Invertebrate community composition differs between invasive herb alligator weed and native sedges

(This is a sample cover image for this issue. The actual cover is not yet available at this time.)

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Original article

Invertebrate community composition differs between invasive herb alligatorweed and native sedges

Imogen E. Bassett a,*, Quentin Paynter b, Jacqueline R. Beggs a

a School of Biological Sciences, University of Auckland, Building 733, Private Bag 92019, Auckland, Tamaki Campus, Auckland, New Zealandb Landcare Research, Auckland, New Zealand

a r t i c l e i n f o

Article history:Received 8 December 2011Accepted 16 April 2012Available online

Keywords:Alligator weedAranaeColeopteraEnemy release hypothesisHemipteraInvasive speciesPlant architecturePlant chemistry

a b s t r a c t

Chemical and/or architectural differences between native and exotic plants may influence invertebratecommunity composition. According to the enemy release hypothesis, invasive weeds should host fewerand less specialised invertebrates than native vegetation. Invertebrate communities were compared oninvasive Alternanthera philoxeroides (alligator weed) and native sedges (Isolepis prolifer and Schoeno-plectus tabernaemontani) in a New Zealand lake. A. philoxeroides is more architecturally and chemicallysimilar to I. prolifer than to S. tabernaemontani. Lower invertebrate abundance, richness and propor-tionally fewer specialists were predicted on A. philoxeroides compared to native sedges, but with greatestdifferences between A. philoxeroides and S. tabernaemontani. A. philoxeroides is more architecturally andchemically similar to I. prolifer than to S. tabernaemontani. Invertebrate abundance showed taxa-specificresponses, rather than consistently lower abundance on A. philoxeroides. Nevertheless, as predicted,invertebrate fauna of A. philoxeroideswas more similar to that of I. prolifer than to S. tabernaemontani. Theprediction of a depauperate native fauna on A. philoxeroides received support from some but not all taxa.All vegetation types hosted generalist-dominated invertebrate communities with simple guild structures.The enemy release hypothesis thus had minimal ability to predict patterns in this system. Results suggestthe extent of architectural and chemical differences between native and invasive vegetation may beuseful in predicting the extent to which they will host different invertebrate communities. However,invertebrate ecology also affects whether invertebrate taxa respond positively or negatively to weedinvasion. Thus, exotic vegetation may support distinct invertebrate communities despite similar overallinvertebrate abundance to native vegetation.

� 2012 Published by Elsevier Masson SAS.

1. Introduction

Invasion of natural communities by exotic species is commonlyregarded as one of the greatest threats to biodiversity (Kolar andLodge, 2001). However, relatively little is known about the effectsof weed invasion on invertebrate communities (Ernst andCappuccino, 2005). Where exotic plants differ in chemistry and/orarchitecture from native plants, they may host differing inverte-brate communities from the pre-existing vegetation. This may bedue to differences in the amount of suitable habitat, food, attach-ment sites or degree of shelter from predators which the plantsoffer to invertebrates.

Invaders which increase structural complexity in an ecosystemoften increase total species density and/or diversity through

provision of more diverse resources (Crooks, 2002; Lawton, 1983).The reverse is true for invaders which decrease structuralcomplexity in invaded systems. Similarly, plant chemistry can becritical in determining invertebrate abundance. Plants with hightissue nitrogen levels and low levels of defensive compoundsrepresent highly palatable food sources for herbivorous inverte-brates compared to low nutrient or heavily defended plant species(McEvoy, 2002). Higher trophic levels may then vary between plantspecies in response to differences in prey abundance, or directly inresponse to structural differences (Hornung and Foote, 2006;Langellotto and Denno, 2004; Pearson, 2009). Those characteristicslisted above represent a trade-off of low structural or chemicaldefences in favour of rapid growth, a strategy characteristic of someinvasive plant species (Cornelissen et al., 1999).

The enemy release hypothesis predicts that introduced plantspecies should host fewer invertebrate herbivores in comparison tothe surrounding native vegetation, as many of the herbivorespresent in the introduced range will be adapted to feed on native

* Corresponding author. Tel.: þ64 9 846 2592; fax: þ64 9 846 2593.E-mail addresses: [email protected] (I.E. Bassett), PaynterQ@

landcareresearch.co.nz (Q. Paynter), [email protected] (J.R. Beggs).

Contents lists available at SciVerse ScienceDirect

Acta Oecologica

journal homepage: www.elsevier .com/locate/actoec

1146-609X/$ e see front matter � 2012 Published by Elsevier Masson SAS.doi:10.1016/j.actao.2012.04.007

Acta Oecologica 41 (2012) 65e73

Author's personal copy

vegetation but not the introduced plant (Keane and Crawley, 2002).However, in a review paper, Colautti et al. (2004) found that whilesome studies did support the prediction of reduced invertebrateabundance on exotic plants, many found higher numbers ofenemies (pathogens as well as invertebrates) on exotic rather thannative plants. This suggests that exotic plants may in fact representnaïve hosts for native herbivores.

Although some invasive plant species may host abundant nativeinvertebrates, the composition of these invertebrate communitiesmay still differ from those on native vegetation (Schooler et al.,2009). The degree of chemical and/or architectural similaritybetween an introduced plant and the native vegetation it invadesmay affect the extent to which this is the case (Douglas andO’Connor, 2003; Harris et al., 2004).

Alternanthera philoxeroides (Mart.) Griseb (Amaranthaceae;alligator weed) is an herbaceous, stoloniferous perennial whichgrows as a terrestrial or emergent aquatic weed. Native to SouthAmerica, it is invasive in many parts of the world. Alligator weedwas first recorded in northern New Zealand in 1906 (Cheeseman,1906). Surveys for potential biological control agents within itsnative range have recorded over 40 species of insects onA. philoxeroides (Maddox et al., 1971). Of those, several have beenintroduced as biocontrol agents in parts of its exotic range. In NewZealand, the most successful of those has been the alligator weedflea beetle (Agasicles hygrophila Selman & Vogt. (Chrysomelidae)),which has been effective in slow-movingwater bodies at restrictingweed growth to near banks. Beyond the presence of introducedbiocontrol agents, little is known about the invertebrate faunahosted by A. philoxeroides outside its native range.

This study sought to examine the invertebrate communitieshosted by A. philoxeroides in lake-margin vegetation in northernNew Zealand, and compare them with invertebrates found on twonative sedge species, Schoenoplectus tabernaemontani (C.C.Gmel.)Palla (Cyperaceae) and I. prolifer (Rottb.) R.Br. (Cyperaceae), in thesame environment. Although A. philoxeroides is herbaceous, sedgeswere chosen for comparison as they are the dominant nativevegetation type at the site, and, like A. philoxeroides, form virtuallymonospecific stands. These two sedge species were chosen as theywere both abundant at the study site and similar ecosystemswithinthe region, and differ considerably from one another in botharchitecture and chemistry. Of the two sedges, I. prolifer is the moresimilar to A. philoxeroides in a number of chemical and architecturalattributes (Table 1).

Overall, it was expected that the three plant species would hostdistinct invertebrate communities. Specifically, the study aimed totest three hypotheses relating to the invertebrate communitieshosted by A. philoxeroides and the sedges. These were: (1) That theabundance of invertebrate herbivores and predators would behigher on the more palatable and architecturally complexA. philoxeroides and I. prolifer than on S. tabernaemontani. (2) That asan exotic plant species, A. philoxeroides would host a lower abun-dance and proportion of native invertebrates than hosted on thenative sedge species. (3) That the abundance and diversity of

specialist herbivores (excluding A. philoxeroides’s introducedbiocontrol agent) would be lower on A. philoxeroides than on eithersedge species, and that this would be associated with a simplifiedherbivore guild composition on A. philoxeroides compared with theherbivore guilds present on the sedges.

2. Methods

2.1. Study site

This study was conducted at Lake Rotokawau East (34�87ʹ06ʺ S,173�32ʹ22ʺ E) on the Karikari Peninsula, Northland, New Zealand.The lake is 21.3 ha in area, with an iron-pan base (Champion et al.,2005) and lake-margin vegetation characterised by small-scalepatchiness, typical of the numerous small, shallow lakes in theregion. A. philoxeroides forms extensive monospecific clumpscovering more than 20% of the lake’s margin (Bassett et al., inpress). I. prolifer and S. tabernaemontani each cover approximately3e10% (seasonally variable) of the lake margin, being amongst themost widespread of the native plants occupying the same elevationzone as alligator weed. The lake is surrounded by terrestrial scrubcomprised of a mixture of native species such as manuka (Lep-tospermum scoparium; Myrtaceae), bracken (Pteridium esculentum;Pteridaceae) and flax (Phormium tenax; Hemerocallidaceae), withsome exotics, such as Acacia longifolia (Mimosoideae). Where theground becomes seasonally inundated, this scrub gives way toa mosaic of native sedges and reeds (e.g. Machaerina juncea(previously Baumea) M. teretifolia, Eleocharis acuta, E. sphacelata(Cyperaceae), Typha orientalis (Typhaceae)), small aquatic herbs(e.g. Myriophyllum propinquum (Haloragaceae), Glossostigma elati-noides (Phrymaceae)) and the larger herb Persicaria decipiens(Polygonaceae). Several other exotics (most notably Paspalum spp.(Poaceae) and Lotus pedunculatus (Fabaceae)) also occupy season-ally inundated ground though in all cases with much lessercoverage than A. philoxeroides. The surrounding landscape is mostlyhuman-modified, predominantly pastoral farming as is the casethroughout most of A. philoxeroides’s current New Zealand distri-bution. The study was conducted at this single lake in a modifiedlandscape as a result of difficulty in obtaining study sites in lessmodified ecosystems due to the strong imperative to controlA. philoxeroides in less degraded ecosystems. During the studyperiod, water depth in the sampling zone ranged from0.24 m � 0.05 m in late summer to 0.69 m � 0.03 m in mid winter.The climate is subtropical, characterised by mild, frost-free wintersand warm summers.

2.2. Sampling methods

The invertebrate fauna on A. philoxeroides and the two sedgespecies was characterised using two complimentary samplingmethods. These were hand removal of invertebrates from excisedvegetation, and collection via emergence traps. Patches for bothsampling methods were selected using criteria of patch size

Table 1Chemical and architectural characteristics of study plants. Chemistry given as percent of dry mass. Fibre is defined as the plant fraction comprised of cellulose, lignin and ash(see Rowland and Roberts, 1994 for methods). Chemistry and biomass taken from Bassett et al. (2010). Range in biomass values given indicates annual variation. Heights areapproximate, taken from Johnson and Brooke (1998) (sedges) and pers. obs. I. Bassett (alligator weed).

Plant species Fibre(%)

Cellulose(%)

Lignin(%)

Nitrogen(%)

Phosphorous(%)

Potassium(%)

Height(m)

Biomass(kg/m2)

Architecture

Alligator weed 25.0 20.3 4.1 2.1 0.29 5.2 0.5 0.1e1.7 Complex (Many branchingleaves and stems)

S. tabernaemontani 42.8 37.2 4.9 1.0 0.12 1.8 2 1e2 Simple (Thick vertical culms)I. prolifer 27.5 25.1 1.6 2.7 0.28 3.4 0.6 0.5e1.5 Complex (Densely inter-

arching culms)

I.E. Bassett et al. / Acta Oecologica 41 (2012) 65e7366

Author's personal copy

4e6 m2, with 2e100 m between patches and as much as possiblealternating among the three plant species such that they wereinterspersed around the lake perimeter.

Excised vegetation involved removal of all vegetation at ground-level from 25 cm � 25 cm quadrats (0.0625 m2 in area).One randomly selected quadrat was excised from each of fivereplicate patches of each vegetation type monthly fromFebruaryeApril 2006 and then again from October 2006eApril2007 (5 replicates � 3 vegetation types � 10 months ¼ 150samples). No quadrat was sampled more than once. In almost allmonths at least part of the vegetation removed from each plot wassubmerged below the water-line, and this was collected in additionto that part of the plant which was emergent above the water. Allexcised vegetation was sealed in ziplock bags and transported inchilled, insulated containers until it could be refrigerated to mini-mise decomposition of plant and animal matter, and to reduceinvertebrate activity and resulting predation. All vegetation wasvisually inspected at up to 450� magnification. All invertebratesdetected were removed, and placed into 70% ethanol.

The second method employed emergence traps placed overvegetation. Emergence traps catch emerging adults of invertebratesthat have spent juvenile life stages in the ground and/or water. Thistrappingmethod was expected to better capture highlymobile taxathat may be disturbed by the collector in the excised vegetationmethod, as well as nocturnal species. It also allowed continuoussampling, rather than the excised vegetation method whichsampled invertebrate fauna present on a single day each month.Traps were 0.5 m � 0.5 m wide (0.025 m2), 1 m tall, and coveredwith 1mmmesh. The tops of the traps were pyramidal, with pipingfrom the apex leading to a removable 400 ml collecting jar partiallyfilled with glycol (50% solution) as a preservative. Jars werereplaced monthly, and invertebrates sorted and transferred to 70%ethanol.

One emergence trap was erected in each of four replicatepatches of each of the three vegetation types, and sampled fromJanuary 2006eMarch 2006, and then again from October2006eMarch 2007, with traps moved to a new position within thepatch between sampling seasons to prevent the second season’scollections being affected by restricted invertebrate access in thefirst season. Therefore we had 4 replicates� 3 vegetation types � 9months (¼108 samples). However, vandalism and extremeweatherresulted in the loss of data for at least one month for most traps.Therefore results for emergence trap data are presented as averageabundance per trap per month (3 vegetation types � 4 traps).

2.3. Focal taxa

Total invertebrate abundance was counted for both samplingmethods. In addition, three focal taxa were used to address theresearch hypotheses, as invertebrates may show taxa-specificresponses to vegetation type (Di Giulio et al., 2001; Gibson et al.,1992). The three taxa chosen were Hemiptera, Coleoptera andAraneae. The order Hemiptera comprises predominantly herbivo-rous species, and is thus likely to be strongly influenced by plantcommunity composition. Hemiptera collected from excised vege-tation were sorted to morphospecies, then identified to specieslevel. Species were categorised as native or exotic based on theChecklist of New Zealand Hemiptera (Larivière, 2005), and classi-fied by feeding habit based on available literature. Four morpho-species could not be assigned biostatus and/or feeding habit due toinsufficient taxonomic resolution.

Araneae were used to examine the effects of vegetation type onpredatory invertebrates. Preliminary analyses suggested thatdiffering abundance of the genus Tetragnatha comprised most ofthe difference in Araneae abundance between vegetation types.

Araneae from excised vegetation were therefore sorted into twocategories, being Tetragnatha spp. and “other” spiders.

Very few Hemiptera were recorded from emergence traps.Instead, we focused on Coleoptera from emergence traps becausethey represented a substantial component of emergence trapcatches (but not excised vegetation catches) from all vegetationtypes. In addition, Coleoptera are awell-described, functionally andtaxonomically diverse order, representing approximately 50% ofdescribed insect species in New Zealand (Watt, 1982). Coleopterahave also been shown to respond to differences in vegetation cover(Brose, 2003; Harris et al., 2004). Coleoptera were sorted to mor-phospecies then identified to species level. Where possible, pub-lished literature was used to categorise beetles as native or exoticand to assign to feeding groups. Six morphospecies could not beassigned biostatus due to inadequate taxonomic resolution, whileavailable information on feeding modes was insufficient for sevenspecies.

2.4. Data analyses

From excised vegetation, preliminary analyses indicated novegetation type � sampling date or season interactions. Further,sampling date effects are peripheral to the hypotheses tested here.Therefore, abundances were averaged for each patch over thewhole sampling period. One-way ANOVA were then used on thesepatch averages to test for differences in abundance betweenvegetation types. ShapiroeWilk and Levene’s tests were used todetermine whether data were normally distributed and varianceshomogeneous. Where necessary, data were transformed to over-come these problems. In all cases, log transformations were used,except for Tetragnatha spp. abundance, which was square-roottransformed. Medians are presented instead of means where datahave been transformed.

To overcome loss of some emergence traps, we determinedmean monthly abundance for each trap (averaged across allsampling months where the trap was operational). We then usedtraps as replicates in one-way ANOVA to examine differences inabundance among vegetation types, using traps as replicates(n ¼ 12). We were able to do this as loss of traps tended to occuracross all vegetation types evenly in the samemonths due to eventssuch as extreme weather, and preliminary analyses indicated nointeractions between vegetation type and sampling month orseason. Tukey’s post hoc tests were used to determine pair-wisedifferences.

For both Hemiptera and Coleoptera, differences in relativeproportions for feeding guilds and biostatus were examined usingChi-squared tests, or Fisher’s Exact tests where Chi-squaredassumptions were breached. These analyses were performed onthe sum of all individuals caught in each vegetation type over theentire sampling period. Thus we tested, for instance, a 2 � 3contingency table for the relative frequency of native vs exoticstatus across the three vegetation types. For Coleoptera guildcomposition, we excluded species which were known to be specificto other neighbouring vegetation, as they were assumed to be‘tourists’ rather than feeding on the vegetation being sampled. Wealso excluded the alligator weed flea beetle from this analysis.Bonferroni-corrected Fisher’s Exact tests were used post hoc todetermine pair-wise differences.

One-way ANOSIM was used to determine the extent of separa-tion between Hemiptera communities on the different vegetationtypes (Clarke and Warwick, 2001). Data were square-root trans-formed to reduce the influence of highly abundant species (Clarkeand Warwick, 2001). A similarity matrix was then created usingthe BrayeCurtis coefficient (Clarke and Warwick, 2001). ANOSIMuses the dissimilarity matrix to generate an R-statistic which

I.E. Bassett et al. / Acta Oecologica 41 (2012) 65e73 67

Author's personal copy

reflects the relative compositional differences in the Hemipteracommunity between groups (alligator weed, S. tabernaemontani,I. prolifer) vs within groups, and varies between �1 and 1, with0 indicating completely random grouping, and 1 indicating totalseparation of groups. SIMPER analysis was then used to determinewhich Hemiptera species were most influential in separatingcommunities (Clarke and Warwick, 2001). Due to low abundance,both ANOSIM and SIMPER were performed using patches asreplicates, including all Hemiptera caught over the duration of theexperiment for each patch. Non-metric multi-dimensional scaling(nMDS) was used to visualise differences in Hemiptera communitycomposition between vegetation types. Statistical analyses wereconducted in PRIMER 6 and R v2.2.1.

3. Results

3.1. Invertebrate abundance

From excised vegetation, average abundance of both spiders andHemiptera was higher on S. tabernaemontani than on eitherA. philoxeroides or I. prolifer (Table 2). For Hemiptera, these differ-ences were primarily due to abundant aphids (Rhopalosiphum padi(Linnaeus)) and mealy bugs (Pseudococcus longispinus (TargioniTozzetti) and Balanococcus sp.) on S. tabernaemontani. In contrast,when only predatory Hemiptera were examined, there were nosignificant differences between vegetation types (Table 2). Highspider abundance on S. tabernaemontani mainly reflected highabundance of the genus Tetragnatha on this structurally simplenative sedge (Table 2). In contrast, “other” spiders and overallinvertebrate abundance showed no differences between vegetationtypes (Table 2).

Emergence traps caught a different suite of invertebratescompared to the excised vegetation samples, and this portion of thecommunity showed a different response to vegetation. Meaninvertebrate abundance was higher for A. philoxeroides trapscompared to S. tabernaemontani traps (Table 3). No differences inColeoptera abundance were detected among vegetation types,although Coleoptera showed the same trend as total invertebrateabundance (Table 3).

3.2. Biostatus

Native Hemiptera were less abundant on A. philoxeroides thanon either sedge species (Table 4), and comprised a lower proportionof total abundance on A. philoxeroides compared to I. prolifer, butnot compared to S. tabernaemontani (c22 ¼ 132.0, P < 0.001; Fig. 1).Exotic Hemiptera were more abundant on S. tabernaemontani thanon either A. philoxeroides or I. prolifer (Table 4). At the species level,no differences were detected among vegetation types in meannative or exotic Hemipteran species richness (Table 4), or therelative proportions of native and exotic Hemiptera species for allmonths combined (Fisher’s Exact test, P ¼ 0.8; Fig. 1).

The proportions of native and exotic Coleoptera caught inemergence traps (all months combined) varied with vegetationtype (c22 ¼ 92.0, P < 0.001; Fig. 1). I. prolifer traps were stronglydominated by native Coleoptera, particularly members of thefamily Scirtidae. Native Coleoptera comprised a lower proportion ofthe total catch from A. philoxeroides than from I. prolifer, but stillcomprised over half of the total catch from A. philoxeroides. Incontrast, over half the Coleoptera collected from S. tabernaemontanitraps were exotic. However, at the species level all vegetation typeswere dominated by exotic Coleoptera (Fisher’s Exact test, P ¼ 0.60;Fig. 1).

3.3. Specialist herbivores

Only two feeding guilds of phytophagous Hemiptera werecaptured; sap-suckers and seed-feeders. I. prolifer hosted a higherproportion of seed-feeders than either of the other vegetation types(Fisher’s Exact test, P < 0.001). This reflected the absence of sap-sucking aphids from I. prolifer, and a high abundance of the seed-feeder Cymus novaezelandiae Woodward (Cymidae) on this sedge.At the species level these differences in guild composition were nolonger evident, with no differences detected among vegetationtypes, probably reflecting low species richness (Fisher’s Exact test,P ¼ 0.4).

After excluding Coleoptera species known to be specific toneighbouring vegetation, species of unknown feeding preferences,and the deliberately introduced biocontrol agent A. hygrophila, nodifferences in Coleoptera phytophage guild composition weredetected among vegetation types (Fisher’s Exact test, P ¼ 0.7). Rootfeeders were most abundant in all vegetation types, with stemfeeders the only other guild represented.

All the phytophagous Hemiptera collected from all vegetationtypes have been recorded on other host plant species (Chinery,1985; Cox, 1987; Lariviere and Larochelle, 2004), and further-more, all those collected from A. philoxeroides were also collectedfrom at least one of the sedges. Agasicles hygrohila was the onlyColeoptera species known to be specific to one of the three studyplant species. Approximately one third of phytophagous Coleopteracollected (67 individuals from 4 species) were tourist species (sensuMoran and Southwood, 1982), known to be specific to otherneighbouring (mainly exotic) plant species (Appendix A). All otherbeetles collected were known to be generalists, or have inade-quately documented biologies to determine their specificity.

3.4. Overall community similarity

All vegetation types hosted distinctly different Hemipteracommunities (Fig. 2). As previously discussed, S. tabernaemontaniwas characterised by abundant aphids (R. padi) and mealy bugs(Balanococcus sp.). These two species contributed 69% of thedifference between S. tabernaemontani and Alligator weed, and 57%of the difference between S. tabernaemontani and I. prolifer (SIMPER

Table 2Differences among vegetation types in faunal abundance of focal taxa. Averages are per 0.0625 m2 of excised vegetation. Medians are given instead of means for data whichwere log transformed (or in the case of Tetragnatha spp. abundance, square-root transformed) to meet assumptions of normality and/or homoscedasticity. Different lettersindicate pair-wise differences at *P < 0.05, **P < 0.01, ***P < 0.001.SEM ¼ standard error of the mean, IQR ¼ inter quartile range.

Mean (SEM) or Median (IQR) Fdf P

A. philoxeroides S. tabernaemontani I. prolifer

Mean invertebrate abundance 40.6 (5.8) 37.8 (5.3) 28.8 (5.0) F2,12 ¼ 1.3 0.3Mean total spider abundance 1 (0.1)a 4.3 (0.4)b*** 1.2 (0.4)a F2,12 ¼ 29.2 <0.001Median Tetragnatha spp. abundance 0.5 (0.2)a 3.2 (0.8)b*** 0.2 (0.3)c* F2,12 ¼ 67.9 <0.001Median “other” spider abundance 0.4 (0.4) 0.8 (0.2) 0.9 (1.3) F2,12 ¼ 1.8 0.2Median Hemiptera abundance 1.2 (0.2)a 15.8 (9)b** 2.5 (3.2)a F2,12 ¼ 20.3 <0.001Median predatory Hemiptera abundance 0.4 (0.4) 0.3 (0.3) 0.7 (1.4) F2,12 ¼ 2.6 0.1

I.E. Bassett et al. / Acta Oecologica 41 (2012) 65e7368

Author's personal copy

analysis). No single species played such a dominant role in differ-entiating Hemiptera communities on alligator weed from those onI. prolifer. C. novaezelandiae was the most influential speciesdifferentiating alligator weed and I. prolifer, contributing 23% of theobserved difference (SIMPER analysis).

4. Discussion

Contrary to expectation, specialist insect herbivoreswere absentfrom all three vegetation types, with the exception of touristspecies and one biocontrol agent on A. philoxeroides. The

invertebrate community was predominantly generalist, with a highproportion of introduced species. Invertebrate abundance anddiversity showed taxa-specific responses to vegetation type, ratherthan consistently lower abundance on A. philoxeroides. Thus theenemy release hypothesis had minimal ability to explain patternsamong these plant species.

4.1. Influence of plant attributes on invertebrate abundance

Invertebrate abundance was predicted to be higher on the morepalatable and architecturally complex A. philoxeroides and I. prolifer

Table 3Differences among vegetation types in abundance from emergence traps. Figures given are per trap, averaged across all sampling months that trap was operative. Differentletters indicate pair-wise differences at *P < 0.05, **P < 0.01, ***P < 0.001.

Mean (SEM) Fdf P

A. philoxeroides S. tabernaemontani I. prolifer

Mean total invertebrate abundance 81.7 (12.4)a* 32.8 (6.8)b 58.2 (12.9)ab F2,9 ¼ 4.9 0.04Mean Coleoptera abundance 15.8 (4.5) 7.3 (1.6) 15.3 (3.6) F2,9 ¼ 1.9 0.2

Table 4Differences among vegetation types in abundance and species richness of native and exotic Hemiptera from excised vegetation. Medians are given instead of means for datawhich were log transformed to meet assumptions of normality and/or homoscedasticity. Averages are per replicate patch for all months combined. Different letters indicatepair-wise differences at *P < 0.05, **P < 0.01, ***P < 0.001. SE ¼ standard error, IQR ¼ inter quartile range.

Mean (SE) or Median (IQR) Fdf P

A. philoxeroides S. tabernaemontani I. prolifer

Median native abundance 2 (6)a 75 (59)b*** 20 (38)b* F2,14 ¼ 15.3 <0.001Median exotic abundance 6 (5)a** 91 (59)b 4 (1)a*** F2,14 ¼ 16.2 <0.001Mean native species richness 1.6 (0.6) 2.2 (0.4) 3.2(0.4) F2,14 ¼ 3.1 0.08Mean exotic species richness 1.0 (0.3) 1.4 (0.3) 1.0 (0.3) F2,14 ¼ 0.6 0.6

Fig. 1. Proportional native and exotic composition of Hemiptera and Coleoptera by abundance and species richness, for all sampling dates combined. Vegetation types with differentletters differ in the proportions of natives and exotics (P < 0.001 in all Bonferroni-corrected pair-wise Fisher’s Exact tests). See Tables 2e4 for significance of differences in abundanceamong vegetation types.

I.E. Bassett et al. / Acta Oecologica 41 (2012) 65e73 69

Author's personal copy

than on the tough culms and simple structure provided byS. tabernaemontani. This was the case for some, but not all, taxa.These taxa-specific results are consistent with other studies ofinvertebrate fauna on invasive plants such as Hymenachneamplexicaulis (Rudge) Nees (Poaceae), Fallopia spp. (Polygonaceae)and Chrysanthemoides monilifera ssp. rotundata (L.) Norlindh(Asteraceae) (Gerber et al., 2008; Houston and Duivenvoorden,2002; Lindsay and French, 2006; Topp et al., 2008). Thus, plantchemistry, architecture and other attributes influence the associ-ated invertebrate community by interacting with characteristics ofthe invertebrates, rather than in isolation. However, for most taxainvertebrate abundance did not differ between A. philoxeroides andI. prolifer, supporting the hypothesis that these more architecturallyand chemically similar plants would support more similar abun-dances of invertebrates than would the more dissimilarS. tabernaemontani.

Our study did not experimentally test the cause of differences inabundance between plant species. It is thus possible that there mayhave been an indirect, trophic effect influencing predator abun-dance. However, both observational surveys and experimentalmanipulation elsewhere found architectural differences directlyaffect spider abundance and community composition (Pearson,2009; Pétillon et al., 2005; Woodcock et al., 2007), and it ispossible that this direct plant e predator effect is occurring at LakeRotokawau as well. Web-building spiders, particularly Tetragnathaspp. are highly correlated with architectural features of vegetation,reflecting specific web attachment requirements (Aiken and Coyle,2000; Gratton and Denno, 2005; Rypstra et al., 1999; Vandergastand Gillespie, 2004). The tall, simple culms of S. tabernaemontaniappeared to provide ideal camouflage for Tetragnathids. In addi-tion, they allowed creation of larger webs spanning open, elevatedspaces, enabling interception of flighted species not possible in theshorter, denser A. philoxeroides or I. prolifer vegetation.

In contrast to Tetragnathids, other spiders, predominantlyPisauridae, did not prefer S. tabernaemontani over the more archi-tecturally complex plant species. Pisauridae construct their char-acteristic, dense “nursery-webs” on scrub, grass or sedgevegetation to protect their young (Forster and Forster, 1999), ratherthan slinging large, prey catchingwebs across open spaces as do theTetragnathids. This is consistent with architectural differencesbeing a key driver of spider abundance in this system. Structurallysimple vegetation such as S. tabernaemontani may also preferen-tially support large, dominant invertebrate predators such as Tet-ragnathid spiders, at the expense of other predatory taxa (Finke andDenno, 2002; Hatley and MacMahon, 1980).

4.2. Influence of plant attributes on invertebrate biostatus

As an exotic plant species, A. philoxeroideswas predicted to hosta lower proportion of native invertebrates than hosted by either ofthe native sedge species studied. Overall, A. philoxeroides hosteda less native-dominated invertebrate fauna than one, but not both,of the native plant species tested in our study. Harris et al. (2004)found comparable dominance of New Zealand native inverte-brates on invasive gorse (Ulex europaeus L. (Fabaceae)) and nativeKanuka (Kunzea ericoides (A.Rich.) Joy Thomps. (Myrtaceae)),indicating that exotic plants do not necessarily represent a poorquality habitat for native invertebrates. However, the speciescomposition of native invertebrates on both gorse andA. philoxeroides differed from that on native vegetation, suggestingthat a subset of native invertebrates may be better than others atutilising exotic plants as habitat. It appears that either enemyrelease plays little role in alligator weed’s invasiveness, or thepressure exerted on alligator weed by invertebrates in its intro-duced range is less than that in its native range. Whether

invertebrates provide biotic resistance against introduced plantsmay depend on invertebrate guild structure, along with plant-specific traits such as life-history and degree of pre-adaptation toherbivory (Maron and Vilà, 2001), and other extrinsic factors whichpromote invasion.

4.3. Influence of plant attributes on specialist herbivores

A. philoxeroides was predicted to host a simplified herbivorefauna, with fewer specialist herbivores than the native sedges.However, the only specialists collected were A. philoxeroides’sbiocontrol agent and species associated with neighbouring plants(tourists). This could be explained by two alternative hypotheses;either these lakes (or specifically these two native sedges) arehistorically generalist-dominated systems, or specialist speciespreviously present have been lost due to factors such as habitatmodification.

There is a paucity of records of host specific herbivores of anyinvertebrate taxa associated with Northland dune lakes in general,or either of the two sedge species studied here. A moth, Bactraoptanias Meyrick, is associated with S. tabernaemontani (Crop andFood Research, 2010). In the United Kingdom, four herbivorousspecies of Hemiptera and one Lepidoptera are known to be hostedby S. tabernaemontani (Biological Records Centre, n.d). All aregeneralist species also recorded from several other host plantspecies. No specific herbivores are recorded for I. prolifer (Crop andFood Research, 2010). World-wide, some taxa within both Hemi-ptera (e.g. some members of Delphacidae and Miridae) and Cole-optera (e.g. some members of Curculionidae) are known for theirclose associations with Cyperaceae and Juncaceae (Bartlett andWheeler, 2007; Biological Records Centre, n.d; Crop and FoodResearch, 2010; Novotný, 1995; Wheeler, 2011). However, manyof these invertebrates utilise multiple host-plants from within thesame plant genus, family or higher taxonomic grouping (Bartlettand Wheeler, 2007; Biological Records Centre, n.d; Crop and FoodResearch, 2010; Novotný, 1995). Insufficient is known about theecology of many other Hemiptera and Coeloptera to determine hostspecificity with any reliability (Bartlett and Wheeler, 2007).Generalist taxa have been found to dominate the benthic fauna of

Fig. 2. Clustering of Hemiptera communities by vegetation type. Points close togetherare more similar in composition than points further apart. AW ¼ A. philoxeroides,S ¼ S. tabernaemontani, I ¼ I. prolifer. ANOSIM differences between vegetation types asfollows; A. philoxeroides e S. tabernaemontani pair-wise R-statistic ¼ 0.79, P < 0.01;A. philoxeroides e I. prolifer pair-wise R-statistic ¼ 0.89, P < 0.01; S. tabernaemontani eI. prolifer pair-wise R-statistic ¼ 1, P < 0.01. R-statistics closer to 1 indicate greaterseparation between vegetation types.

I.E. Bassett et al. / Acta Oecologica 41 (2012) 65e7370

Author's personal copy

South Island lakes and New Zealand streams (Thompson andTownsend, 2000; Wissinger et al., 2006), suggesting that gener-alist taxa may dominate many aquatic systems in New Zealand,although this likely reflects generalists from orders other thanthose focussed on in this study (e.g. Plecoptera, Trichoptera).However, other native plants within this system may host a morespecialised fauna. We collected four Coleoptera species stronglyassociated with native Phormium spp. and other native Agavaceae,although these host-plants represent the more terrestrial edge ofthe study site.

Other ecosystems show a trend towards increasing generalistdominance with increasing habitat loss or modification (Niell et al.,2007; Niemela et al., 2007; Nitterus et al., 2007; Taki and Kevan,2007), and an effect of the surrounding vegetation matrix on hostplant occupancy (Kuschel, 1990; Pawson et al., 2008). Species’persistence in patches of suitable habitat depends on both theamount of habitat and the spatial configuration of the habitat at thelandscape-scale (Hanski, 1998). This suggests habitat modificationlikely contributed to the generalist and exotic dominance of theinvertebrate fauna in our study.

Further research is thus required in a less modified landscape todetermine whether A. philoxeroides would attract an unusuallyexotic and/or generalist fauna relative to native plants wherea greater native invertebrate species pool was available. Given thelack of site-level replication in this study, the hypotheses examinedherewould also benefit by being tested further with either alligatorweed and/or other invasive plant species across replicated invadedsites.

4.4. Conclusion

All three plant species hosted distinctly different invertebratecommunities. However, as predicted based on a higher degree of

architectural and chemical similarity, the invertebrate fauna onA. philoxeroides was more similar to that on I. prolifer than that onS. tabernaemontani. Thus, the extent to which invasive plants differchemically and/or architecturally from the vegetation they invademay be useful in predicting the extent to which they will hostdifferent invertebrate communities, and we can expectA. philoxeroides to have greater impact on invertebrate communitieswhere it invades stands of S. tabernaemontani thanwhere it invadesI. prolifer. However, the direction of the effect was not consistentacross invertebrate taxa, with some more and others less abundanton S. tabernaemontani. Similarly, proportion and abundance ofnative species on the three vegetation types varied across inverte-brate taxa. These taxa-specific responses indicate that plant char-acteristics interact with invertebrate characteristics in determiningpatternsof abundance, andwhileA. philoxeroidesmaybeable tohostconsiderable native invertebrate communities, these communitiesare likely to differ in composition from those on native vegetationwhen it invades sedge-dominated systems. In general, the enemyrelease hypothesis hadminimal ability to predict patterns of feedingguild and specialisation in this system due to the dominance ofgeneralist invertebrates across all three plant species. However lackof basic ecological knowledge fromsuch systems restricts our abilityto attribute the generalist fauna to landscape modification ratherthan historical generalist dominance.

Acknowledgements

This study was funded by Landcare Research, The CooperativeResearch Centre for Australian Weed Management and the Univer-sity of Auckland. Thanks to Richard Green for assistance with field-work, to StephenThorpe, Rich Leschen, GraceHall, David Teulon andRosa Henderson for taxonomic assistance, and to anonymousreferees for comments on earlier drafts of this manuscript.

Appendix A

Hemiptera and Coleoptera species captured from excised vegetation and emergence traps respectively, with their biostatus and feeding classifications, and the host plant(s)from which they were collected.

Invertebrateorder

Invertebratefamily

Invertebe species Biostatus Trophic level Herbivorefeeding guild

Herbivorespecialization(host plant)

Plant(s) collected from

Hemiptera Aphididae Rhopalosiphumpadi Linnaeus

Introduced Herbivore Sap Generalist A. philoxeroides,S. tabernaemontani

Ceratocombidae Ceratocombus sp. Native Predator N/A N/A I. proliferCymidae Cymus novaezaelandiae

WoodwardNative Herbivore Seeds Generalist I. prolifer

Delphacidae Morphospecies 1 Unknown Herbivore Unknown Unknown I. prolifer, S. tabernaemontaniHydrometridae Hydrometra strigosa Skuse Native Predator N/A N/A A. philoxeroides, I. prolifer,

S. tabernaemontaniLygaeidae Nysius huttoni White Native Herbivore Sap and seeds Generalist A. philoxeroidesMiridae Morphospecies 1 Unknown Unknown Unknown Unknown A. philoxeroides, I. prolifer,

S. tabernaemontaniPentatomidae Cermatulus nasalis

nasalis WestwoodNative Predator N/A N/A A. philoxeroides

Dictyotus caenosusWestwood

Introduced Herbivore Sap Generalist A. philoxeroides,S. tabernaemontani

Oechalia schellenbergiiGuérin-Méneville

Native Predator N/A N/A A. philoxeroides

Pseudococcidae Balanococcus sp. Native Herbivore Sap Generalist I. prolifer, S. tabernaemontaniPseudococcus longispinusTargioni Tozzetti

Introduced Herbivore Sap Generalist I. prolifer, S. tabernaemontani

Rhyparochromidae Morphospecies 1 Unknown Herbivore Seeds Unknown A. philoxeroides, I. proliferRemaudiereanainornata Walker

Introduced Herbivore Seeds Generalist I. prolifer

Saldidae Morphospecies 1 Unknown Predator N/A N/A A. philoxeroidesVeliidae Microvelia sp. Native Predator N/A N/A A. philoxeroides, I. prolifer,

S. tabernaemontani

(continued on next page)

I.E. Bassett et al. / Acta Oecologica 41 (2012) 65e73 71

Author's personal copy

(continued )

Invertebrateorder

Invertebratefamily

Invertebe species Biostatus Trophic level Herbivorefeeding guild

Herbivorespecialization(host plant)

Plant(s) collected from

Coleoptera Anthicidae Anthicus ?kreusleri King Introduced Scavenger N/A N/A A. philoxeroides, I. prolifer,S. tabernaemontani

Anthribidae Micranthribus atomus Sharp Native Fungivore N/A N/A I. proliferSharpius brouni Sharp Native Fungivore N/A N/A I. prolifer, S. tabernaemontani

Brentidae Exapion ulicis Forster Introduced Herbivore Seeds Specialist(Ulex europaeus)

A. philoxeroides, I. prolifer,S. tabernaemontani

Carabidae Notagonum sp. Unknown Predator N/A N/A A. philoxeroides, I. proliferHypharpax australis Dejean Introduced Herbivore Seeds N/A I. prolifer, S. tabernaemontaniEuthenarus puncticollis Bates Native Herbivore Seeds Unknown I. prolifer

Chrysomelidae Agasicles hygrophilaSelman and Vogt

Introduced Herbivore Foliage Specialist(A. philoxeroides)

A. philoxeroides

Coccinellidae Coccinella undecimpunctata L. Introduced Predator N/A N/A A. philoxeroides,S. tabernaemontani

Diomus sp. Introduced Predator N/A N/A I. proliferScymnus loewi Mulsant Introduced Predator N/A N/A A. philoxeroides,

S. tabernaemontaniUnidentified sp. Unknown Predator N/A N/A A. philoxeroides, I. prolifer,

S. tabernaemontaniCorylophidae Holopsis sp. Native Fungivore N/A N/A A. philoxeroides, I. prolifer

Sericoderus sp. Introduced Fungivore N/A N/A A. philoxeroides, I. prolifer,S. tabernaemontani

Clypastraea pulchella Lea Introduced Fungivore N/A N/A I. proliferCurculionidae Microcryptorhynchus sp. Native Herbivore Unknown Unknown A. philoxeroides

Microtribus huttoni Wollaston Native Herbivore Stems Generalist A. philoxeroides, I. proliferNaupactus leucoloma Boheman Introduced Herbivore Roots Generalist A. philoxeroidesPhloeophagosomapedatum Wollaston

Native Herbivore Unknown Specialist(Phormium spp.)

A. philoxeroides, I. prolifer

Sericotrogussubaenescens Wollaston

Native Herbivore Unknown Generalist A. philoxeroides, I. prolifer

Sitona lepidus Gyllenhal Introduced Herbivore Roots Specialist(Trifolium spp.)

A. philoxeroides, I. prolifer,S. tabernaemontani

Steriphus ascitus Pascoe Native Herbivore Roots(as juvenile)

Generalist A. philoxeroides

Steriphus diversipeslineatus Pascoe

Introduced Herbivore Unknown Generalist A. philoxeroides

Unidentified sp. (Cossoninae) Unknown Herbivore Stems A. philoxeroides,S. tabernaemontani

Eucossonus sp. Native Herbivore Unknown Generalist S. tabernaemontaniNovitas sp. Native Herbivore Unknown Generalist S. tabernaemontaniSteriphus ascitus Pascoe Native Herbivore Roots Generalist S. tabernaemontaniStoreus albosignatus Blackburn Introduced Herbivore Seeds Specialist

(Acacia longifolia)S. tabernaemontani

Dermestidae Hexanodes vulgate Broun Native HerbivoreElateridae Conoderus exsul Sharp Introduced Omnivore Foliage Generalist A. philoxeroides

Conoderus posticus Eschscholtz Introduced Herbivore Foliage Generalist A. philoxeroidesOchosternus zealandicus White Native Unknown N/A N/A A. philoxeroidesAgrypnus variabilis Candèze Introduced Ominvore Roots

and seedsGeneralist I. prolifer

Hydrophilidae Enochrus maculiceps Macleay Introduced Scavenger N/A N/A A. philoxeroidesCercyon sp. Introduced Scavenger N/A N/A S. tabernaemontani

Latridiidae Melanophthalma sp. 1 Native Fungivore N/A N/A A. philoxeroides, I. prolifer,S. tabernaemontani

Melanophthalma sp. 2 Native Fungivore N/A N/A S. tabernaemontaniAridius sp. Introduced Fungivore N/A N/A S. tabernaemontaniCortinicaria sp. Unknown Fungivore N/A N/A A. philoxeroides, I. prolifer,

S. tabernaemontaniCorticaria sp. Introduced Fungivore N/A N/A S. tabernaemontaniUnidentified sp. 1 Unknown Fungivore N/A N/A I. prolifer

Mycetophagidae Litargus vestitus Sharp Introduced Fungivore N/A N/A A. philoxeroidesSalpingidae Salpingus sp. Native Fungivore N/A N/A I. proliferScirtidae Unidentified spp. Native Detritivore

(as juvenile)N/A N/A A. philoxeroides,

S. tabernaemontaniSilvanidae Cryptamorpha desjardinsi

Guérin-MénevilleIntroduced Fungivore N/A N/A A. philoxeroides, I. prolifer,

S. tabernaemontaniStaphylinidae Pselaphophus atriventris

WestwoodIntroduced Predator N/A N/A I. prolifer

Unidentified sp. (Pselaphinae) Unknown Predator N/A N/A A. philoxeroidesUnidentified species(Aleocharinae: Gyrophaenina)

Native Fungivore N/A N/A S. tabernaemontani

I.E. Bassett et al. / Acta Oecologica 41 (2012) 65e7372

Author's personal copy

References

Aiken, M., Coyle, F.A., 2000. Habitat distribution, life history and behaviour of Tet-ragnatha spider species in the Great Smoky Mountains National Park.J. Arachnol. 28, 97e106.

Bartlett, C.R., Wheeler, A.G., 2007. Kelisia and Stenocranus species (Hemiptera:Fulgoromorpha: Delphacidae): new host-plant associations and distributionalrecords of eight sedge-feeding planthoppers. Proc. Entomol. Soc. Washington109, 400e415.

Bassett, I., Paynter, Q., Hankin, R., Beggs, J., in press. Characterising alligator weed(Alternanthera philoxeroides; Amaranthaceae) invasion at a northern NewZealand lake. New Zeal. J. Ecol.

Bassett, I.E., Beggs, J.R., Paynter, Q., 2010. Decomposition dynamics of invasivealligator weed compared with native sedges in a Northland lake. New Zeal. J.Ecol. 34, 324e331.

Biological Records Centre, n.d.. Database of Insects and their Food Plants. http://www.brc.ac.uk/dbif/hostsresults.aspx?hostid¼5074. (accessed 3.11.11).

Brose, U., 2003. Bottom-up control of carabid beetle communities in early succes-sional wetlands: mediated by vegetation structure or plant diversity? Oecologia135, 407e413.

Champion, P., Wells, R., Matheson, F., Winton, M., 2005. Northland Lakes Assess-ment 2004/05, NIWA Client Report: HAM2005e113, NRC05215. NorthlandRegional Council, p. 265.

Cheeseman, T.F., 1906. Manual of the New Zealand Flora. John Mackay GovernmentPrinter, Wellington.

Chinery, M., 1985. A Field Guide to the Insects of Britain and Northern Europe.Collins, London.

Clarke, K.R., Warwick, R.M., 2001. Change in Marine Communities: An Approach toStatistical Analysis and Interpretation. PRIMER-E, Plymouth.

Colautti, R.I., Ricciardi, A., Grigorovich, I.A., MacIsaac, H.J., 2004. Is invasion successexplained by the enemy release hypothesis? Ecol. Lett. 7, 721e733.

Cornelissen, J.H.C., Perez-Harguindeguy, N., Diaz, S., Grime, P.J., Marzano, B.,Cabido, M., Vendramini, F., Cerabolini, B., 1999. Leaf structure and defencecontrol litter decomposition rate across species and life forms in regional florason two continents. New Phytol. 143, 191e200.

Cox, J.M., 1987. Pseudococcidae (Insect:Hemiptera). Fauna of New Zealand, vol. 11.DSIR Publishing Centre, Wellington.

Crooks, J.A., 2002. Characterizing ecosystem-level consequences of biologicalinvasions: the role of ecosystem engineers. Oikos 97, 153e166.

Crop and Food Research, 2010. Plant-SyNZ Database. http://www.crop.cri.nz/home/plant-synz/database/databasehome.php (accessed 27.03.12).

Di Giulio, M., Edwards, P.J., Meister, E., 2001. Enhancing insect diversity in agri-cultural grasslands: the roles of management and landscape structure. J. App.Ecol. 38, 310e319.

Douglas, M.M., O’Connor, R.A., 2003. Effects of the exotic macrophyte, para grass(Urochloa mutica), on benthic and epiphytic macroinvertebrates of a tropicalfloodplain. Freshwater Biol. 48, 962e971.

Ernst, C.M., Cappuccino, N., 2005. The effect of an invasive alien vine, Vincetoxicumrossicum (Asclepiadaceae), on arthropod populations in Ontario old fields. Biol.Invasions 7, 417e425.

Finke, D.L., Denno, R.F., 2002. Intraguild predation diminished in complex-structured vegetation: implications for prey suppression. Ecology 83, 643e652.

Forster, R., Forster, L., 1999. Spiders of New Zealand and Their World-wide Kin.University of Otago Press, Dunedin.

Gerber, E., Krebs, C., Murrell, C., Moretti, M., Rocklin, R., Schaffner, U., 2008. Exoticinvasive knotweeds (Fallopia spp.) negatively affect native plant and inverte-brate assemblages in European riparian habitats. Biol. Conserv. 141, 646e654.

Gibson, C.W.D., Brown, V.K., Losito, L., McGavin, G.C., 1992. The response of inver-tebrate assemblies to grazing. Ecography 15, 166e176.

Gratton, C., Denno, R.F., 2005. Restoration of arthropod assemblages in a Spartinasalt marsh following removal of the invasive plant Phragmites australis. Restor.Ecol. 13, 358e372.

Hanski, I., 1998. Metapopulation dynamics. Nature 396, 41e49.Harris, R.J., Toft, R.J., Dugdale, J.S., Williams, P.A., Rees, J.S., 2004. Insect assemblages

in a native (kanuka - Kunzea ericoides) and an invasive (gorse - Ulex europaeus)shrubland. New Zeal. J. Ecol. 28, 35e47.

Hatley, C.L., MacMahon, J.A., 1980. Spider community organization, seasonal vari-ation and the role of vegetation architecture. Env. Entomol. 9, 632e639.

Hornung, J.P., Foote, A.L., 2006. Aquatic invertebrate responses to fish presence andvegetation complexity in western boreal wetlands, with implications forwaterbird productivity. Wetlands 26, 1e12.

Houston, W.A., Duivenvoorden, L.J., 2002. Replacement of littoral native vegetationwith the ponded pasture grass Hymenachne amplexicaulis: effects on plant,macroinvertebrate and fish biodiversity of backwaters in the Fitzroy River,Central Queensland, Australia. Mar. Freshwater Res. 53, 1235e1244.

Johnson, P.N., Brooke, P.A., 1998. Wetland Plants in New Zealand. Manaaki WhenuaPress, Lincoln.

Keane, R.M., Crawley, M.J., 2002. Exotic plant invasions and the enemy releasehypothesis. Trends Ecol. Evol. 17, 164e170.

Kolar, C.S., Lodge, D.M., 2001. Progress in invasion biology: predicting invaders.Trends Ecol. Evol. 16, 199e204.

Kuschel, G., 1990. Beetles in a Suburban Environment: A New Zealand Case Study:The Identity and Status of Coleoptera in the Natural and Modified Habitats ofLynfield, Auckland (1974e1989). DSIR, Auckland.

Langellotto, G.A., Denno, R.F., 2004. Responses of invertebrate natural enemies tocomplex-structured habitats: a meta-analytical synthesis. Oecologia 139, 1e10.

Larivière, M.-C., 2005. Checklist of New Zealand Hemiptera (Excluding Sternor-rhyncha). www.landcareresearch.co.nz/research/biosystematics/invertebrates/hemwsite/nzhemlist/ (accessed 20.01.08).

Larivière, M.-C., Larochelle, A., 2004. Heteroptera (Insecta: Hemiptera): Catalogue.Fauna of New Zealand, vol. 50. Manaaki Whenua Press, Lincoln.

Lawton, J.H., 1983. Plant architecture and the diversity of phytophagous insects.Annu. Rev. Entomol. 28, 23e39.

Lindsay, E.A., French, K., 2006. The impact of the weed Chrysanthemoides mon-ilifera ssp. rotundata on coastal leaf litter invertebrates. Biol. Invasions 8,177e192.

Maddox, D.M., Andres, L.A., Hennessey, R.D., Blackburn, R.D., Spencer, N.R., 1971.Insects to control alligator weed, an invader of aquatic ecosystems in the UnitedStates. Bioscience 21, 985e991.

Maron, J.L., Vilà, M., 2001. When do herbivores affect plant invasion? Evidence forthe natural enemies and biotic resistance hypotheses. Oikos 95, 361e373.

McEvoy, P., 2002. Insect-plant interactions on a planet of weeds. Entomol. Exp.Appl. 104, 165e179.

Moran, V.C., Southwood, T.R.E., 1982. The guild composition of arthropodcommunities in trees. J. Anim. Ecol. 51, 289e306.

Niell, R.S., Brussard, P.F., Murphy, D.D., 2007. Butterfly community composition andoak woodland vegetation response to rural residential development. LandscapeUrban Plan. 81, 235e245.

Niemela, J., Koivula, M., Kotze, D.J., 2007. The effects of forestry on carabid beetles(Coleoptera: Carabidae) in boreal forests. J. Insect Conserv. 11, 5e18.

Nitterus, K., Astrom, M., Gunnarsson, B., 2007. Commercial harvest of loggingresidue in clear-cuts affects the diversity and community composition ofground beetles (Coleoptera: Carabidae). Scand. J. For. Res. 22, 231e240.

Novotný, V., 1995. Relationships between life histories of leafhoppers (Auchenor-rhyncha - Hemiptera) and their host plants (Juncaceae, Cyperaceae, Poaceae).Oikos 73, 33e44.

Pawson, S.M., Brockerhoff, E.G., Meenken, E.D., Didham, R.K., 2008. Non-nativeplantation forests as alternative habitat for native forest beetles in a heavilymodified landscape. Biodivers. Conserv. 17, 1127e1148.

Pearson, D.E., 2009. Invasive plant architecture alters trophic interactions bychanging predator abundance and behavior. Oecologia 159, 549e558.

Pétillon, J., Ysnel, F., Canard, A., Lefeuvre, J.-C., 2005. Impact of an invasive plant(Elymus athericus) on the conservation value of tidal salt marshes in westernFrance and implications for management: responses of spider populations. Biol.Conserv. 126, 103e117.

Rowland, A.P., Roberts, J.D., 1994. Lignin and cellulose fractions in decompositionstudies using acid detergent fibre methods. Comm. Soil Sci. Plant Anal. 25,269e277.

Rypstra, A.L., Carter, P.E., Balfour, R.A., Marshall, S.D., 1999. Architectural features ofagricultural habitats and their impact on the spider inhabitants. J. Arachnol. 27,371e377.

Schooler, S.S., McEvoy, P.B., Hammond, P., Coombs, E.M., 2009. Negative per capitaeffects of two invasive plants, Lythrum salicaria and Phalaris arundinacea, on themoth diversity of wetland communities. Bull. Entomol. Res. 99, 229e243.

Taki, H., Kevan, P.G., 2007. Does habitat loss affect the communities of plants andinsects equally in plant-pollinator interactions? Preliminary findings. Biodivers.Conserv. 16, 3147e3161.

Thompson, R.M., Townsend, C.R., 2000. New Zealand’s stream invertebratecommunities: an international perspective. In: Collier, K.J., Winterbourn, M.J.(Eds.), New Zealand Stream Invertebrates: Ecology and Implications forManagement. New Zealand Limnological Society, Christchurch, pp. 53e74.

Topp, W., Kappes, H., Rogers, F., 2008. Response of ground-dwelling beetle (Cole-optera) assemblages to giant knotweed (Reynoutria spp.) invasion. Biol. Inv. 10,381e390.

Vandergast, A.G., Gillespie, R.G., 2004. Effects of natural forest Fragmentation ona Hawaiian spider community. Community Ecosyst. Ecol. 33, 1296e1305.

Watt, J.C., 1982. New Zealand beetles. New Zeal. Entomol. 7, 213e221.Wheeler, A.G., 2011. Four Holarctic plant bugs (Hemiptera: Miridae) associated with

Carex utriculata (Cyperaceae) in Montane grasslands of Valles Caldera, NewMexico: Southernmost U.S. Records and Notes on Seasonality. Proc. Entomol.Soc. Washington 113, 203e217.

Wissinger, S.A., McIntosh, A.R., Greig, H.S., 2006. Impacts of introduced brown andrainbow trout on benthic invertebrate communities in shallow New Zealandlakes. Freshwater Biol. 51, 2009e2028.

Woodcock, B.A., Potts, S.G., Westbury, D.B., Ramsay, A.J., Lambert, M., Harris, S.J.,Brown, V.K., 2007. The importance of sward architectural complexity in struc-turing predatory and phytophagous invertebrate assemblages. Ecol. Entomol.32, 302e311.

I.E. Bassett et al. / Acta Oecologica 41 (2012) 65e73 73