| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
2Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs, Connecticut 06269-3043 USA; 3Department of Biology, Duke University, Durham, North Carolina 27708 USA
Received for publication August 21, 2003. Accepted for publication December 18, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Bryophytes • entomophily • phylogeny • Splachnaceae • Splachnum • spore dispersal syndrome • Tayloria • Tetraplodon
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Examples of rapid diversification following dispersal to islands or access to new habitats occur in various lineages of plants and animals. If dispersal to new open habitats occurs more than once, convergent trends in morphological diversification can be expected (e.g., in sticklebacks [Taylor et al., 1997]
or ciclids [Reinthal and Meyer, 1997
]). Similarly, if the diversification occurs in a new habitat, the selective constraints on previously acquired characters (symplesiomorphies) may be lacking and some of these characters may be lost (e.g., Hawaiian silverswords; Baldwin, 1997
). In either case, reconstructing the relationships among the taxa that have undergone an adaptive radiation may be difficult if only morphological characters are available. Phylogenetic inferences provide the starting point for testing whether patterns in morphological similarities reflect ecological (parallel evolution) or historical (shared ancestry) constraints. Although phylogenetic reconstructions are now routinely used in systematics, debate about the validity of various sources of data continues, especially whether morphological characters should be integrated when testing for phylogenetic hypotheses in groups having undergone a so- called adaptive radiation (Lucknow and Bruneau, 1997
; de Queiroz, 1999
; Lee, 1999
; Grandcolas et al., 2001
). The use of morphological characters can be problematic if most of the characters available are linked to the ecological parameter that is under study. Although molecular data are not free of limitations (e.g., evolution dictated by structural and functional constraints), they do offer a suite of characters that can be considered independent of specific morphological or life history traits.
Diversification following the invasion of a new habitat may have occurred in the moss family Splachnaceae. Various species of this family are coprophilous in a broad sense: they grow on fresh dung or other animal substrates, such as carryon (Koponen, 1990
). Among seedless land plants, the Splachnaceae are unique in that some of its species have their spores dispersed by insects and hence are entomophilous. Because it is the successful germination of the spore that determines the establishment of a new population, the proximal impact of entomophily in the Splachnaceae is on the demography of the populations. Entomophily is functionally analogous to animal seed dispersal in higher plants. Although entomophily is emblematic for the Splachnaceae (thus the common name, dung moss family), not all species share this mode of spore dispersal. Indeed, nearly half of the species are thought to be dispersed by wind (Koponen, 1978
). Entomophilous species differ from their anemophilous (wind-dispersed) congeners by a suite of gametophytic (essentially spore features) and sporophytic characters, that are all putatively linked to the spore dispersal syndromes (Koponen, 1990
). Some of these characters and their putative significance are: (1) reflexed to erect peristome teeth that allow for the insect to come into contact with the spore mass; (2) the ability of the capsule urn to shrink longitudinally and/or radially, leading to the spore mass being pushed upward; (3) similarly, the extension of the pseudocolumella, an axis subtending the actual sporangium, insures that the spore sac shrinks and the spores are pushed out (Rieth, 1957
; Koponen, 1990
; Demidova and Filin, 1994
). Similar to the floral characters for angiosperms, sporophytic (in particular peristome) characters have traditionally been considered good phylogenetic markers in mosses, and much of the classification of mosses has been based on these characters (e.g., Vitt, 1984
). Hence, it is not surprising that supraspecific taxa within the Splachnaceae have historically been defined solely on the basis of sporophytic features (e.g., Brotherus, 1924
; Koponen, 1977
, 1982a
).
The Splachnaceae currently comprise 74 species distributed among seven genera (Crosby et al., 1999
) that have been grouped in three subfamilies (Koponen, 1982a
). The subfamily Splachnoideae, characterized by a differentiated and typically inflated hypophysis, comprises the genera Aplodon, Splachnum, and Tetraplodon. These genera are distinguished by their peristomes (i.e., the teeth lining the sporangium mouth): double and fused in Splachnum, single with eight well-developed exostome teeth in Tetraplodon, and single with 8–16 rudimentary outer teeth in Aplodon. All Splachnoideae are coprophilous (here treated in a broad sense, referring to various types of animal substrates) and, except for T. paradoxus whose capsule is indehiscent, all species are entomophilous. The Voitioideae are coprophilous also, but not entomophilous. This subfamily is distinguished by the capsule lacking a differentiated line of dehiscence (cleistocarpy) and spore dispersal following disintegration of the sporangial wall. The Taylorioideae are composed of three genera. Brachymitrion and Moseniella comprise those species that lack a differentiated hypophysis and have nonsticky spores. Species of Brachymitrion have erect-incurved peristomes, whereas Moseniella is gymnostomous (i.e., the capsule lacks a peristome). Both genera are epiphytic (Koponen, 1983
) and hence are considered strictly wind-dispersed. The genus Tayloria includes both anemophilous and entomophilous taxa. It is therefore not surprising that this genus is the most polymorphic in terms of morphological and chemical characters. Koponen (1982a)
distributed the species among five subgenera (Orthodon, Cyrtodon, Eremodon, Pseudotetraplodon, and Tayloria), which were defined based on the architecture of the peristome and characteristics of the exothecial cells. As pointed out by Koponen (1982a
, p. 242) "if the structure of the peristome is regarded as the main generic character in the Splachnaceae most of Brotherus' (1924)
subgenera of Tayloria [and thus hers] would be good genera." Whether these putative genera represent monophyletic entities is, however, not yet clear. The most dubious of Koponen's subgenera are subg. Pseudotetraplodon and Eremodon. Tayloria gunnii and T. mirabilis (subg. Eremodon) share similar sporophytes (except for the peristome) with T. tasmanica (subg. Pseudotetraplodon), with a whitish, enlarged hypophysis, not known in any other species of Tayloria. Tayloria octoblepharum (subg. Pseudotetraplodon) and T. purpurascens (Eremodon) are morphologically similar, differing most conspicuously in the habit of their peristome teeth (recurved vs. more or less erect). Finally, the monophyly of the genus Tayloria itself is questionable. Subgenus Orthodon is similar in overall gestalt to Brachymitrion, differing only by the slightly differentiated hypophysis and the number of peristome teeth.
Entomophily in the Splachnaceae has attracted much interest since its discovery by Bryhn (1897)
. Previous studies have focused on the systematic and evolutionary significance of putative morphological adaptations (e.g., Koponen, 1978
, 1982a
, b
), reproductive biology (e.g., Bequaert, 1921
; Cameron and Wyatt, 1986
, 1989
), ecological physiology (Webster, 1987
; Cameron and Wyatt, 1989
), chemistry of the volatile attractants (e.g., Pyysalo et al., 1978
, 1983
; Koponen et al., 1990
), and ecological distribution of the species (e.g., Marino, 1988
, 1991
, 1997
). However, the phylogeny of the Splachnaceae and the evolutionary history of entomophily have never been formally addressed. Koponen proposed some evolutionary hypotheses, although never within an explicit phylogenetic hypothesis. She considered the genus Brachymitrion, which comprises epiphytic taxa, to be the most primitive member of the family (Koponen, 1978
). This hypothesis was based on the "basic exostome teeth number 16 in this genus, and ... the verrucate-lirate to spinate sculpturing of spores common among other mosses" (Koponen, 1983
, p. 26). The only other explicit evolutionary hypotheses formulated are that: (1) the scrobiculate spore type, which is shared by the Splachnoideae, Voitioideae, Moseniella, and various species of Tayloria (Koponen, 1978
), arose once and (2) entomophilous taxa are closely related but the anemophilous taxa are diverse and not closely related (Koponen, 1979
). These hypotheses indicate that Tayloria is polyphyletic and lead to an evolutionary scenario wherein entomophily arose once, with the Splachnoideae composing a derived taxon, having the most extreme modifications and full suite of adaptations to entomophily.
The evolution of the Splachnaceae is thus thought to be characterized by the origin of key adaptations (to entomophily). The question though is when did those key innovations arise (e.g., in the ancestor to all Splachnaceae or to a particular lineage within the family?) and when did the radiation occur (early in the evolution of the family or within a strictly entomophilous clade only)? Such questions need to be addressed within a phylogenetic framework. The present study provides the first formal phylogenetic reconstruction of the Splachnaceae and is aimed at testing the following hypotheses: (1) Brachymitrion, which has supposedly ancestral morphological features, is the most primitive member of the family from which entomophilous members were derived; (2) morphological characters traditionally used to define supra-specific taxa within the Splachnaceae are good phylogenetic indicators and hence define homogenous monophyletic entities; and (3) coprophily sensu lato, and hence entomophily, arose once within the family, and entomophilous Splachnaceae are more closely related to each other than they are to anemophilous congeners.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Cox et al., 2000
; Goffinet et al., 2001
) point toward close affinities of the Splachnaceae to the Meesiaceae (sensu Buck and Goffinet, 2000
). Three species of Leptobryum and six taxa of the Meesiaceae sensu stricto were included in these analyses. Splachnobryum, a genus excluded from the Splachnaceae and accommodated in its own family (Koponen, 1981
) has been shown to be only distantly related to the Splachnaceae (Allen and Pursell, 2000
; Goffinet and Cox, 2000
), and hence is not included in the taxon sample. The Splachnaceae are represented by 49 species. Sequences generated during the course of this study are complemented by sequences obtained previously (Cox and Hedderson, 1999
; Goffinet et al., 2001
; Goffinet and Shaw, 2002
). Voucher information for all newly sequenced taxa is presented in the Appendix (see Supplemental Data accompanying the online version of this article).
DNA extraction, PCR amplification, and sequencing
DNA was extracted from 2 to 4 operculate capsules or several stems sampled from herbarium collections. The extraction protocol followed a modification of Doyle and Doyle (1987)
as described in Goffinet et al. (1998)
. Amplification and sequencing procedures are outlined in Buck et al. (2000)
, except that sequencing products were also separated by capillary electrophoresis using the ABI (Applied Biosystems) Prism 3100 Genetic Analyzer. Nucleotide sequences were edited using Sequencher 3.1.1 (Gene Codes Corporation), entered in PAUP*version 4.0b10 for Macintosh-PPC (Swofford, 2002
), and manually aligned. Gaps needed to be included to maximize homology, primarily in the trnL intron. Insertions could typically be linked to duplication of adjacent 5' or 3' regions. In these cases, the inserted fragment was aligned in a way to minimize duplication events and sequence variation of the conserved region among taxa. If insertions occurred in two taxa in the same locations, but the two inserted fragments could be linked to distinct adjacent portions, the inserts were not aligned to each other. Aligned trnL-trnF sequences were partitioned into the trnL intron, the 3' trnL-exon, and the trnL-trnF intergenic spacer based on sequences available from GenBank.
Phylogenetic analyses
Prior to the analyses, the matrix was trimmed of the partial 5' trnL exon and the partial trnF sequences, as well as of the first 30 nucleotides (nts) of the rps4 gene and the 3' intergenic spacer. Furthermore, all indels occurring in a single taxon or those that are shared by two taxa but were invariable were excluded from the analysis. All analyses were run with Leptobryum selected as the outgroup.
Likelihood trees were sampled from the tree space using a Bayesian approach (Huelsenbeck and Ronquist, 2002
) and three heated chains and one cold chain. The data were treated as a single partition (homogeneous Bayesian) or partitioned into six character sets (heterogeneous Bayesian), namely the trnL intron, the 3' trnL exon, the trnL-trnF intergenic spacer, and all three codon positions within the rps4 gene. The optimal model of sequence evolution for each partition was identified using Modeltest 3.06 (Posada and Crandall, 1998
) based on hierarchical likelihood ratio tests. For each partition, parameters corresponding to the appropriate model were estimated. If the model proposed for a partition included 3 (TrN model) or 5 (TVM model), substitution types that cannot be implemented in MrBayes 3.0, the next highest model with nst = 6, was implemented. Parameters were unlinked across all partitions and thus allowed to vary independently. For the homogeneous search, a single tree was saved to a tree file every 50 generations (vs. every 100 generations for the heterogeneous search) for a total of 106 generations. Of the 20 001 and 10 001 trees saved for the homo- and heterogeneous search, respectively, the first 1000 (the "burnin") were ignored for determining posterior probabilities and confidence intervals for model parameters. Although the –ln likelihood of the trees sampled stabilized around its ultimate median earlier in the analysis, a conservative sampling of the trees was preferred. Posterior probabilities (PP) for nodes and estimates of the parameters for the model of molecular evolution were obtained by constructing a consensus tree of the remaining 19 001 trees in the homogeneous search and 9001 in the heterogeneous search. Posterior probability support for bipartitions was considered statistically significant when P 
0.95. The homogeneous and heterogeneous Bayesian analyses were repeated four times to test for the presence of multiple local optima. The sampled trees from all four replicates were ultimately combined and parameter and posteriors estimated based on this set of 76 004 (homogeneous) and 36 004 (heterogeneous) trees.
An heuristic search under the maximum likelihood optimality criterion was implemented in PAUP under the assumption that the evolution of the sites could be approximated by a single model (homogenous likelihood). The GTR + gamma model was defined by the following settings: substitution rates: AC = 0.976830, AG = 3.361470, AT = 0.315430, CG = 0.851400, CT = 4.500510; base frequencies: A = 0.42573, C = 0.12352, G = 0.12375, T = 0.32700; shape parameter alpha for the gamma distribution of rates: 0.32418 (with four rate categories). One hundred random taxon-addition replicates were implemented with tree bisection-reconnection (TBR) branch-swapping. The same settings were used for a bootstrap analysis consisting of 200 pseudoreplicates.
Finally, the complete data set was analyzed under the criterion of maximum parsimony (MP) under the assumption of equal weights for all characters and all transformations, using the program PAUP (v. 4b10, Swofford, 2000). The settings for the heuristic search were as follows: for each of the 200 replicates, 10 trees were kept, following NNI (Nearest-Neighbor Interchange) branch- swapping on trees constructed by random taxon-addition with the steepest descent option in effect. The resulting trees were subsequently swapped to completion using a TBR branch-swapping option. Consensus trees were constructed to reveal the agreement in topology among the optimal trees. Support for the branches was estimated by full heuristic bootstrap analyses (BP; Felsenstein, 1985
) based on 1000 pseudoreplicates and identical search strategies.
Ancestral character-state reconstruction
Empirical observations on the spore dispersal mechanism are lacking for many taxa of the Splachnaceae. Cameron and Wyatt (1986)
demonstrated, however, that the aggregation of the spores is an impediment to wind dispersal, suggesting that the ability to produce small sticky spores can be regarded as an adaptation to insect-mediated dispersal. Although sticky spores may not be a requirement for entomophily (it may simply have improved the efficiency of dispersal), no species with nonaggregated spores has been shown to be entomophilous or is known to occur on animal substrates. In contrast, all known entomophilous species produce sticky spores and grow on dung or other decaying animal material. To grow on animal remains or feces appears indeed as an imperative characteristic for entomophilous species. If the vector is coprophilous, dung and related substrates will be the primary habitat visited by the insects and thus will be the most likely substrate upon which spores are deposited. Hence, transformations in spore stickiness and habitat can be used to approximate the evolution of entomophily.
Based on the characteristics of the spores, species were scored for their spore dispersal syndrome. We have not distinguished between species growing on dung vs. those that grow on decaying animals, blood impregnated soils, bones, etc. We make the assumption that diversification in habitat, and hence niche partitioning, took place after fundamental changes in morphology and chemistry made the attraction of insects possible. Thus, for the remainder of the text, the term coprophily is used in the widest sense possible, referring to all types of animal substrates colonized by entomophilous Splachnaceae. Although they may be reports of some entomophilous Splachnaceae to grow on humus (D. Vitt, University of Southern Illinois, personal communication), our own observations from southern South America and revision of Australia species (Goffinet, in press
) suggests that such potential deviations are rare. Furthermore, it should be noted that Splachnaceae are invariably collected when fertile and hence long (i.e., 1–2 yr) after the colonization of the substrate. Any visual or olfactory traces of the animal substrate are likely to have disappeared by then, especially in the case of dung. The lack of evidence of coprophily for a sample should therefore not be treated as evidence for the species to be polymorphic in its habitat requirement. We have treated all taxa as monomorphic with regard to their habitat preference.
Transformations between entomophily and anemophily were reconstructed using MacClade version 4.05 (Maddison and Maddison, 2002
), under both the ACCTRAN and the DELTRAN assumptions.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
The combined matrix comprised 1296 characters, of which 177 were trimmed at the beginning and end of the loci. Of the remaining 1119 characters, 187 were excluded due to alignment ambiguity or because these regions represent insertions characteristic of a single taxon. The final matrix comprised 932 characters. Parsimony informative characters represent approximately 22% of the sites and are found mainly in the trnL intron and the third codon positions of the rps4 coding region (Table 1).
|
|
|
All analyses converge toward a similar phylogenetic scenario (Figs. 1, 2), wherein (1) monophyly of the Splachnaceae is ambiguous, (2) the Voitioideae are nested within the Splachnoideae, (3) the Taylorioideae are monophyletic, (4) Brachymitrion is sister to Tayloria subg. Orthodon and nested within Tayloria, (5) all subgenera of Tayloria, except for subg. Orthodon, are poly- or paraphyletic, and (6) species of Tayloria (including Brachymitrion) compose three lineages. The main source of topological incongruence among optimal trees retrieved under the different criteria are the positions of the monospecific genera Aplodon (Splachnoideae) and Neomeesia (Meesiaceae). Under maximum parsimony, Aplodon is resolved sister to all other Splachnoideae, a position also shared with many optimal trees recovered under the heterogeneous Bayesian search (Fig. 1). In the most likely tree (Fig. 2), Aplodon is nested within the outgroup taxon Meesiaceae. None of these alternative relationships is well supported: bootstrap proportion under maximum parsimony, BPMP = 62%; bootstrap proportion under maximum likelihood, BPML < 50%; posterior probabilities under a heterogeneous Bayesian search, PPHE = 53%; posterior probabilities under a homogeneous Bayesian search, PPHO < 50%. The monophyly of the Splachnaceae is otherwise ambiguous under maximum parsimony because of the position of Neomeesia, which is resolved as a sister to the Taylorioideae (Fig. 1A) vs. sister to the Splachnaceae and Meesiaceae (Figs. 1B, 2). Here too, none of the alternative hypotheses is well supported (BPMP < 50%; BPM < 50%; PPHE = 50%; PPHO < 50%).
|
|
Members of the Taylorioideae form a monophyletic group (BPMP = 62%; BPML = 84%; PPHE = 100%; PPHO = 100%). The genus Brachymitrion is monophyletic (BPMP = 72%; BPML = 78%; PPHE = 99%; PPHO = 93%) and is nested within the genus Tayloria in all optimal reconstructions. Brachymitrion and Tayloria subg. Orthodon form a clade that is characterized by high posterior probabilities but low MP and ML bootstrap support (BPMP = < 50%; BPML < 50%; PPHE = 97%; PPHO = 93%). The only subgenus of Tayloria to be resolved as a monophyletic lineage in optimal topologies, albeit always with no support (BPMP < 50%; BPM < 50%; PPHE = 51%; PPHO = 54%), is subgenus Orthodon. The subgenera Cyrtodon, Pseudotetraplodon, Eremodon, and Tayloria are poly- or paraphyletic (Fig. 2). Species of these subgenera are, however, consistently arranged in two lineages. The first lineage comprises all species of subg. Tayloria, plus one member of subg. Cyrtodon and subg. Pseudotetraplodon. Monophyly of this clade is well supported (BPMP = 100%; BPML = 100%; PPHE = 100%; PPHO = 100%). The second lineage is similarly well supported (BPMP = 99%; BPML = 100%; PPHE = 100%; PPHO = 100%) and is composed of members of subg. Pseudotetraplodon, all members of Eremodon, and one species of subg. Cyrtodon (Fig. 2). All members of this last clade share a unique single nucleotide insertion in the trnL intron. Note that in all three lineages within the Taylorioideae, members of subg. Cyrtodon occupy basal positions (Fig. 2).
Ancestral character-state reconstruction
Transformations between a humicolous and coprophilous habitat were inferred under parsimony using the tree obtained under maximum likelihood (Fig. 2), with the exception that Aplodon was constrained to form a sister group to the remainder of the Splachnoideae. The inclusion of Aplodon in the Splachnaceae vs. the Meesiaceae is congruent with it sharing the deletion in the trnL intron with other Splachnaceae and with it morphological characters (erect capsule, mitrate calyptrae). Based on this topology, shifts in habitat preferences must have arisen six times (Fig. 3). Under DELTRAN, all changes inferred are transformations from wind to insect dispersal, whereas under ACCTRAN, those changes number only four and are followed by two reversals to anemophily. Although coprophily is likely an imperative condition for entomophily in the Splachnaceae, not all coprophilous mosses are entomophilous. Tayloria grandis and species of Voitia are coprophilous, but their sporangia lack a differentiated line of dehiscence, and spores are dispersed abiotically as the capsule wall is ruptured or disintegrates. We can, however, assume that these taxa arose via reduction from an entomophilous ancestor, because an independent shift to coprophily without adaptations to entomophily is unlikely, as no obvious gains in fitness would derive from such specialization. Hence, our inferences regarding the evolution of coprophily can be extrapolated to the history of transformations in spore dispersal strategies. Consequently, entomophily would have arisen at least six or four times, followed by two or no reversals.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Koponen, 1977
, 1982a
). Similarly, the family itself has been identified by the differentiated neck, which Crum and Anderson (1981)
referred to as an outstanding characteristic of the family. Most of the taxonomic concepts based on morphology are not supported by inferences from analyses of variation in nucleotide sequences. Taxa such as the Splachnoideae, Tayloria, and four of the five subgenera of Tayloria are resolved as para- or polyphyletic entities, the monophyly of Tetraplodon is doubtful, and even the monophyly of the family itself is not unambiguously supported by cpDNA data. Assuming that the phylogenetic signal emanating from the variation in cpDNA sequence data is accurate, these observations reveal that morphological characters linked with dispersal syndromes carry, at least based on our current interpretation of these characters, limited phylogenetic signal. In fact, these observations suggest that the transformations between characters states have occurred more than once and are possibly reversible. The lack of support from the DNA data for the monophyly of the Splachnaceae and for the relationships between major lineages within the family, the absence of morphological synapomorphies for most major lineages within the Splachnaceae, combined with strong molecular support for the monophyly of these lineages, are all features of a phylogeny consistent with a rapid diversification (Givnish and Sytsma, 1997
), which, in the context of a transition to a coprophilous habitat is best described as an adaptive radiation.
Recent phylogenetic studies (Cox and Hedderson, 1999
; Cox et al., 2000
) have revealed that the Splachnaceae are closely related to the Meesiaceae (Bryales) rather than to the Funariales as previously proposed (e.g., Brotherus, 1924
; Vitt, 1984
). Members of the Meesiaceae share with the Splachnaceae general habitat preferences, namely, moist substrates in peatlands and temperate to boreal forests. Unlike the Meesiaceae, the capsule of all Splachnaceae is erect and the calyptra is mitrate, as opposed to a cucullate calyptra covering a curved capsule in the Meesiaceae. Transformations in these characters are likely linked, but these resulting states may be the only features that unambiguously differentiated these two families. Under a scenario wherein the Meesiaceae are plesiotypic these character states may suffice to support the monophyly of the Splachnaceae. Although the monophyly of the Splachnaceae is not well supported by the cpDNA data, support for its polyphyly is also basically lacking. All Splachnaceae (including Aplodon) share a nine-nucleotide deletion in the trnL intron. This deletion is absent in Leptobryum and in the Meesiaceae, except in Meesia muelleri. In the latter taxon, the deletion is larger (i.e., 19 nucleotides), and it is possible that the overlap between the deletions in this taxon and in the Splachnaceae is indicative of a parallel loss. Hence, the Splachnaceae are best retained as a monophyletic family at present.
The Splachnaceae have historically been subdivided into three subfamilies (e.g., Brotherus, 1924
). Brotherus circumscribed the Splachnoideae with the genera Aplodon, Splachnum, and Tetraplodon. The subfamily is defined by an operculate capsule having a well-differentiated hypophysis. As pointed out by Koponen (1982a)
, some austral species of Tayloria (e.g., T. gunnii) that had originally been described within Tetraplodon or Splachnum also have a wide hypophysis. Their transfer to Tayloria was justified based on leaf and peristome characters (e.g., Willis, 1950
). This hypothesis is confirmed here. Consequently, the widening of the hypophysis is homoplastic within the Splachnaceae, and the character hypophysis wider than urn cannot be used as a diagnostic feature of the Splachnoideae. In fact, the subfamily cannot be defined by any one morphological character.
Furthermore, the Splachnoideae sensu Brotherus (1924)
or Koponen (1982a)
do not compose a monophyletic group. Goffinet and Shaw (2002)
have recently demonstrated, based on phylogenetic inferences from variation in the two chloroplast loci used here, that the Voitioideae sensu stricto are nested within the Splachnoideae. Although Voitia may not be nested within Tetraplodon as suggested by Goffinet and Shaw (2002)
, Voitia does appear to share at least a common ancestor with Tetraplodon. The genus Voitia, the sole member of the Voitioideae, comprises two species and is characterized by cleistocarpous capsules, no peristome, no hypophysis, and an enlarged calyptra that covers the capsule completely. Our phylogenetic inferences from cpDNA data suggest that Voitia arose through reduction from a typical Splachnoid ancestor. Such a trend has recently been proposed to account for the affinities of Voitia grandis (Long, 1999
), which shares several sporophytic characters with Voitia but is clearly nested within Tayloria, and in particular the Tayloria subg. Tayloria group (Goffinet and Shaw, 2002
). Furthermore, the loss of dehiscence has also occurred within Tetraplodon; T. paradoxus lacks a differentiated operculum, but has a well-developed peristome. In the absence of a dehiscent operculum, the peristome of T. paradoxus is nonfunctional. The fact that it is still present (unlike in Voitia) suggests that cleistocarpy in T. paradoxus is of recent origin.
Relationships within the Splachnoideae are fairly robust, except for the affinities of the genus Aplodon. Indeed, even when resolved within the Splachnaceae its sister-group relationship to the remaining Splachnoideae is not supported. Koponen (1982a)
and Brotherus (1924)
placed Aplodon within the Splachnoideae on the basis of the inflated hypophysis. The genus is otherwise characterized by a rudimentary peristome of eight or 16 teeth and a hyaline seta, two characters that appear to have no phylogenetic significance. The sole species of the genus grows on animal remains in the arctic (Steere, 1973
), and exclusively on musk-oxen dung on Devon Island (Vitt, 1975
). Further studies are needed to identify the affinities of Aplodon within the Splachnaceae.
The genus Splachnum is best known to bryologists for the umbrella-like hypophysis of S. luteum and S. rubrum. Not all species share such an exaggerated degree of expansion. In fact, the genus is best defined by the presence of an endostome (inner peristome) that is fused to the exostome (outer peristome; Koponen, 1982a
; Schwartz, 1994
). Fused peristomes serve as a synapomorphy for the genus, whose monophyly is well supported by sequence data. The relationships within Splachnum are fully resolved and all significantly supported. The species with an expanded umbrella-shaped hypophysis (i.e., S. luteum and S. rubrum, and to a lesser degree, S. melanocaulon) compose a derived clade within the genus. Species with comparatively narrow hypophyses (e.g., S. pensylvanicum and S. sphaericum) do not, however, occupy a basal position, which is held by S. ampullaceum, a taxon characterized by hypophysis that is large above and tapered below. Hence, we have no evidence for a gradual amplification of the sporophyte in the genus Splachnum.
Tetraplodon differs from Splachnum primarily by the absence of an endostome. The genus is widely distributed and occurs on all continents except in Australia and New Zealand (Koponen, 1982c
; Goffinet, in press
). The genus has not been monographed except for regional treatments (e.g., Frisvoll, 1978
). Three taxa have been described from the Southern Hemisphere: T. lamii, endemic to New Guinea; T. itatiaiae, known only from Brazil; and T. fuegianus, a species endemic to Patagonia. Koponen (1982c
, p. 87) regarded the former as a form of T. mnioides "with an exceptionally large sporophyte," and the latter as a synonym of T. mnioides. Our phylogenetic inferences are not congruent with these concepts because all three taxa from South America and New Guinea compose a monophyletic lineage that is not closely allied to T. mnioides. How these taxa, and in particular the neotropical species, differ from one another in morphology is not clear and warrants critical study.
The genus Tayloria comprises most of the species and hence, most of the morphological diversity in the Splachnaceae. Brotherus arranged the species into five subgenera, which were later complemented by a sixth one (i.e., subg. Pseudotetraplodon) erected by Koponen (1977)
. These subgenera are defined by combinations of characters pertaining exclusively to the sporophyte, including the architecture of the peristome and the morphology of the exothecial cells (Koponen, 1978
). The genus Brachymitrion was resurrected from the subgeneric level by Koponen (1977)
on the grounds that it represented the most primitive group of the family, lacking most if not all sporophytic modifications associated with entomophily. Of these six supraspecific taxa within the Taylorioideae, only Brachymitrion and T. subg. Orthodon appear to compose monophyletic entities. The remaining four subgenera of Tayloria are resolved as either para- or polyphyletic. The argument made by Koponen (1982a
, p. 242) that "if the structure of the peristome is regarded as the main generic character in the Splachnaceae, then most of Brotherus' (1924)
subgenera of Tayloria are good genera" clearly finds no endorsement from our phylogenetic inferences, indicating that these characters offer little phylogenetic signal at this systematic level. In fact, the Taylorioideae are resolved into three major lineages (Fig. 2), none of which can be defined by a single character that is not subsequently lost or modified.
In each of these lineages, one or two members of the traditional subg. Cyrtodon occupy a basal position, albeit poorly supported. The first lineage is centered on Brachymitrion and subg. Orthodon. The probability that all these taxa share a common ancestor is very high, even if the number of characters supporting this hypothesis is small. Brachymitrion and subg. Orthodon differ in the degree of differentiation of the hypophysis (none in Brachymitrion vs. slight in subg. Orthodon), the thickening of the exothecial cells (differentiated thickening of anticlinal and periclinal walls of suboral exothecial cells in subg. Orthodon) and the architecture of the peristome (16 teeth in Brachymitrion vs. eight in subg. Orthodon; Koponen and Weber, 1972
; Koponen, 1982a
). Both taxa comprise only anemophilous species, which are epiphytic or humicolous in montane forests throughout the tropics. Whereas the species of Brachymitrion compose a well-defined monophyletic group, the monophyly of subg. Orthodon of Tayloria is not supported here. Phylogenetic evidence for a basal position of Brachymitrion in the evolution of the Splachnaceae (Koponen, 1983
) is completely lacking. In fact, our inferences point, although with little confidence, toward a derived position of the genus. Given the basal position of T. pseudoalpicola (subg. Cyrtodon) and the possible paraphyly of subg. Orthodon, Brachymitrion, with its overall plesiomorphic characters (e.g., no hypophysis, no shrinking of the capsule upon drying), may indeed be best regarded as a reduced member of the Taylorioideae.
The second well-supported clade within the Taylorioideae comprises all representatives of subg. Tayloria with T. scabriseta (subg. Pseudotetraplodon) nested within it, and T. hornschuchii (subg. Cyrtodon) sister to this composite clade. The geographic distribution of this clade is centered primarily in the temperate zone of the Northern Hemisphere and its ecological counterparts at high altitudes in the Neotropics (T. scabriseta; Churchill and Linares, 1995
) and Africa (T. kilimandscharica; De Sloover, 1973
). Tayloria subg. Tayloria is defined by a peristome of 16 or 32 long and reflexed teeth. Inclusion of T. scabriseta (subg. Pseudotetraplodon) in this group is inconsistent with such a morphological concept for the group, because T. scabriseta has eight short peristome teeth. Tayloria serratavar. tenuis, a circumboreal taxon (Nyholm, 1989
as T. tenuis) and T. scabriseta, an Andean species of Central America and northern South America (Churchill and Linares, 1995
), are the only entomophilous and coprophilous taxa in this group (Koponen, 1978
); surprisingly, they are not sister taxa. Anemophilous members of this clade do, however, bear features reminiscent of their entomophilous relatives, such as reflexed peristome teeth, capsules shrinking upon dehydration (see De Sloover, 1973
), and small scrobiculate spores (Koponen, 1978
).
The final Taylorioid clade includes, in addition to two members of subg. Cyrtodon, namely T. lingulata and T. froelichiana, the remaining taxa of subg. Pseudotetraplodon and all representatives of subg. Eremodon. Except for the former two species, all members of this clade are coprophilous and are therefore assumed to be entomophilous. Subgenus Pseudotetraplodon was erected by Koponen (1977)
to accommodate those taxa of subg. Eremodon sensu Brotherus (1924)
, whose eight peristome teeth are reflexed rather than erect. Clearly, such segregation is not congruent with phylogenetic inferences from cpDNA sequence data. Members of these two subgenera have a well modified hypophysis that is either long and purple or brown or wide and whitish. Neither of these features defines a homogenous clade within this group. Among the species with whitish hypophyses, the South American T. magellanica and T. mirabilis are likely sister taxa, whereas the Tasmanian T. tasmanica and T. gunnii do not share a common ancestor (Fig. 2). Similarly, T. dubyi, T. purpurascens, T. octoblepharum, and T. stenophysata all have a long, dark-purple hypophysis and capsule, that is unlike that of other members of this clade, yet these species are not the closest relative of each other. Clearly, transformations of sporophytic characters putatively linked to spore dispersal syndromes offer less robust phylogenetic signal than previously assumed. The alternation of incurved and recurved teeth, narrow and wide, or white and purple hypophyses suggests that transformations in these characters, regardless of their direction, are not unique and irreversible.
Systematic implications
Variation in nucleotide sequences of the two chloroplast loci sampled here, provides significant evidence that four of the six supraspecific taxa recognized by Brotherus (1924)
and Koponen (1977
, 1982a)
within the Taylorioideae, namely subg. Cyrtodon, Eremodon, Pseudotetraplodon, and Tayloria are polyphyletic. For the classification to reflect monophyletic groupings within this subfamily would require either raising three of the clades within Tayloria (subg. Orthodon, the subg. Tayloria group, and the subg. Eremodon group) to genus level, equivalent to that of Brachymitrion, or to recognize Brachymitrion as a subgenus of Tayloria, along with the aforementioned groups. From a practical point of view, recognizing any of these clades, whether at the generic or subgeneric level, would require the identification of diagnostic morphological characters. The core groups for two of these clades (i.e., the subg. Tayloria group and the subg. Eremodon group) are heterogeneous, and this heterogeneity is further accentuated by the inclusion of taxa formally attributed to subg. Cyrtodon. Members of this subgenus are resolved in basal positions of all three main lineages of the Taylorioideae. None of these clades within this subfamily can currently be diagnosed by morphological synapomorphies that would be shared by all members of the clade. In fact, the same applies to the Splachnoideae and the Taylorioideae. The heterogeneity of the Taylorioideae is accentuated by the presence of the genus Moseniella, which comprises two species endemic to Brazil (Koponen, 1983
). It shares with at least some members of the Brachymitrion-subg. Orthodon clade, serrate leaves, hairy calyptrae, and a short neck (i.e., no differentiated hypophysis), but differs in lacking a peristome (Koponen, 1983
). Moseniella has not been collected since its description (Brotherus, 1918
) and hence could not be included in this study. It is possible that it represents yet another product of morphological reduction within the Taylorioideae. Although we cannot offer a diagnostic feature for the Taylorioideae or Tayloria sensu lato, we propose, based on the phylogenetic evidence presented here, to follow Brotherus (1924)
, who treated Brachymitrion as a subgenus of Tayloria. Such systematic hypothesis requires a single new combination, namely for Brachymitrion immersum Goffinet (The Bryologist 102: 108 [1999], type in NY), which becomes Tayloria immersa (Goffinet) Goffinet, Shaw & Cox comb. nov. The genus Moseniella is tentatively retained as phylogenetically distinct.
The Voitioideae have been defined by a combination of characters: a cleistocarpous or indehiscent capsule, no differentiated hypophysis, and no peristome. None of these features alone is diagnostic of Voitia: cleistocarpy also occurs in Tetraplodon and Tayloria (Goffinet and Shaw, 2002
), gymnostomous capsules also characterize Moseniella, and poorly or undifferentiated hypophyses are known from Brachymitrion and Tayloria subg. Orthodon. The recognition of the Voitioideae Broth. should thus be abandoned, and the genus Voitia should be included in the Splachnoideae.
Consequently, we regard the family as composed of two main natural lineages: the Splachnoideae (i.e., Aplodon, Tetraplodon, Splachnum, and Voitia), which comprises exclusively coprophilous taxa whose spores are dispersed by insects, and the Taylorioideae, which combines anemophilous and entomophilous taxa.
Evolution of entomophily
The reconstruction of shifts from anemophily to entomophily suggests that insect-mediated spore dispersal has arisen more than once in the Splachnaceae. If parallel evolution is favored, six origins of entomophily should be invoked, and no reversal would have taken place. The transition to entomophily appears irreversible. By contrast, if early transitions are preferred, such shift would have arisen three times, and three of these would have been followed by reversal to anemophily. It is a priori impossible to favor one of these scenarios. Reversals from a specialist to a generalist syndrome cannot be dismissed. Among angiosperms, shifts between insect- and wind-pollination syndromes are not rare (Culley et al., 2002
). A reversal to anemophily may be driven by demographic changes in the biotic vector, by changes in the availability of the substrate, or more generally by the severity of the habitat constraints (Cox, 1991
; Culley et al., 2002
). Furthermore, entomophily is not determined by a single character transformation. In fact, several features seem to be correlated with insect being used as vectors: the hypophysis is colored or inflated or both, volatile compounds mimicking decaying animal matter are emitted, the spores are sticky, the capsule contracts upon drying, the teeth are either short or recurved (Koponen, 1990
). All these may be required for efficient dispersal of the spores by insects, insuring proper attraction of the insect, adequate exposure of the spore mass to the insect, and effective dispersal of multiple individuals per dispersal event. The loss of any of these may negatively impact the efficiency of the syndrome and trigger a reversal to wind dispersal. Whether such shifts did indeed take place within the Taylorioideae remains, however, ambiguous.
Conclusions
Insect-mediated spore dispersal is known among land plants only for the moss family Splachnaceae. Like entomophilous flowering plants that use insects to disperse pollen, these Splachnaceae are characterized by a suite of modifications that insure both proper attraction of the insect using olfactory and visual means and attachment of the spore to the insect. All entomophilous Splachnaceae are coprophilous, growing on dung, carrion, bones, and even blood-soaked soil. The classification of the family is built almost exclusively on characters of the sporophyte. Phylogenetic inferences based on sequence variation of two chloroplast loci suggests that most systematic concepts within the family are incongruent with a monophyletic taxon concept and that entomophily has arisen independently in at least three lineages.
|
FOOTNOTES |
|---|
4 E-mail: goffinet{at}uconn.edu
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Baldwin B. G. 1997 Adaptive radiation of the Hawaiian silversword alliance: congruence and conflict of phylogenetic evidence from molecular and non-molecular investigations. In T. Givnish and K. Sytsma [eds.], Molecular evolution and adaptive radiations, 103–128. Cambridge University Press, Cambridge, UK
Bequaert J. 1921 On the dispersal by flies of the spores of certain mosses of the family Splachnaceae. Bryologist 24: 1-4
Brotherus V. F. 1918 Moseniella, un nouveau genre des mousses du Brésil. Arkiv för Botanik utgivet av K. Svenska Vetenskapsakademien 15: 1-3
Brotherus V. F. 1924 Splachnaceae. In A. Engler [ed.], Die natürlichen Pflanzenfamilien, 2nd ed., vol. 10, 333–343. Duncker & Humblot, Berlin, Germany
Bryhn N. 1897 Beobachtungen über das Ausstreuen der Sporen bei den Splachnaceen. Biologisches Centralblatt 17: 48-55
Buck W. R. B. Goffinet 2000 Morphology and classification of mosses. In A. J. Shaw and B. Goffinet [eds.], Bryophyte biology, 71– 123. Cambridge University Press, New York, New York, USA
Buck W. R. B. Goffinet A. J. Shaw 2000 Testing morphological concepts of orders of pleurocarpous mosses (Bryophyta) using phylogenetic reconstructions based on trnL-trnF and rps4 sequences. Molecular Phylogenetics and Evolution 16: 180-198[CrossRef][Web of Science][Medline]
Cameron R. G. R. Wyatt 1986 Substrate restriction in entomophilous Splachnaceae: the role of spore dispersal. Bryologist 89: 279-284[CrossRef][Web of Science]
Cameron R. G. R. Wyatt 1989 Substrate restriction in entomophilous Splachnaceae: II. Effects of hydrogen ion concentration on establishment of gametophytes. Bryologist 92: 397-404[CrossRef][Web of Science]
Churchill S. P. E. L. Linares 1995 Prodromus Bryologiae Novo- Granatensis. Introducción a la flora de musgos de Colombia. Parte 2. Grimmiaceae a Trachypodiaceae. Instituto de Ciencias Naturales—Museo de Historia Natural Bibliotheca José Jeronimo Triana 12: 455-924
Cox C. J. B. Goffinet A. E. Newton A. J. Shaw T. A. J. Hedderson 2000 Phylogenetic relationships among the diplolepideous-alternate mosses (Bryidae) inferred from nuclear and chloroplast DNA sequences. Bryologist 103: 224-241[CrossRef][Web of Science]
Cox C. J. T. A. Hedderson 1999 Phylogenetic analysis among the ciliate arthrodontous mosses: evidence from nuclear and chloroplast DNA nucleotide sequences. Plant Systematic and Evolution 215: 119-139[CrossRef]
Cox P. A. 1991 Abiotic pollination: an evolutionary escape for animal pollinated angiosperms. Philosophical Transaction of the Royal Society of London, Series B 333: 217-224[CrossRef]
Crosby M. R. R. E. Magill B. Allen S. He 1999 A checklist of the mosses. Missouri Botanical Garden, St. Louis, Missouri, USA
Crum H. A. L. E. Anderson 1981 Mosses of eastern North America. Columbia University Press, New York, New York, USA
Culley T. M. S. G. Weller A. K. Sakai 2002 Evolution of wind pollination in angiosperms. Trends in Ecology and Evolution 17: 361-369[CrossRef]
Demidova E. E. V. R. Filin 1994 False columella and spore release in Tetraplodon angustatus (Hedw.) Bruch et Schimp. in B.S.G. and T. mnioides (Hedw.) Bruch et Schimp. in B.S.G. (Musci: Splachnaceae). Arctoa 3: 1-6
de Queiroz A. 1999 Do image-forming eyes promote evolutionary diversification?. Evolution 53: 1654-1664[CrossRef][Web of Science]
De Sloover J. L. 1973 Note de bryologie africaine. I. Brachydontium, Atractylocarpus, Amphidium, Rhabdoweisia, Tayloria, Rhacocarpus, Trachypodopsis. Bulletin du Jardin Botanique National de Belgique 43: 333-348[CrossRef]
Doyle J. J. J. L. Doyle 1987 A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11-15
Felsenstein J. 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791[CrossRef][Web of Science]
Frisvoll A. A. 1978 The genus Tetraplodon in Norway. A taxonomic revision. Lindbergia 4: 225-246
Givnish T. K. Sytsma 1997 Molecular evolution and adaptive radiations. Cambridge University Press, Cambridge, UK
Goffinet B. In press Splachnaceae. Flora of Australia 51
Goffinet B. R. J. Bayer D. H. Vitt 1998 Circumscription and phylogeny of the Orthotrichales (Bryopsida) inferred from rbcL sequence data. American Journal of Botany 85: 1324-1337
Goffinet B. C. Cox 2000 Phylogenetic relationships among basal- most arthrodontous mosses with special emphasis on the evolutionary significance of the Funariineae. Bryologist 103: 212-223[CrossRef][Web of Science]
Goffinet B. C. J. Cox A. J. Shaw T. A. Hedderson 2001 The Bryophyta (Mosses): systematic and evolutionary inferences from a rps4 gene (cpDNA) phylogeny (cpDNA). Annals of Botany 87: 191-208
Goffinet B. A. J. Shaw 2002 Independent origin of cleistocarpy in the Splachnaceae: analyses of cpDNA sequences reveals polyphyly of the Voitioideae. Systematic Botany 27: 203-208[Web of Science]
Grandcolas P. P. Deleporte L. Desutter-Grandcolas C. Daugeron 2001 Phylogenetics and ecology: as many characters as possible should be included in the cladistic analysis. Cladistics 17: 107-110
Huelsenbeck J. P. F. Ronquist 2002 MrBayes: Bayesian inference of phylogeny, version 3.0. Website: http://brahms.ucsd.edu/software.html
Koponen A. 1977 Tayloria subg. Pseudotetraplodon, subg. nov., and new combinations in Brachymitrion, Moseniella, and Tayloria (Splachnaceae, Musci). Annales Botanici Fennici 14: 193-196[Web of Science]
Koponen A. 1978 The peristome and spores in Splachnaceae and their evolutionary and systematic significance. Bryophytorum Bibliotheca 13: 535-567
Koponen A. 1979 Entomophily and the classification of Splachnaceae (Musci). Abstracta Botanica 5: (Supplement 3) 61
Koponen A. 1981 Splachnobryaceae, a new moss family. Annales Botanici Fennici 18: 123-132[Web of Science]
Koponen A. 1982a The generic classification of the Splachnaceae. Beihefte zur Nova Hedwigia 71: 237-245
Koponen A. 1982b On the structure and function of the peristome in the Splachnaceae. Journal of the Hattori Botanical Laboratory 53: 73-98
Koponen A. 1982c The family Splachnaceae in Australia and the Pacific. Journal of the Hattori Botanical Laboratory 52: 87-91
Koponen A. 1983 Studies on the generic concept in the classification of the moss family Splachnaceae. Publications from the Department of Botany, University of Helsinki 11: 1-48
Koponen A. 1990 Entomophily in the Splachnaceae. Botanical Journal of the Linnean Society 104: 115-127
Koponen A. T. Koponen H. Pyysalo K. Himberg P. Mansikkamäki 1990 Composition of volatile compounds in Splachnaceae. In H. D. Zinsmeister and R. Mues [eds.], Bryophytes. Their chemistry and chemical taxonomy, 449–460. Claredon Press, Oxford, UK
Koponen T. W. A. Weber 1972 A revision of the African Tayloriae (Splachnaceae) including Bryomnium. Annales Botanici Fennici 9: 126-134
Lee M. S. Y. 1999 Circularity, evolution, systematics ... and circularity. Journal of Molecular Evolution 12: 724-734
Long D. 1999 Voitia grandis (Splachnaceae, Bryopsdia), a new species from Yunnan, western China, with extensions of range for V. nivalis in eastern Asia. Bryobrothera 5: 133-136
Luknow M. A. Bruneau 1997 Circularity and independence in phylogenetic tests of ecological hypotheses. Cladistics 13: 145-151[CrossRef][Web of Science]
Maddison D. R. W. P. Maddison 2002 MacClade4, version 4.05. Sinauer Associates, Sunderland, Massachusetts, USA
Marino P. 1988 Coexistence on divided habitats: mosses in the family Splachnaceae. Annales Zoologici Fennici 25: 89-98
Marino P. 1991 The influence of varying degrees of spore aggregation on the coexistence of the mosses Splachnum ampullaceum and S. luteum: a simulation study. Ecological Modelling 58: 333-345[CrossRef]
Marino P. 1997 Competition, dispersal and coexistence of Splachnaceae in patchy habitats. Advances in Bryology 6: 241-263
Nyholm E. 1989 Illustrated flora of Nordic mosses. Fasc. 2. Pottiaceae— Splachnaceae—Schistostegaceae. Nordic Bryological Society, Copenhagen and Lund
Posada D. K. A. Crandall 1998 MODELTEST: testing the model of DNA substitution. Bioinformatics 14: 817-818
Pyysalo H. A. Koponen T. Koponen 1978 Studies on entomophily in the Splachnaceae (Musci). I. Volatile compounds in the sporophyte. Annales Botanici Fennici 15: 293-296[Web of Science]
Pyysalo H. A. Koponen T. Koponen 1983 Studies on entomophily in the Splachnaceae (Musci). II. Volatile compounds in the hypophysis. Annales Botanici Fennici 20: 335-338[Web of Science]
Reinthal P. N. A. Meyer 1997 Molecular phylogenetic tests of speciation models in Lake Malawi cichlid fishes. In T. Givnish and K. Sytsma [eds.], Molecular evolution and adaptive radiations, 375–390. Cambridge University Press, Cambridge, UK
Rieth A. 1957 Über die Sporenaustreuung bei einer Splachnacee. Flora 144: 647-654
Schwartz O. M. 1994 The development of the peristome-forming layers in the Funariaceae. International Journal of Plants Sciences 155: 640-657[CrossRef]
Steere W. C. 1973 Observations on the genus Aplodon (Musci: Splachnaceae). Bryologist 76: 347-355[CrossRef]
Swofford D. L. 2002 PAUP*version 4.0b10: phylogenetic analysis using parsimony (and other methods). Sinauer Associates, Sunderland, Massachusetts, USA
Taylor E. B. J. D. McPhail D. Schluter 1997 History of ecological selection in sticklebacks: uniting experimental and phylogenetic approaches. In T. Givnish and K. Sytsma [eds.], Molecular evolution and adaptive radiations, 511–534. Cambridge University Press, Cambridge, UK
Vitt D. H. 1975 A key and annotated synopsis of the mosses of the northern lowlands of Devon Island, NWT, Canada. Canadian Journal of Botany 53: 2158-2197
Vitt D. H. 1984 Classification of the Bryopsida. In R. M. Schuster [ed.], New manual of bryology, vol. 2, 696–759. Hattori Botanical Laboratory, Nichinan, Japan
Webster H. J. 1987 Elemental analysis of dung mosses (Splachnaceae) and their substrates. Memoirs of the New York Botanical Garden 45: 171-178
Willis J. H. 1950 The chequered history of two Tasmanian mosses. Victorian Naturalist 67: 30-35
This article has been cited by other articles:
|
|
 |
|
  A. Vanderpoorten and B. Goffinet Mapping Uncertainty and Phylogenetic Uncertainty in Ancestral Character State Reconstruction: An Example in the Moss Genus Brachytheciastrum Syst Biol, December 1, 2006; 55(6): 957 - 971. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
