Morphological species and cryptic diversity in planktonic foraminifera

Get Complete Project Material File(s) Now! »

Advantages of the GITC*

In addition to preserving the shell (Figure 3), the GITC* protocol has several advantages. It does not require any sophisticated equipment in the field and allows rapid isolation of the specimens on board scientific vessels. Contrary to the classical phenol/chloroform extractions that are time consuming, need much attention during several steps, and require large amounts of material, all steps of DNA extraction (separation-precipitation-dissolution) can be completed in a single tube. This simplifies the handling of large numbers of samples, reduces DNA loss during manipulation, and allows us to adjust the final DNA concentration. For juvenile individuals, PCR reaction may even be carried out in the original extraction tube, although this procedure would not allow shell preservation. In comparative tests (data not shown), the GITC* protocol was also much more effective than other easy DNA extraction methods like proteinase-K digestion (Palumbi et al., 1991), or the DOC protocol (Pawlowski 2000). Finally, the cost is economical compared to the different DNA isolation kits available on the market (about 1 € for 20 samples).

Form and function in pelagic taxa

This discrepancy between morphological and genetic/ecologic differentiation may be the rule rather than an exception in the planktonic realm (Darling et al., 1999; 2004; 2006; de Vargas et al., 1999; 2001; 2002; Stewart et al., 2001), where the evolution of form and function seems to have been controlled by protection from- or interaction with- other organisms, rather than competition for resources and resource space as in most terrestrial and coastal ecosystems (Smetacek 2001). The pelagic realm itself, a three-dimensional and rather homogeneous milieu in perpetual motion, may also act as a force constraining the evolution of size and other morphological features in planktonic taxa (Tappan and Loeblich 1973). In modern foraminifera, this “pelagic selection pressure” on shell design is reflected by the paucity of species morphologically described in the plankton (~50), compared to the benthos (~ 5000) (Armstrong & Brasier, 2005)).
In any case, most common planktonic -and often considered as cosmopolitan- morpho-species analysed so far using PCR based technologies, from the pico to the micro fractions, were found to be composed of a complex of different genetic entities (cyanobacteria and prochlorophytes -Fuhrman and Campbell, 1998-, haptophytes and diatoms -Bucklin et al., 2001; Rynearson and Armbrust, 2004; Amato et al., 2007-, dinoflagellates -Scholin et al., 1995; Bolch et al., 1999; Logares et al., 2007-, foraminifera -Kucera and Darling, 2002; de Vargas et al., 2004; Darling and Wade, 2008-, copepodes -Bucklin et al. 1996-, fishes -Miya and Nishida, 1997). Moreover, it is increasingly obvious that the distinct phylogenetic groups within highly similar morphology often have, independently of geographic distance, different horizontal or vertical biogeographic ranges (Huber et al., 1997; Toledo and Palenik, 1997; de Vargas et al., 1999; 2001; 2002; Morard et al., 2009), physiological features (Huber et al., 1997; Rynearson and Armbrust, 2004), or even radically distinctive genomic compositions (Moore et al., 1998). These results should now spawn a search for estimating the complexity of this hidden world of diversity at the genetic and/or physiological levels is now necessary in order to decipher the functional entities structuring the marine ecosystem, and to draw a valid biogeography of the global Ocean. This is particularly important for the skeleton-bearing planktonic taxa, i.e. the diatoms, coccolitophores, foraminifers, radiolarians ostracods, and pteropods, because (1) they are the principle actors of carbon export to the seafloor, and thus of great importance for the possible biological response to atmospheric changes due to human activities and (2) they have the most complete and well-preserved fossil records in oceanic sediments, a fundamental archive for the study of organismic and climate evolution.

DNA extraction, amplification and RFLP analysis

DNA extractions of 383 specimens were performed using the classical guanidium isothiocyanate (GITC) DNA extraction buffer (for details see de Vargas et al., 1997). For the remaining 306 specimens, the original composition of the GITC extraction buffer was modified by removing EDTA, which has the property to dissolve CaCO3 (Wade & Garcia-Pichel, 2003; de Vargas et al., 2003). We find that, even in the absence of EDTA, the buffer is able to efficiently penetrate the shell of planktonic foraminifera and isolate the nucleic acids. This so-called GITC* protocol makes the DNA extraction step non-destructive for the calcareous shell, and allows us to obtain, from the same individual, both the calcareous shell and the DNA for combined morphological and genetic analyses.
In this study, PCR amplification of a ~1100 base pair (bp) fragment, localized at the terminal 3’ end of the small subunit (SSU) rRNA gene, was carried out for each individual of Orbulina universa. Methods of PCR amplification, PCR product purification and cloning, as well as the foraminifera-specific primers used in this study are described by de Vargas et al. (1997). The length of the 689 O. universa SSU rDNA sequences were compared to those obtained from specimens collected in the Atlantic Ocean by de Vargas et al. (1999).
The ribosomal PCR products were further analysed through restriction fragment length polymorphism (RFLP). Although the RFLP method does not allow the detection of new cryptic species of Orbulina universa within our sampled material (because it fails to recognize small changes characterised by a few substitutions), it allowed us to easily and quickly determine whether our 689 specimens were related to any of the RFLP genotypes described by de Vargas et al. (1999). We used the Sau96I restriction enzyme to cut the nucleotide sequence at a specific pattern for each genotype (for details see de Vargas et al., 1999) and to discriminate between the different genotypes. Distinct patterns for each genotype were UV-detected after migration of the digested PCR products on 1.5% agarose gel and ethidium bromide coloration.

Geographic distribution of genotypes and hydrography

Although the Sargasso species of Orbulina universa is yet to be recognized in the Indian Ocean (Table 1), our dataset is sufficiently detailed to identify the overall distribution of the Caribbean, Mediterranean and Sargasso species. It confirms that the geographic distribution of these cryptic species is not random (Table 1). For example, among the 42 collected stations, 19 yielded the Mediterranean species alone while the three species co-occurred at 7 stations only.
Our data confirm the observation of de Vargas et al. (1999) that the geographic distribution of the cryptic species of Orbulina universa is probably correlated with the primary productivity of the surface waters. The Mediterranean species is more abundant in nutrient-rich waters, where higher levels of chlorophyll-a concentration in the mixed layer are recorded. In the eastern Atlantic, the Mediterranean species is found in frontal zones in the subtropical and equatorial current systems (Figure 3). Its presence off West Africa is likely related to coastal upwelling forced by the trade winds, as pointed out by de Vargas et al. (1999; 2003). In the southern Indian (Figure 4) and Pacific (Figure 5) Oceans, the Mediterranean species is also found in areas of high levels of chlorophyll-a concentration in the mixed layer, within or close to the frontal zones of the subtropical current systems. The Caribbean and Sargasso species of O. universa display similar distributions to one another, and it remains difficult to characterize productivity-related specializations between these cryptic species. Both species are systematically collected in areas of oligotrophic oceanic conditions, typically in stratified water masses of the subtropical gyres (Figs. 3, 4, 5). Co-occurrences of all three cryptic species (Table 1) in AMT-8 stations 12,5.1. Environmental significance of shell porosity in Orbulina universa.
The relationship between shell porosity in planktonic foraminifera and environmental parameters was first recognized by empirical research (Bé, 1968; Frerichs et al., 1972; Bé et al., 1973; 1976) and later confirmed in culture studies (Bijma et al., 1990). These studies have suggested that in most planktonic foraminiferal taxa, shell porosity tracks latitude and is correlated with temperature as a function of oxygen solubility in water. Planktonic foraminiferal shell porosity has been subsequently considered as climatically significant and used as a paleoceanographic index (Colombo and Cita, 1980; Haenel, 1987; Fisher, 2003; Fisher et al., 2003).
Bé et al. (1973) examined the variation in shell size and shell porosity of Orbulina universa specimens from plankton net tow and sediment samples from the Indian Ocean. Size and shell porosity of the final chamber were found to correlate with latitude. According to these authors, the geographic distribution of shell porosity in O. universa displays clinal patterns of higher porosities in the warm waters of the tropical areas, and lower porosities occurring at higher latitudes, where surface waters are cooler. Our results suggest that this relationship originates from collected samples in which Bé et al. (1973) mixed up at least two cryptic species of O. universa (see previous section; Figs. 11 and 12; see also Figure 13c of Bé et al., 1973).
Indeed, the shells of the Caribbean species, found in the Indian Ocean in the oligotrophic and warm waters of the subtropical zone (Figure 4), are characterized by variable but relatively high porosity values linked with larger pores (Figure 13). When coupled with the SST-PSD relationship observed for this cryptic species (Figure 12), this high PSD-variability suggests that it may be actually made of two, yet unseparated distinct cryptic species based on the analysis of rDNA 18S.
15 and 16 (Figure 3), REVELLE stations 11, 12 and 14 (Figure 5) and C-MarZ station 5 (Figure 6A) probably reflect transitional water mass mixing of frontal zones but it is also possible that the species are depth stratified.

READ Beyond the bag-of-words model: using dependency graphs

Cryptic diversity and ecophenotypy in planktonic foraminifera

Since first proposed by Parker (1962), numerous studies have shown that planktonic foraminiferal morphospecies exhibit morphological gradients with latitude. The significance of these observed gradients has not been clearly understood and the various morphotypes were classically considered as ecophenotypic variants (Kennett, 1968a; 1968b; Malmgren and Kennett, 1972; Hecht, 1974; Healy-Williams et al., 1985; Williams et al., 1988; Healy-Williams, 1992). We find that morphological variation in Orbulina universa clearly owes much to the expression of different cryptic species adapted to different environments.
Erroneous attributions of morphological clines to environmental factors may be more common than previously suspected in planktonic foraminifera. In the first genetic study of cryptic diversity in planktonic foraminifera, Huber et al. (1997) discovered genetic differences between two cryptic species of Globigerinella siphonifera that matched differences in their shell size, porosity, spine density and depth-related ecology (see also Bijma et al., 1998). In addition, inter-basin genetic studies of this spinose and symbiont-bearing morphospecies (Darling et al., 1999; de Vargas et al., 2002) have highlighted geographic segregation among cryptic species that is correlated with water mass stratification and productivity. These results strongly suggest that previously described ecophenotypes within G. siphonifera (see for example Parker, 1962; Hecht and Savin, 1972) correspond to different biological species adapted to distinct environmental conditions.
It is premature to speculate about cryptic diversity and ecophenotypy relationships in other morphospecies of spinose planktonic foraminifera because we lack large scale inter-basin genetic analyses and/or direct morphologic characterization of the cryptic diversity within these taxa. However, some of them may constitute good candidates for better understanding the significance of morphological clines in planktonic foraminifera. For example, Globigerinoides ruber and Globigerina bulloides – apparently composed of four (Darling et al., 1997; 1999; Kucera and Darling, 2002) and six (Darling et al., 1999; 2000; Stewart et al., 2001; Kucera and Darling, 2002) distinct cryptic species, respectively – show ambiguous taxonomic status with latitudinally divergent morphotypes (see Parker, 1962; Orr, 1969; Hetcht, 1974, Robbins and Healy-Williams, 1991 for G. ruber; Bandy, 1972; Malmgren and Kennett, 1976; 1978 for G. bulloides). The modest number of specimens surveyed to date for both G. ruber and G. bulloides suggests that there may well be additional undiscovered genotypes in both clades.
Further evidence for relationships between cryptic diversity and ecophenotypy can be found in the non-spinose planktonic foraminifera. Healy-Williams et al. (1985) showed that the population structure of Globorotalia truncatulinoides in the southern Ocean consists of a complex of three sub-populations with distinct morphologies and depth-related specializations whose abundance varies with latitude. Single-cell DNA and morphometric analyses performed on plankton tow specimens from the South Atlantic have revealed that G. truncatulinoides actually corresponds to a complex of at least four cryptic species adapted to particular hydrographic conditions (de Vargas et al., 2001). However, they showed that two of these four cryptic species are characterized by similar morphologies and geographic occurrences. Accordingly, as for Orbulina universa, the morphological cline evidenced by Healy-Williams et al. (1985; see also Kennett, 1968a) is very likely to be the expression of different biological species adapted to different environments (de Vargas et al., 2001).

Table of contents :

Chapitre 1 : Protocole d’extraction ADN
Abstract
1. Introduction
1.1. Excellence and caveats of pelagic fossil records
1.2. Morphological versus genetic species concepts in planktonic foraminifera
1.3. The value of direct morpho-genetic comparisons
1.4. Orbulina universa d’Orbigny
2. Material and methods
2.1. Sample collection
2.2. DNA extraction
2.3. Molecular analyses
2.4. Shell preservation
3. Results
3.1. Efficacy of the GITC*
3.2. GITC* and shell preservation
3.3. Identification of genotypes
3.4. Distribution of genotypes and hydrography
4. Discussion
4.1. Advantages of the GITC*
4.2. Form and function in pelagic taxa
5. Conclusion
Acknowledgments
References
Chapitre 2 : Diversité morphologique des espèces cryptiques d’Orbulina universa (d’Orbigny)
Abstract
1. Introduction
1.1. Morphological species and cryptic diversity in planktonic foraminifera
1.1. Orbulina universa d’Orbigny, 1839
2. Material
3. Methods
3.1. DNA extraction, amplification and RFLP analysis
3.2. Biometry
3.2. Statistical analysis
4. Results
4.1. Identification of genotypes
4.2. Geographic distribution of genotypes and hydrography
4.3. Biometry
5. Discussion
5.1. Environmental significance of shell porosity in Orbulina universa
5.2. Cryptic diversity and ecophenotypy in planktonic foraminifera
5.3. Implications for paleoceanography
6. Conclusion
Acknowledgments
References
Chapitre 3 : Diversité morphologique des espèces cryptiques de Truncorotalia truncatulinoides d’Orbigny
Abstract
1. Introduction
2. Material
3. Methods
3.1. DNA extraction, PCR amplification and RFLP analysis
3.2. Shape Descriptors
3.3. Shape Analyses
4. Results
4.1. Identification of genotypes
4.2. Geographic distribution of genotypes
4.3. Morphological differentiation among cryptic species
4.4. A model for morphological recognition of cryptic species
5. Discussion
6. Conclusion
Acknowledgments
References
Chapitre 4 : Diversité cryptique de Globoconella inflata
Abstract
1. Introduction
2. Results
2.1. Genetic Variation in Globoconella inflata
2.2. Geographic Distribution of Genotypes
2.3. Inter-Hemisphere comparative distribution of G. inflata in surface sediments
2.4. Biometry
3. Discussion
3.1. ITS for Identifying Genetic Variability in Planktonic Foraminifera
3.2. Cryptic Diversity as a Tool for Monitoring Past Migrations of the Antarctic Subpolar Front
4. Material and Methods
4.1. Sample Collection
4.2. DNA Extraction, Amplification and Sequencing
4.3 DataSets, Alignement and DNA Sequence Analysis
4.4. RFLP Analysis.
4.5. Spatial and Environmental Distribution of Globoconella inflata
4.6. Biometrics
Acknowledgments
References
Chapitre 5 : Impact de la diversité cryptique sur les reconstitutions paléocéanographiques
Abstract
1. Introduction
2. Material
2.1. Sample collection
2.2. Core-top and environmental dataset
3. Methods
3.1. Molecular analysis
3.2. Theoretical distribution of genotypes
3.3. Training sets
3.4. Transfer functions
3.5. Temperature estimation robustness
4. Results
4.1. Phylogenetic analysis of Globigerina bulloides
4.2. RFLP identification of genotypes
4.3. Distribution of genotypes
4.4. Modeled abundances
4.5. Improvement of assemblage-based SST reconstructions
4.6. Robustness
5. Discussion
5.1. Improvement of the transfer function
5.2. Contribution of the cryptic species
6. Conclusion
Acknoledgments
References
Conclusions
Bibliographie