OCSPP 810.3100: Treatments for Imported Fire Ants
Response to Public and SAP Comments  

Scope
The SAP recommended clarifying whether the information provided throughout sections (a) through (h) refer to laboratory or field studies.
EPA Response: Where applicable, the Agency will clarify whether the general information provided in sections (a) through (h) refer to laboratory studies, field studies, or both.
Definitions (Section (c))
The SAP had the following comments on the definitions in section (c):
 Add definitions for red, black, and hybrid imported fire ants
 Regarding definitions for mound ((d)l) and colony ((d)c), the SAP stated that mounds may encompass more than just the above ground portion and that the mound usually refers to the structure above and below ground that the ants live in while the colony refers to the individuals that inhabit the mound (Vinson, 1997)
 Revisit the definition of "moribund" and include that after a reasonable time of no recovery, the determination will be made that no recovery will occur at all (this may be different with larvae where the determination may not be failure to recover after a given period, but rather failure to molt to the next instar)
 Add a definition for positive control
One public commenter also had numerous suggestions on the guideline definitions, including:
 changing the definition of "photoperiod" to "the relative amount of time during which an organism receives light" 
 changing the definition of "moribund" to "moribund refers to pests that are exhibiting abnormal behavior to stimuli after becoming intoxicated with a substance; an example is on their backs with only a single appendage twitching" 
 adding "return of a stretched appendage to rest is acceptable and insects exhibiting this response are considered dead" to the definition of mortality
 adding a definition for "site" as "site refers to a small, specific study location where evaluations are done. A site can be made up of various plots" 
 adding a definition for "knockdown" as "knockdown refers to a combined count of ants that are moribund and dead (mortality) at a given time point after exposure to treatments"
 adding a definition for "locality" as "locality refers to a large area where study species are known to exist, from within a locality treatment and controls sites should be selected."
 adding a definition for "active mound"
EPA Response: 
 While specific definitions were not added for red, black, and hybrid imported fire ants, the following language was added to section (a)a: "This guideline provides recommendations for the design and execution of laboratory and field studies, the results of which, when combined, are used to evaluate the performance of pesticide products for the treatment of ants in the Solenopsis saevissima complex, which includes in part, for the purposes of the guideline, only S. invicta, S. richteri, and their hybrids in connection with registration of pesticide products under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) (7 U.S.C. 136, et seq.)."  
 The definition for "mound" was revised to "mound refers to the above-ground and below-ground structure constructed for thermoregulation of a colony; an active mound contains all or part of a colony. A mound is considered active when at least 20 workers exit the mound immediately following disturbance (Oi & Oi 2006)."
 The definition of moribund was deleted because it is not used as an endpoint to determine pesticide product efficacy.
 A definition of positive control was not added as positive controls are not recommended for any of the studies in the guideline.
 The definition of photoperiod comes from Triplehorn et al. (2005) and remained as written.
 The suggested addition was not added to the definition of mortality; the current definition is unambiguous, and the addition would make the endpoint more subjective.
 A definition for locality was not added since the SAP determined that multiple geographic locations were not needed and therefore were removed from the document. However, the Agency  recommended that a minimum of two sites be used. The following definition of site was added to the specific guidance for field studies sections for clarity, "A site refers to a contiguous stretch of land; therefore, a minimum of two separated (i.e., non-contiguous) stretches of land should be used for the study."
 Knockdown was not included as an endpoint in the document and therefore was not added to the definitions.
 A definition for active mound was included within the definition for mound.

Objectives (Section (d))
The SAP agreed that the statement "For products that control red imported fire ants, the objective is to determine if the pesticide application has residual efficacy" should be revised to "For products designed to control fire ants, the objective is to effectively reduce active fire ant colonies and foraging for the length of time claimed on the label. In all cases, the scientific objectives and label claims should be clearly stated, and all treated areas should be compared to areas that have received no treatment or a diluent-only treatment." This is because products (even residual insecticides applied as individual mound treatments) work by killing entire colonies and it takes fire ants time to recolonize a depopulated area. If a residual product that remains in the soil a long time after application does not kill the entire colony, survivors will move a few meters and reestablish. Residual efficacy is primarily important for products claiming season-long or longer control such as certain broadcast residual granular or liquid products. 
The SAP also mentioned that looking only at colonies/mounds/nests is less instructive than numbers of mounds and fire ants in an area since a mound can be vacant because fire ants have merely been displaced and therefore pest control would not have been achieved. Therefore, the SAP recommended mounds per unit area would provide more informative data for broadcast treatments.
Numerous public commenters requested clarification or had suggestions for the Objectives in section (d), including: the true objective would be to determine whether the product kills fire ants and/or the colony/mound/nest at the proposed label rate; residual efficacy may not be the objective and instead the product could be a contact insecticide; it is unclear if the laboratory tests are required in conjunction with the field tests; and whether rate titration data would be needed.
EPA Response: The objective was revised to focus on kill/control of colonies. Typically, if a product kills a colony, this would include controls claims. Therefore, for products that kill/control imported fire ants, the objective of product performance testing is to test the lowest labeled rate to verify kills/controls of a colony. In addition, while already specified within the guideline, it was more clearly stated throughout that laboratory tests should be conducted in conjunction with field tests in order to support any kills/controls claims for the colony by demonstrating mortality of workers, the queen and in some cases brood. The Agency evaluates the time to re-invasion/re-establishment of new colonies only when a registrant specifies a duration of control (e.g., season long control). In all cases, the scientific objective and intended label claims should be stated clearly, and all treated colonies should be compared to colonies that received no treatment or a diluent-only treatment.  
Residual efficacy may be an objective if a duration of control is specified (e.g., season long). Time frames dictating duration of control should be clearly stated on the proposed label and study objectives. Contact insecticides primarily affect foragers whereas the objective of the proposed fire ant guideline (810.3100) is to test products targeting fire ant colonies. Testing contact insecticides is therefore beyond the scope of the proposed guidelines and has been subsumed under the Premises guideline (810.3500). 
Test Materials and Treatments (Section (d))
One public commenter questioned the recommendation that the pesticide product be stored at ambient temperature and humidity for at least a day prior to testing as this does not reflect a real-world use pattern. Another commenter requested clarification on whether test materials (concentrated or pre-diluted final products) should be exposed to outside environmental conditions in the study area or if test materials should be stored in typical residential (air conditioned) and laboratory conditions before use. The SAP mentioned that products do not need to be stored at ambient temperature and humidity at least one day before use because most products are in sealed, moisture-proof containers. Baits remain in better condition if stored in cool, dry conditions.
Regarding the use of the lowest labeled rate in testing, one commenter indicated that the rate should be based on what the label states which could have information about the density of fire ants and the severity of the problem.
EPA Response: Allowing a stored product to acclimate to ambient temperatures prior to testing is meant to standardize testing methods. No change was made to the guideline. Testing of the lowest labeled rate applies when conducting field tests. If needed, one may consult with the Agency for guidance on testing of labeled rates.
Testing Conditions (Section (d))
Public commenters indicated that more flexibility is needed for the temperature and humidity recommendations. Specifically, commenters recommended that the relative humidity range should be revised to 50-80% as keeping the humidity at a maximum of 80% will minimize fungal growth on the test materials and food sources and that the temperature range should be +/- 5 degrees instead of +/- 1 degree. The SAP agreed that maintaining the suggested humidity range is unrealistic, would require a growth chamber, and would result in substances like Fluon(R) breaking down. Instead, the SAP suggested the use of an artificial nest lined with moistened dental stone or plaster of Paris which would mimic nature where there is a nest at relatively high humidity, but foraging is in variable levels of humidity (see Williams 1989). In addition, one public commenter mentioned that a photoperiod of 12:12 L:D is adequate for fire ants because their behavior is more dependent on temperature than photoperiod.
EPA Response: The humidity range recommended by the public commenter includes humidity levels that are too low for field testing against fire ants. The humidity range recommended in the guideline (70-90%) is supported by the literature (Porter and Tschinkel 1987). The temperature range in the guideline remained as is for field testing. However, the guideline was amended to include the following for laboratory tests: "During product performance testing in the laboratory, laboratory testing may be conducted at ambient temperature and RH in the testing arena as long as conditions are optimal for foraging behavior, such as, 82°-91°F (Banks et al. 1981). Conditions that are too cool or dry will inhibit normal colony activity. In addition, temperature and RH within the nest cell should also be maintained at conditions optimal for brood development, such as >= 65 % RH and 82°-91° F (Williams 1989). Environmental conditions for laboratory testing, including temperature, humidity, and photoperiod during testing, should be supported by cited literature." Therefore, the 12:12 L:D photoperiod is acceptable provided the colonies have been acclimated to those laboratory conditions and a reference to testing conditions appropriate for fire ants has been cited.
The following references were added:
Banks, W. A., C. S. Lofgren, D. P. Jouvenaz, C. E. Stringer, P. M. Bishop, D. F. Williams, P. D. Wojcik, and B. M. Glancey. 1981. Techniques for Collecting, Rearing, and Handling red imported fire ants. USDA, SEA, AAT-S-21. 
Williams, D.F. 1989. An improved artificial nest for laboratory rearing of the imported fire ant, Solenopsis invicta (Hymenoptera: Formicidae). Florida Entomologist 72: 705-707.
Choice of Endpoints (Section (d))
Regarding endpoints as described in section (d), public commenters were concerned about observing mortality of queens and presence/absence of brood in the field as this would not be possible while preserving the integrity of the mounds. One commenter also mentioned that mounds may not be present for counting in some Red Imported Fire Ant ( RIFA)-infested areas and can be affected by mowing and rainfall. Finally, the SAP suggested that parentheses be added to indicate which endpoint is appropriate for field versus lab studies (e.g. mortality (lab studies), number of active mounds (field studies), forager numbers (field studies)).
EPA Response: This guideline section provides examples, which may or may not be pertinent to a proposed study. Therefore, mortality of queens may not be applicable to a field-study design, likewise mound counting may not be applicable to a lab-study design; the registrant is responsible for designing appropriate field and laboratory studies to support their product's performance for kill/control of colonies. 
Test Organisms (Section (d))
Regarding the target of the test as described in section (d)a.vi., one public commenter pointed out that field testing cannot exclude immature stages. The commenter also acknowledged that the red imported fire ant is a definitive species, Solenopsis invicta, and that other fire ant species include the black imported fire ant and a hybrid of the red and black imported fire ants. The SAP agreed that black imported fire ants and the hybrids should also be considered and therefore suggested changing references from "red imported fire ants" to "imported fire ants." They also suggested adding the European fire ant (Myrmica rubra) and little fire ant (Wasmannia auropunctata) because they are emerging fire ant pests in the NW/NE and Florida areas, respectively, and stated that S. invicta would not be an equivalent proxy for trials of the tropical fire ant, Solenopsis geminata, or the southern fire ant, Solenopsis xyloni. Finally, most of the SAP members indicated that there were no real differences between the biology of the red and black imported fire ant and their hybrids, so there was no need for a species determination before a field trial was initiated. They mentioned that voucher specimens could be preserved in 70% ethanol so that if species identity is later questioned the voucher specimens could be referenced; this would be particularly useful for new and invasive species.
EPA Response: Field testing may or may not include assessments of immature stages based upon test objectives, therefore section (d)a.vi. remained as written. Details regarding the field observation of immature stages are included in IGR section (i)c.v. Additional endpoints for laboratory testing with immature stages may be found in the IGR section (m) of the proposed guideline. 
The Agency agreed with the public commenter and the SAP that black imported fire ants and the hybrids with red imported fire ants should be included within the scope of the guideline.  The following language was added to section (a) a.: "This guideline provides recommendations for the design and execution of laboratory and field studies, the results of which when combined, are used to evaluate the performance of pesticide products for the treatment of ants in the Solenopsis saevissima complex, which includes in part, for the purposes of the guideline, only S. invicta, S. richteri, and their hybrids in connection with registration of pesticide products under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) (7 U.S.C. 136, et seq.)."  
The Agency disagreed that S. geminata, S. xyloni, or any other congeners beyond the S. saevissima complex, as described for purposes of this guideline, should be included in this guideline due to biological and geographic differences from S. invicta, S. richteri, and their hybrids. Similarly, biological differences among Myrmica rubra and Wasmannia auropunctata would necessitate species-specific protocols that are beyond the scope this guideline. 
The guideline does not require voucher specimens.   
Representative Sampling (Section (d))
The SAP recommended adding a positive control for field tests to account for environmental conditions (drought, excess heat and rain, etc.) that can result in failures not attributed to the product's efficacy which could indicate that the product should be retested; the positive control is particularly important for baits. Multiple commenters did not support the testing of an equal number of treatment and control replicates, citing that this is unnecessary, cost-prohibitive, and that 10 negative control mounds as opposed to 40 would suffice or as one commenter suggested, a single plot. The SAP added that only one replicated control group is needed for studies that test multiple rates/bait ages/products. Another commenter included that the 17-meter minimum distance requirement between plots further adds to the difficulty of having multiple control plots.
EPA Response: Positive controls were not added because comparative data is not used to support product performance. While multiple commenters found a balanced design to be unnecessary, a balanced design is suggested as most robust for this type of study. The option to consult a statistician when developing test protocols is provided in the guideline. Testing of a single plot is not statistically valid. The number of plots should be supported by a power versus sample size analysis.  
The 56 ft (17 m) minimum distance recommendation remained in the guideline because this distance is supported by literature (Stringer et al. 2011) and without a buffer between plots, it is very likely that the same colony will be found on multiple plots within a single site. The SAP commented elsewhere that 200 feet may be an appropriate buffer between plots (see Field Studies for Testing Area-Applied Pesticide Products (Section (i)); however, the Agency recognizes 200 ft (61 m) is excessive when the literature supports 56-66 ft between treatment plots.
No changes were made to the guideline.
Data Analysis (Section (d))
Public commenters were concerned with the complexity and specificity of the statistical analyses recommended in the guideline; specifically, they asked that Abbott's formula be used instead of the analyses proposed and that a more generic recommendation (e.g. "statistical models, when scientifically valid...") be used to allow for more flexibility. Commenters suggested to include how to calculate a power vs. sample size analysis and if that calculation is required per study. In addition, one commenter requested that product efficacy thresholds needed to obtain a claim of kills or controls be added to the guideline.
EPA Response: In the guideline, the Agency provided statistical analysis recommendations. Other methods of analysis may be applicable. A power versus sample size analysis is recommended per study design. A statistician may be consulted for sound design and valid statistical analysis prior to testing. No changes were made to this section of the guideline.  Product Performance Test Guidelines indicate how to test a product's performance for efficacy against target pests; specific thresholds for efficacy are included in the upcoming product performance rule.
GLP and QA Oversight (Sections (d) and (f))
The SAP and public commenters were concerned about the feasibility of conducting the recommended studies, particularly field studies, under GLP, and public commenters specifically mentioned the difficulty of the GLP QA requirement.
EPA Response: No changes were made to the GLP sections of the guideline as section (b)b already indicates that studies that "do not comply with GLP standards may nonetheless be considered if, in the agency's judgment, the design and conduct of the study provide results that are scientifically reliable." This section also indicates that with the submission of a study an applicant must include "[a] statement describing in detail all differences between the practices used in the study and those required by this part."
Reporting of Completed Efficacy Studies (Section (g))
Regarding reporting recommendations (section (g)), two public commenters requested clarification on the submission of raw data, specifically indicating that data is often not recorded on paper anymore, asking if certified copies of the raw data are acceptable instead of the originals, and wondering if summary tables that include data points and individual replicate data is what is being requested as opposed to summarized data that only shows averages for each group.  Another commenter disagreed with the endpoint/reporting (i.e., number of foragers) for area-applied field studies, unless the product is a barrier or repellant, and did not offer an alternative endpoint.
The SAP recommended language which indicates that "should" be reported as opposed to `may' be reported where applicable. 
EPA Response: The guideline specifies to "include legible copies of all raw data." Therefore, "certified copies" are acceptable; the guideline does not request originals. Summary tables are a suitable presentation of data, but should be included in addition to, not a substitute for, raw data. 
EPA thinks the assessment of foragers is an appropriate indicator of product performance when treatment is applied to an area. Mounds (i.e., mound-counts), another possible endpoint, are not always observable in heavily disturbed environments or environments with dense vegetation; therefore, foraging activity is a better indicator of product efficacy of area-applied products. See Field Studies for Testing Area-Applied Pesticide Products (i) for further comment on this topic.
Barriers/repellents were removed from this guideline. 
The Agency agreed with the SAP about identifying the information that should (versus may) be reported and reviewed the text for clarity.
Site Selection for Field Tests (Sections (i) and (j))
All public and SAP comments related to the testing of monogyne and polygyne populations (charge question (2)a) were not in support of using these designations to choose field sites, for the following reasons:
 determining social form is difficult and often inaccurate in the field and requiring genotyping (or even determining worker head size) would add unneeded complexity, time, and cost (public and SAP); 
 previous efficacy testing has shown no difference between these types of colonies (public); 
 the colonies are often intermixed in the field (public); 
 composition of colony types may change over time in a long-term study (public); 
 requiring monogyne and polygyne plots to be the same size is not practical because they have different spatial density patterns, although they reasonably could have the same number of mounds (public); 
 finding suitable test sites is already challenging without this added constraint (public);
 the biology of monogyne and polygyne forms of Solenopsis invicta are identical, and biological factors are extremely close among all of the Solenopsis saevissima complex (includes two fire ants in the United States, S. invicta and S. richteri, and several South American species); the only difference between the social forms of S. invicta is the aggressive interactions between monogyne colonies which would have no impact on field tests because either form is still able to forage effectively whether in a field of monogyne or polygyne colonies (SAP); and
 with the exception of Texas and Florida, polygyne sites are small, ephemeral, and usually not large enough to initiate a field trial, especially with baits (SAP).
One commenter suggested that the testing of both colony types should only be done to support claims such as "kills the queen(s)" or "kills monogyne and/or polygyne colonies." Another commenter acknowledged the importance of having geographically disparate sites, specifically because distribution of the toxicant within the mound structure and persistence of area-applied contact insecticides may differ between soil types, as could the organic material found within the soil which can bind insecticides differently. While the commenter was unsure if there are biological reasons for having more than two sites, they indicated that the season and ambient conditions are very important, particularly for baits since they may perform well one season and not in another, even when exposed to similar temperatures; therefore they suggested it is probably more relevant to ensure an adequate number of seasons are included in the design versus more than two sites as long as the two sites are indeed geographically disparate with varying soil types. In contrast, the SAP did not see a need to conduct tests at geographically disparate sites because the biology and behavior of these ants do not change in different geographical settings; even if the intent is to prevent pseudo-replications, there is still no need to require different geographical sites that include different ecotypes.
Finally, a commenter was concerned about the recommendation of 2 sites with at least 4 plots per site for the field studies because some areas may have inadequate turf quality, difficulty in maintaining plot markers close proximity to high traffic areas or to water bodies, making areas not usable for research; the commenter suggested either reducing the number of sites to 1 when the site is heterogenous or reducing the number of plots per site as needed (e.g. 2 replicates at site A and 3 replicates at site B).
EPA Response: Based on SAP recommendations, suggested guidance to include both monogyne and polygyne populations in testing were removed. Likewise, specific claims about social form will not be supported. In addition, geographically different locations were not needed from a biological perspective and were not included. There was no reduction in the number of plots recommended for testing due to the anticipated replication needed for the statistical model. Testing to support product performance against fire ant colonies, including specific life stages/castes, should be supported by both acceptable field and laboratory data.  
Testing IGRs in the Field (Sections (i) and (j))
Regarding testing IGRs in the field (charge question (2)d), public commenters were concerned that sampling in the field would destroy the nests/be disruptive for the fire ants, rendering them unusable for testing, and would not be supported by landlords of highly maintained areas. One commenter thought a final sampling at the end of the study would be more appropriate. The SAP agreed that there is no need to remove a small portion of the mound because opening a mound with a probe or narrow shovel to observe ant activity is less disturbing and will suffice to determine whether brood is present. They also mentioned that unless speed of activity is being assessed, brood does not need to be assessed at each data collection assuming the study is conducted for a long enough sampling period because the purpose of the study is to determine if the product kills the mound, not to determine the fate of the brood.
Specific to charge question (2)d.i., while some SAP members believed the presence of brood should be confirmed in 100% of the mounds prior to application (the inclusion of up to 10% mounds that did not have brood would be problematic for the assessment of the test by increasing variability and decreasing confidence of the test), overall the SAP agreed that it is reasonable to assume that all mounds large enough to be identified at the beginning of a trial will be mature colonies with brood because fire ants maintain brood throughout the year when a colony is mature. The assumption that all observable mounds have brood is preferred to substantial mound disturbance because even with the disturbance brood might not be apparent because they could be deep within the nest (it would require near complete nest destruction to confirm their presence). In addition, looking for brood increases cost due to the time needed for assessment (assessments could only be done in the cooler parts of the day). Because presence of brood is affected by season and environment and will vary during the year, the test should be run longer than the suggested 60 days. In the field, the determination of efficacy should be based on whether the mound dies, not the absence of brood; the effect on brood should be determined in the laboratory.
A public commenter indicated that if a pre-count is performed, it should not matter if any of the mounds have visible brood as the only impact would be how fast one observes results. The commenter assumed that the question was assuming that the "INDEX" method should be used for evaluation. However, they believe that the consumer is not concerned if a fire ant colony's brood disappears in 30 days; instead they just want the worker/forager ants eliminated. As colony mortality using an IGR can take 3-6 months and is dependent on the adult mortality (affected most by heat and drought), a fall-treated colony could still be a nuisance until the early summer, when the workers finally die. Thus, the commenter believed that brood presence/absence at the beginning of the study may be irrelevant and therefore suggested that the preference could be left up to the study director and how quickly they care about seeing brood reduction/effect of IGR. 
However, in response to charge question (2)d.ii., the public commenter indicated that 60 days is probably enough to see the disappearance of brood but may not be long enough to demonstrate complete colony mortality. Based on the mode of action of the IGR (to block reproduction in the adult and inhibit larval growth and metamorphosis), the SAP indicated that a minimum of 120-days for post treatment assessment should be sufficient to determine the action of the IGR to give time to assure inhibition of reproduction of adults that eclose from pupae that were formed following treatment (the treatment would not stop the development of pupae themselves which do not feed). In 60 days, an IGR could subdue production or development of brood, but if the reproductive ability of the colony is not permanently eliminated, the colony could bounce back. Note that the SAP mentions that 180 days might also be appropriate.
EPA Response: For IGR field studies, because there was no consensus among the SAP as to the need for confirmation of brood in mounds versus assumption of brood in mounds based on mound size (i.e., no visual confirmation), the guidance, as written, did not change. While colony mortality may be an endpoint, it is not typically the endpoint of an IGR-effect claim. Developmental effects of the IGR will be documented from laboratory tests; absence of brood in the field (post treatment) will confirm IGR effect on colony replacement. This reinforces the need for brood in mounds prior to treatment applications, as some SAP members agreed. Also, the Agency tries to avoid using assumptions (i.e., assuming brood is present) when making an efficacy determination and instead prefers to rely on count data or in this case, a visual assessment of brood presence. Confirming 100% brood presence at the start of a study and a 60- day study duration helps to address potential effects of season and environment, a concern of the SAP.  The guideline maintained the study duration at 60 days (unless longer duration is needed for specific claims, such as, colony mortality or 6-month control). 
The Agency agreed with the SAP that the brood does not need to be sampled at each data collection point. The study analyses for field tests of IGRs are based on a pre- and post-assessment of brood presence to minimize disturbance. The guideline was revised for clarification to indicate a single pre- and post-assessment. In addition, to further minimize disruption, the text was revised to indicate that a small portion of the mound should be turned over to sample instead of removing a small portion of the mound. 
Field Studies in General
Regarding the field study designs in general, instead of scratching the surface of the mound with forceps or a similar tool, which is a difficult sampling method for a large field trial, the SAP recommended a probe, up to 3/8 inches in diameter, be used to elicit a response without significantly disturbing the mound. During hotter weather months, the probe may need to be inserted 6-8 inches in depth because fire ants will locate deeper into the soil. Mounds are routinely disturbed in trials by mowers, cattle, vehicles, etc. They noted that disturbance from sampling mounds is much less intrusive than disturbance by mowers, cattle, and vehicles, and it works well to determine activity.
EPA Response: While scratching the mound surface remained in the guideline as the Agency believed it to be a quicker and less intrusive method, the probing method was added to the guideline as an alternative to scratching.
Field Studies for Testing Area-Applied Pesticide Products (Section (i))
Regarding the field test for area-applied products (section (i)), one public commenter was concerned that it would be difficult to find a field with no pesticide application history within the last year. Another indicated that sampling within the proposed temperature range may be prohibitive for some locations depending on the time of year and requested that a wider range be allowed. Numerous commenters were also concerned that disturbing the mound would render it unusable for the study and that brood should only be measured in laboratory tests or only at the end of the study, although one commenter indicated that "presence of brood should be assessed when mound temperature is no greater than 89.6⁰F (Porter 1988). A fist-sized portion of the mound is excessive. Less than a fist-sized amount should be removed from the mound." 
In relation to sampling, one commenter recommended that pre-treatment sampling be allowed from 0-10 days before the study begins because there is no reason to wait another day to start the application, the extra days allow for unexpected weather conditions, and it would burden the researcher to make an additional trip to the site. The commenter also believed that the first sampling should occur when the researcher or label claims control rather than an arbitrary 15 days and therefore the language should be changed to "Sampling should be conducted over a minimum of 60 days with a minimum of four sample collections." 
Commenters also indicated that forager numbers are only necessary for barrier treatments and that mound counts may be unneeded because they're unreliable due to mowing, rainfall effects, and potential lack of mounds in RIFA-infested areas. Other commenters were concerned about the feasibility and cost-prohibitiveness of having equal numbers of control and treatment plots and having identical plot sizes at both sites.
The SAP had a number of general suggestions about section (i), summarized below:
 Separately address the two categories of area applied products, broadcast baits and broadcast residual soil treatments
 Remove suggestion to wait one year after last pesticide application (if fire ants are present and active, there is likely no residual pesticide on the site)
 Sample at intervals of 1, 2, 3, 4 and 6 weeks and monthly thereafter, with earlier sampling performed if specific labeling claims are proposed
 Do not suggest the reporting of inactive mounds
The SAP was concerned about using food attractants for assessment because fire ants recruit foragers to discovered food resources and it is not uncommon to find hundreds of fire ants recruited to one vial, while another nearby vial equidistant from a mound has no or few fire ants. In addition, recruitment to food sources is highly dependent on time of day and being shielded from the sun. The SAP suggested considering effective methods for evaluating fire ant bait products other than foraging, including mound counts and ratings (Drees et al, 2013). These methods, while not perfect, have been used successfully to test fire ant baits and individual mound treatments for many years (Collins et al, 1999). Mound data counts have been the standard for USDA-APHIS protocols when testing products for use in the fire ant quarantine program (Loftin et al, 2016 and 2017; and Callcott, 2014). The SAP indicated that the number of active mounds per plot may be used as a measure of treatment efficacy to supplement forager data and to establish homogeneity among plots at a site. Randomization of field trials should be based on number of mounds, arranging plots according to numbers of mounds and then blocking from highest to lowest numbers (Snedecor and Cochran, 1989). 
Most of the SAP members indicated that using active mound counts alone is sufficient to determine efficacy of a product. However, if foragers are used as a sampling tool, the SAP recommended that vials should be shaded while in the field to prevent reaching internal temperatures that limit foraging and that soil temperatures should be monitored to ensure fire ants are foraging. During the hot summer months, the window for sampling by foraging is usually from sunrise until 10:00 a.m. and from 4:00 p.m. until sunset. In addition, to assure adequate access to the vials, the SAP recommended that vials be placed in contact with the soil and not placed on top of grass. While Drees et al (2013) recommend a 45 to 60-minute timeframe for baiting foragers, baits have been removed and/or eaten by fire ants in trials in Alabama when left longer than 30 minutes, and therefore a sampling period of 30 minutes should be long enough for fire ants to recruit to the bait and not consume or remove the bait. If a longer period is used, one must ensure that enough bait is placed at each station. One SAP member also noted that if a Poisson distribution has not been established as the proper underlying distribution for forager recruitment to hot dog vials, then the presence/absence of fire ants in vials might be a better measure of ant foraging. Several studies have shown that using foraging numbers for a trial is an acceptable method of assessing imported fire ant populations (Jones et al, 1998).
For charge question (2)b.i., the SAP thought that a 60-day duration may not be sufficient in some cases, but the length should depend upon the required endpoint. If colony elimination or complete control is the endpoint, then 60 days may not be sufficient for slow acting baits and IGRs, but if colony reduction for management is the desired endpoint, then 60 days may be sufficient except for IGRs. Specifically, for granular and liquid contact application products, the SAP recommended a minimum of 60 days although 90 days is preferred; for bait products, they recommended 4-6 months with 6 months being the minimum for IGRs. They also noted that products such as fipronil have taken 30-45 days to show reductions in mound numbers and that a sampling period of 30 days might be appropriate for a quick kill product. 
One public commenter noted that it is not reasonable to expect 60- day efficacy for a "kills on contact" claim, due to potential new colony reinfestation post-treatment and that 60-day efficacy should only be required for a product with residual claims. Another commenter thought that a 60-day duration is not a good minimum time frame for area-applied products (depending on the product type) because "1. Consumers will likely want > 60-day control. 2. There is a critical difference between baits and contact insecticides: baits have no residual activity. Once a product has killed as many colonies as it is going to kill, any change in mound numbers is due to reinvasion and seasonal changes within the plot.  Insect Growth Regulator baits need far longer than 60 days to reach full control. Conversely, fast-acting bait test timing should be based on the product's activity. 3. Area-applied contact insecticides DO have true residual activity and it is a critical measurement, but it varies tremendously with temperature, soil type, organic matter content, and reinvasion pressure. Test duration should be based on when the product "breaks" and significant reinvasion can be determined since colonies may go deep below the toxicant layer and emerge when the layer "breaks". They may live for months deep below the soil in this manner. Measuring bait-treated plots over time is primarily measuring the re-invasion pressure." 
Regarding charge question (2)b.ii., the SAP suggested using a minimum of 20 mounds per acre (Drees et al (2013)). Because foraging can occur at a distance of up to 200 feet, small plots are not ideal and a 200-foot barrier should be in between test plots (note that elsewhere the SAP contradicted this buffer recommendation and mentioned that if plots are 0.5 acres, the edges of the center circle should be a minimum of 50 feet apart to prevent fire ants within the circle from foraging into adjacent plots; this is similar to data collection in row crops like cotton, where plots are treated and data are collected from the middle rows beginning at least 10 feet into the plot to reduce edge effects (Graham et al, 1984 and 1987; Darnell et al, 2016)). If areas smaller than one acre are used, the standard of 20 mounds per acre could be extrapolated downward but the number should never be less than would give statistical power to the analysis of the test. For example, in urban areas, the SAP recommended at least 5 mounds per plot, but deferred to Agency statisticians to determine how many mounds would make each plot a significant test. Related to charge question (2)b.ii., but not directly in response to it, the SAP recommended that plots should all be the same size; contradictory to their recommendation above, they also said that a minimum number of 10 mounds per plot should be adequate for an efficacy trial. For granular and liquid contact application products, they recommended a minimum plot size of 0.25 acre with at least 10 mounds included in a central area of each plot. For bait products, they recommended a minimum plot size of 0.5 acres (0.75 to 1 acre is preferable). They noted that in most southern states, a 0.25-acre circle (ca. 58' radius) is usually required to obtain enough plots for a test that has 10 or more mounds in the circle. 
One public commenter recommended that 10 mounds per plot, regardless of plot size, would be reasonable. As mound numbers can decrease during summer, lower numbers could result in the untreated plot having no mounds, which would impact the statistics and may not reflect true treatment efficacy. To determine the appropriate number of mounds per plot, the commenter recommended "consistent size/variable mound number or consistent mound number/variable plot size. The former works best for area applications and the latter for individual mound treatments. 1. A fixed area allows the same quantity of material to be applied to every plot (within treatment) regardless of mound number. It also matches real-world scenario of area treatments. 2. A fixed number of mounds are better for individual mound treatments because you can determine the amount of material you need. This also fits the real-world scenario of homeowner treatments better." 
Specific to charge question (2)b.iii., the SAP thought that the placement of baits at 90% of the radius was problematic due to the foraging ability of fire ants. Reducing the sampling area would help to eliminate false negative results due to untreated foragers from outside the test plot being collected at food lures within the test plot. They suggested the maximum placement of baits to be no more than 50% of the radius of the test plot. Note that a single SAP member suggested that because of the foraging dynamics of this species, baits might not be a useful tool in documenting effects on fire ants within the test plot. 
One public commenter noted that while 9 or more vials may be reasonable in large plots, in small plots only 1-2 vials may suffice as counting fire ants from lures is very time-consuming. Another commenter asked that section be clarified as the language is currently confusing. They believed that the "most important distance to be concerned about would be that from the "treated edge" to the vial/lure/hot dog slice which is dependent upon foraging distance. This may range anywhere from 1 m to >50 m. 12 m from a bait-treated edge is usually enough to isolate the lure from fire ants foraging in from the outside. 12-15 m from the treated edge should be sufficient. Additionally, circular plots are assumed here, but rectangular plots also work well when designed and sampled properly and have advantages over square plots with circular sampling areas."
EPA Response: In response to concerns about test sites being pesticide free for one year prior to treatment, after further consideration, no change was made to the guideline due to currently marketed registered products having residual activity for up to one year.  Because the temperature range (i.e., between 18.3° C and 35° C or 65° F and 95° F) for collecting foraging data provided in the guideline was supported by numerous references and SAP members, the range was not expanded. It was recommended that the timing of field studies involving foragers take into account this temperature range.  Disturbance of mounds during pre- and post-treatment assessments is minimal. Scratching the mound surface remained in the guideline as the Agency believed it to be a quicker and less intrusive method than some other sampling methods. Also, for studies involving IGRs, the Agency agreed with the SAP that the brood does not need to be sampled at each data collection point, and sampling brood at each data collection time point was not recommended in the draft guideline. The study analyses for field tests of IGRs are based on a single pre- and post-assessment of brood presence per mound to minimize disturbance with the post-assessment occurring close to the last day of the study. In addition, to further minimize disruption, the text was revised to indicate that a small portion of the mound should be turned over to sample instead of removing a small portion of the mound. 
The guideline recommends that pre-treatment assessments be conducted 1-7 days prior to treatment application. One commenter proposed expanding pretreatment sampling to 0-10 days. While the Agency agreed that pre-treatment sampling should occur as close as possible to treatment application, we did not agree that it should occur on the same day unless disturbance to the area or mound is required for product application. Disturbing the mound immediately prior to applying the treatment (e.g., bait), may cause the fire ants to associate the treatment with the disturbance and potentially decrease foraging. Also, when conducting pretreatment sampling further in time from the treatment application there is the risk that the number of active mounds will change, sometimes dramatically depending on the time of year (e.g., springtime, after mating has occurred). Therefore, no change was made to the guideline.
The SAP agreed with the guideline that plots should contain a minimum of 20 active mounds per acre.  There was also agreement (Agency, SAP, public) to add that each plot should contain no less than 10 active mounds, regardless of plot size. This was added to the guideline. The Agency concurred with the public comment that a fixed area for all plots within a study ensures the same quantity of material to be applied to every plot. For this reason, the guideline includes equal plot sizes within a study.
The guideline provided recommendations for a minimum of four sample collections during the study duration. One commenter suggested the language for sampling intervals should be changed to, "Sampling should be conducted over a minimum of 60 days with a minimum of four sample collections." The Agency agreed; the wording was revised and added, "with the final sample occurring on the last day (i.e., day 60) or soon thereafter if all samples cannot be collected on the last day."  The Agency concurred with the SAP recommendation that earlier sampling may be performed specific to labeling claims, provided the claim pertains to the colony.  

There were several valid comments provided about duration of studies, ranging from very short, "quick kill" durations to 180 days depending on the type of application, claims, and mode of action. While "kills on contact" claims would not be suitable for tests in this guideline, which focus on colony kill, the Agency agreed that data collection intervals can be selected to support claims, provided the claims pertain to the colony. No change was made to the guideline since it already included text for sampling based on the pesticide's mode of action and desired label claims. Because the guideline included provisions for development of protocols for study designs that differ from those provided, the test durations for area-applied (60 days), mound-applied (30 days) and IGR field tests (60 days) remained; justification of other times should be provided.

Some commenters did not want to assess foragers or mounds but did not propose an alternative indicator of product performance. No change was made to the guideline since the use of mound counts and foragers as indicators of efficacy in fire ant control studies were supported by numerous published studies.  There was also concern about the need for a balanced test design. However, a balanced design is most robust for this type of study. The option to consult a statistician when developing test protocols was provided in the guideline. 

SAP members thought that the guideline should have two categories of area applied products, one for broadcast baits and one for broadcast soil treatment; however, the Agency did not see the need for this distinction since the way product performance is determined would be the same. The SAP did not think that inactive mounds should be reported. The Agency points out that inactive mounds are only important if the registrant chooses to generate data for the optional claim "no new mounds."  Data would be needed for both the number of active and inactive mounds per plot at each sampling interval.

In the SAP discussion about using a food attractant for measurement of forager activity, SAP members suggested using mound counts in conjunction with a rating scale to measure treatment efficacy; they also suggested that the number of active mounds per plot may be used to supplement forager data and to establish homogeneity among plots at a site. Randomization of field trials should be based on number of mounds, arranging plots according to numbers of mounds, and then blocking from highest to lowest numbers (Snedecor and Cochran, 1989). The Agency agreed. The guideline included using mound counts to determine homogeneity among plots during set up of the study, and the reference to randomization of plots was added. The Agency did not see the need for both mound and forager assessments to determine product efficacy of area-applied treatments. In addition, using an index rating scale was advised against, in part due to too many variables being contained within the index. For example, the USDA index for rating mounds provides a single index per mound based on a 2 x 5 categorical rating that includes, presence/absence of brood, size of colony, and mound size. The SAP was concerned that there may be unequal recruitment to vials within a plot and the need for proper application of this sampling method. However, the Agency did not share the concern about recruitment to individual vials because the vial counts are totaled per plot. The Agency thought that foraging activity was a better indicator of product efficacy of area-applied products, particularly when activity within mounds can be influenced by season/weather conditions and time of day that sampling occurs. Foragers from colonies on the plot would still be active even if they were not visible in the mound during sampling. Further, mounds are not always observable in heavily disturbed environments (mowed/grazed) or environments with dense vegetation.  The SAP noted that if attractant food vials were used that they should be placed in contact with the soil and shaded; the Agency revised the guideline to include both along with an example of shading. The guideline includes a 30-60-minute placement window for capture of foragers in vials, which was revised to state "approximately 30-60-minute placement window." The SAP agreed that 30 minutes is long enough for fire ants to recruit to the bait but cautioned about running out of bait if left for longer.

One SAP member suggested a presence/absence determination of fire ants in vials as an alternative to counts if a Poisson distribution has not been established as the underlying distribution for forager recruitment. The guideline recommends that registrants consult with a statistician. A Poisson distribution is most likely appropriate given these are count data. Registrants are encouraged to conduct an evaluation to determine the appropriate statistical methods for the distribution of their data. 

Plot sizes and shapes are to be determined by the registrant; circular or square does not matter as long as all plots within a study are the same and the minimum number of active mounds per plot is met. The guideline was revised to include that sampling of foragers in area-applied field tests may be conducted at 50% of the plot area. One commenter believed that 1-2 attractant vials would be enough sampling per plot; however, the agency disagreed due to the variance that can occur in recruitment to vials.

The following reference was added: Snedecor, G.W. and W. G. Cochran. 1989. Statistical Methods, 8[th] ed. Iowa State University Press. Ames, IA.
  
Field Studies for Testing Mound-Applied Pesticide Products (Section (j))
In response to charge (2)c.i. related to the mound-applied field test (section (j)), the SAP noted that no reason is given for "field tests" (presumably those for area-applied products) having a 60-day test duration and mound treatments having a 30-day test duration. They believed the timing of these tests depend upon the desired endpoint as well as the nature of the active ingredient. If the endpoint is colony elimination or complete control, then a 30-day duration is probably not long enough. The duration for this endpoint would likely need to be more than even 60 days. If the endpoint is to reach at least 90% control, which would suggest that the colonies are declining towards elimination, then a 30-day test duration would be sufficient unless a slow-active ingredient or IGR are being used. One public commenter suggested that mound-applied products should be tested for 30 days or as labeled, while another commenter was concerned that 30 days is too long for a mound-applied contact insecticide test because once the colony or mound is dead it will remain this way and most contact insecticides reach their full effect within hours. Any new mounds formed would be reinvading colonies or colonies that were not completely killed by the treatment that relocated, although it would be difficult to determine this unless the fire ants are observed moving. However, it is possible to infer this by the observation of new mounds appearing within the treated plots as compared to the number of new mounds appearing in the untreated plots, assuming comparable plot sizes between the treatment and control plots. 
For mound-applied IGRs, in relation to charge question (2)d.i., one commenter noted that it would be impossible to avoid disturbing the mound to measure brood presence in the field and that this should only be done in the laboratory.
Other comments related to section (j) include the following:
 Public commenters were concerned with the number of mounds per treatment and equal number of treated and untreated plots as recommended in the guideline. 
 The SAP stated that the definition of a plot wasn't clear and that while plot size might vary, the number of mounds within each replicate (at least 10) should be the same.
 The SAP noted that soil temperature recommendations were listed for collecting foraging data, but not for determining mound activity. They agreed that the air temperatures recommended in Drees et al (2013) for collecting foraging data (between 18.3° C and 35° C or 65° F and 95° F) should be adequate for collecting mound data and recommended that the temperatures be taken at 1" (2 cm) (Porter, 1987).
 The SAP agreed that if a product claims quick kill, sampling data should be taken earlier than seven days and data collection times should depend on the claim.
The SAP cautioned that an individual mound treatment product label should not contain claims about preventing new mounds from forming. The number of all active mounds per plot per treatment within a site should be reported for each assessment to account for possible colony displacement as opposed to colony elimination since more than one satellite mound may form from a colony that was not completely killed if the queen escaped (these will usually reform to one mound with time). They also noted that reporting the number of inactive mounds is not necessary.
EPA Response: In response to the SAP comment that no reason was given for area-applied products having a 60-day test duration and mound treatments having a 30-day test duration, the Agency added the following reason to the guideline for clarity: "Recommendations for field study test durations differ based on product application (i.e., mound vs area-applied) and whether an IGR is included in the product formulation. Field study duration for a mound applied product, where the colony within the mound presumably receives the full labeled rate, typically follows a >= 30-day duration. Area-applied products typically follow a >= 60-day duration because an application is used to treat all colonies within an area. Colonies within the treated area may collect different amounts of the applied product due to variability in sizes of the colonies and their ability to monopolize foraging territory. Therefore, the treatment effect may take longer to be apparent and/or may dissipate quicker (i.e., shorter residual effect) in colonies that did not acquire a lethal dose. Likewise, for products containing an IGR, a >= 60-day duration is recommended to observe an effect unless a longer duration is needed for specific claims, such as colony mortality or a 6-month control claim." The recommended test durations in the guideline  were not changed. Tracking of new mounds that form during a mound-applied field study is optional to support desired claims.
For mound-applied IGRs, one commenter noted that it would be impossible to avoid disturbing the mound to measure brood presence in the field and that this should only be done in the laboratory. The guideline addresses this comment by providing instructions for minimal disturbance of the mounds and only at two times (i.e., pretreatment and at final posttreatment assessment). Because IGRs have been successfully tested in the field for macro-level (i.e., qualitative assessment of brood) evidence of product performance, the impact on brood should be assessed in both field and laboratory tests. 
The Agency agreed with the SAP that a plot should contain >= 10 active mounds and for statistical purposes recommended a balanced design. The SAP also agreed that plot size may vary in the mound-applied study design. The section of the guideline for mound-applied study design ((j) b ii) clearly specifies these parameters; however, for clarity, the Agency defined plot in the mound applied section as, "a plot is simply the area containing at least 10 active mounds. Treatment is applied to each active mound within a plot." However, it should be noted that this would not be the definition of plot in the area applied section because the experimental unit in the mound-applied design is a collection of mounds per plot to which the treatment is applied, but in the area-applied design the experimental unit is the plot that receives the treatment.  

The SAP noted that soil temperature recommendations were listed for collecting foraging data, but not for determining mound activity. The SAP agreed that the air temperatures for collecting foraging data (between 18.3° C and 35° C or 65° F and 95° F) should be adequate for collecting mound data and recommended that the temperatures be taken at 1" (2 cm) (Porter, 1987). The guideline already included reference to the temperature range; the reference for depth of measurement was added.

The Agency concurred with the SAP recommendation that if a product claims quick kill, sampling data should be taken earlier than seven days and data collection times should depend on the claim, provided the claim pertains to the colony. However, the study duration for mound applied treatment would still be a minimum of 30 days; only the frequency of assessments may change. No change was made to the guideline since it already includes text for sampling based on the pesticide's mode of action and desired label claims. 

Laboratory Studies in General (Sections (k), (l), and (m))
For all the laboratory tests, the SAP recommended that bait and the alternate food source be provided ad libitum for the specified exposure time for the trial, instead of applying the lowest labeled rate as currently recommended; the latter is unrealistic because in nature fire ants forage over a large area to collect the food that is needed for the colony until the colony is satiated. The SAP also cautioned against invalidating a trial if one queen out of multiple queens die in a control because queens are occasionally eliminated by the colony and natural mortality is possible. With other queens surviving in the test, the death of one queen is not critical as long as the mortality rate of the entire test colony remains acceptable.
One public commenter believed that laboratory tests should only be needed to provide data for IGRs or for claims such as "kills the queen/brood," and that conducting laboratory tests for repellents could give unreliable results. Another public commenter requested that methodology for compounds that attract fire ants be added to the guideline. This commenter also wondered "does the room need to be maintained at this humidity level or microclimate as is typically done with dental stone or caster nest cells, with room RH at 30-50%." They indicated that during periods of exposure, food and water are not necessary at all times (e.g. if  fire ants are being exposed to treated surfaces for short periods of time). Finally, they recommended that requiring the recording and reporting of relative humidity, photoperiod, temperature, and access to food and water is better than mandating all studies have the same requirements.
Two public commenters did not support the use of two connected foraging areas for laboratory studies because the setup is space limiting and foraging in the two chambers would likely be uneven but necessarily related to the bait; they believed that a single arena with access to both the bait and an alternative food source would be most appropriate and had been used in the past successfully. Another commenter indicated there were more simplistic designs in scientific literature not cited in the guideline. 
Similarly, the SAP responded to charge question (3)b about length of Tygon(R) tubing by questioning whether two arenas connected by Tygon(R) tubing would be appropriate at all since it doesn't represent a natural setting because fire ants can leave a nest and begin foraging without crossing a barrier. Instead they recommended using a single arena with a nest structure with proper nest humidity, where baits would be placed equidistant from the nest. 
In terms of arena size (charge question (3)a), the SAP concluded that specifying a test arena size may not be needed as long as the arena provides adequate space for the colony size being used in the test; for 100 worker fire ants, a 12" x 15" plastic box would be sufficient.
Regarding charge question (3)c, the SAP concurred that the acclimation times provided in the Agency document would be suitable since colonies appear able to organize in this amount of time although one public commenter did not believe that a 24-hour acclimation period from field to lab was sufficient. In response to charge questions (3)e/f, the SAP believed that 100 workers can support brood and a queen for a two-week laboratory test. They noted that using 10,000 workers is excessive and that less than 200 fire ants can maintain a brood for an IGR test; overall they recommended 200-1000 fire ants for IGR studies.
EPA Response: The Agency agreed with the SAP's comments that testing at the lowest labelled rate is unrealistic for laboratory bait studies. In concert with the SAP's recommendations, section (k)(b)(iii) of the guidelines was amended to remove that the bait should be applied at the lowest labeled rate.  Additionally, reporting of "re-baiting" in laboratory studies was also removed from the guideline.
The Agency disagreed that a colony should not be invalidated if the workers outlive their queen. Queen health is a more appropriate indication of colony health. Numerous workers may die in controls, but a queen death negates worker replacement and therefore ensures colony failure. Because previous SAP comments indicated that the social form of fire ants (monogyny or polygyny) was not important for testing, laboratory colonies should be limited to one egg-laying queen. The guideline was updated to remove multiple queens in laboratory studies.
The Agency also disagreed with the public commenter that laboratory studies should be limited to products containing IGRs. Laboratory studies validate specific effects (e.g., mortality) on workers, brood and queen within colonies not visible nor feasible to measure from field research. Field studies compliment laboratory studies by providing macro-scale data on product performance when applied in accordance with the label DFU and at the lowest labeled rate. 
Regarding the testing of fire ant attractants, the Agency encourages registrants to submit protocols for novel testing methods or product types that are not included in the proposed fire ant guidelines.  
A public comment indicated that access to food and water during periods of exposure is not always necessary when conducting studies of fire ants on treated surfaces, and that these studies would only affect the workers/foragers and not directly affect the colony. While not pertinent to this guideline, guidance for residual studies on treated surfaces is provided within the Premises guideline (810.3500). 
Regarding environmental conditions, sections (k) and (l) of the guideline were amended to include the following for laboratory tests: "Laboratory testing may be conducted at ambient temperature and RH in the testing arena as long as conditions are optimal for foraging behavior, such as, 82° - 91° F (Banks et al. 1981). Conditions that are too cool or dry will inhibit normal colony activity. In addition, temperature and RH within the nest cell should also be maintained at conditions optimal for brood development, such as >= 65 % RH and 82° - 91° F (Williams 1989). Also, laboratory colonies should be provided time to acclimate to the photoperiod, as photoperiod is known to affect foraging behavior (Lei et al. 2019). Environmental conditions for laboratory testing, including temperature, humidity, and photoperiod during testing, should be supported by cited literature."    
Regarding the acclimation period for laboratory colonies, the SAP was inconsistent in their responses for a minimum time frame. The Agency agreed with their justification supporting a 72-hour acclimation period (below), "to minimize aberrations that may be due to handling and disruptions in their foraging behavior due to earlier handling".  Therefore, laboratory sections, (k-m), of the guideline were revised for consistency to the 72-hour acclimation period.  
The Agency recognized commenters' concerns that the multi-arena design with connecting tubes may be too complicated and/or take up an inordinate amount of laboratory space. The Agency therefore agreed to amend section (k), of the guideline to include a simpler design using a single plastic arena per test colony (e.g., 12" x 15") containing a nest structure plus bait/food placed equidistant and at opposite ends from the nest. Per the SAP recommendation, the Agency did not specify an arena size, but provided a 12" x 15" arena as an example of minimum arena size for testing of 100-worker colonies. However, section (m) for specific guidance for laboratory studies for testing IGR products required a larger nest arena size to account for testing of larger colonies. The guideline was revised to include the recommendation along with a 16" x 24" nest arena size as an example. In addition, the multi arena design remained unchanged for section (l). Registrants may submit a protocol for review to accommodate specific testing needs or novel application types that may include but are not limited to aerosol mound injections or mound drenches.  
The Agency also agreed with the SAP's recommendations that 100 workers would be suitable to support queen and brood for laboratory test designs within sections (k) and (l). This recommendation was already in the guideline.  For IGR laboratory studies, the SAP advised that 200-1000 workers would be suitable for those extended studies. Section (m)b.ii. of the guidelines was amended to reflect these recommendations. The Agency encourages registrants to submit protocols for novel testing arenas or bioassay designs that are not included in the finalized fire ant guidelines.
Laboratory Studies for Testing Bait Products (Section (k))
One public commenter believed that the experimental unit for the bait laboratory test (section (k)) was too complicated and that the fire ants and bait can be in the same arena, eliminating the need for tubing which could increase likelihood of escape and other complications. Another commenter had the following suggestions: 
 instead of the lab test, a field test using the forager number method could be used and could also account for palatability; 
 it is cost prohibitive to have an equal number of replicates for controls and treatments; 
 the environmental conditions are too restrictive; 
 tests should be conducted for a maximum of 14 days or as labeled; 
 mortality should be conducted at no more than 48 hours each though the duration of the study or as labeled; and
 removal of dead individuals isn't safe or practical.
The SAP had the following comments related to section (k)
 replace "via plastic tubing (e.g. Tygon(R) tubing)" with "via a suitable bridge (e.g., paper strips or Tygon(R) tubing);" and
 acclimation time for the fire ants placed into a test arena should be 72 hours, not 24 hours, to minimize aberrations that may be due to handling and disruptions in their foraging behavior due to earlier handling.
EPA Response: The Agency recognized commenters' concerns that the multi-arena design with connecting tubes may be too complicated and/or take up an inordinate amount of laboratory space. The Agency therefore amended section (k), of the guideline to include a simpler design using a single plastic arena per test colony (e.g., 12" x 15") containing a nest structure plus bait/food placed equidistant and at opposite ends from the nest. Per the SAP recommendation, the Agency did not specify an arena size, but provided a 12" x 15" arena as an example of minimum arena size for testing of 100-worker colonies. The registrant may submit protocols for novel bioassays that may include multiple arenas and suitable bridge designs that are not included in the finalized fire ant guidelines.  
Regarding the acclimation period for laboratory colonies, the Agency agreed that 72h is needed. Section (k).b.iv. was revised accordingly. 
Regarding environmental conditions, section (k)b.ii. of the guideline was amended to include the following for laboratory tests: "Laboratory testing may be conducted at ambient temperature and RH in the testing arena as long as conditions are optimal for foraging behavior, such as, 82° - 91° F (Banks et al. 1981). Conditions that are too cool or dry will inhibit normal colony activity. In addition, temperature and RH within the nest cell should also be maintained at conditions optimal for brood development, such as >= 65 % RH and 82° - 91° F (Williams 1989). Also, laboratory colonies should be provided time to acclimate to the photoperiod, as photoperiod is known to affect foraging behavior (Lei et al. 2019). Environmental conditions for laboratory testing, including temperature, humidity, and photoperiod during testing, should be supported by cited literature."    
Regarding experimental designs that balance equal numbers of treatments and controls, the Agency developed designs based upon power versus sample analyses as they apply to the fire ant's unique biology. Registrants are encouraged to consult their own statisticians and propose/justify appropriate statistical designs for their studies.
The Agency holds that duration of bait studies should remain 14 days with mortality assessments conducted at 48-hour intervals. Deviations from assessment intervals and study duration should be justified by the registrant in the report, and/or submitted as a protocol for Agency review. 
The Agency disagreed with a commenter regarding the exclusive use of field trials for fire ant bait studies. Laboratory studies compliment the field studies by confirming colony-specific endpoints (mortality of queens, workers, etc.) that would be reasonably burdensome to measure or observe in a field trial. 
Finally, the Agency did not understand the concern regarding the removal of dead fire ants during assessments. Dead fire ants are not aggressive. However, the dead fire ants do need to be removed in order to ensure survivorship calculations are accurate. Since fire ants' stinging ability is broadly known by researchers, it is incumbent upon the registrant and/or laboratory managers to determine proper PPE when working with fire ants. Recommendations for PPE are beyond the scope of the 810.3100 guidelines. 
Laboratory Studies for Testing Products Other Than Baits (Section (l))
One public commenter had the following recommendations for barrier or repellent laboratory tests (section (l)): 
 tests should be conducted for a maximum of 14 days or as labeled; 
 the proportion of food consumed in the treated areas should be used as an indirect measure of repellency/barrier efficacy; 
 removal of dead individuals may not be safe or practical; and 
 worker mortality as an endpoint is irrelevant for repellents as not all are contact poisons.
The SAP suggested, if there is volatile component, that there might need to be two test arenas, with some bridging between the arenas; the bridges or tubes should be long enough to minimize the effects of any volatiles from the treatment. They also recommended that the minimum width of the barrier should be specified and that it should be consistent with application outside of the lab. Specifically, in response to charge question 3(f), the SAP concluded that 100 workers per replicate is enough for a 2-week laboratory test whether only workers or a colony structure with brood is used for the test.
EPA Response: Upon further consideration, barrier/repellent products typically do not kill or impact the entire colony and were therefore beyond the scope of the 810.3100 guidelines. Barrier/repellant products should be addressed on a case by case basis per pest species and (when necessary) through a protocol review. The Agency amended the title of this section to "Specific guidance for laboratory studies for testing products other than baits" to accommodate products, such as soil amendments that are intended to kill the colony.
Section (l) b. v. of the guideline was not changed regarding the public commenter's statement that tests should conducted for a maximum of 14 days. 
Because tests for barrier/repellent products have been removed from this guideline (810.3100), the endpoints measuring and analyzing food consumption were also removed. Section (l) b. v. was amended to read as follows:  
"v. Data collection and endpoints: The total number of workers per replicate should be collected.  Mortality should be evaluated at intervals <=48 hours through the duration of the study. Dead individuals should be removed and counted. Controls should be assessed in the same manner as those receiving treatment. Following the final assessment, all nest arenas should be frozen to determine the number of surviving individuals and for calculation of the total number of workers per replicate."
In addition, Section (l) c. i. which discusses the analysis of food consumption from foraging arenas was removed from the guideline. 
The Agency disagreed with the public comment that the removal of dead fire ants may not be safe or practical. Dead individuals should be removed and counted.
The Agency acknowledges the SAP's concern regarding volatiles emanating from treatments and encouraged registrants to submit a protocol if volatile components compromise the proposed design in the guideline. 
The Agency disagreed with the SAP's comment that the guideline should define a minimum width for substrates (treated or untreated) contained within the foraging arenas. However, the guideline was revised for clarity to state that foraging arenas should contain a layer of substrate that completely covers the floor of the arena. The depth of the substrate layer should be consistent and equal in all foraging arenas.  The Agency did not specify a size for the foraging arenas in the guideline to allow for flexibility of the design for spatial constraints. To that end, the text for section (l) b. i. was revised to include, an option for an alternative nest- and foraging-arena design if space is a concern: 
      "An experimental unit consists of an individual nest arena containing fire ant workers, a queen, plus brood affixed to two foraging arenas via plastic tubing (e.g., Tygon(R) tubing) or suitable bridge (Choe & Rust 2008; Knight & Rust 1990). Arenas should consist of an open-top container where interior vertical surfaces are coated with a non-stick material to prevent escape of fire ants. The nest arena should contain water and harborage for colony. Food should be available in the nest arena during the acclimation period, then removed immediately prior to test initiation. One foraging arena should contain substrate (e.g., sand, soil) treated with the product and the other foraging arena should contain an equal amount of untreated substrate or substrate treated with diluent-only. All foraging arenas used in the study should have the same dimensions; the length of the bridges or tubing connecting nest arena to the foraging arenas should be equal. The food sources in the foraging arenas should be palatable, established in the literature, consistent across all trials, and should be available ad libitum to the fire ants through the duration of the study. Foraging arenas should contain equal amounts of food placed at the far end of each arena, so ants travel through the substrate. Testing begins when tubes or bridges from the foraging arenas are connected to the nest arena.
      Alternative nest- and foraging-arena designs may be considered if laboratory space is a concern. Arena design may consist of a larger nest arena (e.g., 24" L x 16" W) within which two separate foraging arenas (e.g., nursery pots 4" L x 4" W x 3.5" H) may be placed (Costa et al. 2005). The nest arena should consist of an open-top box where interior vertical surfaces are coated with a non-stick material to prevent escape of fire ants. The nest arena should contain water and harborage for the colony. Food should be available in the nest arenas during the acclimation period, then removed immediately prior to test initiation. The foraging arenas (e.g., nursery pots) should each be filled with treated or untreated substrate, each with equal amounts of food provided ad libitum and placed in the center of the substrate in order to ensure the fire ants cross over the substrate to access food. Foraging arenas should be placed at equal distances away from the harborage within the nest arena. Testing begins when the foraging arenas are placed into the nest arena."
Section (l) b. ii. of the guideline remains unchanged regarding the SAP's response to question 3(f) that 100 workers are sufficient for a 2-week test. 
The following references were added: 
Choe, D.H. and M.K. Rust. 2008. Horizontal transfer of insecticides in laboratory colonies of the Argentine ant (Hymenoptera: Formicidae). Journal of Economic Entomology 101:1397-1405.
Costa, H.S., L. Greenberg, J. Klotz and M.K. Rust. 2005. Response of Argentine ants and red imported fire ants to permethrin-impregnated plastic strips; Foraging rates, colonization of potted soil, and differential mortality. Journal of Economic Entomology 98:2089-2094.
Knight, R.L. and M.K. Rust. 1990. Repellency and efficacy of insecticides against foraging workers in laboratory colonies of Argentine ants (Hymenoptera: Formicidae). Journal of Economic Entomology 83:1402-1408.
Laboratory Studies for Testing IGR Products (Section (m))
Public commenters believed that the replication recommendation for the IGR laboratory test (section (m)) was not feasible. In addition, one commenter was concerned that the setup described was too complicated; they specifically mentioned that the size of the Petri dish should be similar in size to the colonoid used and that the reasoning for using transparent yellow acetate paper wasn't clear and instead researchers should be given additional options for filtering light such as red cellophane covers and paint. Two public commenters were concerned with the feasibility of using 10,000 worker fire ants per replicate for the study.
In response to charge question (3)d, the SAP believed that 30 days should be sufficient to document some effects on the brood structure of the colony, but not sufficient to show colony collapse or ultimate effects of the IGR since the IGR could stop reproduction only temporarily. They believed that duration is dependent on the desired endpoint that is desired for registration of a product; if that endpoint is colony elimination, then 30 days would not be enough and 120 days might be appropriate, but if the endpoint is showing that the IGR is having an effect on the colony structure, then 30 days would be sufficient.
EPA Response: The Agency acknowledges commenters' concerns about the complexity of the multi-arena design and feasibility of using 10,000 fire ants per replicate for IGR studies. The SAP expressed that 200-1000 workers plus brood and a queen would be suitable for extended IGR studies. The guideline was amended per the recommendation. Section (k) of the guideline was amended to simplify the design to a single plastic testing arena. However, when testing a bait with an IGR, the arena size should be larger (e.g., 24" x 16") to accommodate the larger colony size of 200-1000 workers, brood, and a queen. When testing a product other than a bait, as in section (l), combined with an IGR, a multi-arena design is still recommended.  
The use of an amber cellophane nest cover was cited from the research literature. Although the guideline does not exclude the use of red cellophane, the Agency asserts that paint is not a viable option because it would prevent direct observation of brood without disturbing the nest cell. Registrants may propose other suitable materials or designs to observe qualitative effects of IGRs on brood. No changes were made to the guideline.
The SAP concurred with a minimum 30-d test duration to confirm the effects of IGR as stated in section (m)b.iv. of the guideline. The complimentary field studies are intended to support colony kill as part of the overall product performance. As such, the laboratory endpoints include qualitative effects on brood since IGRs do not directly kill adult fire ants. In the event that a co-formulated product containing IGR and an adulticide kills the colony within 30 days and before IGR effects are visible, it would then be up to the Reviewer to determine if IGR claims are adequately supported. No changes were made to the guideline.





