Biological Invasions

, Volume 15, Issue 11, pp 2429–2442 | Cite as

Imported crazy ant displaces imported fire ant, reduces and homogenizes grassland ant and arthropod assemblages

  • Edward G. LeBrun
  • John Abbott
  • Lawrence E. Gilbert
Original Paper

Abstract

A recently introduced, ecologically dominant, exotic ant species, Nylanderia fulva, is invading the Southeastern United States and Texas. We evaluate how this invader impacts diversity and abundance of co-occurring ants and other arthropods in two grasslands. N. fulva rapidly attains densities up to 2 orders of magnitude greater than the combined abundance of all other ants. Overall ant biomass increases in invaded habitat, indicating that N. fulva exploits resources not fully utilized by the local ant assemblage. At high density, as N. fulva spreads, it eliminates the current ecologically dominant invasive ant, red imported fire ants (Solenopsis invicta). Compared to imported fire ant dominated habitat, N. fulva invasion zones have lower non-ant arthropod species richness and abundance with impacts differing by trophic category. Further, N. fulva reduces abundance and species richness of the remainder of the ant assemblage and does so in a non-random manner: impacting species with small sized workers much less than species with larger workers. In these and other ant assemblages with a large exotic component, the exotics tend to be small bodied species. As a result, N. fulva almost completely eliminates regionally distributed species, but leaves globally distributed species largely unaffected, thereby systematically favoring introduced over native diversity. S. invicta impacts wildlife and arthropod assemblage structure and is nearly ubiquitous in non-forested habitats of the Southeastern United States and Texas. Its displacement by N. fulva has critical implications for the natural systems of this region.

Keywords

Biotic homogenization Tramp ants Nylanderia sp. nr. pubens Invasive ants Tawny crazy ant Hairy crazy ant Rasberry crazy ant Paratrechina fulva 

Introduction

Biological invasions drive diversity loss in the current extinction crisis (Sala et al. 2000), primarily by causing local, although typically not global, extinction of native species (Powell et al. 2011). Secondarily, they homogenize the biota by replacing regionally distinct species assemblages with assemblages comprised of more cosmopolitan taxa (McKinney and Lockwood 1999), thereby making biogeographically distant assemblages more similar.

The Southeastern United States of America (USA) has suffered a series of invasions by ecologically dominant invasive ant species native to central South America. Argentine ants (Linepithema humile Mayr), colonized the USA through New Orleans prior to 1891, and, within 50 years became widespread throughout the Southeastern USA. Argentine ant invasions dramatically reduce diversity of native ant assemblages (Holway et al. 2002). Around 1918, black imported fire ants (Solenopsis richteri Forel), colonized the USA through Mobile, Alabama (Wilson 1951). Red imported fire ants (Solenopsis invicta Buren), also introduced into Mobile (Wilson 1958), followed S. richteri in the 1930s. S. invicta spread and rapidly displaced S. richteri throughout most of the Southeast USA (Tschinkel 2006). S. invicta also apparently displaced Argentine ants (Buren et al. 1974) currently restricted to small scattered populations (Buczkowski et al. 2004). Ecological impacts of the spread of S. invicta were documented in central Texas where it impoverished ant, and non-ant arthropod faunas (Porter and Savignano 1990; Morris and Steigman 1993; Camilo and Phillips 1990), as well as negatively impacting many types of ground nesting birds and reptiles (Allen et al. 2004). In a low disturbance, woodland environment, ant and non-ant arthropod assemblage diversity largely recovered within 12 years of invasion by S. invicta (Morrison 2002). In contrast, in a xeric grassland, evidence of S. invicta impacting ant assemblage structure has persisted in some but not all habitats (LeBrun et al. 2012).

The region is now entering the next chapter in this history of serial invasion by ecologically dominant ants. An exotic ant species, Nylanderia (formerly Paratrechina) fulva (Mayr), colonized the Houston, Texas area around 2002 and, at the same time, dense infestations began to be reported in Florida. Initial attempts to identify this ant did not succeed (Meyers and Gold 2008), which led to it being referred to variously as N. sp. nr. pubens, N. pubens, and Rasberry crazy ant (Cook et al. 2012; Meyers and Gold 2008; Valles et al. 2012). A recent combined genetic and morphological analysis demonstrates that all populations of this ant in the USA are N. fulva, with the potential exception of South Florida that was not tested (Gotzek et al. 2012). By 2012, populations of these ants had established at sites in 21 counties in Texas (Center for Urban and Structural Entomology 2012). Dense, scattered populations of this ant are also present throughout Florida (Valles et al. 2012), in southern Mississippi (MacGown and Layton 2010), and in southern Louisiana (Hooper-Bui et al. 2010).

Nylanderia fulva evolved in central South America where it is a conspicuous, ecologically dominant member of the ant fauna (LeBrun et al. 2007; Wilder et al. 2011; Feener et al. 2008). Introduced N. fulva populations in Colombia have been responsible for widespread ecological and agricultural damage (Zenner de Polania 1990). Similar to most other globally widespread, ecologically destructive invasive ant species (Holway et al. 2002), populations of N. fulva in Texas are supercolonial with no aggression evident between workers from distant nests (Horn 2009). Nests, commonly containing 0–5 queens, males, workers, and brood (LeBrun pers. obs., Horn 2009), occur in soil cavities or in objects on the surface, such as pots, predisposing them to human transport. Its diet appears to be the same as documented in introduced populations in Colombia, consuming insects and the sugary exudates of plants and homopteran insects (Zenner de Polania and Bolaños 1985). No flights of winged reproductives have been reported for this species. Thus, new colonies are likely founded exclusively through colony fission. As a result, isolated populations of N. fulva spread slowly averaging about 20–30 m/month of outward spread (Meyers 2008). The outcome of resource competition with S. invicta appears to depend upon relative density with N. fulva monopolizing the majority of resources in laboratory conditions of equal abundance (Horn 2009). Population abundances of this ant inside invaded areas are tremendous (see “Results”). However, no published information exists on ecological impacts of these dense populations.

In Texas, N. fulva has invaded both urban and natural environments. Natural environments invaded to date include both highly modified and native pastureland habitats as well as highly disturbed wood lots and intact floodplain forest (E. G. LeBrun pers. obs.). The demonstrated capacity of N. fulva to invade wild lands demands an examination of its ecological impacts in natural environments. As a first step in this process, we contrast impacts of N. fulva on co-occurring ant species at two grassland sites on the Texas Gulf Coast.

Employing a natural experiment design, we use pitfall traps to sample areas invaded by N. fulva and adjacent, uninvaded areas to evaluate whether N. fulva impacts diversity and abundance of co-occurring ants and non-ant arthropods. Further, we examine whether N. fulva alters the body size distribution of the resident ant assemblage (McGlynn 1999), and whether, as a consequence of body size biased impacts, it exerts a homogenizing influence by differentially impacting regionally versus globally distributed species.

Methods

Study sites

We sampled two N. fulva invasion fronts at sites separated by 90 km within the Northern Gulf Coastal Prairie Ecoregion (Griffith et al. 2004): Iowa Colony: 29.44°N, 95.44°W, and El Campo: 29.23°N, 96.31°W. The Iowa Colony site supported a higher density N. fulva population than did El Campo (see “Results”). These sites will be referred to respectively as the “high density” (HD) and the “moderate density” (MD) sites. Both sites are disturbed coastal prairie containing a mixture of native and introduced plant species. The HD site has a history of intensive grazing and off road vehicle trails throughout. During periods of heavy rain, this site is prone to shallow flooding. Dominant plant species at the HD site are bushy bluestem (Andropogon glomeratus Walt.), dewberry (Rubus trivialis Michx.), McCartney rose (Rosa multiflora Thunb.), and Sesbania (Sesbania drummondii Rydb.). The MD site is an ungrazed fragment of disturbed prairie vegetation in a matrix of row crop agriculture. Vegetation at the MD site is dominated by little bluestem [Schizachyrium scoparium (Michx.) Nash], Johnson grass [Sorghum halepense (L.) Pers.], dewberry (R. trivialis Michx.), and goldenrod (Solidago cf. altissima). Red imported fire ants (S. invicta) invaded this region around 1967 (Hung and Vinson 1978). Given this 40 year interval, regionally these systems are either at equilibrium or in a process of recovering diversity lost during the initial S. invicta invasion.

Sampling

Sampling was conducted during October 2009, and August 2010. Ant and arthropod assemblages were quantified using grids of pitfall traps. In this natural experiment design, accurate inference from comparisons of invaded and adjacent uninvaded habitat assumes these areas possessed similar arthropod assemblages prior to invasion. For this reason, within both sites, all areas sampled were matched by: tree cover (trees absent), herb and shrub species composition, and known disturbance history. To further protect against spurious conclusions, we carried this work out at two independent sites. Finally, dense imported fire ant populations are a universal characteristic of open, grassland habitats in the Northern Gulf Coastal Prairie ecoregion (E. G. LeBrun pers. obs.), making any random deviation from this norm very unlikely.

Pitfall traps, 250 ml plastic cups with snap cap lids, and 7 cm diameter openings, were installed, charged with a solution of water, odorless laundry soap, and 25 % propylene glycol, and then left closed for 3 days to allow soil disturbance effects to dissipate. Traps were opened for 5 days. Pitfall trapping stations (HDS: 71 stations, MDS: 54 stations) were separated by 30 m, and contained 2 traps each spaced 2 m apart. Contents of both traps were combined for data analysis.

Sample processing

All ants were identified to species and counted. Arthropods were sorted to morphospecies, excluding all morphospecies of Diptera, Lepidoptera, Hymenoptera, and Neuroptera that had wings. These highly mobile groups, likely immigrated to the area of the trap and their numbers do not reflect local conditions. All Collembola were counted but not sorted to morphospecies. Morphospecies, with the exception of spiders apart from Lycosidae and Salticidae, were identified to family level or lower, and, based upon this ID, assigned to a trophic category: herbivore, predator, or detritivore.

The head widths and body masses of ant species were measured. Head width, taken with a dissecting microscope and optical micrometer, was the average of 3 measures from the largest available workers. To assess average worker mass, 10–20 workers chosen randomly from pitfall traps were placed in a drying oven for 48 h and then weighed in batches of 5 on a Mettler-Toledo UMX2 Ultra-microbalance™.

Data analysis

The N. fulva population edge was defined as the line bounding all points where N. fulva was present at that station and at least one other adjacent station. For some analyses, habitat was dichotomized into areas at least 100 m behind the edge of the N. fulva population, here forward “Nf-abundant areas”, and uninvaded areas near and in front of the N. fulva population edge, here forward “Nf-scarce areas”. At both sites, some N. fulva workers were present in isolated stations within the Nf-scarce areas. The distance between a sampling station and the N. fulva population edge was measured on Google Earth (http://www.google.com/earth/index.html). These distances provide a proxy measure for the relative amount of time a station has been occupied by N. fulva.

At the MD site, area uninvaded by N. fulva with habitat matching the invaded area was limited. At this site, to provide sufficient replication for within site comparisons, Nf-scarce stations included all stations in front of the N. fulva population edge plus stations <10 m behind the edge. In this rapidly expanding population, stations 10 m behind the edge had very limited exposure to N. fulva prior to sampling, and N. fulva was present at very low abundance (6.7 ± 9.2 workers/trap) (mean ± SD). No between site statistical comparisons are made, so the lack of a precise between site match in the definition of Nf-scarce stations does not impact the results.

Unless specified otherwise, all analyses were performed using JMP® 9.0 (SAS Institute 2011). Non-parametric, Wilcoxon Rank Sums tests were preferred for analysis of ant abundance data which did not meet assumptions of normality. Contingency tables were used to analyze presence–absence data characterizing ant prevalence.

We compare species density using Ecosim (Gotelli and Entsminger 2004). Species density is the number of species encountered across the array of traps, and provides the best available measure of species richness for each habitat. Estimating species richness requires extrapolating from species encountered to total species present, including those not observed, and relies on prevalence of rare species to estimate the number of unobserved species (Colwell 2011). In comparisons of recently invaded habitat and adjacent uninvaded habitat, species abundances are in flux, making this extrapolation inherently unreliable. To calculate species density, the average and standard deviation in species observed per sample accumulated was calculated by repeatedly aggregating randomly selected stations sampled with replacement (Gotelli and Colwell 2001). To test whether accumulation curves from areas differ, a Z score was calculated using the mean richness for each curve at the maximum observed sample size for the less well sampled area, and the standard deviation from the more intensively sampled area for that level of replication (Zarr 1999). N. fulva were removed from the data set prior to analysis creating some empty traps. To preserve these traps in the data set, a place holder species designated as present in all traps was inserted (LeBrun et al. 2012). Prior to presentation, all means have been back transformed.

To quantify variation in abundance, we define “prevalence” as the fraction of stations where a species was present in the pitfall trap. The “incidence abundance” of a species refers to the number of workers per pitfall trap considering only stations where that species occurred. The product of prevalence and incidence abundance yields the “abundance” of a species: the average workers per pitfall trap across all stations.

Logistic regression was used to evaluate the association between worker body size, and impact of N. fulva on species abundances. Responses of species to N. fulva were dichotomized. If average abundance of a species within a site declined by 25 % or more comparing Nf-scarce to Nf-abundant areas, it was characterized as negatively impacted by N. fulva. Response of all other species were designated as neutral to positive. All species occurring at both sites exhibited the same categorical response to N. fulva at both sites. All species observed at multiple stations at either site were included. Data was taken from the site where a species was most abundant.

To examine whether ant body size dependent impacts alter the biogeographic makeup of the assemblage, we categorized species by their ranges. Species restricted to the Nearctic zoogeographic province (North America north of Central Mexico) were categorized as “regional”. “Global” species are species that occur in more than one zoogeographic province. Finally, “invasive” species are exotic species that penetrate natural environments, reach high densities, and have documented negative ecological impacts on ecosystems they colonize. Species distribution information was obtained by summarizing occurrence information from the website AntWeb (AntWeb 2011), Creighton (1950), and any taxonomic literature published after 1950 that is listed on the Hymenoptera Online database (Hymenoptera Online (HOL) 2011).

Regression analyses of distance from the edge of the expanding N. fulva population and arthropod densities were performed separately for stations ahead of the population edge and stations behind the edge. The processes at work in these two environments were distinct and no uniform functional relationship spanning this boundary was expected. For non-ant arthropod data, a few stations captured extreme numbers of individual morphospecies, distorting the overall analyses of trophic categories. To address this, we reduced the abundance of any morphospecies present in a trap at >5 times its median abundance across all traps in which it occurred to the maximum abundance recorded for this species below that threshold. This constraint limited the effect of extreme outliers while preserving the occurrence and relative abundances of morphospecies. It reduced counts for 2 morphospecies at a total of 3 stations. Due to extreme abundance, Collembola were not included in the detritivore trophic category but were analyzed separately from the other arthropods.

Results

Impacts on ants: high density and moderate density N. fulva sites

Between Fall 2010 and Fall 2011, N. fulva population edges expanded an average of 180 m at the high density site (HDS) and advanced an average of 175 m at the moderate density site (MDS). At both sites, moving from the population edge into the interior, N. fulva abundance increased until roughly 100 m behind the edge (Fig. 1). At stations 100 m or more behind the population edge (Nf-abundant stations), N. fulva alone constituted 4–18 times the biomass of the entire ant assemblage at stations in front of or on the populations edge (Nf-scarce stations) (HDS: Nf-abundant: 110.6 ± 74.5 mg/trap N. fulva, Nf-scarce: 6.1 ± 7.1 mg/trap all ants; MDS: Nf-abundant: 16.8 ± 13.7 mg/trap N. fulva, Nf-scarce: 4.9 ± 3.2 mg/trap all ants) (mean ± SD).
Fig. 1

Abundances of N.fulva and S. invicta on a log scale for stations distributed across expanding N. fulva population fronts at 2 sites. For points behind the edge, distance from the edge of the N. fulva population serves as a surrogate for time since N. fulva arrived at a station. Figure displays results of pitfall traps set in areas in front of the edge of the expanding N. fulva population to well behind the edge. HDS high density N. fulva site, MDS moderate density N. fulva site. Negative distances lie in areas occupied by N. fulva. Dashed lines indicate the edge of the N. fulva population

Remarkably, at both sites, as populations of N. fulva spread, populations of the current ecological dominant, red imported fire ants (S. invicta), declined or were eliminated. At the HD site, by 200 m behind the front, the complete displacement of S. invicta mirrored the increase in N. fulva abundance (Fig. 1 HDS). This displacement was stark comparing Nf-scarce to Nf-abundant stations [Nf-scarce: 25.5 (12–42.25), Nf-abundant: 0 (0–0) S. invicta workers/station] [median (interquartile range)]. At the MD site, abundances of S. invicta declined strongly as one moves into areas occupied by N. fulva, but S. invicta remained a component of the ant assemblage throughout [Nf-scarce: 25 (17–32), Nf-abundant: 4 (1–25) S. invicta workers/station] [median (interquartile range)] (Statistics: Table 1A) (Fig. 1 MDS).
Table 1

Results of analyses contrasting ant abundance between stations in Nf-scarce and Nf-abundant areas at high density (HD) and moderate density (MD) N. fulva sites

 

Group

Variable

Test

HD N. fulva site

MD N. fulva site

Δ

χ2

N

P

Δ

χ2

N

P

A

S. invicta

Abundance

Wilcoxon

41.7

53

0.0001

8.5

52

0.004

B

Regional species

Prevalence

Contingency table

18.0

53

0.0001

8.0

52

0.008

Global species

Prevalence

Contingency table

ND

0.006

53

1.0

ND

0.06

52

1.0

C

Regional species

Incidence abundance

Wilcoxon

4.6

17

0.03

ND

0.006

13

0.94

Global species

Incidence abundance

Wilcoxon

ND

0.10

46

0.75

+

10.5

29

0.001

Abundance: average number of workers per station; Prevalence: % of stations group is present; Incidence abundance: average workers per station for stations where group is present. Δ: direction of change comparing Nf scarce to Nf abundant areas. See figures for details on magnitudes of differences

At the HD site, Nf-abundant areas exhibited significantly lower ant species density than Nf-scarce areas (Z test: N = 22, Z = 3.21, P < 0.002) (Fig. 2). When S. invicta was removed, this difference persisted (Z test: N = 22, Z = 2.27, P < 0.02), demonstrating a wider loss of diversity. At the MD site, the same pattern of lower ant species density in areas behind the N. fulva invasion front held, however, the difference was not significant (Z test: N = 18, Z = 1.76, P = 0.07) (Fig. 2).
Fig. 2

Species density of ant assemblages excluding N.fulva. S. invicta is absent in Nf-abundant areas and abundant in Nf-scarce areas. Symbols are averages of 1,000 species accumulation curves sampled with replacement, with bars presenting standard deviations. Numbers adjacent to curves report total number of species recorded from the habitat. HDS high density N. fulva site, MDS moderate density N. fulva site

Interestingly, N. fulva impacted local ant assemblages in these disturbed grasslands in a non-random manner. At both sites, the abundances of ants other than S. invicta and N. fulva were significantly reduced by the expanding N. fulva population. These impacts were specific to the body sizes of species. Large bodied ant species were considerably more likely to be negatively impacted by N. fulva than were small bodied species (Logistic regression of impact versus: Head width largest worker: N = 16, β1 = 10.07, rL2 = 0.56, P < 0.0005; Dry mass of average worker: N = 16, β1 = 9.05, rL2 = 0.31, P < 0.01). In fact, abundance of all species larger than about 0.7 mm head width declined (Fig. 3).
Fig. 3

Logistic fit of head width of the largest workers for a species and whether abundance of that species did not change/increased (Neutralorpositive response) or decreased (Negative response) comparing Nf-scarce stations with Nf-abundant stations. Species present at both high and moderate density N. fulva sites always exhibited the same qualitative response to N. fulva. Symbols are the first letter of the genus and the first letter(s) of the species epithet for species in Online Resource 1: Table A. Overlapping points are jittered on Y axis. Dotted line demarcates headwidth of N. fulva workers

Globally distributed tramp species comprised the majority of small bodied ants in these assemblages (Online Resource 1: Table A). As a result of this size bias, after invasion, prevalence of regional species taken as a group declined precipitously, almost to elimination, at both the HD and MD sites, comparing Nf-scarce to Nf-abundant stations. In contrast, prevalence of global species did not differ at either site (Statistics: Table 1B). Average abundance of species across stations at which they were present (incidence abundance) was reduced for regional species at the HD site, and unchanged at the MD site. No difference in incidence abundance existed for global species at the HD site. However, global species were more abundant in the presence of N. fulva at the MD site (Statistics: Table 1C) (Fig. 4).
Fig. 4

Comparison of abundances of regionally distributed and globally distributed ant species in Nf-abundant and nearby Nf-scarce areas. RegionalSpecies include all species restricted to the Nearctic. Global Species include all species known to occur in more than one zoogeographic province. S. invicta and N.fulva workers were excluded from this analysis. Points in prevalence plots display the percent of stations where that group occurred with bars displaying 95 % confidence intervals for that value. Points for incidence abundance plots display median abundance for that group across stations where species were present and bars present interquartile range. High density site—plots in top row, moderate density site—plots in bottom row, regional ant species—plots in left column, global ant species—plots in right column

Impacts on non-ant arthropods: high density N. fulva site

Morphospecies richness of non-ant arthropods was reduced by N. fulva invasion. Nf-abundant areas exhibited lower non-ant arthropod morphospecies density than Nf-scarce areas (Z test: N = 25, Z = 2.03, P < 0.043). This difference depended upon trophic category examined, with herbivore diversity strongly reduced by N. fulva (Z test: N = 24, Z = 3.43, P < 0.0006) but predators and detritivore diversity unaffected (Z test: predators: N = 25, Z = 0.86, P = 0.39, detritivores: N = 22, Z = 1.27, P = 0.20) (Fig. 5).
Fig. 5

Impacts of N. fulva upon non-ant arthropod species density at the high density N. fulva site. S. invicta is absent in Nf-abundant areas and abundant in Nf-scarce areas. Points are averages of 1,000 species accumulation curves sampled with replacement. Bars present standard deviations. Numbers report the number of morphospecies observed

Abundances of non-ant arthropods also differed, but with all trophic categories responding. Summed abundance of non-ant, arthropods declined steadily with increasing distance behind the N. fulva invasion front, a surrogate measure for time since invasion. In contrast, in uninvaded habitat, there was no significant relationship. When arthropods are broken up by trophic category, abundances of all categories declined in areas invaded by N. fulva. Among uninvaded stations, no significant relationships existed (Statistics: Table 2). For all categories relationships were best fit by linear functions, indicating that we may not have characterized the full decline (Fig. 6). The responses of specific taxonomic groups are presented in Online Resource 1: Table B.
Table 2

Results of linear regression analyses examining relationships between abundances of non-ant arthropods and the distance from the edge of the N. fulva population at the high density site

Abundances

N. fulva +

N. fulva

r2

N

P

Regression line

r2

N

P

Regression line

All arthropods

0.37

45

0.0001

27.2 + 0.05 × x

0.10

21

0.15

NS

Herbivores

0.21

45

0.001

11.4 + 0.02 × x

0.05

21

0.31

NS

Predators

0.25

45

0.0004

7.5 + 0.009 × x

0.04

21

0.41

NS

Detritivores

0.11

45

0.03

8.3 + 0.02 x x

0.07

21

0.25

NS

Functions for regression lines provided for statistically significant relationships. N. fulva + all stations behind the N. fulva population edge, N. fulva − all stations at or ahead of the N. fulva population edge

Fig. 6

Abundances of non-ant arthropods caught in pitfall traps spanning the edge of the N. fulva population at the high density N. fulva site. For points behind the edge, distance from the edge of the N. fulva population serves as a surrogate for time since N. fulva arrived at a station. Dashed line demarcates the edge of the N. fulva population at time of sampling. Negative distances lie in areas occupied by N. fulva. Functions were fit independently to data in front of and behind the N. fulva invasion front

Discussion

Within a year of colonizing, N. fulva attains numerical abundances up to 2 orders of magnitude greater than other ants in the system combined, including red imported fire ants (S. invicta). Because pitfall traps quantify relative abundance reliably, but tend to oversample fast-moving species, like N. fulva, relative to others (Andersen 1991), how accurately this measures the actual increase in biomass must be interpreted cautiously. Studies of the abundance of some other invasive ant species indicate that these species extreme abundance in the introduced range partly result from more effective exploitation of honeydew and extrafloral nectar relative to the native assemblage (Helms et al. 2011; Tillberg et al. 2007; Wilder et al. 2011). Identifying the factors that underlie the increase in N. fulva biomass requires further study. Understanding how this high biomass is maintained may help to understand the causes of the substantial between site variability in N. fulva abundance, and concomitant between site variability in ecological impacts, documented in this study.

As N. fulva invades, S. invicta suffers strongly. At the HD site, by 200 m behind the front S. invicta is entirely eliminated. Extrapolating from the expansion rate of 180 m/year observed at this site in 2010–2011, N. fulva displaces all fire ant colonies slightly more than a year after spreading into an area. However, S. invicta declines in abundance but persists throughout the MD site. Between fall of 2010 and 2011, the invasion front at the MD site advanced between 150 and 200 m indicating that stations most distant from the N. fulva population edge had been colonized by N. fulva 3–4 years previously. Thus some S. invicta colonies can persist in this environment. Whether S. invicta can establish new colonies and thus co-exist long-term is unknown.

Nylanderia fulva reduces the species richness and abundance of these ant assemblages (Figs. 2, 4). The original invasion by S. invicta reduced the richness and abundance of native ant assemblages regionally (Porter and Savignano 1990; Morris and Steigman 1993; Camilo and Phillips 1990). Recovery of these assemblages has been documented in some systems (Morrison 2002), while in others the evidence of its impact has persisted (LeBrun et al. 2012). Thus the impacts exerted by N. fulva occur against a historical backdrop of repeated assemblage disruption, impoverishment, and partial recovery. The patterns of species responses provide insight into the longer-term trajectory of these assemblages subject to serial invasion.

Interestingly, N. fulva impacts local ant assemblages in these disturbed grasslands in a non-random manner. Comparing HD N. fulva environments with nearby uninvaded habitat, smaller ant species show a mixed response but the majority of species either do not change or increase in abundance. In contrast, large species uniformly decline in abundance. In these ant assemblages, small-bodied ants are disproportionately represented by globally distributed tramp species. The net outcome is a reduction in both biological and functional diversity, accompanied by an intense biotic homogenization of the local ant assemblage (McKinney and Lockwood 1999). By limiting the size distribution of the ant assemblage, N. fulva reduces the functional diversity of ants, the most abundant class of consumers in the leaf litter (Bihn et al. 2010), potentially altering ecosystem function (Kaspari 2005). Further, the preferential removal of a suite of regional species while leaving globally distributed species intact, represents a remarkable process biogeographically homogenizing the local assemblage.

In recently invaded habitat, dense N.fulva populations saturate the environment, and their nests commonly occur in abandoned nest mounds of S. invicta (E. LeBrun pers. obs.). Further, the larvae and brood of other ants was reported as a food source of N. fulva in Colombia (Zenner de Polania and Bolaños 1985). Potentially, the decline of the ant assemblage generally may arise as a consequence of nest raiding by N. fulva either for food or space. Under this proposed mechanism, differential species loss arises because nests of small bodied species, by virtue of having small tunnels, are more resistant to raiding. Suggestively, average head width of the largest workers, a trait intimately related to tunnel diameter, explains more variation in the logistic relationship with impact than does body mass of an average worker. Also consistent with nest raiding, the size threshold above which all ant species suffer negative impacts from N. fulva invasion is slightly larger than the average head width of an N. fulva worker. Testing these and other hypotheses for what drives size biased displacement requires further study.

Is replacement of large-bodied, regional species with small-bodied, global species a general property of ant invasions? Non-native ant species are documented to be smaller than their regionally restricted congeners (McGlynn 1999; Passera 1994; Ness et al. 2004; but see King and Porter 2007). Further, in a study comparing arriving propagules with successfully established ant species, small body size was the most important variable explaining successful establishment (Lester 2005). Thus, in assemblages with many introduced species, a large fraction of small-bodied ants are likely to be global species. Several studies report small bodied ant species resisting displacement by opportunistically nesting, invasive ant species [Anoplolepis gracilipes (Hoffmann and Saul 2010), Pheidole megacephala (Hoffmann et al. 1999), L. humile (Ward 1987; Human and Gordon 1997; Holway 1998)]. These studies attribute resistance of small species to either their foraging niche [hypogeic foragers resist displacement by L. humile (Ward 1987)], or to their functional role [small opportunists and cryptic species resist displacement by P. megacephala (Hoffmann et al. 1999)]. Our results suggest an alternative hypothesis: resistance of small bodied ant species to displacement may be in part driven by resistance to nest raiding, and may represent a general property of invasions by opportunistically nesting ant species.

The suppression by N.fulva of herbivorous arthropod species richness as well as the abundances of all arthropod trophic categories has potential to impact ecosystem function. The functional consequences of the reduction in predator abundance are at least in part offset by the impact of N. fulva on lower trophic levels. The functional consequence of reducing arthropod detritivore abundance, the least impacted group, may be to slow or speed rates of decomposition and nutrient cycling (Gessner et al. 2010). Thus, it is unlikely that N. fulva will strongly alter decomposition rates. However, herbivores of the grass and forb matrix are the most strongly impacted group, declining both in abundance and species richness. In addition to likely direct predation of eggs and early instars, the strength of this reduction may arise because these species need to be stationary to feed, difficult for an insect within an ant population of this density. These changes should translate into reduced rates or patterns of herbivory, potentially altering relative abundances of plant species over time. In grassland systems, insect herbivores tend to enhance plant diversity by preventing dominant species from forming monocultures (Carson and Root 2000). Thus, reduction in herbivore abundance and diversity may reduce long-term co-existence among plant species. Similar impacts have been observed in tropical forests invaded by the ant Wasmannia auropunctata (Roger) which reduced rates of herbivory by leaf-chewing insects (Dunham and Mikheyev 2010).

In habitat uninvaded by N.fulva, we expected no relationship between distance from the edge and arthropod abundance. There was no significant relationship. However, arthropod abundance tended to be highest at the population edge. This trend may result from some unanticipated underlying gradient that N. fulva reverses, or it may arise from insects emigrating from habitat densely occupied by N. fulva and settling in low density areas.

Solenopsis invicta is nearly universal in non-forested habitats in the Southeastern USA west to Central Texas. If N.fulva replaces S. invicta as it spreads, differences between these ants will have critical consequences for natural and agricultural systems regionally. For example, S. invicta is scarce in closed canopy woodlands. However, in addition to thoroughly occupying grasslands, N. fulva reaches high densities in forested environments (E. LeBrun pers. obs.). Thus forest invertebrate and vertebrate fauna that have been less impacted by S. invicta may be under greater threat from this invader. Further, N. fulva exhibits a dramatic seasonal cycle in local abundance. Populations are extraordinarily dense in the summer and fall, but much less so in spring and early summer (E. LeBrun pers. obs.). As a consequence, in contrast with S. invicta (Allen et al. 2004), species that nest in spring may be less directly impacted by these ants. However, because N. fulva reduces general arthropod abundance, food for insectivorous wildlife will be reduced and these animals may suffer greater net impacts as a result.

The long-term environmental consequences of this invasion remain unclear. N.fulva has demonstrated that it can penetrate natural ecosystems. Although undisturbed systems are very rare regionally, N. fulva has invaded intact bottomland forest habitats (E. G. LeBrun pers. obs.) and its success in pasturelands indicates that it will invade intact prairie reserves once it encounters them. The role of disturbance in determining the densities this invader achieves or the impacts it has remains to be investigated. There has been no attempt to eradicate this invader or limit its expansion by regulating transport of goods likely to contain colonies. Thus, in all likelihood, N. fulva will become widespread in urban and non-urban environments in the Gulf Coast region. Lacking data on its aridity and freeze tolerance limits, we cannot predict its potential future range. Equally it is unclear for how long these extremely dense populations will persist, or whether the negative effects we document are transitory (Cooling et al. 2011). However, short-term consequences of this invasion for arthropod communities are clear: reduction or displacement of S. invicta, further degradation of native ant diversity, and a generalized reduction in arthropod abundance. Currently, there is a pressing need for more information about impacts of this invader on other aspects of the habitats it invades, an increased understanding of its physiological tolerances, and its basic ecology both in the USA and in its native range.

Notes

Acknowledgments

We thank Perry Herman and Land Tejas Companies, Ltd for allowing access to their properties. Residents of Iowa Colony provided information on the history of the area. Jerod Romine, Drew Townsend, and Mike Marischen assisted in conducting field work. Alejandro Calixto and Danny Macdonald provided information on the location of established populations. Funding was provided by the Helen C. Kleberg and Robert J. Kleberg Foundation, and the Lee and Ramona Bass Foundation.

Supplementary material

10530_2013_463_MOESM1_ESM.doc (104 kb)
Supplementary material 1 (DOC 104 kb)

References

  1. Allen CR, Epperson DM, Garmestani AS (2004) Red imported fire ant impacts on wildlife: a decade of research. Am Midl Nat 152(1):88–103CrossRefGoogle Scholar
  2. Andersen AN (1991) Sampling communities of ground-foraging ants: pitfall catches compared with quadrat counts in an Australian tropical savanna. Aust J of Ecol 16(3):273–280CrossRefGoogle Scholar
  3. AntWeb (2011) The California Academy of Sciences. www.antweb.org. Accessed 1 May 2012
  4. Bihn JH, Gebauer G, Brandl R (2010) Loss of functional diversity of ant assemblages in secondary tropical forests. Ecology 91(3):782–792PubMedCrossRefGoogle Scholar
  5. Buczkowski G, Vargo EL, Silverman J (2004) The diminutive supercolony: the Argentine ants of the southeastern United States. Mol Ecol 13(8):2235–2242PubMedCrossRefGoogle Scholar
  6. Buren WF, Allen GE, Whitcomb WH, Lennartz FE, Williams RN (1974) Zoogeography of imported fire ants. J N Y Entomol Soc 82(2):113–124Google Scholar
  7. Camilo GR, Phillips SA Jr (1990) Evolution of ant communities in response to invasion by the fire ant Solenopsis invicta. In: Meer RKV, Jaffe K, Cedeno A (eds) Applied myrmecology: a world perspective. Westview Press, Boulder, pp 190–198Google Scholar
  8. Carson WP, Root RB (2000) Herbivory and plant species coexistence: community regulation by an outbreaking phytophagous insect. Ecol Monogr 70(1):73–99CrossRefGoogle Scholar
  9. Center for Urban and Structural Entomology (2012) Rasberry crazy ant Nylanderia sp. near pubens. Department of Entomology, Texas A&M University. http://urbanentomology.tamu.edu/ants/rasberry.html. Accessed 07 Mar 2012
  10. Colwell RK (2011) EstimateS: statistical estimation of species richness and shared species from samples, 8.2.0 edn. University of Connecticut, StorrsGoogle Scholar
  11. Cook SC, Wynalda RA, Gold RE, Behmer ST (2012) Macronutrient regulation in the Rasberry crazy ant (Nylanderia sp nr. pubens). Insect Soc 59(1):93–100CrossRefGoogle Scholar
  12. Cooling M, Hartley S, Dalice AS, Lester PJ (2011) The widespread collapse of an invasive species: Argentine ants (Linepithema humile) in New Zealand. Biol Lett. doi:10.1098/rsbl.2011.1014 PubMedGoogle Scholar
  13. Creighton WS (1950) The ants of North America. Bull Mus Comp Zool Harv Univ 104:1–585Google Scholar
  14. Dunham AE, Mikheyev AS (2010) Influence of an invasive ant on grazing and detrital communities and nutrient fluxes in a tropical forest. Divers Distrib 16(1):33–42CrossRefGoogle Scholar
  15. Feener DH, Orr MR, Wackford K, Longo JM, Benson WW, Gilbert LE (2008) Geographic variation in resource dominance, discovery, and parasitoid-mediated competition in the Brazilian range of the red imported fire ant (Solenopsis invicta). Ecology 89(7):1824–1836 Google Scholar
  16. Gessner MO, Swan CM, Dang CK, McKie BG, Bardgett RD, Wall DH, Hattenschwiler S (2010) Diversity meets decomposition. Trends Ecol Evol 25(6):372–380PubMedCrossRefGoogle Scholar
  17. Gotelli NJ, Colwell RK (2001) Quantifying biodiversity: procedures and pitfalls in the measurement and comparison of species richness. Ecol Lett 4(4):379–391CrossRefGoogle Scholar
  18. Gotelli NJ, Entsminger GL (2004) EcoSim: null models software for ecology. Version, 7th edn. Acquired Intelligence and Kesey-Bear, JerichoGoogle Scholar
  19. Gotzek D, Brady SG, Kallal RJ, LaPolla JS (2012) The importance of using multiple approaches for identifying emerging invasive species: the case of the Rasberry crazy ant in the United States. PLoS One 7(9):e45314PubMedCrossRefGoogle Scholar
  20. Griffith GE, Bryce SA, Omernik JM, Comstock JA, Rogers AC, Harrison B, Hatch SL, Bezanson D (2004) Ecoregions of Texas (color poster with map, descriptive text, and photographs). US Geological Survey Reston, VirginiaGoogle Scholar
  21. Helms KR, Hayden CP, Vinson SB (2011) Plant-based food resources, trophic interactions among alien species, and the abundance of an invasive ant. Biol Invasions 13(1):67–79CrossRefGoogle Scholar
  22. Hoffmann BD, Saul WC (2010) Yellow crazy ant (Anoplolepis gracilipes) invasions within undisturbed mainland Australian habitats: no support for biotic resistance hypothesis. Biol Invasions 12(9):3093–3108CrossRefGoogle Scholar
  23. Hoffmann BD, Andersen AN, Hill GJE (1999) Impact of an introduced ant on native rain forest invertebrates: Pheidole megacephala in monsoonal Australia. Oecologia 120(4):595–604Google Scholar
  24. Holway DA (1998) Effect of Argentine ant invasions on ground-dwelling arthropods in northern California riparian woodlands. Oecologia 116(1–2):252–258CrossRefGoogle Scholar
  25. Holway DA, Lach L, Suarez AV, Tsutsui ND, Case TJ (2002) The causes and consequences of ant invasions. Annu Rev Ecol 33:181–233CrossRefGoogle Scholar
  26. Hooper-Bui LM, Strecker R, Chen X, Aguillard D, Miller A (2010) Super-colonies of crazy ants in Louisiana. In: Proceedings of the imported fire ant and invasive ant conference, pp 13–16Google Scholar
  27. Horn K (2009) Examining competitive interactions between Rasberry crazy ants (Paratrechina sp. nr. pubens) and red imported fire ants (Solenopsis invicta) using laboratory and field studies. Masters of Arts, Rice University, Houston, TX, USAGoogle Scholar
  28. Human KG, Gordon DM (1997) Effects of Argentine ants on invertebrate biodiversity in northern California. Conserv Biol 11(5):1242–1248CrossRefGoogle Scholar
  29. Hung ACF, Vinson SB (1978) Factors affecting the distribution of fire ants in Texas (Myrmicinae: Formicidae). Southwest Nat 23(2):205–214CrossRefGoogle Scholar
  30. Hymenoptera Online (HOL) (2011) Ohio State University. http://hol.osu.edu/. Accessed 01 May 2012
  31. Kaspari M (2005) Global energy gradients and size in colonial organisms: worker mass and worker number in ant colonies. Proc Natl Acad Sci USA 102(14):5079–5083PubMedCrossRefGoogle Scholar
  32. King JR, Porter SD (2007) Body size, colony size, abundance, and ecological impact of exotic ants in Florida’s upland ecosystems. Evol Ecol Res 9(5):757–774Google Scholar
  33. LeBrun EG, Tillberg CV, Suarez AV, Folgarait PJ, Smith CR, Holway DA (2007) An experimental study of competitive interactions between red imported fire ants and Argentine ants in a region of native-range sympatry. Ecology 88(1):63–75PubMedCrossRefGoogle Scholar
  34. LeBrun EG, Plowes RM, Gilbert LE (2012) Imported fire ants near the edge of their range: disturbance and moisture determine prevalence and impact of an invasive social insect. J Anim Ecol 81(4):884–895PubMedCrossRefGoogle Scholar
  35. Lester PJ (2005) Determinants for the successful establishment of exotic ants in New Zealand. Divers Distrib 11(4):279–288CrossRefGoogle Scholar
  36. MacGown J, Layton B (2010) The invasive Rasberry crazy ant, Nylanderia sp. nr. pubens (Hymenoptera: Formicidae), reported from Mississippi. Midsouth Entomol 3:44–47Google Scholar
  37. McGlynn TP (1999) The worldwide transfer of ants: geographical distribution and ecological invasions. J Biogeogr 26(3):535–548CrossRefGoogle Scholar
  38. McKinney ML, Lockwood JL (1999) Biotic homogenization: a few winners replacing many losers in the next mass extinction. Trends Ecol Evol 14(11):450–453PubMedCrossRefGoogle Scholar
  39. Meyers JM (2008) Identification, distribution and control of an invasive pest ant, Paratrechina sp. (Hymenoptera: Formicidae), in Texas. Doctor of Philosophy, Texas A&M University, College Station, TX, USAGoogle Scholar
  40. Meyers JM, Gold RE (2008) Identification of an exotic pest ant, Paratrechina sp.nr. pubens (Hymenoptera: Formicidae), in Texas. Sociobiology 52(3):589–604Google Scholar
  41. Morris JR, Steigman KL (1993) Effects of polygyne fire ant invasion on native ants of a blackland prairie in Texas. Southwest Nat 38(2):136–140CrossRefGoogle Scholar
  42. Morrison LW (2002) Long-term impacts of an arthropod-community invasion by the imported fire ant, Solenopsis invicta. Ecology 83(8):2337–2345CrossRefGoogle Scholar
  43. Ness JH, Bronstein JL, Andersen AN, Holland JN (2004) Ant body size predicts dispersal distance of ant-adapted seeds: implications of small-ant invasions. Ecology 85(5):1244–1250CrossRefGoogle Scholar
  44. Passera L (1994) Characteristics of tramp species. In: Williams PF (ed) Ants: biology, impact and control of introduced species. Westview Press, BoulderGoogle Scholar
  45. Porter SD, Savignano DA (1990) Invasion of polygyne fire ants decimates native ants and disrupts arthropod community. Ecology 71(6):2095–2106CrossRefGoogle Scholar
  46. Powell KI, Chase JM, Knight TM (2011) A synthesis of plant invasion effects on biodiversity across spatial scales. Am J Bot 98(3):539–548PubMedCrossRefGoogle Scholar
  47. Sala OE, Chapin FS, Armesto JJ, Berlow E, Bloomfield J, Dirzo R, Huber-Sanwald E, Huenneke LF, Jackson RB, Kinzig A, Leemans R, Lodge DM, Mooney HA, Oesterheld M, Poff NL, Sykes MT, Walker BH, Walker M, Wall DH (2000) Biodiversity—global biodiversity scenarios for the year 2100. Science 287(5459):1770–1774PubMedCrossRefGoogle Scholar
  48. SAS Institute (2011) JMP®, 9.0th edn. SAS Institute, CaryGoogle Scholar
  49. Tillberg CV, Holway DA, LeBrun EG, Suarez AV (2007) Trophic ecology of invasive Argentine ants in their native and introduced ranges. Proc Natl Acad Sci USA 104(52):20856–20861PubMedCrossRefGoogle Scholar
  50. Tschinkel WR (2006) The fire ants. The Belknap Press of Harvard University Press, CambridgeGoogle Scholar
  51. Valles SM, Oi DH, Yu F, Tan X, Buss EA (2012) Metatransciptomics and pyrosequencing facilitate discovery of pentential viral natural enemies of the invasive Caribbean crazy ant, Nylanderia pubens. PLOS One 7(2):1–9CrossRefGoogle Scholar
  52. Ward PS (1987) Distribution of the introduced Argentine Ant (Iridomyrmex humilis) in natural habitats of the lower Sacramento valley and its effects on the indigenous ant fauna. Hilgardia 55(2):1–16Google Scholar
  53. Wilder SM, Holway DA, Suarez AV, LeBrun EG, Eubanks MD (2011) Intercontinental differences in resource use reveal the importance of mutualisms for invasive species. Proc Natl Acad Sci USA 108(51):20639–20644PubMedCrossRefGoogle Scholar
  54. Wilson EO (1951) Variation and adaptation in the imported fire ant. Evolution 5:68–79CrossRefGoogle Scholar
  55. Wilson EO (1958) The fire ant. Sci Am 198(3):36–41CrossRefGoogle Scholar
  56. Zarr JH (1999) Biostatistical analysis, 4th edn. Simon and Schuster, NJGoogle Scholar
  57. Zenner de Polania I, Bolaños RN (1985) Habitos alimenticios y relaciones simbioticas de la “hormiga loca” Nylanderia fulva con otros artropodos. Rev Colomb Entomol 11(1):3–10Google Scholar
  58. Zenner de Polania I, Cedeno A (1990) Biological aspects of the “Hormiga Loca”, Paratrechina (Nylanderia) fulva (Mayr), in Colombia. In: Vander Meer RK, Jaffe K (eds) Applied myrmecology, a world perspective. Westview Press, Boulder, pp 290–297Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2013

Authors and Affiliations

  • Edward G. LeBrun
    • 1
  • John Abbott
    • 1
  • Lawrence E. Gilbert
    • 2
  1. 1.Brackenridge Field Laboratory, Section of Integrative BiologyUniversity of Texas at AustinAustinUSA
  2. 2.Section of Integrative BiologyThe University of Texas at AustinAustinUSA

Personalised recommendations