An extensive review of mutualistic and similar ecological associations involving tarantulas (Araneae: Theraphosidae), with a new hypothesis on the evolution of their hirsuteness Author Zamani, Alireza Zoological Museum, Biodiversity Unit, University of Turku, Turku, Finland; Author West, Rick C. Sooke, British Columbia, Canada; Author Lamar, William W. College of Sciences, University of Texas at Tyler, Tyler, TX, USA & Zoological Museum, Biodiversity Unit, University of Turku, Turku, Finland; text Journal of Natural History 2024 2024-08-06 58 29 - 32 1107 1143 http://dx.doi.org/10.1080/00222933.2024.2382404 journal article 10.1080/00222933.2024.2382404 1464-5262 13758139 Ants ( Hymenoptera : Formicidae ) Given that both ants and spiders are amongst the most abundant arthropods, especially in tropical regions, interactions between various species of the two groups are common ( Lapinski 2019 ; Figure 7A–F ). However, reports of such interactions in the literature are very rare, with almost all of them describing a predatory behaviour by the ants ( Figure 7F ). Three exceptions to this are mentioned here. Lapinski (2019) documented in detail an interaction in Costa Rica between the army ants Labidus coecus (Latreille, 1802) and a female Stichoplastoris cf. obelix (Valerio, 1980) ( Theraphosinae ) and her juveniles ( Figure 7C, D ). No aggressive behaviour was observed in this association. Instead, the ants were seen entering the tarantulas’ burrow to collect and remove organic material. This caused the spiders to temporarily vacate the burrow, but they returned once the ants had left the area. In another note, Hirschi (1991) reports having observed a small species of Tapinauchenius Ausserer, 1871 (Psalmopoeinae) in Ecuador , exclusively in association with an unidentified, aggressive species of ant that inhabits abandoned termitaria. The author notes that despite provocation, the tarantula remained unharmed by the ants, which otherwise attacked the author himself. Finally, Chomphuphuang et al . (2017) briefly noted that some specimens of Phlogiellus longipalpus Chomphuphuang et al ., 2017 ( Selenocosmiinae ) described in their paper were collected from retreats built in ant colonies. Figure 7. Interactions between tarantulas and ants. A . Avicularia purpurea (marked with an arrow) in its arboreal retreat, living with a colony of an unidentified species of Camponotus , Tena , Napo, Ecuador. B . Tapinauchenius cupreus living with Camponotus femoratus in a tree cavity, Río Momón, Loreto, Peru. C . Stichoplastoris cf. obelix , spiderlings, and Labidus coecus ants not being interested in the spiders, Reserva Biológica Tirimbina, Heredia Province, Costa Rica. D . Same, spiderlings gathered at the maternal burrow entrance, with ants moving close to the entrance. E . Avicularia hirschii displaying an escape strategy against Labidus ants, Madre de Dios, Peru. F . Tapinauchenius plumipes caught, killed and carved up by the army ants Eciton burchellii , Montsinéry-Tonnegrande Commune , French Guiana. D reproduced from Lapinski (2019) . Photo credits: Rick C. West ( A, B, F ), Witold Lapinski ( C, D ), and Emanuele Biggi ( E ). Figure 8. Tarantulas living in termitaria. A . Avicularia juruensis , nr. Iquitos, Loreto, Peru. B . Vitalius dubius , Dona Amélia Farm , Santo Antônio de Possee, São Paulo, Brazil. C . Nhandu coloratovillosus , São Geraldo do Araguaia, Pará, Brazil. D . Nhandu coloratovillosus , Peixe , Tocantins, Brazil. E . Psalmopoeus cambridgei , Tamana Hill , Sangre Grande, Trinidad Island, West Indies. F . Brachionopus sp. with Trinervitermes sp. , Ezemvelo Nature Reserve, Tshwane, Gauteng Province, South Africa. Photo credits: Alexey Yakovlev ( A ), Ivan Sazima ( B ), Fernando J.M. Rojas-Runjaic ( C ), Danté Fenolio ( D ), Sarah Crews ( E ), and Luke Goddard ( F ). Here, we report on a few cases of non-agonistic interactions between tarantulas and ants, mostly based on observations of RCW and WWL in South America. The first cases involve Avicularia avicularia (Linnaeus, 1758) ( Aviculariinae ) and Tapinauchenius plumipes (C.L. Koch, 1842) . In Venezuela , juveniles of the two aforementioned theraphosid species were observed cohabiting with the arboreal ant Camponotus femoratus (Fabricius, 1804) in ‘ant gardens’. Ant gardens are an association between ants and plants, consisting of soil and decayed organic matter, with living epiphytic plants growing out of it and the ant colony itself living inside ( Orivel and Leroy 2011 ). When the ant garden with a juvenile tarantula living on it was bumped, the spider retreated to the protection of its self-made silken retreat while the ants readily swarmed over the garden’s exterior, attacking in defence by biting and spraying formic acid. In about 30 minutes, the ant swarming subsided and the tarantula emerged and resumed its hunting position. By day, the spiders were secure in their individual retreats which, in part, bore into the ant garden. The foraging ants avoided the spider’s silken retreat, the entrance of which was not always silked over to prevent an ant from entering. At night, worker ants were more visibly active and going about their cultivating and food gathering activities on the exterior of the garden and surrounding vegetation. At the same time, the juvenile theraphosid would be out of its silken retreat and stretched out on the side of the ant garden waiting for passing prey. The ants were observed scurrying around the theraphosid, but never interacting with it nor trying to enter its retreat. Additionally, the juvenile theraphosid was never observed trying to predate a passing ant. The ant gardens observed were not large enough or located in such a place that they could support the needs of a subadult or adult-sized arboreal theraphosid of the aforementioned species. As observed in the past, as arboreal theraphosids and their prey demands and sizes grow, they relocate higher towards the canopy of the same tree or on an adjacent tree. In Ecuador and Peru , small arboreal termite mounds were observed constructed on the sides of spiny palm trees and in the fork of living trees that had been taken over by colonies of either C. femoratus or stinging ants, possibly of the genus Pseudomyrmex Lund, 1831 . Juveniles of both Avicularia purpurea Kirk, 1990 ( Figure 7A ) and Tapinauchenius cupreus Schmidt and Bauer, 1996 ( Figure 7B ) were observed in their self-made silken retreats in cracks or crevices of the abandoned termite mound. By day, each juvenile theraphosid remained concealed in its retreat. Small numbers of worker ants foraged on the living tree. At night, the tarantulas would emerge from their retreat and position themselves on the support tree or palm to wait for passing prey. As noted above, ant activity on the side of the support tree or palm was busier at night with passing ants ignoring the juvenile tarantula and vice versa. On another occasion in Peru , WWL and RCW noticed a large active mound of the leafcutter ant Atta sexdens (Linnaeus, 1758) , approximately 6 m across and 1 m in height. There were multiple entry holes into the mound made by the ants, some actively being used by the ants while some of the earlier-made holes in the mound were no longer being used by the ants and were now being utilised as retreats by an unidentified species of Pamphobeteus . On this particular leafcutter ant mound, 12 individuals of Pamphobeteus of varying sizes were found individually occupying unused entry holes into the mound as their retreat. One another occasion, while photographing an adult female Megaphobema velvetosoma Schmidt, 1995 ( Theraphosinae ) eating a bush cricket at its fossorial burrow entrance in rainforest near the Río Nanay, Loreto , Peru , long columns of the army ants Eciton burchellii Westwood, 1842 were observed swarming over the ground and low vegetation, hunting for invertebrates and small vertebrates. Sensing the foraging army ants’ approach, the tarantula retreated to the rear of its short ground burrow while still grasping the bush cricket in its chelicerae. By flashlight, the spider was observed pressed against the rear of its blind end retreat; it had dropped its partially consumed prey and drew its legs up and remained still at the back of its retreat. The army ants entered the burrow and swarmed over the dead bush cricket and tarantula. The ants quickly carved up and carried off pieces of the bush cricket from the retreat. Several ants tried to bite the tarantula but were unsuccessful due to the hirsuteness of the tarsal scopula, palps, legs and abdomen. The majority of ants that entered the burrow, however, appeared disinterested in the large tarantula that would, potentially, have been a major food source for them. The ants eventually gave up trying to subdue the tarantula and left the retreat while the spider remained motionless for about another 20 minutes, before returning to a hunting stance at the entrance of its retreat. During recent fieldwork in the Pará State of Brazil , R. Bertani documented an instance of cohabitation involving a subadult Acanthoscurria geniculata (C.L. Koch, 1841) residing within a subterranean ant colony, likely of the genus Solenopsis Westwood, 1840 . Upon disturbing the colony, the ants aggressively swarmed the observer’s legs, preventing him from capturing photographs (Bertani, pers. comm.). Typically, Solenopsis ants exhibit defensive behaviours and will attack and sting any intruding animal to protect their colony. However, in this particular observation, the theraphosid spider was not attacked by the ants, suggesting a possible association or tolerance between the species. Finally, we also found four photographs on iNaturalist illustrating a potential association between theraphosids and ants in India and Thailand ( Table 3 ), which warrant further investigation.