1 Introduction

Calcareous coralline algae show a high quality and a long fossil record starting about 140 Ma ago (mid-Valanginian, Early Cretaceous) (Chatalov et al. 2015). According to literature, rarefied data of coralline algal diversity show that the group diversified up to a peak in the Early Miocene, some 20 Ma ago, after which diversification slightly decreased with very minor fluctuations (Aguirre et al. 2000). Throughout their evolutionary history, calcareous corallines have been one of the most common and most widely distributed groups of fossil marine benthic algae . They are one of the major algal components both in tropical and in cool-water carbonates.

The oldest known coralline algae are small fragments from the Valanginian (Chatalov et al. 2015) or relatively thin crusts partially coating corals from lower Hauterivian coral reefs from the central Tethys (Arias et al. 1995; Tomás et al. 2007). By the end of the Early Cretaceous (Aptian-Albian) coralline algae formed multispecies rhodoliths in many marine settings (i.e. Beckmann and Beckmann 1966; Lemoine 1970; Simone et al. 2012). Since then, coralline algae have been an evolutionarily successful group of autotrophic marine organisms. However, little is known about the long-term production of rhodolith beds throughout the geological record. Halfar and Mutti (2005) compiled information on rhodolith-dominated deposits during the Miocene. However, no suitable data on the abundance of rhodolith beds for older time intervals are available. Long-term macroevolutionary history of coralline algae, including data of phylogeny, molecular clocks, and the fossil record, indicates that the group underwent radiation events during the Late Cretaceous and then during the Eocene (Aguirre et al. 2010). Whether these diversification events are correlated with widespread rhodolith bed formation remains to be satisfactorily assessed.

During their long geological history , coralline algae colonized almost any marine setting within the photic zone worldwide, as their present-day representatives. Precipitation of high-magnesium calcite in the cell walls, together with specific anatomical traits, makes coralline algae particularly resistant to extremely rough environmental conditions (Adey and Macintyre 1973; Round 1981; Wilson et al. 2004), as well as to high rates of herbivory (Steneck 1983, 1985, 1986; Maneveldt and Keats 2008; Burkepile and Hay 2010). In summary, these physiological attributes, together with their capability to thrive in poorly illuminating settings, account for their adaptability to inhabit a wide variety of marine settings and seem to have contributed to the ecological and evolutionary success of coralline algae (Aguirre et al. 2000).

In the rock record, coralline algae are present in almost all carbonate platform deposits , from coastal to outer shelf, as well as in siliciclastic and mixed carbonate-siliciclastic sediments. They can be major components of fossil assemblages, in the so-called rhodalgal lithofacies (Carannante et al. 1988), occurring as crusts coating both hard and soft substrates, as laminar growths directly on the seabed, or forming rhodoliths. The study of rhodoliths and rhodolith beds (their stratigraphic distribution, the taxonomy of the fossil coralline algal components, the growth morphology of the coralline algae, associated fauna, etc.) can provide key information on the past environmental conditions. In this respect, the fossil record is a unique historical archive that offers the possibility of understanding the foundation conditions for the formation of rhodolith beds and assessing the long-term processes involved in the development and maintenance of these ecosystems. In this chapter, we present a review of fossil rhodoliths and rhodolith beds (see a terminological discussion below), and their use to reconstruct ancient environments . The aims are: (1) to provide an overview of the palaeoecological and palaeoclimatic significance of rhodoliths; (2) to review the environmental factors that catalyse the formation of rhodolith beds, as well as their development through time up to their end; and (3) to discern the sequence stratigraphic contexts in which the formation and development of rhodolith beds are favoured.

2 Nomenclatural Background

2.1 Rhodolith

Non-geniculate coralline algae can occur attached to both hard and soft substrates, either organic or inorganic in origin. While growing, the algal thalli may eventually overgrow the whole settlement area completely covering the substrate in all directions and forming nodular structures freely rolling on the seabed. Coralline algae may also occur as free-living, isolated branches or forming nodular structures without an apparent inner nucleus. In their pioneering paper on the coralline algal nodules from Bermuda Islands, Bosellini and Ginsburg (1971, p. 670) proposed the name rhodolites for those “nodules and detached branched growths with a nodule form composed principally of coralline algae”. Previously, similar structures were identified as pralinés in the French literature (Molinier 1956; Pérès and Picard 1958). These coralline algal nodules were also identified as oncolites (McMaster and Conover 1966; Blanc 1968), a term used mostly for coated grains generated by bacterial activity. Bosellini and Ginsburg (1971) emphasized that the term rhodolite was suitable to differentiate this sort of algal nodule from other similar structures but constructed by different algae or bacteria. Later, Binda (1973) questioned the use of the term rhodolite since it was commonly and traditionally used for a variety of garnet. Consequently, Ginsburg and Bosellini (1973) proposed the name rhodolith, which Barnes et al. (1970, p. 268) had already introduced but had not defined, to substitute rhodolite. Since then, rhodolith has been the most accepted name in the geological literature (Adey and Macintyre 1973; Bosence 1983a, 1991), although the term rhodolite has been sporadically used (i.e., Orszag-Sperber et al. 1977; Montaggioni 1979). Flügel (1978), however, maintained the general name oncolite for any nodular structure biologically constructed, whether algae, animals or cyanobacteria (see also Richter and Sedat 1983).

Despite the general and widespread usage of the term rhodolith, some authors have used different names to refer to the same coralline algal nodular structures. Peryt (1983), citing the rhodolites of Bosellini and Ginsburg (1971), introduced the name rhodoid to refer to unattached nodular structures made up mostly of coralline algae (i.e. Burgess and Anderson 1983). Other authors have used this name to refer to nodular structures formed by other encrusting calcareous rhodophytes , such as peyssonneliaceans (Buchbinder and Halley 1985; Rasser and Piller 2004). Bosence (1983a) discarded the use of rhodoid based on its etymological unsuitability and the general acceptance of rhodolith. Even so, Flood (1983) and Flügel (2004) preferred to maintain the general term rhodoid instead of rhodolith.

Since rhodoliths can be built up by diverse encrusting organisms, further particular names have been proposed for those nodular structures in which other encrusting organisms represent the same proportion, or even greater than coralline algae. Richter and Sedat (1983) described nodules formed by a combination of the cyanobacteria Rivularia haematites and the coralline Lithoporella sp. from Pleistocene terraces of Greece as oncolites. Reid and Macintyre (1988) introduced the name foraminiferal-algal nodules to refer to nodular structures from the eastern Caribbean in which the encrusting benthic foraminifer Gypsina sp. equals to, or is more abundant than coralline algae as a major builder (Fig. 5.1a). Later, Prager and Ginsburg (1989), studying similar Gypsina-coralline algal nodular structures from the Florida shelf called them for-algaliths, a term that has been subsequently used by other authors (i.e. Martín and Braga 1993; Braga and Aguirre 2004; McNeil and Pisera 2010). Aguirre et al. (1993) adopted the term serpulid nodules for those nodular structures in which serpulids are significantly more abundant than corallines (Fig. 5.1b).

Fig. 5.1
figure 1

(a) Macroid of Acervulina, (b) serpulid nodule according to Aguirre et al. (1993)

Full size image

Hottinger (1983) used the term macroid to define any free-moving coated grain made up by encrusting organisms (see also Baarli et al. 2012). He applied this term to nodules dominated by either the benthic foraminifer Acervulina or coralline algae . Thus, a rhodolith would be a macroid built up exclusively or mostly by coralline algae. Nonetheless, Hottinger (1983) also used the term rhodoids for coralline algal-dominated nodules. Peryt (1983) restricted the term macroid to those coated grains (either chemically or biogenically produced) larger than 10 mm in diameter. Bassi et al. (2011, 2012), following Hottinger, used the term macroid for nodules formed mostly by Acervulina (up to 80 % of the nodule).

Steneck (1986) suggested a different nomenclatorial framework to distinguish the wide morphological variation of coralline algal appearances. This author identified three types, depending upon the degree of adherence of the algal thalli (Fig. 5.2): (1) adherent, when the thalli are completely attached to the substrate, having three distinguishable morphological end-points (thick, thin and branched); (2) leafy, when the margin of the algal thallus forms expansions (lateral branches) that are not in contact with the substrate; and (3) free-living or totally nonadherent, when corallines are unattached. The two first categories refer to coralline algal crusts while the latter includes coatings, rhodoliths, nodules, and maërls (= marls or gravels) (Fig. 5.2). This fourfold division of free-living coralline algae was in turn established depending upon the presence/absence of an evident nucleus (coatings and rhodoliths versus nodules and maërls, respectively) or the taxonomic composition (nodules and maerls are monospecific). In coatings, coralline algae completely cover a nucleus but the algae represent less than 50 % of the total structure (including the alga and the nucleus). By contrast, in rhodoliths, the algal cover constitutes more than 50 % of the total structure (Fig. 5.2). For algal structures without a nucleus, maërl refers to monospecific-isolated branches, and nodules are monospecific “more densely branched, spherical-to-ellipsoid plants” (Steneck 1986, p. 280) (Fig. 5.2). The criteria for separating these four categories of free-living corallines might be difficult to apply and consequently should be discarded. On this point, assessing the presence/absence of a nucleus to distinguish nodules from rhodoliths needs sectioning of the algal structures. The same applies to estimate the percentage of algal cover. Finally, the four types of structures might be formed by either a single or several species, and thus the monospecific composition is not exclusive of nodules and maërl despite the definition of Steneck (1986). To avoid terminological confusions and the difficulty of applying nomenclatorial frameworks, we suggest keeping the original definition of Bosellini and Ginsburg (1971) and thereby maintaining the historical stability of the term.

Fig. 5.2
figure 2

Classification of coralline algal appearances depending upon the degree of adherence of the algal thalli (Modified from Steneck 1986). In all pairs of photographs, picture on the right is a recent exemplar and the image on the left is a fossil counterpart

Full size image

2.2 Rhodolith Beds

Rhodoliths can form extensive concentrations, giving rise to the so-called rhodolith beds. In present-day oceans, rhodolith beds occur worldwide, from the Equator to circumpolar latitudes and from the intertidal down to more than 120 m in water depth (Molinier 1956; Pérès and Picard 1958, 1964; McMaster and Conover 1966; Blanc 1968; Adey 1979, 1986; Steneck 1986; Prager and Ginsburg 1989; Littler et al. 1991; Foster 2001; Basso 1998; Riosmena-Rodríguez et al. 2010; Lund et al. 2000; Steller et al. 2003, 2009; Ballesteros 2006; Amado-Filho et al. 2010, 2012a; Matsuda and Iryu 2011; Foster et al. 2013). They constitute hotspots of marine biodiversity in present-day oceans since many marine organisms use rhodolith beds as nurseries, growth habitats or refuges (see a review in Nelson 2009). In addition, some of these rhodolith-bed inhabitants are of particular interest as they are resources for humans (Steller and Cáceres 2009; Steller et al. 2009; Riosmena-Rodríguez et al. 2010).

Lemoine (1910) used the Breton name maërl (also maerl in the literature) to refer to unattached coralline algal concentrations along the Brittany coast. According to OSPAR Commission (2010), “maërl is a collective term for various species of non-jointed coralline red algae (Corallinaceae) that live unattached. These species can form extensive beds, mostly in coarse clean sediments of gravels and clean sands or muddy mixed sediments…. Maërl beds may be composed of living or dead maërl or varying proportions of both”. It, therefore, bears mentioning that in recent literature “rhodolith beds” and “maërl beds” are used indistinctly to refer to extensive concentrations of free-living structures built up mostly by non-geniculate coralline algae. According with the French literature, the term fons à praliné was also used (Molinier 1956).

Coralline algae can also densely occur in coastal to deep-subtidal settings but originating bioconstructions: algal ridges, algal frameworks, coralligènes, trottoirs, and corniches (Pérès and Picard 1958, 1964; Blanc 1968; Adey 1978; Bosence 1985a; Freiwald and Henrich 1994; Freiwald 1998; Macintyre et al. 2001; Nalin et al. 2006; Ballesteros 2006; Georgiadis et al. 2009; Aguirre et al. 2014). In all these cases, encrusting coralline algae form a rigid structure.

In the rock record , rhodoliths can form extensive and thick concentrations (Fig. 5.3). Authors have referred to these accumulations variously as beds, pavements, maërls, banks, algal gravels, mounds, biostromes, rhodolith facies, or rhodolith rudstones-floatstones (Nebelsick et al. 2005; Basso et al. 2012). For simplicity, and as an analogue of the present-day rhodolith beds, we will use the same name.

Fig. 5.3
figure 3

Examples of fossil rhodolith beds. (a) Tortonian deposits of the Almanzora Corridor Basin (Almería, SE Spain), (b) Serravallian rhodolith bed of Alessandria (Piedmont Basin, NW Italy), (c) upper Pliocene deposits of the Cabo de Roche area (Cádiz, SW Spain)

Full size image

3 Classification of Rhodoliths

Bosence (1976, 1983a) summarized a general classification scheme for rhodoliths based on the taxonomic composition (monospecific versus multispecific), the external morphology (spheroidal, ellipsoidal, and discoidal), branch density (I: single branch, II: open branches, III: frequent branches; IV: densely branched), and algal growth forms (laminar, either concentric or boxwork, branching, and columnar). This author further established the methodological procedure to describe the external morphology of the rhodoliths by using the triangular diagram for the classification of pebble shape of Sneed and Folk (1958) (Bosence 1976, 1983a). The three extreme morphologies considered in the diagram are spheroidal, ellipsoidal, and discoidal (Fig. 5.4). These plots are still currently used. However regarding the coralline algal growth form on the surface of the rhodoliths, palaeontologists have been increasingly using the terminological classification of Woelkerling et al. (1993) (Fig. 5.5).

Fig. 5.4
figure 4

Triangular diagram used for the representation of the rhodolith shape. L, I and S indicate the large, intermediate and short axes, respectively, of the rhodoliths

Full size image
Fig. 5.5
figure 5

Names of the different types of coralline algal thallus growth forms. The thalli with some sorts of protuberances are included in the green ellipse, while the thalli encompassing lamellar growths are clustered in the yellow ellipse (Modified from Woelkerling et al. 1993)

Full size image

Branch density has been considered in a different way depending upon the authors. Steller and Foster (1995, p. 203) quantified the branch density as “the mean number of apical tips counted in five haphazardly placed 1 cm2 quadrats”. Basso et al. (2009) preferred the use of protuberance degree in order to avoid the word density since it can have a different meaning. In fossil examples, both branch density and protuberance degree have been used (i.e. Bosence and Pedley 1982; Bosence 1985b; Basso 1998; Basso et al. 2009; Quaranta et al. 2012, Brandano this volume).

Basso (1998) classified rhodoliths within three morphological groups depending on the size, the inner structure, the external shape, the algal growth forms, and the taxonomic composition: (1) Boxwork: multispecific rhodoliths larger than 4 cm across, irregular to elliptical in shape, formed by laminar to columnar algal growth forms, and with a conspicuous nucleus and numerous internal voids filled up by sediment. (2) Prâline: small (2–4 cm in diameter), massive, monospecific rhodoliths irregular to spheroidal in shape built up by laminar to branching or columnar algal thalli, and with an evident nucleus. (3) Unattached branches: small (1–5 cm in length or diameter), non-nucleated (or with a very small nucleus), monospecific rhodoliths of any possible shape made up by branched algal thalli without voids. According to Basso (1998), the unattached branched rhodoliths include the four types of branch density of Bosence (1976, 1983a), being the denser branching (types III–IV) produced in shallow settings.

This scheme proposed by Basso (1998) is difficult to apply for the rapid classification of rhodoliths in the field. Identifying the algal components building up the rhodoliths very often requires observation under the microscope. Thus, the taxonomic composition (whether monospecific or multispecific) should not be a classification criterion. In addition, as commented above, sectioning is needed to ascertain whether rhodoliths are nucleated. In terms of algal growth form, there is a clear overlap between boxwork and prâline since both can be made up of laminar to columnar algal thalli (Basso 1998).

Taking into consideration that the presence of a differentiated nucleus and the taxonomic composition are two criteria that are very difficult to use in the field, we propose to adopt a classification scheme based on morphological characters at three different levels: (1) Rhodolith morphology. This can be determined using the triangular diagram mentioned above with the three end points ellipsoidal, spheroidal, and discoidal (Fig. 5.4). (2) Growth morphology of the algal thalli on the surface of the rhodolith, using the terminology proposed by Woelkerling et al. (1993). (3) Growth form of the algal thalli in the interior of the rhodolith. In latter case, the three categories proposed by Bosence (1983a) would clearly include the whole range of possible algal forms: laminar (concentric or non-concentric [boxwork]), branching (types I to IV), and columnar.

4 Estimation of Rhodolith Abundance

An important issue is to determine rhodolith abundance and density within rhodolith beds. Several approaches have been used. Density, expressed as number of rhodoliths by surface, can be quantitatively estimated directly by counting rhodoliths in a specified area (Ballantine et al. 2000; Amado-Filho et al. 2010). Basso et al. (2012) proposed to quantify rhodolith abundance in rhodolith beds with image analysis, as done by Amado-Filho et al. (2012a). Abundance can be also semi-quantitatively determined by applying charts for visual comparison of the proportion of fossil components (rhodoliths in this case) per rock volume (Kidwell and Holland 1991). Abundance can be further qualitatively approached by estimating rhodolith packing using the three categories of the comparison tables of Kidwell and Holland (1991) (Fig. 5.6): (1) dispersed: rhodoliths are separated from each other by a distance of more than the largest diameter of the rhodoliths; (2) loosely packed: rhodoliths are close each other but not in contact; and (3) densely packed: rhodoliths are in direct contact with each other throughout the rhodolith bed.

Fig. 5.6
figure 6

(a, b) dispersed rhodolith beds, (c, d) loosely packed rhodolith beds, (e, f) densely packed rhodolith beds

Full size image

Within individual rhodoliths, the relative abundance of coralline algal taxa can be estimated by point counting the cross-sectional areas that they occupy in thin sections following the method proposed by Perrin et al. (1995).

5 Rhodoliths as Palaeoenvironmental Indicators

Fossil coralline algae have been known since the nineteenth century and have long been studied intensively from a taxonomic perspective (Aguirre and Braga 2005) but without considering their potential use to make palaeoecological interpretations (Adey and Macintyre 1973; Bosence 1983b). Bosellini and Ginsburg (1971) came to the conclusion, for the first time, that rhodolith morphology, as well as inner algal arrangement and algal growth forms, record valuable information of the environmental conditions in which these nodular structures grow. Therefore, these authors stressed that the study of fossil rhodoliths would be useful to make palaeoenvironmental interpretations.

Adey and Macintyre (1973) updated the general knowledge concerning living rhodoliths and discussed their potential use in geological studies. They further emphasized that taxonomic problems and the limited knowledge of environmental distribution of recent rhodoliths are major hindrances to apply a strict uniformitarian approach to interpret fossil rhodoliths. Bosence (1991) highlighted the necessity of establishing an uniform and consistent taxonomic framework for fossil coralline algae comparable to the taxonomic standards of present-day corallines. Braga et al. (1993) subsequently demonstrated that many of the taxonomic criteria used to identify present-day corallines were preserved in the fossil counterparts, influencing the latter advance of fossil coralline taxonomy.

Adey and Macintyre (1973) also argued that despite the widespread geographic and bathymetric distribution of coralline algae as a whole, particular taxonomic groups occur in characteristic latitudinal and shelf habitats. This was tested in different regions from the tropics to cold-temperate latitudes (Adey 1979, 1986; Adey et al. 1982; Bosence 1983b, 1991; Minnery et al. 1985; Minnery 1990). The realization that groups of coralline algae preferentially inhabit particular environments is critical for palaeoecological and palaeoclimatology-palaeolatitudinal inferences. More recently, the use of geochemical analyses has increased the value of rhodoliths for palaeoenvironmental reconstructions (Kamenos et al. this volume).

5.1 Water Energy

It has often been stated that the morphology of rhodoliths is closely related to water movement. Bosellini and Ginsburg (1971) established five types of rhodolith morphologies that can be found along a water-energy gradient. Spheroidal and ellipsoidal rhodoliths would be characteristic of moderate- to high-energy settings. Under these conditions, water movement promotes a frequent overturning of rhodoliths, and coralline algae grow in all possible directions. Flat, discoidal, and ameboidal rhodoliths would be more frequent in calm waters, as certain stability is required to produce these morphologies. In contrast, Adey and Macintyre (1973, p. 900) concluded that “rhodolith morphology should not be related to energy conditions”, questioning the rhodolith shape-water energy relationship proposed by Bosellini and Ginsburg (1971).

Bosence (1976) experimentally demonstrated using a wave tank that ellipsoidal rhodoliths are more easily transported than spheroidal ones, and that discoidal forms are the most stable morphologies. According with this experiment, discoidal rhodoliths would be expected to be more abundant in quiet settings while ellipsoidal and spheroidal rhodoliths should be dominant in more exposed environments. However, in rhodolith beds from the Mannin Bay (Connemara, Ireland), ellipsoidal and spheroidal rhodoliths did not show any characteristic energy-dependent environmental distribution (Bosence 1976). More recently, observations made in the Gulf of California (Steller and Foster 1995; Steller et al. 2009) and off Fraser Island in the Great Barrier Reef (eastern Australia) (Lund et al. 2000) have also demonstrated that there is no clear relationship between rhodolith morphology and water energy.

Regardless of the lack of correspondence between experimental tests and field observations, morphology of fossil rhodoliths has been usually interpreted according to the original laboratory results of Bosence (1976). Hence, discoidal rhodoliths are often interpreted as formed in calm water settings and ellipsoidal and spheroidal ones are usually considered characteristic of high-energy conditions (e.g. Bassi 1995, 1998; Bassi and Nebelsick 2010; Checconi et al. 2010). Nonetheless, rhodolith shape is not directly correlated with water energy (e.g. Brandano et al. 2005; Bassi et al. 2006). A combined analysis of facies and sedimentary structures, together with rhodolith shapes, is needed to correctly infer the palaeoenvironmental turbulence.

This lack of relationship may be due to different factors. The movement of rhodoliths could be due directly to the activity of vagrant organisms moving around algal nodules and not to currents or waves (Marrack 1999). This author carried out experimental and observational work on rhodolith morphology in rhodolith beds at different sites and water depths from the Gulf of California. Using video camera and scuba surveys, Marrack (1999) showed that bioturbation was the key factor causing displacements of rhodoliths deeper than 12 m, where water energy is low enough or negligible to overturn rhodoliths. At this water depth, rhodoliths were only sporadically moved by currents during severe storms.

Rhodolith morphology can also largely depend upon the inherited shape of the nucleus if the algal cover is poorly developed (Braga and Martín 1988; Aguirre et al. 1993). Large flat nuclei encrusted by corallines producing a thin coating due to a limited time of growth would produce discoidal rhodoliths in shallow, high-energy settings. On the other hand, long-term exposure of rhodoliths on the seafloor with enough time to develop thick algal covers with respect to the nucleus would generate ellipsoidal and spheroidal rhodoliths. This has been shown in recent rhodoliths off Fraser Island (NE Australia) (Lund et al. 2000), as well as in upper Miocene rhodolith beds of the Almanzora Corridor Basin (Almería, SE Spain) (Braga and Martín 1988) and in upper Pliocene rhodolith beds of Cabo de Roche (Cádiz, SW Spain) (Aguirre et al. 1993).

Other features of rhodoliths have been related to water energy and considered reliable tools in palaeoenvironmental analyses. The internal coralline algal growth forms do relate to the degree of water energy (Bosellini and Ginsburg 1971; Adey and Macintyre 1973; Bosence 1983a, b, 1991; Steneck 1986). According to these authors, rhodoliths formed by thin, laminar algal thalli occur in calm waters. Fruticose, delicate branching algal thalli are related to low or intermediate water movement. Under higher turbulence, branch density increases (Bosence 1976; Basso et al. 2009). Additionally, the branch tips in fruticose thalli break and are eroded, producing shorter columns with broadening of tips that can fuse together to better withstand moderate water movements (Steneck 1986). Thus, under moderate energy, algal thalli acquire lumpy-warty growth forms. These variations in thallus morphology related to turbulence have been recorded in different fossil examples (Bosence and Pedley 1982; Braga and Martín 1988; Aguirre et al. 1993; Bassi 1995, 2005; Bassi and Nebelsick 2010).

Growth forms, however, can be genetically controlled. Foliose and fruticose thalli of the rhodolith-forming species Lithophyllum margaritae Riosmena-Rodríguez et al. are actually genetic variants (Schaefer et al. 2002). These authors suggested that these two genetic entities would correspond to different species.

The percentage of constructional voids in rhodoliths is also a good approach to infer water energy (Minnery et al. 1985). Open inner arrangements, with a great proportion of voids among algal thalli, are generally indicative of quiet waters. An increase in turbulence implies a progressive massiveness of the internal rhodolith structure. The relationship between massiveness and water energy has been shown in different fossil examples (Bassi 1995, 2005; Bassi and Nebelsick 2010).

5.2 Water Depth

Coralline algae, like all photosynthetic organisms, depend on the sunlight to survive. Light is progressively absorbed as it penetrates the water column. Coralline algae have phycobiliproteins as accessory pigments (Lobban and Harrison 1994; van der Hoeck et al. 1995; Lee 2008; Graham et al. 2009) that provide them the ability to colonize the deepest settings among all algal groups (Littler et al. 1985).

Light irradiance might be considerably reduced with increasing suspended material, either inorganic (particles of sediment) or organic (food particles). Therefore, turbid waters limit the bathymetric distribution of coralline algae (Adey and Macintyre 1973; Bosence 1983b; Martindale 1992; Koop et al. 2001; Steller et al. 2009). In shallow settings, turbidity might be favoured by resuspension of sediment due to waves and bottom currents, siliciclastic input, and refractory organic matter supply from rivers. Offshore, light penetration might decline in areas of high productivity linked to upwelling centers.

As mentioned above, corallines occupy benthic habitats at any depth within the photic zone. Nonetheless, some taxonomic groups of coralline algae show particular depth distributions (Adey and Macintyre 1973; Adey 1979, 1986; Bosence 1983b, 1991; Minnery et al. 1985; Minnery 1990) (Fig. 5.7). The depth ranges of the taxa, however, vary depending on the latitude (see next Sect. 5.5.3). In low latitudes, shallow-water coralline algal assemblages are mostly dominated by ‘mastophoroids’ (sensu Harvey et al. 2003) followed by lithophylloids (Gordon et al. 1976; Adey 1979; Adey et al. 1982; Bosence 1984; Minnery et al. 1985; Minnery 1990; Verheij and Erftemeijer 1993; Iryu et al. 1995; Payri et al. 2000; Ringeltaube and Harvey 2000; Littler and Littler 2003). Lithophylloids, with subordinate ‘mastophoroids’, dominate in temperate seas, such as in the Mediterranean (Hamel and Lemoine 1953; Comarci et al. 1985; Adey 1986; Di Geronimo et al. 1993; Braga and Aguirre 2009). Members of the family Sporolithaceae are also present in shallow, warm waters but they are scarce and limited to cryptic habitats (i.e. Braga and Bassi 2007). Melobesioids, as well as sporolithaceans, increase with waters depth (Adey 1979, 1986; Adey et al. 1982; Minnery et al. 1985; Minnery 1990; Verheij and Erftemeijer 1993; Iryu et al. 1995; Rasser and Piller 1997; Lund et al. 2000; Payri et al. 2000; Littler and Littler 2003) (Fig. 5.7). In high latitudes, melobesioids is the overwhelmingly dominant group (Adey 1979, 1986; Adey et al. 1982; Steneck 1986).

Fig. 5.7
figure 7

Abundance of typical coralline genera-forming rhodoliths in relation with water depth in low latitudes

Full size image

Taking into consideration the latitudinal variation in relative abundance of taxa with respect to their bathymetric ranges, the depth/coralline-taxa distribution relationship has been used to infer palaeobathymetry. The basic idea is to combine coralline algae that inhabit particular water-depth ranges into ecologically homogeneous groups and then analyse relative abundances of these taxa in geological sections to infer palaeobathymetry, as well as related sea-level changes through time. Qualitative estimations of bathymetrically controlled taxa were used to infer palaeodepth (Buchbinder 1977; Braga and Martín 1988; Aguirre et al. 1993). Braga and Aguirre (2001) used quantitative data for the first time to infer palaeodepth changes in upper Miocene-upper Pliocene temperate and warm-water carbonate deposits in different basins in southern Spain. Subsequent studies have applied this method in different deposits from disparate regions (Braga and Aguirre 2004; Brandano et al. 2005; Kroeger et al. 2006; Braga et al. 2009; Benisek et al. 2009; Aguirre et al. 2012; Nebelsick et al. 2013).

5.3 Palaeolatitude and Palaeoclimate

Corallines were long erroneously considered as warm, low-latitude organisms in the fossil record (Adey and Macintyre 1973). However, as commented above, they are found in warm and cool-temperate carbonates. Despite their cosmopolitan occurrence, particular taxonomic groups of corallines (subfamilies) from shallow-water settings show preferential latitudinal distributions (Fig. 5.8) (Adey and Macintyre 1973; Adey et al. 1982; Adey 1986; Minnery et al. 1985; Bosence 1991; Perrin et al. 1995). Among coralline algae, members of the subfamily ‘Mastophoroideae’ show a preferential tropical distribution (Steneck 1986; Woelkerling 1996a). Representatives of the order Sporolithales show a similar latitudinal dispersal (Verheij 1993; Woelkerling 1996b). Within Sporolithales, the genus Sporolithon has maintained this latitudinal preference through time (Johnson 1963; Fravega et al. 1989; Braga and Bassi 2007). Lithophylloids are abundant mainly in intermediate , warm, and temperate latitudes (Woelkerling 1996c). Finally, the subfamily Melobesoideae is dominant mostly at high latitudes (Woelkerling 1996d).

Fig. 5.8
figure 8

Latitudinal distribution of algal subfamilies in shallow-water assemblages (Modified from Steneck 1986). The subfamily ‘Mastophoroideae’ is here considered according with Harvey et al. (2003)

Full size image

The upper Neogene deposits in different basins from southeastern Spain are characterized by an alternation of reefal and temperate carbonates (Martín and Braga 1994; Brachert et al. 1996). The temperature control of this alternation was firstly demonstrated using δ18O isotope values measured in planktonic and benthic foraminiferal shells collected from the time-equivalent marly sediments (Sánchez-Almazo et al. 2001). ‘Mastophoroid’-rich assemblages (accompanied by Sporolithales) dominated in reefal carbonate deposits while lithophylloid-dominated assemblages did so in the shallow-water temperate carbonates (Braga and Aguirre 2001). In deeper settings , both temperate and reefal carbonates are characterized by abundant melobesioids. Later studies have followed this approach to infer palaeoclimatic changes (Brandano et al. 2005; Kroeger 2007; Nalin et al. 2008).

Geochemical studies have added new insights on the use of corallines as palaeothermometers. Kamenos et al. (this volume) review the applications of these approaches, including elemental analyses as well as oxygen isotopes. Nonetheless, any of these geochemical analyses as palaeoclimatic proxies in deep time (millions of years) have, so far, not been applied. Diagenetic alterations, such as early cementation, early transformation of the high-Mg calcite into low-Mg calcite (Alexandersson 1974, 1977, 1978; Freiwald and Henrich 1994), most likely bias the original geochemical signature, impeding the use of these proxies in palaeontological samples. Stable oxygen isotopes are also difficult to apply as palaeoclimatic/palaeotemperature indicators in fossil coralline algal skeletons due to a strong vital effect, that is, the preference uptake of a specific oxygen isotope due to metabolic (photosynthesis and respiration) and kinetic (isotope diffusion) effects (Wefer and Berger 1991; Halfar et al. 2000; Lee and Carpenter 2001; but see Rahimpour-Bonab et al. 1997).

6 Rhodoliths as Palaeocommunities

Rhodoliths are multispecific structures formed by the successive growth of encrusting organisms, with the accompanying non-sessile ones, that can therefore be considered communities by themselves. Space is a key ecological factor structuring communities of encrusting organisms (Begon et al. 1990; Buss 1990; Hochberg and Lawton 1990; Leigh 1990). Competition for space might be critical for those organisms growing in a limited space during formation of rhodoliths. For rhodoliths with a nucleus, the first step is the colonization of an empty space. Eventually, these colonizers might come into contact during their growth, thus beginning intra- and interspecific competition for the space (Adey and Macintyre 1973; Bosence 1983c, 1985b; Steneck 1985; Prager 1987). Coralline algae grow very slowly in relation with other potential competitors. In this regard, thin thalli of coralline algae that might grow quickly would behave as primary colonizers or generalist epiphytic species interacting with other fast-growing encrusting organisms, such as serpulids, bryozoans or benthic foraminifers (Fig. 5.9) (Adey and Macintyre 1973; Adey and Vassar 1975; Steneck 1985; Woelkerling 1988; Keats et al. 1993; Figueiredo et al. 1997).

Fig. 5.9
figure 9

Thin section through a fossil rhodolith from the nucleus to the outer surface showing interactions among the major builders of the structure. The nucleus is a nodular bryozoan colony. In the outer surface of the nucleus, serpulids (Ser) intergrew with the bryozoan up to an eventual outcompetition of the serpulids. The serpulids were then colonized by thin laminar, fast-growing thalli of Lithophyllum pustulatum (Lamouroux) Foslie (Lp) and L. nitorum Adey and Adey (Ln) that interacted with bryozoans (Bry) and serpulids. Finally, serpulids became progressively less abundant and warty/fruticose thalli of Mesophyllum alternans (Foslie) Mendoza and Cabioch (Ms.a), which are, in turn covered by serpulids, bryozoans and encrusting thalli of Mesophyllum lichenoides (Ellis) Lemoine (Ms.l)

Full size image

However, inferring space competition among organisms forming rhodoliths in fossil examples is difficult, since many factors should be taken into consideration: (1) longevity of the different organisms; (2) the importance of other possible ecological interactions, such as mutualism or parasitism; (3) the role played by non-preserved soft-body epiphytes or epizoos; (4) effect of herbivorous; (5) chemical defenses among competitors; (6) subtle preferences for abiotic factors (light intensity, nutrients, currents, turbulence, etc.); (7) growth rate; and (8) temporal interruption of growth. All these factors might have played a key role structuring rhodolith communities, influencing the distribution and success of all encrusting organisms involved in rhodolith formation.

Analyses of the changes in the growth forms of the coralline algae, as well as in the taxonomic composition from the nucleus to the surface of the rhodoliths can provide valuable information of palaeoenvironmental shifts during their development. Some authors have interpreted these taxonomic transitions as ecological successions by facilitation (Adey and Macintyre 1973; Bosence 1983b, c). Prager (1987), in contrast, highlighted that the taxonomic changes from the nucleus to the outer surface of rhodoliths might be due to physical or biotic changes through time not related to ecological successions (see also Basso et al. 2009). Additionally, rhodoliths do not grow continuously, precluding any inference of ecological succession during rhodolith growth. Bioerosion, erosion, early lithification, or abrasion of rhodolith builders can reflect complex taphonomic histories that record important palaeoenvironmental changes. This is the case of rhodoliths from different regions that consist of two well-differentiated parts, a highly altered, lithified, and bored inner part and an outer pristine one (McMaster and Conover 1966; Focke and Gebelein 1978; Reid and Macintyre 1988; Minnery 1990; Littler et al. 1991). According to radiocarbon dating, the separation between the two parts is characterized by an interruption in the rhodolith growth averaging about 500 years.

Going back into the deep time, when radiometric dating methods are not applicable to rhodoliths, identifying erosive surfaces in the interior of the rhodoliths affecting the whole nodule is the best evidence to detect prolonged interruptions during their formation. Examples of multistory rhodoliths have been described in the fossil record (Checconi et al. 2010; Aguirre and Braga 2012). In rhodolith assemblages from the middle Miocene hemipelagic Orbulina marls from the Southern Apennines (Italy), Checconi et al. (2010) distinguished two rhodolith growth stages separated by an erosive and bored surface: (1) nucleation and growth of the rhodoliths, and (2) a final growth stage before burial. Nucleation is characterized by melobesioids and subordinate ‘mastophoroids’ , with rare sporolithaceans and lithophylloids. The rhodolith growth (main increase in size) is represented by abundant melobesioids and rare to common ‘mastophoroids’; very rare sporolithaceans are also present. The final growth stage is dominated by melobesioids with rare ‘mastophoroids’ and very rare sporolithaceans.

The Pliocene deposits in Cabo de Roche area (Cádiz, SW Spain) consist of three unconformably superimposed units (Fig. 5.10a, b). The upper unit is made up of densely packed concentrations of rhodoliths formed in a relatively deep and sheltered palaeobay (Aguirre 1992; Aguirre et al. 1993; Aguirre and Braga 2012) (Figs. 5.3c, 5.6e, and 5.10b). More than 20 % of these rhodoliths (n = 140 rhodoliths) show two distinct phases separated by an erosion surface (Fig. 5.10b). The inner phase is eroded and bored and consists of thin laminar growths of Lithophyllum pustulatum intergrowing with serpulid worm tubes, which are the dominant components. The outer phase is characterized by warty-lumpy to fruticose growths of Phymatolithon calcareum (Pallas) Adey and McKibbin (Fig. 5.10b). Occasionally, laminar thalli of Mesophyllum lichenoides and/or, to a lesser extent, Lithophyllum incrustans Philippi overgrow P. calcareum branches. The differentiation of the two stages points to a complex history of rhodolith development. These rhodolith-dominated deposits unconformably overlie the middle Pliocene unit (Fig. 5.10b), consisting of bioclastic calcirrudites with nodular structures dominated by serpulids encrusted by thin thalli of L. pustulatum, the so-called serpulid nodules by Aguirre et al. (1993) (Figs. 5.1b, 5.10b). This suggests that the inner parts of the complex rhodoliths of the upper unit are serpulid nodules reworked from these older deposits constituting the nuclei for P. calcareum settlement and growth (Aguirre and Braga 2012). Thus, in this case, an intricate sedimentological and taphonomic history of growth, burial, reworking, and regrowth can be inferred during long-term rhodolith development.

Fig. 5.10
figure 10

(a) location of the Pliocene deposits close to the Cabo de Roche area (Cádiz, SW Spain), (b) synthetic stratigraphic log of the Pliocene deposits of the Cabo de Roche area, consisting of three unconformably overlying units. A bed formed by serpulid nodules intercalates in the middle unit. The upper unit starts with a dense rhodolith bed, up to 4 m thick, in which more than 20 % of the rhodoliths consist of two phases separated by and eroded and bored surface. The inner phase represents serpulid nodules eroded from the middle unit that constituted the nucleus for the algal growths of the outer phase. (Pleist. = Pleistocene)

Full size image

In addition to encrusters contributing to the construction of rhodoliths, endolithic organisms use them to live inside bores, thus promoting the destruction of these hard living substrates. Bassi et al. (2011, 2012) studied this destructive guild in recent nodules of Acervulina-coralline algae from deep fore-reef settings (60–100 m water depth) off Kikai-jima (S Japan). They described an ichnocoenosis dominated by Entobia and Gastrochaenolites, with accessory Trypanites and Meandropolydora as well as unidentified microborings. These ichnogenera preferentially inhabit in settings characterized by low sedimentation rates and high turbulence. This ichnoassemblage has been traditionally considered diagnostic of very shallow waters (less than 10 m) and, therefore, an excellent sea level indicator. This general assumption is patently challenged by the findings of Bassi et al. (2011, 2012). Regarding ecological controls on borers, Bassi et al. (2011, 2012) concluded that while small Entobia, made by the sponge Cliona, is present in nodules of any size, large Gastrochaenolites, produced by bivalves, occurs in the late growth of larger nodules (6–7 cm in diameter).

A decrease in nodule size and thickness of coralline thalli with depth is observed in present-day coralline algal nodules off Fraser Island (eastern Australia) from the inner to the uppermost outer shelf (28–117 m water depth) (Bassi et al. 2013). Such trends most likely account for the decrease in diversity of ichnogenera with depth, as the largest bioeroders are excluded from the ichnoassemblages. Lower coralline growth rates favour higher density of bioerosion in deeper algal nodules.

7 Sedimentological and Sequence Stratigraphic Significance of Rhodolith Beds

Fossil rhodolith beds can be formed by either sedimentological (allochthonous) or biological (autochthonous) processes. Allochthonous concentrations are the result of either offshore or onshore transport from the production areas of rhodoliths. Bassi et al. (this volume) discuss examples of basinward-reworked rhodolith beds in several Neogene basins of Italy (central Mediterranean) associated with submarine lobes and channels. Johnson et al. (2012, 2013, this volume) present study cases of offshore reworked rhodolith-dominated deposits, and onshore export of rhodoliths from deeper settings linked to extremely high-energy events, such as hurricanes or tsunamis, in Macaronesian archipelagos.

Here, we focus on autochthonous accumulations of rhodoliths formed by in situ production. Pure autochthonous rhodolith beds are very unusual because long-term time-averaged processes prevail in the rock record (Kidwell and Bosence 1991). As a consequence, some reworking might be expected in the areas of rhodolith production. As commented above, autochthonous rhodolith beds can potentially form throughout the platform along a depth gradient, from the coast to the outer-shelf. However, some environmental conditions, as discussed in this section, inhibit the development of rhodolith beds.

Growth rates of coralline algae are very low, being greater in the tropics than in the warm and cold-temperate realms (Adey and Macintyre 1973; Kamenos et al. 2008; Halfar et al. 2011). Consequently, they are very sensitive to sediment influx, and long-term rhodolith bed formation requires sediment starvation. Off Brazil, rhodoliths form one of the largest rhodolith bed known so far (Amado-Filho et al. 2012b). However, terrigenous discharges linked to Amazon and Parana rivers inhibit rhodolith bed formation (Milliman 1977, p. 236). Experimental work has demonstrated the deleterious effect of sediment burying rhodoliths, even under a millimetre-thick layer of fine-grained sediment (Riul et al. 2008; Villas-Bôas et al. 2014). Rhodoliths made up of Mesophyllum engelhartii bleached after being buried 75 days and those formed by Lithothamnion sp. died even earlier, after 41 days (Villas-Bôas et al. 2014). Burial of rhodoliths by fine-grained sediment with high content of organic matter might also lead to anoxic conditions that decimate rhodolith growth (Wilson et al. 2004; Hall-Spencer et al. 2006). Sediment supply also muddies the water column thus limiting light penetration and inhibiting the healthy growth of rhodoliths (Adey and Macintyre 1973; Bosence 1983b; Steller and Foster 1995; Foster et al. 1997; Steller et al. 2009). The harmful influence of terrigenous sedimentation on rhodolith bed development has been proven in Pliocene deposits in SE Spain (Aguirre et al. 2012).

As stated above, rhodoliths are able to cope with a wide spectrum of environmental conditions, from shallow high-energy to deep low-energy settings. Nonetheless, successful and continuous development of dense rhodolith beds takes place under moderate water energy (Bosence 1983b; Hottinger 1983; Foster et al. 1997; Steller et al. 2009; Aguirre et al. 2012). In different Neogene basins of SE Spain, rhodolith beds formed in mid to outer ramp settings, in a facies belt nearly parallel to palaeocoast and basinward of a shoal facies belt (Braga et al. 2006; Puga-Bernabéu et al. 2007; Martín et al. 2004; Aguirre et al. 2012). In Miocene warm-temperate to tropical carbonate platforms in central Mediterranean (Italy), rhodolith beds occur from inner to mid platform, both shoreward and offshore of Porites patch reefs (Benisek et al. 2010; Brandano et al. 2010). Under very low water energy, very fine-grained sediment can be deposited in these settings, suffocating coralline algal development. Steller et al. (2009) experimentally removed rhodoliths from a rhodolith bed to deeper areas in the Gulf of California and found that they became buried rapidly and, consequently, were killed off. In contrast, moderate water energy allows to: (1) wash out fine particles and excess of organic matter, reducing fouling; (2) rock the rhodoliths, allowing coralline algae to grow in all directions; (3) permit free movement of vagrant organisms, which slowly displace rhodoliths; and (4) preclude the formation of algal bridging among rhodoliths, preventing the growth of a more rigid structure (Bosence 1983c; Steller and Foster 1995; Foster et al. 1997; Freiwald and Henrich 1994; Harvey and Bird 2008; Peña and Bárbara 2008, 2009; Steller et al. 2009).

The best conditions for rhodolith bed formation and persistence are relatively low-moderate water hydrodynamic, good oxygenation, and low sedimentation rate. The long-term maintenance of these environmental conditions is essential for the formation of thick and dense rhodolith beds like those found in the geological record. In a sequence stratigraphic context, the appropriate conditions are usually reached during transgressive periods (Friebe 1993; Nalin et al. 2008; Leszcynski et al. 2012). When the sea level raises the coastal settings, where high-energy conditions prevail and siliciclastic sedimentation is favoured, move onland widening the areas of carbonate production in the platforms where rhodolith beds can be established.

Although it is largely accepted that thick rhodolith beds form preferentially during the transgressions, rhodolith beds also develop during highstand stages (Aguirre et al. 2012). The required condition for rhodolith bed formation during highstand stages is the absence of terrigenous sedimentation. During lowstand stages, large terrestrial areas are exposed and the base line of erosion in fluvial systems moves downwards. Consequently, erosion and fluvial sediment discharges generally increase, precluding the healthy long-term formation of rhodolith beds. In any case, as stated above, when terrigenous starvation and moderate hydrodynamics concur in a particular area, a sustained rhodolith bed formation is possible.

In the geological record, densely packed rhodolith beds (rhodolith-supported) can be up to several meters thick (Fig. 5.3). Since coralline algal growth is very low (although highly variable from the tropics to high latitudes), the formation of rhodolith beds should imply long time spans. Steller et al. (2009) qualitatively estimated that a rhodolith 3–8 cm in diameter at the Gulf of California would remain on the seabed from years to decades. According to Reid and Macintyre (1988), the 1–3 cm external cover of rhodoliths in different localities, including northern Gulf of Mexico, the Bermuda platform, the eastern Caribbean, and insular shelves of the Canary Islands, has been continuously growing on average for 500 years (see also Littler et al. 1991 for examples in the Florida Keys). A simple calculation implies that to produce a 1-m-thick rhodolith bed due to the continuous accumulation of constantly growing rhodoliths 3 cm across would require about 16,500 year to form. Schäfer et al. (2011, table 6) tabulated growth rates of corallines from different localities and water depths. They range from 0.05 to 5.2 mm year−1 (excluding one value of 0.01 mm year−1). Using these end values, respectively, a rhodolith 3 cm in diameter grow from 600 to 5.8 years. Thus, a 1-m-thick rhodolith bed would be formed between 20,000 and 193 years. Bosence and Wilson (2003) compiled information of accumulation rates of maërl beds from different temperate and tropical areas. Their results show accumulation rates ranging from 0.08 to 1.4 m year−1 (Table 5.1). According to these minimum and maximum values, a 1-m-thick rhodolith bed would form between ~10,000 and 714 years, in agreement with the results estimated above using algal growth rates. A 20-cm thick rhodolith bed on the sea floor on the Alborán Ridge in the western Mediterranean that has been developing for the last 800–1,000 years (Betzler et al. 2011) lies within this range. All calculations imply continuous rhodolith growth and a continuous rhodolith accretion, so that these figures show the minimum time lapse involved in the formation of 1-m-thick rhodolith bed. Obviously, the time lapse involved in the formation of such a rhodolith bed would be longer since neither rhodolith growth nor rhodolith accumulations are continuous. In conclusion, thick fossil rhodolith beds most likely form in hundreds to tens of thousands of years on average.

Table 5.1 Accumulation rates of rhodolith beds from different places
Full size table

In contrast to the geological record, where rhodolith beds of several meters in thickness are found, living rhodolith beds appear to be only superficial features, except for those examples shown by Bosence and Wilson (2003) (Table 5.1). Several hypotheses might be envisaged to account for this “paradox of the present” regarding the long-term rhodolith production in continental shelves: (1) The present-day sea level corresponds to the highstand following the pre-Holocene transgression. As commented above, these are not the best environmental conditions to produce the long-term rhodolith beds and they occur only in areas with low terrigenous input. (2) Thick rhodolith beds developed during the last sea-level rise (Bosence and Wilson 2003) and they are now relic deposits in starved platforms. (3) Present-day rhodolith beds can be thicker than observed (see examples by Bosence and Wilson 2003) but no reliable information on subsurface data is available. Core boring through living rhodolith beds has to be performed.

8 Concluding Remarks

Rhodoliths can provide valuable information to reconstruct past environments, as well as palaeoclimatic conditions. Taxonomic composition of the coralline algal assemblages, growth morphology of algal thalli (both in the interior and surface of the rhodoliths), rhodolith morphology, internal arrangement of rhodolith-builders, taphonomic signatures, and organism interactions (including ichnoassemblages) are key data for the correct palaeoecological interpretations. A uniformitarian approach is important as a source for these inferences. Nonetheless, taphonomic and facies analyses, as derived exclusively from the geological record itself, are required to complement uniformitarian palaeoenvironmental inferences . In this respect, it would be of considerable help to improve taphonomic assessments, both experimental and field-observational analyses, to learn more concerning the preservation behaviour of particular coralline algal taxa, coralline algal growth forms and rhodolith forms. The fossil record also offers the unique possibility of understanding the long-term processes involved in the rhodolith bed dynamics (origination and development). Geological studies demonstrate that the local foundation conditions for the sustained development of rhodolith beds are low sedimentation rates and relatively sheltered settings with intermediate water energy. The maintenance of these stable conditions over time would guarantee the growth of healthy and thick rhodolith beds.

Finally, both short- and long-term studies on coralline algae and rhodolith beds require a precise taxonomic analysis. Bosence (1991), as previously done by Adey and Macintyre (1973), highlighted the necessity of establishing a uniform and consistent taxonomic framework for fossil coralline algae comparable to the taxonomic standards of present-day corallines. Braga et al. (1993) subsequently demonstrated the fact that many of the taxonomic criteria used to identify present-day corallines have been preserved in the fossil counterparts, thus critically influencing the later advance of fossil coralline taxonomy (Braga and Aguirre 1995; Braga 2003). Specific data regarding the fossil coralline taxa forming rhodoliths in Palaeogene and Neogene rhodolith beds are provided by Braga, Brandano and Bassi and coworkers in other chapters of the book.