Trait-mediated indirect interactions in invasions: unique ... Reprints/Savage and Whitney... · Trait-mediated indirect interactions in invasions: unique behavioral responses of an

Trait-mediated indirect interactions in invasions:unique behavioral responses of an invasive ant to plant nectar

AMY M. SAVAGE� AND KENNETH D. WHITNEY

Department of Ecology and Evolutionary Biology, Rice University, 6100 Main Street, Houston, Texas 77005 USA

Citation: Savage, A. M., and K. D. Whitney. 2011. Trait-mediated indirect interactions in invasions: unique behavioral

responses of an invasive ant to plant nectar. Ecosphere 2(9):106. doi: 10.1890/ES11-00145.1

Abstract. Exotic species often form beneficial, facultative associations with indigenous species.

However, we still have a limited understanding of the influences that these positive associations may

have on the dynamics and impacts of species invasions. Highly invasive species may respond differently

than less invasive species to resources that are exchanged in mutualisms, leading to trait-mediated indirect

interactions between native species via invaders that may reshape native communities. In this study, we

tested the hypothesis that the highly invasive ant species, Anoplolepis gracilipes, exhibits stronger trait

changes in response to increasing levels of nectar than co-occurring, less invasive ant species. Across two

islands in the Samoan Archipelago, we located multiple sites dominated by A. gracilipes and multiple sites

dominated by other, less invasive species. At each site, we manipulated nectar levels on a common

extrafloral nectary-bearing shrub and assessed short-term changes in ant worker recruitment and

aggression. We found that the recruitment response of the highly invasive ant species A. gracilipes was not

unique: other dominant ant species also increased recruitment in response to increasing nectar levels.

However, A. gracilipes did show unique changes in aggressive behaviors: as nectar levels increased, the

proportion of prey discovered, attacked and removed by A. gracilipes workers and the speed at which they

performed these aggressive behaviors all increased strongly. Other ant species showed no such responses.

In addition, fewer subordinate ants persisted on plants at sites invaded by A. gracilipes. Finally, plot-level,

simultaneous manipulations of ant access to the plants and nectar availability demonstrated that Morinda

citrifolia-ant mutualisms influenced the b-diversity of local arthropod communities differently when A.

gracilipes dominated local ant assemblages. These results suggest that mutualisms between invasive ants

and native plants can modify interactions between invaders and co-occurring arthropods, possibly leading

to more negative consequences for native communities. They also underscore the importance of

incorporating both positive species interactions and indirect pathways into our studies of both community

ecology and invasion biology.

Key words: aggression; Anoplolepis gracilipes; ant-plant mutualisms; forager recruitment; invasive species; Morinda

citrifolia; nectar; Samoa; trait-mediated indirect interactions.

Received 18 May 2011; revised 22 July 2011; accepted 26 July 2011; final version received 23 August 2011; published 30

September 2011. Corresponding Editor: O. Schmitz.

Copyright: � 2011 Savage and Whitney. This is an open-access article distributed under the terms of the Creative

Commons Attribution License, which permits restricted use, distribution, and reproduction in any medium, provided

the original author and sources are credited.

� E-mail: [email protected]

INTRODUCTION

Although traditionally studied as pairwise

associations, interspecific mutualisms involve

interactions among diverse assemblages in com-

plex, multispecies communities. The outcome of

mutualisms can depend strongly upon the

identity of the interacting partners (Bronstein

1994). For example, Acacia drepanolobium trees

associate with several species of ant bodyguards;

v www.esajournals.org 1 September 2011 v Volume 2(9) v Article 106

these ant species not only differ in the costs andbenefits provided to host trees, but also influenceplant host fitness most strongly at differentphenological stages. Consequently, the fitness ofA. drepanolobium plants is highest when theplants associate with multiple ant speciesthroughout their lifetime (Palmer et al. 2010).Significant variation in the costs and benefitsassociated with different mutualist partners hasbeen reported for other arthropod-plant protec-tion mutualisms (Rudgers and Strauss 2004,Whitney 2004, Miller 2007), mycorrhizal mutu-alisms (Maherali and Klironomos 2007, Manganet al. 2010), pollination mutualisms (Gomez et al.2010), and seed dispersal mutualisms (Whitney2005, Manzaneda and Rey 2008).

Because mutualisms are embedded in complexcommunities, variation in partner identity mayhave significant consequences for communitystructure and function. The community-wideconsequences of mutualisms are still poorlyunderstood. However, recent studies have dem-onstrated that the abundance, diversity, andcomposition of local communities can be stronglyinfluenced by the presence of mutualistic inter-actions (Stachowicz 2001, Bruno et al. 2003, Hayet al. 2004, Rudgers et al. 2007, Lach 2008,Matthews et al. 2009). Furthermore, variation inthe degree and type of benefits exchanged inmutualisms can have cascading effects on thestructure and dynamics of communities. Forexample, Rudgers et al. (2010) demonstrated thatgeographic variation in the benefits provided toplants by ant guards was associated withsignificant differences in the composition andabundance of plant-associated arthropods thatwere not directly involved in the mutualism.

We have a limited understanding of themechanisms that underlie these community-levelresponses to variation in mutualist partneridentity. However, interaction modifications viatrait-mediated indirect pathways likely influencethese differential effects of mutualist partners onecological communities. These effects occur whenthe presence of one species modifies one or moretraits of another species, with consequent effectson other community members through interac-tion modification (Werner and Peacor 2003). Forexample, Bernot and Turner (2001) demonstratedthat periphyton-consuming Physa integra snailssought refuge in the presence of predacious fish,

but not predacious crayfish. Consequently, pe-riphyton abundance was as much as 110% higherin mesocosms with fish than those with crayfish.Similar trait-mediated community-level shiftscould occur when different species are involvedin mutualistic associations.

Species invasions provide ideal systems forinvestigating the importance of species identityin ecological interactions. Invasive species oftendisplay exaggerated traits compared to nativespecies (e.g., more rapid growth rates; Pysek andRichardson 2007, van Kleunen et al. 2010), andthese traits may influence co-occurring commu-nity members. Additionally, novel interactionsbetween invaders and native species provide acontrast to the same associations involving nativespecies that share a long history of interaction.When exotic species invade novel habitats, theymay respond differently than native species tothe resources or services exchanged in mutual-isms. Moreover, it is likely that different exoticspecies differ in their responses to mutualist-derived resources, although such a pattern hasnever been explicitly described, to our knowl-edge. If a highly invasive species displaysdifferent traits (e.g., responses to carbohydrateavailability, see below) when it displaces nativeor less invasive species in facultative mutualismsand this trait change results in community-leveleffects, then trait-mediated pathways may beimportant mechanisms underlying the detrimen-tal impacts of species invasions. However,surprisingly few studies have investigated therole of trait-mediated indirect interactions ininvasions. White et al. (2006) recently surveyedthe ecological literature and found only twostudies that experimentally demonstrated that (1)novel positive interactions between an invaderand a native species influenced co-occurringcommunity members via interaction modifica-tion and (2) these effects were trait-mediated. Inboth cases, the presence of an invasive plantaltered pollinator visits to native plant species(Grabas and Laverty 1999, Chittka and Schurk-ens 2001). Since non-native species frequentlyform positive associations with native species(Bruno et al. 2003), we could be underestimatingthe importance of indirect, trait-mediated mech-anisms in driving the dynamics and impacts ofspecies invasions.

In this study, we investigated positive interac-


SAVAGE AND WHITNEY

tions between ants and extrafloral nectary (EFN)-bearing plants in Samoa. Like many other islandecosystems, Samoa is dominated by non-nativeant species (Wetterer and Vargo 2003), inparticular by the highly invasive species Anoplo-lepis gracilipes. A pan-tropical ‘tramp ant,’ A.gracilipes has strong negative impacts on nativeisland flora and fauna in the Pacific and IndianOceans (Holway et al. 2002, Hill et al. 2003,Lester and Tavite 2004, Abbott 2006, Savage et al.2009). Anoplolepis gracilipes invasions are hypoth-esized to be driven by subsidies from carbohy-drate-excreting plants and insects (Holway et al.2002, Davidson et al. 2003, Lach 2003). In supportof this hypothesis, we observed that variation inthe abundance of EFN-bearing plants was posi-tively associated with the abundances of A.gracilipes across the Samoan Archipelago, butwas unrelated to the abundances of other exoticant species (Savage et al. 2009). However, theprecise mechanisms underlying the invasiveimpacts of A. gracilipes are poorly understood.In a recent experiment, we demonstrated that A.gracilipes responded to increasing levels of plantnectar by both recruiting strongly to EFN-bearing plants and decreasing tending behaviorsof honeydew excreting insects (Savage et al.2011). These results, together with our earliersurveys across the Samoan Archipelago, suggest-ed the hypothesis that unique responses of A.gracilipes to nectar might underlie the particularlystrong negative impact of A. gracilipes onarthropod communities.

Carbohydrate subsidies may facilitate antinvasions through both density and trait-mediat-ed indirect pathways. Here, we focus on theprediction that carbohydrate-rich, mutualist-de-rived resources (such as plant nectar) influencethe behaviors of highly invasive ants morestrongly than the behaviors of co-occurring lessinvasive ants. Nectar availability in the Pacificvaries both naturally (Savage et al. 2009) and inresponse to human activities, e.g., the creation ofplantations of Morinda citrifolia, an economicallyimportant EFN-bearing plant (Potterat and Ham-burger 2007). Therefore, we examined predic-tions of the carbohydrate subsidy hypothesisacross a gradient of nectar availability. Specifi-cally, we hypothesized two separate, but non-mutually exclusive, trait-mediated indirect path-ways whereby these interactions could influence

arthropod community structure. First, we hy-pothesized that (1) as nectar resources wereexperimentally increased, the highly invasiveant species A. gracilipes would recruit moreworkers to EFN-bearing plants than other non-native ant species. We expected that increased A.gracilipes recruitment would be associated withincreased activity outside the nest and increasednegative impacts on co-occurring arthropods (viapredation and/or the threat of predation) relativeto less invasive ant species (dotted lines, Fig. 1A).Second, we hypothesized that (2) as nectarresources were increased, highly invasive A.gracilipes workers would become more aggres-sive toward co-occurring arthropods than lessinvasive ants. Increased levels of aggressioncould also lead to stronger negative communi-ty-wide impacts of A. gracilipes relative to lessinvasive ants (dashed lines, Fig. 1A). Importantly,both of these pathways are not dependent uponchanges in the density of invasive ants throughincreased colony growth (Oliver et al. 2008),because they require only a change in theallocation of workers to tasks (nectar foragingvs. other behaviors) or a change in workeraggression, respectively. Thus, we consider bothof these hypothesized indirect interactions be-tween nectar-producing plants and the arthro-pod community to be trait-mediated. Finally, wepredicted (3) that changes in partner identitywithin M. citrifolia-ant mutualisms (A. gracilipesvs. all other ant species) would cascade to affectco-occurring arthropods, leading to reductions incommunity diversity when A. gracilipes is present(Fig. 1A, solid arrows).

METHODS

Study sitesWe conducted nectar manipulation experi-

ments from 16 June 2009 to 26 September 2009on the islands of Tutuila, American Samoa andSavaii, Independent Samoa (Table 1). In aprevious survey spanning six islands of theSamoan archipelago, we found that A. gracilipeswere least abundant and widespread in Tutuilaand most abundant and widespread in Savaii(Savage et al. 2009). We selected eight sitesdominated by A. gracilipes and six sites dominat-ed by other ant species (with A. gracilipes absent).Data on foraging recruitment from three of the 14


SAVAGE AND WHITNEY

sites were used in a previous study of an A.gracilipes invasion front in northeastern Savaii(Savage et al. 2011). On Savaii, sites wereseparated by 150 m–66.7 km (mean 27.12 6

9.46 km); sites on Tutuila were separated by 150m–35.1 km (mean 16.11 6 3.90 km). We do nothave data on average foraging distances for A.gracilipes or co-occurring ant species in Samoa.However, average foraging distances for A.gracilipes in Australia are ;30 m (Ben Hoffmann,personal communication) indicating that foragersobserved in different sites were likely fromdifferent colonies.

Because it would be logistically difficult andunethical to experimentally manipulate the iden-tity of the dominant ant species at a site (e.g.,adding a highly invasive ant species to a

previously uninvaded site), there is the risk thatobserved differences in ant species responses tonectar manipulations are confounded by differ-ent environmental conditions at sites dominatedby different species. To partially address thisconcern, we examined spatial autocorrelation ofant species responses (forager recruitment andper-worker aggressive responses to experimen-tally manipulated nectar levels, see below). Weconstructed Euclidean distance matrices of (1)geographic distances between sites and (2)dissimilarities between relevant test statisticsper site per response variable. We used betasfrom regressions for the forager recruitmentresponses and chi square values for aggressiveresponses. We then conducted Mantel tests usingthe RELATE function in Primer v. 6.1.10 (Clarke

Fig. 1. Interaction web diagrams depicting (A) hypothesized and (B) actual trait-mediated indirect interactions

(TMII) between plant nectar and herbivores. Solid lines represent direct interactions and dotted lines represent

TMII. Dotted lines on the left side of the diagrams are associated with Anoplolepis gracilipes and those on the right

side are associated with other dominant ant species. Thicker lines represent stronger effects. (A) We predicted

that nectar would affect traits of both highly invasive and less-invasive dominant ant species similarly, but with

stronger effects for highly invasive ant species. (B) Actual patterns detected in our experiments. Both A. gracilipes

and other dominant ant species responded to plant nectar by increasing forager recruitment, but A. gracilipes was

the only species to increase aggression in response to increasing nectar levels.


SAVAGE AND WHITNEY

and Gorley 2007) with Spearman rank correla-tions and 9,999 iterations. We conducted theseanalyses both across all sites and for sites withineach island (for a total of three tests per responsevariable). We interpret a lack of spatial autocor-relation in these responses as an indication thatwe are measuring differences in ant speciesbiology as opposed to differences in environ-mental conditions.

Additionally, due to natural abundance pat-terns, 75% of the sites (six of eight) dominated byA. gracilipes were located on Savaii and 67% ofthe sites (four of six) dominated by other antspecies were located on Tutuila (Table 1), leadingto potential island effects. Therefore, we con-ducted additional tests to examine whether antresponses to our experimental manipulations ofnectar availability (described below) differedbetween islands. For forager recruitment data,we used ANOVAR; for proportional aggression

data, we used logistic regression; and for data onthe time it took ants to perform differentaggressive responses, we used survival analyses.All tests were conducted for both A. gracilipesand for other dominant ant species; modelsincluded island and treatment as factors. Anoplo-lepis gracilipes responses did not significantlydiffer on Savaii versus Tutuila (all P . 0.25),nor did the responses of other ant species differbetween the islands (all P . 0.10). These patternssuggest that, while island and dominant antidentity are partially confounded, island effectsare not sufficiently large to bias our findings.

During the time of the experiments, averagemonthly rainfall was 0.1325 mm (60.19 mm) andaverage daily temperatures ranged between27.088C and 29.868C in Tutuila (Pago PagoWeather station); temperatures ranged between25.838C and 29.178C with no measurable rainfallin Savaii (Avao Weather Station).

Table 1. Descriptions of ant assemblages at sites on the islands of Savaii (S) in Independent Samoa and Tutuila (T)

in American Samoa in 2009. Invasiveness scores (see Table 2 for detailed description) are reported as

percentages of the total possible score. Densities are reported from surveys of ants on the ground and on

Morinda citrifolia plants (mean 6 SE). Dominant species were defined as the species at each site with the highest

numerical abundance. Subordinate species included all species except the numerically dominant species.

SiteDates of

experimentsDominantspecies

Invasive-ness

score (%)

Density ofdominant species

Density of allsubordinate species

Ground M. citrifolia Ground M. citrifolia

Falealupo (S) September Anoplolepisgracilipes

100 23.5 6 6.36 46.50 6 8.66 1.00 6 0.58 2.00 6 2.16

Mauga_West (S) June Anoplolepisgracilipes

100 21.25 6 2.72 46.50 6 4.80 2.00 6 0.91 4.75 6 1.65

Mauga_South (S) September Anoplolepisgracilipes

100 38.75 6 1.32 53.21 6 4.65 0 0

Saleaula_East (S) June Anoplolepisgracilipes

100 28.0 6 1.47 39.25 6 2.14 2.50 6 1.04 1.75 6 0.63

Saleaula_North (S) September Anoplolepisgracilipes

100 17.5 6 2.96 33.00 6 4.14 0 1.00 6 0.41

Saleaula_South (S) September Anoplolepisgracilipes

100 15.38 6 3.99 29.44 6 4.37 0.22 6 0.22 2.33 6 1.0

Futiga (T) August Anoplolepisgracilipes

100 12.50 6 3.23 15.13 6 3.59 3.23 6 0.48 1.38 6 0.63

Masausi_East (T) August Anoplolepisgracilipes

100 8.13 61.03 36.50 6 5.17 0 2.00 6 0.71

Mauga_North (S) September Pheidolemegacephala

83 9.40 6 1.60 11.32 6 0.98 2.31 6 0.86 5.61 6 2.15

Fagatogo (T) July Pheidolemegacephala

83 6.75 6 3.90 8.25 6 2.78 0.25 6 0.25 2.78 6 4.42

Tafuna (T) July Solenopsisgeminata

61 12.25 6 1.75 10.50 6 3.07 1.50 6 0.65 2.50 6 1.19

Gataivai (S) September Paratrechinalongicornis

33 15.75 6 5.11 21.75 6 4.53 8.75 6 5.91 10.0 6 4.08

Masausi_West (T) August Tapinomamelanocephlum

22 11.25 6 3.01 56.25 6 4.00 3.25 6 2.36 3.40 6 1.68

IlliIlli (T) July Tetramoriumbicarinatum

11 2.25 6 0.63 0.75 6 0.48 7.00 6 1.58 4.50 6 2.53


SAVAGE AND WHITNEY

Study organismsNumerically dominant ant species at our study

sites were all exotic and varied their invasiveness(both on EFN-bearing plants and the ground;Table 1). Using the methods of Ward et al. (2008),we calculated invasiveness scores based uponliterature accounts of the biological traits associ-ated with invasiveness and the ecological im-pacts of the invaders. We report each species’score as a percentage (species score/maximumpossible score 3 100). Scores ranged from 11%(Tetramorium bicarinatum) to 100% (A. gracilipes)(Table 2).

Across the Samoan Archipelago, ants haveaccess to multiple resources for carbohydrates,the most common and abundant being theextrafloral nectary (EFN)-bearing plant, Morindacitrifolia. While the timing and nature of arrival ofM. citrifolia to islands throughout the Pacific areunclear (i.e., natural vs. Polynesian-mediateddispersal ;3000 years ago; Whistler 1993, Nelson2006), we consider it to be native. Morindacitrifolia possesses annular disk nectaries clus-tered on inflorescences (hereafter ‘nectary body’;Waki et al. 2007, Fig. 2). At our sites in Samoa,nectary bodies contained 2 to .50 nectaries andreached a maximum size of 80 cm3. Morindacitrifolia plants produce nectary bodies year-round. Ants frequently visit nectaries of thisplant, which are often dominated by the highlyinvasive ant species, A. gracilipes (Savage et al.2009).

Characterization of the ant assemblage at each siteAt each site, we first assessed local ant

assemblages by haphazardly selecting three M.citrifolia plants and counting the number andidentity of ants on each plant. We used acomprehensive count of all ants per plantcompleted in ;5 minutes/plant. We also countedants on the ground ;1 m from each plant, usinga 10 3 10 cm white paper card and counting thenumber of ants that crossed the card in 30 sec(following methods in Abbott 2005). The cardswere laminated with plastic and these assess-ments were conducted on different days (at least2 days between counts). Therefore, it is unlikelythat any pheromones were absorbed by thecards. However, even if this occurred, anypheromones likely dissipated by the time thatcards were reused. We used the card approach

because ;60% of all sites were located on lavafields, making it impossible to use pitfall traps.We also collected foragers at tuna and honeybaits at all sites. However, these responses werenot substantially different from card counts. Weopted to use card counts here, because they areless likely to be influenced by dominancehierarchies than resources.

Experimental designWe established 5 m 3 5 m plots of plants at

each site, focusing on areas dominated by M.citrifolia. Depending upon the size of the site andthe naturally-occurring densities of M. citrifoliaplants, we established 3–5 plots per site. Plotswere separated from each other by ;8 m toincrease the independence of observations. With-in each plot, we haphazardly selected five M.citrifolia plants that were similar in size andwithin 1m of each other to be used for nectarmanipulations. We first counted the number ofnectary bodies, aphids (Aphididae), scale insects(Coccididae), and mealybugs (Pseudococcididae)on each plant and measured the height of themain stem and the diameter of the base of eachplant to the nearest cm. Each plant was thenrandomly assigned to a nectar availability treat-ment (see below).

Hypothesis 1: As nectar levels increase, highlyinvasive ants will recruit more workers to EFN-bearing plants than less invasive ants

We manipulated nectar availability at the plantscale: 0, 50, 100, 150, and 200% of ambient levelsper plant. To reduce access to nectar, we securedbags constructed of lightweight organza materialto the base of nectary bodies with a plastic cabletie. Bags were ;1.5–23 larger than nectary bodiesto minimize contact between nectaries andbagging material. We bagged all nectary bodies,regardless of treatment assignment, to control forany effects of the bags. However, in all treat-ments except 0%, we cut holes (;3–6 cm) in bagsto allow ants to access actual M. citrifolia nectar.Thus, for plants in the 50% treatment, we cutholes in half of the bags, and all bags had holes inthe 100%, 150% and 200% treatments.

We used artificial nectaries (Seal-Rite micro-centrifuge vials, USA Scientific 1605-0000, Ocala,Florida, USA, filled with 500 lL of a 30% sucrosesolution) to supplement nectar levels. In a


SAVAGE AND WHITNEY

previous study, Freeman et al. (1991) demon-

strated that M. citrifolia nectar is dominated by

sucrose, with sucrose contributing an average of

72.6–88.9% to total nectar carbohydrates. To

inform the design of our nectar availability

treatments, we assessed natural nectar produc-

tion and concentration for M. citrifolia. We

excluded insect visitors from nectary bodies for

24 h and collected nectar in microcapillary tubes.

Average nectar production per plant per day was

2249 lL 6 642 SE (range ¼ 645–5226, n ¼ 6

plants). The concentration of M. citrifolia nectar,

using a field refractometer (EZ-Red B1, EZ Red

Co., Deposit, New York), was 28.06% 6 1.04% SE

Table 2. Scorecard of the relative invasiveness of the dominant ant species used in this study. Scores are based

upon literature accounts, following the methods of Ward et al. (2008). Parenthetical numbers are the

invasiveness scores for each trait.

TraitAnoplolepisgracilipes

Pheidolemegacephala

Solenopsisgeminata

Paratrechinalongicornis

Tapinomamelanocephalum

Tetramoriumbicarinatum

Biological

Recruits in largenumbers

Yes1,2 (1) Yes1 (1) Yes3 (0.5) Sometimes1 (0.5) Sometimes4

(0.5)Sometimes5

(0.5)

Monopolizes food Yes1,6 (1) Yes1 (1) Sometimes3,7

(0.5)No1 (0) No8 (0) No5 (0)

Reproductivequeens

Polygyne2,9

(1)Polygyne10 (1) Both11,12 (0.5) Polygyne13 (1) Polygyne14,15

(1)Both16 (0.5)

Supercolonies withreducedintraspecificcompetition

Yes1,2,17 (1) Yes1 (1) Sometimes1,18

(0.5)No19 (0) Unknown (0.5) Sometimes20

(0.5)

Invasive elsewhere .2locations1,21

(1)

. 2 locations1

(1). 2 locations1

(1). 2 locations8

(1)No16 (0) No16 (0)

Subtotal 5/5 ¼ 100% 5/5 ¼ 100% 3.5/5 ¼ 70% 2.5/5 ¼ 50% 2/5 ¼ 40% 1/5 ¼ 20%Ecological impacts

Competitiveadvantage overother ants

Yes1,22 (1) Yes1,23 (1) Sometimes24,25

(0.5)Sometimes26,27

(0.5)Unlikely (0) Unlikely (0)

Detrimental impactson co-occurring(non-ant)invertebrates

Severe1,28 (1) Sometimes29,30,31

(0.5)Sometimes32

(0.5)No (0) No (0) No (0)

Detrimental impactson co-occurringvertebrates

Yes1,33,34 (1) No23,35 (0) Sometimes36

(0.5)No (0) No (0) No (0)

Harms plants Yes1,29,37 (1) Yes38,39 (1) Sometimes40

(0.5)No (0) No (0) No (0)

Subtotal 4/4 ¼ 100% 2.5/4 ¼ 62.5% 1/4 ¼ 25% 0.5/4 ¼ 12.5% 0/4 ¼ 0 0/4 ¼ 0

Total invasivenessscore

9/9 ¼ 100% 7.5/9 ¼ 83% 4.5/9 ¼ 50% 3/9 ¼ 33% 2/9 ¼ 22% 1/9 ¼ 11%

Note: Superscripted numbers refer to sources: 1, Holway et al. 2002; 2, Abbott 2005; 3, Perfecto 1991; 4, Lee 2002; 5, Morrison1996; 6, Drescher et al. 2010; 7, Greenslade 1971; 8, McGlynn 1999; 9, Thomas et al. 2010; 10, Fromhoff and Ward 1992; 11,McInnes and Tschinkel 1995; 12, Adams et al. 1976; 13, Yamauchi and Ogata 1995; 14, Bustos and Cherix 1998; 15, Harada 1990;16, Klotz et al. 2008; 17, Abbott 2006; 18, Hoffmann and O’Conner 2004; 19, Wetterer 2008; 20, Astruc et al. 2001; 21, Lowe et al.2000; 22, Wetterer et al. 2005; 23, Lester and Tavite 2004; 24, Hoffmann 2010; 25, Morrison 2000; 26, Plentovich et al. 2010; 27,Pfeiffer et al. 2008; 28, Wetterer et al. 1999; 29, O’Dowd et al. 2003; 30, Lawrence et al. 2011; 31, Hoffmann and Parr 2008; 32,Heterick 1997; 33, Plentovich et al. 2009; 34, Risch and Carroll 1982; 35, Lach and Thomas 2008; 36, Davis et al. 2008; 37,Plentovich et al. 2011; 38, Gaigher et al. 2011; 39, Handler et al. 2007; 40, Carroll and Risch 1984.


SAVAGE AND WHITNEY

(n ¼ 8 plants). Therefore, we considered the 30%sucrose solution in our artificial nectaries to be areasonable approximation of naturally-occurringM. citrifolia nectar. We inserted a 5 lL micro-capillary tube into the center of each micro-centrifuge vial to allow ants to access sucrose inthe artificial nectaries. Air bubbles occasionallyformed inside the microcapillary tubes. There-fore, all microcapillary tubes were cleared (re-moved and then re-inserted) ;15 minutes beforeeach census. Vials were affixed to plants usingtwist ties, and all plants received 10 vials. The 0,50, 100, 150 and 200% treatment levels contained0, 0, 0, 5 and 10 filled vials, respectively. Thus ineach experiment, we provided local ant assem-blages with 0, 0, 0, 2500 or 5000 ll of artificialnectar per plant over the course of 48hrs. Short-term pulses of nectar availability allowed us todisentangle trait- from density-mediated effectsof nectar on local arthropod assemblages, sinceant population growth responses to nectar wouldhave required much more time.

Response variables.—We conducted six antcensuses per plant: morning (;6:00–8:00), mid-afternoon (;12:00–14:00), and evening (;1600–18:00) over two consecutive days. All censusesoccurred during daylight hours due to Samoancultural restrictions. While this sampling schemeprovided a good estimate of the relative foragingrates for diurnal and crepuscular workers, it didnot account for the nocturnal activities of localant assemblages. Counts took ;5 minutes perplant. We collected specimens of each ant speciesfrom nearby non-treatment plants and identifiedthem using Wilson and Taylor (1967) andShattuck (1999).

Data analyses.—We compared ant responses atsites invaded by A. gracilipes versus sites whereA. gracilipes was absent. At each site, we assignedants to one of two dominance categories:dominant (the most abundant ant species persite) or subordinate (all other co-occurring ants;see Table 1). Although multiple metrics are usedto describe ant dominance (e.g., behavioral,numerical), we focused on numerical dominancein this study (hereafter, we refer to numericaldominance simply as dominance). Subordinateants at sites invaded by A. gracilipes includedspecies that were dominant at A. gracilipes-uninvaded sites and other species that weresubordinate both in the presence and absence of

A. gracilipes. We then conducted repeated mea-sures general linear models on ant density(number of workers per plant) following recom-mendations in von Ende (2001), with separatetests for sites invaded by A. gracilipes vs. notinvaded. Models included the following inde-pendent factors: ant dominance status (dominantor subordinate), nectar treatment (0, 50, 100, 150or 200% ambient levels), site, plot (nested in site),and all interactions with time (Proc GLM; SASInstitute 2003). Ant dominance status, site andplot were treated as categorical factors, and thenectar availability level was treated as a contin-uous factor. Plant size (height 3 diameter atbase), the number of nectary bodies per plant,and the abundance of honeydew excretinginsects per plant were used as covariates. Becausethe repeated factor (elapsed time) had nosignificant influence on ant responses to ourtreatments, we pooled data across time tosimplify data presentation. Statistical analysesmet assumptions of normality of residuals and

Fig. 2. A Morinda citrifolia nectary body, with florets,

extrafloral nectaries, and Anoplolepis gracilipes workers

labeled.


SAVAGE AND WHITNEY

homogeneity of variances at each time followingsquare-root transformation of ant density.

Finally, previous findings suggested that A.gracilipes invasions were associated with signifi-cant declines in co-occurring, plant-associatedant species (Savage et al. 2009). Therefore, wealso used t-tests to determine if the abundance ofsubordinate ants on M. citrifolia plants differedbetween sites invaded or not invaded by A.gracilipes.

Hypothesis 2: As nectar resources increase, highlyinvasive ants will become more aggressive towardco-occurring arthropods than less invasive ants

To examine the relative effects of nectar levelson ant aggression, we followed the recruitmenttrials (above) with aggression trials ;12 h afterthe last recruitment survey. We replenished allfilled tubes, so that each plant’s nectar availabilitywas at the same level at the start of theaggression trials as it was at the start of therecruitment trials (0, 50, 100, 150 or 200%).Approximately 6 h later, we placed an M.citrifolia-feeding larva within 3 cm of a nectarybody (1 larva per plant, 3–5 replicates per nectartreatment, resulting in 15–25 aggression trials persite). In Tutuila, we collected nitidulid (Coleop-tera) larvae from fallen M. citrifolia fruits ;24hours before aggression trials were conducted.These larvae were weighed to the nearest 0.1mgand randomly assigned to nectar availabilitytreatments. Although nitidulid larvae were com-mon in Tutuila, they were rare in Savaii, so weinstead used nectary-body feeding pyralid (Lep-idoptera) larvae there. We used digital calipers tomeasure the length of pyralid larvae (collectedfrom non-treatment plants) to the nearest 0.01mm prior to aggression trials. For both test prey,there were no significant differences among thenectar treatments in the size of the larvae thatwere presented to ants (Nitidulidae: F¼ 0.01, P¼0.9210; Pyralidae F ¼ 0.08, P ¼ 0.7773). To assesspotential differences in ant responses to preytypes, we conducted an additional test usingboth nitiduids and pyralids which co-occurred atone site in Savaii (Saleaula_North). We found nosignificant differences in the responses of ants tothe two different larvae at that site (F¼ 0.71, P¼0.4027). Therefore, we pooled ant responsesacross both target prey types in the analyses(below).

Response variables.—Based upon prior observa-tions of ant behaviors (Savage, unpublisheddata), we determined that the interaction be-tween ants and herbivorous larvae occurs quick-ly; therefore trial length was 150 seconds (unlessthe larva was removed from the plant by antsbefore the elapsed time). For each trial, werecorded the time it took ants to discover, attack,and/or remove each larva from treatment plants.Discovery occurred when ants approached thelarva and tapped it with their antennae. Attackoccurred when ants bit, stung, or sprayed formicacid at the larva. Removal occurred when antsforcibly ejected the larva from the plant, either bythrowing it off the plant or by carrying it awayfrom the plant.

Data analyses.—We examined aggressive be-haviors both in terms of the proportion of preylarvae that were subject to ant aggression and interms of the time required for the aggressivebehaviors to be initiated. We used logisticregression (Proc LOGISTIC; SAS Institute 2003)with a binomial distribution and a chi square testto evaluate proportional data. To examine thetime it took ants to perform aggressive behaviors,we conducted survival analyses (Proc LIFEREG,SAS Inc. 2003, version 9.1, Cary, NC) with aWeibull distribution and Wald Chi square teststatistics. This type of analysis allowed us toaccount for right-censored data.

We first examined the relative aggressiondisplayed by A. gracilipes vs. other dominant,exotic ant species in terms of the overallproportions of prey discovered, attacked, andremoved and the time to perform these behav-iors. To do this, we conducted logistic regres-sions and survival analyses (as described above)with A. gracilipes invasion status and site (nestedin A. gracilipes invasion status) as categoricalfactors. For these tests we used the full datasetand pooled data across all nectar treatmentlevels.

Next, we assessed the influence of the nectartreatment on the relative aggression displayed byA. gracilipes and other dominant ant speciestowards prey larvae. Because we were interestedin the independent effects of nectar on differentstages of ant aggression, we examined preydiscovery, attack and removal in terms of theproportion of all larvae that were discovered, theproportion of discovered larvae that were at-


SAVAGE AND WHITNEY

tacked, and the proportion of attacked larvae thatwere removed from M. citrifolia plants. Similarly,we examined the time to discovery of all prey, thetime to attack of discovered prey, and the time toremoval of attacked prey. In all models (bothlogistic regressions and survival analyses), thefactors included A. gracilipes invasion status, thenectar treatment, and their interaction. We alsoincluded site (nested in A. gracilipes invasionstatus) and the forager recruitment response ofthe dominant ant species for each plant (asestimated in the recruitment trials, above) ascovariates. Inclusion of the latter covariateallowed us to isolate the effects of per-workeraggressiveness from the effects of worker num-ber on the proportions of prey discovered,attacked, and removed and on the time toperform these behaviors.

Finally, when the interaction between A.gracilipes invasion status and the nectar treatmentwas significant, we conducted post-hoc linearand quadratic regressions, with the nectartreatment as the independent variable and theaverage proportion of prey discovered, attacked,and removed and the time to perform thesebehaviors as dependent variables. When bothlinear and quadratic models were significant, weused F-tests to evaluate model fit and presentonly the best-fit model.

Evidence for community-level indirect effectsmediated by ants

In order to demonstrate that TMII could havecommunity-level effects in this system, weexamined dynamics of M. citrifolia-ant mutual-isms at sites dominated by A. gracilipes versusthose dominated by other ant species. To do this,we conducted a plot-level, factorial experiment inwhich we simultaneously manipulated ant accessto (permitted/excluded) and nectar levels of M.citrifolia plants (reduced/ambient). This experi-ment was replicated across sites dominated by A.gracilipes (n ¼ 6) and those dominated by otherant species (n ¼ 5) on the island of Savaii. Threeand six months after treatment application, wesampled arthropod communities using bothsweepnets and surveys on the plants. A detaileddescription of this experiment and responses ofarthropod communities can be found in Savageet al. (in preparation). Here, we present data on b-diversity of plant-associated arthropods (exclud-

ing ants and honeydew-excreting insects). Toassess these arthropod responses, we conducteda permuted dispersion test of among-plot dis-similarity (PERMDISP using Primer-E v. 6.1.13with perMANOVAþ 1.0.3 extension, Clarke andGorley 2007). The factors for this test were A.gracilipes invasion status, the ant access treatmentand the nectar availability treatment. For thepurposes of this study, we were particularlyinterested in determining if the presence of antsinfluenced interactions between the co-occurringarthropod community and plant nectar. Conse-quently, when the three way interaction betweenall three factors was significant, we conductedpost-hoc pairwise tests comparing the effects ofthe ant access treatment within each A. gracilipesinvasion status x nectar availability level.

RESULTS

Lack of spatial autocorrelation in ant responses tonectar

Analyses of spatial autocorrelation demon-strated that the ant responses to nectar describedbelow are unlikely to be driven by spatialvariation in environmental conditions, but arerather the result of species-specific differences inant behaviors. Specifically, there was no signifi-cant spatial autocorrelation in forager recruit-ment responses (all tests: P . 0.21), theproportion of prey discovered (all tests: P .

0.48), the proportion of prey attacked (all tests: P. 0.44), the proportion of prey removed (all tests:P . 0.20), the time to prey discovery (all tests: P. 0.18), the time to prey attack (all tests: P .

0.11) or the time to prey removal (all tests: P .

0.13).

Hypothesis 1: As nectar levels increase, highlyinvasive ants will recruit more workers to EFN-bearing plants than less-invasive ants

We predicted that the highly invasive speciesA. gracilipes would demonstrate significantlystronger recruitment of workers to increasingnectar availability than other ant species (Fig.1A). Within sites invaded by A. gracilipes, thisprediction was supported. Anoplolepis gracilipesworkers recruited strongly to M. citrifolia plantsas experimentally manipulated nectar levelsincreased, with 281% more A. gracilipes workersobserved on plants with the highest nectar levels


SAVAGE AND WHITNEY

(200%) compared to those with no nectar (Fig.3a). In contrast, co-occurring subordinate antswere rarely observed on M. citrifolia plants,regardless of nectar availability level (Fig. 3a,Table 3: Dominance 3 Nectar treatment, P ,

0.0001). The number of nectary bodies per plantalso significantly influenced the abundance ofants per plant (Table 3). However, the response ofants to the nectar treatment was not significantlyinfluenced by this covariate, as evidenced by anon-significant interaction between the numberof nectary bodies and the nectar treatment.Similarly, ant responses to treatments did notvary over time (Table 3) at sites dominated by A.gracilipes.

Contrary to our predictions, other, less inva-sive ant species also recruited strongly to nectarresources in sites where they did not co-occurwith A. gracilipes (Fig. 3b, Table 4). In fact, therewas no significant difference between the re-sponses of A. gracilipes and other ants to nectaravailability (Invasion status 3 Nectar treatment:F ¼ 0.30, P ¼ 0.5857). Subordinate ants at siteslacking A. gracilipes did not display a significantresponse to experimentally elevated nectar levels,as evidenced by a significant Dominance 3

Nectar treatment interaction (Table 4, P ¼0.0024). However, significantly more subordinateants were observed foraging on M. citrifoliaplants at sites without A. gracilipes than at siteswith A. gracilipes (mean number of subordinateants 6 SE; A. gracilipes invaded sites ¼ 0.90 6

2.16; A. gracilipes uninvaded sites¼ 10.36 6 2.50;t-test t¼�7.2, P , 0.0001). At sites uninvaded byA. gracilipes, ant responses were not significantlyinfluenced by any of the covariates, nor did theyvary over time (Table 4).

Hypothesis 2: As nectar resources increase, highlyinvasive ants will become more aggressive towardco-occurring arthropods than less invasive ants

We predicted that highly invasive A. gracilipesworkers would not only display higher overalllevels of aggression, but also increase theiraggressiveness more strongly in response tonectar than co-occurring less invasive ants (Fig.1A). Overall aggression (pooling across nectarlevels) was, in fact, higher for A. gracilipes thanfor other dominant ant species. Compared tosites where other ant species were dominant, theaverage proportion of prey larvae that were

discovered (v2¼ 6.53, P¼ 0.0106), attacked (v2¼7.20, P ¼ 0.0073), and removed (v2 ¼ 6.42, P ¼0.0113) was 27%, 203% and 460% higher,

respectively, at sites where A. gracilipes was

dominant (Fig. 4). On average, A. gracilipes

workers also discovered prey 37% faster (v2 ¼12.0, P ¼ 0.0005), attacked prey 49% faster (v2 ¼24.45, P , 0.0001), and removed prey 44% faster

(v2 ¼ 25.8, P , 0.0001) than the average time

taken by other dominant ants (Fig. 5).

In support of our hypothesis, the amount of

nectar strongly influenced aggressive behaviors

of A. gracilipes, while aggression of other domi-

Fig. 3. Forager recruitment of dominant and

subordinate ants to increasing nectar levels at sites

dominated by invasive Anoplolepis gracilipes (a) and

those dominated by other ant species (b). Error bars

represent 61 SE of the mean. Significant relationships

are depicted with solid lines.


SAVAGE AND WHITNEY

nant ant species was unresponsive to increasing

nectar (Tables 5 and 6; Fig. 6). The number of

foragers recruiting to nectar had a significant

effect on the proportion of and time to prey

discovery and the time it took ants to attack

discovered prey (Table 4). However, the influ-

ence of nectar on ant aggression, particularly for

A. gracilipes, was significant even when account-

ing for these numerical effects, indicating that

nectar availability influenced per-capita worker

aggression. There was a significant effect of the

interaction between the nectar treatment and A.

gracilipes invasion status for all responses with

the exception of the proportion of prey removed

(Tables 5 and 6). Specifically, at the highest nectar

level (200%), A. gracilipes workers discovered

205% more prey in 89% less time (Fig. 6a, b),

attacked 32% more discovered prey in 93% less

Table 3. Results from a repeated measures analysis of covariance of ant density per

plant at sites invaded by Anoplolepis gracilipes.

Source of variation df F P

Dominance (dominant/subordinate) 1, 296 194.27 ,0.0001Nectar treatment 1, 296 35.74 ,0.0001Dominance 3 Nectar treatment 1, 296 139.80 ,0.0001Site 7, 296 4.55 ,0.0001Plot (Site) 28, 296 1.46 0.0730Plant size 1, 296 1.74 0.1878Abundance of honeydew-excreting insects (HEI) 1, 296 1.57 0.2113Number of nectary bodies 1, 296 11.25 0.0009Nectar treatment 3 Abundance of HEI 1, 296 0.00 0.9498Nectar treatment 3 Number of nectary bodies 1, 296 4.65 0.0519Time 5, 292 0.67 0.6459Time 3 Dominance 5, 292 3.28 0.0068Time 3 Nectar treatment 5, 292 1.23 0.2954Time 3 Dominance 3 Nectar treatment 5, 292 0.76 0.5761Time 3 Site 25, 168 6.77 ,0.0001Time 3 Plot (Site) 70, 168 1.34 0.0080Time 3 Plant size 5, 292 0.29 0.9181Time 3 Abundance of honeydew-excreting insects (HEI) 5, 292 0.50 0.7786Time 3 Number of nectary bodies 5, 292 0.77 0.5744Time 3 Nectar treatment 3 Abundance of HEI 5, 292 0.95 0.4465Time 3 Nectar treatment 3 Number of nectary bodies 5, 292 0.63 0.6785

Table 4. Results from a repeated measures analysis of covariance of ant density per

plant at sites uninvaded by Anoplolepis gracilipes.

Source of variation df F P

Dominance (Dominant/Subordinate) 1, 172 1.25 0.2647Nectar treatment 1, 172 3.39 0.0671Dominance 3 Nectar treatment 1, 172 9.47 0.0024Site 5, 172 2.88 0.0159Plot (Site) 14, 172 0.46 0.9521Plant size 1, 172 0.00 0.9892Abundance of honeydew-excreting insects (HEI) 1, 172 1.34 0.2485Number of nectary bodies 1, 172 0.06 0.8015Nectar treatment 3 Abundance of HEI 1, 172 2.95 0.0878Nectar treatment 3 Number of nectary bodies 1, 172 1.31 0.2538Time 5, 168 1.14 0.3387Time 3 Dominance 5, 168 2.19 0.0573Time 3 Nectar Treatment 5, 168 0.67 0.6455Time 3 Dominance 3 Nectar Treatment 5, 168 1.05 0.3887Time 3 Site 25, 76 6.16 ,0.0001Time 3 Plot (Site) 17, 76 0.98 0.5289Time 3 Plant size 5, 168 0.16 0.9753Time 3 Abundance of honeydew-excreting insects 5, 168 0.49 0.7855Time 3 Number of nectary bodies 5, 168 0.61 0.6885Time 3 Nectar treatment 3 Abundance of HEI 5, 168 1.00 0.4214Time 3 Nectar treatment 3 Number of nectary bodies 5, 168 0.53 0.7503


SAVAGE AND WHITNEY

Fig. 4. Behavioral responses of Anoplolepis gracilipes and other dominant ant species to prey larvae introduced

nearMorinda citrifolia nectary bodies. Discovery was defined as antennation with no further aggressive behaviors.

Attacks occurred when workers bit, stung, or sprayed formic acid at larvae, but did not remove them from

plants. Removals occurred when workers either carried prey off plants or physically ejected the larva from the

plant. Error bars represent 61 SE of the mean. Asterisks represent results from logistic regressions, with *

indicating P , 0.05 and ** indicating P , 0.01.

Fig. 5. Time to discovery, attack and removal of prey by Anoplolepis gracilipes versus other dominant ant

species. Shorter bars represent more rapid responses. Error bars represent 61 SE of the mean. Trials lasted a

maximum of 150 seconds. Asterisks represent results from survival analyses, with *** indicating P , 0.001, and

**** indicating P , 0.0001.


SAVAGE AND WHITNEY

time (Fig. 6c, d), and removed attacked prey in76% less time (Fig. 6e, f ) than on plants with thelowest nectar level (0% treatment). Overall, preyremoval by A. gracilipes increased from 19% to81% from the lowest to the highest nectar level.In contrast, the aggressive behaviors of otherdominant ants did not show significant respons-es to nectar manipulations (Fig. 6).

The strongest responses of A. gracilipes toincreasing nectar levels occurred in terms of theproportion of larvae that were discovered (P ¼0.0068, r¼ 0.9987; Fig. 6a, Table 5) and the time ittook workers to discover prey items (P¼0.0037, r¼ �0.9995; Fig. 6b, Table 6). Once larvae werediscovered, there was a marginally significantpositive effect of the nectar treatment on theproportion of larvae attacked by A. gracilipes (P¼0.0594, r ¼ 0.8634; Fig. 6c) and a marginallysignificant negative effect of the nectar treatmenton time to attack (P¼0.0606; r¼�0.8615; Fig. 6d).Once attacked by A. gracilipes workers, 94%(62.4%) of larvae were removed fromM. citrifoliaplants regardless of nectar level (Fig. 6e). Thetime it took A. gracilipes workers to remove

larvae dropped significantly as nectar levelsincreased (P ¼ 0.0481, r ¼�0.8814; Fig. 6f ).

Evidence for community-level indirect effectsmediated by ants

We predicted that there would be cascading,community-wide consequences resulting fromdifferences in the dynamics of M. citrifolia-antmutualisms when A. gracilipes dominated plotsversus when it did not (Fig. 1A). This predictionwas supported in terms of the b-diversity ofplant-associated arthropods. There was a signif-icant three-way interaction between A. gracilipesinvasion status, ant access to the plants, andnectar availability of M. citrifolia (PERMDISP forthree-way interaction: P ¼ 0.009), indicating thatthere was an indirect interaction between theseEFN-bearing plants and local arthropod commu-nities that was mediated by ants. Furthermore,the effect of ants on these indirect interactionswas different when A. gracilipes dominated localant assemblages (Fig. 7). Specifically, at reducednectar levels, arthropod b-diversity did notchange when ants were allowed access to plants,

Table 5. Analyses of proportions of prey that were discovered, attacked or removed across sites invaded and

uninvaded by Anoplolepis gracilipes. Logistic regression with a binomial distribution and a chi square test were

used to compare the proportions of prey discovered, the proportion of discovered prey that were attacked and

the proportion of attacked prey that were removed. Because sites were either invaded or uninvaded by A.

gracilipes, site was used as a nested factor.

Factor

Prey discovered Discovered prey attacked Attacked prey removed

v2 P v2 P v2 P

A. gracilipes invasion status 0.216 0.6418 0.8705 0.3508 0.0023 0.9616Nectar treatment 6.140 0.0132 0.2762 0.5992 0.3366 0.5618Invasion status 3 Nectar treatment 14.86 0.0001 4.8001 0.0285 0.1961 0.6579Forager recruitment 9.000 0.0027 1.2926 0.2556 0.0033 0.9545Site (Invasion status) 12.81 0.2347 8.534 0.5769 3.03 0.9807

Table 6. Analyses of times to discovery, attack and removal of prey items across sites invaded and uninvaded by

Anoplolepis gracilipes. To examine the time it took ants to perform aggressive behaviors, we conducted survival

analyses with a Weibull distribution and Wald v2 test statistics.

Factor

Time to discover preyTime to attack

of discovered preyTime to removalof attacked prey

v2 P v2 P v2 P

A. gracilipes invasion status 2.997 0.0834 0.2994 0.5843 0.7080 0.4001Nectar treatment 7.387 0.0066 7.421 0.0064 0.1309 0.7175Invasion status 3 Nectar treatment 39.90 ,0.0001 24.279 ,0.0001 7.147 0.0075Forager recruitment 16.84 ,0.0001 5.265 0.0218 2.104 0.1469Site (Invasion status) 30.47 0.0007 27.443 0.0022 16.919 0.0762


SAVAGE AND WHITNEY

Fig. 6. Behavioral responses of dominant ant species to experimentally manipulated nectar levels. Filled circles

represent responses of Anoplolepis gracilipes and empty circles represent other dominant ant species. Error bars

represent 61 SE of the mean. All relationships between nectar level and behavioral responses of non-A. gracilipes

ants were ns. Significant relationships (P , 0.05) are depicted with a solid line and marginally significant trends

(0.05 , P , 0.10) are depicted with a dashed line.


SAVAGE AND WHITNEY

regardless of whether the ants were A. gracilipesor other ant species. However, at ambient nectar,the presence of A. gracilipes decreased b-diversity,while the presence of other ant species increasedb-diversity. These b-diversity differences betweenant access treatments were stronger at sitesdominated by A. gracilipes than they were atsites dominated by other ant species (Fig. 7).

DISCUSSION

When exotic species are introduced to novelhabitats, they often form beneficial associations

with indigenous species. However, we knowlittle about the ways that these positive interac-tions affect invaders themselves or their interac-tions with other community members. Trait-mediated indirect interactions via mutualisticassociations may contribute strongly to thenegative impacts associated with species inva-sions. In this study, we predicted that a highlyinvasive ant species would exhibit a strongresponse to increasing nectar levels in terms ofrecruitment and aggressive behaviors. We alsopredicted that these responses would be muchweaker in co-occurring less-invasive ant species.

Fig. 7. Evidence for indirect effects of plants on the arthropod community mediated by ants: among-plot

dissimilarity (b-diversity) of plant-associated arthropods on Morinda citrifolia at sites dominated by Anoplolepis

gracilipes (black bars) and those dominated by other ant species (gray bars). Filled bars represent plots with ants

permitted and empty bars represent plots in which ants were excluded from M. citrifolia plants using a sticky

barrier. There was a significant A. gracilipes invasion status3ant access3nectar level interaction (PERMDISP: P¼0.009); asterisks are from significant post-hoc pairwise comparisons of ant (þ) and ant (�) treatments within each

A. gracilipes status 3 nectar level * P , 0.05, ** P , 0.01.


SAVAGE AND WHITNEY

We expected that differential responses of highlyinvasive and less invasive ants would then leadto different consequences for the plant-associatedarthropod community (Fig. 1A). In fact, wefound that both A. gracilipes and other non-nativedominant ant species responded positively toincreasing nectar levels in terms of foragerrecruitment. However, A. gracilipes uniquelyincreased the degree of aggressiveness as nectarlevels increased (Fig. 1B), resulting in more rapidattacks on and greater removal of plant-associat-ed arthropods. Furthermore, the b-diversity ofthe arthropod community responded to the M.citrifolia-ant mutualism differently when A. gra-cilipes dominated local ant assemblages vs. whenother ant species were dominant. Thus, trait-mediated indirect effects between native nectar-producing plants and plant-associated arthro-pods were present and differed depending on theidentity of the ant partner, and were likely atleast partially responsible for differences in thestructures of communities dominated by A.gracilipes vs. dominated by less-invasive ants.

Of necessity, these experiments were conduct-ed at sites already invaded by exotic ant species,and these species were not uniformly distributedacross islands, e.g., A. gracilipes was moreabundant and widespread on Savaii than onTutuila. Since it would be unethical to transportthis highly invasive species to uninvaded sites inTutuila, we could not create a more balanceddesign by experimentally manipulating the iden-tity of the dominant ant species at these sites.Consequently, there could be confounding effectsof island on these results, although tests withinA. gracilipes or other dominant ant species foundno evidence for behavioral differences betweenislands.

Effects of nectar subsidies on forager recruitmentCarbohydrate-rich resources, such as plant

nectar or hemipteran honeydew, may promoteant invasions by providing a high-energy foodthat fuels greater activity and growth andfurthers the establishment of dominant super-colonies (the ‘carbohydrate subsidy hypothesis’,Lach 2003, Savage et al. 2009). If these carbohy-drate-rich resources are, in fact, important factorsin the progression of ant invasions, then highlyinvasive species should respond more strongly tovariation nectar resources than co-occurring less

invasive species. In a previous study in north-eastern Savaii (Savage et al. 2011), we found thatrecruitment responses of A. gracilipes were muchstronger than a less-invasive ant, Pheidole mega-cephala. In this study, we manipulated plantnectar and tested the recruitment responses of alarger number of ant species (including fourpreviously untested dominant species) across amuch wider array of sites in both Savaii andTutuila. The carbohydrate subsidy hypothesiswas supported at sites invaded by A. gracilipes: A.gracilipes workers recruited strongly to increasingnectar levels, and other ant species were rarelyobserved on the plants. This difference in theresponse of A. gracilipes to variation in carbohy-drates is important, because the availability ofthese resources often varies substantially acrossmultiple spatiotemporal scales. For example,Eubanks (2001) found patchy distributions ofthe red imported fire ant (Solenopsis invicta) inagricultural systems of the South-eastern UnitedStates. He suggested that the presence of ant-tended, honeydew-excreting aphids explainedmuch of this variation—a supposition that waslater supported by manipulative experiments(Kaplan and Eubanks 2005). Similarly, ourbroad-scale surveys across the Samoan Archipel-ago showed that the patchy distribution of A.gracilipes was strongly, positively correlated withthe dominance of EFN-bearing plants (Savage etal. 2009).

However, our prediction that other ant specieswould display weaker recruitment responses tonectar availability was not supported in siteswhere A. gracilipes was absent. In fact, the pooledresponse of other dominant ants to increasingnectar availability was also strong and positive.There is likely variation among these lessinvasive ant species in their recruitment responseto increasing nectar availability. However, thepurpose of this study was to assess differencesbetween A. gracilipes and co-occurring ant speciesbecause (1) our previous surveys (Savage et al.2009) indicated that A. gracilipes was the onlyspecies with a strong positive association withEFN-bearing plants and (2) the carbohydratesubsidy hypothesis explicitly predicts that highlyinvasive ant species respond more strongly tocarbohydrate resources (A. gracilipes is the mostinvasive species in Samoa). Therefore, we did nothave sufficient replication at the level of individ-


SAVAGE AND WHITNEY

ual, non-A. gracilipes species to determine howthese species differed from each other. Despitethese caveats, the strong, positive pooled re-sponses of these other species suggest that whilerecruitment responses likely contribute to dy-namics between highly invasive A. gracilipes,native plants, and co-occurring arthropod com-munities, recruitment, in and of itself, cannotexplain why A. gracilipes is associated withgreater invasive impacts than are the other exoticdominant ant partners of M. citrifolia (Fig. 7;Savage et al. 2009, Savage et al., in preparation).

If long-term responses of ants to nectar differfrom short-term responses, the connections weidentify between EFN, ant behavior, and com-munity-level consequences may be different. Wenote that manipulations of ant access and nectaravailability over six months (Savage et al., inpreparation) resulted in similar patterns of antrecruitment to those reported in this study, thatis, recruitment to supplemented nectar treat-ments stayed high while recruitment to reducednectar treatments stayed low (Savage, unpub-lished data). However, it is likely that sustainedincreases in nectar availability (in contrast to thepulses used in this experiment) would lead toincreased colony growth (also see Oliver et al.2008). Larger colony size could have importanteffects on co-occurring community members (i.e.,density-mediated indirect effects), especially inthe context of invasions. For example, Holwayand Case (2001) found that increased Linepithemahumile (Argentine ant) colony size enhanced bothexploitative and interference competitive abilitiesof this highly invasive ant species.

Additionally, our experimental nectar supple-ments (the 150 and 200% levels) only manipulat-ed one component of M. citrifolia nectar (sucrose)and did not manipulate other nectar componentssuch as amino acids, perhaps leading to unreal-istic results. Some previous studies have indicat-ed that carbohydrate-rich diets lead to increasedactivity in the short-term, but are nutritionallyincomplete and limit the ability of ant colonies togrow (Buschinger and Pfeifer 1988, Porter 1989).However, recent studies have challenged thisassertion (e.g., Kay et al. 2010, Wilder et al. 2010).For example, Byk and Del-Claro (2011) showedthat the abundance and mass of queens, workersand eggs all increased when Cephalotes pusillusants were permitted access to EFN. Some of these

benefits could be due to the presence of traceamounts of amino acids in nectar and honeydew(Bluthgen et al. 2004), although recent evidencesuggests that the effects of amino acids are not asstrong as the effects of carbohydrates. Forexample, Wilder et al. (2011) recently demon-strated that access to carbohydrates (withoutamino acids) reduced worker mortality in L.humile, while access to amino acids (withoutcarbohydrates) increased worker mortality (Wil-der et al. 2011). Most importantly, if amino acidswere driving ant responses to EFN, then wewould expect to see differences in the shape ofant responses moving from reduced to ambientnectar (0 and 50 % to 100%), versus from ambientto supplemented nectar (100% to 150 and 200%).Instead, the slopes of ant responses to nectarmanipulations were consistent across nectarlevels (Figs. 3 and 6), suggesting that thesupplemental nectar treatments manipulatedthe component of M. citrifolia nectar mostrelevant to the ant species in this study.

Effects of nectar subsidies on ant aggressionIn addition to increased forager recruitment,

another trait-mediated pathway whereby carbo-hydrate-rich resources may influence ants andtheir impacts on other species is throughincreased levels of aggression (Fig. 1). Basedupon the carbohydrate subsidy hypothesis, wepredicted that highly invasive ants would re-spond to increasing carbohydrate levels byincreasing the likelihood or speed of attacks onco-occurring arthropods. Again, this prediction iscontingent upon the assumption that less inva-sive ants do not respond in the same way tocarbohydrate availability. In this study, weexplicitly tested this prediction for the first time,by manipulating levels of plant nectar andobserving the aggression displayed by highlyinvasive and co-occurring less-invasive ants.

We found that the highly invasive ant A.gracilipes displayed unique increases in aggres-sion in response to increasing nectar availability,a response not found for other dominant antspecies in Samoa. As nectar levels increased,there was a general increase in the likelihood ofprey discovery and attack and a general reduc-tion in the amount of time it took A. gracilipesworkers to discover, attack and remove prey. Thestrongest responses of A. gracilipes were in terms


SAVAGE AND WHITNEY

of prey discovery, which may be due to higheroverall activity levels when nectar availability ishigh. However, the independent effects of nectaron A. gracilipes attack and removal suggests thatincreased carbohydrate availability can influenceother aspects of aggressive behaviors, at least forthis species. Interestingly, other exotic, dominantant species did not display these nectar-depen-dent aggressive responses, even in the absence ofA. gracilipes. Thus, our findings support theprediction that highly invasive ant speciesrespond more strongly to carbohydrate availabil-ity in terms of increased aggression than lessinvasive ants. Furthermore, this trait differenceapparently has consequences for co-occurringarthropods (Figs. 1B and 7) and thus couldpotentially explain the observed community-level effects of M. citrifolia-A. gracilipes mutual-isms.

At first glance, any changes in aggressiveresponses of A. gracilipes to increasing nectarresources could be simply explained by theforager recruitment responses that we demon-strated in the first experiment. With more antsrecruiting to nectar, the likelihood of preydiscovery should be higher, due to increasedencounter rates. In fact, we found that increasedforager recruitment did explain some of thevariation in ant aggression across nectar levels.However, we also found that significant effects ofnectar on aggression remained even after ac-counting for these numerical effects. Further-more, other ants did not express these aggressiveresponses despite displaying similar recruitmentresponses to A. gracilipes. These patterns indicatethat nectar uniquely influences per-capita workeraggression in A. gracilipes.

Together with our findings, results from otherrecent studies suggest that carbohydrates canstrongly influence ant behaviors with conse-quences for co-occurring arthropods, and thatthese effects can be conditional on ant identity.For example, Grover et al. (2007) found that bothactivity levels and intraspecific aggression of L.humile were higher for lab colonies that were feda diet rich in carbohydrates than under a protein-rich diet. Similarly, Pringle et al. (2010) showedthat native plant-inhabiting ants in a neotropicallowland rainforest were more aggressive to-wards plant-feeding herbivores when fed a dietrich in carbohydrates. However, Kay et al. (2010)

examined the aggression exerted by L. humiletowards heterospecific ants and found no influ-ence of diet, although carbohydrate-rich dietsresulted in greater colony growth (a density-mediated effect). In the deserts of the SouthwestUnited States, Ness et al. (2009) demonstratedthat four species of native ants were moreaggressive towards novel prey when fed supple-mental carbohydrates than when fed protein orgiven no supplements; however, no invasive antspecies were tested in this study. Finally, Lachand Hoffmann (2011) demonstrated that invasiveA. gracilipes workers were more likely to attackprey than workers of one native, dominant antspecies (Oecophylla smaragdina), but only onplants bearing EFN; attack rates of the speciesdid not differ on plants that did not secreteextrafloral nectar.

In most of the cases in which A. gracilipes hasbeen reported to have detrimental impacts on co-occurring community members, this species hasalso associated with plants or insects that secretecarbohydrate-rich food in the form of nectar orhoneydew (Addison and Samways 2000, Holwayet al. 2002, Hill et al. 2003, O’Dowd et al. 2003,Lester and Tavite 2004, Savage et al. 2009). Theamplification of aggressive behaviors that wedetected in this study may provide a mechanismthat underlies this pattern. Similar tests ofaggression in response to carbohydrate availabil-ity across the invaded range of A. gracilipes willhelp to elucidate the generality of these findings.Furthermore, it will be important to conductstudies that manipulate nectar over longer timespans in order to ascertain the relative impor-tance of density and trait-mediated effects ofnectar on A. gracilipes invasions.

ConclusionsMutualisms are common interspecific interac-

tions that can influence the structure anddynamics of communities and the functioningof ecosystems (Bronstein 1994, Stachowicz 2001,Rudgers and Clay 2008). However, we knowsurprisingly little about the mechanisms thatunderlie many of these community-wide effects.Importantly, indirect, trait-mediated pathwaysare likely to be important mechanistic compo-nents of the effects of mutualisms on communi-ties, just as they are for antagonistic interactions.In this study, we demonstrated that (1) resources


SAVAGE AND WHITNEY

exchanged in mutualisms between ants andEFN-bearing plants can change behaviors thatants exert towards other community membersand (2) these changes are more extreme forinvasive A. gracilipes than co-occurring lessinvasive ants. In studies of antagonistic interac-tions, variation in aggression has been shown toinfluence both species co-existence (Frye 1983,Logan 1984, Morrison 1996) and the displace-ment of native species by exotics (Carpintero andReyes-Lopez 2008). Thus, if they represent awidespread pattern, our findings suggest thattrait-mediated indirect interactions associatedwith novel mutualisms between invaders andnative species could contribute to the success anddetrimental impacts of species invasions.

ACKNOWLEDGMENTS

Many thanks to J.A. Rudgers for help with exper-imental design and data analysis and for thoughtfulcomments on the manuscript. We also thank local landowners in Savaii and Tutuila for graciously allowing usto use their land for our experiments. Thanks to MarkSchmaedick for logistical and field assistance. Manythanks to O. Schmitz, B. Hoffmann, and threeanonymous reviewers for thoughtful comments onprevious versions of this manuscript. AMS designedand conducted experiments, analyzed the data, andwas the primary author of the manuscript; KDWprovided logistical support for all phases of the study,including extensive assistance during the data analysisand writing phases of the project. This work wassupported by National Geographic Society grant 8237-07 and by a Rice University Wray-Todd Fellowship toAMS. KDW was supported by NSF DEB 071686.

LITERATURE CITED

Abbott, K. L. 2005. Supercolonies of the invasive ant,Anoplolepis gracilipes, on an oceanic island: forageractivity patterns, activity and biomass. InsectesSociaux 52:266–273.

Abbott, K. L. 2006. Spatial dynamics of supercoloniesof the invasive yellow crazy ant, Anoplolepisgracilipes, on Christmas Island, Indian Ocean.Diversity and Distributions 12:101–110.

Adams, C. T., W. A. Banks, and J. K. Plumley. 1976.Polygyny in the tropical fire ant, Solenopsis geminatawith notes on Solenopsis invicta. The FloridaEntomologist 59:411–415.

Addison, P. and M. J. Samways. 2000. A survey of ants(Hymenoptera: Formicidae) foraging in the west-ern cape vineyards of South Africa. AfricanEntomology 8:251–260.

Astruc, C., C. Malosse, and C. Errand. 2001. Lack ofintraspecific aggression in the ant Tetramoriumbicarinatum: A chemical hypothesis. Journal ofChemical Ecology 27:1229–1248.

Bernot, R. J. and A. M. Turner. 2001. Predator identityand trait-mediated indirect effects in a littoral foodweb. Oecologia 129:139–346.

Bluthgen, N., G. Gottsberger, and K. Fiedler. 2004.Sugar and amino acid composition of ant attendednectar and honeydew sources from an Australianrainforest. Austral Ecology 29:418–429.

Bronstein, J. L. 1994. Conditional outcomes in mutu-alistic interactions. Trends in Ecology and Evolu-tion 9:214–217.

Bruno, J. F., J. J. Stachowicz, and M. D. Bertness. 2003.Inclusion of facilitation into ecological theory.Trends in Ecology and Evolution 18:119–126.

Buschinger, A., and E. Pfeifer. 1988. Effects of nutritionon brood production and slavery in ants Hyme-noptera: Formicidae. Insectes Sociaux 35:61–69.

Bustos, X., and D. Cherix. 1998. Contribution a labiologie de Tapinoma melanocephalum Fabricius.Hymenoptera: Formicidae. Actes des ColloquesInsectes Sociaux 11:95–101.

Byk, J. and K. Del-Claro. 2011. Ant-plant interaction inthe neotropical savannah: direct beneficial effects ofextrafloral nectar on ant colony fitness. PopulationEcology 53:327–332.

Carpintero, S. and J. Reyes-Lopez. 2008. The role ofcompetitive dominance in the invasive abilities ofthe Argentine ant Linepithema humile. BiologicalInvasions 10:25–35.

Carroll, C. R. and S. J. Risch. 1984. The dynamics ofseed harvesting in early successional communitiesby a tropical ant, Solenopsis geminata. Oecologia61:388–392.

Chittka, L. and S. Schurkens. 2001. Successful invasionof a floral market. Nature 411:653.

Clarke, K. R., and R. N. Gorley. 2007. Primer: usermanual and tutorial. Version 6.1.10. Primer-E,Plymouth, UK.

Davidson, D. W., S. C. Cook, R. R. Snelling, and T. H.Chua. 2003. Explaining the abundance of ants inlowland tropical rainforest canopies. Science300:969–972.

Davis, N. E., D. J. O’Dowd, P. J. Green, and R.MacNally. 2008. Effects of an alien ant invasion onthe abundance, behavior and reproductive successof endemic island birds. Conservation Biology22:1165–1176.

Drescher, J., H. Feldhaar, and N. Bluthgen. 2010.Interspecific aggression and resource monopoliza-tion of the invasive ant Anoplolepis gracilipes inMalaysian Borneo. Biotropica 43:93–99.

Eubanks, M. D. 2001. Estimates of the direct andindirect effects of red imported fire ants onbiological control. Biological Control 21:35–43.


SAVAGE AND WHITNEY

Freeman, C. E., R. D. Worthington, and M. S. Jackson.1991. Floral nectar sugar compositions of somesouth and southeast Asian species. Biotropica23:568–74.

Fromhoff, P. C., and P. S. Ward. 1992. Individual-levelselection, colony-level selection, and the associationbetween polygyny and worker monomorphism inants. The American Naturalist 139:559–590.

Frye, R. J. 1983. Experimental evidence of interspecificaggression between two species of kangaroo ratDipodomys. Oecologia 59:74–78.

Gaigher, R., M. J. Samways, J. Henwood, and K.Jolliffe. 2011. Impact of a mutualism between aninvasive ant and honeydew-producing insects on afunctionally important tree on a tropical island.Biological Invasions 13:1717–1721.

Gomez, J. M., M. Abdelaziz, J. Lorite, A. J. Munoz-Pajares, and F. Perfectti. 2010. Changes in pollinatorfauna cause spatial variation in pollen limitation.Journal of Ecology 98:1243–1252.

Grabas, G. P., and T. M. Laverty. 1999. The effect ofpurple loosestrife Lythrum salicaria L. Lythraceae onthe pollination and reproductive success of sym-patric co-flowering wetland plants. Ecoscience6:230–242.

Greenslade, P. J. M. 1971. Interspecific competition andfrequency changes among ants in Solomon Islandcoconut plantations. Applied Ecology 8:323–352.

Grover, C. D., A. D. Kay, J. A. Monson, T. C. Marsh,and D. A. Holway. 2007. Linking nutrition andbehavioral dominance: carbohydrate scarcity limitsaggression and activity in Argentine ants. Proceed-ings of the Royal Society B-Biological Sciences274:2951–2957.

Handler, A. T., D. S. Gruner, W. P. Haines, M. W.Lange, and K. Y. Kaneshiro. 2007. Arthropodsurveys on Palmyra Atoll and insights into thedecline of the native tree Pisonia grandis Nyctagi-naceae. Pacific Science 61:485–502.

Harada, A. Y. 1990. Ant pests of the Tapinomini tribe.Pages 298–309 in R. K. VanderMeer, K. Jaffe, and A.Cedeno, editors. Applied myrmecology: a worldperspective. Westview Press, Boulder, Colorado,USA.

Hay, M. E., J. D. Parker, D. E. Burkepile, C. C. Caudill,A. E. Wilson, Z. P. Hallinan, and A. D. Chequer.2004. Mutualisms and aquatic community struc-ture: the enemy of my enemy is my friend. AnnualReview of Ecology, Evolution and Systematics35:175–197.

Heterick, B. 1997. The interaction between the coastalbrown ant, Pheidole megacephala, (Fabricius) andother invertebrate fauna of Mt. Coot-tha Brisbane,Australia. Austral Ecology 22:218–221.

Hill, M., K. Holm, T. Vel, N. J. Shah, and P. Matyot.2003. Impact of the introduced yellow crazy antAnoplolepis gracilipes on Bird Island, Seychelles.

Biodiversity and Conservation 12:1969–1984.Hoffmann, B. D. 2010. Ecological restoration following

the local eradication of an invasive ant in northernAustralia. Biological Invasions 12:959–969.

Hoffmann, B. D., and S. O’Conner. 2004. Eradication oftwo exotic ants from Kakadu National Park.Ecological Management and Restoration 5:98–105.

Hoffmann, B. D., and C. L. Parr. 2008. An invasionrevisited: the African big-headed ant Pheidolemegacephala in northern Australia. Biological Inva-sions 10:1171–1181.

Holway, D. A. and T. J. Case. 2001. Effects of colony-level variation on competitive ability in theinvasive Argentine ant. Animal Behaviour61:1181–1192.

Holway, D. A., L. Lach, A. V. Suarez, N. D. Tsutsui, andT. J. Case. 2002. The causes and consequences of antinvasions. Annual Review of Ecology and System-atics 33:181–233.

Kaplan, I. and M. D. Eubanks. 2005. Aphids alter thecommunity-wide impacts of fire ants. Ecology86:1640–1649.

Kay, A. D., T. Zumbusch, J. L. Heinen, T. C. Marsh, andD. A. Holway. 2010. Nutrition and interferencecompetition have interactive effects on the behaviorand performance of Argentine ants. Ecology 91:57–64.

Klotz, J., L. Hansen, R. Pospischil, and M. Rust. 2008.Urban ants of North America and Europe: Identi-fication, biology and management. Cornell Univer-sity Press, Ithaca, New York, USA.

Lach, L. 2003. Invasive ants: unwanted partners in ant-plant interactions? Annals of the Missouri Botan-ical Garden 90:91–108.

Lach, L. 2008. Argentine ants displace floral arthro-pods in a biodiversity hotspot. Diversity andDistributions 14:281–29.

Lach, L. and B. D. Hoffmann. 2011. Are invasive antsbetter plant-defense mutualists? A comparison offoliage patrolling and herbivory in sites withinvasive yellow crazy ants and native weaver ants.Oikos 120:9–16.

Lach, L. and M. L. Thomas. 2008. Invasive ants inAustralia: documented and potential consequenc-es. Australian Journal of Entomology 47:275–288.

Lawrence, J. M., M. J. Samways, J. Henwood, and J.Kelly. 2011. Effect of an invasive ant and itschemical control on a threatened endemic Sey-chelles millipede. Ecotoxicology 20:731–738.

Lee, C. Y. 2002. Tropical household ants: pest status,species diversity, foraging behavior and baitingstudies. Proceedings of the Fourth InternationalCongress on Urban Pests 3–18.

Lester, P. J. and A. Tavite. 2004. Long-legged ants,Anoplolepis gracilipes Hymenoptera: Formicidae.,have invaded Tokelau, changing composition anddynamics of ant and invertebrate communities.


SAVAGE AND WHITNEY

Pacific Science 58:391–401.Logan, A. 1984. Interspecific aggression in heratypic

corals from Bermuda. Coral Reefs 3:131–138.Lowe, S., M. Brown, and S. Boudjelas. 2000. 100 of the

world’s worst invasive alien species. Aliens 12:1–12.

Mangan, S. A., E. A. Herre, and J. D. Bever. 2010.Specificity between Neotropical tree seedlings andtheir fungal mutualists leads to plant-soil feed-backs. Ecology 91:2594–2603.

Maherali, H. and J. N. Klironomos. 2007. Influence ofphylogeny on community assembly and ecosystemfunctioning. Science 316:1746–1748.

Manzaneda, A. J., and P. J. Rey. 2008. Geographicvariation in seed removal of a myrmecochorousherb: influence of variation in functional guild andspecies composition of the disperser assemblagethrough spatial and temporal scales. Ecography31:583–591.

Matthews, C. R., D. G. Bottrell, and M. W. Brown.2009. Extrafloral nectaries alter arthropod commu-nity structure and mediate peach (Prunus persica)plant defense. Ecological Applications 19:722–730.

McGlynn, T. P. 1999. The worldwide transfer of ants:Geographical distribution and ecological invasions.Journal of Biogeography 26:535–548.

McInnes, D. A. and W. R. Tschinkel. 1995. Queendimorphism and reproductive strategies in the fireant Solenopsis geminata Hymenoptera: Formicidae.Behavioral Ecology and Sociobiology 36:367–375.

Miller, T. E. X. 2007. Does having multiple partnersweaken the benefits of facultative mutualism? Atest with cacti and cactus-tending ants. Oikos116:500–512.

Morrison, L. W. 1996. Community organization in arecently assembled fauna: the case of Polynesianants. Oecologia 107:243–256.

Morrison, L. W. 2000. Mechanisms of interspecificcompetition among an invasive and two native fireants. Oikos 90:238–252.

Nelson, S. C. 2006. Morinda citrifolia (Noni). Pages 1–13in C. R. Elevitch, editor. Species profiles for PacificIsland forestry. Permanent Agricultural Resources,Holualoa, Hawaii, USA.

Ness, J. H., W. F. Morris, and J. L. Bronstein. 2009. Forant-protected plants, the best defense is a hungryoffense. Ecology 90:2823–2831.

O’Dowd, D. J., P. T. Green, and P. S. Lake. 2003.Invasional ‘meltdown’ on an oceanic island. Ecol-ogy Letters 6:812–817.

Oliver, T. H., T. Pettitt, S. R. Leather, and J. M. Cook.2008. Numerical abundance of invasive ants andmonopolisation of exudate-producing resources—achicken and egg situation. Insect Conservation andDiversity 1:208–214.

Palmer, T. M., D. F. Doak, M. L. Stanton, J. L. Bronstein,E. T. Kiers, T. P. Young, J. R. Goheen, and R. M.

Pringle. 2010. Synergy of multiple partners, includ-ing freeloaders, increases host fitness in a multi-species mutualism. Proceedings of the NationalAcademy of Sciences 107:17234–17239.

Perfecto, I. 1991. Dynamics of Solenopsis geminata in atropical fallow field after ploughing. Oikos 62:139–144.

Pfeiffer, M., H. C. Tuck, and T. C. Lay. 2008. Exploringarboreal ant community composition and co-occurrence patterns in plantations of oil palm Elaeisguineensis in Borneo and Peninsular Malaysia.Ecography 31:21–32.

Plentovich, S., A. Hebshi, and S. Conant. 2009.Detrimental effects of two widespread invasiveant species on weight and survival of nestingseabirds in the Hawaiian Islands. Biological Inva-sions 11:289–298.

Plentovich, S., C. Swenson, N. Reimer, M. Richardson,and N. Garon. 2010. The effects of hydramethylnonon the tropical fire ant, Solenopsis geminate Hyme-noptera: Formicidae, and non-target arthropods onSpit Island, Midway Atoll, Hawaii. Journal ofInsect Conservation 14:459–465.

Plentovich, S., J. Ejizenga, H. Eijzenga, and D. Smith.2011. Indirect effects of ant eradication efforts onoffshore islets of the Hawaiian Archipelago. Bio-logical Invasions 13:545–557.

Porter, S. D. 1989. Effects of diet on the growth oflaboratory fire ant colonies Hymenoptera: Formi-cidae. Journal of the Kansas Entomological Society62:288–291.

Potterat, O. and M. Hamburger. 2007. Morinda citrifoliaNoni. fruit: Phytochemistry, pharmacology andsafety. Planta Medica 73:191–199.

Pringle, E. G., R. Dirzo, and D. M. Gordon. 2010.Indirect benefits of symbiotic coccoids for an ant-defended myrmecophytic tree. Ecological Mono-graphs 92:37–46.

Pysek, P. and D. M. Richardson. 2007. Traits associatedwith invasiveness in alien plants. Biological Inva-sions 193:97–125.

Risch, S. J. and C. R. Carroll. 1982. Effect of a keystonepredaceous ant, Solenopsis geminate, on arthropodsin a tropical agroecosystem. Ecology 63:1979–1983.

Rudgers, J. A. and K. Clay. 2008. An invasive plant-fungal mutualism reduces arthropod diversity.Ecology Letters 11:831–840.

Rudgers, J. A. and S. Y. Strauss. 2004. Selection mosaicin the facultative mutualism between ants and wildcotton. Proceedings of the Royal Society of London,Series B 271:2481–2488.

Rudgers, J. A., J. Holah, S. P. Orr, and K. Clay. 2007.Forest succession suppressed by an introducedplant-fungal symbiosis. Ecology 88:18–25.

Rudgers, J. A., A. M. Savage, and M. A. Rua. 2010.Geographic variation in a facultative mutualism:consequences for local arthropod composition and


SAVAGE AND WHITNEY

diversity. Oecologia 163:985–996.SAS Institute. 2003. SAS version 9.1.3. SAS Institute,

Cary, North Carolina, USA.Savage, A. M., J. A. Rudgers, and K. D. Whitney. 2009.

Elevated dominance of extrafloral nectary-bearingplants is associated with increased abundances ofan invasive ant and reduced native ant richness.Diversity and Distributions 15:751–761.

Savage, A. M., S. D. Johnson, K. D. Whitney, and J. A.Rudgers. 2011. Do invasive ants respond morestrongly to nectar availability than co-occurringnon-invasive ants? A test along an active Anoplo-lepis gracilipes invasion front. Austral Ecology36:310–319.

Shattuck, S. O. 1999. Australian ants: Their biology andidentification. Monographs on invertebrate taxon-omy. Volume 3. CSIRO Publishing, Collingswood,Victoria, Australia.

Stachowicz, J. J. 2001. Mutualism, facilitation, and thestructure of ecological communities. BioScience51:235–246.

Thomas, M. L., K. Becker, K. Abbott, and H. Feldhaar.2010. Supercolony mosaics: two different invasionsby the yellow crazy ant, Anoplolepis gracilipes, onChristmas Island, Indian Ocean. Biological Inva-sions 12:677–687.

van Kleunen, M., E. Weber, and M. Fische. 2010. Ameta-analysis of trait differences between invasiveand non-invasive plant species. Ecology Letters13:235–245.

von Ende, C. N. 2001. Repeated measures analysis:growth and other time dependent measures. Pages134–157 in S. M. Scheiner and J. Gurevitch, editors.Design and analysis of ecological experiments.Oxford University Press, New York, New York,USA.

Waki, J., T. Olkpul, and M. K. Komolong. 2007. Adescriptor list for morphological characterization ofNoni Morinda citrifolia L. The South Pacific Journalof Natural Science 10:61–66.

Ward, D. F., M. C. Stanley, R. J. Toft, A. S. Forgie, andR. J. Harris. 2008. Assessing the risk of invasiveants: a simple and flexible scorecard approach.Insectes Sociaux 55:360–363.

Werner, E. E. and S. D. Peacor. 2003. A review of trait-mediated indirect interactions in ecological com-munities. Ecology 84:1083–1100.

Wetterer, J. K. 2005. Worldwide distribution and

potential spread of the long-legged ant, Anoplolepisgracilipes Hymenoptera: Formicidae. Sociobiology45:77–96.

Wetterer, J. K. 2008. Worldwide spread of the longhorncrazy ant, Paratrechina longicornis Hymenoptera:Formicidae. Myrmecological News. 1:137–149.

Wetterer, J. K. and D. L. Vargo. 2003. Ants (Hyme-noptera: Formicidae) of Samoa. Pacific Science57:409–419.

Wetterer, J. K., S. E. Miller, D. E. Wheeler, C. A. Olson,D. A. Polhemus, M. Pitts, I. W. Ashton, A. G.Himler, M. M. Yospin, K. R. Helms, E. L. Harken, J.Gallaher, C. E. Dunning, M. Nelson, J. Litsinger, A.Southern, and T. A. Burgess. 1999. Ecologicaldominance by Paratrechina longicornis (Hymenop-tera: Formicidae), an invasive tramp ant, inBiosphere 2. The Florida Entomologist 82:381–388.

White, E. M., J. C. Wilson, and A. R. Clarke. 2006.Biotic indirect effects: a neglected concept ininvasion biology. Diversity and Distributions12:443–455.

Whistler, W. A. 1993. Flowers of the Pacific Islandseashore: a guide to the littoral plants of Hawaii,Tahiti, Samoa, Tonga, Cook Islands, Fiji andMicronesia. University of Hawaii Press, Honolulu,Hawaii, USA.

Whitney, K. D. 2004. Experimental evidence that bothparties benefit in a facultative plant-spider mutu-alism. Ecology 85:1642–1650.

Whitney, K. D. 2005. Linking frugivores to thedynamics of a fruit color polymorphism. AmericanJournal of Botany 92:859–867.

Wilder, S. M., A. V. Suarez, and M. D. Eubanks. 2010.The use of simulation modeling to evaluate themechanisms responsible for the nutritional benefitsof food-for-protection mutualisms. Ecological Mod-elling 221:1505–1511.

Wilder, S. M., D. A. Holway, A. V. Suarez, and M. D.Eubanks. 2011. Macronutrient content of plant-based food affects growth of a carnivorousarthropod. Ecology 92:325–332.

Wilson, E. O., and R. W. Taylor. 1967. The ants ofPolynesia (Hymenoptera: Formicidae). Pacific In-sects Monograph 14:1–10.

Yamauchi, K. and K. Ogata. 1995. Social structure andreproductive systems of tramp versus endemic ants(Hymenoptera: Formicidae) of the Ryukyo Islands.Pacific Science 49:55–68.


SAVAGE AND WHITNEY

Documents

Trait-mediated indirect interactions in invasions: unique ... Reprints/Savage and Whitney... · Trait-mediated indirect interactions in invasions: unique behavioral responses of an