In bryozoans (phylum Bryozoa), representative colonial animals mostly found in marine environments, some species possess different types of individuals (heterozooids) specialized in different functions such as defense or structural support for their colonies. Among them, the best-known heterozooids are the avicularia, known to function as defenders. The differentiation processes of heterozooids, including avicularia, should be important keys to understand the evolutionary significance of bryozoans. However, the developmental process of avicularium formation remains to be fully understood. In this study, therefore, in order to understand the detailed developmental process and timing of avicularium formation, extensive observations were carried out in a bryozoan species, Bugulina californica (Cheilostomata, Bugulidae), that possesses adventitious avicularia, by performing stereomicroscopy on live materials, in addition to scanning electron microscopy and histological observations. The whole process can be divided into seven stages based on developmental events. Especially notably, at the earlier stages, there are three major budding events that produce proliferating cell masses corresponding to primordial tissues: (1) budding of the peduncle cushion at the outer margin of the distal part of a young autozooid, (2) budding of the head-part primordium from the peduncle cushion, and (3) budding of the polypide inside the head part. Experimental control of temperature showed that 20°C would be the best to maintain B. californica colonies.
INTRODUCTION
Bryozoans are colonial benthic invertebrates that inhabit mostly marine environments. They are characterized by a ciliated tentacle crown (lophophore) in each zooid, which is used for suspension-feeding (Hyman, 1959; Ryland, 1976; Mukai et al., 1997). Including fossil records, the phylum Bryozoa consists of more than 15,000 species (> 6000 extant species), in which the order Cheilostomata is the largest group, with about 5000 species (Gordon, 1999; Bock and Gordon, 2013). Species belonging to the order Cheilostomata with characteristic calcified skeletons exhibit polymorphism among zooids that represent divisions of labor, such as sexually dimorphic zooids, brooding zooids, structural zooids and defensive zooids (Silén, 1977; Reed, 1991; Harvell, 1994; Lidgard et al., 2012; Schack et al., 2019). By the acquisition of zooid polymorphism, these bryozoans have adapted to diverse environments (Schopf, 1973).
Avicularia are strictly found only in cheilostome bryozoans, although analogs are known in the fossil cyclostome family Eleidae. They have a range of functions, including defense and secretion (Darwin, 1872; Nordmann, 1840; Johnston, 1847; Kaufmann, 1968, 1971; Winston, 1984, 1986, 1991; Carter et al., 2010, 2011; Schack et al., 2019). Although lacking functional intestines and a lophophore, avicularia possess a vestigial polypide, which corresponds to tentacles, tentacle sheath, intestine, musculature and nervous system in autozooids (normal individuals: that are capable of feeding using the lophophore), providing evidence that avicularia have been derived from autozooids (Hyman, 1959; Ryland, 1970; Banta, 1973; Silén, 1977; Winston, 1984; McKinney and Jackson, 1989; Reed, 1991; Carter et al., 2011). Nutrients are provided to avicularia through a duct connecting individual zooids, which is termed a “funiculus” (Nitsche, 1871; Huxley, 1856; Lutaud, 1961; Carle and Ruppert, 1983). The morphologies of avicularia are largely transformed from that of autozooids (Fig. 1A), and are diversified among bryozoan species (Carter et al., 2010). Technical terms of bryozoan biology are very distinctive, so they are listed here as a glossary (Table 1).
Based on the budding position and morphology, avicularia are classified into four types: vicarious avicularia, interzooidal avicularia, adventitious avicularia and vibracula (Carter et al., 2010). Vicarious avicularia, which have enlarged opercula and are similar to autozooids in size, occupy the autozooid positions and interact with other zooids. Interzooidal avicularia are budded between autozooids. Their cystids extend to the substratum and they are always smaller than the parent autozooid. Adventitious avicularia (including birds-head avicularia) are budded on the lateral, frontal or basal walls of autozooids. Vibracula show unique morphologies such as setiform mandibles that can rotate, so that they are considered to be specialized types of avicularia for removal of sediments and larvae that are attached to the surface of bryozoan colonies (Ryland, 1970; Schopf, 1973; Cook and Chimonides, 1978, 1985, 1987; Winston, 1984; McKinney and Jackson, 1989).
Aviculiferous fossil bryozoans show that avicularia have been transformed from autozooids, through vicarious avicularia (Cheetham et al., 2006). This was also proved by the evidence, as mentioned above, that avicularia possess homologous body structures and organs to those in autozooids (Fig. 1A; Cook, 1968; Lidgard et al., 2012).
However, the developmental process of the avicularium, especially in relation to environmental cues such as temperature, has not been extensively investigated. Only a few descriptions have been carried out so far in a few species; immature avicularia in five different developmental stages were observed in Bugula sabatieri (currently synonymized as Bugulina simplex) (Calvet, 1900), and a developmental comparison of polypide and operculum in autozooid and avicularia at four different stages was reported in Flustra foliacea (Silén, 1938).
As for the developmental process of autozooids, only developmental stages of the polypide including the lophophore, digestive tube and ganglion, have so far been described (Borg, 1926; Soule, 1954; Schwaha and Wood, 2011; Schwaha et al., 2011; Schwaha, 2018); the processes of cystid and musculature formation have not been investigated. Besides, the budding process of the avicularium polypide was previously described in an old publication with a few sketches (Calvet, 1900). However, sequential development during avicularium formation has yet to be described.
In this study, to clarify the developmental process of avicularium formation in detail, we established an observation platform in a bryozoan species that constantly produces avicularia and is easy to collect and maintain in laboratories. This study focuses on a bryozoan species, Bugulina californica (Cheilostomata, Bugulidae), that produces birds-head avicularia; namely, each autozooid has an avicularium budding on its surface (Fig. 1B, C; Robertson, 1905).
This species was firstly described from California, but is known to be distributed in East Asian regions (Robertson, 1905; Lutaenko et al., 2013). The budding position of the adventitious avicularia is on the distal half of an autozooid, which is on a side wall of the orifice (opening for the extending lophophore) (Robertson, 1905). This indicates that the budding position of avicularia is autonomously defined in the autozooid body; in other words, it is suggested that the developmental program of avicularia formation is incorporated in the developmental process of autozooids. Thus, since the site of avicularium formation is predetermined, the developmental process can be observed from inception.
Table 1.
Glossary of technical terms in byozoan biology.
Therefore, the developmental stages of adventitious avicularia are defined in this study, based on observations in B. californica. The developmental process of zooid budding in Cheilostomata was described in old reports (Nitsche, 1869; Calvet, 1900), in which the budding of autozooids and avicularia was observed. However, only a few immature zooids at a few developmental stages were included in those data.
In this study, the process of avicularia formation was described in detail, and the developmental stages were defined based on observations made by multiple approaches. We firstly defined the budding position and direction of adventitious avicularia, and then observed the process of avicularium formation over time. Based on the defined developmental stages, outer and inner morphologies are described in detail, throughout the process of avicularium formation.
MATERIALS AND METHODS
Field sampling
Colonies of B. californica were often found in plastic baskets which were hung at a depth of 2–3 meters from a raft for the aquaculture of Ciona intestinalis at the coastal area near the Misaki Marine Biological Station (Moroiso Bay). Stock colonies were kept in the laboratory, in a bucket with running sea water. They were fed marine diatom Chaetoceros calcitrans (Marinetech Co Ltd., Aichi).
Rearing conditions
To observe the developmental process of the avicularia, collected colonies were maintained in the laboratory. Collected colonies that were similar in size were kept in an aquarium with 1.5 L filtered sea water. C. calcitrans was provided as food. The amount of food was 250–500 µl, frequency of feeding was more than once every 2 days, and water was changed once every 3 days. The colonies were observed every 8 hours using a microscope (Olympus BX51) and images were recorded using the attached camera (Olympus DP74). A colony was put on a slide glass with filtered sea water and covered with a cover slip. Five to 10 immature autozooids without avicularia or with developing avicularia that were located at the tip of a colony were observed to follow the developmental process.
Scanning electron microscopy (SEM)
To elucidate the developmental process of avicularia, morphological and histological features were investigated using SEM and histological observations. For SEM, Bryozoan samples were anesthetized with 7% MgCl2 for longer than 10 min, then fixed in FAA fixative (ethanol: formalin: acetic acid = 16:6:1) for 3 hours, and preserved in 70% ethanol until use. The fixed samples were dehydrated using a graded series of ethanol up to 100%, and hardened in HMDS (hexamethyldisilazane) for 1 hour. After soaking in ethanol and t-butanol, samples were freeze-dried using a Freeze Dryer ES-2030 (Hitachi Global, Tokyo, Japan), and coated with gold ions with an Ion Sputter E-1010 (Hitachi Global, Tokyo, Japan). Samples were observed with a JSM-5510LV scanning electron microscope (JEOL Ltd., Tokyo, Japan).
Histological observations
Samples were anesthetized with 7% MgCl2 for longer than 10 min, fixed in FAA fixative for 3 hours, and preserved in 70% ethanol until use. The samples were hydrated in decreasing concentrations of ethanol and then soaked in distilled water, and embedded in agarose gel. After that, the embedded samples were dehydrated in a graded series of ethanol and acetone, then embedded in resin with a kit (Techno-vit8100, Kulzer GmbH, Hanau). Embedded samples were then sectioned at 5 µm, stained with hematoxylin and eosin, and mounted in Malinol (MUTO Pure Chemicals, Tokyo). Sections were observed using a camera-equipped microscope (Olympus BX51, Olympus DP74).
Examination of rearing temperature
To examine the optimal rearing condition (temperature) for B. californica, observations under different temperature conditions were carried out. Collected colonies that were similar in size were kept under three different temperature conditions (15°C, 20°C, 25°C). Other rearing conditions were the same in the three temperature conditions. The period from the stage before the onset of avicularia formation (Stage 0) to the completion of avicularia formation (Stage 7) was recorded for each temperature.
For the statistical test, the homogeneity of variance was examined by the Brown-Forsythe test. The significance of changes in the time for the avicularia formation was examined using one-way analysis of variance (ANOVA) after log-transformation (Zar, 1984). Differences between the means of periods were compared using the Tukey HSD test.
RESULTS AND DISCUSSION
Overall observations on avicularia in B. californica
Branches are biserial, with avicularia formed on the outer distolateral surface (Fig. 1B, C; Fig. 2). Avicularium formation occurs concurrently with autozooid development (Fig. 2A). Immature autozooids at branch tips lack avicularium primordia (Fig. 2A, arrowheads). Immature avicularia were seen mostly on the second or third autozooids from branch tips (Fig. 2A). By the time avicularium formation is complete (Stage 7, see below), autozooids are mature and everting lophophores. However, the situation was more complicated at the bifurcating parts of branches (see Supplementary Figure S1 (zs200143_FigS1.pdf)). Taken together, these findings strongly suggested that a newly budded autozooid at a colony branch produces the next autozooid as it develops, and after that, it starts to bud the primordium of an avicularium.
The position where an avicularium was budded in an associated autozooid was, as described in Robertson (1905), on the outer surface of the distal half of an autozooid, at the side of the autozooid orifice (Fig. 1B, C; Fig. 2). The direction of avicularium budding was orthogonal to the autozooid cystid wall (Fig. 2A, B, F), as described in a previous study (Mawatari, 1971). Based on the position of avicularia, it is hypothesized that the avicularia might protect the lophophore of autozooids, although they might have other functions such as waste removal and providing baits (Darwin, 1872; Nordmann, 1840; Johnston, 1847; Kaufmann, 1968, 1971; Winston, 1984, 1986, 1991; Carter et al., 2010, 2011; Schack et al., 2019). To test the above hypothesis, the correlation between the activities of autozooids and those of avicularia should be examined in future studies. Furthermore, since an avicularium is regularly formed on each autozooid in a predetermined position, the developmental program of avicularia should be incorporated into that of autozooids.
Therefore, at the branch tip of the bryozoan colony, immature avicularia in the course of the developmental process were observed along a colony branch; more-developed avicularia are seen at more proximal parts of a branch (Fig. 2A, B). At the beginning of avicularium formation, a small projection was formed at the margin of an autozooid cystid, and was later shown to become the peduncle cushion (PC) of the avicularium (Fig. 2C, D). Thus, various stages of avicularia can be observed at the distal branch portions, so the developmental stages were defined based on the observations on such avicularia.
Developmental stages of avicularium formation
Based on the observations made using stereomicroscopy, scanning electron microscopy (SEM) and histological sections, the developmental process of avicularia formation was classified into seven stages, i.e., Stage 1 to Stage 7 (Fig. 3). Characteristics of each developmental stage are described below (technical terms for bryozoan body parts are defined in Table 1).
Stage 1
When the size of an autozooid reached approx. 100 µm, a small bud was formed at the outer margin (Fig. 2C, Fig. 3A). Based on observations at later stages, it was concluded that this bud developed into a peduncle cushion, not the entire avicularium (see below). At this time, the budding of the main body of an avicularium from the peduncle cushion had not yet occurred. The surface of the peduncle cushion appeared smooth (Fig. 3B). Histological observations also revealed that, at this stage, the peduncle cushion was filled with proliferating cells.
Stage 2
At Stage 2, the primordial avicularium forming the head part of an avicularium was budding from the peduncle cushion, and then elongating at the latter half of Stage 2 (Fig. 3D). The surface of the primordium was wrinkled. Histological observations showed that the primordium was filled with a dense cell mass coming out from the peduncle cushion and no differentiated structures (Fig. 3F). At this stage, the peduncle cushion was still filled with undifferentiated cells (see Supplementary Figure S2B (zs200143_FigS2.pdf)).
Stage 3
At Stage 3, the primordium continued to elongate, and then formed a constriction when it reached over 50 µm in length (Fig. 3G, arrowheads). SEM observations revealed that there were numerous fine wrinkles in multiple directions with major grooves in the proximal-distal direction at the distal part, while at the proximal part the wrinkles looked coarser in the vertical direction to the P-D axis (Fig. 3H). Histological images showed that the epithelial cells and peritoneal cells on the future-mandible side proliferated to become thickened, and from this thickened cell mass a proliferating cell mass budded inside the primordial cavity (Fig. 3I, arrow). This phenomenon in which the cell mass was budded inside a newly formed polypide was also reported in the development of autozooids, suggesting that the bud is the primordium of a polypide (Soule, 1954; Schwaha and Wood, 2011; Schwaha et al., 2011; Schwaha, 2018). Since a previous study suggested that zooid polymorphism was developed by the alteration of pre-existing structures (Silén, 1977), this similarity between autozooid and avicularium development would imply that the avicularia were evolutionarily derived from autozooids.
Stage 4
Until Stage 4, the elongation of the primordial avicularium continued in the direction perpendicular to the autozooid cystid wall. The proximal part under the constriction formed the peduncle, while the distal side became extremely enlarged (Fig. 3J). Mandible formation began at this stage. The wrinkled structures were uniform at the beginning of Stage 2, but this changed at Stage 4 as the primordium elongated (Fig. 3K). SEM observations showed that a deep groove appeared at the boundary between the mandible and the avicularium head (Fig. 3K). Inside the enlarged primordium, a few cells with some fibrous structures came out from the budded cell mass (Fig. 3L, ad). This is suggested to show the onset of adductor muscle formation. Inside the peduncle cushion, fibroblast-like cells were also observed (see Supplementary Figure S2D (zs200143_FigS2.pdf), arrowhead). This indicates that the muscle formation occurs simultaneously inside the head part and inside the peduncle cushion.
Stage 5
At Stage 5, the apical portion of the enlarged primordium became sharp, to form the rostrum (Fig. 3M, N). The mandibular hook was not formed yet. From Stage 5, the direction of elongation was changed to the direction to the side of the rostrum, i.e., the distal direction of the colony (Fig. 3M–R). SEM observations revealed that some wrinkled structures were seen at the rostrum tip, indicating that the formation of the rostrum tip was not completed, and the calcification was still ongoing at this stage (Fig. 3N). The head part became smooth, and wrinkled structures observed at the previous stages disappeared (Fig. 3N), probably due to calcification that hardened the head structure. It is also possible that the wrinkled epithelial structure enables the abrupt development, as in insect molting/metamorphosis (e.g., Matsuda et al., 2017). Fibers of the adductor muscle that connected the head capsule and the mandible were proliferating, but the connection was not completed (Fig. 3O).
Stage 6
At Stage 6, the mandibular hook was formed and inserted inside the rostrum tip (Fig. 3P). The elongation of the rostrum suggested that the palate, i.e., the distal part of the head part adjacent to the mandible (distal to the vestigial polypide), started to form (Fig. 3Q, asterisk). The formation of this part was also confirmed by histological observations (Fig. 3R, asterisk). The adductor muscle inside the head part was almost completed (see Supplementary Figure S2E (zs200143_FigS2.pdf), arrow). Also, muscles inside the peduncle cushion were formed, showing fibrous structures (see Supplementary Figure S2F (zs200143_FigS2.pdf), arrow). Behavioral observations at this stage showed that the mandible opened slightly, but was not completely opened (see).
This is probably because some membranous structure connected the rostrum and the mandible. The membrane is suggested to be torn off as a result of multiple trial openings of the mandible, resulting in the complete opening of the mandible. At the later period of Stage 6, when cilia were formed from the vestigial polypide between the rostrum and the mandible, the avicularia formation was almost completed (Fig. 3Q, R). The position of the cilia is suggested to correspond to that of the lophophore in autozooids, consistently with a previous study (Carter et al., 2010). The budded cell mass that was formed inside at Stage 3 (i.e., the primordial polypide) later gradually shrunk to become the vestigial polypide in an avicularium.Stage 7
At Stage 7, the formation of an avicularium was completed when the mandible was fully opened (Fig. 3S–U). The adductor muscle was fully developed and the funiculus was connected from the associated autozooid to the vestigial polypide (Fig. 3U). Multiple sensory cilia were observed on the outer surface of the vestigial polypide (Fig. 3U).
As described above, the process of avicularium formation can be divided into seven developmental stages. In the course of development, three major budding events play major roles in development: (1) budding of PC primordium from an autozooid (Stage 1), (2) budding of head-part primordium from PC (Stage 2), and (3) internal budding of the polypide inside a head part (Stage 3). Through these steps, with cell proliferation followed by detailed formation of body structures, the avicularium formation is accomplished.
Avicularia formation after autozooid degeneration
Some avicularia were observed to develop even after polypide regression (at 15°C), raising the question of whether avicularium formation requires the presence of a polypide in the parent zooid. In the case that an autozooid with lophophore less than 100 µm was degenerated, before the onset of avicularia formation, the avicularium formation did not start. After that, the autozooid polypide was absorbed although the zoocium/cystid remained (see Supplementary Figure S3 (zs200143_FigS3.pdf)). This suggests that the avicularia formation requires the associated polypide, probably because nutrition for the development would be provided from the autozooid via the funiculus. However, in the case that the associated autozooid started to degenerate at Stage 3 of the avicularia formation, and completely died at Stage 6 (the brown body was observed), the avicularia formation was completed, although it took longer (128 hours) than in the normal situation (104.00 ± 8.64 hours (mean ± s.e.)) (see Supplementary Figure S4 (zs200143_FigS4.pdf)). This indicates that the later stages of avicularia development can proceed, independently of the associated autozooid. This may suggest that dense cells inside the primordium could be incorporated from the parental autozooid until Stage 2, and then at Stage 3, cell differentiation occurred inside the developing avicularium, when the connection with the autozooid was completed. The independent development of avicularia might have been related to the formation of the polypide of the avicularium from Stage 3, although it required a longer time, probably because the nutritional supply was limited due to the degeneration of the autozooid polypide.
After the completion of avicularia formation, i.e., Stage 7, when the associated autozooid was degenerated and absorbed it did not affect the function of the avicularia during a few days (see).
This suggests that completely developed avicularia can survive without an associated polypide, at least for a few days. Alternatively, nutrition could also be supplied by other autozooids in the same colony, through the funicular network. Our observations showed that polypide regression is easily induced by environmental changes and/or nutritional depletion in a colony. In contrast, interestingly, the recycling of avicularian soft tissues has never been observed, suggesting that unlike autozooids, avicularia cannot be reused as a nutritional source.Optimal temperature condition for laboratory culture
To examine the preferred temperature for rearing, B. californica was incubated at 15, 20 or 25°C (Fig. 5, see Supplementary Figures S5 (zs200143_FigS5.pdf)– S7 (zs200143_FigS7.pdf)). At 15°C, it took 104.00 ± 8.64 hours (mean ± s.e.) from the onset of budding to the completion of avicularium morphogenesis (Fig. 5, and see Supplementary Figure S5 (zs200143_FigS5.pdf)). During this period, even a single stage took more than several tens of hours (see Supplementary Figure S5 (zs200143_FigS5.pdf)). Although the observation was continued after the completion, the size was not increased thereafter. Also, at 15°C, many autozooids died during the development, probably because the temperature was too low.
At 20 and 25°C, avicularium morphogenesis took 72.00 ± 1.33 and 46.86 ± 1.14 hours, respectively (Fig. 5, see Supplementary Figures S6 (zs200143_FigS6.pdf), S7 (zs200143_FigS7.pdf)). At 25°C, the development sometimes proceeded for two stages during an observation interval of 8 hours (see Supplementary Figure S7 (zs200143_FigS7.pdf)). There were statistically significant differences of developmental speed among the three temperature conditions (Fig. 5; P < 0.05, Tukey's HSD test). On the other hand, at 25°C, some autozooids died before the avicularium formation due to proliferation of other microorganisms. Among the three temperature conditions, 20°C was optimal for observing all of the developmental stages during the 8-hour observation period (see Supplementary Figure S6 (zs200143_FigS6.pdf)).
Developmental comparison between autozooids and avicularia
In the case of autozooid formation by budding, a new cystid is budded from the parent autozooid. The inner surface of the cystid is lined by epithelial and peritoneal layers, both of which are budded into the space inside the cystid to form a polypide (Schwaha and Wood, 2011; Schwaha, 2018). Stage 3 of avicularia formation resembled this stage of autozooid development (Fig. 3G, H, I), suggesting that the proliferation and budding of the two layers, i.e., epithelial and peritoneal layers, is conserved between autozooids and avicularia. This may also imply that an avicularium is a zooid, and that avicularia may have been acquired by the modification of autozooid development, as suggested in previous studies (Silén, 1977; Schack et al., 2019).
Evolutionary implications and perspectives
In colonial animals, generally, heterozooids show less individuality in comparison with autozooids, in terms of their independence or autonomy (Beklemishev, 1969). Our observations, however, showed that some avicularia survived and functioned after the loss of associated autozooids, suggesting that avicularia also possess individuality to some extent, although the definition of “individuality” is ambiguous in colonial animals such as bryozoans (cf. Clarke, 2010, 2011). This also suggests that there still could be some nutritional connections to other autozooids, probably through the funiculus.
In the sense that, to form a colony, multiple phenotypes are differentiated in a single species among related or clonal individuals, the coloniality and the heterozooid differentiation are similar to the caste differentiation in social insects (Wilson, 1975), although, in social insects, colony members are not directly connected to each other and possess apparent “individuality”, while bryozoan zooids are connected to one another. Also, in the caste differentiation in social insects, multiple types of individuals, i.e., castes, differentiate during their postembryonic development, depending on various extrinsic factors (Miura, 2019). Furthermore, the soldier caste is a distinctive caste specialized in colony defense, so this is also analogous to the avicularia in bryozoans. The developmental process of soldier differentiation in termites is well documented (Miura and Maekawa, 2020), so similar analyses can also be applied to the study of avicularia formation in bryozoans.
This study has documented the detailed process of avicularia formation in a bryozoan species, B. californica, that possesses adventitious avicularia. It is still unknown whether the process of avicularia formation in B. californica is shared by other bryozoans, including other avicularia types, such as vicarious avicularia, interzooidal avicularia and vibracula. Since the morphological features of other avicularia types are similar to each other (Carter et al., 2010), they have been suggested to share some developmental processes. Also, the number of avicularia acquisitions in the bryozoan evolution is still unclear, although it is suggested that vicarious avicularia evolutionarily derive from autozooids (Silén, 1938; Cheetham et al., 2006). Therefore, it should be important to study the developmental process of vicarious avicularia, although it is difficult to culture the species with vicarious avicularia (Schack et al., 2019). In future studies, extensive phylogenetic analyses together with the investigations of avicularia formation in various bryozoan lineages will reveal the evolutionary process leading to the acquisition of avicularia.
ACKNOWLEDGMENTS
We would like to express our gratitude to M. Kohtsuka, S. Aratake, A. Yoshikawa and M. Kawabata for assisting us at the time of sampling. Thanks are also due to M. Sekifuji for helping us to rear the material species. Y. Togawa, A. Okada and M. Okanishi supported our experimental procedures. M. Yoshida, D. Kurokawa, H. Nozaki and R. Ueshima gave us a number of constructive comments on the current study. This work was supported by a Grant-in Aid for Scientific Research A (No. 18H04006) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
AUTHOR CONTRIBUTIONS
HY, MH, HK and TM designed the whole study. HY, SU, KO, HK and TM collected the study materials from the field. HY reared animals and observed the developmental processes. HY, MN and SU performed the experimental procedures. HY, JS and TM analyzed the data on rearing conditions. All of the authors wrote the manuscript.
SUPPLEMENTARY MATERIALS
Supplementary materials for this article are available online. (URL: https://doi.org/10.2108/zs200143)
Supplementary Figure S1 (zs200143_FigS1.pdf). Distal tip of a colony branch.
Supplementary Figure S2 (zs200143_FigS2.pdf). Process of formation of muscle tissue in peduncle cushion.
Supplementary Figure S3 (zs200143_FigS3.pdf). Observation of a degenerating autozooid that did not produce an avicularium.
Supplementary Figure S4 (zs200143_FigS4.pdf). Observation of an avicularium formed on an autozooid that disappeared during the process of avicularia formation.
Supplementary Figure S5 (zs200143_FigS5.pdf). Developmental process of avicularia formation observed at 15°C.
Supplementary Figure S6 (zs200143_FigS6.pdf). Developmental process of avicularia formation observed at 20°C.
Supplementary Figure S7 (zs200143_FigS7.pdf). Developmental process of avicularia formation observed at 25°C.
Developing avicularium at early St. 6. Developing avicularium at mid St. 6. Developing avicularium at late St. 6. Completed avicularium at St. 7. Completed avicularium at St. 7 whose parental autozooid was dead, but the avicularium barely moved.