Resumen de Benthic diatom metabarcoding: developing new approaches to research and biomonitoring in aquatic ecosystems
- español
Las diatomeas han sido ampliamente estudiadas debido en parte a su gran importancia ecológica y al hecho de ser excelentes bioindicadores. No obstante, la identificación de las diatomeas a nivel de especie es una tarea que requiere de tiempo y conocimientos taxonómicos. Recientemente, el método ADN metabarcoding ha sido contemplado como una alternativa a las identificaciones morfológicas basadas en microscopía óptica (MO), si bien es cierto que existen varios factores que pueden disminuir la eficacia del metabarcoding. El objetivo principal de esta tesis es evaluar la aplicabilidad de la técnica metabarcoding para el seguimiento de sistemas acuáticos bajo la Directiva Marco del Agua, así como para abordar cuestiones ecológicas y biogeográficas. Nuestros resultados indican que el metabarcoding, basado en el marcador cloroplástico rbcL, constituye una alternativa eficaz al MO para el seguimiento de ríos mediterráneos, como los de Cataluña, debido principalmente a que las especies de diatomeas bentónicas más comunes de estos ríos están representadas en la biblioteca de referencia. Por contra, las especies marinas están en menor medida representadas lo cual conlleva que el metabarcoding no sea actualmente una alternativa a la MO en sistemas costeros, pero sí una herramienta complementaria capaz de detectar especies difícilmente identificables por MO como especies pequeñas, delicadas o desprovistas de frústulos. También hemos visto que la elección entre los dos marcadores rbcL usados actualmente en metabarcoding tiene pocas implicaciones para los programas de seguimiento, pero sí para estudios de biodiversidad. Esta tesis también muestra como las variantes genéticas de varios complejos de especies de diatomeas difieren en sus preferencias ecológicas y distribución espacial. Además, el análisis de una base de datos global (rbcL) de diatomeas de agua dulce nos ha permitido observar que las diatomeas pennadas presentan mayor diversidad intraespecífica que las céntricas e identificar cuatro patrones filogeográficos comunes entre las especies analizadas.
- català
Les diatomees han estat àmpliament estudiades en bona part degut a la seva rellevància ecològica i per ser excel·lent bioindicadores. No obstant, la identificació de les diatomees a nivell d'espècie requereix temps i personal amb coneixements taxonòmics. Recentment, el mètode d’ADN metabarcoding s’ha postulat com una alternativa a les identificacions morfològiques basades en microscòpia òptica (MO), val a dir però, que hi ha diversos factors que poden disminuir l'eficàcia del metabarcoding. L´objectiu principal d´aquesta tesi és avaluar l´aplicabilitat de la tècnica metabarcoding per al seguiment dels sistemes aquàtics sota la Directiva Marc de l´Aigua, així com per abordar qüestions ecològiques i biogeogràfiques de les diatomees. Els nostres resultats indiquen que el metabarcoding, basat en el marcador cloroplàstic rbcL, constitueix una alternativa eficaç al MO per al seguiment de rius mediterranis, com els que hi ha a Catalunya, bàsicament gràcies a que les espècies de diatomees bentòniques més comunes d'aquests rius estan representades a la biblioteca de referència. Per contra, les espècies marines hi són molt menys representades i això comporta que el metabarcoding no pugui ser encara una alternativa a la MO en sistemes costaners, però sí una eina complementària capaç de detectar espècies difícilment identificables per MO com són espècies petites, delicades o desproveïdes de frústuls. També hem vist que l'elecció entre els dos marcadors rbcL usats actualment en metabarcoding té poques implicacions per als programes de seguiment, però sí per a estudis de biodiversitat. Aquesta tesi també mostra com les variants genètiques de diversos complexos d'espècies de diatomees difereixen a les seves preferències ecològiques i distribució espacial. A més, l'anàlisi d'una base de dades global (rbcL) de diatomees d'aigua dolça ens ha permès observar que les diatomees pennades presenten més diversitat intraespecífica que les cèntriques i alhora identificar quatre patrons filogeogràfics comuns entre les espècies analitzades.
- English
Diatoms have been one of the most studied groups of protists, partly because they are rich in morphological characters, relative to many flagellate and amoeboid groups, with a siliceous cell wall varying greatly in size, shape and patterning. They are also abundant in aquatic systems and important in biogeochemical cycles. They have applications in biotechnology and stratigraphy and, particularly relevant to this thesis, they are excellent biological indicators. However, estimates suggests that only a small proportion of the total number of extant species have been described so far and many aspects of their ecology remain unknown. One problem in ecological studies and biomonitoring is that they require identification of hundreds of individuals at the species level, which is time-consuming task requiring expert knowledge and considerable microscopical skills. Furthermore there is increasing evidence of cryptic or pseudocryptic species, which differ in few orno discernible morphological characteristics; consequently their geographical distributions and ecological preferences will remain unclear until identification is practica!.
DNA metabarcoding (high-throughput sequencing [HTS] of a particular short marker) has recently emerged as an alternative to species identifications based on light microscopic examination (LM). This technology is transforming the way protist diversity can be studied, as thousands of DNA strands can be sequenced in parallel at once, allowing entire communities to be characterised from environmental samples in a relatively simple procedure, to complement the less extensive but richer data provided by microscopy. A key question, however, is the extent to which metabarcoding data faithfully reflect the natural communities from which they are derived. The answer to this question depends on a multitude of factors, including the genetic marker selected, the communities being studied, the molecular processing of the samples and the bioinformatics pipeline used, among others. AII of these steps can introduce biases.
The main objective of this thesis was to evaluate the potential and difficulties of using of DNA metabarcoding for the characterisation of sorne benthic diatom communities in freshwater and coastal environments. There is a special focus on the applicability of the method for Water Framework Directive (WFD) bioassessment of Mediterranean rivers, but we also address ecological and biogeographical questions. The work is organized into 5 chapters, 3 of them represent independent papers that have been published, 1 is under revision and 1 in preparation; these are: 1) Evaluation and sensitivity analysis of diatom DNA metabarcoding for WFD bioassessment of Mediterranean rivers. 2) Assessment of marine benthic diatom communities: insights from a combined morphological-metabarcoding approach in Mediterranean shallow coastal waters. 3) Evaluation of two short and similar rbcl markers for diatom metabarcoding of environmental samples: effects on biomonitoring assessment and species resolution. 4) DNA metabarcoding reveals differences in distribution patterns and ecological preferences among genetic variants within sorne key freshwater diatom species. 5) Phylogeographical patterns in freshwater diatoms revealed by DNA metabarcoding of a short rbcl marker.
Our results for WFD biomonitoring of Mediterranean rivers in Catalonia (NE Spain) builds on previous research, especially in France and the UK, and indicates that rbcl metabarcoding of benthic diatoms constitutes an efficient and reliable alternative to LM. One reason for this is that WFD ecological assessments in Catalonia are driven by a relatively small number of common species, for which rbcl data are available; i.e. the DNA reference library (essential for converting DNA data into relative abundances of named taxa) is adequate in this region, in contrast to sorne other regions, e.g. Fennoscandia. However, metabarcoding cannot be considered as an alternative to LM in coastal environments, because of the low coverage of marine benthic diatom species in the reference library. On the other hand, metabarcoding has especial advantages in coastal environments because of its ability to capture information on very delicate or small diatoms and diatoms that exist as endosymbionts in foraminifera and dinoflagellates. In addition, we found that a useful by-product of the rbcl metabarcoding protocol is records of other Ochrophyta and Chlorophyta that are co-amplified with diatoms; these can include rarely recorded groups and species.
There are biases in metabarcoding assessments of diatom communities, relative to microscopical ones, which need to be taken into account when interpreting results. One bias is caused by interspecific variation in rbcl copy number per cell and this seems to have been a factor explaining sorne discrepancies in the relative abundance of species between LM and DNA metabarcoding in both freshwater and coastal systems. In one case, a major difference between outcomes was the result of failure of the LM approach to faithfully record the abundance of a diatom, Fistulifera saprophila, that is destroyed by harsh cleaning methods. We also investigated the relative advantages of two rbcl markers that have been proposed (263-bp and 331-bp target regions), finding that the choice has few implications for WFD biomonitoring programmes, but sorne implications for biodiversity analyses, because of the higher resolution of the 331-bp marker; this allows identification at the species level of certain genetic variants that cannot be separated by the 263-bp marker. In addition, use of the longer marker seems to be more efficient for classifying when using a na"fve Bayesian classifier.
Combining metabarcoding and environmental data for Catalan and French rivers, we created a dataset to investigate whether different Amplicon Sequence Variants (ASVs) within species have the same or different ecological preferences, showing that in Achnanthidium minutissimum and Fistulifera saprophila, but not in e.g. Nitzschia inconspicua, ASVs could be separated into different ecological groupings, sorne of which deviated from the generally accepted characterizations of the species.
We reanalysed available N American, European and Asían HTS outputs to assemble a large combined dataset of 263-bp Amplicon Sequence Variants (truncating any 331-bp ASVs to 263 bp). We used this to address ecological and biogeographical questions, finding high intraspecific heterogeneity in many cases. Four common phylogeographic patterns were distinguished, which correlate to sorne extent with biological characteristics: centric diatoms (which are predominantly oogamous and have multiple chloroplast per cell) tend to show lower intraspecific diversity than pennates. 263-bp rbcl variants from many species were widely distributed in Europe, North America, the lndian Ocean and Asia, supporting rapid dispersion of diatoms relative to rbcl divergence. We concur with other recent assessments of the huge opportunities offered by DNA metabarcoding, even in its current state, which will increase further with technical adjustments and roll-out of long-read technologies.
