15 Toxicity of Cyanide to Aquatic-Dependent Wildlife Jeremy M. Clark, Rick D. Cardwell, and Robert W. Gensemer CONTENTS 15.1 Distribution of Cyanide in the Environment 286 15.2 Exposure Pathways 287 15.3 Mechanisms of Toxicity 287 15.4 Literature Review Methods and Scope 287 15.4.1 Data Quality Determination 288 15.4.2 Data Normalization 288 15.4.2.1 Normalization to mg/kg Body Weight 288 15.4.2.2 Normalization of Toxic Dose to Cyanide Ion 289 15.4.2.3 Endpoint Normalization 289 15.5 Data Analysis 289 15.6 Discussion 290 15.6.1 Route of Exposure — Mammals 290 15.6.2 Route of Exposure — Birds 303 15.6.3 Comparisons between Birds and Mammals 304 15.6.3.1 Drinking Water 304 15.6.3.2 Food 304 15.6.3.3 Direct Injection 304 15.6.4 Simple Cyanide versus Complex Cyanide Compounds 304 15.7 Bioaccumulation of Cyanide 305 15.8 Toxicity Thresholds for Cyanide 305 15.9 Summary and Conclusions 306 Acknowledgments 306 References 307 The U.S. Environmental Protection Agency’s (USEPA) ambient water quality criteria (AWQC) for cyanide were developed in 1984 [1] and have been used extensively to develop local water quality standards for protection of aquatic life. New knowledge on the relative toxicity of bioavailable cyanide species, and the measurement of cyanide species [2] have prompted a reevaluation of the However, AWQC for protection of aquatic life do not necessarily represent concentrations that would be protective of the entire aquatic ecosystem. Consideration also should be given to the sensitivity of wildlife species whose primary habitats are aquatic or are dependent on aquatic life as a food source. Aquatic-dependent wildlife is comprised of waterfowl, shorebirds (e.g., sandpipers), and aquatic mammals (e.g., otter, beaver). 285 © 2006 by Taylor & Francis Group, LLC aquatic toxicity data that serve as the basis of the current national criteria [3; see also Chapter 14]. 286 Cyanide in Water and Soil Here, we review the toxicity of cyanide compounds to aquatic-dependent wildlife exposed via drinking water and food. Our focus was to evaluate the bioavailability 1 of cyanide from different exposure pathways and the degree to which toxicity changes when different cyanide compounds pass from the intestinal tract into the bloodstream. More specifically, the purpose of this review is to evaluate the following questions: • What is the relative toxicity of cyanide compounds to aquatic-dependent wildlife? • Does cyanide toxicity to birds and mammals differ materially by route of exposure (e.g., drinking water versus dietary exposure)? • What is the range of toxicity of cyanide compounds and do simple cyanide compounds differ significantly from complex cyanides in toxicity? • Does normalization of the toxic dose of a cyanide compound to free cyanide (HCN and CN − ) concentration provide a more accurate and comparable estimate of a toxicity threshold or reference value? • Which no-effect concentrations appear protective of birds and mammals generally, and aquatic-dependent wildlife specifically? Because toxicologic data for aquatic-dependent wildlife species are extremely limited, data for birds and mammals commonly tested in the laboratory also were used. Testing of surrogate animal species is standard practice in wildlife risk assessments, as relatively few species have been tested, compared to the large number of bird and mammal species of concern [5]. 15.1 DISTRIBUTION OF CYANIDE IN THE ENVIRONMENT Cyanide compounds are used for a wide variety of private and industrial processes and formed as (SCN − ) is produced in plants from the family Brassicaceae [2,6] 2 . Anthropogenic sources include mining operations, manufacture of synthetic fabrics and plastics, pesticides, and production interme- diates in agricultural chemical production [6,7]. Formation of cyanide compounds during treatment onment include free cyanide, simple cyanides, metallocyanide complexes, thiocyanate, synthetic Free cyanide (CN − and HCN) appears to be the primary toxic form in the aquatic environment [8,9]. In aqueous solution below pH 9.2, the majority of free cyanide exists as hydrogen cyanide, HCN [6]. Simple cyanides typically refer to water-soluble salts of free cyanide such as sodium or potassium cyanide (NaCN and KCN), respectively. In water, NaCN and KCN completely dissociate to produce free cyanide, which is a pH-dependent combination of CN − Metallocyanide salts produce variable fractions of free cyanide upon dissolution in water, the con- centrations of which depend on pH and the metal’s affinity for the CN − ion (e.g., CdCN − , Cu(CN) − 2 , Ni(CN) 2− 4 , Zn(CN) − 3 , Fe(CN) 4− 6 , etc.; see Chapters 2 and 5). Of the metal–cyanide complexes, iron– cyanide complexes often predominate in surface waters because of the abundance of iron and the high affinity of CN − for Fe 2+ and Fe 3+ (Chapter 5). Most environmentally important complexes associated with mining and mineral extraction (e.g., gold) are classified as “weak acid dissociable” (WAD) cyanides [10]. Exceptions are cobalt and iron cyanides, which are not quantified by the WAD 1 The term “bioavailability” is defined in this context as the degree to which a chemical can be taken up by an organism, subsequently interacting with a biologically important site of action [4]. 2 This family includes cauliflower, cabbage, and turnips. © 2006 by Taylor & Francis Group, LLC a result of certain chemical reactions (Chapter 4). In addition, they are formed naturally by certain plants (Chapter 3); for example, cyanogenic glycosides are produced in cassava and thiocyanate of municipal wastewater can also occur [2; and Chapter 25]. Chemical forms of cyanide in the envir- nitriles, and organic cyanides (Chapter 2). and HCN [9; and Chapter 5]. cyanide analytical method [2; and Chapters 5 and 7]. Toxicity of Cyanide to Aquatic-Dependent Wildlife 287 Biogenic sources of cyanide consist of various species of bacteria, algae, fungi, and higher plants in many food plants and forage crops, and may represent the greatest sources of cyanide expos- ure to terrestrial mammals [10]. In this regard, cassava (Manhot esculenta) has received the most study because of its elevated content of organic cyanide compounds (glycosides) and because of its importance as a major food staple in Asia, Africa, South America, and the Caribbean Islands [11]. 15.2 EXPOSURE PATHWAYS Animals may be exposed to cyanide or cyanide compounds via a number of pathways. They may ingest food or water containing natural or anthropogenic cyanide. Toxicity from cyanide-producing (cyanogenic) plants is believed to result from enzymatic release of HCN from the ingested organic cyanide compound. Hydrocyanic acid is readily absorbed by the guts of birds and mammals [10]. Secondary poisoning 3 of terrestrial vertebrates from feeding on cyanide-poisoned invertebrates and fish is unlikely, as free cyanide is neither bioaccumulated nor persistent in the environment [1,6,10]. Because secondary poisoning is unlikely, reported anthropogenic cyanide poisonings of wildlife are usually acute events resulting from water exposure. 15.3 MECHANISMS OF TOXICITY Toxicity in animals results from the binding of cyanide to the ferric heme form of cytochrome c oxidase, which is the terminal oxidase in the mitochondrial respiratory chain [6]. This blocks electron transfer from cytochrome c oxidase to molecular oxygen, thereby inhibiting cellular respiration. This results in cellular hypoxia even in the presence of normal, oxygenated hemoglobin [6]. Hypoxia concomitantly causes a shift from aerobic to anaerobic metabolism, resulting in lactate acidosis that lowers blood pH, and depresses the central nervous system, leading to respiratory arrest and death [6]. In vivo, the majority of cyanide not complexed with heme iron can be detoxified by combining with thiosulfate to produce thiocyanate, which is excreted in the urine over a period of several days [6]. More minor detoxification pathways include exhalation of HCN and conjugation with cystene or hydroxocobalamin (vitamin B 12 ) [6]. Cyanide is readily absorbed into the bloodstream and binds to hemoglobin forming methemoglobin, which is considered one of the better indicators of cytotoxicity [6]. 15.4 LITERATURE REVIEW METHODS AND SCOPE Studies on cyanide toxicity to animals were obtained using both literature databases and Internet search strategies. The terms (wildlife, bird ∗ , avian, shorebird ∗ , waterfowl, amphibian ∗ , “marine mammal,” “marine mammals”) and (toxic ∗ , ecotoxic ∗ , sensit ∗ ) and (cyanid ∗ , metallocyanid ∗ , organocyanid ∗ ) were used to search literature databases: ASFA, BIOSIS, CC Search ® 7 Editions, Water Resources Abstracts, and Zoological Record in January–February 2003. Various search engines were used to scan the Internet for relevant articles using the keywords and phrases. These searches returned 224 records. Records were retrieved and abstracts or titles screened to judge relevance and utility, yielding 49 records. Of these, 24 were available and reviewed, and 10 were accepted as adequate studies according to the criteria described in the following section. No data for marine aquatic-dependent wildlife were found. 3 Secondary poisoning represents toxicity to organisms that consume a cyanide-containing plant or animal. © 2006 by Taylor & Francis Group, LLC producing and excreting cyanide compounds (Chapter 3). Elevated concentrations of cyanide occur 288 Cyanide in Water and Soil 15.4.1 DATA QUALITY DETERMINATION Reported data were screened according to the following criteria. In some instances, these cri- teria could not be applied, and in some instances where data were accepted, qualifications were identified. • Primary publications were used when possible, rather than review papers. • The complete study design had to be detailed in the paper. • Multiple doses had to be tested with evidence of a satisfactory dose–response relationship. • Studies had to report either a lethal dose for 50% of a population (LD50), or no observable adverse effect level (NOAEL) calculated using an acceptable statistical method for each endpoint measured. 15.4.2 DATA NORMALIZATION Data were normalized from the units reported in the original study to dose in units of milligrams [mg] of cyanide ion [CN] per kilogram [kg] body weight [BW] to facilitate comparison between studies. The calculations performed are outlined below. 15.4.2.1 Normalization to mg/kg Body Weight The concentration or doses of cyanide compound (CC) were converted to a standard dose of mg tested compound (TC) per kg BW using the following equations: • Dietary food concentration reported in ppm: ppm (mg CC/kg food) ×average food consumption (kg food/day) average body weight (kg BW) = mg TC/kg BW/day (15.1) • Drinking water or injection concentration reported in mmol/kg: mmol CC/kg ×(1 mol/1000 mmol) × molec. wt.(g/mol) × 1000 mg/g = mg TC/kg BW (15.2) In some subchronic and chronic tests, the doses tested changed during the study, requiring an assumption about the average dose tested. For example, one study commenced with one-day-old chicks and lasted for nine weeks, during which time the concentration of cyanide in the food remained unchanged, but the ration consumed and, hence, dose changed with time [12]. Sample et al. [13] proposed a solution for this situation in their derivation of widely-used toxicological benchmarks for wildlife. They proposed using the animal’s average body weight for the test period to calculate average food consumption using an accepted allometric equation from USEPA [14]: food consumption rate (g/day) = 0.648(BW [g]) 0.651 (15.3) These values were expressed as kg food/day by multiplying by 0.001 g/kg. Sample et al. [13] note that this method over- and under-estimates food consumption (and hence dose) for younger and older chicks, respectively, but is an acceptable estimate of the average dose. © 2006 by Taylor & Francis Group, LLC Toxicity of Cyanide to Aquatic-Dependent Wildlife 289 TABLE 15.1 Cyanide Compounds Tested and Percentage Cyanide Contents Assumed in Normalizing Doses to CN Compound name Formula Formula weight CN molecular weight Percent CN Acetone cyanohydrin C 4 H 7 NO 85.12 26.02 30.57 Acetonitrile C 2 H 3 N 41.06 26.02 63.37 Acrylonitrile C 3 H 3 N 53.07 26.02 49.03 CN of cassava NA NA 26.02 100 Hydrocyanic acid CHN 27.03 26.02 96.26 Malononitrile C 3 H 2 N 2 66.07 26.02 39.38 n-butyronitrile C 4 H 7 N 69.12 26.02 37.64 Potassium cyanide KCN 65.12 26.02 39.96 Propionitrile C 3 H 5 N 56.10 26.02 46.38 Sodium cyanide NaCN 49.01 26.02 53.09 Succinonitrile C 4 H 4 N 2 80.10 26.02 32.48 NA = not available. 15.4.2.2 Normalization of Toxic Dose to Cyanide Ion After normalizing dosages based on total chemical concentrations, the data were normalized for CN dose (mg CN/kg BW) by accounting for the percentage of cyanide in the test compound (Table 15.1). Dosages normalized in these two manners were then compared. 15.4.2.3 Endpoint Normalization The objective of this analysis was to express all test results in terms of NOAEL values normalized to mg CN/kg BW. However, different studies reported toxicities in various ways, often hindering comparison. The methodology used by the European Commission [15] was adopted; it estimates NOAEL values by applying an uncertainty factor of 10 to the lowest observable adverse effect level (LOAEL) for a chronic endpoint, and an uncertainty factor of 100 for an LD50. These assessment factors are not well researched and are thus uncertain, especially for fast-acting gases like the free and simple cyanide compounds, which appear to possess a single mode of action. 15.5 DATA ANALYSIS Results are expressed as cumulative frequency distributions, which allowed interpretation of the data in terms of: • Relative sensitivity of birds versus mammals • Relative toxicity of exposure routes • No-effect levels protecting each organism group and exposure pathway • Data variability Normalized data for mammals exposed via drinking water (DW), food, and injection pathways © 2006 by Taylor & Francis Group, LLC are shown in Figure 15.1. Raw data are provided in Tables 15.2 to 15.4. Data for birds are shown with the mammalian data in Figure 15.2, and raw data are listed in Tables 15.5 to 15.7. 290 Cyanide in Water and Soil 0.001 0.01 0.1 1 10 100 mg CN/kg BW 10 20 30 40 50 60 70 80 90 Cumulative percent 0 100 FIGURE 15.1 Toxicity of cyanide to mammals as a function of exposure pathway, with endpoints normalized to NOAELs expressed as mg CN/kg BW (see text for details regarding normalization). Data are plotted using a cumulative distribution function of the ranked NOAELs. Data points corresponding to specific cyanide exposure pathways are denoted by (♦) for drinking water, () for food, and () for direct injection. Normalized cyanide NOAELs ranged from 0.005 to 80 mg CN/kg BW. The lowest estimated no-effect levels and, hence, the most sensitive endpoints were injection studies with mammals (Figure 15.1). The latter exhibited no-effect concentrations ranging from 0.005 to 1.4 mg CN/kg representing injection studies with complex cyanides fell into the upper portion of the dataset; although the lowest NOAEL for a complex cyanide was 0.027 mg CN/kg BW, the majority of NOAELs for complex cyanides were greater than 0.13 mg CN/kg BW (Table 15.4 and Figure 15.1). Only one avian injection study was found with a NOAEL of 0.16 mg CN/kg BW based on mortality, Dietary exposures of complex cyanides appeared much less toxic to birds and mammals than those with simple cyanides. The estimated NOAELs for complex cyanides introduced via the diet studies. The remaining avian food ingestion studies used sodium cyanide, which exhibited much greater toxicity with estimated NOAELs ranging from 0.014 to 0.11 mg CN/kg BW (Table 15.6). NOAELs estimated from drinking water studies for both birds and mammals fell within the ranges of most other NOAELs except for mammalian food ingestion studies (Figure 15.2). Estimated 15.6 DISCUSSION Data presented in the previous section will be discussed first in terms of the influence of exposure route on the relative toxicity of cyanide to mammals and birds, and then in terms of differences between mammals and birds for each exposure route. 15.6.1 ROUTE OF EXPOSURE —MAMMALS Two studies examined effects of drinking water exposure to two different wildlife species from three simple cyanide compounds [16,17]. Ballantyne’s [17] study with rabbits calculated very similar © 2006 by Taylor & Francis Group, LLC BW, although the majority (approximately 85%) was below 0.1 mg CN/kg BW (Table 15.4). Data which ranks it within the upper range of estimated mammalian NOAELs (Table 15.7 and Figure 15.2). ranged from 5.9 to 79.6 mg CN/kg BW (Table 15.3 and Figure 15.2), and included the single bird food ingestion study with cassava (Table 15.6) along with all of the mammalian food ingestion NOAELs for mammals ranged from 0.02 to 4.3 mg CN/kg BW (Table15.2), and those for birds ranged from 0.01 to 0.8 mg CN/kg BW (Table 15.5). Toxicity of Cyanide to Aquatic-Dependent Wildlife 291 TABLE 15.2 Toxicity of Cyanide to Mammals Exposed via Drinking Water Reference Species Age/ body weight Cyanide compound tested Exposure type Endpoint Effect Concentration as reported Concentration normalized to NOAEL mg compound/kg BW Concentration normalized to NOAEL mg CN/kg BW Comments [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown KCN Water gavage Mortality LD50 0.09 mmol/kg 0.059 0.023 Study design not described in detail [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown HCN Water gavage Mortality LD50 0.092 mmol/kg 0.025 0.024 Study design not described in detail [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown NaCN Water gavage Mortality LD50 0.104 mmol/kg 0.051 0.027 Study design not described in detail [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown KCN Water gavage Mortality LD50 0.115 mmol/kg 0.075 0.030 Study design not described in detail [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown NaCN Water gavage Mortality LD50 0.117 mmol/kg 0.057 0.030 Study design not described in detail [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown HCN Water gavage Mortality LD50 0.156 mmol/kg 0.042 0.041 Study design not described in detail [16] Canis latrans (Coyote) Unknown/ 7to15kg NaCN Water gavage Mortality NOAEL 4 mg/kg 4 2.124 [16] Canis latrans (Coyote) Unknown/ 7to15kg NaCN Water gavage Mortality LOAEL 8 mg/kg 8 4.247 © 2006 by Taylor & Francis Group, LLC 292 Cyanide in Water and Soil TABLE 15.3 Toxicity of Cyanide to Mammals Exposed via Food Ingestion Reference Species Age/ body weight Cyanide compound tested Exposure type Endpoint Effect Concentration as reported Concentration normalized to NOAEL mg compound/kg BW Concentration normalized to NOAEL mg CN/kg BW Comments [19] Cricetomys gambianus Waterhouse Rat (African giant) Weaning/ 87 g HCN Dietary (concentration in cassava parts) Growth rate NOAEL 110 mg/kg* food 6 5.917 HCN measured in cassava tuber [18] Sus sp. (Pig) Unknown 16.1 kg CN of cassava Dietary Mortality Unbounded NOAEL 400 ppm 17.25 17.250 [18] Sus sp. (Pig) Unknown/ 16.1 kg CN of cassava Dietary Daily weight gain Unbounded NOAEL 400 ppm 17.25 17.250 [19] Cricetomys gambianus Waterhouse Rat (African giant) Weaning/ 87 g HCN Dietary (concentration in cassava parts) Mortality Unbounded NOAEL 597 mg/kg* food 33 32.236 HCN measured in cassava peel [32] Canis familiaris (Dog) Unknown/ unknown NaCN Dietary Food consumption, blood chemistry, behavior, or organ histology NOAEL 150 mg/kg 150 79.635 As cited by [6] [19] Cricetomys gambianus Waterhouse Rat (African giant) Weaning/ 87 g HCN Dietary (concentration in cassava parts) Growth rate LOAEL 150 mg/kg* food 8 8.068 HCN measured in cassava tuber © 2006 by Taylor & Francis Group, LLC Toxicity of Cyanide to Aquatic-Dependent Wildlife 293 TABLE 15.4 Toxicity of Cyanide to Mammals Exposed via Direct Injection Reference Species Age/ body weight Cyanide compound tested Exposure type Endpoint Effect Concentration as reported Concentration normalized to NOAEL mg compound/kg BW Concentration normalized to NOAEL mg CN/kg BW Comments [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown HCN Intramuscular Mortality LD50 0.018 mmol/kg 0.005 0.005 Female value used, male not significantly different. Study design not described in detail [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown HCN Intravenous Mortality LD50 0.022 mmol/kg 0.006 0.006 Study design not described in detail [21] Oryctolagus cuniculus (Rabbit) Unknown/ unknown HCN IV injection Mortality LD50 0.66 mg/kg 0.007 0.006 [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown NaCN Intravenous Mortality LD50 0.025 mmol/kg 0.012 0.007 Study design not described in detail [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown KCN Intravenous Mortality LD50 0.029 mmol/kg 0.019 0.008 Study design not described in detail [21] Felis catus (Cat) Unknown/ unknown HCN IV injection Mortality LD50 0.81 mg/kg 0.008 0.008 [21] Rattus norvegicus (Rat) Unknown/ unknown HCN IV injection Mortality LD50 0.81 mg/kg 0.008 0.008 [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown NaCN Intramuscular Mortality LD50 0.033 mmol/kg 0.016 0.009 Male value used, female not significantly different. Study design not described in detail © 2006 by Taylor & Francis Group, LLC 294 Cyanide in Water and Soil TABLE 15.4 Continued Reference Species Age/ body weight Cyanide compound tested Exposure Type Endpoint Effect Concentration as reported Concentration normalized to NOAEL mg compound/kg BW Concentration normalized to NOAEL mg CN/kg BW Comments [21] Mus musculus (Mouse) Unknown/ unknown HCN IV injection Mortality LD50 0.99 mg/kg 0.010 0.010 [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown HCN Transocular injection Mortality LD50 0.039 mmol/kg 0.010 0.010 Study design not described in detail [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown KCN Intramuscular Mortality LD50 0.047 mmol/kg 0.031 0.012 Male value used, female not significantly different. Study design not described in detail [21] Macaca mulatta (Monkey) Unknown/ unknown HCN IV injection Mortality LD50 1.3 mg/kg 0.013 0.013 [21] Canis familiaris (Dog) Unknown/ unknown HCN IV injection Mortality LD50 1.34 mg/kg 0.013 0.013 [21] Cavia porcellus (Guinea Pig) Unknown/ unknown HCN IV injection Mortality LD50 1.43 mg/kg 0.014 0.014 [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown KCN IP injection Mortality LD50 0.055 mmol/kg 0.036 0.014 Study design not described in detail [17] Oryctolagus cuniculus (Rabbit) Unknown/ unknown NaCN IP injection Mortality LD50 0.057 mmol/kg 0.028 0.015 Female value used, male not significantly different. Study design not described in detail © 2006 by Taylor & Francis Group, LLC [...]... tolerance, increased metabolism, and reduced bioavailability can be seen proceeding from the injection to water ingestion to dietary exposure pathways However, this generalization should be viewed with caution owing to the paucity of food ingestion and drinking water data compared to injection data 15. 6.2 ROUTE OF EXPOSURE — BIRDS The avian drinking water data were obtained from a single study, and the... BW) 15. 6.3 COMPARISONS BETWEEN BIRDS AND MAMMALS 15. 6.3.1 Drinking Water Differences in sensitivity between birds and mammals exposed to cyanide in drinking water are difficult to ascertain from the data analyzed here, as all bird NOAELs and LOAELs were unbounded and, hence, more uncertain than bounded values Also, the majority of mammalian data was estimated from LD50s and, hence, used the uncertain... Eisler [6] indicates that they suggested that . occur [2; and Chapter 25]. Chemical forms of cyanide in the envir- nitriles, and organic cyanides (Chapter 2). and HCN [9; and Chapter 5]. cyanide analytical method [2; and Chapters 5 and 7]. Toxicity. owing to the paucity of food ingestion and drinking water data compared to injection data. 15. 6.2 ROUTE OF EXPOSURE —BIRDS The avian drinking water data were obtained from a single study, and. LLC producing and excreting cyanide compounds (Chapter 3). Elevated concentrations of cyanide occur 288 Cyanide in Water and Soil 15. 4.1 DATA QUALITY DETERMINATION Reported data were screened according