Effect,of,indaziflam,on,microbial,activity,and,nitrogen,cycling,processes,in,an,orchard,soil

Amir M.GONZÁLEZ-DELGADO,Pierre-AndréJACINTHE and Manoj K.SHUKLA

1Department of Plant and Environmental Sciences,New MexicoState University,Las Cruces NM 88003(USA)

2Department of Earth Sciences,Indiana UniversityPurdue University,Indianapolis IN 46202(USA)

ABSTRACT Indaziflam is a preemergent herbicide widely used for the control of weeds in pecan(Carya illinoinensis)orchards in the southwestern region of the United States.Given the paucity of data regarding the effect of indaziflam on the biochemical properties of soils supporting pecan production,this study was conducted to evaluate the effects of different application rates of indaziflam on soil microbial activity,diversity,and biochemical processes related to nitrogen(N)cycling.During two consecutive growing seasons(2015 and 2016),soil samples were obtained from experimental mesocosms consisting of soil-filled pots where pecan saplings were grown and treated with indaziflam applied at two different rates(25 and 50 g active ingredient(ai)ha−1,with the higher rate being slightly lower than the recommended field application rate of 73.1 g ai ha−1).Soil samples were collected approximately one week before and one week after herbicide application for determination of soil microbial biomass and diversity,N mineralization,and β-glucosaminidase activity.Soil samples collected from the control mesocosms without herbicide application were treated in the laboratory with two rates of indaziflam(75 and 150 g ai ha−1)to determine the immediate effect on microbial activity.No significant effect of herbicide treatment on soil respiration and microbial biomass was detected.The results showed a slight to moderate decrease in microbial diversity(7%in 2015 and 44%in 2016).However,decreased β-glucosaminidase activity with herbicide treatment was observed in soils from the mesocosms(33%)and soils treated with indaziflam in the laboratory(45%).The mineral N pool was generally dominated by ammonium after indaziflam application,which was consistent with the drastic decrease(75%)in nitrification activity measured in the laboratory experiment.The results of this study indicate that indaziflam,even when applied at higher than recommended rates,has limited effects on soil microbial activity,but may affect N cycling processes.

KeyWords:Carya illinoinensis,β-glucosaminidase activity,mesocosm experiment,microbial biomass,microbial diversity,mineral nitrogen speciation,nitrification

Pecan(Carya illinoinensis(Wangenh.)K.Koch)is an important agricultural commodity in the sunbelt region of the southern United States,contributing>85%of the global production of pecan(Onunkwo and Epperson,2000).To maintain the productivity of pecan orchards and preserve the quality of pecan crop,appropriate application of agrochemicals is needed for pest control.Indaziflam(N-((1R,2S)-2,3-dihydro-2,6-dimethyl-1H-inden-1-yl)-6-((1R)-1-fluoroethyl)-1,3,5-triazine-2,4-diamine)(Alion®Herbicide,Bayer CropScience,USA),a phytosanitary product,is a cellulose biosynthesis inhibitor herbicide.It was registered in 2010 by the United States Environmental Protection Agency(USEPA)for preemergence control of annual grass and broadleaf weeds(Alonsoet al.,2011;Kaapro and Hall,2012;Brabhamet al.,2014),and it has been extensively used in pecan orchards since then.Indaziflam is also registered for agricultural and forestry uses in Brazil,Canada,Indonesia,and South and West African countries(Guerraet al.,2014;APVMA,2015;Torreset al.,2018).

Indaziflam exhibits long residual activity in soils without causing deleterious effects on turf grass and growing crops.Research has shown that even after 28 d of application,indaziflam remains highly effective(93%–100%)against the preemergence of annual bluegrass(Brosnanet al.,2012).These attributes have contributed to its wide adoption in high-value horticultural crop production.Several studies comparing indaziflam with other preemergence herbicides(e.g.,oxadiazon,pendimethalin,oryzalin,dithiopyr,prodiamine,dimethenamid,atrazine,S-metolachlor,dithiopyr,flumioxazin,oxadiazon,pendimethalin,and simazine)have documented the superior weed control performance of indaziflam(Becket al.,2013;Begitschkeet al.,2018).However,it should be noted that indaziflam performance(i.e.,weed control without crop damage)can be affected by application rate,crop species(e.g.,sunflower being a very tolerant species),crop developmental stage,and environmental stress(Guerraet al.,2014).Soil properties also play a role in determining indaziflam mobility and potential injury to crops.Previous studies have reported decreased injuries to bermudagrass with increased clay and organic matter contents of soils(Joneset al.,2013a,b;Schneideret al.,2015).

In a laboratory study using Brazilian Oxisols and American Mollisols,Alonsoet al.(2011)concluded that indaziflam was low to moderately mobile in the environment.Leonet al.(2016)reported that indaziflam had greater lateral movement potential than other commonly used preemergent herbicides.Previous laboratory studies have found that the rate of indaziflam leaching is proportional to its application rate and amount of rainfall(Jhala and Singh,2012;Jhalaet al.,2012).Soil dissipation studies have reported indaziflam half-life values of 30–86 d(González-Delgadoet al.,2017),22–176 d(personal communication),and>150 d(USEPA,2010).The three main breakdown products of indaziflam are indaziflam-triazine indanone,indaziflam-carboxylic acid,and indaziflam-triazinediamine,with the latter two being generally more mobile in soils than indaziflam(Alonsoet al.,2015).

Currently,there is limited published information on the interactions of indaziflam and its breakdown products with terrestrial biota.Specifically,there is limited information on the effect of indaziflam on soil biochemical processes,especially when considering the persistence of indaziflam in soils and its potential interactions with soil microorganisms.Torreset al.(2018)reported an increase in soil respiration following treatment with 100 g active ingredient(ai)ha−1of indaziflam but no significant effect on soil microbial biomass or amylolytic microorganism activity.Similarly,Cook and Carr(1985)reported that soil nitrification was not negatively affected by application at the recommended rate of dichlobenil,another cellulose biosynthesis inhibitor herbicide.Several previous studies have reported no adverse effect or even increased soil microbial activity when herbicides were applied at recommended rates(Dinelliet al.,1998;Busseet al.,2001;Haneyet al.,2002b;Bhowmicket al.,2014).In contrast,some studies have reported substantial suppression of microbial activity with increased application rates of some herbicides(Fuscaldoet al.,1999;Crouzetet al.,2010;Sireeshaet al.,2012;Sofoet al.,2012;Longet al.,2014;Pose-Juanet al.,2017).In light of these variable findings,laboratory and field-scale studies are required to evaluate the effect of indaziflam application rate on soil microbial activity and soil biochemical processes.Therefore,this study aimed to evaluate the effects of different application rates of indaziflam on soil microbial activity,microbial diversity,and biochemical processes related to nitrogen(N)cycling in a greenhouse mesocosm experiment with cultivation of saplings of pecan and in an experiment in the laboratory.

Greenhouse experiment and soil sampling

In the present study,soil samples were collected from a greenhouse experiment that was previously set up at the Fabian Garcia Science Center,New Mexico State University,USA to investigate the mobility and fate of indaziflam in large mesocosms(volume:35 L;surface area:972 cm2)filled with a sandy loam soil(thermic Typic Torripsamments)and planted with 2-year-old pecan trees.The soil was sandy loam in texture(760 g kg−1sand,150 g kg−1silt,and 90 g kg−1clay),with a moderately alkaline pH(8.01±0.15)and an organic carbon(C)content of 8.72±1.1 g kg−1soil.

A hand-held sprayer was used to apply indaziflam to the mesocosms at 25 and 50 g ai ha−1on July 26,2015 and July 25,2016.The high application rate(50 g ai ha−1)used in the greenhouse mesocosms was set to be slightly lower than the recommended field application rate(73.1 g ai ha−1),considering the lower dissipation rate of herbicides in the greenhouse than in the field.The soil surface in the indaziflam-treated mesocosms was either left undisturbed or homogenized to a depth of 5 cm using a hand tool;hence,two soil conditions were used.Three pots filled with homogeneous soil,without herbicide application,were used as a control.Each treatment(including the control)was set up in triplicate.The pots were irrigated with 7 L water 24 h after indaziflam application and every two weeks during the growing season.During the non-growing season(December–early May),the mesocosms were placed outside and irrigated every other week with 2 L water.Pecan trees were fertilized(one teaspoon per pot)with 20-20-20(Nphosphorus-potassium)granular Helena Brand fertilizer.A detailed description of the greenhouse experiment was provided by González-Delgadoet al.(2017).

During the 2015 and 2016 growing seasons,soil samples up to 15 cm depth were collected from the mesocosms with a push probe approximately one week before and one week after indaziflam application.Soil samples were stored in plastic bags and transported to the laboratory in ice-filled containers.Except for sieving(1-mm mesh),the soil samples were kept refrigerated(4◦C)until use.

Laboratoryassessment of the immediate effect of indaziflam on soil microbial processes

For this assessment,we used composite soil samples collected from the control mesocosms in July and August of 2016 without exposure to indaziflam.Therefore,this assessment would provide information on the soil microbial community response to initial exposure to indaziflam.Each composite sample was divided into three portions,and each portion(approximately 40 g of soil)was transferred into glass jars.One jar received 2 mL water(untreated control),and the other two jars were treated with 2 mL indaziflam solution to yield equivalent application rates of 75 and 150 g ai ha−1to evaluate the influence of application rates greater than the recommended one on microbial activity.The jar content was thoroughly mixed and left at 24◦C.

Determination of soil microbial activity,microbial diversity,and biochemical processes

Within 3–4 d of collection,the soil samples were extracted with 0.5 mol L−1KCl(1:2 soil to solution ratio),and the extract was analyzed for concentrations of NH4-N and NO3-N using EPA methods 350.1(USEPA,1993)and 353.1(USEPA,1978),respectively,on an Aquakem 20 photometric analyzer(EST Analytical,USA).For determination of soil respiration,sieved field-moist soil sample(20 g)was placed in a glass jar and incubated at 23◦C.The concentration of CO2in the headspace was measured three times over a 10-d period using a Varian CP3800 gas chromatograph(Varian Analytical Instruments,USA)equipped with a thermal conductivity detector(Jacinthe,2015).The rate of CO2production during the incubation period was used as a measure of soil respiration.Following the 10-d incubation,the soil was extracted with 0.5 mol L−1KCl(40 mL)for 1 h with shaking and filtered,and the filtrate was analyzed for the final concentration of mineral N(NH4-N+NO3-N).The rate of N mineralization was computed as the net amount of mineral N(final concentration–initial concentration)produced during the 10-d incubation period.

The activity ofβ-glucosaminidase was assayed as described by Parham and Deng(2000).Briefly,1 g soil was weighed into a Pyrex glass tube and treated with 4 mL acetate buffer(0.1 mol L−1,pH 5.5)and 1 mL 10 mmol L−1p-nitrophenyl-N-acetyl-β-D-glucosaminide solution as the substrate.The suspension was mixed and placed in an incubator at 37◦C for 1 h.At the end of the incubation,1 mL CaCl2(0.5 mol L−1)and 4 mL NaOH(0.5 mol L−1)were added to stop the reaction and allow the development of a yellow color associated with the release ofp-nitrophenol.The suspension was filtered,and the yellow color intensity of the filtrate was measured at 405 nm using a spectrophotometer.To account for background soil color andp-nitrophenol produced from non-enzymatic pathways(e.g.,hydrolysis),controls were performed using the procedure above,either without soil or with the addition of substrate after the reaction was quenched.The amount ofp-nitrophenol in the filtrate was determined using a standard curve forp-nitrophenol,andβ-glucosaminidase activity is reported asµgp-nitrophenol g−1soil h−1.

Microbial biomass C(MBC)was determined using the substrate-induced respiration technique(Anderson and Domsch,1975;Kaiseret al.,1992).Moist soil sample(10 g)was weighed into a glass beaker and amended with glucose(6 mg g−1soil).The beaker was then placed into a glass jar(450 mL)and incubated for 3 h in a temperature-controlled(23◦C)chamber.Air sample was withdrawn from the jar headspace and analyzed for CO2by gas chromatography(Jacinthe,2015).The rate of CO2production in the glucoseamended soil was used to determine the MBC,following the computation approach described by Kaiseret al.(1992).

Community-level physiological profiles(CLPPs)were assessed using BIOLOG ECO microplates(Biolog Inc.,USA).Each microplate contained 31 C sources comprising six categories of substrates:amines,amino acids,carbohydrates,carboxylic acids,phenolics,and polymers.The bottles and solutions used in this assay were autoclaved.For each treatment,duplicate 1 g soil sub-samples were suspended and serially diluted in 100 mL autoclaved phosphate buffer(10 mmol L−1,pH 7.4)to a final dilution of 10−3.The suspension was vortexed for several minutes between each dilution.Finally,150µL aliquots of the final dilution were added to each of the 96 microplate wells.The plates were incubated at 23◦C,and absorbances at 595 nm were read after 120 h using a VersaMax microplate reader(Molecular Devices,USA).Finally,Shannon’s diversity index(H)was computed as follows(Garland,1996;Graystonet al.,1998):

wherepiis the ratio of the absorbance of substrateito the sum of absorbances of all wells in the microplate.

On the day following indaziflam treatment,1 g soil was taken from each jar and serially diluted to assess microbial diversity during the laboratory assessment using the CLLP approach described above.In addition,10 g soil was also transferred into a 50-mL glass beaker to assess nitrification activity.The beaker was then amended with NH4Cl(80 mg NH4-N kg−1soil)and incubated at 24◦C for 10 d.Nitrification rate was computed as the rate of net NO3-N production(NO3-N at the end minus NO3-N at the beginning)during the incubation period.The indaziflam-treated soil was also assessed forβ-glucosaminidase activity as described above(Parham and Deng,2000)and for urease activity as described by Tabatabai and Bremner(1972).Briefly,soil(5 g)was incubated with tris(hydroxymethyl)aminomethane(THAM)buffer,toluene,and urea solution(0.1 mol L−1)at 37◦C for 2 h,and the net amount of NH4-N produced was taken as a measure of urease activity.Gravimetric moisture content was determined by oven drying of moist soil samples at 105◦C for 48 h.All results are reported on a dry weight basis.

Statistical analyses

Data were first analyzed for normality using the Shapiro-Wilk test prior to conducting analysis of variance(ANOVA).For most of the parameters(except soil respiration and N mineralization),data transformation was not necessary.For these two parameters,a log transformation(log(x+1),wherexis the original value)was satisfactory to meet the normal distribution requirement of ANOVA.The effects of soil condition(homogenizedvs.undisturbed),indaziflam application rate,and time(beforevs.after application)on soil biological parameters were analyzed using ANOVA.Analysis was conducted separately for each year using the General Linear Models procedure in SAS.When a statistically significant effect was detected,Fisher’s protected least significant difference test was used for comparison of the means.Unless otherwise noted,statistical significance was considered atP<0.05.

Biological activityand mineral N in soils from the greenhouse experiment

None of the soil biological parameters considered in this experiment was affected by soil condition(Table I).In other words,soil mixing and homogenization to achieve greater contact between soil microbes and the active ingredient of indaziflam had no significant effect on soil biological processes.Therefore,data from undisturbed and homogenized mesocosms were combined for subsequent data analysis and presentation.However,soil sampling time(before or after)relative to indaziflam application had a significant effect on some parameters,especially in 2015(Table I).

TABLE ISummary of analysis of variance(ANOVA)for the effects of soil condition(undisturbed vs.homogenized),indaziflam application rate(0,25,and 50 g active ingredient ha−1),and soil sampling time(before vs.after indaziflam application)on soil biological propertiesa)and mineral N speciation

In 2015,soil MBC and respiration rate were generally higher before indaziflam application than after application(Fig.1).Variation in soil respiration rate was noted in 2015(hence,the statistically significant effect of sampling timedetected by ANOVA,Table I),with averages of 4.1±1.4 and 0.7±0.1 mg C kg−1soil d−1before and after indaziflam application,respectively.However,as a similar decline also occurred in soils from the control mesocosms,this trend was probably not due to indaziflam application.In 2016,soil MBC and respiration rate were consistently similar before and after indaziflam application.Regardless of the rate of indaziflam application,there was no significant difference between the control and indaziflam-treated mesocosms regarding these parameters.The activity ofβ-glucosaminidase was sensitive to indaziflam exposure,as indicated by the statistically significant effect detected by ANOVA in both 2015 and 2016(Table I).Overall,β-glucosaminidase activity was on average(across application rate and year)1.5-fold lower in the indaziflam-treated soils than in the control soil.In 2015,Shannon’s diversity index showed a slight decline(7%)in the soil samples collected from indaziflam-treated mesocosms compared to the control,but the decline(44%)was more pronounced in 2016.

Fig.1 Soil microbial biomass C(MBC),respiration rate,β-glucosaminidase(β-GL)activity,and Shannon’s diversity index(H)one week before(July)and one week after(August)indaziflam application at 0,25,and 50 g active ingredient(ai)ha−1 on July 26,2015 and July 25,2016 in the greenhouse mesocosm experiment with cultivation of saplings of pecan.Error bars represent standard deviations of the means(n=15).PNP=p-nitrophenol.

The results also showed a significant effect of indaziflam treatment on mineral N speciation(P<0.01),but the intensity of this effect varied with year(Tables II and III).In 2015,it was difficult to relate the difference in mineral N speciation to the indaziflam treatment because the NH4/NO3ratio was similar between the treated and control mesocosms(Table II).In 2016,there was a similar representation of NH4-N and NO3-N in the mineral N pool in all treatments(NH4/NO3ratio of 3)before indaziflam application,but the mineral N was dominated by NH4-N following application of indaziflam(NH4/NO3ratio of 6–10)(Table III).The N mineralization rate(measured only in 2015)was not significantly affected by indaziflam application(Table II).

Immediate effect of indaziflam on soil biological processes

Indaziflam(applied at rates up to 150 g ai ha−1)had no immediate adverse effects on soil microbial diversity and urease activity(Fig.2),as revealed by soil samples collected one month apart(July and August).It should be noted that,across treatments,urease activity increased significantly from 43.2±0.9 to 63±0.2 mg NH4-N kg−1soil h−1between these two sampling dates.However,drastic decreases in both nitrification rate andβ-glucosaminidase activity were observed on the day immediately following laboratory exposure to indaziflam.Across sampling dates,βglucosaminidase activity and nitrification rate were decreased by 45%and 75%,respectively.

Fig.2 Shannon’s diversity index(H),urease activity,β-glucosaminidase(β-GL)activity,and nitrification rate of soils one day after exposure to indaziflam in the laboratory.Indaziflam application rates were equivalent to 0,75,and 150 g active ingredient(ai)ha−1.The soils used in this experiment had no prior exposure to indaziflam and were collected in July and August,2016 from the control without herbicide application in the greenhouse mesocosm experiment with cultivation of saplings of pecan.Error bars represent standard deviations(n=9).PNP=p-nitrophenol.

TABLE IIEffects of indaziflam application rate on soil NH4-N and NO3-N contents and N mineralization rate one week before and one week after indaziflam application on July 26,2015 in the greenhouse mesocosm experiment with cultivation of saplings of pecan

TABLE IIIEffects of indaziflam application rate on soil NH4-N and NO3-N contents one week before and one week after indaziflam application on July 25,2016 in the greenhouse mesocosm experiment with cultivation of saplings of pecan

This study was conducted to assess the impact on soil biological processes of indaziflam,a widely used herbicide in orchard production systems in the southwestern region of the United States.A previous study(González-Delgadoet al.,2017)examined the mobility and residence time of indaziflam in soils,but questions related to soil biology have not been examined.Overall,our results(a greenhouse mesocosm study and short-term laboratory assays)have shown limited or no effect of indaziflam on the size,diversity,and respiratory activity of soil microorganisms.However,the results indicate a potential impact on N cycling processes.

Soil MBC was not negatively impacted by two successive applications(2015 and 2016)of indaziflam at the recommended rate(25 and 50 g ai ha−1)(Fig.1).Similar results were obtained by Torreset al.(2018),who exposed soil microorganisms to an even higher rate(100 g ai ha−1)of indaziflam.Several previous studies have also reported no negative effects of several herbicides on soil microbial biomass(Haneyet al.,2002a;Weaveret al.,2007;Das and Dey,2013;Barman and Das,2015).Some studies have even reported instances of increased microbial biomass after herbicide treatments,and these observations have been ascribed to the capacity of some soil microbial species to use the herbicide as a source of nutrients and energy(Haneyet al.,2002b;Bhowmicket al.,2014).Reports of negative effects on soil microbial biomass have also been published,most often when the herbicide application rate exceeds the recommended rate.Sofoet al.(2012)reported a decrease in soil microbial biomass following exposure to four triazinyl-sulfonylurea herbicides(cinosulfuron,prosulfuron,thifensulfuron methyl,and triasulfuron),but negative effects were most noticeable when the application rate was 10-fold the recommended field dose rate for these herbicides.Similar detrimental effects were reported with triasulfuron(Pose-Juanet al.,2017)and with imazethapyr and quizalofop-p-ethyl(Sahaet al.,2016)following exposure of soil microorganisms to high doses of these products.

Soil respiration was not affected by indaziflam application across all four sampling occasions(Fig.1).Although variations in the rate of soil respiration were detected in 2015(Table I),these fluctuations are probably not related to indaziflam but to other factors.Pose-Juanet al.(2015)reported no effect of the herbicide mesotrione on soil microbial respiration,whereas Busseet al.(2001)reported an increase in microbial respiration with glyphosate when applied in excess(up to 100-fold)of the recommended field dose(although no effect on respiration was detected at the recommended dose).The respiratory response of soils to the addition of herbicides can be related to the C/N ratio of these products,given the general understanding that organic materials with low C/N ratios are degraded faster than those with high C/N ratios(Haneyet al.,2002b).Therefore,witha C/N ratio of 3,glyphosate is more easily biodegraded and used by soil microorganisms as a C source than mesotrione(C/N ratio of 12)(Walpolaet al.,2004).The same can be said regarding the microbial degradation of atrazine(C/N ratio of 1.4)compared to glyphosate(Haneyet al.,2002b).However,in the present study,microbial respiration rate and microbial biomass were not affected by the application of indaziflam(C/N ratio of 3)at 25 and 50 g ai ha−1.The lack of a stimulatory effect on soil respiration is not surprising considering the small amount of C injected into the soil system as herbicide relative to the large background of soil organic matter.More than indaziflam treatment,favorable soil moisture(Tables II and III)may have contributed to the relatively higher respiration rate measured in the soils collected in July 2015 compared to the other sampling occasions(Fig.1).In contrast,in an incubation study with vineyard soils and indaziflam doses 4–16 times greater than the recommended application rate,Koçaket al.(2021)reported a stimulation of C mineralization but a major decline in N mineralization after 42 d of incubation.Therefore,the N mineralization results in this study are in agreement with those reported by Koçaket al.(2021)and confirm that indaziflam application can potentially impact N cycling processes.The results showed limited to no impact of indaziflam application on microbial diversity during the laboratory experiment(Fig.2).In the greenhouse study,the effect on microbial diversity was barely noticeable in 2015 but significant in 2016(Fig.1).Variable effects of herbicides on soil microbial diversity have been reported.In a study involving 12 different herbicides,Lupwayiet al.(2004)found that none of the herbicides,when used at recommended rates,altered microbial diversity relative to the control treatment.In contrast,in an evaluation of the biological effects of methamidophos,Wanget al.(2006)reported an increase in microbial diversity.However,a limited impact on microbial biomass can be observed,which does not exclude the possibility of a shift in the soil microbial community structure.Such a shift could occur without a deleterious impact on microbial biomass and microbial diversity.Here,at least two observations support this line of reasoning.While values remained constant in the control,microbial diversity in the herbicide-treated soils progressively decreased from 2015 to 2016(Fig.1).These trends suggest that the size of the soil microbial community may not be affected by repeated applications of indaziflam,which may induce subtle alterations in microbial community structure and activity of specific groups of soil microorganisms,as suggested by the results of the N cycling assays.

Information on the long-term effects of herbicides on soil microbial activity,diversity,and biochemical processes is relevant for the selection of agricultural management practices that protect soil quality.Alguacilet al.(2014)reported the lowest arbuscular mycorrhizal fungi diversity in fields treated with the herbicide oxyfluorfen(three applications per year)compared to other agricultural management practices(i.e.,natural cover,oats straw application,and plowing).Similar to indaziflam,oxyfluorfen was shown to have a dissipation half-life of 34–52 d under natural rainfall and irrigated conditions(Alisteret al.,2009).Therefore,these preemergent herbicides can potentially interact with soil microorganisms for at least one month under field conditions.Consistent with this reasoning,soil microbial diversity showed a decreasing trend one week after indaziflam application during the greenhouse mesocosm study in 2015 and 2016(Fig.1).The lowest Shannon’s diversity index values observed in 2016 could possibly be indicative of the long-term effect of indaziflam on soil microbial diversity.

The activity ofβ-glucosidase was another soil biological property impacted by indaziflam applications in 2015 and 2016.Borowiket al.(2017)reported inhibition ofβglucosidase activity in soil samples 60 d after treatment with the herbicide Lumax,a mixture of terbuthylazine,mesotrione,and S-metolachlor(half-life:1–2 months).Similar results were obtained by Tejada(2009)in a study evaluating the degradation and soil biological effects of the herbicides glyphosate,diflufenican,and their mixture.As an enzyme responsible for the degradation of chito-oligosaccharides into amino sugars,β-glucosaminidase is important for C and N cycling in soils(Ekenler and Tabatabai,2004;Ponder and Eivazi,2008;Tautgeset al.,2016).Consequently,βglucosaminidase activity has been shown to correlate with indices of N mineralization in soils(Ekenler and Tabatabai,2004;Haddadet al.,2013).In the present study,N mineralization was not influenced by herbicide treatment in 2015.In both the short-term laboratory assessment(Fig.2)and the greenhouse study(Fig.1),β-glucosaminidase activity was almost always lower in the indaziflam-treated soils than in the controls.These results contrast with those of Meanset al.(2007)but agree with the findings of Ponder and Eivazi(2008),who reported lowerβ-glucosaminidase activity in soils treated with glyphosate and simazine compared to weedy controls.Additionally,in the present study,the difference inβ-glucosaminidase activity was present even before new herbicide application was made during a given year,suggesting a remanent effect of indaziflam onβ-glucosaminidase.Indaziflam may also have some impact on N mineralization,although the assessment of the Ncycling index in 2015 did not show a clear effect(Table II).However,mineral N speciation results in 2016(Table III)and results of N nitrification assays(Fig.2)indicate that soil microorganisms involved in the conversion of NH4-N to NO3-N(nitrification)may have been strongly impaired by indaziflam exposure.Nitrification is carried out by specific groups of bacteria(NitrosomonasandNitrobacter),whichare very sensitive to various toxicants including herbicides(Oortset al.,2006;Jacinthe and Tedesco,2009).This interpretation is supported by the observation that NH4-N content was consistently greater than NO3-N content in every treatment before and after herbicide application in 2016(results were mixed in 2015).These results suggest that,in the presence of indaziflam,N mineralization is dominated by the ammonification process with a parallel repression of nitrification.Using a laboratory experiment to evaluate the effect of various herbicides on soil microbial activity and nutrient availability,Das and Dey(2013)verified that ammonification proceeded at a faster rate than nitrification.This interpretation applies to the results of the present study.

Soil microbial biomass and respiration were not affected by indaziflam application(25 and 50 g ai ha−1)in the greenhouse experiment.A limited impact was noted with regard to soil microbial diversity.However,indaziflam treatment had a noticeable effect onβ-glucosaminidase activity and mineral N speciation,with a dominance of NH4-N in the mineral N pool in indaziflam-treated soils.These results were further confirmed in the laboratory experiment examining the short-term impact of indaziflam exposure on N cycling processes,namely,an apparent inhibition of nitrification.Agrochemicals(nitrification inhibitors)are often purposely added to N fertilizers to block NH4-N to NO3-N conversion in soils to reduce N loss and improve N-use efficiency of agro-ecosystems.If this effect can be demonstrated in future studies,it would represent a novel and undocumented property of this herbicide.

The authors thank the New Mexico State Agricultural Experiment Station,National Institute of Food and Agriculture and Nakayama Endowment,USA.The authors also thank Bayer CropScience,USA for financial support.