Abstract
Background and aims
Trichomes are epidermal outgrowths generally associated with protection against herbivores and/or desiccation that are widely distributed from ferns to angiosperms. Patterns of topological variation and morphological evolution of trichomes are still scarce in the literature, preventing valid comparisons across taxa. This study integrates detailed morphoanatomical data and the evolutionary history of the tribe Bignonieae (Bignoniaceae) in order to gain a better understanding of current diversity and evolution of trichome types.Methods
Two sampling schemes were used to characterize trichome types: (1) macromorphological characterization of all 105 species currently included in Bignonieae; and (2) micromorphological characterization of 16 selected species. Individual trichome morphotypes were coded as binary in each vegetative plant part, and trichome density and size were coded as multistate. Ancestral character state reconstructions were conducted using maximum likelihood (ML) assumptions.Key results
Two main functional trichome categories were found: non-glandular and glandular. In glandular trichomes, three morphotypes were recognized: peltate (Pg), stipitate (Sg) and patelliform/cupular (P/Cg) trichomes. Non-glandular trichomes were uniseriate, uni- or multicellular and simple or branched. Pg and P/Cg trichomes were multicellular and non-vascularized with three clearly distinct cell layers. Sg trichomes were multicellular, uniseriate and long-stalked. ML ancestral character state reconstructions suggested that the most recent common ancestor (MRCA) of Bignonieae probably had non-glandular, Pg and P/Cg trichomes, with each trichome type presenting alternative histories of appearance on the different plant parts. For example, the MRCA of Bignonieae probably had non-glandular trichomes on the stems, prophylls, petiole, petiolule and leaflet veins while P/Cg trichomes were restricted to leaflet blades. Sg trichomes were not present in the MRCA of Bignonieae independently of the position of these trichomes. These trichomes had at least eight independent origins in tribe.Conclusions
The patterns of trichome evolution indicate that most morphotypes are probably homologous in Bignonieae and could be treated under the same name based on its morphological similarity and common evolutionary history, in spite of the plethora of names that have been previously designated in the literature. The trichome descriptions presented here will facilitate comparisons across taxa, allowing inferences on the relationsthips between trichome variants and future studies about their functional properties.Free full text
Trichome structure and evolution in Neotropical lianas
Abstract
Background and Aims
Trichomes are epidermal outgrowths generally associated with protection against herbivores and/or desiccation that are widely distributed from ferns to angiosperms. Patterns of topological variation and morphological evolution of trichomes are still scarce in the literature, preventing valid comparisons across taxa. This study integrates detailed morphoanatomical data and the evolutionary history of the tribe Bignonieae (Bignoniaceae) in order to gain a better understanding of current diversity and evolution of trichome types.
Methods
Two sampling schemes were used to characterize trichome types: (1) macromorphological characterization of all 105 species currently included in Bignonieae; and (2) micromorphological characterization of 16 selected species. Individual trichome morphotypes were coded as binary in each vegetative plant part, and trichome density and size were coded as multistate. Ancestral character state reconstructions were conducted using maximum likelihood (ML) assumptions.
Key Results
Two main functional trichome categories were found: non-glandular and glandular. In glandular trichomes, three morphotypes were recognized: peltate (Pg), stipitate (Sg) and patelliform/cupular (P/Cg) trichomes. Non-glandular trichomes were uniseriate, uni- or multicellular and simple or branched. Pg and P/Cg trichomes were multicellular and non-vascularized with three clearly distinct cell layers. Sg trichomes were multicellular, uniseriate and long-stalked. ML ancestral character state reconstructions suggested that the most recent common ancestor (MRCA) of Bignonieae probably had non-glandular, Pg and P/Cg trichomes, with each trichome type presenting alternative histories of appearance on the different plant parts. For example, the MRCA of Bignonieae probably had non-glandular trichomes on the stems, prophylls, petiole, petiolule and leaflet veins while P/Cg trichomes were restricted to leaflet blades. Sg trichomes were not present in the MRCA of Bignonieae independently of the position of these trichomes. These trichomes had at least eight independent origins in tribe.
Conclusions
The patterns of trichome evolution indicate that most morphotypes are probably homologous in Bignonieae and could be treated under the same name based on its morphological similarity and common evolutionary history, in spite of the plethora of names that have been previously designated in the literature. The trichome descriptions presented here will facilitate comparisons across taxa, allowing inferences on the relationsthips between trichome variants and future studies about their functional properties.
INTRODUCTION
A good understanding of the evolutionary patterns of individual morphological features represents the basis of comparative biology (MacLeod and Forey, 2002). Such studies combine information derived from phylogenetic analyses with detailed morphological data to improve understanding of the patterns of morphological change (e.g. Fougère-Danezan et al., 2010). The identification of a common origin for individual morphological features represents a critical step for the establishment of homology hypotheses (Scotland and Pennington, 2000; West-Eberhard, 2003), which is essential for comparisons accross taxa (e.g. Fahn, 1986), and for the test of adaptive hypotheses (Larson and Losos, 1996).
Trichomes are epidermal outgrowths that are widely distributed in land plant groups (Levin, 1973; Payne, 1978). These structures are often associated with protection against herbivores and/or desiccation (Wagner et al., 2004), and are of extreme importance for the maintenance of various plant groups in herbivore-rich environments and dry areas. Current patterns of trichome variation are thought to have resulted from natural selection, with individuals that have trichomes showing greater advantages (e.g. leaf damage, loss of water or high temperature) when compared with individuals lacking those structures (e.g. Levin, 1973; Elle et al., 1999; Valverde et al., 2001; Romero et al., 2008; Johnson et al., 2009; Kaplan et al., 2009).
Directional selection (Valverde et al., 2001) and stabilizing selection (Elle et al., 1999) have often been used to explain the patterns of trichome variation encountered in natural populations. Such models have gained further support from genetic studies that have characterized DNA regions responsible for the patterns of trichome variation of model organisms [e.g. quantitative trait locus (QTL) studies by Mauricio, 2005; Symonds et al., 2005]. Such studies have allowed the use of trichomes in population-level quantitative evolutionary studies. Unfortunately, little is still known about the general morphology of trichomes and about their patterns of topological variation (Theobald et al., 1979), preventing valid comparisons across taxa. In addition, macroevolutionary patterns of trichome types based on robust phylogenetic analyses are still scarce (but see Beilstein et al., 2006; Chauveau et al., 2011) and little is still known regarding the mode, timing and pattern of evolution of trichome density in plants as a whole.
Trichomes in Bignoniaceae and tribe Bignonieae
Bignoniaceae exhibit notable variation in terms of trichome structure, position, abundance and size among species (e.g. Seibert 1948). Indeed, various studies have reported different trichome types in representatives of Bignoniaceae (see Fig. 1), and species exhibit a variety of glandular and non-glandular trichomes on different vegetative plant parts (e.g. Seibert, 1948) and reproductive organs (see Souza et al., 2010). However, detailed structural studies are still lacking for most of these trichomes. Though glandular trichomes have been reasonably well documented in some species, sampling is still incomplete in the family (Fig. 1).
Previous morphological studies of glandular and non-glandular trichomes in Bignoniaceae and the phylogenetic relationships described by Olmstead el al. (2009), in which the clades 2, 3, 4 and 5 are Tabebuia alliance, Palaeotropical clade, Tecomeae and Jacarandeae, respectively. The shaded area represents the anatomical studies in the tribe Bignonieae. All taxonomic names are shown in the Supplementary Data Table S1.
The lack of structural studies on different trichome types has prevented the establishment of hypotheses of common origin and/or meaningful comparisons among trichome morphotypes in Bignoniaceae. Furthermore, a wide array of terms have been used to describe the various types of glandular trichomes in Bignonieae (Fig. 1), complicating the distinction among trichome morphotypes and preventing evolutionary studies with trichomes in this group. For example, the patelliform glandular trichomes distributed across the leaf surface, petioles, prophylls of axillary buds, interpetiolar regions, calyces, corollas and fruits of representatives of the tribe have been called ‘nectaries’ or ‘extrafloral nectaries’ (e.g. Elias and Gelband, 1976; Oliveira and Leitão-Filho, 1987; Nogueira et al., 2012a), ‘scale-like trichomes as morphotype of EFNs’ (e.g. Díaz-Castelazo et al., 2005; Machado et al., 2008) and ‘glands’ (e.g. Seibert, 1948; Gentry, 1974, 1980; Laroche, 1974; Lohmann, 2006), making a standardization of the trichome terminology extremely necessary in this group.
Apart from showing great variation in trichome structure and morphology, Bignonieae represents the most abundant and diverse clade of Bignoniaceae and also of Neotropical lianas (Lohmann, 2006). The tribe includes 21 monophyletic genera and approx. 400 species (nearly half of all species of Bignoniaceae) (Lohmann and Taylor, 2013), 25 % of which have been sampled in a molecular phylogenetic analysis based on plastid (ndhF) and nuclear (PepC) markers (Lohmann, 2006). The wide variation found in trichome morphology, topology and other quantitative traits in representatives of Bignonieae and the robust phylogenetic tree available make this tribe an excellent group for addressing trichome evolution. A uniform trichome terminology in Bignonieae would greatly improve comparative taxonomic, ecological and evolutionary research in this morphologically diverse clade of lianas.
We reviewed the literature on the morphology and anatomy of trichomes in representatives across Bignoniaceae to understand better the diversity of trichomes of the focal group, tribe Bignonieae. Additionally, we characterized trichome types in representatives of all 21 genera currently recognized in Bignonieae to fill the data gaps currently encountered in the literature on the family. For those taxa, detailed anatomical studies, macromorphological descriptions and quantitative analyses of trichome size and density were carried out. This information was then used as a basis for understanding macroevolutionary patterns of non-glandular and glandular trichomes in Bignonieae using a robust molecular phylogenetic tree for the group (Lohmann, 2006). The following questions were addressed. (1) How are the trichome morphotypes distributed in vegetative parts of representatives of Bignonieae, what is their overall structure and how are they characterized morphologically? (2) Are the morphological variants of each trichome morphotype homologous, or has each trichome morphotype evolved independently during the phylogenetic history of the group? (3) What evolutionary patterns are evident for each trichome morphotype on different plant parts? (4) What is the evolutionary pattern of trichome size and density in Bignonieae?
MATERIALS AND METHODS
Trichome characterization
To obtain a clear picture of the patterns of morphological variation encountered in trichome types in representatives of Bignonieae, we first reviewed all currently available trichome studies of Bignoniaceae (Fig. 1). We then used two complementary sampling schemes to characterize the individual trichome types present in the vegetative portions of representatives of Bignonieae. For that, we used a stereomicroscope and a broad-scale sampling scheme to characterize the macromorphology of trichomes in the 105 species of Bignonieae included in the phylogenetic anlysis of the tribe. This first characterization resulted in the definition of the main trichome morphotypes in the tribe. We then conducted detailed anatomical studies using scanning emission microscopy (SEM) and light microscopy on a smaller number of taxa (16 species of Bignonieae), selected to characterize the micromorphology of all major trichome morphotypes. These two sampling schemes led to a detailed description of the overall morphology of the individual trichome types in Bignonieae.
Macromorphological characterization
We used herbarium specimens to characterize macromorphological patterns of variation of trichome types in representatives of Bignonieae. For that, three specimens of each of the 104 species of Bignonieae included in the combined molecular phylogenetic analysis of Lohmann (2006) plus Callichlamys latifolia and Perianthomega vellozoi were sampled. This sampling scheme included representatives of all 21 genera of Bignonieae and a sample of taxa across different locations and habitats. Herbarium specimens were analysed using a stereomicroscope (Olympus SZ60; zoom range of 6·3:1). The following variables were recorded for each trichome type: (a) presence/absence; (b) size; and (c) density on stems, prophylls of axillary buds, petioles, petiolules, and adaxial and abaxial leaf surfaces (basal, middle and upper portions). Trichome density was sampled in three squares of 1 mm2 each, located in the basal, middle and upper portions of leaflets. Larger trichomes such as the patelliform/cupular glandular trichomes were sampled in larger squares of 1 cm2. The number of patelliform/cupular glandular trichomes (abundance data) was also recorded for the interpetiolar region of stems and the prophylls of the axillary buds, petioles and petiolules.
Micromorphological characterization
We used fixed materials in a micromorphological characterization of trichome types in 16 species of Bignonieae. The following taxa were sampled: (1) Adenocalymma pedunculatum (Vell.) L.G.Lohmann; (2) Amphil-ophium crucigerum (L.) L.G.Lohmann; (3) Amphilophium parkeri (DC.) L.G.Lohmann; (4) Anemopaegma album Mart. ex DC.; (5) Anemopaegma scabriusculum Mart. ex DC.; (6) Bignonia prieurei DC.; (7) Cuspidaria sceptrum (Cham.) L.G.Lohmann; (8) Dolichandra unguis-cati (L.) L.G.Lohmann; (9) Fridericia triplinervia (Mart. ex DC.) L.G.Lohmann; (10) Lundia nitidula DC.; (11) Mansoa difficilis (Cham.) Bureau & K.Schum.; (12) Perianthomega vellozoi Bureau; (13) Pleonotoma albiflora (Salzm. ex DC.) A.H.Gentry; (14) Pyrostegia venusta (Ker Gawl.) Miers; (15) Stizophyllum riparium (Kunth) Sandwith; and (16) Tanaecium pyramidatum (Rich.) L.G.Lohmann. Field collections were made in areas of savanna (Parque Estadual de Grão Mogol, Minas Gerais State/Brazil and Parque Nacional da Chapada Diamantina, Bahia State/Brazil) and rain forest (Reserva Florestal Adolpho Ducke, Amazonas State/Brazil).
Mature and young vegetative plant parts were fixed in 50 % FAA (Johansen, 1940) and stored in 70 % ethanol. For light microscopy, the material was dehydrated in an ethanol series and then embedded in 2-hydroxyethylmethacrylate (Leica Microsystems Inc., Heidelberg, Germany). Longitudinal sections of 6 µm were cut with a rotary microtome and stained with 0·05 % toluidine blue (pH 4·3) (O'Brien et al., 1964). For SEM studies, the fixed material was dehydrated in an ethanol series and critical point dried. Samples were mounted on stubs and sputter-coated with gold. Observations were made with a Quanta 200 scanning electron microscope (Fei Company, Hillsboro, OR, USA).
Overall anatomical descriptions and trichome classification were based on Theobalde et al. (1979). Specific terminology used for the characterization of the patelliform/cupular glandular trichomes followed Elias (1983). Herbarium vouchers used in this study are listed in Supplementary Data, Specimens.
Phylogenetics and trichome evolution
Phylogenetic tree
A moderately to well-supported combined molecular phylogenetic analysis of Bignonieae [i.e. with 92 % of the nodes receiving ≥50 % maximum likelihood (ML) bootstrap support] based on plastid (ndhF) and nuclear (PepC) sequences is currently available for the group (Lohmann, 2006). The tree used here was the same as the combined molecular phylogenetic analysis of Lohmann (2006) except that Pyrostegia dichotoma was treated as a synonym of Pyrostegia venusta and excluded from the tree, and Callichlamys latifolia and Perianthomega vellozoi were included in order to represent the only two genera of Bignonieae that were absent in the combined analysis of Lohmann (2006). The phylogenetic position of Callichlamys and Perianthomega followed relationships recovered in the ndhF phylogenetic analysis of Lohmann (2006), which included a broader sampling of taxa than the combined molecular (ndhF plus PepC) analysis of Lohmann (2006). This phylogenetic tree was combined with fossil data to estimate the time of divergence of all nodes in Bignonieae (Lohmann et al., 2013), and the tree used here is the same as that used in the earlier study of evolution of extrafloral nectaries of Bignonieae (Nogueira et al., 2012b).
Character coding
Each of the four trichome morphotypes recorded for Bignonieae were coded as present/absent (binary coding) according to their occurrence on each plant part considered (stems, prophylls of axillary buds, petioles, petiolules, and adaxial and abaxial sides of leaflets). In addition, continuous measurements of trichome densities were transformed into discrete character states and coded as multistate characters. The total range of trichome density among the 105 species of Bignonieae was divided into three classes equally. Those three classes of the range were used as the three states of trichome density: (1) low; (2) intermediate; and (3) high. The same approach was applied to trichome sizes.
Ancestral character state reconstructions
We initially used information on trichome position, structure and ontogeny (gathered from the literature, Fig. 1) to generate hypotheses of homology for the various trichome types. We subsequently combined all the morphological information with data on the phylogenetic history of Bignonieae through ML ancestral character state reconstructions in order to verify whether the morphological similarity was based on common ancestry. The ML ancestral character state reconstructions of trichome morphotypes were implemented in Mesquite 2·74 (Maddison and Maddison, 2007). In all analyses, trichome characters were equally weighted and considered unordered. For binary traits, two evolutionary models were considered as applied by Ree and Donoghue (1999): an Mk1 model with one parameter, in which any particular change is considered equally likely, and an Mk2 model with two parameters that represent forward and backward rates of changes between character states. Models were chosen using the likelihood ratio test (LRT). The evolutionary history of multistate character states (resulting from the discretization of quantitative characters) was only reconstructed for trichomes for which ancestral character state reconstructions based on binary characters (coded as presence/absence) suggested that the trait being considered was already present in the most recent common ancestor (MRCA) of the tribe. These analyses considered the presence of each trichome type in each plant part separately. As a result, ancestral state reconstructions of quantitative characters were not conducted for homoplasious trichome types. These analyses were performed exclusively with the Mk1 model with one parameter. In addition, ML reconstructions were used to assess the rates of transition of trichome traits in all ancestral character reconstructions.
RESULTS
Morphological characterization of trichome types
Two functional categories of trichomes were observed on vegetative plant parts of representatives of Bignonieae: (1) non-glandular and (2) glandular. Both trichome categories had variable sizes and cell numbers. All variants of non-glandular trichomes were included under the same category due to the similarities of these trichomes. On the other hand, three distinct morphotypes of glandular trichomes were recognized. These four trichome morphotypes (Fig. 2) were recorded on vegetative plant parts of Bignonieae species: (1) non-glandular trichomes (including unicellular simple, multicellular simple and multicellular branched – ‘dentritic/stellar’ – Fig. 2A–D); (2) peltate glandular trichomes (Fig. 2B, D, E–G); (3) stipitate glandular trichomes (Fig. 2H); and (4) patelliform/cupular glandular trichomes (Fig. 2I–K).
SEM micrographs of trichome morphotypes distributed on vegetative plant parts of Bignonieae. (A–D) Non-glandular trichomes: (A) simple trichomes on the abaxial side of leaflets of Anemopaegma scabriusculum; (B) branched trichomes and peltate glandular trichomes on the leaflet vein of Amphilophium crucigerum; (C) simple trichomes on the leaflet vein of Tanaecium pyramidatum; (D) simple trichomes and peltate glandular trichomes (Pg) on the stems of Periathomega vellozoi. (E–G) Peltate glandular trichomes: (E) Anemopaegma album; (F) Amphilophium crucigerum; (G) Adenocalymma pedunculatum. (H) Stipitate glandular trichomes with adhesive secretion on the petiole of Cuspidaria sceptrum. (I) Patelliform/cupular glandular trichomes on the leaflets of Adenocalymma pedunculatum with one mature cupular trichome and two young trichomes (upper–central/left) on the abaxial side of leaflets. (J) Anemopaegma album with clusters of patelliform trichomes on the basal portion of the abaxial side of leaflets. (K) Perianthomega vellozoi with one patelliform trichome in the abaxial side of the leaflet blade. Pg, peltate glandular trichome. Scale bars: (A) = 1 mm; (B, D, K) = 100 µm; (C) = 150 µm; (E–H) = 50 µm; (I) = 250 µm; (J) = 200 µm.
Non-glandular (Ng) trichomes
This trichome morphotype was present in all 105 species and 21 genera of Bignonieae sampled (Fig. 3). The distribution of this morphotype varied in terms of the position, density and plant size. These trichomes were uni- to multicellular, uniseriate, simple or branched, and formed by 1–25 cells, characterizing short and long trichomes. Trichome cells were covered by a thick, smooth or rough cuticule. Cuticular warts were present in some cases (e.g. Anemopaegma). Approximately 90 % of the species sampled (95/105) had simple trichomes with one to multiple cells (Fig. 3A–L) with variable position on the plants among species; the remaining 10 % of the species had branched trichomes with multiple cells (Fig. 3M–R) also with variable position on the plants among species. Multicellular and branched trichomes were encountered in some representatives of six different genera (i.e. Adenocalymma, Amphilophium, Callichlamys, Fridericia, Manaosella and Pleonotoma), not characterizing any particular genus. This trichome type occurred more widely in Amphilophium (Fig. 3N, O).
Distribution and structure of non-glandular trichomes in Bignonieae. (A–L) Simple (unbranched) non-glandular trichomes on different plant parts. Unbranched trichomes were distributed on the abaxial blade of leaflets in Anemopaegma setilobum (A) and Cuspidaria sceptrum (B), and on the adaxial blade of Adenocalymma pubescens (C). Long trichomes of Amphilophium askersonii on petiole (D), leaflets (E) and stems (F). Arrows in C and E indicate one simple trichome. Photomicrograph of microtome sections of simple trichomes on the petiole of Adenocalymma pedunculatum (G); with two cells, and unicellular trichomes on the edge of leaflets of Perianthomega vellozoi (H). Simple trichomes on the leaflet vein and petiole + petiolule of Pyrostegia venusta (I) and Anemopaegma scabriusculum (J), respectively; and trichomes on the abaxial blade of leaflets in Fridericia lasiantha (K). Photomicrograph of a microtome section of simple multicellular trichomes of Cuspidaria sceptrum (L). (M–R) Branched non-glandular trichomes on different plant parts. Branched trichomes were distributed on the abaxial blade of leaflets in Fridericia cinnamomea (M), leaflet veins in Amphilophium frutescens (N, O), and with variable densities on the leaflets between Fridericia nigrescens (P) and Fridericia dispar (Q). Arrows in I and N indicate the leaflet vein; the arrow in Q indicates a branched trichome. Photomicrograph of microtome section of the branched multicellular trichomes on the leaflets of Amphilophium parkerii (R). Scale bars: (D, J) = 5 mm; (A–C, F, K, O, P, Q) = 0·25 mm; (G, H, L) = 50 µm; (M, N) = 2 mm; (I) = 1 mm; (E, R) = 0·5 mm.
Although Ng trichomes were encountered in all sampled species, their presence was highly variable on the different plant parts among species (topological/positional variation). Only in 39 % of the species (41 out of 105) were these trichomes encountered on all vegetative parts, although different trichome densities were observed. In 7·6 % of the species (8/105), trichomes were distributed across all plant parts examined except leaflets. Overall, 82 % of the species (86/105) had trichomes on both sides of leaflets. In plants with trichomes on the abaxial side of leaflets, 19·3 % of the species (18/93) had trichomes growing exclusively over the leaflet veins. In the remaining 80·7 % of the species (75/93), trichomes were found across the abaxial side of the leaflet blade (Fig. 3A, K, M). In plants with trichomes on the adaxial side of leaflets, 44·2 % of the species (38/86) had trichomes growing exclusively over the leaflet veins (Figs. 3I, N). Approximately one-third of all species sampled (35/105) had Ng trichomes that were restricted to individual plant parts. In general, these species had trichomes that were restricted to the leaflet margin (e.g. Periatomega vellozoi) and/or to the ridges of the interpetiolar region (e.g. Mansoa difficilis), and/or to the apex of the prophylls of the axillary buds (e.g. Lundia nitidula). In addition, 27·6 % of the species (29/105) had high trichome densities (>10 trichomes mm−2) on the abaxial surface of the leaflet blades. For species with Ng trichomes on both sides of the leaflets, the abaxial surface included approx. 16·5 times more trichomes per mm2 on average than the abaxial side (n = 63).
Peltate glandular (Pg) trichomes
This trichome morphotype was present in all 105 species and 21 genera of Bignonieae sampled (Fig. 4), with variable density and size among plants. Pg were multicellular, non-vascularized, stalked (a short stalk) and convex or rounded at earlier stages of development (sometimes becoming flattened with age). Each trichome had three distinct cell types. (1) Head cells: the head was composed of 6–24 thin-walled cells that were covered by a thin and smooth cuticle, but lacked visible pores. The trichome head also had a sub-cuticular space that accumulated secretions in many species (Fig. 4H2, K2). (2) Stalk cells: the stalk of trichomes was generally composed of a single cell in most species (Fig. 4F3, H3, K2) but included two or three cells in some taxa (e.g. Stizophyllum). These cells are rectangular (Fig. 4J4) or in the shape of an inverted trapezoid in most species (Fig. 4G3, H4). (iii) Foot cells. The foot was formed by two cells and generally is in the same plane as other epidermal cells. The Pg trichomes were similar among different species, with the number of head cells varying with development (Fig. 4F–M). Probably, in the post-secretory stage, the head of the Pg trichome disintegrates with cell death (more visible in Fig. 4G3, L3).
Structure and development of peltate glandular trichomes in Bignonieae. (A–E) Traditional ‘lepidote’ indumentum in the abaxial side of leaflets showing the typical trichomes (arrows) of Adenocalymma cymbalum (A), Amphilophium paniculatum (B), Anemopaegma chamberlaynii (C), Pleonotoma jasminifolia (D), Stizophyllum perforatum (E). (F–M) Developmental stages of peltate glandular trichomes on leaflets: Adenocalymma pedunculatum (F), Bignonia prieurei (G), Amphilophium crucigerum (H), Anemopaegma scabriusculum (I1–I3) and A. album (I4), Fridericia triplinervia (J; inset in J3, frontal view), Pleonotoma albiflora (K), Tanaecium pyramidatum (L) and Stizophyllum riparium (M). Abbreviations: h., head cells; s., stalk cells; f., foot cells; and sb., sub-cuticular space with secretion. Scale bars: (A, D) = 0·25 mm; (B, C, E) = 0·50 mm; (F–M) = 25 µm.
The Pg trichomes had a more homogeneous distribution across vegetative plant parts, and a higher variation in the density and size of trichomes. The diameter of the glandular head varied between 0·02 and 0·10 mm2 (median = 0·048, n = 103), with larger heads generally having more cells. Species of Pyrostegia and Stizophyllum had the largest glandular heads (up to 0·1 mm2), and species of Amphilophium had the highest trichome densities (up to 216 trichomes mm−2 in Amphilophium paniculatum). Trichomes of Pyrostegia, Amphilophium and Stizophyllum (14/105 species sampled) reflected light when observed under the microscope (Fig. 4B, E). In some cases, peltate glandular trichomes were difficult to visualize without magnification lenses either because of their small size (glandular head diameter <0·028 mm; e.g. Tynanthus), or because trichomes were encountered in low densities (0·6 trichomes mm−2; e.g. most species of Adenocalymma; Fig. 4A). In 42 % of the species sampled (44 out of 105), the leaflet blade had high or intermediate densities of Pg trichomes (>5 trichomes mm−2) (Fig. 4B, C, E). Similar to the patterns found for the Ng trichomes, the abaxial side of leaflets had 5·3 times more Pg trichomes by mm2 on average than the adaxial side (n = 97). Furthermore, Pg trichomes were generally more easily visualized on the abaxial side of the leaflet blades than on other vegetative plant parts (Fig. 4A–E). Trichomes were rarely observed on older portions of the interpetiolar regions, petioles and petiolules, suggesting that this trichome type is generally lost with age and/or secondary growth.
Stipitate glandular (Sg) trichomes
This trichome morphotype occurred only in nine of the 105 species sampled, and was encountered in six of the 21 genera of Bignonieae currently recognized (Fig. 5). In all species, the Sg trichomes had similar anatomical features that included a secretory head, and a multicellular, uniseriate, simple, non-vascularized stalk (generally long-stalked) with a variable number of cells among species. In general, the foot cells are at the same level and have the same morphology as other epidermal cells around the Sg trichome. (1) Head cells. A group of 8–12 thin-walled secretory cells. (2) Stalk cells. Include 3–12 cells, with the relative length of each stalk cells being greater than its width (2–3 times longer in mature trichomes), differentiating this trichome type from the Pg trichomes (Fig. 5E).
Structure and development of stipitate glandular trichomes in Bignonieae (arrows). (A–E) Developmental stages of stipitate glandular trichomes of Cuspidaria sceptrum. (F–K) Glandular trichomes on the petiole of Adenocalymma adenophorum (F), leaflets of Fridericia erubescens (G), petioles of Manaosella cordifolia (H), stems of Adenocalymma trichocladum (I), in young leaves near the inflorescence of Mansoa hirsuta (J) and in petioles of Martinella obovata (K). Abbreviations: h., head cells; s., stalk cells; f., foot cells; Scale bars: (A–E) = 50 µm; (F, G, I, K) = 0·5 mm; (H, J) = 0·25 mm.
The Sg trichomes were rare and restricted to a few species. Plants with Sg trichomes showed high levels of variation in the distribution and density of trichomes among species (Fig. 5F–K). In some taxa, these trichomes were associated with the inflorescence (e.g. Lundia densiflora and Manaosella cordifolia). In others, trichomes were dispersed throughout the plant, independently of the inflorescence position (e.g. Adenocalymma adenophorum). Cuspidaria sceptrum (dry habitat) and A. adenophorum (forest habitat) produced a sticky secretion that detained insects (e.g. Diptera). However, no ants were observed visiting these plants (A. Nogueira, pers. obs.). Developmental analyses of the Sg trichomes (i.e. C. sceptrum; Fig. 5A–E) and Pg trichome (Figs. 4F–M) corroborated the categorization of these trichomes into two distinct morphotypes.
Patelliform/cupular glandular (P/Cg) trichomes
This trichome morphotype was present in all 105 species and 21 genera of Bignonieae (Fig. 6), distributed in different positions, with variable abundance and size among plants. The P/Cg trichomes were multicellular, non-vascularized, disc-shaped (patelliform) or cup-shaped (cupular), with a concave surface, and three distinct cell layers. (1) Secretory layer. This layer was formed by 30–68 thin-walled columnar cells that were covered by a thin and smooth cuticle (Fig. 6F, P), but lacked visible pores. The secretory cells were organized in a palisade-like arrangement and had a densely staining cytoplasm. (2) Intermediate layer. This layer had two different morphologies depending on the taxa. Some taxa had an intermediate layer composed of large, ellipsoid, vacuolated cells, surrounded by thick and highly suberized anticlinal walls (Fig. 6C, D, F, I, J, L, M), whereas others had an intermediate layer that was composed of one row of thick, highly suberized cells that connected the secretory cells to the epidermis (Fig. 6G, O, P). (3) Foot or basal layer. This layer was multicellular, and consisted of small quadrangular cells.
Morphoanatomical variation of patelliform/cupular glandular trichomes in Bignonieae. (A–D) Patelliform/cupular trichomes of Adenocalymma pedunculatum: (A) SEM micrograph of mature prophylls with aggregated cupular trichomes; (B, C) photomicrograph of microtome section of young prophylls with developing trichomes (B), trichomes infected by fungi (C) and cupular trichomes on leaflets (D). (E–G) Patelliform glandular trichomes in Amphilophium: (E) SEM micrograph of abaxial side of young leaflets of A. crucigerum; (F, G) photomicrograph of microtome section of patelliform trichomes on the leaflets of A. crucigerum (F) and A. parkerii (G). (H) SEM photomicrographs of aggregated patelliform trichomes at the base of the abaxial side of leaflets of Mansoa difficilis (inset: detail of individual trichome). (I, J) Photomicrograph of microtome section of patelliform trichomes at the base of young leaflets (abaxial side) of Anemopaegma album (I) and on the prophylls (adaxial side) of Bignonia prieurei (J). (K, L) Aggregated trichomes on the young stems (interpetiolar region) of Lundia nitidula (K; SEM photomicrograph) and Fridericia triplinervia (L; microtome sections; inset: detail of individual trichome). (M) Patelliform trichome on the abaxial side of the leaflets of Cuspidaria sceptrum (microtome section). (N) SEM photomicrographs of trichomes on the abaxial side of the leaflets of Tanaecium pyramidatum (inset: detail of individual trichome). (O, P) Photomicrograph of microtome section of trichomes on the adaxial side of the leaflets (O) and prophylls of Pleonotoma albiflora (P). Arrows in A, E, H, K, N indicate each trichome type, where these are most evident. Abbreviations: s.l., secretory layer; i.l., intermediate layer; f.l., foot or basal layer; and fg., fungi growing on the trichome secretory tissue. Scale bars: (A) = 350 µm; (B, I, L, N) = 100 µm; (D, F, G, J, M, O, P) = 50 µm; (E) = 200 µm; (H) = 150 µm, inset = 50 µm; (K) = 110 µm; (N) = 60 µm (inset).
Overall, P/Cg trichomes are flattened structures (Fig. 6E, F, H–K, M, N–P). Although most species of Bignonieae had flattened trichomes, most species of Adenocalymma had cupular trichomes, the cupular shape being caused by cell proliferation and expansion in the upper and central direction at the periphery of the secretory layer, creating a pocket-like structure (Fig. 6A, D, I). In some cases (e.g. A. cymbalum), the expansion of the surrounding epidermis was very pronounced, leading to a ‘volcano-like’ structure (sensu Lohmann, 2006). A few species (e.g. Amphilophium parkerii, Bignonia prieurei) had trichomes with morphology intermediate between the cupular and patelliform variants (Fig. 6G, J). These intermediate morphologies were also observed in a single species (e.g. Adenocalymma pedunculatum, Fig. 6C).
The abundance of P/Cg trichomes was highly variable among different plant parts, varying from sparsely distributed in some species to highly clustered in others. Although all 105 species sampled had P/Cg trichomes, 24·8 % of the species sampled (26/105) had a few P/Cg trichomes scattered on the leaflet blades (e.g. Adenocalymma impressum, Manaosella cordifolia, Periathomega vellozoi, Pyrostegia venusta and Tanaecium pyramidatum). In contrast, 79 of the species sampled had clusters of P/Cg trichomes (>5 trichomes mm−2) in different plant parts (Fig. 6A, H, K, L). In particular, 37·1 % of the species (39 out of 105) had clusters of P/Cg trichomes at the base of leaflets (abaxial side); this feature was particularly common in representatives of Mansoa (Figs 6H, I). Furthermore, 35·2 % of the species (37 out of 105) had clusters of P/Cg trichomes at the interpetiolar region of stems; this feature was particularly common in representatives of Lundia and Fridericia (Fig. 6L). In addition, 22·8 % of the species sampled (24 out of 105) had clusters of P/Cg trichomes on the prophylls of the axillary buds; this feature was particularly common in representatives of Adenocalymma, Bignonia and Pleonotoma (Fig. 6A, B). Clusters of P/Cg trichomes were rarely documented on petioles and petiolules, found only in 11/105 species sampled; clustered P/Cg trichomes were particularly common in representatives of Pachyptera and in the monotypic Callichlamys. P/Cg trichomes clustered on the adaxial side of the leaflet blades were rare and restricted to Bignonia corymbosa. In the field, P/Cg trichomes are yellowish-green and turgid when active, but brownish and dry when inactive. Active trichomes secreted hyaline and viscous nectar, which accumulated in the concave surface of trichomes and was eaten by insects, mostly ants.
Ancestral character state reconstructions
Initially we used information of positional and structural similarity to generate hypotheses of common origin. Four similar structural trichome types were described (see details before) on different vegetative plant parts. Thus, ML optimizations of the four trichome morphotypes (i.e. Ng, Pg, Sg and P/Cg) were performed on each plant part separately (i.e. interpetiolar region of stems, prophylls of the axillary buds, petiole, petiolule, and adaxial and abaxial surface of leaflets) (Table 1). Further details on each reconstruction are presented below.
Table 1.
ML ancestral character state reconstructions of different trichome types in the MRCA of tribe Bignonieae (Bignoniaceae)
| Trichome characters | Two-rate model | One-rate model | Estimated ancestral character states for the tribe Bignonieae (proportional likelihood of trichome presence) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Types | Plant parts | Rate of evolution (gain, loss of trichomes) | Likelihood | Estimated rate of evolution | Likelihood | Was a two-rate model favoured? | |||
| 1 | Ng | Interpetiolar regions (stems) | 0·124, 0·026 | 47·85 | 0·011 | 51·30 | Yes | Presence (0·82) | |
| 2 | Ng | Prophylls | 0·007, 0·00005 | 5·07 | 0·00005 | 5·63 | No | Presence (0·99) | X |
| 3 | Ng | Petioles | 0·051, 0·004 | 28·98 | 0·005 | 30·53 | No | Presence (0·99) | X |
| 4 | Ng | Petiolules | 0·067, 0·009 | 34·89 | 0·006 | 36·43 | No | Presence (0·99) | X |
| 5 | Ng | Leaflet vein (adaxial side) | 0·059, 0·016 | 49·08 | 0·011 | 50·33 | No | Presence (0·95) | |
| 6 | Ng | Leaflet blade (adaxial side) | 0·022, 0·035 | 70·28 | 0·029 | 71·32 | No | Absence (0·31)* | |
| 7 | Ng | Leaflet vein (abaxial side) | 0·027, 0·011 | 54·03 | 0·014 | 55·40 | No | Presence (0·81)* | |
| 8 | Ng | Leaflet blade (abaxial side) | 0·021, 0·022 | 67·53 | 0·022 | 67·53 | No | Absence (0·30)* | |
| 9 | Pg | Interpetiolar regions (stems) | 0·058, 0·005 | 29·06 | 0·005 | 33·53 | Yes | Presence (0·87) | |
| 10 | Pg | Prophylls | 0·536, 0·006 | 5·56 | 5·607 | 7·21 | No | Presence (0·99) | X |
| 11 | Pg | Petioles | 0·041, 0·003 | 20·41 | 0·003 | 23·26 | Yes | Presence (0·93) | X |
| 12 | Pg | Petiolules | 0·044, 0·003 | 22·81 | 0·003 | 26·91 | Yes | Presence (0·90) | X |
| 13 | Pg | Leaflet blade (adaxial side) | 0·035, 0·001 | 18·42 | 0·003 | 23·08 | Yes | Presence (0·79) | |
| 14 | Pg | Leaflet blade (abaxial side) | 0·060, 0·001 | 11·44 | 0·001 | 12·88 | No | Presence (0·99) | X |
| 15 | Sg | Interpetiolar regions (stems) | 0·026, 0·273 | 30·69 | 0·005 | 34·76 | Yes | Absence (0·09)* | X |
| 16 | Sg | Petioles | 0·026, 0·273 | 30·69 | 0·005 | 34·76 | Yes | Absence (0·09)* | X |
| 17 | Sg | Petiolules | 0·026, 0·273 | 30·69 | 0·005 | 34·76 | Yes | Absence (0·09)* | X |
| 18 | Sg | Leaflet blade (adaxial side) | 0·024, 0·290 | 28·27 | 0·004 | 31·90 | Yes | Absence (0·08)* | X |
| 19 | Sg | Leaflet blade (abaxial side) | 0·020, 0·240 | 28·26 | 0·004 | 31·55 | Yes | Absence (0·08)* | X |
| 20 | P/Cg | Interpetiolar regions (stems) | 0·006, 0·018 | 48·55 | 0·011 | 50·72 | Yes | Absence (0·35)* | |
| 21 | P/Cg | Prophylls | 0·005, 0·014 | 43·54 | 0·008 | 45·39 | No | Absence (0·03)* | X |
| 22 | P/Cg | Petioles | 0·014, 0·032 | 52·81 | 0·013 | 53·15 | No | Absence (0·05)* | X |
| 23 | P/Cg | Petiolules | 0·003, 0·027 | 20·69 | 0·003 | 21·21 | No | Absence (0·01)* | X |
| 24 | P/Cg | Leaflet blade (adaxial side) | 0·043, 0·009 | 44·95 | 0·009 | 47·48 | Yes | Presence (0·78)* | |
| 25 | P/Cg | Leaflet blade (abaxial side) | 0·481, 0·009 | 9·86 | 0·001 | 12·18 | Yes | Presence (0·98) | X |
Likelihoods are reported as negative logarithms and the ancestral character state is presented in the last column. Estimates of the rates of evolution are also presented.
Ng, non-glandular trichomes; Pg, peltate glandular trichomes; Sg, stipitate glandular trichomes; P/Cg, patelliform/cupular glandular trichomes.
*Multiple independent gains in Bignonieae.
X = most likely states according to a decision threshold T.
The estimated parameters of the evolutionary model favoured are presented in bold.
Non-glandular (Ng) trichomes
The ML ancestral character state reconstrutions of Ng trichomes indicate that this structure was already present in the MRCA of Bignonieae in almost all vegetative plant parts (Table 1, Fig. 7A). In particular, ancestral state reconstructions indicated that Ng trichomes were already present on the stems (82 %), prophylls (99 %), petioles (99 %), petiolules (99 %) and leaflet veins (adaxial side with 95 %, and abaxial side with 81 %) in the MRCA of Bignonieae. On the leaflet vein of the abaxial side, the Ng trichome seems to have been lost at least once, and gained at least seven times (Fig. 7A). Furthermore, the MRCA of Bignonieae probably did not bear Ng trichomes on the adaxial (31 %) or abaxial (30 %) side of leaflets, with this trait appearing on the blade of the adaxial side of the leaflets at least 20 times, and on the blade of the abaxial side at least 16 times (Fig. 7A). When the two morphological variants of Ng trichomes were considered (unbranched and branched), ML reconstructions showed eight independent origins of the unbranched to branched trichomes (Fig. 7B). Ancestral character state reconstructions of trichome density and size on the abaxial blade of leaflets suggested that the MRCA of Bignonieae had a low density (40 %) of intermediate trichome size (37 %) (Table 2).
Table 2.
ML ancestral character state reconstructions of trichome quantitative traits (size and density/abundance) in the MRCA of tribe Bignonieae (Bignoniaceae)
| Trichome characters | One-rate model | ||||
|---|---|---|---|---|---|
| Types | Quantitative traits (and position on the plants) | Estimated rate of evolution | Likelihood | Estimated ancestral character state of tribe Bignonieae (Proportional likelihood) | |
| 1 | Ng | Density (adaxial side of leaflets) | 0·020 | 107·19 | Small (0·40) |
| 2 | Ng | Size (adaxial side of leaflets) | 0·031 | 80·32 | Intermediate (0·37) |
| 4 | Pg | Density (abaxial side of leaflets) | 0·036 | 109·20 | High (0·35) |
| 5 | Pg | Size of glandular head (abaxial side of leaflets) | 0·023 | 106·35 | Intermediate (0·38) |
| 6 | P/Cg | Number of EFNs (adaxial side of leaflets) | 0·024 | 110·67 | Small (0·49) |
| 7 | P/Cg | Number of EFNs (abaxial side of leaflets) | 0·022 | 109·02 | Intermediate (0·41) |
| 8 | P/Cg | Size of the glandular head (abaxial side of leaflets) | 0·031 | 110·73 | Intermediate (0·34) |
Likelihoods are reported as negative logarithms and the ancestral character state is presented in the last column, based on a one-rate model of evolution (Mk1).
Estimates of the rates of evolution are also presented.
Ng, non-glandular trichomes; Pg, peltate glandular trichomes; P/Cg, patelliform/cupular glandular trichomes.
The size of the glandular head is the diameter of the glandular head.
Evolution of non-glandular trichomes. (A) The evolutionary appearance of trichomes (proportional likelihood of appearance >80 %) is shown by rectangles with different colours, according to the appearance of non-glandular trichomes on different vegetative plant portions. Losses of these characters and the detailed maximum likelihood (ML) ancestral character state reconstructions are presented in Supplementary Data Fig. S1. (B) ML ancestral character reconstruction of non-glandular trichomes coded as ‘simple’ and ‘dendritic’ trichomes. Black portions of each pie chart represent the proportional likelihood of the appearance of dendritic non-glandular trichomes. The scale on the right is millions of years ago.
Peltate and glandular (Pg) trichomes
Ancestral character state reconstructions of Pg trichomes indicate that the MRCA of Bignonieae already had Pg trichomes in all plant parts (Table 1, Supplementary Data Fig. S1), the most widespread type of trichome in Bignonieae. In particular, this trichome type was probably present on the stems (87 %), prophylls (99 %), petioles (93 %), petiolules (90 %), and adaxial (79 %) and abaxial (99 %) side of leaflets in the MRCA of Bignonieae. Even though Pg trichomes were probably present in the MRCA of Bignonieae on the adaxial side of leaflets, these trichomes seem to have been lost at least once in the ancestor of the clade Adenocalymma + Neojobertia, and re-gained five times in this clade. Ancestral character state reconstructions of quantitative features of Pg trichomes suggested that the MRCA of Bignonieae probably had a high density (35 %) and intermediate size (38 %) on the abaxial side of leaflets (Table 2, Fig. 8). The largest and densest peltate glandular trichomes were observed in Amphilophium, Pyrostegia and Stizophyllum.
Evolution of peltate glandular trichomes. Right: ML ancestral state reconstructions of trichome density. Left: ML ancestral state reconstructions of the diameter of the trichome glandular heads. Asterisks indicate the lineages in which peltate glandular trichomes are ‘larger’ and ‘denser’ in the tribe.
Stipitate glandular (Sg) trichomes
The MRCA of Bignonieae did not bear stipitate glandular trichomes on all vegetative plant parts sampled (Table 1, Fig. 9). Nine independent gains of Sg trichomes were hypothesized in Bignonieae (Fig. 9), in the following genera: Adenocalymma (two gains), Fridericia (one gain), Xylophragma (one gain), Cuspidaria (two gains), Lundia (one gain), Manaosella (one gain) and Martinella (one gain). The morphological differences observed in Sg among taxa and the lack of a historical connection between those trichomes in the various lineages suggest that the various Sg trichomes encountered in Bignonieae did not have the same origin in this plant group. Thus, the evolution of quantitative features related to these trichomes was not explored further.
Evolution of stipitate glandular trichomes. ML ancestral character state reconstruction indicated nine independent evolutions of this trichome morphotype in that lineage. Black portions of each pie chart represent the proportional likelihood of stipitate glandular trichome appearances. The scale on the right is millions of years ago.
Patelliform/cupular glandular (P/Cg) trichomes
The MRCA of Bignonieae probably only had P/Cg trichomes on the abaxial side of leaflets (Table 1, Fig. 10). In particular, this trichome type was probably not present on stems (35 %), prophylls (3 %), petioles (5 %), petiolules (1 %) or the adaxial side of leaflets (78 %) of the MRCA of Bignonieae. In addition, P/Cg trichomes probably evolved multiple times in each plant part (Fig. 10). More specifically, at least six independent evolutionary events of P/Cg trichomes were hypothesized for the interpetiolar region of stems of Bignonieae, including an origin in the MRCA of the ‘Tanaecium + Fridericia + Xylophragma + Cuspidaria + Tynanthus + Lundia’ clade, an origin in the MRCA of Dolichandra and an origin in the MRCA of Martinella. Similarly, at least six origins of P/Cg trichomes were hypothesized for the prophylls of the axillary buds, including one in the MRCA of Adenocalymma, one in the MRCA of Pleonotoma and one in the MRCA of the ‘Amphilophium + Anemopaegma + Pyrostegia + Mansoa + Bignonia’ clade. Ancestral state reconstructions of quantitative features of P/Cg trichomes suggested that the MRCA of Bignonieae showed a low density of P/Cg trichomes on the adaxial side of leaflets (49 %), a high density on the abaxial side (41 %) and small glandular heads (34 %) (Table 2).
Evolution of the patelliform/cupular glandular trichomes (P/Cg). The evolutionary appearance of trichomes (proportional likelihood of appearance >80 %) is shown by rectangles with different colours, according to the appearance of P/Cg trichomes on different vegetative plant portions. Losses of these characters and the detailed ML ancestral character state reconstructions are presented in Supplementary Data Fig. S1. The scale on the right is millions of years ago.
DISCUSSION
In this study we encountered four main trichome morphotypes (i.e. non-glandular, peltate glandular, stipitate glandular and patelliform/cuputar glandular) with a highly variable position on the vegetative plant parts (i.e. stems, prophylls, petiole, petiolule, and adaxial and abaxial blade of leaflets) among species of Bignonieae. Three of the four trichome morphotypes (Ng, Pg and P/Cg) were already present in the MRCA of Bignonieae, but in general these trichome types only occur on some plant parts, with secondary occupations of each trichome type on new plant parts during the history of diversification of Bignonieae. At least for some plant parts, we corroborate the hypothesis of a single evolutionary origin for the different trichome types (e.g. Ng, Pg and P/Cg trichomes), suggesting that these structures are best treated under the same name based on their morphological similarity and common evolutionary history of such structures.
Integrating morphological data and ancestral character state reconstructions
Trichomes often demonstrate variable macro- and micromorphological features, making it difficult to determine the exact trichome type that is being referred to in the literature (Theobald et al., 1979). In the specific case of Bignoniaceae, a wide range of terms has been used to describe the same trichome type (Fig. 1). Below we summarize major morphoanatomical features and the evolutionary history of each trichome type.
Non-glandular (Ng) trichomes
The Ng trichomes of representatives of Bignonieae have only rarely been described in detail (e.g. Ogundipe and Wujek, 2004). Instead of being described in terms of their structure, Ng trichomes have been traditionally described in terms of their density and overall appearance. For example, ‘villose leaves’ have been described for Anemopaegma velutinum, referring to a hairy cover of dense, straight, long and soft trichomes (Bureau and Schumann, 1897). However, no information is available on the anatomy and developmental sequence of Ng trichomes in Bignoniaceae or other plant groups (e.g. Lamiaceae; Naidu and Shah, 1981). Even though Ng trichomes have generally been treated as morphologically homogeneous structures, the variable number of cells (and size) and variable patterns of distribution of this trichome morphotype on different plant parts indicate the need for more detailed morphoevolutionary studies of trichomes in general.
Ancestral state reconstructions of Ng trichomes in Bignonieae indicated that simple and unbranched trichomes were present in the ancestor of Bignonieae, with branching trichomes having evolved at least eight times. This transition from simple to branched trichomes was similar to that observed in Brassicaceae, in which the pattern of trichome evolution indicated numerous innovations of trichome branching (Beilstein et al., 2006). Furthermore, a hierarchical pattern of occupation of trichomes over different vegetative plant parts was also observed during the evolutionary history of Bignonieae. More specifically, the MRCA of the tribe probably already had Ng trichomes on the branches, prophylls of the axillary buds, and petioles. However, it was only later that trichomes covered the surfaces of leaflet blades. In most species, the density of trichomes remained low and Ng trichomes remained unbranched. A few lineages recently acquired higher trichome densities (e.g. species of Amphilophium and Tynanthus) and/or have acquired branched trichomes (e.g. species of Amphilophium and Fridericia).
Peltate glandular (Pg) trichomes
The Pg trichomes were first recorded in representatives of Bignoniaceae in the 19th century in taxonomic treatments of the Prodromus Systematis Naturalis Regni Vegetabilis (Candolle, 1845) and Flora Brasiliensis (Bureau and Schumann, 1897). For example, in the Flora Brasiliensis (Bureau and Schumann, 1897), Pg trichomes were documented in Cydista aequinoctialis (‘densius lepidota’, p. 32), Pithecoctenium dolichoides (‘conspicue lepidota obsolete pellucide punctulata’, p. 22), Pyrostegia venusta (‘lepidibus minutis inspersa et ope eorum plus minus manifeste pellucide punctate’, p. 23) and Stizophyllum inaequilaterum (‘pellucide punctate’, p. 29). Seibert (1948) used two distinct terms to refer to the variants of Pg trichomes: ‘pellucid glands’ and ‘glandular scales’. This author considered Amphilophium, Pithecoctenium (= Amphilophium sensu Lohmann, 2003; Lohmann and Taylor, 2013), Pyrostegia and Stizophyllum to have ‘pellucid glands’, and compared these structures with the translucent pellucid dots in leaves of species of Rutaceae. However, in Rutaceae, such structures are present in almost all members of the family and represent secretory cavities (Engler, 1931; Kubitzki et al., 2011) that are immersed in the mesophyll, and not epidermal projections as in Bignoniaceae.
Seibert (1948) considered the majority of the representatives of Bignoniaceae to have ‘glandular scales’ (and a few ‘pellucid glands’) that he described as ‘minute scales found on stems, petioles, leaves, calyx, corolla, ovary and fruit, and responsible for the ‘lepidote’ condition frequently encountered in the Bignoniaceae'. Gentry (1980), on the other hand, preferred to call these structures ‘glandular-lepidote trichomes’ instead of ‘glandular scales’. The study of Pg trichomes conducted here revealed that even though these trichomes are identical in structure in all studied species, they are much denser in Amphilophium. All these structures are composed of a glandular convex or rounded head of a variable number of cells. Furthermore, this trichome type has a single origin in Bignonieae. The morphological similarity of Pg among species and the phylogenetic pattern indicated that this secretory structure is best treated uniformly under the same terminology. We here propose that these structures are best treated morphologically as peltate glandular trichomes.
Ancestral character state reconstructions indicate that Pg trichomes were already present on both sides of the leaflet blades and all other vegetative portions of the MRCA of Bignonieae. However, Pg trichomes probably originated prior to the origin of Bignonieae, given that this trait has been documented in other representatives of Bignoniaceae, in three clades defined by Olmstead et al. (2009): the Palaeotropical clade (Ogundipe and Wujek, 2004), Tecomeae (Parija and Samal, 1936) and Jacarandeae (Zatta et al., 2009). This trichome morphotype is widely distributed in Bignonieae, but has only been reported in a few representatives of the family in general, probably reflecting the lack of detailed studies of trichome structure in other lineages of Bignoniaceae.
The density and size of the glandular heads represent the most variable features of Pg trichomes, differently from the other trichome types studied here. Ancestral state reconstructions of the individual features of this trichome type indicate that the MRCA of all Bignonieae probably had high densities of Pg trichomes with intermediate head size on the abaxial portion of leaflet blades. Even though the MRCA of the ‘multiples of four’ clade and ‘Arrabidaea and allies’ clade (sensu Lohmann, 2006) had Pg trichomes with intermediate size heads and densities, the Pg trichomes increased in size and density in the first clade and decreased in size and density in the second one. Trichomes with large heads and high densities were relatively rare in Bignonieae, with huge large heads being restricted to two lineages (Pyrostegia and Stizophyllum), and huge high densities only present in 11 lineages, and being particularly common in Amphilophium.
Stipitate glandular (Sg) trichomes
This is the first record of Sg trichomes on the vegetative portions of representatives of Bignonieae. This kind of glandular trichome generally lacks a sticky secretion, and has been associated with alternative biotic defences or direct defence against herbivores. This trichome type has never been associated with ant attraction, which has more generally been associated with nectar-secreting trichomes. For example, adhesive glandular trichomes, a kind of Sg trichome, have been associated with spider bodyguards that often feed on carcases of insects that adhere to these trichomes (Morais-Filho and Romero, 2010). Ancestral character state reconstructions of this trichome morphotype indicate that Sg trichomes were not present in the MRCA of the group, independent of the plant portions analysed, and suggest that this trichome type has evolved at least nine times in the group. These results indicate that even though there is general structural similarity between the Sg trichomes encountered in the various lineages of Bignonieae, these similarities are not based on a common ancestry, suggesting a homoplastic pattern of evolution of these secretory structures.
Patelliform/cupular glandular (P/Cg) trichomes
The P/Cg glandular trichomes were observed on different plant parts, but only on the abaxial side of the leaflet blade in nearly all species of the group. These trichomes are associated with nectar secretion in Bignoniaceae, and have been commonly called extrafloral nectaries (EFNs) as they are not thought to be involved in pollination (Elias, 1983). However, the specific function of P/Cg trichomes has only been studied in five representatives of Bignoniaceae: Campsis radicans (Elias and Gelband, 1975), Catalpa speciosa (Stephenson, 1982), Catalpa bignonioides (Ness, 2003), Anemopaegma album and Anemopaegma scabriusculum (Nogueira et al., 2012a). This terminology (EFNs) was widely used in Bignoniaceae in spite of the lack of functional evidence for this structure, i.e.‘EFN’ has been used in the literature solely to refer to a particular trichome morphotype, without direct evidence for its function.
In Bignonieae, P/Cg trichomes were documented in all 105 species sampled in the tribe, and also in the 16 species of Bignonieae trichome whose morphoanatomy was analysed in detail. Notably, this trichome type has been the most widely reported in all previous anatomical studies conducted in Bignoniaceae (Fig. 1). A mixture of terminologies has been used to designate these structures, obscuring the phylogenetic signal of this trichome type among species. For example, Seibert (1948) called P/Cg trichomes ‘glands’, whereas Elias (1983) called them ‘extrafloral nectaries’, terminologies that are more directly associated with the function than the overall morphological structure. On the other hand, terminologies that are more directly associated with the overall morphological structure of trichomes have also been proposed. For example, Díaz-Castelazo et al. (2005) treated P/Cg trichomes as ‘scale-like trichomes’, and Laroche (1974) called P/Cg trichomes ‘patelliform glands’. In addition, Rivera (2000) used the topographical terminology proposed by Fahn (1979) and characterized P/Cg trichomes as belonging to ‘type 12’ (i.e. trichomes located in the outer epidermis of the calyx and/or corolla).
Our results corroborate earlier observations that P/Cg trichomes are more often present in the leaflet blades of representatives of Bignonieae (Elias, 1983). However, P/Cg trichomes have also been observed on the petals, sepals and fruits (Parija and Samal, 1936; Seibert, 1948; Laroche, 1974; Elias and Prance, 1978; Stephenson, 1982; Subramanian and Inamdar, 1989). The variation found in the structure of P/Cg trichomes of Bignonieae has been used to distinguish different species (Elias and Prance, 1978; Elias, 1983). Our study corroborates the earlier suggestion that the large and cupular morphotype of P/Cg that is marginally surrounded by the expansion of epidermis is typical of Adenocalymma (‘volcano’-shaped glands as proposed by Lohmann, 2006). Furthermore, this study also corroborates the observation that the patelliform and cupular trichomes have a uniform anatomical structure (Rivera, 2000). Indeed, these variants of P/Cg trichomes have been shown to be generally homogeneous morphologically, varying only in terms of their shape and showing a differential cell proliferation and expansion, which is larger in cupular trichomes. These results thus suggest that cupular and patelliform trichomes are only superficially different and best treated under the same type of trichome called here P/Cg trichomes.
Ancestral character state reconstructions indicated that the MRCA of Bignonieae already bore P/Cg trichomes on both surfaces of the leaflet blades (most probably just on the abaxial side of leaflet blades). However, similar to what was found for Pg trichomes, this trait has also been documented in some representatives of other tribes/clades of Bignoniaceae (Parija and Samal, 1936; Seibert, 1948), indicating that their origin probably precedes the origin of the tribe. It has been hypothesized that P/Cg trichomes may simply represent an increase in complexity of Pg trichomes (Parija and Samal, 1936; Elias and Newcombe, 1979). In this case, the structural changes resulting in the evolution of these two trichome types would have occurred before the diversification of Bignonieae, since both types of trichomes were already present in the MRCA of the tribe. However, this hypothesis remains to be tested in a broader systematic context, in combination with detailed ontogenetic and evolutionary studies of both trichome types. Ontogenetic studies conducted with other representatives of Lamiales demonstrated an ontogenetic resemblance between the patelliform glandular trichomes of the Acanthaceae (McDade and Turner, 1997) and those of Bignoniaceae (Inamdar, 1969; Subramanian and Inamdar, 1986a, b). In both cases, the trichomes are formed from a single initial protodermic cell, suggesting that the trichomes found in both families might have a common origin, which would imply a more ancient origin for this structure.
Ancestral character state reconstructions of the density of P/Cg trichome morphotype per leaflet suggest that the ancestor of tribe Bignonieae probably had low densities of P/Cg trichomes on the adaxial side of leaflets and intermediate densities of trichomes on the abaxial side, with intermediate secretory head size. Furthermore, ancestral character state reconstructions also suggest that the ancestor of Bignonieae did not have P/Cg trichomes in the interpetiolar portions of stems nor in the prophylls of the axillary buds, with this feature evolving much more recently in other representatives of the tribe. Since these trichomes have been shown to be associated with ant attraction, and species with higher numbers of nectar-secreting trichomes had higher abundances of ants on the plants (Nogueira et al. 2012a, b), portions of the plant with high concentrations of these trichomes probably have a better protective role against plant herbivores. Thus, the clustering of such secretory structures on particular plant regions, as observed on the prophylls of axillary buds and interpetiolar portions of stems, suggests that these trichomes could be indirect protecting plant tissues that are costly to produce (i.e. the meristems of the axillary buds; Heil and McKey, 2003).
Concluding remarks
Despite the presumed functional importance of trichomes in different plant species and the importance for microevolutionary studies (Levin, 1973), little is still known about the macroevolutionary patterns of those structures in different plant families and on different plant organs. Traditionally, trichomes have been thought to evolve multiple times in the history of angiosperms. Theobald et al. (1979, p. 53) noted: ‘it is evident that the trichomes have had infinite independent origins and as a result no ‘phylogenetic’ sequence, except on a very limited scale within certain small groups'. However, this does not seem to be the case in Bignonieae, at least in some parts of the plants, in which three major trichome morphotypes (non-glandular, peltate glandular and patelliform/cupular glandular) probably had a single origin in the group. Only stipitate glandular trichomes seem to have evolved multiple times independently of the position of their occurrence on vegetative organs. Our results indicate that most trichomes found in representatives of Bignonieae had the same origin and were similar morphologically, and are thus best treated under the same name (in each morphotype). A standardization of trichome terminology greatly facilitates comparisons among taxa, allowing inferences on relationships as well as inferences on their eco-evolutionary consequences.
Although we have improved the understanding of trichome evolution in a clade of angiosperms, it still remains unclear how trichomes interact ontogenetically and ecologically. For example, trade-off between trichome types could explain patterns of evolution in some clades (e.g. trade-off between Sg and P/Cg trichomes; Nogueira et al., 2012b) or at least clarify the independent evolutionary trajectories of each trichome type. No information is available about the genetic basis of these trichome types and relationships with the patterns of position, density and size of trichomes. It is possible that putative heterotopic events (as discussed by Baum and Donoghue, 2002), with modifications of expression of some types of trichomes in specific plant parts, may have occurred during the evolutionary history of Bignonieae. Data directly related to the function of different trichomes are of importance for understanding the role of selection agents such as herbivores, sunlight and water for the evolution of trichome types. Experimental approaches that test functional hypotheses of trichomes in Bignonieae may be able to address the interpopulation variation of trichome characters in the future.
ACKNOWLEDGEMENTS
This paper is part of the doctoral thesis of A.N. This project was funded by the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP Fellowship to A.N. 2007/54917-1 and FAPESP grant to L.G.L. 2007/55433-8) and by a Pq-2 Grant from the Conselho Nacional de Pesquisas to LGL (CNPq, Brazilian Government). Logistic support was provided by IB-USP, IB-UNESP, Parque Estadual de Grão Mogol (Minas Gerais State/Brazil), Parque Nacional da Chapada Diamantina (Bahia State/Brazil) and Reserva Florestal Adolpho Ducke (Amazonas State/Brazil). We are grateful to the technical team of the Centro de Microscopia Eletrônica from IB-UNESP for the preparation of samples for MEV, and to Yve Canaveze for the anatomical assistance throughout this project. We also thank Alexandre Zuntini, Benoit Loeuille, Juliana G. Rando, Luciano P. de Queiroz, Miriam Kaehler, Paulo Inácio de Prado, Thiago J. Izzo, Thomas M. Lewinsohn, Robin Burnham and one anonymous reviewer for fruitful discussions and comments on this manuscript.
LITERATURE CITED
- Baum DA, Donoghue MJ. Transference of function, heterotopy and the evolution of plant development. In: Cronk Q, Bateman R, Hawkins JA, editors. Developmental genetics and plant evolution. London: Taylor & Francis; 2002. pp. 52–69. [Google Scholar]
- Beilstein MA, Al-Shehbaz IA, Kellogg EA. Brassicaceae phylogeny and trichome evolution. American Journal of Botany. 2006;93:607–619. [Abstract] [Google Scholar]
- Bureau E, Schumann K. Bignoniaceae. In: Martius CFPV, editor. Flora Brasiliensis. 1897. [Google Scholar]
- Candolle D. Prodromus Systematis Naturalis Regni Vegetabilis. Paris: Sumptibus Sociorum Treuttel et Würtz; 1845. [Google Scholar]
- Chauveau O, Eggers L, Raquin C, et al. Evolution of oil-producing trichomes in Sisyrinchium (Iridaceae): insights from the first comprehensive phylogenetic analysis of the genus. Annals of Botany. 2011;107:1287–312. [Europe PMC free article] [Abstract] [Google Scholar]
- Díaz-Castelazo C, Rico-Gray V, Ortega F, Ángeles G. Morphological and secretory characterization of extrafloral nectaries in plants of Coastal Veracruz, Mexico. Annals of Botany. 2005;96:1175–1189. [Europe PMC free article] [Abstract] [Google Scholar]
- Duarte MR, Jurgensen I. Diagnose Morfoanatômica de folha e caule de Pyrostegia venusta (Ker Gawl.) Miers, Bignoniaceae. Latin American Journal of Pharmacy. 2007;26:70–75. [Google Scholar]
- Elias TS. Extrafloral nectaries: their structure and distribution. In: Bentley B, Elias TS, editors. The biology of nectaries. New York: Columbia University Press; 1983. pp. 174–203. [Google Scholar]
- Elias TS, Gelband H. Morphology and anatomy of floral and extrafloral nectaries in Campsis (Bignoniaceae) American Journal of Botany. 1975;63:1349–1353. [Google Scholar]
- Elias TS, Newcombe LF. Foliar nectaries and glandular trichomes in Catalpa (Bignoniaceae) Acta Botanica Sinica. 1979;21:215–228. [Google Scholar]
- Elias TS, Prance GT. Nectaries on the fruit of Crescentia and other Bignoniaceae. Brittonia. 1978;30:175–181. [Google Scholar]
- Elle E, van Dam NM, Hare JD. Cost of glandular trichomes, a ‘resistance’ character in Datura wrightii Regel (Solanaceae) Evolution. 1999;53:22–35. [Abstract] [Google Scholar]
- Engler A. Rutaceae. In: Engler A, Prantl K, editors. Die natürlichen Pflanzenfamilien. 2nd edn, 19a. Leipzig: Wilhelm Engelmann; 1931. pp. 187–359. [Google Scholar]
- Fahn A. Nectaries. In: Fahn A, editor. Secretory tissues in plants. London: Academic Press; 1979. pp. 51–111. [Google Scholar]
- Fahn A. Structural and functional properties of trichomes of xeromorphic leaves. Annals of Botany. 1986;57:631–637. [Google Scholar]
- Fougère-Danezan M, Herendeen PS, Maumont S, Bruneau A. Morphological evolution in the variable resin-producing Detarieae (Fabaceae): do morphological characters retain a phylogenetic signal? Annals of Botany. 2010;105:311–325. [Europe PMC free article] [Abstract] [Google Scholar]
- Gentry AH. Coevolutionary patterns in Central American Bignoniaceae. Annals of the Missouri Botanical Garden. 1974;61:728–759. [Google Scholar]
- Gentry AH. Bignoniaceae: part I (Crescentieae and Tourrettieae) Flora Neotropica. 1980;25:1–130. [Google Scholar]
- Heil M, McKey D. Protective ant–plant interactions as model systems in ecological and evolutionary research. Annual Review of Ecology, Evolution, and Systematics. 2003;34:425–453. [Google Scholar]
- Hyakutake S, Grotta AS. Contribuição ao estudo morfológico e anatômico de Anemopaegma arvense (Vell.) Stellfeld var. petiolata Bur. Bignoniaceae. Revista da Faculdade de Farmácia e Bioquímica da Universidade de São Paulo. 1965;3:51–78. [Google Scholar]
- Inamdar JA. Structure and ontogeny of foliar nectaries and stomata in Bignonia chamberlaynii. Proceedings of the Indian Academy of Sciences. 1969;70:232–241. [Google Scholar]
- Johansen DA. Plant microtechnique. New York: McGraw-Hill Book Company; 1940. [Google Scholar]
- Johnson MTJ, Agrawal AA, Maron JL, Salminen JP. Heritability, covariation and natural selection on 24 traits of common evening primrose (Oenothera biennis) from a field experiment. Journal of Evolutionary Biology. 2009;22:1295–307. [Abstract] [Google Scholar]
- Kaplan I, Dively GP, Denno RF. The costs of anti-herbivore defense traits in agricultural crop plants: a case study involving leafhoppers and trichomes. Ecological Applications. 2009;19:864–72. [Abstract] [Google Scholar]
- Kubitzk K, Kallunki JA, Duretto M, Wilson PG. Rutaceae. In: Kubitzki K, editor. Flowering plants. Eudicots: Sapindales, Cucurbitales, Myrtaceae (the families and genera of vascular plants) Hamburg: Springer; 2011. pp. 276–356. [Google Scholar]
- Laroche RC. Consideration of the calyx of Adenocalymma comosum (Cham.) A.P.DC. Annals of the Missouri Botanical Garden. 1974;61:530–533. [Google Scholar]
- Larson A, Losos JB. Phylogenetic systematics of adaptation. In: Rose MR, Lauder GV, editors. Adaptation. San Diego: Academic Press; 1996. pp. 187–220. [Google Scholar]
- Levin DA. The role of trichomes in plant defense. Quarterly Review of Biology. 1973;48:3–15. [Google Scholar]
- Lohmann LG. Phylogeny, classification, morphological diversification and biogeography of Bignonieae (Bignoniaceae). Louis, EUA: University of Missouri-St; 2003. PhD Dissertation. [Google Scholar]
- Lohmann LG. Untangling the phylogeny of neotropical lianas (Bignonieae, Bignoniaceae) American Journal of Botany. 2006;93:304–318. [Abstract] [Google Scholar]
- Lohmann LG, Taylor CM. A new generic classification of Bignonieae (Bignoniaceae) based on molecular phylogenetic data and morphological synapomorphies. Annals of the Missouri Botanical Garden. 2013 in press. [Google Scholar]
- Lohmann LG, Bell CD, Calió MF, Winkworth RC. Pattern and timing of biogeographical history in the Neotropical tribe Bignonieae (Bignoniaceae). Botanical. Journal of the Linnean Society. 2013;171:154–170. [Google Scholar]
- MacLeod N, Forey PL. Morphology, shape and phylogeny. Boca Raton, FL: CRC Press; 2002. Introduction; pp. 1–7. [Google Scholar]
- Machado SR, Morellato LPC, Sajo MG, Oliveira PS. Morphological patterns of extrafloral nectaries in woody plant species of the Brazilian cerrado. Plant Biology. 2008;10:1–14. [Abstract] [Google Scholar]
- Maddison W, Maddison D. Mesquite: a modular system for evolutionary analysis. 2007 http://mesquiteproject.org . [Google Scholar]
- Mauricio R. Ontogenetics of QTL: the genetic architecture of trichome density over time in Arabidopsis thaliana. Genetica. 2005;123:75–85. [Abstract] [Google Scholar]
- Morais-Filho JC, Romero GQ. Plant glandular trichomes mediate protective mutualism in a spider–plant system. Ecological Entomology. 2010;35:485–494. [Google Scholar]
- McDade LA, Turner MD. Structure and development of bracteal nectary glands in Aphelandra (Acanthaceae) American Journal of Botany. 1997;84:1–15. [Google Scholar]
- Naidu AC, Shah GL. Observation on the cotyledonary stomata and trichomes and their ontogeny in some genera of Lamiaceae. Phyton. 1981;21:137–152. [Google Scholar]
- Ness JH. Catalpa bignonioides alters extrafloral nectar production after herbivory and attracts ant bodyguards. Oecologia. 2003;134:210–218. [Abstract] [Google Scholar]
- Nogueira A, Guimarães E, Machado Silvia Rodrigues, Lohmann LG. Do extrafloral nectaries present a defensive role against herbivores in two species of the family Bignoniaceae in a Neotropical savanna? Plant Ecology. 2012a;213:289–301. [Google Scholar]
- Nogueira A, Rey PJ, Lohmann LG. Evolution of extrafloral nectaries: adaptive process and selective regime changes from forest to savanna. Journal of Evolutionary Biology. 2012b;25:2325–2340. [Abstract] [Google Scholar]
- O'Brien TP, Feder N, McCully ME. Polychromatic staining of plant cell walls by toluidine blue. Protoplasma. 1964;59:368–373. [Google Scholar]
- Ogundipe OT, Wujek DE. Foliar anatomy on twelve genera of Bignoniaceae (Lamiales) Acta Botanica Hungarica. 2004;46:337–361. [Google Scholar]
- Oliveira P, Leitão-Filho H. Extrafloral nectaries: their taxonomic distribution and abundance in the woody flora of cerrado vegetation in southeast Brazil. Biotropica. 1987;19:140–148. [Google Scholar]
- Olmstead RG, Zjhra ML, Lohmann LG, Grose SO, Eckert AJ. A molecular phylogeny and classification of Bignoniaceae. American Journal of Botany. 2009;96:1731–1743. [Abstract] [Google Scholar]
- Parija P, Samal K. Extra-floral nectaries in Tecoma capensis Lindl. Proceedings of the Indian Academy of Sciences. 1936:241–246. [Google Scholar]
- Payne WW. A glossary of plant hair terminology. Brittonia. 1978;30:239–255. [Google Scholar]
- Rao L. A short note on the extrafloral nectaries in Spathodea stipulata. Journal of the Indian Botanical Society. 1926;5:113–116. [Google Scholar]
- Ree RH, Donoghue MJ. Inferring rates of change in flower symmetry in Asterid Angiosperms. Systematic Biology. 1999;48:633–641. [Google Scholar]
- Rivera GL. Nectarios extranupciales florales en especies de Bignoniaceae de Argentina. Darwiniana. 2000;16:125–126. [Google Scholar]
- Romero GQ, Souza JC, Vasconcellos-Neto J. Anti-herbivore protection by mutualistic spiders and the role of plant glandular trichomes. Ecology. 2008;89:3105–3115. [Abstract] [Google Scholar]
- Scotland R, Pennington RT. Homology and systematics: coding characters for phylogenetic analysis. Boca Raton: FL: CRC Press; 2000. Systematics Association Special Volumes. [Google Scholar]
- Seibert RJ. The use of glands in a taxonomic consideration of the family Bignoniaceae. Annals of the Missouri Botanical Garden. 1948;35:123–137. [Google Scholar]
- Souza LA, Santos GDO, Moscheta IS. Morfoanatomia floral de espécies lianescentes de Bignoniaceae. Iheringia. 2010;5790:5–15. [Google Scholar]
- Stephenson AG. The role of the extrafloral nectaries of Catalpa speciosa in limiting herbivory and increasing fruit production. Ecology. 1982;63:663–669. [Google Scholar]
- Subramanian RB, Inamdar JA. The structure, secretion and biology of nectaries in Tecomaria capensis Thunb (Bignoniaceae) Phytomorphology. 1989;39:69–74. [Google Scholar]
- Subramanian RB, Inamdar JA. Nectaries in Bignonia illicium L. – Ontogeny, structure and functions. Proceedings of the Indian Academy of Sciences. 1986a;96:135–140. [Google Scholar]
- Subramanian RB, Inamdar JA. Occurrence, structure, ontogeny and biology of nectaries in Kigelia pinnata DC. The Botanical Magazine. 1986b;98:67–73. [Google Scholar]
- Symonds VV, Godoy AV, Alconada T, et al. Mapping quantitative trait loci in multiple populations of Arabidopsis thaliana identifies natural allelic variation for trichome density. Genetics. 2005;169:1649–1658. [Europe PMC free article] [Abstract] [Google Scholar]
- Theobald WL, Krahulik JL, Rollins RC. Trichome description and classification. In: Metcalfe CR, Chalk L, editors. Anatomy of the dicotyledons – Volume 1. Oxford: Oxford University Press; 1979. pp. 40–53. [Google Scholar]
- Thomas V, Dave Y. Structure and biology of nectaries in Tabebuia serratifolia Nichols (Bignoniaceae) Botanical Journal of the Linnean Society. 1992;109:395–400. [Google Scholar]
- Tresvenzol LMF, Fiuza TS, Rezende MH, et al. Morfoanatomia de Memora nodosa (Silva Manso) Miers, Bignoniaceae. Revista Brasileira de Farmacognosia Brazilian Journal of Pharmacognosy. 2011;20:833–842. [Google Scholar]
- Valverde PL, Fornoni J, Nunez-Farfan J. Defensive role of leaf trichomes in resistance to herbivorous insects in Datura stramonium. Journal of Evolutionary Biology. 2001;14:424–432. [Google Scholar]
- Wagner GJ, Wang E, Shepherd RW. New approaches for studying and exploiting an old protuberance, the plant trichome. Annals of Botany. 2004;93:3–11. [Europe PMC free article] [Abstract] [Google Scholar]
- West-Eberhard MJ. Homology. Developmental Plasticity and Evolution. 2003:485–497. [Google Scholar]
- Zatta DT, Oliveira FNM, Bara MTF, Rezende MH. Morfoanatomia Foliar e Parâmetros de Qualidade da Jacaranda decurrens (Bignoniaceae) Latin American Journal of Pharmacy. 2009;28:358–365. [Google Scholar]
Articles from Annals of Botany are provided here courtesy of Oxford University Press
Full text links
Read article at publisher's site: https://doi.org/10.1093/aob/mct201
Read article for free, from open access legal sources, via Unpaywall:
https://academic.oup.com/aob/article-pdf/112/7/1331/17009128/mct201.pdf
Citations & impact
Impact metrics
Citations of article over time
Alternative metrics
Article citations
Contrasting patterns of genetic and phenotypic divergence of two sympatric congeners, Phragmites australis and P. hirsuta, in heterogeneous habitats.
Front Plant Sci, 14:1299128, 12 Dec 2023
Cited by: 1 article | PMID: 38162310 | PMCID: PMC10756910
An attempt at fixing the oversimplification of Nightshades' (genus Solanum) epidermal hair complexity.
Front Plant Sci, 14:1176674, 28 Sep 2023
Cited by: 0 articles | PMID: 37841602 | PMCID: PMC10569493
Taxonomic revision of Martinella Baill. (Bignonieae, Bignoniaceae).
PhytoKeys, 177:77-116, 13 May 2021
Cited by: 0 articles | PMID: 34040488 | PMCID: PMC8137638
Glandular trichomes of the flowers and leaves in Millingtonia hortensis (Bignoniaceae).
Planta, 253(1):13, 03 Jan 2021
Cited by: 5 articles | PMID: 33389109
An updated synopsis of Tanaecium (Bignonieae, Bignoniaceae).
PhytoKeys, 132:31-52, 26 Sep 2019
Cited by: 2 articles | PMID: 31598067 | PMCID: PMC6776558
Go to all (14) article citations
Data
Data behind the article
This data has been text mined from the article, or deposited into data resources.
BioStudies: supplemental material and supporting data
Similar Articles
To arrive at the top five similar articles we use a word-weighted algorithm to compare words from the Title and Abstract of each citation.
Secondary phloem diversity and evolution in Bignonieae (Bignoniaceae).
Ann Bot, 116(3):333-358, 01 Sep 2015
Cited by: 4 articles | PMID: 26311709 | PMCID: PMC4549963
Evolution of extrafloral nectaries: adaptive process and selective regime changes from forest to savanna.
J Evol Biol, 25(11):2325-2340, 27 Sep 2012
Cited by: 10 articles | PMID: 23013544
Phylogeny and biogeography of Tynanthus Miers (Bignonieae, Bignoniaceae).
Mol Phylogenet Evol, 85:32-40, 07 Feb 2015
Cited by: 2 articles | PMID: 25659336
Molecular Mechanisms of Plant Trichome Development.
Front Plant Sci, 13:910228, 01 Jun 2022
Cited by: 8 articles | PMID: 35720574 | PMCID: PMC9198495
Review
Free full text in Europe PMCPlant Glandular Trichomes: Natural Cell Factories of High Biotechnological Interest.
Plant Physiol, 175(1):6-22, 19 Jul 2017
Cited by: 98 articles | PMID: 28724619
Review