Invasive plant architecture alters trophic interactions by changing predator abundance and behavior

As primary producers, plants are known to influence higher trophic interactions by initiating food chains. However, as architects, plants may bypass consumers to directly affect predators with important but underappreciated trophic ramifications. Invasion of western North American grasslands by the perennial forb, spotted knapweed (Centaurea maculosa), has fundamentally altered the architecture of native grassland vegetation. Here, I use long-term monitoring, observational studies, and field experiments to document how changes in vegetation architecture have affected native web spider populations and predation rates. Native spiders that use vegetation as web substrates were collectively 38 times more abundant in C. maculosa-invaded grasslands than in uninvaded grasslands. This increase in spider abundance was accompanied by a large shift in web spider community structure, driven primarily by the strong response of Dictyna spiders to C. maculosa invasion. Dictyna densities were 46–74 times higher in C. maculosa-invaded than native grasslands, a pattern that persisted over 6 years of monitoring. C. maculosa also altered Dictyna web building behavior and foraging success. Dictyna webs on C. maculosa were 2.9–4.0 times larger and generated 2.0–2.3 times higher total prey captures than webs on Achillea millefolium, their primary native substrate. Dictyna webs on C. maculosa also captured 4.2 times more large prey items, which are crucial for reproduction. As a result, Dictyna were nearly twice as likely to reproduce on C. maculosa substrates compared to native substrates. The overall outcome of C. maculosa invasion and its transformative effects on vegetation architecture on Dictyna density and web building behavior were to increase Dictyna predation on invertebrate prey ≥89 fold. These results indicate that invasive plants that change the architecture of native vegetation can substantially impact native food webs via nontraditional plant → predator → consumer linkages.     

Guild mobility affects spider diversity: Links between foraging behavior and sensitivity to adjacent vegetation structure

The movement patterns of species may affect their susceptibility to modified habitat structure. It is likely that sedentary species perceive habitat features at smaller spatial extents compared to mobile species, but there is a lack of experimental research on the effects of fine-scale habitat characteristics on organisms of differing mobility. Spiders display two basic mobility levels based on foraging behavior: web-building species are restricted to specific sites whereas active hunters are mobile. We collected spiders inhabiting sagebrush shrubs with a structurally enhanced, unmodified, or removed understory, to examine (1) whether habitat structure in the immediate vicinity of shrubs affected cursorial and web spiders differently in terms of abundance and species richness and (2) which genera most contributed to changes in community composition. Shrubs without understory had reduced cursorial spider densities and species richness compared to shrubs with added and unmodified understories, whereas web spiders lacked significant responses to treatments. Community-level differences based on relative abundance of genera were detected in cursorial spiders but not in web spiders, despite a strong contribution of the web-building genus Theridion to community dissimilarities. Our results support the hypothesis that sedentary organisms may be sensitive to contiguous habitat at finer spatial scales than cursorial organisms, and highlight the risks associated with only collecting local habitat information when studying mobile species.     

Web-Building Flexibility Differs in Two Spatially Constrained Orb Spiders

Shelter and trap-building animals that compete for limited space and/or face costly relocations benefit from being flexible in their construction behavior. Orb spiders are good examples of this and their easily quantifiable two-dimensional webs allow us to analyze the behavioral adaptations and costs in terms of higher error levels or less precision resulting from building webs in sub-optimal conditions. Here I study behavioral flexibility in spatially constrained spiders by analyzing a wide range of web parameters including measures that indicate errors during web-building. I compare the geometry of laboratory webs of two orb spiders, Cyclosa caroli and Eustala illicita, built in differently shaped experimental frames and report two major findings. i) The two species differ in their ability to build webs in constrained spaces. ii) E. illicita adjusted a range of parameters including shape, area utilization and mesh height in response to spatial constraints, but kept other parameters constant, most notably the length of anchor threads and the shape of the auxiliary spiral. I furthermore found that constrained spiders did not make significantly more errors during web-building than when they had amble space available.     

The Effect of Wind Exposure on the Web Characteristics of a Tetragnathid Orb Spider

Studies on spiders in their natural habitats are necessary for determining the full range of plasticity in their web-building behaviour. Plasticity in web design is hypothesised to be important for spiders building in habitats where environmental conditions cause considerable web damage. Here we compared web characteristics of the orb spider Metellina mengei (Araneae, Tetragnathidae) in two different forest habitats differing in their wind exposure. We found a notable lack of differences in web geometry, orientation and inclination between webs built along an exposed forest edge and those built inside the forest, despite marked differences in wind speed. This suggests that M. mengei did not exhibit web-building plasticity in response to wind in the field, contrasting with the findings of laboratory studies on other species of orb spiders. Instead, differences in prey capture and wind damage trade-offs between habitats may provide an explanation for our results, indicating that different species employ different strategies to cope with environmental constraints.     

Aged pesticide residues are detrimental to agrobiont spiders (Araneae)

Spiders are among the most abundant arthropods in agroecosystems, playing an important role as natural enemies of various pests. In this study we evaluated residual activity of selected pesticides on the mortality and behaviour of four spider species (Dictyna uncinata, Pardosa palustris, Philodromus cespitum and Theridion impressum). We used three pesticides: a herbicide Command (clomazone), and insecticides Decis (deltamethrin) and Nurelle (chlorpyrifos and cypermethrin). Mortality was recorded after exposure of spiders to fresh (2-h), 5, 10, 15 and 20-day-old residues. For each residue mortality was evaluated after 24–72 h. Residual effect differed between preparations and, in some cases, between spiders. All of the Nurelle residues (fresh to 20-day-old) caused 100% mortality in all spider species. Residues of Command were rather harmless (causing <20% mortality) to all spider species as the herbicide activity declined with age. Residues of Decis had species-specific effects as the mortality varied between 0 and 90%. In Dictyna the mortality gradually declined with the age of residues, while in Pardosa the mortality increased. In Philodromus and Theridion the mortality declined up to 10-day-old residues and then increased so that 20-day-old residues caused almost as high mortality as the new ones. Exposure to pesticide residues immediately affected the movement of Pardosa spiders. Residues of Decis and Nurelle decreased spider locomotion, while those of Command increased locomotion in comparison with the control. In another experiment, we studied repellence of fresh pesticide residues to Pardosa spiders. In comparison with the control, spiders stayed a similar time on the surface treated with Command, but four times less on Decis and nine times less on Nurelle-treated surfaces, respectively. In conclusion, aged insecticide residues possess a high activity and can cause long-term decline in the abundance and prolonged behavioural disturbance of spiders in agroecosystems.     

Synergy between two invasive species,redback spiders and rabbits,threaten the endangered Cromwell chafer beetle

Interactions between invasive species can be difficult to predict and can result in unanticipated impacts of significance for native fauna. Here we show that introduced European rabbits (Oryctolagus cuniculus) create habitat that enables invasive redback spiders (Latrodectus hasselti Thorell, 1870) to establish and prey upon the nationally endangered, endemic Cromwell chafer beetle (Prodontria lewisii Broun, 1904). We examined the spatial relationship between rabbit holes and redback spider occurrence, recorded all prey caught in redback spider webs over a 4-month period, and tested the role rabbit holes play in providing habitat for redback spiders experimentally, by filling in rabbit holes in areas used by spiders and monitoring subsequent occupation of the areas over four months. Redback spiders predominately resided in old rabbit holes, with the highest densities of spiders coinciding with high densities of rabbit holes. Cromwell chafer beetles were commonly caught in webs. Filling in rabbit holes eliminated the presence of redback spiders at all treated sites and reduced prey capture in those areas. Conservation management to protect Cromwell chafer beetles should focus on eliminating rabbits and their holes from beetle habitat.     

Body size and tree species composition determine variation in prey consumption in a forest‐inhabiting generalist predator

Trophic interactions may strongly depend on body size and environmental variation, but this prediction has been seldom tested in nature. Many spiders are generalist predators that use webs to intercept flying prey. The size and mesh of orb webs increases with spider size, allowing a more efficient predation on larger prey. We studied to this extent the orb‐weaving spider Araneus diadematus inhabiting forest fragments differing in edge distance, tree diversity, and tree species. These environmental variables are known to correlate with insect composition, richness, and abundance. We anticipated these forest characteristics to be a principle driver of prey consumption. We additionally hypothesized them to impact spider size at maturity and expect shifts toward larger prey size distributions in larger individuals independently from the environmental context. We quantified spider diet by means of metabarcoding of nearly 1,000 A. diadematus from a total of 53 forest plots. This approach allowed a massive screening of consumption dynamics in nature, though at the cost of identifying the exact prey identity, as well as their abundance and putative intraspecific variation. Our study confirmed A. diadematus as a generalist predator, with more than 300 prey ZOTUs detected in total. At the individual level, we found large spiders to consume fewer different species, but adding larger species to their diet. Tree species composition affected both prey species richness and size in the spider''s diet, although tree diversity per se had no influence on the consumed prey. Edges had an indirect effect on the spider diet as spiders closer to the forest edge were larger and therefore consumed larger prey. We conclude that both intraspecific size variation and tree species composition shape the consumed prey of this generalist predator.     

Viability selection on female fly finery in the wild

Female ornaments have evolved in a few taxa in which females compete for access to important resources provided by their mates. However, the effects of these sexually selected traits on survival have not been studied. Elaborate leg‐scale and/or abdominal ornaments are displayed by females of some Rhamphomyia dance flies (Diptera: Empididae) to flying males carrying prey gifts (females do not hunt). Previous analyses have shown significant sexual selection on these female traits. We studied viability selection on the traits by sampling the webs of two spider species and comparing prey R. longicauda females to survivors. We also investigated viability selection from one of the spiders over two seasons. We found that the direction of viability selection on R. longicauda from sticky Tetragnatha spider webs was consistent over two seasons. For abdominal ornaments the form of viability selection was positive and primarily directional (linear). Viability selection also favoured shorter tibiae but there was no significant selection on the size of residual tibial scale area. However, with the addition of dance fly kills from the non‐sticky, leaf‐covering webs of a Dictyna spider, abundant in only one of the seasons, the overall direction of viability selection favoured larger tibial ornaments. While noting that this viability selection on tibial scale ornaments may be a statistical artefact of the fewer traits in the two‐predator analysis (abdominal structures were missing from most Dictyna prey), we suggest that simple differences in the natural history of selective agents causing mortality may partly explain the variation in whether sexual traits are under viability selection. Viability selection on ornamental traits may vary greatly between seasons with changes in the abundances of different natural enemies so that net directional selection on these traits over many generations may be weak.     

Habitat edges affect patterns of artificial nest predation along a wetland-meadow boundary

Wetland habitats are among the most endangered ecosystems in the world. However, little is known about factors affecting the nesting success of birds in pristine grass-dominated wetlands. During three breeding periods we conducted an experiment with artificial ground nests to test two basic mechanisms (the matrix and ecotonal effects) that may result in edge effects on nest predation in grass-dominated wetland habitats. Whereas the matrix effect model supposes that predator penetrate from habitat of higher predator density to habitat of lower predator density, thus causing an edge effect in the latter, according to the ecotonal effect model predators preferentially use edge habitats over habitat interiors. In addition, we tested the edge effect in a wetland habitat using artificial shrub nests that simulated the real nests of small open-cup nesting passerines. In our study area, the lowest predation rates on ground nests were found in wetland interiors and were substantially higher along the edges of both wetland and meadow habitat. However, predation was not significantly different between meadow and wetland interiors, indicating that both mechanisms can be responsible for the edge effect in wetland edges. An increased predation rate along wetland edges was also observed for shrub nests, and resembled the predation pattern of real shrub nests in the same study area. Though we are not able to distinguish between the two mechanisms of the edge effect found, our results demonstrate that species nesting in wetland edges bordering arable land may be exposed to higher predation. Therefore, an increase in the size of wetland patches that would lead to a reduced proportion of edge areas might be a suitable management practice to protect wetland bird species in cultural European landscapes.     

Urbanization Impacts on Mammals across Urban-Forest Edges and a Predictive Model of Edge Effects

With accelerating rates of urbanization worldwide, a better understanding of ecological processes at the wildland-urban interface is critical to conserve biodiversity. We explored the effects of high and low-density housing developments on forest-dwelling mammals. Based on habitat characteristics, we expected a gradual decline in species abundance across forest-urban edges and an increased decline rate in higher contrast edges. We surveyed arboreal mammals in sites of high and low housing density along 600 m transects that spanned urban areas and areas turn on adjacent native forest. We also surveyed forest controls to test whether edge effects extended beyond our edge transects. We fitted models describing richness, total abundance and individual species abundance. Low-density housing developments provided suitable habitat for most arboreal mammals. In contrast, high-density housing developments had lower species richness, total abundance and individual species abundance, but supported the highest abundances of an urban adapter (Trichosurus vulpecula). We did not find the predicted gradual decline in species abundance. Of four species analysed, three exhibited no response to the proximity of urban boundaries, but spilled over into adjacent urban habitat to differing extents. One species (Petaurus australis) had an extended negative response to urban boundaries, suggesting that urban development has impacts beyond 300 m into adjacent forest. Our empirical work demonstrates that high-density housing developments have negative effects on both community and species level responses, except for one urban adapter. We developed a new predictive model of edge effects based on our results and the literature. To predict animal responses across edges, our framework integrates for first time: (1) habitat quality/preference, (2) species response with the proximity to the adjacent habitat, and (3) spillover extent/sensitivity to adjacent habitat boundaries. This framework will allow scientists, managers and planners better understand and predict both species responses across edges and impacts of development in mosaic landscapes.     

Removal of secondary compounds increases invertebrate abundance in lichens

We investigated how lichen carbon-based secondary compounds (CBSCs) affect abundance of invertebrates in five lichen species growing on the forest floor (Cladonia rangiferina, Cladonia stellaris) or on tree trunks (Evernia prunastri, Hypogymnia physodes, Pseudevernia furfuracea). To do this, CBSCs were removed by rinsing lichen thalli in acetone (which has no adverse effects on the lichens) and the lichens were re-transplanted in their natural habitat. After 4 months there was higher abundance of mites, springtails and spiders in the three epiphytic lichens that had their CBSC concentrations reduced. The increase in predatory spiders following CBSC reduction suggests that the compounds have multitrophic consequences. The acetone treatment reduced the number of nematodes in four of the lichen species. Given that lichens serve as important habitats for a diverse range of invertebrates, increased knowledge of how lichen CBSCs may regulate their abundance helps us to better understand the role that lichens and their defence compounds play in structuring forest food webs.     

Conspecific and Heterospecific Attraction: A Mechanism of Web-site Selection Leading to Aggregation Formation by Web-building Spiders

This study investigated whether conspecific and/or heterospecific attraction to silk is a mechanism of web-site selection leading to aggregation formation by two species of web-building spiders, Hypochilus thorelli Marx (Araneae: Hypochilidae) and Achaearanea tepidariorum (C.L. Koch) (Araneac: Theridiidae). We determined that the spatial distribution of these two spiders was clumped and that H. thorelli had a greater tendency to aggregate than did A. tepidariorum. To determine the mechanism responsible for this spatial pattern, we conducted three field experiments. We examined web-site selection by H. thorelli in three contexts: no spiders or webs present (cue removed), vacant webs present, and occupied webs present. In the case where no webs were present, there was no tendency for spiders to choose previously occupied sites as web sites. When vacant webs were present, spiders chose to occupy the vacant webs. When occupied webs were present, spiders either invaded webs and evicted the owners, or settled adjacent to and attached their webs to those of residents. Various microhabitat variables (height, angle, temperature, humidity, and substrate character of the rock surface) of randomly selected unoccupied sites and the web sites chosen by spiders were compared. There were no detectable differences between web sites and unoccupied sites with respect to any of the variables measured. This leads us to conclude that web-site choice by immigrating spiders was based on the presence of silk rather than other features of the site.     

Substrate Availability May Be More Important than Aquatic Insect Abundance in the Distribution of Riparian Orb-web Spiders in the Tropics

Spiders that are abundant along streams may depend on energy subsidies across land–water ecotones, but the effects of season and habitat structure on this trophic linkage remain poorly understood in the tropics. We carried out surveys and a manipulative experiment to investigate the effects of season and substrate availability on the distribution of riparian orb-web spiders in Hong Kong, southern China. In the surveys, spider abundance, prey, substrate use, and web orientation were recorded. The experiment involved installation of in-stream artificial substrates (ropes and bamboo poles) to increase substrate availability for web attachment. We found no seasonal difference in web abundance, but seasonal differences were observed for the prey on webs: aquatic insects (mostly Ephemeroptera and chironomid midges) contributed 69 percent of total prey collected during the wet season, but only 38 percent during the dry season. Most webs (50–80%) were < 0.5 m above the water and 45–51 percent of them tended to be orientated horizontally to the water surface and supported by overhanging vegetation and boulders. The addition of artificial substrates resulted in a 23–34 percent increase in the number of webs at the four treatment sites compared to controls, indicating that availability of web-building substrates is a critical determinant of the spider distribution. Our results suggest that riparian zones are potential 'hotspots' of food availability for spiders, and that the aquatic insect subsidy allows this habitat to support increased densities of spiders when the constraint of substrate availability is relaxed.     

Upside-down spiders build upside-down orb webs: web asymmetry,spider orientation and running speed in Cyclosa

Almost all spiders building vertical orb webs face downwards when sitting on the hubs of their webs, and their webs exhibit an up–down size asymmetry, with the lower part of the capture area being larger than the upper. However, spiders of the genus Cyclosa, which all build vertical orb webs, exhibit inter- and intraspecific variation in orientation. In particular, Cyclosa ginnaga and C. argenteoalba always face upwards, and C. octotuberculata always face downwards, whereas some C. confusa face upwards and others face downwards or even sideways. These spiders provide a unique opportunity to examine why most spiders face downwards and have asymmetrical webs. We found that upward-facing spiders had upside-down webs with larger upper parts, downward-facing spiders had normal webs with larger lower parts and sideways-facing spiders had more symmetrical webs. Downward-facing C. confusa spiders were larger than upward- and sideways-facing individuals. We also found that during prey attacks, downward-facing spiders ran significantly faster downwards than upwards, which was not the case in upward-facing spiders. These results suggest that the spider''s orientation at the hub and web asymmetry enhance its foraging efficiency by minimizing the time to reach prey trapped in the web.     

Invasive Canadian goldenrod (<Emphasis Type="Italic">Solidago canadensis</Emphasis> L.) as a preferred foraging habitat for spiders

Invasions of alien species are a great threat to biodiversity and native species communities. There are many examples in the literature on how the invasive plants affect the natural environment. Beside reports on negative effects of these invasions, there are also several studies indicating a positive impact of the invaders. Canadian goldenrod (Solidago canadensis) is one of the most invasive plant species all over Europe. Earlier studies showed that the goldenrod affects natural plant communities and has a negative impact on many animals, both vertebrates and arthropods. However, all the studies were conducted during goldenrod flowering. In contrast, this study has tested the novel hypothesis that Canadian goldenrod has a positive effect on native spider hunting success and increases spider abundance in farmland outside the growing season. Observations were made during spring on 13 meadows: 7 invaded and 6 non-invaded by goldenrod. All tall plants from experimental plots (1 m² each) were examined for the presence of spiders, their nets and prey. Prey items were counted only in spider webs. The results support the hypothesis that S. canadensis is a favourable foraging habitat for spiders: 14.6 spiders/m² on invaded plots versus 2.2 on non-invaded ones. Many spiders of the families Theridiidae and Araneidae were found on goldenrod plants, but on native plants only the Araneidae were found. Moreover, on invaded plots, much more prey items/m² were present in spiders webs than on non-invaded plots (155.3 vs. 13.8). The study is a rare example of a positive influence of invasive plants on the native arthropod community. This is also a novel approach that shows the importance of dry goldenrod stems in invaded ecosystems.     

Carcass Decoration Changes Web Structure and Prey Capture Rate in an Orb-Web Spider, Cyclosa mulmeinensis (Araneae, Araneidae)

Cyclosa spiders attach prey carcasses as decorations to their webs, but the functions of the carcasses are unclear and controversial. We investigated distinctive features of these webs in the field and conducted prey-capture experiments in the lab using the orb-web spider Cyclosa mulmeinensis. Webs with attached decoration had a significantly narrower mesh width than those without decoration and a higher degree of vertical asymmetry in the web’s shape. In the laboratory, webs without decorations trapped significantly more prey, even though other features of the webs were nearly identical. These results suggest that web decorations do not attract prey in this species, but might play other roles such as blinding predators to the spider’s presence.     

Signaling by decorating webs: luring prey or deterring predators?

Many organisms convey false signals to mislead their prey orpredators. Some orb-weaving spiders build conspicuous structureson webs called decorations. Web decorations and spider colorationsare both suggested to be important signals involved in interactionsbetween spiders and other organisms. There are several hypothesesabout the functions of signaling by decorations, among whichprey attraction had received much support, but empirical evidenceregarding predator defense is controversial. In this study,we conducted field experiments to investigate the effects ofspider decoration and coloration on insect interception ratesof webs built by Argiope aemula and to evaluate whether presenceof decorations may decrease predation risk of spiders. Decoratedwebs with spiders present had the highest prey interceptionrate, followed by undecorated webs with spiders, and then undecoratedwebs without spiders. Such results indicated that decorationsof Argiope spiders functioned as visual lures, and so did spiders'bright body colorations. In the field, almost all wasp attackevents occurred on medium-sized spiders rather than on largeones. Moreover, medium-sized Arg. aemula on decorated webs receivedfar more attacks than those on undecorated webs. Results ofthis study thus show that the signals conveyed by decorationscan visually lure prey but at the cost of an increased predationrisk. Received 20 March 2007; revised 3 August 2007; accepted 5 August 2007.     

Comparative biology of Portia africana,P. albimana,P. fimbriata,P. labiata,and P. shultzi,araneophagic, web-building jumping spiders (Araneae: Salticidae): Utilisation of webs,predatory versatility,and intraspecific interactions

Abstract

Portia is a behaviourally complex and aberrant salticid genus. The genus is of unusual importance because it is morphologically primitive. Five species were studied in nature (Australia, Kenya, Malaysia, Sri Lanka) and in the laboratory in an effort to clarify the origins of the salticids and of their unique, complex eyes. All the species of Portia studied were both web builders and cursorial. Portia was also an araneophagic web invader, and it was a highly effective predator on diverse types of alien webs. Portia was an aggressive mimic, using a complex repertoire of vibratory behaviour to deceive the host spiders on which it fed. The venom of Portia was unusually potent to other spiders; its easily autotomised legs may have helped Portia escape if attacked by its frequently dangerous prey. Portia was also kleptoparasitic and oophagic when occupying alien webs. P. fimbriata from Queensland, where cursorial salticids were superabundant, used a unique manner of stalking and capturing other salticids. The display repertoires used during intraspecific interactions were complex and varied between species. Both visual (typical of other salticids) and vibratory (typical of other web spiders) displays were used. Portia copulated both on and away from webs and frequently with the female hanging from a dragline. Males cohabited with subadult females on webs, mating after the female matured. Adult and subadult females sometimes used specialised predatory attacks against courting or mating males. Sperm induction in Portia was similar to that in other cursorial spiders. Portia mimicked detritus in shape and colour, and its slow, mechanical locomotion preserved concealment. Portia occasionally used a special defensive behaviour (wild leaping) if disturbed by a potential predator. Two types of webs were spun by all species (Type 1, small resting platforms; Type 2, large prey-capture webs). Two types of egg sacs were made, both of which were highly aberrant for a salticid. Responses of different species and both sexes of Portia were quantitatively compared for different types of prey. Many of the trends in behaviour within the genus, including quantitative differences in predatory behaviour, seemed to be related to differences in the effectiveness of the cryptic morphology of Portia in concealing the spider in its natural habitat (‘effective crypsis’). The results of the study supported, in general, Jackson & Blest’s (1982a) hypothesis of salticid evolution which, in part, proposes that salticid ancestors were web builders with poorly developed vision and that acute vision evolved in conjunction with the ancestral spiders becoming proficient as araneophagic invaders of diverse types of webs.     

The Trochidae and Turbinidae of the Kermadec Ridge (Mollusca: Gastropoda)

Abstract

Taieria erebus (Gnaphosidae) was found to be a versatile predator: it captured insects both cursorially (away from webs) and kleptopar-asitically (on alien webs); it captured spiders in both the presence and absence of webs; and it also ate the eggs of host spiders (oophagy). When T. erebus invaded webs, it was as an aggressive mimic — it performed a repertoire of vibratory behaviours to lure the host spider. Although T. erebus pursued and captured spiders on diverse web-types, it was more effective as a predator when invading densely (rather than sparsely) woven cribellate and non-sticky webs, and was especially effective on non-cribellate sticky webs. Gnaphosids are traditionally referred to as hunting spiders, but T. erebus built a small prey-capture web. T. erebus also preyed on segestriid spiders, then used their webs to catch more prey, this being an unusual example of a spider using, as a tool for predation, the spinning-work of another species from an unrelated family. T. erebus used specialised behaviours to prey on nesting cursorial spiders. Prey was either grasped or stabbed; the venom of T. erebus was highly potent against spiders. Experiments indicated that vision was of little or no importance in the predatory behaviour of T. erebus. The behaviour of T. erebus is compared to that of Portia, a web-building salticid spider which is very versatile in its predatory behaviour and has acute vision. T. erebus is discussed in relation to hypotheses concerning gnaphosid and salticid evolution.