(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



The most polyphagous insect herbivore? Host plant associations of the Meadow spittlebug, Philaenus spumarius (L.) [1]

['Vinton Thompson', 'Division Of Invertebrate Zoology', 'American Museum Of Natural History', 'New York', 'United States Of America', 'Claire Harkin', 'School Of Life Sciences', 'University Of Sussex', 'Falmer', 'Brighton']

Date: 2023-10

A comprehensive list of all known host plant species utilised by the Meadow Spittlebug (Philaenus spumarius (L.)) is presented, compiled from published and unpublished sources. P. spumarius feeds on 1311 host plants in 631 genera and 117 families. This appears, by a large margin, to be the greatest number of host species exploited by any herbivorous insect. The Asteraceae (222 species) and Rosaceae (110) together account for 25% of all host species. The Fabaceae (76) and Poaceae (73), are nearly tied for third and fourth place and these four families, combined with the Lamiaceae (62), Apiaceae (50), Brassicaceae (43) and Caprifoliaceae (34), comprise about half of all host species. Hosts are concentrated among herbaceous dicots but range from ferns and grasses to shrubs and trees. Philaenus spumarius is an “extreme polyphage”, which appears to have evolved from a monophage ancestor in the past 3.7 to 7.9 million years. It is also the primary European vector of the emerging plant pathogen Xylella fastidiosa. Its vast host range suggests that it has the potential to spread X. fastidiosa among multiple hosts in any environment in which both the spittlebug and bacterium are present. Fully 47.9% of all known hosts were recorded in the Xylella-inspired BRIGIT citizen science P. spumarius host survey, including 358 hosts new to the documentary record, 27.3% of the 1311 total. This is a strong demonstration of the power of organized amateur observers to contribute to scientific knowledge.

Funding: The BRIGIT project was funded by UK Research and Innovation ( https://www.ukri.org/ ) through the Strategic Priorities Fund, by a grant from the Biotechnology and Biological Sciences Research Council ( https://www.ukri.org/councils/bbsrc/ ), with support from the UK Department for Environment, Food and Rural Affairs and the Scottish Government to A.J.A.S and C.H. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Information on P. spumarius host plants is scattered across many languages and continents, from the primary scientific literature to amateur entomology publications, agricultural tracts and unpublished sources. Schmidt [ 25 , 26 ], apparently an amateur, published the earliest extensive host lists in 1914, totaling 137 species, in a German civic booster journal promoting the progress of science and industry. The next major contributions came from Hawaii [ 27 , 28 ], following the introduction and proliferation of P. spumarius on Big Island in the 1940s. This work was followed shortly by the 1950 publication of 97 California hosts observed by DeLong and Severin [ 29 ] in the course of the first experiments demonstrating that P. spumarius and other spittlebugs can transmit what is now known to be X. fastidiosa. A series of host lists in American PhD theses [ 30 – 33 ] culminated in 1954 in the 383 species compilation of Weaver & King [ 5 ]. Metcalf’s encyclopedic catalog of the Aphrophoridae [ 34 ] includes annotated references to host information in P. spumarius publications through 1955, a very useful source of otherwise obscure observations. In 1965 Noury [ 35 ] compiled a list of about 167 hosts from France and Middle and Northern Europe. Halkka et al. [ 36 ] provide a list of 165 hosts based on 1960s field observations in Finland. A 1993 PhD thesis by Booth [ 37 ] includes records of 90 hosts in natural areas in Wales and New Zealand. Jennifer Owen [ 38 ] and Denis Owen [ 39 ] record observing 143 host species in 47 plant families in their late twentieth century English suburban garden, detailed in notes that we recovered from J. Owen’s papers in the Leicester Museum & Art Gallery. More recently, concern over X. fastidiosa has stimulated a golden age of P. spumarius host studies in the Mediterranean region (references below). We have collected, evaluated and distilled these sources and all others available to us into a single, readily accessible summary table in searchable format.

Philaenus spumarius first became a major subject of study in the late 1940s and early 1950s in the United States and Canada for its role as a serious non-native pest of alfalfa (Medicago sativa) and other forage legumes [ 5 ]. From about 1960 to 2010 it was primarily of interest to entomologists and ecologists working on insect population dynamics and energetics [ 13 – 16 ] and to researchers interested in its remarkable colour polymorphism [ 17 – 20 ]. More recently, it has attracted considerable attention from applied entomologists and pest management scientists, after it was shown to be the primary European vector of X. fastidiosa, a pathogen that causes Olive Quick Decline Syndrome (OQDS), which has devastated olive groves in the Apulia region of Italy [ 1 , 21 ]. This threat, along with the recent association of P. spumarius with Almond leaf scorch disease (ALSD) in the Alicante and Balearic Islands regions of Spain [ 22 , 23 ], has generated intense research activity focused on the meadow spittlebug, reflected in an exponential recent rise in publications and citations (175 papers and 2,699 citations since 2010; Web of Science, accessed 21 May 2023). Xylella fastidiosa was introduced from the New World to Europe relatively recently [ 23 ]. It has now been detected in over 500 plant species, including crops, ornamentals and trees, across many plant families [ 24 ]. To predict which agricultural and natural plant species and ecosystems are at risk, it is crucial to understand the host range of the vector.

The present work has multiple aims. First, we bring together disparate sources in a publicly available, comprehensive, documented compilation of P. spumarius hosts. This will be of use to investigators studying the spread of X. fastidiosa and broader biological phenomena, such as the evolution of xylem feeding insects and the evolution of extreme polyphagy. Second, we analyse the broad patterns of P. spumarius host exploitation, place it in the context of other highly polyphagous insects, and examine the implications of host patterns for X. fastidiosa spread. Third, we suggest some useful guidelines for future surveys of spittlebug host plants. Finally, we assess the efficacy of the BRIGIT project in corroborating existing host records and establishing new ones. We believe this is the first time a citizen science project of this scope has been tested against the historical record.

a) Frothy spittle mass on one of its common host plants, the legume Lotus corniculatus. bird’s-foot trefoil. The nymph is hidden within. b) Spittle foam removed to expose nymph. The dark red dot is its right eye. Note the swollen sucking pump anterior to the eyes. c) Adult P. spumarius. This is the most common of several colour forms. Photographs by Claire Harkin.

The Meadow spittlebug Philaenus spumarius (L.) (Hemiptera: Aphrophoridae) is one of the world’s most widespread and abundant insects ( Fig 1 ). It is also the major European vector of Xylella fastidiosa Wells et al., an emerging bacterial plant pathogen that threatens crops as diverse as grapes, almonds, citrus and olives [ 1 – 4 ]. Despite its importance, the last comprehensive review of P. spumarius host plants is almost 70 years old [ 5 ]. Concern with X. fastidiosa has led to a proliferation of recent studies adding new P. spumarius hosts (e.g. [ 6 – 11 ]), including the BRIGIT citizen scientist initiative in Britain that enlisted amateurs to identify P. spumarius host plants [ 12 ].

The geographic units chosen for this report include several with natural boundaries. New Zealand (NZ) and Hawaii (HI) represent discrete island areas where P. spumarius has been introduced. The Azores (AZ), where P. spumarius is likely but not certainly recently introduced [ 51 – 53 ] are a special island case, but one with very few host records. In North America, where P. spumarius is introduced, it has a disjunct distribution [ 54 ], divided by the dryer midsection of the continent into discrete eastern (ENA) and western areas (WNA). Britain and Ireland (B&I) represent another natural division. We somewhat arbitrarily divide continental Europe into four areas: Finland and Scandinavia (F&S), Western Europe (WE), Eastern Europe (EE), and the Mediterranean Basin (MED). See the footnotes to Table 1 for the detailed boundaries. A number of Noury’s [ 35 ] listings are from a source covering “Middle and Northern Europe”, a category that does not fit our arbitrary divisions. These records are recorded as M&NE. A few North American references do not specify which section of the continent the records refer to. These are recorded as EorWNA. Although P. spumarius is frequent in Kyrgyzstan [ 55 ] and is reported across Asia to China and Japan (refs. in [ 34 ]), we found only one host record east of European Russia, in Uzbekistan [ 56 ]. P. spumarius has been reported once, on strawberries, on the island of Réunion in the Indian Ocean [ 57 ] but, in the absence of later reports, seems not to have become established.

In cases in which we have multiple records for the same host species, we have given preference in the following order: formal scientific publications (journal articles, books, stand-alone scientific publications), followed by theses and dissertations, followed by more ephemeral Internet sources, followed by unpublished observations by cited observers. Where we have records for multiple geographic areas, we list all areas and cite at least one record for each. For hosts with records for both nymphs and adults, we cite at least one record for each stage. The goal is to keep the list of references for each host species compact, while documenting the known geographical distribution and observed stages for each host. The 358 citizen science host observations new to science are marked ●●●, while those new to the UK are marked ●●, and those confirming earlier local observations are marked ●.

We have included all naturally occurring hosts for which we have found at least one usable record. We exclude records based solely on laboratory experiments. Most records are reported to species level. In instances in which we encountered a mix of records that were generic-only and species-specific for a given genus, we include only species-specific records, to avoid the possibility of double counting. This makes our host number estimates more conservative, as, in all likelihood, some of the excluded generic records represent species distinct from but congeneric to those included. Holopainen & Varis [ 49 ] report that using this criterion decreased their host total for Lygus rugulipennis from 437 to 402, a reduction of 8%. Application of the same criteria to underlying data for the scales Aspidiotis nerii and Hemiberlesia lataninae [ 50 ] reduced the number of species records by about 12.5% in each case. As noted above, we also exclude records of adult hosts based on single specimens.

Botanical nomenclature poses additional hurdles. Some of our sources are over 100 years old and plant taxonomy has moved on. Wherever possible, we have updated species names to conform with current usage. Where we encountered nomenclatural ambiguity, we used two sources to determine currently valid names: the Integrated Taxonomic Information System [ 46 ] and Kew Plants of the World Online [ 47 ]. We have updated plant family-level taxonomy to be consistent with contemporary usage, following the template of Christenhusz & Byng [ 48 ]. This pared down the number of families by about 9%. In many cases, particularly those involving agricultural or popular sources, we had to make reasonable inferences from common names in multiple languages, an endeavor with its own hazards. We are confident that most of our botanical names are correct and up to date, but in a work of this scale, from such diverse sources, some errors are inevitable. We also recognize the possibility that some of the original plant identifications may have been in error, representing another, hopefully small, source of noise in the data.

Validation of BRIGIT citizen science host plant records adopted a highly conservative approach: 1) all records based on adult P. spumarius were rejected due to the difficulty in distinguishing between sitting and feeding behaviours, as previously described; 2) nymph host records were accepted only where P. spumarius identification could be confirmed via accompanying photographs; 3) host records based on the presence of spittle were attributed to P. spumarius only when the reported host was an herbaceous dicot.

The citizen science survey benefitted from the specific character of the British spittlebug fauna. In many other areas there are common spittlebug species frequenting herbaceous dicots that might easily be misidentified for P. spumarius based on observation of spittles alone. This is less the case in Britain, but two British genera, Neophilaenus and Aphrophora, do include spittlebugs that might be mistaken in the spittle stage for P. spumarius if the nymphs themselves are not examined carefully. The four British Neophilaenus species are restricted almost entirely to monocots and can be screened out by eliminating ambiguous records from grasses and sedges. The four Aphrophora species are primarily insects of trees and shrubs, but the nymphs of the most common and widespread species, Aphrophora alni Fallén, also occasionally feed on herbaceous plants. Thus, although the nymphs are distinct, identification based solely on spittles runs the risk of conflating P. spumarius and A. alni. In practice, however, extensive field collections across the UK undertaken during the BRIGIT project found very few A. alni on herbaceous plants; consequently, this risk is judged to be minimal.

We distinguish among records for nymphal hosts, adult hosts and plants that host both stages. Nymphal host records are unambiguous because nymphs ( Fig 1b ) cannot produce spittles ( Fig 1a ) without feeding. Adult records are more problematic. Adult P. spumarius ( Fig 1c ) are active insects and move about from plant to plant. Unless accompanied by direct observation of feeding in the form of regularly expelled droplets of excreta, adult records may represent “sitting” rather than feeding individuals. This produces inevitable “noise” in adult host records. To minimize this noise, we have omitted instances in which adult host records are clearly unreliable, including those based on single collected specimens. Nevertheless, some of our adult host records are likely in error, or, at a minimum, insecurely documented. On the other hand, there is no doubt that adults do feed on a wide variety of plants. Aside from eggs, P. spumarius does not have an inactive or dormant stage. They must feed to live, and feed in quantity. When there are a lot of adults on a particular plant or when adults associate with multiple individuals of the same plant species, chances are they are feeding. This is particularly true of the many adults found on trees and shrubs in dry climate summers, when plants in the herbaceous understory have browned out for the season [ 22 , 44 , 45 ].

More disconcertingly, and for no obvious reason, Weaver and King omit 39 hosts out of 85 listed in Schmidt’s second 1914 paper [ 25 ] and misstate the date of that publication as 1915. Given these anomalies and inconsistencies, we have included only hosts that could be corroborated in the original cited works and incorporate, where appropriate, hosts listed in the original sources that Weaver and King omitted. We include the host records Weaver and King report as their own observations.

Weaver and King also inexplicably omit about two dozen nymphal hosts listed in Teller [ 32 ], while incorporating 33 others. Likewise, their treatment of Marshall [ 33 ] is problematic. They include most of his records, despite the fact that he does not specify life stage, but attribute one host record to him that is not in his work (Rumex acetosa). They also attribute two of Marshall’s records (Pinus strobus and Rumex occidentalis) to Krauss [ 42 ], making it appear that they came from Hawaii rather than New York State. In another oddity, they misattribute a work by Davis [ 27 ] to the authors of a preceding work by Davis & Mitchell [ 28 ] and get the title of the paper by Davis wrong. These mistakes have the look of clerical errors. In the case of Licent [ 43 ], they include three of the hosts listed, but omit three others, perhaps a case of oversight, since the three unlisted hosts are in a different section of a very long treatise.

Prior to this work, Weaver & King’s 1954 compilation [ 5 ] has been the primary source of P. spumarius host records. It originated as a shorter host list in a preceding thesis by King [ 30 ]. Ostensibly, it is an amalgam of 26 cited sources, combined with personal observations in Ohio by Weaver and King themselves. We have tracked down and examined all of the original sources cited. This revealed several anomalies. For example, Weaver and King state that all the records in their compilation are for nymphal hosts and go on to say that they omitted the long 1950 California list compiled by DeLong & Severin [ 29 ] because those authors did not distinguish between nymphal and adult hosts. Nevertheless, they still cite four of the 97 hosts reported in that work, while omitting the rest. They list hosts from Osborn’s 1916 work [ 41 ] without reservation, though Osborn too omits information on whether his observations are for nymphs or adults. Unlike Osborn, DeLong & Severin provide dates and locales for their observations. All were made in Alameda County, on the sunnier, warmer side of San Francisco Bay, in April or the first three weeks of May, or in San Francisco, on the foggier, cooler side of the bay, in April, May or the first week of July. In these areas, for the periods in question, most or all P. spumarius individuals are still in the nymphal stage (VT observations), so it may be reasonably inferred that DeLong & Severin observed nymphs for all listed hosts.

The BRIGIT project [ 12 ] ran from 2019 to 2021 with the aim of improving surveillance and response capacity for X. fastidiosa should it be introduced into the UK. A key objective of the project was to develop a greater understanding of the distribution and host plant preferences of P. spumarius, identified as the primary vector of the bacterium in Europe [ 40 ]. Citizen scientists were encouraged to submit sightings via a national website for natural history observations or a bespoke portal, supported by identification aids on the BRIGIT website.

This work is based on the published literature, both formal and informal, host records associated with museum specimens, personal observations, private communications from colleagues, and the BRIGIT citizen science effort carried out in Britain. The first four sources are encompassed in an unpublished world database of spittlebug host plants built and maintained by VT. Ironically, it contains most published host records for most spittlebugs (Hemiptera: Cercopoidea) except for P. spumarius, which was initially exempted from full incorporation because the number of host records was enormous and because there seemed to be little point in recording the hosts of an insect that appeared to feed on almost any available plant. That changed with increasing concern about the role of P. spumarius as a X. fastidiosa vector. Thousands of P. spumarius records have been added, increasing the database from about 5,000 to 9,000 records over the past two years. Even with these additions it is not comprehensive for P. spumarius, missing for example, all of the BRIGIT citizen science records reported here, which are in a separate database maintained by CH and AJAS. The hosts presented here include all known to us as of May 2023.

A large majority of host records, 1113 (84.9%), are for nymphs only. Eighty-eight (6.7%) are for nymphs and adults, and 6.6% (86) for adults alone (in 24 cases the life stage could not be determined or reasonably inferred from the information available). We include about 1890 geographical area records. Most hosts have been recorded from a single geographical area, but many are recorded from two or more. Multiple area hosts are typically widespread weedy herbs or common garden plants. Europe and North America account for most geographical records, 73.2% and 19.6% respectively, but Hawaii (4.1%) and New Zealand (2.8%) are well represented for their size. Britain and Ireland loom especially large, alone accounting for 37.2% of all records, in major part a reflection of the BRIGIT program. BRIGIT citizen science records include 358 hosts that are not duplicated in preexisting sources, a full 27.3% of all recorded hosts. Another 198 (15.1%) represent records that are new to Britain and Ireland. Seventy-two (5.9%) represent confirmation of hosts recorded for Britain and Ireland in preexisting sources. In total, BRIGIT citizen science records include 628 (47.9%) of the 1311 recorded P. spumarius hosts.

Also included are the number of genera for each host family, the percent of host species from that family among all host species, the cumulative percent of all host species going down the ranking, the number of species in each host family, and an index of the occurrence of host species by family weighted for family size. The weighted occurrence index is the number of host species in each family divided by the total number of species in each family and expressed as a percentage. It provides a crude measure of the relative prominence of each host family, taking into account the large differences in numbers of species per family.

The headline result is that at 1311 species P. spumarius has far more documented host plants than any other herbivorous insect ( Table 2 ). They include ferns, herbs, shrubs, vines and trees, annuals and perennials, grasses and forbs, plants of the tropics, subtropics, temperate and boreal zones, conifers–just about every imaginable kind of vascular plant except those living submerged in aquatic environments. This extraordinary species level diversity is reinforced in the higher order taxonomic diversity, 117 families and 631 genera. Table 3 summarizes the distribution of host species by family for all families represented by 10 or more species.

Discussion

Philaenus spumarius appears to be the most polyphagous insect herbivore We start by comparing P. spumarius with other insects in terms of the number of host species exploited. At 1311 species, P. spumarius has, to our knowledge, more documented hosts than any other herbivorous insect. For comparison, Table 2 lists some of the serious contenders. Following Normark and Johnson [209], we limit comparisons to insects that feed directly on plant tissues. This excludes organisms like bees and syrphid flies that feed on pollen and nectar, flies that feed on rotting fruit, and leafcutter ants that attack plants but actually eat fungi that they cultivate on the plant material. None of the comparable insect species for which we have found data approach P. spumarius in number of host species. Hyphantria cunea (Drury), the fall webworm caterpillar weighs in closest, with a bit less than half the P. spumarius host numbers, though the webworm data include an unspecified number of artificial feeding tests [201], which we omitted in our compilation for P. spumarius. The closest arthropod competitor we have found is an arachnid, the red spider mite, Tetranychus urticae Koch, which is said to have more than 1100 hosts in over 140 plant families [210]. This puts T. urticae at the same order of magnitude as P. spumarius in host species number and substantially greater in host family number, the latter probably due to higher representation of plant families confined to the tropics.

Why is P. spumarius so polyphagous? Our results substantiate Ossiannilsson’s undocumented 1981 assertion that P. spumarius has more than 1000 hosts [211], an informed guess that had become embedded in the literature (cf. [4, 54, 212]), despite a lack of supporting evidence. A 1977 statement by Halkka & Mikkola [213] that there are “nearly 4000 recorded food-plant species” is clearly a typographical error. The documented 1311 species in 117 families put P. spumarius squarely in the category of “extreme polyphage”, defined by Normark & Johnson [209] as species that feed across more than 20 plant families. Like many extreme polyphages, it is a geographically widespread and invasive pest species, with very high population sizes. However, it does not exhibit other characteristics that Normark & Johnson [209] associate with extremely polyphagous insects, such as flightless females, larval dispersal, parthenogenesis or partiality to woody plants. What characteristics have, in fact, contributed to the extraordinarily broad host range? Two factors are probably paramount. The first is xylem sap feeding, a nutritional mode that permits access to a food source that is similar across a wide range of host plants and not chemically defended (refs. in [214, 241]). Xylem feeding apparently permits P. spumarius to feed on almost any plant it can penetrate with its mouth parts. The second factor is wide geographical range and the ability to thrive in climates from Hawaii, just south of the Tropic of Cancer [28], to within 65 km of the Arctic Circle in Finland [36]. Most or all extreme polyphages have cosmopolitan or invasive distributions [209]. Among xylem feeding insects, which include spittlebugs, cicadas and one subfamily of leafhoppers, P. spumarius is singular in its occupation of most of the Holarctic plus multiple distant islands. By that standard, other xylem feeders have been modest travelers. In addition to P. spumarius, four other spittlebug species have been introduced from Europe to North America [215], two others to Hawaii [216], and one other to New Zealand [217]. Two xylem feeding leafhoppers have been introduced from North America to Europe [218, 219], and three cicada species have hopped from New Zealand’s North Island to South Island ([220] & C. Simon, personal communication). None have achieved anything approaching the reach of P. spumarius. Perhaps not coincidentally, the other well-documented spittlebug extreme polyphage, Aphrophora alni (Table 2), is among the four other spittlebugs introduced from Europe to North America. It is not clear whether wide distribution is a cause or effect of extreme polyphagy. Each is clearly predisposed to promote the other [209]. Philaenus spumarius polyphagy seems to be a recent evolutionary development. It is one of a cluster of eight closely related Philaenus species living around the Mediterranean Basin [221]. Five are narrow monophages as nymphs, four feeding exclusively on the lily Asphodelus ramosus L. (or its close relatives A. aestevus or A. microcarpus) and one on Eryngium [222]. Two, P. spumarius and its very closely related sister species Philaenus tesselatus, are broad polyphages, though the extent of polyphagy is much less studied in P. tesselatus [223, 224]. The host status of the eighth species, Philaenus arslani, is uncertain. It has been collected from a modest variety of hosts, including three thistles, Cistus and “diverse shrubs” [150], all apparently but not explicitly adult hosts. Maryańska-Nadachowska et al. [221, 225] propose that the line leading to P. spumarius originated from an Asphodelus-feeding ancestor between 7.9 and 3.7 Mya. If so, P. spumarius broke out of the Mediterranean monophage pack and spread to an enormous variety of hosts in a relatively short geological time period, exhibiting what Normark & Johnson [209] describe as a “niche explosion”. Why? One answer might be the evolution of a more extensive arsenal of gene families involved in digestion, detoxification and transport of xenobiotics, as suggested for the red spider mite [210], the only arthropod we have found with a comparable host range, and for the green peach aphid [226] and corn earworm [205], among the runners up for most polyphagous insect herbivore (Table 2). On the other hand, the fact that xylem feeders encounter so few xenobiotics may make heroic detoxification capacity unnecessary. Ongoing work to sequence the complete P. spumarius genome [227] should provide data for a comparative analysis of the evolution of feeding versatility-related genes in relationship to niche explosion. Whether or not accompanied by extensive changes in the food assimilation related genome, the rapid evolution of the P. spumarius line from narrow monophagy to extreme polyphagy may have been facilitated by the feeding ecology of the adults. Asphodelus lilies die back in the Mediterranean summer dry season. Philaenus species dependent on Asphodelus as nymphs move to alternative hosts as adults [222], typically ectomycorrhizal trees and shrubs [228], a broadening of host range that may have set the stage for the evolution of polyphagy in the P. spumarius line. Although it has been stated that Mediterranean climate P. spumarius aestivate on these summer hosts [44, 222], there is no evidence for a state of summer torpor or hibernation ([69] & VT observations in California). The apparent evolution of an extreme polyphage from monophagic ancestors in a relatively short evolutionary interval is highly unusual. Polyphages are rare among herbivorous insects ([229, 230] and references therein), extreme polyphages even more so [209]. Had it been included in the most recent world survey of insect host plant breadth by Forister et al. [231], Philaenus spumarius would have been, in the most literal sense, off the charts. It seems to be one of a kind. It is also a clear counterexample to the suggestion that extreme polyphagy is an illusion based on multiple indistinguishable cryptic species feeding on different hosts [204, 226]. Extensive work on mitochondrial haplotype distribution in P. spumarius rules out multiple unrecognized cryptic species, although its mitochondrial lineages are bifurcated into two distinct clades [51, 52, 232] and one study suggested the presence of an unrecognized cryptic species in Anatolia and the Caucasus [233].

Patterns in host plant usage Given that P. spumarius seems able and willing to feed on almost any available host, what patterns in host usage can we discern? In sheer species numbers (Table 3) the Asteraceae (222) win hands down, with over twice as many hosts as the runner up Rosaceae (110), followed by the Fabaceae (76) and the Poaceae (73). The latter two high ranking groups merit special comment. Spittlebugs have a demonstrated affinity for nitrogen-fixing hosts, including many Fabaceae [214] (Fig 1a). In P. spumarius this is reflected in its pest status in legume forage crops in North America. Though it occurs on greater host species numbers in Asteraceae, it achieves highest densities on Fabaceae, up to 1280 nymphs/m2 in on M. sativa [14]. The large numbers of Poaceae hosts (71) are surprising in the other direction. It has long been recognized that P. spumarius favors herbaceous dicots and is relatively rare on grasses [36]. While grasses as a group are not preferred hosts, the present results demonstrate that there are relatively large numbers of grass host species, which contribute markedly to total host diversity, and it is clear that P. spumarius is sometimes locally common on grasses. Booth [37], for example, found P. spumarius plentiful on grasses at some open sites in New Zealand and shaded sites in Wales, while Lester et al. [234] found P. spumarius to be relatively common on grasses during the Scottish professional BRIGIT survey. At the high end of the host spectrum, P. spumarius is found not only on large numbers of Asteraceae species, but occurs in large numbers and high density on some individual species, including several Solidago spp. and a number of thistles. Among the Rosaceae, Filipendula ulmaria by itself accounted for 22% of 40,737 nymphal host records collected by Halkka and his colleagues in Finland [213]. This highlights a major limitation of species lists as a measure of host diversity. They count occurrence but not frequency, though local frequency is often recorded in the underlying sources. Another way to look at host attraction is to compare the ratio of P. spumarius hosts in a given family to the total number of species in that family. Table 3 includes a weighted occurrence index, the percentage of P. spumarius host species among all species in a plant family. This corrects, in a rough and ready way, for the fact that some plant families are small and some are enormous. Given that P. spumarius has an essentially temperate distribution, however, it should be noted that the index will be highly conservative for plant families that have a large proportion of species in the tropics. Among families with at least ten P. spumarius hosts, the index ranges from a low of 0.12 for Rubiaceae to a high of 7.19 for Betulaceae. In relation to total species numbers, P. spumarius occurs on a small proportion of Rubiaceae and a high proportion of Betulaceae. The Pinaceae (4.82), Caprifoliaceae (4.00), Rosaceae (3.42) and Onagraceae (3.05) are also high scoring. In contrast, two of the top three host families, Asteraceae (0.86) and Fabaceae (0.38), are knocked out of this competition by their enormous species numbers. The high rankings of two families comprised solely of trees and shrubs, Betulaceae and Pinaceae, might seem counterintuitive for an insect that clearly favors herbs, but both groups are ectomycorrhizal and their high scores are consistent with the general overrepresentation of this category among spittlebug hosts [228]. Plant morphology also clearly plays an important role in host plant selection. Early instar nymphs are especially attracted to plants with rosette form or other forms of growth that favor closely apposed leaf surfaces [5], no doubt because compact leafy clusters in close proximity to soil moisture form an advantageous early-instar nymphal microhabitat. Later instars tend to favor tall and robust perennial herbs [213]. On the other hand, plant features like abundant trichomes and lignification of tissues clearly deter P. spumarius feeding [90, 235]. When P. spumarius nymphs feed on woody plants it is invariably on new, unlignified growth, such as saplings, adventitious shoots of trees, or growing areas at the tips of branches ([5, 121] and our observations). Are there any otherwise apparently suitable plants on which P. spumarius does not feed? The most intriguing possibility is crownvetch, Coronilla varia L. (synonym Securigera varia (L.) Lassen), a Eurasian legume widely planted for forage and roadside erosion control in Eastern North America. Wheeler [236] reports that he found P. spumarius “in small numbers as adults only” on crownvetch but excludes it from his extensive list of arthropods collected on crownvetch in Pennsylvania. He adds that F.V. Grau, the founder of crownvetch studies in the USA, reported that he had never seen spittlebugs on this species in 28 years of work. We found no other records for crownvetch, a particularly unexpected result because C. varia is a nitrogen-fixing forage legume, the category of host on which P. spumarius otherwise reaches greatest densities in the USA. This suggests, subject to experimental verification, that there may be something exceptional about its biology that repels P. spumarius. If so, it might be a candidate species for understory plantings in orchards and groves where there is a desire to suppress P. spumarius vector populations [237]. It also appears that P. spumarius nymphs may not occur on Asclepias, species of which are notoriously well-defended chemically. Beirne [65] reports that P. spumarius nymphs do not feed on Asclepias species and the only P. spumarius Asclepias record we have found is for adults in Maryland (Table 1). There are, however, Asclepias nymphal records for two Lepyronia species [126, 238], demonstrating that this genus is not off-limits to all spittlebugs. Schmidt [25] says that he never observed spittles on Chenopodium or Atriplex. Chenopodium album, the species to which he is most likely referring, has been widely recorded as a P. spumarius host, both nymphal and adult, but Atriplex has only been observed as a host once, and only for adults (Table 1), suggesting that this genus may not be hospitable to nymphs. In general, our results sustain the early observations of Schmidt [25] and Fabre [83] that nymphs feed successfully on many plants that are chemically well-defended. Although P. spumarius has been recorded on several fern species (Table 1), it has not been recorded on bryophytes. Press and Whittaker [239] illustrate a spittle of the grass-feeding spittlebug Neophilaenus lineatus on a moss (Polytrichum commune), demonstrating that mosses are within the realm of plausible hosts. Notable plant categories on which P. spumarius records are rare in relationship to their numbers are Orchidaceae, Bromeliaceae, and CAM plants as a group. This is not surprising for orchids and bromeliads, the large majority of which are tropical epiphytes, putting them largely out of the geographical and ecological range of P. spumarius. It is more surprising for CAM plants, many in the Crassulaceae (eight species in Table 1), which are diverse and widely distributed in areas and habitats that P. spumarius frequents. CAM plants (which overlap to include many species in the Orchidaceae and Bromeliaceae) maintain close control of daytime transpiration, perhaps interfering with the accessibility of xylem sap.

Implications for Xylella fastidiosa management The most important lesson to be drawn from this review is that P. spumarius can and does feed on an extremely diverse array of plants, including, it appears, almost any vascular plant that comes its way with sufficiently accessible xylem vessels, the only apparent exceptions being, as noted above, Coronilla varia and Asclepius species. Philaenus spumarius nymphs are relatively sessile, moving infrequently, if at all, among hosts. By analogy with their leafhopper vector counterparts [240], they probably lose any X. fastidiosa infection upon molting, which occurs five times during nymphal development. In consequence, nymphs are not effective vectors. In contrast, adults are known to be effective vectors [110] but their exact host range is much less certain, with many fewer hosts documented (Table 1), this in turn being substantially due to the difficulty noted in distinguishing between a functional host plant and one on which the insect is merely positioned. Although efficacy of transmission varies greatly with host species [1], P. spumarius adults are exceptionally well positioned to vector X. fastidiosa quickly and widely wherever the two co-occur, contingent on local conditions that favor the propagation of the bacterium within and among host plants [22, 40, 106]. Potential counter measures include: 1) local elimination or population reduction of P. spumarius by management of agriculturally adjacent host plants [2, 106, 212, 237], 2) reduction or elimination of plant sources of Xylella fastidiosa infection [21], and 3) reduction of target plant susceptibility, through selection of cultivars with genetic resistance or other measures that reduce plant vulnerability to transmission and/or infection [2, 21, 241, 242]. In turn, P. spumarius can be used as a sentinel organism to detect and monitor the presence of X. fastidiosa in local environments [243]. Going forward, we suggest several guidelines for future studies of P. spumarius on host plants. First and foremost, investigators should always specify life stage. A substantial number of past reports, especially from the agricultural sector, have omitted this important information. Second, to the degree possible, quantify the results. Counts are best, but even simple qualitative observations, such as “rare” or “abundant” are helpful, especially for observations on adults. Third, where P. spumarius nymphs are abundant, record the plants on which they are apparently absent, especially in instances in which the uninfested plant species are frequent and apparently suitable for feeding. This will assist in the ongoing search for alternative understory plants suitable for reducing P. spumarius numbers in agricultural settings. Recent screening of nymphs in the Basilicata Region of Italy by Trotta et al. [11] is a model for this approach. The authors include data on 48 plant species that hosted nymphs and on 17 species that did not.

[END]
---
[1] Url: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0291734

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.