Fish assemblage of Cedar Fork Creek, Ohio, unchanged for 28 years (1).

Berra, Tim M. ; Petry, Paulo

ABSTRACT. A 270 m section of Cedar Fork Creek, a clear, gravel-bottomed, headwater stream of the Ohio River System in north-central Ohio, was sampled 48 times from 1975-2003. During the 28 year period 32,237 individuals of 10 fish families and 44 species were collected. Eight taxa made up 86% of the total number collected. Cyprinids (Luxilus cornutus, Campostoma anomalum, Pimephales notatus, Semotilus atromaculatus, and Notropis buccatus) accounted for 65% of the individuals. Two darter species (Etheostoma caeruleum, E. nigrum) made up 17.4% of the total, and the white sucker, Catostomus commersoni, accounted for 3.3%. The same common species were abundant throughout the 28 years, and the same rare species were consistently present in small numbers. Species richness averaged 23 species per year. Margalef's index of diversity varied only slightly from 2.8 to 3.6 during the study indicating the constancy of species composition. Two jackknife estimators (nonparametric resampling procedures) suggested that the collections detected 97-100% of the species present. Exotic species failed to make inroads into the Cedar Fork community except for carp, Cyprinus carpio, that have been in Ohio since 1879, and brown trout, Salmo trutta, that were recently stocked by the Ohio Department of Natural Resources.

INTRODUCTION

Cedar Fork Creek is an unpolluted, clear, warm-water, gravel-bottomed, fourth-order stream in Richland County near the village of Bellville in north central Ohio at 40[degrees]37.019 N, 82[degrees]33.923 W. The Cedar Fork is 15.4 km long with a gradient of 3.5 m/km and drains an area of 124 [km.sup.2]. Various measurements of its environmental quality such as the Index of Biotic Integrity, Modified Index of Well Being, Invertebrate Community Index, and the Qualitative Habitat Evaluation Index range from good to exceptional (Ohio EPA 2000). This locality is just a few km south of the generally accepted late Wisconsinan boundary (Totten 1973). Cedar Fork is a tributary of the Clear Fork of the Mohican River that is part of Ohio's largest watershed, the Muskingum River, which flows into the Ohio River at Marietta (Sanders 2002).

The fishes of Cedar Fork Creek, OH, have been sampled more or less regularly since 1975. In a study on the incidence of black spot disease in 4,175 fishes belonging to 29 taxa, pool-dwelling species were more likely than riffle-dwelling species to be infected by the fluke, Uvulifer ambloplitis (Berra and Au 1978). Berra and Au (1981) examined 18,361 fish specimens from 34 species for incidences of teratological fishes in Cedar Fork Creek. Only 47 defective fishes (0.26%) were found. The incidence of deformed fishes from heavy-metal polluted streams in the nearby Rocky Fork of the Mohican River was 0.53% (Reash and Berra 1989). Mohican watershed drainage maps, water chemistry, and a comparison of polluted and unpolluted stream fish populations were given by Reash and Berra (1986, 1987). Gleason and Berra (1993) studied the hybridization of Luxilus cornutus and L. chrysocephalus in Cedar Fork Creek. They reported a stable ratio of 91% L. cornutus, 7% L. chrysocephalus, and 2% hybrids, and demonstrated selection against young-of-year hybrids.

Matthews (1998, p 104-129) provided a review of the extensive literature of stream fish assemblages. By considering fluctuation and stability as part of the same long-term picture, Matthews' analysis reconciled the apparent contradiction between two seemingly mutually exclusive statements, namely that fish populations fluctuate in abundance and composition annually and yet one can collect essentially the same species in the same numbers at the same locality from year to year. Grossman and others (1998) demonstrated that variability in mean and peak flows has a stronger effect on fish assemblage structure and use of spatial resources in a North Carolina stream than either interspecific competition for space or predation. Fish assemblage structure in a Texas prairie stream was determined more by differences in environmental conditions among sites than by seasonal variation in environmental conditions (Ostrand and Wilde 2002).

Meffe and Berra (1988) examined the persistence (constancy of species composition) and stability (constancy of relative species abundance) of the fish assemblage in Cedar Fork Creek based on 38 collections made from July 1975-July 1984. They concluded that the assemblage was both persistent and stable over the 9-year period. The purpose of the present paper is to extend the analysis to 28 years.

MATERIALS AND METHODS

Collection Site and Methods

A total of 48 samples from July 1975-August 2003 was collected from the same 270 m section of Cedar Fork Creek. Collections were made in 15 years of the 28-year period. All collections were made with a 1.6 mm mesh, 3.0 m seine, with the exception of 3 samples obtained by pulsed direct-current electrofishing. Most collections (75%) were made during the warm months May-September, when water temperature averaged 19[degrees] C (range = 7.5-23.5[degrees] C). No collections were made in January and February when ice lined the banks. The stream was very active geomorphologically especially during spring flooding. Erosion and deposition constantly rearranged the pools and riffles, but the locations of the downstream and upstream boundaries of the study area were the same for each collection. Land usage in the rural rolling hills of the watershed was recreational and agricultural, consisting of corn, soybeans, and dairy cattle, and changed little during the course of the study. Water depth varied from a few cm in riffle areas to nearly 2.0 m in deep pools. The stream ranged from 520 m wide during normal flow. The dominant aquatic vegetation was attached filamentous algae. There was little floating or emergent plant cover. One or occasionally two teams of seiners covered the study area for each sample. Sampling times averaged 127 min (60-170 min), usually in the afternoon. If two teams were utilized, their times were additive. Three collections were made with multiple teams of students. Those times are not included in the 127 min average. Specimens were fixed in 10% formalin and stored in 40% isopropyl alcohol for later identification using Trautman (1957, 1981). Because of the difficulty of positively identifying some small scaleless juveniles and hybrids, Luxilus cornutus and L. chrysocephalus were lumped as Luxilus spp., Moxostoma dusquesnei and M. erythrurum were lumped as Moxostoma spp., and Lepomis cyanellus, L. gibbosus, and L. macrochirus were lumped as Lepomis spp. (Table 1). Voucher specimens were retained in the fish collection of the first author.

Statistical Methods

Inter-annual species diversity differences were analyzed using Margalef's diversity index [D.sub.Mg] = (S-1)1n[N.sup.-1] where S is the number of species in sample i, and N = total number of specimens in sample i. This index is a function of species captured accounting for the sum of all specimens in each sample. It also accounts for the difference in sampling effort, thus providing an unbiased comparison among years. Samples from the same year of collecting were pooled for analysis. Total species richness for Cedar Fork Creek was estimated using two Jackknife estimators, which are nonparametric resampling procedures (Palmer 1990, 1991) implemented in PCORD (McCune and Mefford 1997).

RESULTS

Forty-four fish species and 32,237 individuals belonging to 10 families were recorded in 48 collections from Cedar Fork Creek between 1975 and 2003 (Table 1). Of these, 8 taxa numbered greater than 1000 individuals each and made up 86% of the total. Minnows (Cyprinidae) were the dominant family accounting for 65% of the individuals. Two darter species (Percidae) made up 17.4% of the total catch, and white suckers, Catostomus commersoni, composed 3.3%. Eleven species, belonging to 7 families, were represented by 6 or fewer individuals each and accounted for less than 0.1% (Table 1). The remaining 14% of the fish fauna was composed of 21 taxa from 6 families whose individuals numbered from 11-778.

Eleven taxa were represented in all 15 years of the study in which collections were made, and 10 taxa were present in at least 46 of the 48 collections (Table 1). Most of these species were numerically abundant but others were always present at lower levels. Species richness varied among years from 17 to 31 species, with an average of 23 species per year. Margalef's index of diversity varied from 2.8 to 3.6 over the 28-year period of sampling, indicating that the fish community remained largely constant with little variation in species composition once sampling effort was accounted for (Fig. 1). The highest observed value was in 1987 after a major flood that apparently introduced some farm pond species into the system raising the number of species detected during that year's sampling. Estimates of total number of species were 44 and 45 for the first and second order Jackknife estimators, respectively, suggesting that the sampling procedure detected 97-100% of the estimated number of species present in Cedar Fork Creek.

[FIGURE 1 OMITTED]

DISCUSSION

The fish assemblage of Cedar Fork Creek has changed very little in the 28 years since collections began in 1975. Species richness remained largely unchanged between years for the entire sampling period (Fig. 1). The same core group of taxa that dominated the fauna between 1975-1984 (Meffe and Berra 1988) is still dominant today, and the same rarely collected species are still rarely, but consistently, encountered. The mean percent of a given taxon/sample closely mirrors its percent of the total number of specimens over the 28-year study (Table 1). For example, Campostoma anomalum composed 17.3% of the total number of specimens, and its representation in each of the 48 samples averaged 18.4%. This indicates that the relative proportion of the various species did not change much over time. Invasive species other than carp, Cyprinus carpio, have made little headway in this stream. Carp have been in Ohio since 1879 (Trautman 1981). Brown trout, Salmo trutta, have been stocked by the Ohio Department of Natural Resources in the Clear Fork and have thereby gained access to the Cedar Fork. The presence of brown trout was confirmed by electrofishing. This species did not appear in the samples obtained by seining. The Clear Fork and its tributary, Cedar Fork, are two of only a few "trout streams" in Ohio. The emergence of cool spring water from the bottom of the streambed makes trout survival possible, but summer temperatures are too warm for natural reproduction. Other Ohio trout streams include the Mad River (Great Miami River drainage) and Clear Creek (Hocking River drainage) (Sanders 2002). Only one catfish specimen was taken. A small black bullhead, Ameurus melas, was seined after a large flood in July 1987. This fish almost certainly was washed into the Cedar Fork from farm ponds in the area. The two species of crappies, Pomoxis (Table 1), are also most likely escapees from farm ponds during floods.

We make no pretense that our sampling effort was identical for each collection. This would be impossible due to the constantly changing stream morphology, water depth, temperature changes, seasonal variation, time spent, number of helpers, and so on. However, the overall trend is abundantly clear and the samples are representative of the system. The fish species composition has not changed significantly in 28 years. In an analysis of stream fish assemblages from 50 undegraded northern watersheds Wang and others (2003) wrote that a diverse assemblage of cyprinids, catostomids, centrarchids, and percids, like we report in Table 1, existed in warm, wide, gravel-bottomed streams similar to those in our study area.

It should be noted that the 32,237 fishes collected in the 48 samples were removed from the population, yet the community consistently reassembled. This is reminiscent of the repopulation studies conducted by Berra and Gunning (1970) who demonstrated that longear sunfish (Lepomis megalotis) populations removed from Louisiana stream segments could recover within one year. Meffe and Sheldon (1990) reported a similar post-defaunation recovery of fish assemblages in South Carolina streams. Longitudinal movement within the overall stability of the larger system is thought to be responsible for the repopulation in both of the above reports and the current study.

Even though numbers change from year to year, the overall pattern of persistence and stability remains the same over an extended period of time. This apparent paradox can be explained by viewing the assemblage as a whole. In the absence of anthropogenic factors (pollution, habitat degradation, invasive species), most stream fish assemblages oscillate about stable conditions and are relatively resistant to change (Matthews 1998). The year-to-year variation of relative abundance among the dominant species is most likely the effect of inter-annual variation of recruitment success. It is reassuring that the Cedar Fork fish assemblage has remained intact for 28 years. Barring any human disturbance, it should persist with only minor natural variation for the foreseeable future.

ACKNOWLEDGMENTS. The comments of G. Meffe, W. Matthews, and S. Spotte greatly improved an early draft of the manuscript. We are grateful to R.-J. Au, R. Berra, and C. Gleason who participated in the early field work, and to the various students who helped with the more recent collections.

LITERATURE CITED

Berra TM, Au R-J. 1978. Incidence of black spot disease in fishes in Cedar Fork Creek, Ohio. Ohio J Sci 78:318-22.

Berra TM. Au R-J. 1981. Incidence of teratological fishes from Cedar Fork Creek, Ohio. Ohio J Sci 81:225-9.

Berra TM, Gunning GE. 1970. Repopulation of experimentally decimated sections of streams by longear sunfish, Lepomis megalotis megalotis (Rafinesque). Trans Am Fish Soc 99:776-81.

Gleason CA, Berra TM. 1993. Demonstration of reproductive isolation and observation of mismatings in Luxilus cornutus and L. chrysocephalus. Copeia 1993:614-28.

Grossman GD, Ratajczak RE Jr, Crawford M, Freeman MC. 1998. Assemblage organization in stream fishes: effects of environmental variation and interspecific interactions. Ecological Monographs 68:395-420.

Matthews WJ. 1998. Patterns in freshwater fish ecology. New York: Chapman & Hall and International Thompson Publ. 756 p.

McCune B, Mefford MJ. 1997. PC-ORD. Multivariate analysis of ecological data. Version 3.08. Gleneden Beach (OR): MjM Software Design.

Meffe GK, Berra TM. 1988. Temporal characteristics of fish assemblage structure in an Ohio stream. Copeia 1988:684-90.

Meffe GK, Sheldon AL. 1990. Post-defaunation recovery of fish assemblages in southeastern blackwater streams. Ecol 71:657-67.

Ohio EPA. 2000. 1988 Biological and water quality study of the Black Fork, Clear Fork, Rocky Fork and Jerome Fork of the Mohican River and selected tributaries. Columbus (OH): Environmental Protection Agency, Div of Surface Water. 157 p.

Ostrand KG. Wilde GR. 2002. Seasonal and special variation in a prairie stream-fish assemblage. Ecol of Freshwater Fish 11:137-49.

Palmer M. 1990. The estimation of species richness by extrapolation. Ecol 71:1195-8.

Palmer M. 1991. Estimating species richness: the second-order jackknife reconsidered. Ecol 72:1512-3.

Reash RJ, Berra TM. 1986. Fecundity and trace-metal content of creek chubs from a metal contaminated stream. Trans Am Fish Soc 115:346-51.

Reash RJ, Berra TM. 1987. Comparison of fish communities in a clean-water stream and an adjacent polluted stream. Am Midl Nat 118:301-22.

Reash RJ, Berra TM. 1989. Incidence of fin erosion and anomalous fishes in a polluted stream and a nearby clean stream. Water, Air, and Soil Pollut 47:47-63.

Sanders RE, editor. 2002. A guide to Ohio streams. Columbus: Ohio Chapter Am Fish Soc. 110 p.

Totten SM. 1973. Glacial geology of Richland Co., Ohio. Columbus: Ohio Div Geol Surv. Rept on Investigations No. 88.

Trautman MB. 1957. The Fishes of Ohio. Columbus: Ohio State Univ Pr. 683 p.

Trautman MB. 1981 (rev ed). The Fishes of Ohio. Columbus: Ohio State Univ Pr. 782 p.

Wang L, Lyons J, Rasmussen P, Seelbach P, Simon T, Wiley M, Kenehl P, Baker E, Niemela S, Stewart P. 2003. Watershed, reach, and riparian influences on stream fish assemblages in Northern Lakes and Forest Ecoregion, USA. Can J Fish Aquat Sci 60:491-505.

(1) Manuscript received 5 May 2005 and in revised form 15 September 2005 (#05-08).

(2) Corresponding author: e-mail: berra.l@osu.edu

(3) e-mail: fishnwine@charter.net

TIM M. BERRA (2) AND PAULO PETRY (3), Department of Evolution, Ecology & Organismal Biology, The Ohio State University, Mansfield, OH 44906; Museum of Comparative Zoology, Department of Ichthyology, Harvard University, Cambridge, MA 02138

TABLE 1
Taxa collected in Cedar Creek in a 28 year period between 1975
and 2003. Number of samples = 48

 Mean %/
Family/Species n % Total Samples

Petromyzomtidae
 Lampetra aepyptera 2 0.006 0.012
 (least brook lamprey)

Clupeidae
 Dorosoma cepedianum 1 0.003 0.006
 (gizzard shad)

Cyprinidae
 Campostoma anomalum 5582 17.316 18.421
 (central stoneroller)
 Clinostomus elongatus 62 0.192 0.311
 (redside dace)
 Cyprinella spiloptera 4 0.012 0.027
 (spotfin shiner)
 Cyprinus carpio (common carp) 6 0.019 0.028
 Hybopsis amblops (bigeye chub) 83 0.257 0.236
 Luxilus cornutus (1) 7827 24.280 22.618
 (common & striped shiners)
 Notropis buccatus 1251 3.881 3.157
 (silverjaw minnow)
 Notropis photogenis 220 0.682 0.850
 (silver shiner)
 Notropis rubellus 206 0.639 0.835
 (rosyface shiner)
 Notropis stramineus 135 0.419 0.779
 (sand shiner)
 Notropis volucellus 38 0.118 0.167
 (mimic shiner)
 Phoxinus erythrogaster 20 0.062 0.066
 (southern redbelly dace)
 Pimephales notatus 3888 12.061 9.967
 (bluntnose minnow)
 Pimephales promelas 51 0.158 0.199
 (fathead minnow)
 Rhinichthys atratulus 674 2.091 2.993
 (eastern blacknose dace)
 Semotilus atromaculatus 2396 7.432 7.958
 (creek chub)

Catostomidae
 Carpiodes cyprinus (quillback) 190 0.589 0.476
 Catostomus commersoni 1073 3.328 3.135
 (white sucker)
 Hypentelium nigricans 492 1.526 1.619
 (northern hog sucker)
 Moxostoma spp. (2) 272 0.844 0.644
 (black and golden redhorse)

Ictaluridae
 Ameurus melas (black bullhead) 1 0.003 0.005

Salmonidae
 Salmo trutta (brown trout) 4 0.012 0.059

Gasterosteidae
 Culaea inconstans 11 0.034 0.042
 (brook stickleback)

Cottidae
 Cottus bairdii (mottled sculpin) 778 2.413 2.479

Centrarchidae
 Ambloplites rupestris 52 0.161 0.166
 (rock bass)
 Lepomis spp. (3) 161 0.499 0.722
 (green, pumpkinseed, bluegill)
 Micropterus dolomieui 58 0.180 0.153
 (smallmouth bass)
 Micropterus punctatus 1 0.003 0.002
 (spotted bass)
 Micropterus salmoides 51 0.158 0.172
 (largemouth bass)
 Pomoxis annularis 1 0.003 0.005
 (white crappie)
 Pomoxis nigromaculatus 2 0.006 0.016
 (black crappie)

Percidae
 Etheostoma blennoides 614 1.905 1.854
 (greenside darter)
 Etheostoma caeruleum 3310 10.268 11.350
 (rainbow darter)
 Etheostoma flabellare 297 0.921 1.155
 (fantail darter)
 Etheostoma nigrum 2287 7.094 6.879
 (johnny darter)
 Etheostoma zonale 127 0.394 0.412
 (banded darter)
 Percina caprodes (logperch) 4 0.012 0.011
 Percina maculatas 5 0.016 0.012
 (blackside darter)

Total 32237

 No. Years No. Coll.
Family/Species Present Present

Petromyzomtidae
 Lampetra aepyptera 2 2
 (least brook lamprey)

Clupeidae
 Dorosoma cepedianum 1 1
 (gizzard shad)

Cyprinidae
 Campostoma anomalum 15 48
 (central stoneroller)
 Clinostomus elongatus 8 15
 (redside dace)
 Cyprinella spiloptera 2 3
 (spotfin shiner)
 Cyprinus carpio (common carp) 3 3
 Hybopsis amblops (bigeye chub) 11 26
 Luxilus cornutus (1) 15 48
 (common & striped shiners)
 Notropis buccatus 15 46
 (silverjaw minnow)
 Notropis photogenis 12 25
 (silver shiner)
 Notropis rubellus 12 32
 (rosyface shiner)
 Notropis stramineus 10 16
 (sand shiner)
 Notropis volucellus 2 7
 (mimic shiner)
 Phoxinus erythrogaster 6 10
 (southern redbelly dace)
 Pimephales notatus 15 47
 (bluntnose minnow)
 Pimephales promelas 5 12
 (fathead minnow)
 Rhinichthys atratulus 15 44
 (eastern blacknose dace)
 Semotilus atromaculatus 15 48
 (creek chub)

Catostomidae
 Carpiodes cyprinus (quillback) 6 12
 Catostomus commersoni 15 46
 (white sucker)
 Hypentelium nigricans 14 43
 (northern hog sucker)
 Moxostoma spp. (2) 9 24
 (black and golden redhorse)

Ictaluridae
 Ameurus melas (black bullhead) 1 1

Salmonidae
 Salmo trutta (brown trout) 1 2

Gasterosteidae
 Culaea inconstans 8 9
 (brook stickleback)

Cottidae
 Cottus bairdii (mottled sculpin) 15 48

Centrarchidae
 Ambloplites rupestris 9 20
 (rock bass)
 Lepomis spp. (3) 11 28
 (green, pumpkinseed, bluegill)
 Micropterus dolomieui 8 16
 (smallmouth bass)
 Micropterus punctatus 1 1
 (spotted bass)
 Micropterus salmoides 10 13
 (largemouth bass)
 Pomoxis annularis 1 1
 (white crappie)
 Pomoxis nigromaculatus 3 2
 (black crappie)

Percidae
 Etheostoma blennoides 15 46
 (greenside darter)
 Etheostoma caeruleum 15 48
 (rainbow darter)
 Etheostoma flabellare 14 37
 (fantail darter)
 Etheostoma nigrum 15 48
 (johnny darter)
 Etheostoma zonale 11 35
 (banded darter)
 Percina caprodes (logperch) 3 4
 Percina maculatas 2 4
 (blackside darter)

Total

(1) includes 91% Luxilus cornutus and 7% L. chrysocephalus.

(2) includes Moxostoma dusquesnei and M. erythrurum.

(3) includes Lepomis cyanellus, L. gibbosus, and L. macrocbirus.