When Cockroaches Replace Ants in Trophobiosis: A New Major Life-Trait Pattern of Hemiptera Planthoppers Behaviour Disclosed When Synthesizing Photographic Data

Abstract

The mutualistic interspecific relationships of trophobiosis between trophobiont planthoppers (Hemiptera, Fulgoromorpha) providing food to the host called xenobiont, are reviewed. The degree of interspecific relationships between these symbionts varies from occasional or short time duration (a few hours to a few days) to longer ones, with trophobionts left free to escape (optobiotic type) by the xenobiont, or maintained enclosed in nests or ant shelters (cryptobiotic type). Of 267 collected cases, 126 are new illustrated observations. Occasional trophobiosis is documented in 13 families of planthoppers and appears to be quite general in Fulgoromorpha, although it is reported for the first time for Dictyopharidae, Eurybrachidae, and Nogodinidae. Xenobionts associated with planthoppers are reported from ants and other Hymenoptera, Lepidoptera, and Blattodea, but also from Mollusca and even small gekkonid vertebrates. Tettigometridae appear to be exclusively tended by ants, while Fulgoridae significantly more often by cockroaches (40%) than by ants (27%). Long-time trophobiosis occurs always with ants, cryptobiotic ones reported in Cixiidae, Delphacidae, Tettigometridae, Meenoplidae, Flatidae and Hypochthonellidae, while optobiotic ones remain restricted to tettigometrids. A particular focus on Tettigometridae attended by ants is provided with new etho-ecological observations of 92 currently described tettigometrids species, 32 different species (35%) are now known to be able to be ant-attended. In Bulgaria, where fourteen species occur, trophobiosis occurs with at least five species of them (36%). In tettigometrids, subsociality, sessility, and underground life appear to be key factors allowing more complex relationships with ants. However, the planthopper size and thus the amount of food (drops of honeydew) is probably also an important factor. This might explain many new observations in large-sized and often isolated fulgorids with cockroaches. Tapping of trophobiont forewings by cockroaches, moths, or of the bark subtrate by geckos has been observed, but antennal palpation behaviours by ants are the most commonly observed with tettigometrids, although not with larger planthoppers. In tettigometrids, specific tegumentary glands secretions (allomones) of the abdomen pleurites might also mediate their long-term mutualistic associations, even possibly completing honeydew kairomones actions mediating planthopper trophobiosis in general.

1. Introduction

Trophobiosis is defined as a disjunctive symbiotic association based on an interspecific relationship between two symbiotic organisms: one member, the host or trophobiont, providing food to the second. This particular type of interspecific interaction ranges from irregular and facultative relationships to mutualistic long-term associations, some being permanent, where both symbiotic organisms gain benefit. Trophobiosis is best known between ants (Insecta, Hymenoptera, Formicidae) and hemipterans (Insecta, Hemiptera) [1,2]. It also occurs between some other taxa, such as in well-known cases of ants and Lycaenidae (Insecta, Lepidoptera) [3]. Among Hemiptera, aphids and coccids (Sternorrhyncha) are commonly known to enter in mutualistic associations with ants, but this type of association occurs more rarely also with Auchenorrhyncha species of both suborders, Cicadomorpha and Fulgoromorpha [4,5,6].

With Hemiptera, ants might exploit the relationship on a short term (occasional trophobiotic interactions) or on a longer term (trophobiosis evolving into mutualistic interactions) by few to several or many individuals. In exchange, ants obtain drops of honeydew excreted by the host, a sugar-rich source, resulting from feeding from phloem vessels by these phytophagous hoppers. The secreted liquid metabolic waste remains, however, rich in nutrients (various sugars, amino acids) for the attending host [7]. Its chemical composition depends on several biological factors of the hopper (instar, symbiotic bacteria present in the gut) but also of the host plant and the pressure of its infestation (review in [7]). When non-attended, the feeding planthopper regularly produces a drop of honeydew at the end of the anal tube and, by a kicking movement of the abdomen and legs, projects the drop in the near environment of the insect. Honeydew drops act both as a volatile and contact pheromone due to the natural weathering (fermentation, oxidation, etc.) of the sugars and amino acids contained in this excretion [7] and might attract predators and parasitoids. However, it is also an important source of nutrition for attending hosts and particularly for ants which might enter into a mutualistic interaction with the trophobiont providing it. In exchange, protection from predators and parasitoids, transport when environmental conditions are unfavourable, waste removal, and protection of eggs and nymphs are provided by the ants [5]. Costs and benefits of these associations to both hemipterans and ants was recently reviewed in [2].

Most observations and investigations on trophobiosis between ants and Auchenorrhyncha have been made within members of the Cicadomorpha group (mainly Membracidae from the tropics and subtropics), but less so with Fulgoromorpha where it has been reported sporadically in a few planthopper families only: Caliscelidae, Cixiidae, Delphacidae, Issidae, Flatidae, Fulgoridae, Meenoplidae, and Hypochthonellidae (reviewed in [4,6,8]). However, among Fulgoromorpha most mutualistic ant-attended records concern the family Tettigometridae for which both some morphological and behavioural characters (sedentary and gregarious habits) were considered as facilitating ant attendance [4].

In the frame of a wider project about interspecific relationships of the planthoppers with other associated species, we provide here the review of the different aspects of trophobiosis from 13 Fulgoromorpha families (267 cases) with a special focus on tettigometrids, which include 40.8% of the observations.

2. Materials and Methods

Our bibliographic survey on trophobiosis was supplemented both by our personal photographic documents and by those collected for several years by data mining on the Internet. This approach has also been successfully reinforced by the involvement of citizen science—in particular, thanks to the project Knowledge of the Biodiversity of Fulgoridae in Cambodia [9]. This heuristic strategy makes it possible to rely on persistent and verifiable information provided by photographic data by objectively documenting a trophobiotic relationship and by confirming, as precisely as possible, the taxonomic determination of the trophobiont-xenobiont pair.

Additionally, ant-associated tettigometrids were particularly studied in surveys of the Auchenorrhyncha fauna, mainly in Bulgaria, but also in some other Southern European countries—Albania, North Macedonia, Greece, Malta, Italy, and France between 2001 and 2019. Observations of ant visits or co-habitation were recorded and photographed and voucher specimens of trophobionts, ants and host plants were sampled and identified. The tettigometrids were collected by direct sampling or by using an entomological net. In most cases, both tettigometrids and ants were stored in ethyl acetate and transported to the laboratory, while plants were dried using a standard method and stored in an herbarium. During field observations, the life stages, the location on the host plant or another type of microhabitat, the size of aggregations, the presence of direct contact and stimuli, and various type of ant care—antennating, egg protection, transfer of nymphs and pushing of adults—were also documented.

In all symbiotic pairs when the trophobiont could be identified (n = 108), its body length was estimated by the average of the minimum and maximum lengths provided with its original description or taxonomic revisions, or directly from specimens in collection. For larger fulgorid species, it was counted from the top of the cephalic capsule in dorsal view to the end of the tegmina. When the trophobiont was only identified to genus, an approximation to the average length of the known species in the genus and in the distribution area where it was reported was used (Appendix A).

3. Results

Two hundred and sixty-seven trophobiont-xenobiont pairs that have been collected from the published literature and internet resources (

Table 1. Numbers of trophobiosis observations (distinct trophobiont-xenobiont pairs) reported by planthoppers families, for identified planthopper taxa (at least to genus level), and by xenobiont taxa. For Tettigometridae, figures are also reported by subfamilies. In brackets: 2 probably non-true trophobiosis cases.

Planthopper Families		Nb Observ.		Nb ≠ Identif.		Attended by Xenobionts
						Ants		Cockroaches	Moths	Gastropods	Gekkonids	Others
Undetermined family		3		0		1		0	0	0	2	0
Caliscelidae		2		2		2		0	0	0	0	0
Cixiidae		20		11		17		3	0	0	0	0
Delphacidae		29		4		29		0	0	0	0	0
Dictyopharidae		1		1		1		0	0	0	0	0
Eurybrachidae		2		2		1		1	0	0	0	0
Flatidae		10		4		9		1	0	0	0	0
Fulgoridae		70		31		19		28	8	5	8	(2)
Hypochthonellidae		1		1		1		0	0	0	0	0
Issidae		13		10		12		1	0	0	0	0
Meenoplidae		2		2		2		0	0	0	0	0
Nogodinidae		3		2		2		1	0	0	0	0
Ricaniidae		2		2		0		0	0	0	1	1
Tettigometridae	Tettigometrinae	109	66	35	19	109	66	0	0	0	0	0
	Egropinae		4		4		4
	Nototettigometrinae		39		12		39
Total		267		107		205		35	8	5	11	1

), 126 of which (47.2%) are newly reported, with most being illustrated here. Two observations remain doubtful as being true trophobiosis relationships and were not included in our calculations (Table 1). One hundred and seven different planthoppers were identified to species.

Ant-attended planthoppers are the most frequent cases with 204 observations (76.8%) and have been reported for all planthopper families except for Ricaniidae. However, other groups are also reported, particularly cockroaches (35 cases, 13.1%), but also gekkonids (16 cases, 6%), moths (8 cases, 3%) and gastropods (5 cases, 1.9%).

If tettigometrids appear to be exclusively tended by ants, fulgorids are more often tended by cockroaches (31 cases, 40%) than by ants (19 cases, 27%), while the other observations also concern moths (8 cases, 11.4%), gekkonids (8 cases, 11.4%) and Gasteropods (5 cases, 7%). Trophobiosis involving cockroaches are mainly reported for fulgorids (80%) but are also known from Cixiidae (8.6%) and Eurybrachidae, Flatidae, Issidae, and Ricaniidae (2.8% each). For 178 symbiont pairs in which the trophobiont was identified at least to the genus level (147 involving ants and 31 cockroaches), a statistical two-sample Student t-test on the average size of the trophobionts shows significantly (p = 0.0054, threshold 5%) that cockroaches xenobionts are more likely to enter in trophobiose with larger sized trophobionts than ants (Figure 1).

All data were extracted either from literature references or from original photos collected on the web or newly provided here, and carefully checked to confirm trophobiont identification to species or genus. Thirteen planthoppers families of twenty-one currently recognized [10] are concerned: Caliscelidae (Figure 2), Cixiidae (Figure 3 and Figure 4), Delphacidae (Figure 5), Dictyopharidae and Eurybrachidae (Figure 6), reported for the time, Flatidae (Figure 7), Fulgoridae (Figure 8, Figure 9, Figure 10, Figure 11, Figure 12, Figure 13, Figure 14, Figure 15, Figure 16, Figure 17, Figure 18, Figure 19, Figure 20, Figure 21 and Figure 22), Nogodinidae (Figure 23) also reported for the first time, Tettigometridae: Nototettigometrinae (Figure 24, Figure 25, Figure 26 and Figure 27), Tettigometridae: Tettigometrinae (Figure 28, Figure 29, Figure 30, Figure 31, Figure 32, Figure 33, Figure 34 and Figure 35), and a few indetermined planthoppers (Figure 36). For the most important families, 109 cases (40.8%) involve 35 different tettigometrid species and 72 cases (26.9%) involve 36 different fulgorid species; delphacids remain poorly documented for 11% of the observations, followed by cixiids (7%) and issids (5%).

4. Discussion

4.1. Trophobiosis: Terminology

In most specific symbioses, the two partners of the interspecific relationships are just called symbionts, a generic term preventing the ability to differentiate them precisely. In trophobiosis, the first member, which provides some food to the second, is called the trophobiont [11]. The name is derived from the Ancient Greek τροφή, trophē, meaning “nourishment” and—βίωσις -biosis. In order to better identify the second symbiont in this particular symbiotic behaviour from symbionts active in others kind of behaviours occurring in other types of interspecific relationships, we propose to call the symbiont benefiting from the food a xenobiont. The name is derived from the Ancient Greek noun ξένος, xénos, meaning “foreigner, or guest” and—βίωσις-biosis. Accordingly, we use the term trophobiosis in the following definition: a symbiotic association between a pair of organisms or symbionts, where one member, the xenobiont, modifies its usual behaviour, waits, and collects after or without stimulation of it, food provided by the second member, the trophobiont. With this definition we exclude cases of species collecting occasionally and randomly drops of honeydew kicked around by the planthopper, or circumstantial cases where high amounts of honeydew produced by great numbers of planthopper individuals in the same place attracts unusual opportunistic species not developing particular trophobiosis behaviours.

Two main types of trophobiosis, mostly ant attendance, have been reported in the Fulgoromorpha literature [4]. The first type is an occasional one and established for a short time (from some hours to a few days), generally concerning one or few individuals; the second reports for mutualistic long-time attending periods on several weeks to months and involve several individuals (one adult and/or young instars to several/many adults and nymphs). Long-time trophobioses are divided into cryptobiotic and optobiotic ones according to whether the trophobiont is maintained confined or free to escape [4]. This division approaches Donisthorpe’s one [12] who used the terms “extranidal” to denote ant guests that occur and interact with ants outside the ant nest, and “intranidal” for those living in an ant colony, and actively seek out ants. However, Donisthorpe’s division is too restrictive, applying especially for ant-attending cases and active “guests”.

In general, trophobionts seem little affected in their usual behaviours, but they are especially more active in their food intake being solicited by the xenobiont. In reverse, xenobionts display more specific behaviours when initiating a trophobiosis relationship: (1) waiting for drops of honeydew, (2) collection of them on the surrrounding substrate, (3) interception of honeydew stream when they are emitted after kicking, or (4) before, directly at the anal extremity of the trophobiont, (5) guidance of larvae and adults to ant nests with (6) possibly phoresy or transportation of individuals (including eggs) by ants called trophophoresy [13], and (7) active protection against predators or parasites. Opportunistic behaviour by ants collecting honeydew already actively collected by other xenobionts or kleptotrophobiosis has also been reported [14].

Table 2. Different types of trophobiosis behaviours observed with Fulgoromorpha (Hemiptera).

No trophobiosis	Disjunctive trophobiosis (facultative)
No change in symbionts behaviour	Xenobiont and trophobiont change their behaviour xenobiont: - waits close to the hopper, - avoids making it escape, and does not try to predate it, - collects on the surrounding substrate or intercepts the kicked honeydew drops, - or directly obtains them at the extremity of the trophobiont’s anal tube, - often maintains active contacts (with antennae, palps, legs, …) stimulating the trophobiont, - might guide and protect the trophobiont from predators trophobiont: - increases rythm of honeydew drop production, - does not escape
Non-attendance Occasional collection of honeydew drops kicked by the planthopper, randomly deposited on the surrounding substrate	Short time attendance occasional	Long-time attendance: - xenobiont adopts guiding and protective behaviours - possibly even trophophoresy
	Optobiotic (open life)	Cryptobiotic (confined life)
	free trophobionts, sessile, often subsocial or gregarious	Underground: trophobionts in ant nests feeding on roots	Under shelter: trophobionts in aerial ant nests feeding on aerial parts of the plant

summarizes the different behaviours observed.

4.2. Planthopper Xenobionts

Until the end of the 19th and beginning of the 20th century, trophobiosis was only occasionally reported for planthoppers, mostly for Tettigometridae. In the last forty years, it appeared to be much more common than expected. Since the review of Bourgoin [4] several new examples have been documented [8,9,10,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39] and reported in FLOW [10]. In addition, the development of digital macrophotography and internet tools allowing sharing and dissemination of photos and videos more easily (iNaturalist, Facebook groups, Flickr, etc.), have provided an important new source of very valuable natural history data [9]. They are, however, scattered on the internet, difficult to find, and the species are often not identified. Appendix A summarizes all available information for attended planthoppers and trophobiosis found on the internet together with the papers previously published.

Most published planthopper xenobionts belong to ants, which appear to be engaged in opportunistic or occasional behaviours. From all available information (Appendix A), and particularly for Tettigometridae, which remain the best documented, attending ants belong to one of the five Formicidae subfamilies: Dolichoderinae Forel, 1878, Formicinae Latreille, 1809, Myrmicinae Lepeletier de Saint-Fargeau, 1835, Ponerinae Lepeletier de Saint-Fargeau, 1835, and Dorylinae Leach, 1815, and almost twenty different genera are concerned. However, this pattern appears to be not specific for tettigometrids or planthoppers, but is general for sap sucking Hemiptera [40].

However, several unexpected planthopper attendances by other xenobionts taxa have also recently been reported with other insect groups. A well-known case in Australia is the occasional consumption of poisoned honey for man when domestic honey bees (Hymenoptera, Apidae) collect honeydew produced by the passion-vine ricaniid planthopper, Scolypopa australis (Walker, 1851), when feeding on the poisonous shrub tutu (Coriaria arborea Lindsay) (Cucurbitales, Coriariaceae D.C.) [41,42,43]. Recently, high concentrations of the spotted lanterfly, Lycorma delicatula (White, 1845) (Fulgoridae), newly invasive in the USA, have shown various insects such as hornets, yellow jackets and paper wasps (Hymenoptera, Vespidae), Diptera and gastropods (slugs) to be attracted by the vast amount of honeydew production (Figure 10b,c, Figure 11 and Figure 12a). If most of these cases should be reported as opportunistic collection of honeydew drops on the substrate around the fulgorids (not true trophobiosis), and in some cases with short-time occasional attendance. They may also attract more regularly domestic honey bees, which produce a “spotted lanternfly dark honey” of a complex distinct flavor, now commercialized in Philadelphia Pennsylvania USA [44]. Similar observations were made in Madagascar and Vietnam with high concentrations of flatids feeding on the same plant (personal observations by the authors). More unusual and also only recently reported, are Blattodea [9,14,23,33], and various families of Lepidoptera—Tortricidae, Nolidae, Nymphalidae and Erebidae [14,45]. Several new cases involving Blattodea and Lepidoptera are reported and illustrated in this paper (Figure 3, Figure 4, Figure 6, Figure 7, Figure 8, Figure 9, Figure 10, Figure 12, Figure 13, Figure 14, Figure 15, Figure 16, Figure 17, Figure 18, Figure 20, Figure 21 and Figure 23).

More spectacular are trophobiosis cases of medium-sized Flatidae planthoppers tended by small Gekkonidae in Madagascar [24,34], although geckos are well known predators mainly feeding on insects, spiders, snails, and small vertebrates [34]. In Thailand Kemal and Koçak [31] reported 20–30 large-sized adult fulgorids Pyrops candelaria (Linnaeus, 1758) attended by 4–5 geckos Hemidactylus platyurus (Schneider, 1797) on tree trunks (Figure 16d) near the bottom of the tree in the morning, but in the upper parts in the noon and afternoon, probably due to the daily heat. While trophobiotic behaviour could not be directly observed in this case, a similar gecko-attendance observation with trophobiosis was however reported later by Constant [33] in Borneo.

Another rather unexpected case of a short-time trophobiosis is the one of the molluscan land snail Euglandina aurantiaca Angas (now Pittieria aurantiaca (Angas, 1879)) (Spiraxidae), reported by Naskrecki and Nishida [14], intercepting droplets of honeydew stream kicked away by three large-sized fulgorids Enchoptera sanguinea Distant, 1887, Enchophora stillifera (Stål, 1862) and Phrictus quinquepartitus Distant, 1883, in Costa Rica. The speed of the ejected droplets of honeydews by the hopper varies with species from 0.796 ms⁻¹ to 1.695 ms⁻¹ (±3–6 km/h) [14]. New cases with molluscs are reported here (Figure 8d, Figure 15c and Figure 20c,d). One very spectacular and up-to-now unique case was also recently documented in Borneo (Figure 19d,e) of a large-sized species of Fulgoridae, Scamandra polychroma Gerstaecker, 1895, totally surrounded by a wall of undetermined small ants. According to P. Soltys, who took this fantastic picture, there was no contact at all between the hopper and the ants; the association disappeared the day after, while probable, direct trophobiosis behaviour was not observed (Soltyspers.com).

4.3. Xenobiont Behaviour

Excluding the dubious record from Lichtenstein [46] reporting of mutilated and even lost wings in ant-attended Tettigometra, the old literature reports mostly tettigometrid species explicitly found in ant nests under stones with the ants but without mentioning any associated host plant or elaborated xenobiont behaviours. Food plant information appears in recent reports only because they have a crucial role in trophobisis, allowing the hoppers to feed and excrete honeydew.

While most are well-known predators, xenobionts are able to approach the planthopper, adult or nymph, without it escaping. They manage to collect or even intercept the honeydew stream produced by the planthopper after a kicking movement, or they can collect the drop of honeydew that the trophobiont produces directly at the anal tube opening, before any planthopper kicking. Elaborated specific behaviours generally appear during long-time trophobiotic relationships with ants. These behaviours provide protection from predators and parasitoids, transport when environmental conditions are unfavourable (trophophoresy [13], waste removal, and protection of eggs and nymphs [5]).

In Africa, Weaving [47] reported Nototettigometra (Nototettigometra) patruelis (Stål, 1855) sometimes traveling down the stem of its host plant and entering tunnels constructed by the ants around the roots. Similar behaviour for Tettigometra (Tettigometra) sulphurea Mulsant and Rey, 1855 was reported by Bourgoin [48], although the specimens never went deeply underground, staying a few centimeters beneath the surface. We report here new observations about trophophoresy and active guiding of tettigometrids by ants (Appendix A). Trophophoresy is well known in Hemiptera: Coccoidea [49], but it remains exceptionally reported with planthoppers, although it was also observed once with nymphs of the meenoplid Phaconeura sp. [21].

Antennal palpations by ants, most often on the side of the trophobiont abdomen and on the forewings, are commonly reported with tettigometrids. However, in those cases they are not necessarily followed by the production of a drop of honeydew [48]. In Cameroon, the delphacid Peregrinus maidis (Ashmead, 1890) has been reported having trophobiotic associations with 16 different ant species, several of them having enclosed the trophobiont into natural shelters or built shelters with vegetal fibres around them [16,18,19,20].

In Blattodea, Naskrecki and Nishida [14] reported physical contacts with various species of Fulgoridae that occur also quite often, either with the cockroach palpating the planthopper’s wings with maxillary palps, or even by resting its front legs on the wings. However, they noted that this contact did not appear to elicit any specific response from the planthopper. In Lepidoptera, the same authors also reported the moth Elaeognatha argyritis Hampson, 1905 (Nolidae) tapping the wings of Enchophora sanguinea Distant, 1887 with its antennae, which resulted in an immediate production of a honeydew drop. With Gastropoda, they reported Pittieria aurantiaca extending the anterior portion of their foot, forming a wide hood above the planthopper’s anal tube, allowing the snail to intercept and catch the stream of honeydew droplets, but there was never a physical contact between the symbionts [14]. They also observed several times Camponotus sp. ants climbing the head of the snail and licking honeydew from its body while the snail was attending Enchophora sanguinea and Phrictus quinquepartitus Distant, 1833. They described this extreme case as kleptotrophobiosis. Finally, and although this observation bears repeating, it should be mentioned the particular behaviour of a gecko tapping the bark of the substrate with its head, as a stimulus [24]—perhaps vibrational—inducing the production of honeydew by the planthopper. The different elaborated behaviours adopted by the xenobionts are summarized in

Table 3. Interaction diversity of xenobionts behaviours according to their taxonomic group and related to the size of the trophobiotic planthopper.

Trophobionts		Large-sized planthoppers (mostly Fulgoridae)				Small-sized planthoppers	Tettigometridae, some Cixiidae, Delphacidae, …
Xenobionts		Gekkonids	Gastropods	Lepidoptera (Moths)	Cockroaches and ants	Ants, rarely cockroaches	Ants
Xenobiontelaborated behaviours	Guidance Trophoresy	no	no	no	no	no	Eggs, nymphs and adults trophophoresy. Nymphs and adults guidance
	Active protection	no	no	no	ants: ?	Active protection against predators and parasitoids	Active protection against predators and parasitoids
	Honeydew collecting	Collect of honeydew drops kicked by the planthopper	Collect of honeydew drops kicked by the planthopper or interception of honeydew stream	Direct collect of honeydew drops at the anal opening. No kickingby the planthopper	Collect of kicked honeydew drops or direct collect at the anal opening without kicking by the planthopper	Direct collect of honeydew drops at the anal opening. No kicking by the planthopper	Direct collect of honeydew drops at the anal opening. No kicking by the planthopper
	Physical contacts	No physical contact with the trophobiont Tappering bark (?)	No physical contact with the trophobiont	Antennal palpations on trophobiont wings	Maxillary palps palpations of wings; Prolegs resting on wings	Antennal and palp palpations	Antennal and palp palpations, often of lateral sides of the abdomen
	Duration	Short-time opto- trophobiosis	Short-time opto- trophobiosis	Short-time opto- trophobiosis	Short-time optotrophobiosis, kleptotro-phobiosis	Short-time opto- trophobiosis	Crypto- and optobiotic long time-period trophobiosis

.

4.4. Planthopper Trophobionts

Trophobiosis is now reported from 13 planthopper families (Appendix A), and all cases reported in planthoppers should be classified as facultative trophobiosis. A few exceptions might be considered as permanent (but not obligatory) mutualistic behaviour: the monospecific family Hypochthonellidae [50], the hypogeic delphacid Notuchus kaori Hoch and Asche, 2006 [25] in New Caledonia, and perhaps the meenoplid taxon Phaconeura sp. in Australia [21], the last two found with Paratrechina sp. (Formicidae). These species exhibit morphological adaptations to hypogeic habitats (interstitial or deeper in caves habitats) that are difficult to separate from possible myrmecophilous morphological adaptations, but that might at least favor attendance by ants.

Underground attendance for long period in ant nests (cryptobiotic) is documented in members of Cixiidae, Delphacidae, Hypochthonellidae and Tettigometridae [4,6,48] and also recently reported in Flatidae: Budginmaya eulae Fletcher, 2009 [51] and for four undetermined flatid species [8]. Cryptobiotic attendance under aerial shelters built by ants is reported in Delphacidae only [4,5,16,18,19,20,26]. Optobiotic ant attendance is reported here for the first time in the families Dictyopharidae, Nogodinidae and Eurybrachidae and several new cases are reported for Caliscelidae, Ricaniidae and Issidae, taxa for which such behaviour remained exceptional (Appendix A).

In most optobiotic cases, trophobionts are adult planthoppers only, but in cryptobiotic situations probably also nymphs (Delphacidae, Tettigometridae, Flatidae). However, in a few observations only nymphs seem to act as trophobionts. Cixiidae nymphs of Reptalus sp. in Bulgaria for instance were regularly found in the soil nests of Aphaenogaster subterranea (Latreille, 1798) (Figure 3c,d) but adult individuals of Reptalus inhabiting and feeding on the above-ground parts of the plants were never observed attended by ants. Lörinczi [29] reported a similar observation with Reptalus panzeri (Löw, 1883) in Hungary, and it is probably also the case for the New Guinea flatid nymphs found in nests of the ant Pseudolasius Emery, 1887 [8]. In Madagascar, nymphs of Cixiidae—probably Oliarus sp.—were observed feeding on roots and attended by Pheidole sp. ants (Figure 4e). They were collected in a subterranean environment of the caves in the Tsingy of Namoroka Strict Nature Reserve, a limestone karst region in northwest Madagascar, but adults were not observed.

4.5. Ant Attendance in Tettigometridae

-

Tettigometrid behaviour as trophobiont.

Tettigometridae form a small group of planthoppers, for which major life-traits still remain very poorly known [52], but for which several cases of long-time attendance by ants were reported since the 19th and at the beginning of the 20th century. According to Silvestri [53], the first observations of ants with tettigometrids seems to be from Savi who mentioned the occurrence of a “petite cicadelle brunâtre” [small brownish leafhopper] in an ant nest, probably Tettigometra (Tettigometra) impressifrons Mulsant and Rey, 1855 and the xenobiont Tapinoma nigerrimum (Nylander, 1856). However, the earliest direct observation of trophobiosis behaviour recorded was published in France by Rouget [54] for Tettigometra laeta Herrich-Schäffer, 1835 with ants Tapinoma erraticum (Latreille, 1798). Successively new observations were quickly published [46,53,55,56,57,58,59,60,61,62,63,64,65,66,67]. These early publications were found in single published notes, concentrated in a short period some hundred years ago, and provided records for three Tettigometra species only (T. (Eurychila) bifoveolata Signoret, 1866, T. (Mitricephalus) macrocephala Fieber, 1865 and T. (Tettigometra) virescens Panzer, 1799). A careful study of this old literature allowed the retention of only about 25 original observations [4,48]. It is only recently that new observations were added, showing new interest in these behaviours, probably boosted by the development of macrophotography.

To date, 106 observations of tettigometrids attended by ants have now been published or attested by macrophotographic images published on the web. They concern 18 tettigometrid species ant-attended, representing some 20% of the family species. All major suprageneric tettigometrid taxa are concerned except the Phalixinae Ghauri, 1964, Plesiometrini Bourgoin, 2018 and the Cyranometrini Bourgoin, 1987 [68]. We provide and document here several new data on ant attendance in Tettigometridae (Appendix A, Figure 24, Figure 25, Figure 26, Figure 27, Figure 28, Figure 29, Figure 30, Figure 31, Figure 32, Figure 33 and Figure 34), including more specific field observations on twelve European species: Tettigometra (Tettigometra) atra Hagenbach, 1825 in Malta and Bulgaria (Figure 28a), T. (Tettigometra) brachycephala Fieber, 1865 in Croatia (Figure 28b), T. (Tettigometra) fusca Fieber, 1865 in Ukraine (Figure 28c and Figure 29a), T. (Tettigometra) impressifrons Mulsant and Rey, 1855 in Italy (Figure 29b,c), T. (Tettigometra) laeta Herrich-Schäffer, 1834 in Bulgaria (Figure 30a) and Albania (Figure 30b), T. (Tettigometra) picta Fieber, 1865 in Malta (Figure 30c), T. (Tettigometra) sulphurea Mulsant and Rey, 1855 in Bulgaria (Figure 31a), Greece (Figure 31b), including Lesbos Island (Figure 32a), Croatia (Figure 31c) and Serbia (Figure 32b), T. (Eurychila) bifoveolata Signoret, 1866 in France (Figure 32c), T. (Mitricephalus) griseola Fieber, 1864 in Italy (Figure 33a), T. (Mitricephalus) leucophaea (Preyssler, 1792) in Bulgaria (Figure 33b,c), T. (Mitricephalus) macrocephala Fieber, 1864 in Bulgaria (Figure 34a,d and Figure 35a) and France (Figure 34b,c), T. (Metroplaca) longicornis Signoret, 1865 in Bulgaria (Figure 35b) and France (Figure 35c).

Trophobiosis is reported for the first time for T. griseola and T. picta. The Mediterranean species Tettigometra picta (nymphs and adults) was found to be tended by Tapinoma nigerrimum (s.l.) on all studied host plants (about 20) of Lavatera arborea L. (Malvaceae). Besides aggregation of individuals on host plants, transportation by ants was also observed.

Of all the Tettigometra species we observed with ants, three were found in early spring—T. atra, T. laeta and T. (M.) longicornis; the earliest one was T. atra in March (Figure 28a). All of them were observed in ant nests, located under stones, although T. (M.) longicornis (Figure 35c) was also found during summer on the host plant Eryngium campestre L. (Apiales, Apiaceae). The presence of adults during the spring months suggests that they have most likely overwintered in the ant nests.

Ant attendance was observed on different parts of plants, sometimes depending on their growth. For instance, in the beginning of April when the plant leaves are in an early stage of development and they form a rosette close to the ground, the adults of T. (T.) sulphurea were amongst the leaves or on the ground, under the leaves. Other Tettigometra species mentioned in this study were observed over an extended period of time—from April to September.

-

Xenobiont preferences in tettigometrids.

The current study increases the known range of ant species partnering in trophobiotic interactions with four further Tettigometra species. Comparing the study results with literature data, some relationships between Tettigometra species and ant species are documented for the first time: T. atra in nests of Tapinoma simrothi Krausse, 1911 and Lasius platythorax Seifert, 1991, T. macrocephala in trophobiotic interactions with Formica cinerea Mayr, 1853 and Camponotus aethiops (Latreille, 1798), T. longicornis in soil nests of Crematogaster sordidula (Nylander, 1849), T. sulphurea tended by Tetramorium chefketi Forel, 1911, Camponotus piceus (Leach, 1825), Formica pratensis Retzius, 1783, Lasius niger (Linnaeus, 1758) and L. paralienus Seifert, 1992.

However, trophobiotic relationships appear non-specific, and one tettigometrid species can be tended by different ant species. In our newly reported observations, six ant genera were involved: Tapinoma (Dolichoderinae), Formica, Lasius, and Camponotus (Formicinae), Tetramorium and Crematogaster (Myrmicinae). Interestingly, some plurispecific situations were found: T. laeta and T. atra (Figure 30a, Appendix A) together tended by Tapinoma erraticum in an ant nest under a stone, and T. impressifrons (Figure 29b,c) and T. griseola (Figure 33a) located together on the host plant Cakile maritima Scop. (Brassicales, Brassicaceae), both also tended by Tapinoma erraticum. In reverse, literature and new observations show that Reptalus (Cixiidae) seems closely associated only with Aphaenogaster subterranea.

-

Aggregations and ethological notes in tettigometrids.

The formation of groups of individuals is one of the most characteristic features of the ant attendance in tettigometrids [4]. The lower ones in number were usually those that were found under stones (usually not more than five individuals).

In T. (T.) macrocephala, nymphs and adults form small groups (Figure 34a–c and Figure 35b) (up to 15 specimens) on the host plant Eryngium campestre around the nodes, leaf veins, and branches of the plants while entrances to the ant nest are situated at the base of the plant. Often, nymphs and adults from the hidden entrances in the ant nest go up on the plant in the evening. In case of unfavourable conditions, the planthoppers were herded back into the underground area of the ant nest. During autumn, several adults are often found on the plant base, close to entrances of ant nests.

Most frequent cases observed were with T. (T.) sulphurea (Figure 31 and Figure 32a,b), always on Onopordum acanthium L. (Asterales, Asteraceae), visited by five different ant species. Females lay their eggs in large groups close to each other at the base of plants. Eggs are pale yellow with an elongated oval form (Figure 31b). In their interaction with Tetramorium chefketi, we observed the building of a shelter around the base of plants, where the female planthopper inserts the eggs and stays almost all the time. In other observations, T. (T.) sulphurea females are most often found near the entrances of the ant nests on the basal parts of the host plant or on the ground surface. Young nymphs may live in dense large groups close to the base of plants. Older nymphs and adults may climb higher on the plant, but always stay near the base of the leaves. In addition, some nymphs were hiding in holes of unknown origin. For T. (M.) longicornis in France, small groups of about 10 individuals of nymphs and adults were found on the host plant Eryngium campestre. In Bulgaria, several adults were also found in a soil nest of Crematogaster sordidula in the early spring.

Unlike all other tettigometrid observations, only single adults and nymphs standing on the leaves of the plant Ammophila arenaria (L.) (Poales, Poaceae) were always observed for T. leucophaea (Figure 33b,c). In only two species (T. atra and T. laeta), direct trophobiotic relationships was not observed, probably due to their specific habitat (nests under stones), but when being disturbed, they used to go alone or were transported by ants inside the nest. In all other cases, ant-antennating, transport of individuals, guiding of adults, and direct collection of honeydew from tettigometrids anal openings were observed.

4.6. Cockroach-Attendance in Fulgoridae

We report trophobiosis involving cockroaches in six planthopper families. This symbiotic relation appears as a new major life-trait pattern of Hemiptera planthoppers behaviour, unknown until now. According to the collected observations, when medium and large planthoppers species are involved, cockroach xenobionts enter more often than ants in trophobiose. Interestingly, 80% of the cases involve the Fulgoridae. At this stage, we cannot say whether this observation corresponds to a particular behaviour of the Fulgoridae, due to the fact that they are generally larger in size than the other families of planthoppers, and therefore more attractive in terms of the quantity of food excreted, or whether this is a simple observation bias (species more easily observable, more collected, etc.). However, it should be noted that these large planthoppers are often observed on the same trees over months, showing slight aggregating behaviour that favours the deposit of kicked honeydew drops on the surrounding substrate and that subsequently might attract attending cockroaches.

5. Conclusions

The new observations collected and reported here put into focus the unexpected importance of trophobiosis as a more general and important life trait of fulgoromorph planthoppers in the interspecific relationships they involve. Trophobiosis not only concerns specifically ant-attended tettigometrids, but also cockroach attended fulgorids, and it appears as a potential behavioural trait probably general for most planthoppers. These observations raise several interesting questions that still need more investigations before starting to answer them.

-

Trophobiosis is an interspecific interaction between a trophobiont and a xenobiont. In tettigometrids and probably most planthoppers (except perhaps for Hypochthonella caeca and Notuchus kaori for which we have too little data), these interactions are non-permanent. Several factors need to be considered for such a behaviour taking place. Feeding and honeydew production is a regular physiological activity for the planthopper. However, if this criterion is necessary, it is not sufficient as this is general for all phytophagous Hemiptera but not all of them develop trophobiotic relationships. A priori, the interaction starter is given by the xenobiont that stops its other activities, stays around, avoids making the trophobiont escape, and does not predate it. The interaction could be maintained all the time the trophobiont regularly excretes honeydew and even longer in cryptobiotic conditions. Honeydew, which is rich in sugars and amino acids, constitutes a nutritional resource for xenobionts and acts as a volatile and a contact kairomone, stimulating some particular behaviours of predators and parasitoids: searching, preys/hosts locations and attacks [7]. As these behaviours are induced, might trophobiosis be therefore initiated by the honeydew production, alerting the xenobionts around?

-

With ants, such interactions seem facilitated by sessility and subsocial traits of the tettigometrid trophobionts that guarantee higher benefits for the xenobiont: easier and abundant source of feeding at a smaller cost [4]. However, all recent new observations of larger fulgorid species show that if sessility and sub-social behaviours are facilitating factors (several of the large fulgorids are often observed in small groups of individuals on the same tree), how also is the amount of food produced an important criterion?

-

Absence of trophobiosis interactions reported with Achilidae, which are often found grouped (few adults and nymphs at different instars) under tree bark, but which are though to feed from fungal hyphae [43], also raise the issue of honeydew quality as an additional criterion. Do Achilids produce honeydew which is less nutrient-rich than other sap-sucking planthopper taxa and perhaps without semiochemical kairomone stimulating xenobionts to start any trophobiotic behaviour? Our current knowledge about achilid natural history does not allow any better hypothesis.

-

Another aspect of planthopper-ant interactions is that the starter (or their maintenance) could also, at least in the case of the tettigometrids, be an active action of the trophobiont itself. Indeed, the old observations, confirmed by our recent ones, show that tettigometrids are often found in ant nests, under stones, probably overwintering for some species, but without food plants mentioned around, neither roots. In these cases, the regular production of honeydew should be clearly reduced or even stopped for a while. How then are interactions with ants maintained, without the ants starting to predate the hoppers? Bourgoin [4] has mentioned antennal palpations by ants that might encourage the honeydew drop production (as observed with non-ant xenobionts), are also located in the body parts that are rich in pore tegumentary glands [69]. The hypothesis that they might use allomones to chemically manage ant behaviour to their benefit cannot be discarded. It is probable that trophobiosis and mutualistic interactions with ants in tettigometrids are ruled by both allomones and kairomones and are much more widespread than thought until now, and occur with most of the species of the family.

-

Trophobiosis, which seems very common although facultative in tettigometrids, also raises other interesting questions which are difficult to answer at this time: how far is the behaviour general in the family? Why have tettigometrids not yet been observed attended by cockroaches? Are they kept away by ants? How important are ants in maintaining tettigometrid populations? Indeed, there is a demonstrated tendency in some species such as Nototettigometra (N.) patruelis, Tettigometra (M.) fusca, T. impressopunctata Dufour, 1846 or T. (M.) leucophaea of drastic and sudden downsizing of their populations, up to totally disappearing from places where they were previously very commonly known [47,70,71]. There is suspicion of cryptic species in the family [52]: would precise observations of such sister or closely related species, allow a smart comparative analysis to better understand how trophobiosis takes place and is maintained over time?

-

In large-sized fulgorids, xenobionts belong to other insect orders than ants, such as cockroaches and moths, and also to mollusca and even small gekkonid vertebrates. Obviously, if ant attendance seems relatively common with tettigometrids, xenobiontic Blattodea appear as the most common cases observed after ant attendances, and the main one for the Fulgoridae. What behaviours are occurring at night with these nocturnal xenobionts? How tiny is the frontier between trophophosis and predation for all predator xenobionts?

-

Ants, currently dated from the Early Cretaceous from around 110 Mya [72,73] to 140 Mya [2] represent a much younger insect lineage than cockroaches, which dates back to the Permian [74], more than 100 My older. If plants (including fungi) represent the main framework for the interspecific interactions of planthoppers, which are obligate phytophagous insects, the role of these mutualistic behaviours in food webs remains to be further analyzed. For instance, Compton and Robertson [75] have shown a net beneficial effect of the association of the tettigometrid N. (N.) patruelis and its associated ants by protecting fig seeds and pollinators despite the phytophagous action of the planthopper on fig trees. During the evolution of fulgoromorphs, to what extent have these interactions, and in particular a new step in the mode of trophobiosis from cockroach xenobionts to ant xenobionts, also supported diversification of planthoppers, as previously suggested by Grimaldi and Engel [76] and more generally for Hemiptera and their associations with ants?

To answer all these questions, it is clear that beside the few primary biodiversity data related about which is present, where and when (taxonomical data and their associated biological data), we just know almost nothing about life traits in planthoppers! Many more observations are yet needed before we could start investigating all these questions, and how, in the evolutionary framework of the phylogeny, they could help us to better understand how important trophobiosis has been for the evolution of planthoppers.