CAUFDRNIAI FISH-GAME I "CONSERVATION OF WILDLIFE THROUGH EDUCATION" f VOLUME 66 APRIL 1980 NUMBER 2 J u ^joy m-- ^■Jl jjI iJ California Fish and Game is a journal devoted to the conservation of wild- life. If its contents are reproduced elsewhere, the authors and the California Department of Fish ond Game would appreciate being acknowledged. Subscriptions may be obtained at the rate of $5 per year by placing an order with the California Department of Fish and Game, 1416 Ninth Street, Sacramento, California 95814. Money orders and checks should be made out to California Department of Fish and Game. Inquiries regarding paid sub- scriptions should be directed to the .Editor. Complimentary subscriptions are granted, on a limited basis, to libraries, scientific and educational institutions, conservation agencies, and on exchange. Complimentary subscriptions must be renewed annually by returning the post- card enclosed with each October issue. Please direct correspondence to: Kenneth A. Hashagen, Jr., Editor California Fish and Game 1416 Ninth Street Sacramento, California 95814 u D VOLUME 66 APRIL 1980 NUMBER 2 Published Quarterly by STATE OF CALIFORNIA THE RESOURCES AGENCY DEPARTMENT OF FISH AND GAME —LDA— STATE OF CALIFORNIA EDMUND G. BROWN JR., Governor THE RESOURCES AGENCY HUEY D. JOHNSON, Secretary for Resources FISH AND GAME COMMISSION ELIZABETH L. VENRICK, President Cardiff ABEL GALLETTI, Vice President SHERMAN CHICKERING, Member Los Angeles San Francisco NORMAN B. LIVERMORE, JR., Member RAYMOND DASMANN, Member San Mateo Nevada City DEPARTMENT OF FISH AND GAME E. C. FULLERTON, Director 1416 9th Street Sacramento 95814 CALIFORNIA FISH AND GAME Editorial Staff KENNETH A. HASHAGEN, JR., Editor-in-Chief Sacramento DARLENE A. OSBORNE, Editor for Inland Fisheries Sacramento RONALD M. JUREK, Editor for Wildlife Sacramento J. R. RAYMOND ALLY, Editor for Marine Resources Long Beach DAVID A. HOOPAUGH, Editor for Salmon and Steelhead Sacramento DONALD E. STEVENS, Editor for Striped Bass, Sturgeon, and Shad Stockton KIM McCLENEGHAN, Editor for Environmental Services Rancho Cordova CONTENTS 67 Page Estimated Safe Copper and Zinc Levels for Chinook Salmon, Oncorhynchus tshawtscha, in the Upper Sacramento River California B. J. Finlayson and K. M. Verrue 68 Age and Growth of White Sturgeon Collected in the Sacra- mento-San Joaquin Estuary, California: 1965-1970 and 1973-1976 .... David W. Kohlhorst, Lee W. Miller, and James J. Orsi 83 Diets of Mountain Whitefish, Prosopium w////amson/ {G\rard) , and Brook Trout, Salvelinus fontinalis (Mitchell), in the Little Walker River, Mono County, California John Paul Ellison 96 The Use of Road-killed Deer for Assessing Reproductive Potential and Winter Condition of the Devil's Garden Interstate Mule Deer Herd Hal Salwasser, Douglas j. Thayer, and D. Lewis Young 105 The Littoral Bottom Fauna of High Elevation Lakes in Kings Canyon National Park Timothy P. Taylor and Don C. Erman 112 Notes A Blue Catfish from the Sacramento-San Joaquin Delta Thomas L. Taylor 120 First California Record: The Scalloped Hammerhead Shark, Sphyrna lewini, in Coastal Santa Barbara Waters Craig Fusaro and Shane Anderson 121 The Bigscale Goatfish, Pseudupeneus grandisquamis (Gill, 1863), Added to California's Marine Fauna John E. Fitch 123 Notes on the Establishment of the Rainwater Killifish, Lucania parva, in California Michael J. McCoid and James A. St. Amant 124 Book Reviews 126 68 CALIFORNIA FISH AND GAME Calif. Fish and Game 66 ( 2 ) : 68-82 1 980 ESTIMATED SAFE ZINC AND COPPER LEVELS FOR CHINOOK SALMON, ONCORHYNCHUS TSHAWYTSCHA, IN THE UPPER SACRAMENTO RIVER, CALIFORNIA ^ B. J. FINLAYSON AND K. M. VERRUE California Department of Fish and Came Water Pollution Control Laboratory 2005 Nimbus Road Rancho Cordova, California 95670 Long-term (83-d) and short-term (96-h) toxicity tests with combined copper and zinc solutions were conducted on chinook salmon, Oncorhynchus tshawytscha, eggs, alevins, and swim-up fry to determine tolerance levels to copper and zinc in the acid-mine waste from Spring Creek, a tributary to the Sacramento River. The ratio of copper-to-zinc in the waste can vary between 1:2 and 1:12. Eggs were just as tolerant or more tolerant than alevins and fry to combined concentrations of copper and zinc. The 83-d LC50 values (median lethal) for the period of eggs-to-fry to 1:3, 1:6, and 1:11 copper-to-zinc ratio stock solutions were 44 jag/l Cu and 160 /ag/l Zn, 27 ju.g/1 Cu and 206 jug/l Zn, and 17 jag/l Cu and 253 /ng/l Zn, respectively, demonstrating that the toxicities of copper and zinc were "additive". The 96-h LC50 values for fry not previously exposed to the stock solutions were 37 \x%l\ Cu and 132 /Ag/I Zn, 29 jag/l Cu and 213 ju.g/1 Zn, and 20 /xg/l Cu and 279 jag/l Zn, for the same copper-to-zinc ratio stock solutions, respectively, demonstrating that an acclimation to these metals had occurred in the 83-d tests during embryonic and alevin development at the higher (1:3) copper-to-zinc ratio but not at the lower (1:6 and 1:11) copper-to-zinc ratios. Based on reduced growth of exposed fry for the period of eggs-to-fry, estimated "safe" levels of copper and zinc for chinook salmon would be below 11 and 83 jag/ 1, respectively. INTRODUCTION The acid-mine waste from the Spring Creek drainage (Figure 1 ) has caused numerous kills of anadromous and resident salmonid fishes in the upper Sacramento River between Keswick Reservoir and Cottonwood Creek, California (U.S. Fish and Wildlife Service 1959; Prokopovich 1965; Nordstrom 1977). Of the fishes affected by the waste, anadromous chinook salmon are of particular concern since they are both a recreationally and commercially valuable species, and the carrying capacity of the spawning areas in the upper Sacramento River below Keswick Dam fail to reach their anticipated levels. Since 1963, the waste has collected in Spring Creek Reservoir and been metered into Keswick Reservoir. The flow of the waste through the Spring Creek Diversion Dam is determined by the amount of "dilution" water originating from Shasta Lake. From 1963 through 1978 this water quality management program was based on dilution factors calculated by Lewis (1963) using waste copper concentrations as criteria. Finlayson and Ashuckian (1979) listed six theoretical reasons why these dilution factors had not provided for adequate protection of salmonid resources in the upper Sacramento River and further demonstrated that these factors could potentially allow very toxic concentrations of zinc to enter the Sacramento River. The factors did, however, provide borderline protection for steel head trout, Salmo gairdneri, against acute copper toxicity. ^ Accepted for publication September 1979. ESTIMATED SAFE COPPER AND ZINC LEVELS 69 BOULDER CREEK SHASTA LAKE SCALE FIGURE 1. Location of Spring Creek drainage, upper Sacramento River basin, California Finlayson and Ashuckian (1979) investigated the toxicity of the Spring Creek waste to steelhead trout eggs, alevins, and swim-up fry, and concluded that the toxicity of the waste was entirely due to its copper and zinc content and not to the other metals (aluminum and iron) present. Much information is available 70 CALIFORNIA FISH AND GAME regarding the toxicities of zinc (Nehring and CoettI 1974; Sinley, GoettI, and Davies 1974; Lorz and McPherson 1977; Chapman 1978; Chapman and Stevens 1978; Lorz, Williams, and Fustish 1978), copper (Hazel and Meith 1970; McKim and Benoit 1971; Chapman and McCrady 1977; Lorz and McPherson 1977; Chapman 1978; Chapman and Stevens 1978; Lorz et al. 1978), and their combinations (Lloyd 1961; Sprague 1964; Sprague, Elson, and Saunders 1965; Sprague and Ramsey 1965; Thomsen, Hazel, and Meith 1970; Lorz et al. 1978) to salmonids. However, no one has investigated the combined toxicities of copper and zinc to chinook salmon eggs, alevins, or swim-up fry. It is the dissolved fractions of zinc (Sinley et al. 1974) and copper (Chapman and McCrady 1977) (Zn ++ and Cu ++ in acid waters; ZnOH+ and CuOH+ in soft waters), which are believed to produce toxic effects in fish (Davies et al. 1979). Much of the copper contained in the Spring Creek waste now can be removed by several copper cementation plants located at the mine-adit openings (Nordstrom 1977). However, during periods of precipitation, the leachate from exposed tailings results in an increase in the ratio of copper to the other metals present. Hence, the copper-to-zinc ratios of the waste leaving the Spring Creek Diversion Dam fluctuate between 1:2 and 1:8 for wet and dry periods, respectively (D. Wilson, Calif. Dept. of Fish and Game, pers. commun.). Successful operation of the copper cementation plants could result in ratios as low as 1:12 during future dry periods (D. Wilson, pers. commun.). The other metals in the waste remain in the same ratio to each other throughout the year even though their concentrations decrease during storm periods and increase during dry periods (B. Finlayson, unpublished data, 1979). Wilson (1978) developed new interim release schedules for the Spring Creek diversion dam to replace those of Lewis (1963). The new schedules use ratios of copper-to-zinc as criteria for releases and should eliminate the potential of toxic zinc concentrations in the upper Sacramento River. Recent investigations (Finlayson and Ashuckian 1979) have supported this new approach. However, additional toxicological information is needed before new Sacramento River Basin Plan copper and zinc objectives and permanent Spring Creek release schedules can be calculated and implemented by the State Regional Water Quality Control Board. The objectives of this study were to gather accurate and relevant toxicological information on copper-to-zinc ratios that were not considered by Lewis (1963), to evaluate the new release schedules of Wilson (1978), and to complement the toxicity data on steelhead trout (Finlayson and Ashuckian 1979). MATERIALS AND METHODS We tested the toxicities of combined copper and zinc solutions rather than the acid-mine waste from Boulder Creek (Figure 1) which was done by Finlayson and Ashuckian (1979). This alternative was toxicologically valid because the toxicity of the waste had been demonstrated to be due entirely to its copper and zinc content. The combined copper and zinc stock solutions were prepared using reagent grade copper (CUSO4.5H2O) and zinc (ZnS04-7H20) sulfates in de-ionized water. All of the metals present in the Spring Creek acid-mine waste are in the form of sulfates (Nordstrom 1977). Continual-flow acute and chronic toxicity tests using the methods of Peltier (1978) and Finlayson and Ashuckian (1979) were carried out at the ESTIMATED SAFE COPPER AND ZINC LEVELS 71 Department's Water Pollution Control Laboratory (WPCL) with freshly fertilized chinook salmon eggs and the resulting alevins and swim-up fry. All concentrations were tested in replicate. Water from the American River, a tributary to the Sacramento River, was sand-filtered and used for dilution. Since the American and Sacramento rivers are of similar pH and hardness, similar dissolved fractions of both metals would be expected in the Sacramento River and in our toxicity test water. Three copper-to-zinc ratio (1:3, 1:6, and 1:11 ) stock solutions were used in both chronic (83-d) and acute (%-h) toxicity tests. The concentrations of zinc in the solutions were held constant at approximately 1525 mg/l and the copper concentrations were adjusted by adding copper to the appropriate ratio. The solutions were acidified to pH 1 .8 with sulfuric acid ( H2SO4) to keep the metallic ions in solution and to keep the metals from plating-out on the walls of the 20-1 glass toxicant reservoirs. During the tests, the stock solutions were replenished every 7 to 14 d, and each new solution was analyzed for copper and zinc as a quality control measure (Table 1 ). TABLE 1. Chemical Composition of Copper and Zinc Stock Solutions Used in Toxicity Tests. Copper-to-zinc ratios Metal 1:3 1:6 1:11 Copper (mg/l) 548 + 21° 266±12 145±10 n=8 n==13 n = 15 Zinc (mg/l) 1503 + 61 1543±101 1543±75 n=8 n = 13 n = 15 " Mean + SD The solutions were diluted in a geometric series of six concentrations (100, 56, 32, 18, and 10%, and control) and continually delivered to the developing eggs, alevins, and fry using the predilution systems and modified Mount and Brungs (1967) type proportional diluters designed and constructed by Finlayson and Ashuckian (1979). Immediately prior to entry into the proportional diluters, the concentrated stock solutions were prediluted to 0.042, 0.049, and 0.081% of original strength for the 83-d chronic toxicity tests and to 0.035, 0.053, and 0.052% of original strength for the 96-h acute test for the 1:3, 1:6, and 1:11 copper-to-zinc ratio solutions, respectively. These initial dilutions constituted the 100% concentrations used in the geometric series of dilutions. The 83-d chronic toxicity tests began on 6 December 1978 with fertilized chinook salmon eggs obtained by dry spawning four adult female and two male fish at the Department's Nimbus Hatchery on the American River and ended on 27 February 1979 when the individuals were at the swim-up fry stage. At the end of the 83-d tests, the control fry had a mean weight of 400 mg, a mean total length of 36 mm, and had the yolk-sac absorbed to the diameter of 1 to 2 mm. The 96-h acute tests began on 5 March 1979 with the 90-d old unexposed swim-up fry of similar size and weight and ended on 9 March 1979. Continuous low-level illumination was provided by two, 40 W incandescent lamps during the 83-d tests. A continual 8-h light and 16-h dark photoperiod was provided during the 96-h tests by eight, 60 W fluorescent lights which were switched on and off with a rheostat. 72 CALIFORNIA FISH AND CAME Following fertilization, the eggs were allowed to water-harden in the different concentrations of the combined copper and zinc solutions. The control eggs were water-hardened in filtered Annerican River water. Water-hardening lasted 3 h and took place in 4-1 glass jars that previously had been acid-rinsed, followed by several rinses with de-ionized water. Following hardening, the eggs were volunnetrically measured and floated into perforated, polyethylene egg baskets (50 eggs per basket; 3 baskets per replicate) contained inside of the 12-1 plexiglass troughs used by Finlayson and Ashuckian (1979). Eggs not used in the chronic toxicity tests were water-hardened in filtered American River water, deposited in several square, 10-1, overflowing plexiglass aquaria, placed in the dark with a supply of filtered flowing water, and allowed to hatch and develop into swim-up fry. These unexposed swim-up fry were used in the acute toxicity tests, with 25 fry placed in each 1 2-1 plexiglass trough ( 50 fry per concentration ) . During the tests, water samples from the troughs were collected twice a week for analyses of both total and dissolved copper and zinc. Water samples ( 10 ml) were collected in 15-ml polycarbonate Nalgene® centrifuge tubes which were first rinsed twice in 1^ nitric acid (HNO3) followed with two rinses of de-ionized water and then followed by a sample rinse. Samples (10 ml) for dissolved metals were collected in a SESI® nylon syringe and filtered through a Gelman®type A-E 25-mm glass fiber filter held in a Nuclepore® 25-mm plastic valve. Before taking each sample, the syringe, filter, and valve were rinsed with 1/V nitric acid and then with de-ionized water. All samples were preserved with 0.3 ml of 6/V nitric acid and capped until analyzed. Copper concentrations > 50 jag/ 1 and all zinc concentrations were determined by air-acetylene flame atomic absorption analysis; copper concentrations <50 jug/l were determined by carbon-rod atomization atomic absorption analysis. Analytical precision (2o-) was determined from the modified Shewhart equation: o- = a/2 (x - x)VN-1 where the absolute value of x = [A, — A2]/[A, + A2], and Ai and A2 are paired observations. Analytical precision for zinc was ± 4.4%, for copper concentrations > 50 ju,g/l it was ± 3.5%, and for copper concentrations < 50 jag/l it was ± 18%. Dissolved oxygen, pH, and temperature were measured in each test trough three times weekly, whereas water hardness was measured once every 3 wk. The hardness of American River water varies throughout the year between 19 and 26 mg/l CaCOa (California Department of Fish and Game, unpublished data, 1970-78). Deaths were not recorded in the chronic tests until the eggs had become "eyed", approximately 3 wk after fertilization. The dead (opaque, white) eggs were recorded weekly and carefully suctioned away to avoid disturbing the other developing embryos. When hatching began 7 wk after fertilization, dead (absence of ventilation and movement) alevins and fry were recorded and removed three times weekly. Mortalities were calculated for the life-history periods of eggs-to-hatch (Me), hatch-to-swim-upfry (Ma), and for the total 83-d exposures of eggs-to-swim-up fry (M,). A sample of 20 fry or those remaining (whichever was less) was measured and weighed in each replicate at the end of the 83-d test to determine the effects of copper and zinc on fry growth. Mortalities were also calculated for the 90-d old unexposed swim-up fry (Mf) ESTIMATED SAFE COPPER AND ZINC LEVELS 73 for the 96-h exposures. Log-logit (metal concentrations vs. mortality estimates) analyses (Finley 1971) were used to calculate the LCSO., LCSOo, LC50„ and LC50, (50% lethal concentration) and the LClOe, LCIO,, LCIO,, and LC10, (10% lethal concentration) by least square regression. The mortality estimates used in log-logit analyses were first normalized using Abbott's formula: m = (1 — SJS e) 100 where m = normalized mortality, Sx = survival in concentration x, and Sc = survival of controls. RESULTS AND DISCUSSION The toxicity test concentrations had a pH range of 6.6 to 7.4 with the lowest and highest values corresponding to the highest and lowest solution concentrations, respectively. Water hardness of the test concentrations varied with time between 21 and 27 mg/l CaCOa. Water temperatures during the toxicity tests varied with time between 8.5 and 11.0 C, and dissolved oxygen concentrations varied with the time between 8.5 and 12.2 mg/l. Mortalities of the eggs (M^), alevins (Ma), and for the total 83-d exposures (M,) increased with increased total and dissolved copper and zinc concentrations in the 1:3 ratio (Appendix 1), 1:6 ratio (Appendix 2), and 1:11 ratio (Appendix 3) toxicity tests. For the period of eggs-to-fry (M,), 100% mortality occurred in the highest concentrations while minimal mortality (means = 8.5 to 19.5%) occurred in the lowest concentrations tested. Mortalities of the controls from eggs-to-fry were also minimal (means = 5.5 to 8.6%). Dissolved copper concentrations in the controls were consistently <2 jLig/l and dissolved zinc concentrations ranged from 2 to 5 jag/l. All copper-to-zinc ratio solution concentrations tested produced 83-d old fry which had lower mean weights and lengths than the control fry. Mortalities of the unexposed 90-d old fry (Mf) increased with total and dissolved copper and zinc concentrations in the 1 :3 ratio, 1 :6 ratio, and 1 :1 1 ratio tests (Appendix 4). Complete mortality occurred at the highest concentrations while minimal mortality (means =0.0 to 2.6%) occurred at the lowest concentrations tested. Mortalities of the control swim-up fry were also minimal (means = 0.0 to 5.3%). Dissolved copper and zinc concentrations in the control concentrations ranged from 1 to 2 /xg/l and < 5 to < 7 jag/l, respectively. The toxicity test data indicate that the swim-up fry at the end of the 83-d exposures (Table 2) were more tolerant to copper and zinc in the 1:3 copper-to-zinc ratio than the previously unexposed swim-up fry in the 96-h exposures (Table 3), almost as tolerant to copper and zinc in the 1:6 copper-to-zinc ratio, and less tolerant to copper and zinc in the 1:11 copper-to-zinc ratio. For example, the LC50, values of the 1:3, 1:6, and 1:11 copper-to-zinc ratio solutions during the 83-d exposures were 44 jag/l Cu and 160 jLig/l Zn, 27 jag/l Cu and 206 jbtg/l Zn, and 17 jag/l Cu and 253 jLtg/l Zn, respectively, while the LC50f values during the 96-h exposures were 37 jag/l Cu and 132 /xg/l Zn, 29 jag/l Cu and 213 /xg/l Zn, and 20 /xg/l Cu and 279 /xg/l Zn, respectively. The apparent acclimation to copper and zinc during the 83-d test in the higher ( 1 :3) copper-to-zinc ratio, lack of this acclimation in the middle ( 1 :6) ratio, and the greater toxicity of copper and zinc during the 83-d test in the lower (1:11 ) ratio may indicate that chinook salmon were able to acclimate to copper with relatively less amounts of zinc present (1:3 ratio) but that acclimating to zinc 74 CALIFORNIA FISH AND GAME with relatively less amounts of copper present (1:11 ratio) was not possible. Chinook salmon have been reported to acclimate to copper during early life-history stages but copper acclimation has not been investigated in the presence of zinc (C. Chapman, Research Aquatic Biologist, U. S. Environmental Protection Agency, pers. commun.). TABLE 2. Summary of Statistics and 95% Confidence Limits (In Parentheses) for Chinook Salmon Exposed for 83 Days to 1:3, 1:6, and 1:11 Copper-to-Zinc Stock Solutions. Eggs-to-hatch Metal Dissolved zinc (p.g/1) Total zinc (/i.g/1) Dissolved copper (;ig/l) Total copper (jm.g/1) Metal Dissolved zinc (fig/l) .... Total zinc (ju.g/1) Dissolved copper 'fj,g/l) Total copper (ftg/l) Metal Dissolved zinc (|ig/l) Total zinc ()xg/l) Dissolved copper (/xg/l) Total copper ()ig/l) LC50, LC10, 1:3 1:6 1:11 1:3 1:6 /.// 174 280 4,77 145 224 396 (141-207) (210-380) (377-631) 207 326 524 175 254 437 (169-242) (248^38) (418-681) 48 37 31 40 29 26 (39-58) (28-50) (25-41) 84 63 58 70 50 49 (69-99) (48-85) (46-78) Hatch-to-swim-up fry LC50a LCWa 1:3 1:6 1:11 1:3 1:6 /.•// 182 205 253 119 156 208 (159-234) (146-298) (187-314) 214 243 283 146 187 234 (190-266) (173-351) (212-347) 49 27 17 32 20 14 (43-63) (20-38) (13-20) 89 45 32 56 37 27 (75-136) (33-63) (24-41) Total LC50, LC10, 1:3 1:6 1:11 1:3 1:6 ;.•/; 160 206 253 131 165 200 (112-196) (157-256) (180-317) 192 245 283 170 197 226 (148-232) (187-301) (206-351) 44 27 17 38 21 13 (34-54) (21-33) (12-21) 77 46 32 64 36 25 (59-94) (36-56) (22-39) In toxicity test exposures of both 83 d and 96 h, it was apparent that the toxicities of copper and zinc were additive. The additive nature of copper and zinc was expressed by the fish becoming more tolerant of zinc concentrations as the copper concentrations decreased. This additive nature of copper and zinc toxicity is probably not of a completely additive concentration nature (i.e. V^ Cu LC50 plus ^2 Zn LC50 producing 50% mortality), although we have no separate copper and zinc LC50 values to substantiate this suspicion. The 83-d LC50, values indicate that as the copper-to-zinc ratio decreased from 1:3 to 1:11, there was a 58% increase in zinc tolerance coupled with 61% reduction in copper tolerance. This is in comparison to the 96-h LC50, values which indicate there ESTIMATED SAFE COPPER AND ZINC LEVELS 75 was a 1 1 1% increase in zinc tolerance coupled with a 46% reduction in copper tolerance between the same copper-to-zinc ratio solutions. Finlayson and Ashuckian ( 1 979) found additive toxicity between copper and zinc to steelhead trout. Because there is substantial evidence that the toxicities of copper and zinc are additive, and by reducing the copper-to-zinc ratio from 1 :3 to 1 :1 1 will make the combined copper and zinc solutions 58% to 111% more tolerable to Chinook salmon, the continued and efficient operation of the copper cementation plants in the Spring Creek drainage, which have the capability to decrease the copper-to-zinc ratio of the acid-mine waste, are of beneficial water quality value. TABLE 3. Summary of Statistics and 95% Confidence Limits (In Parentheses) for Chinook Salmon Swim-up Fry Exposed for % Hours to 1:3, 1:6, and 1:11 Copper-to-Zinc Ratios of Stock Solutions. Metal Dissolved zinc (fig/l) Total zinc (\i%l\) Dissolved copper (ju.g/1) . Total copper (jxg/l) 90-d swim-up fry LC50, LOO, 1:3 1:6 1:11 1:3 1:6 /.■// 132 213 279 111 180 234 (103-160) (181-245) (243-321) 144 245 318 120 210 271 (112-173) (209-282) (274-373) 37 29 20 30 25 17 (25-48) (25-35) (18-24) 48 40 31 40 33 26 (37-58) (35^5) (2^38) The Chinook salmon eggs in our tests were more tolerant to copper and zinc concentrations in the two lower (1:6 and 1:11) copper-to-zinc ratio solutions than were the resulting alevins and fry (Table 2). However, the alevins and fry were as tolerant to copper and zinc as the eggs in the higher (1:3) copper-to-zinc ratio solution. Finlayson and Ashuckian (1979) found steelhead trout eggs were more tolerant of copper and zinc than the resulting alevins and fry in two copper-to-zinc ratios (1:4 and 1:12), and Hazel and Meith (1970) found Chinook salmon eggs were more tolerant of copper than the resulting alevins and fry. Hence, the eggs of both chinook salmon and steelhead trout are as tolerant or more tolerant to solutions containing copper and zinc than are the resulting alevins and fry. Lethal levels of copper and zinc for our chinook salmon agree reasonably well with those of other researchers. Our LC50a and LClOo values varied from 17 to 49 jLLg/l Cu and 14 to 32 juig/l Cu and from 182 to 253 jug/l Zn and 119 to 208 jag/ 1 Zn, respectively while Chapman (1978), in water hardness of 22-24 mg/l CaCOa, found the separate copper and zinc 200-h LC50 and LC10 values for non-acclimated chinook alevins to be 20 and 15 jag/l Cu and >661 jag/l and 364 to 661 jLLg/l Zn, respectively. Hazel and Meith (1970) found about a 10% mortality to chinook salmon alevins and fry at 21 jug/l Cu. Our 96-h LCI Of values for previously unexposed fry varied from 17 to 30 jag/l Cu and 111 to 234 jag/l Zn while Chapman (1978) found the separate copper and zinc 200-h LC10 values for non-acclimated fry at 14 jug/l Cu and 68 jag/l Zn. Chapman (1978) concluded that chinook salmon and steelhead trout alevins were more tolerant to copper and zinc than were swim-up fry. Additionally, he found that chinook 76 CALIFORNIA FISH AND GAME salmon and steelhead trout tolerances to copper and zinc increased with development past the swim-up fry stage. Therefore, we must conclude that swim-up fry of chinook salmon and steelhead trout are probably the most sensitive life-history stage to copper and zinc, and hence, recommended "safe" levels based on this life-history stage should be protective for all developmental stages. Based on our acute tests, estimated safe levels for copper and zinc at copper-to-zinc ratios of 1:3 to 1:11 for chinook salmon would be below 17 ja g/l and 111 |u,g/l, respectively. However, other criteria besides death affect the success of an organism. When compared to our control fish, exposed fry showed reduced growth (weight and length) at the end of the 83-d exposures in all concentrations of all copper-to-zinc ratio solutions. The lowest mean concentrations in the 83-d tests were 24 jag/l Cu and 88 jag/l Zn, 12 jug/l Cu and 86 fxg/l Zn, and 11 /xg/l Cu and 161 /xg/l Zn for the 1:3, 1:6, and 1:11 copper-to-zinc ratio solutions, respectively. The lowest copper (11 jag/l) and zinc (86 jag/1) concentrations affecting a reduction of growth in chinook salmon fry were below the lowest copper (17 jag/l) and zinc (111 jag/l) concentrations at the LCI Of values for previously unexposed chinook salmon swim-up fry. Since our response criteria to combined copper and zinc solutions consisted of growth from fertilized eggs to the most sensitive life-history stage (swim-up fry) for chinook salmon, the copper and zinc concentrations of 11 and 86 jag/l, respectively, should approximate the maximum acceptable toxicant concentrations (MATC) of Mount and Stephan (1967). MATC values are between the toxicant concentration having no effect and the toxicant concentration having a measurable effect on the organism. However, even lower copper concentrations have been reported to have an effect on anadromous salmonids. For example, Lorz and McPherson (1977) found that as little as 5 jug/l Cu (7% of Cu 96-h LC50), in water hardness of 68-78 mg/l CaCoa, caused a reduction (below that of control fish) in the percentages of downstream yearling migrants of coho salmon, Oncorhynchus kisutch. Additionally, Lorz et al. (1978) found a greater effect on downstream migration when copper and zinc were present together at sublethal levels (10 jug/l Cu and 800 jLtg/l Zn) than when copper (10 ju,g/l) or zinc (800 jag/l) were present alone. We would recommend safe levels for chinook salmon in the Sacramento River below 11 jag/l Cu and 86 jag/l Zn until further information is obtained on the possible chronic effects of copper and zinc concentrations on smoltification and downstream migration. Our toxicity data and those of Finlayson and Ashuckian (1979) are based on dissolved concentrations of copper and zinc. The term "dissolved" refers to copper and zinc concentrations which can pass through a 0.30 jam glass fiber filter. The filtering process removes 90% of the metal-containing particles and complexes which are greater than 0.30 jam in diameter and therefore, probably removes all suspended particulate matter containing copper and zinc. These particles and complexes are bound by adsorption onto particulate matter and complexed in large carbonate, hydroxide, and oxide colloids or precipitates, and are believed to be biologically unavailable. In natural waters relatively free of suspended particulate matter, the fractions of copper and zinc concentrations which compose these biologically unavailable complexes would be expected to increase with increased water pH and alkalinity. The relationship of these water ESTIMATED SAFE COPPER AND ZINC LEVELS 77 quality parameters to copper (Chapman and McCrady 1977) and zinc (Sinley et al. 1974) toxicity has been expressed by fish becoming increasingly more tolerant of both metals with increased water pH and alkalinity. Since our toxicity data are based on dissolved concentrations of copper and zinc, they should better reflect the concentrations of these metals actually biologically available to cause fish toxicity. There are two advantages with presenting copper and zinc toxicity levels in dissolved concentrations. First, it allows for a more accurate comparison of the toxicities of various metals to various organisms and secondly, it allows for the toxicity data to be applied to a variety of waters regardless of suspended particulate matter content and possibly water pH, alkalinity, and hardness. Other investigators (Sinley et al. 1974; Chapman 1978) have instead based their toxicity data on total metal concentrations and have listed the pH, hardness, and alkalinity (sometimes the suspended particulate matter content) of the water in which the data were obtained. Finlayson and Ashuckian (1979) calculated that the old release schedules of Lewis (1963) would allow for theoretical maximum dissolved copper concentrations below Keswick Dam of 10 and 17 jag/l during periods May through December, and January through April, respectively. Likewise, these schedules allowed for maximum dissolved zinc concentration ranges of 20 to 910 jag/ 1, and 40 to 1600 jag/ 1 for the same periods, respectively. Our present study indicates that safe levels of copper and zinc for chinook salmon are approximately 1 1 and 86 jag/l, respectively. Thus, the release schedules of Lewis (1963) provided borderline protection from copper toxicity and little, if any, protection from zinc toxicity. The new interim release schedules proposed by Wilson (1978) allow for maximum dissolved copper and zinc concentrations of 5 |xg/l and 64 jag/l, respectively. Both the maximum copper and zinc criteria proposed by Wilson (1978) are below the levels determined by us to have a small adverse effect on the growth of chinook salmon fry. In conclusion, this investigation has demonstrated that the chinook salmon eggs are as tolerant or more tolerant to combined concentrations of copper and zinc than alevins and fry, and that chinook salmon appear to be more easily acclimated to copper rather than to zinc during the period of eggs-to-fry. Additionally, this investigation has demonstrated that the toxicities of copper and zinc are additive to chinook salmon eggs, alevins, and fry. Finally, this investigation suggests safe copper and zinc levels below 1 1 jag/ 1 and 86 jag/ 1 for chinook salmon in the upper Sacramento River. ACKNOWLEDGMENTS We thank N. Morgan and B. Castle for assistance with the metal analyses, C. Riley and his crew at the Nimbus Hatchery for the chinook salmon eggs used in our study, and D. Langdon for the preparation of the final manuscript. Suggestions that improved the manuscript were provided by G. Chapman, H. Lorz, D. Wilson, K. McCleneghan, and K. Hashagen. The investigation was partially supported by funds from the California Regional Water Quality Control Board— Central Valley Region (I.A. No. S-1296). REFERENCES Chapman, C. 1978. Toxicities of cadmium, copper, and zinc to four juvenile stages of chinook salmon and steelhead. Am. Fish. See, Trans., 107:841-847. 78 CALIFORNIA FISH AND GAME Chapman, C, and J. McCrady. 1977. Copper toxicity: A question of form; Pages 134-151 in: Recent advances in fish toxicology: A symposium held in Corvallis, Oregon on January 13-14, 1977. Oregon State University, Corvallis, Oregon and Oregon Dept. of Fisheries and Wildlife, July 1977. U.S. Environ. Prot. Agency, Ecological Res. Ser. EPA-600/3-77-035: 203 p. Chapman, C, and D. Stevens. 1978. Acutely toxic levels of cadmium, copper, and zinc to adult male coho salmon and steelhead. Am. Fish. Soc, Trans., 107:837-840. Davies, P., A. Calabrese, J. CoettI, P. Hodson, J. Kent, J. McKim, J. Pearce, D. Reish, J. Sprague, and W. Sunda. 1979. Cadmium, pages 51-66 in: Revew of the EPA Red Book: Quality criteria for water. R. V. Thurston, R. C. Russo, C. M. Fetterolf, T. A. Edsall, and Y. M. Barber eds. Water Quality Section, American Fisheries Society, Bethesda, MD, Finlayson, B., and S. Ashuckian. 1979. Safe zinc and copper levels from the Spring Creek drainage for steelhead trout in the upper Sacramento River, California. Calif. Fish Game, 65(2): 80-99. Finley, D. 1971. Probit analysis. 3rd ed. Syndics of Cambridge University Press, New York. 265 p. Hazel, C, and S. Meith. 1970. Bioassay on king salmon eggs and sac fry in copper solutions. Calif. Fish Game, 56 (2):121-124. Lewis, R. 1963. Recommended flow releases from Spring Creek Debris Dam for the protection of salmonid fishes in the Sacramento River. Calif. Dept. Fish and Game, Region 1, Redding, CA. Unpublished MS. 8 p. Lloyd, R. 1961.Thetoxicity of mixtures of zinc and cooper sulfates to rainbow trout (Salmogairdnerl, Richardson). Ann. Appl. Biol., 49:535-538. Lorz, H., and B. McPherson. 1 977. Effects of copper and zinc on smoltification of coho salmon. U.S. Environ. Prot. Agency, Ecological Res. Ser. EPA-600/3-77-032: 69 p. Lorz, H., R. Williams, and C. Fustish. 1978. Effects of several metals on smolting of coho salmon. U.S. Environ. Prot. Agency, Ecological Res. Ser. EPA-600/3-78-090: 85 p. McKim, J., and D. Benoit. 1971. Effects of long-term exposures to copper on survival, growth, and reproduction of brook trout (Salvelinus fontinalis). Can., Fish. Res. Bd., J. 28:655-662. Mount, D. and W. Brungs. 1967. A simplified dosing apparatus for fish toxicology studies. Water Res. 1:21-29. Mount, D., and C. Stephan. 1967. A method for establishing acceptable toxicant limits for fish-malathion and the butoxyethyanol ester of 2,4-D. Am. Fish. Soc, Trans., 96(2):185-193. Nehring, B., and J. Goettl. 1974. Acute toxicity of a zinc-polluted stream to four species of salmonids. Bull. Environ. Contam. and Tox. 12:464-^69. Nordstrom, D. 1977. Hydrogeochemical and microbiological factors affecting the heavy metal chemistry of an acid mine drainage system. Ph.D. Dissertation. Stanford University, Palo Alto, CA. 210 p. Peltier, W. 1978. Methods for measuring the acute toxicity of effluents to aquatic organisms. U.S. Environ. Prot. Agency, Environ. Monitoring Ser. EPA-600/4-78-012: 52 p. Prokopovich, N. 1965. Siltation and pollution problems in Spring Creek, Shasta County, California. J. Am. Water Works Assoc. 57:986-995. Sinley, J., J. CoettI, and P. Davies. 1974. The effects of zinc on rainbow trout in hard and soft water. Bull. Environ. Contam. and Tox. 12:193-201. Sprague, J. 1964. Lethal concentrations of copper and zinc for young Atlantic salmon. Can., Fish. Res. Bd., J. 21:1-17. Sprague, J., P. Elson, and R. Saunders. 1965. Sub-lethal copper-zinc pollution in a salmon river-field and laboratory study. Int. J. of Air and Water Poll. 9:531-543. Sprague, J., and A. Ramsey. 1965. Lethal levels of mixed copper-zinc solutions for juvenile salmon. Can., Fish. Res. Bd., J. 22:425^32. Thomsen, W., C. Hazel, and S. Meith. 1970. Toxicity of copper and zinc ions to steelhead in Mokelumne River water. Calif. Dept. Fish and Game, Water Projects Branch, Sacramento, CA. Lab. Rep. 50 p. U.S. Fish and Wildlife Service. 1959. Effects of mine-waste pollution on anadromous and resident fish in Upper Sacramento River, Shasta County, California. 15 p. + Attachments. Wilson, D. 1978. Proposed interim release flows from Spring Creek Debris Dam for the protection of salmonid fishes in the Sacramento River. Calif. Dept. Fish and Game, Region 1 , Redding, CA. Memorandum to California Regional Water Quality Control Board— Central Valley Region, August 31, 1978. 5 p. -I- Attachments. ESTIMATED SAFE COPPER AND ZINC LEVELS 79 APPENDIX 1 — Zinc and Copper Concentrations in Toxicity Testing Troughs, Mortalities for Chinook Salmon Eggs Raised to Swim-up Fry Stage for 83 Days, and Weights and Lengths of Surviving Swim-up Fry at 83 Days in Five Concentrations and the Controls of the 1:3 Copper-to-Zinc Stock Solutions. Zn Cu Percent mortality Wt Eggs A levins Total Z Replicate (M8/I) (M8/I) (MJ {MJ (M,) (mg) (mm) 100-1 700±50'' (751) 220 ±27 (310) 100.0 100.0 — — 100-2 710±58 (760) 210±25 (310) 100.0 "^~ 100.0 " " 56^1 383 ±48 (430) 120±19 (220) 100.0 — 100.0 — — 56-2 382 ±24 (421) 111±19 (182) 100.0 " 100.0 32-1 214±29 (240) 58±9 (110) 70.7 63.9 89.4 330 29 32-2 200±14 (242) 52±10 (92) 65.0 63.4 87.2 330 29 18-1 150±20 (178) 41 ±6 (70) 37.3 21.7 50.9 340 31 18-2 141+15 (173) 39±6 (67) 35.7 40.3 61.6 340 31 10-1 88±14 (111) 24±4 (45) 2.4 3.3 5.6 350 34 10-2 89±0 (112) 24±3 (40) 4.9 8.6 13.1 350 34 Control-1 4±3 (23) 2±2 (4) 2.4 3.3 5.7 400 37 Control-2 4±4 (22) 1±2 (3) 6.7 1.8 8.4 395 37 ' Mean dissolved zinc and copper concentrations ± SD; mean total zinc and copper concentrations in parentheses. 80 CALIFORNIA FISH AND CAME APPENDIX 2 — Zinc and Copper Concentrations in Toxicity Testing Troughs, Mortalities for Chinook Salmon Eggs Raised to Swim-up Fry Stage for 83 Days, and Weights and Lengths of Surving Swim-up Fry at 83 Days in Five Concentrations and the Controls of the 1:6 Copper-to-Zinc Stock Solution. Zn Cu Percent mortality Wt Eggs Alevins Total L Replicate (fig/l) (/ig/l) (M.) (MJ (M,) (mg) (mm) 100-1 800±82'' (900) 110±15 (190) 100.0 — 100.0 — — 100-2 820 ±90 (900) 122±17 (216) 100.0 " 100.0 56-1 480 ±67 (530) 64±6 (112) 91.9 100.0 100.0 — — 56-2 447 ±40 (510) 63±13 (111) 88.2 86.7 98.4 300 27 32-1 263 ±33 (302) 32±5 (53) 12.3 30.7 39.2 300 32 32-2 250 ±50 (300) 27±6 (46) 9.3 29.9 36.4 300 32 18-1 147±16 (180) 20±14 (33) 3.2 17.4 20.0 330 34 18-2 150±31 (180) 19±4 (33) 9.0 13.1 20.9 330 34 10-1 86±14 (106) 12±2 (19) 2.4 2.5 4.8 360 35 10-2 91 ±25 (112) 12±3 (21) 9.4 3.4 12.2 360 35 Control-1 3±3 (6) 1±0 (2) 2.4 3.3 5.6 400 36 Control-1 5±4 (10) 1±0 (2) 3.9 1.6 5.5 390 36 ' Mean dissolved zinc and copper concentrations parentheses. SD; mean total zinc and copper concentrations in ESTIMATED SAFE COPPER AND ZINC LEVELS 81 APPENDIX 3 — Zinc and Copper Concentrations in Toxicity Testing Troughs, Mortalities for Chinook Salmon Eggs Raised to Swim-up Fry Stage for 83 Days, and Weights and Lengths of Surviving Swim-up Fry at 83 Days in Five Concentrations and the Controls of the 1:11 Copper-to-Zinc Stock Solution. Zn Cu Percent mortalit] V Wt Eggs Alevins Total L Replicate (jug/l) (fig/l) (M.) (MJ (M,) (mg) (mm) 100-1 1240 ±7/" (1340) 87±9 (143) 100.0 — 100.0 — 100-2 1250±121 (1340) 81 ±13 (149) 100.0 100.0 56-1 662 ±62 (701) 37±6 (80) 82.7 100.0 100.0 — — 56-2 643 ±59 (702) 41 ±7 (78) 79.9 100.0 100.0 " 32-1 383 ±33 (442) 21±5 (42) 11.9 49.5 55.6 330 33 32-2 383 ±39 (426) 24±5 (49) 12.0 40.0 47.2 330 33 18-1 225 ±24 (224) 17±2 (31) 0.8 9.2 9.8 360 35 18-2 225±18 (268) 16±2 (30) 4.6 25.8 29.2 360 35 10-1 161 + 16 (184) 11±2 (20) 15.9 5.2 20.3 390 35 10-2 161±16 (190) 11±2 (20) 13.9 15.3 18.6 390 35 Control-1 2±1 (7) 1±0 (2) 5.9 4.5 10.0 400 36 Control-2 4±3 (5) 1±2 (2) 6.4 0.9 7.2 400 36 ' Mean dissolved zinc and copper concentrations ± SD; mean total zinc and copper concentrations in controls. 2 — 80190 82 CALIFORNIA FISH AND CAME APPENDIX 4 — Zinc and Copper Concentrations and Mortalities for 90-Day Old Chinook Salmon Swim-up Fry Exposed for % Hours to Five Concentrations and the Controls of 1:3, 1:6, and 1:11 Copper-to-Zinc Stock Solutions. Copper-to-zinc ratios 1:3 1:6 1:11 Percent Percent Percent Zn Cu mortility Zn Cu mortality Zn Cu mortalit} Replicate (fig/l) l/ig/l) (M,) (fig/l) (fig/l) (M,) (fig/l) (Mg/I) (M,) 100-1 53216° 150114 100.0 741127 12010 100.0 813161 5017 100.0 (543) (205) (850) (160) (873) (71) 100-2 513121 180114 100.0 796129 131121 100.0 806135 6513 100.0 (571) (200) (817) (162) (844) (99) 56-1 302 + 22 8418 100.0 430123 6015 100.0 42517 3515 95.0 (348) (140) (442) (94) (439) (51) 56-2 298128 7318 100.0 439119 63110 100.0 42019 3015 93.3 (306) (115) (487) (89) (449) (51) 32-1 16817 5611 85.0 24310 31111 73.7 243113 1814 68.4 (174) (63) (277) (48) (269) (27) 32-2 156113 4719 80.0 248111 3616 72.7 232112 1713 68.4 (183) (69) (293) (47) (259) (25) 18-1 113110 3411 57.2 15116 2212 31.2 145114 1013 5.3 (126) (50) (183) (32) (181) (15) 18-2 11819 3411 25.0 160118 2211 5.0 16019 1213 11.8 (121) (40) (208) (33) (176) (15) 10-1 7513 1813 5.3 8913 1111 0.0 134134 711 0.0 (83) (21) (109) (13) (136) (12) 10-2 78111 2414 0.0 9210 1211 0.0 10716 711 5.0 (84) (24) (100) (14) (120) (14) Control-1 715 212 0.0 615 111 0.0 713 211 10.5 (10) (4) (8) (1) (7) (2) Control-2 613 211 0.0 513 211 0.0 514 211 0.0 (12) (3) (10) (4) (8) (4) ' Mean dissolved zinc and copper concentrations ± SD; mean total zinc and copper concentrations in controls. WHITE STURGEON AGE AND GROWTH 83 Calif. Fish and Game 66 ( 2 ) : 83-95 1 980 AGE AND GROWTH OF WHITE STURGEON COLLECTED IN THE SACRAMENTO-SAN JOAQUIN ESTUARY, CALIFORNIA: 1965-1970 AND 1973-1976 ^ DAVID W. KOHLHORST, LEE W. MILLER, JAMES J. ORSI Bay-Delta Fishery Project California Department of Fish and Game 4001 North Wilson Way Stockton, California 95205 Ages of white sturgeon, Acipenser transmontanus, collected in the Sacramento-San Joaquin Estuary from 1965 to 1970 and from 1973 to 1976 were estimated from transverse sections of pectoral fin rays. Sturgeon were difficult to age and there was considerable disagreement between individuals interpreting ages. Estimates of male and female growth rates were not significantly different. A von Bertalanffy growth curve was calculated for ages 0-21 from the 1973-1976 collection. A length-weight relationship was estimated from 1965-1970 data. Estimated growth rate was similar in 1965-1970 and 1973-1976, but was lower than in 1954. It was not possible to determine if a recent decrease in growth rate was real or was due to different aging techniques. Possible causes of reduced growth are discussed. INTRODUCTION White sturgeon occur mainly in estuaries and rivers along the Pacific Coast of North America. Presently they support a small but important sport fishery in the Sacramento-San Joaquin Estuary and its tributaries. Sturgeon were common here in the mid-1 800's but commercial exploitation severely reduced the population by 1 900. Sturgeon fishing was prohibited from 1 901 to 1 91 0 and from 1917 to 1954, when sport fishing was again permitted. The sport fishery expanded with the discovery in 1964 that shrimp [Crangon spp. and Palaemon macrodactylus) were effective bait. Both catch and effort peaked in 1967 and have generally declined since (unpublished data). Initiation of the sport fishery in 1954 stimulated investigations of sturgeon biology to provide information to manage the fishery. Pycha (1956) reported on sturgeon age, growth, size composition, migrations, and abundance. Chadwick (1959) and Miller (1972^, 1972/?) estimated harvest rates and described migrations. Young (Schreiber 1962) and adult (McKechnie and Fenner 1971) sturgeon food habits also were described. Stevens and Miller (1970) and Kohlhorst (1976) determined time and location of spawning. This report presents new information about white sturgeon age composition and growth from 1965-1970 and 1973-1976 and compares the recent growth rate to that estimated by Pycha (1956). METHODS Data were gathered during two collection periods: by the junior authors from spring 1965 to spring 1970 and by the senior author from spring 1973 to summer 1976. Samples from the two periods were analyzed separately since we wanted to compare growth rates between periods. In general, sampling was not systematic; we examined most available fish. ' Accepted for publication November 1979. 84 CALIFORNIA FISH AND GAME In any age group, some fish in our samples may have as much as one more season's growth than others, increasing the variability in length at each age. This problem was most significant in the 1965-1970 data. It was minimized in the 1973-1976 data because about two-thirds of the sample was collected from September to March after the growing season (June through October) was essentially complete. Transverse sections of the first ray of the pectoral fin were used for age determination. Although pectoral fins were obtained from several sources, most samples were collected from the sport fishery, particularly commercial passenger fishing boats (CPFB's), and from sturgeon captured in trammel nets during tagging in fall 1974 (Table 1). All samples obtained from CPFB's in 1 965-1 970 and from creel censuses in both periods were > 1 01 .6 cm total length (tl), the minimum legal size for sturgeon. Fin rays from sturgeon of all sizes were collected by a Department of Fish and Game employee riding on CPFB's in 1975-1976. TABLE 1. White Sturgeon Collection Methods in the Sacramento-San Joaquin Estuary. Collection Number of Collection method period sturgeon Sport angling Commercial passenger fishing boat 1965-70 579 1975-76 502 Creel census 1967, 48 1969-70 1973-75 346 Trammel net 1974 422 Gill net 1966 1 1973-76 145 Surface trawl 1975-76 1 1 1 Beam trawl 1974-75 32 Otter trawl 1975 4 Fish screens at U. S. Bureau of Reclamation Tracy Pumping Plant 1965 4 Fyke trap 1970 1 1973 2 Unknown 1969-70 6 Total 1965-70 639 1973-76 1,564 ' Only legal-sized fish (101.6 cm tl) were collected. Methods of removing fin rays varied. Some fish were subsequently released alive; on those, the first ray was severed near its proximal end with a fine-toothed saw or heavy-duty wire cutter and then cut from the rest of the fin. The entire fin was removed from dead fish. Both fin rays and whole fins were thoroughly air-dried before sectioning. Two 7.6-cm diameter diamond lapidary saw blades mounted 0.9 mm apart in 1965-1970 and 0.5 mm apart in 1973-1976 were used to section the rays. The water cooled saw operated at about 5000 rpm. Two to four transverse sections were cut between 1 .3 and 2.5 cm from the proximal end of the ray. Age is most easily differentiated in this region and loss of early annuli is minimized (Sunde 1961 ). In 1965-1970 sections were polished with fine sandpaper and mounted on microscope slides with clear fingernail polish. In 1973-1976 the narrower spacing of the saw blades eliminated the need to sand the sections before mounting. WHITE STURGEON AGE AND GROWTH 85 The sections were examined under a binocular dissecting microscope using transmitted light. Annuli were visible as narrow light or translucent bands ( Figure 1 ). Periods of more rapid growth appeared dark or opaque. FIGURE 1. Cross-section of the first pectoral fin ray of a 21-yr-old white sturgeon collected in October 1974 from the Sacramento-San Joaquin Estuary. Five of the annuli are numbered for reference. Sturgeon are difficult to age because they live long, annuli often are difficult to interpret, and known age fish are lacking for validation of aging technique. To assess the reliability of aging white sturgeon, five individuals familiar with fish aging techniques (but not necessarily with sturgeon aging) independently and 86 CALIFORNIA FISH AND GAME without training or guidelines interpreted 40 fin rays randomly selected from the 1965-1970 collection. There were many disagreements (see Results). We established two criteria in an attempt to increase consistency: (1 ) Only annuli that were distinct in the lobes of the rays and also in the dorsal and ventral portions were counted. Often, annular rings were very closely spaced in the lobes and did not appear in the other areas. Our criterion was intended to eliminate the possibility that some readers would interpret them as false annuli while other readers would interpret them as true annuli with little growth between. (2) A marginal annulus was assigned if one was not visible between 1 April, approximately the midpoint of the spawning season (Kohlhorst 1976), and 1 July. By 1 July that year's annulus was usually visible near the edge of the section and no additional annulus was assigned. This criterion was necessary because annuli were formed in spring but were not usually apparent until more rapid growth resumed. After we defined these criteria, three new readers estimated ages of the same 40 randomly selected fish. Their performance was only slightly better than that of the first group of individuals, but for consistency we still adopted the criteria. Subsequently, three trained readers independently examined all ray sections. If at least two age determinations agreed, that was accepted as the true age and was used in our age-growth analysis. If all three disagreed, the sample was omitted. Length, to the nearest 2.5 cm in 1965-1970 and to the nearest centimetre in 1973-1976, and capture date were recorded when fins were collected. Sex was recorded when dissection was practical. All lengths reported are total lengths (tl). However, only fork length (fl) was measured on 68 fish. These were converted to tl using the geometric mean functional relationship (Ricker 1973) tl =5.06 + 1.08 fl (r^ = 0.98, n = 366) that we obtained from sturgeon on which both tl and fl were measured. In 1965-1970 age-length data from many of the youngest and oldest age groups were either biased by gear selectivity or unreliable due to small sample sizes. Therefore, growth was estimated only for ages 11-17. The data were fit to the linear equation: I, = a -F bt where: I, = centimetres tl at age t t = age in years using the method of least squares. Growth of sturgeon collected in 1973-1976 was adequately described by fitting all age-length data to the von Bertalanffy growth curve: I, = Uo[l-e -K(t-t,)J where: I, = centimetres tl at age t Loo = asymptotic tl in centimetres K= constant determining the rate of decrease in growth t = age in years to — hypothetical age at 0 length using the least squares computer program BCC-2 (Tomlinson and Abramson 1961, Abramson 1971). A separate sample of sturgeon collected in 1965-1970 was measured and WHITE STURGEON AGE AND GROWTH 87 weighed to the nearest 0.2 kg. The length-weight relationship was estimated by fitting the equation: log w = log a + b(log I) where: w = kilogram weight I = centimetre tl using a geometric mean functional regression (Ricker 1973). RESULTS Evaluation of Aging Technique Before we adopted specific aging criteria, we did not obtain complete agreement among the five technicians on the age of any of the 40 randomly selected fin ray samples. Agreement between pairs of readers 1, 3, 4, and 5 averaged 1 2.5 specimens (Table 2) . Reader 2 averaged only 2.5 agreements with the other readers. Readers 1, 3, 4, and 5 disagreed mostly by 1 or 2 yr. Reader 2 often disagreed by 3 yr or more. Reader 2 generally assigned younger ages by ignoring annuli which he considered false. TABLE 2. Variation Among Readers in Age Determination of 40 Randomly Selected Stur- geon Fin Rays Collected in the Sacramento-San Joaquin Estuary in 1965-70. Disagreement Readers (number of years) Total compared Agreement 12 3 4 5 disagreements 1&2 3 2 6 9 10 10 37 1&3 13 12 6 4 3 2 27 1&4 11 17 7 2 2 1 29 1&5 10 16 8 2 2 2 30 2&3 5 10 5 9 7 4 35 2&4 1 5 7 12 8 7 39 2&5 1 3 7 12 4 13 39 3&4 15 14 7 4 0 0 25 3&5 12 14 9 2 1 2 28 4&5 14 16 7 1 2 0 26 Mean 1-5 8.5 10.9 6.9 5.7 3.9 4.1 31.5 Mean 1, 3, 4, 5 with each other 12.5 14.8 7.3 2.5 1.7 1.2 27.5 Meanof 2 with 1, 3, 4, 5 2.5 5.0 6.3 10.5 7.3 8.5 37.5 The specific criteria for identifying annuli did not have much effect on results (Table 3). Agreement between pairs of readers averaged 12.7 specimens. However, two of the three readers agreed on ages of 26 of the 40 sturgeon, meaning that ages could be assigned to 65% of these fish. TABLE 3. Variation Among Trained Readers in Age Determination of 40 Randomly Selected Sturgeon Fin Rays Collected in the Sacramento-San Joaquin Estuary in 1965-70. Disagreement Readers (number of years) Total compared Agreement 12 3 4 disagreements 1&2 6 22 9 2 1 34 1&3 16 12 n 0 1 24 2&3 16 16 6 2 0 24_ Mean 1-3 12.7 16.7 8.7 1.3 0.7 27.3 88 CALIFORNIA FISH AND GAME In the 1965-1970 sample, 15-19 yr old sturgeon were the most troublesome group, two of three readers agreed on the age of only 56% of them (Table 4). Agreement in other age groups, from 0-4 yr to >20 yr was consistent, ranging from 65 to 67%. TABLE 4. Effect of Age on the Rate of Age Agreement Between Two of Three Trained Readers for White Sturgeon Collected in the Sacramento-San Joaquin Estuary in 1965-70 and 1973-76. 1965-1970 Age Agree Disagree Total % agreement (M 2 13 67 5-9 56 30 86 65 10-14 282 152 434 65 15-19 58 46 104 56 >20 8 __4 12 _67 Total 406 233 639 64 1973-1976 Age Agree Disagree Total % agreement {yA 286 9 295 97 5-9 525 92 617 85 10-14 234 90 324 72 15-19 147 119 266 55 >20 _J2 _30 __62 _52 Total 1224 340 1564 78 In the 1973-1976 sample aging difficulty increased with age. At least two readers agreed on the age of 97% of 0-4-yr-old fish and agreement consistently decreased to 52% for fish > 20 yr (Table 4). Age-Length Frequency We examined fin rays from 639 white sturgeon collected from 1965 to 1970 (Table 1 ) and assigned ages to 406 of them. They ranged from 2 to 27 yr and from 44 to 203 cm tl. From 1973 to 1976, 1,564 white sturgeon were examined (Table 1). We assigned ages to 1,224 of these fish. They ranged from 0 to 24 yr and from 24 to 201 cm tl Age frequencies and length frequencies for combined 1965-1970 and 1973-1976 data both have two prominent modes. Two and 9-yr old fish dominated the collections, as did 50-54-cm and 100-104-cm length groups (Figure 2). Mean lengths of these modal ages, 54 cm and 103 cm, respectively, fall within the modal length groups, suggesting that age interpretation was reasonably consistent throughout the sample processing. Growth Rate Growth rates of the two sexes were compared to determine if they could be combined for an unbiased estimate of average growth. In making this comparison, only part of the available age range was used because: ( 1 ) sample sizes were small ( <4) at ages >18 in 1965-1970; (2) in both samples, sexed sturgeon were obtained primarily from the sport fishery which has a minimum legal TL of 101 .6 cm, and data from sport caught fish are biased by selection for the fastest growing individuals among the partially recruited age groups. This bias is apparent in the length-age relationship for all ages (Figure 3). WHITE STURGEON AGE AND GROWTH 89 <_3 or CO cr> 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 AGE (YEARS) 22 32 62 72 82 92 102 112 122 132 142 152 162 172 162 192 202 MIDPOINT OF 5 cm LENGTH GROUP FIGURE 2. Age and length frequency distributions of white sturgeon collected in the Sacramento-San Joaquin Estuary in 1965-1970 and 1973-1976 combined. (A) Age distribution. (B) Length frequency grouped in 5-cm increments. Growth rate was estimated separately for each sex for ages 11-17 in 1965-1970 and ages 10-20 in 1973-1976. In these age groups, linear regression coefficients for males and females were not significantly different in 1965-1970 (t=1.29, P = 0.20) or 1973-1976 (t = -0.34, P = 0.50) (Figure 3). Therefore, in subsequent analyses we combined the data. 90 CALIFORNIA FISH AND CAME The following linear growth equation describes the 1965-1970 growth data from ages 11-17: 1, =44.86 + 6.21 2t Length and age data from age 0-21 generated the following von Bertalanffy growth equation for the 1973-1976 data: I, = 261.2 [|.e-00^27(, + 3.638)-| The von Bertalanffy curve agrees well with mean lengths calculated for each age; therefore, it appears to provide a good description of overall white sturgeon growth (Figure 4). The curve probably overestimates first year growth (35.6 cm) very slightly. Mean length of fish collected after one growing season (33.2 cm) is probably the best estimate. Subsequently growth estimates slowly declined, from 8.9 cm during the second growing season to 4.0 cm during the 22nd growing season. Length increased from 44.5 cm at age 1 to 164.4 cm at age 21 (Figure 4B). Length-Weight Relationship We determined both length and weight of 209 sturgeon collected in 1965-1970; the sex of 11 of these was known. Since the length-weight relationships for males and females were not significantly different (t — 0.52, P >0.50), all data were combined to determine the functional relationship log W = 3.348 log L -5.927 (Figure 5). Weight increased from 6.2 kg at 101.6 cm to 50.4 kg at 190 cm. DISCUSSION Estimates of sturgeon size varied considerably at each age. This finding is characteristic of sturgeon growth studies ( Magnin 1 964, Cuerrier 1 966, Semakula and Larkin 1968). Aging inaccuracies are a factor, but size variation at each age may actually be large. The validity of age determinations from sturgeon pectoral fin rays has been discussed by several authors, including Cuerrier (1951 ), Roussow (1957), and Sunde (1961). However, they did not determine consistency in age interpretation among several readers. In our analysis, two of three trained readers agreed on the age of 64% of the fish in the 1965-1970 sample and 78% in the 1973-1976 sample. In 1973-1976, agreement decreased with age, indicating increased difficulty in interpreting fin ray sections with age. The growth rate calculated for age 11-17 sturgeon in 1965-1970 was similar to the rate for that age range in the 1973-1976 sample. The rates were not compared statistically because different methods were used to estimate growth in the two samples. A comparison over a wider age range was not possible because of weaknesses inherent in the 1965-1970 data, specifically: (1 ) Mean lengths from age 6 to at least age 1 0 are overestimated since essentially the entire sample consisted of fish larger than the minimum legal length (101.6cm) in the sport fishery. Hence, the fastest growing individuals from these younger, partly recruited age groups were unavoidably selected. ( 2 ) The data for ages > 1 7 may not be representative because of small sample sizes. We could not identify any important biases affecting the 1973-1976 data since sample sizes generally were adequate and fish were collected by a variety of methods. Hence, we believe that the 1973-1976 data adequately represent recent white sturgeon growth in the Sacramento-San Joaquin Estuary. WHITE STURGEON AGE AND GROWTH 91 E o i80r 170 160 150 140 130 120 _ A 1965-1970 X 110 o 5 100 < 180 »- -, 1701- < UJ 160- 150 140 130 120 110 100 O FEMALES A ^A TL= 38.8+ 6.73 (Age) r2 = 0.41 n = 128 o MALES O- TL = 56.5+5.28 (Age) r2 = 0.24 n =124 o ^ A ^ O O O A "I 1 \ 1 1 1 1 1 1 \ 1 1 1 1 r" 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 B 1973-1976 FEMALES A A TL= 51.8 + 6.00 (Age) r2 = 0.72 n = 104 % C. % & MALES O O TL= 52.5+ 5.8! (Age) r2 = 0.69 n =90 n 1 \ 1 1 1 1 1 1 1 1 \ 1 1 1 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 AGE (YEARS) FIGURE 3. Growth rates of male and female white sturgeon from the Sacramento-San Joaquin Estuary. Lines are least squares linear regressions. (A) 1965-1970. (B) 1973-1976. Assuming that Pycha ( 1 956) used aging criteria similar to ours, it appears that white sturgeon growth has decreased since 1954 (Figure 6). However, growth here apparently is still more rapid than in the Fraser River, British Columbia (Semakula and Larkin 1968). Lower water temperatures and a shorter growing season in the Fraser River due to a 12° difference in latitude probably account for the difference. 92 CALIFORNIA FISH AND CAME 200r Range 95% Confidence Interval 0 I 2 3 4.5 6 7 8 9 10 1 1 12 13 14 15 16 17 18 19 20 21 AGE (YEARS) FIGURE 4. Crowth curves fitted by the method of least squares for white sturgeon collected in the Sacramento-San Joaquin Estuary. Mean total lengths, confidence intervals, and length ranges are shown for comparison. Sample sizes are given in parentheses. (A) Linear growth between ages 11 and 17 of fish collected in 1965-1970. (B) Von Bertalanffy growth curve for age 0-21 fish collected in 1973-1976. The apparent reduction in growth rate since 1954 has several possible explanations: (1 ) Genetic — Presently, the fastest growing individuals are vulnerable to the sport fishery (which started in 1954) at a younger age and for a longer period than slower growing sturgeon. Hence, proportionately more slow growing fish remain in the population and cause an apparent reduction in population growth rate when sampled by nonselective methods. WHITE STURGEON AGE AND GROWTH 93 60 _ / 50 log W= 3.348 logL- r2 = 0.9! n = 209 -5.927 / 40 - // " 30 - * .^''"''^^^ ' ' * 20 . : ■ . yj-^ 10 n 1 1 1 — 1 1 1 1 1 1 1 — 100 FIGURE 5. no 120 130 140 150 160 LENGTH (cm) 170 180 190 200 Length-weight relationship of white sturgeon collected in the Sacramento-San Joaquin Estuary from 1965 to 1970. The curve is a geometric mean functional regression. (2) Intraspecific competition for food— '[here is evidence that sturgeon abundance increased between 1954 and 1967 (Miller 1972a), and that it declined between 1967 and 1974 (unpublished data). Increased abundance may decrease growth by intensifying intraspecific competition for a limited food supply. If food is limiting, effects of competition between 1954 and 1967 may still be apparent in our data since age-length relationships determined from a discrete sample reflect growth conditions over a number of previous years. Improved growth due to less competition more recently may not be detected for several years. (3) Decreased food supply — White sturgeon feed primarily on benthic invertebrates in the San Francisco Bay area ( McKechnie and Fenner 1 971 ) . One of these, the bay shrimp, is fished commercially and California Department of Fish and Game records indicate a substantial decrease in mean annual landings from 307,320 kg in the 1940's to 34,940 kg in the 1970's. Market demands and processing costs account for part of this change, but the magnitude of the reduction suggests that shrimp abundance actually decreased. Factors affecting shrimp abundance may have caused undocumented decreases in populations of other benthic organisms also. (4) Effects of environmental contaminants — White sturgeon in the estuary accumulate polycholorinated biphenyls (PCB's) in their tissues. Means levels in muscle tissues of legal-sized fish collected in San Pablo and Suisun bays in 1975 were 2.0 ppm (SD = 2.5, n = 10) in males and 3.5 ppm (SD = 5.0, n = 12) in females. Concentrations in the gonads were much higher, 49.3 ppm (SD = 24.8, n = 8) in males and 23.7 ppm (SD = 27.8, n = 12) in females. Sublethal levels of PCB's inhibit growth in channel catfish, Ictalurus punctatus (Hansen, Wiekhorst, and Simon 1976) and might affect sturgeon in the same manner. 94 CALIFORNIA FISH AND GAME 200 180 160 140 o X 120 h- O 2 UJ "* 100 _i < o 80 60- 40- 20 Q 1954 A 1965-70 © 1973-76 0 Froser River I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 AGE (YEARS) FIGURE 6. Growth of white sturgeon collected in the Sacramento-San Joaquin Estuary in 1954, 1965-1970, and 1973-1976, and in the Eraser River, British Columbia. Eraser River fish ages are adjusted to our method of age designation (age 0 is the youngest age class). Facts are not available to evaluate the relative significance of the various potential causes of reduced growth. ACKNOWLEDGMENTS Many individuals have contributed to our study. We thank Robert McKechnie, Richard Fenner, Salvatore Mercuric, David Ramczyk, Nileta Schaller, Gregg Schmidt, Tim Morefield, John Grant, Ronald Rogers, Susan Smith, Jerry Webb, and Nancy Holzhauser for technical assistance. John Severa diligently collected samples and allowed us to sample aboard his fishing boat. Jean-Paul Cuerrier, staff limnologist of the Canadian Wildlife Service, and Ivan Donaldson, biologist with the U. S. Army Corps of Engineers, generously gave their time to corroborate our age estimates in 1965-1970. John Horton assisted in data analysis and Nanci Dong drafted the figures. Gary Smith, Donald Stevens, and Harold Chadwick reviewed the manuscript and provided editorial assistance. WHITE STURGEON ACE AND GROWTH 95 Funds for this work were provided by Dingell-Johnson project California F-9-R, "A Study of Sturgeon, Striped Bass, and Resident Fishes", supported by Federal Aid to Fish Restoration funds. REFERENCES Abramson, N, ), 1971. Computer programs for fish stock assessment. FAO Fish. Tech. Paper No. 101. Chadwick, H. K. 1959. California sturgeon tagging studies. Calif. Fish Game, 45(4):297-301. Cuerrier, J-P. 1951. The use of pectoral fin rays for determining age of sturgeon and other species of fish. Can. Fish. Cult., 11:10-18. 1966. L'esturgeon de lac Acipenser fulvescens Raf. de la region de lac St. Pierre au cours de la periode du frai. Naturaliste Canadian, 93(4):279-334. Hansen, L. C, W. B. Wiekhorst, and L. Simon. 1976. Effects of dietary Aroclor 1242 on channel catfish (Ictalurus punctatus) and the selective accumulation of PCB components. Can., Fish. Res. Bd., )., 33:1343-1352. Kohlhorst, D. W. 1976. Sturgeon spawning in the Sacramento River in 1973, as determined by distribution of larvae. Calif. Fish Came, 62(1):32^0. Magnin, E. 1964. Croissance en longueur de trois esturgeons d'Amerique du Nord: Acipenser oxyrhynchus Mitchell, Acipenser fulvescens Rafinesque et Acipenser brevirostris LeSueur. Verh. Internat. Verein Limnol., XV:968-974. McKechnie, R. J., and R. B. Fenner. 1971. Food habits of white sturgeon, Acipenser transmontanus , in San Pablo and Suisun bays, California. Calif. Fish Came, 57(3):209-212. Miller, L. W. 1972d . White sturgeon population characteristics in the Sacramento-San Joaquin Estuary as measured by tagging. Calif. Fish Game, 58(2):94-101. 1972/?. Migrations of sturgeon tagged in the Sacramento-San Joaquin Estuary. Calif. Fish Game, 58(2);102-106. Pycha, R. L. 1956. Progress report on white sturgeon studies. Calif. Fish Game, 42(1):23-35. Ricker, W. E. 1973. Linear regressions in fishery research. Can., Fish. Res. Bd., J., 30(3):409-434. Roussow, G. 1957. Some considerations concerning sturgeon spawning periodicity. Can., Fish. Res. Bd., J., i4(4):553-572. Schreiber, M. R. 1962. Observations on the food habits of juvenile white sturgeon. Calif. Fish Game, 48(1 ): 79^0. Semakula, S. N., and P. A. Larkin. 1968. Age, growth, food, and yield of the white sturgeon (Acipenser transmontanus) of the Fraser River, British Columbia. Can., Fish. Res. Bd., J., 25(12):2589-2602. Stevens, D. E., and L. W. Miller. 1970. Distribution of sturgeon larvae in the Sacramento-San Joaquin River system. Calif. Fish Game, 56(2):80-86. Sunde, L. A. 1961 . Growth and reproduction of the lake sturgeon (Acipenser fulvescens Rafinesque) of the Nelson River in Manitoba. MS Thesis. Univ. of British Columbia. 60 p. Tomlinson, P. K., and N. J. Abramson. 1961 . Fitting a von Bertalanffy growth curve by least squares including tables of polynomials. Calif. Dept. Fish and Game, Fish. Bull. (116):1-69. 96 CALIFORNIA FISH AND CAME Calif. Fish and Game 66 ( 2 ); 96-1 04 1 980 DIETS OF MOUNTAIN WHITEFISH, PROSOPIUM WILLIAMSON! (GIRARD), AND BROOK TROUT, SALVELINUS FONTINALIS (MITCHILL), IN THE LITTLE WALKER RIVER, MONO COUNTY, CALIFORNIA ' JOHN PAUL ELLISON* Biology Department California State Universiy, Long Beach Long Beach, California 90840 Diets of brook trout, Salvelinus fontinalis (Mitchill), and mountain whitefish, Prosopium williamsoni (Girard), in July are described and compared to other studies. Diets were compared statistically and shown to be significantly different. Stomachs of brook trout contained insects of both terrestrial and aquatic origin, indicating a possible subsistence on the drift, whereas whitefish stomachs contained mostly immature aquatic insects, suggesting a bottom feeding habit. Differences between the diets of brook trout and whitefish are probably the result of differential feeding habits. INTRODUCTION Mountain whitefish, Prosopium williamsoni (Girard), a fresh water salmonid (Subfamily Coregoninae), range from the eastern slopes of the Rocky Mountains in the headwaters of the Saskatchewan and Missouri rivers to the Cascades and Columbia drainage in the northwest. It is also abundant in the Great Basin, with the Truckee River and the Lahontan Basin being the southernmost limit of its range (Eddy 1969). P. williamsoni occurs sympatrically with some species of endemic salmonids such as Salmo clarki and Salmo gairdneri (Sigler 1951; Eddy and Surber 1960). Because the mountain whitefish is treated as a game fish in some states, including California, and also occurs sympatrically with other salmonids of interest to sport fishermen, it has been the subject of several studies. Brown (1952) investigated its reproduction and development, while McHugh (1940, 1941 ) described the food and growth of this species. Sigler (1951 ) studied the life history of whitefish and made recommendations for its management and McAfee (1966) provided a literature survey. More recent investigations have treated the seasonal migration (Pettit and Wallace 1975) and food habits (Pontius and Parker 1973) of mountain whitefish. There is no published literature available for the food habits of mountain whitefish in California. Prosopium williamsoni has a poor reputation as a food species, and some biologists believe it is detrimental to populations of the more popular trout by feeding on trout eggs and young and offering direct competition for food (McHugh 1940). This paper compares the diet of mountain whitefish with the diet of brook trout, Salvelinus fontinalis, in July to determine the degree of similarity between their diets and possibly the extent of any competition. No attempt is made in this paper to comment on the supposed oophagus tendencies ' Accepted for publication July 1979 ^ Current address: California Department of Fish and Came, Post Office Box 47, Yountville, California 94599 DIETS OF MOUNTAIN WHITEFISH AND BROOK TROUT 97 of mountain whitefish as the fish used in this study were collected in July when few salmonid eggs are available to potential predators. STUDY AREA Fish used in this study were collected from the Little Walker River in Mono County, California. Located on the eastern slope of the Sierra Nevada Mountains, it is a tributary of the West Walker River, which flows east into Walker Lake. This drainage system is part of the Lahontan system; its geological morphology and history are discussed by Snyder (1917). The Little Walker River drains an area of 163 km^ and has a mean yearly discharge of 1.5 m^/sec (51 cfs). The mean discharge in the month of July for ihe years 1965 to 1972 was 3.4 m^/sec (118 cfs). Vegetation at the study site was primarily sage, with dense stands of pine and aspen in the immediate vicinity of the River. The study section was 21 km northwest of Bridgeport, California. This section of the river was 5 to 9 m in maximum width, 0.3 to 1.5 m in maximum depth, 1 km long, and with an average gradient of 52 m/km. Elevation of this section was between 2,182 m and 2,555 m (U.S.G.S. topographic map). Bottom types ranged from rubble, with a current velocity in excess of 1 m/sec (35 cfs), to a sand and gravel bottom silted in areas, with a current velocity of about 0.2 m/sec (7 cfs) . Latter mentioned bottom type was found in diversions created by beaver dams and accompanying snags. Other native fish found in the Little Walker River include Lahontan cutthroat trout, Salmo clarkii henshawi) longnose dace, Rhinichthys cataractae; Lahontan redside, Richardsonius egregius; Piute sculpin, Cottus beldingi; mountain sucker, Catostomus platyrhynchus; and Tahoe sucker, Catostomus tahoensis. Introduced species are rainbow trout, Salmo gairdneri; brown trout, Salmo trutta; and brook trout, Salvelinus fontinalis. MATERIALS AND METHODS On 12 and 26 July 1973, 38 brook trout and 64 mountain whitefish were collected, using an electrofishing apparatus (Smith-Root Mark V Electroshocker). Previous to these dates fish were collected with hook and line in order to assess their distribution in the study section. Fish collected by hook and line were not used in the diet analysis because this collecting method resulted in the introduction of prey not usually found in the fishes' diets (viz., grasshoppers). All fish were collected between 1100 hours and 1800 hours. At the time of collection, weight to the nearest gram and standard length to the nearest millimetre were recorded and scale samples taken. As the survivability of the fish collected was not an imporatant factor, the electrofishing device was set to maximize fish attraction. Most fish collected by this method experienced extreme muscle contraction and usually died as a result. Fish not killed immediately with the electrofishing device were placed in a bucket and allowed to suffocate. No regurgitation of stomach contents was observed. Stomachs were removed by cutting the digestive tract at the esophagus and at the beginning of the intestine just posterior to the pyloric caecae. After removal, stomachs were individually wrapped in cheesecloth, numbered with a tag, and fixed in 10% formalin for 48 hours, and then rinsed and transferred to 70% isopropyl alcohol. If stomachs were not removed within an hour after capture. 98 CALIFORNIA FISH AND CAME the fish were injected with 20% formalin into the body cavity to arrest the digestive process. Organisms found in stomachs were examined with a variable power dissecting microscope and identified to the lowest taxon feasible using Usinger (1956), Ross (1965), and Borror and White (1970. Lengths ( excluding anal ceri) and weights of each taxonomic group were measured using a dissecting microscope with an ocular micrometer disc and a Roller-Smith Precision Balance, respectively. All insects were weighed to the nearest 0.2 mg and their lengths were measured to the nearest 0.1 mm. The mean length of each taxon was calculated from the measurements of each entire individual in the group or an aliquot of 10 if there were more than 20 organisms present. At the time of the first collection of fish, a reference collection of benthic, drift, and terrestrial insects was made. Insects were preserved in 70% isopropyl alcohol. Weights and lengths of preserved insects in the reference collection were used to determine reconstructed weights of insects found in stomachs of fish (Ricker 1937). Data from the diet analysis were analyzed with an IBM 376 computer using an STSS (Nie, etal. 1970) program for discriminate analysis. The variables in the analysis were calculated by multiplying the number of each prey by its reconstructed weight. RESULTS AND DISCUSSION Diet Composition Results in this paper are presented in two ways: (1 ) the type, number, and percent composition by number of principal prey species are listed because it is commonly used and permits comparison with other studies; and (2) diet composition, as represented by reconstructed weights of principal prey species, is also given. The stomachs of most fish were full or mostly full; only one whitefish had an empty stomach. The data were not segregated on the basis of size or age class because it has been demonstrated that mountain whitefish of the size range sampled in my study do not show a significant difference in diet with an increase in size or age. Laakso (1950) found that fingerling whitefish use the same food as adults, but feed on smaller animals (probably the earlier instars of the same aquatic larvae). Sigler (1951 ) stated that there appears to be no great difference between the food habits of different sizes of mountain whitefish. Pontius and Parker (1973), in their investigation of the food habits of P. williamsoni, grouped the fish into the following size classes based on the size range and number of fish in their samples: I, less than 230 mm; II, 230 to 330 mm; and III, 330 mm or larger. They were able to show a quantitative difference in diets only between size classes II and 111. Of the fish collected for my study, 92% fell within Pontius' and Parker's size class 1 and the remainder within size class II. Fish representing size class I were not adequately sampled in their study and fish greater than 330 mm were inadequately sampled in my study of the Little Walker River population. When the data I obtained from the stomach analysis of whitefish from the Little Walker River ^Table 1) were adjusted by removing fish representing Sigler's (1951 ) 0+ age class, there was no appreciable change in the order of abundance of the food groups. The only noticeable change was in the position of larval chironomids, which dropped from 15.4 to 10.9%. This result is DIETS OF MOUNTAIN WHITEFISH AND BROOK TROUT 99 consistent with previous studies. McHugh (1940) reported that the difference he found between diets of 0+ and older fish was due to larger proportion of larval dipterans (chiefly chironomid larvae) in the stomachs of smaller fish. Pontius and Parker (1973) noted that the difference in diets of larger whitefish was due to a consistently lower percentage of chronomids in the stomachs of older fish. TABLE 1. Principal Prey in the July Diets of Brook Trout and Mountain Whitefish as Indicated by Number of Occurrence and Percent of Total. Brook Trout Whitefish Order Family Number Percent Number Percent Ephemeroptera Baetidae (Baetis) (N) ' 40 3.0 1367 25.7 Baetidae (Ephemerella) (N) 85 6.5 1139 21.4 Heptageniidae (N) 9 0.7 562 10.5 Baetidae (A) 77 5.9 185 3.5 Trichoptera Lepidostomatidae (L) 283 21.6 47 0.9 Brachycentridae (L) 73 5.6 141 2.6 Rhyacophilidae (L) 23 1.8 128 2.4 Diptera Chironomidae (L, P) 89 6.8 823 15.4 Simulidae (L, P) 8 0.6 547 10.3 Miscellaneous (A) 85 6.5 16 0.3 Coleoptera Miscellaneous, nonaquatic (A) 50 3.8 4 0.1 Hymenoptera Formicidae (A) 214 16.3 13 0.2 Miscellaneous (A) 37 2.8 — — Plecoptera Peltoperlidae (N) 25 1.9 48 0.9 Other prey (N, L, P, A) 215 16.4 310 5.8 Stage: N = nymph; L = larvae; P = pupae; A = adult As for brook trout, Griffith (1974) found that the only difference in diet between age classes of fish was between fingerling (age class 0+) and older fish. None of the brook trout collected in my study were 0+ individuals. In addition to insects and other animals, the stomachs of fish examined contained organic and inorganic debris. This debris was variously composed of sand, gravel, and pieces of terrestrial and aquatic vegetation that had been incorporated into pupal and larval cases of those caddisfly which were eaten. Also found were pieces of terrestrial vegetation the approximate size and shape of insect larvae and larval cases, which may have been mistaken by fish for food items. This material, which comprised an estimated 20 to 30% of the volume of the stomach contents, will be disregarded in this paper. Also disregarded are cases of trichopteran larvae and pupae removed before the insects were weighed. Brook Trout Organisms found in the stomachs of brook trout collected from the Little Walker River in July were largely terrestrial and aquatic insects. The single most important prey species, as determined by reconstructed weight (Table 2), were larval brachycentrids. The second most important food item were salmonid fry (probably whitefish), followed in descending order of importance by terrestrial adult coleopterans, formicids, lumbricids, larval lepidostomatid trichopterans, 100 CALIFORNIA FISH AND GAME nymphal baetids {Ephemerella spp.), adult heptageniids, terrestrial adult hymenopterans other than formicids, and terrestrial and aquatic adult dipterans. The use of reconstructed weight instead of simple occurrence by number elevated in relative importance those food organisms which would otherwise have been considered an insignificant portion of the diet (e.g., salmonid fry). TABLE 2. Principal Prey in the July Diets of Brook Trout and Mountain Whitefish as Indicat- ed by Reconstructed Weight of Prey (Gram) and Percent of Total. Brook Trout Whitefish Order Family Cram Percent Cram Percent Ephemeroptera Baetidae (5deto) (N)' 0.07 0.5 3.00 9.4 Baetidae (Ephemerella) (N) 0.75 5.1 7.52 23.7 Heptageniidae (A) 0.71 4.8 7.14 22.5 Baetidae (A) 0.24 1.6 0.37 1.2 Trichoptera Lepidostomatidae (L) 1.02 6.9 0.21 0.7 Brachycentridae (L) 2.48 16.7 8.38 26.4 Hydropsychiidae (L) 0.04 0.3 0.55 1.7 Rhyacophilidae (L) 0.09 0.6 1.32 4.2 Miscellaneous (P) , 0.27 1.8 0.41 1.3 Diptera Simuliidae (L, P) 0.01 0.1 0.49 1.5 Miscellaneous (A) 0.46 3.1 0.06 0.2 Coleoptera Miscellaneous, nonaquatic (A) 1.50 10.1 0.02 0.1 Hymenoptera Formicidae (A) 1.22 8.2 0.01 0.0 Miscellaneous (A) 0.55 3.7 — — Opisthopora Lumbricidae 1.17 7.9 0.06 0.2 Osteichthyes Miscellaneous (eggs) — — 0.76 2.4 Miscellaneous (fry) 2.00 13.5 — — Other prey (N, L, P, A) 2.27 15.3 1.48 4.7 ' Stage: N = nymph; L = larvae; P = pupae; A = adult Griffith (1974) found that brook trout fed largely on Ephemeroptera, Coleoptera, Diptera, Trichoptera, and Plecoptera. He reported that these aquatic insects comprised 97% of the number of organisms in the drift and 92% of the number of organisms eaten by the trout. Insects of terrestrial origin were poorly represented in the drift of the Idaho streams he studied. Aquatic insects made up about 72% of the number of organisms eaten by brook trout in the Little Walker River (Table 1 ) . This difference may be expected as trout in streams and rivers largely feed off the drift (Jenkins, Feldmuth and Elliott 1970, Griffith 1974), and the terrestrial component of the drift from the Little Walker River is greater than that found by Griffith at his study site in Idaho. Whitefish Mountain whitefish collected in July from the Little Walker River were found to be feeding largely on immature aquatic insects. The following food groups, listed in descending order of importance as determined by reconstructed weight (Table 2), comprised 86% of the stomach contents: larval brachycentrid trichopterans, nymphal baetids (Ephemerella spp.), nymphal heptageniids, nymphal baetids (Baetis spp.), and larval rhyacophilids. Terrestrial organisms were almost entirely absent from the diet of whitefish and made up only 1.9% of the reconstructed weight of the organisms in the stomachs. The immature DIETS OF MOUNTAIN WHITEFISH AND BROOK TROUT 101 ephemeropterans and trichopterans accounted for 91.1% of the whitefish diet. Laakso (1950), using a volunnetric method, found whitefish, Prosopium williamsoni cismontanus, of the Gallatin and Yellowstone rivers in Montana to have differing summer diets. Whitefish from the Yellowstone River had a high percentage of ephemeropterans and plecopterans nymphs in their diet. The Ephemeroptera were largely represented by nymphs of the family Baetidae, and Trichoptera were almost entirely of the family Brachycentridae. Whitefish from the Gallatin River had a summer diet composed principally of plecopteran nymphs, annelid worms, and simuliid larvae. Laakso describes P. williamsoni s^s essentially being a benthic feeder that will utilize surface organisms during summer and spring to supplement its diet. Mountain whitefish of the Little Walker River had a diet roughly comparable to the summer diet of whitefish in the Yellowstone River. Both groups of fish utilize immature Ephemeroptera to a large extent (53.4%, by volume, in Yellowstone River whitefish, and 55.6%, by reconstructed weight, in Little Walker River whitefish). As the diet of any benthic-feeding fish is directly related to the composition of the benthic community in its habitat, no valid comparison can be made of allopatric populations of such fish without also comparing the quantitative aspects of the benthos. Because such data are not available in this instance, no further comparison will be made of these populations. McHugh (1940) also described P. williamsoni diS a basically bottom-feeding fish with the most important food being aquatic insect larvae. In spite of its usual bottom-feeding behavior, McHugh found that mountain whitefish taken from waters poor in aquatic insect fauna had a large percentage of terrestrial insects in their stomachs. Those whitefish he collected from rivers showed a high percentage of immature Ephemeroptera, Trichoptera, and chironomid larvae and pupae in their diets. A large proportion of the fish McHugh examined were collected in summer. Sigler (1951 ) listed insect orders found in mountain whitefish from the Logan River, Utah, in descending order of importance by volume as Trichoptera, Diptera, Ephemeroptera, and Plecoptera. However, he made no mention of the exact date of collection of the fish. His data serve to illustrate the relative importance of the larval dipterans (mostly chironomids). He, too, indicated that the mountain whitefish is a bottom feeder. Mountain whitefish collected during the summer by Pontius and Parker (1973) from the Snake River in Wyoming fed on large numbers of larval Trichoptera and chironomids. The authors stated that it was impossible to determine to what extent the fish fed on the bottom or drift by comparing stomach contents to bottom and drift samples. However, they did suspect that P. williamsoni was primarily a bottom-feeder, due to the high incidence of benthic forms in the stomachs and the whitefish's subterminal mouth, which suggested that it is adapted for this type of feeding. High numbers of benthic insects found in the stomachs of whitefish from the Little Walker River tended to support the findings of other studies that P. williamsoni is primarily a bottom-feeder. However, some whitefish were captured with a baited hook suspended in the water column, suggesting that they occasionally feed in the drift. Results of the stomach analysis showed that whitefish in the Little Walker River feed more extensively on immature ephemeropterans in summer and depend less on chironomid larvae and pupae 102 CALIFORNIA FISH AND GAME during this period than whitefish in other rivers. This may simply have been an indication of the relative abundance or availability of these two food items. The incidence of plecopteran nymphs in the diet of these whitefish was not as high as encountered in other studies. The relatively high importance of Trichoptera, particularly the family Brachycentridae, was consistent with other populations of whitefish. Comparison of Diets Four major aquatic food groups ( Ephemeroptera, Trichoptera, Diptera, and Plecoptera) as determined by reconstructed weight account for a little over one-third (37.7%) of the brook trout's summer diet, with mostly terrestrial prey species making up the remainder of the diet ( Figure 1 ) . These same food groups (Table 1 ) made up nearly the entire diet of whitefish in July (94.6%). ■§50 +-> u D c o 0 30 cr 05 +-> o \- 20 £ 10 u (D a- 0 ^ y KL [~] Ephemeroptera Hlrichoptera H Diptera ^Plecoptera 171 y y ^iS ^ Brook trout Whitefish FIGURE 1. Relative abundance of benthic insect food groups in the diets of mountain whitefish and brook trout from the Little Walker River in July 1973. Larval Trichoptera, particularly of the family Brachycentridae, were the one group of food organisms of high relative importance to both whitefish and brook trout. Larval brachycentrids were among the largest aquatic insect larvae available in the Little Walker River and may have provided the fish the greatest amount of caloric and nutritional value per unit of energy expended in their capture of the prey organisms present. The trichopterans' large size and cumbersome larval case would have made them an obvious and vulnerable portion of the benthos or drift. Griffith (1974) determined that brook trout selected immature Trichoptera and Ephemeroptera, as indicated by positive electivity indices (Ivlev 1961 ). Pontius and Parker (1973) found that mountain whitefish actively select larval Trichoptera. DIETS OF MOUNTAIN WHITEFISH AND BROOK TROUT 103 It was probably in the selection of larval Trichoptera that the greatest degree of competition occurred between brook trout and whitefish during the sumnner in the Little Walker River. It may have been possible that this competition was not as direct as indicated by the results. Whitefish may have been largely selecting larval Trichoptera from the benthos and brook trout taking them from the drift. This is not to suggest that either brook trout or whitefish were strictly limiting their feeding to any one portion of the environment. Another factor to be considered is that the cropping of any benthic form will ultimately affect its abundance in the drift and vice versa. However, it did appear that whitefish were largely benthic in their feeding habits, whereas brook trout utilized drift organisms to a large extent. I used discriminate analysis to analyze the data from my study to determine if brook trout and whitefish have statistically different diets. Discriminate function analysis is a multivariate tool which forms linear combinations of discriminating variables that maximize separation between groups (Sokal and Rohlf 1969). A feature of the computer program I used is the derivation of an f value based on the within group variance. An F oi 13.15 with degrees of freedom of 22 and 77 was determined for the diets of brook trout and whitefish, yielding a significant difference in their diets at the 99.99% level of confidence. Those food groups which contributed most to the difference between the diets of the two species were immature Ephemeroptera, adult Hymenoptera, adult Diptera, and adult nonaquatic Coleoptera. It was primarily the fishes' differential utilization of these food groups which allowed the necessary partitioning of the environment for the coexistence of these two species with respect to diet. SUMMARY 1. Mountain whitefish and brook trout were collected from the Little Walker River, Mono County, California, and their stomach contents analyzed. A reference collection of insects and aquatic invertebrates was used as an aid in the classification and determination of the reconstructed weights of the prey species. Diets of whitefish and brook trout in July were qualitatively and statistically compared and the results contrasted with other studies. 2. In July brook trout fed on both terrestrial and aquatic insects. The brook trout's diet was composed largely of larval brachycentrids, terrestrial adult coleopterans, formicids, lumbricids, larval lepidostomatids, nymphal baetids {Ephemerella spp.), adult heptageniid ephemeropterans, terrestrial adult hymenopterans other than formicids, and terrestrial and aquatic adult dipterans. 3. Brook trout in the Little Walker River were similar in their utilization of drift fauna to populations of brook trout in other states. 4. In July whitefish fed mostly on immature aquatic insects, especially larval brachycentrids, nymphal baetids {Ephemerella spp.), larval heptageniid ephemeropterans, larval baetid ephemeropterans {Baetis spp.), and larval rhyacophilids. 5. Whitefish from the Little Walker River had summer diets similar to whitefish populations investigated by others. They utilized benthic fauna and ate immature aquatic insects, predominately. 6. Brook trout and whitefish in July have diets which are significantly different. I attribute this difference to the predominantly benthic-feeding habits of whitefish as opposed to the predominantly drift-feeding habits of brook trout. 104 CALIFORNIA FISH AND GAME Based on this study mountain whitefish did not offer brook trout serious competition for food during July in the Little Walker River. ACKNOWLEDGMENTS My grateful appreciation is extended to C. W. Hill, E. L. Sleeper, and R. C. Clover of California State University, Long Beach, and to Phil Pister of the California Department of Fish and Game for their editorial and technical assistance and encouragement throughout this study. REFERENCES Borror, D. J., and R. E. White. 1970. A field guide to the insects. Houghton Mifflin Co., Boston. 404 pp. Brown, C. J. D. 1952. Spawning habits and early development of the mountain whitefish, Prosopium williamsoni, in Montana. Copiea, 1952:109-113. Eddy, S. 1969. The fresh water fishes. William C. Brown Co., Dubuque, Iowa. 286 pp. Eddy, S., and T. Surber. 1960. Northern fishes. Charles T. Brandford Co., Nexton Centre, Massachusetts. 276 pp. Griffith, J. S. 1974. Utilization of invertebrate drift by brook trout (Salvelinus fontinalis) and cutthroat (Salmo clarkf) in small streams in Idaho. Amer. Fish. Soc, Trans., 103(3):440-448. Ivlev, V. S. 1961. Experimental ecology of the feeding of fishes. Yale University Press, New Haven. 302 pp. Jenkins, T. M., C. R. Feldmeth, and G. V. Elliott. 1970. Feeding of rainbow trout (Salmo gairdnerf) in relation to abundance of drifting invertebrates in a mountain stream. Can., Fish. Res. Bd., J., 27(12):2356-2361. Laakso, M. 1950. Food habits of the Yellowstone whitefish, Prosopium williamsoni cismontanus (Jordon). Amer. Fish. Soc, Trans., 80:99-109. McAfee, W. R. 1966. Mountain whitefish. Pages 299-303 in A. Calhoun, ed. Inland fisheries management. Calif. Resource Agency, Calif. Fish and Game. 546 pp. McHugh, ). L. 1940. Food of the Rocky Mountain whitefish, Prosopium williamsoni (Girard). Can., Fish. Res. Bd., J., 5(2):131-137. 1941. Growth of the Rocky Mountain whitefish. Can., Fish. Res. Bd., J., 5(4):337-343. Nie, N. H., C. H. Hull, ). C. Jenkins, K. Steinbrenner, and D. H. Bent. 1970. Statistical programs for the social sciences. McGraw-Hill Book Co., New York. 675 pp. Pettit, S. W., and R. L. Wallace. 1975. Age, growth, and movement of mountain whitefish, Prosopium williamsoni (Girard), in the North Fork Clearwater River, Idaho. Amer. Fish. Soc, Trans., 1 04 ( 1 ) :68-76. Pontius, R. W., and N. Parker. 1973. Food habits of the mountain whitefish, Prosopium williamsoni (Girard). Amer. Fish. Soc, Trans., 102(4):764-773. Ricker, W. E. 1937. The food and the food supply of sockeye salmon (Oncorhynchus nerka Walbaum) in Cultus Lake, British Columbia. Can., Biol. Bd., J., 3(5):450-468. Ross, H. H. 1965. A textbook of entomology. John Wiley and Sons, New York. 539 pp. Sigler, W. F. 1951. The life history and management of the mountain whitefish Prosopium williamsoni (Girard) in Logan River, Utah. Agr. Exp. Sta., Utah State Agr. Coll., Logan, Utah, Bull., (347:1-21. Snyder, J. 0. 1917. The fishes of the Lahontan system of Nevada and Northeastern California. U. S. Bur. Fish., Bull., (35):33-86. SokaL R. R., and F. J. Rohlf. 1969. Biometry. W. H. Freeman and Co., San Francisco. 775 pp. Usinger, R. L. 1956. Aquatic insects of California. University of California Press, Berkeley. 508 pp. THE USE OF ROAD-KILLED DEER 105 Calif. Fish and Came 66 ( 2 ): 1 05-1 1 1 1 980 THE USE OF ROAD-KILLED DEER FOR ASSESSING REPRODUCTIVE POTENTIAL AND WINTER CONDITION OF THE DEVIL'S GARDEN INTERSTATE MULE DEER HERD^ HAL SALWASSER Department of Forestry and Conservation University of California, Berkeley ^ and DOUGLAS I. THAYER California Department of Fish and Came P.O. Box 1623, Alturas, CA 96101 and D. LEWIS YOUNG Modoc National Forest, U.S. Forest Service ^ This paper illustrates some inventory uses of highway-killed deer and describes a simple method for gathering biological data from these deer. Data on reproductive potentials, herd age structure, and winter fat reserves are evaluated to show some ecological conditions of the herd. The Devil's Garden Interstate herd has declined in size since 1964 due to poor fawn survival. Fawn production at birth is high. We postulate that winter depletion of fat reserves could retard fetal growth. This condition, related to fall and spring nutrition, may be an important factor in high neo-natal fawn mortality in the herd. Road-killed deer can serve the biologist/manager as an important inventory tool in assessing some aspects of deer herd ecology. INTRODUCTION Recent recruitment declines in many western mule deer, Odocoileus hemionus, herds have resulted in dwindling herds and reduced hunter success. The problem is acute on the Devil's Garden Interstate herd of northeastern California and south central Oregon. A 40% drop in recruitment since 1964 has led to a 60% herd reduction (Salwasser 1976, 1979). The reported buck harvest has declined by 80% over the period. Local speculation regarding the factors causing poor recruitment include too few bucks for total breeding, an abundance of "barren does," and excessive coyote predation on fawns. Obviously, the only way to reverse current recruitment trends is to alter the ecological factors actually responsible. To examine some possible factors, biologists from the Interstate Deer Herd Technical Committee initiated an investigation of winter deer condition and reproductive potentials through the use of road-killed deer. The sampling period was October 1974 to April 1975. Evaluation of diets was also included in the investigation but is not included in this report. This paper discusses the acquisition of information on age structure, reproductive potential, and carcass condition through necropsy of road-killed deer, the analysis of such data, and the ecological and management implications of the findings. Literature concerning the use of data from road-killed deer is limited. The ' Assistance to this investigation was provided by Federal Aid in Wildlife Restoration projects W-51-R "Big Came Studies" and W-52-R "Wildlife Investigations Laboratory." ^Current address, U.S. Forest Service, Tahoe National Forest, Nevada City, CA 95959. 'Current address, U.S. Forest Service, Dakotas Planning Team, Pulver Hall, Dickinson, ND 58601. 106 CALIFORNIA FISH AND CAME Kansas Forestry, Fish, and Came Commission has used road-killed deer to gather reproductive data for both mule deer and white-tailed deer, O. virginianus, for several years (Peabody and Hiavachick 1970). Their largest sample of mule deer in recent years included 13 does. South Dakota biologists used 17 road-killed does to gather reproductive data for both mule deer and white-tailed deer as part of a study to determine mortality of fawns in the Black Hills (Rice 1977). In Colorado, Gill (1971 ) used road-killed deer as one part of the estimate of total mortality in the Middle Park mule deer herd. Goodwin and Ward (1976) reported the sex ratios of road-killed mule deer were similar to those ratios observed on the adjacent winter range along part of Interstate 80 in Wyoming. The North Dakota State Game and Fish Department examined road-killed does to obtain fetus counts during the spring of 1978 in an attempt to obtain reproductive data after a severe winter, but the sample size obtained was small (J. McKenzie, North Dakota State Fish and Game Department, pers. commun.). In past years, the Colorado Division of Wildlife has utilized road-killed deer to gather various data, but none have been published (A. Anderson, Colorado Division of Wildlife, pers. commun.). METHODS Our field method for necropsy of road-killed deer was designed with several criteria in mind. First, road kills are found in a variety of conditions, ranging from severely dismembered to essentially whole carcasses. The data we collected were obtainable from most carcasses regardless of condition. Second, eight biologists participated in data collection, so the methods had to be standardized to reduce human judgment errors. Third, the biologists collected data as a part-time addition to regular duties; therefore, the necropsy method had to be simple and time efficient. Each cooperating biologist was equipped with sampling kits (one per deer) which included: 1 — quart (or litre) plastic ice cream carton for rumen contents 2 — 20 cm X 25 cm (8 in. x 10 in.) plastic food storage bags 1 — 150-ml glass jar (6 oz baby food jar) for gonads and eyeballs 2— "sale tag" labels 1 — large plastic bag 46 cm x 61 cm (18 in. x 24 in.) for fetuses 1 — data card An accessory kit was also provided to cooperators: 1 — gallon 10% formalin for preservation of all samples 1 — metric tape measure 1 — small metric rule 1 — permanent ink marker 1 — box extra plastic food storage bags extra labels Each carcass was processed as follows: 1. Check for external parasite load or abnormalities, reported qualitatively. 2. Measure hindfoot length (Hall 1962) with cloth tape. 3. Remove jaw for later aging. 4. Remove eyeballs, preserve in formalin in glass jar. THE USE OF ROAD-KILLED DEER 107 5. Open body cavity (Cowan and Karstad 1971). 6. Remove handful of solid rumen contents and preserve in equal amount of formalin in bag in carton. 7. Measure rib fat at thickest depth on sternum with small metric rule. 8. Estimate percent of kidney surface area covered by fat, record highest value of either kidney. 9. Remove and preserve ovaries (testes if male) in glass jar. 10. Remove and preserve fetuses (or entire reproductive tract, if small) in large plastic bag. 11. Check for any internal abnormalities. All data were recorded on a standard data form. Specimen containers, labels, and the data form were pre-labeled to avoid errors of omission. Specimens were forwarded to the senior author for further evaluations. Ages up to 2 years old were assigned by tooth eruption and replacement (Robinette et al, 1957); ages of older deer were based on annular ring counts on the first incisor. Ovaries were sectioned and corpora lutea counted using the criteria established by Teer, Thomas, and Walker (1965). Fetuses were sexed and weighed to the nearest gram. The left hindfoot was measured to the nearest millimetre. While not all data are reported in this paper, the method described provides information on age structure, parasite conditions, skeletal size, diets, fat reserves, reproductive potentials, fetal growth and conception dates, and gross abnormalities. Bled carcass weights, a desirable parameter to measure under special collections, should not be measured unless most carcasses are whole. Data collection required about 15 minutes per carcass after cooperating biologists were trained on two or three carcasses. All desired data and samples were obtained from each deer, except badly mutilated ones. Additional tasks, such as detailed parasite checks and quantitative kidney fat measurements, can be incorporated according to time, supplies, and training. A detailed presentation of the necropsy method and data analysis procedures is available upon request from the senior author. RESULTS AND DISCUSSION Data were collected from 43 road-killed deer. Fat reserve information was obtained from all carcasses, and reproductive performance was ascertained from 30 female deer taken on California Highway 139 through the Interstate winter range. Female sample age structure reflects recent poor recruitment and indicates a declining population (Table 1 ). Nearly two-thirds of the doe sample was composed of 3-year-old and older animals. Interestingly, fawn road-kills showed an even ratio of bucks to does. TABLE 1. Age Structure of Winter Road-Killed Deer in the Devil's Garden Interstate Mule Deer Herd, October 1974-April 1975. Male Female Age sample sample Fawn 6 7 Yearling 1 3 2 year 0 4 3 year 0 10 4 + year 1 11 TOTAL r sF Ovulation Fetal Percent Fertility Twin Male: rate rate pregnant % rate % female 0.00 0.00 0 - - - 1.33 1.00 100 75 0 0:3 2.00 1.75 100 88 75 6:1 2.00 1.67 100 83 67 3:7 2.00 1.90 100 95 90 14:5 108 CALIFORNIA FISH AND GAME The road-kill sample reflects true population age and sex structure. The composition of the winter road-kill sample, 7 bucks (BB): 100 does (DD): 46 fawns ( FF ) , is essentially identical to the fall composition count on the herd, 6BB: lOODD: 48FF. From our data, although an admittedly small sample, there appears to be little sex or age differential in vulnerability to highway mortality. Additional yearly comparisons or larger samples are needed, however, before making stronger conclusions on such vulnerability. Ovulation and fetal rates (Table 2) show that potential fawn production is not a factor in low recruitment. Fetal rates for a herd are a function of: i) percent prime breeding age does, ii) doe condition at rut, iii) sufficient bucks to breed all does, and iv) abortive diseases. TABLE 2. Reproductive Performance of Winter Road-Killed Does from the Devil's Garden Interstate Mule Deer Herd, October 1974-April 1975. Age class Sample Fawn 7 Yearling 3 2 year 4 3 year 6 4+ year 10 Yearling+ 23 1.91 1.70 100 89 70 23:16 2 year + 20 2.00 1.80 100 90 80 23:13 The high fetal rate for Interstate does (1.7 fetuses per yearling and older doe), 100% pregnancy, and high fertility indicate no problem with the first three factors. Recent low post-harvest buck ratios (6BB : 100DD) are still biologically sufficient for reproductive success. Furthermore, fetal ages showed a 25-day breeding period from early November to early December with a peak during the last 2 weeks in November. Low buck ratios are not resulting in extended breeding periods or conception on the second or third estrus cycle. We saw no evidence of abortion, resorption, or fetal abnormalities. FHowever, none of our samples were from late gestation, and abortion cannot be ruled out by these data. Chattin (1948) found a 1.75 fetal rate in 2-year-old and older Interstate does during 1946-1947. The current 1.80 rate in the same age group indicates little change in this potential. The slight increase may be due to small sample variation, or it may be a function of a higher percentage of 3-year-old and older does presently in the herd. Interstate fetal rates are lower than those reported from the highly productive Sublett herd in Utah during the mid-1 950's; Juiander, Robinette, and Jones (1961 ) found fetal rates of 1.96 in mature does compared with our 1 .80, and 1 .56 in yearlings, compared with our 1 .00. Interstate fetal rates are, however, higher than for many wild herds (Robinette and Gashwiler 1950, FHudson 1959, Juiander et al. 1961; Nellis 1968; Browning, Schulenberg, and Brunetti, 1973; Salwasser, Holl, and Ashcraft 1978). The influence of nutrition on fetal rates and fawn survival has long been understood by deer biologists. Robinette et al. (1973) reported significant increases in fawn production and survival when they improved the nutritional plane of their penned mule deer. Verme (1969) found significant survival differences in white-tailed deer fawns born to does on differing nutritional planes. THE USE OF ROAD-KILLED DEER 109 Verme ( 1 967 ) also documented the phenomenon of compensatory breeding, which explains why does on low quality diets and with poor fawn survival still show high fawn production potentials. Doe condition at rut is influenced by both summer and fall habitat quality and the number of fawns nursing the doe during that period. If fawns suffer high neo-natal mortality, as has been shown to occur in the wild (Trainer 1975, Smith and LeCount 1976, Salwasser et al. 1978) the doe is able to use all available nutrition to rebuild her own tissue reserves. Hence, does not nursing fawns are able to show high ovulation and fertility rates on what is actually a poor nutritional plane for total reproductive success. Ransom (1967) has shown the extreme of compensatory breeding; on the northern fringe of white-tailed deer distribution, does skip a year in breeding if they are successful in raising a fawn. It appears that compensatory breeding may be responsible for the relatively high fetal rates in the Interstate herd. The Interstate herd's fetal sex ratio shows a slight preponderance (59%) of males. Verme (1965) observed a higher male fawn ratio from does on poor diets. Robinette et al. (1973), however, did not find a significant difference in sex ratios as a result of doe diets. Our sample is probably too small to make meaningful conclusions as to the nutritional implications of Interstate fetal sex ratios. The sex ratio of road-killed fawns indicates that by the following winter the ratio is likely to be 50 : 50. Evaluation of carcass fat showed the Interstate deer entered winter with moderate energy reserves (Table 3). Fawns carried the lowest fat stores going into winter, with 3 mm along the sternum and a 20% covering of kidney surfaces. By late February they had depleted their reserves. Mature does retained some fat reserves until early April. Winter conditions in 1975 were quite stressful to the herd. There was no fall "greenup" that year. Consequently, the herd was forced to subsist on dry grass, sagebrush, and juniper through a long cold winter. Green forage did not appear on the ranges until mid-April, some 6 to 8 weeks after fawns had exhausted their fat reserves and 2 to 3 weeks after adults had depleted their reserves. Late winter fawn mortality and reduced fetal growth would be expected outcomes of the 1975 interaction between individual deer condition, habitat quality, and winter weather. TABLE 3. Fat Reserve Trends in Road-Killed Mule Deer from the Devil's Garden Interstate Mule Deer Herd, October 1974-April 1975 Fawn Yearling 2 Year 3+ Year n Rib Kidney n Rib Kidney n Rib Kidney n Rib Kidney Period (mm) (%) (mm) (%) (mm) (%) (mm) (%) 17 Oct— 24 Dec 3 3 20 3 13 75 16 50 4 11 85 (5-25)^ (75-80) (75-100) 1 Feb— 7 Feb -- - -- - -- - 29 70 (50-90) 24 Feb— 20 Mar 90 5 32 15 13 5 14 4 35 (0-10) (0-25) (0-100) 30 Mar— 13 Apr 40 3 21 5 23 5 41 10 (0-10) (0-10) (0-10) (0-25) ' Rib fat measured at thickest depth on sternum. ^ Percent of kidney surface area covered by fat. ■* Range of values in parenthesis. 110 CALIFORNIA FISH AND GAME SUMMARY AND CONCLUSIONS Road-killed deer can be a valuable addition to the manager's inventory arsenal. A sample of 43 deer yielded reproductive data on 23 breeding females, an adequate sample for most management purposes. Other data can be obtained on herd age structure and deer condition with a minimum of manpower, time, and supplies. This information can be used to indicate animal and herd response to habitat conditions. When integrated with other inventories, such as herd composition counts, it can give valuable clues to which factors are or are not responsible for observed population trends. The age structure of the Interstate sample indicates the results of ever-declining recruitment: a population dominated by older age classes. Herd size will continue to decline because there are fewer females in young age classes to replace oldei deer as they succumb to natural factors. Yet, fawn production remains relatively high because most females in the herd are in the prime reproductive ages of 3 years and older. Even at the low buck ratio of 6 per 100 does at rut, all reproductive age does were pregnant, and they conceived during the optimum breeding period. These facts show conclusively that fawn production is not responsible for poor recruitment in the herd. Factors contributing to fawn mortality are the problem. Other work on mule deer in California has indicated that spring nutrition, summer nutrition, and summer cover are probably the most important factors in neo-natal fawn survival (Salwasser 1976, 1979; Salwasser et al. 1978; Holl, Salwasser, and Browning 1979). The carcass condition analysis performed on these road-killed deer further indicates that current fall habitats may not provide sufficient nutrition to prepare the herd for a stressful winter. Fawns in 1975 depleted their energy reserves with 4 weeks of winter remaining; adults by winter's end. It is likely that fetal growth would be impaired under such conditions. Indeed, Verme (1977) has recently shown that the small size of fawns born after such circumstances is correlated with high fawn mortality. Correction of the recent trends in recruitment must focus on spring, summer, and fall habitats. Conceivably, improvement of the fawning and summer habitats (which the Interstate herd uses from 4 weeks prior to fawning to the start of the fall migration ) and of fall migration ranges will significantly alter the nutrition and cover problems that are causing high fawn mortality and early winter depletion of fat reserves. On the Interstate ranges, this will require a commitment by managers to use fire, logging, and livestock grazing to create and maintain serai brushfields on forest soils. Road-killed deer can be used to further monitor deer herd trends and responses to management actions. This study indicates a sample of 40 to 50 deer is sufficient to gather valuable information for assessing the success or failure of deer management. ACKNOWLEDGMENTS We wish to acknowledge the efforts of those assisting in the collection and analysis of samples: Ed Styskel, Mike Bauer, Leon Fisher, Ron Escanoand Mickey Server of the U.S. Forest Service. Frank Grogan of the Oregon Wildlife Commission collected samples, and Ken Rutherford of the Oregon Wildlife Commission assisted in data analysis. California Department of Transportation employees provided information on the location of road-killed deer. THE USE OF ROAD-KILLED DEER 1 1 1 REFERENCES Browning, B., R.W. Schulenburg, and O. Brunetti. 1973. Railroad Flat deer study. California Departnnent of Fish and Came, P-R, W-52-R Project Report, 49 pp. Chattin, ).E. 1948. Breeding season and productivity in the Interstate deer herd. Calif. Fish Game, 34(1): 25-31. Cowan, I. McT., and L. Karstad. [1969] 1971. Post-mortem examinations. Pages 251-258 in R. H. Jr. Giles, ed. Wildlife management techniques, 3d ed. The Wildlife Society, Washington, D.C. 633 pp. Gill, R.B. 1971. Middle Park deer study— productivity and mortality. P-R Job Prog. Rep., Proj. No. W-38-R-25, Work Plan 14, Job 3: 189-207. Goodwin, G.A., and A.L. Ward. 1976. Mule deer mortality on Interstate 80 in Wyoming: causes, patterns, and recommendations. USDA, Forest Service, Res. Note Rm-332. 4 pp. Hall, E.R. 1962. Collecting and preparing study specimens of vertebrates. Mus. Nat. Hist., Univ. Kansas, Misc. Publ. 46 pp. Holl, S.A., H. Saiwasser, and B. Browning. 1979. The diet composition and energy reserves of California mule deer during pregnancy. Calif. Fish Game, 65(2):68-79. Hudson, PA. 1959. Fetal recoveries in mule deer. J. Wildl. Manage., 23(2):234-235. Julander, O., W.L. Robinette, and D.A. Jones. 1961. Relation of summer range condition to mule deer herd productivity. J. Wildl. Manage., 25(1):54-60. Nellis, C.H. 1968. Productivity of mule deer on the National Bison Range, Montana. J. Wildl. Manage., 32(2):344-349. Peabody, W.C, and D.B. Hiavachick. 1970. Deer biological data. Kansas Forestry, Fish and Game Commission. P-R Job Prog. Rep. Proj. No. W-23-R-8, 14 pp. Ransom, A.B. 1967. Reproductive biology of white-tailed deer in Manitoba. J. Wildl. Manage., 31 (1 ):144-123. Rice, L.A. 1977. Mortality rates of fawn age class in South Dakota deer populations, 1976-77. P-R Job Prog. Rep., Proj. No. W-75-R-19, Job No. 1 and 3. 9 pp. Robinette, W.L., and J.S. Gashwiler. 1950. Breeding season, productivity, and fawning period of the mule deer in Utah. J. Wildl. Manage., 14(4):457^69. Robinette, W.L., D.A. Jones, G. Rogers, and J.S. Gashwiler. 1957. Notes on tooth development and wear of Rocky Mountain mule deer. J. Wildl. Manage., 21 (2):134-153. Robinette, W.L., C.H. Baer, R.E. Pillmore, and C.E. Knittle. 1973. Effects of nutritional change on captive mule deer. J. Wildl. Manage., 37(3):312-326. Saiwasser, H. 1974. Coyote scats as an indicator of time of fawn mortality in the North Kings River deer herd. Calif. Fish Game, 60(2): 84-87. 1976. Man, deer and time on the Devil's Garden. Proc. West. Assoc. State Came and Fish Comm., 1976:295-318. 1979. The ecology and management of the Devil's Garden Interstate deer herd and its range. Unpubl. Ph.D. dissertation. Univ. California, Berkeley. 377 pp. Saiwasser, H., S.A. Holl, andC.A. Ashcraft. 1978. Fawn production and survival in the North Kings River deer herd. Calif. Fish Game, 64(1):38-52. Smith, R.H., and A. LeCount. 1976. Factors affecting survival of mule deer fawns. Arizona Game and Fish Dep., Fed. Aid in Wildl. Restor. Proj. W-78-R. 15 pp. Teer, J.G., J.W. Thomas, and E.A. Walker. 1965. Ecology and management of white-tailed deer in the Llano Basin of Texas. Wildl. Monogr., 15. 62 pp. Trainer, C. 1975. Direct causes of mortality in mule deer fawns during summer and winter periods on Steens' Mountain, Oregon ... a progress report. Proc. West. Assoc. State Came and Fish Comm., 1975:163-170. Verme, L.J. 1965. Reproduction studies on penned white-tailed deer. J. Wildl. Manage., 29(1):74-79. ' 1967. Influenceof experimental diets on white-tailed deer reproduction. Trans. N. Am. Wildl. Nat. Res. Conf., 32:405^20. 1969. Reproductive patterns of white-tailed deer related to nutritional plane. J. Wildl. Manage., 33(4):881-887. 1977. Assessment of natal mortality in upper Michigan deer. J. Wildl. Manage., 42(4):70O-708. 112 CALIFORNIA FISH AND CAME Calif. Fish and Game 66 ( 2 ): 1 1 2-11 9 1 980 THE LITTORAL BOTTOM FAUNA OF HIGH ELEVATION LAKES IN KINGS CANYON NATIONAL PARK ' TIMOTHY P. TAYLOR and DON C. ERMAN Department of Forestry and Resource Management University of California, Berkeley, CA 94720 Benthic macroinvertebrates were collected with skin diver-operated coring tubes from the littoral zone of 11 high elevation lakes in Kings Canyon National Park. Macroinvertebrate density averaged 5418 individuals/m^ was fairly uniform (SD = 1716/m'), and was positively correlated with human use around the lakes. Chironomidae, Oligochaeta, or Pisidium were the most common invertebrate organisms found. INTRODUCTION The high mountain lakes of California's Sierra Nevada are major attractions to hikers in wilderness areas of national parks. Many of these alpine lakes now contain introduced trout (Christensen 1977), for which benthic macroinvertebrates are an important dietary component (Elliott and Jenkins 1972). But their attractiveness as well as the general increased human use of remote areas has led to so many people visiting and camping at the lakes that managers of wilderness areas are concerned about deteriorating natural features. In Kings Canyon National Park heavy visitor use of remote high elevation lakes has prompted the Park Service to restrict use and to sponsor research on the possible impacts of people on these lakes. It was our broad objective to evaluate the effects of people on a series of high elevation lakes in Kings Canyon National Park. Over three summers, studies were conducted on 1 1 lakes close to the John Muir Trail in a heavily used part of Kings Canyon N. P. (lat 118^25'N, long 36048'W; elevation 3161-3350 m; Figure 1). In 1975 and 1976 data on basic limnology were obtained (Silverman and Erman 1979; Taylor and Erman, in press) . These data suggested a pattern among the lakes that pointed to an impact by people. Lakes with greatest use had more rooted aquatic plants and less dissolved nitrate than lakes with low human use. We hypothesized that conditions in the sediment may have been the key factors in reflecting past use levels, and so we returned to the lakes for an intensive, short term sampling of bottom features in all lakes in August 1977. The effects of past use on benthic plants and other features are reported elsewhere (Taylor and Erman, in press). The purposes of the study reported here were to obtain information on the composition and abundance of benthic invertebrates of these little studied lakes and to determine if invertebrates were responding to different levels of human use around the lakes. METHODS In the last week of August 1977, 11 oligotrophic lakes (all Kearsarge Lakes — Upper, Middle, Main, and East; Bullfrog and Charlotte lakes; Upper and Lower Rae lakes; Dragon Lake, and 2 of the 7 major Sixty lakes — Upper and ^ Accepted for publication September 1979. BOTTOM FAUNA OF HIGH ELEVATION LAKES 113 Cedar Grove Trailhead — 27 km 1 ff ' Q "To «." ij' y. o 4 '. o I \ ^ »- J \ A W ) \ ® A Yl% ^ Jyy/\ \\ ^t^ ^^P^^c/lW V Sixty V^ V Lower rs V^^Lower ^ \^ ^ r ,Rae Sixty y ^ " /^* yi N Upper Rae Dragon CharloMeN^-'^-^i'l-'fr°3 "»^ I km Onion Valley 8 km > Kearsarge Lakes FIGURE 1 . Location of study lakes, major streams, and main trails in Kings Canyon National Park. 114 CALIFORNIA FISH AND GAME Lower), representing the range of lake sizes and levels of visitor use characteristic of the area, were sampled by skin divers. These are the same lakes for which previous data on limnological conditions were collected with the exception of Dragon Lake, which was added in 1977; morphological features were not obtained. The late August date was chosen so that bottom plants would be at peak annual growth and any seasonal differences among the lakes would be minimized by a short sampling period. All of the lakes contained trout but no data on their relative abundance was obtained. We allocated the total amount of diving effort to each lake based on the proportion of bottom area shallower than 6 m (lake morphology was surveyed in 1975-76). This depth was the limit for free diving. Within this zone divers sampled benthic plants at 2 m and 4 m (Taylor and Erman, in press) and bottom fauna at the mid point of the zone, 3 m. Points around the shore were spaced evenly, and a pair of samples were taken 4 m apart on the 3 m contour at each point. In Bullfrog Lake, where as many as 22 pairs were required, sampling was reduced to only 10 pairs because of the rigor of diving at these elevations. Samples were obtained by pushing a 2.8-cm diameter by 15-cm coring tube into the bottom, capping it, and extracting it. Cores from mud and sand bottoms were about 12 cm in length, but cores from coarser bottoms were as short as 3 cm. Gravel or rock substrates could not be sampled by core. Uneven sample numbers (Table 1 ) reflect a missed sample because of these conditions. We did not have a comparable method for quantitatively sampling gravel or rock. The cores were preserved in 5% formalin and mixed thoroughly; a small amount (1-2 g) of the core was set aside for other analysis, and silt was washed from the remainder through a 0.21 -mm mesh screen. All macroinvertebrates were removed, identified as far as possible ( Edmondson 1 959; Mason 1 968; Brinkhurst and Jamieson 1971; Burch 1972), and the data used to compute a Shannon diversity index (Shannon and Weaver 1949) for each lake: H'= S^iog.^ where n, is the number of individuals of species i, and N is total individuals of all species. TABLE 1. Physical Features and Human Impact Index Values of the Study Lakes Mean depth Volume Surface area Human impact Elevation (m) (m^xlOV (ha) index (m) Upper Kearsarge 3.1 38 1.2 57 3324 East Kearsarge 4.2 196 4.6 23 3383 Middle Kearsarge 4.3 65 1.5 68 3322 Main Kearsarge 8.7 664 7.5 98 3322 Bullfrog 6.0 595 8.7 106 3231 Charlotte 4.8 465 9.6 179 3161 Upper Rae 14.0 3,421 24.5 86 3200 Dragon N.A.^ N.A. N.A. 32 3383 Lower Rae 6.3 1,424 22.5 244 3200 Lower Sixty 5.5 317 5.7 36 3250 Upper Sixty 8.2 457 5.5 27 3250 ' in part from Taylor and Erman, in press. ^ Data not available. Dragon Lake was not surveyed in previous work. On the remaining subsample, organic content was determined by combustion of dry samples ( Black et al. 1 965 ) . The effect of formalin on the organic content was negligible when checked against blanks of unpreserved soil. BOTTOM FAUNA OF HIGH ELEVATION LAKES 1 1 5 Relative levels of human impact upon the study lakes were determined from data compiled by the Park Service. Investigators identified campsites around the lakes and ranked each campsite from 1 to 5 (5 = greatest impact) based on impacts on vegetation, leaf litter, duff, soil; on campsite area and development; and on human trails in the area. The sum of all rankings around a lake gave a composite score for each lake ( human impact index, Taylor and Erman, in press; Table 1). RESULTS AND DISCUSSION Chironomids, oligochaetes, and Pisidium (fingernail clams) were the dominant benthic organisms found in the lakes (Table 2 ) . Upper and Lower Rae lakes and the three upper Kearsarge lakes were dominated by oligochaetes, but in the lower lakes of the Kearsarge basin. Main Kearsarge and Bullfrog, Pisidium became progressively more important. Upper and Lower Sixty and Charlotte lakes were dominated by chironomids. These three taxa are also dominant in the Convict Creek basin lakes (Reimers, Macioiek, and Pister 1955) and are dominant in many other oligotrophic lakes (Pillion 1967; Sarkka 1972; Pennak 1 977) . However, Lower Rae, Lower Sixty, and Upper Sixty had no Pisidium, and, surprisingly, we found no chironomids in Dragon Lake. However, the distribution of these organisms is patchy, and our relatively light sampling effort at some lakes may have missed some groups. Chironomid density averaged 1437 individuals/m ^ in our study lakes; oligochaetes, 2135/m ^ and Pisidium, 1567/m 2. The chironomid density was about the same as densities found in 3 2440-m elevation, northern Sierra lakes (Needham and Sumner 1941; Calhoun 1944; Friehofer 1952) and 10 southern Sierra lakes at 2311-3377 m elevations (Reimers, Macioiek, and Pister 1955). By contrast, we found much higher densities of oligochaetes (primarily tubificids of which Limnodrilus hoffmeisteri snas most common) and Pisidium than other investigators reported from northern and southern Sierra lakes. Precise comparisons with other studies are difficult because of differences in sampling techniques. Apparently corers collect more individuals per unit area than Ekman grabs (Paterson and Fernando 1971), and the other studies cited used an Ekman grab. However, small corers may underestimate some taxa (Kajak 1963). In addition, many small organisms slip through screen openings larger than 0.2 mm (Jonasson 1958). We used a 0.21 -mm mesh screen; most studies in the past used even larger mesh openings. The total invertebrate density ranged from 2599/ m^ in Lower Rae Lake to 7308/m^ in Dragon Lake and Bullfrog Lake (Table 2). Mean density of all lakes was 541 8/m^ and variation was moderate (SD = 1716/m^). Oligotrophic mountain lakes in the Sierra and elsewhere generally have densities less than 2500/m^ (Needham and Sumner 1941; Calhoun 1944; Friehofer 1952; Erman 1969; Sarkka 1972), although densities up to about 4700/ m^ were found in lakes of the nearby Convict Creek basin ( Reimers et al. 1 955 ) . By contrast, the littoral zone of eutrophic Lake Esrom in Denmark had 10 810 individuals/m' (Jonasson 1969). Elsewhere in North America chironomids from the profundal zone have been used to classify trophic lake types (Saether 1975). Orthocladiinae characterize oligotrophic lakes, while the genus Chironomus characterizes eutrophic lakes. These types occur together in the littoral zone of the study lakes (Table 2); 116 CALIFORNIA FISH AND GAME however, samples from the littoral zone may reflect only local conditions rather than conditions in the lake as a whole (Erman 1969; Saether 1975). These differences in species composition do not reflect any obvious trend in response to the human impact index. The low number of samples from several lakes (Middle, Kearsarge, Upper Rae, Dragon, and Upper Sixty) and the small core size restrict any strong conclusions; nevertheless we observed an apparent correlation between invertebrate density (based on log transformations to correct for clumped distributions, Elliott 1971 ) and the human impact index (Figure 2). Lower Rae was an exception; it had the lowest density but the highest human impact index. We are uncertain why this Lake was so different. If Lower Rae is excluded, the positive correlation is significant (r = 0.64, P < 0.05). In a related study we found a highly significant positive correlation between the human impact index and the frequency of occurence of benthic plants in these lakes (Taylor and Erman, in press), but invertebrate density and benthic plant frequency were not significantly correlated. Physical and chemical factors (organic carbon, nitrate nitrogen, total iron, substrate type, lake dimensions, and lake flushing rate) also were not significantly correlated with invertebrate density. Diversity indices calculated for the study lakes ranged from 0.61 in Dragon Lake to 3.29 in Charlotte Lake (Table 2), but diversity was not significantly correlated with any of the above factors. Bullfrog Lake is interesting because it has a high human impact index (Table 1 ), a high frequency of bottom plants (Taylor and Erman, in press), and a high macroinvertebrate density (Table 2). This Lake once had obvious overuse from campers and grazing stock and was closed to use over a decade earlier (D. Parsons, National Park Service, pers. commun.). Nevertheless, upstream and nearby use and perhaps slow recovery continue to place Bullfrog Lake in the class of lakes showing effects of human use. 50 100 HUMAN 150 IMPACT 200 INDEX 250 FIGURE 2. The correlation of macroinvertebrate density to human use around lakes in Kings Canyon National Park, California. BOTTOM FAUNA OF HIGH ELEVATION LAKES 117 CO ^_ ^_ o LO f ■«r ro r~v. LD LH CO ■* '^ n e 'S z e o >. c n U w M C — I 1-^ r'3 ^ Cr^ -^ ^ OO CO u-1 1 — r— aO y— O ro CO i-n O r^ OO vD r^ CO o in fN ? ^ s o o CM rv< en rn a> o Lit CM o ^ t-M v£) CO cn s CO -— o o >9 1^ CO 01 Tj- O^ o LTl CO a^ >— (--4 r^ a^ Ln rvi ■* CTi 1-^ .— CT^ f^ o m rvi 0.4 5 ^ T— rr. »— > Z «^ o u c o N o <^ ■^ < E o Q O I .?- :§ ? ^ CO 0 ■^ 00 rn ^ rn rsl 0 0 in 0 ^~~ 5 o CO ■Xi CO y— -^ ss >— .— CM O CT^ ^ r-4 Ln on m »— 0 Ln o^ 0 CO *— CO fN CT^ 00 u-l rsi ■^ in -^ 00 0 00 CO 0 m 0^ 0 !-s| m o^ .— m t-^ CO o ^ o § ^ CO O -"^ vX- CO o rs O *— (^ vX" '"^ <^ O 0 CO rsi rsi 0 in 00 •^ rn in CT^ m in in r>i in l\ (-si !-M r^4 u-l o ? iX> ^ "(5 3 k at > e u c V es ca < ■j^ 1*^ Uj la o I c5 ^. ■XI (3^ rvi hx 0 CO — r>i •^ ■* 00 t^ m ,_ 00 0 in i-n 0 r^ m r»i est CM 'S- O 0^ § s •sD rn XI ^ O O Q t-M T^ e^ ^ X) o o) ■_ "O -el IS ■9- P 4; '^ E Q Jj U .§ fO ro S ' ro 3 -3 : = •^ rq ni tj <~> Q O O n -o 'E o c o u ^ e _ a; rO ™ ra y rs( XI CT^ CM X> 0^ x> r^ "I ^ rO ro -Q U ^ -g (•1 in i/i ^ Q; a; a> a; o "3 u "^ ■" I - ^ C C — :^ ^ t— T^ 3 (TJ O) •3 - O - 1 1 8 CALIFORNIA FISH AND GAME We could detect no local effects on total number of invertebrates or individual taxa close to areas of most camping. But further studies that concentrate samples at locations near traditional camping areas would be useful to see if localized effects of people have occurred. Our data on benthic plants indicated that increased human use was producing lake-wide stimulation of plant growth in these relatively small, rapidly circulating lakes (Taylor and Erman in press). We hypothesized that stimulation was from input of plant nutrients from the greater human activity. Possibly invertebrates are also benefiting from a lake-wide increase in organic and inorganic nutrients, although the organic carbon in sediment samples was not correlated with invertebrate density. An increase in benthic standing crop is typical of enrichment. The benthic amphipod Hyalella azteca was absent from all samples, although in a previous study this species was the predominant animal in plankton tows from Rae and Sixty lakes (Silverman and Erman 1979). Its absence is probably due to hand sampling, which may disturb the amphipod, and the small corer, which may bias the samples (Kajak 1963). Hyalella azteca (= H. knicker- bockeri) was introduced into the Rae Lakes, along with benthic alga Nitella, in 1919 (Coleman 1925) as fish forage. In summary, we present information on abundance and composition of littoral bottom fauna from 11 high elevation lakes. Macroinvertebrates tend to follow the same pattern we observed for benthic plants; they are more abundant in the lakes with historically high levels of human use. Concern is justified about use levels changing the characteristics of these remote alpine Ia1r= 29.95 mm). The specimens are deposited with California Fish and Game (Region 5), Texas Cooperative Wildlife Collection (TCWC 2499.1 ), and Los Angeles County Museum of Natural History (LACM 38462-1). During the course of the African clawed frog study. Vail Lake and all tributaries were routinely surveyed. Arroyo Seco Creek was the only locality where L. parva was encountered. At this site, L. parva was restricted to a small outflow of an impoundment of Arroyo Seco Creek. Lucania parva was taken in running water, while the mosquitofish, Gambusia affinis, was taken in both running and still waters. NOTES 125 Hubbs and Miller analyzed a number of characters to find which could be used to deternnine parent localities for the introduced populations. They concluded that the number of supraorbital pores (SOP), the preopercular pores (POP), and gill rakers on the first arch (GR) are the characters best used for parental population determination. Comparisons were made of our counts of SOP (male: range 6-7; x = 6.916; female: range 7, x = 7), POP (male: range 7, X = 7; female: range 7-8, x — 7.05), and GR (male: range 6-10, x = 7.916; female: range 6-9, x = 7.75) with those given in Hubbs and Miller. These counts most closely resemble those given for the Florida/Gulf Coast populations by Hubbs and Miller. Vail Lake is an artificial impoundment and the fishes (with the exception of one species of those encountered ) are not native. During the course of the study, Micropterus coosae, redeye bass; Pomoxis nigromaculatus, black crappie; Lepomis cyanellus, green sunfish; C. affinis, mosquitofish; and Gila orcutti, arroyo chub; were encountered. It seems likely that the population of L. parva was introduced with the game fishes during stocking. Lucania parva was also collected by St. Amant in Irvine Lake, Orange County in 1963 as reported by Hubbs and Miller (1965). This introduction is probably the result of shipments of game fishes made in the 1940's from the Dexter Federal Fish Hatchery along the Pecos River, New Mexico. Additional collections of L. parva in Lake Irvine were made in 1964 (Hubbs and Miller 1965). ACKNOWLEDGMENTS The authors thank James Dixon and John McEachran for reviewing an earlier draft of the manuscript. We also thank R. R. Miller and Camm Swift in confirming identifications. The senior author thanks Thomas Fritts for his patience. REFERENCES Hubbs, C. L., and R. R. Miller. 1965. Studies of cyprinodont fishes. XXII. Variation in Lucania parva, its establishment in western United States, and description of a new species from an interior basin in Coahuila, Mexico. Univ. Mich. Mus. Zool. Misc. Pub. 127:104pp. Moyle, P. B. 1976. Inland fishes of California. Univ. Calif. Press, Berkeley. viii + 405pp. — Michael J. McCoid. Department of Herpetology, Natural History Museum, San Diego, CA 92112 and James A. St. Amant. Inland Fisheries, California Department of Fish and Game, 350 Golden Shore, Long Beach, CA 90802. Mr. McCoid's current address: Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, TX 77843. Accepted for publication October 1979. 126 CALIFORNIA FISH AND GAME BOOK REVIEWS Pacific Coast Subtidal Marine Invertebrates: A Fishwatchers' Guide By Daniel W. Gotshall and Laurence L. Laurent. Sea Challengers Los Osos, California 1979; 107 P.; illustrated; $12.50/hardback; $9.50/softback. If a criterion by which books are judged is the proximity of the final product to the authors' goals, then Pacific Coast Subtidal Marine Invertebrates must be rated very highly. In their introduction, Gotshall and Laurent state their purpose is "to enable the sport diver and others to identify most of the larger invertebrates in the area covered by the Guide". In this, they have succeeded admirably. Lising the same style and format as Gotshall's popular Inshore Fishes of the Pacific Coast, this second book in the Fishwatchers' Guide series covers 161 of the invertebrates commonly encountered by divers from Alaska to Baja California. Each species is represented by an excellent, and in some instances outstanding, photograph and a brief text that includes identifying characteristics, size, range, and a few natural history notes. Other aids to identification are included. I was particularly intrigued by the artificial key to the invertebrates and spent some time trying, in vain, to confound it. This key and accompanying drawings function to direct the user to the correct group of organisms. The photographs and descriptions serve to pin down the identification. Two or three of the photographs are not of the same high standards as their cohorts and a few spelling errors occur. However, most of these latter have been noted in an errata sheet and any lack in the photographs would not be as noticeable in lesser company. In general, I enthusiastically commend this book to anyone with a desire to stick their heads below the sea surface. The uninitiated will be educated and old hands will view fine portraits of old friends. — John J. Grant Pacfic Salmon By R. J. Childerhose and Marj Trim; University of Washington Press, Seattle, WA; 1979; 166 pp; illustrated; $25.95. My preliminary evaluation, after first thumbing through Pacific Salmon, was that it was an attractive coffee table book, but probably without much substance. Wrong. To its intended general public audience, it could be titled All You Wanted to Know About British Columbia Salmon. This book is divided into four major sections: history, biology, environment, and enhancement. Through this format the author, science writer R. J. Childerhose, has done an excellent job of summarizing what is known about the five northeast Pacific salmon: chinook, coho, sockeye, pink, and chum. He addresses just about every facet of salmon biology and management, from estuarine life to fish marking to the 2(X)-mile limit. The subjects are treated with enough detail for understanding, but not so much that the reader gets bogged down. A highlight of Pacific Salmon is the excellent color photography of salmon, fishermen, rivers, lakes, hatcheries, habitat, etc., in British Columbia by Marj Trim. As I indicated previously, the book emphasizes British Columbia stocks, with occasional mention of conditions or studies in Oregon, Washington, and Alaska. California populations are dismissed with the statements that ". . .[the] great Sacramento [River] runs were decimated by new fishing methods" and that because of this and hydraulic gold mining, "in a few years the salmon runs of the Sacramento River were gone." These, plus the absence of California from the species distribution maps, incorrectly lead the reader to believe that salmon are gone from the Golden State. My only other complaint is the lack of even a modest reference section which could introduce the interested reader to the scientific literature. Pacific Salmon presents the results of the white man's ecological arrogance with cautious optimism for the future of today's stocks. It closes with a quotation from 1855 of Chief SeathI of the Duwamish tribe, in part: "What is man without the beasts? If all the beasts were gone, man would die from great loneliness of spirit, for whatever happens to the beast also happens to the man. All things are connected. Whatever befalls the earth befalls the sons of the earth." — Dave Hoopaugh Field Guide to the Fish of Puget Sound and the Northwest Coast By David Somerton and Craig Murray; University of Washington Press, Seattle, WA; 1976; 70 pp.; $4.95, paperback (printed on v/ater-resistont stock). This booklet has been described by the publishers as a "long awaited guide to the fishes of Puget Sound and the Northwest waters, from northern California to Alaska, [which] will be welcomed by REVIEWS 127 amateur naturalists, scuba divers, and sports fishermen alike." I am afraid that we are still waiting! In the preface the authors state the purpose of their effort was to produce a field guide "which used features for identification that were easily recognized under a variety of viewing conditions, and which could be taken underwater as well as used on land or in a boat." The authors did not intend to include all Puget Solund fishes but cover 99 species which they "feel will be at least occasionally seen in Puget Sound or the Strait of Juan de Fuca." Twenty groupings of fishes are considered — i.e., (a) salmon, trout, char; (b) rockfish; (c) cartilaginous fish, etc., and the species for each group are then delineated. Depth range and relative abundance (for Puget Sound) is indicated for each species. The authors do state, "don't be discouraged when unable to identify many fish — it takes time." A certain knowledge is initially assumed as only a brief explanation or characterization of the larger groupings is given. I am confused by the authors' varying use of the singular and plural in indicating family groups — i.e., rockfish and seaperch vs. flounders and sculpins. They might have followed the American Fisheries Society's "List of Common and Scientific Names of Fishes" for standardization. The first group of fishes discussed is the salmonids. Eight species are considered (five salmon, two trout, and one char) and the authors point out that these fishes are more often encountered by the angler than the diver. A key is given here, the only one contained, to aid the angler in identifying his fish. It might have been preferable to place the herring and smelt (two species considered) after the salmonids rather than relegating them a rather cryptic space midway through the text between the clingfish and sand lance. The second group, rockfishes, includes 11 of 24 species recorded in Puget Sound. This genus is difficult even to the trained biologist and I wonder about identifications made from diagrammatic drawings without the aid of a key. Distinguishing characters are given but the task still remains difficult. The use of red snapper as a secondary common name for the yelloweye rockfish is unfortunate. This appellation has been applied to all of the reddish rockfishes as well as a number of brownish and greenish forms. It is confusing and should be avoided. Sixteen of 36 sculpins are included and again several of the species would be difficult to identify without more complete information. The greenlings, family Hexagrammidae, are characterized by having several lateral lines. This is correct only for members of the genus Hexagrammos; the lingcod and painted greenling have only one lateral line. The booklet concludes with a list of "common fishes" which "represents most of the fishes likely to be encountered in Puget Sound and along the Washington coast." The inclusion of rosy rockfish is based upon historic literature records which cannot be substantiated and should be changed to rosethorn rockfish. Vermilion rockfish is incorrectly spelled. Although the list is not intended to be complete, several species are not included which should be: northern anchovy, graveldiver, and Pacific pompano, among others. The booklet is printed on water-resistant stock and will serve as a handy addition in the fisherman's tackle box or the diver's gear bag, the audience which will find it most useful. It will be of limited value to the marine biologist. — Robert N. Lea The Abalone Book By Peter C. Howorth; Naturegraph Publishers, Inc., CA; 1979; 77 p.; illustrated; $3.50. The Aba/one Book by Peter C. Howorth is an interesting and educational little text covering most aspects of the abalone from pre-historical usage to contemporary gastronomical preparation. The book appears to be intended for the sport-diving and shore-picking public who are descending on the abalone populations in ever increasing numbers. Of particular interest to these will be the eight pages of color photographs portraying west coast species. Howorth, in cautioning pickers to be concerned over such matters as picking technique, legal size, and bag limits, serves to reinforce vital conservation practices. Indeed, the entire book has a concerned tone with the author demonstrating a good understanding of the problems confronting the abalone resource. In a clear, concise style, Howorth presents sections on abalone evolution, historic and pre-historic usage, life history, species descriptions, and current status. The final section describes the preparation of abalone for food and includes some excellent recipes. A minor error in a figure identification does not mar the overall presentation of an interesting and well-illustrated text that will allow the average abalone hunter to better understand his prey and better care for it once captured. — John j. Crant Photoelectronic composition by 80190 — 800 1-80 4,500 LDA californu office of state pmnting INSTRUCTIONS TO AUTHORS EDITORIAL POLICY The editorial staff will consider for publication original articles and notes dealing with the conservation of the fauna and flora of California and its adjacent ocean waters. Authors may submit two copies, each, of manuscript, tables, and figures for consideration at any time. MANUSCRIPTS: Authors should refer to the CBE SUjle Manual (fourth edition) for general guidance in preparing their manuscripts. Some major points are given below. 1. Typing — ^^All material submitted, including headings, footnotes, and references must be typewritten double-spaced on white bond paper. Papers shorter than 10 typewritten pages, including tables, should follow the format for notes. 2. Citations — All citations should follow the name-and-year system. The "library style" will be followed in listing references. 3. Abstracts — Each paper will be introduced by a short, concise ab- stract. It should immediately follow the title and author's name and be indented at both margins to set it off from the body of the paper. 4. Ahhreviations and numerals — Use approved abbreviations as listed in the CBE Style Mammal. In all other cases spell out the entire word. TABLES: Each table should be typewritten double-spaced throughout wdth the heading centered at the top. Number tables with arable numerals and place them together in the manuscript following the references. Use only horizontal rules. See a recent issue of California Fish and Game for format. FIGURES: Submit figures at least twice fi.nal size so they may be reduced for publication. Usable page size is 4f inches by 7| inches. All figures should be tailored to this proportion. Photographs should be submitted on glossy paper with strong contrasts. All figures should be identified with the author's name in the upper left corner and the figure number in the upper right corner. Markings on figures should be in blue pencil or grease pencil, as this color does not reproduce on copyfilm. Figure captions must be typed on a separate sheet headed by the title of the paper and the author's name. PROOF AND REPRINTS: Galley proof will be sent to authors approxi- mately 60 days before publication. Fifty reprints will be provided free of charge to authors. Additional copies may be ordered through the editor at the time the proof is submitted. Al O -0 i^ r" o ■" m CA 01 ^ S -t Z z 3 (/> -1 I rs C3 :& jp U) m r— -i •X) — 3> J & "'1 m ' TO O -1 w -< SO C' > n z X Z 3> ^ > -* -n o 2 n z 2 «^ ^ a: 23. ■H en m 0 0 n c: > 5 « #*■ 01 S z 1^ CD o m -^ ■'H £> m fii S m CO ;?" C (5 T li in W p o "O O > -H *■ Q o ™ ^ r: m -t>