They mostly come at night: Predation on sleeping insects by introduced candy-striped spiders in North America.

Abstract

Candy-striped spiders (two closely related and nearly indistinguishable species: Enoplognatha ovata (Clerck, 1757) and Enoplognatha latimana Hippa & Oksala, 1982) are, at first glance, unremarkable cobweb weavers in the family Theridiidae. Common across Europe, they were presumably accidentally introduced to North America through human transport. First recorded on the East and West Coasts around 1900, they now occur across most of the United States and Canada (Oxford & Reillo, 1994). Though much effort has gone into studying the striking color polymorphism of these spiders, whose abdomens may be plain yellow, red-striped, or red (e.g., Oxford & Reillo, 1993), little work has focused on their behavior (but see Greco & Kevan, 1999). Bristowe (1931) concluded a brief paper about Enoplognatha ovata by stating that it “is a very common spider whose habits have never attracted anyone's serious attention.” This statement remained true for much of the following century, but the candy-striped spider's predatory behavior captured our attention in the summer of 2016 when we noticed it behaving in a rather extraordinary way. One of us (Sean McCann) is an avid macro photographer for whom sleeping insects are particularly rewarding subjects. In the early morning, one can often find sleeping bees, wasps, and sometimes other insects perched on dead vegetation (Rau & Rau, 1916). The stillness of these subjects in the cool temperatures of the predawn is very convenient for photography—later in the day these same insects fly rapidly and are difficult to capture. One morning during a photography session on the coastal dunes of southern Vancouver Island in 2016, we came upon a scene much like the one in Figure 1: A candy-striped spider (either E. ovata or E. latimana—both occur at this site) had massacred a group of sleeping wasps during the night. The grisly tableau sparked our curiosity—was this an example of a marauding spider exploiting the same weakness that facilitates photography (cold-induced stillness and lethargy) to capture large and well-defended prey without the aid of a capture web? We first turned to the literature to search for information about the predatory behavior of candy-striped spiders. English naturalist W. S. Bristowe provided delightfully rich descriptions of the natural history of E. ovata in an early paper and one of his books (Bristowe, 1931, 1958). He described candy-striped spiders (as Theridion ovatum) hunting from small, inconspicuous cobwebs built in low vegetation or under flower heads (Figure 2a): In spite of its flimsy appearance Theridion ovatum is an intrepid warrior. If an insect which comes to sip the honey of the ragwort, Umbellifer, or other flower beneath which a Theridion is lurking happens to touch one of its threads the spider in an instant hurls viscid threads at it with its hind legs. It does this regardless of the size of the insect, and I have watched numerous battles in which the spider was ultimately victorious over flies many times its own bulk. (Bristowe, 1931) Sometimes a commotion in an adjoining web belonging to another spider attracts her attention and she then invades her neighbour's web by way of the connecting threads […] Poaching can lead to darker deeds and I once saw [an E. ovata] eating a [Phylloneta sisyphia] in the latter's web. Candy-striped spiders can also forage on plants without the aid of a web. They prey on leafhoppers by eavesdropping on their vibratory communication signals (Virant-Doberlet et al., 2011), and we observed predation events on grass stems and other plants where no capture web was evident (Figure 2e; Appendix S1: Table S2). Long-jawed orb weavers in the genus Tetragnatha (Araneae: Tetragnathidae) are also known to forage off their webs, but such active hunting is uncommon among web-building spiders (Gould, 2021). Already it was clear that candy-striped spiders had an unusually broad and flexible repertoire of foraging behaviors. Nonetheless, we were unable to find any records of the final predation tactic that we witnessed: actively locating and capturing sleeping insects perched on dead vegetation, which we call marauding (Figures 1 and 2f,g; Appendix S1: Table S2). In an effort to document and better understand nighttime marauding by candy-striped spiders, we searched for sleeping insects in the coastal sand dune habitat where we were working in the summers of 2016 and 2017. We witnessed some natural attacks in progress (Figure 2f; https://doi.org/10.6084/m9.figshare.c.6329204, Video1_natural_marauding.mp4, Video2_natural_marauding.mp4), and we also introduced spiders to stems of dead plants where sleeping insects were perched to observe their predation behavior in detail (https://doi.org/10.6084/m9.figshare.c.6329204, Video3_staged_marauding.mp4, Video4_staged_marauding.mp4, 22_marauding_staged_Prionyx_gumweed.jpg). Candy-striped spiders readily climbed dead vegetation and, upon detecting a sleeping insect (apparently by making contact with the tip of a foreleg), turned their bodies and began to wrap them extensively with sticky silk. Despite efforts by prey insects to defend themselves by kicking, stinging, and/or flying away, the spiders were ultimately successful in subduing their prey in five of the six interactions we observed. Like other theridiid spiders, candy-striped spiders produce strong silk studded with sticky glue droplets, and their ability to fling it with their hind legs while remaining at a safe distance from prey make them very effective predators of large and well-defended insects. We suspect, however, that the time it takes a sleeping insect to warm up before being capable of active defense and flight is a key contributor to the success of marauding. Indeed, during one video recording of attempted predation on an Ammophila in which we used very bright lights that emitted significant heat, the warmed-up wasp was able to quickly respond to and repel the approaching spider by kicking it with one leg (https://doi.org/10.6084/m9.figshare.c.6329204, Video3_staged_marauding.mp4). To determine how common marauding might be relative to previously recorded foraging tactics (web-based ambush predation, web invasion for kleptoparasitism and/or araneophagy, and active hunting on living plants), we engaged in observations and sampling over 14 days between 12 July and 8 August 2017 (detailed methods in Appendix S1). Briefly, we collected data (photographs and/or specimens) to determine prey taxa and the tactics used to capture them. If we did not see the predation event in progress, we inferred the tactic based on time of day, prey taxon and habit (e.g., nocturnal vs. diurnal, nectar-feeding), presence or absence of a capture web, plant type, and/or whether the plant was alive. We are confident that diurnal prey captured on dead plants resulted from marauding because (1) diurnal nectar-feeding insects would have little reason to visit dead plants during the day, (2) a dead plant would provide a poor site for a candy-striped spider web because the spider would be highly visible to predators and prey, (3) we never observed spiders on dead plants during the day unless they were actively feeding on a large prey insect (presumably captured the night prior), and (4) all diurnal prey taxa that we observed being consumed on dead plants in daylight were taxa that we had also observed sleeping perched on dead plants. Conversely, we interpreted prey remains or active feeding on prey in candy-striped spider webs on live flowering plants as ambush predation based on (1) the presence of capture webs and (2) because we only very rarely observed insects sleeping on live plants. This survey yielded 250 identifiable prey items and revealed that the candy-striped spiders in this habitat have a broad diet dominated by hymenopterans (Figure 3; Appendix S1: Table S2). The successful predation events we documented were most often the result of marauding (42%) or web-based ambush (38%), with active hunting and web invasion much less common (Figure 3). In North America, candy-striped spiders often occur in very high densities (e.g., Tomascik, 2015), and they thrive in a variety of habitats, suggesting that they may have strong impacts on insect communities. At our field site, the most common prey are in orders that contain pollinators (Figure 3), and these insects are taken both by marauding and web-based ambush. When hunting on webs under flowers, candy-striped spiders operate much like crab spiders, which are sit-and-wait predators of flower visitors that affect the mortality and behavior of pollinators, even at low densities and prey-capture rates (Benoit & Kalisz, 2020). In vulnerable ecosystems like the coastal sand dunes where we made our observations, there may be conservation implications of predation by these introduced spiders on insects that are habitat specialists or pollinators of threatened species. Candy-striped spiders are certainly worthy of further study, both to determine their ecological impacts and to understand the evolution of their remarkable predatory versatility (Curio, 1976). This study highlights the value of curiosity-driven observational natural history to inspire countless research questions that span behavioral, evolutionary, and classical ecology and the impact of invasive predators across temporal and physical scales. Indeed, one of us (Catherine E. Scott) has built their current research program on the foundations reported here and could easily spend an entire career studying these spiders. The value of observing organisms in their natural environments to spark good research questions is immense, but we too seldom take the time to simply watch. If it took almost 100 years before anyone reported on the remarkable marauding behavior of an extremely common spider, how many other intriguing but as yet undescribed ecological interactions are just waiting for a curious observer to take the time to record them? We are very grateful to the Tsawout First Nation for allowing us to work on their lands. This work was supported by a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery grant (to M. C. B. Andrade), a NSERC CGS-D and Toronto Entomologists' Association Eberlie Grant (to C. E. Scott), and many generous donors to the Team Black Widow crowdfunding campaign. We also thank M. D. Jackson, C. M. Buddle, and two anonymous reviewers for helpful and encouraging comments on the manuscript. The authors declare no conflicts of interest. Data (Scott & McCann, 2023) are available in Figshare at https://doi.org/10.6084/m9.figshare.c.6329204. Voucher specimens are deposited in the entomology collection of the Royal British Columbia Museum (https://royalbcmuseum.bc.ca/collections/natural-history/entomology); catalogue numbers are provided in Appendix S1. Appendix S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.