*8.1.2 Lophophytum mirabile*


#### **8.2 Host/parasite relationship**

In the study area, *L. leandri* was found parasitizing exclusively on specimens of *Parapiptadenia rigida* and it was found that it carries out its entire cycle connected to it. Likewise, *L. mirabile* was found parasitizing roots of *Anadenanthera colubrina* var. *cebil*, on which it fulfills its entire life cycle [31]. Both are obligate parasites.

Based on the above descriptions, a schematic representation of the reproductive cycle of the Argentine species of *Lophophytum* has been established. The morphological changes of the structures present in both the pistillate and the staminate flowers are correlated with data on their embryology at progressive stages of development. In **Figure 8** it can be seen that the upper part shows anther and microgametophyte formation. In the central region the ontogeny of the pistillate flower and development of the megagametophyte are represented up to fruit formation. The lower region shows the tuber and inflorescence stages. The vertical lines link the developmental stages between the staminate and pistillate flowers on the same inflorescence. The reproductive cycles of both species are completed in 90 days, developing between the months indicated on the timelines for each species.

*Anatomy, Embryology and Life Cycle of* Lophophytum*, a Root-Holoparasitic Plant DOI: http://dx.doi.org/10.5772/intechopen.99981*

**Figure 8.** *Schematic representation of the reproductive cycle of species of Lophophytum over time.*

The flowering of *L. leandri* is concentrated between July and August (**Figure 8**). During this period, the host is at the end of its winter dormancy, with practically no foliage and with pods still attached to the branches. These pods open and drop seeds simultaneously with the flowering of *L. leandri*. Seeds of *P. rigida* were found germinating among mature inflorescences already fruiting and with seeds of *L. leandri*, which are generally distributed at the foot of *P. rigida* trees. The seeds of *P. rigida* germinate immediately upon dispersal; they have no dormancy period. In mid-November *P. rigida* resumes vegetative growth, by which time the inflorescences of *L. leandri* are completely disintegrated. The weather throughout this period is humid and conducive to the development of the host seedlings.

In September, specimens of *L. mirabile* have been found with a tuber and the scaly peduncle of the inflorescence still underground, without any developed reproductive structures (**Figure 8**). In November A. colubrina is in full bloom and with regrowth. *L. mirabile* starts flowering at the end of November which may continue until the end of February. Although the seeds of A. colubrina var. cebil are disseminated before the flowering of *L. mirabile*, they have to become scarified in order to germinate, so the appearance of *A*. *colubrina* var. *cebil* seedlings coincides with the rainy season and with the flowering of *L. mirabile*. In this case, too, the pods and seeds of the host are usually found very close to the plants of *L. mirabile*, and seedlings are even found around the inflorescences of the parasite.

It has been recorded that the seeds of both the host trees (*P. rigida* and *A. colubrina*) fall underneath the canopy of the tree, directly onto the mature infructescences of the parasites. The legume embryo germinates rapidly and it is common to see the young legume seedlings growing directly on the decayed *Lophophytum* infructescences, maximizing the possibility of contact between the roots and the *Lophophytum* embryo. In both pairs of species (*Parapiptadenia*/*L. leandri* and *Anadenanthera*/*L. mirabile*) the time of germination of the legume seeds coincides with the time of the parasite fruit production which would facilitate close contact between the taxa for the establishment of parasitism. It also shows a process of coevolution of each pair of species in relation to the environment in which they live. Unfortunately, it has not been possible to collect material that shows the morphogenesis of this process, and more collections are needed.

#### **Figure 9.**

*Phenological stages of the parasites, their hosts and the climatic data of their respective environments over the year (Ll: L. leandri, Lb: L. mirabile, Ac: A. colubrina, Pr: P. rigida; Precip: Precipitation, T°: Temperature, H°: Humidity).*

During the six years of observations [31], phenological data were collected on both the parasitic species and their hosts. Climatic data were also collected for the two regions of Argentina where this parasitic association occurs. Graphs were made comparing the temperature, humidity, and precipitation data with the main phenological stages of each parasite and its host (**Figure 9**). Thus, it was possible to interpret the phenological events of the plants (both parasite and host) better in relation to the climate. In all the years of observations, we have never detected a host tree showing any symptoms of damage caused by the holoparasites. We have collected specimens that were on trees many years old (more than 30 years) and that were vigorous, fulfilling their reproductive cycle with apparently normal flowering and fruiting.

#### **9. Conclusions**

In the vegetative bodies of *Lophophytum*, like other members of the Balanophoraceae, organs such as leaves, stems or roots are completely absent [1, 3, 20, 21, 31, 52, 53]. The reductions are also extreme when considering the anatomical structures, such as typical meristems or epidermis, and even stomata are absent. The vegetative body is covered by several layers of tanniniferous parenchyma and sclereids, which are progressively detached, similar to a peridermis. The interface between the parasite and the host has a choraloid design, which facilitates the exchange of both water and photosynthates from the vascular tissues of the host legume towards the *Lophophytum*.

In contrast to the strong reductions in the vegetative body, the staminate flowers show reductions only in their sterile floral parts. The development processes of the anthers, microsporogenesis and microgametogenesis occur normally and in correspondence with the antecedents of the majority of the angiosperms studied, and there are no substantial differences between the two species analyzed [32]. The secretory and uninucleate tapetum characteristics are shared with other genera of the family, such as *Helosis* [47–49], and *Corynaea* crassa [45].

The pistillate flowers are another example of the absence of a perianth, but here the reduction of parts also extends to the reproductive structures [31, 33, 34]. The absence of integuments determines the presence of ategmic ovules. The lack of differentiation of the chalaza, funiculus and vascularization makes it very difficult to establish a concrete boundary between the placenta and nucella. The terms

#### *Anatomy, Embryology and Life Cycle of* Lophophytum*, a Root-Holoparasitic Plant DOI: http://dx.doi.org/10.5772/intechopen.99981*

anastroph/orthotroph are not applicable in *Lophophytum*. Differentiation of the megaspore mother cell is the feature used to establish the micropylar and chalazal poles. The megagametophyte is of the tetrasporic type, with 8-nucleate nuclei at the maturity of an *Adoxa*-like organization, with the typical egg apparatus. *Lophophytum* is a clear case of megagametophyte inversion, confirmed by ontogenetic studies of numerous flowers and not as isolated cases [34]. The embryo sac develops aggressively within the nucellus and rotates during megasporogenesis and megagametogenesis, finally acquiring a "J" shape with the egg apparatus oriented towards the chalazal pole and the antipodes towards the micropylar pole. This would favor the proximity of the egg cell and the pollen tubes that will eventually enter through the styles. However, double fertilization has not been recorded, suggesting the existence of parthenogenesis.

The development of the endosperm is nuclear in nature, but has particularities, as it is possible to divide it into several stages culminating in a fully cellular endosperm [34]. In *Lophophytum*, similar events to those observed in the apomictic species of *Balanophora* have been recorded, such as the fact that the endosperm develops autonomously without fertilization, that it develops enveloping the zygote, which starts dividing much later than the endosperm. The mature embryo is globular and undifferentiated as in other holoparasites, such as *Pilostyles* and *Orobanche* [54–57]. They lack a seed coat, due to the absence of integuments in the ovule. The term seed in the strict sense could not be applied to the Balanophoraceae, as it has been shown that the structure is derived entirely from the embryo sac and is completely devoid of teguments, as already mentioned by Holzapfel [42].

The dispersal unit of *Lophophytum* is a uniseminated achene [16]. The diaspores of *Lophophytum* are dispersed by rodents that feed on them, separating the achenes from the inflorescence axis. Dispersal is favored by the previous action of Oxycorynus larvae on the axes of the secondary rachis [35].

From the comparative analysis of reductions and fusions in the gynoecium of the Balanophoraceae with the results observed in *Lophophytum*, a line of possible successive steps is proposed which includes several profound modifications: i) gradual loss of identity of the ovule and placenta: both structures are still recognizable in *Lophophytum*, whereas in the other Balanophoraceae they are not distinguishable. ii) gradual reduction of integuments, loss of the chalaza and funiculus and absence of vascular supply in the ovules. iii) progressive fusion of the placenta/ ova/carpels, with a consequent reduction of the ovarian cavity until its complete disappearance in *Balanophora*. *Lophophytum* is the only genus of Balanophoraceae in which the ovules are still clearly identifiable from the placenta. In the other genera of the family the boundary between the ovules and placentas is blurred, so the term placental-nucellar complex is still used for these cases.

The family Balanophoraceae is an excellent example of how knowledge of embryological data expands the possibility of establishing their phylogenetic relationships [3, 4, 16, 56, 57]. Given the lack of vegetative characteristics due to the particular structure of these plants, the importance of floral characteristics for taxonomic identification is emphasized. All data acquired from the flower structure and anatomy make species identification possible.

In the family Balanophoraceae, knowledge of the germination process and the initiation of the host relationship (establishment of the initial haustorium) was only achieved in *Balanophora abbreviata* [50]. One of the unfinished points of this review is the germination process and the connection of the haustorium with the host root. Host and host specificity are revalidated, at least for the Argentine species of *Lophophytum*, where *L. leandri* spends its entire life cycle on *Parapiptadenia rigida*, whereas *L. mirabile* spends its entire life cycle on *Anadenanthera colubrina* [31]. There is a direct relationship between the life cycle of the hosts and that of the

#### *Parasitic Plants*

parasites with respect to the coincidence of their reproductive phases. The seeds of the parasites only mature when the host seeds are ready for germination. The flowering period of the Argentine species is concentrated in contrasting periods, with *L. mirabile* flowering in midsummer, and *L. leandri* in winter.

The studies carried out here are a clear example of the process of co-evolution between a holoparasite and its host. Each species of *Lophophytum* develops its reproductive stages at the time of year when its seeds can come into contact with the seedlings of its host, so the chances of establishing a parasitic relationship are optimized.
