General characteristics of Myrtaceae

de Candolle (Reference de Candolle1828) divided the Myrtaceae into three tribes: Myrteae (with fleshy berries), Leptospermeae (dry dehiscent loculicidal capsules) and Chamelaucieae (dry indehiscent capsules). The genera Heteropyxis and Psiloxylon have been placed in the Heteropyxidaceae and Psiloxylaceae, respectively (e.g. Johnson and Briggs, Reference Johnson and Briggs1984). However, Scott (Reference Scott1980) concluded that Psiloxylon belonged to the Myrtaceae, and Tobe and Raven (Reference Tobe and Raven1987, Reference Tobe and Raven1990) found embryological evidence that Heteropyxis and Psiloxylon shared a single ancestor and suggested that the two genera should be included in the Myrtaceae, or close to it. Based on a matK phylogeny that included 66 genera from all alliances and suballiances of core Myrtaceae, Heteropyxis natalensis and Psiloxylon mauritianium, Wilson et al. (Reference Wilson, O'Brien, Heslewood and Quinn2005) concluded that there are two subfamilies of Myrtaceae: Psiloxyloideae with tribes Heteropyxideae and Psiloxyleae and Myrtoideae with 15 tribes. Three additional tribes have been distinguished for the Myrtoideae, making a total of 18 in this subfamily (Wilson et al., Reference Wilson, Heslewood and Tarran2022): Backhousieae, Chamelaucieae, Cloezieae, Eucalypteae, Kanieae, Leptospermeae, Lindsayomyrteae, Lophostemoneae, Melaleuceae, Metrosidereae, Myrteae, Osbornieae, Syncarpieae, Syzygieae, Tristanieae, Tristaniopsideae, Xanthomyrteae and Xanthostemoneae.

The Myrtaceae has 126 accepted genera and c. 6000 species (POWO; Landrum, Reference Landrum2021) and is the ninth-largest family of angiosperms (Govaerts et al., Reference Govaerts, Sobral, Ashton and Barrie2008). Fourteen tribes with 70 genera and c. 1700 species occur in Australia (Thornhill et al., Reference Thornhill, Ho, Külheim and Crips2015; Hardstaff et al., Reference Hardstaff, Sommerville, Funnekotter, Bunn and Offord2022). In contrast, species richness in the Neotropics is due mostly to the tribe Myrteae with 51 genera and c. 2500 species (Wilson et al., Reference Wilson, O'Brien, Heslewood and Quinn2005; Lucas et al., Reference Lucas, Harris, Mazine, Belsham, Nic Lughadha, Telford, Gasson and Chanse2007; Vasconcelos et al., Reference Vasconcelos, Proença, Ahmad, Aguilar, Aguilar, Amorium, Campbell, Costa, Mazine, Peguero, Prenner, Santos, Soewarto, Wingler and Lucas2017). At least 15 genera (Archirhodomyrtus, Austromyrtus, Decaspermum, Gossia, Lenwebbia, Lithomyrtus, Lophomyrtus, Myrtella, Myrtuastrum, Neomyrtus, Octamyrtus, Pilidiostigma, Rhodamnia, Rhodomyrtus and Uromyrtus) and c. 450 species of Myrteae occur outside the Neotropics, including Southeast Asia, northeastern Australia and Pacific Islands (Wilson, Reference Wilson and Kubitzki2010). Outside the Neotropics, Eugenia is found in Africa, Madagascar and Mauritius (Snow, Reference Snow2000; van der Merwe et al., Reference van der Merwe, van Wyk and Botha2005).

Various species of this family are used as ornamentals, medicines or food by humans (Hardstaff et al., Reference Hardstaff, Sommerville, Funnekotter, Bunn and Offord2022). Species in about 20 genera of Myrtaceae have been introduced into parts of the world beyond their natural range and are considered to be invasive (Mbobo et al., Reference Mbobo, Richardson, Lucas and Wilson2022). Both dry- (e.g. Eucalyptus) and fleshy- (e.g. Psidium) fruited species can be invasive, and fleshy-fruited species are more likely to be invasive on islands than dry-fruited species (Mbobo et al., Reference Mbobo, Richardson, Lucas and Wilson2022).

With the exception of Myrtus in northern Africa and the Mediterranean region of southern Europe (Wilson, Reference Wilson and Kubitzki2010), the Myrtaceae is mostly a tropical family with high species richness in South America and Australia; it also occurs in Africa, Southeast Asia, India and on various Pacific Islands (Wilson, Reference Wilson and Kubitzki2010; Thornhill et al., Reference Thornhill, Ho, Külheim and Crips2015). The family consists of trees, shrubs and a few subshrubs and strangling (viny) epiphytes (e.g. some species of Metrosideros). The monotypic shrub/small tree Osbornia octodonta is a mangrove, but it does not have pneumatophores. The leaves and stems of Myrtaceae have secretory cavities and lysigenous glands that produce ethereal oils, making plants (e.g. Eucalyptus) aromatic (Wilson, Reference Wilson and Kubitzki2010).

Flowers of Myrtaceae are actinomorphic and (0-) 4 or 5 (-7)-merous with numerous (10–270) stamens that may be in fascicles opposite the petals. Flowers of Eucalyptus and Corymbia have an operculum (bud cover) that opens at anthesis and then falls from the flower (Mabberley, Reference Mabberley2017). In some species, the stamen display attracts pollinators that collect pollen, but in other species, the thick and sweet petals are the attraction for pollinators. Flowers have a hypanthium, and the ovary may be superior, inferior or half-interior. There is one pistil, and the ovary is 1-6[-18] locular and carpellate. Placentation is axile, basal or parietal with 2–300(-500) ovules in the ovary (Wilson, Reference Wilson and Kubitzki2010; Vasconcelos et al., Reference Vasconcelos, Prenner and Lucas2019; Landrum, Reference Landrum2021).

Wilson (Reference Wilson and Kubitzki2010) described six general kinds of fruits/dispersal units for the Myrtaceae: (1) three-loculed soft fruit or berry developed from a superior ovary, Psiloxylon; (2) fruit with an inferior ovary and fleshy hypanthium, usually called a ‘berry’, Syzygieae; (3) drupe-like fruits with a thin fleshy covering over a mass of seeds with bony seed coats, Myrtella and Lithomyrtus; (4) indehiscent, leathery fruit, Osbornia; (5) dry, indehiscent fruits (‘nut-like’), Chamelaucium, Corynanthera and Thryptomene and (6) dry, dehiscent capsule, Leptospermeae and Melaleuceae. Seeds of Myrtaceae have little or no endosperm. The embryo in mature seeds is starchy or oily, fully developed (does not grow inside the seed prior to initiation of germination) and may be straight, coiled or folded (Zomlefer, Reference Zomlefer1994; Snow, Reference Snow2000; Simpson, Reference Simpson2006; Wilson, Reference Wilson and Kubitzki2010; Retamales et al., Reference Retamales, Cabello, Serra and Scharaschkin2014; Ribeiro et al., Reference Ribeiro, Nascimento, Cruz and Gurgel2021; Neto et al., Reference Neto, Santos, Lucas, Veto, Barrientos-Diaz, Staggemeier, Vasconcelos and Turchetto-Zolet2022). The seed coat may be membranous, bony or somewhat leathery, depending on the tribe/genus (Corner, Reference Corner1976; Landrum and Sharp, Reference Landrum and Sharp1989; Retamales et al., Reference Retamales, Cabello, Serra and Scharaschkin2014; Ribeiro et al., Reference Ribeiro, Nascimento, Cruz and Gurgel2021; Sbais et al., Reference Sbais, Machado, Valdemarin, Thadeo, Mazine and Mourão2022), but it does not have a palisade layer of Malpighian cells, i.e. specialized macrosclereids with a light line that are found in water-impermeable seeds (Corner, Reference Corner1976; Werker, Reference Werker1997). Seeds are 0.5–20 mm in length, depending on the species (Kirkbride et al., Reference Kirkbride, Gunn and Dallwitz2006).

Palaeohistory of Myrtaceae

Berger et al. (Reference Berger, Kriebel, Spalink and Sytsma2016) reported that the Myrtales diverged from the Geraniales c. 124 Ma with a crown age of c. 116 Ma and that the Myrtales originated in West Gondwana (i.e. South America and Africa). However, Zhang et al. (Reference Zhang, Landis, Wang, Zhu and Wang2021) concluded that the Myrtales differentiated from the Geraniales c. 111.5 Ma with a crown age of c. 104.9 Ma. Laurasia had separated from Gondwana at 116–104.9 Ma, and the southern and central parts of South America and Africa separated between 135 to 105 and 119 to 105 Ma, respectively (McLoughlin, Reference McLoughlin2001). Based on data from studies on molecular phylogeny, the divergence of the Myrtaceae has been placed at 85 (Thornhill et al., Reference Thornhill, Ho, Külheim and Crips2015; Berger et al., Reference Berger, Kriebel, Spalink and Sytsma2016) to 80 Ma (Sytsma et al., Reference Sytsma, Litt, Zjhra, Pires, Nepokroeff, Conti, Walker and Wilson2004), which was before the final separation of South America from Antarctica and the separation of Australia from Antarctica c. 30 Ma with opening of the Drake Passage and Tasman Straight, respectively (Scotese and Golonka, Reference Scotese and Golonka1992; Lawver and Gahagan, Reference Lawver and Gahagan2003; Scotese, Reference Scotese2021). However, Gonçalves et al. (Reference Gonçalves, Shimizu, Ortiz, Jansen and Simpson2020) used data for 78 plastid protein-coding genes from 125 species representing 8 myrtalean families, including 8 genera and 51 species of Vochysiaceae, and fossils from 4 myrtalean families and estimated the crown age of Myrtales as 125.5 Ma. These authors placed the divergence of Myrtaceae from its closest relative the Vochysiaceae at c. 100 Ma with a stem age of 115 Ma. Based on these dates, the Myrtaceae diverged before the beginning of the separation of West Gondwana and South America.

Hill and Scriven (Reference Hill and Scriven1995) and McLoughlin (Reference McLoughlin2001) concluded that the disturbance caused by continental rifting may have provided new environmental conditions that promoted the diversification and dispersal of angiosperms, including the Myrtaceae. Jordan et al. (Reference Jordan, Barraclough and Rosindell2016) thought that continental movements probably do not explain the increase in a number of terrestrial species but that changes in climates due to the movements of continents may have promoted increased speciation.

Thornhill et al. (Reference Thornhill, Ho, Külheim and Crips2015) reported that Myrtaceae had a Gondwanan origin and that at least 6 of the 22 sister groups of this family included in their study may be a product of vicariance. Three of the 22 sister groups had evidence of overland dispersal events, while the other 13 had undergone transoceanic long-distance dispersal. Some researchers, e.g. Sytsma et al. (Reference Sytsma, Litt, Zjhra, Pires, Nepokroeff, Conti, Walker and Wilson2004), have suggested that the origin of extant Myrtaceae was in Australasia since tribes such as Chamelaucieae, Eucalypteae, Leptospermeae, Lindsayomyrteae, Lophostemoneae, Melaleuceae and Xanthostemoneae are not found in South America or Africa. Thornhill et al. (Reference Thornhill, Ho, Külheim and Crips2015) suggested that radiation of subfamily Myrtoideae occurred in the part of Gondwana that eventually became Australia. Berger et al. (Reference Berger, Kriebel, Spalink and Sytsma2016) found a significant increase in diversification rates in Myrtaceae at c. 75 Ma, and speciation was 0.32 species Ma⁻¹ and extinction 0.15 species Ma⁻¹. These authors determined that extensive radiation of Myrtaceae occurred in Australia from the Eocene into the Miocene, as the cooling and drying of the climate increased.

The crown age of subfamily Myrtoideae is c. 75 (Biffin et al., Reference Biffin, Lucas, Craven, Costa, Harrington and Crisp2010) to 71.5 Ma (Thornhill et al., Reference Thornhill, Ho, Külheim and Crips2015), and it is c. 39.7 Ma for subfamily Psiloxyloideae (Thornhill et al., Reference Thornhill, Ho, Külheim and Crips2015). The divergence of Heteropyxis and Psiloxylon was c. 18 Ma (Berger et al., Reference Berger, Kriebel, Spalink and Sytsma2016). The crown age of the tribe Myrteae is 50.7 Ma (Thornhill et al., Reference Thornhill, Ho, Külheim and Crips2015), and its likely ancestral area is eastern Gondwana (Australia, New Caledonia, New Guinea and New Zealand) (Vasconcelos et al., Reference Vasconcelos, Proença, Ahmad, Aguilar, Aguilar, Amorium, Campbell, Costa, Mazine, Peguero, Prenner, Santos, Soewarto, Wingler and Lucas2017; Estrella et al., Reference Estrella, Buerki, Vasconcelos, Lucas and Forest2019). The divergence of Australasian and South American Myrteae was 43.9 Ma, after which much radiation occurred in both regions (Thornhill et al., Reference Thornhill, Ho, Külheim and Crips2015). Dispersal events between Australia and South America were possible in the Tertiary via Antarctica (Sytsma et al., Reference Sytsma, Litt, Zjhra, Pires, Nepokroeff, Conti, Walker and Wilson2004).

Fossilized parts of Myrtaceae plants of various ages have been found: flowers, Early Eocene, Argentina (Zamaloa et al., Reference Zamaloa, Gandolfo and Nixon2020); flowers and fruits, Eocene, Australia (Basinger et al., Reference Basinger, Christophel, Greenwood and Wilson2007); fruits and seeds, Eocene, British Columbia (Canada) and Palaeocene, North Dakota (USA) (Pigg et al., Reference Pigg, Stockey and Maxwell1993; Manchester, Reference Manchester1999); leaves, Early Miocene, Australia (Tarran et al., Reference Tarran, Wilson, Paull, Biffin and Hill2018); leaves, Middle Eocene, Argentina (Panti, Reference Panti2016); pollen, Cretaceous–Eocene, Sarawak (Malaysia) (Muller, Reference Muller1968); pollen, Palaeogene–Neogene, Australia (Thornhill and Macphail, Reference Thornhill and Macphail2012); wood, Late Cretaceous–Early Tertiary, Antarctica (Poole et al., Reference Poole, Mennega and Cantrill2003) and wood, Late Cretaceous, India (Shukla et al., Reference Shukla, Mehrotra and Tyagi2012). Fossils can be helpful in dating a phylogeny, but in the case of Myrteae differences in crown mode have resulted, depending on the kind of fossils considered. For example, Vasconcelos et al. (Reference Vasconcelos, Proença, Ahmad, Aguilar, Aguilar, Amorium, Campbell, Costa, Mazine, Peguero, Prenner, Santos, Soewarto, Wingler and Lucas2017) using macrofossils and fossil pollen of Myrteae obtained a crown node for Myrteae of 65.55 Ma (Cretaceous–Palaeocene boundary) and 40.76 Ma (mid-late Eocene), respectively.

Radiation of Myrtaceae resulted in tribes with dry (capsular) fruits and those with fleshy fruits (Thornhill et al., Reference Thornhill, Ho, Külheim and Crips2015). Sytsma et al. (Reference Sytsma, Litt, Zjhra, Pires, Nepokroeff, Conti, Walker and Wilson2004) found that fleshy, indehiscent fruits have originated at least three times in the Myrtaceae: Myrtoid group (Myrteae), Acmena group (Syzygieae) and Osbornia (Osbornieae). Compared with other lineages of Myrtaceae, tribes Syzygieae and Myrteae have had high rates of diversification, and the increased rate is associated with a shift from dry to fleshy fruits, which occurred independently in both tribes (Biffin et al., Reference Biffin, Lucas, Craven, Costa, Harrington and Crisp2010). Nine of the 13 long-distance dispersal events proposed by Thornhill et al. (Reference Thornhill, Ho, Külheim and Crips2015) involved taxa with fleshy fruits that could be dispersed by birds or bats. The presence of Myrtus in the Mediterranean Region perhaps is due to a long-distance dispersal event from East Gondwana to the Mediterranean via northern Africa during the Eocene (Thornhill et al., Reference Thornhill, Ho, Külheim and Crips2015).

Two large genera of Myrtaceae with dry (capsular) fruits are Eucalyptus (tribe Eucalypteae) and Metrosideros (tribe Metrosidereae). The Australasian eucalypt group includes Allosyncarpia, Angophora, Arillastrum, Corymbia, Eucalyptus, Eucalyptopsis and Stockwellia (Ladiges et al., Reference Ladiges, Udovicic and Nelson2003), with Eucalyptus being the largest with 712 species (POWO, 2024). Some species diversification of the Eucalyptus group is related to the cooling and drying of Australia and increased fire frequency (Ladiges et al., Reference Ladiges, Udovicic and Nelson2003). Crisp et al. (Reference Crisp, Burrows, Cook, Thornhill and Bowman2011) reported that the sclerophyllous woodlands and savannas in Australia are dominated by species of Eucalyptus, many of which can resprout after fire. Using trait mapping on a dated phylogeny of Myrtaceae, these authors found that epicormic resprouting (from buds on the stem) in Myrtaceae was correlated with the development of fire-prone Eucalyptus-dominated habitats beginning 60–62 Ma.

Metrosideros with c. 60 species has high richness in Australia, New Caledonia and New Guinea, and it occurs on various Pacific islands such as Bonin, Fiji, Hawaii, Marquesas, New Zealand and Samoa (Wilson, Reference Wilson, Keast and Miller1996; Mabberley, Reference Mabberley2017; Wright et al., Reference Wright, Liddell, Lacap-Bugler and Gillman2021). Metrosideros angustifolia is the only species of Myrtaceae with capsular fruits in Africa (Sytsma et al., Reference Sytsma, Litt, Zjhra, Pires, Nepokroeff, Conti, Walker and Wilson2004; Mabberley, Reference Mabberley2017), and M. stipularis is the only one in the New World (Sytsma et al., Reference Sytsma, Litt, Zjhra, Pires, Nepokroeff, Conti, Walker and Wilson2004). Seeds of Metrosideros are wind dispersed and can be lifted by wind speeds of 5–18 km h⁻¹ (Wright et al., Reference Wright, Yong, Dawson, Whittaker and Gardner2000). Thus, long-distance dispersal by wind may help account for the occurrence of this genus on widely separated Pacific islands.

Except for M. stipularis with dry fruits in Chile and Argentina, all Myrtaceae in the Neotropics belong to the tribe Myrteae and have fleshy fruits (Lucas et al., Reference Lucas, Belsham, Nic Lughadha, Orlovich, Sakuragui, Chase and Wilson2005, Reference Lucas, Matsumoto, Harris, Nic Lughadha, Benardini and Chase2011; Wilson et al., Reference Wilson, O'Brien, Heslewood and Quinn2005; Neto et al., Reference Neto, Santos, Lucas, Veto, Barrientos-Diaz, Staggemeier, Vasconcelos and Turchetto-Zolet2022). Species diversification in Myrteae accelerated in the Neotropics compared with that of Myrteae in the Old World (Vasconcelos et al., Reference Vasconcelos, Proença, Ahmad, Aguilar, Aguilar, Amorium, Campbell, Costa, Mazine, Peguero, Prenner, Santos, Soewarto, Wingler and Lucas2017). The development of new embryo traits [e.g. large storage cotyledons or large leaf-like folded cotyledons) (Landrum, Reference Landrum1986; Landrum and Stevenson, Reference Landrum and Stevenson1986), polyploidy (Costa et al., Reference Costa, Pinto, Morais, Oliveira and Barreto2017) and bony seed coats as in Psidium (Landrum and Stevenson, Reference Landrum and Stevenson1986)] have been suggested as new adaptive advantages associated with the increased rates of speciation of fleshy-fruited species.

Eugenia (Myrteae) with 1218 species (POWO) is the largest genus of Myrtaceae in the Neotropics (Mazine et al., Reference Mazine, Faria, Giaretta, Vasconcelos, Forest and Luca2018). After the ancestors of Eugenia migrated to southern South America, there was species diversification and dispersal to northern South America and the Caribbean region. The highest numbers of Eugenia species in South America are in the Atlantic Forest, Amazon Forest and Cerrado (Brazilian savanna) with 250, 91 and 74 species, respectively (Bünger et al., Reference Bünger, Mazine, Forest, Bueno, Stehmann and Lucas2016). For E. uniflora, there are two evolutionary lineages in the Atlantic Forest, one in the north and another in the south (Turchetto-Zolet et al., Reference Turchetto-Zolet, Salgueiro, Turchetto, Cruz, Veto, Garros, Segatto, Freitas and Margis2016). Eugenia was dispersed from South America to Southeast Asia and Africa (van der Merwe et al., Reference van der Merwe, van Wyk and Botha2005; Lucas et al., Reference Lucas, Harris, Mazine, Belsham, Nic Lughadha, Telford, Gasson and Chanse2007; Mazine et al., Reference Mazine, Faria, Giaretta, Vasconcelos, Forest and Luca2018). Two clades of Eugenia occur in southern Africa: one related to New World Eugenia and one related to Old World Eugenia (van der Merwe et al., Reference van der Merwe, van Wyk and Botha2005).

Using chloroplast and nuclear DNA sequences of the genus Myrceugenia (Myrteae), Murillo-A et al. (Reference Murillo-A, Stuessy and Ruiz2016) determined that four lineages of the genus had diverged in South America by the early Miocene: three in Chile and one in southeastern Brazil. One Chilean lineage dispersed northward, and species became part of the subtropical montane flora; part of this lineage subsequently migrated southward. The other two Chilean lineages migrated south, and species became part of the cool-temperate rainy forest flora. The lineage in southeast Brazil diversified, with species now growing in the Paraná (Araucaria angustifolia) forest, tropical semi-deciduous forests, pampas and Cerrado.

Myrcia (Myrteae) is a large genus with c. 800 species, and it is divided into nine sections (Lima et al., Reference Lima, Goldenberg, Forest, Cowan and Lucas2021). Santos et al. (Reference Santos, Lucas, Sano, Buerki, Staggemeier and Forest2017) concluded that Myrcia originated in the Montane Atlantic Forest of eastern Brazil in the late Eocene to early Miocene, after which some lineages diversified in the region. Other lineages migrated northward to the Amazon, Guyana and Caribbean regions, where diversification occurred. Also, lineages of Myrcia dispersed from the Atlantic Forest to regions with Cerrado, Yungas (subtropical cloud forest) and savanna vegetation, which was followed by diversification of new species (Amorim et al., Reference Amorim, Vasconcelos, Souza, Alves, Antonelli and Lucas2019). In fact, Myrcia section Aguava seems to have originated in the Cerrado in the mid-Miocene (Lima et al., Reference Lima, Goldenberg, Forest, Cowan and Lucas2021).

Syzygium (Syzygieae) with 1231 species (POWO) is an Old World tropical/subtropical genus of trees or rarely shrubs, many of which are cultivated for their edible fleshy fruits (Uddin et al., Reference Uddin, Hossain, Reza, Nasrin and Alam2022). Using data from molecular phylogenetic studies of Syzygium, Low et al. (Reference Low, Rajaraman, Tomlin, Ahmad, Ardi, Armstrong, Athen, Berhaman, Bone, Cheek, Cho, Choo, Cowie, Crayn, Fleck, Ford, Forster, Girmansyah, Goyder and Gray2022) determined that the genus originated in Sahul, which was a land mass consisting of Australia, New Guinea and the Aru Islands that was connected due to low seas levels during the Last Glacial Maximum, e.g. c. 23,000–19,000 years ago (Clark and Mix, Reference Clark and Mix2002). Migration of Syzygium from Sahul to the Sunda Islands (Brunei, East Timor, Indonesia, Malaysia and Singapore) has occurred at least 12 times, and each dispersal event was followed by species diversification. Dispersal and diversification have resulted in various species of Syzygium growing in the Northern Pacific, India and Africa (Low et al., Reference Low, Rajaraman, Tomlin, Ahmad, Ardi, Armstrong, Athen, Berhaman, Bone, Cheek, Cho, Choo, Cowie, Crayn, Fleck, Ford, Forster, Girmansyah, Goyder and Gray2022). These authors note that dispersal to a new region often has resulted in rapid speciation.

Background information on reproductive biology

Kinds of embryos in seeds of Myrtaceae

Martin (Reference Martin1946) listed bent, folded and linear embryos for Myrtaceae, but investing and spatulate embryos also occur in this family (e.g. Dawson, Reference Dawson1970; Landrum and Kawasaki, Reference Landrum and Kawasaki1997; Wilson, Reference Wilson and Kubitzki2010). Thus, five morphologically distinct kinds of embryos are found in Myrtaceae (Fig. 1). A linear-full embryo (i.e. a fully developed linear embryo) is long, usually curved with small, recurved cotyledons and an enlarged hypocotyl that is greatly swollen in some species (Fig. 1A). A spatulate embryo has spoon-shaped cotyledons attached directly above a straight (easily visible) hypocotyl/radicle (Fig. 1B). A bent embryo has rounded cotyledons and a hypocotyl that curves sharply around the end of the cotyledons (Fig. 1C). A folded embryo has large foliaceous cotyledons that are folded together and an obviously protruding hypocotyl that may be somewhat curved (Fig. 1D). An investing embryo has thick fleshy cotyledons that cover most, or all, of the embryo axis (Fig. 1E). The investing embryo of Myrtaceae has been described as massive but undivided, i.e. undifferentiated (McVaugh, Reference McVaugh1956). However, Justo et al. (Reference Justo, Alvarenga, Alves, Guimarães and Strassburg2007) clearly showed that the investing embryo of Eugenia pyriformis was differentiated and had an embryo axis c. 1.0 mm in length located between the cotyledons.

Figure 1. Embryo types of Myrtaceae. (A) Linear-full (myrtoid); (B) spatulate; (C) bent; (D) folded (myrcioid) and (E) investing (eugenioid). A, axis (which is covered by cotyledons); C, cotyledons; H, hypocotyl.

Investing, folded and linear-full embryos in tribe Myrteae have been called eugenioid, myrcioid and myrtoid embryos, respectively (e.g. Landrum and Stevenson, Reference Landrum and Stevenson1986; Lucas et al., Reference Lucas, Harris, Mazine, Belsham, Nic Lughadha, Telford, Gasson and Chanse2007; da Silva and Mazine, Reference da Silva and Mazine2016). However, in this review, we refer to all embryos using the embryo classification system of Martin (Reference Martin1946) as modified by Baskin and Baskin (Reference Baskin and Baskin2007).

In seeds of Myrtaceae with a bent, folded, investing or linear-full embryo, there is a shoot apical meristem between the cotyledons and a root apical meristem at the lower end of the hypocotyl. The hypocotyl is the first structure to emerge from seeds with a bent, folded, investing and linear-full embryo (see spatulate embryo below), after which the primary root and a shoot are produced by the apical meristems (Beltrati, Reference Beltrati1978; Landrum and Stevenson, Reference Landrum and Stevenson1986; Aronne and de Micco, Reference Aronne and de Micco2004; Meza and Bautista, Reference Meza and Bautista2007; Rego et al., Reference Rego, Cosmo, Gogosz, Kuniyoshi and Nogueira2011; Bardales et al., Reference Bardales, Pisco, Flores, Mashacuri, Ruíz, Correa and Gómez2014; Cosmo et al., Reference Cosmo, Gogosz, Rego, Nogueira and Kuniyoshi2017; Freitas et al., Reference Freitas, de Lucena, Bonilla, da Silva and Sampaio2018). The amount (length) that the hypocotyl extends from seeds before the primary root is visible varies, resulting in some seedlings with a relatively long hypocotyl (between the seed and root) (Aronne and de Micco, Reference Aronne and de Micco2004) and others with a short hypocotyl (Beltrati, Reference Beltrati1978; Nacata and Andrade, Reference Nacata and Andrade2020). This kind of seed dormancy was placed in Subclass 4 (Hypocotylar) of Class physiological dormancy (PD) by Baskin and Baskin (Reference Baskin and Baskin2021). The dormancy formula they suggested for seeds is $C_{xb}^{{m}^{\prime}}$, where C is class physiological dormancy (PD), × level 1 (nondeep), 2 (intermediate) or 3 (deep) of PD, subscript b that warm temperatures are required to break PD and superscript m′ that the root and shoot arise from meristematic tissue on opposite ends of the hypocotyl after the hypocotyl has emerged from the seed. For nondormant (ND) seeds, the formula is $C_{{\rm nd}}^{{m}^{\prime}}$, where subscript nd means nondormant (Baskin and Baskin, Reference Baskin and Baskin2021).

Information about the germination morphology of seeds with a spatulate embryo is not clear. Some authors (e.g. Ladiges et al., Reference Ladiges, Foord and Willis1981; Robinson et al., Reference Robinson, Boon, Sawtell, James and Cross2008; Baumann and Hewitt, Reference Baumann and Hewitt2023) have said that radicle emergence was the criterion for germination. However, the photographs of Melaleuca alternifolia seeds in various stages of germination show the first stage as having a hypocotyl that is twice as long of the radicle (Pinheiro et al., Reference Pinheiro, de Medeiros, Hilst, Pinheiro and dos Dias2020), causing us to wonder if germination morphology in seeds with a spatulate embryo is like that in the other kinds of Myrtaceae seeds, i.e. the radicle does not grow until the hypocotyl has emerged from the seed.

A literature search was conducted to increase the size of our embryo database for the Myrtaceae. In total, we have information on embryo morphology for 240 species in 123 genera and 20 tribes of Myrtaceae (Supplementary Table S1). Some tribes of Myrtaceae have only one kind of embryo, e.g. Chamelaucieae has only linear-full and Heterophyxideae, Lindsaomyrteae, Lophostemoneae, Psiloxyleae, Syncarpieae and Syzygieae have only investing embryos (Table 1). Xanthomyrteae and Xanthostemoneae have only bent embryos; Tristaniopsideae only folded embryos; and Cloezieae, Leptospermeae, Melaleuceae, Metrosidereae, Osbornieae and Tristanieae only spatulate embryos. Backhousieae (bent, linear-full), Eucalypteae (folded, investing) and Kanieae (linear-full, spatulate) have two kinds of embryos, while the Myrteae have four kinds: bent, folded, investing and linear-full.

Table 1. Number of genera with different kinds of embryos in each tribe of Myrtaceae and occurrence of nondormancy (ND) and physiological dormancy (PD) in each tribe

Note: +, yes; −, no information.

According to Martin (Reference Martin1946), bent, folded, investing, linear and spatulate embryos have a central (axile) position inside the seed. Martin's family tree of seed phylogeny shows the linear embryo as being in about the middle of the tree, and the upward progression of embryos on the tree is linear → spatulate → bent → folded → investing. That is, the investing embryo is at the top of Martin's tree. However, when compared to the crown age of the various tribes of Myrtaceae (Thornhill et al., Reference Thornhill, Ho, Külheim and Crips2015), no clear pattern of phylogenetic relationships between the various kinds of embryos in Myrtaceae is evident. For example, the crown age of Lophostemoneae and Syzygieae with an investing embryo is 41.3 and 29.3 Ma, respectively, while that of Melaleuceae and Metrosidereae with spatulate embryos is 55.5 and 24.9 Ma, respectively. The crown age of Myrteae with bent, folded, investing and linear-full embryos is 50.7. The crown age of Xanthostemoneae and Backhousieae with a bent embryo is 55.6 and that of Backhousieae with bent and linear-full embryos is 18.5 Ma.

Vochysiaceae is the closest relative of Myrtaceae (Gonçalves et al., Reference Gonçalves, Shimizu, Ortiz, Jansen and Simpson2020), and living species of Vochysiaceae have seeds with folded, investing or spatulate embryos (Niembro, Reference Niembro1983; Garwood, Reference Garwood1998; Ferreira et al., Reference Ferreira, Davide and Tonetti2001; Kirkbride et al., Reference Kirkbride, Gunn and Dallwitz2006). Thus, it is not surprising to find these three kinds of embryos in various positions on molecular phylogeny trees of Myrtaceae based on either combined plastid or combined nuclear data (Wilson et al., Reference Wilson, Heslewood and Tarran2022). The presence of two or more kinds of embryos in four tribes of Myrtaceae (Table 1) suggests that there are some possible evolutionary relationships between kinds of embryos in this family that merit research, e.g. what is the origin of the bent and linear-full embryos in the Myrtaceae? Bent and linear-full embryos occur in Backhousieae and linear-full and spatulate embryos in Kanieae; however, the Myrteae have linear-full, bent, folded and investing embryos but no spatulate embryos.

Occurrence of seed dormancy: tribes, life forms and vegetation regions

Since the five kinds of embryos in seeds of Myrtaceae are fully developed (Fig. 1) and seeds are water permeable (e.g. Pérez-Fernández et al., Reference Pérez-Fernández, Lamont, Marwick and Lamont2000; Auld and Ooi, Reference Auld and Ooi2009; Hue et al., Reference Hue, Abdullah, Sinniah and Abdullah2013), freshly matured seeds are either ND or have PD. If seeds tested over a range of temperatures and in light and in dark germinate in less than about 4 weeks and show no increase in germination percentages when given a dormancy-breaking treatment, they are ND. On the other hand, if fresh seeds either do not germinate in about 4 weeks or exhibit a widening of range of environmental conditions (e.g. temperature) over which they germinated after receiving a dormancy-breaking treatment, they have PD (Baskin and Baskin, Reference Baskin and Baskin2014).

To enhance our database for seed dormancy/germination of Myrtaceae in Baskin and Baskin (Reference Baskin and Baskin2014), a literature search was conducted using the name of each tribe of Myrtaceae, seeds, semillas, sementes, germination, germinação and germinación. In total, information on seed dormancy/germination was found for 571 species of Myrtaceae (Supplementary Table S2). Some species in all 20 tribes of Myrtaceae have ND seeds, and species in 8 tribes have seeds with PD (Table 1). For many tribes, however, the absence of PD in the tribe may be due to a lack of detailed germination studies for members of that tribe.

Each of the 571 species was recorded by life form (tree or shrub) in the vegetation region in which it grew (Supplementary Table S2). Then, the proportion of tree and shrub species in each vegetation region with ND seeds or with PD was calculated to create a seed dormancy profile for the Myrtaceae (Table 2). Overall, seeds of 55.6% of the Myrtaceae species had ND seeds, and the other 44.4% had seeds with PD. The highest number of tree species was recorded for tropical rainforest and semi-evergreen rainforest, and that for shrubs was the matorral (sclerophyllous woodlands with winter rain). The relative importance of ND and PD trees and shrubs varied with the vegetation region. In rainforests, the percentage of trees and shrubs with either ND or PD was almost equal (c. 50%). In semi-evergreen rainforests and savannas, both trees and shrubs had a higher percentage of ND than PD. In the matorral and broad-leaved evergreen forests, trees had a higher percentage of ND than shrubs, while shrubs had a higher percentage of PD than trees. In hot deserts (trees) and grasslands (shrubs), all species had ND seeds. Tropical montane trees had a higher percentage of PD (68.4) than ND (31.6), but 50% of shrubs had ND seeds and 50% seeds with PD. In dry deciduous forests, trees had 42.9 and 57.1% ND and PD, respectively, but all shrubs had seeds with PD.

Table 2. Seed dormancy profile for trees and shrubs of Myrtaceae in different vegetation regions

Not only did the rainforest and semi-evergreen rainforest have the highest number of species of trees (93 and 104, respectively), but they also had 12 and 9 tribes of Myrtaceae, respectively (Table 3). Shrubs in the rainforest and semi-evergreen rainforest were represented by 3 and 12 tribes of Myrtaceae, respectively. In the other seven vegetation regions, both trees and shrubs were represented by 0–5 tribes. Trees were represented by 0 and 5 tribes in grassland and savanna, respectively, and shrubs by 0 and 5 tribes in hot desert and broad-leaved evergreen forest, respectively.

Table 3. Tribes of Myrtaceae, vegetation regions (1–9)^a and life form (tree or shrub)

^a Vegetation regions: 1, rainforest; 2, semi-evergreen rainforest; 3, tropical montane; 4, tropical dry deciduous; 5, savanna; 6, hot desert; 7, matorral; 8, broad-leaved evergreen and 9, temperate grassland.

Dormancy-break in seeds of Myrtaceae

Since Myrtaceae is mostly a tropical family, it comes as no surprise that the breaking of PD in seeds of many members of this family occurs at warm conditions suitable for germination. Thus, in many studies, seeds have been sown in nurseries or greenhouses without receiving any dormancy-breaking treatments (Supplementary Table S2), and the number of days until seeds germinated was monitored. Methods used in laboratories to break seed dormancy and promote germination include treatment with GA₃ (Cochrane et al., Reference Cochrane, Kelly, Brown and Cunneen2002; Scalon et al., Reference Scalon, Filho and Rigoni2004; Liang et al., Reference Liang, Liu, Huang, Wei, Ye, Luo and Xiong2013; Saldías and Velozo, Reference Saldías and Velozo2014; Damiani et al., Reference Damiani, da Silva, Goelzer and Déo2016; Griebeler et al., Reference Griebeler, Araujo, Rorato, Turchetto, Tabaldi, Barbosa and Berghetti2019; Mali et al., Reference Mali, Jain, Yadav, Sharma and Singh2021; Santos et al., Reference Santos, Santos, Pinto, Dresch, Scalon and Torales2022), potassium nitrate (Liang et al., Reference Liang, Liu, Huang, Wei, Ye, Luo and Xiong2013), smoke-infused water (Cochrane et al., Reference Cochrane, Kelly, Brown and Cunneen2001), sodium nitrite and potassium cyanide (Bardales et al., Reference Bardales, Pisco, Flores, Mashacuri, Ruíz, Correa and Gómez2014). Germination also has been promoted by mechanical scarification (Gentil and Ferreira, Reference Gentil and Ferreira1999; Martinotto et al., Reference Martinotto, Paiva, Santos, Soares, Nogueira and Silva2007; Tafarel et al., Reference Tafarel, Silvestre, Pansera, Rodrigues and Sartori2021) and by removing the seed coat from the embryo (Rizzini, Reference Rizzini1970; Gentil and Ferreira, Reference Gentil and Ferreira1999). In a study of the cold hardiness of 15 species of Eucalyptus being considered for possible introduction into Ireland, 4 weeks of cold stratification increased germination (compared to fresh seeds) for only one species (Afroze et al., Reference Afroze, Douglas and Grogan2021). The positive response of seeds to treatments such as GA₃ and scarification indicates that seeds have nondeep PD.

Seed germination of Myrtus communis is promoted by soaking seeds in water, treatment with GA₃ and cold stratification (Benvenuti and Macchia, Reference Benvenuti and Macchia2001). Ballesteros et al. (Reference Ballesteros, Meloni and Bacchetta2015) recommended 3 months of cold stratification to break dormancy of M. communis seeds. After cold stratification, Benvenuti and Macchia (Reference Benvenuti and Macchia2001) obtained higher germination percentages at 25–30°C than at 10–20°C. We conclude that seeds of this species also have nondeep PD.

If GA₃ and scarification do not promote germination and if seeds require a long period of time (40–150 d) to germinate (Rizzini, Reference Rizzini1970; Smith-Ramírez et al., Reference Smith-Ramírez, Armesto and Figueroa1998; Santos et al., Reference Santos, Ferreira and Aquilla2004; Scalon et al., Reference Scalon, Filho and Rigoni2004; Masetto et al., Reference Masetto, Davide, Faria, da Silva and Rezende2009; Simpson, Reference Simpson2011; Saldías and Velozo, Reference Saldías and Velozo2014), they may have intermediate or deep PD. Unfortunately, no studies have been done to determine if the long-germinating seeds of Myrtaceae have intermediate or deep PD.

There are six types of nondeep PD, and they can be distinguished by the changes in temperature requirement for germination during the dormancy-breaking treatment (Types 1, 2 and 3) or by the temperature range over which seeds will germinate when dormancy is broken (Types 4, 5 and 6) (Baskin and Baskin, Reference Baskin and Baskin2014; Soltani et al., Reference Soltani, Baskin and Baskin2017). In the early stages of dormancy-break, seeds with Types 1, 2 and 3 dormancy germinate at low, high and intermediate temperatures, respectively. As dormancy-break continues, seeds with Types 1 and 2 dormancy exhibit an increase in the maximum temperature for germination and a decrease in the minimum temperature for germination, respectively, while seeds with Type 3 dormancy exhibit an increase in the maximum and a decrease in the minimum temperatures for germination. In the early stages of dormancy-break, seeds with Type 6 dormancy germinate over a range of low to high temperatures. During the continuation of dormancy-break, seeds with Type 6 dormancy do not exhibit an increase in the range of temperatures over which they can germinate, but germination percentages may increase. Seeds with Types 4 and 5 dormancy gain the ability to germinate only at high and low temperatures, respectively.

Seeds of Myrtaceae with PD have been tested at 20, 25 and 30°C (Hossel et al., Reference Hossel, Hossel, Júnior, Mazzaro and Fabiane2017; Paim et al., Reference Paim, Avrella, Emer, Caumo, Alver and Fior2018; Souza et al., Reference Souza, Souza and Panobianco2018; Leão-Araújo et al., Reference Leão-Araújo, Souza, Peixoto and Gomes-Júnior2019); 25, 30 and 30/20°C (Mugnol et al., Reference Mugnol, Quintáo, da Silva, Elisa and Mara2014) and 20, 25, 30, 35 and 30/20°C (Maeda et al., Reference Maeda, Bovi, Bovi and do Lago1991; Masetto et al., Reference Masetto, Davide, Faria, da Silva and Rezende2009). Among these studies, the ability of seeds to germinate at 20°C varied between the species. Seeds of Psidium cattleianum germinated to low percentages at 20°C (Hossel et al., Reference Hossel, Hossel, Júnior, Mazzaro and Fabiane2017), while those of Campomanesia adamantium (Leão-Araújo et al., Reference Leão-Araújo, Souza, Peixoto and Gomes-Júnior2019), C. guazumifolia (Souza et al., Reference Souza, Souza and Panobianco2018) Eugenia pleurantha (Masetto et al., Reference Masetto, Davide, Faria, da Silva and Rezende2009), Myrceugenia myrtoides (Paim et al., Reference Paim, Avrella, Emer, Caumo, Alver and Fior2018) and Syzygium aromaticum (Maeda et al., Reference Maeda, Bovi, Bovi and do Lago1991) germinated to high percentages. Seeds of P. guineense germinated at 25, 30 and 30/20°C, and after both 20 and 42 d of incubation, the highest percentage was at 30/20°C. After 42 d of incubation, the germination percentage had increased at 30°C (Mugnol et al., Reference Mugnol, Quintáo, da Silva, Elisa and Mara2014).

Some studies have tested seeds with PD at 5, 10 and 15°C. No seeds of Rhodomyrtus tomentosa germinated at 5 or 10°C (Liang et al., Reference Liang, Liu, Huang, Wei, Ye, Luo and Xiong2013) or at 10 or 15°C (Hue et al., Reference Hue, Abdullah, Sinniah and Abdullah2013). There was little or no germination of seeds of Acca sellowiana, Campomanesia xanthocarpa, Eugenia involucrata or E. pyriformis at 15°C (Gomes et al., Reference Gomes, Oliveira, Ferreira and Batista2016), but seeds of Psidium guineense germinated to 35% at this temperature (Santos et al., Reference Santos, Queiróz, Bispo and Dantas2015). Seeds of Darwinia species and Melaleuca species (with PD when freshly matured) germinated at 15°C (Cochrane et al., Reference Cochrane, Kelly, Brown and Cunneen2002), but the full range of temperatures for germination of these seeds was not determined.

The general conclusion from studies in which seeds of Myrtaceae were tested over a range of temperatures is that the highest germination percentages were at high temperatures. These results suggest that seeds have Type 4 nondeep PD. However, due to a lack of detailed studies on the temperature requirements for germination during the period of dormancy-break, we cannot rule out the possibility that some species have Type 6 nondeep PD (or other types of nondeep PD) with a temperature range of 20 to about 35°C for germination after PD is broken. Much more research needs to be done on the temperature requirements for germination during the dormancy-breaking period of seeds of Myrtaceae.

In Supplementary Table S2, we have recorded the temperatures (or conditions such as nursery or greenhouse) at(in) which a high percentage of the seeds of each species germinated. For 320 species listed in tropical vegetation regions, only 5 species (1.6%) have 15°C (often along with temperatures >15°C) listed as suitable for high germination; 3 of the species are in the Myrteae and one each in the Eucalypteae and Melaleuceae. For 251 species listed for temperate vegetation regions, 59 species (23.5%) have 15°C listed as a temperature for high germination. All 59 species occur in the matorral, and they belong to the Chamelaucieae, Eucalypteae or Melaleuceae, which are dry-fruited tribes. The ability of seeds of Myrtaceae to germinate at relatively low temperatures, e.g. 15°C, especially in the matorral indicates that germination can be delayed until the onset of the cool, wet season in winter. That is, dormancy-break occurs in summer and seeds germinate when the cool, wet season begins; however, no studies have been done that inform us as to what type of nondeep PD these seeds have.

Seed germination requirements

Many seed germination studies of Myrtaceae species have been conducted in nurseries, shade houses and greenhouses at near natural temperature regimes (Supplementary Table S2). The temperature at which a high germination percentage was obtained is available for 246 species (Supplementary Table S2), and the mean (±SE) of these temperatures is 22.5 ± 0.2°C. Determinations of the light (L)–dark (D) requirements for seed germination have been made for 34 species (Supplementary Table S2): 15 species, L > D; 14, L = D; 3, D > L; and 2 species with mixed results, i.e. one paper reported L = D and another D > L. Germination of Eucalyptus marginata seeds was significantly lower in white light than in darkness or in light with peak wavelengths of 430, 450, 490, 520, 570, 640 and 720 nm, as transmitted through Kodak Wratten photographic coloured filters (Rokich and Bell, Reference Rokich and Bell1995). Germination of E. calophylla seeds was significantly reduced in white light and at wavelengths of 570 and 640 nm.

Seeds of Myrtaceae differ in their ability to germinate during and after water stress, and not surprisingly the ND, desiccation-sensitive seeds of Campomanesia pubescens (Dousseau et al., Reference Dousseau, Alvarenga, Guimarães, lara, Custódio and Chaves2011) and Eugenia pyriformis (Andrade and Ferreira, Reference Andrade and Ferreira2000) are dispersed at the onset of the rainy season in the seasonally dry Cerrado of Brazil (Escobar et al., Reference Escobar, Rubio de Casas and Morellato2021). The ND seeds of Myrcia guianensis and M. splendens, which also grow in the Cerrado of Brazil, are dispersed at the onset of the rainy season and mid-rainy season, respectively. Thus, seed dispersal in these four species occurs at a time when soil moisture would be adequate for seed germination.

Seeds of Eugenia brasiliensis, E. involucrata, E. pyriformis and E. uniflora germinated to 100, 84, 66 and 91% at 0.0 MPa, respectively, but germination of each species decreased with water stress, e.g. 44, 8, 0 and 39% at −1.5 MPa, respectively; and 0, 2, 0 and 0%, respectively, at −2.0 MPa (Inocente and Barbedo, Reference Inocente and Barbedo2019). Seeds of E. umbelliflora were dispersed during the dry season in restinga vegetation in Brazil (Braz and de Mattos, Reference Braz and Mattos2010). At 0 and −0.37 MPa, seeds of E. umbelliflora germinated to 88 and 38%, respectively, and 52 and 54 d, respectively, were required for seeds to reach 50% of final germination. Further, moisture content (MC) of fresh seeds was 45–50%, and after drying at 68% relative humidity for c. 17 d, it was 28%, at which point only 30% of them germinated. At −0.1, −0.4 and −0.7 MPa, seeds of Melaleuca nematophylla germinated to 58.4, 36.3 and 0.5%, respectively (Merino-Martín et al., Reference Merino-Martín, Courtauld, Commander, Turner, Lewandrowski and Stevens2017). Seeds of Eucalyptus caesia subsp. caesia, E. ornata, and E. salubris germinated to 95–100% at −0.1 MPa; to 95, 95 and 70%, respectively, at −0.4 MPa and to 70, 35 and 8%, respectively, at −0.7 MPa (Rajapakshe et al., Reference Rajapakshe, Turner, Cross and Tomlinson2020). However, seeds of E. salmonophloia germinated to 50, 4 and 0% at −0.1, −0.4 and −0.7 MPa, respectively. Thus, seeds of the range-restricted E. caesia subsp. caesia and E. ornata were more tolerant of water stress during germination than those of the widely distributed E. salmonophloia.

Seeds of Eucalyptus macrocarpa and E. tetragona from deep sand habitats and those of E. loxophleba and E. wandoo from lateritic loam habitats germinated to c. 100% at a soil moisture potential of −0.1 MPa (Schütz et al., Reference Schütz, Milberg and Lamont2002). At −0.5 MPa, seeds of E. tetragona germinated to c. 95% and those of the other three species to 70–75%. At −1.0 MPa, however, the only germination (c. 5%) was for seeds of E. tetragona. Seeds of E. todtiana incubated on a moist substrate for 24 h at 15°C reached an MC of c. 50%. However, when seeds at 50% MC were placed on dry filter paper in an ‘air-blown cabinet’ at 23° and 55% relative humidity for 48 h MC decreased to c. 10% (Pérez-Fernández et al., Reference Pérez-Fernández, Lamont, Marwick and Lamont2000). Seeds dried for 48 h germinated to 100% and reached 50% germination in c. 9 d. Thus, seeds of E. todtiana recovered from dehydration and germinated.

Seeds of Eucalyptus brassiana, E. camaldulensis, E. grandis, E. saligna, E. tereticornis and E. urophylla germinated to 8, 39, 9, 4, 46 and 45%, respectively, at a water stress of −0.6 MPa but to only 0, 5, 0, 0, 5 and 9%, respectively, at −0.8 MPa (de Sá-Martins et al., Reference de Sá-Martins, Cleiton-José, Rocha-Faria and de Melo2019). In a NaCl solution with an osmotic potential of −1.5 MPa, seeds of E. brassiana, E. camaldulensis, E. grandis, E. saligna, E. tereticornis and E. urophylla germinated to 9, 18, 4, 4, 18 and 6%, respectively.

Some species of Myrtaceae grow in habitats that are submerged in water for part, or all, of the year. Seeds of Leptospermum lanigerum and Melaleuca squarrosa germinated under water, but flooding reduced seedling growth and survival (Zacks et al., Reference Zacks, Greet, Walsh and Raulings2018). The recalcitrant seeds of Eugenia stipitata submerged in 6 cm of water began to germinate after 2 months and after 1 year 87% had germinated (Calvi et al., Reference Calvi, Anjos, Kranner, Pritchard and Ferraz2017). Most seeds of Melaleuca ericifolia did not germinate while flooded, but even after 3–4 weeks of flooding, seeds germinated to high percentages when transferred to moist germination pads in Petri dishes (Ladiges et al., Reference Ladiges, Foord and Willis1981). Seedlings from the few seeds that germinated under water did not grow past the cotyledon stage while flooded. Fleshy fruits of Blepharocalyx cruckshanksii and Luma apiculata growing in forested wetlands of south-central Chile floated for 37 and 53 d, respectively (Mora and Smith-Ramírez, Reference Mora and Smith-Ramírez2017). After 90 d in water, seeds of both species germinated (c. 80%) inside the fruits, when fruits were removed from water and placed in moist soil. The authors concluded that the fleshy fruits promoted dispersal, but after fruits became lodged on moist soil as water receded seeds inside them could germinate readily.

There is concern that global warming will modify the environment to the extent that dormancy-break, germination and seedling survival will be negatively impacted. In a study of 100 plant species growing in Western Australia that included 37 Myrtaceae species/taxa, Eucalyptus kruseana, E. nigrifunda, E. pimpiniana, E. jimberlanica and Rhodanthe pyrethrum germinated to higher percentages when incubated at temperatures lower than those in the field during the wet season (Cochrane, Reference Cochrane2020). One species germinated to the highest percentages at field temperature during the wet season, while 31 species germinated to higher percentages at temperatures higher than those in the field during the wet season. Thus, based on this small sample of Myrtaceae species, it appears that increased temperatures due to global warming may not significantly impede the regeneration of species of Myrtaceae from seeds. For 26 species of Eucalyptus in Western Australia, modelling of seed germination response to temperature revealed that the majority of species will be able to germinate in the future, especially in the cool winter months (Cochrane, Reference Cochrane2017).

Although temperatures may be favourable for seed germination of many Myrtaceae species in the future, the question is will there be adequate precipitation for seedling survival? In other words, if precipitation decreases in the driest months will there be enough soil moisture to sustain the seedlings? Some modelling has been done to look at future precipitation patterns, but more is needed. Barrientos-Díaz et al. (Reference Barrientos-Díaz, Báez-Lizarazo, Enderle, Segatto, Reginato and Turchetto-Zolet2024) predicted that future temperature and rainfall conditions will be favourable for species of Myrteae to live in the Atlantic Forest of Brazil. However, predictions for the future distribution of Eugenia uniflora in South America suggest that Argentina and Paraguay will not have suitable habitat for this species, but populations of it may increase on the Brazilian Plateau (Turchetto-Zolet et al., Reference Turchetto-Zolet, Salgueiro, Turchetto, Cruz, Veto, Garros, Segatto, Freitas and Margis2016).

Tribes and species of Myrtaceae with desiccation-sensitive seeds

Information was found for desiccation sensitivity of 58 species of Myrtaceae, and all of them belonged either to the Myrteae (33 species) or Syzygieae (25 species) (Table 4). According to Wilson (Reference Wilson and Kubitzki2010), all the genera of Myrteae and Syzygieae represented in Table 4 have fleshy fruits. Seven species of Myrteae have intermediate seed storage behaviour, and the other 26 have recalcitrant seeds. The species of Myrteae with intermediate storage behaviour have seeds with linear-full embryos, and those that are desiccation-sensitive have seeds with investing, folded or linear-full embryos. The 25 species of Syzygieae have desiccation-sensitive seeds, and all have investing embryos.

Table 4. Species of Myrtaceae with desiccation-sensitive (recalcitrant [R] or intermediate [I]) seed storage behaviour

Among five Brazilian species of Eugenia with desiccation-sensitive seeds, the speed of seed drying and the speed of germination help explain differences in geographical range (Rodrigues et al., Reference Rodrigues, Silva, Ribeiro, Loaiza-Loaiza, Alcantara, Komatsu, Barbedo and Steiner2022). For seeds of E. uniflora, E. involucrata, E. pyriformis, E. brasiliensis and E. astringens, the water content threshold that decreased germination to c. 50% was 0.44, 0.33, 0.33, 0.25 and 0.25 g H₂O (g DW)⁻¹, respectively, and the species occurred in four, three, two, one and one morphoclimatic domain(s) (based on temperature and precipitation data) in Brazil, respectively. However, under laboratory conditions, seeds of E. uniflora had the second highest rate (speed) of germination and the slowest rate of water loss compared with the other species. The authors concluded that rapid germination and slow seed drying help explain why E. uniflora has a wider geographical distribution than the other four species.

Plantlet production from seed fragments (totipotency)

One consequence of seed predation is that the predator may not consume the whole seed, especially in the case of large seeds (Vallejo-Marín et al., Reference Vallejo-Marín, Domínguez and Dirzo2006; Pérez et al., Reference Pérez, Shiels, Zaleski and Drake2008; Loayza et al., Reference Loayza, Gachon, García-Guzmán and Carvajal2015). In some species, if fragments of seeds have an intact embryonic axis, there is a possibility that a plantlet will be produced. For example, the large recalcitrant seeds of Myrcianthes coquimbensis, a threatened Myrtaceous shrub in the Atacama Desert of Chile, will produce a plantlet if up to 75% of seed mass is removed from either mature or immature seeds (Loayza et al., Reference Loayza, Gachon, García-Guzmán and Carvajal2015).

Normal plantlet development occurred when seeds of Eugenia stipitata subsp. sororia (Anjos and Ferraz, Reference Anjos and Ferraz1999; Calvi et al., Reference Calvi, Anjos, Kranner, Pritchard and Ferraz2017), E. brasiliensis, E. involucrata and E. uniflora (Silva et al., Reference Silva, Bilia and Barbedo2005) were cut into two parts. Seeds of E. pyriformis cut in half longitudinally or transversally and those cut transversally into two parts, i.e. one-fourth of the seed and three-fourths of the seed, produced normal plantlets (Silva et al., Reference Silva, Bilia, Maluf and Barbedo2003). Normal plantlet development occurred for seeds of E. cerasiflora, E. pruinosa and E. umbelliflora cut in half transversally or longitudinally and from three-fourths of a seed (Delgado et al., Reference Delgado, Melo and Barbedo2010). When seeds were cut into four equal parts (in a linear fashion), normal plantlets developed from one-fourth (external/end part) of a seed for E. cerasiflora and E. pruinosa but not for E. umbelliflora. Also, normal plantlet development occurred from one-fourth of a seed (internal/central part) for E. cerasiflora and E. umbelliflora but not for E. pruinosa.

In some species, the percentage of plantlet development/germination was higher for pieces of seeds than for intact seeds, e.g. E. cerasiflora and E. pruinosa (Delgado et al., Reference Delgado, Melo and Barbedo2010), but often the percentage for seed fragments and intact seeds did not differ significantly, e.g. E. umbelliflora (Delgado et al., Reference Delgado, Melo and Barbedo2010) and E. uniflora (Silva et al., Reference Silva, Bilia and Barbedo2005). However, plantlet formation from E. pyriformis seeds cut into 0, 2 and 4 pieces was 97, 73 and 62%, respectively (Costa et al., Reference Costa, Pinto, Morais, Oliveira and Barreto2017). The production of normal plantlets from seeds of E. uniflora cut into two parts was higher for seeds produced from cross- than from self-pollinated flowers (Fidalgo et al., Reference Fidalgo, Cécel, Mazzi and Barbedo2019).

The regeneration of roots and plantlets from fragments of Eugenia seeds occurs in seeds taken from both immature and mature fruits (Teixeira and Barbedo, Reference Teixeira and Barbedo2012; Amador and Barbedo, Reference Amador and Barbedo2015; Delgado and Barbedo, Reference Delgado and Barbedo2020). However, there is a reduction in the success of plantlet formation from E. pyriformis seed fragments with an overall decrease in seed size, which is related to a number of seeds in fruit (Prataviera et al., Reference Prataviera, Lamarca, Teixeira and Barbedo2015). Although seed fragments of E. pyriformis (Amador and Barbedo, Reference Amador and Barbedo2011), E. brasiliensis, E. uniflora (Amador and Barbedo, Reference Amador and Barbedo2015) and E. stipitata (Calvi et al., Reference Calvi, Anjos, Kranner, Pritchard and Ferraz2017) can form a plantlet, only one plantlet is produced if the incision does not completely separate the seed parts. That is, if there is a connection between two seed parts only one of them produces roots and a plantlet.

Normal plantlet development occurred when seeds of Syzygium myrtifolium were cut into two parts longitudinally, but when seeds were cut transversally into two parts only one part produced a plantlet (Tsan, Reference Tsan2023). When a seed was cut into four parts longitudinally, at least one of the parts produced a plantlet, with the central and right fractions being the most likely to do so. However, when cut into two longitudinal and two transverse (i.e. four) parts, only the top left and top right parts produced a plantlet.

Effects of fire (heat and smoke) on seed germination

The appearance of Myrtaceae seedlings in the field after a fire (e.g. Mount, Reference Mount1969; Williams, Reference Williams2000; Wright, Reference Wright2018; Wright et al., Reference Wright, Albrecht, Silcock, Hunter and Fensham2019) has prompted people to conduct experiments on the effects of heat and smoke on seed germination. Heat treatments on seeds of various species have revealed that temperatures simulating those in/at the soil surface during a fire can increase, decrease or have no effect on germination percentages (Table 5). Smoke, in general, either increases or decreases germination percentages of Myrtaceae seeds (Table 6). Seeds of Baeckea utilis did not respond to heat in the absence of smoke, but they responded to smoke in the absence of heat (Thomas et al., Reference Thomas, Morris and Auld2007). For seeds of Kunzea ambigua and K. capitata, there was an interaction between incubation temperature (15 and 25°C), water stress (0 and −0.9 MPa) and fire cues (heat and smoke) (Thomas et al., Reference Thomas, Morris, Auld and Haigh2010). Fire cues increased germination percentages at 15 and 25°C across the range of water stress, i.e. fire cues increased seed tolerance to water stress.

Table 5. Effect of dry heat treatments on seed germination of species of Myrtaceae

All studies were conducted in Australia unless otherwise noted.

^a Study conducted in New Zealand.

^b Increase in number of seedlings in soil seed bank samples after soil heating.

Table 6. Effect of smoke and/or smoke extracts on seed germination of species of Myrtaceae

^a Seeds tested at 26/30 and 33/18°C, respectively.

Another aspect of fire in a plant community is that the smoke contains various compounds, including cyanohydrin (glyceronitrile, which in the presence of water releases cyanide), ethylene, karrikins (especially karrikin-1), nitrate and nitric oxide, that are known to promote seed germination (Flematti et al., Reference Flematti, Merritt, Piggott, Trengove, Smith, Dixon and Ghisalberti2011, Reference Flematti, Waters, Scaffidi, Merritt, Ghisalberti, Dixon and Smith2013; Soós et al., Reference Soós, Badics, Incze and Balázs2019; Cao et al., Reference Cao, Schöttner, Halitschke, Li, Baldwin, Rocha and Baldwin2021, Reference Cao, Baskin, Baskin and Li2023; Kępczyański and Kępczyańska, Reference Kępczyński and Kępczyńska2023). However, only karrikins and cyanohydrins can persist in the upper layers of the soil after a fire (Flematti et al., Reference Flematti, Waters, Scaffidi, Merritt, Ghisalberti, Dixon and Smith2013). Smoke also contains compounds that are structurally similar to karrikins, i.e. contain a butanolide ring that inhibits germination (Light et al., Reference Light, Burger, Staerk, Kohout and Van Staden2010; Burger et al., Reference Burger, Pošta, Light, Kulkarni, Viviers and Van Staden2018). Soós et al. (Reference Soós, Badics, Incze and Balázs2019) suggested that after a fire both germination inhibitors and promotors are in the surface layers of soil. However, the promotors cannot be effective in stimulating germination until rain water has removed the inhibitors.

If a fire occurs while seedlings/juveniles are relatively small, they may not be robust/resilient enough to tolerate fire and are killed (e.g. Fordyce et al., Reference Fordyce, Eamus, Duff and Williams1997; Tozer and Bradstock, Reference Tozer and Bradstock1997; Wardell-Johnson, Reference Wardell-Johnson2000; Fujita, Reference Fujita2021; Plumanns-Pouton et al., Reference Plumanns-Pouton, Swan, Penman, Collins and Kelly2023). Thus, regeneration from seeds is not successful if the fire interval is more frequent than the time required for young plants to reach a fire-tolerant size.

Various species of Myrtaceae produce new stems from buds if the aerial portion of the plant is damaged/destroyed, as for example by fire. Epicormic buds (dormant buds located under the bark), arise from meristematic cells (epicormic strands) that are inside the bark, often at the junction of the bark and vascular cambium, e.g. Eucalyptus and Melaleuca (Burrows, Reference Burrows2002; Clarke et al., Reference Clarke, Lawes, Midgley, Lamont, Ojeda, Burrows, Enright and Knox2013). Lignotubers which form at the base of the stem, e.g. in many species of Eucalyptus, have many buds that can grow if the stem is killed. The development of a lignotuber begins shortly after seedling emergence, and it increases in size as the plant grows (Fordyce et al., Reference Fordyce, Eamus and Duff2000; Nicolle, Reference Nicolle2006). Lignotubers were initiated on seedlings of Eucalyptus cinerea by the time plants were 6 weeks old (Graham et al., Reference Graham, Wallwork and Sedgley1998). After 9 months of growth, the size of the lignotubers on Eucalyptus obliqua juveniles derived from seeds from 13 provenances in Australia was inversely related to mean annual precipitation in the original habitat (Walters et al., Reference Walters, Bell and Read2005). Interestingly, a molecular marker for lignotuber formation (Elig) has been identified in Eucalyptus (Bortoloto et al., Reference Bortoloto, Fuchs-Ferraz, Kettener, Rubia, González, de Souza, Oda, Rossini and Marino2020).

Buds that can replace damaged aerial stems also occur on rhizomes. For example, Eugenia dysenterica and E. pumicifolia, which grow in the fire-prone Cerrado of Brazil, have a woody rhizome covered with periderm. Silva et al. (Reference Silva, Ferraro, Ogando, Aguiar and Appezzato-Da-Gloria2020) found numerous buds (254–517 per plant) on the upper surface of the rhizome of each species.

Global warming is having significant effects on fire regimes, especially in fire-prone habitats (Ooi et al., Reference Ooi, Tangney, Auld, Baskin and Baskin2022). There are increases in the severity and frequency of fires, as well as changes in the time of year when fires occur (Ooi et al., Reference Ooi, Tangney, Auld, Baskin and Baskin2022). Even if species of Myrtaceae resprout after fire, increased fire intensity and frequency can cause shifts from wet- to dry-plant communities (Furland et al., Reference Furlaud, Prior, Williamson GJ and Bowman2021; Fensham et al., Reference Fensham, Laffineur and Browning2024). Further, the pathogenic fungus Austropuccinia psidii may retard the regrowth of new stems and leaves following a fire. The impact of A. psidii on regrowth of tissues on nine Myrtaceae species following fire in a coastal heathland in New South Wales (Australia) varied between species and ranged between minor leaf damage to die-back and eventual death of the tree (Pegg et al., Reference Pegg, Entwistle, Giblin and Carnegie2020).

For some species, e.g. Eucalyptus pauciflora in the subalpine zone in Victoria, Australia, increased fire frequency has raised serious concerns about the ability of the species to persist in its natural habitat (Coates, Reference Coates2015). Increased temperatures, decreased precipitation and increased fire frequency are expected to negatively impact the tropical montane species Melaleuca uxorum and could result in its extinction as well as the loss of the local specialized flora in the habitat of this species being replaced by widely distributed species (Ford and Hardesty, Reference Ford and Hardesty2012).

Formation of soil seed banks

In 170 soil seed bank studies, in which soil samples were collected after the seed germination season but before newly matured seeds were dispersed, i.e. samples potentially had at least a short-lived persistent soil seed bank, we found nine papers that contained information for species of Myrtaceae. Seven species in six tribes of Myrtaceae were listed in these nine papers (Table 7). The number of seeds per species in the seed bank ranged from 1 to 32 m⁻². However, for Eucalyptus grandis, the number of seeds was not given (Gonçalves et al., Reference Gonçalves, Martins, Marins and Felfili2008), and, for Tristaniopsis sp., none of the seeds found in the soil samples germinated (Graham and Page, Reference Graham and Page2018). However, some soil seed bank studies conducted in habitats where species of Myrtaceae were growing did not contain any seeds of Myrtaceae (e.g. Vlahos and Bell, Reference Vlahos and Bell1986; Yates et al., Reference Yates, Taplin, Hobbs and Bell1995; Sem and Enright, Reference Sem and Enright1996; Wang, Reference Wang1997; Hamilton-Brown et al., Reference Hamilton-Brown, Boon, Raulings, Morris and Robinson2009; Bechara et al., Reference Bechara, Salvador, Ventura, Topanotti, Gerber, Cruz and Simonelli2020; Neto et al., Reference Neto, Martins and Silva2021; Kraaij et al., Reference Kraaij, Geerts and Malan2024).

Table 7. Soil seed bank of Myrtaceae

To help determine how long seeds of Myrtaceae live in the soil, seeds of various species have been placed in mesh bags and buried in the field. At intervals over a 1- to 3-year period, some of the buried seeds were exhumed, and the number of viable seeds was determined (Table 8). In general, seeds do not live for long periods of time in the soil. Of the 11 species listed in Table 8, only Kunzea ambigua and K. capitata had >50% viable seeds after 2 years (Auld et al., Reference Auld, Keith and Bradstock2000).

Table 8. Longevity of seeds of Myrtaceae placed in mesh bags and buried in soil in the field

^a All buried seeds had germinated after 1 month; thus, no nongerminated viable seeds were present at the end of the study.

^b No information.

Yates et al. (Reference Yates, Taplin, Hobbs and Bell1995) buried seeds of Eucalyptus salmonophloia in soil at a depth of 1 cm in 3-cm-diameter areas in Western Australia in summer, autumn, winter and spring. Then, 2, 6 and 12 months after burial in each season, soil cores (from inside the 3-cm-diameter areas) were removed and germination of seeds was monitored. Some seeds buried in summer survived and germinated after 12 months of burial, but few or no seeds buried in autumn, winter or spring survived and germinated after 12 months of burial.

Seeds of 10 species of Myrtaceae were sown in moist soil in a shade house in Western Australia, and seed viability was determined initially and after 1 year: Agonis linearifolia, 80 (initially) and 5% (after 1 year) viable seeds; Astartea fascicularis, 47 and 14%; Baeckea camphorosmae, 52 and 30%; Calytrix breviseta var. breviseta, 63 and 18%; C. depressa, 54 and 20%; Verticordia aurea, 60 and 40%; V. chrysantha, 63 and 38%; V. densiflora, 58 and 23%; V. eriocephala, 20 and 11% and V. huegelii, 26 and 3% (Roche et al., Reference Roche, Dixon and Pate1997a).

Seeds of Baeckea gunniana and B. utilis were buried in the Ginini Flats subalpine bog complex of the Brindabella Mountains (Australian Alps) in southeastern Australia (Guja and Brindley, Reference Guja and Brindley2017). After 27 months of burial, the exhumed seeds of both species germinated to c. 97%. The germination of B. utilis seeds at 25/15 and 20/10°C exhumed after 3, 6, 9, 12, 21 and 27 months revealed that dormancy cycling was occurring with seeds exhumed after 3, 12 and 27 months having the highest germination percentages and those exhumed after 9 and 21 months the lowest percentages.

The application of smoke to field sites has been shown to promote the germination of seeds in the soil. For example, smoke fumigation in Banksia woodland in Western Australia significantly increased seed germination of 15 plant taxa, but none of them were Myrtaceae (Dixon et al., Reference Dixon, Roche and Pate1995). Aerosol smoke and smoke-infused water applied to soil seed bank samples collected in Western Australia promoted seed germination of native grasses, sedges, herbs and woody species, including a few seeds of Eucalyptus sp., as well as seeds of weedy herbs in families other than Myrtaceae (Cochrane et al., Reference Cochrane, Monks and Lally2007). In Queensland (Australia), soil seed bank samples from nonburned forest/woodland/shrubland habitats at four sites were subjected to heat and/or smoke treatments (Page, Reference Page2009). Following a fire in the four sites, additional soil samples were collected and germination of seeds in them was compared with that of seeds in the treated, nonburned samples. The number of seedlings (m⁻²) in the heat and smoke-treated samples was higher than that in the control samples with no treatments. However, the number of species and the number of seedlings that emerged from the four sites after the fire were higher than in the nonburned control, except in the burned mixed Eucalyptus forest with lower numbers than in the control.

Smoke treatments have been applied in the field in relation to using the soil seed bank from Eucalyptus/Banksia woodlands in Western Australia as a source of seeds for rehabilitation of surface-mined sites (Roche et al., Reference Roche, Koch and Dixon1997b). Following smoke treatments of soil in the field, the number of species and the number of seedlings increased significantly, but often the density of Myrtaceae seedlings was low. For example, the density of Eucalyptus marginata seedlings increased from 0 to 1.67 m⁻² after smoke treatments of soil in Western Australia (Roche et al., Reference Roche, Koch and Dixon1997b). Soil samples collected in a plant community dominated by Eucalyptus cneorifolia in South Australia were subjected to heat (80° for 60 min), smoke (from burning barley hay) and heat + smoke treatments in a greenhouse. Compared with the control, all treatments increased the germination of Thryptomene ericaea; and heat + smoke increased the germination of Baeckea crassifolia, Calytrix glaberrina and E. cneorifolia (Rawson et al., Reference Rawson, Davies, Whalen and MacKay2013).

Formation of aerial seed banks

Various species of Myrtaceae growing in fire-prone habits, e.g. in southwestern Australia (Lamont et al., Reference Lamont, Le Maitre, Cowling and Enright1991) retain seeds on the mother plant for extended periods of time (Table 9). Prolonged storage of viable seeds in the canopy is called serotiny, and it is more likely to be found in fire-killed, nonsprouting species than in species capable of resprouting after fire (Lamont et al., Reference Lamont, Le Maitre, Cowling and Enright1991, Reference Lamont, Pausas, He, Witkowski and Hanley2020). The seed-holding structures in serotinous Myrtaceae are woody capsules (Wellington and Noble, Reference Wellington and Noble1985) or infructescences of capsules (Whelan and Brown, Reference Whelan and Brown1998; Kim et al., Reference Kim, Walck, Hidayati, Merritt and Dixon2009). These capsules can provide some protection of seeds from the heat of fires (Judd and Ashton, Reference Judd and Ashton1991; Judd, Reference Judd1994; Whelan and Brown, Reference Whelan and Brown1998; Battersby et al., Reference Battersby, Wilmshurst, Curran and Perry2017b).

Table 9. Aerial seed bank (serotiny) in species of Myrtaceae

M-S, moderately to strongly serotinous; W, weakly serotinous; W-M, weakly to moderately serotinous; W-S, weakly to strongly serotinous; –, no information; mo, months; yr, years.

Aerial seed banks have advantages for the species, including protection of seeds from granivores on the soil surface such as ants, and the continuous supply of viable seeds although few or no seeds are produced in some years (Lamont and Enright, Reference Lamont and Enright2000). When the seeds are dispersed, they can germinate immediately because they are ND (e.g. Kim et al., Reference Kim, Walck, Hidayati, Merritt and Dixon2009). However, habitat substrate moisture and temperatures must be favourable for germination at the time of seed dispersal because seeds of serotinous species quickly lose viability on/in the soil (Cowling and Lamont, Reference Cowling and Lamont1987; Enright and Lamont, Reference Enright and Lamont1989), or they may be eaten by predators (Wellington and Noble, Reference Wellington and Noble1985).

In some serotinous species, capsules slowly open throughout the year, resulting in a low rate of seed dispersal, e.g. Eucalyptus luehmanniana (Tozer and Bradstock, Reference Tozer and Bradstock1997) and Melaleuca quinquenervia (Baumann and Hewitt, Reference Baumann and Hewitt2023). In fire-prone habitats in New Zealand, capsules of Leptospermum scoparium remain attached to the parent plant and do not open for 1 year or longer, but in other kinds of habitats capsules split and release the seeds within 1 year (Harris, Reference Harris2002). Mature fruits remain alive on plants of Callistemon rigida for 3–4 years or longer, but they die and open when their water supply is stopped (Ewart, Reference Ewart1907). Fruits of Melaleuca parvistaminea dried and opened when plants were cut (Jacobs et al., Reference Jacobs, Richardson and Wilson2014). In M. ericifolia, a wetland species, the peak of annual seed dispersal is in April, at which time water levels in the habitat are low (Hamilton-Brown et al., Reference Hamilton-Brown, Boon, Raulings, Morris and Robinson2009). Fire is an important and reliable seed-releasing factor because it promotes massive capsule opening and seeds are released into sites where fire has removed the standing vegetation (e.g. dos Santos et al., Reference dos Santos, Matias, Deus, Águas and Silva2015; Hewitt et al., Reference Hewitt, Holford, Renshaw, Stone and Morris2015), i.e. fire prepares a good seed bed (Lamont and Enright, Reference Lamont and Enright2000). However, if a fire occurs after seed dispersal, it kills seeds on the soil surface (dos Santos et al., Reference dos Santos, Matias, Deus, Águas and Silva2015).

Future challenges to maintain species richness of Myrtaceae

Concluding thoughts

In our comparison of the highly speciose, widely distributed plant families, the Asteraceae, Rubiaceae and Myrtaceae, which are the first, fourth and ninth most speciose angiosperm families, respectively (Mabberley, Reference Mabberley2017), have been considered. The Asteraceae has trees, shrubs, lianas and herbs, with number of life forms decreasing with distance from the Equator, resulting in only herbs in the tundra (Baskin and Baskin, Reference Baskin and Baskin2023). Only one kind of embryo (spatulate) is found in cypselae (seeds) of Asteraceae, and seeds may be ND or have PD. The six known types of nondeep PD occur in the Asteraceae, and, depending on habitat/vegetation region, PD is broken by warm summer or cold moist winter conditions. Thus, in Asteraceae, the great species richness is related to seed dormancy-breaking and germination requirements that closely coincide with a wide range of habitats throughout the world, except Antarctica.

The Rubiaceae has trees, shrubs, lianas/climbers and herbs. The highest species richness is in moist tropical forests, but some shrubs and herbs grow in temperate regions and a few herbs in the tundra. The Rubiaceae has five kinds of embryos, and seeds are ND or have morphological, physiological or morphophysiological dormancy. The greatest species richness in Rubiaceae is related to the diversity of seed dormancy, especially among tropical rainforest trees and semi-evergreen rainforest shrubs (Baskin and Baskin, Reference Baskin and Baskin2024).

The Myrtaceae has only trees, shrubs and a few viny epiphytes but no herbs. The distribution of the family outside the tropics is in regions with a Mediterranean climate, e.g. Australia, South Africa and southern Europe/North Africa, and to a limited extent in temperate vegetation regions such as broad-leaved evergreen forests and grasslands. Five kinds of fully developed embryos are found in seeds of Myrtaceae; however, seeds are either ND or have PD, regardless of tribe, habitat/vegetation region or kind of fruit produced. Great species richness is found in fleshy-fruited Myrtaceae that grow in moist tropical regions. Seeds of fleshy-fruited species are either ND or have PD that is broken during exposure to relatively high temperatures, after which seeds germinate at high temperatures. The only known exception is for seeds of fleshy-fruited Myrtus communis that become ND during cold stratification. However, after cold stratification, seeds germinated to high percentages at 25–30°C than at 10–20°C (Benvenuti and Macchia, Reference Benvenuti and Macchia2001). Also, seeds of Baeckea utilis buried in the subalpine of the Brindabella Mountains (Australian Alps) in southeastern Australia germinated to higher percentages when exhumed in spring (3, 12 and 27 months of burial) than those exhumed in winter or autumn (6, 9 and 21 months of burial) (Guja and Brindley, Reference Guja and Brindley2017). These results suggest that cold stratification during winter was breaking seed dormancy.

Great species richness is also found in dry-fruited Myrtaceae species that grow in seasonally dry tropical vegetation regions and in habitats with a Mediterranean climate, e.g. the matorral with hot, dry summer and cool, moist winters. For dry-fruited species, dormancy-break during the hot, dry season is followed by germination when the wet season begins. In tropical vegetation regions, temperatures are high when the wet season begins; thus, both dormancy-break and germination occur at high temperatures. In the matorral, seeds germinate over a range of low to high temperatures that include the temperatures (e.g. c. 15°C) of the cool, rainy season. Thus, in the Myrtaceae, we find many fleshy-fruited species in which both seed dormancy-break and germination occur during exposure to warm, wet conditions (e.g. rainforest) and many dry-fruited species in which dormancy-break at warm, dry conditions are followed by germination at either warm, wet (e.g. savannas) or cool-to-warm, wet (e.g. matorral) conditions.