was evaluated by determining if the sediment sample actually was
toxic to one or more aquatic organisms, as indicated by the results of
various sediment toxicity tests (Ingersoll and MacDonald 1999). The
following responses of aquatic organisms to contaminant challenges
(i.e., toxicity test endpoints) were used as indicators of toxicity in this
assessment (i.e., sediment samples were designated as toxic if one or
more of the following endpoints were significantly different from the
responses observed in reference or control sediments), including amphipod
(Hyalella azteca) survival, growth, or reproduction; mayfly
(Hexagenia limbata) survival or growth; midge (Chironomus tentans
or Chironomus riparius) survival or growth; midge deformities; oligochaete
(Lumbriculus variegatus) survival; daphnid (Ceriodaphnia
dubia) survival; and bacterial (Photobacterium phosphoreum) luminescence
(i.e., Microtox). In contrast, sediment samples were designated
as nontoxic if they did not cause a significant response in at least
one of these test endpoints. In this study, predictive ability was
calculated as the ratio of the number of samples that were correctly
classified as toxic or nontoxic to the total number of samples that were
predicted to be toxic or nontoxic using the various SQGs (predictive
ability was expressed as a percentage).
The criteria for evaluating the reliability of the consensus-based
PECs were adapted from Long et al. (1998). These criteria are intended
to reflect the narrative intent of each type of SQG (i.e.,
sediment toxicity should be observed only rarely below the TEC and
should be frequently observed above the PEC). Specifically, the individual
TECs were considered to provide a reliable basis for assessing
the quality of freshwater sediments if more than 75% of the sediment
samples were correctly predicted to be not toxic. Similarly, the individual
PEC for each substance was considered to be reliable if greater
than 75% of the sediment samples were correctly predicted to toxic
using the PEC. Therefore, the target levels of both false positives (i.e.,
samples incorrectly classified as toxic) and false negatives (i.e., samples
incorrectly classified as not toxic) was 25% using the TEC and
PEC. To assure that the results of the predictive ability evaluation were
was evaluated by determining if the sediment sample actually wastoxic to one or more aquatic organisms, as indicated by the results ofvarious sediment toxicity tests (Ingersoll and MacDonald 1999). Thefollowing responses of aquatic organisms to contaminant challenges(i.e., toxicity test endpoints) were used as indicators of toxicity in thisassessment (i.e., sediment samples were designated as toxic if one ormore of the following endpoints were significantly different from theresponses observed in reference or control sediments), including amphipod(Hyalella azteca) survival, growth, or reproduction; mayfly(Hexagenia limbata) survival or growth; midge (Chironomus tentansor Chironomus riparius) survival or growth; midge deformities; oligochaete(Lumbriculus variegatus) survival; daphnid (Ceriodaphniadubia) survival; and bacterial (Photobacterium phosphoreum) luminescence(i.e., Microtox). In contrast, sediment samples were designatedas nontoxic if they did not cause a significant response in at leastone of these test endpoints. In this study, predictive ability wascalculated as the ratio of the number of samples that were correctlyclassified as toxic or nontoxic to the total number of samples that werepredicted to be toxic or nontoxic using the various SQGs (predictiveability was expressed as a percentage).The criteria for evaluating the reliability of the consensus-basedPECs were adapted from Long et al. (1998). These criteria are intendedto reflect the narrative intent of each type of SQG (i.e.,sediment toxicity should be observed only rarely below the TEC andshould be frequently observed above the PEC). Specifically, the individualTECs were considered to provide a reliable basis for assessingthe quality of freshwater sediments if more than 75% of the sedimentsamples were correctly predicted to be not toxic. Similarly, the individualPEC for each substance was considered to be reliable if greaterthan 75% of the sediment samples were correctly predicted to toxicusing the PEC. Therefore, the target levels of both false positives (i.e.,samples incorrectly classified as toxic) and false negatives (i.e., samplesincorrectly classified as not toxic) was 25% using the TEC andPEC. To assure that the results of the predictive ability evaluation were
การแปล กรุณารอสักครู่..