Portal de Periódicos da CAPES

Neglected plant parasites: Mitrastemonaceae

2022; Wiley; Volume: 5; Issue: 1 Linguagem: Inglês

10.1002/ppp3.10322

2572-2611

Luiza Teixeira‐Costa, Kenji Suetsugu,

Plant and animal studies

Mitrastemonaceae are a remarkable family of parasitic flowering plants. The two species that make up this family have a disjunct distribution and a peculiar appearance. Both are devoid of roots, stems, and regular leaves and live inside the roots of their host plants for most of their life cycle. Their white-pink flowers and berry-like fruits become visible on the forest floor only during their reproductive phase. This cryptic growth form, allied to their patchy distribution, and the muted color of their flowers have contributed to making Mitrastemonaceae one of the least collected and analyzed families of parasitic plants. On the other hand, recent findings regarding the phylogenetic position of the family have advanced our understanding of the family's biogeography and prompted a reexamination of their ecology and evolution. Here we review key aspects of the life cycle, reproductive ecology, and host relationships in the family Mitrastemonaceae to identify gaps in our current knowledge and instigate future research. Mitrastemonaceae are a small, yet remarkable family of parasitic flowering plants. They are composed of a single genus, Mitrastemon Makino (sometimes misspelled as "Mitrastema" or "Mitrastemma"), for which a total of six taxa were described since the early 1900s (Meijer & Veldkamp, 1993). Most of these taxa have been synonymized and at present only two species are recognized: Mitrastemon matudae Yamam., known from Mexico, Central America, and Colombia; and Mitrastemon yamamotoi Makino, found across Southeast Asia, Taiwan, Japan, and parts of India (Nickrent, 2020). Both species are devoid of roots, stems, and regular leaves. They appear as cream-colored to whitish-pink, solitary, hermaphrodite flowers that form aggregated patches on the forest floor (Figure 1a–d). Their vegetative body is reduced to an endophyte ("endo" = internal + "phyte" = plant) that spreads within the roots of host trees, from which these achlorophyllous plants derive all their nutrients (Watanabe, 1936a). This cryptic growth form in which the parasitic plant is entirely hidden within host tissues lasts for most of these plants' life cycle. Allied to this peculiar lifestyle, the muted colors and small flower size ranging from 3 to 5 cm tall (Makino, 1911; Meijer & Veldkamp, 1993) make these plants easy to overlook among the leaf litter (Jochems, 1928). This may explain the small number of herbarium records (<150 specimens) available online for both M. matudae and M. yamamotoi in collections worldwide (Teixeira-Costa et al., 2022). In turn, this scarcity of records could be why populations appear to be fragmented across the disjunct geographical distribution of the family (Figure 1e). Difficulties in collecting and interpreting the morphology and development of Mistrastemon also led to long-standing challenges in resolving the phylogenetic position of the family. Associations with other parasitic plant families, such as Rafflesiaceae and Hydnoraceae, and even an affiliation with the Nepenthaceae, a tropical family of carnivorous pitcher plants, have been suggested (Makino, 1911). It was only in recent decades that Mitrastemonaceae were placed with confidence in the Ericales (Barkman et al., 2004; Nickrent et al., 2004), an order including commercially important species such as tea, persimmon, blueberry, and azalea. Later, Mitrastemonaceae were resolved as sister to the Lecythidaceae, the Brazil nut family, suggesting an ancient Neotropical origin for the clade (Rose et al., 2018). These recent findings have advanced our understanding of historical biogeography for the family and prompted a reexamination of the ecology of Mitrastemonaceae. Here we review key aspects of the life cycle, reproductive ecology, and host relationships in Mitrastemonaceae to identify gaps in our current knowledge about the biology of one of the most overlooked families of parasitic flowering plants. The lifestyle in parasitic plants like Mitrastemonaceae, known as endoparasites, is associated with a series of developmental adaptations. In many parasitic plants, the absence of one or more organs is caused by processes of reduction or abortion at some stage during the course of development. In Cuscuta (Convolvulaceae) and Cassytha (Lauraceae), for instance, the rudimentary root system withers and dies once host attachment has been secured (Weber, 1981; Yuncker, 1932). On the other hand, roots, shoots, and expanded leaves are entirely absent in Mitrastemonaceae. Unlike other achlorophyllous root parasites, such as those in the family Balanophoraceae, no underground organs such as tuber or rhizomes are formed. The peculiar body plan of Mitrastemonaceae plants is likely to be due to abortion of the embryonic shoot apex soon after germination (Teixeira-Costa & Davis, 2021). This hypothesis is based on the biology of other endoparasitic plants, such as Viscum minimum (Viscaceae; Kuijt, 1986) and Tristerix aphyllus (Loranthaceae; Mauseth et al., 1985). Still, further confirmation is required for Mitrastemonaceae, as the initial phases in the life cycle of these plants have yet to be observed and described. It is only at the onset of their reproductive development that Mitrastemon plants become externally visible (Figure 2a). This reproductive phase follows a long vegetative growth period that extends for 3 to 4 years (Watanabe, 1933a). During this initial period, the body of the parasite consists of a network of parenchymatous threads that grows exclusively within the host roots, ultimately colonizing the host bark (Figure 2b,c). This reduced vegetative body later differentiates vessel elements that establish vascular connections with the host wood (Figure 2d; Teixeira-Costa et al., 2021). These direct, vessel-to-vessel connections provide an abundant supply of water, nutrients, and host-derived chemicals, which, in turn, may allow the parasite to undergo faster cell proliferation and differentiation. As flower buds develop, they appear on the host root as small knobs, which continue to grow for approximately 5 months until they eventually break through the host bark and become exposed (Watanabe, 1936b). Multiple flower buds can sprout in proximity (Figure 2e) due to the frequent anastomosing of individual endophyte clusters (Figure 2f; Watanabe, 1936a, 1936b). Nevertheless, not all flower buds mature simultaneously, and neighboring host roots often bear plants at different developmental stages (Jochems, 1928). This strategy might contribute to offering resources to pollinators and seed dispersers over a longer period. It is noteworthy that the extensive proliferation of parasite endophytic tissue within the host bark can lead to severe modifications of the root structure. In infested hosts, high vessel density and reduced vessel lumen size is apparent, especially in the area promoting an intersection between the parasite flower buds and the host wood (Teixeira-Costa et al., 2021). An increase in vessel proliferation could then induce the initiation of multiple lateral roots (Aloni, 2021), leading to a bifurcated branching pattern with roots that run closer to soil surface (Suetsugu & Yamashita, in press; Watanabe, 1933b, 1934a). In other host–parasite systems, structural modifications have been related to changes in host cambial activity promoted by the parasite (Aloni, 2015; Spallek et al., 2017; Teixeira-Costa & Ceccantini, 2015), which appears to be a common feature of host–parasite interactions (Teixeira-Costa, 2021). Despite of these structural alterations, no significant pathogenicity is reported in the association between Mitrastemon species and their hosts. Moreover, considering the slow and prolonged life cycle of these parasites, it is unlikely that severe damaged is caused to the host. The flowers of both species of Mitrastemonaceae are similar regarding their morphology and development. Differences between the two species relate mainly to size, with M. matudae typically having longer and wider flowers (Meijer & Veldkamp, 1993). Additionally, differences between the two recognized varieties, M. yamamotoi Makino var. yamamotoi and M. yamamotoi var. kanehirai (Yamam.) Makino are attributed to the number of bracts per series (Flora of China Editorial Committee, 2003; Watanabe, 1934b), although intermediary forms have been described for the same area with no clear correlation to host identity or plant size (Meijer & Veldkamp, 1993). Like many plants in the Ericales, the androecium is particularly distinctive (Chartier et al., 2017). In both Mitrastemon species, flowers are protandrous, meaning that the androecium mature before the gynoecium. During the initial male phase, the connate stamens form a miter-shaped tube (Figure 3a), which caps the style and stigma (Nickrent et al., 2004). This characteristic staminal tube gives the genus its name (Meijer & Veldkamp, 1993). As development progresses, the flower enters an intermediary stage marked by the circumscissile rupture of the deciduous tube (Figure 3b). This leads to the female phase of the flower, as the dehiscence of the androecial tube reveals the mature gynoecium (Figure 3c). Nectaries can be found across the flower during both stages, being present on the ovary wall, on outer surface of the tepal tube, and at the base of the uppermost pair of scale leaves, which store most of the nectar due to their cistern-like shape (Watanabe, 1934c). A large volume of nectar is produced (~600 μl in 20 h per flower, according to Watanabe, 1934b), which attracts a variety of animals. Atypical flower morphology, combined with the non-photosynthetic habit, which allows Mitrastemon to thrive in the shaded forest floor, suggests that these species may rely on unusual pollinators. Indeed, the dark environment inhabited by these plants is rarely visited by typical pollinators such as bees and butterflies (Suetsugu, 2015, 2018). Nevertheless, a wide diversity of animals ranging from flies (Drosophila sp. and Chrysomyia sp.) to a bee (Apis sp.) and a bird species (Zosterops sp.) have been occasionally reported to forage for nectar among M. yamamotoi flowers (Matuda, 1947; Watanabe, 1934a). It was only recently that Suetsugu (2019) unequivocally showed that hornets (Figure 3d), crickets (Figure 3e), and cockroaches (Figure 3f) effectively pollinate Mitrastemon. The investigated population of M. yamamotoi in southern Japan was observed to be diurnally pollinated by hornets (Vespa sp. and Vespula sp.), while cockroaches (Opisthoplatia sp. and Onychostylus sp.) and a cricket species (Diestrammena sp.) were among the most frequent nocturnal floral visitors, carrying pollen grains on their legs, heads, and mouthparts (Suetsugu, 2019). Suetsugu (2019) also showed that M. yamamotoi is a self-compatible species; crickets and cockroaches contribute to its geitonogamous pollination, whereas hornets facilitate outcrossing. No data on the pollination of M. matudae appear to be available. Contrary to anecdotal reports (e.g., Matuda, 1947), birds were rarely seen visiting M. yamamotoi flowers for nectar (Suetsugu, 2019). On the other hand, species such as Zosterops palpebrosus have frequently been suggested to forage the berry-like fruits of Mitrastemon species (Figure 3g; Watanabe, 1936c). Indeed, several of the minute Mitrastemon seeds have been observed in bird droppings (Figure 3h,i), indicating that birds do act as seed dispersers at least for M. yamamotoi (Suetsugu, unpublished data). Regarding seed fate, initial research by Watanabe (1933a) remarked on the low seed germination rate of M. yamamotoi and suggested asexual reproduction should play a larger role in the effective propagation at a local scale. Years later, this hypothesis was strengthened after a series of careful germination experiments yielded no positive results (Watanabe, 1936c, 1937a). In some parasitic plants, Mitrastemon species included, mycobionts have been suggested to play a role in the process of germination (Baird & Riopel, 1986; de Vega et al., 2010; Watanabe, 1936c; Wicaksono et al., 2021; Williams & Zuck, 1986). If so, both root exudates from Fagaceae hosts, as well as fungal-specific molecules might be necessary as germination cues. These hypotheses remain untested, however, and no record of successful ex situ germination is available for Mitrastemon species (Baskin & Baskin, 2022). Mitrastemonaceae are distributed across a wide geographical area, occurring in tropical and subtropical forests with equatorial or warm-temperate climates with fully humid to winter-dry precipitation regimes (FAO, 2020; Kottek et al., 2006). In populations located in areas with more seasonality, Mitrastemon is generally observed to flower during the dry season (Meijer & Veldkamp, 1993). However, these observations lack quantitative details and often contradict each other. In the same locality in India, for instance, M. yamamotoi is recorded to have been collected in different seasons (Rao & Balakrishnan, 1972), or to occur only during the dry season (Mir et al., 2016). In southern Taiwan, M. yamamotoi var. kanehirai is reported to flower from February to April and produce fruits in October (Flora of China Editorial Committee, 2003; Matuda, 1947), while video recordings for the same variety growing in central Taiwan show that flowering occurs from October to November and seed dispersion occurs from early December to February (Jeng et al., 2017). On the other hand, observations from southern Japan appear to be more consistent, indicating that the flowering season extends from late October to December (Matuda, 1947; Suetsugu, 2019; Watanabe, 1936b). Finally, in North Sumatra and in southern Mexico, flowering and seed dispersion are reported to occur all year round for both M. yamamotoi and M. matudae (Jochems, 1928; Matuda, 1947). In at least some of the conflicting cases mentioned above, discrepancies can be related to factors such as differences in elevation or even recent climatic changes. It is, however, more likely that fluctuations in the parasite reproductive phenology are correlated or influenced by host identity and biology. Recently, the parasitic vine Cuscuta australis (Convolvulaceae) was shown to lack autonomous flowering and to use host-derived signals to regulate its own reproduction (Shen et al., 2020). Considering that flowering signals are mobile proteins transported through the phloem (Kondhare et al., 2021), a similar mechanism could have evolved in Mitrastemon. At the same time, asynchronous flower production has also been reported in other endoparasitic species, such as Rafflesia spp. (Rafflesiaceae; Beaman et al., 1988) and Bdallophytum bambusarum (Cytinaceae; García-Franco & Rico-Gray, 1996). In these endoparasites, population genetics revealed that multiple individuals of the same parasite species can infest a single host and that the endophytic tissue of these plants can extend for up to 14 m within host tissues in the case of Rafflesiaceae (Barkman et al., 2017; García-Franco et al., 1998). Therefore, additional factors beyond the influence of host phenology could be involved in controlling flowering time in Mitrastamon and other endoparasites. Host range can be defined as the diversity of species effectively used by a parasitic plant throughout its geographical distribution (Heide-Jørgensen, 2008). For many parasitic species, this topic has received renewed attention in recent decades. Mitrastemon yamamotoi and M. matudae were originally described from Southern Japan and Southern Mexico as parasites growing on Quercus cuspidata and Quercus boqueronae (Fagaceae), respectively (Makino, 1909; Yamamoto, 1936). Years later, as both species were collected in several new localities, M. matudae was consistently observed only on Quercus spp. (Meijer & Veldkamp, 1993) while additional hosts were reported for M. yamamotoi, including Castanopsis, Lithocarpus, and Trigobalanus (Fagaceae; Matuda, 1947; Togashi, 1974; Watanabe, 1937b). Since then, Mitrastemonaceae have been traditionally interpreted to be a family of host-specialist parasites restricted to the Fagaceae. This form of strict host specificity is also reported for other endoparasites, such as Rafflesiaceae and Apodanthaceae (Teixeira-Costa & Davis, 2021). More recently, however, a wide variety of other species have also been claimed to be hosts for M. yamamotoi, including Asteraceae, Fabaceae, and Myrtaceae plants (Amoroso et al., 2018; Mir et al., 2016; Rao & Balakrishnan, 1972). These observations have then raised questions regarding the host range of Mitrastemonaceae. Host specificity is crucial for understanding many aspects of the natural and evolutionary history of parasitic plants, including diversification, historical biogeography, and conservation (Schneider et al., 2016). Therefore, we analyzed herbarium specimens of M. yamamotoi indicated to be growing on Elaeocarpus lanceifolius (Elaeocarpaceae; ASSAM 38224), Eugenia sp. (Myrtaceae; L 1666884), and Vernonia volkameriifolia (Asteraceae; ASSAM 38225). For each specimen, a segment of the host root at some distance of the parasite reproductive structures was sampled, rehydrated, manually sectioned, and prepared for anatomical analysis (for methodological details, see Smith & Smith, 1942; Ruzin, 1999). In all three cases, the structure of the roots was clearly consistent with the wood anatomy of most Quercus and Castanopsis species (Fagaceae), which is remarkably different from the wood anatomy of the plants indicated in the specimen labels. Thus, despite including a limited number of samples, our analysis corroborates the interpretation of Mitrastemonaceae as specific to hosts in the Fagaceae. Future analyses using methods such as DNA barcoding will be especially useful in cases where host identity is uncertain due to insufficient collection of host material (Pelser et al., 2016; Teixeira-Costa et al., 2022). These methods are of increasing importance in the case of parasite species associated with the roots of their host plants, including not only Mitrastemonaceae but also euphytoid or obligate root parasites, for which careful examination and extensive excavation of roots are required for the correct host identification. As shown by the examination of haustorium connections, nearly 40% of the species occurring in close proximity to the euphytoid parasite Thesium chinense (Thesiaceae) did not serve as host plants (Suetsugu et al., 2008). Finally, it is worth noting that while Mitrastemonaceae have been estimated to have diverged from their sister lineage around 104 million years ago (Rose et al., 2018), the stem-age of Fagaceae is estimated at around 99.3 million years (Xiang et al., 2014). Given that branching age estimation is subject to blurring due to age calibration and other factors (Arbogast et al., 2002), it is reasonable to assume that the branching age estimates of the two lineages are in close agreement. In this case, it is possible that the Mitrastemonaceae have evolved in parallel with the Fagaceae since their divergence. On the other hand, this slight discrepancy in the estimated ages of parasite and host lineages might indicate that Mitrastemonaceae might have been associated with other host plants during their evolutionary history. This apparent contradiction has also been observed for other parasite lineages with high host specificity, leading to the hypothesis of temporal specialization, whereby parasitic plants evolved to become more efficient in infesting a narrow range of hosts while losing the genetic information required for successful infestation of rarely encountered host species (Naumann et al., 2013). Host range expansion is also possible, and it has been reported to take place over shorter time frames (Bruce & Gressel, 2013). Continued research and careful analyses are needed, however, to determine if that is the case for Mitrastemonaceae. Furthermore, once the entire genome of Mitrastemon is sequenced and analyzed, past host association will likely be elucidated by detecting any additional horizontal gene transfers (Cai et al., 2021) beyond what has been suggested between Mitrastemon and Fagaceae (Barkman et al., 2007; Nickrent et al., 2004). Several aspects of the biology of Mitrastemonaceae remain unknown. As with most endoparasitic lineages, seed germination and initial host penetration have yet to be documented. Similarly, the processes of nutrient uptake by the parasite during the early stages of its development remain undescribed. Penetration of the host xylem, as well as vessel element differentiation are only observed in the parasite endophytic tissue upon the initiation of flower bud development (Teixeira-Costa et al., 2021). In this scenario, host-derived nutrients are hypothesized to enter the parasite endophyte via either symplast or apoplast pathways. The former implicates the formation of secondary plasmodesmata, complex structures connecting the protoplast of cells derived from different origins. On the other hand, capturing nutrients via the apoplast would require the presence of specialized transmembrane transporters in the cells of the parasite endophyte. Both processes have been documented in the interactions between other parasite species and the phloem of their hosts (Fischer et al., 2021; Péron et al., 2017). Thus, elucidating the resource uptake mechanisms in plants like Mitrastemon will not only advance the knowledge about these peculiar species but also shed light into the dynamics of the intimate association between endoparasitic plants and their hosts. In the context of their reproductive ecology, the interactions between Mitrastemon and its pollinators and seed dispersers also warrant further investigation. For instance, the way in which M. yamamotoi attracts its insect pollinators remains untested. One hypothesis is that as insects like hornets, crickets, and cockroaches usually forage fermented sap, the unique assemblage of M. yamamotoi pollinators could be attracted by volatiles produced by nectarivorous yeasts (Suetsugu, 2019). This would explain the sour, fermented-like smell reported for the nectar of M. yamamotoi flowers (Jochems, 1928). The discovery made by Suetsugu (2019) also suggests that pollination systems involving unusual and unexpected taxa might be more widespread than previously thought, especially in non-photosynthetic plants with highly modified floral morphology. It is noteworthy that reproductive systems can differ among populations, even within the same species (Whitehead et al., 2018), and the population studied by Suetsugu (2019) is near the northern limit of its distribution. Moreover, specific information on pollination and seed dispersal interactions in M. matuade remain unavailable. Therefore, further research is needed to clarify the reproductive biology of the genus Mitrastemon as a whole. Data on interactions between Mitrastemonaceae and non-host species, including mutualists, competitors, and herbivores, is also critical for conservation efforts. Both M. matudae and M. yamamotoi inhabit forest areas considered as global biodiversity hotspots (Figure 1e), but which are also severely threatened by land-use and climate changes, as well as over-exploitation (Dinerstein et al., 1995; Hughes, 2017). In fact, M. yamamotoi has already been reported as threatened in parts of India and China (Botanic Gardens Conservation International, 2022; Mir et al., 2016). Future research focusing on seed dormancy, germination, and initial host penetration in Mitrastemonaceae should provide important information also to ex situ cultivation efforts carried out by institutions with prominent roles in conservation and public engagement, such as botanical gardens (Primack et al., 2021; Thorogood et al., 2022). Herbarium specimens used for the analysis of host specificity in Mitrastemon yamamotoi were provided by the Botanical Survey of India, Eastern Regional Centre (ASSAM 38224, 38225) and the Naturalis Biodiversity Center (L 1666884). The authors are grateful to the curators and staff members at these institutions, as well as to Anthony Brach and Michaela Schmull (Harvard University Herbaria), for their valuable support. LTC acknowledges Alexander Vázquez, Daniel Gómez, Fernando Vázquez, Juan Martínez, Juvenal Hernández, and Martín Castillo who greatly helped collecting Mitrastemon matudae at the Reserva Biológica La Sepultura, in Mexico. Figure 1c is under a Creative Commons license type Attribution Non-Commercial 4.0 International (creativecommons.org/licenses/by-nc/4.0/). LTC was supported by a postdoctoral fellowship from the Harvard University Herbaria, and a junior fellowship from the Hanse Wissenschaftskolleg - Institute for Advanced Studies. KS is supported by PRESTO from Japan Science and Technology Agency (JST) (JPMJPR21D6). The authors have no conflict of interests to disclose. L.T.C. led the writing of the manuscript, prepared the figures, and carried out the analysis of herbarium specimens. K.S. provided most of the photographs, contributed significantly to the development of the manuscript, and wrote most sections on reproduction ecology. Both authors contributed critically to the drafts and gave final approval for publication. Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.