Caryologia. International Journal of Cytology, Cytosystematics and Cytogenetics 72(4): 105-119, 2019 Firenze University Press www.fupress.com/caryologiaCaryologia International Journal of Cytology, Cytosystematics and Cytogenetics ISSN 0008-7114 (print) | ISSN 2165-5391 (online) | DOI: 10.13128/caryologia-670 Citation: C.F. Crane (2019) Megaga- metophyte Differentiation in Zephyran- thes drummondii D. Don and Zephy- ranthes chlorosolen (Herb.) D. Dietr. (Amaryllidaceae). Caryologia 72(4): 105-119. doi: 10.13128/caryologia-670 Published: December 23, 2019 Copyright: © 2019 C.F. Crane. This is an open access, peer-reviewed article published by Firenze University Press (http://www.fupress.com/caryologia) and distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distri- bution, and reproduction in any medi- um, provided the original author and source are credited. Data Availability Statement: All rel- evant data are within the paper and its Supporting Information files. Competing Interests: The Author(s) declare(s) no conflict of interest. Megagametophyte Differentiation in Zephyranthes drummondii D. Don and Zephyranthes chlorosolen (Herb.) D. Dietr. (Amaryllidaceae) Charles F. Crane USDA-ARS Crop Production and Pest Control Research Unit, West Lafayette, IN 47907 and Department of Botany and Plant Pathology, Purdue University *Corresponding author: Charles.Crane@ars.usda.gov; ccrane@purdue.edu Abstract. Megagametophyte differentiation was examined in cleared ovules from emergent buds and open flowers of Zephyranthes drummondii D. Don and Z. chlo- rosolen (Herb.) D. Dietr., two highly apomictic species that exhibit the Antennaria type of megasporogenesis and hemigamy. Stages from binucleate megagametophytes through early endosperm divisions were sampled. A large central vacuole appears after the megasporocyte divides mitotically, and this vacuole persists until the endosperm becomes cellular. Upon cellularization of the egg and antipodal apparati, a central col- umn of cytoplasm develops longitudinally across the central vacuole, and both polar nuclei move into it before moving in unison to the chalazal end. The mature megaga- metophyte is organized conventionally, with one egg, two synergid, two polar nuclei, and three antipodal cells. Endosperm development is helobial, but there are few divi- sions in the chalazal chamber of the endosperm. The behavior of fertilized and unferti- lized ovules was also studied in response to pollination. Although synergids degenerate autonomously, pollination accelerates synergid degeneration in unfertilized as well as fertilized ovules relative to unpollinated flowers of the same age. Tabulation of numeri- cal abnormalities suggests that progressive imprinting is involved in megagametophyte differentiation; the data do not support a strictly zonal specification of nuclear fate, but instead a role for nuclear polarization before mitotic divisions. This study demon- strated the value of cleared ovules in gathering statistically and temporally meaningful observations of megagametophyte differentiation, relating in particular to the move- ment of polar nuclei and the response of the megagametophyte to pollination. Keywords. Megagametophyte, embryo sac, apomixis, hemigamy, polar nuclei. INTRODUCTION Zephyranthes drummondii D. Don and Z. chlorosolen (Herb.) D. Dietr. are two congeneric amaryllid species that share apomictic reproduction and a flowering response to rainfall, hence the common name “rain lilies”. The buds differentiate within the bulb for several months and then emerge 106 Charles F. Crane in response to wetting of the roots. The flowers open at sunset, most frequently on the fourth day following rainfall. The flowers remain open for one or two days, depending on temperature. In nature, self-pollination usually occurs soon after the anthers dehisce inside the bud at mid-morning on the day of anthesis. The incon- gruous combination of large, showy, sweetly scented flowers and self-pollination has motivated several studies of reproduction in these species and the related Habran- thus tubispathus (L’Her.) Traub (Pace, 1913; Brown, 1951; Coe, 1953). These studies indicate obligate apomixis by mitotic megasporogenesis, i.e., the Antennaria type (first described in Antennaria alpina (L.) Gaertn. by Juel, 1900), and hemigamy (synonym: semigamy; Battaglia, 1945), which is development of the zygote after plasmog- amy but without fusion of egg and sperm nuclei within the cytoplasm of the egg cell. Various details of the stag- es from the free-nuclear embryo sacs through fertiliza- tion remain to be documented, such as the path taken by the polar nuclei to reach the chalazal end of the central cell and whether the central cell experiences triple fusion to initiate the endosperm. Many apomictic species require pollination for seed set, but the mechanisms vary. In Potentilla (Gustafsson, 1946, p. 31), the embryo can begin to develop autono- mously in unpollinated flowers, and a sperm nucleus fertilizes only the central cell, thus initiating endosperm development. The exact fate of the other sperm nucleus is rarely known; in the Ranunculus auricomus species complex it has been reported also to fertilize the cen- tral cell frequently (Nogler, 1972), leading to expectedly 6n endosperm with the 2:1 maternal to paternal genome ratio usually found in sexually produced endosperm. Facultative double fertilization of the central cell has also been indicated with f low cytometry in apomic- tic Crataegus (Talent and Dickinson, 2007). In most apomictic panicoid and eragrostoid Poaceae (Brown and Emery, 1957; Voight and Bashaw, 1972), there is but one polar nucleus, and fusion with only one sperm would produce the 2:1 maternal:paternal ratio. Never- theless, Bashaw and Hanna (1990) reported frequently observing a sperm in the central cell of the panicoid grass Cenchrus ciliaris L., but none near the egg, possi- bly indicating that both sperms usually enter the central cell but only one fuses with the lone polar nucleus. In most species with nucellar (adventitious) embryos, such as apomicts in the genus Citrus, the megagametophyte is reduced and sexual fertilization is more or less unaf- fected; the nucellar embryos then outcompete the sexu- ally produced embryo (Gustafsson, 1946, p. 35; Nygren, 1967, p. 559). As defined above, nearly obligate hemiga- my is known in nature only in certain species of Rud- beckia (Asteraceae; Battaglia, 1945) and Habranthus and Zephyranthes (sister genera in the Amaryllidaceae). Also, a dominant, incompletely penetrant hemigamous mutant Se has been recovered in cotton (Turcotte and Feaster, 1969); unlike hemigamy in Zephyranthes, it readily pro- duces maternal haploid, paternal haploid, and hybrid sectors in chimeric embryos or maternal-paternal twin embryos. Similar behavior has been observed at ca. 1% frequency in wild-type Theobroma cacao, where it has been exploited as a source of haploids (Lanaud, 1988). In contrast to broadly descriptive classical stud- ies, modern research (mostly in Arabidopsis) has taken advantage of a battery of transposon insertion mutants, a finished genome sequence, fluorescent reporter mol- ecules, and informatic tools, to accumulate a body of concepts and literature dealing with signaling and gene interactions during fertilization in sexual species (Zhou and Dresselhaus, 2019). No form of apomictic repro- duction is understood in comparable detail, although apomictic behaviors can spur insights into aspects of sexual reproduction. For example, the facultative double fertilization of the central cell in Crataegus (Talent and Dickinson, 2007) suggests that the polar nuclei briefly remain attractive or receptive to a second sperm nucleus after fusing with the first one, that the central cell can attract both sperm nuclei, and that the egg more strong- ly attracts one and only one sperm nucleus (else trip- loids and haploids result); once framed, all three of these hypotheses can be tested experimentally in an amenable species. Unfortunately, to date there apparently is no reported Arabidopsis mutant that exactly duplicates the hemigamous behavior seen in Zephyranthes. Previous embryological studies of apomictic Zephy- ranthes and Habranthus (Pace, 1913; Brown, 1951; Coe, 1953) have produced limited evidence about megasporo- genesis, because this stage occurs within the bulb where the bud length is not visible without destroying the plant and each plant produces zero to five buds per year. The evidence in favor of the Antennaria type is mostly negative: dyads expected from first-meiotic restitution (the Taraxacum type) or complete omission of the first meiotic division (the Blumea type [Chennaveeriah and Patil, 1971] or the syndrome in Elymus rectisetus (Nees in Lehm.) A. Love et Connor [Crane and Carman, 1987]) have not been observed, while enlarging, vacu- olate, undivided megasporocytes are frequent. The posi- tive evidence is a single image of a mitotic metaphase in an enlarged, vacuolate megasporocyte in Habran- thus tubispathus (Brown, 1951). Indirect evidence is the occurrence of the ordinary, monosporic Polygonum type in sexual Zephyranthes candida (Ao et al., 2016), which militates against the occurrence of the Ixeris type (mei- 107Megagametophyte differentiation in Zephyranthes otic first-division restitution in the tetrasporic Fritillaria type) in Zephyranthes. The present study took advantage of the readily accessible later stages in bud development to understand the maturation of the megagametophyte in two apomictic species of Zephyranthes, with particu- lar interest in three processes shared with related sexual species: the movements of polar nuclei, the senescence of the megagametophyte with and without fertilization, and the regulation of nuclear fates as the megagameto- phyte matures. These are universal aspects of angio- sperm reproduction that are easily observed in apomic- tic Zephyranthes. MATERIALS AND METHODS The ovules of maturing flowers in Zephyranthes are particularly suited for clearing studies. At this stage, the ovules are easily removed from the ovary before fixa- tion, and the cleared ovules are readily pipetted onto a microscope slide. The ovules are flat and thus present the embryo sac in sagittal optical section. The embryo sac is very large and thus the egg apparatus and polar nuclei are not close to the plane of overlying and under- lying nucellar cells. The nucellus is free of birefringent calcium oxalate crystals at the stages examined, facili- tating differential interference-contrast microscopy. Finally, the refractive index of methyl salicylate is close to optimal to resolve nuclei and yet see through many cell layers, which has not been the case in other species like Nothoscordum bivalve (L.) Britton in N.L.Britton & A.Brown (Crane, 1978) and Elymus rectisetus (Nees) A.Love & Connor (Crane and Carman, 1987). Flowers were examined in Z. drummondii D. Don and Z. chlorosolen (Herb.) D. Dietr. at stages from emer- gence from the bulb through 48 hours post-anthesis. Zephyranthes drummondii was collected from two sites in Austin, Texas: at the intersection of 27th Street and Speedway, and at a hilltop on the east side of Interstate 35 just south of its interchange with U.S. 183. Zephy- ranthes chlorosolen was collected along the entrance ramp of U.S. 183 onto southbound Interstate 35. Excised ovules of both species were fixed overnight in FPA50, which is 37% formalin: glacial propionic acid: 50% etha- nol, 1:1:18 v:v:v (Herr, 1971), and dehydrated through 70%, 95%, and absolute ethanol. The ovules were infil- trated with methyl salicylate in three steps: 2:1 absolute ethanol: methyl salicylate, 1:2 absolute ethanol: methyl salicylate, and pure methyl salicylate. The dehydration and infiltration steps were minimally one hour. Ovules were viewed under differential interference contrast (Nomarski) as whole mounts in methyl salicylate with cover slips at the side to support an overlying cover slip. Variations of this method have been used subsequently by Young et al. (1979), Stelly et al. (1984), and Zeng et al. (2007); a recent application appeared in Kwiatkowska et al. (2019). Buds of Z. drummondii were collected during the spring of 1976 at stages from emergence through early endosperm development. Buds of Z. chlorosolen were collected in triplicate in seven specific groups from 7 July 1976 through 10 July 1976 after rainfall on 4 July 1976, in order to survey development in pollinated and unpollinated flowers before, during, and shortly after the usual time of self-pollination. The Z. chlorosolen groups are detailed in Table 1 (below). All the flowers in the first six rows of Table 1 were emasculated at sunset one or two days before opening. Pollinations were performed within an hour with pollen from opening flowers else- where in the population. The stigma was removed from unpollinated flowers to prevent pollination. The stigma was slightly exserted above the anthers in the three flow- ers of the seventh row in Table 1, and these flowers were self-pollinated upon anthesis. After a delay of 48 to 72 Table 1. Batches of Z. chlorosolen flowers used in this study. Codea Emasculation date Pollination status Date of anthesis Date picked Age relative to anthesis when picked -2e+72 7 July Unpollinated 9 July 10 July +1 day -2pol+72 7 July Pollinated 9 July 10 July +1 day -1e+48 7 July Unpollinated 8 July 9 July +1 day -1pol+48 7 July Pollinated 8 July 9 July +1 day -1e+72 7 July Unpollinated 8 July 10 July +2 days -1pol+72 7 July Pollinated 8 July 10 July +2 days 0pol+48 8 July Pollinated 8 July 10 July +2 days aCodes consist of number of days relative to anthesis (0, -1, -2), pollination status (pollinated or emasculated and unpollinated, and the approximate number of hours after pollination or emasculation when the flower was picked for fixation. 108 Charles F. Crane hours, picked flowers were brought indoors, with ovule excision and fixation commencing immediately for the first flower processed. The other two flowers per treat- ment were held at 4C until earlier flowers had been pro- cessed. About 30 of the 50 to 80 total ovules were ran- domly sampled per flower. Mature embryo sacs of Z. chlorosolen were classified as normal or abnormal on the basis of nuclear and cell count. A normal embryo sac consisted of one egg, two synergids, three antipodals, and a binucleate central cell whose nuclei ultimately fused. Abnormal embryo sacs differed in count, usually as a result of non-division of a nucleus at an earlier stage. Synergids were classified as having a filiform apparatus, which usually coincid- ed with a micropylar-end position of their nucleus and a chalazal-end vacuole. The egg did not have a filiform apparatus and had a chalazal or lateral nuclear position and a micropylar-end vacuole. Free nuclei in the central cell were classified as polar nuclei. Sometimes the antip- odal cells resembled a second egg apparatus in nuclear and vacuolar positions, such that one antipodal had a nucleus closer to a polar nucleus. Abnormalities were tabulated in an attempt to discern if there was a pattern of successive determination of nuclear fates. Synergid and egg degeneration were also followed in relation to ageing and pollination. Degeneration was indicated by cytoplasmic collapse, nuclear shrinkage, and general loss of visible cellular content. RESULTS Gametophytic maturation in Zephyranthes drummondii Most fertile ovules had reached the four-nucleate stage as the bud emerged from the neck of the bulb at the soil surface, but a few were still binucleate. The embryo sac was discoid at this stage, and its micropy- lar end directly abutted the nucellar epidermis. There was a large central vacuole, which persisted throughout further development until the endosperm cellularized in fertilized ovules. A tapering cytoplasmic strand, nar- rowest at the middle, traverses the vacuole after the first mitosis, but this strand disappears before the second mitosis. A prominent hypostase was fully developed at the chalazal end of the ovule, and it persisted well into endosperm development. Four-nucleate embryo sacs (Fig. 1A) and eight-nucle- ate embryo sacs lacked any visible cytoplasmic strands that span the central vacuole. In most ovules, the last mitosis occurred on the third day before flower opening. Mitosis at the chalazal end of a four-nucleate embryo sac was observed to precede mitosis at the micropylar end, but it is not known if this is generally the case. The post- telophase daughter nuclei at the micropylar end already occupied the positions expected of nuclei in the egg apparatus (Fig. 1B), and their flattened shape indicated that the two mitotic spindles had been perpendicular to each other. The prospective egg nucleus and micropylar polar nucleus were already larger than the prospective synergid nuclei on the second day before flower open- ing. Meanwhile, the divisions at the chalazal end of the embryo sac were more difficult to see because of the thick, birefringent walls of the hypostase and the pres- ence of up to 20 cell layers in the light path. Neverthe- less, the last mitotic spindles there appeared to be mutu- ally perpendicular as they were at the micropylar end, Figure 1. Early development and migration of polar nuclei in Z. drummondii. The micropylar end is to the left or down in each pic- ture. A. Tetranucleate embryo sac with nuclei side by side at each end. B. Flattened nuclei soon after telophase at the micropylar end, indicating orthogonal spindles; e, predicted egg nucleus; mp, predicted micropylar polar nucleus; s, predicted synergids, based on frequently occurring positions in the mature egg apparatus. C. More mature, fully cellular egg apparatus with egg nucleus (at left) unusually close to the micropylar end of the egg cell. The filiform apparatus of a synergid (right) appears feltlike or fibrillar. D. Mature antipodal apparatus in contact with the hypostase, whose walls are thickened and birefringent. The migrated but unfused polar nuclei lie immediately to the left. E. Inception of the central column. There are also smaller, variously oriented cytoplasmic strands that appear (with light microscopy) to have intruded into a previously uninterrupted central vacuole. The egg apparatus of this embryo sac appeared in C. F. The central column has reached full thickness. G. The micropylar polar nucleus has entered the central column first. H. Both polar nuclei have entered the central column. I. Striations between the approaching polar nuclei are possibly cytoskeletal ele- ments. J. The polar nuclei can meet near the egg apparatus, or move there temporarily after meeting at the center. 109Megagametophyte differentiation in Zephyranthes and a candidate chalazal polar nucleus was evident far- thest from the hypostase before cellularization. Cellularization occurred by the day before flower opening. The filiform apparatus began to develop (Fig. 1C) in the synergids. The egg nucleolus developed a nucleolar vacuole (Fig. 1C). After cellularization, one of the antipodal nuclei was larger than the other two (Fig. 1D), just as the egg nucleus was larger than the syner- gid nuclei, and this distinction was even more evident in unfertilized ovules post-anthesis. The polar nuclei migrated on the day before open- ing, after the initial cellularization of the egg and antip- odal apparati. Before migration, both polar nucleoli had formed a nucleolar vacuole. Thin strands of cytoplasm began to intrude into the central cell vacuole (Fig. 1E), and the central strand soon spanned the vacuole. The central strand continued to thicken and developed a granular appearance (Fig. 1F). The polar nuclei migrated into the central strand; in Fig. 1G the micropylar nucleus has moved first. The polar nuclei approached each oth- er in the strand (Fig. 1H), and faint striations indicated the presence of cytoskeletal elements (microtubules and/ or microfilaments) between them (Fig. 1I). Although the polar nuclei could meet relatively near the egg appara- tus (Fig. 1J), they finally moved in unison to the chalazal end of the central cell (Fig. 1D). Then the central cyto- plasmic strand began to separate into separate strands and disappear (Figs. 2A and 2B; also Fig. 1D) on the day of flower opening. Over time the polar nuclei became appressed (Fig. 2C) and then fused, as indicated by fusion of their nucleoli (Fig. 2D). In one abnormal case with extra polar nuclei in lieu of synergids, all four polar nuclei became trapped within the middle of the central strand and maintained their distinctness through three days past flower opening (Fig. 2E). The abnormality with four trapped polar nuclei was seen several times also in cleared ovules of Hippeastrum xjohnsoni (H. reginae (L.) Herb. x H. vittatum (L’Her.) Herb.). Fertilization occurred on the second day post-open- ing. Figure 2F possibly depicts plasmogamy of the egg and sperm cells. Later on the smaller sperm nucleus was visible within the egg, and it could divide before being walled off from cb of the bicellular embryo (Figs. 2G and 2I). Some eggs partially collapsed after plasmogamy (Fig. 2H), but most maintained an expanded, semicircu- lar shape leading to elongation toward the chalaza. The initial division plane was usually transverse. The pollen tube was sometimes visible. In one ovule that had been punctured during excision and handling, where the surrounding synergid debris had been lost, there was an apparent pore at the hooked pollen tube tip where sperms had been released. Karyogamy appeared to be necessary for endosperm development. In one abnormal instance, karyogamy was incomplete on the tenth day after floral opening, and the partially decondensed, distended sperm nucleus was still appressed to the polar fusion nucleus, which had not divided. In another ovule, the sperm and polar fusion nucleus were close to each other, and both were degen- erating. Further evidence for the necessity of karyogamy in hemigamous Zephyranthes and Habranthus comes from the results of interspecific pollinations, which often result in an increased frequency of endosperm failure and empty seeds. In H. tubispathus x Z. candida (Lindl.) Herb., for example, all the seeds were empty in spite of seed setting of more than 95%. The primary endosperm nucleus divided transversely and thus produced a small chalazal cell covering the antipodals and a far larger Figure 2. Later development in Z. drummondii, except J, which is from Habranthus robustus pollinated with Z. macrosiphon. A, B. Dissipation of the central column, which splits into parallel strands. C. Appressed polar nuclei next to antipodals. D. Completely fused polar nuclei with fused nucleoli. E. Persistent central column in abnormal embryo sac with four polar nuclei and no synergids. F. Egg cell during or shortly after plasmogamy with a sperm cell. G. Bicellular embryo; one of two sperm-derived daughter nuclei (arrow) lies below the maternal nucleus in cb. H. Zygote shrink- age or degeneration. I. Both sperm-derived nuclei (s) have been walled off from cb in the same embryo as G. J. Intranuclear meta- phase of endosperm nucleus in the micropylar chamber of the helo- bial endosperm. K. Highly endopolyploid, multinucleolate nucleus in failing endosperm. L. Sterile ovule with uninucleate embryo sac whose nucleus resembles a polar nucleus and occupies the chalazal position of migrated polar nuclei. 110 Charles F. Crane micropylar cell containing the rest of the volume of the central cell. Free-nuclear divisions ensued in both cells, but no more than five nuclei were observed in the cha- lazal cell. Abundant free-nuclear mitoses in the micro- pylar cell resulted in a multinucleate shell that laid down cell walls centripetally as it began to fill in the central vacuole. The first mitoses in the micropylar cell were synchronized and resulted in 2, 4, 8, 16, or 32 nuclei at a time before synchrony broke down, whereas mitoses in the chalazal cell were not synchronized. The early endosperm nuclei were large enough to hold an entire spindle apparatus within the old nuclear membrane (Fig. 2J). The nuclei of mature, fully cellular endosperms were smaller and spherical. Lobate, multinucleolate endosperm nuclei sometimes appeared, but they seemed to be associated with endosperm failure. They appeared to be under traction by attached spindle fibers (Fig. 2K; this example is from Habranthus robustus Herb. ex Sweet pollinated with Z. carinata Herb.). The fate of unfertilized embryo sacs was also exam- ined. Neither the egg nucleus nor the fused polar nuclei ever divided. Both synergids usually degenerated one or two days before the egg did. The degenerating egg appa- ratus lost evident vacuoles as the nuclei faded out. The cytoplasm collapsed, i.e., the egg shrank and its bound- ary became crenulate. The antipodals degenerated at the about same time as the synergids, and they could be crushed by proliferating nucellar cells near the incom- pressible hypostase. The central cell nucleus was usually the last nucleus to degenerate in the embryo sac. From five to 20% of the ovules were sterile, lacking a mature embryo sac. Although both integuments were of normal size, the nucellus was smaller. Sterile ovules fell into four main types: those with a uninucleate embryo sac in a hypodermal position, those with a slightly enlarged megasporocyte in a hypodermal or subhypo- dermal position, those with no enlarged cells at all, and those with the megasporocyte surrounded by thickened cell walls within the hypostase. Sterile ovules were most frequent toward the base of the ovary, but they could occur anywhere. Figure 2L shows a uninucleate embryo sac with an enlarged nucleus and prominent nucleolar vacuole. Pollination response and senescence in Zephyranthes chlo- rosolen The experimental layout appears in Table 1. Obser- vations centered on timing of fertilization relative to pollination, synergid degeneration, sequence of degen- eration of cells in unfertilized embryo sacs, timing of first division in the egg versus the endosperm, frequency of spermatic division in the egg, and numerically abnor- mal embryo sacs. In total, 796 ovules were examined. They were classified as unfertilized normal, fertilized normal, unfertilized abnormal, or sacless, depending on presence of an embryo sac, number of components in the embryo sac, and evidence of fertilization such as dividing endosperm or presence of a pollen tube at the micropyle or presence of a sperm nucleus within the egg. There were 524 unfertilized normal, 122 fertilized nor- mal, 63 unfertilized abnormal, and 87 sacless ovules in all. Table 2 gives the numbers of instances for all condi- tions of embryo sacs and their components versus treat- ment. Supplemental Table 1 gives the same information divided among the 21 individual flowers sampled. No embryo or endosperm developed in emasculated, unpollinated flowers. Pollen tubes reached the ovules about 48 hours after pollination, but their growth rate depended on the time of pollination, and their arriv- al continued over a period of many hours. Thus in the “Egg” rows of Table 2, most of the fertilizations occurred between 48 and 72 hours after pollination in buds pol- linated the day before natural anthesis, whereas most of the fertilizations had occurred less than 48 hours after pollination in flowers pollinated upon opening. No pol- len tubes reached the ovules after pollination two days before opening. In spite of the slower growth rate, pollen tubes reached the ovules sooner after early pollination, as evidenced by the presence of multicellular embryos only after early pollination. At the times sampled, 110 ovules had received one pollen tube, seven had received two, and one had received three; the number was not noted or pollen tubes were not seen in the other evident- ly fertilized ovules. If only the early pollinated buds col- lected 72 hours later are considered, these totals became 65, two, and one. When more than one pollen tube had reached an ovule, they followed different paths toward the embryo sac. No instances of two sperm nuclei were seen in undivided egg cells. The synergids degenerated in both fertilized and unfertilized ovules, but pollination accelerated degen- eration (Table 2, “Synergid” rows). At least one of the synergids had begun to degenerate in every fertilized embryo sac, whereas both synergids were intact in a sub- stantial minority of embryo sacs in unpollinated flowers of the same age. Synergids were more degenerated in the unfertilized ovules of pollinated flowers than they were in unpollinated f lowers. Although the two synergids generally did not degenerate exactly synchronously, the combination of intact and fully degenerated synergids was relatively uncommon and appeared mostly in ferti- lized ovules. The independence of degeneration was test- ed with the chi-squared test. For ovules collected from 111Megagametophyte differentiation in Zephyranthes Table 2. Classification of Z. chlorosolen embryo sacs and their components, merged by time of emasculation, pollination, and collection. Treatmenta -2e -2pol -1e -1pol -1e -1pol 0pol Hours post 72 72 48 48 72 72 48 Unfertilized 96 91 94 91 96 22 33 Fertilized 0 0 0 6 0 70 47 Abnormal 9 8 15 4 3 7 17 Sacless 8 13 12 3 9 10 32 Total 113 112 121 104 108 109 129 Frac.abnorm.b 0.08 0.071 0.124 0.038 0.028 0.064 0.132 Frac.sacless 0.071 0.116 0.099 0.029 0.083 0.092 0.248 Frac.(ab+sa) 0.15 0.188 0.223 0.067 0.111 0.156 0.38 unf unf unf unf fert unf unf fert unf fert Egg OK 95 89 87 83 5 91 20 59 33 47 Embryo 0 0 0 0 0 0 0 8 0 0 Deging 1 0 2 5 1 4 1 3 0 0 Deg 0 2 5 3 0 1 1 0 0 0 Synergids OK-OK 22 16 19 16 0 10 4 0 0 0 OK-+/-OK 4 4 11 8 0 6 2 0 3 0 OK-deging 10 2 2 4 0 7 0 1 1 3 OK-deg 9 13 4 2 0 3 0 9 1 11 +/-OK-+/-OK 3 5 17 11 0 7 0 0 2 1 +/-OK-deging 5 6 6 8 1 5 1 1 4 1 +/-OK-deg 9 7 10 14 2 12 2 9 1 5 Deging-deging 9 10 3 4 0 11 5 2 6 2 Deging-deg 8 4 7 8 1 16 0 14 8 9 Deg-deg 17 24 15 16 2 19 8 34 7 15 Polar nuclei OOP 5 8 0 1 0 0 0 0 0 1 Deging 9 10 2 3 0 11 0 1 3 6 Ch appress 63 54 69 69 3 12 2 0 4 7 Ch fusing 14 11 19 13 2 28 6 5 10 5 Ch fused 5 5 4 4 1 45 14 9 15 17 Mult. PEN 0 0 0 0 0 0 0 3 0 5 >=2 endosp 0 0 0 0 0 0 0 50 0 6 1 mc + 1 ch 0 3 0 1 0 0 0 1 1 0 Antipodals 3: all OK 75 72 59 68 4 47 12 18 9 13 3: all+/-OK 5 1 3 2 0 7 1 7 1 2 3: all deging 2 0 1 2 0 2 3 9 5 4 3: all deg 1 1 0 0 0 3 2 16 4 13 2+/-OK 1deging 3 6 7 7 0 11 3 4 5 4 1+/-OK 2deging 3 1 5 2 0 11 1 9 2 2 5 any 0 1 0 2 0 0 0 0 0 0 4 any 3 4 3 2 1 2 0 1 0 1 2 any 3 3 14 5 0 8 0 3 5 4 1 any 1 0 1 1 1 5 0 3 2 3 none seen 0 2 1 0 0 0 0 0 0 1 Code UUUU 16 13 11 9 0 6 3 0 0 0 UDUU 59 56 50 55 5 44 9 13 10 13 112 Charles F. Crane unpollinated flowers 24 hours after natural opening (the first and third columns of Table 2, “Synergid” rows), there were 76 in which neither synergid had greatly degenerated, 55 in which only one synergid had degen- erated, and 59 where both had degenerated. This gave a probability of 0.5447 of not having degenerated versus 0.4553 of having degenerated. If degeneration occurred at random, one would expect 56.373 with neither syn- ergid degenerated, 94.241 with one degenerated, and 39.387 with both degenerated. The resulting chi-squared value was 32.94 with two degrees of freedom, and thus the synergids did not degenerate independently. The polar nuclei had failed to migrate in six embryo sacs, and they were out of their usual chalazal-end posi- tion in 15 more embryo sacs (Table 2, “Polar nuclei” rows). Eighteen of these instances occurred in flowers emasculated two days before opening, even though the flowers were collected at the same age as those emas- culated one day before opening. Most polar nuclei were appressed at 24 hours after natural opening. While the polar nuclei were aligned along the long axis of the embryo sac during migration, they assumed other align- ments upon reaching the chalazal end, and about a 20:4 ratio of perpendicular to parallel alignments was observed relative to the long axis. Most polar nuclei had begun to fuse or had completely fused, as indicated by complete fusion of their nucleoli, at 48 hours past natural opening. Pollination accelerated nuclear and nucleolar fusion of polar nuclei in both fertilized and unfertilized ovules. The primary endosperm nucleus developed multiple nucle- oli before dividing. The large number of multinucleate endosperms after pollination one day early further indi- cated the earlier time of fertilization in that group. One or two extra antipodal cells occurred in 18 embryo sacs, but the divisions that produced them were not seen (Table 2, “Antipodal” rows). Fertilization, especially endosperm development, accelerated antipo- dal degeneration. The unfertilized ovules of pollinated flowers also had more degenerated antipodals than did ovules in unpollinated flowers of the same age. Asyn- chronous degeneration was more likely to be observed in unpollinated flowers. The condition of each embryo sac was encoded in Table 2 in four letters, all either U for intact or D for degenerating and degenerated. The letters respectively denoted the egg, synergids, polar nuclei, and antipo- dals, and any instance of degeneration merited a “D” for synergids and antipodals. Completely intact embryo sacs were most common at 24 hours past natural open- ing, and pollination two days early did not affect them. Later pollination accelerated synergid and antipodal degeneration. Completely intact embryo sacs were never seen in fertilized ovules, since at least one synergid had degenerated. Antipodal degeneration usually began soon after synergid degeneration had begun. In 20 ovules, the entire egg apparatus was degenerating before the polar nuclei or antipodals began to degenerate. The egg and the polar nuclei were about equally likely to be the last intact component in senescing embryo sacs. However, all possible degeneration sequences were observed multiple times. Endosperm division clearly preceded egg division (Table 2, ninth column). The egg usually divided after at least eight nuclei existed in the micropylar cham- ber of the helobial endosperm. In four out of 122 fer- tilized ovules, the egg degenerated without dividing; this is a typical frequency of mature seeds with plump endosperm but no embryo in Z. chlorosolen. The number of sperm nuclei was tabulated in 83 fertilized egg cells. Three of them had experienced division of both sperm Treatmenta -2e -2pol -1e -1pol -1e -1pol 0pol UDUD 9 7 19 11 0 28 7 51 20 27 UDDU 1 7 2 0 0 2 0 0 1 0 UDDD 4 1 0 0 0 7 0 2 1 6 UUUD 2 2 4 5 0 3 1 0 0 0 UUDU 2 1 1 2 0 0 0 0 0 0 UUDD 2 0 0 1 0 1 0 0 1 0 DDUU 1 2 4 8 1 3 0 2 0 0 DUUU 0 0 3 0 0 0 0 0 0 0 DDUD 0 0 0 0 0 2 2 1 0 0 aTreatment and “hours post” refer to pollination versus emasculation and the number of hours after either when picked. bLeft column abbreviations are: Frac.sacless, fraction of all ovules that lacked an embryo sac; Frac.abnorm, fraction of all embryo sacs that were abnor- mal; Frac.(ab+sa), sum of Frac.sacless and Frac.abnorm.; deg, degenerated; deging, degeneration in progress; OK, intact; OOP, out of posi- tion; ch, chalazal; mc, micropylar; mult. PEN, multinucleolate primary endosperm nucleus; appress, appressed; any, any condition. The UD codes follow a system given in the Results section in reference to Table 2. 113Megagametophyte differentiation in Zephyranthes and egg nuclei. The remainder contained only one sperm nucleus. There were 63 embryo sacs that differed from the conventional organization of one egg, two synergids, two polar nuclei, and three antipodals. These are described and enumerated in Table 3, whose columns give the row number in the table, the number of instances observed, the total number of nuclei, the counts of eggs, syner- gids, free nuclei in the micropylar part of the central cell, free nuclei in the chalazal part of the central cell, and antipodals. Because most of the embryo sacs were examined after the polar nuclei had met and migrated into the chalazal end of the central cell, the last two col- umns give a best guess as to the numbers of free nuclei at each end of the central cell prior to migration. Most abnormal embryo sacs contained fewer than eight nuclei. The most common abnormalities were misdifferentiation of an antipodal nucleus as a third polar nucleus (row 1), and absence of the egg apparatus (rows 10 and 15). There were 13 embryo sacs with only one antipodal cell, and in four of these (rows 8 and 17) there appeared to be only two chalazal-end nuclei prior to cellularization. The number of functioning polar nuclei ranged from one to four; the latter occurred when both expected syner- gid nuclei instead became polar nuclei and behaved as in Fig. 2E. Although no instances were observed in Z. chlo- rosolen, four polar nuclei and no synergids (row 5) were seen multiple times in Hippeastrum xjohnsonii. Eight embryo sacs could not be described as com- binations of egg, synergids, polar nuclei, and antipo- dals. In the first, the antipodal nuclei were not walled off from the central cell in an otherwise normal embryo sac. Instead, they remained separate from the two fused polar nuclei. In the second and third, the synergids lacked a filiform apparatus and the polar nuclei were still separate at the ends of the central cell 24 hours after normal anthesis. In the fourth, only the cell walls of the egg apparatus persisted, and all nuclei had degenerated. In the fifth, there was but one antipodal, and it was unu- sually large. The two polar nuclei flanked this antipodal, and the egg apparatus had degenerated. In the sixth, a transverse partition near the micropylar end divided the sac into two cells, each with four free nuclei seemingly randomly arranged near the partition. In the seventh, the synergid and egg nuclei occupied the same, egglike cell. In the eighth, the egg nucleus had been walled off partially, but then the egg nucleus had migrated to the chalazal end of the central cell as a polar nucleus. The embryo sac was otherwise normal. DISCUSSION Development of the embryo sac from the megaspo- rocyte in apomictic Zephyranthes is similar to develop- ment from the surviving megaspore in Ornithogalum caudatum (Tilton and Lersten, 1981). The biggest differ- ences are the larger size and more discoid shape of the embryo sac and its central vacuole in Zephyranthes, such that the four-nucleate stage has two nuclei side-by-side at each end, rather than arranged linearly as is usual in Ornithogalum. A second difference is the stated move- ment of only the micropylar polar nucleus through the central column to the chalazal end in Ornithogalum, versus migration of both polar nuclei into the column in Zephyranthes before migration of both to the chala- zal end. Otherwise, Tilton and Lersten (1981) observed the initiation, thickening, and splitting of the column Table 3. Component counts in numerically abnormal embryo sacs in Z. chlorosolena. row N total eggs syn cent mc cent ch antip inf mc inf ch 1 7 8 1 2 0 3 2 1 2 2 1 8 0 2 0 3 3 2 1 3 1 8 1 1 0 3 3 2 1 4 1 8 1 1 0 4 2 2 2 5 0 8 1 0 0 4 3 3 1 6 2 7 1 2 0 1 3 0 1 7 1 6 1 0 0 2 3 1 1 8 3 6 1 2 0 2 1 1 1 9 1 6 0 2 0 2 2 1 1 10 6 5 0 0 0 2 3 1 1 11 1 5 1 0 0 1 3 0 1 12 2 5 1 2 0 0 2 0 0 13 1 5 1 0 1 2 1 2 1 14 2 5 1 2 0 1 1 1 0 15 4 4 0 0 0 1 3 0 1 16 2 4 0 0 0 2 2 0 2 17 3 4 1 0 0 2 1 1 1 18 1 4 1 2 0 1 0 1 0 19 2 3 0 0 0 1 2 0 1 20 4 3 0 0 0 3 0 1 2 21 1 3 0 0 0 2 1 1 1 22 1 3 1 0 1 1 0 1 1 23 3 2 0 0 0 1 1 0 1 24 3 2 1 0 0 1 0 1 0 25 1 1 0 0 0 1 0 0 1 asyn, synergids; cent mc, free nuclei in micropylar half of the cen- tral cell; cent ch, free nuclei in chalazal half of the central cell; antip, antipodal cells; inf mc, inferred to come from the micropylar end of the central cell; inf ch, inferred to come from the chalazal end of the central cell. 114 Charles F. Crane much as in Zephyranthes, although they did not spe- cifically order these stages in time. Tilton and Lersten mentioned other examples of the central column from various taxa, including Crocus, Hordeum, and in gener- al any species with helobial endosperm. They also cited the notion, from a 1902 paper by Ikeda, that at least one polar nucleus moves through the column. More recently, Zeng et al. (2007) and Hu et al. (2009) also observed in rice the migration of polar nuclei through a transient, de-novo formed cytoplasmic column, where they met at the center and then moved to the micropylar end of the central cell. Tilton and Lersten (1981) suggested that in Ornithogalum the second sperm nucleus might reach the fused polar nuclei via the central column, but in Zephy- ranthes drummondii this column has dissipated prior to arrival of the pollen tube at a synergid, and therefore the sperm must take some path through the peripheral cyto- plasm of the central cell. Embryo sac development in emergent buds of Z. drummondii and Z. chlorosolen is more rapid than in the related Zephyranthes candida, at four days in the former and up to seven days in the latter. As in Z. drummon- dii, Ao (2018) reported (on the basis of 10-μm paraffin sections) the formation of a central cytoplasmic column that spanned the central cell and conveyed the micro- pylar polar nucleus to the chalazal end. However, Ao (2018) reported that the column persisted until fertili- zation and that sometimes the polar nuclei and sperm nucleus fused within the chalazal part of it, as depicted in his Figure 2D. Ao did not distinguish a stage where both polar nuclei moved into the column and met near its midpoint. Ao (2018, 2019) also noted that the antipo- dal cells developed callosic walls when the polar nuclei approached them, that the antipodal cells later fused when the callose had disappeared, and that the antipo- dal nuclei ultimately fused into one. In contrast, in Z. drummondii no callose or cellular fusion was observed in the antipodal apparatus before it degenerated. Also, Ao (2019) noted a nuclear endosperm in Z. candida, and this in conjunction with the claimed fusion of antipo- dal cells and claimed increase in antipodal nuclei to nine is consistent with a misinterpretation of a helobial endosperm where the small chalazal chamber has been mistaken for fused antipodal cells. The inception, thickening, and eventual dissipation of a central column differ markedly from the descrip- tion of polar-nuclear migration in Arabidopsis thaliana. There the polar nuclei are depicted as traveling through a peripheral cytoplasmic shell surrounding the central vacuole, and they migrate while the egg and antipodal apparati form cell walls (Yadegari and Drews, 2004). In rice, the photographs of Zeng et al. (2007) and Hu et al. (2009) were not suitable to detect incipient cell walls, since the nuclei were stained preferentially, but it is pos- sible that the migration began before cellularization began. The egg apparatus appears to develop complete cell walls in Zephyranthes. This is supported by Fig. 1C, which is the egg apparatus from the embryo sac whose pre-migration central cell appears in Fig. 1E. A Zephy- ranthes egg cell tends to maintain its hemispherical shape even when plasmolysis pulls the central cell cyto- plasm away from it, which is consistent with a rigid egg-cell wall. Also, Tilton and Lersten (1981) noted the appearance of a complete cell wall around the egg in Ornithogalum caudatum, but commented that the chala- zal end of the wall might be reticulate at ultrastructural magnification. Synergid degeneration appears to be autonomous in Zephyranthes, but pollination accelerates degeneration in unfertilized as well as fertilized ovules. Degeneration of one synergid is positively correlated with degeneration of the other synergid. Although up to three pollen tubes were observed at the micropyle, there is no evidence in this study that intact synergids attracted pollen tubes. It is possible that penetration necessarily destroyed the receiving synergid. The situation is concordant with the conclusions of Leydon et al. (2015) and cited references therein, that synergids can degenerate autonomously, that approach of a pollen tube promotes synergid degen- eration, and that discharge of a pollen tube is observed only in a degenerated synergid. The embryo-sac abnormalities observed in Zephy- ranthes are not consistent with a random assignment of function at the eight-free-nucleate stage, since combina- tions such as three egg cells or one synergid and three micropylar-end polar nuclei were not observed. On the other hand, there was insufficient evidence to support a rigidly progressive imprinting of nuclear fate, with con- sistent defaults for undivided nuclei from each of the three rounds of mitosis. For example, embryo sacs with only one synergid were rare (two out of 63 abnormals), and thus the nuclei that normally become synergid nuclei tended to share the same fate, either becoming synergids or becoming extra, functioning polar nuclei. For this reason, one might conclude that the commit- ment to becoming synergid nuclei has already happened at the four-nucleate stage of the embryo sac and that it is distinct from the commitment to divide. Yet the com- bination of no synergids and three polar nuclei was not observed, so the usual fate of an undivided synergid precursor is unclear. For another example, an undi- vided micropylar end nucleus (one having skipped the second and third mitotic divisions) tended to function 115Megagametophyte differentiation in Zephyranthes as a polar nucleus, moving to the chalazal end of the central cell upon cellularization. For that matter, even an undivided megasporocyte nucleus could develop the chalazal position and single, large, vacuolated nucleolus typical of migrated polar nuclei (Fig. 2L). In contrast, if the second round of mitosis were skipped, each daughter of the third round could form a polar nucleus and one of something else. Furthermore, failure to wall off one antipodal resulted in an extra free nucleus that was not observed to become appressed to or fuse with the legiti- mate polar nuclei. Thus very careful, systematic evalua- tion of a much larger sample of properly timed, abnor- mal embryo sacs, with attention to the filiform appara- tus, nuclear position, and cellular and nucleolar vacu- olization, will be needed to establish if the spectrum of abnormalities can support the inference of checkpoints and default fates in differentiating embryo sacs. The spectrum of abnormalities varies among plant taxa. Yudakova (2009) observed a 10- to 100-fold high- er frequency of extra eggs and polar nuclei in embryo sacs of the Hieracium type in Poa chaixii and Poa prat- ensis than in embryo sacs of the Taraxacum (or pos- sibly Elymus rectisetus) type in Poa badensis. Yudakova (2009) also stated that, “No significant cases of extra polar nucleus formation from the synergid nuclei were recorded among numerous analyzed embryo sacs of the studied bluegrass species, while female gametophytes with additional egg cells instead of synergids occurred in all studied plants.” This situation markedly contrasts with the greater frequency of expected synergid nuclei becoming extra polar nuclei in Zephyranthes and Hip- peastrum. In semifertile hybrids of indica × japonica rice, Zeng et al. (2007) observed a wide range of abnor- malities over stages from degenerating megaspores through mature embryo sacs. At maturity, frequent abnormalities included small size, a misplaced (lateral) egg apparatus, absence of the egg apparatus, misplaced (lateral) antipodals, and migration of polar nuclei to the chalazal end (which would be normal in Zephyranthes). If the egg apparatus was absent, the micropylar-end polar nucleus was either present or absent. Otherwise, Zeng et al. (2007) did not detail numerical abnormalities in the manner of Table 3. The failure to observe multiple egg cells in Zephy- ranthes (Table 3) could reflect observational bias, since the filiform apparatus develops over time and is not always clearly visible in whole mounts, and the position of the nucleus varied from lateral to chalazal-end within egg cells. An indirect line of evidence comes from twin embryos. Spontaneous diploid-haploid twins are relative- ly easily found in Lilium (Cooper, 1940) and in grasses, where they are a source of spontaneous haploids apart from indeterminate gametophyte mutants. Within Zeph- yranthes, identical (maternal) twins are about 10 times more frequent in Z. pulchella J.G. Smith than in Z. chlo- rosolen. An extra egg would be the most facile source of such twins, although postzygotic cleavage is also possi- ble, and experimental distinction would be difficult in an apomictic species. An extra egg could account for frequent maternal haploid production from indetermi- nate gametophyte (ig) mutants in maize and rice (Evans, 2007; Zhang et al., 2015), although ig maize also pro- duces androgenetic haploids when used as a female par- ent with other maize genotypes (Kindiger and Hamann, 1993). Yudakova (2009) followed the four-zone hypothesis of Enaleeva (2002) to account for abnormalities in Poa embryo sacs. From the micropylar end, the consecutive zones are synergid, egg, central, and antipodal. Yudako- va (2009) conjectured that the zones were established in response to signals from adjacent nucellar cells and that each zone determined the fate of its contained nuclei. Abnormalities would arise when a free nucleus was situ- ated out of its proper zone, and a piece of evidence for this was the prevalence of an extra egg rather than polar nucleus in place of a synergid, since the synergid and central zones are not adjacent. However, in Zephyranthes and Hippeastrum the relative abundances are reversed, with a tendency of both prospective synergid nuclei to function as extra polar nuclei. Also, in Zephyranthes, the egg and synergid nuclei can begin at the same dis- tance from the micropylar extremity of the embryo sac (Fig. 1B). Furthermore, in rice (Zeng et al., 2007) and Habranthus tubispathus (Pace, 1913), an otherwise nor- mal egg apparatus sometimes appears on the side of the embryo sac rather than at the micropylar end. A lat- eral egg apparatus would require that the zones would respond to particular cells or cell groups in the nucel- lus and that free nuclei would move in response to some attraction from those nucellar cells. Alternatively, a lat- eral egg apparatus could easily arise from a misoriented spindle of the first mitotic division, if the nuclei move only slightly thereafter. Nuclear movement seems to occur only at particu- lar times in the differentiation of embryo sacs, first after the division of the megasporocyte (which functions as the surviving megaspore in apomictic Zephyranthes), then upon migration of the polar nuclei, then upon fertilization, and finally during the first few rounds of mitosis in the developing endosperm. The nuclei appear to remain nearly stationary apart from these four epi- sodes. The nature of the fertilization block is particularly interesting in Zephyranthes, since the sperm nucleus can approach within its own diameter the egg nucleus with- 116 Charles F. Crane out fusing with the egg nucleus. Also remarkable is how the other sperm nucleus moves or is moved across more than 100 micrometers of the peripheral cytoplasm of the central cell to reach the fused polar nuclei. Karyogamy of the sperm and fused polar nuclei appears to be necessary for endosperm development in Z. drummondii and Z. chlorosolen, as evidenced by an invariably undivided central cell nucleus in unfertilized ovules and the similar types of endosperm failures in interspecific crosses of these species and interspecific crosses of sexual Z. traubii (W. Hayw.) Moldenke. How- ever, this necessity contrasts with the situation described by Tandon and Kapoor (1962) in Zephyranthes cv. ‘Ajax’ (Z. candida (Lindl.) Herb. x Z. citrina Baker), where 4n chromosome counts were reported for 366 out of 430 endosperm metaphases after self-pollination. This sug- gests that hemigamy also operates in the central cell, assuming that this seed-propagatable cultivar is apom- ictic like its Z. citrina parent (Howard, 1996). Apomic- tic Zephyranthes possibly vary in their requirement for karyogamy for endosperm inception. If so, there is a very unanswered question as to how the usual maternal- paternal imprinting mechanism (endosperm balance number) is circumvented in Z. citrina and Z. pulchel- la. The problem merits further investigation with flow cytometry and paternal markers in all of these species. STUDY LOCATION The Z. drummondii flowers were collected at lati- tude 30°17’37.5”N, longitude 97°44’11.5”W (30.293736, -97.736538). The Z. chlorosolen f lowers were collect- ed at latitude 30°20’16.4”N, longitude 97°42’04.4”W (30.337887, -97.701208). Both populations have been destroyed by subsequent building and road construction. ACKNOWLEDGEMENTS I performed the experiments in partial fulfillment of the requirements for the Ph.D. degree at the University of Texas at Austin. I thank my now deceased major pro- fessor, Walter V. Brown, for stimulation and tolerance of this and related projects during my graduate research. I thank Stefan Kirchanski for suggesting methyl salicylate as a clearing agent. I thank Edwin T. Bingham, David A. Sleper, John G. Carman, David M. Stelly, and Stephen B. Goodwin for support over subsequent decades. Finally, I thank my wife, Yan M. Crane, for reformatting the pho- tomicrographs and for her steadfast support. DECLARATION The author declares no conflicts of interest regarding this research. REFERENCES Ao, C.Q. 2019. The endosperm development and the vari- ations of structures of embryo sacs: unraveling the low fecundity of Zephyranthes candida (Amarylli- daceae). Plant Biosystems. 153: 673-678. Ao, C.Q. 2018. Double fertilization in Zephyranthes can- dida, with special notes on the second fertilization and the behavior of the primary endosperm nucleus. Phyton. 58: 135-138. Ao, C.Q., Wang, L.Y., Sun, H., Lin, J.T., Chai, Y., and Chen, C.C. 2016. Megasporogenesis and megagame- togenesis in Zephyranthes candida (Amaryllidaceae), with special notes on the behavior of the synergids, the central cell and the antipodal cells. Phyton. 56: 91-101. Bashaw, E. C., and Hanna, W. W. 1990. Apomictic repro- duction. In: Chapman GP, ed. Reproductive versatility in the grasses. Cambridge, UK: Cambridge University Press, 100–130. Battaglia, E. 1945. Fenomeni citologici nuovi nella embri- ogenesi (semigamia) e nella microsporogenesi (dop- pio nucleo di restituzione) di Rudbeckia laciniata L. Nuovo Giorn. Bot. Ital. N. S. 52:34-38. Brown, W. V. 1951. Apomixis in Zephyranthes texana Herb. American Journal of Botany 38: 697-702. Brown, W. V., and Emery, W. H. P. 1957. Apomixis in the Gramineae, tribe Andropogoneae: Themeda triandra and Bothrochloa ischaemum. Botanical Gazette 118: 246-253. Chennaveeriah, M.S., and Patil, R.M. 1971. Apomixis in Blumea. Phytomorphology 21: 71-76. Coe, G. E. 1953. Cytology of reproduction in Cooperia pedunculata. American Journal of Botany 40: 335-343. Cooper, D. C. 1943. Haploid-diploid twin embryos in Lilium and Nicotiana. American Journal of Botany 30(6): 408-413. Crane, C. F. 1978. Apomixis and crossing incompatibili- ties in some Zephyrantheae. Ph.D. dissertation, The University of Texas at Austin. Crane, C. F., and Carman, J. G. 1987. Mechanisms of apomixis in Elymus rectisetus from eastern Australia and New Zealand. American Journal of Botany. 74: 477-496. Enaleeva, N.K. 2002. On the cytological mechanism of “normal” and “abnormal” cell differentiation in angi- 117Megagametophyte differentiation in Zephyranthes osperm megagametophytogenesis. Reproduktivnaya biologiya, genetika i selektsiya (Reproductive Biology, Genetics, and Selection). Saratov: Izd-vo un-ta, 2002, p. 47–54. Evans, M.M.S. 2007. The indeterminate gametophyte1 gene of maize encodes a LOB domain protein required for embryo sac and leaf development. The Plant Cell. 19: 46-62. Gustafsson, A. 1946. Apomixis in higher plants. I. The mechanism of apomixis. Lunds Universitets Arsskrift 42: 1-67. Herr, J. M. 1971. A new clearing-squash technique for the study of ovule development in some angiosperms. American Journal of Botany. 58: 785-790. Howard, T. 1996. Two new Zephyranthes species from Mexico. Herbertia. 51: 38-41. Hu, C. Y., Zeng, Y. X., Lu, Y. G., Li, J. Q., and Liu, X. D. 2009. High embryo sac fertility and diversity of abnormal embryo sacs detected in autotetraploid indica/japonica hybrids in rice by whole-mount eosin B-staining confocal laser scanning microscopy. Plant Breeding 128: 187-192. Juel, H.O. 1900. Vergleichende Untersuchungen über typische und parthenogenetische Fortpflanzung bei der Gattung Antennaria. Kongliga Svenska Vetenska- ps-Akademiens Handlingar 33: 1-59. Kindiger, B., and Hamann, S. 1993. Generation of hap- loids in maize: a modification of the indeterminate gametophyte (ig) system. Crop Science. 33: 342–344. Kwiatkowska, M., Kadluczka, D., Wedzony, M., Dedicova, B., and Grzebelus, E. 2019. Refinement of a clearing protocol to study crassinucellate ovules of the sugar beet (Beta vulgaris L., Amaran- thaceae). Plant Methods 15: 71. Lanaud, C. 1988. Origin of haploids and semigamy in Theobroma cacao L. Euphytica 38: 221-228. Leydon, A., Tsukamoto, T., Dunatunga, D., Qin, Y., Johnson, M., and Palanivelu, R. 2015. Pollen tube discharge completes the process of synergid degen- eration that is initiated by pollen tube-synergid interaction in Arabidopsis. Plant Physiology. 169: 485-496. Nogler, G. 1972. Genetik der Aposporie by Ranunculus auricomus s.l. W. Koch. II. Endospermzytologie. Ber- ichte Schweiz. Botanische Gesellschaft. 82:54-63. Nygren, A. 1967. Apomixis in the angiosperms. In Ruh- land, W., ed., Handbuch der Pflanzenphyiologie, vol. 18: 551-596. Pace, L. 1913. Apogamy in Atamosco. Botanical Gazette. 56: 376-394. Stelly, D.M., Peloquin, S.J., Palmer, R.G, and Crane, C.F. 1984. Mayer’s hemalum-methyl salicylate: A stain- clearing technique for observations within whole ovules. Stain Technology 59: 155-161. Talent, N., and Dickinson, T.A. 2007. Endosperm for- mation in aposporous Crataegus (Rosaceae, Spirae- oideae, tribe Pyreae): parallels to Ranunculaceae and Poaceae. New Phytologist 173: 231-249. Tandon, S.L, and Kapoor, B.M. 1962. Contributions to the cytology of endosperm in some angiosperms -- I. Zephyranthes Ajax Sprenger. Caryologia. 15: 21-41. Tilton, V. R., and Lersten, N. R. 1981. Ovule development in Ornithogalum caudatum (Liliaceae) with a review of selected papers on angiosperm reproduction. III. Nucellus and megagametophyte. New Phytologist 88: 477-504. Turcotte, E.L, and Feaster, C.V. 1969. Semigametic pro- duction of haploids in Pima cotton. Crop Sci. 9: 653- 655. Voight, P.W., and Bashaw, E.C. 1972. Apomixis and sexu- ality in Eragrostis curvula. Crop Science 12: 843-847. Yadegari, R., and Drews, G.N. 2004. Female gametophyte development. The Plant Cell. 16: S133-S141. Young, B.A., Sherwood, R.T, and Bashaw, E.C. 1979. Cleared-pistil and thick-sectioning techniques for detecting aposporous apomixis in grasses. Canadian Journal of Botany. 57: 1668-1672. Yudakova, O.I. 2009. Abnormalities of female gameto- phyte development in apomictic bluegrass forms. Russian Journal of Developmental Biology 40: 150– 156. Zeng, Y. X., Hu, C. Y., Lu, Y. G., Li, J. Q., and Liu, X. D. 2007. Diversity of abnormal embryo sacs in indica/ japonica hybrids in rice demonstrated by confocal microscopy of ovaries. Plant Breeding 126: 574-580. Zhang, J., Tang, W., Huang, Y., Niu, X., Zhao, Y., Han, Y., and Liu, Y. 2015. Down-regulation of a LBD-like gene, OsIG1, leads to occurrence of unusual double ovules and developmental abnormalities of various floral organs and megagametophyte in rice. Journal of Experimental Botany 66: 99-112. Zhou, L.Z., and Dresselhaus, T. 2019. Friend or foe: Sign- aling mechanisms during double fertilization in flow- ering seed plants. Current Topics in Developmental Biology 131: 453-496. 118 Charles F. Crane Su pp le m en ta l T ab le 1 . C la ss ifi ca tio n of Z . c hl or os ol en e m br yo s ac s an d th ei r co m po ne nt s, it em iz ed in c ol um ns b y in di vi du al fl ow er . C od es a pp ea r as in T ab le 2 , e xc ep t F IX D , w hi ch is tim e of d ay w he n ov ul es w er e fix ed . C O N D -2 e7 2 -2 e7 2 -2 e7 2 -2 p7 2 -2 p7 2 -2 p7 2 -1 e4 8 -1 e4 8 -1 e4 8 -1 p4 8 -1 p4 8 -1 p4 8 -1 e7 2 -1 e7 2 -1 e7 2 -1 p7 2 -1 p7 2 -1 p7 2 0p 48 0p 48 0p 48 FI X D 20 :0 0 22 :0 0 1: 00 21 :0 0 0: 00 2: 00 22 :0 0 22 :0 0 23 :0 0 22 :0 0 23 :0 0 - 20 :0 0 23 :0 0 1: 00 21 :0 0 23 :0 0 1: 00 23 :0 0 - - U N FE 33 33 30 28 30 33 32 31 31 33 31 27 31 33 32 11 8 3 12 7 14 FE R T 0 0 0 0 0 0 0 0 0 0 1 5 0 0 0 21 22 27 7 23 17 A BN O 5 3 1 5 3 0 5 6 4 0 0 4 1 2 0 1 6 0 9 4 4 SA C L 4 1 3 7 4 2 2 4 6 2 1 0 4 1 4 4 4 2 15 8 9 T O TA 42 37 34 40 37 35 39 41 41 35 33 36 36 36 36 37 40 32 43 42 44 FR A B 0. 11 9 0. 08 1 0. 02 9 0. 12 5 0. 08 1 0 0. 12 8 0. 14 6 0. 09 8 0 0 0. 11 1 0. 02 8 0. 05 6 0 0. 02 7 0. 15 0 0. 20 9 0. 09 5 0. 09 1 FR SC 0. 09 5 0. 02 7 0. 08 8 0. 17 5 0. 10 8 0. 05 7 0. 05 1 0. 09 8 0. 14 6 0. 05 7 0. 03 0 0. 11 1 0. 02 8 0. 11 1 0. 10 8 0. 1 0. 06 2 0. 34 9 0. 19 0. 20 5 FR A S 0. 21 4 0. 10 8 0. 11 8 0. 3 0. 18 9 0. 05 7 0. 17 9 0. 24 4 0. 24 4 0. 05 7 0. 03 0. 11 1 0. 13 9 0. 08 3 0. 11 1 0. 13 5 0. 25 0. 06 2 0. 55 8 0. 28 6 0. 29 5 un f un f un f un f un f un f un f un f un f un f un f fe rt un f fe rt un f un f un f un f fe rt un f fe rt un f fe rt un f fe rt un f fe rt un f fe rt Eg g IN TA 33 33 29 27 29 33 30 31 26 31 29 1 23 4 29 31 31 9 18 8 21 3 20 12 7 7 23 14 17 EM BR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 6 0 0 0 0 0 0 D EG I 0 0 1 0 0 0 1 0 1 2 1 0 2 1 1 2 1 1 1 0 1 0 1 0 0 0 0 0 0 D EG E 0 0 0 1 1 0 1 0 4 0 1 0 2 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 Sy ne rg id s IT IT 5 10 7 3 7 6 14 2 3 6 3 0 7 0 0 2 8 0 0 3 0 1 0 0 0 0 0 0 0 IT PI 1 3 0 1 3 0 2 5 4 5 1 0 2 0 0 4 2 1 0 1 0 0 0 2 0 0 0 1 0 IT D N 3 6 1 0 2 0 0 1 1 2 0 0 2 0 2 2 3 0 0 0 1 0 0 1 0 0 1 0 2 IT D G 0 6 3 5 4 4 0 1 3 0 0 0 2 0 1 0 2 0 0 0 7 0 2 0 2 1 7 0 2 PI PI 1 1 1 1 2 2 3 6 8 2 6 0 3 0 3 0 4 0 0 0 0 0 0 2 0 0 0 0 1 PI D N 4 1 0 6 0 0 4 2 0 3 4 0 1 1 0 3 2 1 1 0 0 0 0 3 0 0 0 1 1 PI D G 4 2 3 1 4 2 1 6 3 6 6 1 2 1 6 4 2 0 2 2 2 0 5 0 1 0 2 1 2 D N D N 6 3 0 3 3 4 1 1 1 1 0 0 3 0 3 6 2 5 1 0 0 0 1 0 1 3 1 3 0 D N D G 5 1 2 2 1 1 3 4 0 4 3 0 1 1 9 5 2 0 8 0 3 0 3 3 1 2 6 3 2 D G D G 4 0 13 6 4 14 4 3 8 4 8 0 4 2 7 7 5 4 9 2 9 2 16 1 2 1 6 5 7 Po la r nu cl ei O O PO 2 0 3 4 0 4 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 D EG I 3 5 1 4 3 3 2 0 0 2 1 0 0 0 3 6 2 0 0 0 0 0 1 2 1 1 5 0 0 C A PP 17 23 23 14 25 15 22 27 20 25 26 0 18 3 4 3 5 1 0 1 0 0 0 2 0 1 6 1 1 C FU S 7 4 3 3 2 6 7 4 8 4 3 1 6 1 11 7 10 2 0 2 4 2 1 4 0 2 1 4 4 C FU D 4 1 0 1 0 4 1 0 3 2 0 0 2 1 13 17 15 8 4 5 2 1 3 4 4 2 1 9 12 M PE N 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 1 0 1 0 4 0 0 EN D 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 15 0 15 0 20 0 0 0 6 0 0 M C C H 0 0 0 2 0 1 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 A nt ip od al s 119Megagametophyte differentiation in Zephyranthes 3I N T 24 29 22 22 22 28 17 22 20 20 26 1 22 3 16 12 19 4 2 7 11 1 5 5 1 0 4 4 8 3P IP 1 3 1 1 0 0 1 0 2 1 0 0 1 0 0 3 4 1 3 0 1 0 3 0 0 0 1 1 1 3D G N 1 0 1 0 0 0 1 0 0 1 0 0 1 0 1 1 0 2 5 1 3 0 1 4 0 0 3 1 1 3D EG 1 0 0 1 0 0 0 0 0 0 0 0 0 0 2 1 0 1 7 0 3 1 6 1 2 2 10 1 1 2P D N 1 0 2 2 1 3 2 3 2 3 3 0 1 0 6 2 3 2 1 0 0 1 3 1 1 2 1 2 2 1P D N 2 1 0 1 0 0 4 0 1 2 0 0 0 0 1 7 3 1 3 0 2 0 4 1 1 0 0 1 1 5A N Y 0 0 0 0 1 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4A N Y 2 0 1 0 2 2 0 1 2 1 1 0 0 1 2 0 0 0 0 0 1 0 0 0 0 0 0 0 1 2A N Y 0 0 3 1 2 0 5 5 4 3 1 0 1 0 3 3 2 0 0 0 1 0 2 0 0 2 3 3 1 1A N Y 1 0 0 0 0 0 1 0 0 1 0 0 0 1 0 4 1 0 0 0 0 0 3 0 1 1 1 1 1 U N SE 0 0 0 0 2 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 St at us U U U U 3 8 5 2 6 5 7 2 2 1 2 0 6 0 0 1 5 0 0 3 0 0 0 0 0 0 0 0 0 U D U U 22 18 19 15 18 23 9 23 18 18 23 1 14 4 17 11 16 4 1 4 9 1 3 4 1 2 4 4 8 U D U D 5 2 2 5 1 1 7 6 6 7 3 0 1 0 9 14 5 5 19 1 11 1 21 6 4 4 14 10 9 U D D U 0 1 0 4 1 2 2 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 1 0 0 0 0 0 U D D D 1 2 1 0 0 1 0 0 0 0 0 0 0 0 3 3 1 0 0 0 1 0 1 0 1 1 5 0 0 U U U D 0 0 2 1 0 1 4 0 0 3 0 0 2 0 0 0 3 0 0 0 0 1 0 0 0 0 0 0 0 U U D U 0 2 0 0 1 0 1 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 U U D D 2 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 D D U U 0 0 1 1 1 0 0 0 4 2 2 0 4 1 1 1 1 0 1 0 1 0 0 0 0 0 0 0 0 D U U U 0 0 0 0 0 0 2 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 D D U D 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 2 0 0 0 0 1 0 0 0 0 0 0 Substantia An International Journal of the History of Chemistry Vol. 2, n. 1 - March 2018 Firenze University Press