Rafflesia (Rafflesiaceae) is a genus of parasitic plants that is endemic to Southeast Asia. Many species in this genus are of high conservation concern in many countries due to their rare and threatened status. However, the ecology of Rafflesia species is not yet fully understood, particularly in terms of flower predation. Here, we present new observations of predation on a Rafflesia patma flower bud using a camera-trap study, together with the results of a survey of flower damage in the three Rafflesia species that occur on Java, Indonesia. We found that two small mammals predated a flower bud of R. patma in Pangandaran (West Java, Indonesia), one of which has not previously been identified as a predator of Rafflesia flowers. We also detected wasp infection in Rafflesia rochussenii and predation by another animal on Rafflesia zollingeriana. Overall, flower bud predation by animals damaged up to 10.65% of the surveyed flowers.
Introduction
Plants in the genus Rafflesia (Rafflesiaceae) are not only famous among botanists and ecologists but are also attractive to the public due to their exceptionally large flowers that smell unpleasant. Rafflesia species are endoparasitic, infecting vines of members of the grape family (Vitaceae), in the genus Tetrastigma (Meijer, 1997). The plants are only visible during the generative stage, being entirely contained inside the host during the vegetative stage (Nikolov & Davis, 2017). This genus, which is endemic to Southeast Asia (Meijer, 1997), split from Rhizanthes approximately 74 MyBP and experienced rapid species diversification between 12 and 5 MyBP (Bendiksby et al., 2010).
There are three known species of Rafflesia in Java, Indonesia: Rafflesia patma Blume, Rafflesia zollingeriana Koords., and Rafflesia rochussenii Teijm. & Binn. (Jafarsidik & Meijer, 1983). Although The Plant List (2013) treated R. zollingeriana as a synonym of R. patma and adopted R. horsfieldi R. Br. as the accepted name, we have retained these as separate species in this study based on their different morphological characteristics (Nais, 2001) and phylogenetic analysis (Bendiksby et al., 2010). R. patma is mainly distributed in the central to western part of Java, particularly in lowland forests (Meijer, 1997), while R. zollingeriana specifically occurs in lowland forests on the southern coast of East Java (Jafarsidik & Meijer, 1983). By contrast, R. rochussenii is the only Rafflesia species in Java that is confined to mountainous areas at altitudes >1,000 m above sea level (a.s.l.) and is restricted to West Java (Meijer, 1997). All three species are currently threatened with extinction, primarily due to forest degradation and habitat loss (Mursidawati et al., 2015).
Rafflesia plants are known to have unisexual flowers (Nikolov & Davis, 2017), so pollinators play a significant role in successful reproduction, with fruit set ranging from 0% to 50% (Nais, 2001). Plants in this genus also have a long flower development time of up to 4 years (Hidayati & Walck, 2016). Therefore, any disturbance to the reproductive process will interfere with the regeneration cycle of these species.
Flower predation is defined as any type of damage that is caused by consumption of any part of the flower, including the bracts, sepals, petals, stamens, pistils, pollen, or ovules (McCall & Irwin, 2006) and can be detrimental to the fitness of a plant population (Irwin, Brody, & Waser, 2001; McCall & Irwin, 2006; Washitani, Okayama, Sato, Takahashi, & Ohgushi, 1996). Although there have been few reports of flower predation in Rafflesia species (e.g., see Hidayati, Meijer, Baskin & Walck, 2000; Nais, 2001; and references therein), it is important to their ecology because, when it does occur, the impact can be severe. Therefore, in this study, we investigated flower predation in the three Javanese Rafflesia species and used the findings to gain an insight into its consequences for the conservation of these species.
Methods
Study Area
Java is one of the largest islands in Indonesia, with an area of 132,000 km2 (Whitten, Soeriaatmadja, & Afiff, 1996). It has been estimated that 0.8 million hectares of primary forest remain on Java (Margono, Potapov, Turubanova, Stolle, & Hansen, 2014) and there are more than 80 conservation areas (Ministry of Environment and Forestry, 2016). We selected six conservation areas on Java as study sites: Bojong Larang Jayanti Nature Reserve (NR), Cipeucang Protected Forest, Gunung Gede Pangrango National Park (NP), Leuweung Sancang NR, Meru Betiri NP, and Pangandaran NR (Figure 1).
In general, Java has an annual rainfall of approximately 1,500 mm per year with a distinct wet and dry season, a minimum temperature of 22°C to 24°C, and a maximum temperature of 31°C to 33°C. Tropical lowland and montane forests are the dominant habitat (Whitten et al., 1996).
Camera Trap
We set up a camera trap (Bushnell Trophy Cam HD; Bushnell Outdoor Products, Overland Park, Kansas) in the center of the forest in Pangandaran NR (7°43′13.18″S, 108°39′46.87″E) approximately 1.5 m from an R. patma flower bud. The camera was operated continuously for 22 days from September 3 to September 25, 2014, and was automated using a passive infrared sensor, which triggered recording for 30 s upon the detection of movement. The camera was also programmed to automatically record using night vision during the nighttime.
Population Survey and Flower Damage Inventory
From March 2017 to October 2017, we visited the six conservation areas and randomly searched for Tetrastigma species and the three Rafflesia species. All Rafflesia plants encountered were recorded and enumerated. We also carefully examined each living flower for the presence of damage caused by animals.
Results
A total of 51 videos were recorded, 15 of which captured sightings of wild animals that were attracted to the R. patma flower bud (Figure 2(a)). A total of six species were recorded: the Asian palm civet (Paradoxurus hermaphroditus), the Javan langur (Trachypithecus auratus), the Java mouse deer (Tragulus javanicus), the long-tailed monkey (Macaca fascicularis), the Sunda pangolin (Manis javanica), and the Sunda porcupine (Hystrix javanica). Eleven of these videos were recorded during the nighttime, and two of these videos showed animals exhibiting direct predatory activities (Table 1): The first video (Figure 2(b), Online Supplementary Movie S1) showed a Sunda porcupine was observed digging the ground and foraging on the flower bud, while the second video (Figure 2(c), Online Supplementary Movie S2) recorded a Java mouse deer consuming the flower parts of R. patma (Figure 2(d)).
Table 1.
List of Animals Captured With Camera Trap.
During the field survey, we discovered that flower bud predation also occurred in the other two species of Rafflesia on Java. We observed wasp infection of R. rochussenii flower buds and signs of animal predation for instance scattered flower parts of R. zollingeriana. In total, up to 10.65% of the flowers (18 of the 169 living buds and blooms surveyed) were damaged by animal predation, which was observed at three of the six study sites, namely, Gunung Gede Pangrango NP, Meru Betiri NP, and Pangandaran NR.
Discussion
Flower predation has not been well studied in Rafflesia spp., partly due to the suspected predators being cryptic mammals that are easily alarmed by humans. One study directly observed treeshrews (Tupaia tana) and squirrels (Callosciurus notatus) licking and scooping out the pulp of R. keithii fruit in Borneo (Emmons, Nais, & Briun, 1991), while another, which presumably used indirect observation, reported that treeshrews (Tupaia javanica) and Sunda porcupines consumed R. patma in Pangandaran, Java (Hidayati et al., 2000). In addition, Hidayati et al. (2000) suspected that some other animals, such as Asian leopard cats (Felis bengalensis), wild boars (Sus scrofa), Sunda sambar (Cervus timorensis), barking deer (Muntiacus muntjak), banteng (Bos javanicus), and crows (Corvus spp.), also damage the flower buds of R. patma at the same site, while Nais (2001) suggested that long-tailed mountain rats (Niviventer rapit) predate the flower buds of Rafflesia species in Borneo.
The use of a camera trap in this study verified some of these reports and also showed that the Sunda porcupine and Java mouse deer predated a flower bud of R. patma, representing the first report of Rafflesia flower predation by the Java mouse deer. Sunda porcupines and Java mouse deer are considered to be predominantly frugivorous, consuming mainly fruits (Corlett, 1998). However, this study showed that these mammals are also florivorous on Rafflesia species, particularly R. patma.
Flower bud predation can cause minor impairment to severe damage to the flower. We found that mammalian predation caused severe damage to small Rafflesia flower buds in the early growth stage. Indeed, in some cases, animals almost entirely consumed the flower buds of R. patma and R. zollingeriana, leaving only a few pieces of flower debris (Figure 3(a) and (b)). By contrast, damage to the blooms was very rare and often only caused minor damage, having no significant effect on the blooming stage (Figure 3(d)). However, this was also very serious in some cases (Figure 3(c)). Flower predation by insects, such as the wasp infection we observed in R. rochussenii flowers in Gunung Gede Pangrango NP, can also be detrimental to the flower buds. Although we only detected this in one flower bud out of the 54 living flowers surveyed, Nais (2001) showed that approximately 6.7% of all bud mortality in Rafflesia species in Sabah (Borneo) was associated with wasp infection.
Flower predation during the bud stage is expected to have direct effects on the flowers of Rafflesia species (McCall & Irwin, 2006), as the animals either consume the whole flower or severely damage all of the reproductive parts, including the pollen and ovules. In addition, flower predation may have indirect effects on the plants (McCall & Irwin, 2006), such as decreasing the pollinator visitation rate through flower damage, although this has not been confirmed. Both direct and indirect effects can negatively impact on the maternal fitness of Rafflesia species (Kelly, Ladley, Robertson, & Crowfoot, 2008). The flower mortality rate in Rafflesia has been estimated to be between 50% and 90% and is mostly attributed to flower bud predation (Nais, 2001), indicating that flower predation is one of the most important factors that affects the persistence of Rafflesia populations in the wild. Therefore, it is crucial that predator populations are considered when developing conservation strategies for Rafflesia species.
Implications for Conservation
The predation of Rafflesia flowers during their development can affect the reproductive success of the plants, particularly in terms of the blooming process, pollination rate, and fruit set. Furthermore, it may have negative consequences on the regeneration of Rafflesia populations in the wild. In this study, we observed flower predation on all three of the species of Rafflesia that occur on Java. Therefore, comprehensive and continuous monitoring of Rafflesia populations and any potential predators is necessary. In addition, further research is required to explore the consequences of flower predation on Rafflesia species to improve our understanding of the ecology of this genus and to develop appropriate management strategies.
Acknowledgments
We thank all staffs at Bojonglarang Jayanti Nature Reserve, Gunung Gede National Park, Leuweung Sancang Nature Reserve, Meru Betiri National Park, and Pananjung Pangandaran Nature Reserve. We also thank Yoshikawa san for valuable comments on earlier version of the manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by JASTIP (The Japan-ASEAN Science, Technology and Innovation Platform). Research permit was granted from the Ministry of Environment and Forestry, Republic of Indonesia (SK.140/KSDAE/SET/KSA.2/3/2017).