Molecular systematics, historical ecology, and phylogeography of Halimeda (Bryopsidales)

Abstract

Halimeda (Bryopsidales), a genus of calcified, segmented green seaweeds, abounds in reefs and lagoons throughout the tropics. To investigate phylogenetic, phylogeographic, and historic ecological relationships of the genus, the nuclear rDNA including the SSU and both ITS regions were sequenced. A maximum likelihood tree revealed the following: (1) there were anatomical and morphological synapomorphies for five well-supported lineages; (2) the last common ancestor of one lineage invaded sandy substrata; those of two other lineages established in wave-affected habitats, whereas the cenancestor of the remaining two lineages occupied sheltered rocky slopes. Yet, several species adapted to new habitats subsequently, resulting in several cases of convergence; (3) all lineages separated into Atlantic and Indo-Pacific daughters, likely resulting from the rise of the Panamanian Isthmus. Each daughter pair gave rise to additional convergent species in similar habitats in different oceans; (4) Halimeda opuntia, the only monophyletic pantropical species detected so far, dispersed from the Indo-Pacific into the Atlantic well after the closure event; (5) minor SSU-sequence differences across species and phylogeographic patterns of vicariance indicated a relatively recent diversification of the extant diversity. Cretaceous and Early Tertiary fossil look-alikes of modern species must then have resulted from iterative convergence.

Introduction

The segmented, calcified thalli of the seaweed genus Halimeda (Lamouroux, 1812) are among the most phenotypically complex within the bryopsidalean algae (Vroom et al., 1998). As in other members of this order (sensu Van den Hoek et al., 1995), thalli consist of branched siphons without cross walls, permitting unimpeded migration of protoplasmic components across the thallus superstructure (Drew and Abel, 1990; Littler and Littler, 1999). Segments are composed of an inner medulla and a peripheral cortical layer. Medullary siphons string the segments together or branch into the cortex where they rebranch into layers of short, often inflated siphons called utricles. Peripheral utricles, which are always inflated, adhere and enclose the segment's intersiphonal spaces (Barton, 1901). There, calcium carbonate precipitates as aragonite (Borowitzka and Larkum, 1977). Some medullary siphons surface in weakly calcified regions of the segment's distal perimeter where they adhere and may fuse. Their tips can develop new segments (Hay et al., 1988), secondary holdfasts (Hillis-Colinvaux, 1980; Walters and Smith, 1994) or gametangia (Gepp, 1904; Kamura, 1966). Thalli are holocarpic; with the onset of sexual reproduction the protoplasm amasses in gametangia where it is converted into gametes (Meinesz, 1980). The latter are released in concert (Drew and Abel, 1988) in species-specific short intervals (Clifton, 1997; Clifton and Clifton, 1999).

At present, 33 extant Halimeda species are recognized (Table 1; Ballantine, 1982; Dong and Tseng, 1980; Drew, 1993, Drew, 1995; Hillis-Colinvaux, 1980; Noble, 1986). They abound in a range of reef habitats (Gilmartin, 1960; Hillis-Colinvaux, 1980, Hillis-Colinvaux, 1986; Roberts et al., 1987; Taylor, 1950, Taylor, 1960; Tsuda and Kamura, 1991; Tsuda and Wray, 1977; Van den Hoek et al., 1978). Several morphological traits seem to be linked to particular habitats, e.g., unconsolidated substrata or wave exposed sites (Hillis-Colinvaux, 1980); these traits must have been acquired once or more depending on whether adaptation to these environments occurred once or multiple times.

Although most species are confined to either the tropical Atlantic or the tropical Indo-Pacific (Hillis-Colinvaux, 1980) many have close look-alikes in the other ocean. Their similarity may result from (sub)-recent dispersal. Alternatively, look-alike pairs could result from vicariant events separating the tropical Atlantic from the tropical Indo-Pacific (Coates and Oblando, 1996; Rögl and Steininger, 1984). In the latter case the phenotypically similar species may be nearest neighbors (geminates) or polyphyletic entities (cognates) coming about through convergence.

The empty “ghost”-thalli that are left behind following gamete shedding fall apart in loose segments adding their casts to the sediment (Drew, 1993; Freile et al., 1995; Payri, 1988; Roberts et al., 1987). Therefore, the genus has an extensive fossil record (Badve and Nayak, 1983; Braga et al., 1996; Elliott, 1965; Flügel, 1988; Mankiewicz, 1988) that goes back with certainty to the Early Cretaceous (Bucur, 1994). Many Late Cretaceous and Early Tertiary fossils show similarity to extant groups of species (Morellet and Morellet, 1922, Morellet and Morellet, 1941; Schlagintweit and Ebli, 1998) suggesting that the extant diversity consists of living fossils. Alternatively, similar phenotypes may have appeared through iterative evolution (Cifelli, 1969; Newell, 1967).

The issues raised can be addressed by analysis of data that are independent of environmental influence. Many authors have used nuclear rDNA (nrDNA) sequences for phylogenetic inferences in green algae (Bakker et al., 1994, Bakker et al., 1995a, Bakker et al., 1995b; Coleman and Mai, 1997; Durand et al., 2002; Famà et al., 2000; Jousson et al., 1998; Nakayama et al., 1998; Olsen et al., 1994, Olsen et al., 1998). Different mutation rates among nrDNA regions permit phylogenetic resolution at several taxonomic levels (Jorgensen and Cluster, 1988). In their preliminary phylogenetic assays of Halimeda, Hillis et al. (1998) and Kooistra et al. (1999) revealed tentative agreement between sequence phylogenies and changes among some phenotypic characters. However, their trees were inferred from only a part of the SSU nrDNA. In this paper, we surveyed a larger portion of the nrDNA, including the internal transcribed spacer regions (ITS-1 and ITS-2), from a greater number of species in the genus. The extensive phylogeny is used to examine the evolution of phenotypic and life-history traits. We also evaluate how biogeography (Avise, 2000; Veron, 1995) and reef ecology (Wanntorp et al., 1990) affected evolution and how all the information corresponds with evidence from the fossil record.