Effects of Oil Field Brines on Biological Integrity of Two Tributaries of the Little Muskingum River, Southeastern Ohio

Two headwater tributaries of the Little Muskingum River were compared to assess possible effects of oil field brines on biological integrity of the streams. Diatom, macroinvertebrate, and fish communities were analyzed during the summer, 1990, for possible changes in community-structure caused by brines. Dissolved chemicals associated with brines and natural surface waters were quantified monthly between May and October, 1991. Cranenest Fork, with a lower density of wells producing brines than Straight Fork, had a slightly more integrated benthic macroinvertebrate community and lower proportions of salt-tolerant diatoms. Twelve of fifteen measures of benthic macroinvertebrate community structure were consistent with expected effects of greater brine enrichment in Straight Fork. Both streams, however, conformed to benthic macroinvertebrate criteria for protection of aquatic life established in Ohio Water Quality Standards. The major difference between streams was the larger (13:1) percentage of salt-tolerant diatoms such as Navicula salinarum,Navicula tripunctata, andNavicula viridula v. avenacea in Straight Fork than in Cranenest Fork. Fish communities were similar between streams. Cranenest Fork and Straight Fork shared relatively similar geomorphological characteristics and land-use patterns. Chemical-physical quality was similar except for conductivity and levels of sodium and chlorides, all of which were significantly greater in Straight Fork, as would be expected with the probability of greater brine enrichment of this stream. The maximum concentration of total dissolved solids (<300 mg/L) was very low compared to most surface water in Ohio. No chemical-physical water quality violations of Ohio Water Quality Standards were noted in either stream. Ohio J. Sci. 92 (5): 139^146, 1992 INTRODUCTION The Little Muskingum River is a small tributary of the Ohio River located on the unglaciated plateau of the Appalachian Highlands of southeastern Ohio. The river has been assigned special status as a State of Ohio Resource Waterway and designated as a warmwater aquatic habitat, suitable for agricultural and industrial uses, and safe for primary body-contact recreation (Ohio Environmental Protection Agency 1987). The watershed is one of the most heavily forested and sparsely populated areas in Ohio. There are no large municipalities. Only a few scattered villages, rural homes, and small farms occur in the area. The watershed is relatively free of major water pollutants associated with mining activities, heavy industry, and municipal sewage, common in other sections of Ohio (U.S. Forest Service 1990). The primary water quality problems affecting aquatic life in the river are siltation from eroding soils, organic wastes from households and farm animals, and brines from the production of oil and gas within the watershed. The effects of the latter pollutants were the special focus of the present investigation. Accidental and deliberate discharge of brines were known to occur and were suspected of having an impact on aquatic life. Of special interest was the possible effects of brines on aquatic life from so-called Exempt Mississippian (oil/ gas) Wells (EMWs). An Exempt Mississippian Well (EMW) as defined by the State of Ohio, Division of Oil and Gas, must have been 'Manuscript received 3 August 1992 and in revised form 3 November 1992 (#92-20). drilled and completed before 1 January 1980; located in the unglaciated part of the state; drilled no deeper than the Mississippian Big Injun Sandstone in the entire area with Pennsylvanian age bedrock or drilled no deeper than the Mississippian Berea Sandstone in the area with Permian age bedrock; and the well must be permanently connected from the wellhead to a residence for domestic fuel supply. Owners of EMWs must comply with all Ohio laws and rules regulating oil and gas production, but are exempt from paying a fee if brines are not disposed into an injection well, filing a certificate of liability insurance, and filing a plan for storage and disposal of brine with the Ohio Division of Oil and Gas. Otherwise brine disposal practices must follow methods approved by the Division of Oil and Gas of the Ohio Department of Natural Resources. Actual brine disposal practices were a matter of controversy between oil/gas producers and government regulators. In some areas brines are released from EMWs into riparian zones or directly into the river or its tributaries. The objective of the present study was to determine whether these brine releases significantly affected the biotic integrity of the river. We report here the results of chemical-physical ambient water analyses and of assessments of biological integrity for two headwater streams in the Little Muskingum River system. One stream, Cranenest Fork, with a low density of EMWs within the watershed, is compared to Straight Fork, a stream with a relatively high density of EMWs. The streams are otherwise quite similar with respect to geomorphology and land-use patterns. No reference streams completely free of EMWs could be located. 140 EFFECT OF BRINES—L. MUSKINGUM RIVER VOL. 92 MATERIALS AND METHODS Study Area The Little Muskingum River originates from several tribu-taries at elevations of 250-300 m in Monroe County and flows generally southwest for approximately 100 km to the Ohio River near Marietta, OH. Two headwater tributaries of the Little Muskingum River, Cranenest Fork and Straight Fork, were selected for comparative analysis (Fig. 1). These streams share relatively similar geomorphometric characteristics and land-use patterns, except for numbers of EMWs per unit area of watershed (Table 1). Straight Fork has more than twice the density of EMWs as Cranenest Fork. Otherwise the streams are similar in watershed area, stream length and gradient, and bedrock composition. Upper Pennsylvanian and Permian Dunkard Systems form the bedrock (Bownocker 1920, Stauffer and Schroyer 1920, Thornbury 1965). Bedrock exposures include Green and Washington Formations of the Permian System and the Monongahela Formation of the Pennsylvanian System. Though shales are the dominant bedrock, sandstones and limestones affect steepness of valley sides, local stream gradients, and ground water supplies. Springs are more numerous in areas underlain by limestone bedrock (Hayhurst et al. 1974). Riparian vegetation is similar for both streams, consisting mostly of aggrading hardwood forests. Land-use, except for intensity of oil/gas production, is limited primarily to widely dispersed rural homes and small farms. No municipal sewage systems discharge to either stream. Brines from EMWs in the study area have been analyzed by the Ohio Division of the Geological Survey. Based on 11 samples, mean chloride levels of approximately 61,000 mg/L were noted. A typical analysis of a filtered sample was (in mg/L): Na 26,500, potassium 176, magnesium 1,440, calcium 9,000, strontium 138, lithium 2, iron 156, iodine 10, bromine 622, chloride 61,200, total dissolved solids 100,600, pH 5.7 (Ohio Geological Survey 1991). Chemical/Physical Water Analyses Water samples for selected chemical-physical analyses were collected monthly between 10 May and 11 September 1990 at single sites on each stream (as shown in Fig. 1). Sampling events included an over bank flood (17 May) and low flow of one-tenth mean annual discharge (2 August). Sample collection, initial processing, and preservation methods were in accordance with Standard Methods (American Public Health Association 1989). Water samples were analyzed in the laboratory using United States Environmental Protection Agency (1984) protocols and Standard Methods (American Public Health Association 1989). Analyses 'were completed for chloride, bromide, sulfate, bicarbonate, nitrate, sodium, potassium, calcium, magnesium, iron, manganese, zinc, chromium, lithium, strontium, lead, barium, and arsenic. These elements and ions were selected for analysis because of the possibility that a "chemical fingerprint" of EMW brines could be detected (Breen et al. 1985). Quality control of chemical analyses was assured by using chemical standards provided by the United States Environmental Protection Agency and by arranging for duplicate analyses of split water samples with the Ohio Environmental Protection Agency water analytical laboratory. Results between laboratories were nearly identical. Water temperatures, dissolved oxygen, pH, and conductivity were determined for ambient water at the time of each sample collection. Water temperature was measured with a thermistor; dissolved oxygen with a YSI Model 57 meter; pH with a Corning Autocal Model 108 pH meter; and conductivity with a YSI Model 33 S-C-T meter. Benthic Macroinvertebrates Benthic macroinvertebrates were sampled from HesterDendy artificial substrate samplers (Hester and Dendy 1962) and from natural substrates. Hester-Dendy samplers were exposed for approximately six weeks (Ohio Environmental Protection Agency 1988). Natural substrates were sampled with a modified shovel sampler, a device designed to retain all organisms within the top 10-20 cm of sediments, so that accurate quantitative results could be obtained (Prater et al. 1977). Samples of natural sub-strates with associated invertebrates were placed into a 10 L bucket half-filled with stream water. The organisms were dislodged from the substrates with a stiff bristled brush and by agitation. Dislodged organisms were screened from the water with a U.S. Standard No. 30 soil sieve and placed into 500 ml plastic jars. The jars of organisms were retained temporarily on ice while transported to the laboratory. Within 24 hours the living organisms were hand-picked from the collection, sorted into major taxonomic groups, and preserved in 80% ethanol. All organisms were identified to the lowest practical level using taxonomic keys primarily from Merritt and Cummins (1984) and from Peckarsky et al. (1990). Individuals of each species (or lowest taxon) were counted. From these data a variety of measures of community composition were determined. These included the total number of taxa; the number of taxa and the proportion of selected groups including ephemeropterans, plecopterans, trichopterans, and dipterans; and the proportion of organic pollution-tolerant organisms. These measures of benthic community composition have been widely used as indexes of water quality (Hellawell 1986, Plafkin et al. 1989). The Ohio Environmental Protection Agency recently has incorporated these measures into a composite index known as an Invertebrate Community Index (ICI). Values for ICI have been calibrated for various regions of Ohio and incorporated into Ohio Water Quality Standards (Ohio Environmental Protection Agency 1987). In addition to these measures, the density of benthic macroinvertebrates was determined. This measure of benthic community composition is not widely used as an index of water quality because of difficulties in making accurate estimates. In the present study, estimates of density, even though scaled over a wide range, were considered important because decreased densities of benthic invertebrates were an expected major effect of brine enrichment. Three composite indexes of water quality, Shannon-Weiner diversity (Wilhm and Dorris 1968), Belgian Biotic (Depauw and Vanhooren 1983), and Trent Biotic (Woodiwiss 1964) were calculated from the inventory of taxa. These indexes are well-known and have been used widely to assess ambient water quality in OHIO JOURNAL OF SCIENCE J. H. OLIVE ET AL. 141