1!". f__ A I: ~ ' 4' ,w a . £3: -. ...r:,; ,1 .P: ‘12 :3: "r4: .. . I 77:. ' 'zrgn‘ ~ L“ =9? — .J “*“9 52”. .2 -n ,,. a: «.7 .. ’W.” I . ‘ 5;. ‘ ‘I‘st‘trfi‘ttfl «it {55 ‘ Mn N“ "(I \‘Uu ‘ *3 311 , s A 41".“ t " “1'.ng 7-4;" 1-“); 4 ’ ‘ :4.“ ‘ tag—:55 . '55: ,x .4 5.1L" *1. ,. t 3.‘ '§ ‘ , 382M“? mg g. “r“ 4* .4 . . ix. . 'L ‘ . 1‘5} " 3‘ 312:; ;", WT?" {332%} {$5 wk ff. $35533 ' ' ' ~11: "55$" ,. dwxgfiqtfi a, -, J54 ., ‘37 733.”.r1‘3-3‘Ir‘ - . _ ‘33. SW ~ w 51.nv :7 "3% ‘ 1:" ,| in”: , .1“? , "H‘la‘jhgaqfi 45?: #Wk *3 w: W m i, \ 11;?“ it 31““ .ll-l: . m;q} ‘ u ' - [I 3"?” (: Kg; _‘:‘ 1;.“ 53:7“ ( . W 't m , my. “14., ,L .3 ‘ *1 ik* Ms. 3? L. ’ 335415ww~wh "‘fx‘" V“ ‘* 2h: ‘5‘: '. w * .. ”31:53“ F “504113“! a , ‘Lhw ; '~.~x~§17iiili;!1f=h§ 1 fihfigg‘ifit‘l "MW: I. 454-31 'Pvfi‘fih‘i‘: ‘E‘U ‘Ci‘i‘lP-lgv “1‘1?” My .1“. xxx 5N: LAME 1 1“ ( fiYlemrlehgfifik‘Efig‘l 4 L «. gram“: ( t . ~ L: l l .e LL‘L ‘ A 'l hinflnfih Qi‘fimmififi, W“ .il flux » 1:56;"! M“gasi«:8«ias$§’3~%§& “A! if! 3‘ m" u L‘ gfififi 1. i . aw: - » ~- ‘ v t..- n, 1‘ WI “1 'u t ; ' 'v «A r» "if 5g H. 1 I”; L > y .0 I 3h, ‘5“? ‘8 ~b ‘ 5%“ s! r! ‘V Ill]: 437‘ i“. . ' ”12-4. 1“ ~-"~;€.¢{&=n§dm;“ ‘E ‘ T figfig'fi" . . n 7'7 6 . ‘ 57“.“ w“ 1., ‘_-,‘.~'.- 'm H“; a: ‘2‘ W‘s Zimghi‘zgmi‘ ’ $353 13"»: ‘ WEE” Maggi “k. ’ ‘ v ', ‘il:_ Eras" r} r; 11“ v j . If: . 112%“ ‘1 ‘11: $1 1 " 4%:6 .~ ! ‘ 5. *1 ¢ ‘1 _ k } ft « '1‘ ’1. 'o 2"? [i v’ .1 vagrwmjnvnvn‘t'fli“ » 71.91292, .1: 14' . £51913?" .. .14.. 4C . rs jéiwfi o ‘3‘ flash .7. ar‘ mm» l]! i“ 'IMH‘IWHIIMWM L _. 35A 3 1293 01094 8127 flaggeesity dissertation entitled Photoperiodic Effects On The Avian Dietary LC50 With Bobwhite (Colinus Virginianus) And Mallards (Anus Platxrhxnchosl presented by William J. Breslin has been accepted towards fulfillment of the requirements for Ph.D. degeem Animal Science 1 i M o > Major professo Robert K. Rin 7-27—84 ger Date MSU i: an Affirmative Anion/Equal Opportunity Institution 0.12771 _ .[.{.{ .q_,wrtr 1 Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. LIBRARIES “ i 30!? 5.’ 3 g'gqs‘ -.{-;.~.: 555% iL..\G|02 56.8% mg 4‘1 2.4-"- ocr {from William J. Breslin A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Animal Science and Center for Environmental Toxicology TABLE OF CONTENTS ABSTRACT .............................................. INTRODUCTION . ....... ..... ............ . ..... . .......... REVIEW OF LITERATURE .................................. Regulation of Food Consumption .................... Dose-Response Relationships ....... . ............... Chemicals ......................................... ANTU ...................................... .... Fenthion ...................................... Endrin ........... ' ............................. Sodium Secobarbital ........................... Strychnine .................................... PROCEDURE ............................................. Design and Chronology of Study .................... Experimental Design ............................... Range Finding - Preliminary Tests ................. Facilities ........................................ Animals ........................................... Diets ....................................... t ..... Observations of Record ............................ Statistical Analysis .............................. RESULTS ............................................... Litchfield and Wilcoxon Analysis .................. Direct Regression of Dose on Mortality ............ Inverse Prediction of the Regression of Mortality on Dose ............................... DISCUSSION ............................................ CONCLUSIONS ........................................... REFERENCES ............................................ APPENDIX .............................................. 12 l7 l7 l8 19 20 21 22 22 24 24 26 27 30 31 33 35 35 48 48 74 79 8O 83 lung-Ioduh-IUIIII-nb -.- _ ..... ..'---" - I' .. SI o--.--o.-n..- Illa... ' _ :H ' '1' LIST OF TABLES Table # Page 1 Chronology of LC50 trials ................. 23 2 Percent mortality for bobwhites and mallards fed ANTU, secobarbital, and strychnine during preliminary range finding tests ............................. 25 3 Brooder and room light temperature ........ 28 4 Room light intensities for the 14D photOperiod during the brightening and dimming sequence .......... 29 5 Dietary concentrations (ppm) used during LC50 testing ....................... 32 6 LC50 values for ANTU, fenthion, endrin, secobarbital, and strychnine in bobwhites and mallards calculated by the probit analysis method of Litchfield and Wilcoxon (1949) ............ 36 7 Percent mortality and LC50 values for bobwhites fed dietary ANTU ............ 37 8 Percent mortality and LC50 values for bobwhites fed dietary fenthion ........ 38 9 Percent mortality and LC50 values for bobwhites fed dietary endrin .......... 39 10 Percent mortality and LC50 values for bobwhites fed dietary seco— barbital .................................. 40 11 Percent mortality and LC50 values for bobwhites fed dietary strychnine ................................ 41 12 Percent mortality and LC50 values for mallards fed dietary ANTU ............. 43 13 Percent mortality and LC50 values for mallards fed dietary fenthion ......... 44 II Table # 14 15 l6 17 18 19 20 21 22 23 24 Percent mortality and LC50 values for mallards fed dietary endrin ........ Percent mortality and LC50 values for mallards fed dietary seco— barbital ............................... Percent mortality and LC50 values for mallards fed dietary strychnine ............................. LC50 values for ANTU, fenthion, endrin, secobarbital, and strychnine in bobwhites and mallards calculated by log- probit analysis using a direct regression of dose on mortality ........ LC 0 values for ANTU, fenthion, en rin, secobarbital, and strychnine in bobwhites and mallards calculated by a log— profit analysis using an inverse prediction of a regression analysis of mortality on dose .......... Mortality patterns of bobwhite and mallards during LC50 testing with secobarbital ...................... Mortality patterns of bobwhite and mallards during LC50 testing with strychnine ........................ Mortality patterns of bobwhite and mallards during LC50 testing with fenthion ............................... Mortality patterns of bobwhite and mallards during LC50 testing with endrin ................................. Mortality patterns of bobwhite and mallards during LC50 testing with ANTU ................................... Observed symptoms of toxicosis for bobwhite and mallard LC50 trials ....... III ..... ..... ..... Page 46 47 49 50 52 53 55 56 57 58 Table # 25 26 27 28 29 3O 31 32 Mean feed consumption (grams/bird/day) by photoperiod of bobwhites and mallards fed dietary ANTU, fenthion, endrin, secobarbital, and strychnine at day five of study Mean feed consumption (grams/bird/day) by photoperiod of bobwhites and mallards fed dietary ANTU, fenthion, endrin, secobarbital, and strychnine (three day recovery period) Initial mean body weights (grams) by photoperiod of bobwhites and mallards fed dietary ANTU, fenthion, endrin, secobarbital, and strychnine Mean body weights (grams) by photoperiod of bobwhites and mallards fed dietary ANTU, fenthion, endrin, secobarbital, and strychnine at day five of study Mean body weights (grams) by photoperiod of bobwhites and mallards fed dietary ANTU, fenthion, endrin, secobarbital, and strychnine at day eight of study Mean feed consumption (grams/ bird/day) by treatment of bobwhites and mallards fed dietary ANTU, fenthion, endrin, secobarbital, and strychnine for the five-day treatment period Mean feed consumption (grams/ bird/day) by treatment of bobwhites and mallards fed dietary ANTU, fenthion, endrin, secobarbital, and strychnine for the three—day recovery period Initial mean body weights (grams) by treatment of bobwhites and mallards fed dietary ANTU, fenthion, endrin, secobarbital, and strychnine IV ooooooooooooooooooooo oooooooooooooo ............................... oooooooooooooooooooooooooooooooooooo .................................... ooooooooooooooooooooooooo -------------------------- Page 60 61 63 65 66 68 69 71 ¢.-.a.'-‘..I oo-honor-cocoooo-cacooloacoroqiln-i." éfl' “ ' Mean body weights (grams) by treatment of bobwhites and mallards fed dietary ANTU, - fenthion, endrin, secobarbital, :- and strychnine (day eight of -- - study) 73 LIST OF APPENDIX Page Appendix A. Composition of quail starter ...... ....... 83 Appendix B. Composition of duck starter . ............. 84 Appendix C. Room temperature, brooder temperature, and room relative humidity during LC50 trating ............................. 85 VI '1 .I' _"(I". ." in aiming-no .A rttuseq 912.115 ?-: ---.'.-.- '.' comm: . *- 'i F. ABSTRACT PHOTOPERIODIC EFFECTS ON THE AVIAN DIETARY LC50 WITH BOBWHITE (COLINUS VIRGINIANUS) AND MALLARDS (ANAS PLATYRHYNCHOS) by William J. Breslin Eight-day (five-day treatment and three-day recovery period) dietary LC50 trials were conducted with bobwhite (Colinus virginianus) and mallards (Anas platyrhynchos) using a-napthylthiourea(ANTU), fenthion, endrin, sodium secobarbital, and strychnine under 24 hours lightzo hours dark (24), 14 hours lightle hours dark (14), and 14 hours lightle hours dark with dimming and brightening of lights between light and dark periods to simulate dusk and dawn (14D), to determine the effect of photoperiod on the avian dietary LC50. No signifi- cant differences in LC50 values between the three photoperiods could be established for a particular chemical in bobwhites or mallards. Mortality patterns and symptoms of toxicosis were generally similar between the 24, 14, and 14D photoperiods for each chemical. Photoperiod significantly affected feed consumption during the five-day treatment and three—day reco- very periods. The 24—hour photoperiod resulted in significantly greater feed consumption than either of the two 14-hour photo— periods. Significant differences in feed consumption between the birds on the 14 and 14D photoperiods were less frequent ed consumption c: birds (D ifferences between (1: than significant INTRODUCTION The Environmental Protection Agency has established testing guidelines for the registration of pesticides in the United States under the Federal Insecticide Rodenticide and Fungicide Act (FIFRA) (U.S. EPA, 1982) and for the determination of safety of industrial chemicals under the Toxic Substance Control Act (TSCA) (U.S. EPA, 1983). Portions of both acts deal with wildlife toxicity testing. At present, wildlife toxicity testing guidelines are not final and information required for a particular pesticide is determined on a case- by-case basis (Federal Register, 1978a). The oral LD50 and _ dietary LC50 are two tests frequently required early in the registration process, supplying basic lethality and species susceptibility data for a particular chemical. The avian dietary LC50 protocol, having no similar mammalian counterpart as does the LDSOI allows considerable variation in the methodo— logy of experimentation, resulting in the potential for significant variations in test results and thus making it difficult for interpreting and categorizing a chemical’s toxicity. One section of the proposed avian LC50 protocol allowing considerable experimental variation deals with photo- period (Federal Register, 1978b). Photoperiod, the relative length of light and dark periods, is one of many important factors which control the behavior and physiological state of most domestic and wild bird species (Morris, 1967; Van Tyne and Berger, 1976; Tucker and Ringer, on: so! aeoliab:up m‘lfil .': .s-t-e” 5.1." ._ -‘-" '5' 5c nel’lfltflrsdob ed- "2‘ U.r I - - ”’i --.‘-‘II‘EII.I'-lf .. -. l.- :.- . 1982). Of particular significance to toxicity testing is the effect that light duration has on the bird's activity, feeding behavior, reproductive success, and the biochemical inter— actions which occur during these events. Since in LC50 testing the quantity of toxic compound ingested is dependent on food consumption, the effects of photoperiod on feeding behavior is of critical concern. In studies with domestic chickens, increasing photoperiod lengths from 12 hours light:12 hours dark to 24 hours 1ight:0 hours dark significantly altered the number of feedings, feeding duration, and feed consumption (Squibb and Collier, 1979). Overall, chickens reared under 24 hours light consumed signi— ficantly greater amounts of feed than chickens reared under diurnal lighting schemes. The birds reared under 24 hours light consistently (at short intervals) consumed small amounts of feed throughout most of the 24 hour period (Squibb and Collier, 1979), while birds raised on diurnal lighting schemes consumed large quantities of feed just prior to darkness and shortly after the onset of lighting (Dingle, 1971; Squibb and Collier, 1979). These differences in feeding patterns could possibly result in significant variations in toxicity. Birds which consume large quantities of treated feed at the initia— tion or cessation of the lighting periods would be exposed to greater doses of toxicants over a shorter period of time compared to birds consuming smaller quantities of feed over an extended period. Thus, chemicals which are more acutely toxic and less persistent may be more toxic when administered to birds on diurnal lighting schemes. Conversely, less acutely toxic or more persistent chemicals may be more toxic when administered to birds under continuous light due to the over— all increased feed consumption and increased chemical intake. In addition to feeding behavior, significant increases in feeding efficiencies have been reported in chickens exposed to continuous light, resulting in increased early growth rates and heavier body weights (Squibb and Collier, 1979). This increase in feeding efficiency has been attributed to a continuous period of food processing, which is interrupted in birds exposed to alternate lightzdark periods. Test birds housed under continuous light for the standard five-day pre- test acclimation period may show significantly greater body weights than diurnally acclimated birds at the onset of testing, making test results incompatable between studies using different lighting schemes. Another potentially important effect of photoperiod on the avian LC50 is diurnal fluctuations in the level and activity of microsomal mixed function oxidase (MFO) enzymes. Fluctuating MFO enzyme activities may result in significant diurnal changes in the toxification—detoxification reactions producing oscilations in the levels of toxic compounds within the animal. The purpose of this study was to assess the potential effects various lighting schemes have on dietary LC50 deter- minations in bobwhite (Colinus virginianus) and mallards (Anas u-i mart-m Milli-isle .':-:..z-.—.= - Elatyrhynchos). The study was funded by the Environmental Protection Agency on a contract basis in order to help develop a more uniform testing protocol for subacute dietary toxicity testing with avian wildlife species. REVIEW OF LITERATURE Regulation of Food Consumption Sturkie (1976) states "Animals eat to satisfy energy requirements and volume receptors or to attain a state of fullness or satiety". The amount of feed consumed is depen- dent on many factors, some of which include photoperiod, activity, age, size, environmental temperature, reproductive stage, appearance and taste of food, and availability of water. In mammals, it is generally recognized that the satiety and appetite centers are located in the ventromedial and late- ral nucleus of the hypothalamus, respectively. Studies in various species of birds utilizing hypothalamic lesion techniques also indicate that the ventromedial nucleus and the lateral hypothalamus play an important role in the regulation of feed intake (Sturkie, 1976). Lesions in the ventromedial nucleus at the base of the third ventricle above the optic chiasma caused hyperphagia in chickens (Gallus domesticus) and white—throated sparrows (Zonotrichia albicollis) while lesions in the lateral hypothalamus just caudal to the ventromedial nucleus produced aphagia (Smith, 1969; Kuenzel, 1972). The exact mechanism by which the hypothalamus controls feed intake has not been determined. Certain investigators theorize that adipose tissue stores and energy requirement regulate appetite and feed intake. Mu gt 31. (1968) proposed that there is a "set point" for tissue fat concentration. If tissue lipid concentrations rise above this "set point", feed intake decreases and lipolysis is accelerated. Conversely, when the tissue lipid concentrations drop below the "set point" feed intake increases and lipogenesis and fat deposition are stimulated. When fat deposits approach or equal the "set point" an equilibrium forms stabilizing feed intake and body weight (Lepkovsky, 1973). Feed intake and indirectly the size of the fat droplets are regulated by the hypothalamus. Long term changes in eating behavior and fat deposits caused by hyperphagia or hypophagia are thought to be due to the develop- ment of new set points. Lepkovsky (1973) supported the "set point" theory of feeding regulation by reporting that continued force feeding of chickens at twice their normal ad libitum intake and nutri— ent requirements made the birds obese. After the force feeding was terminated and the birds were allowed to feed ad libitum, they refused to eat for 7 to 10 days, lost weight, and depleted their abnormally high fat stores, theoretically to a level low enough (set point) to trigger the feeding center. Hill and Dansky (1954) also provided data which support energy requirements as a controlling factor in feed intake. When these authors varied the caloric content and nutrient density of feed, chickens altered their food consumption. Hill and Dansky also noted that the differences in energy intake between birds of the same weight and egg laying performance could be attributed to other factors such as the volume of feed eaten or the stimulation of hypothetical volume receptors in the crop and esophagus. Polin and Wolford (1973) conducted studies in chickens, the results of which contradict the "set point" hypothesis while supporting the theory that volume receptors are impor— tant in controlling the feeding behavior of birds. The authors fed groups of chickens dd libitum, dd libitum and force fed, and force fed 150 percent of the ddlibitum group. The data showed a direct relationship between increasing blood lipids and adipose tissue with increasing feed intake, but failed to show any difference in feed consumption 2 to 3 days after the termination of force feeding even though there was considerable differences in fat deposits among the groups. Polin and Wolford concluded that their study provided evidence that volume receptors, which are influenced by rate of filling, capacity, and discharge of feed, control the feeding process and suggested that hormones or other factors such as energy requirements regulate the "set point" at which these receptors operate. They also pointed out that it is well known that the emptiness, fullness, or distention of the digestive tract in mammals influence food intake. Squibb and Collier (1979) studied the individual eating patterns of broiler chicks from day one to 20 days of age under three different lighting schemes (12 hours light:12 hours dark (LD); continuous light (LL); and continuous dark (DD)). The LL and LD lighting regimes were initiated on the first day after hatch while the DD regime started with a light dark period on day one after hatch, followed by a gradual decrease in light intensity until total darkness was reached on day 8. Previous studies showed that when chicks were immediately placed in total darkness upon hatching 75 percent of the birds died due to their inability to find food and water. As early as day 2 of the trial, the chicks on conti— nuous lighting were eating roughly twice as many meals per day and spending approximately 40 percent more total time eating per 24 hour period than the LD chicks. Although the LL chicks consumed more meals and spent a greater amount of total time eating per 24 hour period early in the study, the LD chicks spent more time eating per meal. At day 3 the LD chicks ate for approximately 13 minutes per hour; the LL chicks ate 6 minutes per hour; and the DD chicks 11 minutes per hour. At day 11 the LD chicks had increased their hourly eating time to 25 minutes while the LL and DD chicks ate an average of 12 and 15 minutes per hour, respectively. The differences between the LL and LD lighting schemes in the number of meals and the time spent eating per meal continued and inten- sified throughout the study. During days 8 through 20, the birds on the continuous lighting schedule ate an average 1.6 times the number of meals than the birds on the diurnal photo- period but spent approximately 50 percent less time eating per meal. The total time spent eating per 24 hour period was 10 greatest in the chicks under the diurnal lighting scheme at the completion of the study while the chicks reared in conti— nuous light or dark spent roughly an equal amount of time eating per 24 hour period. Throughout the study, the LL and DD chicks ate each hour while the chicks on the diurnal photo— period ate every hour during the 12 hour light period. The LD chicks did not attempt to feed at any hour during the 12 hours of darkness. In this same study by Squibb and Collier (1979), at day 3, hourly feeding duration within a lighting regime remained relatively constant under all lighting schemes, except for a sharp surge during the hour the caretaker entered the room. Individual variability in minutes eating per hour at day 3 was greatest in the LD chicks. These birds were unable to anticipate "lights out" at this time, as the time spent feeding in the 2 hours prior to the beginning of the dark period was decreasing. By day 11, the LD chicks still showed a high degree of variability in hourly feeding time, but were able to anticipate lights out, as indicated by increasing time spent feeding during the four hours prior to darkness. At day 20, the chicks had developed a strong sense of timing in anticipation of lights out as shown by a very low indivi- dual variability in hourly feeding time and by a steep rise in the hourly feeding time just prior to the 12 hour dark period. There was no significant difference in the total 20 day feed consumption between the LL and DD chicks but the LD chicks ate significantly less feed than either the LL or DD "ad m”:- m"- Jet-mg 1mm! M'. 10:: pan» B-dflflbmfi' 1.21mi -. um! Ila-Lu: 9.1:. ain'rn'n GU MT .h‘i-Mt'! fair-g" -: rm! - . . - . '- - .. 3-! 11 chicks. Twenty day body weights of chicks on continuous light were significantly heavier than body weights of the LD chicks, while the DD Chick's 20 day body weights were intermediate to the LL and LD birds. Efficiency of feed utilization was high— est in the LL chicks, intermediate in the LD chicks, and lowest in the DD birds. Water to feed ratios were significantly lower in the DD group, but similar in the LL and LD birds. An earlier study by Weaver and Siegel (1968), in which commercial broiler crosses were reared under continuous diurnal photoperiods, supports the feeding behavior conclu- sions of the Squibb and Collier study. Weaver and Siegel reported that the average percentage of birds feeding per hour during the light period was consistently lower under conti- nuous light than that for the flock under the diurnal lighting regime. These researchers identified a highly significant time-light regime interaction in feed consumption in all photoperiods studied. Although the birds on the continuous light schedule showed feeding rhythms, the rhythms were less uniform and dramatic than those under the light-dark photo— period, and produced no evidence of a day—night pattern. The birds on the diurnal lighting regime showed 2 large peaks in the percent of birds feeding; one just after the onset of light and the other just prior to darkness. These increases in percent feeding were due to the birds consuming large quantities of feed after the onset of lighting and prior to darkness, and showed that the chickens were able to condition themselves to the anticipated dark period. m m .mm 8.1 has 3.! m “1.4 saith-amok? nit-1.1:; 3.: and: 31 fins 'r 2v (:1 ' 2' iaéiiélnrfl E ‘E -' 12 Dose—Response Relationships One of the major fundamental principles of pharmacology and toxicology is that of the relationship between the inten- sity of a response caused by a drug or xenobiotic and the dose administered. This dose—response relationship can be best summarized with three basic principles. (1) There are mole- cular or receptor sites in which the chemical reacts to produce a response. (2) The production and intensity of the response are related to the concentration of the chemical or toxic agent at the reactive site or receptor. (3) The concentration of the agent at the receptor site is related to the dose administered or exposure level. In addition, the intensity of the response elicited by the agent is thought to be propor— tional to the number of receptors occupied by the agent, with maximal response occurring when all receptors are occupied. This receptor occupancy is a function of drug concentration and its ability to react or combine with the receptor. The ability of a compound to react with a receptor or the drugs affinity for the receptor is generally considered to be constant (Klaassen and Doull, 1980). In general, the toxic effects of a chemical in an organism are not produced unless the agent, its metabolites, or conver- sion products reach and react with receptor sites or macro- molecules in concentrations and for a length of time to cause an alteration in normal function. Thus, the chemical's inherent toxicity, or its inherent ability to produce harm in an organism, and the degree of exposure of an organism to the -..I r. cr- ns oldfifl’l "; .- - - - - . 'uL- 13 toxin are two important factors which determine toxicity. In addition, many factors contribute to the degree of inherent toxicity of a chemical and the exposure situation. Two important aspects of exposure which play a role in the develop- ment of toxicosis are the duration and frequency of exposure. Many agents show quite different effects between acute and chronic exposure. Acute exposure to agents that are rapidly absorbed tends to produce immediate toxic effects, while chronic exposure to a toxin may produce both an acute effect, after each successive dose, and a long term effect due to the chronic presence of low levels of the compound. Fractionization of the dose can also reduce the intensity of the chemical's effect. Administering a specific quantity of chemical in two doses over a period of several hours may result in less than one half of the effect that would have been obtained had the same amount of chemical been given in a single dose. This "fractionization effect" occurs due to the compound being metabolized or excreted between successive doses, thus reducing the peak tissue or blood concentrations or by the ability of the organism to partially or fully reduce the injury caused by the first dose prior to the second. In certain incidences, the production of a toxic response can be totally dependent on the frequency rather than the duration of exposure (Klaassen and Doull, 1980). The absorption or elimination of most drugs and xenobio- tics follow expotential (first order) kinetics. A constant proportion of the compound administered is absorbed or 1h manure ‘o usages erasing-l - ME. 35135114' .--="' 11".:- . ';‘.-'.'- H“ .‘-'.I'.-r.: has a '1. 2: .".-‘.'!E*\-'1'..":§ 3?"; i"-- l4 excreted per unit of time, since the chemical concentration within the organism usually does not saturate the mechanism responsible for absorption and excretion. In certain cases where the elimination process is restricted due to the limited quantity of carrier molecules, drug metabolizing enzymes, or cofactors in the metabolism pathway or the concentration of the toxic compound saturates the excretion mechanism, zero order elimination (constant quantity per unit time) kinetics results. In first order kinetics an absorption or elimination rate constant (K) which expresses the fractional change per unit of time can be calculated. From this constant, a half—time, (.693/K) or the time required for 50 percent of the admini— stered dose to be absorbed or excreted, can be determined. Both the rate constant (K) and the half—time are independent of drug concentration and dose. From these formulas, it can be determined that the effect of a single dose can be charac- terized by its latency, time to peak effect, time to peak concentration, magnitude of peak concentration, and the dura- tion of effect. As the dose increases, the latency is reduced and the peak effect is increased without altering the time of peak effect. The duration of effect increases with increasing dosage but proportionally less than peak effect. In first order kinetic models, 4 half-times are required for near complete elimination (94%) of a compound from the organism, and any dosing interval shorter than this results 15 in the accumulation of the compound. The accumulation of the compound continues during successive doses until the rate of elimination equals the rate of absorption. Once this equili- brium has been reached the body concentration becomes a func— tion of the maintenance dose (dose/dose interval) or exposure rate and the half-time for elimination. Thus, if two dosing intervals are used for a compound but the total drug admini- stered within a 24 hour period is equal between the two dosing schedules, the plateau levels of tissue stores would be equal, but the peak tissue concentrations would be different. Light has been known to alter the metabolism and response to drugs in many animal models. In general, light can alter the organism's response to drugs or xenobiotics by two mechanisms; photochemical reactions or changes in the levels and activity of drug metabolizing enzymes. Photochemical reactions can be either deleterious or beneficial. Chemicals, such as selected sulfonamides, tetracyclines, nalidixic acid, sulfonylureas, thiazides, phenothiazines, and coal tars, when chronically ingested, result in photosensitivity after expo— sure to sunlight or artificial fluorescent light (Klaassen, 1980). Two types of photosensitive reactions have been obser— ved; a phototoxic reaction which is characterized by the fundamental dose-response relationship and results from the production of a reactive chemical metabolite, and the photo— allergic reaction which is an idiosyncratic response charac- terized by the promotion of a chemical reaction between the drug and subcutaneous proteins resulting in the formation of 16 a photoantigen (Klaassen, 1980). Conversely, photochemical reactions such as that observed in the phototherapy of pre- mature infants with neonatal hyperbilirubinemia result in the formation of less toxic photometabolites. When infants with hyperbilirubinemia are exposed to blue or visible light, the radiation penetrates the skin photooxidizing bilirubin to a more hydrophilic product which can be easily eliminated by both biliary and urinary excretory pathways (Sanvordeker and Lambert, 1974). Similarly, alterations in drug metabolizing enzymes can be deleterious or beneficial depending on the nature of the chemical and its metabolites. Light affects the drug metabolizing enzyme system by promoting rhythmic changes in the activity and production of enzymes. In rat studies conducted by Joir SE Ei' (1971) using female Long-Evan rats kept under 12 hours light:12 hours dark, liver microsomal drug metabolizing enzyme levels peaked during the dark phase of the diurnal photoperiod, then declined to minimum levels during the 12 hour light period. Nair and Casper (1969) reported similar diurnal fluctuations in hepatic drug metabo— lizing enzyme actitivy in male Sprague—Dawley rats raised under a 12 hour light:12 hour dark schedule. In addition to the diurnal photoperiod, Nair and Casper also studied the effects of continuous light and continuous darkness on the hepatic drug metabolizing enzymes of rats. Both of the continuous lighting schedules abolished the daily rhythms in enzyme activities. The rats raised in total darkness main- tained significantly higher enzyme levels than the rats raised ' - . ' 5..1 J u. d 1 “.|- f I. I '(_ u ‘1- u {.4 . 17 under continuous light. When phenobarbital sleeping times were measured in these two groups of rats, the animals exposed to continuous light slept significantly longer than the rats on the dark schedule indicating light can indirectly alter the response to drugs by affecting drug metabolizing enzyme levels. Chemicals The chemicals used in this study were selected on the basis of their ability to affect the nervous system. The five chemicals represent agents that stimulate, inhibit, or have no effect on the nervous system. ANTU Synomyms: a-napthylthiourea; a-naphylthiocarbamide; Krysid Description: A colorless to gray, odorless, bitter tasting, crystalline compound with a melting point of 198°C and solubilities of 0.06 g/100 ml water and 2.43 g/100 ml acetone. Molecular weight: 202.27 Molecular formula: CllHlONZS Use, mode of action, and clinical signs: ANTU is a rodenticide that is highly specific for the adult Norway or brown rat (Rattus norvegicus) while being less toxic to other Rattus species and safe to most domestic animals. Of the domesticated animals, dogs, cats, and swine are the most sus— ceptable. ANTU is a fast acting toxicant which causes an increase in pulmonary capillary permeability leading to plural "when“: Sumter-Iii . entreoibnl slnheflss swab an! llodsaam put! n ' w'. "a : . . 'H“th7 18 effusions, pulmonary edema, and anoxia. Clinical signs include salivation, vomitation, diarrhea (general gastric irritation), dyspnea, coughing, tachycardia, and pulmonary rales. Post- mortem examinations reveal cyanosis, hydrothorax, inflammation of the trachea, bronchi, and gastrointestinal tract, and hyperemia of the kidneys and liver. Chemical purity: 95% Source: K and K Laboratories - ICN, 121 Express St., Plainview, NY. Lot number: 45898A Fenthion Synonyms: Baytex, Baycid, ENTEX, Lebaycid, Mercap- tophos Description: A yellow, oily liquid with a slight odor of garlic, a boiling point of 87°C (commercial product boiling point 105°C) and a vapor pressure of 3.0 x 10"5 mmHg at 20°C. It is readily soluble in methanol, ethanol, ether, and other organic solvents. Its solubility in water is 5.5 mg/100 ml water. Molecular weight: 278.34 Molecular formula: C10H15O3P82 Use, mode of action, and clinical signs: Fenthion is an organophosphate insecticide which can be readily absorbed through the skin, mucous membranes, lungs, or digestive tract. Its toxicity results from irreversible binding to acetyl- cholinesterase causing acetylcholinesterase inhibition which 19 leads to overstimulation of the parasymphathetic nervous system. Clinical signs include lacrimation, salivation, tracheal congestion, gastrointestinal disturbances (hyper- motility), muscle stimulation, convulsions, ataxia, immobility, and dyspnea. Post—mortem findings after acute exposure may be minimal and nonspecific while subacute or chronic exposures result in excessive secretions within the respiratory tract. Chemical purity used: 94% Source: Mobay Chemical Corp.; Agricultural Chemicals Division; P.O. Box 4913; 8400 Hawthorn Road; Kansas City, MO. Batch number: 8030130 Endrin Synonyms: Mendrin, Nendrin, Hexadrin, Compound 269 Description: A white crystalline solid or powder (technical powder tan) with a melting point of 226-230°C, a vapor pressure of 2.0 x 10'7 mmHg at 25°C and solubilities of 17 g/100 ml acetone, 13.8 g/100 ml benzene, and 7.1 g/100 ml hexane. Molecular weight: 380.93 Molecular formula: C12H8Cl6O Use, mode of action, and clinical signs: Endrin is an organochlorine insecticide which was used mainly on field crops, particularly cotton. Its specific mechanism of action has yet to be established, but toxicity is thought to result primarily from the effects endrin and endrin metabolites have on the central nervous system. Chlorinated hydrocarbon insec— ticides in general act as diffuse stimulants or depressants 20 of the central nervous system. Clinical signs due to endrin toxicity are: convulsions, tremors, vomiting, abdominal distress, drowsiness, lethargy, and ataxia. Respiratory failure is considered to most common cause of death. The liver, being a secondary site of action, shows signs of fatty infiltration, hypertrophy, and proliferation of smooth endo- plasmic reticulum. Chemical purity used: 99% Source: The Anspec Company, Inc.; P.O. Box 7730; 122 Enterprise Drive; Ann Arbor, MI. Lot number: 5—171 Sodium secobarbital Synonyms: Sodium Meballymal, Sodium Seconal, Sodium Bipinal, Immenoctal, Pramil, Quinalspan, Sebar, Sedutain Description: A white, bitter tasting, hygroscopic powder with a melting point of 100°C. It is very soluble in water, soluble in alcohol, and practically insoluble in ether. Molecular weight: 260.27 Molecular formula: C12H17N2Na03 Use, mode of action, and clinical signs: Sodium seco— barbital is a short acting barbituate. Barbituates in general act to reversibly depress the activity of all excitable tissue. Depression at sensitive CNS synapses may result from a pre— synaptic decrease in transmitter release or by enhancing the postsynaptic GABA responsive membranes. In the peripheral nervous system, barbituates depress the threshold of spinal 21 reflexes by diminishing the response of sympathetic ganglia to preganglionic stimulation. Respiratory complications result from the depression or elimination of the hypoxic and chromoreceptor drives controlling respiration. Barbituates also have various effects on the liver, kidneys, gastrointes— tinal tract, and skeletal muscles. Symptoms of poisoning include drowsiness, confusion, loss of coordination, lethargy, and ataxia. Death may result from cardiovascular collapse or respiratory arrest. Chemical purity used: 99% Source: Sigma Chemical Co.; P.O. Box 14508; St. Louis, MO . Lot number: 059C0207 Strychnine Synonyms: Mole death, Mouse—nots Description: A colorless to white crystalline powder with a melting point of 270~280°C and solubilities of 1.56 x 10'6 g/100 ml water, 6.67 x 10‘1 g/100 ml alcohol, and 5.56 x 10"1 g/ml benzene. Molecular weight: 334.4 Molecular formula: C21H22N202 Use, mode of action, and clinical signs: Strychnine is an indole alkaloid used as a pesticide, ruminatoric, and stimulant. Its primary use as a pesticide is in bird and mammal control. Strychnine causes neural stimulation by antagonizing spinal and medullary post—synaptic inhibition, 22 preventing the moderating and controlling effects in normal reflexes. Early poisoning signs of nervousness, tenseness, and stiffness are followed by tetanic seizures, convulsions, muscle tremors, ataxia, and rigor. Convulsive seizures can be initiated by external stimuli such as light, sound, and touch (hypersensitivity). Clinical signs may appear within ten minutes of exposure and remain for up to two hours if untreated. Death usually results from anoxia brought about by convulsions. Chemical purity used: 96% Source: Pfaltz and Bauer, Inc.; Division of Aceto Chemical Co., Inc.; 375 Fairfield Ave.; Stamford, CT. Lot number: 24763 PROCEDURE Design and Chronology of Study The study consisted of 30 individual LC50 tests conducted over an 11 month period from March 1982 through January 1983. The 30 tests were run in ten separate trials each consisting of three tests per trial (see Table l for chronology of trials). During each trial one of five chemicals (ANTU, fenthion, endrin, secobarbital, or strychnine) was tested in bobwhite or mallards. Individual tests within a trial were identical with the exception of photoperiod. The three photo— periods used were: 24 hours light:0 hours dark (24); 14 hours light:10 hours dark (14); and 14 hours light:10 hours dark with dimming and brightening of lights between light and dark 23 Nm\m \N Nm\a \k -Nm\s \N me\s NM -Nm\_ \k Nm\_ NM Ne\_ \N Mmm\nwws w:_:;uxtnm Nm\kw\__ Nm\aw\__-wm\ew\__ Nm\em\__-we\a_\__ Ne\a_\__ Nw\a_\l_-we\e_\__ .es_srmsouam ~m\_m\m Nm\_w\m -Nm\m_\m Nm\e_\m -Ne\m_\m Ne\m_\m Nm\m_\m -~e\m \@ =_rs:L Nm\mw\a Nm\mm\a -Na\om\k Nm\om\k -Nm\m_\k ~e\m_\k Nm\m_\k -Nm\o_\a :o_;3:wl Nm\mm\o_ ~e\am\o_-mm\o~\o_ Ne\em\o_-Ne\_N\o_ Nm\_m\o. Nm\_m\o_-ww\e_\o_ :_z< Seawmm: ~w\e_\m Nm\e_\m -~e\__\m Nm\__\m -Nn\e \m Ne\e \m Ne\o \m -Nm\_ \m w:_:;ustsm mm\e_\_ mm\e_\_ -mm\__\_ mm\__\_ -ms\o \_ me\e \_ mm\e \_ -mm\_ \_ _ao_2tesouam Ne\mw\e ~w\mw\e -Nm\ow\e Ne\oN\e -Ne\m_\e Na\m_\e Nm\mi\e -Ne\o_\e :_rs:u Nw\m \e ~m\m \e -~w\e \e Nm\e \e -Nm\_ \4 Nm\_ \4 ~w\_ \e -Na\am\m =o_;s:at mm\m \_ mw\m \_ -Nm\_M\N_ Nm\_M\N_-Ns\oN\N. Nm\o~\m_ Nw\em\~_-~w\_~\w_ =_z< amazemmm :owum:_snoh uo_noa :o_umtom_mM5cm broom : uo_gwa - . .mwmmmem ato>ooom _ao_eon do eo_noa amok :owume__oo< 24 periods to simulate dusk and dawn (14D). Eight treatments (two controls and six chemically-treated groups) of ten birds per treatment were used within each LC50 test. Experimental Design Photoperiods = 3 (24L:0D; 14L:10D; l4L:lOD with dimming and brightening) Chemicals/photo. = 5 (ANTU, fenthion, endrin, secobarbital, strychnine) Species/chemical = 2 (bobwhite, mallard) Treatments/chem. = 8 (2 controls, 6 treated diets) Birds/treatment = 10 Range Finding — Preliminary Tests A literature review of the chemicals used in this study indicated that dietary LC50 values for fenthion and endrin had been previously determined in bobwhite and mallards. The reported LC50 estimates for these two compounds (Heath 3E dl., 1972) were used as the approximate median concentrations in the LC50 trials. ANTU, secobarbital, and strychnine trial concentrations were determined, in part, through preliminary testing using three widely spaced dietary concentrations per chemical with ten birds per dietary concentration. The mortality results of these preliminary tests were used to set the median dietary concentrations and the concentration spacing for the LC50 trials (see Table 2 for preliminary test dietary concentra- tions and mortality). 25 Table 2. Percent mortality for bobwhitesand mallards fed ANTU, seco— barbital, and strychnine during preliminary range finding tests. Dietary Mortality Chemical concentration (%) Bobwhite ANTU 2 ,0001 10 5,000 20 l0,000 30 Secobarbital 2.000 0 5,000 O l0,000 3O Strychnine lOO 0 375 20 l,000 3O Mallard ANTU 2,000 30 5,000 60 l0,000 50 Secobarbital l,000 0 3,000 l0 5,000 20 Strychnine lOO 0 375 30 l,000 80 1 Parts per million. 26 Facilities All birds were housed in the Michigan State University Poultry Research and Teaching Center during the acclimation and testing period. The three rooms used, one for each photo- period, were entirely enclosed and measured 2.95 m X 4.50 m X 2.54 m (W X L X H). Each room contained one Petersime 250—24 battery brooder in which the birds were maintained. The brooders were divided into six levels, each of which was separated into four equally sized sections measuring 34.3 cm X 99.4 cm X 24.1 cm (W X L X H). Only two levels (eight sec- tions) with ten birds per section, allowing a minimum of 134.2 cm2 floor space per bird, were used during each test. The galvanized steel wire mesh of the sides and floors measured 1.3 cm X 1.3 cm and 0.8 cm, respectively, for bobwhite and 1.3 cm X 2.4 cm and 1.7 cm X 1.7 cm, respectively, for mallards. Each section was provided with an individual feeder, measuring 5.7 cm X 61.0 cm X 1.9 cm (W X L X H) for bobwhite and 8.8 cm X 63.5 cm X 6.1 cm (W X L X H) for mallards, and waterer, one quart mason jar for bobwhite and 8.8 cm X 33.0 cm X 5.1 cm (W X L X H) trough for mallards. Mallard feeders and waterers were attached to the outside of each section while bobwhite feeders and waterers were placed within each section. Photoperiod was controlled by Paragon model 41005—D auto- matic timers that were set for 24 hours light:0 hours dark or 14 hours light:10 hours dark. In addition, a voltage regula— tor constructed by the Department of Agricultural Engineering, Michigan State University, was used in one of the 14 hours 27 light rooms to create gradual dimming and brightening sequences of 17 min duration before dark and light periods. Light was provided by a 150 watt incandescent light bulb centered in the ceiling of each room. Light intensities for each room and battery level are given in Table 3 and 4. A thermometer and humidity gauge were located on the wall inside each testing room. One 1,500 watt convection floor heater with variable temperature controls was used to heat each room when temperatures dropped below 21°C (70°F). Animals Bobwhite chicks were reared at Michigan State University from eggs collected from an established breeding population of second generation wild bobwhite. The original stock of birds was obtained from the Illinois State Game Farm, Mt. Vernon, IL. Upon hatching, all bobwhite were wing banded and placed in quarantine until five days of age. All mallard ducklings were purchased from Whistling Wings Duck Farm, Hanover, IL. Prior to each trial, approximately 300 one-day old ducklings were shipped (shortly after hatch- ing) and received at Michigan State University, Department of Animal Science within 36 hours, whereupon they were wing banded and placed in quarantine until five days of age. During the quarantine periods, all birds (bobwhite and mallards) were observed and all deformed, sick, injured, or abnormal birds were removed from the flock. At five days of age all birds in apparent good health were transported to the 28 Table 3. Brooder and room light intensities]. Photoperiod 242 143 1404 Room 69.9 69.9 69.9 Brooder upper level 32.3 32.3 32.2 Brooder lower level 30.l 30.l 30.l 1 Light intensity in luxes. 2 24 hours lightzo hours dark. 3 l4 hours light l0 hours dark. 4 l4 hours lightle hours dark with dimming and brightening between light and dark periods. 29 Table 4. Room light intensities1 for the MD2 photoperiod during the brightening and dimming sequence. Bri htenin se uence Dimmin se uence Time min ntensity ime (min) Intensity 0 0 0 69.9 (maximum) 1 0 l 64.6 2 2.2 2 59.2 3 5.4 3 53.8 4 8.6 4 48.4 5 l0.8 5 43.8 6 l6.l 6 37.7 7 l9.4 7 34.4 8 2l.5 8 30.l 9 26.9 9 26.9 l0 32.3 l0 2l.5 ll 37.7 ll l6.l l2 43.8 l2 l2.9 l3 48.4 l3 l0.8 l4 56.0 l4 5.4 l5 64.6 l5 0 (minimum) l6 66.7 l6 0 l7 69.9 (maximum) l7 0 Light intensity in luxes. 2 l4 hours lightle hours dark with brightening and dimming between dark and light periods. 30 testing facility and placed in their testing batteries for a five day acclimation period preceding each trial. At ten days of age the birds were randomly assigned to their testing sections. Prior to the assignment of birds to sections, any abnormally under-weight or over—weight birds were removed from the flock to increase flock uniformity and prevent bias caused by unequal body weights. All birds used for testing were ten days of age at the onset of testing. At trial terminations, all birds were killed by either cervical dislocation or chloroform asphyxiation. Diets All test diets were prepared by the Department of Animal Science, Michigan State University using a mash as the basal diet (Appendix A and B). Control diets consisted of basal mash plus any carriers used in the preparation of treated diets. Premixes treated with ANTU, endrin, secobarbital, or strychnine were prepared by adding the powdered form of these compounds to 1,500 grams of fine mash which had been filtered through a number 20 sieve. The entire 1,500 grams of mash and chemical were mixed and filtered through a number 20 sieve to prevent chemical clumping and obtain an equal distri— bution of the compound throughout the feed. The unfiltered feed which failed to pass through the sieve in obtaining 1,500 grams of fine mash, and any additional feed required to bring the premix to the desired concentration, were added to the chemical—fine mash mixture and tumbled for ten minutes on a mechanical tumbler. Treated diets used during the test were 31 prepared by dilutions of the premix diet. Specific quantities of premix and untreated feed were combined to reach the desired dietary concentration and tumbled for ten minutes to assure mixing. The fenthion premix was prepared by dissolving liquid Baytex in corn oil and adding this oil—chemical mixture to a specific quantity of mash. The corn oil content of the premix did not exceed one percent of the total feed—oil—chemi— cal mixture. The premix was tumbled for ten minutes on a mechanical tumbler after the oil-chemical mixture had been thoroughly mixed manually into the mash. Fenthion test diets were prepared by the dilution method described for ANTU, endrin, secobarbital, and strychnine. Nominal dietary con- centrations of all chemicals tested are listed in Table 5. Eight treatments per test were used in each trial; two controls plus six concentrations of the test chemical. Chemical concentrations within the six treated diets were geometrically spaced at intervals of 1.3X to 1.6x, depending on the chemical used. Feed and water were provided dd libitum throughout the quarantine, acclimation, and testing periods. Observations of Record Feed consumption, body weights, mortality, signs of intoxi- cation, room and brooder temperatures, and relative humidity were recorded throughout the testing period. Average feed consumption and feed wastage per battery section (dietary concentration) were measured on days 2, 4, 5, 7, and 8 during 32 0px 0cm 00m._ 0N0 0mg mmm 0.0 0.0 0mm.0_ m00.m 00¢.m 0N0.m 000.N 000.N 0.0 0.0 v.0m 0.0N 0.NN m.m— m.m— v.0_ 0.0 0.0 00m mvp mop me mm NM 0.0 0.0 0m~.0_ 000.K 00v.m 0N0.m 000.N 000.N 0.0 0.0 vw_.0 0¢0.m 00v.N 00m._ 0M0 00m 0.0 0.0 0m~.0_ mmo.~ 00v.m 0N0.m 000.N 000.N 0.0 0.0 «.mm m.0~ 0.0— 0.N_ w.0 _.N 0.0 0.0 0.0m 0.Nv 0.0m e.FN m.m~ 0.0_ 0.0 0.0 0mm.0_ mmw.m 00v.m 0N0.m 00m.m 000.N 0.0 —0.0 a r e .;m m. .. ..ll-.....fl.w-fl..w N T .:o____e to; worse udoEHmotL xcmuo_m _ w:_:;thom _mo_ntmcooom :_to:L :o_:o:od zez< era__ms m:_czoxtom Fauwncmaooom :_tucw :o_zo:ol sez< amlezsom .m:_ommo 0m0L m:_t=u com: Asaev meowoacocoocoo zimoo_0 .m m_ame 33 all bobwhite trials. Any significant amount of feed spilled into the litter pans beneath bobwhite sections was separated from feces and weighed. Mallard feed consumption was measured at identical times during the testing periods but feed wastage could not be accurately determined due to the diffi- culty in separating feed from feces and water which had collected in the litter pans. All birds were weighed at days 0, 5, and 8 or on their day of death. All weighings (feed consumption and body weights) within each trial were conducted at similar times of day in order to maintain consistency in weighing intervals. Observations were made three times during the first day of exposure and twice daily thereafter until study termination. All signs of intoxication and the number of birds showing specific signs were recorded for each dietary concentration within a test. Band numbers and body weights of birds which died during the trial were also recorded at these times. Brooder temperature, room temperature, and room relative humidity were measured daily during the first observation period of each day. Statistical Analysis LC50 values with 95% confidence limits and the slope of the concentration response curves were determined for each test run using three different established methods; the probit analysis method described by Litchfield and Wilcoxon (1949), a log—probit analysis using a direct regression of dose on 34 mortality (Montgomery and Peck, 1982), and a log-probit analy— sis using an inverse prediction of the regression of mortality on dose (Gill, 1978). Statistical contrasts of LC50 values between photoperiods for a specific compound were accomplished by confidence limit comparisons. If the LC50 : confidence limits of two photoperiods were separated (nonoverlapping) by > 10% of the smallest confidence limit, the LC50 values of the two photoperiods were considered significantly different at a < 0.05 (Browne, 1979). Initially, a two—way analysis of variance (photoperiod by dietary concentration), which compensates for unequal replica- tion and missing blocks when appropriate, was run on all body weight and feed consumption data on a trial basis. The two— way analysis of variance on body weight was run at day 0, 5, and 8 of every trial, while the feed consumption analysis was run on the five-day treatment and three—day recovery periods. Feed consumption was measured three times during the five—day treatment period and twice during the three-day recovery period. Sampling units for body weights and feed consumption were individual bird weights and mean feed consumption per dietary concentration (expressed as grams/bird/day), respec— tively. Means of body weights and feed consumption for each of the photoperiods were compared by use of Tukey's all pair- wise comparisons test, while means of body weights and feed consumption for each treatment were compared using Dunnett's two-tail t—test (Gill, 1978). The analysis of variance pro- cedures used were run on the Michigan State University Cyber 35 750 Computer using the SPSS (Nie SE d£., 1975) and JENSTAT (Alvey SE d£., 1977) statistical packages. RESULTS Three trials resulted in insufficient mortality to pro- perly determine LC50 values. The bobwhite—ANTU, bobwhite— secobarbital, and mallard—secobarbital trials failed to pro— duce greater than 50% mortality at a dietary concentration exceeding 10,000 ppm, in at least two of the three photo— periods. Although insufficient data were obtained from these trials, LC50 values were calculated with the available infor— mation and are presented along with the other trials. Litchfield and Wilcoxon Analysis No significant differences in LC50 values between the three photoperiods could be established in bobwhite or mallards using the probit analysis method of Litchfield and Wilcoxon (Table 6). The photoperiod effects on LC50 determinations varied considerably among the two species and five chemicals tested. Bobwhite mortality on the 24 photoperiod produced lower LC50 values than either of the two 14—hour light photo— periods when chicks were exposed to fenthion, endrin, and strychnine, but produced higher and intermediate LC50 values compared to the LC50 values for the 14-hour photoperiods when chicks were exposed to ANTU and secobarbital, respectively (Tables 6 through 11). Mallard mortality under the 24 photo- period resulted in higher LC50 values than the 14—hour photo- periods when ducklings were exposed to fenthion and endrin, 36 . mcomuo; View HEM uzmfla :wwzbw: ”3:7: ”.0 c:_:o:_m.:n «Em ESE—EC ”315 «Imp muzo£ 07:3: mid: ivmm lemm 0 ode Acme Iowa . Amqe.mlom_.nc oo~.~ Aenv.~-em~.: imam.o~lmmm.m c mmo.e leve.~omumo~.mac Aeem.mv-mmm.o~0 ooo.mm Avom.mm immfi.a . “1.8m-v.mmc o.mm Ad.mm-v.-c Ae.onlm.q~c e.n~ Am.m~-a.m . Am.mein.mvc o.oo .~.mm-m.mv. no.vm-m.eac o.- Am.m~-m.m3. Aomm.o -ome.m. oom.v Amom.v -mem.~c Loom.mm-mom.m. ooo.mn Aemo.mn-mom.m. 1.4.o ammo omQu A.a.o ammo smog Le\2 muse; «H muse: : ohm CON. _~ coo.mm :cm.- b.0m mg: 0.00 SAN oov.m oo~.m coflumils. .2... .Amvmz .Nec -qmm 0 ohm A_m_.m lemm._c cmo._ .emm.e~1mev.o. oce.c~ iamm.wmiome.nc oom.e~ 8.2. ANA; Iw.v~v 06m Im.m 0 $.04 :.mo~xh.~8 mm :.mm 10.0: .m.oN norm.m -mmn.~ . oco.v Aceo.oe-moe.~H. ooo.m~ A.A.U awe Ombq muse: em .4107 H >:n:; 1.51:9: C :«00 u >....::; ~m0._:_..:_.c m. J00 n >325; _m.ut=.\:_..c_\ .wwm H CECE— 1.61.5.3» .wma H ‘32:; ~mc..=...:_;w 25.30.53,... I): . l ltll . i.l . I. . . . ._ CH C_ of: To: a: A: o _ 23:3: wwtimmm flown» MW. CH A: Una; :wz Cfi C_ Outta—Cc Hougljnmowmwm m A: CA p.22 To: c. A: on Lmswckml :CEE v OH cm one; in: OM OM c He Gucci: :0: :3. m . _ ,_ OH R: 6521“: OH a: mowzzpcizl. N392< :cwu Am>m3 modumim Imuucoocoo 604 mafiflzflm \hppd. .02 .5552: pct 03:55:.— 0C @922: Emlyn—E. DESK of >2 couoiigoo mph—badge Cam wcuflzzooo a; o:1_:o>uom 0.5 abomnumoooom :iutco.::____:.:.:EZ< paw mo:_o> emu.— . c @730. 37 Table 7. Percent mortality and LC50 values for bobwhites fed dietary ANTU]. Percent mortalit Dose 24 hrs. )4 hrs. 14 hrs. (ppm) n light light light w/d2 0.0A l0 0 0 o 0.05 lo 0 o 0 2.000 l0 l0 0 30 2,800 l0 lo 20 0 3,920 10 l0 20 40 5,488 lo 20 20 lo 7,683 10 l0 so 50 10,755 l0 40 50 20 LC503 28.000 3,200 l5,000 95:. C.L. (ll,755-55,540) (5,563-l2,087) (5,865-38.365) Slope 2,205 244 3,072 0504 l3,836 5,745 5,470 95% C.L. ( 3,020-53,24l) (4,853- 9,375) (Z,805-lO,7l5) Slope 0.004 0.0l0 0.02l 1.0505 4l .495 7 ,534 2l ,749 95:; c.L.5 --- (2,05l-m) Slope 47l l2l 333 l 2 Chemical purity = 95% Fourteen hours light:10 hours dark with dimming and brightening of lights between light and dark periods. LC50 values, 95% C.L., and Slopes were detennined by the method described by Litchfield and Wilcoxon (l948). LC50 values, 95% C.L., and Slopes were detennined by a log-probit analysis using a direct regression of dose on mortality (Montgomery and Peck, l982). UI LC50 values, 95% C.L., and Slopes were determined by a log-probit analysis using an inverse prediction of mortality on dose (Gill, l978). Lack of confidence limits resulted from the inability to reject the hypothe— sis of slope equaling zero. 38 l Table 8 . Percent mortality and LC50 values for bobwhites fed dietary fenthion . Percent mortality Dose 4 hrs. )4 hrs. l4 hrs.“ (ppm) h light light light w/de 0.0A l0 0 0 0 0.08 lo 0 O O 10.9 l0 0 0 0 lS.3 lo 20 10 l0 2l.4 lo 60 30 60 30.0 10 80 80 90 42.0 l0 l00 l00 lOO 58.8 lo 100 lOO 9O LC503 20.5 23.7 2l.0 95% C.L. (l6.8-25.l) (19.5-28.8) (l7.9-24.6) Slope 0.25 0.22 0.l2 L0504 2l .2a 22.2b 24.0C 95% C.L. (2l.l-2l.3) (2l.9-22.5) (23.0-25.l) Slope 3.04 2.93 2.35 L0505 2l.l 22.2 23.5 95% C.L. (l2.9-34.5) (l5.2-32.9) ( 3.4-l22.7) Slope 0.3l 0.3l 0.48 l 2 Chemical purity = 94% Fourteen hours light l0 hours dark with dimming and brightening of lights between light and dark periods. LC50 values, 95% C.L., and Slopes were detenhined by the method described by Litchfield and Wilcoxon (l948). LC50 values, 95% C.L., and Slopes were determined by a log-probit analysis using a direct regression of dose on mortality (Montgomery and Peck, l982). LC50 values, 95% C.L., and Slopes were determined by a log-probit analysis u5lng an inverse prediction of mortality on dose (Gill, l978). 39 Table 9 . Percent mortality and L050 values For bobwhites fed dietary endrinl. Percent mortalit Dose :3 hrs. 4 hrs. l4 hrs. (ppm) n light light light w/dz 0A l0 0 0 0 08 lo 0 0 0 7.l lo 30 30 0 9.2 lo 40 4O 30 l2.0 l0 l00 40 40 l5.6 lo 70 80 50 20.3 lo 80 l00 90 26.4 l0 lOO l00 90 LC503 l0.8 l0.8 13.5 95% C.L. (8.2-l4.2) (8.7-l3.3) . (ll.3-l6.4) Slope 0.23 0.l6 0.l7 1.0504 11.131D 10.4a 14.3b 95% C.L. (7.6-16.4) (8.9-l2.2) (l4.0-l5.6) Slope 7.4l 5.50 6.42 LC505 7.7 9.9 15.1 95% C.L.6 ( 6.l-44.9) Slope 0.34 0.l5 0.20 1 Chemical purity = 99% 2 Fourteen hours ligntle nours dark with dimming and brightening of lights between light and dark periods. 3 L050 values, 95% C.L., and Slopes were detennined by the method described by Litchfield and Wilcoxon (l948). 4 LC50 values, 95% C.L., and Slopes were determined by a log-probit analysis using a direct regression of dose on mortality (Montgomery and Peck, l982). 5 L050 values, 9 % C.L., and Slopes were determined by a log-probit analysis 6 using an inverse prediction of mortality on dose (Gill, l978). Lack of confidence limits resulted from the inability to rejeCt the hypothe— sis of slope equaling zero. 40 Table l0. Percent mortality and L050 values for bobwhites fed dietary seco— barbital). Percent mortalit Dose 24 hrs. l4 hrs. l4 hrs.2 (ppm) n light light light w/d 0.0A 10 0 0 103 0.08 10 0 103 0 2,000 10 0 10 0 2,800 10 0 0 0 3.920 10 20 20 0 5,488 10 0 100 10 7.683 10 0 50 10 lO,756 10 40 30 30 L050“1 16,300 12,300 23,000 957. C.L. (7,739-34,329) (7,153-22,904) (l0.892—48.567) Slope 1,279 529 550 L0505 7,211 4,943 11,614 951; C.L. (2,018-25,763) (4,140- 5,902) ( 5,420-24,831) Slope 0.0l8 0.023 0.005 LC505 41,933 6,366 l2,892 957; c.1..7 --- Slope 1,560 110 495 Chemical purity = 99% 2 Fourteen hours lightle hours dark with dimming and brightening of lights between light and dark periods. Percent mortality values used to compute LC50 values were adjusted to com- pensate for control mortality. L050 values, 95% C.L., and Slopes were detennined by the method described by Litchfield and wilcoxon (l948). LC50 values, 95% C.L., and Slopes were determined by a log-probit analysis using a direct regression of dose on mortality (Montgomery and Peck, l982). U1 LC50 values, 95% C.L., and Slopes were determined by a log-probit analysis using an inverse prediction of mortality on dose (Gill, l978). Lack of confidence limits resulted From the inability to reject the hypothe- sis of slope equaling zero. 41 Tablell . Percent mortality and L050 values for bobwhites fed dietary strychninel. Percent mortalit Dose 24 hrs. l4 hrs. l4 hrs. (ppm) n light light light w/d2 0.0A 10 0 0 0 0.08 10 0 o 0 586 10 0 0 10 938 10 10 20 10 1,500 10 20 40 50- 2.400 10 100 70 60 3.840 10 100 100 100 6,144 10 100 100 100 LC503 1,650 1,700 1,700 9 % C.L. (l,284-2,l2l) (1,197-2,414) (l,l8l-2,448) Slope 2l.O 37.l 40.3 L0504 1,309a 1,549b 1,371a 95% C.L. (l,l83-l,449) (1,521-1,574) (1,259-1,493) Slope 0.051 0.028 0.025 LC505 3.122 1,552 -1,337 95% C.L.5 ----- ( 674-3,573) ( l4-23,659) Slope 18.2 32.5 33.3 1 Chemical purity = 96%. 2 UI Fourteen nours ligntle hours dark with dimming and brightening 3f lights between light and dark periods. L050 values, 9 % C.L., and Slopes were determined by the method described by Litchfield and Nilcoxon (l948). LC50 values, 95% C.L., and Slopes were detenmined by a log-probit analy- sis using a direct regression of dose on mortality (Montgomery and Peck, l982). LC50 values. 95% C.L., and Slopes were determined by a log-probit analy- sis using an inverse prediction of mortality on dose (Gill, l978). Lack of confidence limits resulted from the inability to reject the hypothesis of slope equaling zero. 42 however, mortality on the 24 photoperiod resulted in inter— mediate LC50 values to the 14—hour photoperiods when exposed to strychnine, ANTU, and secobarbital (Tables 6 and 12 through 16). Similarly, mortality between the two 14—hour light photo- periods varied with species and chemicals (Table 6). Bobwhite LC50 values in the 14 photoperiod were lower than LC50 values of the 14D photoperiod when chicks were exposed to ANTU, endrin, or secobarbital (Tables 6, 7, 9, and 10). The LC50 values for bobwhite fed strychnine were equal for the two 14-hour photoperiods, while the 14 photoperiod produced a slightly higher LC50 value than the 14D photoperiod when the chicks were fed fenthion (Tables 6, 8, and 11). Mallard mortality resulted in higher LC50 values under the 14 photo- period compared to the 14D photoperiod when ducklings were exposed to fenthion and secobarbital, but produced lower LC50 values under the 14D photoperiod when ducklings were exposed to ANTU and strychnine (Tables 6, 12, 13, 15, and 16). LC50 values for mallards between the two 14-hour photoperiods were equal for endrin (Tables 6 and 14). In summary, the 24 photoperiod produced the highest LC50 values in three of the 10 trials while the 14 and 14D photo— periods produced the highest LD50 values in three and four of the trials, respectively. The lowest LC50 values between the three photoperiods showed a similar pattern of being equally distributed. 43 Table 12. Percent mortality and LC50 values for mallards fed dietary ANTU]. Percent mortalit Dose 24 hrs. l4 hrs. l4 hrs. (ppm) h light light light w/d2 0.0A 10 0 0 0 0.08 10 0 0 0 2,000 10 273 30 10 2,800 10 70 50 40 3,920 10 564 40 40 5,488 10 70 60 70 7,683 10 50 90 60 10,756 10 80 90 80 1.0505 4,000 3,400 4,900 95% C.L. (l,786-8,960) (2,378-4,862) (3,490—6,880) Slope l06 l04 l24 1.0505 3,855 3,589 4,742 953201. (2,8l8-5,284) (3,105-4,159) (4,677-4,797) Slope 0.0106 0.0086 0.0096 LC507 2,931 3,436 4,753 951401.8 --- (l,l6l-7,586) (1 ,879-l2,388) Slope 268 l38 136 l 2 Chemical purity = 95% Fourteen hours lightle hours dark with dimming and brightening of lignts between light and dark periods. Mortality percentage based on ll animals (n = ll). Mortality percentage based on nine animals (n = 9). 01 pm L050 values, 95% C.L., and Slopes were determined by a log-probit analysis described by Litchfield and Wilcoxon (l948). LC50 values, 9 % C.L., and Slopes were determined by a log-probit analysis using a direct regression of dose on mortality (Montgomery and Peck, l982). L050 values, 95% C.L., and Slopes were determined by a log-probit analysis using an inverse prediction of mortality on dose (Gill, l978). Lack of confidence limits resulted from the inability to reject the hypothe- sis of slope equaling zero. 44 Table l3. Percent mortality and LC50 values for mallards fed dietary fenthion . Percent mortaliv Dose 24 hrs. l4 hrs. l4 hrs. (ppm) h light light light we? 0.0A l0 0 O 0 0.08 10 0 0 0 37 10 lo 30 lo 52 lo 40 4O 60 73 lo 30 40 60 l02 lO 60 70 70 l43 10 80 80 l00 200 l0 l00 9O l00 LC503 82 69 50 95% C.L. (6l.7-l09.l) (48.5- 98.2) (45.l- 79.3) Slope l.57 2.09 l.l2 LC504 74.5 68.5 61.0 95% C.L. (66.8- 82.8) (63.4- 74.l) (5l.3- 72.4) Slope 0.46 0.38 0.78 LC505 71.4 67.6 57.9 95% C.L.6 (l4.6-23l.7) (40.6-l05.4) ( O -3l0.0) Slope l.7 2.6 l.2 l 2 pl Chemical Purity = 9 m. Fourteen hours light:l0 hours dark with dimming and brightening of lights between light and dark periods. LC50 values, 95% C.L., and Slopes were detennined by the method described by Litchfield and Wilcoxon (l948). LC50 values, 95% C.L., and Slopes were detennined by a log-probit analy- sis using a direct regression of dose on mortality (Montgomery and Deck, l982). LC50 values, 95% C.L., and Slopes were determined by a log-probit analy— sis using an inverse prediction of mortality on dose (Gill, l978). Lack of confidence limits resulted from the inability to reject the hypothesis of slope equaling zero. 45 Tablel4 . Percent mortality and LC50 values for mallards fed dietary endrin]. Percent mortalit Dose 24 hrs. l4 hrs. l4 hrs. (ppm) h light light light w/dZ 0.0A 10 0 0 0 0.08 10 0 0 0 10.4 10 0 0 10 13.5 10 10 20 o 17.5 10 30 10 10 22.8 10 10 40 30 29.6 10 40 50 50 38.4 10 60 60 7 L0503 34.0 29.6 29.6 95% C.L. (24.6—46.9) (22.4-39.1) (22.4-39.1) Slope 0.68 0.60 0.59 Lcso4 27.9 26.5 25.9 95% C.L. (20.8-37.5) (20.9-33.4) (l8.9-35.5) Slope 2.72 3.09 3.66 L0505 31.9 29.4 31.5 95% C.L.6 (l2.0-l.786) (ll.2-0.48) ---— Slope 0.4 0.3 0.4 l 2 Chemical purity = 99%. Fourteen hours light:l0 hours dark with dimming and brightening of lights between light and dark periods. LC50 values, 95% C.L., and Slopes were determined by the method described by Litchfield and Wilcoxon (l948). LC50 values, 95% C.L., and Slopes were determined by a log-probit analy- sis using a direct regression of dose on mortality (Montgomery and Peck, l982). LC50 values, 95% C.L., and Slopes were detennined by a log-probit analy- sis using an inverse prediction of mortality on dose (Gill, l978). UI Lack of confidence limits resulted from the inability to reject the hypothesis of slope equaling zero. 46 Table l5. Percent mortality and LC50 values for mallards fed dietary seco- barbital). Percent mortalit Dose 24 hrs. l4 hrs. )4 hrs. (ppm) h light light light w/d2 0.0A l0 0 0 O 0.08 l0 0 0 0 2,000 l0 0 0 l0 2,800 l0 0 0 l0 3,920 l0 20 20 l0 5,488 l0 20 0 l0 7,683 lo 50 lo 50 l0,756 l0 40 20 70 LC503 10,600 98,000 7,683 95% C.L. (6,463—l7,384) (l9,l03-502,740) (5,583-l0,573) Slope 3S4 l04,309 l53 L0504 7,656 8,250 7,656 95% C.L. (S,023-ll,64l) ( 2,698- 25,35l) (4, 32-ll,885) Slope 0.0098 0.0066 0.0080 LCSOS 8,959 20,502 9,354 95% C.L.6 --- --- Slope lZl 53l 209 l 2 Chemical purity = 99% Fourteen hours lignt2l0 nours darK with dimming and brightening of between light and dark periods. lights U] LC50 values, 9 % C.L., and Slopes Litchfield and Wilcoxon (l948). LC50 values, 9 % C.L., and Slopes using a direct regression of dose LC50 values, 95% C.L., and Slopes were determined by the method described by were determined by a log-probit analysis on mortality (Montgomery and Peck, l982). were detennined by a log-probit analysis using an inverse prediction of mortality on dose (Gill, l978). Lack of confidence limits resulted from the inability to reject the hypothe— sis of slope equaling zero. 47 Table l6. Percent mortality and LC50 values for mallards fed dietary stry- chninel. Percent mortalit Dose 24 hrs. l4 hrs. l4 hrs. (ppm) h light light light w/d2 0.0A l0 0 0 0 0.08 lo 0 0 0 323 lo 50 40 30 420 lo 50 80 60 546 10 90 90 80 710 lo 80 9O 70 923 lo 90 80 90 l.200 l0 lOO l00 l00 LC503 370 270 410 95% C.L. (284-482) (l69-430) (320-524) Slope 5.97 8.43 7.l3 LC504 432 424 460 95% C.L. ‘ (3l9-585) (285-628) (367—577) Slope 0.054 0.067 0.070 LC505 ~ 379 343 421 951:. 01.5 ( l8-927) -—- ( 69-955) Slope , l3.6 l5.9 ll.6 l 2 03 Chemical purity = 9 % Fourteen hours lightle hours dark with dimming and brightening of lights between light and dark periods. LC50 values, 95% C.L., and Slopes were determined by the method described by Litchfield and Wilcoxon (l948). LC50 values, 95% C.L., and Slopes were determined by a log-probit analysis using a direct regression of dose on mortality (Montgomery and Peck, l982). LC50 values, 95% C.L., and Slopes were determined by a log-probit analysis using an inverse prediction of mortality 0n dose (Gill, l978). Lack of confidence limits resulted from the inability to reject the hypothe- sis of slope equaling zero. 48 Direct Regression of Dose on Mortality Predicting LC50 values using the log—probit analysis with a direct regression of dose on mortality produced similar LC50 values to the Litchfield and Wilcoxon method in those trials in which sufficient mortality was obtained (Tables 7 through 17). The two analytical methods also produced similar fre— quencies in the relative toxicity (magnitude of LC50 values) of the chemicals between the three photoperiods. Although the LC50 values between the two methods were similar, the confi- dence limits for the LC50 values were more narrow for the direct regression of dose on mortality procedure (Tables 7 through 16), resulting in the determination of significant photoperiodic effects in the bobwhite—fenthion, bobwhite— endrin, and bobwhite-strychnine trials. In the bobwhite- fenthion trial all three LC50 values were significantly differ— ent (P < 0.05) between the photoperiods. The bobwhite—endrin trial resulted in the 14D photoperiod producing a significantly (P < 0.05) higher LC50 value than the 14 photoperiod while the bobwhite—strychnine trial resulted in the 14 photoperiod producing a significantly (P < 0.05) lower LC50 value than the 24 and 14D photoperiods. Inverse Prediction of the Regression of Mortality on Dose As was seen with the probit analysis of Litchfield and Wilcoxon, no significant differences in LC50 values between the three photoperiods could be established (Table 18 and 7 through l6). The 24 photoperiod resulted in the lowest toxi— city (highest LC50 values) in 50% of the trials. The 14 and 49 .3530; thc Ccm uzmMA .503qu mbzm: no ?;:vb:mwun «Er. 7.l—EEG :23 0:3. 925; Gigi:— w..:3: :oozzca .1lllllllll‘lllllllli iilil )lll 1) 1 l. I .I ll. ll . lull illlillil.1lllillll'll4llullillllllnliiv. 1~1 188m leom a see Ammo imam . ewe Amam loan a mac 3. o. c.a__ez Amae.~-omm.bva~em.~ Avem.H--m._v.oem._ .mve.a1mm~.svaocm.~ o. o. 6._;3;:m odeoxim Ammm._~-mmo.vv wmo.e Aamm.m~1mmo.mc com.m .aeo.bfi-mmo.ma wmo.e 0. ca c_:__mz A_mm.em-oNv.mc vaw.aa Amom.m Ice..vv mem.e Amo>.mmlmao.mc _-.e ca ca 63_;:;::. an _ 17:59.33 Am.mm1o.m: m.mN Av.nmlm.oN. m.m.~ Am.hm1m.omv mKN a: A: _ct_c:zz Aw.m_io.v:nm.va AN.N~1m.m V144: AviwqioK Iiém CH CH 9,1533— 5:195. Simelmémv 0.1» 3.vn1v.mcv m.mo Am.mm1m.©o. m.vn OH OH mic—:02 :.mmio.mmvoo.vm Am.NNIm.Hmv£N.NN Am.HN1~.~Nvmm._m CH CH olfi;3._::1 :s 2. :3; 1885.5 leeo.v. mvn_e AmmH.vlmo~.m. mwm.m .emm.m -m~m.~. mmm.m ca o. eteaaez .m_~.calmom.mv obv.m Aoem.m-mmm.qv mee.o A_e~.mw1o~o.m. cmw.mb on Ca assesses Maze 74.0 ammo. omUA A570 wmmv:mU4 A470 mmmv emf: :0: 70.5%: mwmumim RUE») musoz Va 9.25; a; muse: vm imhbcoocoo oa< alqulahma poioéscza lll:l1ll .. \mpch .02 .>um~m?_2= cc wmcc w: :Cfimmwhmwx 30.2p 0 95m: mmm>~ozm ”320.17%: >2 twbmasodmo £723.35 nucm mwuazznon cw 0:253»qu 6:0 1.3 qaumnouum fight—523:9. .952 no» mm:~m> amou— . : 3:0... 50 .ouoN no::~m5_cm maofim mo m_mo:uoa>: 0.: €5.78...“ Ou 43:13.: 0.: ECC cmafizmou mbtz: oucwmquc; .._3 VA..::_ N .mpoflxw; x1220 ES. 33: :mwBAwn wait: “.3 3:150:0qu Ccc 355:2: :23 xump muso; 371.0: mnzoz :vmj.:_:..__ Ammo Imov Hmv A A 2% Abmmlc: Ohm CA A: “21:12 A©m0.mmlv: hmm.~ 3508458: mom; A A mm...” A: :— 04:25.3: 3:1...ch An. A 8-58:3 «.26 A . 25.5 A a «had 2 A: 321.2 A a 392 A a 82.8 A . So...» on A: 31.23: Almwmmiluvw A . 9: 362.44: 7% Aoanloa: 8.: 3 3 35:82 8.3. 14.8 73 A a To A a 8.8 S 2 322.2: .:.m Alma 4.. 8.2:) 8 95 3.37931 8.5 Aeémmloé: Y: A: 3 E252 £82-72 8.3 8.2 Ah: ~.- 3.3., -m.~: «.3 3 3 3:315. :A.____:;.._ 83.21.83,: STAN $3.731: mm: A a Amo.~ 0A A: 23...: A a 81.3 A8 130.3 EQK A a met; 2 A: 31:33: N l.lli :._.Z< Add $8 83 Add $3 30; Add $3 33 .53 no}: muso; «A muse; v~ mgsc; em Imuacwozou Am>mg moTJCLm 1 .. . 1 1. Unfluhafilxiszl... . :1 lllllll \mU.:.n .Cz wv< ._,.m..w_=_w;..o .9806 :o >uw~puuoa mo mqm>~mcm :CAmmoumwu c ya :oquompou; omega; :m vim: mdm>~m:m DCAEATan m >2 coumazofipo mound—ME can mouflszncn :_ o:«:;o>.3m "Em Hmufinumscuom .:Ct:o:_:_;b:£.:._.z< EC m2£p> amp: .m. 27...... 51 14D photoperiods produced the lowest toxicity values in 10 and 40% of the trials, respectively. The confidence limits on the LC50 values calculated by the inverse prediction of the regression of mortality on dose were greatly widened in com— parison to the other two analytical methods used. (LC50 max In comparing the variation LC50 min ) of LC50 values for a particular chemical and species over the three photoperiods tested, a range of 1.00:1.03 to 1.00:1.52, 1.00:1.08 to 1.00: 1.42, and 1.00:1.09 to 1.00:2.34 was obtained for the Litch— field and Wilcoxon, direct regression, and inverse prediction methods, respectively. If the variation in LC50 values for a particular photoperiod, chemical, and species is contrasted .by analytical method, a range of 1.00:1.03 to 1.00:2.39 is obtained. The average variations in LC50 values between photo- periods for a particular chemical and species were 27.6, 18.6, and 64.3% for the Litchfield and Wilcoxon, direct regression, and inverse prediction methods, respectively, while the average variation for individual LC50 determinations between analytical methods was 27.3%. Mortality patterns of the three photoperiods for any particular chemical were similar regardless of the chemical's mechanism or speed of action. Secobarbital and strychnine caused high mortality during the first 24 hours of exposure followed by reduced rates of death in the remaining treatment period in both species and throughout all photoperiods (Tables 19 and 20). Conversely, the fenthion and endrin treatments resulted in minimal or no death during the first 24 hours of 52 .vm_c HE: mvctbao c2252 : ._a_tb to s5: 5 A ena.c_ 1 — 1 1 1 1 1 1 1 _ 1 a mxo.x _ N 1 1 1 1 1 1 1 1 _ 1 1 1 A5: 1 1 A 1 _ 1 1 1 1 1 1 1 0%; 1 1 1 1 1 1 1 1 1 1 1 1 1 25.x 1 1 1 1 1 1 1 1 1 1 1 1 1 A526. -111llllllllllli.llltl.:l I: 5539 1 1 F _ 1 1 1 1 1 1 N N omx.:_ m 1 1 1 1 1 1 1 1 mwo.N 1 1 n N 1 1 1 1 1 1 1 1 31:4. 1 1 A _ 1 1 1 1 1 1 1 am oNa.m 1 1 1 1 1 1 1 1 1 1 1 1 oom.~ 1 1 1 _ 1 1 1 1 1 1 1 1 9:9.N oa_:zcam co_mmm:soza . . . n w m Q m N mfi Aszav c_1lllllllll11.11 11.111..1.llkmllll :o_bo;b:uu:oo AcsbuA: m. 1m 1 p ._au_oceaoomm ;u_3 a:_umma snub a:_c:v mocc__es ecu wu_:3aon co mzcobba: >b__cbcoz .m_ a_:ek 53 .coAv AszA mccA; a: a ;_:z : .Aa_;A a: As: 1 1 1 1 A m N e 1 1 1 1 1 1 N w 1 1 1 A 1 e A v ocN.A 1 1 A N A N m 1 1 1 1 1 N A m 1 1 1 A A 1 A o mNa 1 1 1 A N N N 1 1 1 1 N A 1 m 1 1 1 1 A A N v oAN 1 A A m A A A 1 1 1 A N A 1 m 1 1 1 1 m N N N own 1 1 A A 1 N N 1 1 1 1 N A 1 m 1 1 1 A A N 1 A ONe 1 1 1 A A A 1 1 1 1 1 A 1 1 m 1 1 A 1 A 1 N A an 5.33: 1 1 1 1 1 A m 1 1 1 1 1 N 1 w 1 1 1 1 1 A A m qvA.@ 1 A A 1 A N m 1 1 1 1 1 N e v 1 1 1 1 1 1 N w 3¢2.m 1 1 1 A 1 A e 1 1 1 1 A A N m 1 1 1 A m A A v cav.N 1 1 A A 1 N A. 1 1 1 1 1 1 A m 1 1 1 1 1 A 1 A ocm.A 1 1 1 1 A 1 1 1 1 1 1 1 A A 1 1 1 1 1 1 A 1 1 a; 1 - 1 1 2A 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 £3 vA_;23:n 11)- 1lA 11.81- 1.4.1.18 .111m111N1: A-.- m1: .A. 1 .3 15.1.18 5 111N. 1A a N 118 mite 1 m N CA A53 .1 1. 1 ..1va 111- 11 11 111.1. 1 1,1 .1 wA .1111: 1 «N 1.1 .1 . :oAAaAA:28:05 wcAcz;cbe;; 1 1 1 11111. 1- xteAuA: .ocAczoxcbm :sA: acAsnyA :nun ::Ac:c mUAcAAsE use obAzzaoa Ao m;;oAbc; NAAAuAcaz .ON 8.:eA 54 treatment but, mortality sharply increased after three or four days of exposure (Tables 21 and 22). Also, as the dietary concentration of fenthion increased, the time required for the onset of mortality decreased. Thus, the higher dietary con- centrations produced mortality during the second day of expo- sure while the lower dietary concentrations did not produce mortality until the third or fourth days of exposure. Morta- lity patterns due to ANTU ingestion were similar between photoperiods, but bobwhite and mallard reactions were somewhat different. The majority of bobwhite deaths occurred during day two or day four through six, while mallard mortality began during the first 24 hours of exposure and continued through day six at a relatively constant rate (Table 23). In general, bobwhites were more sensitive than mallards to the more toxic chemicals fenthion and endrin, but showed less sensitivity than mallards to the less toxic chemicals ANTU, secobarbital, and strychnine (Table 6). Clinical observations of the symptoms of toxicosis for each trial are reported in Table 24. There were no significant differences in slopes between photoperiod for any of the trials. The slopes of the dose— mortality curves were relatively uniform between photoperiod for a particular trial with the exception of the three trials in which insufficient mortality was obtained for proper analysis. The maximum difference between the dose-response slopes for the three photoperiods of a partucilar trial was 2.17-fold. 55 I 1 '\ ,— m N l l I I N m N N F-Nr—V I I I I I I Nle— F—Nmme I I I I I I I I I I NAA xeu Ae msAssoAsAm use .AeAAssesoomm .sAAusm..5AsAswA.:Az< eo..mmm.o;a AceuwAu umA muAeAAeE use mmAAszaos As qucesvoss As Axeu\ucAs\msecaA soAAsEsmsou uweA sew: .mN eAseA 61 .Amo.o A av AswchAAu AAAseoAAAsaAm Ass use AsAcomssm asem ms. sAAz 30. e s. mseos 0:. As< .wqusms Aceu use usmAA swwzums mssmAA As asAswAsaAAs use msAssAu 2AA: Aces muses 8.11.1 A.A..Am A. 8.. N AANA AA AN. H AN.A.A A. AA. 1+1 amAm 8. mA. H AZN 8. AAA. N SAN 8. SAN aAeA 0. SAN ANNA e. 3.7.. ems... e. is N 8Nem . N. .AN 1N ANAN e. 8N N 8.12 m. seememsm e. 85“ as? e. Sew ANAA 8. .N... H as... NA mN... H was NA ANA H em... o. AN... H 8.8 8. AN... H 8.8 AA AN... H was 8. mm... H A? 8. Am... A 89A NA NA... H eeA N. Am... A 89A A. AA... H A... N. A...“ AAA NA AN...“ 99A e. ANeHaeAe 8. AN...“ me... 2 .1.1A.m.11.1..A1 1 - 1... .NaHA e .53., e 11 .11Au\z mcsos1AA11 11 1 1mcswa,%Am1 111 1 1 msso; AN quAwsoAssa 1111111111llll 11 .Aqust xsw>ooec xeu ems .sAz< AceAmAu um» museAAes use mquszsos As quce N atbsaAA mcsos seobssss o._.AA..Ao.A1_Am AeAAsse20osn sAsu2N soAsAsmA :Az< as. e wsAssoANAm AeAAscesoosm :Acu:A soAsAsms :Az< ws_;:ss= us:os§oo moAuosw sAV msAssochm use .AeAAssesoomm .sAsusmasoAsusmA assess As Axeu\ucAs\mEesaA soAAsEsmsoo uweA see: . 0N oAseA 62 14 photoperiods consumed significantly greater (P < 0.05) quantities of feed than ducklings on the 14D photoperiod (Table 26). Generally, there was no trend toward increased feed con- sumption in the 24 photoperiod during the recovery period. Only three trials resulted in continuous lighting having the highest feed consumption value of the three photoperiods during recovery (Table 26). Significant differences (P < 0.05) in initial mean body weights between photoperiods occurred in three of the ten trials (Table 27). For the trial in which ANTU was admini- stered to bobwhite, the chicks on the 24 photoperiod were significantly heavier at the initiation of the trial than chicks on the 14 or 14D photoperiods (Table 27). Also during the ANTU-bobwhite trial, chicks on the 14 photoperiod had significantly greater initial mean body weights than chicks on the 14D photoperiod. In the mallard trials, ducklings exposed to ANTU and the 14 photoperiod were significantly heavier at the study onset than ducklings exposed to ANTU and the 24 or 14D photoperiods (Table 27). Ducklings fed the fenthion treated diets and maintained on continuous lighting had significantly lower initial body weights than birds exposed to fenthion on the 14 photoperiod (Table 27). The birds on the 24 photoperiod had the highest initial mean body weights of the three photoperiods in 50% of the trials run, while the 14 and 14D photoperiods produced the heaviest weights in 30 and 10% of the trials, respectively (Table 27). 63 .Amo.o A av AswsmAAAu AAAseQAAAsaAm Ass use AsAcumssm usem msb 2AA: zos e sA mseus oz. As< .mqusws Aceu use usmAA swwzumn mAsmAA as msAswAsaAAs use msA=_Au 2AA: Aceu muss; qusesoAOsa N. H 888 A... N.. H 8A.... 8 N.. H 8N8 .AN 8.. H 8A8 A... 8.. N 8N8 - A... 8.. m 8.8 A... A... H 8.... A... e. m 8A.: A... A... N 88: A... 8.. H 88Ne . A... .8... H 8A.. A... .AN 8: A... A.N1... 8.... A... .NN EN. 2. .NN 8ON. 8 3.388.: 8 AN...“ 8A.: A... AN...“ 88.: A... AN...“ 88.... A... 8N...“ 8...... A... 8N...“ 8...... A... NN... H 8A.... A... NN... H 8A.... A... NN... .1 88.8. A... em...“ 88.: A... cm...“ 8A.: A... an...“ 88.: A... AN...“ 8.....N A... AN...H8..AN A... AN...“ MANN oe 1111.1..m1fl1.x. 8 8.. 1... A 8 8.. N A. a 11 .331 8.518811... 111111 18111.5; A. 8.18.. AN N oAAAscAA mess: swo.ssoA wsAssoAsAm AeAAssesoosw :AcusA scAsAsms :Az< .81.. 8 A . 81: wsAssuxs.m AeAAssesobsm sAsu:A soAsAses qu< ussossso meruzw .wsAssoAAAm use .AeAAssesooem .sALusm.scAsAsmA .qu< aseAmAu um» museAAeE use mwAAszsos Ao quseQOAoss As Amsesov musuAmz Auos sees AeAAAsA . AN sAse. 64 On day five of the LC50 studies, photoperiod caused a significant (P < 0.05) alteration in body weight in seven of the trials (Table 28). In all seven trials, the 24 photoperiod produced significantly greater mean body weights than one or both of the l4—hour lighting regimes. Bobwhite on the 24 photoperiod were significantly heavier at the end of the five- day treatment period than birds under the 14 and 14D photo- periods when exposed to ANTU, fenthion, endrin, or secobarbi— tal (Table 28). In addition, bobwhite exposed to ANTU and the 14 photoperiod were significantly heavier than birds exposed to ANTU on the 14D photoperiod (Table 28). Mallards maintained on the 14D photoperiod had significantly lighter five-day body weights than ducklings exposed to the 24 and 14 photoperiod regimes during the ANTU and strychnine trials (Table 28). During the mallard—endrin trial, ducklings under the 14 photo— period had significantly lower body weights than ducklings under the 24 and 14D photoperiods (Table 28). Mean body weights at the study termination (day eight) of birds exposed to the 24 photoperiod were significantly (P < 0.05) greater than at least one of the 14 hour photoperiods in five of the ten trials (Table 29). Bobwhite fed ANTU, endrin, and secobarbital treated diets produced greater body weights at day eight under the 24 photoperiod than those under either of the l4—hour lighting regimes (Table 29). The 14D photo- period also produced significantly lower body weights than the 14 photoperiod at day eight of the fenthion and secobarbital trial. Mallards fed ANTU showed significantly higher day 65 .36... A .3 20.8.3... 3.....U.C....a..m .o: w... .5523 9:3 as. 5.... 3.... m s. 28... oz. b.< N .376; .....u usm Em... swwzums 3...... .0 3:25:37... usm ms_.___....u 5.... .28.. 9.3.... 95.3... 3...... 58...... . m.m w 58— mm m6 H mm: mm 5m .+ um: um osEsoxbm m... H 8...: a. .e H 8.... m. e... H 8...; B .8...:8..88 q... H 88. 8 m... H .3. B m... H 8...: m. 5.3.. a... H 8... N. m. H .N: a. Z H 8.... e. 52...... m...H.....N. m... ....H 8... S .....H 8.... 2. 2:... @1839. om... H cmém as G... H umbm 3 S... H uabm 3 8.5.96.3 mm... H no.3 3 3... H 55$ mm mm... H umdm .K Fuupasaooowm 2.. H Aa... N. ....... H 2.....N 2. :1. H a... S. :75... 2.... H a... ...V 3... H a... a. S... H 8....N 3. 5:35.. 2.... H 0...... S 2.... H a... B 2.... H mm... o. .E... 82.3% fitmflw Tim -.m_.mHM s .m.mH..w. s us...........o - _.u\.. 95...; E 3...... u. 1 3...... cm mfluimmw . T . .5|wombamaswmtilt l Sui. .o 2.... .mu .2. 8.2:??? us.V ..8.....8..oumm .sCusw £0.55. .Sz< zsfiwwu um. muss—.2: usm 3:55.... we 8.7.9.39... .8 3553 3.3.6.. zuos saw: . mm 3...; 66 .Amo.o A a. ucmgm.._v >_H:mu_._:o_m go: wga .;_LUm2:m 05cm oz. :u.3 30. a c. m:mys oz. x:< N .mvo_.w; x.mn ncm gnaw. :mw3uw2 m.;@.. .o m:_:m.;a_.a ccm m:_EE_v c._3 x.c: m.:o: o...;m.. mgso: :wu..:o. _ N... H m... R N... H mm: Nm o... H H...... 3 2:53.... 3 H .N: .3 TH eNm. m. N. H .5... 3 2.2.38... m... H GEN m... o... H cmoN N. m. H QEN 3 5...... ..... H a... . 2. q... H mm... m... N. H U... .2. 5......5. E H .6: o... M... H fiNoN S m. H USN 3 2:... 3...}. 2.... H no.2 t. E... H of... 3 . 8... H caNm 3 825.-.; N... H 0...; E E... H 3N: .3 N... H m........ E .3555... 3.... H .58 o... I... H 6.3 S E... ,- mimN mm 5...... B... H U.V...N m... m... H .NaN .3 9.... H «.NN 3. 8.2.2... 8... H .QNM 3 S... H .9: E B... H M93 o. 2.... 13......3... -ww.m H R : .m.m H m. - : .m.m H M : szsozscu - - .3: $3.... 3 ‘ 3...... 3. [$69. .NN .. Hal. flaw Vmwhmzmwozm .xczum we .;a_w xmc Hm m:_:;ozgum vcm ._a..a.mnoumm .:_.v:m .:o_;.:m» .zhz< A.mum.u we. mv.m__cs vcm mmu_;3acn we uo..wao.o;a >2 .mscgmv muzmwwz >con cam: . mm o.;:. 67 eight body weights on the 24 photoperiod than on the 14D photoperiod, while ducklings exposed to secobarbital weighed significantly more on the 14D photoperiod than ducklings on the 14 photoperiod (Table 29). Overall, eight—day body weights were heavier on the 24 photoperiod than on either the 14 or the 14D photoperiods in seven trials, while the 14 photoperiod produced heavier day eight body weights than the 14D photoperiod in seven trials (Table 29). Treatment effects on feed consumption for bobwhite and mallards during the five-day exposure period were highly significant (P < 0.001) in all trials except the bobwhite— secobarbital trial. The trial in which bobwhite were fed secobarbital resulted in fairly uniform feed consumption throughout all treatments, while all other trials produced a general trend for decreasing feed consumption with increasing dietary concentrations (Table 30). The treatment effect on feed consumption for the three— day recovery period did not result in the level of signifi- cance produced during the treatment period. For the recovery period, only four trials resulted in significant differences in feed consumption between levels (Table 31) indicating feed consumption of birds exposed to the various chemicals returned to control values in most cases. Of the four trials producing significant treatment effects, only seven diets resulted in significantly (P < 0.05) lower feed consumption than the control (Table 31). 6 8 _a:Hoa god mac—Hm.uzmo:oo agauu_v a:_muwtuc. :— who mucoEHuw.H esp o_.N H m_.¢ Ne.. . m.2: Ho.N H MN.NN mo.. H aN. qq.. QN.H o \ mm.N N.N oa.o ee.o Q_._ Ne.o H mN.o H N_.o 0N.o . .m.o H .Amo.o v sv m_ot.=ou zen. Hcmtwcc.u >.H:ou_u_:a_m wtc mHa_;uWa=m ;._x memo: m .m e_n~_ com mco_aut.:wuzcu .o.t. .vumz :o—“atacuucou Hmmco_z mgu m=_chmosauc e azu Am:o OH=H uw:.aeou wtuz mucm=Hmo.. _ot.:cu ex. oguv .ctdcoU oz“ m:_H:~wwtaoN o :H_z ._a_t. coco a:.c:c vwm: «ca.dctH=au=ou study—u Hzm.u uzd oH Louwt o za=otgd o macy=dmwt_ o oo,N H av.N N 9N.. H .o.N a Na._ H cm.a m oN._ H 6MN. 9 eN.. H 6N.: a mN._ H N.NN a. a Nm._ H ue.q. a N».. H ._.e. o NN.. H um.NN a Nm.. H e.eN a NN.. H o.NN o mN.. H a..n a. a Ha.N H um.oN a mm.N H 092 a Ha.N H G9: a Ho,N H N.NN a Hm.N . o.NN m No.N H «.mm a. a mo._ H ca.N a mo._ H av.m a mo._ H no.N a me._ H n.2: a me.. H a..._ a .N.o H _.NN N. 9 cc._ H m_.m a v... H note a q... H ”cam 0 ve.. H as.q_ a 99.. H N.m. a No._ H H.Nm N. m qq.o H mm.N a om.o H aa.N a on.o H _.q a om.e H o,v a sale . a.. a NN.o H e.v e. a mN.c H c.« o .m.o H m.e e NN.: H _.m o aN.o H o.¢ m mN.e H N.. a .N.o H N.e o. a m_.e H as. m e_.o H 0N... a c_.o H not. a a..o . em.N m N..e N me.N e N_.o H N.m m. N NN.c H gm.o m mN.o H ~N.. a mN.o . me._ a mN.o H aN.N a NN.c H ”o.m N N..e H N.e N. a .fioHNNN o .fioHugN a .ficwoem a .fichfiv a .7:. 9m 9 NNcha a. (F LVH t ._ a H N : _. M H F .. Lm Cr “““ m, um CF! .w -NM‘H‘FIJ.» m ‘(\:\‘\:;(‘\“N till: i.-:oxm-(\\s\r1;\ (l..wv ‘‘‘‘‘‘ y‘tplt. ‘!l»:,(1;:‘ouytxl —..:w:; cw...— .=o_tm; H:9=Howtg >u:-~>_c w;~ to. m:_=;oxtum v:m ._au_ALuaoowm .=_5u=a.=o_;.:uu.=pz< xtaHm_e tom m=;u__ce ecu me.z::cn No H:us.cwca x; Asmu\=t.n\msatov coHHssamzou cow. cam: ~:_::oxt.m _uH_:Ne;ooom ENE; :o__::u. =.=< mum..az w:_:;oxtgm _cH_atc;odum :_.v:u ::_:H:m_ =_z< m..;rso= =c::;..; .:m o_am_ 6 9 .306 w a. 35:8 :5: 1552:. 32.3226? a: 33.62.; 5.: £3: .m 2;: wow 23:35:39 .2... m .23.); 6.. 6mm: :EHubcwucou «3:3; 2: 9.2.3333 c 2:. AME: 3.... vac—4:5... 0.5: 3:953: 35:8 2.: 9.: 35:8 of 9.3.53.3? c 5.: 263323.50 330:... 9.33:...— =_ 0.; 3:25an 2: ..2: .33 05:6 «.3: 2533.525... :3va 2...: u... 3 .33 e .3335 o 3.5.53: _ ..- o ma.. H a..NN 9 mm.. H N.em e mm.. H 6a... o NN._ H N... o =N.. H 9.0N a NH._ H ..oN N. o:_=;....m em.N H N.mm e cN.N H N.NN c 0N.N H ..NN a om.N H ¢.aN c wN.N H v.NN o eN.N . m.om o sN.N H m..m N. .m..s.usoumm oe.m H o.Nm e on.» H N.qm a om.m H a.Nv e om.m H «.NN o om.m H ..m. o o=.m H e.oe o om.m H «.mq N. =..1=. 0N.e H aN... N Nv.m H o.oN v Nq.. H m.sN o Nq.q H ..em 0 N... . ”.NN e Nv.e H N.NN a m..N H N.en N. =o.;.=a. ee.m H Q.NH w «c.m H N.cN o qc.m H c.9N e we.m H ..NN o qo.m H N.Nm e v=.m H n.mv c Nm.N H N.qe N. =.=< H.H.Hmz --. c --- o av.c H N.e N mN.c H N.s o mn.o H e.c o mN.o H 0.0 o mN.e H N.@ N. a=.:;u...m am.o H m.e o on.e H m.e a a¢.o H v.c N an.o H N.c o an.o H 0.9 o an.o H q.e e NN.o H N.a N. .N..a.usauum Ne.o H aNN N mm.o H m.. N Ne.c H q.e o mm.o H o.v . NN.o H N.. o NN.o H v.9 9 ca.: H o.m N. =..s=. .m.o H N.e N --- o ON.c H «.9 a aN.o H v.v a aN.o . c.m e aN.o H q.¢ a cN.o H N.m N. =a.;.=u. eq.o H No.v a o..° H N.. o o«.o H 80.9 e ee.c H N.m e c..c . c.o a ev.c . e.m a NN.o HN..o N. =.z< m._;=.os E H u. 1.... a H w; .. is H... . 1:, HMH .N. 1 ._ MPH til... : NM Cr; __ me M115 2.3.5 1‘ . . 1 m. u-.- I .m illlil «(lli|.13mlit|(11 .m 1,151 : zl-IeIl.iii blii. .2525 no: 4.3.3; .2339. H3535 9... .3 9.2.5»th v.3 43.93983 .:T_E_w.:o__::wu.=.z< rim—v v3 «Ea—.5: was 32:39.: He Hawfimw: N3 :~3E.::=.f3 co::=5m:oo «EN: :32 ..H 2:2 70 Significant differences in initial mean body weights between treatments occurred in the bobwhite—ANTU trial, mallard—ANTU trial, and mallard-secobarbital trial (Table 32). The LC50 tests in which bobwhite were exposed to ANTU resulted in the number three and four (3,920 and 5,488 ppm, respectively) treatment birds being significantly (P < 0.05) lower in initial body weights than the control birds (Table 32). Conversely, in the mallard trial using ANTU, the control birds were significantly (P < 0.05) lower in initial body weights than the number one, two, three, and six treatment (2,000, 2,800, 3,920, and 10,756 ppm, respectively) birds (Table 32). The mallard—secobarbital trial started with the number four treatment (5,488 ppm) significantly (P < 0.05) lower in initial mean body weight than the control (Table 32). All ten LC50 trials resulted in significant treatment effects on body weights at day five and day eight of the tests (Table 33 and 34), with most chemically—treated diets produ- cing significantly lower body weights than the control diets. Significant photoperiod-dietary concentration interaction occurred in initial, five—day, and eight—day body weights of bobwhite fed endrin and in feed consumption during the three- day recovery period of mallards fed strychnine. The inter- actions were attributed to inconsistent effects of dietary treatments across dietary concentrations and between photo— periods. Body weights and feed consumption were highly 7 l .ezHum to; ;._1 m:o.Hn:H:wu:Ou xtuHQHu a:.muw:u:_ :_ who mgcy=HmeH ugp M? N? cm‘ 07 v v o — c m a; wv m m H\ ¢ -( -\ -\ ‘ - COCO -, Na ea m.. we umN. o.~. 0.0— o.o_ «.N— v._m on on an an on an an an am am o.N H OH ..N H so c.m H o.. Q.N H Hm ..N H NN. n..o H m... Nchne em.o H e.e. ae.c H ..a. om.o H N..N LTD} m On an em on On On On an an an .Amo.o w =. m_o:H:ou est. H:w:o.__u A_H=ao_._:m_m o.~ mHa_:uma=m ;H.: mcamz N .m 0.2:. wow m:c_Hc:H:wu:oo _c_.H .vwm: :OHHutH:mu:cu >.ado.v Haw;o_; as. a:_.=ommtaw. o s:e Auco ca:— uw:.a§ou w.w: mH:w:Hew:H .otHcoo oz. w:.v .o:H:ou qu mc_H:omw.;m. c ._w_;d zucw mc—Lzfi vww: mco.dnsu:wv:0u x;caw_fi “cm—0 uca Dd Loth m 1020;:u o wacwéumu._ _ c.N H on on o.N H as on o.N H .o an c.N H on on w.. H mm as o=.=;a...m ..N H «mm on ..N H .N am ..N . s: cm .,H H as am N.N H no so .c..s.e;ouum o.” H ... cm o.n H c.. cm o.N H ... an o.N H N.. an ..N H o.. oo =..==. ..N H as an v.N H es an ..N H as on e.N H em cm ... H a. as =o.;.:u. ..N H a.. cm ..N H .HN. aN ..N H a..N. cm N.m H m0N. .m ..N H N.. on =.z< .HHH..Hz a..o H m... on Ne.o H a,.. an m..o H e... an M¢.= H a... on on.o H q... ow ~=.=;u...m N..c H N.m. on .v.o H c.c. an ...o H a.m. cm .e.: H a.o. on NN.o H m.a. as .a._a.maauym oN.c H N.a. em em.o H e.u. on en‘s H v.@. cm wH.a H q.o. an mN.o H N,o. cs =..==. o..c H N.:. on av.c H o.m. am ov.c . m.c. ca ae.: H a... an cm.o H HTN. as =c_;.:¢. aN.o H a...QN on am.c H ue.cN cm am.c H ...N on om.= . N..N an mN.o HNm..N so =.z< H..:rsaa LEI; .. lumwh ‘ m um): 1.. : m: F 3 Hr .. ...__s_____3 lifz‘l‘v “‘11,m (SJ: ;,m(ilfr‘ iiiii l‘)1?l11urfi .H:Q=.mu.— .w:_::uxth u:u ._~H_n.a30uwm .:..v=w.=o.zH:w..=_z< >.~H~.z no. mugs—.NE wzu moH.;x::: Ho HawEHmth x: Aman.o. wuzanz sacs :uwc .o.g.:_ .mm o_2u. 7 2 .Amo.o v c. m_o:H:Ou 50.. H:w.o___v x_H:uu_H_:aHm mtg “Ha—tomazm zHH: m:uzz N .m 013—, wwn mac—dugdcwuzou —c—.: —c:.~wu to. .uwmz :oHHutH:wo:Oo Hmw;mHz oz. o:_H:wmm.;w. o 3:: cho =.:_ swcHaaou w:u1 m.:9=.c~.H _o..:ou czH wzuv .ota:9u mzH m:_H:ummtam: o ....x m:o_aw‘_u:®u:ou ALnamJ—u D=_wmmhuc— Ev wsfl mecca—dew; 9: AMT: SUMO ac—L3U flwwa w:5_u~x_u=uucov >._n..u:v ‘50—0 05.. Du Lemmi— o 55:01:: O mu=mx=umw._ _ 11- o N.o.H Nos N N.m H .NN e o.m H Nos m m.o H a...: .. =.m H use. a. N.N H .o. co m=_=;oNr.w ..N . n.o. N. N.N H a... a. N.w H u.... NN N.N H 99. mN o.a H He. 0N s.e . as. aN o.v H NH. so .m..n.u;04um N.a H cN: N. ..N H Na.. N. ..N H uuN. HN o.N H c3. mN N.o H see. NN m.c H QB. 0N m.e H :oN as =.rc=. ..a.H NH. N ....H aNN m ..c H ~.N c. N.m H NNN oN N.m H NNN oN ... . mam aN N.N H Ne. so =c.;.=m. c :.H cNN .. e.a H ..N N. ..N H NNN N. ..N H mma N. N.m H .... a. N.s H M2. mN N.¢ H mm. as =.z< m.m_.mz -.- o ma.N H aNa. N mm._ H ....N . cm.o H a..NN m. Nm.o H ..oN 9N NN.= H ..NN oN em.a H N.NN ca ~=.=;uNt.m Nm.o . N..mN ON NN.c H m.NN HN aa.o H e.NN a. mm.o H N.NN sN oN.c H ..NN an oc.c H H.om aN Nm.o H «.NN am .N..arcaoawm NN.N H H.v. N Nm.. H a...: c ca.c H a.... N. Na.a H aN... N. .o.o H no.0. «N 90.: H Nq.m. QN eq.c H ..HN cc =.t=:. .1 H ..N. . -.- o no.. H a...... :. om.o H uq.N. o. No.o H No.m. wN Ne.c . NH.¢N oN vq.o H «.NN so :o.;.:w. NN.: . MN m. .N ea.o H fl....N. o. .N.c H em.cN cN mN.c.H .0.eN mN NN.a . ‘62 NN 9N.o . mo.on oN o..o HNV.¢N so =_z< HH.H;;H= Hm ..N 1... 1%.... . .. NM: =1. 1-1wa.? 1:: 1pm.. .... .. . .m 5.11%. 11mmwxl1fi 3.9.5: . l @111 11 m . 1.1 n .1111111 .m-..l-xl1|1 N1 -ypy1111111 .1. 1:: _H:9=Ham.— . xvaHm Ho m>.H Hazy w:.=;u>:.m ucm ..uu.o:mnOuom .:.:c=o.:c_;H::u.=_z< athoH: aw. mutm__ce 5:: mm..;zaca Ho H:fl=.~o.H A: AmEctmv mH;o_mz to: :mm: .mm ~_:a_ 7 3 .Amo.c w :. m_o.H:0o Eat. H:s.o_H_v H_H:ou_H.:o.m o.c mH:_:uma=m z..z m:aoz N .m u_nn_ wwm m:o_HN:H:wu:oo _a_.H _a:Hoa to. .vwma :oHHatchu:cu umwza_; ozd m:.H=wmo.aut a u:~ Auco 0H:_ uu:_cscu w:ux mH:wEHeu:. _o:H:ou oz. mgH. .o..c0u qu a:_H:wmm.:w. o :..z mco.HNtH:wo:Oo >ua-_s o=_mmo.u:_ :_ at: mucwEHNw.H 0:. ..N... zoom a:_N=u vow: m:o_HNtH:wocco x.a.w.u “cm—o wzH 0H thm: o gasotzH o m.:i=.mm.__ 1.1 a N.H. H amN. q o.N. H a...: o ..N. H .m. e o.c H .N. .. a.o H 6.3 a. N.N H =N. so ~=_=;UN_.H N.o. H NNH. N. m.o H Nm. a. N.N H NNH. NN m.m H oN. mN o.N H Ho. NN N.N H No. aN 01m H m.. as .N..a.m;cvmm N.N. H N.oN .. c... H eooN m. ..o . aN... NN e,m H .NN. mN m.m H cNN NN N.m H n8N mN N.m H ceN as =..1:. -1 H No. . c.N. H QNo. H H... H aNN. o. N.N H NNN. N. c.o . mom. a. N.N H 6o: mN s.¢ H on. so =3.;.:;. H.NN H I.2. m m.o. H flN: o. m.m. H ea... c. ..N. H 63. o. c.e. H ace. v. N.o. H NNN. HN N.e H .NN oe =.=< HHH..uz -11 c -11 o N..N H o.aN N Nm.. H m..n a. n... H o.mm eN Nc.. H a¢..m aN HN.o H m.vN as o:.=;UN_.m cN.. . es.en aN ao.. H m.¢N «N NNH. H ..NN a. mo.. H s.sN NN em.= . N..N ON om.o H m.mm 9N oN.e H N.NN mm .N..s.msoaum 1- . N..N . Ns.N H ae.a. N vv.. H N...N a. .N.. H N..NN N. Hc.. H ~H.HN a. Hm.o H av.mN NN am.o H o.oN ow =..==. .1 H ..a. . -1- o No.N H aNNN m a... . NN.VN m. o:.: H eq.eN sN N=.c H aN..N an mm.o H m.qm co =3.;.=._ N... H aNAN a. o... H am.qN a. mo.: . NN.cN mN cc.. H a...NN NN No.c H em.om NN ea.o H fl..N... oN No.o HNN NH ow =.z< ~H_::;:m 31.1.11? ..va1 1:1. ,. 13. . H = . ..mHL. ... . um 1H .. m . 1um11H1N1m1 111NMH11N1 1:11 ..__2.,____3 1 11.1111mi111 11 1m1111 1 1. 1F 1.111.111m111. . 1. 1. 1.111111 111111111 11b11 _.:Q=How:_ . HuaHm H: .;a_w Has. w:.:;ux.Hm vcu ..aHHQLmOOUum .:.:v:w.:c_:H:m_.=_z< >.m.~_v uwH.mvtc__¢§ 1:: me_:z;:; .: .cmsgmth a; .mam.mv mHza.wz Mao; :mmz .e. u_;c_ 74 irregular and did not follow the decreasing weights with increasing dietary concentration noted in all other studies. Mean room temperature, brooder temperature, and room relative humidity are presented in Appendix C. DISCUSSION The avian subacute dietary LC50 is a fundamental para- meter which may be required by the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) and the Toxic Substance Control Act (TSCA) as a routine data point in evaluating the hazards of pesticides and other toxic chemicals when dietary exposure is the most probable route of intoxication. In addition to the subacute lethality and species susceptibility data, the LC50 provides a method for the comparisons of the short-term dietary toxicity of chemicals to a given species or between various species. Some authors have devised a toxicity ranking system, based on the LC50, to categorize the inherent toxicity of compounds in an attempt to establish and predict the relative chemical toxicities by and within chemi— cal classes (Heath et al., 1972). In addition, the slopes of the dose—response curves can be used to establish the rate of increase of mortality resulting from a proportional increase in exposure. The greater the slope, the more rapid is the increase in mortality and the lower the margin of safety. When chemical contamination in the bird's natural diet can be established by analytical techniques, the degree of hazard may be predicted by comparing the concentration of the 75 chemicalin the natural diet to the lethality of dietary concentrations in laboratory studies. When the contamination of natural diets cannot be determined, the potential letha— lity of the compound may be estimated by comparing the chemical in question with a chemical in which the field mortality is known for specific application rates. For example, if the LC50 for compound A and B are 10 and 20 ppm, respectively, and the field mortality of a species for com- pound B is known at various application rates or general residue levels, the potential mortality caused by compound A should be predictable. Certain authors use a similar analogy when they attempt to develop analytical methods of predicting lethality values, for a particular species, on the chemical and physical properties of toxic compounds (Rekker, 1980). Of key importance to the value of any classification or prediction scheme is the reliability and repeatability of the results. There are several established and acceptable methods of analysis of the dose—mortality response. Three accepted and frequently used methods are the probit analysis of Litchfield and Wilcoxon (1948), the analysis of the regression of percent mortality on dose (Boyd, 1958; Gill, 1978), and the regression of dose on percent mortality (Boyd, 1965; Gill, 1978; Montgomery and Peck, l982). Boyd (1972) states that values for the LDSO determined by one method have been found to be repeatedly significantly different from values calculated by another method with experiments run in his laboratory. 76 In looking at the analysis of data from this experiment using the Litchfield and Wilcoxon procedure, the variation between LC50 values of the three photoperiods in a particular trial, in which sufficient mortality was obtained to accurately calculate an LC50 value, was relatively low. The LC50 ratios between photoperiods for these trials ranged from 1.00:1.15 to 1.00:1.52, which was somewhat smaller than the range in LC50 values of intratrial replications (1.00:1.02 to 1.00:1.89) reported by Hill and Camardese (1981) in studies exposing Japanese quail (Coturnix coturnix japonica) to carbofuran, dicrotophos, and thionazin. LC50 variations between photo- periods were also considerably lower than the two-fold and four-fold variations in LC50 values of intertrial LC50 repli- cations exposing bobwhite and mallards, respectively, to diel— drin (Heath gt $1., 1972). The variations between photoperiod for a specific chemical and species using the direct regres- sion method ranged from 1.00:1.08 to 1.00:1.42 which was also considerably less than the intratrial replication of Hill and Camardese (1981) and the intertrial replication of Heath et 31. (1972). The inverse prediction of the regression of mortality on dose resulted in greater photoperiodic variation 1.00:1.09 to 1.00:2.34 in LC50 values within a trial than the intratrial variation of Hill and Camardese but produced intermediate variations to the intertrial replications of Hill et 31. (1975). The variation in LC50 values between analytical methods ranged from 1.00:1.03 to 1.00:2.39 and was also greater than, and intermediate to the intratrial 77 and intertrial variance, respectively, of Hill and Camardese (1981) and Hill et 31. (1975). Although there were considerable variations in LC50 determinations between analytical methods, the LC50 values calculated in this study for endrin and fen— thion were in agreement with values published in the literature with the exception of the LC50 of fenthion for mallards which was somewhat lower in this study than that published in the literature. The slope of the dose—response curves for a particular trial also showed less variation between photoperiod than the variation of intra and intertrial replication reported by Hill and Camardese (1981). The maximum variation in slopes of the dose—response curves between photoperiods of a trial was 2.17-fold in comparison to the 2.64 and 4.30-fold variations obtained in intra and intertrial replications, respectively, of Hill and Camardese (1981). Due to the considerable variation in photoperiod effects on mortality between the two species and five chemicals tested and the general inability to obtain significant differences in LC50 values between photoperiods, it must be concluded that photoperiod did not affect LC50 determinations in this study. Although feed consumption and body weights, especially during the five—day treatment period, proved on occasion to be signi- ficantly greater on the 24 photoperiod than the 14-hour photo— periods, the effects of increased chemical intake or altered feeding patterns were not strong enough to significantly affect mortality in these studies. Also, the significant differences in feed consumption and body weight between the two 14—hour 78 photoperiods were less frequent than the significant differences between the 24—hour and l4—hour photoperiods. Additional studies utilizing intratrial replication of photoperiod or alternate chemcials may prove otherwise, but based on this study and the data supplied by Heath 33 31. (1972), Hill 33 31. (1975), and Hill and Camardese (1981), it appears that intertrial variability of LC50 testing and analytical methods of analysis causes greater problems in evaluating and classi— fying chemical toxicities than does photoperiodic effects. 4. 79 CONCLUSIONS Photoperiod significantly affected feed consumption during the five—day treatment and three—day recovery periods. The 24-hour photoperiod tended to result in significantly greater feed consumption than either of the two 14-hour photoperiods. Significant differences in feed consumption between the 14 and 14D photoperiods were less frequent than significant differences between the 24 and 14 or 24 and 14D photoperiods. Photoperiod significantly affected body weights at day 0, 5, and 8 of the trial. The 24-hour photoperiod tended to produce significantly greater body weights than the 14— hour photoperiods. Again, the frequency of significant differences between the l4—hour photoperiods was less than the frequency of significant differences between the 24- hour and 14-hour photoperiods. Mortality patterns and symptoms of toxicosis were gener— ally similar between the 24, 14, and 14D photoperiods. Photoperiod did not significantly affect LC5os. 80 REFERENCES Alvey, N.G., C.F. Banfield, R.I. Baxter, J.C. Gower, W.J. Krzanowski, P.W. Lane, P.K. Leech, J.A. Nelder, R.W. Payne, K.M. Phelps, C.E. Rogers, G.J.S. Ross, H.R. Simpson, A.D. Todd, R.W.M. Wedderburn, and G.N. Wilkinson, 1977. Genstat a General Statistical Program. The Statistical Department Rothamsted Experimental Station. Boyd, E.M., 1958. The acute oral toxicity of spiramycin. Canadian Journal of Biochemistry and Physiology 36:103-110. Boyd, E.M., 1965. Toxicological studies. In: Clinical Testing of New Drugs. ed. Herrick, A.D. and Cattell, M. Revere Publishing Company. Boyd, E.M., 1972. Predictive Toxicometrics. Scientechnica LTD. p. 408. Browne, R.H., 1979. A visual assessment of the significance of a mean difference. Biometrics 35:657—665. Dingle, J.G., 1971. Feeding activity of caged layers. Poultry Sci. 50:1520-1521. Federal Register, August 22, l978a. U.S. Environmental Protection Agency, Vol. 43, No. 163. p. 37336. Federal Register, July 10, l978b. U.S. Environmental Protec— tion Agency, Vol. 43, No. 132. pp. 29727-29728. Gill, J.L., 1978. Design and Analysis of Experiments in the Animal and Medical Sciences. Vol. 1. Iowa State Univer— sity Press, Ames, IA pp. 409. Heath, R.G., J.W. Spann, E.F. Hill, and J.F. Kreitzer, 1972. Comparative dietary toxicities of pesticides to birds. USDI., U.S. Fish and Wildlife Service, Special Scientific Report - Wildlife No. 152, Washington, D.C. p. 57. Hill, E.F. and M.B. Camardese, 1981. Subacute toxicity testing with young birds: Response in relation to age and intest variability of LC50 estimates. Avian and Mammalian Wild— life Toxicology: Second Conference. ASTM STP 757. D.W. Lamb and E.E. Kenaga, Eds., Am. Soc. for Testing and Materials, pp. 41—65. Hill, P.W. and L.M. Dansky, 1954. Studies of the energy requirements of chickens. I. The effects of dietary energy level in growth and feed consumption. Poultry Sci. 33:112. 81 Hill, E.F., R.G. Heath, J.W. Spann, and J.D. Williams, 1975. Lethal dietary toxicities of environmental pollutants to birds. USDI., U.S. Fish and Wildlife Service, Special Scientific Report - Wildlife No. 191, Washington, D.C. p. 61. Joir, A., E. Di Salle, and V. Santini, 1971. Daily rhythmic variation and liver drug metabolism in rats. Biochemical Pharmacology 20:2965—2969. Klaassen, C.D., 1980. Absorption, distribution, and excretion of toxicants. In: Toxicology: The Basic Science of Poisons. J. Doull, C.D. Klaassen, and M.O. Amdur, Eds., Macmillan Publishing Co. pp. 778. Klaassen, C.D. and J. Doull, 1980. Evaluation of safety: Toxicologic evaluation. In: Toxicology: The Basic Science of Poisons. J. Doull, C.D. Klaassen, and M.O. Amdur, Eds., MacMillan Publishing Co. pp. 778. Kuenzel, W.J., 1972. Dual hypothalamic feeding system in a migratory bird (Zonotrichia albicollis). American Journal of Physiology 223:1138. Lepkovsky, S., 1973. Hypothalamic adipose tissue interrela- tionships. Federation Proceedings 31:1705. Litchfield, J.T., Jr. and F. Wilcoxon, 1949. A simplified method of evaluating dose—effect experiments. J. Pharmacol. Exp. Ther. 96(2):99-113. Montgomery, D.C. and E.A. Peck, 1982. Introduction to Linier Regression Analysis. John Wiley and Sons. pp. 504. Morris, T.R., 1967. Light requirements of the fowl. Symp. In: Environmental Control in Poultry Production. T.C. Carter, Ed., Brit. Egg Market Ed., pp. 1-35. Mu, J.Y., T.H. Yin, C.L. Hamilton, and J.R. Brodbeck, 1968. Variability of body fat in hyperphagic rats. Yale Journal of Biology and Medicine. 41:133. Nair, V. and R. Casper, 1969. The influence of light on daily rhythm in hepatic drug metabolizing enzymes in rat. Life Sciences 8:1291-1298. Nie, N.H., C.H. Hull, J.G. Jenkins, K. Steinbrenner, and D.H. Brent, 1975. Statistical Package for the Social Sciences. 2nd Edition. McGraw-Hill, pp. 675. Polin, D. and J.H. Wolford, 1973. Factors influencing food intake and caloric balance in chickens. Federation Proceedings. 32:1720. Rekker, R.F., 1980. LD50 values: are they about to become predictable? Trends in Pharmacological Sciences. October, pp. 383—384. Sanvordeker, D.R. and H.J. Lambert, 1974. Environmental modification of mammalian drug metabolism and biological response. Drug Metabolism Reviews. 3:201-229. Smith, C.J.V., 1969. Alterations of food intake of chickens as a result of hypothalamic lesions. Poultry Sci. 48:475. Squibb, R.L. and G.H. Collier, 1979. Feeding behavior of chicks under three lighting regimes. Poultry Sci. 58: 641-645. Sturkie, P.D., 1976. Alimentary canal: Anatomy, prehension, deglutition, feeding, drinking, passage of ingesta, and motility. In: Avian Physiology. P.D. Sturkie, Ed., Springer—Verlag. pp. 399. Tucker, H.A. and R.K. Ringer, 1982. Controlled photoperiodic environments for food animals. Sci. 216:1381-1386. U.S. EPA, 1982. Environmental Effects Test Guidelines. Office of Toxic Substances, Washington, D.C., NTIS Pub. PB82—232992. U.S. EPA, 1983. Pesticide Assessment Guidelines. Subdivision E: Wildlife and Aquatic Organisms. NTIS No. TB83-153908. Van Tyne, J. and A.J. Berger, 1976. Fundamentals of Ornitho- logy. John Wiley and Sons, NY, pp. 808. Weaver, W.D. and P.D. Siegel, 1968. Photoperiodism as a factor in feeding rhythms of broiler chickens. Poultry Sci. 47:1148—1154. 83 APPENDIX A Composition of guail starter. Ingredient Parts per kg Corn, #2 yellow 375.2 Soybean meal, dehulled (49% protein) 420 Dist. dried grains solubles, corn 40 Fish meal 40 Alfalfa meal, dehy. (l7%protein) 50 Animal fat, stabl. 37.6 Dicalcium phosphate 20.0 Choline chloride (50%) 3.0 Methionine hydroxy analogue 0.7 Salt 3.5 Mineral mixa 5.0 Vitamin mixb 5.0 a Mineral mix: Supplies per kg diet: Cobalt, 50 mcg; Manganese, 55 mg; Magnesium, 500 mg; Iron, 80 mg; Copper, 4 mg; Zinc, 80 mg; Selenium, (from sodium selenite), O.l mg; Carrier (dist. dried sol., corn with % tallow) to 5.0 g. b Vitamin mix: Supplies per kg diet: Vitamin A, l5,000 I.U.; Vitamin D3, l,500 I C.U.; Vitamin E, l5 I.U.; Vitamin K (menadione sodium bisul- fite complex), 2.7 mg; Thiamine, 6.0 mg; Riboflavin, l0.0 mg; Niacin, loo 0 mg; Pyridoxine, l0.0 mg; Biotin, 220 mcg; Folacin, 5.0 mg; Vitamin 812, ll.0 mcg; Carrier (dist. dried sol., corn with % tallow) to 5.0 g. 84 APPENDIX B Composition of duck starter. Ingredient Parts per kg Corn, #2 yellow 503.l Soybean meal (48% protein) 3l0 Alfalfa (l7% protein) 50 Wheat bran 6O Corn oil, stabl.a 40 dl-methionine 0.9 Limestone Dicalcium phosphate 22 Salt Choline Clz, 50% b C Vitamin mix Mineral mix Selenium mixd a Ethoxyquin add Vitamin mix: C . . Mineral mix: From Calcium C oowww UTU‘IOOO ed at l25 mg/kg diet. Supplies per kg diet: Vitamin A, l5,000 I.U.; Vitamin D3, l,500 I.C.U.; Vitamin E, l5 I.U.; Vitamin K (menadione sodium bisul- fite complex), 2.7 mg; Thiamine, 6.0 mg; Riboflavin, l0.0 mg; Niacin, lO0.0 mg; Pyridoxine, l0.0 mg; Biotin, 220 mcg; Folacin, 5.0 mg; Vitamin Biz, ll.O mcg; Carrier (dist. dried sol., corn with % tallow) to 3.0 g. Supplies per kg diet: Cobalt, 50 mcg; Manganese, 55 mg; Magnesium, 500 mg; Iron, 80 mg; Copper, 4 mg; Zinc, 80 mg; Selenium, (from sodium selenite), O.l mg; Carrier (dist. dried sol., corn with % tallow) to 0.50 g. arbonate Co. at recommended levels. 85 1|. .Loccw cgmvcaum H cam: m .uchme N .HHmccwczce wwwcmwo _ N.. H NN m... H a H... H NN N... H .N m... H ... a... H NN a... H N ...... H Na .... H .N 8.56.23 N.. H mm m... H .... G... H ..N .... H mm H... H ..a .... H NN .... H S ...... H .... N. H ..N .3538an ....HNH. m...H.... m...H .N m...HN.. .....H ... w...HNN ....Hz. m...H..m .....HMN 5...... m...HmN m...H ... ....HGN ...onN N...H ... ...NH NN N.HaN N...H..a N..H NN 8:25. ......HQ .....H ... ....H .N ....me .....H ... ...H .N N.NHS m...H .... ....HHN 2.... 22% .... H .... N... H ... N... H NN N.. 1- 3. m... H a a... H NN .... H .... m... H ... .... H NN mists: .1... H Nm w... H ... .... H ..N H... H on N... H .... H... H E ...... H .... N... H a N... H ..N 35.28% m..Hem N...H ... .....HHN eoHQ .....H ... m...H NN ....H ... m...HNs m...H..N E16... 0.... H ... m... H ... H... H ..N a... H NM m... H a m... H «N a... H 2 m... H a m... H MN 8.55. ... H Nm N... H a m. H N m. . H NH N... H a N.. H ..N m... H mm .... H a a... HNHN HE... a..;zsom AHHcHE:; qswu mama HuHuHszc aEwH game AHHvHszg mama mama .muwswcu m>Humpmm cwuoocm Eooz m>HHm_wm Lmvoocm scam w>HHmHmz cmuoogm zoom ...: S ..N i i I- SENSE... .mcwummu omen mcwcsu quHuHEzg m>HHHHwL Eco; ucm .Hwisumcmasmu concern .Hwtzumtquww sooz u xHozmma< HICHIGQN STRTE UNIV. LIBRQRIES lllllll 1| 1 31293010948127