Chapter II: Part 2

The habitat preferences of this species and of the related species _N. volucellus_ have been described as follows by Hubbs and Ortenburger (1929b:68): "It seems probable that _volucellus_ when occurring in the range of _buchanani_ occupies upland streams, whereas _buchanani_ is chiefly a form of the large rivers and adjacent creek mouths." The results of this survey and impressions gained from other collections, some of which are unpublished, are in agreement with this view. A collection on the Verdigris River at Independence, Kansas, directly downstream from the mouth of the Elk River, showed _N. buchanani_ to be common while _N. volucellus_ was not taken. At station E-5 upstream from E-4, however, _N. volucellus_ was taken but _N. buchanani_ was not found.

In the upper Neosho basin, Cross (1954a:310) took _N. volucellus_ but not _N. buchanani_. Other collections have shown _N. buchanani_ to be abundant in the lower Neosho River in Kansas. Moore and Paden (1950:85) observe that _N. buchanani_ was found only near the mouth of the Illinois River in Oklahoma and was sharply segregated ecologically from _N. volucellus_ that occupied a niche in the clear main channels in contrast to the more sluggish waters inhabited by _N. buchanani_.

#Notropis camurus# (Jordan and Meek): Stations C-3, C-4, C-5, C-6, C-7, C-8, C-9, C-10, C-11, C-12, C-13, E-1, E-5 (C-131).

Highest concentrations of the bluntface shiner were found close to the mouths of two tributaries of Big Caney River: Rock Creek and Otter Creek. On Rock Creek (Station C-4) this shiner was abundant in a shallow pool below a riffle where water was flowing rapidly. Many large males in breeding condition were taken (June 3). The species formed 20.2 per cent of the fish taken.

On Otter Creek (Station C-13) the species was common in shallow bedrock pools below riffles. It formed 12.1 per cent of the fish taken.

At station C-5, _N. camurus_ was characteristically found in an area of shallow pools and riffles. At station C-10 it was found in clear flowing water over rubble bottom and in small coves over mud bottom. At C-11 (July 26) _N. camurus_ was taken only in one small pool with rapidly flowing water below a riffle. In this pool _N. camurus_ was the dominant fish. At station C-12, on April 2, _N. camurus_ was abundant in the stream, which was then clear and flowing. On August 24, it was not taken from the same pool, which was then turbid and drying.

The frequent occurrence of this species in clear, flowing water seems significant. Cross (1954a:309) notes that the bluntface shiner prefers moderately fast, clear water. Hall (1952:57) found _N. camurus_ only in upland tributaries east of Grand River and not in lowland tributaries west of the river. Moore and Buck (1953:22) took this species in the Chikaskia River, which was at that time a clear, flowing stream. They noted that in Oklahoma it seems to be found only in relatively clear water.

_N. camurus_ did not seem to ascend the smaller tributaries of Big Caney River as did _N. rubellus_ and _N. boops_ even when these tributaries were flowing.

#Notropis deliciosus missuriensis# (Cope): Stations A-1, A-2, A-3, W-1, W-2, W-3 (C-136).

Sand shiners seemed to be abundant in the Arkansas River, rare in the Walnut River and absent from other streams surveyed. This shiner was most abundant in shallow, flowing water in the Arkansas River; in backwaters, where _Gambusia affinis_ prevailed, _N. deliciosus_ formed only a small percentage of the fish population.

#Notropis girardi# Hubbs and Ortenburger: Stations A-2 and A-3.

At station A-2 the Arkansas River shiner made up 14.6 per cent of all fish taken. At A-2, it was found only in rapidly-flowing water over clean sand in the main channels. It was absent from the shallow, slowly-flowing water where _N. deliciosus missuriensis_ was abundant. At A-3 _N. girardi_ made up 22 per cent of the total catch, and again preferred the deeper, faster water over clean-swept sand. Failure to find _N. girardi_ at station A-1 is not understood.

Females were gravid in both collections (August 25 and 27). In neither collection were young-of-the-year taken. Moore (1944:210) has suggested that _N. girardi_ requires periods of high water and turbidity to spawn. Additional collecting was done at station A-3 on December 22, 1957. A few adults were taken in flowing water but no young were found.

In this area, _N. girardi_ showed no tendency to ascend tributaries of the Arkansas River. Not far to the west, however, this pattern changes as shown by Hubbs and Ortenburger (1929a:32) who took this fish at seven of ten stations on the Cimarron, Canadian, and Salt Fork of the Arkansas. _N. girardi_ was taken only in the lowermost stations on both Stillwater Creek (Cross, 1950:136) and the Chikaskia River (Moore and Buck, 1953:22). In the next major stream west of the Chikaskia, the Medicine River, _N. girardi_ seems to occur farther upstream than in the Chikaskia. (Collection C-5-51 by Dr. A. B. Leonard and Dr. Frank B. Cross on Elm Creek near Medicine Lodge on July 20, 1951.)

#Notropis lutrensis# (Baird and Girard): Stations A-1, A-2, W-1, W-2, W-3, W-4, G-1, G-2, G-4, G-5, G-8, G-9, G-10, G-11, G-12, G-13, G-14, G-15, G-16, B-1, B-2, B-3, C-1, C-2, C-3, C-4, C-5, C-6, C-9, C-10, C-11, C-12, C-13, C-14, M-1, E-1, E-2, E-4, E-7 (E&F, C-131, C-133, C-136).

The red shiner was taken in every stream surveyed. The relative abundance seemed to be greatest in two types of habitat which were separated geographically. The first habitat was in large rivers such as the Arkansas and Walnut. In the Arkansas River the red shiner consistently made up 20 per cent to 25 per cent of the catch. On the Walnut River percentages ranged from 10 per cent (station W-3) to 45 per cent (station W-2).

The second habitat in which numbers of _N. lutrensis_ reached high proportions was in the upper parts of the most intermittent tributaries. At the uppermost station in Silver Creek this species formed 30 per cent of the fish taken. In Crab Creek the following percentages were taken in six collections from mouth to source: 20.6%, 26.1%, 25%, 85%, 14.6%, and 1%. In the mainstream of Grouse Creek the highest percentage taken was 19.27 near the mouth at station G-1. In middle sections of Grouse Creek this species was either absent or made up less than 2 per cent of the fish taken.

At no station on Big Caney River was the red shiner abundant. The smallest relative numbers were found at upstream stations, in contrast to collections made on tributaries of Grouse Creek. This distributional pattern possibly may be explained by the severe conditions under which fish have been forced to live in the upper tributaries of Grouse Creek. Water was more turbid, and pools were smaller than in Big Caney. These factors possibly decimate numbers of the less hardy species permitting expansion by more adaptable species, among which seems to be _N. lutrensis_. In the upper tributaries of Big Caney River conditions have not been so severe due to greater flow from springs and less cultivation of the watershed in most places. Under such conditions _N. lutrensis_ seems to remain a minor faunal constituent.

#Notropis percobromus# (Cope): Stations A-1, A-2, W-1, W-2, W-3, G-1.

At station W-1 the plains shiner constituted 20 per cent of the fish taken. The river was flowing rapidly with large volume at the time of this collection, and all specimens were taken near the bank in comparatively quiet water over gravel bottom. At station W-3, below Tunnel Mill Dam at Winfield, _N. percobromus_ comprised 18.7 per cent of the fish taken, second only to _Lepomis humilis_ in relative abundance. Immediately below the west end of the dam, plains shiners were so concentrated that fifty or more were taken in one haul of a four-foot nylon net. The amount of water overflowing the dam at this point was slight. Water was shallow (8-12 inches) and the bottom consisted of the pitted apron or of fine gravel. At the east end of the dam where water was deeper (1-3 feet) and the flow over the dam greater, large numbers of _Lepomis humilis_ were taken while _N. percobromus_ was rare.

In the Arkansas River smaller relative numbers of this shiner were obtained. At station A-2, it formed 4.68 per cent of the total. At this station _N. percobromus_ was taken with _N. lutrensis_ in water about 18 inches deep next to a bank where the current was sluggish and tangled roots and detritus offered some shelter.

At station G-1 on Grouse Creek the plains shiner made up 7.68 per cent of the fish taken. The habitat consisted of intermittent pools with rubble bottoms at this station, which was four miles upstream from the mouth of the creek. The plains shiner seems rarely to ascend the upland streams of the area.

#Notropis rubellus# (Agassiz): Stations C-3, C-5, C-6, C-7, C-8, C-10, C-11, C-12, C-13, C-14 (J&J).

No fish in these collections showed a more persistent preference than _Notropis rubellus_ for clear, cool streams. All collections of the rosyface shiner were in the Big Caney River system, but at only four stations in this system was it common. At station C-11 the highest relative numbers (10.6 per cent) were obtained. This site possessed the most limpid water of any station on the mainstream of Big Caney. Aquatic plants (_Myriophyllum heterophyllum_ and _Potamogeton nodosus_) were common. Other fishes that flourished at this station were _N. boops_, _N. camurus_, _Campostoma anomalum_, and _Etheostoma spectabile_. The water temperature was 86 deg. at surface and 80 deg. at bottom whereas air temperature was 97 deg..

_N. rubellus_ was common at all stations in Otter Creek, the clear, upland character of which has been discussed. In May and June only adults were found. On September 1, examination of several pools in upper Otter Creek revealed numerous young-of-the-year in small spring-fed pools.

Literature is scarce concerning this shiner in Kansas. Cross (1954a:308) stated that it was abundant in the South Fork of the Cottonwood River and was one of those fishes primarily associated with the Ozarkian fauna, rather than with the fauna of the plains. Elliott (1947) found _N. rubellus_ in Spring Creek, a tributary of Fall River which seems similar to Otter Creek in physical features. Between the Fall River and Big Caney River systems is the Elk River, from which there is no record of the rosyface shiner. Perhaps its absence is related to the intermittent condition of this stream at present. The Elk River is poor in spring-fed tributaries, which seem to be favorite environs of the rosyface shiner.

_N. rubellus_ was taken by Minna Jewell and Frank Jobes in Silver Creek on June 30, 1925 (UMMZ 67818). The shiner was not found in any stream west of the Big Caney system in my collections.

In Oklahoma, Hall (1952:57) found _N. rubellus_ in upland tributaries on the east side of Grand River and not in the lowland tributaries on the west side. Martin and Campbell (1953:51) characterize _N. rubellus_ as preferring riffle channels in moderate to fast current in the Black River, Missouri. It is the only species so characterized by them which was taken in my collections. Moore and Paden (1950:84) state "_Notropis rubellus_ is one of the most abundant fishes of the Illinois River, being found in all habitats but showing a distinct preference for fast water...."

#Notropis topeka# (Gilbert): Two specimens (formerly Indiana University 4605) of the Topeka shiner labeled "Winfield, Kansas" are now at the University of Michigan Museum of Zoology. Collector and other data are not given. Evermann and Fordice (1886:185) noted that two specimens of _N. topeka_ were taken from Sand Creek near Newton in Harvey County, but do not list it from Cowley County near Winfield. They deposited their fish in the museum of Indiana University.

#Notropis umbratilis# (Girard): Stations G-1, G-3, G-4, G-7, G-8, G-9, G-12, G-14, B-2, B-3, C-1, C-2, C-3, C-4, C-5, C-6, C-7, C-8, C-9, C-10, C-11, C-12, C-13, C-14, C-15, C-16, E-1, E-2, E-4, E-5, M-1, M-2 (J&J, C-131, C-132).

The redfin shiner flourished in all the streams surveyed except the Arkansas and Walnut Rivers. _N. umbratilis_ has been found in upland tributaries of the Walnut River, some of which originate in terrain similar to that in which Elk River, Big Caney River, and Grouse Creek originate. (Collection C-26-51 by Cross on Durechon Creek, October 7, 1951.) This suggests downstream reduction in relative numbers of this species, a tendency which also seemed to exist on both Big Caney River and Grouse Creek. _N. umbratilis_ was the most abundant species in Big Caney River except at the lowermost stations where it was surpassed in relative abundance by _N. lutrensis_ and _Gambusia affinis_.

_N. umbratilis_ was a pool-dweller, becoming more concentrated in the deeper pools as summer advanced. In May and early June, large concentrations of adult _N. umbratilis_ were common in the shallow ends of pools together with _N. rubellus_, _N. boops_, _Pimephales notatus_, and _Pimephales tenellus_. By July and August, only young of the year were taken in shallow water, and adults were scarcely in evidence.

#Notropis volucellus# (Cope): Stations G-5, G-8, C-3, C-5, C-7, C-8, C-9, C-10, M-1, E-4, E-5.

The mimic shiner was a minor element in the fauna, 2.02 per cent at station C-5 being the largest percentage taken. In the Big Caney River system _N. volucellus_ was taken only in the main stream. In the Grouse Creek drainage it was found at two stations in the upper part of the watershed, where water is clearer, gradient greater, and pools well-shaded and cool.

In the Elk River the mimic shiner was taken only in the upper part of the main stream. The dominant shiner in situations where _N. volucellus_ was taken was, in all cases, _N. umbratilis_. Elliott (1947) found _N. volucellus_ in Spring Creek, a tributary of Fall River. Farther north in the Flint Hills region, _N. volucellus_ was reported by Cross (1954a:310).

#Notemigonus crysoleucas# (Mitchell): Station W-5.

This isolated record for the golden shiner consisted of nine specimens collected on June 6 in Timber Creek, a tributary of the Walnut River. Most of the creek was dry. _N. crysoleucas_ was taken in one pool with dimensions of 8 feet by 4 feet with an average depth of 4 inches. This creek is sluggish and silt-laden, even under conditions of favorable precipitation. Hubbs and Ortenburger (1929b:89) observed that the golden shiner prefers sluggish water. Hall (1952:58) took the golden shiner only in the lowland tributaries west of Grand River and not east of the river in upland tributaries.

#Phenacobius mirabilis# Girard: Stations W-3, C-3.

In no case was the suckermouth minnow common; it never comprised more than 1 per cent of the fish population.

#Pimephales notatus# (Rafinesque): Stations W-4, G-5, G-7, G-9, G-12, G-13, B-3, C-1, C-2, C-3, C-4, C-5, C-6, C-7, C-8, C-9, C-10, C-11, C-12, C-13, C-14, C-15, C-16, C-17, C-18, M-1, M-2, E-1, E-2, E-4, E-5, E-7 (J&J, C-131, C-132, C-133).

This was much the most abundant of the four species of _Pimephales_ in this area. It was taken at 33 stations as compared with 10 for _P. tenellus_, 8 for _P. promelas_, and 3 for _P. vigilax_.

The bluntnose minnow was taken almost everywhere except in the main stream of the Arkansas and Walnut rivers and in lower Grouse Creek. _P. notatus_ seemed to prefer clearer streams of the Flint Hills part of my area. There was a marked increase in percentages taken in the upland tributaries of both Caney River and Grouse Creek. In the Elk River, too, higher concentrations were found upstream.

The highest relative numbers of bluntnose minnows were taken at station G-12 on Crab Creek, station C-12 on Cedar Creek and station C-16 on Spring Creek. At G-12, this minnow was abundant in the deeper isolated pools. Males in breeding condition were taken on June 9. In Cedar Creek the population of bluntnose minnows was observed periodically in one pool in which they were dominant. This pool was 100 feet by 50 feet, shallow, and with bedrock bottom. At its upper end, however, there was a small area of heavily-shaded deeper water. Throughout the spring bluntnose minnows were found in large schools in the shallow area. As the summer progressed they were no longer there, but seining revealed their presence in the deeper, upper end.

At station C-16 on Spring Creek on July 9 male _P. notatus_ were taken in extreme breeding condition, being light brick-red in color and with large tubercles.

#Pimephales tenellus# (Girard): Stations G-1, C-2, C-3, C-5, C-6, C-7, C-8, M-1, E-2, E-4 (C-131 C-133).

The mountain minnow was never taken far from the mainstream of Big Caney, Middle Caney, or Elk River. In this respect it differed from _P. notatus_, which reached large concentrations in the small upland tributaries. On the other hand, _P. tenellus_ was not so abundant as _P. vigilax_ in the silty larger streams. In no collection was the mountain minnow common. The highest percentages were 2.4 per cent (Station C-5), and 2.1 per cent (Station C-7) on Big Caney River. These stations consisted of clear, flowing water over rubble bottoms. Males at C-7 (June 16) were in breeding condition.

Moore and Buck (1953:23) reported finding this species among rocks in very fast water rather than in the quiet backwaters frequented by _P. vigilax_. Other records of the mountain minnow from the Flint Hills indicate that it seeks areas of maximum gradient and flow; in this distributional respect it is like _Notropis camurus_. The two species are recorded together from other streams in this region such as the Chikaskia (Moore and Buck, 1953:23), Cottonwood (Cross, 1954a:310), and Spring Creek, tributary of Fall River (Elliott, 1947). It is conceivable that a preference for flowing water might explain its restriction to the medium-sized, less intermittent streams in this area. The only tributary which the species seemed to ascend to any extent was Otter Creek, which is seldom intermittent downstream.

#Pimephales vigilax perspicuus# (Girard): Stations A-3, C-1, C-4.

The parrot minnow was found only in downstream habitats. Collection C-4 (June 3) on Rock Creek was made about 1/2 mile from the mouth of this tributary of Big Caney and the creek here had almost the same character as the river proper. The presence of other channel fishes such as _Ictiobus bubalus_ indicates the downstream nature of the creek. Some males of _P. vigilax_ in breeding condition were taken in this collection.

At C-1, only one specimen was found in a turbid, isolated pool with bedrock bottom. At A-1 only one parrot minnow was taken; it was in deep, fairly quiet water near the bank.

Other collections outside the three-county area revealed the following: In the Neosho River, several parrot minnows were found in quiet backwaters and in shallow pools. In the Verdigris River three were taken directly under water spilling over the dam at this station, while others were found, together with _P. promelas_, in the mouth of a small creek that provided a backwater habitat with mud bottom.

Cross and Moore (1952:405) found this species only at stations in the lower portion of the Poteau River. Farther west the minnow may ascend the smaller sandy streams to greater distances. Moore and Buck (1953:23) took parrot minnows at six of 15 stations on the Chikaskia River and found the species as far upstream as Drury, Kansas. Elliott (1947), in comparing the South Ninnescah and Spring Creek fish faunas, found only _P. vigilax_ and _P. promelas_ on the sandy, "flatter" Ninnescah and only _P. notatus_ and _P. tenellus_ on Spring Creek, an upland, Flint Hills stream in Greenwood County.

#Pimephales promelas# Rafinesque: Stations A-2, A-3, W-3, W-4, G-9, B-1, M-1, E-4 (E&F, C-136).

Occurrences of the fathead minnow were scattered, but included all streams sampled except Big Caney.

Three of the collections were in small intermittent streams where conditions were generally unfavorable for fishes and in one instance extremely foul. Two of these stations had turbid water and all suffered from siltation.

In Middle Caney Creek the species was rare but in the Elk River (June 28) more than 100 specimens, predominantly young, were taken. This station consisted of a large isolated pool with a variety of bottom types. Water was turbid and the surface temperature was high (93 deg. F.). In different parts of the pool the following numbers of specimens were taken in single seine-hauls: 15 over shallow bedrock; 35 over gravel (1-1/2 feet deep); 50 over mud bottom (1 foot deep).

_P. promelas_ was found also in the large, flowing rivers: Arkansas, Walnut, Verdigris, and Neosho. The species was scarce in the Arkansas River, and was found principally in muddy coves. In the Walnut (W-3), this minnow comprised 7.65 per cent of the fish taken and was common in quiet pools.

#Campostoma anomalum# Rafinesque: Stations W-4, G-4, C-1, C-3, C-5, C-6, C-7, C-8, C-9, C-10, C-11, C-12, C-13, C-14, C-15, C-16, C-17, C-18, B-3 (E&F, C-131, C-136).

Although the stoneroller was found in most streams surveyed, it was taken most often in the Big Caney system, where it occurred at 16 of the 18 stations. In contrast, it was represented at only one of 17 stations on Grouse Creek. High percentages were found in three creeks--Cedar, Otter, and Spring. As noted above, these streams are normally clear, swift and have steep gradients and many rubble and gravel riffles. On these riffles young stonerollers abounded. Station C-16 on Spring Creek typifies the habitat in which this species was most abundant. The stream has an average width of 10 feet and depth of a few inches. The volume of flow was less than 1 cubic foot per second but turbulence was great. Water was clear and the bottom was gravel and rubble. Following rains in June, stonerollers quickly occupied parts of Spring Creek (upstream from C-16) that had been dry throughout the previous winter.

On April 2 many _C. anomalum_ and _Etheostoma spectabile_ were taken in shallow pools and riffles in an extensive bedrock-riffle area on Cedar Creek near station C-12. Most of the females were gravid and the males were in breeding condition. On June 6 these pools were revisited. Flow had ceased and the pools were drying up. Young-of-the-year of the two species were abundant, but only a few mature stonerollers were taken. On August 24, prolonged drought had drastically altered the stream and all areas from which stonerollers and darters had been taken were dry. Seining of other pools which were almost dry revealed no stonerollers.

Collections on May 31, June 15, and June 16 in Otter Creek revealed large numbers of stonerollers. They were found in riffle areas, in aquatic vegetation, and especially in detritus alongside banks. Most of the specimens were young-of-the-year.

#Anguilla bostoniensis# (Le Sueur): An American eel was caught by me in Grouse Creek in 1949.

#Gambusia affinis# (Baird and Girard): Stations A-1, A-2, A-3, W-1, W-2, W-3, W-4, W-5, G-1, G-2, G-3, G-4, G-5, G-7, G-8, G-9, C-1, C-2, C-3, C-4, C-6, C-15, E-1.

Mosquitofish occurred widely but in varied abundance. Huge populations were in the shallow sandy backwaters and cut-off pools of the Arkansas River. In the shallow pools of several intermittent streams such as station G-8 on Silver Creek this fish also flourished.

_G. affinis_ was taken at every station in the Arkansas, Walnut and Grouse systems except those stations on two upland tributaries of Grouse Creek (Crab Creek and Grand Summit Creek). The mosquitofish was not observed in the clear upland tributaries of Big Caney, nor on upper Big Caney River itself in May, June, and July. On September 3, however, _Gambusia_ were taken at station C-15 on Otter Creek and others were seen at station C-14 on the same date.

Hubbs and Ortenburger (1929b:99) and Cross and Moore (1952:407) observed that _G. affinis_ usually was absent from small upland tributaries, even though it was abundant in lower parts of the same river systems.

#Fundulus kansae# (Garman): Stations A-2, A-3, Evermann and Fordice as _Fundulus zebrinus_.

At station A-2, seven plains killifish were taken together with a great many _Notropis deliciosus_ and _Gambusia affinis_ in a shallow, algae-covered channel with slight flow and sand bottom. At station A-3 many young killifish were taken in small shallow pools on December 22. _Fundulus kansae_ has been found in the lower part of the Walnut River Basin, especially where petroleum pollution was evident. Eastward from the Walnut River plains killifish have not been taken.

#Fundulus notatus# (Rafinesque): Stations B-1, G-1, G-2, G-3, G-4, G-5, G-7, G-8, G-10, G-11, G-14, C-1, M-1, E-1, Evermann and Fordice as _Zygonectes notatus_.

The black-banded topminnow was not taken in the Arkansas River but was common in the Walnut and Grouse systems. It was common also in Middle Caney, but in Big Caney and Elk River it was taken only at the lowermost stations.

This species did not seem to ascend far into smaller tributaries of Grouse Creek. In Crab Creek it was taken at the lower two of six stations and in Grand Summit Creek at the lower of two stations.

The highest relative numbers were taken at stations G-3 (17.5 per cent), G-4 (24 per cent), G-10 (25.75 per cent) and G-11 (41.52 per cent), on Crab Creek and Grouse Creek. Both upstream and downstream from these stations, which were within five miles of each other, the relative abundance dropped off sharply. The bottoms at these stations were mostly rubble and mud, and water was turbid at three of the stations. At G-10 (June 24) and G-11 (July 16) young-of-the-year were abundant.

#Ictalurus melas# (Girard): Stations W-2, W-3, W-4, W-5, B-1, B-2, B-3, G-1, G-2, G-3, G-4, G-5, G-8, G-9, G-11, G-12, G-13, G-14, G-17, C-1, C-9, C-11, C-12, C-14, C-15, C-17, C-18, E-1, E-2, E-4, E-5, E-6, N-1, Evermann and Fordice as _Ameiurus melas_ (C-133).

The black bullhead was taken at slightly more than half of the stations, and probably was present at others. Larger numbers were taken in Grouse Creek than in any other stream system. In many small, shallow pools in the Grouse Creek system young black bullheads shared dominance with _Gambusia affinis_ in the late summer. _I. melas_ was also abundant in isolated pools at the extreme upper ends of Crab Creek, Beaver Creek and Grand Summit Creek. _I. melas_ was most common in areas with silty bottoms. The species seemed scarce in the main stream of Big Caney River but was common in some of its tributaries.

#Pylodictis olivaris# (Rafinesque): Stations A-3, G-1, C-5.

Flathead catfish were taken by angling at stations A-3 and C-5. At station G-1 (September 5) a flathead catfish five inches long was taken in the four-foot nylon net.

#Ictalurus punctatus# (Rafinesque): Stations A-3, W-2, W-3, G-2, C-5, E-4.

Channel catfish from stations W-3, A-3, and C-5 were taken on hook and line. At station G-2 (August 29) twenty young-of-the-year were seined from the shallow narrow end of a large pool. All collections of both _I. punctatus_ and _P. olivaris_ were in the larger streams surveyed.

#Ictalurus natalis# (LeSueur): Stations G-1, G-2, G-3, G-4, G-8, G-11, C-12, C-14, C-15.

The yellow bullhead was taken at only 9 stations, compared with 33 stations for the black bullhead. _I. natalis_ was represented in 7 of 17 stations in the Grouse Creek system but in only 3 of 18 stations in the Big Caney system. Of the seven records from Grouse Creek four were from the main stream. At every station where yellow bullheads were taken, black bullheads were found also and were abundant, usually several times more abundant than _I. natalis_.

At G-11 on Crab Creek (July 16), _I. natalis_ made up 3.8 per cent of the fish taken. All were young-of-the-year, existing in a tiny, gravelly pool containing not more than five gallons of water, and were the only fish present. Young yellow bullheads were also found in small pools with gravel bottoms at station G-4 on September 7.

#Labidesthes sicculus# (Cope): Stations G-1, G-2, G-3, G-7, G-10, B-2, C-1, C-2, C-3, C-4, C-5, C-12, E-1, E-2, E-3, E-7, M-1 (E&F, C-131).

The brook silversides was taken, sometimes abundantly, in all stream systems except the Walnut and Arkansas. At station G-7 on July 8, 41.8 per cent of the fish taken were of this species. _L. sicculus_ was most abundant in large pools where the bottom was predominantly bedrock and gravel. The highest concentrations were in the mainstreams of Big Caney, Grouse, and Elk Rivers. Brook silversides were taken rarely in the smaller tributaries of these streams.

#Percina phoxocephala# (Nelson): Stations C-2, C-3, C-5, G-1 (C-133).

Slenderhead darters were scarce, and were found only over gravel bottoms. Specimens were taken from flowing and quiet water, and from both shallow and deep water.

Larger numbers of _P. phoxocephala_ were taken by the writer in other collections made during 1956 on the Neosho and Verdigris Rivers over bottoms of rubble or gravel. Restriction of this darter to the larger streams follows a pattern observed by Cross (1954a:313) who noted it was absent from smaller riffles in minor tributaries. Elliott (1947), however, took one specimen of _P. phoxocephala_ in Spring Creek, a tributary of Fall River.

#Percina caprodes carbonaria# (Baird and Girard): Stations G-3, G-4, G-7, G-12, C-5, C-6, C-7, C-9, C-12, C-13, C-14 (J&J, C-131, C-133).

The logperch was generally distributed in the Caney, Elk, and Grouse systems. This species usually comprised less than 1 per cent of the fish taken; however, at station G-12 it formed 3.76 per cent of the total.

In many instances the logperch was taken over submerged gravel bars, often along the edges of the larger pools. At 8 of 13 stations where the logperch was taken, the golden redhorse was also found. At every station where logperch were found, _Notropis umbratilis_ was taken and _Pimephales notatus_ also occurred at all but three of these stations.

#Percina copelandi# (Jordan): Stations C-4, C-5, C-6, C-8, G-1 (C-131, C-133, J&J).

Channel darters were collected over bottoms of rubble or gravel, both in flowing streams and in isolated pools. Although _P. copelandi_ was found only in Big Caney River and at the lowermost station on Grouse Creek (G-1) in this survey, this species has been taken previously from Elk River (K. U. 3463 and K. U. 3197) and from Silver Creek. _Notropis camurus_ occurred everywhere that _H. copelandi_ was found. In several instances the two species were taken in the same seine-haul.

#Etheostoma spectabile pulchellum# (Girard): Stations W-4, G-1, G-4, G-5, C-6, C-9, C-11, C-12, C-13, C-14, C-15, C-16, C-17, C-18, E-1, E-5. Evermann and Fordice as _Etheostoma coeruleum_ (C-131, C-132).

The habitat preferences of the orangethroat darter seemed similar to those of _Campostoma anomalum_. There were sixteen stations at which both species were taken, seven where only _E. spectabile pulchellum_ was taken and six where only _C. anomalum_ was taken. The largest relative numbers of both species were found in the same small, clear upland tributaries of Big Caney River. On May 31, collections from riffles at station C-15 (upper Otter Creek) consisted almost entirely of these two species. On September 1 at this station the stream was intermittent, but even the tiniest pools abounded with young darters and stonerollers.

Gravid females and males in breeding condition were taken in riffles in Cedar Creek on April 2. During June numerous young and adult orangethroat darters were taken in Cedar Creek, in partly decayed leaves which lined the banks. On June 15 in Otter Creek young darters were abundant in streamside detritus and in clear, shallow, rubble riffles. At station C-11 a few darters were taken on rubble riffles; however, large numbers were found inhabiting thick mats of _Potamogeton foliosus_ Raf., which grew in shallow water. Many darters (_Etheostoma spectabile pulchellum_ and _Percina phoxocephala_) were taken in September along gravelly banks at stations C-2 and C-3 by disturbing small rocks and leaf-litter along the shores. Young orangethroat darters seemed to seek out sheltered areas and in some cases were found in sluggish, even foul, water (Stations W-4, B-1 and G-12). Moore and Buck (1953:26) note that the orangethroat darter is able to thrive in Oklahoma in rather sluggish and even intermittent waters which reach quite high summer temperatures.

Unlike other darters taken in this survey, the orangethroat darter was common to abundant at several stations and was found at a great many more stations than any other darter. The comparatively great tolerance of this species to varying habitats, suggested by this survey, is also reflected by its widespread distribution in Kansas.

#Micropterus salmoides salmoides# (Lacepede): Stations B-1, G-4, G-5, G-7, G-12, C-1, C-3, E-1, E-2, E-3.

Most of the largemouth bass taken were young-of-the-year. In Big Caney River this species seemed rare, being found at only two downstream stations compared with eight stations at which _M. punctulatus_ was taken.

Many ponds in the Flint Hills have been stocked with largemouth bass. At present largemouth bass are frequently caught by hook and line in Crab Creek (Station G-12); however, Mr. A. C. Metcalf, who has fished this stream for approximately 45 years, states that he took no bass in the creek prior to the building and stocking of large ponds on nearby ranches.

#Micropterus punctulatus# (Rafinesque): Stations C-4, C-5, C-6, C-7, C-8, C-10, C-14, C-15, E-2, E-5 (C-133).

The spotted bass was taken only in tributaries of the Verdigris River, where it seemed more numerous than the preceding species. It has been reported from other Verdigris tributaries such as Fall River (Elliott, 1947) and is common eastward from the Verdigris Basin. A spotted bass (K. U. 3467) was taken by Cross on the Little Walnut River in Butler County on April 5, 1955. This seems to be the only record of this species from the Walnut River Basin at the present time.

#Pomoxis annularis# (Rafinesque): Stations W-3, W-5, G-1, G-2, G-5, G-10, G-11, G-12, C-1, C-2, C-4, C-5, C-6, M-1, E-1, E-2, E-4, E-5 (C-136).

White crappie were found in almost all habitats and were taken in all rivers except the Arkansas. The relative abundance of this species was greater at downstream than at upstream stations on Grouse Creek, Big Caney, and Elk River. Schools of young crappie were frequently found and the factor of chance in taking or failing to take a school of crappie prevented confident appraisal of abundance. White crappie usually sought quiet waters. Often they were found in backwaters and many times schools were taken over bottoms where mud and detritus had been deposited. It was not uncommon to take _Pomoxis annularis_ and _Ictalurus melas_ in the same seine-haul in such areas.

#Pomoxis nigromaculatus# (LeSueur): Station C-1.

Black crappie were taken in Otter Creek on May 29 and September 3. Several ponds in eastern Cowley County are stocked with black crappie, but none was taken from streams into which these ponds drain.

#Lepomis cyanellus# (Rafinesque): Stations W-3, W-4, W-5, B-1, B-2, B-3, G-1, G-2, G-3, G-4, G-5, G-8, G-9, G-10, G-12, G-13, G-14, G-15, G-16, G-17, C-1, C-2, C-3, C-4, C-5, C-6, C-7, C-8, C-9, C-10, C-11, C-12, C-13, C-14, C-16, C-17, C-18, E-1, E-4, E-5, E-6, E-7, M-2 (C-131, C-132, C-133, C-136, E&F).

The green sunfish was taken at 45 of 60 stations, which is the greatest number recorded for one species. The only stream from which it was not obtained was the Arkansas River. Green sunfish constituted a minor but consistent part of the fauna in Big Caney River except for some intermittent pools on small tributaries, where it was high in relative abundance. It usually comprised approximately 4 per cent of the fish taken at stations on Grouse Creek. In some intermittent tributaries of Grouse Creek and Elk River percentages also were high.

Funk and Campbell (1953:74) observed that _L. cyanellus_ held a definite but minor place in all collections made on the Black River in Missouri. This pattern was also observed by the writer in collections made on the Neosho and Spring Rivers in southeastern Kansas. This seems to indicate that the Big Caney River populations (exclusive of the upstream stations in intermittent streams) follow a pattern commonly found in southeastern Kansas and probably in the Ozark region.

#Lepomis humilis# (Girard): Stations A-3, W-2, W-3, W-4, W-5, G-1, G-2, G-3, G-4, G-5, G-7, G-8, G-9, G-10, G-11, G-12, G-14, G-15, C-1, C-2, C-3, C-4, C-5, C-6, C-7, C-8, C-9, C-10, C-11, C-12, C-13, C-16, C-17, B-2, B-3, E-1, E-2, E-4, E-5, E-6, E-7, M-1. (C-131, C-132, C-133, C-136, J&J, E&F.)

The orangespotted sunfish was found in every stream surveyed, although only one specimen was taken from the Arkansas River.

The largest relative number of this species (44.6) was taken at station G-1. Percentages at other stations on Grouse Creek and its tributaries progressively declined in an upstream direction.

In Big Caney River representation of _L. humilis_ in collections varied from 1.56 per cent at station C-1 to 23.47 per cent at station C-7. This sunfish was usually the dominant species in collections made from the Elk River, where the relative abundance ranged from 10 to 30 per cent.

The orangespotted sunfish is widespread in Kansas and seems to be a diagnostic constituent of the Plains Fauna. Moore and Buck (1953:26) found it "very common" in the Chikaskia River in Kansas and Oklahoma. Cross (1950:140) noted that in Stillwater Creek it seemed to be the most tolerant and consequently the most abundant of the stream's cent rarchids. Moore and Paden (1950:91) note that _L. humilis_ is most common in muddy waters and found in overflow pools, backwaters, and oxbow lakes. This species is frequently found in farm ponds in the area surveyed, which further suggests a wide range of habitat tolerance.

#Lepomis megalotis breviceps# (Baird and Girard): Stations W-3, W-4, W-5, B-1, B-2, G-1, G-4, G-5, G-7, G-8, G-9, G-10, G-11, G-12, G-13 (all Big Caney River stations except C-18), E-1, E-2, E-3, E-4, E-5, E-6, M-1, M-2 (C-131, C-132, C-133, J&J, E&F).

In Big Caney River the longear sunfish shared dominance with the redfin shiner (_Notropis umbratilis_) at almost every station. The average of its relative abundance at all stations in the Big Caney system was 16.5 per cent. It was also abundant at several stations on Grouse Creek and made up 43.25 per cent of all fish taken at station G-4.

Cross (1950:140) observed that _L. megalotis breviceps_ increased in Stillwater Creek probably as a result of clearer water and stabilized water level.

In collections made west of the area treated here (Moore and Buck, 1953:26; Elliott, 1947) the longear sunfish is less abundant than in Big Caney River and Grouse Creek.

#Lepomis macrochirus# (Rafinesque): Stations W-3, G-3, G-4, G-5, C-3, C-5, E-1, E-2 (C-131, C-132, C-133).

The bluegill was, in all cases, a minor constituent in the fish fauna. No clear pattern of habitat preference can be deduced. In the Verdigris River at Independence (collection AM-53, August 22, 1956) bluegills were common in quiet pools and coves below a low-water dam. Moore and Paden (1950:91) note that _L. macrochirus_ prefers quiet waters and Hubbs and Lagler (1947:94) state that it is "generally restricted to the quieter pools."

The bluegill is widely-stocked in impoundments of the area treated here.

#Aplodinotus grunniens# (Rafinesque): Stations C-4, E-2.

The dearth of stations from which the freshwater drum is reported may indicate difficulty in taking this species with seines, rather than scarcity. Both collections were at downstream stations. At station C-4 three half-grown drum were taken. Fishermen take "drum" at least as far upstream as station C-5 on Big Caney River. In the Elk River one specimen was taken in a 20-foot seine below a dam at Elk Falls.

FISHES OF DOUBTFUL OR POSSIBLE OCCURRENCE

In addition to the species listed above, the following species have been reported nearby and may occur within the area surveyed.

_Lepisosteus productus_ (Cope)--This gar has not been reported from Kansas. It has been taken at several points in the northern half of Oklahoma and as far west as Canton Reservoir by Buck and Cross (1951). A specimen of the spotted gar was taken by Elkin (1954:28) in Salt Creek in Osage County, Oklahoma.

_Polyodon spathula_ (Walbaum)--The paddlefish has never been reported from the Arkansas River system in Kansas. Several reports by fishermen were traced by the writer, but authentication was not achieved. One mounted specimen was examined in a sporting goods store in Arkansas City. This fish was said to have been taken on the Arkansas River south of Arkansas City but information on the date and method of capture were vague. Mr. Darrell Wheat of Arkansas City reported taking four paddlefish below a dam at Oxford, Kansas, in 1948 and 1949.

_Hiodon alosoides_ (Rafinesque)--One specimen (K. U. 3095) of the goldeye was taken in 1953 on the Arkansas River near Oxford in Sumner County. Fishermen also report taking this fish occasionally in the Walnut River in Cowley County.

_Noturus flavus_ (Rafinesque)--The stonecat was taken in the Verdigris system by R. D. Lindsay in 1911 (K. U. 2058) and more recently by Cross in Montgomery County (C-120) and Schelske (1957:46) in Wilson and Montgomery Counties. The close proximity of these collection areas to lower portions of the Elk River indicate probable occurrence in Elk River and other Verdigris tributaries.

_Noturus nocturnus_ (Jordan and Gilbert)--The freckled madtom has been taken on all sides of the area studied making its occurrence therein highly probable. This madtom has been taken in Beaver Creek in Osage County, Oklahoma (OAM 4771); from a tributary of the Walnut River in Sedgwick County by Cross (1954); from the Chikaskia River (Moore and Buck, 1953:24); and from several localities on the Verdigris River (Schelske, 1957:47).

_Etheostoma cragini_ (Gilbert)--One Cragin's darter (K. U. 3470) was taken by Cross in the Arkansas River near the Sumner-Cowley county line (Sec. 25, T31S, R2E). Records of this darter are few and widely scattered geographically. Several collections from north-eastern Oklahoma are noted by Moore and Cross (1950:144).

_Etheostoma whipplii_ (Girard)--Schelske (1957:38) reports the redfin darter from the Verdigris River three miles southeast of Benedict, Kansas. Dr. George Moore of Oklahoma A. & M. College states that it has been taken in the Verdigris drainage in Oklahoma at several locations.

_Etheostoma zonale arcansanum_ (Jordan and Gilbert)--Two banded darters (K. U. 3213) have been reported by Schelske (1957:49) from Fall River near Neodesha, Kansas. Because a tributary of Fall River enters Elk County its presence in this and other Verdigris tributaries in the area seems possible. This darter has been reported from only one other stream in Kansas, Shoal Creek in Cherokee County, where it has been collected often.

_Roccus chrysops_ (Rafinesque)--The white bass has been stocked in Hulah Reservoir on Big Caney River in Oklahoma. To date it has not been reported from the Big Caney in Kansas. White bass are common in many reservoirs of Kansas and Oklahoma and have been taken in rivers in both states. Mr. Clement Gillespie of Arkansas City, Kansas Forestry, Fish and Game Commission wildlife protector for the area, states that two hundred young of _R. chrysops_ were released in Grouse Creek several years ago under auspices of the Commission. The fish has not been reported by fishermen since that time to the knowledge of Mr. Gillespie or of the writer.

_Lepomis microlophus_ (Gunther)--One redear sunfish was taken on Salt Creek in Osage County, Oklahoma, by Elkin (1954:28). Because this species has been stocked widely in Oklahoma its eventual occurrence in Kansas seems probable.

_Chaenobryttus gulosus_ (Cuvier)--The warmouth has been taken south of the collection area in Osage County on Salt Creek by Elkin (1954:28).

FAUNAL COMPARISONS OF DIFFERENT STREAMS

The faunas of Elk River, Big Caney River, and Grouse Creek were generally similar. These streams and most of their tributaries originate in the same hilly area of eastern Cowley County and western Elk and Chautauqua counties; their similarities and differences have been pointed out.

The following species were taken in all of these streams:

_Lepisosteus osseus_
_Dorosoma cepedianum_
_Ictiobus bubalus_
_Moxostoma erythrurum_
_Minytrema melanops_
_Cyprinus carpio_
_Campostoma anomalum_
_Notropis boops_
_Notropis lutrensis_
_Notropis umbratilis_
_Notropis volucellus_
_Pimephales notatus_
_Pimephales tenellus_
_Fundulus notatus_
_Gambusia affinis_
_Ictalurus melas_
_Ictalurus punctatus_
_Etheostoma spectabile_
_Percina caprodes_
_Micropterus salmoides_
_Pomoxis annularis_
_Lepomis cyanellus_
_Lepomis humilis_
_Lepomis megalotis_
_Lepomis macrochirus_
_Labidesthes sicculus_

No species was found in Elk River to the exclusion of Big Caney and Grouse Creek. Fish taken exclusively in Grouse Creek were _Ictiobus cyprinella_ at station G-2 and _Notropis percobromus_ at station G-1. The following species were taken only in Big Caney River: _Ictiobus niger_, _Notropis rubellus_, _Phenacobius mirabilis_, _Pimephales vigilax_, and _Pomoxis nigromaculatus_.

_Notropis buchanani_ and _Pimephales promelas_ were taken in Grouse Creek and Elk River, but not in Big Caney River, although the watershed of Big Caney lies largely between these two streams. Three species, _Notropis camurus_, _Micropterus punctulatus_, and _Aplodinotus grunniens_, were found in Elk River and Big Caney but not in Grouse Creek. _Ictalurus natalis_, _Pylodictis olivaris_, and _Percina phoxocephala_ were taken in Big Caney River and Grouse Creek but not in Elk River. _Percina copelandi_ was taken by Cross on Elk River in 1954 and 1955 (K. U. 3464 and K. U. 3197).

Forty species were taken in Big Caney River, 35 in Grouse Creek and 31 in Elk River. Collections were made from only six stations on Elk River as compared with 18 from Big Caney and 17 from Grouse Creek.

Twenty-four species were taken in the Walnut River system, only one of which (_Notemigonus crysoleucas_) was taken exclusively there.

In the Arkansas River 18 species were found, four of which did not occur elsewhere. These were _Hybopsis aestivalis_, _Notropis blennius_, _N. girardi_, and _Fundulus kansae_.

Table 5 lists the number of stations in each of the streams surveyed from which each species was taken.

TABLE 5.--SPECIES OF FISHES COLLECTED AND NUMBER OF STATIONS IN
EACH STREAM SYSTEM AT WHICH EACH SPECIES WAS FOUND.

KEY:
A: Arkansas River 3 stations
B: Walnut River 5 stations
C: Grouse Creek 17 stations
D: Big Caney River 18 stations
E: Elk River 6 stations
F: Middle Caney 2 stations
G: Beaver Creek 3 stations

============================================================
Total number | | | | | | |
of stations | A | B | C | D | E | F | G
--------------------+-------+---+----+----+--------+---+----
_L. osseus_ | 1 | 3 | 3 | 6 | Seen | |
_D. cepedianum_ | Seen | 1 | 1 | 3 | 2 | 1 |
_Carpiodes carpio_ | 2 | 1 | 1 | 1 | | |
_I. bubalus_ | | 1 | 2 | 4 | 3 | |
_I. cyprinella_ | | | 1 | | | |
_I. niger_ | | | | 2 | | |
_M. erythrurum_ | | | 4 | 10 | 3 | |
_M. melanops_ | | | 1 | 3 | 1 | |
_Cyprinus carpio_ | 1 | 4 | 4 | 2 | 1 | |
_C. anomalum_ | | 1 | 1 | 14 | 2 | | 1
_H. aestivalis_ | 1 | | | | | |
_N. blennius_ | 2 | | | | | |
_N. boops_ | | | 2 | 14 | 2 | 2 |
_N. buchanani_ | | | 1 | | 1 | |
_N. camurus_ | | | | 13 | 2 | |
_N. deliciosus_ | 3 | 3 | | | | |
_N. girardi_ | 2 | | | | | |
_N. lutrensis_ | 3 | 4 | 13 | 14 | 5 | 1 | 3
_N. rubellus_ | | | | 11 | | |
_N. percobromus_ | 3 | 3 | 1 | | | |
_N. umbratilis_ | | | 8 | 18 | 4 | 2 | 2
_N. volucellus_ | | | 2 | 5 | 2 | 1 |
_N. crysoleucas_ | | 1 | | | | |
_H. placita_ | 3 | 2 | | | | |
_P. mirabilis_ | | 1 | | 1 | | |
_P. notatus_ | | 1 | 6 | 18 | 5 | 2 | 1
_P. promelas_ | 2 | 2 | 1 | | 1 | 1 | 1
_P. vigilax_ | 1 | | | 3 | | 1 |
_P. tenellus_ | | | 1 | 7 | 1 | 2 |
_F. notatus_ | | 4 | 10 | 1 | 1 | | 1
_F. kansae_ | 2 | | | | | |
_G. affinis_ | 3 | 5 | 8 | 8 | 1 | |
_I. melas_ | 1 | 4 | 12 | 9 | 5 | | 3
_I. natalis_ | | | 6 | 3 | | |
_I. punctatus_ | 1 | 2 | 1 | 5 | 1 | |
_P. olivaris_ | 1 | | 1 | 1 | | |
_E. spectabile_ | | 1 | 4 | 17 | 2 | | 1
_P. copelandi_ | | | 1 | 5 | | |
_P. phoxocephala_ | | | 1 | 4 | | |
_P. caprodes_ | | | 5 | 8 | 1 | |
_M. salmoides_ | | | 4 | 2 | 3 | | 1
_M. punctulatus_ | | | | 7 | 1 | |
_P. annularis_ | | 2 | 7 | 1 | 4 | 1 |
_P. nigromaculatus_ | | | | 1 | | |
_L. cyanellus_ | | 3 | 14 | 17 | 5 | 1 | 3
_L. humilis_ | 1 | 4 | 13 | 17 | 6 | 1 | 2
_L. megalotis_ | | 3 | 9 | 18 | 6 | 2 | 2
_L. macrochirus_ | | 1 | 3 | 3 | 2 | |
_A. grunniens_ | | | | 1 | 1 | |
_L. sicculus_ | | | 5 | 7 | 4 | 1 | 1
------------------------------------------------------------

DISTRIBUTIONAL VARIATIONS WITHIN THE SAME STREAM

An analysis of faunal variations in different parts of the same stream system was made for Big Caney River and Grouse Creek. Collecting was more extensive in these streams, and sampling was done over a wider range of habitat, than in the Arkansas and Walnut rivers.

The fish taken in the first five seine hauls at each station were counted and the number of each species was recorded as a percentage of the total number of fish taken. These percentages were calculated for the main stream and for each tributary in an attempt to discern possible intra-stream faunal patterns. In Table 6 lower, middle, and upper segments of each stream have been segregated and the average of all stations within each segment is shown.

The results are subject to several sources of error, some of which are discussed below:

(1) Seining techniques could not be entirely standardized. One station might present a series of long narrow riffles and narrow, shallow pools in which only a small seine could be used effectively; another station might consist of a large, deep, isolated pool in which a larger seine was needed for effective sampling. In practice, the five seine hauls were made with any of several seines ranging from ten to twenty feet in length.

(2) Seines are species-selective, due partly to the preference of certain fishes for special habitat niches. Fishes that are often found under stones or in weedy pools require special collecting techniques and frequently were not represented in the initial five hauls. If work subsequent to the first five hauls indicated that such fish were a prominent part of the fauna at a particular station, these results were considered before percentages were calculated.

(3) Temporal variations occur in populations at the same station. There were both seasonal and diurnal differences in relative numbers of species taken in these collections. This was noted especially at station C-5 where collecting was done both at night and by day. Spawning by certain species during the course of the study complicated estimates of their relative abundance.

(4) In tabulating percentages of fishes obtained an arbitrary element is often unavoidable in deciding whether a station, especially a station on a tributary, should be considered as part of the lower, middle, or upper segment of a river system.

Despite these disadvantages it is felt that table 6 has factual basis permitting some reliable interpretation.

TABLE 6.--RELATIVE ABUNDANCE IN PER CENT OF FISHES IN
COLLECTIONS FROM THREE STREAM SEGMENTS.