Arthropod-Plant Interactions

, Volume 2, Issue 1, pp 1–8

Utilization on extrafloral nectaries and fruit domatia of Canavalia lineata and C. cathartica (Leguminosae) by ants

Authors

    • Graduate School of Advanced Technology and ScienceThe University of Tokushima
  • Tadashi Yamashiro
    • Department of Life Sciences, Faculty of Integrated Arts and SciencesThe University of Tokushima
Original Paper

DOI: 10.1007/s11829-008-9030-y

Cite this article as:
Yamashiro, A. & Yamashiro, T. Arthropod-Plant Interactions (2008) 2: 1. doi:10.1007/s11829-008-9030-y

Abstract

Utilization of the extrafloral nectaries (EFNs) and fruits of Canavalia lineata and C. cathartica by ants was investigated at 30 sites in Japan. The fruits of C. lineata and C. cathartica were inhabited by five and eight ant species, respectively. Ant nesting periods and their utilization of EFNs differed between C. lineata and C. cathartica. Canavalia lineata flowers once a year, and periods of EFN-utilization and fruit-nesting by ants do not overlap. The fruit-nesting ants on C. lineata seem to invade the plant from the holes made by moth larvae or breaches made by decay. The ants nesting on the fruits of C. lineata may defend the plant against seed herbivores because they feed on moth larvae. Canavalia cathartica flowers several times over a year, and fruits are found throughout the year; therefore, periods of EFN-utilization and fruit-nesting by ants are overlapped. Canavalia cathartica offers year-round nesting sites and food for ants, and therefore may receive a higher defensive effect from ants than C. lineata.

Keywords

Canavalia catharticaCanavalia lineataDomatiaEFNsFruitNestPlant–ant interaction

Introduction

Mutual relationships between ants and plants are extremely diverse. Plants have many strategies of utilizing ants; the most common is to provide nectar secreted from extrafloral nectaries (EFNs) as a reward (Martin and McKey 2003). Extrafloral nectar secretion recruits a diverse array of ants such as genuinely protective ants (e.g. Bently 1976; O’Dowd 1979; Stephenson 1982; Willmer and Stone 1997), opportunists (Apple and Feener 2001), and parasites (Yu and Pierce 1998; Stanton et al. 1999; Raine et al. 2004). However ants attracted to EFNs seem to generally protect the plant from harmful forms of herbivory and seed predation (e.g. Bently 1976; O’Dowd 1979; Stephenson 1982) and many studies have demonstrated that ants protect plants with EFNs from herbivore attacks and that this protection results in increased reproductive success (Horvitz and Schemske 1984; Kelly 1986; Oliveira et al. 1999). EFNs have been confirmed at least for 332 genera across 93 plant families (Koptur 1992), being more common in tropical than in temperate regions (Oliveira and Freitas 2004). Since EFNs represent only a part of the nutritional requirements of an ant colony, ant-guard systems mediated solely by the supply of nectar are usually very generalized (Schemske 1983).

Myrmecophytes develop a special hollow structure (domatia) facilitating ant nesting on the plant body, and almost all mutualistic ants fiercely defend their nests (Beattie 1985). Approximately 400 species in the tropics have been identified as myrmecophytes, and they have been thought to have evolved independently in many plant lineages (Davidson and McKey 1993).

Most myrmecophyte domatia develop on vegetative organs such as petioles, stems, leaf bases, and roots (Beattie 1985). Recently, ants nesting in the propagating organs have been reported on Mucuna interrupta (Leguminosae) (Kato et al. 2004). Although fruit-nesting ants may have different protective effects on host plants to ants that nest on plant vegetative organs, little is known about their symbiotic relationships. We recently found a similar example of an ant nest in the fruits of Canavalia lineata and C. cathartica (Leguminosae). Both Canavalia lineata and C. cathartica also have EFNs on their inflorescence, and we hypothesize that, as common elsewhere, nectar is provided as a reward for ant-guards. We address herein the following two questions (1) How many ant species utilize fruits of these two Canavalia species as nesting sites? (2) What specific expectations are associated with the interaction between two Canavalia species and ants?

Materials and methods

The genus Canavalia comprises approximately 50 species distributed in both the new and old world tropics (Sauer 1964). In Japan, three species, Canavalia lineata, C. cathartica and C. rosea are known to occur (Ohashi 2001). Of these species, C. lineata and C. cathartica develop cavities in the fruits and ant nests were observed in the course of our study.

Canavalia lineata is a perennial vine distributed in Japan, Taiwan, China, and Indochina (Ohashi 2001). This species occurs on sandy or rocky beaches, seaside cliffs, or the margins of thickets facing the sea in the region from the Boso peninsula to the Ryukyu Islands in Japan (Fig. 1). The flowering time of the species is from June to September. EFNs are formed on swollen mounds on the axis of the pseudoraceme (McKey 1989) (Fig. 2). The fruits are flat-ellipsoid, 9–11 cm long, 2–3 cm wide, and indehiscent or tardily dehiscent (Fig. 2) (Sauer 1964).
https://static-content.springer.com/image/art%3A10.1007%2Fs11829-008-9030-y/MediaObjects/11829_2008_9030_Fig1_HTML.gif
Fig. 1

Geographic distribution of sampling sites. Close circle indicates Canavalia lineata and open circle indicates C. cathaltica. The number in parentheses indicates the number of inflorescence and fruit

https://static-content.springer.com/image/art%3A10.1007%2Fs11829-008-9030-y/MediaObjects/11829_2008_9030_Fig2_HTML.jpg
Fig. 2

(a) Flower of Canavalia lineata, (b) flower of C. cathartica, (c) nectar (arrow) secretion from EFNs on inflorescence of C. cathartica, (d) pods of C. lineata (left-hand) and C. cathartica (right-hand), (e) inside of the fruit of C. lineata nested by Ochetellus glaber (Dolichoderinae)

Canavalia cathartica is distributed in Japan, Taiwan, Indochina, Malesia, Polynesia, Australia and East Africa, and occurs primarily in coastal thickets (Ohashi 2001). The morphology of the inflorescence is similar to that of C. lineata. This species flowers throughout the year in the tropics and primarily from May to September in the subtropical region of Japan. The fruits are wider than those of C. lineata, 7–10 cm long, 3–4 cm wide, and indehiscent or tardily dehiscent (Fig. 2) (Sauer 1964).

During December 2005 and November 2006, we collected fruits of C. lineata from 20 sites and those of C. cathartica from 10 sites (Fig. 1). Fruits were examined for the presence or absence of ants inside, and nesting ants were killed with chloroform if they were present. Identification of nesting ants and observation of their colony constitution were conducted under a dissecting microscope. To investigate the invasion characteristics of ant species in fruit, we examined the morphology of the pericarps of ant-nested fruits. We calculated the percentage of damaged fruit per individual plant. The difference in the percentage of damaged fruit for each EFN-visiting or fruit-nesting ant was compared using a Kruskal–Wallis test. The ants visiting EFNs were collected at 24 of the 30 sites (Fig. 1). EFNs visited by ants were collected for up to 12 inflorescences per plant individual, and they were fixed with 70% ethyl alcohol. Collected ants were identified under a dissecting microscope.

To examine the feeding behavior of nesting ants, fruits containing the nests of four ant species, Camponotus nawai (L4 and L7), C. bishamon (L20), Ochetellus glaber (L4, L8 and C8), and Technomyrmex albipes (C6 and C10) were collected and carried back to the laboratory with a plastic bottle. We provided sugar water to the ant colonies until the forage experiments were conducted. In the course of our research, we found larvae of two moth species, Cryptophlebia ombrodelta and C. repletana (Torticidae) that were dominant seed feeders of two Canavalia species. The moth larvae were collected from the same sites where ant colonies were collected and used as bait. Larvae (10–13 mm in length) were placed into the plastic bottles where the ant colonies were kept (20–25°C), and ant-foraging behavior was monitored for two weeks in a laboratory.

Results and discussion

The characteristics of fruit on C. lineata and C. cathartica

In C. lineata, we collected a total of 3075 fruits from 84 plants and found 131 ant-nested fruits from 49 plants (4.2% and 58.3% per fruit and plant, respectively). Five ant species belonging to three genera of three families, Ochetellus glaber (Dolichoderinae), Camponotus bishamon (Formicinae), Camponotus vitiosus (Formicinae), Camponotus nawai (Formicinae), and Crematogaster nawai (Myrmicinae), utilized the fruits of C. lineata as a nest (Appendix). Ochetellus glaber occupied 73 (55.7%) of the 131 ant-nested fruits (Appendix, Fig. 2). In C. cathartica, we collected a total of 860 fruits from 45 plants and found 76 ant-nested fruits from 23 plants (8.8 and 51.1% per fruit and plant, respectively) (Appendix). Eight nesting ant species belonging to eight genera in three families, Technomyrmex albipes (Dolichoderinae), Camponotus vitiosus (Formicinae), Tapinoma sp. (referred to as Tapinoma sp. 3 in JADG 2003) (Dolichoderinae), Ochetellus glaber (Dolichoderinae), Monomorium floricola (Myrmicinae), Tetramorium bicarinatum (Myrmicinae), Pheidole megacephala (Myrmicinae), and Cardiocondyla yamauchii (Myrmicinae), were found. Technomyrmex albipes occupied 27 (35.5%) of the 76 ant-nested fruits. Three ant species, Camponotus bishamon, Ochetellus glaber, and Technomyrmex albipes, were common.

The pericarps of C. lineata and C. cathartica were thick and lignified after maturing. We did not observe chewing or drilling behaviors in ants, which were observed in myrmecodomatia such as those studied on neotropical acacias (Janzen 1966, 1967), African acacias (Hocking 1970), and Macaranga species (Beattie 1985) on both immature and mature fruits of C. lineata and C.cathartica. We found two possible invasion patterns of nesting ants in the fruits of C. lineata and C. cathartica: (1) from a crack in the pericarps, and (2) from holes excavated by moth larvae such as Cryptophlebia ombrodelta and C. repletana (Appendix). While the majority of fruit nests of Ochetellus glaber, Camponotus bishamon, and C. vitiosus existed in the holes excavated by moth larvae in the pericarps, those of three ants species, Crematogaster nawai, Technomyrmex albipes, and Tetramorium bicarinatum, existed in the cracks in the pericarps. In C. cathartica and C. lineata, when moth larvae invaded non ant-nesting fruits, 6.1% (23 of 375 fruits) and 7.9% (13 of 148 fruits) of fruits, respectively, had more than one intact mature seed. In contrast, rather large number of the fruits contained intact mature seeds when ants were nesting in the fruits damaged by moth larvae (Appendix). In a laboratory experiment, we observed that the moth larvae were killed by workers of four ant species, Camponotus bishamon, C. vitiosus, Ochetellus glaber, and Technomyrmex albipes (Table 1). Therefore, the fruit-nesting ants may kill seed herbivores and defend the seeds from these endophagous seed-feeders.
Table 1

Foraging behavior of fruit nesting ants

Plant species

Site

No. of colony

Moth species*

No. of larvae fed

No. of larvae killed

    Ant species

C. lineata

    Camponotus bishamon

L20

2

R

2

2

    C. vitiosus

L4

5

O

7

7

L7

7

O

12

12

    Ochetellus glaber

L4

4

O

6

3

L8

3

O

6

5

L16

2

R

2

2

C. cathartica

    Camponotus bishamon

C6

2

R

3

3

    Ochetellus glaber

C8

2

R

3

3

    Technomyrmex albipes

C6

4

R

7

6

C10

1

R

1

0

* O and R indicate Crytophlebia ombrodelta and C. repletana, respectively

Stem domatia on myrmecophytes have been hypothesized to originate from the secondary ant occupation of cavities on stems made by wood-boring insects (Ward 1991; Davidson and McKey 1993). In such relationships, ants have been considered to protect host plants from wood-boring insects (Ward 1991). Although the cases observed here were of ant nests in woody fruits, our observations give plausibility to this hypothesis regarding the origins of stem domatia.

EFN, fruit utilization and ant-guards

In C. lineata, during the secretion by EFNs from May to September, 14 ant species utilized EFNs (Appendix). Of these ants species, Crematogaster nawai and Pheidole megacephala were found to be highest in the frequency of EFN visitations per plant individual, at sites L1–14 and sites L15–20, respectively (Appendix). By contrast, EFN visitations by Ochetellus glaber and Camponotus bishamon, who frequently nest in the fruits, were low (5.8 and 0.3%, respectively). In C. lineata, three nesting ants, Camponotus bishamon, C. vitiosus, and Ochetellus glaber, became dominant on the plant after the EFN’s occupant species, Crematogaster nawai and Pheidole megacephala, decreased at the end of the flowering season. These ant species began to nest in the fruits from autumn until the following spring, when most of the fruits were dehisced or decayed.

In C. cathartica, nine ant species aggregated on EFNs, of which Technomyrmex albipes and Tetramorium bicarinatum were the most frequently found. These two ant species simultaneously utilized both EFNs and fruits on the same plant. In particular, Technomyrmex albipes was found to be the dominant ant on both EFNs and fruits of C. cathartica, compared with other ant species. Rates of EFN visitation by other nesting ant species, Camponotus bishamon, Ochetellus glaber, Tapinoma sp., and Cardiocondyla yamauchii, were very low (0.5, 0.0, 0.0, and 0.0%, respectively).

Apparently, many ant species were attracted by EFNs of both C. lineata and C. cathartica. However, their protection effects for seeds were different. In Crematogaster nawai on C. lineata and Technomyrmex albipes on C. cathartica, the percentage of damaged fruits was significantly lower than those of the other nesting ants when they utilized both EFNs and fruits (Fig. 3, P < 0.005, Kruskal–Wallis test). In contrast, three Camponotus ants and Ochetellusglaber showed a high percentage of damaged fruits (Fig. 3), suggesting that they have non-protective effects. Our observations are similar to the report that Camponotus planatus workers are never seen removing eggs or disrupting feeding phytophagous insect on Acacia mayana (Raine et al. 2004). A non-protective role for these three Camponotus species is further suggested by the fact that they colonise fruits at the beginning of winter, after the fruits have matured. We suggest that these ants use the fruits as hibernation sites.
https://static-content.springer.com/image/art%3A10.1007%2Fs11829-008-9030-y/MediaObjects/11829_2008_9030_Fig3_HTML.gif
Fig. 3

Percentage of damaged fruits (mean ± 1 SE). Number of plants examined is shown at the top of each bar. * Represents significant difference (P < 0.005, Kruskal–Wallis test)

Two possible interactions exist between ants and two Canavalia species; one is opportunistic utilization of EFNs or fruit as nesting sites such as with three Camponotus species, Ochetellus glaber, Cardiocondyla yamauchii, and Pheidole megacephala. The others involve specialized EFNs and fruit utilization, such as with Crematogaster nawai in C. lineata and Technomyrmex albipesin C. cathartica.

Effects of temporal matching of EFN secretion and fruit production

Our results suggest that Crematogaster nawai and Technomyrmex albipes offer strong protective effects for C. lineata and C. cathartica, respectively. However, the utilization of Crematogaster nawai on C. lineata differs from that of the other ant species, in that they utilize the remaining fruits of the plant at flowering time, and their colonies are composed only of workers, while those of the other ants are composed of workers, eggs, larvae, and pupae (Table 2). We observed that Crematogaster nawai frequently built shelters around the EFNs (their nests are usually found under a stone or inside of dead twigs). These shelters are chiefly made of plant debris and fibers, and are considered to protect nectaries from the sun (Jolivet 1996). It seems that Crematogaster nawai utilizes the fruits of C. lineata as satellite stations to enhance their utilization of EFNs. However, the frequency of the utilization of fruits as satellite stations by Crematogaster nawai on C. lineata was found to be low. In many sites, the frequency of EFN visitation by Crematogaster nawai was also decreased with fruit maturation as was also observed in other ant species.
Table 2

Colony composition of the ants nesting on the fruit of C. lineata and C. cathartica

Species

No. of fruit

Queen

Late female

Male

Worker

Egg and Larva

Pupa

Camponotus bishamon

16

0

0

0

8–254

0–483

0–25

C. vitiosus

3

0

0

0

1–69

0–80

0

C. nawai

1

0

0

0

90

0

48

Crematogaster nawai

8

0–1

0–2

0–57

24–71

0

0

Monomorium floricola

2

1–4

0

0

39–143

16–201

9–10

Ochetellus glaber

45

0–4

0–5

0–92

22–810

1–1052

0–131

Tapinoma sp. 3

7

1–15

0

0

88–240

0–29

8–43

Technomyrmex albipes

14

0

0

0

38–965

0–284

0–25

Tetramorium bicarinatum

2

0–1

0

0

25–63

0

0

With regard to Technomyrmex albipes on C. cathartica, its colonies were found to be composed of workers, eggs, larvae, and pupae during year-round nesting. This ant species has been known to inhabit caulinary domatia on Humboldtia brunonis (Leguminosae) in western India and to provide efficient anti-herbivore protection for host plants (Gaume et. al. 2005). Therefore, EFN and fruit utilization by Technomyrmex albipes is expected to result in higher anti-herbivore protection than that provided by other ants utilizing EFNs and fruits of C. cathartica.

In both C. lineata and C. cathartica, initial ant recruitment seems to depend solely on EFN activity. Therefore, the longevity of EFN secretion becomes a crucial factor in the construction of the ant-guard system. Canavalia lineata flowers once a year, and EFN activity decreases as the fruits mature. Thus, EFN are not secreted during the period of ant nesting in fruits. In this species, opportunistically recruited ants such as the Camponotus species may utilize fruits as hibernation sites. In contrast, C. cathartica flowers several times a year and even if older inflorescences reduce EFN activity as their fruits mature, EFNs may become active on other young inflorescences. Canavalia cathartica is able to continuously attract the specific guarded ants by providing their nesting sites. Plants offering year-round nesting sites are known to attract more stable populations of ants (Carroll and Janzen 1973). Furthermore, the provision of year-round housing, nectar, and food bodies is considered to bring about the highest defensive effect for host plants (Beattie 1985). Therefore, C. cathartica, which exhibits temporal matching of EFN activity and fruit production, may receive a higher defensive effect from ants than C. lineata.

Acknowledgements

The authors are grateful to Drs. Y. Tateishi and M. Maki for their useful comments. This work was partly supported by the grant from Sasagawa Science foundation to AY.

Copyright information

© Springer Science+Business Media B.V. 2008