TPWD 1961 F-5-R-8 #642: Fisheries Investigations and Surveys of the Waters of Region 3-B: Basic Survey and Inventory of Species in the Rio Grande River of Texas

JOB COMPLETION REPORT As Required By FEDERAL AID IN FISHERIES RESTORATION ACT TEXAS . Federal Aid Project No. F—5—R-8 Fisheries Investigations and Surveys of the Waters of Region 3-B Job No. BulS Basic Survey and inventory of Species in the Rio Grande River of Texas in Region 3wB Project Leader Lawrence S. Campbell H. D. Dodgen ExecutiveSecretary Texas Game and Fish Commission Austin,Texas Marion Toole Eugene A. Walker D-J Coordinator ~Director, Program Planning February 14, 1962 ABSTRACT Sampling of fish populations included 71 gill net collections and 31 sein- ing collections from the Rio Grande River between El Paso and the mouth of the Devil's River. Examination of thousands of fish obtained by seining and 1,523 fish captured in gill nets established occurrence of #2 species of fish belonging to 11 families. Distribution and relative abundance of fish were found to con- form to three basic ecological associations: turbid intermittent pool, permanent flow clear water, and turbid indolent flow. The basic distribution patterns were as outlined for the Chihuahuan Biologic Province by Dr. Clark Hubbs. Floods, as they occur in the Rio Grande, are believed to be detrimental to propagation of many game fish. This, plus water quality, limit production of some game fish such as centrarchids. The upper third of the stream, from El Paso to Presidio, is dominated by rough and forage species that were 9h.27 per cent of the netted sample. The clear water associations produced 33.h9 per cent usable game fish and represent the most productive stream areas in the region. Data on the lower third of the stream, from Regan canyon to the mouth of the Devil's River, are insufficient and tends to be biased by the conditions that prevailed when field ‘work was done. Generally, the Rio Grande is the most productive stream in the region, and is the longest and most desirable fishing area that is now open to public access without charge. For that reason, it is afforded priority for fishery management. A neW'reservoir, now under construction near Del Rio, should be investigated to determine what may be done to promote game fish production. Reconnaissance studies in the future should include work to provide additional data on the lower third of the stream. Inquiries should be made to determine if flood retention structures on tributaries of the stream can be utilized for public fishing. The canal system in El Paso County should be re-examined in an effort to find means of employing these sturctures for that purpose. JOB COMPLETION REPORT State of Texas Project No. FeS-R-B Name: Fisheries Investigations and Surveys of the waters of Region 3—B Job No. 3—15 Title: Basic Survey and Inventory of Species, as well as Their Distribution in the Rio Grande River of Texas in Region 3—B Period Covered: April 1, 1960, to February 28, 1961 Objectives: To gather fundamental data on the above water in regard to physical, chemiu cal and biological asPects, and to determine the distribution of species present and their relative abundance. To also determine ecological factors influencing species distribution. Techniques Used: Standard procedures were used which supplied the information necessary to complete stream survey forms similar to those given by Lagler in his ”Handbook of Freshwater Biology". Inquiries and investigations to locate means of access to the river, and reconnaissance trips on the stream by boat or raft were necesm sary to locate netting and seining stations. Seventy—one overnight sets were made with gill nets at 28 locations. Nylon gill nets were used that measured 125 feet in length and 8 feet in depth. The nets were made up of 25-foot sections with mesh sizes beginning with lwinch square mesh and increasing in each succeeding section by onewhalfwinch intervals to a 3—inch square mesh. An attempt was made to use hoop and fyke nets for sampling; however, unexm pected increases in stream flow rendered this effort non—productive. Seining collections were obtained at 31 locations. Each location or col» lection consisted of several drags with either a 12- or 20-foot oneefourthminch mesh nylon straight seine. A sample of the fish captured at each location was examined in detail and the relative abundance of various Species was noted. Individuals whose identity was not positively established when examined in the field were preserved and identified in the laboratory. Detailed water quality data were obtained for six locations on the river. Data included determination of pH, total solids, chlorides, sulphates, sodium chloride, dissolved oxygen and biochemical oxygen demand. This detailed analysis was provided by the International Boundary and Water Commission and by the State Health Department. Hydrology records were obtained from those organizations and from the United States Geological Survey. Water temperatures, turbidity and pH were taken at 21 stations. Rotenone sampling of pools was planned but not completed because the desired and necessary conditions were not encountered during field work. Introductory Statement The Rio Grande River rises on the eastern face of the continental divide in southern Colorado. At the river's cold water source, temperatures average about 14° F. in January, and near 50° F. in July. Precipitation averages about 5 inches per year. At the semi—tropical mouth, temperatures average near 50‘ F. in January, and about 68° F. in July. Precipitation there averages between 20 and 30 inches. From source to estuary, the stream falls nearly three.miles, and meanders nearly 1,900 miles, making the Rio Grande the second longest river in the United States. The river commonly does not carry a great volume of water, and in many places, especially where it traverses desert, the Stream is sometimes dry. Local storms create flood conditions that subside in a few hours, and most of the annual run-off, which averages nearly 4,000,000~acre—feet, enters the river from this sporadic source. Where flow is permanent, it is usually sustained by springs and seepage. The topographic basin, which includes 335,500 square miles, is semi—arid and only about onemhalf of that area, 171,900 square miles, is contributive. After the river leaves the mountain streams, it intermittently cleaves through mountain canyOns, wide level plains, and desert terraces. Pastoral valleys, varying in width from a few hundred yards to a few miles, are inter— Spersed along the stream for three-fourths of its length. The river and these valleys are the artery for biologic and economic life within the river basin. The river receives comparatively few tributaries, and many are entirely dry except for sporadic flow. Man's influence and use of the river pre—dates recorded history, but from the year 1600, when Spanish colonization encouraged settlement, irrigation in the rich alluvial bottom land has progressed toward total usage of the river“s water. Today, over 2,500,000 acres are irrigated from the river's watershed. Primarily to control irrigation, a series of reservoirs have been constructed. These include Rio Grande, Continental, Santa Maria, Terrace, Mountain Home, Sanchez, Costilla, E1 Vado, Bluewater, Elephant Butte, Caballo, Almogordo, McMillan, Avalon, Red-Bluff, Willacy, McClure, Devil‘s Lake and Lake Walk on the United States watershed. In Mexico they are Boquilla, La Colina, Rosetilla, Madero, Venustiano Carranza, Marte Gomez, Culebron, and Palito Blanco. These structures, with International Falcon Reservoir, effectuate storage for 8,000,000 acre feet of water. Work included_in this survey was restricted to the Rio ' Grande River and does not include tributaries. Survey reports for the Devil‘s River are included under Project No. F—9—R~2, Job No. 8-4, for the Pecos River under Project No. F—S—R—S, Job No. 8—13, and for Terlingua Creek under Project No. F—5-Rm2, Job No. Bull. In all, 680.1 miles of stream, beginning with the American Dam in El Paso and ending at the mouth of the Devil's River, were in» cluded in this investigation. Findings: Most of this mid-section of the Rio Grande can arbitrarily be classified as belonging to, or being dominated by, one of three basic ecological associae tions. (See Figure 1.) These are: turbid intermittent pool, permanent flow clear water, and turbid and indolent flow associations. -3- In an attempt to present a clear picture of the stream and its fishery potential, some materials and data obtained will be grouped and included under each association as it is taken up in discussion. This is also done so that the influence of the various associations on species distribution and relative abundance of species will be more apparent, and to attempt to prevent mislead- ing data that would result from overall averages. Intermittent Pool Associations: Nearly all of the stream that should be classified in this category is located above the river°s confluence with the Concho River of Mexico. For over 293.5 miles, from El PaSo to the mouth of the Conchos, the stream and its included fish life are adversely affected and regu~ lated by intensive irrigation demands. As is illustrated in hydrology data in Tables 1 through 9, flow expands to great proportions only to return to a mini» mum flow within a few days. In that area of the state, over 1,200 miles of canals and laterals are constructed that carry water from the river to fulfill irrigation commitments for more than 155,000 acres annually. .0ther irrigation by wells from subsurface sources directly influence stream flow and are of eXm treme consequence during nonmflow periods to residual pools. Irrigation demands are greatest in late spring and through the summer. During that time the stream is usually rapidly diminished until it is reduced to pools. These are then decreased byéevaporation and seepage as water percolates into sub“ surface sands to fill vacancies created by withdrawals through wells for irrigation. Through these processes, the chemical composition of the remaining water is affected, and salinity, total dissolved solids, insoluable carbonates and pH increase. The stream bed is wide throughout most of this valley and flow velocities, except during occasional flooding, are rarely sufficient to adequately carryciflfsuspended silt and organic debris. This debris decays rapidly in warm waters which sometimes exceed 90° F. In many instances, dis~ solved oxygen-is reduced below critical requirements for many game fish, while carbon dioxide content nears saturation. Turbidity usually increases. More specific and detailed dataartaavailable in Tables 1 through 11, and as shown therein, oxygen deficiencies occasionally occur even when flow is above average. As a result of these conditions, most game fish and/or their progeny (if any) are destroyed. At the same time, undesirable tenaceous fish survive and squ cessfully reproduce. .Even when game fish of advanced sizes are released in these waters, their survival is at best temporary. At all gaging stations from the American Dam to Upper Presidio, lack of flow frequently occurs and usually lasts for ten or more succeeding days within each year. At the E1 Paso»Hudspeth County line, about midmpoint of this stream area, flow was rem corded for only 13 days in 1952, and for only 4 days in 1954. A few spring areas, such as the radioactive waters of Indian Springs in Hudspeth County, are present. Within these few widely separated Spring areas, a few game fish survive. In El Paso County, the terrain is uninhibiting and most of the stream is available and accessable to the public for fishing. However, from the Hudpseth County line to Presidio, access is often difficult or near impossible. Canals and a few small control reservoirs supply nearly all of the fishing for this .area, and apparently offer the only potential means of meeting the public demand for recreation with the presently existing facilities. (See Figure 2.) Permanent Flow Clear Water Associations: Beginning a few miles below the mouth of the Concho River, the stream bed narrows and the average fall of the stnmmnincreases from 3.89 feet per mile to 4.78 feet per mile. The frequency of -4- walled canyons increases and these constrict flow and increase velocity of the stream. From here, and continuing for many miles, a series of steps in the stream bed give momentum to flow between pools, and this movement is reduced as the water passes through the more level gradient of the pools. A buildwup of ground water increases and tends to stabilize flow as irrigation demands decrease. Within 78 miles, a rapid transition takes place. The pools become more frequent and less turbid as flow from the subsurface_is forced upward after appearing and dissappearing many times in the stream bed. Water movement through pools is more pronounced and a permanent flow finally establishes between pools. Stream ecology changes so radically that it appears to be virtually unrelated to the previously described intermittent pools. Periodic floods and confined permanent flow of the stream are retained in narrow canyon walls such as Santa Elena, Mariscal, Boquillas and others, and the increased scouring action of the current sweeps gravel and rock rubble clear of debris that would otherwise be_deposited from above. Clear waters are protected by canyon walls during much of the day from direct exposure to sunlight, and as a result, extreme temperatures are less common. As shown in Tables 12 through 14, the water quality improves and food production and other factors are conducive to game fish survival. This stream area, from a few miles below the mouth of the Concho River to a few miles below the mouth of Regan Canyon, is about 200 stream=miles long and is the largest and most desirable area of stream within this region that is now available to the However, this stream area is largely inaccessable and it was necessary to work much of the stream by floating trips that often included portages. (See Figure 3.) Future construction of access roads should make clear water associaw tions more usable to the public. Turbid Indolent Flow Associations: A few miles below the mouth of Regan Canyon the stream again broadens to fill the expanded canyon walls, velocity of flow decreases as gradient is reduced to 3.02 feet per mile. Meanders are broader and turns are longer and more gradual. Much of the silt and debris carried from above by flood waters is deposited in the stream bed, on bars at turns in the river, on occasional flood plains and along narrow shelves at the foot of canyon walls. The stream”s basic direction shifts to align itself almost directly with the daily travel of the sun during the summer and stability of flow reduces as springs exercise less influence. The stream is usually very turbid, and in summer, water temperatures sometimes build up to and exceed 90° F. As shown in Tables 15 through 18, water quality decreases as salinity, total dissolved solids, and pH increase beyond desirable limits. Dissolved oxygen deficiencies are common, and excesSive carbon dioxide content is frequently recorded. Flucn tuatidns in the stream level prevents reliable spawning activity of many game fish and all factors tend to favor production of many undesirable Species. Where infrequent Springs or clear water flow enters the stream, such as the mouth of the Pecos and Devil“s Rivers, Goodenough Springs, San Felipe Springs and a few others, game fish congregate and are sustained by those conditions. The 109 miles of stream between Regan Canyon and the mouth of the Devil”s River are largely in~ accessable. This is partly because of steep canyon walls that extend, virtually unbroken on the United States side, along the entire stream course. Also, landm owners are not inclined to permit access to the river by persons who are unknown to them. Since this stream is an international boundary, their reasoning is obvious. This final association may be said to epitomize the popular public concept of the lazy Rio Grande. (See Figure 4.) Other Basic Factors that Directly Igfluence Fish hifg: As previously stated, local floods of a sporadic nature occur every year. However, more extensive ’ 55-9] flooding is not uncommon. From 1902 to 1960, flooding that affected at least one or more of the described ecological areas occurred on 19 occasions. During that same period, seven major floods affected the entire length of the river. The effect of floods on fish life is not completely understood. However, tremendous current velocities are built up as tributaries fall into valleys. As these water levels exceed stream banks, the inundated loose soils of intensively cultivated fields are picked up and carried with the stream in great quantity. The stream load is probably greater than is common for many rivers, and is primarily made up of adobe, other adhesive clays, sand, gravel and organic debris. A smothering affect on some species of game fish or their progeny is believed to occur. It is also possible that chemicals toxic to fish, such as insecticides commonly used in agricultural practices, enter the stream during flood conditions. Many in» stances of fish kills have been reported during and immediately after flooding. (See Tablesl9 and 26.) Pollution is known to enter the stream from several sources on the United States watershed. Parts of Devil“s River and the Pecos River are known to be polluted by oil field operations and lesser tributaries probably receive similar pollutants.The extent and total detrimental effect of this factor is not known, but hydrology and water quality data indicate that this damage may be extensive. Virtually nothing is known of sources of pollution from Mexico. However, the lack of soil conservation techniques, the lack of proper facilities for disposal of polluting substances, and public apathy where such problems are concerned is obvious. In spite of the prevailing conditions, the production of invertebrate food for sustaining fish life is often excellent. Many beetle larvae {Colepge£a), primarily of the genus Berosgg, are present in the coarse gravel of clear water stream beds. Other insects and larvae encountered in clear water associations are larvae of crane flies, Diptera, mayflies and their nymphs, phemerida, and c; damselflies and dragonflies, Odonata. Aquatic Hemiptera are also present. In intermittent pools and in turbid slow-flow waters, only_piptera.and their larvae were abundant. A few Hemiptera were also observed. Mollusca were common but seldom abundant through the entire river. Aquatic vegetation is not regarded as constituting a major problem in any of these stream associations. This is primarily the result of scouring action of periodic floods, the normal fluctuation of the stream and in many instances constant turbidity. A complete and accurate inventory of plants present was not possible under the conditions that prevailed When field work for this survey was done. However, some of the more important species that are known to occur are pondweeds (Potomogeton diversifiora) with additional Species of that genus probably occurring, arrowhead (§2§§EEEEEE“SP.), tapegrass QVallis: neria Sp.), yellow waterlily tfluphat advena;, and muskgrass (Chara Sp.). Positive Species identification of most of these plants was impossible since such identification is often dependent upon flowering parts of the plants. It was necessary, because of other work commitments, that the survey work be done at a time when these parts of the plants were absent or were undeveloped. Along shores, cattails (Typha latifolia) were common and were especially numerous in sunny marshy areas near canals and in other places of permanent moisture. Other shore plants included were bulrushes (Scirpus americanus), rushes {guncus nodosus)with other plants of this same genus also probably occurw ring, sedges (Carex sp.), smartweed {Polygonigm sp.), desert willow (Chilopsis linearis), dock (Rumex_mexicanus), alfalfa (Medicago saliva) that had escaped —6— from domestic cultivation, Dipetalis subulata which has no accepted common name, umbrellagrass (Fuirena hispidal), common reeds (Phragmites communis), and giant reedgrass (Arundo donax). These latter reeds and reed-like plants are common on bars in the river, along shelves at the foot of steep canyon walls, and tend to border much of the stream on both sides as soil accumulations will permit. The most common trees along the Rio Grande are the cottonwood (Populus palmeri). quaking aspen (Populus tremuloides), and several Species of willows (Salix nigra; Salix interior; Salix gooddingii; and Salix taxifolia). Ferns, Polypodiacae, are common in a few spring areas. Beaver (Castor canadensis) are common along the clear water associations, and were very detrimental to nets used in sampling fish populations. Nutria (Myocasto; coypus) are reported in lower regions of the watershed. Species Present and Their Distribution: Several problems were encountered in establishing distribution of species. Decisions, involving the inclusion in the river‘s fauna of species that appeared to be endemic to springs near the mouths of tributaries, were required. In other instances, a particular species appeared to have temporarily expanded its distribution as a result of specific and unusual conditions that could not be eXpected to prevail for any extended period. Conformity of distribution, as governed by basic stream ecology, had to be considered. Consideration was also given to whether Species introduced by hatcheries and bait dealers Should be included. Field work was adversely affected by several circumstances. Flooding pre» vented effective seining during much of the time spent making collections. No accurate and reliable meanscxftxedetermining,stream conditions were discovered. This was because of the inaccessable nature of the stream, the lack of knowledge of the extent of rainfall in Mexico, and because of the length of time required to reach the stream from the regional headquarters. Flooding occurred during all field work with one exception. When the frequency of occurrence of flooding is taken into account, the degree of misfortune can be apparent. Much seining effort was rendered unproductive by the expansion of the stream during flooding on shore areas that were heavily vegetated. Counts of fish taken under the limiting circumstances are regarded as being entirely misleading. Because of this, the abundance of minnows and several other species is best expressed in discussion. The seining Stations worked and a description of their location are included in Figure 8. Annotated Checklist of Species of Fish Occurring in the Chihuahuan or Mid=Rio Grande River ‘Lepisosteidae - Cars Alligator gar (Lepisosteus Spatula) were common but not abundant and were taken as far west as Presidio. They were much more common in lower stream areas after permanent turbidity is apparent. Spotted gar (L, oculatus), whose presence was questioned and identity unw confirmed, was represented by a single fish from the mouth of the Pecos River. It was too large to be preserved with the ordinary equipment present when this collection was made. Longnose gar (L, osseus) was the most-common species of this group through= out the stream, and was taken in abundance as far west as the mouth of the Concho River. This Species probably also occurs in a few localities in the intermittent pools above, but confirmation from netting is not yet obtained. Clupeidae n Herrings Gizzard shad (Dorosoma cepedianum) probably were the most common and rela- tively abundant species of the stream. They were less numerous but of greater size in clear water associations. Caracidae — Tetras Mexican tetra (Astyanax mexicanus) were very abundant near the mouth of the Devil's River, in San Felipe Springs, near the mouth of the Pecos River, and often abundant in other Spring areas. They are probably located throughout the stream and tributaries as a result of introductions by bait dealers. None were taken west of the mouth of the Pecos River. Catostomidae a Suckers and Buffalofishes Blue sucker (Cycleptus elongatus) was found primarily in clear water assom ciations as far west as Presidio. They were relatively abundant in a few locale ities, but never as dominating as other species of this group. Smallmouth buffalo (Ictiobus bubalus) were taken as far west as Presidio. Young fish are abundant in BiglknuiNational Park, but never abundant in the other stream areas. Reproduction was indicated in riffle areas. River carpsucker (Carpiodes carpio) were very abundant and the dominating species in intermittent pool associations and in turbid waters below Regan Canyon. It was encountered less frequently in clear waters with rapid flow where Similar species appear to dominate. It was the most problematical species in the stream. Gray redhorse (Moxostoma congestum) was common but seldom found in abundance, and was taken as far west as Presidio. It cedes dominance to other suckerutype fish in intermittent pool and turbid associations, but may be regarded as being in contention in clear rapid flow waters. It may occasionally be dominant in this latter association. Cyprinidae w Shiners and Minnows Carp (Cyprinus carpio) was found throughout the watershed but much more abundant in intermittent pools and in turbid waters. Possibly the most adapts able single Species, it is regarded as being the second most problematic Species in the watershed. Golden Shiner (Notemigonus crysoleucas)was present and known to be intro» duced from hatcheries. It was prevalent in small reservoirs such as Lake Ascarate and those of the Diablo tributary group. Speckled chub (Hybopsis aestivalis) was apparently rare and not actually taken during this inventory. It was included in this region by Hubbs. Rio Grande chub {Gila pandora), a riffles species, was found only in clear water associations near the mouth of the Devil's River and the Pecos River. Rio Grande Shiner (Notropis jemezanus) was common but of questionable abundn ance, and apparently located only east of the Big Bend. Chihuahua (E, chihuahua) abounds from the mouth of the Concho River to the mouth of Santa Helena canyon. Tamaulipas Shiner (E, braytoni) was common and in localities relatively abundant, especially in Big Bend National Park. Red Shiner (E, lutrensis), the most widely spread species of minnow, was usually much less numerous than is common in tributaries. Roundnose minnow (Dionda episcopa) was common and very abundant in clear water tributaries, but not collected west of the Pecos. Devil's River minnow (Q, diaboli) was found exclusively in the Devil‘s River. ' Fathead minnow (Pimephales promelas) was common and abundant as far west as the Big Bend. Mexican stoneroller (Campostoma ornatum), common but never numerous, was taken primarily near the mouth of Terlingua Creek and near Lajitas. Ameiuridae - Freshwater Catfishes Channel catfish (Ictalurus punctatus) was common but rarely abundant, and appears to dominate other species of this genus in intermittant pools and occasionally in very turbid waters. It.yields dominance in clear water associa- tions. Blue catfish (l, furcatus) was probably the most important game Species in the river and definitely So in clear water associations of the stream. This species tends to concentrate in clear pools below riffles and is particularly abundant near the mouth of the Pecos River and near the mouth of SantaElena Canyon. It was regarded as being the most beautiful and desirable game fish of the catfish group. Flathead catfish (Pylodictis olivaris) was common and relatively abundant for the Species in a few localities. It is considered important commercially. Cyprinodontidae ~ Killifishes and Topminnows Rainwater killifish (Lucania parva) is primarily a Pecos River species, but believed to also be present in saline waters of other tributaries of the Rio Grande, and possibly in a few other specific areas of the river. Rio Grande killifish (Fundulus zebrinus) occurs in a few backwaters in the Big Bend area, and may occur near the mouth of other tributaries that offer saline waters. Poeciliidae - Mosguitofishes Big Bend gambusia (Gambusia gaigei) was found only in Graham Ranch Spring in the Big Bend National Park.