Jounrer,oF THEArr,rnnrcat Moseurro CoNtRor, AssocrerroN VoL. 8, No. 1
IMPACT OF THE BLACK FLY (DIPTERA: SIMULIIDAE) CONTROL AGENT BACILLUS THURINGIENSIS VAR. /SRAETENS/S ON CHIRONOMIDS (DIPTERA: CHIRONOMIDAE)AND OTHER NONTARGET INSECTS: RESULTS OF TEN FIELD TRIALS
DANIEL P. MOLLOY
Biological Suruey,New York State Museum, Thc State Ed.watinn Departmcnt, Cultural Eduration Center, Albany, NY 12230
ABSTRACT. Except for moderate mortality among fiIter-feeding chironomids, Rheotanytarsusspp., the results of 10 field trials with Bacillus thuringiensis var. israelcnsis(B.t.i.) indicated a wide margin of safety to the chironomid community and other stream nontarget insects. Mayflies, caddisflies and 2 other types of chironomids, i.e., tube-dwelling (Chironominae) and surface-dwelling,caseless larvae (mix of Chironominae,Diamesinae, Orthocladiinae, Tanypodinae),did not appearto be affected.The suscep- tibility of filter-feeding chironomids was considerably less than black flies; for example, 4 months of data collected during an operational black fly control program indicated a mean (+95% Cl\ mortality amongRheotanytarsus larvae of 23(15-32)% vs. 98(97-99)% among black flies. Although clearly dem- onstrating the potential of adverseimpact on filter-feeding chironomids in operationalblack fly programs, these trials also confirmed the narrow impact of B.t.i. on the overall stream insect community.
INTRODUCTION of Molloy and Struble (1988). Formulations were applied simply by spreading aqueoussus- The bacterium Ba.cillusthuringiensis Berliner pensionsacross a stream using a plastic cup. var. isro.elpnsisde Barjac (B.t.i.) is consideredto Somefield trials required one application, while have low nontarget impact when applied to 480 applications were made in the operational streams and rivers for the control of black fly program (field trial 10). Water temperature (Diptera: Simuliidae) Iarvae (Dejoux and range during the study was 0-23'C. Field trials Elouard 1990, Lacey and Mulla 1990, Molloy 1-9 represented relatively small-scale, short- 1990).Lethality has been clearly demonstrated term (one wk) experiments in low-discharge against only a small group of nontargets, almost (320-20,740 liters/min), 1-5 m wide streams. all of which are flies in the superfamily Culicoi- Field trial 10, in contrast, was conductedas part dea (Diptera: Nematocera).Within this super- of a 4-month (April-July) operationalblack fly family, the family Chironomidae is the most control program (Molloy and Struble 1989); it Iikely nontarget group to suffer some mortality involved stream retreatments every 3-4 wk (Car and de Moor 1984, Pistrang and Burger along 32 km of stream length and included 1984,Back et al. 1985,de Moor and Car 1986). streamsranging from rivulets (50 liters/min) to This paper presentsthe results ofan extensive 10 m wide (76,100liters/min). Depending on seriesof field trials in small streams to further existing stream conditions, resourcesavailable investigate the safety of B.t.i. formulations to and degreeof quantification desired, a variety nontarget insects, with the major focus on chi- of techniques were used to measurethe impact ronomids. These trials were conducted to pro- on chironomids and other nontargetsas outlined vide the agencythat issuespermits for black fly below. control within New York, the New York State Department of Environmental Conservation, FIELD TRIAL I with additional data on the impact of B.t.i. in stream communities. Objectiue:Assessing impact on nontargets by recording change in their densities:Submerged plastic ribbons (2.8 x 30 cm) were used as arti- MATDRIALS AND METHODS ficial substrates.Eighty were placed in both an Ten field trials were conducted in northern upstream control area and 2,000m downstream New York State during April-August in streams of the treatment point at 4 days pretreatment. with moderate to high densities of black fly Each ribbon was numbered to allow for random Iarvae.Test streamslacked extensivevegetation removal and was held in place at one end by a and had riffles of cobble/gravelinterspaced with 100 g lead weight. Half of these ribbons were silt bottom pools. Test variables included for- removedand placed in alcohol at each sampling mulation, dosage,number of applications per area at t h pretreatment and the remaining half stream, temperature and stream discharge, as at 4 daysposttreatment. Thesepreserved insects outlined in Table 1. Prior to stream treatments, were subsequentlyidentified and their densities dischargewas measuredfollowing the methods (no./m' ) calculated. MARCH 1992 NoNTARGET IMPACT oF B.'r.t
d u.l FF @(0 ON oH60 (or OJ rti @i(oc{ I 9<. 6-a r6l bNdm a Eg cor dro .9 9l Ni HO b! d d a e
a B, 3. B, 3. = ET ;F>::>i:)*iEr+ "a ^ iE:\.: E?+sd? i? ffdzffdzffaz3oz?F B E ? Ef;'i !.E !.E ;{s;+s;is;+sffi tso ct) ;: g.: i: P.: I 9.: i:9.: !; H {> as lri !!!!!lri I !.'i ! !r'i ! &A .4 2> oSZZ*#=sz Bs
.j $ ri O I 6lF-rlii .i Hiii o B u o (6 e F
E ".!! i.g a = ".$ tr c .!i H E Io i trx trx trH 'F tr E o ^L^ *c a <; .l a o. p. VHAH o 8F F F tr q. q. F F SB.E {1 B t } b S @.. o@".*H €ct 6. E 6 3 H H @* F E XE:F SEE E E € € € : o I o-O-r o 9.Y+.!EEa -q € $tEF ooFo"-- fif f fl g # E F a L) bo a F o €9€! o€ @aaa f> mbobobo : k'= --h bo - b a,* hx I "dF F- Y ' = 6 C^ F F X+ 6o ctRS 5 5 vo e€ai;5FE E E @ o iiE e 6 6 aa 2F E:d E o tr: 5-X .Y.=oE = = .YE EAFR!o ui; -o E:o E =E E E !q o @ F F 9q E tr 3 e 9.6 U: :t > q- -h(; .9tr F F F =:E;EHE IAA =E: tr o o a a*3 EEr'- I +fl Es F6O).' dq F .E .E AN C ar O > o.= (f) FCEEF?F3?E P E N6 A ro (t) (n il.+ o k L atr d7 ed F =to.i E! 26 Jounner,oF THE Aunnrcex Moseutro CoNrnol Assocrerron VoL.8,No. I Table 2. Field trial 1: changesin densities of nontarget insects on artificial substratesfollowine treatment with Bactimos@WP.' Surface-dwellins ,. Mayfly nymphs Caddisfly Iarvae .hil;;;Ji;; Total Location and time of sampling No./m' Vo change No./m2 Vo change No./m' Vo c}range No./m2 % change Upstream control t h pretreatment 1,991 368 28 2,387 4 days posttreatment 6,508 ?05 +92 61 +118 7,274 2,000m downsiream t h pretreatment 3,025 558 36 3,618 4 days posttreatment 5,683 835 +50 72 +100 6,590 +82 " Changein no./m2 of black flies at upstream control and 2,000m site, respectively,was +70% and -76%. Table 3. Field trial 2: changesin mean number of nontarget insects in 5 Surber samplesfollowing treatment with Teknar@WDC.' Surface-dwell- Mayfly Caddisflv ing chiron- uther lnsects" I otal nymphs larvae omid larvae Location and time of sarnpling No. % change No. % change No. % change No. % change No. 7o change Upstream control t h pretreatment 724 126 54 40 944 5 days posttreatment 958 +32 r23 -z 65 +20 49 +23 1,195 lZt 400 m downstream t h pretreatment 151 105 34 19 309 5 days posttreatment 341 +726 172 +64' 69 +103' 33 +74 615 +99" 'Change in mean number of black flies at upstream control and 400 m site, respectively, was +380% and -\007o. b Includes stonefly nymphs, Iarval and adult elmid beetles, empid larvae and tipulid larvae. " Statistically significant (t-test at 5% level). Objectiue:Assessing impact on chironornidsby FIELD TRIAL 2 recording changein their body size: In addition Objectiue:Assessing impact on nontargets by to the above-mentionedquantitative assessment recording change in their densities an method, a qualitative method was also used in following intentional ouerdose:An overdose(equivalent of field trial 1 to determine if any chironomids had 201 ppm/l min applied as a maximum challenge been killed. Since earlier instar chironomids are to stream nontargets. Five Surber sampleswere more susceptibleto B.t i. intoxication (Ali et al. taken at t h pretreatment and at 5 days post- 1981),the mean larval body size within chiron- treatment at 400 m downstream of the treat- omid populations would increase following ment point and in an upstream control. Each B.t i.-induced mortality. meas- Consequently, sample was placed in a jar containing 70% al- urement of larval size, specifically postgenal cohol. Insects were subsequently sorted and head Iength on capsules,was used to determine counted at the laboratory. Mean numbers per whether mortality had occurred in the chiron- sample were compared for statistical signifi- (a omid larval community. Fifty caselesslarvae canceby t-test (P < 0.05). mix of Chironominae, Diamesinae, Orthocladi- inae, Tanypodinae) were collected with an art- ist's brush from artificial substratesat 2,000 m FIELD TRIALS 3-1O downstreamof the treatment point immediately Objectiue:Assesslng impact on chirornmids by pretreatment and again at 4 daysposttreatment. exarninationof liuing and deadlaruae: In these 8 As an experimental control, an equal number field trials, 3 types of late-instar chironomid also was collected concurrently from upstream larvae were checkedfor mortality both pretreat- of the treatment point. Larvae were preserved ment and posttreatment; these were caseless in7lVo alcohol,and their postgenallengths sub- surface-dwellers,filter-feeders and tube dwell- sequently measured at the laboratory. Mean ers. Larvae which did not move after being sizesat these sites were then compared for sta- touchedwith a probe were considereddead, with tistical significanceby t-test (P < 0.05). magnification (10x) usedas neededin this proc- MARCH 1992 NoxreRcnr IMPAcTot B.r.t ess. Caselesslarvae were collected from stream substrates(stones, wood debris, vegetation) and O were usually immediately examined in the field. B =a Filter-feeding larvae Iive in cases attached to @ Ytr roorNO. a in the 10 trials (Tables 2-8). Mayflies, caddis- flies, and the other 2 types of chironomids, i.e., tube-dwelling (Chironominae) and surface- @-! x- dwelling,caseless larvae (mix of Chironominae, Diamesinae,Orthocladiinae, Tanypodinae), did a a =a aimO@@ unbi.9a q iNNT not appear to be affected. o a o =;.E F In field trial 1, no statistically significant dif- EF9tr ference was found between the size of chiron- a di" omid larvae pretreatment vs. 4 days posttreat- ment at the 2,000 m site: mean (+957 61; postgenal lengths before and after treatment were, respectively, 171(153-188)pm and 162(143-181)pm.Upstream controls similarly Bts showedno significant changefrom pre- to post- N@mooN iNi(l)ri treatment, i.e., 120(108-132)pmand 140(120- 160)pm. These data indicated that surface, d dwelling, caselesschironomids were unaffected a at the 2,000m site where black fly densities had beenreduced by 76%.Among susceptibleinsect c! populations, morbality due to B.t.i. ingestion is 2, o .+^^^^^ almost invariably higher among early instars. m NNTOOO Table 5. Field trial 5: mortality among chironomids present on stream substratesat one day following treatment with Vectobac@WP.' Samplingsites (m % mortality (+9570 Cl)b downstreamfrom the initial treat- Surface-dwelling Filter-feeding Tube-dwelling ment point) chironomids chironomids chironomids Upstream controls 0 0 0 40 0 65 (43-84) 0 140 2 (0-10) 91 (64-100) 0 240 0 80 (66-91) 0 ' 100%black fly rnortality at all sampling sites. o Mean of 3 sampleswhere n : 3O,20 and 20, respectively,for surface-dwelling,fiIter-feeding and tube- dwelling larvae. Table 6. Field trials 6, 7 and 8: mortality among chironomid larvae present on stream substratesat one day followins treatment with VectobacWP@ and Teknar'DWDC." % mortality (+95% CDb Trial 6 Trial 7 Trial 8 Vectobac@WP Vectobac'nWP TeknartDWDC Distance downstream from the treatment Filter-feeding Surface-dwelling Tube-dwelling Filter-feeding point (m) chironomids chironomids chironomids chironomids Upstream control 3 (2-4) 4 (0-1e) 3 (0-21) 4 (0-23) 15" 6 (2-13) 9 (2-18) 8 (1-18) 45 (37-52) " Black fly mortality was 0% at upstream control in all trials and was 94, 100 and 84% at the 15 m sites, respectively,in trials 6, 7 and 8. bMean of 3 sampleswhere n - 34, 30 and 20, respectively,for trials 6, 7 and 8. " Correcteddata nresented(Abbott's formula). Table 7. Field trial 9: mortality among black flies and chironomids present on stream substratesat one day followingtreatment with Vectobac"WP. Sampling sites (m % mortallLy (number sampled) downstreamfrom the initial treat- Surface-dwelling Filter-feeding ment point) Black flies chironomids chironomids Upstream control 0 (180) 0 (17) 0 (38) 40 79 (29) 0 (e) 21 (61) 95 r00(44) 0 (5) 38 (21) L25 loo(18) og) 30 (71) 150 100(59) 32(111) 200 100(275) 2r (75) 240 100(138) 0 (25) 6 (16) 315 100(154) 18 (8e) 380 100(183) 10 (81) 545 100(220) o t+ei 565 100(231) 3 (33) 75 (20) 590 r00(197) 6 (4e) 82 (40) oJa 99(176) 0 (8) 61 (18) 680 100(138) o9 56 (50) ?80 r00 (74) 60 (81) 880 1oo(54) 80 (41) 960 1oo(68) 43 (46) 1,045 ee(r27) 57 (35) MARCH 1992 Norrencrr IMPACToF B.rt. I No evidence of adverse impact was likewise apparent from Surber sample data in field trial 6t a tr ri ^O 2. Posttreatment increases in density down- o rol srtss lro stream of the treatment were greaterthan those , :66 coi a a recorded in the upstream control for all insect u+l$o a groups(Table 3). Field trials 3-10 focused exclusively on chi- o ronomidlarvae (Tables 4-8). While caselesssur- face-dwellersand tube-dwellers appearedunaf- u kxtr =>e fected in all trials, some filter-feeders consist- :Pkc o l{cdd ently died following all B.t.i. treatments. In o a .o F:- f a general,when black fly mortality was very high (e.g.,9l-t00%), filter-feedingchironomid mor- a .i tality wasmoderate (ca.9-60%) (Tables 4, 6 and @co 7). An adverseimpact amongfilter-feeding chi- Cd m ronomids was not limited to a particular for- lJ mulation or temperature (Tables 4-8); mortality occurredwhen either of 2 formulations was used and across a wide temperature range (i.e., 0- a 23"C). a cg o +l DISCUSSION bo Chironomids were the most carefully moni- tored nontarget group in these trials, and the 60 hi LXtr data clearly indicated a wide margin of safety to @6 cO o :YHC all exceptthe filter-feeders,Rheotanytorsus spp. tcd6 N oA; B Almost a decadeago, filter-feeding chironomids a were predicted to be the nontarget insect most likely to be affectedby .B.t.i.applications (Gaug- 33 oq ler and Finney a tr: 1982). Rheotanytarsus Iawae $@ construct silken IO nets which filter-out particles drifting downstream; a thus, these chironomids, h{i t4 !;N as well as black flies and mosquitoes, have a feeding behavior which serves to concentrate B.t.i. from the water column-thereby increas- ^o) N .+l ing the likelihood of their intoxication and AS .: No lr death. Thesecurrent trials have provided exten- orl i@ sive evidence that the Rheotanytarsrzsspp. in- habiting small streams can be killed at dosages U a within operationalcontrol levels in temper;te o climates (5-50 ppm/1 min). The mean filter- al :., o 6@ H;ic c06 feeding chironomid ^l = !'i <@ mortality during field trial C6 :!kq 10 was 23% (Table al l- from a North American study (Back et al. 1985) rotting insects become detached. Since some in which an intentional-overdoseof ca. 85 ppm/ deadinsects usually detachbefore an assessment 1 min had been applied. In this latter study, it is performed, this method typically underesti- was theorized that these nonfilter-feeding ne- mates true mortality, thereby losing some accu- matoceransingested B.t.l. cry'stalsfrom benthic racy. Another inherent problem with this as- surfaceswhere the crystals had adhered or set- sessmentapproach is selectionof the insects; tled, and this is a highly probable scenario. when examining a substrate for insects, it is Mortality among nonfilter-feeding chironom- easier to be distracted by and count a moving ids did occur in trial 2-a test designedto assess insect as comparedwith a motionless dead one. the affect of B.t.i. on nontargets at an extreme This technique,however, is the best quantitative overdose(equivalent of 201ppm/1 min or ca. 10 assessment method possible for nontargets times higher than that typically used in opera- which live in attached cases,such as Rheotany- tional programs).Although no reductionin sur- tarsus.In contrast to black fly larvae which drift face-dwelling, caseless chironomids was re- off in the current when they are dead, B.t i.- corded at the 400 m site (Table 3), dead larvae kiI\ed Rheotanytarsus larvae remain in their were observed on vegetation at 4 h posttreat- caseswhich are permanently attached to sub- ment at 40 m downstreamof the treatment point strates. Measurement of their mortality by di- (% mortality undetermined).This indicated that rect examination of unpreservedIarvae in their surface-dwelling,caseless chironomid larvae can casesis thus a very simple and highly accurate be killed close to the sourceof such an extreme procedurewhich can easily be incorporated into overdose. Similarly, Rutschke and Grunewald a nontarget assessmentprogram. (1984) observeddead chironomids in a stream Determining pre- and posttreatment densities trial using a dosage17 times higher than normal. of benthic insects using benthic samplers,such These 10 field trials indicate a very low risk as the Surber used in field trial 2, is a traditional to 2 groups of chironomids,i.e., tube-dwelling and an excellent way to quantitatively measure and surface-dwelling (nonfilter-feeding) larvae. impact in the benthic community. This method, In the test streams, populations of these latter however, requires a relatively large labor input 2 chironomids were far more abundant than to sort out the nontarget organisms from the filter-feeders, which typically represented less sample'sdebris. than 1% of the chironomid community. Thus, the overall impact of B.t.i. applicationson chi- ACKNOWLEDGMENTS ronomid communities in the test streams ap- pearedvery low. The assistanceof the following technical staff A variety of assessmenttechniques were em- is gratefully acknowledged:Bob Struble, Bar- ployed in this study and the following comments bara Griffin, Lora Couture, Bruce Kaplan, Judy Petteys, Jim are offered regardingtheir value and usefulness. Gerardi, Karen Ungeheuer, Ed Appre- Looking for changes in a nontarget group's Ash, Tracy Meyers and Paul Lundberg. landowners who mean body size, as used in field trial 1, was a ciation is also expressedto all granted permission these studies. qualitative method of impact assessment.With to conduct This investigation was supportedin part by the appropriate upstream controls, a significant in- UNDP/World Bank/WHO Special Programme crease in a nontarget population's mean size for Researchand Training in Tropical Diseases, from pre- to posttreatment indicates that mor- Abbott Laboratories, and Sandoz,Inc. This pa- tality, albeit unquantified, has occurred due to per is contribution number 678 of the New York selective elimination of younger instars. While State ScienceService. only qualitative, this assessmenttechnique is a very cost-effectiveone, requiring little labor and equipment, and it is highly recommended for REFERENCES CITED initial nontarget assessments.Having identified Ali, A., R. D. Baggsand J. P. Stewart.1981. Suscep- a nontarget group with a significant posttreat- tibility of someFlorida chironomids and mosquitoes ment increase in mean size, quantitative tests to various formulationsof Bacillusthuringiensis can be subsequentlyconducted focusing on that serovar.israelensis. J. Econ. Entomol. 7 4:672-677. nontarget. Back,C., J. Boisvert,J. O. Lacoursidreand G. Char- Posttreatment counting of attached live and pentier. 1985. High-dosage treatment of a Quebec insects, as used in field trials 3-10, is an stream with Bacilhts thuringiensis serovar. israelen- dead (Diptera: quantitative assessment method. sis: efficacy against black fly larvae economical, Simuliidae) and impact on non-target insects. Can. The timing of this assessmenttechnique is crit- Entomol. ll7 :1523-1534. ical for most nontargets. The counting must Car, M. and F. C. de Moor. 1984. The response of Vaal occur afber the B.t.i. treatment has had an op- River drift and benthos to Simulium (Diptera: portunity to inflict mortality, yet before the Nematocera) control using Bacillus thuringiensis MARCH 1992 NoNTARGEThupecr oF B.r:t. var. israelensis(H-14). Onderstepoort J. Vet. Res. nontarget effects on invertebrates and fish. J. Am. 51:155-160. Mosq.Control Assoc.5:397-415. Dejoux, C. and J. M. Elouard. 1990.Potential impact Molloy, D. 1990.Progress in the biological control of of microbial insecticideson the freshwater environ- black flies wlth BaciLlusthuringiercis var. israplen- ment, with special referenceto the WHO/UNDP/ srs,with emphasison temperate climates. pp. 161- World Bank, OnchocerciasisControl Programme. 786. In: H. de Barjac and D. Sutherland (eds.), pp. 65-83. In: M. Laird, L. A. Lacey and E. W. Bacterial control of mosquitoesand black flies: bio- Davidson (eds.), Safety of microbial insecticides. chemistry, geneticsand applicationsof Bacillus thu- CRC Press,Inc., Boca Raton,FL. ringiensis israelensis and Bacillus sphaerirus. Rut- de Moor, F. C. and M. Car. 1986.A field evaluation of gers Univ. Press,New Brunswick, NJ. Bacillus thuringiensis var. israelensisas a biological Molloy, D. and H. Jamnback.1981. Field evaluation control agent for Simulium chutteri (Diptera: of Bacilhrs thuringiensis var. isroelensisas a black (Diptera: Nematocera) in the Middle Orange River. Onder- fly Simuliidae) biocontrol agent and its stepoort J. Vet. Res. 53:43-50. effect on nontarget stream insects. J. Econ. Ento- Gaugler,R. and J. R. Finney. 1982.A review of Barillus mol. 74:314-318. Molloy, D. P. thuringiensis var. israelensis(serotype 14) as a bio- and R. H. Struble. 1988.A simple and inexpensive logical control agent ofblack flies (Simuliidae). pp. method for determining stream dis- charge from a streambank. J. Freshwater l-17 . In: D. Molloy (ed.),Biological control of black Ecol. 4:477-487. flies (Diptera: Simuliidae) with Bacillus thuringien- Molloy, D. P. and R. H. Struble. 1989.Investigation sis var. israelensis(serotype 14): A review with rec- of the microbial control of black flies (Diotera: ommendations for laboratory and field protocol. Simuli idae) with B aci llus t huring iz nsisv ar. israz Le n - Misc. Publ. Entomol. Soc.Am. 12(4). srsin the Adirondack Mountains of New York. Bull. Lacey, L. A. and M. S. Mulla. 1990.Safety of Bacillus Soc. Vector F,col. 14:266-2'16. thuringicnsis ssp. israelensk and,Bacillw sphaericu.s Pistrang, L. A. and J. F. Burger. 1984. Effect of to nontarget organisms in the aquatic environment. Bacillus thuringicnsis var. israelensison a geneti- pp. 169-188.1n: M. Laird, L. A. Lacey and E. W. cally-defined population of blackflies (Diptera: Si- Davidson (eds.), Safety of microbial insecticides. muliidae) and associatedinsects in a montane New CRC Press,Inc., Boca Raton, FL. Hampshirestream. Can. Entomol. 116:975-981. Merritt, R. W., E. D. Walker, M. A. Wilzbach. K. W. Rutschke,J. and J. Grunewald.1984. The control of Cummins and W. T. Morgan. 1989.A broad evalu- blackflies (Diptera: Simuliidae) as cattle pests with ation of B.t.i. for black fly (Diptera: Simuliidae) Bacillus thuringiensis H-14. Zentralbl. Bakteriol. control in a Michigan river: efficacy, carry and Mikrobiol. Hyg. Ser.A 258:413.