Distribution and Relative Abundance of Fish in Ruth Reservoir, California, in Relation to Environmental Variables

Great Basin Naturalist

Volume 42 Number 4 Article 5

12-31-1982

Distribution and relative abundance of fish in Ruth Reservoir, California, in relation to environmental variables

Steven Vigg Desert Research Institute, Reno, Nevada

Thomas J. Hassler Humboldt State University, Arcata, California

Follow this and additional works at: https://scholarsarchive.byu.edu/gbn

Recommended Citation Vigg, Steven and Hassler, Thomas J. (1982) "Distribution and relative abundance of fish in Ruth Reservoir, California, in relation to environmental variables," Great Basin Naturalist: Vol. 42 : No. 4 , Article 5. Available at: https://scholarsarchive.byu.edu/gbn/vol42/iss4/5

This Article is brought to you for free and open access by the Western North American Naturalist Publications at BYU ScholarsArchive. It has been accepted for inclusion in Great Basin Naturalist by an authorized editor of BYU ScholarsArchive. For more information, please contact [email protected], [email protected]. DISTRIBUTION AND RELATIVE ABUNDANCE OF FISH IN RUTH RESERVOIR, CALIFORNIA, IN RELATION TO ENVIRONMENTAL VARIABLES'

Steven Viifg- Hassler' and Thomas J.

Abstract.— The fish population of Ruth Reservoir, Cahfornia, was sampled every two weeks with variable mesh gill nets from May 1974 through May 1975. Fish were captured in the following order of numerical abundance: Humboldt sucker (Catostomus hurnboldtianus), golden shiner {Notemigonus crysoleucus), brown bullhead [Ictahtrus nebulosus), white catfish (/. catus), rainbow trout {Sahno gairdneri), and largemouth bass {Micwpterus sabnoides). The three most abundant species made up about 95 percent of total numbers and weight. All species exhibited a similar cyclic temporal availability pattern: catch rates increased to a maximum during summer and fall and de- creased during winter and spring. Environmental variables with the most pronounced relationships to fish catches were temperature (direct) and turbidity (inverse).

Information on Ruth Reservoir fish ecology Study Area collected prior to this study was limited; data consisted of stocking records, yearly creel Ruth Reservoir is impounded behind R. W. survey data on opening weekends of the fish- Matthews Dam, near the headwaters of the ing season, five gill net sets during November Mad River in Trinity County, California (Fig.

1968, and the results of a reward tagging pro- 1). This water supply reservoir, about 127 km gram for salmonids during May 1972 (Ruth by river from the Pacific Ocean, provides Reservoir file, California Department of Fish municipal and industrial water for the Hum- and Game, Eureka). Management measures boldt Bay Area. The dam was completed in have consisted primarily of stocking hatch- 1961 and is operated by Humboldt Bay Mu- ery-reared salmonids. Unauthorized in- nicipal Water District (HBMWD). The reser- troductions of exotic species into the reservoir has a maximum surface area of 445.2 ha, voir have also been made. a maximum storage capacity of 63.9 million Alterations to the dam have been proposed cubic meters, a mean depth of 14.4 m at that would affect the physical and chemical maximum pool, and a minimum discharge of characteristics of the lake and thus the aquat- 142 liters per second. ic organisms, specifically the fish populations. Annual water level fluctuations have The present dam may be modified or re- ranged from 9.8 to 15.5 m, with a mean fluc- placed by a larger structure to meet future tuation of 12.6 m (HBMWD, unpublished water needs (U.S. Army Corps of Engineers data). The water level is usually lowest in 1973). Air-induced circulation and a multi- November and highest in January. The high- level discharge structure have been proposed est recorded water level, 5.8 m above to reduce downstream turbidity (Winzler and spillway elevation, was on 22 December Kelly 1975). 1964, and the lowest, 13.7 m below spillway The objectives of this study were to deter- elevation, on 29 November 1967. Water level mine the relative abundance and distribution fluctuated 14.4 m during the study. of fish in Ruth Reservoir and determine their The reservoir is 11.3 km long at full pool relation to environmental variables. and has a mean width of 0.6 km (Winzler and

'Cooperators include the California Department of Fish and Game and the United States Fish and Wildlife Service. 'Bioresources Center, Desert Research Institute, P.O. Box 60220, Reno, Nevada 89506. 'California Cooperative Fishery Research Unit, Humboldt State University, Areata, California 95521.

529 530 Great Basin Naturalist Vol. 42, No. 4

California DAM ! N

RUTH RESERVOIR

SCALE

KILOMETERS

MAD RIVER Fig. 1. Gill net sampling stations in Ruth Reservoir, California

Kelly 1975). Several small tributaries flow ishes by late spring as the suspended particles into the reservoir, but the major inflow is settle and thermal stratification confines sus- from the Mad River, which has a watershed pended sediments to the bottom zone, thus of 30,822 ha above the dam (Iwatsubo et al. developing a turbid density current (Winzler 1972). The dominant geological feature of and Kelly 1975). Persistent turbidity occurs the watershed is the Franciscan Formation. in the reservoir and downstream from the Heavy precipitation, steep slopes, and un- reservoir's bottom discharge. stable geology have resulted in high erosion Surface water temperatures range from in this area (Yoimg 1971). The primary influx to 26.7 C; the minimum generally occurs in of sediment corresponds to major precipi- December or January, and the maximum in tation from November through April. During July or August. Bottom temperatures range the rainy season, large amoimts of fine sus- from less than 4 to as high as 17 C. The reser- pended sediment are distributed throughout voir has characteristic spring and fall over- the mixed reservoir; surface turbidity dimin- turns of a dimictic lake. Dissolved oxygen December 1982 ViGG, Hassler: Ruth Reservoir Fish 531

(DO) ranges from saturation to seasonal an- Limnological data were obtained during aerobic bottom deficits (California Depart- each sampling period at each station. Tem- ment of Water Resources, Red Bluff, 1969). perature, turbidity, conductivity, and DO were measured at limnetic stations corre- Methods sponding to the gill net stations. Water samples were taken with a 2-1 water bottle, 1 m Sampling was conducted at two-week in- below the surface, at middepth, and 1 m tervals from May 1974 to May 1975; 26 sam- above the bottom. Immediately upon bring- ples— 7 each in summer and fall and 6 each in ing the sample to the surface we measured winter and spring—were analyzed by season. temperature with a mercury bulb thermome- The seasons were defined as follows: summer, ter, or the thermistor of the DO meter. A 1 June to 31 August; fall, 1 September to 1 bathythermograph was used to measvire December; winter, 2 December to 2 March; depth-temperature profiles. Turbidity, Jack- and spring, 3 March to 31 May. Five gill net son Turbidity Units (JTU), was measured sampling stations were established from the with a Hach Model 1860 Turbidimeter. A dam to reservoir headwaters (Fig. 1). Beckman Solu Bridge was used to measure Fish populations were sampled with bot- electroconductivity, recorded as micro mhos tom set, variable mesh gill nets 1.83 X 54.86 per centimeter (/imho/cm) at 25 C. Dissolved m comprising six 9.14-m panels of the follow- oxygen determinations were made with a ing mesh sizes (bar measure): 1.27, 1.91, 2.54, Hach Model CA-10 DO kit (June through 3.18, 3.81, and 6.35 cm. All mesh sizes were October) and a Delta Scientific Model 85 DO made of nimiber 104 multifilament white ny- Meter (November through May). Surface and lon except the 6.35-cm mesh, which was discharge temperature, reservoir surface ele- number 139. vation, inflow, and discharge data were ob- The nets were set in the late evening and tained from HBMWD records. fished overnight for 12 to 16 hours. Fish To detect significant differences in hori- catch was adjusted to a standard 12-hour set. zontal and seasonal fish distribution, and in- The net was anchored in approximately 2 m teraction between reservoir area and season, of water at the inshore end and set per- we analyzed the catch-per-unit-of-effort data pendicular to shore. The end of the net by using a two-way nested analysis of vari- placed closest to shore was randomized. Each ance design computer program. Fish relative gill net panel was marked with a painted ver- abundance was analyzed by sampling station tical stripe to give two replicates for each and season. One-way analysis of variance set. Fish catch from the right and left halves (Sokal and Rohlf 1969) was used to analyze of each mesh size was recorded separately seasonal differences in mean fish catch at Sta- and randomly assigned to one of two derived tion 5. This station was dewatered by sea- replicates. The data could thus be treated as sonal low water and was not included in the overall replicate 27.43 m variable mesh nets in each analysis. Fish catch data were transformed: location at each time, enabling the use of a nested analysis of variance design. (logio(Y + 1)) where Y = fish catch.

Table 1. Gill net catches at five sampling stations in Ruth Reservoir from May 1974 through May 1975. 532 Great Basin Naturalist Vol. 42, No. 4

CUBIC REGRESSION EQUATION O CALCULATED VALUES

Y=-IOI.2IO-h79.743X -6.477XX*-l-0.-1-0. 14 1 X • ACTUAL DATA {r^= ..792)

300 I I WINTER SPRING

200

I 13 24 9 23 6 23 JUNE JULY 1974

Fig. 2. Relationship between time and total fish catch in Ruth Reservoir, June 1974 through May 1975. December 1982 ViGG, Hassler: Ruth Reservoir Fish 533

Humboldt Sucker 75 r Golden Shiner

Brown Bullhead

White Catfish

Sumnner Fall Winter Spring

Fig. 3. Mean fish catch by species and season in Ruth Reservoir, 1974-1975.

about 45 percent of the fish were taken at with respect to season and station. The mean the lower three stations. Catches of Hum- catch of white catfish differed significantly boldt suckers, golden shiners, and brown bull- by season (?< 0.001) but not by station. A heads illustrate this trend of higher abun- significant interaction for catches of brown dance at upper reservoir stations. Catches of bullhead by season and station (P<0.01) in- white catfish were more evenly distributed dicated that seasonal distribution was not throughout the reservoir; but were slightly consistent on a spatial basis. There was a higher in the middle and lower than in the large difference in seasonal mean catches of upper reservoir areas. bullhead in the upper end of the reservoir, Himiboldt suckers were generally more but catches in the lower end were con- abundant in catches than other species during sistently low and not greatly different. There all seasons. However, during summer and fall was no significant interaction for Humboldt catches of brown bullhead were highest. sucker, golden shiner, white catfish, and total Golden shiner catches were relatively high catch. Thus, mean seasonal catch for these during summer at Station 5 and during spring species was independent of lake area effects at Stations 2 and 4. White catfish were the and, conversely, differences in area fish catch second or third most abundant species in the were significant, regardless of season. catch at the lower end of the reservoir (Sta- There was a significant difference in rep- tions 1 and 2) during all seasons except win- licates within season and station for all spe- ter, when only one was captured in the entire cies (P< 0.005). Seasons were not biologically lake. discrete units of time, i.e., temporal trends of Temporal and spatial fish distribution pat- fish catch existed within seasons. terns were summarized by analysis of vari- At station 5 total species mean catch was ance. Mean catches of Humboldt suckers, significantly higher for the summer-fall than golden shiners, brown bullheads, and total for winter-spring (P< 0.001, Table 3). There species were significantly different (F<0.01) was a significant difference in mean catch of 534 Great Basin Naturalist Vol. 42, No. 4 brown bullhead (P< 0.001) and golden shiner upper end of the reservoir, bottom DO de- (P<0.05) between the two time periods. creased there to 2 mg/1. Bottom DO was de- Catches of both species were higher during pleted in midreservoir in late August and in summer and fall than during winter and the lower end in September (Fig. 4). This DO spring. There was no significant difference in depletion trend probably indicated either the catches of Humboldt sucker and white movement of the low-oxygen water mass catfish between the two time periods. Tem- down reservoir or differential in-place bot- poral catch trends within the time periods tom DO depletion, or both. Destratification are indicated by the significant differences in and mixing in the upper end of the reservoir replicates for all species except the white cat- in late August resulted in high DO concen- fish, which was scarce in catches during both trations (10 mg/1) throughout the water col- periods. umn. Destratification and reoxygenation of the midreservoir area took place during late September, and by October the entire reser- Environmental Variables voir was well mixed. Dissolved oxygen con- Fish catch and limnological data were tab- centrations were near saturation levels for ulated by sample period, station, and depth the rest of the year. (Vigg 1979). Seasonal variations of environ- Definite seasonal variation in turbidity oc- mental parameters were pronounced (Table curred. Surface turbidity was highest (max-

4). Temperature was highest during August imum of 79 JTU during February at Station (maximum 27.0 C), and lowest in late De- 4) during the winter and spring, when high cember (1.0 C). The surface DO concentra- rainfall, runoff, and erosion resulted in large tion was never below 8.0 mg/1. Bottom DO amounts of suspended sediments in the lake. deficits occurred during August and Septem- Inorganic suspended sediments persisted in ber. During early August, when maximum the bottom zone of the lake throughout sum- annual water temperatures occurred in the mer. Turbidity was lowest during fall, when

Table 2. Percentage fish catch (adjusted to 12-h set) by station and season in Ruth Reservoir, June 1974 through May 1975.

Season and December 1982 ViGG, Hassler: Ruth Reservoir Fish 535

all suspended sediments had been flushed total fish catch. This pattern was consistent from the reservoir. Turbidity was highest at for all major fish species. Turbidity and tem- the bottom and lowest at the surface during perature accounted for 72.2, 53.5, and 58.8 all seasons. Lake area effects also introduced percent of the catch variation for Humboldt considerable variation in turbidity— i.e., spa- suckers, brown bullheads, and white catfish, tial trends in turbidity occvirred as storm nm- respectively. A significant (P<0.05) relation- off moved through the reservoir. ship also existed between the turbidity-tem- Conductivity varied with time and with perature system and the catch of golden shi- vertical and horizontal lake area. However, ners. However, the proportion of catch variation was not great, the values ranging variation explained by temperature and tur- only from 80 to 200 jumho/cm (mean, about bidity—about 30 percent— was substantially 125 jumho/cm). less for the golden shiner than for the other species. In all tests, inclusion of additional environmental variables did not account for a Simple and Multiple statistically significant proportion of inde- Linear Correlations pendent variation. The catches of largemouth bass and rainbow trout were so small that Correlations between fish catch and con- correlation analyses would not be current measurements of environmental vari- meaningful. ables at specified stations indicated that temperature and turbidity had major effects on Discussion fish catches (Table 5). Consistent significant direct temperature and inverse turbidity rela- The fish population dynamics of the reser- tionships with the catches of Humboldt suck- voir have not been continuously monitored er, brown bullhead, white catfish, and total since the dam was completed in 1961. How- species occurred. ever, current evidence does suggest that es- Although a significant inverse relationship tablishment of nonnative species in the late existed between fish catches and DO, there 1960s was associated with a decline in the was no discernible biological basis for a rainbow trout population. Introductions of cause-effect relationship of this type; i.e., in- the golden shiner, brown bullhead, white cat- creased DO concentrations would not be ex- fish, and largemouth bass were unauthorized. pected to cause a decrease in fish catches. Golden shiners were first observed during The range of DO saturation variation was not great, and DO concentrations measured at Table 3. Gill net catches (adjusted to 12-h set) at Sta- the water depths of net sets were not limiting tion 5 during summer-fall (s-f) and winter-spring (w-s) in to fish. Since there was generally a high cor- Ruth Reservoir June 1974 through May 1975. relation (r>0.90) between temperature and

DO concentration, it is reasonable to assume Species, and that the fish-DO correlation is a result of the seasonal period indirect temperature effect. Conductivity and fish catches were not consistently related. Environmental variables with biologically explainable effects on fish catch were included in multiple linear correlations with

fish catch (Table 6). Significant (P<0.01) multiple linear correlations existed between total and individual species catch and the turbidity-temperature environmental system. Surface turbidity and bottom temperature accounted for 80.5 percent of the variation in total fish catch. Time of year, depth of sampling station, and Mad River inflow explained very little additional variation in No. 4 536 Great Basin Naturalist Vol. 42,

factor detrimental to the summer and fall 1968 (La Faunce 1968), and could have been a brown bullhead and white catfish during fall reservoir trout population. Erman (1973) re- of {Catostomus ta- 1968. Largemouth bass are believed to have ported that populations been introduced later— possibly in 1970. hoensis) and (C. platijrhynchus) in Sagehen from 17.8 percent Changes in relative abundance of adult fish Creek increased impoundment to 41.3 of different species are apparent from com- (1952-1961) before Reservoir and 79.2 per- parisons of gill net samples taken in 1968 percent in Stampede impound- with those taken during the present study. cent upstream (1970-1972) after stream-reser- Rainbow trout composed 31 percent of the ment. This illustrates that this sucker populations. catch in 1968, but less than 1 percent in voir system favored the 1974-1975. The relatively high trout catch in Since Humboldt suckers spawn in Mad season as rain- 1968 probably represents a population that River during the same (spring) young of the remained from the stocking of hatchery- bow trout, it is likely that the for space and food. At reared fish in the previous May and the resi- two species compete reproduction of trout in dent river population entrapped by the dam. present, the natural Reservoir appears Corresponding to the dramatic difference in the Mad River above Ruth trout catches were the substantial differences negligible. trout fishery of the reservoir in catches of golden shiners (from 15 to 33 The rainbow stocking fingerling percent), brown bullheads (from to 20 per- has been maintained by catchable-size fish cent), and white catfish (from a trace to 3 (1962-1968) and these hatchery-raised percent). Humboldt suckers made up 54 per- (1969-1975). Since unsuccessfully with other res- cent of the catch in 1968 and 43 percent in trout compete stocking 1974-1975. The largemouth bass maintains a ervoir species, the recent strategy of times of heavy angler naturally reproducing population and sup- catchable trout during the opening weekend of ports a sizable fishery in the reservoir. It was effort (i.e., before holidays) on a probably more abundant in 1974-1975 than the fishing season and before basis is logical. However, if the gill net samples indicated (<1 percent) put-and-take were plan- because centrarchids are typically difficult to trout of a more predaceous strain size, they would be able to capture in nets. Crayfish, which were very ted at a larger shiners. abundant in the 1968 sample, were present forage on golden availability and spatial dis- only in trace amounts in 1974-1975. Both temporal Reservoir were Western suckers and golden shiners are tribution of the fish in Ruth environmental properties two of the most successful competitors of associated with basis. A cyclic trend rainbow trout in terms of reduced trout pro- that varied on a seasonal the warm summer and duction in California reservoirs (Inland Fish- of high catches during catches during the cold and eries Branch 1971). Humboldt sucker and fall, and low spring was apparent for all golden shiner composed over 75 percent of rainy winter and environmental-fish relationships the sample in numbers and 83 percent in species. The weight during 1974-1975. Thus, the quantified during this study were simple; i.e., and turbi- 1974-1975 species composition and relative temperature was directly related, catches. Both temper- abundance of nongame fish in Ruth Reservoir dity inversely, to fish

below the surfaces (S), mid-depth (M), and Im above Table 4. Mean seasonal environmental measurements at Im 1975. the bottom (B) at Stations 1-4 in Ruth Reservoir, June 1974 through May December 1982 ViGG, Hassler: Ruth Reservoir Fish 537

Temperature Dissolved Oxygen s= Surface

M=Middepth Station I. Station 2 O B= Bottom 25 Station 3. 25 Station 4. a> o

Q. E 20 20

15 15

- 10 _ 10 E c Q) o» OX

a> _> o (O CO

August August September September 6 23 4 19 Fig. 4. Bottom-dissolved oxygen concentration and vertical temperature profile bv station in Ruth Reservoir Au- gust and September 1974. atiire and turbidity have known biological ferred temperature of the two most abundant relationships affecting the survival, phys- fish species—Humboldt sucker and golden iology, and behavior of fish. shiner (Reutter and Herdendorf 1974). The The mean summer temperature in Ruth maximum surface temperature of Ruth Reser- Reservoir (22.2 C) approximates the pre- voir (27 C) exceeds the upper lethal threshold 538 Great Basin Naturalist Vol. 42, No. 4

environmental variables measured at Table 5. Significant correlation coefficients between fish catch and selected 1-4 Reservoir from 1974 through May 1975 (26 surface (S), mid-depth (M), and Bottom (B) at Stations in Ruth June observations per station, n= 104). December 1982 ViGG, Hassler: Ruth Reservoir Fish 539

Table 5 continued. I

540 Great Basin Naturalist Vol. 42, No. 4

Table 6. Multi