The Project Gutenberg eBook of Fish Populations, Following a Drought, In the Neosho and Marais des Cygnes Rivers of Kansas, by James Everett Deacon.

Location	Date	Number	Average total length (inches)	Range
Middle Neosho	1958	6	14.9	13.9-15.5
Middle Neosho	1959	9	13.6	11.0-16.0
Lower Neosho	1958	3	21.0	20.3-21.6
Lower Neosho	1959	5	21.3	18.0-24.5

Gizzard shad declined in abundance from 1957 to 1959. The largest population occurred at the middle station on the Marais des Cygnes in 1957. Shad were mainly in quiet water; often, when the river-level was high, I found them predominately in backwaters or in the mouths of tributary streams. Examination of nine individuals, ranging in size from seven inches to 13.5 inches T. L., indicated that maturity is reached at 10 to 11 inches T. L. Spawning probably occurred in late June in 1959 ("ripe" female caught on 26 June); young-of-the-year were first recorded in mid-July.

The blue sucker was taken rarely in the Neosho River and not at all in the Marais des Cygnes in my study. Cross (personal communication) obtained several blue suckers in collections made in the mainstream of the Neosho River in 1952; both young and adults occupied swift, deep riffles. The species seemingly declined in abundance during the drought, and at the conclusion of my study (1959) had not regained the level of abundance found in 1952.

Big-mouthed buffalo were found in quiet water at all stations, but were rare. A ripe female, 21.5 inches long, was taken at the lower station on the Neosho on 16 June, 1959.

Ictiobus niger (Rafinesque)
Black Buffalo
and
Ictiobus bubalus (Rafinesque)
Small-mouthed Buffalo

Black buffalo were not taken at the upper station on the Neosho and were rare at other stations. Small-mouthed buffalo were taken at all stations and were common in the lower portions of the two streams. While the shocker was being used, buffalo were often seen only momentarily, thereby making specific identification impossible; both species were frequently taken together, and for this reason are discussed as a unit. Both species maintained about the same level of abundance throughout my study.

The two species were taken most often in the deeper, swifter currents of the mainstream, but were sometimes found in pools, creek-mouths and backwaters. On several occasions in the summer of 1959, buffalo were seen in shallow parts of long, rubble riffles, with the dorsal or caudal fins protruding above the surface. Ernest Craig, game protector, said buffalo on such riffles formerly provided much sport for gig-fishermen. He stated that the best catches were made at night because the fish were less "spooky" then than in daytime. In my collections made by use of the shocker, buffalo were taken more frequently at night (Table 9, p. 402).

On 19 June, 1959, I saw many buffalo that seemed to be feeding as they moved slowly upstream along the bottom of a riffle. The two species, often side by side, were readily distinguishable underwater. Small-mouthed buffalo appeared to be paler (slate gray) and more compressed than the darker black buffalo. To test the reliability of underwater identifications, I identified all individuals prior to collection with a gig. Correct identification was made of all fish collected on 19 June. The smallest individual obtained in this manner was 18.5 inches T. L. On 26 August, 1959, 16 small-mouthed buffalo were captured and many more were seen while the shocker was in use in the same riffle for one hour and ten minutes. One small-mouthed buffalo was caught while the shocker was being used in the pool below that riffle for one hour and fifty minutes. No black buffalo were taken on 26 August.

Spawning by buffalo was not observed but probably occurred in spring; all mature fish in my earliest collections (mid-June of each year) were spent. Small-mouthed buffalo reach maturity at approximately 14 inches T. L.

River carpsucker were abundant throughout the study at all stations. Adults were taken most frequently in quiet water, but depth and bottom-type varied. The greatest concentrations occurred in mouths of creeks during times of high water; occasionally, large numbers were taken in a shallow backwater near the head of a riffle at the middle Neosho station. River carpsucker feed on the bottom but seem partly pelagic in habit. They were taken readily by means of the shocker and gill nets at all depths. The population of C. carpio in the Neosho River probably was depleted by drought, although many individuals survived in the larger pools.

When stream-flow was restored, carpsucker probably moved rapidly upstream but had a scattered distribution in 1957. Trautman (1957:239) states that in the Scioto River, Ohio, river carpsucker moved upstream in May and downstream in late August and early September. Numbers found at the middle and lower Neosho stations suggest similar movements in the Neosho River in 1957. In midsummer they were common at the middle station but rare at the lower station; however, they became abundant at the lower station in November. The abundance in late fall at the lower Neosho station might have resulted either from downstream migration or from continued upstream movement into thinly populated areas. No indication of seasonal movement was found in 1958 or 1959.

River carpsucker reach maturity at approximately 11 inches T. L., and spawning occurs in May or June. A ripe male was taken from a gravel-bottomed riffle, three feet deep, at the middle station on the Neosho station on 10 June 1959.

Fig. 2. Length-frequency of river carpsucker in the Neosho River, 1958 and 1959. Fig. 2.

The size-distribution of individuals taken at the middle Neosho station is presented in Fig. 2. The collection in early July of 1958 indicates that one size-group (probably the 1957 year-class) had a median length of approximately seven inches. The modal length of this group was nine inches in June, 1959. A second, predominant size-group (Fig. 2) seemed to maintain almost the same median size throughout all the collection periods, although specimens taken in the spring of 1959 were slightly smaller than those obtained in 1958. This apparent stability in size may have been due to an influx of the faster-growing individuals from a smaller size-group, coupled with mortality of most individuals more than 14 inches in length.

Young-of-the-year were taken at every station. Extensive seining along a gravel bar at the lower Neosho station indicated that the young are highly selective for quiet, shallow water with mud bottom. In these areas, young-of-the-year carpsucker were often the most abundant fish.

River carpsucker were collected more readily by use of the shocker after dark than in daylight (Table 9, p. 402).

A specimen of Carpiodes velifer taken at the lower station on the Neosho in 1958 provided the only record of the species in Kansas since 1924. Many specimens, now in the University of Kansas Museum of Natural History, were taken from the Neosho River system by personnel of the State Biological Survey prior to 1912. The species has declined greatly in abundance in the past 50 years.

The short-headed redhorse occurred at all stations. It was common at the middle and lower stations on the Neosho, rare at the upper station on the Neosho, abundant at the upper station on the Marais des Cygnes in 1957, and rare thereafter at all stations on the Marais des Cygnes. Short-headed redhorse typically occur in riffles, most commonly at the uppermost end where the water flows swiftly and is about two feet deep. An unusually large concentration was seen on 13 June, 1959, in shallow (six inches), fast water over gravel bottom at the middle station on the Neosho River.

Thirty-nine individuals were marked by clipping fins at the middle Neosho station in 1959. Four were recovered from one to 48 days later: two at the site of original capture (one 48 days after marking), one less than one-half mile downstream, and one about one mile downstream from the original site of capture.

At the middle Neosho station in 1958, this species was taken more readily by use of the shocker at night than by day (Table 9, p. 402).

The golden redhorse was abundant at the upper Neosho station, rare at the middle Neosho station, and did not occur in collections at other stations. This species was taken most frequently over gravel- or rubble-bottoms in small pools below riffles, and was especially susceptible to collection by means of the shocker.

Twenty-nine golden redhorse of the 1957 year-class, taken at the upper Neosho station on 9 September 1958, were 6.2 to 8.6 inches in total length (average 7.4 inches); 26 individuals of the same year-class caught on 21 August 1959 were 9.3 to 13.5 inches in total length (average 10.9 inches).

The carp decreased in abundance from 1957 to 1959 at the upper and middle Marais des Cygnes station and at the middle and lower Neosho stations. Carp were more abundant in the Marais des Cygnes than in the Neosho, although the largest number in any single collection was found in one pool at the upper Neosho station in 1958.

Carp were taken most commonly in quiet water near brush or other cover. At the middle Neosho station, collecting was most effective between the hours of 6:30 a.m. and 12:30 p.m. and least effective between 12:30 p.m. and 6:30 p.m. (Table 9, p. 402). Ripe males were taken as early as 19 April (16.1 inches, 19.4 inches T. L.) and as late as 30 July (16 inches T. L.) at the middle Neosho station. Ripe females were taken as early as 19 April at the middle Neosho station (19.2 inches T. L.) and as late as 7 July at the lower Neosho station (16 inches T. L.). Young-of-the-year were taken first at the middle Marais des Cygnes on 8 July 1957. They were recorded on later dates at the upper Marais des Cygnes and at the lower and middle Neosho stations.

The golden shiner was taken rarely at the upper Marais des Cygnes station in 1958 and 1959 and at the middle Marais des Cygnes station in 1957 and 1958. At the middle Neosho station Notemigonus was seined from a pond that is flooded frequently by the river, but never was taken in the mainstream.

The creek chub was taken only at the upper stations on both rivers. It increased in abundance at the upper Neosho station from 1957 to 1959, and was not taken in the upper Marais des Cygnes until 1959.

A single specimen from the lower Marais des Cygnes station provides the only record of the species from the Marais des Cygnes system in Kansas, and is the only silver chub that I found in either river in 1957-1959. The species is taken often in the Kansas and Arkansas rivers.

The gravel chub, present only at the lower and middle Neosho stations, occupied moderate currents over clean (free of silt) gravel bottom. The gravel chub was not taken in 1957, was rare at both Neosho stations in 1958, became common at the lower Neosho station in part of 1959, but was never numerous at the middle Neosho station. Dr. F. B. Cross recorded the species as "rare" in 1952 at a collection site near my middle Neosho station, but larger numbers were taken then than in any of my collections at that station. The population was probably reduced by drought, and recovery was comparatively slow following restoration of flow.

Young-of-the-year and adults were common in collections from riffles at the lower Neosho station from 1 July through 8 July, 1959. I obtained only one specimen in intensive collections in the same area on 25, 26, and 27 August. Seemingly the species had moved off shallow riffles into areas not sampled effectively by seining.

The sucker-mouthed minnow was common at the middle Marais des Cygnes station but was not taken at the upper and lower stations until 1959, when it was rare. At the middle and lower Neosho stations this fish increased in abundance from 1957 to 1959; at the upper station, sucker-mouthed minnows were not taken until 1959 when collections were made on the White farm. There, the species was common immediately below a low-head dam, but was not taken in extensive collections on the Bosch Farm in 1959.

The species was most common immediately below riffles, or in other areas having clean gravel bottom in the current. On 5 June, 1959, many individuals were taken at night (11:30 p.m.) on a shallow gravel riffle (four inches in depth) where none had been found in a collection at 5:00 p.m. on the same date.

Young-of-the-year were taken at the lower Neosho station on 24 June, 1959, and commonly thereafter in the summer.

In 1958, the rosy-faced shiner was taken rarely at the lower stations on both streams. This species is common in smaller streams tributary to the lower portions of the two rivers, and probably occurs in the mainstream only as "overflow" from tributaries. Possibly, during drought, rosy-faced shiners found suitable habitat in the mainstream of Neosho and Marais des Cygnes rivers, but re-occupied tributary streams as their flow increased with favorable precipitation, leaving diminishing populations in the mainstream.

The red-finned shiner, most abundant at the upper Neosho station, occurred at all stations except the upper Marais des Cygnes. This fish seems to prefer small streams, not highly turbid, having clean, hard bottoms. It is a pool-dwelling, pelagic species.

The blunt-faced shiner was taken only in 1957, at the middle Neosho station, where it was rare. This species, abundant in clear streams tributary to the Neosho River (field data, State Biological Survey) may have used the mainstream as a refugium during drought. The few specimens obtained in 1957 possibly represent a relict population that remained in the mainstream after flow in tributaries was restored by increased rainfall.

The red shiner, abundant in 1952 (early stage of drought), was consistently the most abundant fish in my collections in the Marais des Cygnes and at the lower and middle Neosho stations. However, the abundance declined from 1957 to 1959 at the two Neosho stations. At the upper Neosho station the species was fourth in abundance in 1957, and third in 1958 and 1959 (Table 12).

The red shiner is pelagic in habit and occurs primarily in pools, though it frequently inhabits adjacent riffles. Collections by seining along a gravel bar at the lower station showed this fish to be most abundant in shallow, quiet water over mud bottom, or at the head of a gravel bar in relatively quiet water. At the lower end of the gravel bar in water one to four feet deep, with a shallow layer of silt over gravel bottom and a slight eddy-current, red shiners were replaced by ghost shiners or river carpsucker young-of-the-year as the dominant fish.

Fifty-nine dyed individuals were released in an eddy at the lower end of a gravel bar at the middle Neosho station on 5 June, 1959. Some of these fish still were present in this area when a collection was made 30 hours later. No colored fish were taken in collections from quiet water at the upper end of the gravel bar. A swift riffle intervening between the latter area and the area of release may have impeded their movement. Forty-six individuals, released at the head of the same gravel bar on 10 June, 1959, immediately swam slowly upstream through quiet water and were soon joined by other minnows. These fish did not form a well-organized school, but moved about independently, with individuals or groups variously dropping out or rejoining the aggregation until all colored fish disappeared about 50 feet upstream from the point of release.

Evidence of inshore movement at night was obtained on 8 June, 1959, in a shallow backwater, having gravel bottom, at the head of a gravel bar at the middle Neosho station. A collection made in the afternoon contained no red shiners, but they were abundant in the same area after dark.

In Kansas, red shiners breed in May, June, and July. Minckley (1959:421-422) described behavior that apparently was associated with spawning. Because of its abundance, the red shiner is one of the most important forage fishes in Kansas streams, and frequently is used as a bait minnow.

The mimic shiner was taken only rarely at the two lower Neosho stations. This species, like N. camurus, is normally more common in clear tributaries than in the Neosho River, and probably frequents the mainstream only during drought.

Field records of the State Biological Survey indicate that the ghost shiner was common in the mainstream of the lower Neosho River during drought. In 1957, the species was abundant at the lower and middle stations on the Neosho River and at the lower Marais des Cygnes station.

Collections at all stations show that the species has a definite preference for eddies—relatively quiet water, but adjacent to the strong current of the mainstream rather than in backwater remote from the channel. The bottom-type over which the ghost shiner was found varied from mud to gravel or rubble.

The sand shiner was taken rarely in the Neosho and commonly in the Marais des Cygnes in 1952. In my study the species occurred at all stations, but not until 1959 at the upper and lower Neosho stations. Sand shiners were found with equal frequency in pools and riffles. Spawning takes place in June and July.

The mountain minnow was common at the lower and middle Neosho stations throughout the period of study, and increased in abundance from 1957 to 1959. It was taken only in 1959 at the upper Neosho station, where it was rare. This species does not occur in the Marais des Cygnes River. The largest numbers were found in 1959 at the lower Neosho station, where this fish occurred most commonly in moderate current over clean gravel bottom. The mountain minnow, like Hybopsis x-punctata, was common in late June and early July but few were found in late August, 1959. The near-absence of this species in collections made in late August is responsible for the apparent slight decline in abundance from 1957 to 1959, as shown in Table 11. Metcalf (1959) found mountain minnows most commonly in streams of intermediate size in Chautauqua, Cowley and Elk counties, Kansas. The predilection of this species for permanent waters resulted in an increase in abundance during my study. With continued flow, this species possibly will decrease in abundance in the lower mainstream of the Neosho River. I suspect that the species is, or will be (with continued stream-flow), abundant in tributaries of intermediate size in the Neosho River Basin.

The parrot minnow was not taken in the Marais des Cygnes River and was absent at the upper Neosho station until 1959. This species was common at the lower and middle Neosho stations throughout the period of study and increased in abundance from 1957 to 1959.

At the lower Neosho station, this fish preferred slow eddy-current over silt bottom, along the downstream portion of a gravel bar. The parrot minnow was taken less abundantly in the latter part of the summer, 1959, than in early summer, but the decline was less than occurred in the mountain minnow.

The blunt-nosed minnow was common, and increased in abundance in both rivers from 1957 to 1959. The largest numbers were found at the upper Neosho station in 1959, and a large population also was present at the lower Neosho station in 1959.

Pools having rubble bottom, bedrock, and small areas of mud were preferred at the upper Neosho station. At the lower Neosho station the fish was most common in quiet water at the lower end of a gravel bar. The parrot minnow also was common in this general area; nevertheless, these two species were seldom numerous in the same seine-haul, indicating segregation of the two. The blunt-nosed minnow was taken frequently in moderate current over clean gravel bottom, especially in late summer, 1959, when P. notatus increased in abundance as the mountain minnow decreased.

The fat-headed minnow was taken at all stations except at the lower one on the Marais des Cygnes, and was most abundant at the upper Neosho station. Intensive seining at the lower Neosho station indicated that this species preferred quiet water and firm mud bottom.

In the Neosho River in 1957 to 1959, habitats of the species of Pimephales seemed to be as follows: Pimephales tenellus (mountain minnow) occurred primarily in moderately flowing gravel riffles in the downstream portions of the river. Pimephales vigilax (parrot minnow) was mostly in the quiet areas having mud bottom at the downstream end of gravel bars, and less commonly on adjacent riffles, at the lower station. Pimephales notatus (blunt-nosed minnow) had a wider range of habitats, occurring in quiet areas and moderate currents both upstream and downstream. Pimephales promelas (fat-headed minnow) occurred throughout both rivers but was most abundant in the quiet water at the uppermost stations.

The stoneroller was most abundant at the upper Neosho station and was not taken at the lower Marais des Cygnes station. This fish increased in abundance from 1957 to 1959, but was never common at the middle Marais des Cygnes or the middle and lower Neosho stations.

The stoneroller prefers fast, relatively clear water over rubble or gravel-bottom.

The abundance of channel catfish was greatly reduced as a result of the drought of 1952-1956. With the resumption of normal stream-flow in 1957, the small numbers of adult channel catfish present in the stream produced unusually large numbers of young. These young of the 1957 year-class, which reached an average size of about nine inches by September 1959, will provide an abundant adult population for several years.

The reduction in number of channel catfish in streams can be related to the changed environment in the drought. When stream levels were low in 1953 (Tables 1-4), fish-populations were crowded into a greatly reduced area. An example of these crowded conditions was observed by Roy Schoonover, Biologist of the Kansas Forestry, Fish and Game Commission, in October, 1953, when he was called to rescue fish near Iola, Kansas. The Neosho River had ceased to flow and a pool (less than one acre) below the city overflow dam was pumped dry. Schoonover (personal communication) estimated that 40,000 fish of all kinds were present in the pool. About 30,000 of these were channel catfish, two inches to 14 inches long, with a few larger ones. Fish were removed in the belief that sustained intermittency in the winter of 1953-1954 would result in severe winterkill. These conditions almost certainly were prevalent throughout the basin.

In addition to winterkill, crowding probably resulted in a reduced rate of reproduction by channel catfish, and by other species as well. This kind of density-dependent reduction of fecundity is known for many species of animals (Lack, 1954, ch. 7). In fish, it is probably expressed by complete failure of many individuals to spawn, coupled with scant survival of young produced by the adults that do spawn. Reproductive failure of channel catfish in farm ponds, especially in clear ponds, is well known, and is often attributed to a paucity of suitable nest-sites (Marzolf, 1957:22; Davis, 1959:10).

In the Neosho and Marais des Cygnes rivers, the intermittent conditions prevalent in the drought resulted in reduced turbidity in the remaining pools. Many spawning sites normally used by channel catfish were exposed, and others were rendered unsuitable because of the increased clarity of the water. In addition, predation on young channel catfish is increased in clear water (Marzolf; Davis, loc. cit.), and would of course be especially pronounced in crowded conditions. The population was thereby reduced to correspond to the carrying capacity of each pool in the stream bed.

The return of normal flow in 1957 left large areas unoccupied by fish and the processes described above were reversed. The expanded habitat favored spawning by nearly the entire adult population, and conditions for survival of young were excellent. As a result, a large hatch occurred in the summer of 1957. (Several hundred small channel catfish were sometimes taken by use of the shocker a short distance upstream from a 25-foot seine, set in a riffle). Subsequent survival of the 1957 year-class has been good. By 1959, few of the catfish spawned in 1957 had grown large enough to contribute to the sport fishery, but they are expected to do so in 1960 and 1961.

The 1957 year-class was probably the first strong year-class of channel catfish since 1952. Davis (1959:15) found that channel catfish in Kansas seldom live longer than seven years. The 1952 year-class reached age seven in 1959. The extreme environmental conditions to which these fish were subjected in drought caused a higher mortality than would occur in normal times. The adult population in the two rivers probably was progressively reduced throughout the drought, and the reduction will continue until the strong 1957 year-class replenishes it. For these reasons, fishing success was poor in 1957-1959.

Juvenile channel catfish were more abundant in the Neosho than in the Marais des Cygnes in 1958 and 1959, although both streams supported sizable populations. In the Marais des Cygnes the upper station had fewer channel catfish than the middle and lower stations. In the Neosho, populations were equally abundant both upstream and downstream. The habitat of channel catfish in streams has been discussed by Bailey and Harrison (1948).

I found adults in various habitats throughout the stream, but most abundantly in moderately fast water at the lower and middle Neosho stations. At the upper Neosho station where riffles are shallow, yearlings and two-year-olds were numerous in many of the small pools over rubble-gravel bottom. Cover was utilized where present, but large numbers were taken in pools devoid of cover. Young-of-the-year were nearly always taken from rubble- or gravel-riffles having moderate to fast current at both upstream and downstream stations.

Collections showed that young of 1957 were abundant on riffles throughout the summer and until 17 November, 1957. Subsequent collections were not made until 11 May, 1958, at which time 1957-class fish still were abundant on riffles at the lower Neosho station; on that date, the larger individuals were in deeper parts of the riffles than were smaller representatives of the same year-class.

In a later collection (2 June, 1958), numbers present on the riffles were greatly reduced and the larger individuals were almost entirely missing. Some of the smaller individuals were still present in the shallower riffle areas. Table 7 compares sizes of the individuals obtained on 2 June with sizes collected from deep riffles at the middle Neosho station on 7 June, 1958. The larger size of the group present in deep riffles is readily apparent. The yearlings almost completely disappeared from subsequent collections on riffles.

A bimodal size-distribution of young-of-the-year was noted also in 1958 and 1959; but, no segregation of the two sizes occurred on riffles in summer. Marzolf (1957:25) recorded two peaks in spawning activity in Missouri ponds. Two spawning periods may account for the bimodal size distribution of young-of-the-year observed in my study.

In 1959, young-of-the-year began to appear in the latter part of June and became abundant by the first part of July. Individuals as small as one inch T. L. were taken in gravel-bottomed riffles on 1 July, 1959.

Yearling individuals at the lower and middle Neosho stations showed a pronounced tendency to move into shallow, moderately fast water over rubble or gravel bottom at night, where they were nearly ten times more abundant than in daytime (Table 9). Adults probably have the same pattern of daily movement as yearlings, except that at night the adults move to deeper riffles. Bailey and Harrison (1948:135-136) demonstrated that channel catfish feed most actively from sundown to midnight.

Channel catfish (especially two-year-olds and adults) were abundant on a rubble-riffle during the day in some collections at the lower Neosho station in 1959.

Table 7. Length-frequency of Channel Catfish from the Neosho River, 1957, 1958 and 1959. (Numbers in Vertical Columns Indicate the Number of Individuals of a Certain Size Collected on That Date.)

Length in inches	Nov. 2 1957	June 2 1958 (shallow riffle)	June 7 1958 (deep riffle)	Sept. 9 1958	Sept. 11 1959
1.5				1
2.0	3
2.5	13	2		1	2
3.0	4	11		3	4
3.5	3	21	7	1	14
4.0		11	12		9
4.5		4	10	1
5.0		2	11	2
5.5		1	7	26
6.0				58	2
6.5			1	32	5
7.0				16	5
7.5			1	4	5
8.0					22
8.5					45
9.0					81
9.5					41
10.0					21
10.5					8
11.0					4
11.5					1
12.0					3
12.5					1
13.0					1

Near the end of the spawning season in 1959, I found spawning catfish at the lower Neosho station. Ripe females were taken between 9 June and 30 June, 1959; and, on 19 June I found a channel catfish nest with eggs (water temp. 79° F.). The nest-site was a hole in the base of a clay bank; the floor was clean gravel with a small mound of gravel at the entrance. The nest-opening, five to six inches in diameter, widened almost immediately into a chamber about two and one-half feet long and one foot wide. Normally the water was about six inches deep in the mainstream as it ran over a riffle adjacent to the catfish nest. When I put my hand into the opening the fish bit vigorously, but became quiescent when I stroked its belly. I then felt the rounded gelatinous mass of eggs on the bottom of the nest. On June 22 (water temp. 86° F.) the fish was removed, struggling, from the nest, and returned to the stream. The next day (23 June 1959, water temp. 84° F.) the eggs had hatched and the young were in a swarm in the nest. The adult did not attempt to bite but left as soon as I put my hand into the hole.

Marzolf (1957:25) reports that young remain in the nest from seven to eight days after hatching. My seining records show a marked increase in abundance of small young-of-the-year on the first of July. Probably the time of hatching of the nest described above correlated well with hatches of other nests.

One and sometimes two channel catfish were found in other holes in the stream-bank or bottom. The fish occasionally attacked my hand vigorously, but at other times remained quiet or left without attacking. No other channel catfish eggs were found, although one hole under a rock in the middle of the river had one or two individuals in it each time it was checked until 11 July, 1959. A local fisherman informed me of his belief that these holes are occupied only in the spawning season.

Observations that I made in a pond owned by Dr. E. C. Bryan of Erie indicated that channel catfish, when disturbed in the early stages of guarding the eggs, either eat the eggs and abandon the nest or leave the nest exposed to predation by other animals. In the later stages of nesting, the fish, if removed, will return to guard the nest. After the eggs hatch the guarding response probably diminishes and the fish leaves the nest readily.

At the lower Neosho station, several "artificial" holes were dug into the clay bank and two pieces of six-inch pipe were forced into the bank. Nearly all these holes were occupied by catfish for a short period in June; many of the holes were enlarged, either by the current or by fish. I suspect that fish enlarged some holes, because in the spawning season several males were observed that had large abrasions atop their heads, around their lips, and to a lesser extent on their sides. These could have been caused by butting and scraping the sides, roof and floor of a hole. I found it possible to enlarge the holes by rapidly moving my hand while it was inside a hole.

The growth-rate of channel catfish in the Neosho was approximately the same at all stations, and the large 1957 year-class grew to an average size of about nine inches by mid-September, 1959 (Table 7). Channel catfish mature at a total length of 12 to 15 inches. Thus, most individuals of the 1957 year-class in the Neosho River probably will mature in their fourth or fifth summer (1960 or 1961 spawning season).

The sizes attained by young-of-the-year in 1957 differed in the two rivers. Six hundred and thirty-three young taken in the Marais des Cygnes River attained an average size of 4.7 inches (range two to six inches) by mid-September. (Age was determined by length-frequency and verified by examining cross-sections of fin-spines from the larger individuals). One hundred and fifty young from the Neosho River averaged 3.0 inches (range 2 to 3.7 inches) on 2 November. Gross examination of the riffle-insect faunas indicated a larger standing crop in the Neosho than in the Marais des Cygnes River. Thus, the slower growth of young channel catfish in the Neosho seemed not to be correlated with food supply. Bailey and Harrison (1948:125-130) found that young channel catfish in the Des Moines River, Iowa, fed almost exclusively on aquatic insect larvae. My observations indicate that this is true in the Neosho and Marais des Cygnes rivers also.

Young produced in 1958 in the Neosho River attained an average total length of three inches by 26 August, and young produced in 1959 attained an average size of 3.5 inches by 11 September. Both groups probably continued growth until October, and may have averaged four inches total length at that time.

The 1958 and 1959 year-classes were much less abundant than were the 1957 young. Therefore, it seems likely that the growth of the 1957 young in the Neosho River was depressed because of crowding. The 1959 year-class was larger than the small 1958 year-class, thus conforming to a general expectation that strong year-classes will be followed by weak year-classes.

Reproduction by channel catfish in 1957 seemed greater in the Neosho River than in the Marais des Cygnes River (Table 10); this coincided with a greater change in volume of flow in the Neosho River than in the Marais des Cygnes River (Tables 1-4). The 1957 year-class seemed more crowded, and grew more slowly, in the Neosho than in the Marais des Cygnes River.

Yellow bullhead were taken only at the middle station on the Marais des Cygnes and upper station on the Neosho. The yellow bullhead is more restricted to streams than is the black bullhead. Both species decreased in abundance during a period of continuous flow (1957 to 1959) following drought at the upper Neosho station. Collections in 1958-'59 indicated an increase in average size. Of four individuals marked and released at the upper Neosho station in 1959, one was recaptured about three hours after being released. It had not moved from the area of release.

The black bullhead was abundant at the upper stations on each river, especially in backwaters having mud-bottom. The species was not taken in the mainstream of the lower and middle Neosho stations, but was taken at the middle Neosho station in a pond that is often flooded by the river. Although the fish was common or abundant in nearly all pools at the upper Neosho station, it was most abundant in one pool that had a bottom predominately of mud.

At the middle Marais des Cygnes station, 109 individuals were collected and fin-clipped on 8, 9 and 24 July 1957. Three of the 19 marked on 8 July were recaptured in the same area on 9 July. The area was poisoned on 13 September, 1957, and 130 black bullhead were taken, none of which had been marked.

In 1959, 96 black bullhead were taken at the upper Neosho station (five in Area 1 and 91 at the White Farm). In these collections, 25 were marked (fin-clipped or dyed) and six were recaptured. Four of the six had not left the area of capture one and two days after being released. The fifth fish recaptured was one of five individuals that had been displaced one pool downstream. When recaptured seven days later, this fish had moved upstream over two steep riffles (two to three inches deep, 75 feet and 166 feet long) past the site of original capture to the next pool. The sixth fish, marked at the same time but returned to the original pool, was recaptured nine days after original capture and had moved upstream over a long riffle (two to three inches deep, 166 feet long) and a short riffle into the second pool above the original site of its capture.

Rotenone was applied to a small (.04 acre-feet) backwater ditch having a soft mud bottom at the upper Marais des Cygnes station on 25 July, 1957; 1526 black bullhead, one green sunfish and one white crappie were collected. A sample of 60 bullhead averaged 4.6 inches T.L. (range 3.5 to 6.6 inches) and 540 individuals averaged .7 ounce each. These fish probably represented the 1956 year-class.

The upper Neosho station had a large population of black bullhead, strongly dominated by fish less than four inches T. L. (range 1.5 to 3.8 inches), in the spring of 1957. Most were approximately two inches T. L. and probably represented the 1956 year-class. Growth, according to length-frequency, following restoration of stream-flow, shows a regular increase in length of this dominant 1956 year-class (Fig. 3). A scarcity of young, especially in 1958 and 1959, is apparent in Fig. 3. This may be due to the fact that a strong year-class usually is followed by one or several weak year-classes. However, it more probably reflects the fact that black bullhead are characteristically pond fish, and as such are not so well adapted to reproduction in flowing streams as are many other species. Metcalf (1959) found this species most abundantly in the intermittent headwaters of Walnut River and Grouse Creek in Cowley County, Kansas.