Insect-Plant Interactions Among Several Species of Piper in the Rio Abajo Forest Preserve, Puerto Rico

Selbyana 33(1): 1–15. 2019.

EATING OUT OR DINING IN: INSECT-PLANT INTERACTIONS AMONG SEVERAL SPECIES OF PIPER IN THE RIO ABAJO FOREST PRESERVE, PUERTO RICO

1 2 3 CHRISTA WISNIEWSKI , ALLAN J. BORNSTEIN *, DIANE L. WOOD

1Burns & McDonnell Engineering Company, Inc, 9400 Ward Parkway, Kansas City, MO 64114, USA 2Department of Biology, Southeast Missouri State University, One University Plaza, Cape Girardeau, MO 63701, USA; E-mail: [email protected] 3Department of Biology, Southeast Missouri State University, One University Plaza, Cape Girardeau, MO 63701, USA

Abstract Pollination activity was investigated for Piper glabrescens, Piper hispidum, Piper jacquemontianum, and Piper umbellatum, common species in the Rio Abajo Forest Preserve in Puerto Rico. During observation periods, the main floral visitors were syrphid and drosophilid flies, which suggests they play a prominent role as potential pollen vectors for these four species. In addition, two ant species (Linepithema iniquum and Wasmannia auropunctata) were observed to nest in the stems of three Piper species (Piper abajoense, Piper aduncum, and Piper umbellatum). In the case of Piper umbellatum, this interaction is the first documentation of an ant-plant association in Piper section Pothomorphe.

Key words: Piperaceae, Pollination, Syrphidae, Ant-plant interaction, Formicidae * Corresponding Author

Wisniewski et al.: Insect-plant interactions of Piper 2

the Atlantic forest of Brazil (Vargas-Rojas Introduction and Vieira 2017). Multiple studies have addressed From a morphological perspective, insect-plant interactions involving Neo- individual flowers of Piper L. are small tropical pipers (genus Piper L., family and lack the perianth parts (or nectar) that Piperaceae), including herbivory (e.g., traditionally function to attract insects. Marquis 1990, 1991; Dyer et al. 2003, Therefore, insects visit Piper flowers to 2004, 2010; Letourneau 2004), pollination gather pollen, undoubtedly attracted by (e.g., Semple 1974; Fleming 1985; de the usually white to yellowish flowers Figueiredo and Sazima 2000, 2004; clustered together in large numbers to Kikuchi et al. 2007; Vargas-Rojas and produce elongate, spike-like inflores- Vieira 2017), and ant habitation (e.g., cences. With rare exception, Neotropical Risch et al. 1977; Letourneau 1998; species of Piper have bisexual flowers, Fischer et al. 2002; Tepe et al. 2004, which may explain their ability to self- 2007a, 2007b, 2009, 2014). pollinate, but those species studied in With respect to pollination biology, detail display some degree of dichogamy, available research suggests that insects, either protandry or protogyny (Martin and wind, and selfing play a role for Gregory 1962; de Figueiredo and Sazima Neotropical Piper. For example, Semple 2000; Kikuchi et al. 2007; Valentin-Silva (1974) found that several bee species of et al. 2015; Vargas-Rojas and Vieira the genus Trigona (Jurine) and some 2017), which presumably limits selfing to beetles were pollinators of various Piper some extent. spp. in Costa Rica. Similarly, Fleming Holldobler and Wilson (1990) (1985) discovered that Trigona bees, indicate that the predominant evidence for along with syrphid flies, were the main myrmecophily is the presence of plant pollinators in his study of five Piper structures that serve as domatia, but not species in Costa Rica. During his study of all ant affiliations are necessarily P. arieianum C. DC. in Costa Rica, symbiotic due to the opportunistic nature Marquis (1988) observed pollinators that of ants. A total of 11 Piper species from included 15 species of Diptera and Central and South America are known to Hymenoptera. For some pipers in harbor nesting ants (Burger 1972; Risch et southeastern Brazil, de Figueiredo and al. 1977; Tepe et al. 2004, 2009, 2014). Sazima (2000) observed that hoverflies Several ant species have been noted to and bees were common visitors and inhabit Piper spp. (Risch et al. 1977; therefore likely pollinators, but concluded Holldobler and Wilson 1990; Tepe et al. that wind was also a factor. In their study 2004, 2014; Muñoz et al. 2012), but of Piper dilatatum L.C. Rich. on Barro Pheidole (Westwood) appears to be the Colorado Island, Panama, Kikuchi et al. predominant genus, with Pheidole (2007) also determined that syrphids, bicornis (Forel) the principal species halictids, and stingless bees (e.g., Trigona (Letourneau 2004). Many of the nesting spp.) were the most likely pollinators, and ants establish colonies in petiolar cavities, that their activity, especially the bee stems, or both (Burger 1972; Risch et al. species, leads to significant levels of 1977; Letourneau 2004; Tepe et al. 2004, geitonogamy. Various bees [e.g., Apis 2009, 2014). In Piper cenocladum C. DC., mellifera (L.) and Melipona (Illiger) spp.] P. fimbriulatum C. DC., and P. and syrphids [especially Ocyptamus sagittifolium C. DC., Pheidole spp. (Macquart) spp.] were common visitors initially established nests in the petiolar based on a study of P. caldense C. DC. in cavities and, as the colony increased in

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 3

size, entered and excavated the stem at the Materials and Methods base of the petiolar cavity (Risch et al. Study Site 1977). Obviously, Pheidole spp., as well This study was conducted in the Río as other ant species, are benefiting from Abajo Forest Reserve in Puerto Rico, nesting sites provided by various Piper located south of Arecibo, in the karst species. Additionally, several Piper spp. region. The reserve was established in are known to provide nutrients in the form 1935 on land that had previously been of single-celled food (pearl) bodies (Risch deforested (Acevedo-Rodriguez and et al. 1977; Fischer et al. 2002), which Axelrod 1999). The main ecosystems that occur, depending on the species, either on have been re-established in the reserve are the inner surface of the petiolar sheath tropical and subtropical forests (Acevedo- (Risch et al. 1977) or on the surface of the Rodriguez and Axelrod 1999). stem cavity (Tepe et al. 2009). There are 11 known Piper species in The occurrence of ants in young this area, including P. abajoense Bornst, Piper stems is facilitated by the general P. aduncum L., P. amalago L., P. lack of woodiness, with only the blattarum Spreng., P. glabrescens (Miq.) outermost, peripheral ring of vascular C. DC., P. hispidum Sw., P. jacque- bundles producing any noticeable montianum Kunth, P. marginatum Jacq., secondary tissues (Tepe et al. 2007a). This P. peltatum L., P. swartzianum (Miq.) C. means that a significant portion of the DC., and P. umbellatum L. As currently young stems consists of soft, circumscribed, most of these species are parenchymatous tissue, including the pith widely distributed in the Neotropics, region excavated for occupancy. Although except for the endemic P. abajoense there are some medullary vascular bundles (Puerto Rico), P. blattarum (Puerto Rico), in the pith, as well as a sclerenchymatous and P. swartzianum (Dominican Republic ring that separates the pith from the cortex + Puerto Rico). Although these 11 species (Tepe et al. 2007a), these features do not co-occur in the area, and previous appear to significantly deter insect collection labels indicate that many of activity. In addition, myrmecophytic them should be in flower simultaneously species of Piper studied anatomically at various times throughout the year, the have fewer raphides (needle-shaped pollination study was limited to those crystals) and starch grains in the pith species in flowering condition and present region (Tepe et al. 2007a), further in sufficient abundance during the time of facilitating excavation activity. the study; these species were P. To date, all of the studies concerning glabrescens, P. hispidum, P. jacque- Neotropical Piper pollination and ant montianum, and P. umbellatum. interactions have involved mainland Sampling Protocol species. However, insect (and plant) The first field season occurred from species often have limited distributions, so 5–10 January 2013, and prior to the first it is important to examine possible inter- full field day the best methods, times, and actions in different regions of the world. locations for examining species of Piper Therefore, the intent of this study was to and collecting potential pollinators were observe insect visitors to the inflores- determined. A second field season cences of Piper, and to document any ants occurred from 1–5 June 2013 to see if nesting in the stems of available Piper different pollinators might be active at species to compare the Puerto Rican insect that time. During both observation -Piper interactions with those of the periods, the time of day for peak insect mainland species.

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 4

activity was determined to be from 10 insect visitors because field identification a.m. to 3 p.m. Collecting was conducted for many taxa to the level of genus or during these times to ensure that species is difficult. Additionally, some representative insect visitors were insects were purposely not collected in sufficiently documented. order to observe their basic behavior (e.g., how they moved along a given During the January collection period, inflorescence). P. glabrescens and P. umbellatum were in flower and abundant; a total of ca. 60 Plants were also visually inspected hours of collecting/observing occurred for for the presence of ants nesting in the the former species and ca. 30 hours for the stems. If ants were actively moving into latter. In June, P. hispidum, P. jacque- and out of openings in a stem, then that montianum, and P. umbellatum were stem was cut one internode above and studied for the same reasons; a total of ca. below the nest openings. The length of the 30 hours of collecting/observing occurred collected stem was then sectioned and all for each species during this period. sections were placed in a labeled, wide- mouth, 1000 ml Nalgene® container of Specimens of the representative 85% ethanol. To confirm proper field Piper species involved in the pollination identification of the plant, representative project were collected during both study flowers and/or fruits were collected and periods to serve as herbarium vouchers for placed in the container, as well. confirmation of identification. One Lab Analysis voucher will remain in the herbarium at Southeast Missouri State University After both collecting events, insect (SEMO), and any duplicates will be specimens were taken back to the distributed to herbaria in the United States laboratory at Southeast Missouri State (e.g., MO) and Puerto Rico (e.g., University (SEMO) for processing. All UPRRP). insects were identified to the lowest possible taxonomic level and sent to During both field periods, an area ca. specialists for confirmation or further 2×20 meters, in which Piper spp. were in identification. A few representative flower, was established for each insects were examined via light and researcher (three plot areas in January, scanning electron microscopy to detect two of which contained Piper glabrescens the presence of Piper pollen. Given that only and one that contained Piper ethanol preservation could potentially umbellatum; two plot areas in June, one of wash some pollen from the insect bodies, which contained both Piper jacque- random vials (ca. 20) were also examined montianum and Piper hispidum, and the for pollen within the ethanol fluid. All other contained just Piper umbellatum). remaining voucher specimens were There were several individuals of a given deposited in the insect collection at species within each observation area, and SEMO, and any duplicates will be sent to each individual displayed multiple the University of Puerto Rico-Rio Piedras, inflorescences at various stages of and the Smithsonian Natural History maturation. Representative individuals of Museum, to be included in their the insects visiting Piper inflorescences collections. were collected using an insect net or aspirator. Each individual was initially Results placed in a kill-jar, and then transferred to Pollination Biology a labeled vial of 85% ethanol for preservation. These stored individuals A total of twelve different insect taxa served as references of the diversity of were collected while visiting the

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 5

of Piper jacquemontianum and P. inflorescences of at least one of the four umbellatum. species of Piper during the two collection periods. During the January survey, nine During peak activity, syrphids were different species were observed as visitors observed to alight on a given inflores- to P. glabrescens and P. umbellatum with cence, probe the flowers with their mouth syrphids as the most common/diverse parts, and move slowly up and down a (Table 1; Figure 1A). Leucopodella given spike, but rarely for more than one gracilis (Williston), Ocyptamus neo- minute. On occasion it was possible to parvicornis (Telford), and the drosophilid follow a single individual as it traveled fly Leucophenga (Mik) sp. were from one spike to another, most often on especially prevalent visitors to P. the same plant. glabrescens based on field observations Light, and especially scanning and representative collecting effort, electron, microscopy confirmed the whereas the most frequent visitor to P. presence of Piper pollen on the syrphid umbellatum was the weevil, Peridinetus and drosophilid flies, especially around signatus (Rosenschöld) (Table 1). Other their mouth region, but occasionally syrphid visitors during this observation among the setae on their thorax and legs period included Leucopodella incompta (Figure 1B, C). Examination of storage (Austen), Ocyptamus parvicornis (Loew), vials indicated only a few pollen grains of Ocyptamus ricus (Curran), Toxomerus Piper in less than 20% of the samples luna (Hull), and Toxomerus violaceus surveyed, indicating that ethanol (Curran), all of which were noted on P. preservation had minimal effect on pollen glabrescens, but only O. ricus occurred on adherence. P. umbellatum. Ant Habitation Insects collected from the inflores- cences of Piper hispidum, P. Two ant species, Linepithema jacquemontianum, and P. umbellatum iniquum (Mayr) and Wasmannia during the June survey were primarily auropunctata (Roger) were observed syrphids (Table 1), which coincides with nesting in the stems of three Piper results from January. Again, based on species. Linepithema iniquum nests were field observations and representative collected from P. aduncum and P. collecting effort, the most common umbellatum stems, whereas Wasmannia species were Leucopodella gracilis and auropunctata nests were collected from P. Ocyptamus neoparvicornis. Other species abajoense Bornst. stems. of syrphids were observed and collected, A colony of Linepithema iniquum but generally in lower numbers compared was collected from two live stems of to the January collection period. Piper umbellatum (Figure 1D), and one Individuals of Peridinetus signatus and dead stem of P. aduncum. The colony in Leucophenga sp. were also collected from P. umbellatum occupied a 146.6 cm the inflorescences of P. hispidum and/or section, consisting of 100.7 cm in the P. jacquemontianum during this time. The main (larger) stem and a 45.9 cm branch honey bee, Apis mellifera, was also (smaller) arising from that main stem. On observed on the inflorescences of P. the main stem, six of the internodes hispidum in June, but to a very limited contained adult ants and in two adjoining extent. internodes adults, eggs, and immatures On multiple occasions, pairs of the were present. Fourteen other internodes weevil, Peridinetus signatus, were did not contain any individuals, but there observed in copulo on the inflorescences was evidence of potential ant occupation

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 6

Figure 1. Various images of Piper-insect associations in Puerto Rico. A. Syrphid fly, Ocyptamus parvicornis, visiting an inflorescence of Piper umbellatum. B. Scanning electron micrograph from mouth-part region of a drosophilid fly, Leucophenga sp., with Piper pollen indicated (yellow circles). C. Scanning electron micrograph showing Piper pollen (yellow circles) among setae on leg of the syrphid fly, Ocyptamus neoparvicornis. D. Individuals of the ant, Linepithema iniquum, within cauline nest of Piper umbellatum. E. Close-up of resinous exudate bodies within excavated stem of Piper umbellatum. F. Entrance holes to cauline nests occupied by the ant Linepithema iniquum, located both above and below a node (arrowheads) in Piper umbellatum, with several ants visible nearby.

SELBYANA Volume 33(1) 2019

Wisniewski et al.: Insect-plant interactions of Piper 7

) ) -

10

–

cates cates

- - - - + - + + + - + -

ndi

Jun

+ + - - - - + - + - - -

Jan

, , a minus sign (

P. umbellatum

Piper

- - - - -

+ + + + + + +

Jun

* * * * * * * * * * * *

Jan

P. jacquemontianum

species

- - - - -

+ + + + + + +

Jun

Piper Piper

* * * * * * * * * * * *

Jan P. hispidum

* * * * * * * * * * * *

Jun

spp. spp. inflorescences at the Rio Abajo Forest Reserve in Puerto Rico (5

- - - -

+ + + + + + + +

Jan

P. glabrescens

Piper

(Telford) ǂ (Telford)

(Austen) ǂ (Austen)

(Loew)ǂ

(Williston) ǂ (Williston)

(Curran) ǂ (Curran)

(Rosenschöld)

sp.

(Curran) ǂ (Curran)

(Hull) ǂ (Hull)

(Linnaeus)

(Mik)

5 5 June 2013). A plus sign (+) indicates individuals visited inflorescences of that species of

–

Peridinetus signatus Peridinetus Leucophenga gracilis Leucopodella incompta Leucopodella neoparvicornis Ocyptamus parvicornis Ocyptamus ricus Ocyptamus luna Toxomerus violaceus Toxomerus mellifera Apis

1. 1. Diversity of insect taxa collected from

Curculionidae Drosophilidae Heleomyzidae Phoridae Syrphidae Apidae

Coleoptera Diptera Hymenoptera

Insect Taxa Insect

Table January and 1 indicates no visits, and an asterisk (*) indicates that a given species was not in flower during that observation specialist. a by verified period; identification ǂ i

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 8

based on excavation of the pith. In the internodes did not appear to have been excavated internodes, a thin pith layer completed by the ant colony based on the lined the walls, and had a golden to light degree and irregularity of pith excavation brown coloration. It is assumed that this (i.e., entirely removed or thin layer was due to exposure to air, because remaining). The excavated area was undisturbed Piper pith tends to be solid continuous between the three internodes. white (E. Tepe pers. comm.). In The two rounded entrances, one at the mid internodes with slightly larger diameters, a -point of the basal internode and the other double tunnel system existed, with a at the middle node were 3.9 mm and 4.0 central partition of pith separating the mm, respectively. No resinous exudate adjacent tunnels. In several internodes structures or pearl bodies were present. there were small, shallow excavations directly off the main tunnel. Small, The second colony of Wasmannia rounded, resinous exudate at the surface occurred in a single dead stem of Piper of the remaining pith was noted in many abajoense and occupied a 32.5 cm section of the excavated internodes (Figure 1E), composed of four internodes of a younger but pearl bodies were absent. In all but side branch from the plant’s main stem. one internode, entrances were located at The first and second internodes contained or very near the nodes (Figure 1F), and low numbers of immatures with few internodes were not interconnected as a adults. The excavation patterns and result of excavation. The opening sizes entrances were similar to that described ranged from 1.2 to 4.2 mm in diameter, above for the ant colony found in the live with an average of 1.99 mm. stems of P. abajoense, except that adjacent internodes were not connected In the dead stem of Piper aduncum, via continuous tunneling. In all but one the colony occupied a 72 cm section internode, entrances were located at or composed of five internodes on the main near the node, and the rounded openings stem and a sixth internode branching off ranged from 1.3 to 8.5 mm in diameter, of that main stem between internodes two with an average of 3.1 mm. No resinous and three. This sixth internode contained exudate structures or pearl bodies were some immatures, but no adults. It is present. assumed they had left the stem after being placed in the container filled with ethanol. Discussion Due to the degraded condition of the stem, Insects were using six species of excavation patterns could not be easily Piper (P. abajoense, P. aduncum, P. discerned because the pith, where present, glabrescens, P. hispidum, P. jacque- had separated from the stem wall. Of the montianum, and P. umbellatum) in 12 openings, 10 were located at or near various ways. They were observed feed- the node. The opening sizes ranged from ing on pollen within the inflorescences of 0.8 to 3.8 mm in diameter, with an P. glabrescens, P. hispidum, P. jacque- average of 2.37 mm. montianum, and P. umbellatum. Two ant Two Piper abajoense individuals species, Linepithema iniquum and each contained a colony of Wasmannia Wasmannia auropunctata, were found auropunctata. One colony occurred in a nesting in the stems of P. abajoense, P. single live stem and occupied a 21.7 cm aduncum, or P. umbellatum. section composed of three internodes. The Pollination Biology most basal internode contained immatures, while the middle internode The predominant insect visitors contained a few adults. Excavation of the (potential pollinators) for all plants during the two collecting events were hoverflies

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 9

in the family Syrphidae (Table 1), which Unlike the bees of Puerto Rico, there agrees with the results from de Figueiredo is diversity of syrphid fly species located and Sazima (2000, 2004, 2007) in their on the island (Thompson 1981), which is study of Piper and Peperomia species in reflected in the seven species collected southeastern Brazil. The syrphid fly genus (Table 1). The majority of these species Ocyptamus, commonly collected in this in Puerto Rico have a Neotropical origin, study, was also identified as a pollinator in with only 18% of the total species of Brazil (de Figueiredo and Sazima 2000, syrphid fly being endemic to the island 2004; Vargas-Rojas and Vieira 2017) and (Thompson 1981). Costa Rica (Fleming 1985). Syrphids were The diversity and abundance of also frequent visitors (18%) to Piper syrphids clearly enables them to be dilatatum Rich. in Panama (Kikuchi et al. important potential pollinators in Puerto 2007), but identification to the genus level Rico. Combined with their generalist was not performed in that study so a more feeding activity and stable population specific comparison is not possible. sizes throughout the year (Owen and The analyses conducted on Costa Gilbert 1989; de Figueiredo and Sazima Rican pipers by Semple (1974) and 2000), they appear to be common and Fleming (1985), and Piper dilatatum from efficient pollinators (Gilbert et al. 1985; Panama (Kikuchi et al. 2007), indicated Shi et al. 2009; Willmer 2011). Relying that Trigona bees were the predominant on generalist pollinators is especially pollinating group. Similarly, Vargas-Rojas important for the plants during the cold and Vieira (2017) discovered that bees and dry seasons in tropical forests (Freitas [Apis mellifera and Melipona spp.] were and Sazima 2003; Dyer et al. 2004), or much more frequent visitors to when bee populations have declined inflorescences of P. caldense in Brazil (Willmer 2011). than syrphid flies (84% vs. 16%). During In order to facilitate collection of an examination of the Puerto Rican bee pollen, some species of syrphid flies have fauna, Genaro and Franz (2008) did not modified curly or branched hairs, modi- record any Trigona or Melipona bees on fied bristles, or elongated mouth parts the island, and these bee genera were not (Holloway 1976; Willmer 2011). These collected during this study. Apparently, flies may also have pollen scrapers Puerto Rico has limited hymenopteran located on their legs, facilitating the biodiversity due to long-distance dispersal transfer of pollen from hind to front legs; required from the mainland to establish this occurs when the individuals are populations on the island (Genaro and hovering in the air (Holloway 1976) and Franz 2008). The few endemic bee species allows the pollen to be cleaned off of their present are in low abundance and exhibit bodies (Freitas and Sazima 2003). restricted distributions across the island Although morphological details were not (Genaro and Franz 2008). The loss of recorded for the various syrphids habitat, due to agriculture, deforestation, collected in this study, microscopy did and increasing urban sprawl, appears to be confirm the presence of Piper pollen the main cause of the current decline of among the setae on their legs and thorax, bee species found on the island (Genaro indicating that they could be effective and Franz 2008). The low bee diversity in pollen vectors. In addition, several Puerto Rico may be why potential Piper syrphids were observed to fly between pollinators collected in this study differed inflorescences (occasionally on different from those collected in Central and South neighboring individuals of the same Piper America; quite simply, they have different species), suggesting that pollen flow could pollinator pools. occur. Additional studies would be

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 10

required to confirm that pollen is actually Puerto Rico (Wheeler 1908; Wild 2007), being deposited between inflorescences, whereas Wasmannia auropunctata is perhaps using fluorescent dye as was done described as a ground foraging species by Kikuchi et al. (2007). that often occupies disturbed sites (e.g., floodplains along creeks) (Hahn and The weevil species, Peridinetus Wheeler 2002; Orivel et al. 2009) and is signatus, was also collected from Piper most likely an introduced species to the jacquemontianum and P. umbellatum island (Wetterer and Porter 2003). inflorescences (Table 1). This weevil was seen mating on the inflorescences of P. Previous researchers have noted that umbellatum. It is certainly known that occupation of dead stems is common for species of Peridinetus in the adult stage Linepithema iniquum (Wheeler 1908; are frequent herbivores of Piper leaves in Wild 2007, 2009; Escárraga and Guerrero Costa Rica (Marquis 1991), so their 2016). However, based on excavation occurrence on pipers in Puerto Rico is not patterns and entrance openings, it appears especially surprising. Additionally, it is that L. iniquum is capable of establishing not uncommon for weevils to mate and a colony through their own excavation oviposit on a host plant (R. Anderson efforts within the stems of Piper pers. comm.), thus these pipers may also umbellatum. The presence of resinous serve as a host for Peridinetus signatus exudate (Figure 1E) is likely a wound larval development. response and not comparable to the Piper Myrmecophytes presence of pearl bodies, as in other pipers with true cauline domatia (E. Previously, known ant-plant associa- Tepe, pers. comm.). tions involving cauline occupation in Piper have been limited to Central and In the Piper aduncum dead stems, South America, and in most cases the the condition of the collected sample was associated species were within the genus too degraded to ascertain whether Line- Pheidole (e.g., Risch et al. 1977; Tepe et pithema iniquum was responsible for the al. 2007a, 2009, 2014). Although excavation. If entrance opening size is Linepithema iniquum and Wasmannia indicative, then it is likely that this auropunctata have been documented to colony simply took advantage of a nest in Piper cernuum Vell. in the Atlantic vacated stem that was previously occupied by a larger, stem-boring insect, forests of Brazil (Muñoz et al. 2012), this is the first documented occurrence of possibly of some curculionid larva. these species actually nesting within Piper The excavation patterns and stems in Puerto Rico, and the first record relatively large entrance openings suggest of such relationships beyond the that Wasmannia auropunctata mainland. In the case of P. umbellatum, established nests within Piper abajoense this interaction is the first documentation stems that had been previously excavated of ant-plant associations in Piper section by other stem-boring insects. This seems Pothomorphe (Miq.) C. DC., representing reasonable to assume given its known another independent evolutionary biology as an opportunistic species, development of Piper plants inhabited by whether in its native (Dejean et al. 2003; ants, and expands the number to four Wetterer and Porter 2003; Longino and clades (see Figure 2). Fernández 2007) or introduced range Comparatively, the two ant species (Wetterer and Porter 2003; Orivel et al. are ecologically different in several 2009; Muñoz et al. 2012; Mikissa et al. respects. Linepithema iniquum is 2013). considered to be arboreal and native to Conclusions

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 11

Figure 2. Summary phylogeny of Piper based on Bayesian analysis of nuclear ITS and plastid psbJ-petA sequence data (originally redrawn from Jaramillo et al. 2008, and subsequently modified from Tepe et al. 2009). Numbers above branches indicate Bayesian posterior probabilities, and numbers in triangles represent an estimation of species in each section. Piper abajoense and Piper aduncum in section Radula are newly documented to harbor ants, while Piper umbellatum represents the first documentation of an ant-plant association from section Pothomorphe. Each of the ant-associated species is nested within a clade of species not linked with ants, indicating a minimum of four independent evolutionary origins of ant-plant associations among New World Piper.

remains to be investigated in Caribbean Overall, this study documents novel pipers. Numerous studies of mainland ecological interactions involving Puerto taxa support the role of frugivorous bats Rican pipers and insects. In particular, the as the primary dispersal agents (see results confirm that insects, especially Heithaus et al. 1975; Fleming et al. 1977; syrphids (Table 1), likely contribute to Fleming 1981; Fleming and Heithaus pollination of the four Piper species 1981; Palmeirim et al. 1989; Charles- examined, in contrast to the prevalent role Dominique 1991; Marinho-Filho 1991; of bees among the few mainland taxa Bizerril and Raw 1998; Thies and Kalko studied. Regardless of previous results, to 2004), but comparative analyses with better understand the differences in island taxa will first require appropriate natural history between island and mainland species of Piper, more studies field projects. need to be completed concerning possible With respect to ants establishing pollen vectors, especially given the cauline nests, it is not surprising that both apparent difference in pollinator pools Linepithema iniquum and Wasmannia between these regions. In essence, a auropunctata colonized what appear to be significant part of the sexual life history abandoned stems that were previously (i.e., pollination) of most pipers is excavated by a different wood-boring unknown. Similarly, the fate of successful insect. However, the excavation and sexual reproduction (i.e., fruit dispersal) establishment of a colony by L. iniquum

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 12

Funding Committee at Southeast Missouri in Piper umbellatum stems is significant State University. because it involves an ant species outside of the genus Pheidole, the most common Literature Cited ant inhabitant of pipers, and represents an Acevedo-Rodriguez, P. and F.S. Axelrod. interaction with a species from a 1999. Annotated checklist for the Trach- completely different lineage of Piper. eophytes of Río Abajo Forest Preserve, Relatively few representatives (< 20 Puerto Rico. Caribb. J. Sci. 35: 265–285. species) from four distantly related Bizerril, M.X.A. and A. Raw. 1998. Feeding lineages of Neotropical pipers (see Figure behavior of bats and the dispersal of Piper 2) are known to have ants nesting within arboreum seeds in Brazil. J. Trop. Ecol. their stems and/or petioles, along with 14: 109–114. several, perhaps closely related, Old Burger, W.C. 1972. Evolutionary trends in World taxa (Jaramillo and Callejas 2004; Central American species of Piper Gardner 2006). Clearly, further research (Piperaceae). Brittonia 24: 356–362. efforts concerning Piper-ant interactions Charles-Dominique, P. 1991. Feeding strategy are needed to determine whether 1) these and activity budget of the frugivorous bat relationships exist among other lineages, Carollia perspicillata (Chiroptera: Phyl- 2) perhaps involve additional species lostomidae) in French Guiana. J. Trop. within lineages already documented, and Ecol. 7: 243–256. 3) might occur within other distributional de Figueiredo, R.A. and M. Sazima. 2000. regions. These studies will allow Pollination biology of Piperaceae species in southeastern Brazil. Ann. Bot. 85: 455– researchers to better address important 460. questions such as why these associations exist in distantly related taxa, how these ———. 2004. Pollination ecology and resource partitioning in Neotropical associations have evolved, and why Pipers. Pp. 33–57 in L.A. Dyer and different plant organs (stems, petioles, and A.D.N. Palmer, eds. Piper: A Model leaves) are used by the ant inhabitants. Genus for Studies of Phytochemistry, Acknowledgments Ecology, and Evolution. Kluwer Academic/Plenum Publishers, New York. We thank Dr. Frank Axelrod for his ———. 2007. Phenology and pollination help in acquiring the necessary permits to biology of eight Peperomia species collect specimens in Puerto Rico and (Piperaceae) in semideciduous forests in initially pointing out this wonderful field southeastern Brazil. Plant Biol. 9: 136– site. Dr. F. Christian Thompson kindly 141. provided verification of the syrphid Dejean, A., S. Durou, I. Olmsted, R.R. Snel- species we collected. We also thank the ling, and J. Orivel. 2003. Nest site selec- Department of Natural and Environmental tion by ants in a flooded Mexican Resources of Puerto Rico for allowing us mangrove, with special reference to the to conduct this project in the Rio Abajo epiphytic orchid Myrmecophila christ- inae. J. Trop. Ecol. 19: 325–331. Forest Reserve. Dr. Eric Tepe kindly examined the stems of Piper umbellatum, Dyer, L.A., C.D. Dodson, J.O. Stireman, M.A. Tobler, A.M. Smilanich, R.M. Fincher, provided useful incite concerning Piper- and D.K. Letourneau. 2003. Synergistic ant interactions, and allowed us to modify effects of three Piper amides on generalist the phylogenetic diagram from his 2009 and specialist herbivores. J. Chem. Ecol. paper. Dr. E. O. Wilson kindly provided 29: 2499–2514. relevant communication concerning the Dyer, L.A., J. Richards, and C.D. Dodson. proper use of the term “domatia”. Funds 2004. Isolation, synthesis, and evolution- to support travel to Puerto Rico were ary ecology of Piper amides. Pp. 117–139 provided by the Grants and Research in L.A. Dyer and A.D.N. Palmer, eds.

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 13

Piper: A Model Genus for Studies of Hahn, D.A. and D.E. Wheeler. 2002. Seasonal Phytochemistry, Ecology, and Evolution. foraging activity and bait preferences of Kluwer Academic/Plenum Publishers, ants on Barro Colorado Island, Panama. New York. Biotropica 34: 349–356. Dyer, L.A., D.K. Letourneau, G.V. Chavarria, Heithaus, E.R., T.H. Fleming, and P.A. Opler. and D.S. Amoretti. 2010. Herbivores on a 1975. Foraging patterns and resource dominant understory shrub increase local utilization in seven species of bats in a plant diversity in rain forest communities. seasonal tropical forest. Ecol. 56: 841– Ecol. 91: 3707–3718. 854. Escárraga, M. and R. . Guerrero. 2016. The ant Holldobler, B., and E.O. Wilson. 1990. The genus Linepithema (Formicidae: Doli- Ants. Springer-Verlag, Berlin. choderinae) in Colombia. Zootaxa 4208: Holloway, B.A. 1976. Pollen-feeding in 446–458. hoverflies (Diptera: Syrphidae). New Fischer, R.C., A. Richter, W. Wanek, and V. Zeal. J. Zool. 3: 339–350. Mayer. 2002. Plants feed ants: food Jaramillo, M.A. and R. Callejas. 2004. Current bodies of myrmecophytic Piper and their perspectives on the classification and significance for the interaction with phylogenetics of the genus Piper L. Pp. Pheidole bicornis ants. Oecologia 133: 179–198 in L.A. Dyer and A.D.N. 186–192. Palmer, eds. Piper: A Model Genus for Fleming, T.H. 1981. Fecundity, fruiting Studies of Phytochemistry, Ecology, and pattern, and seed dispersal in Piper Evolution. Kluwer Academic/Plenum amalago (Piperaceae), a bat-dispersed Publishers, New York. tropical shrub. Oecologia 51: 42–46. Jaramillo, M.A., R. Callejas, C. Davidson, J.F. ———. 1985. Coexistence of five sympatric Smith, A.C. Stevens, and E.J. Tepe. 2008. Piper (Piperaceae) species in a tropical A phylogeny of the tropical genus Piper dry forest. Ecol. 66: 688–700. using ITS and the chloroplast intron psbJ- Fleming, T.H. and E.R. Heithaus. 1981. petA. Syst. Bot. 33: 647–660. Frugivorous bats, seed shadows, and the Kikuchi, D.W., E. Lasso, J.W. Dalling, and N. structure of tropical forests. Biotropica Nur. 2007. Pollinators and pollen 13: 45–53. dispersal of Piper dilatatum (Piperaceae) Fleming, T.H., E.R. Heithaus, and W.B. on Barro Colorado Island, Panama. J. Sawyer. 1977. An experimental analysis Trop. Ecol. 23: 603–606. of the food location behavior of Letourneau, D.K. 1998. Ants, stem-borers, frugivorous bats. Ecol. 58: 619–627. and fungal pathogens: experimental tests Freitas, L. and M. Sazima. 2003. Daily of a fitness advantage in Piper ant-plants. blooming pattern and pollination by Ecol. 79: 593–603. syrphids in Sisyrinchium vaginatum ———. 2004. Mutualism, anti-herbivore (Iridaceae) in southeastern Brazil. J. defense, and trophic cascades: Piper ant- Torrey Bot. Soc. 130: 55–61. plants as a mesocosm for experiment- Gardner, R.O. 2006. Piper (Piperaceae) in the ation. Pp. 5–32 in L.A. Dyer and A.D.N. Philippine Islands: the climbing species. Palmer, eds. Piper: A Model Genus for Blumea 51: 569–586. Studies of Phytochemistry, Ecology, and Evolution. Kluwer Academic/Plenum Genaro, J.A. and N.M. Franz. 2008. The bees Publishers, New York. of Greater Puerto Rico (Hymenoptera: Apoidea: Anthophila). Insecta Mundi 40: Longino, J.T. and F. Fernández 2007. 1–24. Taxonomic revision of the genus Wasmannia. Mem. Amer. Entomol. Inst. Gilbert, F.S., E.F. Harding, J.M. Line, and I. 80: 271–289. Perry. 1985. Morphological approaches to community structure in hoverflies Marinho-Filho, J.S. 1991. The coexistence of (Diptera, Syrphidae). Proc. R. Soc. two frugivorous bat species and the phen- London, Ser. B 224: 115–130. ology of their food plants in Brazil. J. Trop. Ecol. 7: 59–67.

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 14

Marquis, R.J. 1988. Phenological variation in species of tropical Piper (Piperaceae) and the Neotropical understory shrub Piper their ant inhabitants. Am. Midl. Nat. 98: arieianum: causes and consequences. 433–444. Ecol. 69: 1552–1565. Semple, K.S. 1974. Pollination in Piperaceae. ———. 1990. Genotypic variation in leaf Ann. Mo. Bot. Gard. 61: 868–871. damage in Piper arieianum (Piperaceae) Shi, J., Y.-B. Luo, P. Bernhardt, J.-C. Ran, Z.- by a multispecies assemblage of herbi- J. Liu and Q. Zhou. 2009. Pollination by vores. Evol. 44: 104–120. deceit in Paphiopedilum barbigerum ———. 1991. Herbivore fauna of Piper (Orchidaceae): a staminode exploits the (Piperaceae) in a Costa Rican wet forest: innate colour preferences of hoverflies diversity, specificity, and impact. Pp. 179 (Syrphidae). Plant Biol. 11: 17–28. –208 in P.W. Price, T.M. Lewinsohn, Tepe, E.J., W.A. Kelley, G. Rodriquez - G.W. Fernandes, and W.W. Benson, eds. Casteñada, and L.A. Dyer. 2009. Charact- Plant-Animal Interactions: Evolutionary erizing the cauline domatia of two newly Ecology in Tropical and Temperate discovered Ecuadorian ant plants in Regions. John Wiley and Sons, Inc. New Piper: an example of convergent York. evolution. J. Insect Sci. 9: 27. doi: https:// Martin, F.W. and L.E. Gregory. 1962. Mode doi.org/10.1673/031.009.2701. of pollination and factors affecting fruit Tepe, E.J., G. Rodriquez-Casteñada, A.E. set in Piper nigrum L. in Puerto Rico. Glassmire, and L.A. Dyer. 2014. Piper Crop Sci. 2: 295–299. kelleyi, a hotspot of ecological inter- Mikissa, J.B., K. Jeffery, D. Fresneau, and J.L. actions and a new species from Ecuador Mercier. 2013. Impact of an invasive alien and Peru. Phytokeys 34: 19–32. ant, Wasmannia auropunctata Roger, on a Tepe, E.J., M.A. Vincent, and L.E. Watson. specialized plant-ant mutualism, Barteria 2004. Phylogenetic patterns, evolutionary fistulosa Mast. and Tetraponera aethiops trends, and the origin of ant-plant F. Smith in a Gabon forest. Ecol. associations in Piper section Macro- Entomol. 38: 580–584. stachys: Burger’s hypothesis revisited. Muñoz, A., V.O. Becker, and J.H.C. Delabie. Pp. 156–178 in L.A. Dyer and A.D.N. 2012. Ant (Formicidae) assemblages Palmer, eds. Piper: A Model Genus for associated with Piper spp. (Piperaceae) in Studies of Phytochemistry, Ecology, and the undergrowth of an Atlantic rainforest Evolution. Kluwer Academic/Plenum remnant in southeastern Bahia, Brazil. Publishers New York. Sociobiology 59: 741–754. ———. 2007a. Stem diversity, cauline Orivel, J., J. Grangier, J. Foucaud, J. le Breton, domatia, and the evolution of ant-plant F.-X. Andrès, H. Jourdan, J.H.C. Delabie, associations in Piper sect. Macrostachys D. Fournier, P. Cerdan, B. Facon, A. (Piperaceae). Amer. J. Bot. 94: 1–11. Estoup, and A. Dejean. 2009. Eco- ———. 2007b. The importance of petiole logically heterogeneous populations of the structure on inhability by ants in Piper invasive ant Wasmannia auropunctata sect. Macrostachys (Piperaceae). Bot. J. within its native and introduced ranges. Linn. Soc. 153: 181–191. Ecol. Entomol. 34: 504–512. Thies, W.A. and E.K.V. Kalko. 2004. Owen, J. and F.S. Gilbert. 1989. On the Phenology of neotropical pepper plants abundance of hoverflies (Syrphidae). (Piperaceae) and their association with Oikos 55: 183–193. their main dispersers, two short-tailed Palmeirim, J.M., D.L. Gorchov, and S. fruit bats, Carollia perspicillata and C. Stoleson. 1989. Trophic structure of a castanea (Phyllostomidae). Oikos 104: neotropical frugivore community: is there 362–376. competition between birds and bats. Thompson, F.C. 1981. Flower flies of the Oecologia 79: 403–411. West Indies (Diptera: Syrphidae). The Risch, S., M. McClure, J. Vandermeer, and S. Entomological Society of Washington, Waltz. 1977. Mutualism between three Washington, DC.

SELBYANA Volume 33(1) 2019 Wisniewski et al.: Insect-plant interactions of Piper 15

Valentin-Silva, A., V.P.D.M. Coelho, M.C. Ventrella, M.F. Vieira. 2015. Timing of pollen release and stigma receptivity period in Piper viscosanum: New insights into sexual reproduction of the genus. Amer. J. Bot. 102: 626–633. Vargas-Rojas, D.L. and M.F. Vieira. 2017. Sex expression, breeding system and pollinators of Piper caldense (Piperaceae) in the Brazilian Atlantic forest. Acta Biol. Colomb. 22: 370–378. Wetterer, J.K. and S.D. Porter. 2003. The little fire ant, Wasmannia auropunctata: distri- bution, impact and control. Sociobiology 41: 1–41. Wheeler, W.M. 1908. The ants of Puerto Rico and the Virgin Islands. Bull. Am. Mus. Nat. Hist. 32: 117–158. Wild, A.L. 2007. Taxonomic Revision of the Ant Genus Linepithema (Hymenoptera: Formicidae). University of California Press, Berkeley. Wild, A.L. 2009. Evolution of the Neotropical ant genus Linepithema. Syst. Entomol. 34: 49–62. Willmer, P. 2011. Pollination and Floral Ecology. Princeton University Press, Princeton.

SELBYANA Volume 33(1) 2019