B M C Habitat Afﬁnity of Resident Natural Enemies …ccag.tamu.edu/files/2012/03/2010-brewernomaJEEsbabio...BIOLOGICAL AND MICROBIAL CONTROL Habitat Afﬁnity of Resident Natural

BIOLOGICAL AND MICROBIAL CONTROL

Habitat Affinity of Resident Natural Enemies of the Invasive Aphisglycines (Hemiptera: Aphididae), on Soybean, With Comments on

Biological Control

MICHAEL J. BREWER1 AND TAKUJI NOMA2

J. Econ. Entomol. 103(3): 583Ð596 (2010); DOI: 10.1603/EC09332

ABSTRACT We integrated a natural enemy survey of the broader landscape into a more traditionalsurvey for Aphis glycines Matsumura (Hemiptera: Aphididae), parasitoids and predatory ßies onsoybeanusingA.glycines-infested soybean,Glycinemax(L.)Merr., placed incroppedandnoncroppedplant systems to complement visual Þeld observations. Across three sites and 5 yr, 18 parasitoids andpredatory ßies in total (Hymenoptera: Aphelinidae [two species] and Bracondae [seven species],Diptera: Cecidomyiidae [one species], Syrphidae [seven species], Chamaemyiidae [one species])were detected, with signiÞcant variability in recoveries detected across plant system treatments andstrong contrasts in habitat afÞnity detected among species. Lysiphlebus testaceipes Cresson was themost frequently detected parasitoid, and no differences in abundance were detected in cropped(soybean, wheat [Triticum aestivum L.], corn [Zea mays L.], and alfalfa [Medicago sativa L.]) andnoncropped (poplar [Populus euramericana (Dode) Guinier] and early successional vegetation)areas. In contrast, Binodoxys kelloggensis Pike, Stary & Brewer had strong habitat afÞnity for poplarand early successional vegetation. The low recoveries seasonally and across habitats of AphelinusasychisWalker, Aphelinus sp., and Aphidius colemoni Viereck make their suitability to A. glycines onsoybean highly suspect. The widespread occurrence of many of the ßies reßects their broad habitatafÞnityandhost aphid ranges.Theconsistent lowÞeldobservationsofparasitismandpredation suggestthat resident parasitoids and predatory ßies are unlikely to contribute substantially to A. glycinessuppression, at least during the conventional time period early in the pest invasion when classicalbiological control activities are considered. For selected species that were relatively well representedacross plant systems (i.e., L. testaceipes andAphidoletes aphidimyzaRondani), conservation biologicalcontrol efforts may be fruitful. The additional information gained from expanding the natural enemysurvey into the broader landscape was essential in making these distinctions relevant to conservationbiological control, while adding agroecosystem-speciÞc information valuable to classical biologicalcontrol.

KEY WORDS biocontrol, soybean aphid, parasitoids, predatory ßies

The soybean aphid, Aphis glycines Matsumura(Hemiptera: Aphididae), was Þrst detected on soy-bean, Gyclines max (L.) Merr., in Wisconsin in 2000(Ragsdale et al. 2004). It rapidly spread throughout thenorth central United States, causing yield loss of up to40% (Ragsdale et al. 2007). Heteroecious holocylicpopulations were conÞrmed. It has a narrow host plantrange in this invasion zone, with soybean as the sum-mer host and buckthorn (Rhamnus L.), as the winterhost (Vennette and Ragsdale 2004).

In Asia, a much lower frequency of A. glycinesoutbreaks on soybean was attributed to natural en-emies and the ability of soybean to tolerate a fewhundred A. glycines per plant without economic

harm (Liu et al. 2004, Wu et al. 2004). Rates ofparasitism of A. glycines as high as 75% and substan-tial predation were reported in China (Gao 1994,Wu et al. 2004). In the north central and northeast-ern United States, a lady beetle and true bug con-tributed to A. glycines suppression (Fox et al. 2004,Rutledge et al. 2004), whereas parasitism of A. gly-cines rarely exceeded 10% (Nielsen and Hajek 2005,Kaiser et al. 2007, Noma and Brewer 2008). A pred-atory ßy fauna also attackedA. glycines in the UnitedStates and China (Wu et al. 2004, Kaiser et al. 2007,Noma and Brewer 2008). Although the number ofspecies ofA. glycinesnatural enemies in soybean wascomparable to that found in Asia (Kaiser et al. 2007),economic outbreaks of A. glycines occurred in thenorth central United States since its initial invasion(Ragsdale et al. 2007). These Þeld data were used aspart of the justiÞcation for an importation biologicalcontrol program (Heimpel et al. 2004).

1 Corresponding author: Texas AgriLife Research and ExtensionCenter, 10345 State Hgwy. 44, Corpus Christi, TX 78406 (e-mail:[email protected]).

2 Department of Entomology, Michigan State University, East Lan-sing, MI 48824.

0022-0493/10/0583Ð0596$04.00/0 � 2010 Entomological Society of America

Such Þeld data of the resident natural enemy faunaalong with literature review of the detected enemiesare recommended to discern need for additional nat-ural enemies and priority taxa to introduce (Van Dri-escheandBellows1996, seechapter8).Natural enemyÞeld surveys are common on the threatened crop, asdone for A. glycines on soybean in the United States(e.g., Rutledge et al. 2004), where they arguably de-tect resident natural enemies with the greatest poten-tial to suppress the pest. But natural enemy surveyslimited to the threatened crop may miss potentiallyeffective natural enemies located in the broader land-scape, as ecological and physiological barriers of thenatural enemy using the invasion pest on the threat-ened crop subside over time through host and habitatrange shifts and other mechanisms (Cornell andHawkins 1993, Rutledge and Wiedemann 1999). Andnatural enemy surveys on the threatened crop alonedo not allow depth in considering conservation bio-logical control strategies involving noncrop vegeta-tion.

Using A. glycines as a case example of an invasivepest with a narrow host plant range, we considerwhether knowledge of natural enemies of A. glycinesoutside of soybean may be relevant to decision-mak-ing in biological control. There is considerable vari-ation in the composition and structure of the land-scape where soybean is found, varying from asoybean and corn rotation (and sometimes wheatadded to the rotation) dominating the landscape tothis rotation nested in a matrix of crops and non-cultivated areas (Noma et al. 2010). We hypothesizethat integrating a natural enemy survey of thebroader landscape into a more traditional surveyfor A. glycines natural enemies on soybean willstrengthen guidance on classical and conservationbiological control approaches being considered dur-ing the invasive phase of the pest.

Methods and Materials

NaturalEnemySurveys inSoybeanand theBroaderLandscape. Habitat afÞnity of parasitoids and preda-tory ßies of A. glycines was estimated from recoveriesobtained from A. glycines-infested soybean placed insoybean and other cropped and noncropped plots.Using aphid-infested plants complements visual Þeldobservations of parasitoids and predatory ßies on thethreatened crop by conÞrming capability to completedevelopment to the adult stage using the aphid as host.The method has previously been used in soybean todetect A. glycines natural enemies (Noma and Brewer2008) and in other crops (Milne 1995).

Using this technique and selected visual Þeldobservations, we monitored for parasitoids andpredatory ßies at two locations: Michigan State Uni-versity Kellogg Biological Station Long Term Eco-logical Research plots in Hickory Corners (42� 24�N, 85� 23� W) and the Michigan State Universityfarms in East Lansing (42� 41� N, 84� 29� W). The twoexperimental sites used at the Kellogg BiologicalStation were the large main site that was subdivided

into replicated 1-ha plots, and small biodiversity sitewhich was subdivided into replicated 10- by 30-mplots. The experiments were set out as randomizedcomplete blocks. There were four plant systemtreatments used at the large main site: 1) alfalfa,Medicago sativa L.; 2) early successional vegetation,3) poplar [Populus euramericana (Dode) Guinier],and 4) a soybean-wheat (Triticum aestivumL.)-corn(Zea mays L.) rotation with each crop occurringonce every 3 yr (Table 1, see archived data sets fordetails of plant composition and plot maintenance[Kellogg Biological Station 2006]). These large plotswere designed to mimic agriculturally managed andminimally manipulated plant systems. The smallbiodiversity site contained three treatments ofthe soybeanÐwheatÐcorn rotation. Each crop waspresent each year by shifting the crop entry point ofthe rotation among the three treatments, allowingwithin-year crop to crop comparisons (Table 1, seearchived data sets for details [Kellogg BiologicalStation 2006]). Four replications of each treatmentand no insecticides were used at both sites. TheMichigan State site contained two treatments (soy-bean and buckthorn). Soybean plantings of 0.5 haand larger were selected that were adjacent to buck-thorn found along wood lots and railroad right-of-ways. The treatments were set in a randomizedcomplete block design and replicated three times in2006 and 2007.

At the Kellogg sites, A. glycines-infested soybeanwas placed in selected treatments three times in

Table 1. Plant system treatments and natural enemy samplingmethodsa used across 5 yr (2003–2007) of natural enemy surveysat Kellogg Biological Station (KBS) large main site and small biodi-versity site, Hickory Corners, MI, and the Michigan State site, EastLansing, MI

Plant systemtreatmentb

2003 2004 2005 2006 2007

KBS large main site

Alfalfa I, V I, V I, V I, VEarly successional I I IPoplar ISoybeanÐwheatÐ

cornSoybean Wheat Corn SoybeanI, V I, V I, V I, V

KBS small biodiversity site

SoybeanÐwheatÐcorn

Soybean WheatI, V I, V

WheatÐcornÐsoybean

Wheat CornI, V I, V

CornÐsoybeanÐwheat

Corn SoybeanI, V I, V

Michigan State site

Soybean I IBuckthorn I I

a I, A. glycines-infested soybean; V, visual Þeld observation. For therotational systems, the crop available for sampling is indicated.b Alfalfa,M. sativa; early successional, early successional vegetation

of idled crop land; poplar, Populus � euramericana with a heavilyshaded ground cover; soybean, G. max; wheat, T. aestivum.; corn, Z.mays; buckthorn (Rhamnus). Archived data sets of treatment de-scriptions and plot maintenance can be found at Kellogg BiologicalStation (2006).

584 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 103, no. 3

2003, four times each in 2004 and 2005, and two timesin 2006 between June and September. The timeperiods were chosen to represent a range of soybeangrowth stages (Fehr and Caviness 1977) and sea-sonal A. glycines population shifts: 1) mid-June, dur-ing early vegetative growth of soybean (V2ÐV3)

when aphids may Þrst migrate into soybean; 2) mid-July, during ßowering (R1ÐR2) when aphids may bemultiplying on soybean; 3) mid-August, during soy-bean pod Þll (R3ÐR5) when aphids may be reachingpeak densities; and 4) early September, during plantsenescence (R6) when aphids may be declining in

Fig. 1. Generalized seasonal recoveries of parasitoids and predatory ßies across plant system treatments and study yearsusing theA. glycines-infested soybean method at the Kellogg sites, Hickory Corners, MI, JuneÐSeptember 2003Ð2006. Numberafter the month is the number of samples used to generate the average (x, mean number recovered per pot) shown by linethickness (no line, x � 0; dashed line, 0 � x �0.1; thin line, 0.1 � x �1; regular line, 1 � x �5; thick line, 5 � x).

June 2010 BREWER AND NOMA: HABITAT AFFINITY OF NATURAL ENEMIES 585

density. At the Michigan State site, A. glycines-in-fested soybean was set out in the two treatments 12times in 2006 and nine times in 2007, between Mayand October to include time periods when A. gly-

cines switches between winter and summer hostplants.

Preparation of A. glycines-infested soybean, its Þeldexposure to predation and parasitism, and the process

Fig. 2. Generalized seasonal recoveries of parasitoids and predatory ßies across plant system treatments and study yearsusing theA. glycines-infested soybean method at the Michigan State site, East Lansing, MI, 2006Ð2007. Number after the monthis the number of samples used to generate the average (x, mean number recovered per pot) shown by line thickness (noline, x � 0; dashed line, 0 � x �0.1; thin line, 0.1 � x �1; regular line, 1 � x �5; thick line, 5 � x).

Table 2. Analysis of variance comparing natural enemy recoveries among plant system treatments (plant system) and sampling periods(date) for six species detected using the A. glycines-infested soybean method at the Kellogg sites, Hickory Corners, MI, 2004 and 2005

Taxa Yr Site Factor/interactiona F df P

L. testaceipes 2004 Large Plant system by date 2.19 7, 39 0.05Small Date 20.37 3, 33 �0.0001

2005 Large Date 3.85 3, 24 0.02A. asychis 2004 Large Plant system by date 3.86 7, 39 0.003

Small Plant system by date 5.10 6, 36 0.00072005 Large Plant system by date 33.54 4, 27 �0.0001

Small Date 601.54 3, 33 �0.0001B. kelloggensis 2004 Large Plant system 12.71 3, 42 �0.0001

Date 5.00 3, 42 0.005Small Date 11.63 3, 36 �0.0001

A. aphidimyza 2004 Large Plant system by date 4.96 7, 42 0.0004Small Plant system by date 3.39 6, 36 0.009

2005 Large Plant system 5.14 2, 25 0.01Date 7.59 3, 24 0.001

Small Plant system 6.11 2, 36 0.005Date 9.21 3, 36 0.0001

A. obliqua 2004 Large Date 16.39 3, 39 �0.0001Small Date 12.12 3, 33 �0.0001

2005 Large Plant system by date 4.20 3, 18 0.02Small Date 29.21 2, 24 �0.0001

S. contigua 2004 Large Plant system by date 4.67 7, 42 0.0006Small Date 3.91 3, 36 0.02

2005 Large Date 8.41 3, 27 0.0004Small Plant system by date 2.53 6, 36 0.04

a SigniÞcant factors and interactions are presented.


of incubating, extracting, and identifying capturednatural enemies were fully described in Kaiser et al.(2007). In brief, round plastic pots (17 cm in heightand 15 cm in diameter) each contained �12 V2 stagesoybean plants infested with �1,000 A. glycines toensure an abundant resource of aphids to mitigateexpected loss from aphid movement (Noma andBrewer 2008). Five pots were set out in each repli-cated treatment. At the Kellogg sites, pots were placedat preset stations per plot at the large main site andwere randomly placed in the plot interior at the smallbiodiversity site. At the Michigan State site, pots wererandomly placed in the soybean and buckthorn plots.In all cases, pots were placed at least 5 m from a plotedge.AnadditionalÞvepotsof infested soybeanplants

were kept in the laboratory during each Þeld exposureto check for possible natural enemy contamination inthe laboratory (no contamination was detected).

The A. glycines-infested soybean was exposed tonatural enemies for three to four days, with a consis-tent exposure period for each sampling date. The ex-posure period was limited to 3 to 4 d to reduce intra-guild predation. In addition, any ladybugs, lacewings,and true bugs were removed as the pots were broughtto the laboratory. Intraguild predation has been doc-umented in this system, but Þeld studies have shownthe effect to be minimal (Costamagna et al. 2008) ornot substantially interfering with parasitism (Chaconet al. 2008). In the laboratory, the potted plants wereincubated at room temperature under a photoperiod

Fig. 3. Relative recoveries of L. testaceipes using the A. glycines-infested soybean method at the Kellogg large main site(left) and small biodiversity site (right), Hickory Corners, MI, during years of signiÞcant differences among plant systemtreatments and sampling dates. Different letters on the bars indicate signiÞcant (� � 0.05) differences within the samplingdate when a plant system by date interaction was detected. Different letters under the dates indicate signiÞcant samplingdate differences when there was no interaction detected. Error bars are SEs. Ø, no data collected.

Table 3. Analysis of variance comparing natural enemy recoveries among plant system treatments (plant system: buckthorn vs.soybean) and sampling periods (date) for five species detected using the A. glycines-infested soybean method at the Michigan State site,East Lansing, MI, 2006 and 2007

Taxa Yr Factor/interactiona F df P

L.testaceipes 2007 Plant system by date 10.70 1, 41 0.002Date 12.93 5, 41 �0.0001

Aphelinus sp. 2007 Plant system 25.24 1, 41 �0.0001Date 6.02 5, 41 0.0003

A. colemani 2006 Plant system by date 13.69 3, 23 �0.0001A. aphidimyza 2006 Date 4.73 3, 23 0.01

2007 Plant system 7.54 1, 41 0.009Date 26.11 5, 41 �0.0001

E. americanus 2007 Plant system 4.16 1, 41 0.05Date 3.26 5, 41 0.01

a SigniÞcant factors and interactions are presented.


of 16:8 (L:D) h for 1 wk. Plants were then clipped,placed into emergence canisters, and incubated for 2wk to allow immature parasitoids and predatory ßiesto complete development. Emerged parasitoids andpredatory ßies were recovered from the collection vialand the inner lining of the canister.Concurrent Visual Field Observation.On the same

dates of setting out A. glycines-infested soybean at theKellogg sites, 20Ð25 plants from the corn, soybean,wheat, and alfalfa plots (Table 1) were nondestruc-tively inspected for aphids, mummiÞed aphids, andpredatory ßy larvae and pupae. Counts and identity ofaphids were recorded in each of the crops. On soy-bean, a series of count ranges were used to estimateA.glycines density per plant: up to 50 aphids werecounted, after which ranges of 51Ð100, 101Ð500, 501Ð1,000, and 1,001Ð5,000 aphids per plant were used andmean densities were estimated by taking the mid-points of ranges. In 2003, aphid mummies (aphelinidand braconid types) and immature predatory ßies(cecidomyiids and syrphids) were counted and iden-

tiÞed to family in the Þeld. From 2004 to 2006, allmummies and immature ßies were counted, broughtto the laboratory, and reared to adults in gelatin cap-sules for species identiÞcation. Visual observationswere not done at the Michigan State site.

Species determinations of parasitoids, predatoryßies, and aphids were made using keys and otheridentiÞcation aids (Heiss 1938, Gagne 1981, Vockeroth1992, Blackman and Eastop 2000, Pike et al. 2007).Specimens previously identiÞed by experts were usedas additional reference (Kaiser et al. 2007). In regardto our collections of Aphelinus not identiÞed to spe-cies, these populations are part of the Aphelinusvaripes complex (Heraty et al. 2007). The populationmay be derived from a Minnesota and Wisconsin re-distribution of a ÔWyomingÕ strain of this species pre-viously referred to as Aphelinus albipodus Hayat &Fatima (a member of theA. varipes complex) (Breweret al. 2001). Or the population may beAphelinus certusYasnosh (also part of this complex), which has re-cently found to be widespread in the eastern United

Fig. 4. Relative recoveries of aphid parasitoids and predatory ßies using the A. glycines-infested soybean method at theMichigan State site, East Lansing, MI, during years of signiÞcant differences among plant system treatments and samplingdates. Different letters on the bars indicate signiÞcant (� � 0.05) differences within the sampling date when a plant systemby date interaction was detected. Different letters next to the plant systems or under the dates indicate signiÞcant differencesin these factors when there was no interaction detected. Error bars are SEs. Graphs shown for cases where signiÞcant habitator date effects were detected.


States and possibly introduced with A. glycines (He-impel et al. 2010). Voucher specimens of parasitoidsand predatory ßies were deposited at Michigan StateUniversity (VC 2006-01).Data Analyses. To maintain experimental integrity

and to acknowledge the potential effect of differentplot sizes, data were analyzed separately for theKellogg large main site, small biodiversity site, and theMichigan State site. From data collected using A. gly-cines-infested soybean, the number recovered foreach taxon was averaged across the pots set in eachplot. Then, mean number of natural enemies recov-ered per pot was calculated for each plant systemtreatment and sampling period. For each taxon, year,and site, relative numbers of recoveries were com-pared among the different plant system treatmentsusing analysis of variance (ANOVA) following a ran-domized complete block design with repeated mea-sures (date) (PROC MIXED, SAS Institute 2004). Theplant system treatments were considered Þxed effectsin the model (Littell et al. 1991). The count data weretransformed into a logarithmic scale (log10 [100� �1/6]) to satisfy the assumption of normality forANOVA (Mosteller and Tukey 1977). When differ-ences were detected, TukeyÕs multiple comparisontest at � � 0.05 was used to separate plant systemmeans either across dates if only the plant systemtreatment effect was signiÞcant or within dates whena plant system by date interaction was detected.

For the visual Þeld observations at the Kellogg sites,mean number of aphids per plant, percentage of aphidparasitism per plant (we used the conservative Þeld

estimate based on mummies and live aphids at the timeof inspection), and the number of immature ßies perplant were estimated for each of the crops. Percentageof parasitism and density of immature ßies per plantwere separated at family or species level based onability to identify taxa.

Results and Discussion

Across sites and years, 18 parasitoids and predatoryßies in total were detected on soybean, other crops, anduncroppedplotswhenA.glycines-infested soybeanwereplaced in these habitats. They were Aphelinus asychisWalker,Aphelinus sp. of theA. varipes complex (Hyme-noptera: Aphelinidae), Aphidius ervi Haliday, Aphidiuscolemani Viereck, Binodoxys kelloggensis Pike, Stary� &Brewer,Diaeretiella rapae (MÕIntosh), Ephedrus sp., Ly-siphlebus testaceipes Cresson, Praon sp. (Hymenop-tera: Braconidae), Aphidoletes aphidimyza Rondani(Diptera: Cecidomyiidae), Allograpta obliqua Say,Eupeodes americanus Wiedemann, Eupeodes voluc-ris Osten Sacken, Sphaerophoria contigua Macquart,Syrphus sp., Toxomerus marginatus Say, Paragus hem-orrhous Meigen (Diptera: Syrphidae), and LeucopisglyphinivoraTanasijtshuk(Diptera:Chamaemyiidae).D. rapae and a few specimens within the Praon genuswere recovered only in the soybean system treatment.One species (P. hemorrhous) was not found on soy-bean but was detected in very low numbers when theA. glycines-infested soybean was placed in other cropsand noncropped plots (Figs. 1 and 2). Kaiser et al.(2007) summarized the previous literature on host

Fig. 5. Relative recoveries of B. kelloggensis using the A. glycines-infested soybean method at the Kellogg large main site(left) and small biodiversity site (right), Hickory Corners, MI, during years of signiÞcant differences among plant systemtreatments and sampling dates. Different letters next to the plant system legend or under the dates indicate signiÞcantdifferences (� � 0.05) in these factors when there was no interaction detected. Error bars are SEs. Ø, no data collected.


aphid and plant habitat ranges for the soybean aphidparasitoids and predatory ßies encountered in thisstudy.

ANOVA detected signiÞcant variability in naturalenemy recoveries using A. glycines-infested soybean

across plant system treatments, either as a main effector in interaction with sampling dates. Differenceswere detected at the Kellogg sites for six species thataveraged greater than one recovery per pot during atleast one sampling date during 2004 and 2005 (Table

Fig. 6. Relative abundance of aphids and natural enemies from visual Þeld observations on soybean and other cropvegetation at the Kellogg large main site (left) and small biodiversity site (right), Hickory Corners, MI, 2004. Bars representdetections from soybean (s), corn (c), wheat (w), and alfalfa (a), and each bar is partitioned into contributions of the differenttaxa to the total overall aphid count per plant (top graphs), percentage of parasitism (middle graphs), and predatory ßy countper plant (bottom graphs). Error bars are standard errors. ScientiÞc names not listed in text: A. pisum, Acyrthosiphon pisumHarris; T. trifolii, Therioaphis trifolii (Monell); S. avenae, Sitobion avenue (F.); M. euphobiae, Macrosiphum euphobiae(Thomas); and R. maidis, Rhopalosiphum maidis (Fitch).


2). At the Michigan State site, two of these species andthree additional ones showed signiÞcant differences inrecoveries between soybean and buckthorn from Mayto October for both years of sampling (Table 3). Dueto their infrequent recovery, signiÞcant variability was

rare in the analyses for the other species, and thesespecies were not considered further.Habitat Affinity of Parasitoids. There were consid-

erable between-species contrasts in habitat afÞnitydetected which were relevant to biological control. L.

Fig. 7. Relative abundance of aphids and natural enemies from visual Þeld observations on soybean and other cropvegetation at the Kellogg large main site (left) and small biodiversity site (right), Hickory Corners, MI, 2005. Bars representdetections from soybean (s), corn (c), wheat (w), and alfalfa (a), and each bar is partitioned into contributions of the differenttaxa to the total overall aphid count per plant (top graphs), percentage of parasitism (middle graphs), and predatory ßy countper plant (bottom graphs). Error bars are SEs.


testaceipes was the most frequently recovered parasi-toid, with most detections occurring in May and againin September and October (Figs. 1 and 2). L. testa-

ceipeswas recovered fromA.glycines-infested soybeanin all plant system treatments with few differencesdetected at the Kellogg and Michigan State sites (Figs.

Fig. 8. Relative recoveries ofA. asychis using theA. glycines-infested soybean method at the Kellogg large main site (left)and small biodiversity site (right), Hickory Corners, MI, during years of signiÞcant differences among plant system treatmentsand sampling dates. Different letters on the bars indicate signiÞcant (� � 0.05) differences within the sampling date whena plant system by date interaction was detected. Different letters under the dates indicate signiÞcant sampling date differenceswhen there was no interaction detected. Error bars are SEs. Ø, no data collected.

Fig. 9. Relative recoveries of A. aphidimyza using the A. glycines-infested soybean method at the Kellogg large main site(left) and small biodiversity site (right), during years of signiÞcant differences among plant system treatments and samplingdates, Hickory Corners, MI. Different letters on the bars indicate signiÞcant (� � 0.05) differences within the sampling datewhen a plant system by date interaction was detected. Different letters next to the plant systems and under the dates indicatesigniÞcant differences in these factors when there was no interaction detected. Error bars are SEs. Ø, no data collected.


3 and 4a), indicating its ready ability to occupy manyhabitats in the landscape where soybean occurs. Incontrast, B. kelloggensis was most commonly recov-ered from A. glycines-infested soybean placed in theearly successional and poplar plots. B. kelloggensis re-covery was at times high in these treatments, but therewere no instances that this increased presence corre-sponded to an increased presence in the soybeantreatment (Fig. 5). There were few detections of B.kelloggensis at the Michigan State site (Fig. 2), and nosigniÞcant differences in recoveries between soybeanand buckthorn were detected. In companion visualÞeld observations, parasitism by these and others spe-cies occurred in all cropped treatments (Figs. 6c andd and 7c and d), but unfortunately percentage of Þeldparasitism of A. glycines was very low in soybeanwhenever A. glycines occurred (Figs. 6b and 7b).A. asychis recoveries were quite variable, with a

tendency for it to occur more in the cropped treat-ments, including soybean, than in the noncroppedtreatments at the Kellogg sites (Fig. 8). There werefew detections at the Michigan State site (Fig. 2), andno signiÞcant differences in recoveries among treat-ments were detected. A. asychis contribution to par-asitism of A. glycines and other aphids observed in theplots was minor (Figs. 6cd and 7cd).Aphelinus sp. wasinfrequently recovered using A. glycines-infested soy-bean placed in all plant system treatments in Þeldobservations at the Kellogg site (Fig. 1), andAphelinussp. was found parasitizing aphids only on corn on onedate (Fig. 7c, mainly Rhopalosiphum padi L., personal

observations). Aphelinus sp. was detected more fre-quently in soybean than buckthorn in 2007 at theMichigan State site (Fig. 4b), attacking A. glycinesduring the sensitive soybean growth stages (Figs. 2aand 4b). A. colemani was infrequently recovered atthe Kellogg site (Fig. 1). A. colemani was found insoybean and not buckthorn after soybean matura-tion in October 2006 at the Michigan State site (Figs.2 and 4c).Habitat Affinity of Predatory Flies. In comparison

to parasitoids, predatory ßies as a group had muchbroader habitat afÞnity. A. aphidimyza was readilydetected on A. glycines-infested soybean placed insoybean and most other plant system treatments. Itwas limited only in vegetation of the early successionaland poplar treatments at the Kellogg site, and it wasdetected more in buckthorn than soybean in 2007 atthe Michigan State site (Figs. 4e and f and 9). A.obliqua tended to be detected across all plant systemsat the Kellogg sites, with only seasonal differencesnoted (Fig. 10). S. contigua occurred in low to modestnumbers at the Kellogg site, with few plant systemdifferences detected (Fig. 11). A. obliqua and S. con-tiguawere infrequentlydetectedat theMichiganStatesite (Fig. 2), and no signiÞcant differences in recov-eries between soybean and buckthorn were detected.E. americanus was rarely detected at the Kellogg site(Fig. 1). E. americanuswas the most common syrphidspecies found at the Michigan State site in 2007 andwas detected more in buckthorn than soybean (Fig.4d). In visual Þeld observations, these ßies were de-

Fig. 10. Relative recoveries ofA. obliquausing theA. glycines-infested soybean method at the Kellogg large main site (left)and small biodiversity site (right), during years of signiÞcant differences among plant system treatments and sampling dates,Hickory Corners, MI. Different letters on the bars indicate signiÞcant (� � 0.05) differences within the sampling date whena plant system by date interaction was detected. Different letters under the dates indicate signiÞcant sampling date differenceswhen there was no interaction detected. Error bars are SEs. Ø, no data collected.


tected in low numbers, rarely exceeding 0.2 ßy larvaeand pupae per plant, cumulative across species (Figs.6e and f and 7e and f). The uncommon visual Þeldobservations of A. glycines predation on soybean is incontrast with their widespread detection using A. gly-cines-infested soybean in many habitats. This differ-ence may have occurred because of egg and smalllarval stages that were easily missed in visual Þeldobservations. Predation of A. glycines by small pred-ators has been conÞrmed by molecular techniques(Harwood et al. 2007). Predatory ßies are well knownaphid predators in agricultural systems and havestrong host searching capability (Vockeroth 1992),which is consistent with our Þnding of common de-tections in the broader landscape.Relevance to Importation and Conservation Bio-logical Control. Expanding the traditional natural en-emy Þeld survey in the threatened soybean crop intothe broader landscape substantiated that L. testaceipeswaswell suited topreyuponA.glycines. Itswidespreadoccurrence in the landscape and its well documentedbroad host aphid range (Kaiser et al. 2007) indicatepotential for it to bridge across habitats and into soy-bean when it is present. This is desirable given the hostplant limitations of soybean as the one annual host forA. glycines. But its seasonal ßuctuations and reducedpresence mid-season are problematic for biologicalcontrol (Figs. 1a and 2a). The data lend support of L.testaceipes as a candidate for conservation biologicalcontrol, especially targeting increases in early andmid-season abundance (Fig. 2a). In contrast, the

strong habitat afÞnity of B. kelloggensis points to a lowlikelihood of it usingA. glycines on soybean. PlacingA.glycines-infested soybean in the broader landscapewas essential in this case to distinguish the capabilityofB. kelloggensis to parasitizeA. glycines from its likelyecologically low capability to contribute to suppres-sion of A. glycines on soybean. Although longer termadaptation resulting in aphid host and habitat expan-sion cannot be discounted (Cornell and Hawkins1993), the low recoveries seasonally and across hab-itats of A. asychis, Aphelinus sp., and A. colemonimaketheir ecological suitability to A. glycines suspect, cer-tainly in regard to biological control decision-makingthat often occurs during the early invasion phase ofpest species in contemporary U.S. agriculture (VanDriesche and Bellows 1996). The widespread occur-rence of many of the ßies reßects the broad host aphidranges of species detected (Kaiser et al. 2007). Giventheir common presence in the landscape, there ispotential for conservation biological control effortstargeting this group to contribute to A. glycines sup-pression. We advise conservation biological controlnot be attempted in isolation of classical biologicalcontrol, as the current parasitism and predation fromall species detected in this study were unable to retardA. glycines in soybean from exceeding a common eco-nomic threshold of 250 aphid per plant (Ragsdale et al.2007) at the Kellogg sites in 2003 and 2005. These wereyears when A. glycines infestation was commonthroughout the regionof this study(NomaandBrewer2008).

Fig. 11. Relative recoveries of S. contigua using the A. glycines-infested soybean method at the Kellogg large main site(left) and small biodiversity site (right), during years of signiÞcant differences among plant system treatments and samplingdates, Hickory Corners, MI. Different letters on the bars indicate signiÞcant (� � 0.05) differences within the sampling datewhen a plant system by date interaction was detected. Different letters under the dates indicate signiÞcant sampling datedifferences when there was no interaction detected. Error bars are SEs. Ø, no data collected.


Overall, the new information on habitat afÞnityfrom A. glycines-infested soybean placed in thebroader landscape arguably did not change the inter-pretation related to classical biological control derivedfrom observations in soybean alone in this case exam-ple (Heimpel et al. 2004). The consistent low impacton A. glycines on soybean based on observed parasit-ism and predation combined with the habitat afÞnityinformation derived fromA. glycines-infested soybeansuggest that parasitoids and predatory ßies are notlikely to contribute substantially to A. glycines sup-pression on soybean, at least during the conventionaltime period early in the pest invasion when classicalbiological control activities are considered in contem-porary U.S. agriculture.

But the differences in range of habitat afÞnity be-tween and within parasitoid and predator taxa ob-tained using the A. glycines-infested soybean was sub-stantial. It added plant system speciÞc informationrelevant to soybean production to the known prey andhabitat range of these parasitoids and predatory ßiesfound in the literature (Kaiser et al. 2007). Expandingthe natural enemy survey beyond the threatened soy-bean crop added speciÞc information relevant to con-servation biological control for species found acrossplant systems (i.e., L. testaceipes and A. aphidimyza).In contrast, another species (i.e., B. kelloggensis) wascapable of parasitizingA. glycinesbut was probably toohabitat restricted to contribute to A. glycines suppres-sion in soybean regardless of habitat management ef-forts. The additional information gained from ex-panding the natural enemy survey into the broaderlandscape was essential in making these distinctionsrelevant to conservation biological control, whileadding vigor and agroecosystem-speciÞc informa-tion valuable to classical biological control decision-making.

Acknowledgments

We are grateful to Michigan State University students M.Kaiser, S. Langley, A. Jenkins, J. Natzke, M. Minakawa, andP. Thomas for assistance during this study. We also thankDrew Corbin and the research leaders of the Kellogg Bio-logical Station Long Term Ecological Research project foraccess to the large and small plots, as supported by theNational Science Foundation Long-Term Ecological Re-search Program. Additional support for this study was pro-vided by the Michigan Agricultural Experiment Station andthe USDA CSREES RAMP (2004-51101-02210).

References Cited

Blackman, R. L., and V. F. Eastop. 2000. Aphids on theworldÕs crops: an identiÞcation and information guide.Wiley, New York.

Brewer, M. J., D. J. Nelson, R. G. Ahern, J. D. Donahue, andD. R. Prokrym. 2001. Recovery and range expansion ofparasitoids (Hymenoptera: Aphelinidae and Braconidae)released for biological control of Diuraphis noxia (Ho-moptera: Aphididae) in Wyoming. Environ. Entomol. 30:578Ð588.

Chacon, J. M., D. A. Landis, and G. E. Heimpel. 2008. Po-tential for biotic interference of a classical biological

control agent of the soybean aphid. Biol. Control 46:216Ð225.

Cornell, H. V., and B. A. Hawkins. 1993. Accumulation ofnative parasitoid species on introduced herbivores: acomparison of hosts as natives and hosts as invaders. Am.Nat. 141: 847Ð865.

Costamagna, A. C., D. A. Landis, and M. J. Brewer. 2008.The role of natural enemy guilds in Aphis glycines sup-pression. Biol. Control 45: 368Ð379.

Fehr, W. R., and C. E. Caviness. 1977. Stages of soybeandevelopment. SR 80. Iowa State University, Ames, IA.

Fox, T. B., D. A. Landis, F. F. Cardoso, and C. D. DiFonzo.2004. Predators suppress Aphis glycinesMatsumura pop-ulation growth in soybean. Environ. Entomol. 33: 608Ð618.

Gagne, R. J. 1981. Cecidomyiidae, pp. 257Ð292. In J. F.McAlpine, B. V. Peterson, G. E. Shewell, H. J. Teskey, J. R.Vockeroth, and D. M. Wood [eds.], Manual of NearcticDiptera. Agriculture Canada, Ottawa, Ontario, Canada.

Gao, J. 1994. Biological characteristics and control effect ofLysiphlebia japonica (Hym.: Braconidae) on Aphis gly-cines (Hom.: Aphididae). Chin. J. Biol. Control 10: 91Ð92.

Harwood, J. D., N. Desneux, H.J.S. Yoo, D. L. Rowley, M. H.Greenstone, J. J. Obrycki, and R. J. O’Neil. 2007. Track-ing the role of alternative prey in soybean aphid preda-tion byOrius insidiosus: a molecular approach. Mol. Ecol.16: 4390Ð4400.

Heraty, J.M., J. B.Woolley, K. R.Hopper, D. L.Hawks, J.-W.Kim, and M. Buffington. 2007. Molecular phylogeneticsand reproductive incompatibility in a complex of crypticspecies of aphid parasitoids. Mol. Phylogenet. Evol. 45:480Ð493.

Heimpel, G. E., D. W. Ragsdale, R. Venette, K. R. Hopper,R. J. O’Neil, C. E. Rutledge, and Z. Wu. 2004. Prospectsfor importation biological control of the soybean aphid:anticipating potential costs and beneÞts. Ann. Entomol.Soc. Am. 97: 249Ð258.

Heimpel, G. E., L. E. Frelich, D. A. Landis, K. R. Hopper, K.Hoelmer, Z. Sezen, M. K. Asplen, and K. Wu. 2010. Eu-ropean buckthorn and Asian soybean aphid as compo-nents of an extensive invasional meltdown in North Amer-ica. Biol. Invasions (DOI: 10.1007/s10530-010-9736-5).

Heiss, E. M. 1938. A classiÞcation of the larvae and pupariaof the Syrphidae of Illinois exclusive of aquatic forms. Ill.Biol. Monogr. 16: 1Ð142.

Kaiser, M. E., T. Noma, M. J. Brewer, K. S. Pike, J. R. Voc-keroth, and S. D. Gaimari. 2007. Hymenopteran parasi-toids and dipteran predators found using soybean aphidafter its midwestern United States invasion. Ann. Ento-mol. Soc. Am. 100: 196Ð205.

Kellogg Biological Station. 2006. Experimental design.(http://lter.kbs.msu.edu/experimentalDesign.htm,http://www.lter.kbs.msu.edu/datatables/40).

Littell, R. C., R. J. Freund, and P. C. Spector. 1991. SASsystem for linear models, 3rd ed. SAS Institute, Cary, NC.

Liu, J., K. Wu, K. R. Hopper, and K. Zhao. 2004. Populationdynamics of Aphis glycines (Homoptera: Aphididae) andits natural enemies in soybean in northern China. Ann.Entomol. Soc. Am. 97: 235Ð239.

Milne, W. M. 1995. Use of trap plants as a means of mea-suring the activity of cereal aphid parasitoids in the Weld.Agric. Ecosyst. Environ. 52: 31Ð34.

Mosteller, F., and J. W. Tukey. 1977. Data analysis and re-gression: a second course in statistics. Addison-Wesley,Reading, MA.

Nielsen, C., and A. E. Hajek. 2005. Control of invasive soy-bean aphid, Aphis glycines (Hemiptera: Aphididae), pop-ulations by existing natural enemies in New York state,


with emphasis on entomopathogenic fungi. Environ. En-tomol. 34: 1036Ð1047.

Noma, T., and M. J. Brewer. 2008. Seasonal abundance ofresident parasitoids and predatory ßies and correspond-ing soybean aphid mortality, with comments on classicalbiological control of soybean aphid in the U.S. Midwest.J. Econ. Entomol. 101: 278Ð287.

Noma, T., C. Gratton, M. Colunga-Garcia, M. J. Brewer,E. E. Mueller, K.A.G. Wyckhuys, G. E. Heimpel, andM. E. O’Neal. 2010. Relationship of soybean aphid(Hemiptera: Aphididae) to soybean plant nutrients, land-scape structure, and natural enemies. Environ. Entomol.39: 31Ð41.

Pike, K. S., P. Stary, M. J. Brewer, T. Noma, S. Langley, andM.Kaiser. 2007. A new species ofBinodoxys (Hymenop-tera: Braconidae: Aphidiinae), parasitoid of the soybeanaphid, Aphis glycines Matsumura, with comments on bio-control. Proc. Entomol. Soc. Wash. 109: 359Ð365.

Ragsdale, D. W., D. J. Voegtlin, and R. J. O’Neil. 2004. Soy-bean aphid biology in North America. Ann. Entomol. Soc.Am. 97: 204Ð208.

Ragsdale,D.W.,B.P.McCornack,R.C.Venette,B.D.Potter,I. V.Macrae, E.W.Hodgson,M.E.O’Neal, K.D. Johnson,R. J.O’Neil,C.D.DiFonzo,T.E.Hunt, P.A.Glogoza, andE. M. Cullen. 2007. Economic threshold for soybean

aphid (Hemiptera: Aphididae). J. Econ. Entomol. 100:1258Ð1267.

Rutledge, C. E., and R. N. Wiedemann. 1999. Habitat pref-erences of three congeneric braconid parasitoids: impli-cations for host-range testing in biological control. Biol.Control 16: 144Ð154.

Rutledge, C. E., R. J. O’Neil, T. B. Fox, and D. A. Landis.2004. Soybean aphid predators and their use in inte-grated pest management. Ann. Entomol. Soc. Am. 97:240Ð248.

SAS Institute. 2004. SAS/STAT userÕs guide, version 8. SASInstitute, Cary, NC.

Van Driesche, R. G., and T. S. Bellows, Jr. 1996. Biologicalcontrol. Chapman & Hall, New York.

Vennette, R. C., and D. W. Ragsdale. 2004. Assessing theinvasion by soybean aphid (Homoptera: Aphididae):where will it end? Ann. Entomol. Soc. Am. 97: 219Ð226.

Vockeroth, J. R. 1992. The ßower ßies of the subfamily Syr-phinae of Canada, Alaska, and Greenland. Diptera: Syr-phidae. The insects and arachnids of Canada Part 18.Agriculture Canada, Ottawa, Ontario.

Wu, Z., D. Schenk-Hamlin, W. Zhan, D. W. Ragsdale, andG. E. Heimpel. 2004. The soybean aphid in China: ahistorical review. Ann. Entomol. Soc. Am. 97: 209Ð218.

Received 6 October 2009; accepted 19 February 2010.


Documents

B M C Habitat Afﬁnity of Resident Natural Enemies …ccag.tamu.edu/files/2012/03/2010-brewernomaJEEsbabio...BIOLOGICAL AND MICROBIAL CONTROL Habitat Afﬁnity of Resident Natural