of our study area. The potential threat in the south-central region is from local drawdowns of the water table near springs or from other small-scale perturbations, such as pollution from hog farms or other confined livestock operations, that might extirpate specific populations.
There also seems to be a natural constraint on the distribution of the Arkansas darter within the southern portion of the Red Hills in the lower Cimarron, Salt Fork Arkansas, and Medicine Lodge river systems along the Kansas-Oklahoma border (Fig. 2). The absence of the Arkansas darter in this area might be due to the naturally high concentrations of ions in these streams. The Arkansas darter apparently has some tolerance to high chloride levels (as sodium chloride; monovalent ions), based on the presence of a few individuals captured from the lower segment of Rattlesnake Creek (Ernsting and Eberle 1989, Eberle et al. 1996). However, calcium sulfate (divalent ions), leached from abundant gypsum deposits in the Red Hills, is largely responsible for the high conductivities measured in streams draining this region. The ions contributing to the higher conductivities in Rattlesnake Creek and the Red Hills streams (primarily monovalent ions versus divalent ions, respectively) might affect Arkansas darters differently. Perhaps the Arkansas darter can tolerate the high levels of sodium chloride in Rattlesnake Creek but not the high levels of calcium sulfate in the southern drainages of the Red Hills; however, this has not been tested. Thus, it is possible that elevated levels of calcium sulfate restrict the distribution of the Arkansas darter to the northern headwater portions of drainages in the Red Hills. Regardless of the contributing ions, we know of no large populations in streams with conductivities greater than 2700 µS/cm.
Given this general distribution of the Arkansas darter in the south-central portion of our study area, we were puzzled by its absence from the headwater reaches of Bluff Creek, a tributary of the Chikaskia River (Fig. 2). Cross (1967) mapped a single collection from the upper Bluff Creek drainage, but the record on which this was based is unknown. We sampled creeks in this drainage that had the cool water and aquatic vegetation typical of Arkansas darter habitat, but the only darter we collected was the orangethroat darter. The reason for the apparent absence of the Arkansas darter in the Bluff Creek drainage is uncertain given that there are records from the upper Chikaskia River and its tributaries; however, it is possible that the Arkansas darter was introduced recently into the Chikaskia River Basin.
The only confirmed records of the Arkansas darter from the upper portion of the Chikaskia River Basin in Kansas are based on unpublished records of collections made from 1978 through 1988. In 1992, Luttrell (1993) extended the known range of this species downstream in the Chikaskia River Basin in Kansas. The existing collection data suggest that the Arkansas darter might have been introduced into the upper Chikaskia River or its tributaries, perhaps from the nearby South Fork Ninnescah River Basin (Fig. 1), and it is now dispersing downstream. This could explain its apparent absence from the Bluff Creek drainage, if one discounts the single report (Cross 1967), because Bluff Creek does not flow into the Chikaskia River until both streams enter Oklahoma (Fig. 1). We are aware of no records of the Arkansas darter in the lower reach of the Chikaskia River from which it could disperse upstream through Bluff Creek (Fig. 2).
The possibility that the Arkansas darter is not native to the Chikaskia River and its tributaries in Kansas is supported circumstantially by differences between the fish communities in the Chikaskia River Basin and the nearby Medicine Lodge, Salt Fork Arkansas, and Ninnescah river basins. The Medicine Lodge and Chikaskia rivers are both tributaries of the Salt Fork Arkansas River, with confluences in Oklahoma (Fig. 1). The Ninnescah River is a tributary of the Arkansas River, and it is located entirely within Kansas (Fig. 1). The native fauna of the Medicine Lodge, Salt Fork Arkansas, and Ninnescah river basins includes (or included) peppered chub (Macrhybopsis tetranema; Eisenhour 1999), Arkansas River shiner (Notropis girardi), and plains minnow (Hybognathus placitus), which apparently were absent from the Chikaskia River Basin in Kansas (Moore and Buck 1953, Cross 1967). These three species of minnows, particularly the plains minnow, sometimes moved into the lower Chikaskia River from the Salt Fork Arkansas River in Oklahoma (Moore and Buck 1953). Based on their likely distributions in south-central Kansas prior to 1850, there seems to be an intuitive correlation between the presence of these three species of minnows in the warm, open water of the river main stems and the occurrence of the Arkansas darter in the cool, vegetated tributaries. All four species apparently were native to the Medicine Lodge, Salt Fork Arkansas, and Ninnescah river basins, and all four, except for the previously mentioned records of the Arkansas darter, were absent from the Chikaskia River Basin in Kansas (Cross 1967).
Alternatively, the presence of the Arkansas darter in only the upper portion of the Chikaskia River Basin might be indicative of the general southern limit to its range (Fig. 2). The relatively recent range extensions in the Chikaskia River Basin might reflect changes in stream habitats similar to those in the upper Cimarron River that have allowed the Arkansas darter to more readily move into the main stem of the river (Cross et al. 1985, Pigg et al. 1985). The relatively isolated population in Bluff Creek might now be extirpated, or we might have failed to capture individuals from a sparsely populated stream reach. The missing elements from the local ichthyofauna in the upper portion of the Chikaskia River Basin also might be an artifact of limited faunal surveys, as summarized by Cross (1967) and Moore and Buck (1953). Additional surveys in the Chikaskia River Basin to monitor the distribution and potential dispersal of the Arkansas darter in this drainage might help to answer this question. A thorough study of water quality attributes, including quantitative data on the various ions, in streams throughout our study area also would be useful.
The overall condition of Arkansas darter populations in south-central Kansas seems to be reasonably stable, but threats remain. As land-use patterns continue to change in this region, steps to regularly monitor the status of the Arkansas darter should be implemented. The area of greatest concern is in the Ninnescah River Basin, where large-scale groundwater declines are possible. If these declines occur, the result is likely to be similar to the situation in the western portion of our study area, where the Arkansas darter is on the verge of regional extirpation. Groundwater appropriations for agricultural, industrial, or municipal use in Ninnescan River Basin should be monitored by appropriate wildlife agencies so that they can be in a position to take effective preventive actions. Less dramatic, but equally important, is the possibility that local extirpations could occur throughout the range of the Arkansas darter. In this event, we currently have no knowledge of the genetic attributes of the various populations. Yet this information is essential to the proper management of the Arkansas darter. This information could be used to protect unique genetic components of the species and to determine appropriate sources of individuals used to replace extirpated populations.