Abstract

The Solanaceae contains many species of agricultural importance. Several of these are cultivated for their fruits, such as the tomato, the pepper and the aubergine. The family is very diverse in fruit type with capsules, drupes, pyrenes, berries, and several sorts of dehiscent non‐capsular fruits occurring in the 90+ genera. In this paper, recent work on fruit type evolution in angiosperms is reviewed in relation to dispersal agents and habitat ecology. Defining fruit types in the Solanaceae in a simple five state system, then mapping them onto a previously published molecular phylogeny based on chloroplast DNA allows discussion of the evolution of these fruit types in a phylogenetic framework. Capsules are plesiomorphic in the family, and although berries are a synapomorphy (shared derived character) for a large clade including the genus Solanum (tomatoes and aubergines), they have arisen several times in the family as a whole. Problems with homology of drupes and pyrenes are discussed, and areas for future investigation of fruit structure homology identified. The distribution of fruit types in the large and diverse genus Solanum is also discussed in the light of monophyletic groups identified using chloroplast gene sequences. This variety is related to recent advances in the understanding of the molecular biology of fruit development. Finally, several key areas of future comparative, phylogenetic investigation into fruit type evolution in the family are highlighted.

Received 26 March 2002; Accepted 2 July 2002

Introduction

Today’s terrestrial habitats are dominated by the angiosperms – flowering plants, one of whose key features is the possession of fruits of a wide range of forms and types. Angiosperms today exhibit an enormous range of fruit types, from dry single‐seeded achenes like sunflowers (Asteraceae), to huge fleshy fruits like those of Annona (the cherimoya; Annonaceae) or Artocarpus (the breadfruit; Moraceae). Complex classifications of this huge variety of fruit types have been suggested (Spujt, 1994), with almost every variation being given a special name. However, mere description of fruits, though important for the accurate recording of variation in nature, is perhaps of less interest than an understanding of the evolution of such a variety, particularly in the light of an increased understanding of the genetic and developmental control of various floral features of angiosperms (Schwarz‐Sommer et al., 1990; Weigel and Meyerowitz, 1994; Reichmann and Meyerowitz, 1997; Theissen et al., 2000; Becker et al., 2000; Ng and Yanofsky, 2001). The evolution of such a wide array of fruit types and the degree to which the pattern of fruit types seems highly homoplasious has long intrigued both systematists and evolutionary biologists.

The diversification of the angiosperms in the Cretaceous and Tertiary was accompanied by explosive diversification in a variety of features (Friis et al., 1987). Rapid diversification in co‐evolved pollination systems in the Cretaceous occurred concomitantly with a rapid diversification of the insects (Friis and Crepet, 1987; Wing and Boucher, 1998). Other reproductive features of angiosperms, such as fruit and seed diversity, however, remained relatively unchanged through the Cretaceous (Eriksson et al., 2000). Early angiosperm fruits were apocarpous, with several free carpels. Syncarpous fruits, with fused carpels, only appear in the middle Cretaceous (Albian‐Cenomanian), some 97 million years ago. Fossil fruits from the early to middle Cretaceous were all dry and show no obvious modifications for dispersal (Friis and Crepet, 1987). Drupes and berries, the classic fleshy fruits, first appeared in the late Cretaceous or early Tertiary (Friis and Crepet, 1987). Tiffney (1984) suggested that the relatively rapid evolution of a wide variety of fruit and seed types and sizes in the Paleocene (65–56 Ma) and Eocene (56–35 Ma) was correlated with the rise of mammals and birds, primary dispersers of today’s angiosperms. Fruits are the mechanism by which seeds are dispersed, and so both fruit type (mainly fleshy versus dry) and seed size have been discussed in the context of the diversification of angiosperm fruits (Eriksson et al., 2000). Seed sizes were relatively small through the Cretaceous, with a drastic increase at approximately the Cretaceous‐Tertiary boundary (about 65 Ma), both in median seed size and in the range of seed sizes (Eriksson et al., 2000). This increase has been interpreted as a response to the increasing dominance of vertebrates, mostly mammals, and thus to coevolution with seed dispersers (Tiffney, 1984). The proportion of fleshy, and presumably animal dispersed fruits increased through the Cretaceous, and the proportion of wind dispersed (dry) fruits fluctuated, showing several peaks (see Fig. 2B in Eriksson et al., 2000). The combination of these two results led Eriksson et al. (2000) to reject the hypothesis that fruit and seed size were driven by co‐evolution with the dispersers, but instead to suggest that the trends in overall fruit type and seed size were driven by climate and vegetation change around the Cretaceous–Tertiary boundary. Plants with larger seeds tend to be correlated with wet forest habitats (Westoby et al., 1996) and the warm and increasingly wet climates of the Eocene supported the development of extensive closed canopy forests. The changing recruitment environment for seedlings, with the dry climates of the Cretaceous favouring small seeded plants, and the wetter and more closed canopy forests of the Eocene favouring larger seeded plants, is seen by Eriksson et al. (2000) as the primary factor driving the evolution of seed size and fruit in angiosperms as a whole. Plants with large seeds that were favoured in closed canopy environments would have a reduced dispersal capacity unless secondary animal dispersal evolved (Eriksson et al., 2000). The evolution of specialized frugivores such as bats, birds and rodents during the Tertiary was a secondary phenomenon, related to the abundance of a large‐seeded resource, rather than the primary cause of the increase in fruit and seed size among angiosperms (Eriksson et al., 2000). Variation in the characters related to dispersal effectiveness are perhaps too limited to strongly affect the intensity of selection (Schemske, 1983) and thus are not the primary driving force in the evolution of broad patterns of fruit and seed size in angiosperms. It has been suggested that the interaction between plants and their animal dispersers is manifested at a level higher than the species (Howe, 1984; Herrera, 1985), so patterns of coevolution might be expected to occur at the family level in flowering plants.

The use of fruit types as important taxonomic characters has, to a certain extent, confounded the rigorous study of patterns of fruit type evolution. Fruit types are sometimes considered to be evolutionarily conservative (Spujt, 1994) and are thus given a high taxonomic weight. Recent studies, however, using phylogenetic methods, have shown that, in many families of angiosperms, fruit type is highly homoplasious (i.e Apiaceae, Plunkett et al., 1996; Spalik et al., 2001; Cunoniaceae, Bradford and Barnes, 2001; Lamiales, Wagstaff and Olmstead, 1997; Myrtaceae, Johnson and Briggs, 1994; Ranunculaceae, Hoot, 1995; Rosaceae, Morgan et al. 1994), and fleshy fruits in particular have evolved repeatedly in a wide variety of clades (Amaryllidaceae, Snijman and Linder, 1996; Rubiaceae, Bremer and Eriksson, 1992; Bremer et al., 1995; Melastomataceae, Clausing et al., 2000; Malpighiaceae, Davis et al., 2001). Using trees derived from the sequences of various chloroplast and/or nuclear genes (see Soltis et al., 1998, for a discussion of DNA sequencing and its application to angiosperm phylogeny), it has been demonstrated that, in Rubiaceae for example, fleshy fruits have arisen 12 times independently (Bremer et al., 1995). Clausing et al. (2000) have shown that previous classifications of the family Melastomataceae that relied heavily on fruit type (dry capsules considered plesiomorphic or ancestral and fleshy true berries as apomorphic or derived) were misleading, since the fleshy fruits in the family were not all equivalent or homologous. Homology is of critical importance in the assessment of characters for phylogenetic analysis, for example, in the Melastomataceae, soft berries are anatomically quite variable and have arisen from the development of fleshiness of different parts of the developing ovary (Clausing et al., 2000) and very slightly different developmental trajectories in some Rosaceae can lead to what appear to be radically diffferent fruit types (Evans and Dickinson, 1999). Thus, the character ‘fruit a soft berry’ is not the same thing in every clade where it occurs. The accurate interpretation of fruit types is also important, in some groups it has been found that fruits previously thought to be berries (i.e. indehiscent) were incorrectly designated as such (Barnes and Rozefelds, 2000), and correct assessment of fruit type has resulted in a better understanding of the relationships and patterns of character evolution.

Fruits are more than simply characteristics of angiosperms—they are the means by which seeds are dispersed and as such are clearly under strong selective pressure. Fruit type, however, is not the only important fruit trait that potentially affects seed dispersal. It has been suggested that tight co‐evolution between plants and their seed dispersers would result in fruit quality being directly related to seed dispersal quality (McKey, 1975), with specialist seed dispersers feeding on nutritionally higher quality fruits. This paradigm had much in common with that thought to be controlling plant–pollinator interactions, which do indeed show such tight and specific relationships. It has sometimes been assumed that differences in quality of seed dispersal can fully explain variation in fruit traits (Schupp, 1992). Wheelwright and Orians (1992), however, felt that expectations of such tight co‐evolution were highly misguided, as the target for seeds (a good germination site) was much less obvious than that for pollen (a con‐specific stigma) and that frugivores are ‘paid in advance’ for future services. Studies on plants and their seed dispersers undertaken since these paradigms were suggested have shown that if co‐evolution does exist it is a much more diffuse phenonomen, with convergence on broad suites of fruit traits and fruit‐feeding behaviours occurring in natural communities (Murray et al., 2000). Recent studies have also failed to corroborate the assumption that differences in quality determine fruit trait variation (Cipollini and Levey, 1997a). Fruit quality, however, is important to frugivores, and will affect seed dispersal in some way. Quality can be assessed in a variety of ways using protein or fat content (McKey, 1975), caloric value or levels of secondary metabolites (Cipollini and Levey, 1997a). This latter measurement of fruit quality is often not considered, but toxins in fruit are of great importance not only to natural frugivores, but to humans who cultivate fruits for their own purposes.

Solanaceae contain many taxa of importance, both agronomically (potatoes, tomatoes, peppers) and medicinally (mandrake, tobacco, deadly nightshade, henbane). Members of the family occur worldwide, but the highest species diversity is found in the Neotropics. The family is very diverse, both in terms of life form, with species ranging from ephemeral herbs (Leptoglossis and Schizanthus of the Chilean deserts) to large forest trees (Duckeodendron of the Amazon), and in flower morphology (Knapp, 2002a). Estimates of species diversity in the family range from 9000–10 000 species, with about 2000 of those being species of the large cosmopolitan genus Solanum (see Knapp, 2002a, for a review of genera in the family). Taxonomic work in the family largely proceeded at the generic level and the last complete treatment was that of Dunal (1852). Recently, Hunziker (2001) has provided a complete treatment of the family at the generic level, including a new classification system.

In recent years, three advances have contributed significantly to the understanding of the relationships of flowering plants: (1) an explicit framework for interpreting phylogeny (cladistics; Hennig, 1966; Kitching et al., 1998), (2) the incorporation of data from the DNA sequences of a variety of regions of the genome, both from the nucleus and the chloroplast, to construct phylogenetic hypotheses in the form of trees, and (3) computer technology that is able to handle such large data sets. Systematic studies in the Solanaceae have benefited greatly from these advances, with phylogenetic studies being published for a variety of genera and species groups (Knapp, 1989; Mione et al., 1994; Bruneau et al., 1995; Axelius, 1996; Knapp and Helgason, 1997; Estrada and Martínez, 1999; Bohs, 2001; Peralta and Spooner, 2001) and the resolution of the relationships of several enigmatic genera (i.e. Duckeodendron, Fay et al., 1998; Schizanthus, Olmstead et al., 1999; Sclerophylax, L Bohs, personal communication) whose inclusion in the family has been the subject of much debate (D’Arcy, 1991). Molecular studies have also shown that the tomatoes, previously recognized as the segregate genus Lycopersicon Mill., are deeply nested with the genus Solanum (Spooner et al., 1993; Peralta and Spooner, 2001), as are the tree tomatoes (ex Cyphomandra Sendtn.; Bohs, 1995, 2002; Bohs and Olmstead, 1997). Some disagreement still exists as to the utility of these new classifications (Hunziker, 2001; Chetelat, personal communication), but as more gene sequences from both the chloroplast and the nucleus are analysed, the results are being shown to be robust (L Bohs, personal communication; RG Olmstead, personal communication). Although differences in taxonomic arrangement are perceived as being disruptive in the short term, the robustness of phylogenetic hypotheses allows the accurate and rigorous examination of character evolution, thus leading to a better understanding of how members of the family have evolved and are related.

Solanum is by far the largest and most diverse genus in the family, with between 1000 and 2000 species growing in all habitats from rainforests to the world’s driest deserts. Floral morphology in Solanum follows a basic radially symmetrical plan (but see Knapp, 2001, 2002a), and solanums in flower are remarkably easy to recognize. The genus was last treated taxonomically by Dunal (1852), and has traditionally been divided into two major groups, the spiny solanums (with attenuate anthers and usually spiny foliage and stems) and the non‐spiny solanums (with oblong anthers and no spines) (D’Arcy, 1972; Bohs and Olmstead, 1999; Knapp, 2001). The non‐spiny solanums are clearly a paraphyletic group (a group containing some, but not all of the descendants of a common ancestor), and recent work using a variety of chloroplast and nuclear genes by L Bohs (University of Utah) has gone a long way towards defining monophyletic groups within the genus (Bohs, 2002). Several segregate genera traditionally treated, apart from Solanum, have recently been found to be nested within the genus and those species have been formally transferred (Spooner et al., 1993; Bohs, 1995). In the discussions here, these species are treated as part of the larger, monophyletic Solanum.

In this paper, the patterns of fruit evolution in the economically important family Solanaceae are examined using a recently published molecular phylogeny (Fay et al., 1998; Olmstead et al., 1999). This examination in a phylogenetic framework is used specifically to address the following questions: (1) what is the range of fruit types found in the family? (2) how many independent origins of distinct fruit characters can be inferred from the phylogeny of the family? (3) are there any instances where homology of a particular fruit type is questionable? and (4) are there trends in fruit type that can be associated with habitat or dispersers? In addition, there will be a close examination of fruit type diversity and fruit chemistry in the large genus Solanum, where greater understanding of the molecular mechanisms of fruit development gained from studies of agronomically important crops may help in the understanding of evolutionary patterns, particularly in relation to fruit dispersal modes.

Materials and methods

Solanaceae: evolution and phylogeny

Olmstead et al. (1999) produced a phylogenetic classification of the Solanaceae using two types of cpDNA data, (1) restriction site analysis of the entire chloroplast genome, and (2) DNA sequencing of two chloroplast genes, rbcL and ndhF (for details see Fay et al., 1998; Olmstead et al., 1999). Fay et al. (1998) included in this large data set the genus Duckeodendron, previously segregated as the family Duckeodendraceae (Kuhlmann, 1947; Thorne, 1992: Fay et al.’s study was completed after that of Olmstead et al., but published before it). Duckeodendron is included here in the Olmstead et al. (1999) classification, at the rank of tribe (as suggested by Fay et al., 1998; but see Bremer et al., 2001). The phylogenetic classification presented in Olmstead et al. (1999) unfortunately used Latinized tribal and subfamilial ranks, but without validation. In Table 2 the Olmstead et al. (1999) classification is presented, with the names used in parentheses to indicate their use here as sensu Olmstead et al. only. Each of the monophyletic groupings has also been given an informal name (i.e. the Petunia clade) to facilitate discussion. These informal clade names will be used throughout this paper. Genera are listed in each clade in alphabetic order in Table 2, following Olmstead et al. (1999). Several segregate genera have been described recently (Hunziker, 2000a, b) and are included in the clade with the genus from which they have been distinguished. The Olmstead et al. (1999) classification has been used here rather than the more recent Hunziker (2001) scheme, not because molecular data are necessarily a better indicator of relationship, but because Hunziker’s classification recognizes many paraphyletic and some polyphyletic groups and in so doing becomes less useful for the examination of evolutionary patterns than the Olmstead et al. classification, which is based on monophyly. Monophyletic groups are natural taxa, containing all the descendants of a common ancestor, while paraphyletic groups contain only some of the descendants of a common ancestor (birds are a classic example; Wiley, 1981) and polyphyletic groups contain taxa who do not share a close common ancestor. Monophyletic groups are more informative for the examination of character evolution.

For each genus, the possession of one or many of five fruit types has been recorded, defined a priori as relatively broad categories and in line with usage in other phylogenetic frameworks (Bremer and Eriksson, 1992; Clausing et al., 2000; Bradford and Barnes, 2001). Table 1 provides definitions for the fruit types identified here for the Solanaceae, with an indication of some of the terms used by Spujt (1994) that have been subsumed here under these categories (the list is not by any means exhaustive). Berries (with and without stone cells, see below and Table 1; Figs 3E, F, 4B, C, D), capsules (Figs 1B, C, 3A–D), dehiscent non‐capsular fruits (Fig. 2), mericarps (Fig. 4A), and drupes or pyrenes (Fig. 1A) have been used as the five basic fruit types in the family based on initial homology assessments in the literature and from examination of both live and preserved specimens. Fruiting specimens have been examined of all genera in the herbaria of the Natural History Museum, London (BM), the Missouri Botanical Garden (MO) and the Royal Botanic Gardens, Kew (K), leading to some differences with fruit types published in the literature (e.g. the fruit of the rare Nevada endemic Oryctes is considered to be a non‐capsular dehiscent fruit, while Hunziker, 2001, states that it is a capsule). Table 2 lists all the genera in the family and the fruit types found within them; if a genus has several different fruit types, it is indicated here. Figures 1–4 illustrate some of the fruit type diversity within the family.

In order to assess the distribution of fruit types in the family, the framework molecular phylogeny of Olmstead et al. (1999) has been used here, using as the terminals the monophyletic groups of their classification (Figs 5–7). Olmstead et al. (1999) did not have molecular data for all the genera of the family, but subsequent sampling has by and large confirmed the monophyletic groupings they identified (RG Olmstead, personal communication). Where newly sampled genera have been found to be members of different groups, this has been indicated in the notes to Table 2. Fruit characters have been mapped onto the framework tree using MacClade version 3.06 (Madison and Madison, 1996) showing all most parsimonious states at each node. The topology of this tree should not be interpreted as the true phylogeny of the family, but instead as a heuristic device with which to examine the distribution of fruit types in the family. Those interested in the more complex trees resulting from the molecular analyses are referred to the original literature (Olmstead et al., 1999; Fay et al. 1998).

The genus Solanum: evolution and phylogeny

To examine the distribution of fruit types with relation to phylogeny at another hierarchical level in the same set of data, fruit types in the genus Solanum have been recorded using the monophyletic groups identified by Bohs (2002) using sequences of the chloroplast gene ndhF. Although these groups are preliminary, they represent the first attempt to assess monophyly in the genus and are largely supported by both morphological (S Knapp, personal observation) and other molecular data, both chloroplast (Bohs and Olmstead, 1997, 1999; Olmstead and Palmer, 1997) and nuclear (L Bohs, personal communication). Since the relationships of these monophyletic groups are not yet robustly supported, the consensus tree has not been reproduced (Bohs, 2002; for a tree produced with a much smaller sample see Bohs and Olmstead, 1997, 1999). Several group relationships, however, are robustly supported by all these molecular studies; (1) the segregate genera Cyphomandra and Lycopersicon form part of a monophyletic Solanum and should be included, (2) the spiny solanums form a robustly supported monophyletic group and are derived within the genus, and (3) traditional subgeneric classifications of the genus were largely composed of paraphyletic and polyphyletic groups (see Bohs and Olmstead, 1997, for a discussion). Species of Solanum as delimited here all possess berries (Table 2; Fig. 5), but within that category diversity is immense. Mature Solanum berries are of several basic kinds; (1) soft, brightly coloured, juicy berries (e.g. S. dulcamara; Figs 8D, 9A, D), (2) brightly coloured relatively dry berries (e.g. S. pseudocapsicum; the Jerusalem cherry), (3) soft, greenish berries (e.g. S. mite; Knapp and Helgason, 1997; Figs 8A, E, 9B), (4) hard, greenish or yellowish berries (e.g. S. carolinense; Cipollini and Levey, 1997a; Fig. 8B), or (5) dry berries enclosed in an accrescent calyx (e.g. S. rostratum; Fig. 9E, F). This last category includes those berries identified as censer fruits (Symon, 1984; Lester and Symon, 1989); dry berries borne on long flexible stalks where seeds are shaken out by wind or impact.

Results

Solanaceae: evolution and phylogeny

The sister group (most closely related family) for the Solanaceae is the Convolvulaceae (Savolainen et al., 2000), where capsular fruits are the norm (Cronquist, 1981). Outgroup comparison with the sister group thus would indicate that capsular fruits were the plesiomorphic state in the Solanaceae, and capsules occur in most of the basal clades of the tree (Fig. 5A). The only other occurrence of capsular fruits in the family is in the Datura clade, where the genus Datura has large, usually spiny capsules (Fig. 3E, F) with relatively large, ant‐dispersed seeds (Persson et al., 1999). From the phylogeny this can be seen to represent a secondary derivation of the capsule from a berry‐fruited ancestor (see below).

Two basal clades however, Duckeodendron and the Goetzea clade both possess drupes (Fig. 5B). Similarities of these drupaceous fruits have been noted by Carlquist (1988), but significant differences in the large size and fibrous mesocarp of Duckeodendron perhaps mean they are derived differently. Baehni (1943) used the fact that the drupe of Goetzea was derived from a single locule, rather than from two locules, to exclude it and its close relatives from the Solanaceae. Both Lycium and Grabowskia (the Lycium clade) have fruits that are defined as pyrenes—having one or two seeds enclosed in indurate endocarp and surrounded by fleshy exocarp. This clade is not closely related to either the Goetzea clade or to Duckeodendron, so it is perhaps important to assess the homology of this fruit type in the family. Further to the distribution of drupes/pyrene‐type fruits is the case of Lycianthes, where one subgenus (Eulycianthes Bitter) has few (c. 8, as opposed to the many in the rest of the genus) seeds, each of which is enclosed in a bony endocarp superficially analogous to the situation in Lycium (Symon, 1987). The fruit of Saracha (in the Iochroma clade), although scored here as a berry, has most of the seeds surrounded by a bony substance, apparently derived from the copious stone cells (personal observation; Alvárez, 1996).

Berries are clearly derived in the family and seem to have three separate origins (Fig. 6A). Two equally parsimonious possibilities exist for this pattern. Either berries define a monophyletic clade and they have been lost three times, or berries have three separate origins, in Cestrum of the Cestrum clade, in Duboisia of the Anthocercis clade and in the subfamily ‘Solanoideae’, where berries do occur in all terminal taxa (Table 2; Fig. 6). The first of these two scenarios involves four steps, while the second involves only three and is thus preferable using the criterion of parsimony. This is somewhat like the case in the Rubiaceae (Bremer and Eriksson, 1992) or Melastomataceae (Clausing et al., 2000), where soft berries have clearly evolved many times independently. Latua, an unusual and rare Chilean endemic which would represent an independent origin of berries (Table 2), has not been analysed using molecular methods and Olmstead et al. (1999) provisionally included it in the Petunia clade. The relationships of Latua are unclear and it has been excluded from the analysis until more detailed molecular and morphological studies are undertaken. The berries of Cestrum (Cestrum clade) are unusual in the family in that in a few taxa they are occasionally somewhat capsular at the apex (Francey, 1935; Benítez de Rojas and D’Arcy, 1998). Sessea, the sister group of Cestrum, has strictly capsular fruits with winged seeds. The berries in the most derived clade (Olmstead et al.’s subfamily ‘Solanoideae’) are all morphologically similar, with two carpels, axile placentation and mostly lenticular seeds (Symon, 1987). Many of the genera in this large clade contain stone cells (Fig. 6B), accretions of sclerenchyma that occur in the fleshy part of the berry (Bitter, 1911, 1914). In some cases these stone cells can be quite numerous, Solanum multivenosum Symon (an Australian species of the Archaeosolanum clade, see below) has an average of 180 (range: 162–1110) stone cells per fruit (Symon, 1987, 1994). Bitter (1911, 1914) suggested that these structures were the remnants of a once stony endocarp (i.e. that the ancestral fruit type in the family was a drupe). Bohs (1994) found that stone cells in the Cyphomandra clade (including Solanum betaceum Cav., see below) were lignified with high concentrations of sodium and calcium. From the distribution of stone cells in berries (Fig. 6B), it is clear that they are a derived character relative to the possession of berries, and have apparently been either gained or lost several times.

The unusual sclerified mericarps of Nolana are an autapomorphy (unique derived character) of that clade (Fig. 7B; Tago‐Nakazawa and Dillon, 1999). These peculiar fruits, which are generally composed of 5–15 carpels (Di Fulvio, 1971; Bruno, 1994) are the principal reason some authors have maintained the genus at the family level (Nolanaceae; Mesa, 1981; Hunziker, 2001). Nicandra, however (the Nicandra clade; Fig. 7B; Table 2), also has a 5‐carpellar ovary and the fruit is a fleshy berry. The mericarps of Nolana are variously connate (Fig. 4A) when immature, and vary from laterally united and multi‐seeded to completely free and single‐seeded (Tago‐Nakazawa and Dillon, 1999).

Non‐capsular dehiscent fruits occur in the Hyoscyamus clade, where most of the genera (with the exception of Atropa; Fig. 4B; Table 2) have pyxidia—fruit of this sort has been usually classified as a type of capsule (Roth, 1977; Spujt, 1994), but in the Solanaceae immature fruits of these genera are indistinguishable from immature berries (Fig. 2A, B). The dehiscence line around the circumference develops only as the pericarp dries out, and the entire structure is held tightly or loosely in an accrescent calyx (Fig. 2C, D). Although these sorts of dehiscent fruits have traditionally been considered to be capsular (Roth, 1977; Spujt, 1994), it is clear at least in the case of the Solanaceae that pyxidia are highly modified berries. The distribution of this fruit type on the phylogeny shows that it has arisen three times independently, in the Hyoscyamus clade, in Oryctes of the Physalis clade and several times (see below) in the genus Solanum.

The genus Solanum: evolution and phylogeny

Despite the apparent uniformity of fruit type in this large genus, fruits in Solanum show a remarkable range of subtle variation. The colour of berries in Solanum varies from black, through red and orange to yellow and green (Table 3; Figs 8, 9). In general, brightly coloured berries tend to be juicy and extremely soft, while those berries that are green at maturity are harder, and are occasionally extremely woody (Symon, 1987). Many of the green berries suddenly become yellower and soft over a period of several hours, at which point they are taken by birds or bats (Knapp, 2002b). Bright colour is derived in the tomatoes (Peralta and Spooner, 2001), with the cultivated tomato Solanum lycopersicum L. and its close relatives the only taxa in the group with red or orange fruits (Marshall et al., 2001). In other clades of Solanum, patterns of fruit colour evolution have not been specifically investigated and are not well understood. As in the Solanaceae as a whole, stone cells (Bitter, 1911, 1914) are widely distributed in the genus Solanum, and appear to be confined to particular monophyletic groups (Table 3).

Fruit size in Solanum also varies considerably; cultivated varieties of Solanum lycopersicum (the tomato) and S. melongena L. (the aubergine) can be extremely large (Lippman and Tanksley, 2001; Passam et al., 2001) and weigh up to several kilograms, but wild species also have a wide range of fruit sizes. Some of the smallest solanum fruits occur in the morelloid/dulcamaroid clade, with berries of less than 1 cm in diameter, while the largest wild solanum berry is that of the wolf‐fruit, S. lycocarpon A. Rich. of the Brazilian cerrados, which can be up to 15 cm in diameter! Placentation patterns in the genus are also quite variable, but all are based upon the 2‐carpellate, axillary condition (Nee, 1986; Symon, 1987). Placental enlargement and expansion and an increase in locule number have occurred in isolated species in all clades, not only in those taxa cultivated by humans for their fruits.

Discussion

Discussions of ecologically relevant character evolution must be framed in a phylogenetic context for testable hypotheses to be constructed. Fruit morphology, although traditionally of great importance taxonomically, is one such character of ecological importance. Fruit represent the vehicle for seed dispersal, a critically important stage of a plant’s life cycle that is certainly under strong selective pressure. It is important not only to examine the distribution of the range of fruit types in the group in question, but also to do this using a phylogeny that does not use fruit type in its construction (Bremer and Eriksson, 1992). This is not to say that fruit type and fruit characters are never features that define monophyletic groups. Fruit anatomical characters are important for recognizing generic level relationships in the Oleaceae (Rohwer, 1996) and in the dogwoods (Cornus, Cornaceae) the two major clades were congruent with fruit colour and the presence of particular fruit chemicals (Xiang et al., 1996). Members of the family Solanaceae have a relatively restricted suite of fruit types, and homoplasy (structural similarity due to independent origin rather than common ancestry) appears to be less common than in other groups studied (Bremer and Eriksson, 1992; Clausing et al., 2000). In the family, capsules are plesiomorphic, and berries—soft indehiscent fruits—have either evolved once and been lost three times, or have evolved three times independently. Distinguishing between these two hypotheses depends upon the homology of the character ‘fruit a berry’ in the groups concerned. As mentioned above, Latua, while possessing a soft berry structurally very similar to those found in the diverse berry‐fruited clade (Olmstead et al.’s subfamily ‘Solanoideae’), has traditionally been placed in the Petunia clade (Olmstead et al., 1999), based on its alkaloid chemistry. Hunziker (2001), however, places Latua in a monotypic tribe related to Cestrum. The genera Cestrum and Duboisia represent berry‐fruited members of clades otherwise possessing capsular fruits (Table 2), arguing for the independent (homoplastic) origin of soft, indehiscent fruits in those genera. Some species of Cestrum have semi‐capsular fruits, where the locules appear to separate at the apex (Francey, 1935). The anatomy of these putatively transitional species will prove of interest, as will structural studies of the vasculature of the berries of Cestrum and Duboisia compared with that of their capsular‐fruited close relatives. In the Melastomataceae, Clausing et al. (2000) showed that the anatomical origin of the fleshy portion of the fruit differed in several clades, thus showing that traditional coding of a fruit type character as ‘fleshy’ or ‘dry’ was not tenable.

The possession of berries is a synapomorphy of the large, derived clade identified as subfamily ‘Solanoideae’ by Olmstead et al. (1999). This clade (comprising the terminals between the Lycium clade and the Physalis clade; Fig. 6) represents the vast majority of species diversity within the family. Within this clade berries have been lost several times; in the Hyoscyamus clade, where fruits are pyxidia, in the genus Datura with capsular fruits with large seeds, and in several other genera where variously dry and modified berries have arisen (Fig. 7A, B; Table 2; Symon, 1979). Homoplastic evolution of derived fruit types in the Solanaceae has involved several losses of soft indehiscent fruits, rather than the reverse as is hypothesized to be the case in other families (Bremer and Eriksson, 1992; Clausing et al., 2000).

The presence of stone cells in the berries of members of this clade is scattered on the tree (Fig. 5B). Interestingly, stone cells never occur in the berries of Cestrum or Duboisia, lending support to the independent origins of berries in those genera. In both Lycium (Lycium clade) and Lycianthes (Capsicum clade), stone cells are commonly found in fleshy fruits and a few species (Lycianthes lycioides (L.) Hassl., Lycium section Sclerocarpellum C. Hitchc.; Bernardello, 1986; Miller, 2002) have pyrenes, where the few (usually 2–8) seeds are completely surrounded by a hard, bony endocarp. It is tempting to suppose that the bony endocarp is derived from the stone cells, or vice versa, as suggested by Bitter (1911, 1914). His suggestion that stone cells were the remnants of the bony endocarp of an ancestral drupaceous fruit is not supported by the distribution of fruit types on the molecular phylogeny. Bernardello (1983) examined the ontogeny of fruits of Lycium and Grabowskia and found that the sclerotic granules from the mesoderm gradually enclosed the seeds, beginning at the apex and proceeding to the base. He concluded that the drupe was the most ‘advanced’ fruit type in the Solanaceae and that the presence of sclerotic granules (stone cells) was an indicator of fruit specialization. Although Duckeodendron and the members of the Goetzea clade do possess drupaceous fruits, recent evidence appears to suggest that they form a single, albeit rather poorly supported, clade (RG Olmstead, personal communication) thus suggesting that true drupes have evolved once as an autapomorphy in that clade. The structural homologies of these drupes and pyrenes have never been assessed (Carlquist, 1988), and the fibrous endocarp of Duckeodendron is anatomically quite distinct from the more bony endocarp of the members of the Goetzea clade. The pyrenes found in Lycium and Lycianthes most probably have an entirely separate origin to the drupes of Duckeodendron and Goetzea. In Lycium, section Sclerocarpellum has been shown to be polyphyletic, pyrenes have independently arisen at least twice in the genus (Miller, 2002), in addition to having evolved independently in Grabowskia. In the genus Saracha (Physalis clade) stone cells are abundant and usually completely enclose a few to most of the seeds in the berry (personal observation; Alvárez, 1996).

Pyrenes in Lycium and Lycianthes, both members of clades with stone cells in the berries, are derived fruit types, rather than ancestral as suggested Bitter (1911, 1914). It is tempting to speculate that a reduction in seed number coupled with accretion of stone cells around the seed led to the pyrene in all cases where it occurs in the family, but detailed anatomical and developmental studies are lacking (Bernardello, 1983). The mericarps of Nolana are also a fruit type probably derived from a berry (Tago‐Nakazawa and Dillon, 1999). In young flowers all the carpels are connate at the base (Di Fulvio, 1971; Bruno, 1994), and each mature mericarp represents a single carpel rather than half a carpel as is the case in Lamiaceae and Boraginaceae (Mesa, 1981). Some structures, perhaps like the stone cells of Lycium or the sclerites composing the endocarp of the pyrene of Grabowskia, may be involved in mericarp structure and development and it is clear from the sister group relationship of Nolana and the Lycium clade (Fig. 7B) that structural homologies should be explored.

The fossil record for Solanaceae is relatively recent, with the earliest fossil known from the Eocene (Collinson et al., 1993) and no inferences can be made based on first occurrence in the Solanaceae of particular fruit types in the fossil record. This Eocene London Clay fossil Cantisolanum daturoides EM Reid and MEJ Chandler consists of two seeds, similar in morphology to those occurring in the berry‐fruited clade, indicating that fleshy fruits had probably evolved at least by that date. Some European fossils attributable to modern genera (Physalis, Solanum) have been dated from the middle to late Miocene (Collinson et al., 1993), indicating that these are at least minimum dates for the occurrence of modern fleshy fruits in the Solanaceae. Timings obtained from molecular studies indicate that the Solanales (the inclusive group including the Convolvulaceae) had an origin some 30 million years before the first fossil evidence (82–86 mya versus 53 mya; Wikström et al., 2001). The genus Nolana has been estimated to be some 30 million years old (Eocene to Miocene; Tago‐Nakazawa and Dillon, 1999), thus development of the apomorphic mericarps of that clade is concomitantly ancient. It is clear from molecular evidence and dating that rapid diversification and speciation in the family occurred in the late Cretaceous to early Tertiary, similar to the pattern of relatively recent evolution found for other species‐rich derived clades of angiosperms (Magállon et al., 1999; Magállon and Sanderson, 2001).

Fruit type evolution, while constrained by phylogeny, is also clearly influenced by habitat, including dispersal agents. Several broad generalizations, however, can be made. Capsular fruits, plesiomorphic in the family, are found in taxa that are generally herbaceous or shrubby inhabitants of dry areas. Species of Nicotiana for example, are weedy shrubs or herbs found in secondary habitats (Goodspeed, 1954) and are often in dry areas such as the southwestern USA or central Australia. In the Solanaceae, taxa with capsular fruits also tend to have very small seeds (e.g. Nicotiana, with thousands of seeds <0.25 mm long in a single capsule) or strongly winged seeds (e.g. Sessea and Metternichia). Small seeds and capsular fruit have been associated with dry, open habitats, while larger seeded fruits have been associated with closed forests (Eriksson et al., 2000). To a certain extent this holds true for the Solanaceae, Duckeodendron is a large canopy tree of Amazonia, while Schizanthus is a herb of dry coastal Chile, but all the members of the Goetzea clade, also with drupes, occur in dry areas of the Caribbean.

Seed dispersal by animals has also been implicated in the evolution of fleshy fruits in angiosperms as a whole (Tiffney, 1984; Eriksson et al., 2000). Janson (1983) suggested that fruit size, colour and morphology are adapted to the characteristics of animals that eat that fruit, but his study did not take into account the effects of phylogeny on the distribution of fruit morphologies in the forests he studied. In the Solanaceae, fleshy berries are eaten and the seeds dispersed by a wide variety of vertebrates, including birds, bats and small rodents (Symon, 1979; Cipollini and Levey, 1997a, b, c; Tewksbury and Nahban, 2001; Knapp, 2002b).

It is perhaps more illuminating to look at habitat characteristics associated with the evolution of dry, dehiscent fruits in the large berry‐fruited clade than to attempt to explain the origin of fleshy fruits as an adaptation to animal dispersal. In the Hyocyamus clade, with the exception of Atropa with a fleshy berry (Fig. 4B), the dry fruits open within the calyx (Fig. 2). In the endemic Tibetan genus Przewalskia, the unit of inflated calyx plus dry fruit sometimes acts as a tumbleweed, scattering seeds as it blows through the habitat (M Gilbert, personal communication). In many species of Hyoscyamus, seeds are scattered by wind as the dry inflorescence is moved about, and in Scopolia, the entire plant dries up and falls to the ground, ripe dehiscent fruit and all (Hoare and Knapp, 1997). Species of Nolana all occur in arid to semi‐arid habitats along the west coast of South America, and the mericarps are, in general, passively distributed. A persistent accresent calyx serves as a cup to contain the loose mericarps, and in one annual species, the entire plant curls up at anthesis, breaks free and acts as a tumbleweed (MO Dillon, in litt.). Species of Datura also all occur in dry or seasonally dry habitats, where their seeds, with an oily elaiosome (Persson et al., 1999) are collected by ants. All of the species of Solanum that possess derived, dry and dehiscent fruit types occur in dry habitats (Symon, 1984; Lester and Symon, 1989; Whalen, 1979). Members of Solanum section Androceras (Whalen, 1979) are all dispersed as tumbleweeds, and the seeds drop out of the enclosing calyx as the plant is blown about (Fig. 9E). The Australian species with censer mechanisms and the Mexican species with a splash cup fruit (Fig. 9F) all have seed dispersal assisted by knocking or rain. That all these dry and variously dehiscent fruit types derived from fleshy berries are found on species occurring in arid zone habitats suggests that environmental factors have been important in the convergent and multiple evolution of these fruit types in the Solanaceae.

Secondary chemistry of fruits is an important factor influencing seed dispersers and predators. Many ripe fleshy fruits whose primary function is to attract seed dispersers also contain unpleasant chemicals that deter consumption by vertebrates. This apparent paradox has drawn the attention of those interested in the adaptive roles of fruit traits (Cipollini and Levey, 1997a, b, c), but this phenomenon has rarely been addressed in the light of phylogenetic constraints (Cipollini et al., 2002). Theoretically, differences in seed dispersal quality should influence the evolution of fruit traits (Janson, 1983), but this has not been supported for fruit type using null phylogenetic models (Bremer and Eriksson, 1992; Herrera, 1992). In the Solanaceae, secondary chemistry has been studied in the chile peppers (Capsicum) and in the diverse genus Solanum, where wide variation exists. Chile peppers have high concentrations of the extremely pungent chemicals, capsaicinoids, in both ripe and unripe fruits (Kosuge and Furata, 1970; Contreras‐Padilla and Yahia, 1998). As pepper fruits ripen, pungency decreases due to peroxidase degradation (Contreras‐Padilla and Yahia, 1998), but ripe bird peppers (Capsicum annuum L.), although attractive to and eaten by birds, still contain enough pungent compound to deter mammals (Tewksbury and Nabhan, 2001). Tewksbury and Nabhan (2001) found that the hypothesis of directed deterrence—where chemicals in ripe fruit function selectively to discourage seed predators while not deterring seed dispersers—was supported in Capsicum using populations in the southwestern USA. The small mammal seed predators were put off by the pungency of pepper fruits, while dispersers, birds, were not. Selective pressures influencing the acquisition and retention of toxic chemicals in fruits are likely to be the results of trade‐offs associated with differing ecological requirements at different life stages of the plant.

The directed deterrence hypothesis has also been tested in the genus Solanum, where fruit secondary chemistry is well known. Solanum fruits have widely varying concentrations of toxic steroidal glycoalkaloids such as solasodine (Carle, 1981; Cipollini et al., 2002). In the morelloid clade, the concentrations of these compounds decrease dramatically as fruit matures, while in many members of the spiny solanums, concentrations remain high at fruit ripeness (Cipollini and Levey, 1997c). It has been suggested that these compounds act as antifungal defences (Cipollini and Levey, 1997b) or act to deter seed predators (as in Capsicum above; Cipollini and Levey, 1997a). Cipollini and Levey (1997a, b, c) found that the directed deterrence hypothesis was not supported in Solanum; seed predators and dispersers alike were deterred by fruit secondary chemicals. Intriguingly, Wahaj et al. (1998) found that alkaloids in the ripe fruits of S. americanum (a member of the morelloid clade) affected gut retention time of seeds, suggesting that chemicals may influence seed deposition patterns by frugivores. Amazonian maned wolves expel worms after eating the fruits of S. lycocarpon, suggesting they are perhaps using the secondary chemicals contained therein as a vermifuge (S Knapp, personal observation; Courtenay, 1994; Cipollini et al., 2002). A multiplicity of purposes may be served by fruit secondary chemicals, some of which have simple adaptive explanations, but others are more complex. Examining the phylogenetic context in which these fruit chemistry characteristics are found led Cipollini et al. (2002) to conclude that phylogeny does not have an important influence on fruit chemical or morphological trait variation within the species they studied. Their preliminary data are intriguing because they suggest that rather than seeking explanations using just phylogeny or ecology, it will be more profitable to investigate the interaction of phylogeny and ecology with respect to the evolution of fruit traits. Real understanding of the evolution of a wide variety of fruit traits, however, will only come with an integration of the study of phylogeny, ecology and development, as it is their interaction that ultimately determines phenotype.

With the exception of the crucifer Arabidopsis thaliana (L.) Heynh., members of the Solanaceae are among the best understood of flowering plants in terms of their genomes, largely due to their importance to agriculture. The understanding of the genetic regulation of fruit shape and size in the tomato (Solanum lycopersicum) has benefited greatly from a comparative approach using a few closely related species (Grandillo et al., 1996; Frary et al., 2000; Lippman and Tanksley, 2001), but such studies are rarely put into a wider phylogenetic context. The molecular underpinnings of both fruit development (Gu et al., 1998; Liljegren et al., 2000; Ferrándiz et al., 1999, 2000) and fruit ripening (Giovannoni, 2001) have been intensively studied in recent years, and it is clear that the commonalities of genetic control systems and regulatory mechanisms will allow new questions to be asked about the evolution of fruit types and fruit traits such as colour or secondary chemistry. Studies undertaken in tomatoes with respect to fruit size and shape (see references above; Gautier et al., 2001; and for peppers, Aloni et al., 1999), chemical composition (Klann et al., 1996; D’Aoust et al., 1999), coloration changes during ripening (Giovannoni, 2001) and locule number (Lippman and Tanksley, 2001) will all be of interest in light of the variety of fruit traits such as colour (brightly coloured to green at maturity), fruit texture (soft and juicy to hard and woody) and fruit size (small, e.g. Solanum americanum to extremely large, e.g. S. lycocarpon) that have evolved via natural selection in Solanum.

At the family level recent studies of the molecular regulation of fruit development will be of interest as well. The variety of dehiscence mechanisms present in the Solanaceae (septicidal and loculicidal capsules, dehiscent non‐capsular fruit) may allow the elucidation of how the SHATTERPROOF MADS‐box genes (Ferrándiz et al., 2000) influence a variety of dehiscence types in closely related organisms. Lignification, perhaps like that involved in the dehiscence mechanism of Arabidopsis (Ferrándiz et al., 2000; Liljegren et al., 2000), is also involved in the development of the hard berries found in many clades of Solanum (Symon, 1987), but neither developmental nor anatomical homology of these processes is well understood.

Conclusions and possible directions for future work

Because several important fruit crops are found in the family Solanaceae and molecular developmental studies on these taxa have been extensive, the family represents an ideal testing ground for the comparative study of the development and evolution of fruit traits in a phylogenetic context. Important too is the study of plants in their native environments—the selective pressures that, over time, have resulted in the huge variety of fruit types observed today had their origins in interaction with the environment, including dispersers and predators. The phylogeny of the family is currently being studied by a variety of workers, using both molecular (Olmstead et al., 1999; Bohs and Olmstead, 2001; Bohs, 2002) and morphological characters (Knapp et al., 1998; Estrada and Martínez, 1999; Bohs, 2001; Knapp, 2002b). This analysis, though a preliminary one, has suggested some areas for further investigation into fruit evolution in the Solanaceae. (1) Cestrum and Duboisia berries—do they have independent origins? Developmental trajectories of berries in the clades in which they occur may shed light on this question. (2) Drupes of Duckeodendron and Goetzea clade—morphology and development—are they homologous? (3) Datura capsules—how are they derived from berries? (4) Evolution of pyxidia—the mechanics of dehiscence mechanisms. (5) Molecular biology of rapid ripening in bat‐dispersed Solanum berries. (6) The distribution of ‘fruit ripening genes’ in Solanum—is there a correlation with berry type?

These are by no means the only fruit traits of interest in the family, but they are clearly areas where phylogeny and development could profitably interact to shed new light on previously under‐investigated problems. Such future study will require the collaborative efforts of taxonomists, developmental biologists, anatomists and molecular biologists. Using the comparative framework provided by phylogeny, insights gained from the study of fruit traits in single species important for agriculture or genomic studies will have a wider evolutionary relevance.

Acknowledgements

I thank Graham Seymour for inviting me to speak at the Fruit Development and Ripening symposium and to write this paper, which has shown me a new perspective on the Solanaceae; the Photographic Unit at NHM for help with the photographic plates; P Kenrick for advice on the palaeobotany of Solanaceae; A Newton for help with and use of her copy of MacClade; M Gilbert and MO Dillon for permission to use photographs; the curators of herbaria cited in the text for permission to examine the specimens in their care; I Peralta, J Mallet, L Bohs, M Nee, and M Dillon for the time they spent with me discussing things solanaceous; and two anonymous reviewers for helpful comments on the manuscript. The recent death of Ing. AT Hunziker (December 2001) has left a great hole at the heart of Solanaceae taxonomy, his work inspired a generation of systematists and he will be sorely missed.

Fig. 1. Representative fruit types in the Solanaceae. (A) Drupe of Duckeodendron cestroides Kuhlm. (base of fruit in upper right corner). (B) Capsules of Schizanthus pinnatus Ruiz and Pav. (C) Capsules of Sessea corymbiflora Taylor and Phillips. Scale bars=1 cm.

Fig. 1. Representative fruit types in the Solanaceae. (A) Drupe of Duckeodendron cestroides Kuhlm. (base of fruit in upper right corner). (B) Capsules of Schizanthus pinnatus Ruiz and Pav. (C) Capsules of Sessea corymbiflora Taylor and Phillips. Scale bars=1 cm.

Fig. 2. Representative fruit types in the Solanaceae: dehiscent non‐capsular fruit (pyxidia). (A) Physochlaina orientalis, unripe fruit. (B) Anisodus luridus, unripe fruit with the calyx cut away. (C) Przewalskia tangutica Maxim., ripe fruit showing line of dehiscence. (D) Fruit after dehiscence showing seeds (photographs courtesy of M Gilbert).

Fig. 2. Representative fruit types in the Solanaceae: dehiscent non‐capsular fruit (pyxidia). (A) Physochlaina orientalis, unripe fruit. (B) Anisodus luridus, unripe fruit with the calyx cut away. (C) Przewalskia tangutica Maxim., ripe fruit showing line of dehiscence. (D) Fruit after dehiscence showing seeds (photographs courtesy of M Gilbert).

Fig. 3. Representative fruit types in the Solanaceae. (A) Nicotiana tomentosa Ruiz and Pav., unripe capsules. (B) Nicotiana tabacum L., ripe and opened capsules. (C) Brunfelsia lactea, bright orange ripe fruit on plant. (D) Brunfelsia lactea, fruit showing dehiscence and dry seeds. (E) Datura stramonium L., unripe spiny capsule. (F) Datura ceratocaula Ortega, unripe green capsule. (G) Cestrum fragile Francey, white ripe berry. (H) Markea panamense Standl., ripe juicy berry.

Fig. 3. Representative fruit types in the Solanaceae. (A) Nicotiana tomentosa Ruiz and Pav., unripe capsules. (B) Nicotiana tabacum L., ripe and opened capsules. (C) Brunfelsia lactea, bright orange ripe fruit on plant. (D) Brunfelsia lactea, fruit showing dehiscence and dry seeds. (E) Datura stramonium L., unripe spiny capsule. (F) Datura ceratocaula Ortega, unripe green capsule. (G) Cestrum fragile Francey, white ripe berry. (H) Markea panamense Standl., ripe juicy berry.

Fig. 4. Representative fruit types in the Solanaceae. (A) Immature mericarps of Nolana carnosa (Lindl.) Miers ex Dunal (photograph courtesy of MO Dillon). (B) Atropa belladonna L., ripe fleshy berry. (C) Capsicum fruits. (D) Solanum lycopersicum L., ripe berries.

Fig. 4. Representative fruit types in the Solanaceae. (A) Immature mericarps of Nolana carnosa (Lindl.) Miers ex Dunal (photograph courtesy of MO Dillon). (B) Atropa belladonna L., ripe fleshy berry. (C) Capsicum fruits. (D) Solanum lycopersicum L., ripe berries.

Fig. 5. (A) Distribution of capsular fruits on the framework molecular phylogeny. (B) Distribution of drupes and pyrenes on the framework molecular phylogeny.

Fig. 5. (A) Distribution of capsular fruits on the framework molecular phylogeny. (B) Distribution of drupes and pyrenes on the framework molecular phylogeny.

Fig. 6. (A) Distribution of berries on the framework molecular phylogeny. (B) Distribution of berries with stone cells on the framework molecular phylogeny.

Fig. 6. (A) Distribution of berries on the framework molecular phylogeny. (B) Distribution of berries with stone cells on the framework molecular phylogeny.

Fig. 7. (A) Distribution of non‐capsular dehiscent fruit on the framework molecular phylogeny. (B) Fruit types coded as states (see legend in figure) and mapped onto the framework molecular phylogeny.

Fig. 7. (A) Distribution of non‐capsular dehiscent fruit on the framework molecular phylogeny. (B) Fruit types coded as states (see legend in figure) and mapped onto the framework molecular phylogeny.

Fig. 8. Fruit types in Solanum. (A) S. angustialatum Bitter (potato clade) green berry with unusual verrucose pericarp. (B) S. tuberosum L. (potato clade), green relatively hard berry. (C) S. jasminoides Paxt. (Dulcamaroid clade), shiny black juicy berry. (D) S. nitidum Ruiz and Pav. (Dulcamaroid clade), berries turing from green to bright red at maturity. (E) Solanum sp. (Morelloid clade), green unripe fruits showing darker lines at locule margins. (F) S. betaceum Cav. (Cyphomandra clade) brightly colored berry with juicy pulp.

Fig. 8. Fruit types in Solanum. (A) S. angustialatum Bitter (potato clade) green berry with unusual verrucose pericarp. (B) S. tuberosum L. (potato clade), green relatively hard berry. (C) S. jasminoides Paxt. (Dulcamaroid clade), shiny black juicy berry. (D) S. nitidum Ruiz and Pav. (Dulcamaroid clade), berries turing from green to bright red at maturity. (E) Solanum sp. (Morelloid clade), green unripe fruits showing darker lines at locule margins. (F) S. betaceum Cav. (Cyphomandra clade) brightly colored berry with juicy pulp.

Fig. 9. Fruit types in Solanum. (A) S. argentinum Bitter and Lillo (Geminata clade), bright orange berry (B) S. falconense S. Knapp (Geminata clade) yellow soft berry. (C) S. oblongifolium Dunal (Geminata clade) woody, hard green berry. (D) S. candidum Dunal (spiny solanum clade) bright orange densely pubescent berries. (E) S. rostratum Dunal (spiny solanum clade), spiny accrescent calyces holding dry berries. (F) S. tridynamum Dunal (spiny solanum clade), splash cup or censer fruit (dry berry).

Fig. 9. Fruit types in Solanum. (A) S. argentinum Bitter and Lillo (Geminata clade), bright orange berry (B) S. falconense S. Knapp (Geminata clade) yellow soft berry. (C) S. oblongifolium Dunal (Geminata clade) woody, hard green berry. (D) S. candidum Dunal (spiny solanum clade) bright orange densely pubescent berries. (E) S. rostratum Dunal (spiny solanum clade), spiny accrescent calyces holding dry berries. (F) S. tridynamum Dunal (spiny solanum clade), splash cup or censer fruit (dry berry).

Table 1.

Fruit type definitions used for the Solanaceae, with equivalents in the systems of Spujt (1984) in parentheses

Fruit type category Definition 
Capsule A multicarpellate fruit opening into more or less equal valves along longitudinal sutures; septicidal capsules open along ventral suture, each valve is a single carpel+placenta; loculicidal capsules open along dorsal suture, each valve is composed of half of two carpels (septicidal capsules, loculicidal capsules) 
Berry A fleshy fruit with a pulpy interior, no stony layer and many seeds (bacca; carcerulus; diclesium) 
Berry with stone cells As above, but with sclerified inclusions in the fleshy portion of the fruit 
Drupe or pyrene A fleshy fruit with one or more stones; seeds enclosed by a stony layer, the endocarp (drupe) 
Non‐capsular dehiscent fruit   A fruit that is dry at maturity, opening variously, but not along longitidunal sutures (foraminicidal capsule; pyxidium)  
Mericarps (nutlets) A usually dry fruit with single‐ or multi‐seeded units that are separate at maturity (achenarium; drupetum?) 
Fruit type category Definition 
Capsule A multicarpellate fruit opening into more or less equal valves along longitudinal sutures; septicidal capsules open along ventral suture, each valve is a single carpel+placenta; loculicidal capsules open along dorsal suture, each valve is composed of half of two carpels (septicidal capsules, loculicidal capsules) 
Berry A fleshy fruit with a pulpy interior, no stony layer and many seeds (bacca; carcerulus; diclesium) 
Berry with stone cells As above, but with sclerified inclusions in the fleshy portion of the fruit 
Drupe or pyrene A fleshy fruit with one or more stones; seeds enclosed by a stony layer, the endocarp (drupe) 
Non‐capsular dehiscent fruit   A fruit that is dry at maturity, opening variously, but not along longitidunal sutures (foraminicidal capsule; pyxidium)  
Mericarps (nutlets) A usually dry fruit with single‐ or multi‐seeded units that are separate at maturity (achenarium; drupetum?) 
Table 2.

Fruit type diversity in the Solanaceae arranged by genus Generic arrangement follows that of Olmstead et al. (1999), except where footnoted.

Clades referred to in text and figures; names in parentheses are those fide Olmstead et al. (1999); Fay et al. (1998) Genus Capsule Berry Berry with stone cells Non‐capsular dehiscent Drupe or pyrene Mericarps (nutlets) 
Schwenkia clade (Schwenkioideae) Schwenkia L.      
 Protoschwenkia Soleredera      
 Melananthus Walp.      
 Heteranthia Nees and Mart.      
Schizanthus (Schizanthoideae) Schizanthus Ruiz and Pav.      
Duckeodendron (Duckeodendreae) Duckeodendron Kuhlm.      
Goetzea clade (Goetzeoideae) Goetzea Wydl.      
 Espadaea A.Rich.      
 Coeloneurum Radlk.      
 Henoonia Griseb.      
Petunia clade (Petunioideae) Benthamiella Speg.      
 Bouchetia Dunal      
 Brunfelsia L.      
 Calibrachoa La Llave and Lex.      
 Combera Sandw.      
 Fabiana Ruiz and Pav.      
 Hunzikeria D’Arcy      
 Latua Phil.b      
 Leptoglossis Benth.      
 Nierembergia Ruiz and Pav.      
 Pantacantha Speg.      
 Petunia Juss.      
 Plowmania Hunz. and Subils      
Salpiglossis clade Salpiglossis Ruiz and Pav.      
 (Cestroideae: Salpiglossideae) Reyesia Gay      
Browallia clade (Cestroideae: Browallia L.      
 Browallieae) Streptosolen Miers      
Cestrum clade (Cestroideae:  Cestrum L.      
 Cestreae) Sessea Ruiz and Pav.      
 Metternichia Mikanc      
 Vestia Willd.      
Anthocercis clade  Anthocercis Labill.      
 (Nicotianoideae: Anthocercideae) Anthotroche Endl.      
 Crenidium Haegi      
 Cyphanthera Miers      
 Duboisia R. Br.      
 Grammosolen Haegi      
 Symonanthus Haegi      
Nicotiana (Nicotianoideae:  Nicotianeae) Nicotiana L.      
Capsicum clade (Solanoideae:  Capsicum L.     
 Capsiceae) Lycianthes (Dunal) Hassl.    
Datura clade (Solanoideae: Datureae) Datura L.      
 Brugmansia Pers.      
Hyoscyamus clade  Anisodus Link      
 (Solanoideae: Hyoscyameae) Atropa L.      
 Atropanthe Pascher      
 Hyoscyamus L.      
 Physochlaina G. Don      
 Przewalskia Maxim.      
 Scopolia Jacq.      
Jaborosa (Solanoideae: Jaboroseae) Jaborosa Juss.      
Solandra clade (Solanoideae:  Dyssochroma Miers      
 Solandreae) Juanulloa Ruiz and Pav.      
 Markea A. Rich.      
 Merinthopodium Donn. Sm.      
 Schultesianthus Hunz.      
 Solandra Sw.      
 Trianaea Planch. and Linden      
Lycium clade (Solanoideae: Lycieae) Grabowskia Schltdl.      
 Lycium L.     
 Phrodus Miers     
Mandragora (Solanoideae:  Mandragoreae) Mandragora L.      
Nicandra clade (Solanoideae:  Exodeconus Raf.     
 Nicandreae) Nicandra Adans.     
Nolana (Solanoideae: Nolaneae) Nolana L. (incl. Alona     
Iochroma clade (Solanoideae:  Acnistus Schott     
 Physaleae: Iochrominae) Dunalia Kunth     
 Iochroma Benth.     
 Saracha Ruiz and Pav.     
 Vassobia Rusby     
Physalis clade (Solanoideae:  Brachistus Miers     
 Physaleae: Physalinae) Chamaesaracha (A. Gray) Benth.      
 Leucophysalis Rydb.     
 Oryctes S. Watson      
 Physalis L. (incl. Quincula Raf.)     
 Witheringia L’Her.      
Salpichroa clade (Solanoideae:  Nectouxia Kunth      
 Physaleae: Salpichroinae) Salpichroa Miers      
Withania clade (Solanoideae:  Athenaea Sendt.      
 Physaleae: Withaninae) Aureliana Sendt.     
 Cuatresia Hunz.      
 Darcyanthus Hunz.d     
 Deprea Raf.     
 Eriolarynx (Hunz.) Hunz.e     
 Larnax Miers     
 Physalistrum Makino      
 Tubocapsicum (Wetts.) Makino      
 Withania Pauq.(incl. Mellissia Hook. f.)     
Solanum clade (Solanoideae:  Discopodium Hochst.      
 Solaneae) Jaltomata Schltdl.      
 Nothocestrum A. Gray      
 Solanum L. (incl. Cyphomandra Mart. ex Sendtn., Lycopersicon Mill., Normania Lowe, Triguera Cav.)    
Clades referred to in text and figures; names in parentheses are those fide Olmstead et al. (1999); Fay et al. (1998) Genus Capsule Berry Berry with stone cells Non‐capsular dehiscent Drupe or pyrene Mericarps (nutlets) 
Schwenkia clade (Schwenkioideae) Schwenkia L.      
 Protoschwenkia Soleredera      
 Melananthus Walp.      
 Heteranthia Nees and Mart.      
Schizanthus (Schizanthoideae) Schizanthus Ruiz and Pav.      
Duckeodendron (Duckeodendreae) Duckeodendron Kuhlm.      
Goetzea clade (Goetzeoideae) Goetzea Wydl.      
 Espadaea A.Rich.      
 Coeloneurum Radlk.      
 Henoonia Griseb.      
Petunia clade (Petunioideae) Benthamiella Speg.      
 Bouchetia Dunal      
 Brunfelsia L.      
 Calibrachoa La Llave and Lex.      
 Combera Sandw.      
 Fabiana Ruiz and Pav.      
 Hunzikeria D’Arcy      
 Latua Phil.b      
 Leptoglossis Benth.      
 Nierembergia Ruiz and Pav.      
 Pantacantha Speg.      
 Petunia Juss.      
 Plowmania Hunz. and Subils      
Salpiglossis clade Salpiglossis Ruiz and Pav.      
 (Cestroideae: Salpiglossideae) Reyesia Gay      
Browallia clade (Cestroideae: Browallia L.      
 Browallieae) Streptosolen Miers      
Cestrum clade (Cestroideae:  Cestrum L.      
 Cestreae) Sessea Ruiz and Pav.      
 Metternichia Mikanc      
 Vestia Willd.      
Anthocercis clade  Anthocercis Labill.      
 (Nicotianoideae: Anthocercideae) Anthotroche Endl.      
 Crenidium Haegi      
 Cyphanthera Miers      
 Duboisia R. Br.      
 Grammosolen Haegi      
 Symonanthus Haegi      
Nicotiana (Nicotianoideae:  Nicotianeae) Nicotiana L.      
Capsicum clade (Solanoideae:  Capsicum L.     
 Capsiceae) Lycianthes (Dunal) Hassl.    
Datura clade (Solanoideae: Datureae) Datura L.      
 Brugmansia Pers.      
Hyoscyamus clade  Anisodus Link      
 (Solanoideae: Hyoscyameae) Atropa L.      
 Atropanthe Pascher      
 Hyoscyamus L.      
 Physochlaina G. Don      
 Przewalskia Maxim.      
 Scopolia Jacq.      
Jaborosa (Solanoideae: Jaboroseae) Jaborosa Juss.      
Solandra clade (Solanoideae:  Dyssochroma Miers      
 Solandreae) Juanulloa Ruiz and Pav.      
 Markea A. Rich.      
 Merinthopodium Donn. Sm.      
 Schultesianthus Hunz.      
 Solandra Sw.      
 Trianaea Planch. and Linden      
Lycium clade (Solanoideae: Lycieae) Grabowskia Schltdl.      
 Lycium L.     
 Phrodus Miers     
Mandragora (Solanoideae:  Mandragoreae) Mandragora L.      
Nicandra clade (Solanoideae:  Exodeconus Raf.     
 Nicandreae) Nicandra Adans.     
Nolana (Solanoideae: Nolaneae) Nolana L. (incl. Alona     
Iochroma clade (Solanoideae:  Acnistus Schott     
 Physaleae: Iochrominae) Dunalia Kunth     
 Iochroma Benth.     
 Saracha Ruiz and Pav.     
 Vassobia Rusby     
Physalis clade (Solanoideae:  Brachistus Miers     
 Physaleae: Physalinae) Chamaesaracha (A. Gray) Benth.      
 Leucophysalis Rydb.     
 Oryctes S. Watson      
 Physalis L. (incl. Quincula Raf.)     
 Witheringia L’Her.      
Salpichroa clade (Solanoideae:  Nectouxia Kunth      
 Physaleae: Salpichroinae) Salpichroa Miers      
Withania clade (Solanoideae:  Athenaea Sendt.      
 Physaleae: Withaninae) Aureliana Sendt.     
 Cuatresia Hunz.      
 Darcyanthus Hunz.d     
 Deprea Raf.     
 Eriolarynx (Hunz.) Hunz.e     
 Larnax Miers     
 Physalistrum Makino      
 Tubocapsicum (Wetts.) Makino      
 Withania Pauq.(incl. Mellissia Hook. f.)     
Solanum clade (Solanoideae:  Discopodium Hochst.      
 Solaneae) Jaltomata Schltdl.      
 Nothocestrum A. Gray      
 Solanum L. (incl. Cyphomandra Mart. ex Sendtn., Lycopersicon Mill., Normania Lowe, Triguera Cav.)    

aProtoschwenkia may be more closely related to members of the Cestrum and Browallia clades (RG Olmstead, personal communication).

bLatua has been excluded from the character mapping analysis, as I am extremely doubtful as to its inclusion in the Petunia clade based on both flower morphology and chemistry. The relationships of Latua are of great interest, not only due to its use as a hallucinogen (Plowman et al., 1971), but due to its enigmatic status (Cosa Gastiazoro, 1991).

cMetternichia may be related to Duckeodendron and members of the Goetzea clade (Fay et al., 1998; RG Olmstead, personal communication).

dDarcyanthus was previously considered a species of Physalis (P. spruceana Hunz.; see Hunziker, 2000a).

eEriolaryx was previously part of Vassobia (section Eriolaryx Hunz.; see Hunziker, 2000b).

Table 3

. Monophyletic clades in Solanum as defined by Bohs (2002) with the distribution of fruit types in each See text for discussion of fruit types.

Clade No. of species, distribution Fruit types Berry colour 
Archaesolanum clade c. 8 spp, Australia, New Guinea, New Zealand Fleshy berries with copious stone cells Red or purple 
Normania clade c. 3 spp, Macaronesia, Mediterranean Fleshy berries with no stone cells Bright red 
African non‐spiny clade c. 7 spp, Africa Fleshy berries with no stone cells Red 
Potato clade (incl. Solanum tuberosum L., S. lycopersicum L.) c. 200–250 spp, New World Hard or fleshy berries with no stone cells Mainly green, brightly coloured only in the tomatoes 
Morelloid/dulcamaroid clade (incl. Solanum nigrum L., S. duclamara L.) c. 130 spp, worldwide Morelloids: fleshy berries with stone cells Morelloids: brightly coloured, black red or orange 
  Dulcamaroids: fleshy berries without stone cells Dulcamaroids: bright red, black 
Wendlandii/Allophyllum clade c. 10 spp, New World Hard and fleshy berries Green or yellowish 
Cyphomandra clade (incl. Solanum betaceumc. 50 spp, New World Fleshy berries with copious stone cells Red, yellow or green 
Geminata clade c. 140–150 spp, New World (mostly) Mostly hard berries, a few fleshy berries, all without stone cells Green or yellowish, occasionally red or orange 
Brevantherum clade c. 60 spp, New World Soft fleshy berries, with stone cells Green or yellowish, occasionally orange 
Lepstemonum clade (the spiny solanums, incl. Solanum melongena L.) c. 450 spp, worldwide Soft fleshy berries, hard berries, all without stone cells; variously modified non‐capsular dehiscent fruits (censer fruits, tumbleweeds) Green or yellowish, orange or red; dry fruits brownish 
Clade No. of species, distribution Fruit types Berry colour 
Archaesolanum clade c. 8 spp, Australia, New Guinea, New Zealand Fleshy berries with copious stone cells Red or purple 
Normania clade c. 3 spp, Macaronesia, Mediterranean Fleshy berries with no stone cells Bright red 
African non‐spiny clade c. 7 spp, Africa Fleshy berries with no stone cells Red 
Potato clade (incl. Solanum tuberosum L., S. lycopersicum L.) c. 200–250 spp, New World Hard or fleshy berries with no stone cells Mainly green, brightly coloured only in the tomatoes 
Morelloid/dulcamaroid clade (incl. Solanum nigrum L., S. duclamara L.) c. 130 spp, worldwide Morelloids: fleshy berries with stone cells Morelloids: brightly coloured, black red or orange 
  Dulcamaroids: fleshy berries without stone cells Dulcamaroids: bright red, black 
Wendlandii/Allophyllum clade c. 10 spp, New World Hard and fleshy berries Green or yellowish 
Cyphomandra clade (incl. Solanum betaceumc. 50 spp, New World Fleshy berries with copious stone cells Red, yellow or green 
Geminata clade c. 140–150 spp, New World (mostly) Mostly hard berries, a few fleshy berries, all without stone cells Green or yellowish, occasionally red or orange 
Brevantherum clade c. 60 spp, New World Soft fleshy berries, with stone cells Green or yellowish, occasionally orange 
Lepstemonum clade (the spiny solanums, incl. Solanum melongena L.) c. 450 spp, worldwide Soft fleshy berries, hard berries, all without stone cells; variously modified non‐capsular dehiscent fruits (censer fruits, tumbleweeds) Green or yellowish, orange or red; dry fruits brownish 

References

AloniB, Pressman E, Karni L.
1999
. The effect of fruit load, defoliation and night temperature on the morphology of pepper flowers and on fruit shape.
Annals of Botany
 
83
,
529
–534.
AlvárezA.
1996
. Systematics of Saracha (Solanaceae). MSc thesis, University of Missouri, St Louis.
AxeliusB.
1996
. The phylogenetic relationships of the physaloid genera (Solanaceae) based on morphological data.
American Journal of Botany
 
83
,
118
–124.
BaehniC.
1943
. Henoonia, type d’une famille nouvelle? Boissiera
7
, 346.
BarnesRW, Rozefelds AC.
2000
. Comparative morphology of Anodopetalum (Cunoniaceae).
Australian Systematic Botany
 
13
,
2667
–282.
BeckerA, Winter K‐U, Meyer B, Saedler H, Theissen G.
2000
. MADS‐box gene diversity in seed plants 300 million years ago.
Molecular Biology and Evolution
 
17
,
1425
–1434.
Benítez de RojasCE, D’Arcy WG.
1998
. The genera Cestrum and Sessea (Solanaceae) in Venezuela.
Annals of the Missouri Botanical Garden
 
85
,
273
–351.
BernardelloLM.
1983
. Estudios en Lycium (Solanaceae ). III. Estructura y desarrollo de fruto y semilla en Lycium y Grabowskia.
Boletin de la Sociedad Argentina de Botánica
 
22
,
147
–176.
BernardelloLM.
1986
. Revision taxonómica de las especies sudamericanas de Lycium (Solanaceae).
Boletin de la Academia Nacional de Ciencias, Cordoba
 
57
,
173
–356.
BitterG.
1911
. Steinzellkonkretionen im Fruchtfleisch beerentragender Solanaceen und deren systematische Bedeutung.
Botanisches Jahrbucher
 
45
,
483
–507.
BitterG.
1914
. Weitere Untersuchungen über das Vorkommen von Steinzellkonkretionen im Fruchtfleisch beerentragender Solanaceen.
Abhandlungen Naturwissenschaften Vereine Bremen
 
23
,
114
–163.
BohsL.
1994
. Cyphomandra (Solanaceae).
Flora Neotropica
 
63
,
1
–175.
BohsL.
1995
. Transfer of Cyphomandra (Solanaceae) and its species to Solanum.
Taxon
 
44
,
583
–587.
BohsL.
2001
. A revision of Solanum section Cyphomandropsis (Solanaceae).
Systematic Botany
 
Monographs
 
61
,
1
–85.
BohsL.
2002
. Major clades in Solanum based on ndhF sequence analysis. Monographs in Systematic Botany from the Missouri Botanical Garden (in press).
BohsL, Olmstead RG.
1997
. Phylogenetic relationships in Solanum (Solanaceae) based on ndhF sequences.
Systematic Botany
 
22
,
5
–17.
BohsL, Olmstead RG.
1999
. Solanum phylogeny inferred from chloroplast DNA sequence data. In: Nee M, Symon DE, Lester RN, Jessop JP, eds. Solanaceae IV. Richmond, Surrey: Royal Botanic Gardens, Kew,
97
–110.
BohsL, Olmstead RG.
2001
. A reassessment of Normania and Triguera (Solanaceae).
Plant Systematics and Evolution
 
228
,
33
–48.
BradfordJC, Barnes RW.
2001
. Phylogenetics and classification of Cunoniaceae (Oxalidales) using chloroplast DNA sequences and morphology.
Systematic Botany
 
26
,
354
–385.
BremerB, Eriksson O.
1992
. Evolution of fruit characters and dispersal modes in the tropical family Rubiaceae.
Biological Journal of the Linnean Society
 
47
,
79
–95.
BremerB, Andreasen K, Olsson D.
1995
. Subfamilial and tribal relationships in the Rubiaceae based on the rbcL sequence data.
Annals of the Missouri Botanical Garden
 
82
,
383
–397.
BremerK, Backlund A, Sennblad B, Swenson U, Andreason K, Hjertson M, Lundberg J, Backlund M, Bremer B.
2001
. A phylogenetic analysis of 100+ genera and 50+ families of euasterids based on morphological and molecular data with notes on possible higher level morphological synapomorpies.
Plant Systematics and Evolution
 
229
,
137
–189.
BruneauA, Dickson EE, Knapp S.
1995
. Congruence of chloroplast DNA restriction site characters with morphological and isozyme data in Solanum section Lasiocarpa.
Canadian Journal of Botany
 
73
,
1151
–1167.
BrunoGB.
1994
. Organización y vasculatura del gineceo de Nolana crassulifolia y N. rostrata.
Boletin de la Sociedad Argentina de Botánica
 
30
,
51
–57.
CarleR.
1981
. Investigations on the content of steroidal alkaloids and sapogenins within Solanum sect. Solanum (= sect. Morella) (Solanaceae).
Plant Systematics and Evolution
 
138
,
61
–71.
CarlquistS.
1988
. Wood anatomy and relationships of Duckeodendraceae and Goetzeaceae.
IAWA Bulletin
 
9
,
3
–12.
CipolliniML, Levey DJ.
1997
a. Why are some fruits toxic? Glycoalkaloids in Solanum and fruit choice by vertebrates.
Ecology
 
78
,
782
–798.
CipolliniML, Levey DJ.
1997
b. Antifungal activity of Solanum fruit glycoalkaloids: implications for frugivory and seed dispersal.
Ecology
 
78
,
799
–809.
CipolliniML, Levey DJ.
1997
c. Secondary metabolites of fleshy vertebrate‐dispersed fruits: adaptative hypotheses and implications for seed dispersal.
American Naturalist
 
150
,
346
–372.
CipolliniML, Bohs LA, Mink K, Paulk E, Böhning‐Gaese K.
2002
. Secondary metabolites of ripe fleshy fruits: ecology and phylogeny in the genus Solanum. In: Levey DJ, Silva WR, Galetti M, eds. Seed dispersal and frugivory: ecology, evolution and conservation. London: CABI Publishing (in press).
ClausingG, Meyer K, Renner SS.
2000
. Correlations among fruit traits and evolution of different fruits within Melastomataceae.
Botanical Journal of the Linnean Society
 
133
,
303
–326.
Contreras‐PadillaM, Yahia EM.
1998
. Changes in capsaicinoids during development, maturation, and senescence of chile peppers and relation with perioxidase activity.
Journal of Agricultural and Food Chemistry
 
46
,
2075
–2079.
CollinsonME, Boulter MC, Holmes PL.
1993
. Magnoliophyta (‘Angiospermae’). In: Benton MJ, ed. The fossil record Vol. 2. London: Chapman and Hall,
809
–841.
CosaGastiazoroMT.
1991
. Estudio morfoanatómico de órganos vegetativos en Cestroideae (Solanaceae). I. Nicotianeae.
Kurtziana
 
21
,
111
–152.
CourtenayO.
1994
. Conservation of the Maned Wolf: fruitful relationships in a changing environment.
Canid News
 
2
,
1
–5.
CronquistA.
1981
. An integrated system of classification of flowering plants. New York: Columbia University Press.
D’AoustM‐A, Yelle S, Nguyen‐Quoc B.
1999
. Antisense inhibition of tomato fruit sucrose synthase descreases fruit setting and the sucrose unloading capacity of young fruit.
The Plant Cell
 
11
,
2407
–2418.
D’ArcyWG.
1972
. Solanaceae studies. II. Typification of subdivisions of Solanum.
Annals of the Missouri Botanical Garden
 
59
,
262
–278.
D’ArcyWG.
1991
. The Solanaceae since
1976
, with a review of its biogeography. In: Hawkes JG, Lester RN, Nee M, Estrada R. N, eds. Solanaceae III: taxonomy, chemistry, evolution. Kew, Richmond, Surrey: Royal Botanical Gardens, Kew,
75
–137.
DavisCC, Anderson WR, Donoghue MJ.
2001
. Phylogeny of Malpighiaceae: evidence from chloroplast ndhF and trnl‐F nucleotide sequences.
American Journal of Botany
 
88
,
1830
–1846.
Di FulvioTE.
1971
. Morfología floral de Nolana paradoxa (Nolanaceae), con especial referencia a la organización del gineceo.
Kurtziana
 
6
,
41
–51.
DunalMF.
1852
. Solanaceae. In: DeCandolle, AP, ed. Prodromus systematis naturalis regni vegetabilis, Vol. 13. Paris: Victor Masson,
1
–690.
ErikssonO, Friis EM, Löfgren P.
2000
. Seed size, fruit size, and dispersal systems in angiosperms from the early Cretaceous to the late Tertiary.
American Naturalist
 
156
,
47
–58.
EstradaE, Martínez M.
1999
. Physalis L. (Solanaceae: Solaneae) and allied genera. I. A morphology‐based cladistic analysis. In: Nee M, Symon DE, Lester RN, Jessop JP, eds. Solanaceae IV. Richmond, Surrey: Royal Botanic Gardens, Kew,
139
–159.
EvansRC, Dickinson TA.
1999
. Floral ontogeny and morphology in subfamily Amygdaloideae T. and G. (Rosaceae).
International Journal of Plant Sciences
 
160
,
955
–979.
FayMF, Olmstead RG, Richardson JE, Santiago E, Prance GT, Chase MW.
1998
. Molecular data support the inclusion of Duckeodendron cestroides in Solanacaeae.
Kew Bulletin
 
53
,
203
–212.
FerrándizC, Pelaz S, Yanofsky MF.
1999
. Control of carpel and fruit development in Arabidopsis.
Annual Review of Biochemistry
 
68
,
321
–354.
FerrándizC, Liljegren SJ, Yanofsky MF.
2000
. Negative regulation of the SHATTERPROOF genes by FRUITFULL during Arabidopsis fruit development.
Science
 
289
,
436
–438.
FranceyP.
1935
. Monographie du genre Cestrum.
Candollea
 
6
,
46
–398;
Candollea
 
7
,
1
–132.
FraryA, Nesbitt TC, Frary A, Grandillo S, van der Knaap E, Cong B, Liu J, Meller J, Elber R, Alpert KB, Tanksley SD.
2000
. fw2.2: a quantitative trait locus key to the evolution of tomato fruit size.
Science
 
289
,
85
–88.
FriisEM, Crepet WL.
1987
. Time of appearance of floral features. In: Friis EM, Chaloner WG, Crane PR, eds. The origins of angiosperms and their biological consequences. Cambridge: Cambridge University Press,
145
–179.
FriisEM, Chaloner WG, Crane PR.(eds)
1987
. The origins of angiosperms and their biological consequences. Cambridge: Cambridge University Press.
GautierH, Guichard S, Tchamitchian M.
2001
. Modulation of competition between fruits and leaves by flower pruning and water fogging, and consequences on tomato leaf and fruit growth.
Annals of Botany
 
88
,
645
–652.
GiovannoniJ.
2001
. Molecular biology of fruit maturation and ripening.
Annual Review of Plant Physiology and Plant Molecular Biology
 
52
,
725
–749.
GoodspeedTH.
1954
. The genus Nicotiana.
Chronica Botanica
 
16
,
1
–536.
GrandilloS, Ku H‐M, Tanksley SD.
1996
. Characterization of fs8.1, a major QTL influencing fruit shape in tomato.
Molecular Breeding
 
2
,
251
–260.
GuQ, Ferrandiz C, Yanofsky MF, Martienssen R.
1998
. The FRUITFULL MADS‐box gene mediates cell differentiation during Arabidopsis fruit development.
Development
 
125
,
1509
–1517.
HennigW.
1966
. Phylogenetic systematics (translated from the German). Urbana IL: University of Illinois Press.
HerreraCM.
1985
. Determinants of plant‐animal co‐evolution: the case of mutualistic dispersal of seeds by vertebrates.
Oikos
 
44
,
132
–144.
HerreraCM.
1992
. Interspecific variation in fruit shape: allometry, phylogeny and adaptation to dispersal agents.
Ecology
 
73
,
1832
–1841.
HoareA, Knapp S.
1997
. A phylogenetic conspectus of the tribe Hyoscyameae: Solanaceae.
Bulletin of the Natural History Museum, London (Botany)
 
27
,
11
–29.
HootSB.
1995
. Phylogeny of Ranunculaceae based on preliminary atpB, rbcL and 18S nuclear ribosomal sequence data. In: Jensen U, Kadereit J, eds. The systematics and evolution of the Ranunculiflorae. Vienna: Springer Verlag,
241
–251.
HoweHF.
1984
. Constraints on the evolution of mutualism.
American Naturalist
 
123
,
764
–777.
HunzikerAT.
2000
a. Darcyanthus nom. nov. substitutes Darcya (Solanaceae).
Boletin de la Sociedad Argentina de Botánica
 
35
, 345.
HunzikerAT.
2000
b. Two novelties of the tribe Solaneae (Solanaceae).
Kurtziana
 
28
,
65
–68.
HunzikerAT.
2001
. The genera of Solanaceae. Ruggell, Germany: ARG Gantner Verlag KG.
JansonCH.
1983
. Adaptation of fruit morphology to dispersal agents in a neotropical forest.
Science
 
219
,
187
–1889.
JohnsonLAS, Briggs BG.
1994
. Myrtales and Myrtaceae—a phylogenetic analysis.
Annals of the Missouri Botanical Garden
 
71
,
700
–756.
KitchingI, Forey PL, Humphries CJ, Williams DM.
1998
. Cladistics: the theory and practice of parsimony analysis, 2nd edn. Oxford, NewYork: Oxford University Press.
KlannEM, Hall B, Bennett AB.
1996
. Antisense acid invertase (TIV1) gene alters soluble sugar composition and size in transgenic tomato fruit.
Plant Physiology
 
112
,
1321
–1330.
KnappS.
1989
. A revision of the Solanum nitidum species group (section Holophylla pro parte: Solanaceae).
Bulletin of the British Museum, Natural History, Botany
 
19
,
63
–102.
KnappS.
2001
. Is morphology dead in Solanum taxonomy? In: van der Werden G, Barendse G, van den Berg R, eds. Solanaceae V. University of Nijmegen; Nijmegen, The Netherlands,
23
–38.
KnappS.
2002
a. Floral diversity and evolution in the Solanaceae. In: Cronk QCB, Bateman RM, Hawkins JA, eds. Developmental genetics and plant evolution. Taylor and Francis, London,
267
–297.
KnappS.
2002
b. Solanum section Geminata.
Flora Neotropica
 
84
,
1
–595.
KnappS, Helgason T.
1997
. A revision of Solanum section Pteroidea: Solanaceae.
Bulletin of the Natural History Museum, London (Botany)
 
27
,
31
–73.
KnappS, Persson V, Blackmore S.
1998
. A phylogenetic conspectus of the tribe Juanulloeae.
Annals of the Missouri Botanical Garden
 
84
,
67
–89.
KosugeS, Furata M.
1970
. Studies on the pungent principle of Capsicum. Part XIV. Chemical constitution of the pungent principle.
Agricultural and Biological Chemistry
 
34
,
248
–256.
KuhlmannJG.
1947
. Duckeodendraceae Kuhlmann (nova familia).
Arquivos do Servico Florestal
 
3
,
7
–8.
LesterRN, Symon DE.
1989
. A Mexican Solanum with splash‐cup or censer fruits.
Solanaceae Newsletter
 
3
,
72
–73.
LiljegrenSJ, Ditta GS, Eshed Y, Savidge B, Bowman JL, Yanofsky MF.
2000
. SHATTERPROOF MADS‐box genes control seed dispersal in Arabidopsis.
Nature
 
404
,
766
–770.
LippmanZ, Tanksley SD.
2001
. Dissecting the genetic pathway to extreme fruit size in tomato using a cross between the small‐fruited wild species Lycopersicon pimpinellifolium and L. esculentum var. Giant Heirloom.
Genetics
 
158
,
413
–422.
MadisonWP, Madison DR.
1996
. MacClade: analysis of phylogeny and character evolution (version 3.06). Sunderland, MA: Sinauer Associates.
MagallónS, Crane PR, Herendeen PS.
1999
. Phylogenetic pattern, diversity, and diversification of eudicots.
Annals of the Missouri Botanical Garden
 
86
,
297
–372.
MagallónS, Sanderson MJ.
2001
. Absolute diversification rates in angiosperm clades.
Evolution
 
55
,
1762
–1780.
MarshallJA, Knapp S, Davey MR, Power JB, Cocking EC, Bennett MD, Cox AV.
2001
. Molecular systematics of Solanum section Lycopersicum (Lycopersicon) using the nuclear ITS DNA region.
Theoretical and Applied Genetics
 
103
,
1216
–1222.
MesaA.
1981
. Nolanaceae.
Flora Neotropica
 
26
,
1
–197.
MillerJS.
2002
. Phylogenetic relationships and the evolution of gender dimorphism in Lycium (Solanaceae).
Systematic Botany
 
27
,
416
–428.
MioneT, Olmstead RG, Jansen RK, Anderson GJ.
1994
. Systematic implications of chloroplast DNA variation in Jaltomata and selected physaloid genera (Solanaceae).
American Journal of Botany
 
81
,
912
–918.
McKeyD.
1975
. The ecology of coevolved seed dispersal systems. In: Gilbert LE, Raven PH, eds. Coevolution of animals and plants. Austin TX: University of Texas Press,
159
–191.
MorganDR, Soltis DE, Robertson KR.
1994
. Systematic and evolutionary implications of rbcL sequence variation in Rosaceae.
American Journal of Botany
 
81
,
890
–903.
MurrayKG, Kinsman S, Bronstein JL.
2000
. Plant–animal interactions. In: Nadkarni NM, Wheelwright NT, eds. Monteverde: ecology and conservation of a tropical cloud forest. New York, Oxford: Oxford University Press,
245
–267.
NeeM.
1986
. Placentation patterns in the Solanaceae In: D’Arcy WG, ed. Solanaceae: biology and systematics. New York: Columbia University Press,
169
–175.
NgM, Yanofsky MF.
2001
. Function and evolution of the plant MADS‐box gene family.
Nature Reviews Genetics
 
2
,
186
–195.
OlmsteadRG, Palmer JD.
1997
. Solanum: implications for phylogeny, classification and biogeography from cpDNA variation.
Systematic Botany
 
22
,
19
–29.
OlmsteadRG, Sweere JA, Spangler RE, Bohs L, Palmer JD.
1999
. Phylogeny and provisional classification of the Solanaceae based on chloroplast DNA. In: Nee M, Symon DE, Lester RN, Jessop JP, eds. Solanaceae IV. Richmond, Surrey: Royal Botanic Gardens, Kew,
111
–137.
PassamHC, Baltas C, Boyiatzoglou A, Khah EM.
2001
. Flower morphology and number of aubergine (Solanum melongena L.) in relation to fruit load and auzin application.
Scientia Horticulturae
 
89
,
309
–316.
PeraltaIE, Spooner DM.
2001
. Granule‐bound starch synthase (GBSSI) gene phylogeny of wild tomatoes (Solanum L. section Lycopersicon (Mill. Wettst.) subsection Lycopersicon).
American Journal of Botany
 
88
,
1888
–1902.
PerssonV, Knapp S, Blackmore S.
1999
. Pollen morphology and the phylogenetic analysis of Datura and Brugmansia. In: Nee M, Symon DE, Lester RN, Jessop JP, eds. Solanaceae IV. Richmond, Surrey: Royal Botanic Gardens, Kew,
171
–187.
PlowmanT, Gyllenhaal LO,Lindgren JE.
1971
. Latua pubiflora, magic plant from southern Chile.
Botanical Museum Leaflets
 
23
,
610
–92.
PlunkettGM, Soltis DE, Soltis PS.
1996
. Evolutionary patterns in the Apiaceae: inferences based on matK sequence data.
Systematic Botany
 
21
,
177
–195.
ReichmannJL, Meyerowitz EM.
1997
. MADS domains proteins in plant development.
Biological Chemistry
 
378
,
1079
–1101.
RohwerJ.
1996
. Die Frucht‐ und Samenstrukturen der Oleaceae—eine vergleichend‐anatomicsche Untersuchung.
Biblioteca Botanica
 
148
,
1
–177.
RothI.
1977
. Fruits of angiosperms. Berlin: Gebrüder Borntraeger.
SavolainenV, Fay MF, Albach DC, et al.
2000
. Phylogeny of the eudicots: a nearly complete familial analysis based on rbcL gene sequences.
Kew Bulletin
 
55
,
257
–309.
SchemskeDW.
1983
. Limits to specialization and co‐evolution in plant–animal mutualisms. In: Nitecki M, ed. Coevolution. Chicago IL: University of Chicago Press,
67
–109.
SchuppEW.
1992
. Quantity, quality, and the effectiveness of seed dispersal by animals.
Vegetatio
 
107
,
15
–30.
Schwarz‐SommerZ, Huijser P, Nacken W, Saedler H, Sommer H.
1990
. Genetic control of flower development by homeotic genes in Antirrhinum majus.
Science
 
250
,
931
–936.
SnijmanDA, Linder HP.
1996
. Phylogenetic relationships, seed characters, and dispersal system evolution in Amaryllideae (Amaryllidaceae).
Annals of the Missouri Botanical Garden
 
83
,
362
–386.
SoltisDE, Soltis PS, Doyle JJ.(eds)
1998
. Molecular systematics of plants. II. DNA sequencing. Boston, MA: Kluwer Academic Publishing.
SpalikK, Wojewodzka A, Downie SR.
2001
. The evolution of fruit in Scandiceae subtribe Scandicinae (Apiaceae).
Canadian Journal of Botany
 
79
,
1358
–1374.
SpoonerDS, Anderson GJ, Jansen RK.
1993
. Chloroplast DNA evidence for the interrelationships of tomatoes, potatoes and pepinos (Solanaceae).
American Journal of Botany
 
80
,
676
–688.
SpujtRW.
1994
. A systematic treatment of fruit types.
Memoirs of the New York Botanical Garden
 
70
,
1
–181.
SymonDE.
1979
. Fruit diversity and dispersal in Solanum in Australia.
Journal of the Adelaide Botanical Garden
 
1
,
321
–331.
SymonDE.
1984
. A new form of Solanum fruit.
Journal of the Adelaide Botanical Garden
 
7
,
123
–126.
SymonDE.
1987
. Placentation patterns and seed numbers in Solanum (Solanaceae) fruits.
Journal of the Adelaide Botanical Garden
 
10
,
179
–199.
SymonDE.
1994
. Kangaroo apples: Solanum sect. Archaeosolanum. Adelaide, Australia: Published by the author.
Tago‐NakazawaM, Dillon MO.
1999
. Biogeografía y evolución en el clado Nolana (Nolaneae—Solanaceae).
Arnaldoa
 
6
,
81
–116.
TewksburyJJ, Nabhan GP.
2001
. Directed deterrence by capsaicin in chillies.
Nature
 
412
,
403
–404.
TheissenG, Becker A, Di Rosa A, Kanno A, Kim JT, Münster T, Winter K‐U, Saedler H.
2000
. A short history of MADS‐box genes in plants.
Plant Molecular Biology
 
42
,
115
–149.
ThorneRF.
1992
. Classification and geography of flowering plants.
Botanical Review
 
58
,
225
–348.
TiffneyBH.
1984
. Seed size, dispersal syndromes, and the rise of the angiosperms: evidence and hypothesis.
Annals of the Missouri Botanical Garden
 
71
,
551
–576.
WagstaffSJ, Olmstead RG.
1997
. Phylogeny of Labiatae and Verbenaceae inferred from rbcL sequences.
Systematic Botany
 
22
,
165
–179.
WahajSA, Levey DJ, Sanders AK, Cipollini ML.
1998
. Control of gut retention time by secondary metabolites in ripe Solanum fruits.
Ecology
 
79
,
2309
–2319.
WeigelD, Meyerowitz EM.
1994
. The ABCs of floral homeotic genes.
Cell
 
78
,
203
–209.
WestobyM, Leishman M, Lord J.
1996
. Comparative ecology of seed size and dispersal.
Philosophical Transactions of the Royal Society
 
of London, Series B
 
351
,
1309
–1318.
WhalenMD.
1979
. Taxonomy of Solanum section Androceras.
Gentes Herbarum
 
11
,
359
–426.
WheelwrightNT, Orians GH.
1992
. Seed dispersal by animals: contrasts with pollen dispersal, problems of terminology, and constraints on coevolution.
American Naturalist
 
119
,
402
–413.
WikströmN, Savolainen V, Chase MW.
2001
. Evolution of the angiosperms: calibrating the family tree.
Philosophical Transactions of the Royal Society of London,
 
Series B
 
268
,
2211
–2220.
WileyEO.
1981
. Phylogenetics: the theory and practice of phylogenetic systematics. New York: John Wiley and Sons.
WingSL, Boucher LD.
1998
. Ecological aspects of the Cretaceous flowering plant radiation.
Annual Review of Earth and Planetary Sciences
 
26
,
379
–421.
XiangQ‐Y, Brunsfeld SJ, Soltis DE, Soltis PS.
1996
. Phylogenetic relationships in Cornus based on chloroplast DNA restriction sites: implications for biogeography and character evolution.
Systematic Botany
 
21
,
515
–534.

Comments

0 Comments