Characteristics of the episodic toxicity of six metals, 44 pesticides, four physical water parameters, and 27 other assorted stressors (Table 1) are summarized in the database. It is not possible to provide a summary of biological responses to each stressor here; however, an overview of discernible responses to particular characteristics of episodic stressor exposure is provided. Generally, increased harm to the organism was observed when the number of pulse exposures and the length of each pulse exposure were increased and when a decrease in the recovery time between pulses occurred. However, there were exceptions, most often caused by differences in the sensitivity of exposed organisms, probably as a consequence of morphological differences 20, that is, significant species- and age-related differences in the uptake and depuration rates of specific chemicals.
Naturally, the magnitude of stress (concentration of the stressor) is the primary determinant of toxic effect. At low concentrations (below the threshold of effect level), recovery from exposure to the stressor was observed regardless of the number of pulses and the length of exposure. However, at higher concentrations, the complexity of episodic exposures (the relationship between the length of exposure and the number of pulses) became apparent. In the text that follows, time durations of 24 h or less are expressed in hours and durations greater than 24 h in days.
Effect of repeated pulse exposures
Increasing the number of pulsed exposures to a toxicant generally resulted in increased toxicant effects on the organism, probably because of a threshold tissue burden being exceeded. For example, Pynnonen 21 showed that repeated pulse exposures reduced the ability of freshwater clams (Anodonta anatina and Unio pictorum) to eliminate aluminium.
With exposure concentration kept constant, Diamond et al. 22 determined that the survival of P. promelas was significantly lowered when the number of copper pulses increased from one to two. In addition, the biomass of P. promelas was negatively correlated with the number of pulses 22. Although Bearr et al. 23 reported that P. promelas survival was not significantly different between the single- and the double-pulse treatments to copper (with 4 d of recovery time between), the addition of a third pulse (4 d after the second pulse) did result in a significant decrease in fish survival. A similar response was observed when Andersen et al. 24 exposed D. magna to dimethoate, an organosphosphate insecticide. After the first exposure pulse, recovery from immobility was seen for all pulse durations. However, after the second pulse, mortality occurred and increased significantly during the recovery period for pulse durations greater than 2 h.
Even if no mortalities occur as a result of increasing the number of pulse exposures, sublethal effects are sometimes significant. For example, in an experiment testing the effects of the exposure frequency of ammonia (eight or 24 pulses of 0.2–0.4 mg/L over a 53-d period) on brown trout (Salmo trutta), Milne et al. 25 recorded no mortalities. However, lower fish weights were recorded in some instances, and growth, gill condition, organ weights, and hematocrit were all significantly affected by repeated exposures, particularly at the higher exposure frequency 25.
Effect of the duration of the pulse exposure
Increasing the length of exposure to a cadmium pulse was found by Gama-Flores et al. 26 to reduce the population growth rate of cladocerans (Moina macrocopa) and rotifers (Brachionus calyciflorus). A similar response was observed for P. promelas biomass, which was negatively correlated with the length of copper pulse exposure 22. However, the relationship between exposure duration and biological effect can be complicated. Bearr et al. 23 report that P. promelas fry exposed to two copper pulses of 3, 6, or 24 h had significantly lower mortality than fry exposed to two pulses of 12 h in length, regardless of the recovery time between pulses. Bearr et al. 23 suggest that the 3- and 6-h exposure durations might have been too brief to elicit adverse effects on survival, whereas the 24-h exposure initiated acclimation of sorts in surviving organisms.
For D. magna, Hoang and Klaine 27 found that daphnids exposed to a single 4- to 24-h pulse of between 800 and 2,000 µg/L selenium showed no mortality during exposure but latent mortality postexposure, with mortality increasing with pulse exposure duration and exposure concentration. In the case of zinc, daphnids were more sensitive to 24-h pulses of 250 to 1,000 µg/L (with continued mortality even postexposure), whereas 3- to 6-h pulses at high concentrations resulted in no effects on daphnid survival or reproduction 8.
Effect of recovery time between pulse exposures
The literature indicated that longer recovery times between multiple exposure pulses led to greater survival in D. magna8 and amphipod (Hyalella azteca) 6 exposed to copper, and D. magna exposed to selenium 27 and zinc 8. A recovery period of at least 3 d between pulses of the organophosphate insecticide chlorpyrifos was necessary for daphnids to recover from a 0.5-µg/L pulse, while a longer recovery period of 4 d was necessary for a 1.0-µg/L pulse 28. Milne et al. 25 report that rainbow trout (Oncorhynchus mykiss) and S. trutta juveniles exposed to repeated pulses of potentially lethal ammonia concentrations were able to survive if enough time for recovery was allowed. However, it appears that if the length of exposure exceeds a certain duration or the stressor concentration exceeds a certain threshold, recovery time between pulses becomes irrelevant. For example, Naddy et al. 28 show that when daphnids were exposed to two 12-h pulses of 0.5 µg/L chlorpyrifos, >85% mortality was observed, regardless of the interval between pulse exposures, which was 0, 3, 7, or 14 d.
Bearr et al. 23 reported a more complex relationship between postexposure mortality and length of recovery between pulses. Pimephales promelas exposed to 24-h copper pulses of 30 to 40 µg/L had significantly higher mortality when pulses were spaced farther apart in time (to a threshold) than when pulses of the same magnitude were spaced more closely, that is, exposures having a 2- to 4-d recovery time between pulses resulted in less mortality than did treatments with shorter (12–24 h) or longer (5–6 d) recovery times 23. Diamond et al. 8 report on an experiment in which P. promelas's biochemical defense system was activated by copper exposure for approximately 2 to 4 d, after which it ceased if copper was removed from the media, leaving the fish susceptible to a new pulse.
Effect of organism age on toxicity of a pulse exposure
Hoang and Klaine 29 investigated the effects of a 12-h pulse of either arsenic, copper, selenium, or zinc on D. magna of various ages (3 h to 10 d old) and monitored the effects after 20.5 d of recovery. They found that the 21-d mortality increased with organism age from 3 h to 2 d and then decreased with age from 2 to 10 d for arsenic and selenium. For copper and zinc however, the 21-d mortality increased with organism age from 3 h to 4 d old and then decreased with age from 4 to 10 d. No difference in the 21-d growth was detected by the study for the metals tested. Reproduction was affected, however, with 21 d cumulative reproduction for arsenic decreasing with age from 3 h to 3 d old and then increasing with age from 3 h to 3 d. For copper and zinc, the 21-d cumulative reproduction decreased with age from 3 h to 4 d and then increased with age from 4 to 10 d, and for selenium, the 21-d cumulative reproduction decreased with age from 3 h to 2 d and then increased with age from 2 to 10 d. Hoang and Klaine 29 concluded that D. magna are particularly susceptible to these four metals at the time of first moulting and for a short period thereafter.
However, Andersen et al. 24 exposed D. magna aged <24 h and 3 d to a single pulse of dimethoate, an organophosphate insecticide, and found no significant difference in recovery based on age. In addition, Hosmer et al. 30 exposed D. magna of varying ages (<24 h, 4–6 d, 8 d, 11 d) to varying concentrations of the insecticide fenoxycarb and monitored effects for 21 d. There were no significant effects on survival or time to first brood of first- and second-generation daphnids in any age group at any exposure concentration.
Australian crimson-spotted rainbowfish (Melanotaenia fluviatilis) larvae were exposed by Reid et al. 31 to a 2-h cyanazine pulse at concentrations between 1.9 and 43.2 mg/L, and recovery was monitored for a further 94 h. Five age groups were exposed (0, 3, 6, 9, and 12 d posthatch). Toxicity was found to decrease with larval age. There was a considerable reduction in cyanazine toxicity between 0 and 3 d posthatch, with no further reduction in toxicity for 6-, 9-, and 12-d-old fish. The authors proposed this as evidence of hepatic xenobiotic metabolism, suggesting that the liver of larval rainbowfish becomes increasingly more functional with age. However, results could also be influenced by the greater respiration rates and small body surface area of smaller larvae 31.
Finally, Parker and McKeown 32 exposed eggs and alevin of Kokanee or Sockeye salmon (Oncorhynchus nerka) to a 24-h pulse of pH 4 and reported variation in sensitivity by developmental stage, the most sensitive stage being early embryonic development and newly hatched alevins. The most significant effect on survival and median hatching time was noted when the eggs were episodically exposed during early development, and exposure at later stages had no apparent effect on egg survival 32.
Consequently, when attempting to determine the hazard of an episodic pollution event, the specific situation must be investigated, that is, the specific affected species involved and its developmental stage, the specific chemical stressor (not just chemical grouping) and its concentration, and naturally, the number of pulses, length of pulse exposure, and recovery time between pulses.