Application of water quality modeling technology to investigate the mortality of lobsters (Homarus americanus) in western Long Island Sound during the summer of 1999.

ABSTRACT A phased approach was taken for applying previously developed numerical models to address quantitatively whether application of 4 pesticides (i.e., methoprene, malathion, resmetbrin and sumithrin) to combat mosquitoes carrying West Nile virus could alone have caused the massive die-off of lobster observed in western Long Island Sound during 1999. Model results show that even with an overly conservative model input assumption (i.e., that the entire mass of pesticides applied in the watershed reached the open waters of Long Island Sound without any attenuation or decay in either the watershed or the Sound) the calculated 24-h average ambient levels of methoprene in the Sound were less than 0.0005 [micro]g/L and well below the lowest reported ecologic endpoint of lobster stress (i.e., 2.8 [micro]g/L stage 2 larvae LC50). Under the assumed conservative model loadings, results for malathion were highest in the East River (maximum 24-h average = 10.3 [micro]g/L) and were much lower in the western Sound area of the lobster die-off. These levels are below the lowest reported ecologic endpoint of lobster stress (i.e., 4.1 [micro]g/L larvae LC50). Model calculations for resmethrin and sumithrin were compared with ecologic endpoints for lobster stress measured for resmethrin. The lowest reported ecologic endpoints reported for resmethrin were 0.01 [micro]g/L and 0.095 [micro]g/L for reduction in adult phagocytosis after weekly exposure and larval 96-h LC50, respectively. Calculated resmethrin levels on a 24-h average basis were as high a 0.225 [micro]g/L, but were significantly lower in western Long Island Sound. For sumithrin, the maximum calculated 24-h average concentration, 0.151, occurred in Eastchester Bay in near surface waters. The calculated levels of neither resmethrin nor sumithrin reached the LC50 value for adult lobsters of >1 [micro]g/L. Malathion, resmethrin and sumithrin were also modeled with a less conservative and more realistic set of assumptions that included decay of the pesticide within the receiving water. Based on a 24-h average malathion concentrations calculated by the model were < 1 [micro]g/L in near bottom waters throughout Long Island Sound and probably did not represent a stress to the lobsters. Calculated resmethrin levels throughout near bottom waters of Long Island Sound were <0.005 [micro]g/L and, therefore, did not represent a stress to lobsters. Calculated sumithrin concentrations in near bottom waters were as high as 0.08 [micro]g/L in portions of western Long Island Sound. Assuming that resmethrin endpoints are applicable to sumithrin, it is unlikely that sumithrin could have caused mortality in adult lobsters, however, we cannot fully rule out the possibility that sumithrin may have been a stressor at sublethal levels.

KEY WORDS: Homarus americanus, lobster, Long Island Sound, model, mortality, pesticides, methoprene, malathion, resmethrin, sumithrin, West Nile virus

INTRODUCTION

A massive die-off of American lobster (Homarus americanus H. Milne Edwards, 1837) populations was observed in western Long Island Sound during the summer and fall of 1999. Numerous potential causative agents for the die-off were considered by the Long Island Sound lobster research steering committee including: pathogens (i.e., bacterial and/or viral), parasitic paramoebae, high temperature, low dissolved oxygen, prolonged hydrogen sulfide exposure, oceanic conditions and pesticide applications in the watershed surrounding Long Island Sound, to combat mosquitoes carrying West Nile virus. It is not clear whether the lobster die-off had begun prior to the initiation of West Nile virus related pesticide applications. Previously developed (Blumberg et al. 1999, HydroQual 2001a, 2001d, 2001e, 2001f) and applied (HydroQual 2004) (i.e., the hydrodynamic portion of the system wide eutrophication model [SWEM] developed for New York City Department of Environmental Protection [NYCDEP] and the contaminant assessment and reduction program's [CARP] hydrodynamic and contaminant fate and transport models) numeric water quality models may provide useful information for identifying the causative agents. Described later in detail is a phased approach for applying the models to address quantitatively whether pesticide application to combat mosquitoes carrying West Nile virus could alone have caused the massive die-off of lobster observed in Long Island Sound during 1999.

MODELING APPROACH

General Modeling Approach--Simulation Period

The modeling investigations considered the period June 1999 through November 1999. This period was selected for several reasons: (1) the decline in lobster populations was first noticed by the public during late August and early September of 1999; (2) the observed lobster mortality roughly coincided with pesticide spraying in the fall of 1999; (3) a mixing event occurred in late August 1999, which elevated near bottom water temperatures; (4) Hurricane Floyd, which occurred in September 1999, is believed to be the major mechanism for washing pesticides out of the watershed and into Long Island Sound and (5) detailed calculations of hydrodynamic transport in Long Island Sound, New York/New Jersey Harbor and the New York Bight were readily available for the 1998 to 1999 and 1999 to 2000 water years (i.e., October through September) from CARP (HydroQual 2004).

General Modeling Approach--Pesticides of Concern

Four different pesticides were applied by local jurisdictions to combat mosquitoes potentially carrying West Nile virus, including methoprene, malathion, resmethrin and sumithrin. The quantities of these pesticides applied were tracked on a weekly basis by postal zip code and were supplied by personnel from New York State Department of Environmental Conservation and Connecticut Department of Environmental Protection on a weekly, county-by-county basis. The data provided by the states were used for development of model loadings. Each of these four pesticides was modeled as an inert (i.e., nondecaying) substance for the June 1999 through November 1999 period. Based on the active ingredient mass in the total volume of pesticide used, 9.4 kg of methoprene, 4,741 kg of malathion and 319 kg of pyrethroids were applied in 1999 by various jurisdictions within the Long Island Sound drainage area and were included in the model as direct inputs. Model inputs were specified uniformly over the 7 days in each week and evenly over the shoreline of each county.

Phase I Modeling Approach

This phase of modeling work attempted to answer several fundamental questions for each of the pesticides separately: (1) If the total amount of pesticide applied, assuming no loss mechanisms, were to enter the estuary, what would the resultant concentrations in the water column of Long Island Sound be at various locations throughout the Sound and at different times during the period June 1999 through November 1999? and (2) How do the model-calculated concentrations of pesticides in Long Island Sound compare with relevant endpoint values developed by various researchers for lobster toxicity to water only exposures?

In this phase of modeling a calibrated/validated hydrodynamic transport model was used in conjunction with a rigorously tested conservative tracer/initial dilution model (HydroQual 2004). In this phase of work, each pesticide was run in the model as a conservative substance at a loading rate of the estuary set equal to the rate of pesticide application in the communities surrounding the estuary. Any retention or breakdown of the pesticides in the watershed and in the receiving water was ignored. For purposes of these calculations, the pesticide loadings were treated as a direct input to the Sound. No attempt was made to breakout modes of entry (e.g., runoff, CSO, STP, etc.) to the Sound. Calculated levels of the pesticides in the receiving water were compared with toxicity endpoints as noted earlier and are summarized in Table 1.

The Phase 1 modeling approach, by design, is overly conservative in that it addresses the upper limit, or maximum amount possible, of pesticide reaching the Sound. Further, it assumes no decay or transformations within the Sound. This phase of work results in one of two scientific, quantitative and technically defensible conclusions. The possible conclusions from this phase of modeling are that either the pesticides could not possibly be the cause of the mortality because calculated receiving water concentrations are nowhere near the relevant endpoints even under the maximum possible loading conditions assumed, or that the pesticides could be the cause of the mortality but further, more detailed and less conservative analysis would be warranted to get a more accurate depiction of the receiving water concentrations that would result under a more realistic loading condition and a consideration of transformation and decay processes.

Phase II Modeling Approach

Phase II modeling was conducted for malathion and the pyrethroids resmethrin and sumithrin to gain a more definitive understanding of the possible role those pesticides played in the Long Island Sound lobster die-off. Phase II modeling is intended to develop a more realistic representation of the chemical conditions. This phase of modeling included a first cut at representing how the chemicals behave in the environment. As in phase I, the hydrodynamic model in conjunction with the conservative tracer/initial dilution model is used; however, each chemical is modeled as a substance that follows first-order decay based on its known half life. Calculated levels of the pesticides in the receiving water were then compared with the toxicity endpoints developed by other lobster researchers. The possible outcome of this phase of modeling was to determine whether malathion and the pyrethroids should be considered a cause of lobster mortality.

It was not possible to determine whether malathion, resmethrin and sumithrin reached Long Island Sound by sources other than runoff from the land (i.e., it is not known whether any of these pesticides entered the sewer system and reached Long Island Sound by way of treatment plant effluents or combined sewer overflows rather than runoff from the land). On this basis, for phase II modeling, like phase I modeling, we conservatively assumed that all pesticides applied in the watershed reached the estuary. For purposes of phase II modeling, however, the pesticides were assumed to follow first order decay on reaching the receiving water.

Based on reported information (DeGuise et al. 2004), a first-order decay rate of 0.88 per day (i.e., after the relationship that the fraction of pesticide remaining at any time was represented by the expression [e.sup.-0.88 x time]) was assigned to malathion in the phase II model. This corresponds to a half-life of approximately 19 h. The development of the decay rate included a regression analysis to concentration data collected over time (DeGuise et al. 2004) in artificial seawater at 14[degrees]C. The regression analysis was constrained such that at time zero 100% of the chemical was present. Ignoring the time zero constraint in the regression analysis would have yielded a different relationship, 0.75 x [e.sup.-0.76 x time], which corresponds to a half-life of approximately 12 h as reported by DeGuise et al. (2004).

DeGuise et al. (2004) also reference a slower decay rate for malathion, a half-life on the order of 1 w, for raw river water. Phase II model calculations for malathion using this slower decay rate were also performed. It is not clear that this slower decay rate is appropriate for Long Island Sound because of differences in temperature, pH and organic content between river water and Long Island Sound water. Another researcher (Brownawell, B., SUNY Stony Brook, personal communication with NYSDEC) reports that based on his survey of recent fate and rate information and his familiarity with older studies for malathion, the half-life falls within the range 2 +/- 1 d. This is in general agreement with the half life of approximately 19 h developed for this modeling analysis from data in DeGuise et al. (2004) and falls between the two half-life values for which the model was run.

With regard to the pyrethroids, it is believed that they would have decayed at different rates in near surface waters than in deeper waters because of photolysis near the surface. It was reported (Brownawell, B., SUNY Stony Brook, personal communication with NYSDEC) that based on a survey of fate and rate information in the literature, appropriate half lives to use in the top 10% of the water column are 1 h for resmethrin and 10 h for sumithrin. In the remainder of the water column, a half life of 120 h was suggested for resmethrin and sumithrin. These recommended half lives were incorporated into the phase II model.

It is noted that photolysis could have been more rigorously modeled. SWEM accurately calculates the incident solar radiation and its attenuation with depth as a function of time and location for the summer/fall 1999 period (HydroQual 2001b, 2001c, Landeck Miller in press). If a relationship were known between the decay of the pyrethroids and solar radiation, it could be incorporated into the phase II model to more accurately represent the loss of pyrethoids caused by photolysis.

Hydrodynamic Model Description

The hydrodynamic transport model used in phase I and II modeling (HydroQual 2004, 2001a, 2001d, 2001e, 2001f, Blumberg et al. 1999) is based on the Estuarine, Coastal and Ocean Model (ECOM) (Blumberg & Mellor 1987) source code. The model is driven by water level, meteorological forcing, spatially and temporally varying surface heat flux and freshwater fluxes from the numerous rivers, wastewater treatment plants, combined sewer overflows and runoff from the land that enter the waters of Long Island Sound, New York/New Jersey Harbor and the New York Bight. The model solves a coupled system of differential, prognostic equations describing conservation of mass, momentum, heat and salt.

RESULTS

Phase I Modeling Results

Receiving water levels of the pesticides on a maximum 24-h average basis calculated in phase 1 modeling were compared with lobster stress endpoints (based on 48-h to 96-h exposures) and are summarized in Table I. The maximum methoprene concentration calculated by the phase 1 model was 0.0005 [micro]g/L, well below the larval LC50 of 2.8 [micro]g/L and other lobster stress endpoints. The maximum malathion concentration calculated by the model was 10.3 [micro]g/L, above the larval LC50 of 4.1 [micro]g/L and other endpoints of lobster stress. Calculated resmethrin and sumithrin maximum 24-h average concentrations were 0.225 [micro]g/L and 0.151 [micro]g/L, respectively. Both exceeded the larval LC50 of 0.095 [micro]g/L for resmethrin. Malathion, resmethrin and sumithrin were considered further with phase II modeling to gain a more definitive understanding of the possible role those pesticides played in the Long Island Sound lobster die-off that occurred in the summer/fall of 1999.

Phase II Modeling Results

Phase II modeling results are presented both as maximum concentrations over the whole model domain and in terms of spatial and temporal distributions. Assuming that all of the malathion applied to the watershed reached the receiving water, and that once in the receiving water malathion would decay at a rate of 0.88 per day, the highest 24-h average model calculated concentration of malathion was 5.4 [micro]g/L and occurred in near surface waters of Flushing Bay (Table 2). However, calculated concentrations in near bottom waters throughout the remainder of the East River and Long Island Sound were <1.0 [micro]g/L (Fig. 1). In near bottom waters of Flushing Bay, calculated malathion concentrations were <1.6 [micro]g/L (Table 2). Because the LC50 values for malathion are 4.1 [micro]g/L for larvae and 38 [micro]g/L for adults, the model results, which include a conservative loading assumption, indicate that malathion should not be considered a causative agent of the lobster die-off. Using a considerably slower decay rate (i.e., half-life of ~1 wk), model calculations showed near bottom maximum daily average malathion concentrations of 5.0 [micro]g/L in Flushing Bay and <1 [micro]g/L in Long Island Sound (Fig. 2).

[FIGURES 1-2 OMITTED]

Assuming that all of the resmethrin and sumithrin applied to the watershed reached the receiving water, and once in the receiving water the pyrethroids would decay at the depth-dependent rates described above, the highest 24-h average model-calculated concentrations of resmethrin and sumithrin occurring in near surface waters of Flushing and Eastchester Bays, respectively, were 0.034 and 0.099 [micro]g/L (Table 2). In the case of resmethrin, however, concentrations occurring in near bottom waters throughout Long Island Sound were less than 0.01 [micro]g/L (Fig. 3).

[FIGURE 3 OMITTED]

For near bottom waters of portions of western Long Island Sound, calculated sumithrin concentrations exceeded 0.01 [micro]g/L, but were less than 1 [micro]g/L. Sumithrin results for near bottom waters on a maximum 24-h average basis are shown in Figure 4,. Figure 5 and Figure 6 show selected time series results for calculated concentrations of resmethrin and sumithrin.

[FIGURES 4-6 OMITTED]

Phase II modeling summary results for malathion and the pyrethroids are presented in Table 2.

DISCUSSION

Modeling Limitations

One limitation on the modeling results obtained from each of the two phases of modeling completed is that our ability to calculate concentrations of the pesticides methoprene, malathion, resmethrin and sumithrin has not been demonstrated through a traditional model calibration/validation process because of the lack of extensive measurements of these pesticides in the open waters, sediment beds and organisms of Long Island Sound. The models, however, have undergone extensive testing and rigorous comparisons with observed data. First, The hydrodynamic transport model used in phases I and II was calibrated against measurements of salinity, temperature, water elevation, and current velocity collected in 1994 to 1995 throughout Long Island Sound, New York/ New Jersey Harbor and the New York Bight (HydroQual 2001d). Validations and skill assessments have been performed for data collected during several additional water years (i.e., October 1 through September 30) including 1988 to 1989, 1998 to 1999, 1999 to 2000, 2000 to 2001 and 2001 to 2002 (HydroQual 2004, 2001f). Second, the conservative substance/initial dilution simulation model applied for phases I and II has been used successfully for the water years 1994 to 1995, 1988 to 1989, 1998 to 1999, 1999 to 2000, 2000 to 2001 and 2001 to 2002 for calculating initial dilution results for present day loadings of several contaminants of concern including PCB homologs, dioxin/furan congeners, the metals mercury and cadmium, many PAH compounds and the organochlorine pesticides chlordane, and DDT as well as several of their daughter products (unpublished data).

Data Limitations

In the absence of more detailed information on the actual loadings of pesticides to the estuary, the modeling work performed was based on the assumption that all of the pesticides applied in the watershed reached the estuary without any attenuation or delay. Further, it was assumed that the pesticides reaching the estuary were not bound to particles or associated with dissolved macro-molecules. Lastly, to the extent practical, the photoloysis of resmethrin and sumithrin could be more accurately modeled if the decay of these pesticides were known as a function of solar radiation. Overall, the assumptions made in the absence of data tend to result in an overestimation rather than an underestimation of the pesticide levels in the receiving water to which lobster were exposed.

Definitive measurement of what the potential synergistic toxicological effects of the four pesticides acting in concert may have been, whether the toxicity of sumithrin to lobsters differed from that of resmethrin, or how the toxicity of the four pesticides might have deviated from the reported endpoints for lobster stress as a result of the effects of other site specific environmental conditions in Long Island Sound during the fall of 1999 (e.g., temperature, dissolved oxygen, hydrogen sulfide, sediment ammonia fluxes, paramoebae, oceanic conditions, etc.) was not available. Model calculations show that the concentrations of the individual pesticides in Long Island Sound had different spatial (Fig. 1, Fig. 4) and temporal (Fig. 5, Fig. 6) patterns. For this reason, it is unlikely that potential synergistic toxicologic effects of the pesticides played a role in the lobster mortality.

Interpretation of Results

The results of the modeling analysis were intended to potentially overstate the role of the pesticides in Long Island Sound lobster mortality. Highest 24-h average model outputs of pesticide receiving water concentrations were compared with endpoints of lobster stress based on 48- to 96-h exposures and pesticide loadings to the receiving water were specified as the mass of pesticide active ingredient applied in the watershed. Model calculations of pesticide receiving water concentrations on a 48- or 96-h average basis (i.e., the exposure periods for the stress experiments) are lower than those presented on a 24-h average basis. It is unlikely that the entire mass of pesticide applied in the watershed reached the receiving water. Taking into consideration the results of our modeling analysis and the model and data limitations described, the four pesticides modeled are not believed to have been a direct cause of lobster mortality in Long Island Sound but we cannot fully rule out the possibility at this time that sumithrin may have acted as a stressor. It is likely that environmental conditions in Long Island Sound during the fall of 1999 other than West Nile related pesticide concentrations are responsible for the lobster mortality.

Potential Follow-up Investigations

More research related to refining the numerical model application and improving the interpretation of the numerical model results for the four pesticides is being considered by the Long Island Sound Lobster Research Steering Committee. Specifically, gaining a better understanding of the toxicity of sumithrin to lobsters, and better understanding the sensitivity of the model calculations to the temporal signal associated with the pesticide application are under consideration.

CONCLUSION

Based on measured lethal and sublethal endpoints for lobsters, modeling results demonstrate that the pesticide methoprene was not a causative agent in the lobster die-off that occurred in Long Island Sound during the summer and fall of 1999. It is noted that methoprene has the potential to bioaccumulate. Bioaccumulation effects were not considered in the modeling analysis. Similarly, a comparison of measured lethal and sublethal endpoints for lobsters and modeling results demonstrates that the pesticide malathion was not a causative agent or contributing stressor in the lobster die-off that occurred in Long Island Sound during the summer and fall of 1999. Based on measured lethal and sublethal resmethrin endpoints for lobsters and assuming that the toxicity of sumithrin is similar to that observed for resmethrin, modeling results demonstrate that the pyrethroid pesticides, resmethrin and sumithrin, were not likely the causative agents in the Long Island Sound lobster decline of summer and fall 1999. There is not sufficient information on the toxicity of sumithrin to fully rule out the possibility that sumithrin may have been a stressor.

ACKNOWLEDGMENTS

The pesticide modeling analyses summarized in this article were supported by New York and Connecticut Sea Grant. We also acknowledge the New York City Department of Environmental Protection (under subcontract agreement with Greeley and Hansen in association with Hazen and Sawyer, P.C., and Malcolm Pirnie, Inc.) and the Port Authority of New York and New Jersey (under subcontract agreement with the Hudson River Foundation on behalf of the New York/New Jersey Harbor Estuary Contaminant Assessment and Reduction Program) for their support of the initial development, calibration, validation, and verification of the hydrodynamic and conservative tracer/initial dilution models applied for the pesticide modeling analyses.

LITERATURE CITED

Blumberg, A. F., L. A. Khan & J. P. St. John. 1999. Three-dimensional hydrodynamic model of New York Harbor region. J. Hydr. Engrg. ASCE 125(8):799-816.

Blumberg, A. F. & G. L. Mellor. 1987. A description of a three-dimensional coastal ocean circulation model. In: N. Heaps, editor. Three-dimensional coastal ocean models. Coastal and estuarine sciences, vol. 4. Washington, DC: American Geophysical Union. pp. 1-16.

DeGuise, S., J. Maratea & C. Perkins. 2004. Malathion immunotoxicity in the American lobster (Homarus americanus) upon experimental exposure. Aquat. Toxicol. 66:419-425.

DeGuise, S., J. Maratea, E. S. Chang & C. Perkins. 2005. Resmethrin immunotoxicity and endocrine disrupting effects in the American lobster (Homarus americanus) upon experimental exposure. J. Shellfish Res. 24(3):781-786.

HydroQual, Inc. 2004. A model for the evaluation and management of contaminants of concern in water, sediment, and biota in the NY/NJ harbor estuary. Hydrodynamic sub-model. Draft report prepared for the Hudson River foundation on behalf of the contaminant assessment and reduction program (CARP). 552 pp.

HydroQual, Inc. 2001a. Newtown Creek water pollution control project East River water quality plan/task 10.0 system-wide eutrophication model (SWEM)/subtask 10.1 construct SWEM. Report prepared for the City of New York, Department of Environmental Protection; New York, New York under subcontract to Greeley and Hansen. 75 pp.

HydroQual, Inc. 200lb. Newtown Creek water pollution control project East River water quality plan/task 10.0 system-wide eutrophication model (SWEM)/subtask 10.2 obtain and reduce loading/water quality data. Report prepared for the City of New York, Department of Environmental Protection; New York, New York under subcontract to Greeley and Hansen. 111 pp.

HydroQual, Inc. 2001c. Newtown Creek water pollution control project East River water quality plan/task 10.0 system-wide eutrophication model (SWEM)/subtask 10.4 calibrate SWEM water quality/subtask 10.6 validate SWEM water quality. Report prepared for the City of New York, Department of Environmental Protection; New York, New York under subcontract to Greeley and Hansen. 141 pp and appendices.

HydroQual, Inc. 2001d. Newtown Creek water pollution control project East River water quality plan/task 10.0 system-wide eutrophication model (SWEM)/subtask 10.6 Validate SWEM Hydrodynamics. Report prepared for the City of New York, Department of Environmental Protection; New York, New York under subcontract to Greeley and Hansen. 312 pp.

HydroQual, Inc. 2001e. Technical Memorandum--Response to: "Report on the calibration and validation of the system-wide eutrophication model hydrodynamic model" by Connecticut Department of Environmental Protection. Report prepared for the City of New York, Department of Environmental Protection; New York, New York under subcontract to Greeley and Hansen. 59 pp.

HydroQual, Inc. 2001f. Newtown Creek water pollution control project East River water quality plan/task 10.0 system-wide eutrophication model (SWEM)/sub-task 10.3 calibrate SWEM hydrodynamics. Report prepared for the City of New York, Department of Environmental Protection-New York, New York under subcontract to Greeley and Hansen. 261 pp.

Landeck Miller, R. E. & J. P. St. John. 2005. Modeling production in the Lower Hudson River Estuary. In: J. Levinton, editor. The Hudson River ecosystem. New York: Cambridge University Press. In Press.

Walker, A. N., P. Bush, T. Wilson, E. Chang, T. Miller & M. Horst. 2005. Metabolic effects of acute exposure to methoprene in the lobster, Homarus americanus. J. Shellfish Res. 24(3):787-794.

Zulkosky, A. M., J. P. Ruggieri, S. A. Terracciano, B. J. Brownawell & A. E. McElroy. 2005. Acute toxicity of resmethrin, malathion, and methoprene to larval and juvenile American lobsters (Homarus americanus) and analysis of pesticide levels in surface waters after Scourge, Anvil, and Altosid application. J. Shellfish Res. 24(3):795-804.

ROBIN E. LANDECK MILLER, (1) JAMES R. WANDS, (1) KAREN N. CHYTALO (2) AND RICHARD A. D'AMICO (2)

(1) HydroQual, Inc., 1200 MacArthur Boulevard, Mahwah, New Jersey 07430; (2) New York State Department of Environmental Conservation, 205 N. Belle Mead Road, E. Setauket, New York 11733

* Corresponding author. E-mail: [e-mail]

TABLE. 1.
Summary of phase 1 model results and comparisons to endpoints.

  Parameter                   Methoprene             Malathion

Lowest sub-lethal                 5 (2)                 5 (3)
  endpoint, [micro]g/L      Post-larval molting    Effects on adult
                              effects                phagocytosis
Lowest larval lethal              2.8 (6)               4.1 (5)
  endpoint, [micro]g/L      Stage 2 larvae LC50    Larvae LC50
Lowest adult lethal               >221 (7)              38 (3)
  endpoint, [micro]g/L      Adult LC50             Adult LC50
Maximum phase I model             0.0005                10.3
  calculation, [micro]g/L
Location of maximum         Sunken Meadow State    Flushing Bay near
  calculation                 Park near surface      surface

  Parameter                   Resmethrin          Sumithrin (1)

Lowest sub-lethal               0.01 (4)            0.01 (4)
  endpoint, [micro]g/L      Effects on adult    Effects on adult
                              phagocytosis        phagocytosis
Lowest larval lethal            0.095 (5)           0.095 (5)
  endpoint, [micro]g/L      Larvae LC50         Larvae LC50
Lowest adult lethal             >1 (4)              >1 (4)
  endpoint, [micro]g/L      Adult LC50          Adult LC50
Maximum phase I model           0.225               0.151
  calculation, [micro]g/L
Location of maximum         Flushing Bay near   Eastchester Bay near
  calculation                 surface             surface

(1) Endpoints reported for pyrethroids are based on resmethrin.

(2) Walker et al. 2005.

(3) DeGuise et al. 2004.

(4) DeGuise et al. 2005.

(5) Zulkosky et al. 2005.

(6) S. DeGuise, personal communication.

(7) M. Horst, personal communication.

TABLE 2.
Summary of phase II model results and comparisons to endpoints.

Parameter                        Malathion            Resmethrin

Lowest sub-lethal
  endpoint, [micro]g/L            5 (2)                0.01 (3)
                            Adult phagocytosis   Adult phagocytosis
Larval lethal endpoint,
  [micro]g/L                4.1 larval LC50 (4)  0.095 larval LC50 (4)
Adult lethal endpoint,
  [micro]g/L                38 adult LC50 (2)    >1.0 adult LC50 (3)
Location of maximum
  calculation               Flushing Bay         Flushing Bay
Maximum phase II model
  Calculation, [micro]g/L          5.4                   0.034
Maximum near bottom
  phase II model
  calculation, [micro]g/L          1.6                   0.030
Timing of near bottom
  maximum calculation             9/12/9                9/15/99

Parameter                       Sumithrin (1)

Lowest sub-lethal
  endpoint, [micro]g/L            0.01 (3)
                            Adult phagocytosis
Larval lethal endpoint,
  [micro]g/L                0.095 larval LC50 (4)
Adult lethal endpoint,
  [micro]g/L                >1.0 adult LC50 (3)
Location of maximum
  calculation               Eastchester Bay
Maximum phase II model
  Calculation, [micro]g/L           0.099
Maximum near bottom
  phase II model
  calculation, [micro]g/L           0.082
Timing of near bottom
  maximum calculation              9/18/99

(1) Endpoints reported for pyrethroids are based on resmethrin.

(2) DeGuise et al. 2004.

(3) DeGuise et al. 2005.

(4) Zulkosky et al. 2005.