As genomic and ecological data sets grow larger in size, researchers are flooded with far more information than was available when many conventional model-based approaches were designed. To deal with these massive amounts of data, many researchers have turned to machine learning techniques, which promise the ability to help find signals within the noise of the complex data sets generated by modern sequencing approaches. Applications for machine learning in molecular ecology are broad and include global studies of biodiversity patterns, species delimitation studies, and studies of the genomic architecture of adaptation, among many others. Here at Molecular Ecology Resources, we are excited to highlight research that applies supervised and unsupervised machine learning algorithms to answer questions of interest to the readership of molecular ecology. This special issue will also highlight the nuances and limitations of machine-learning techniques. Rather than focusing on the supposed differences between machine-learning and model-based approaches, this issue would aim to highlight the broad spectrum of machine-learning approaches, many of which can incorporate model-based expectations and predictions.
We are soliciting original research
that applies novel robust applications of machine learning methods on molecular
data to address questions across ecological disciplines.
Details
Manuscripts should be submitted in the usual way through the Molecular Ecology Resources website. Submissions should clearly state in the cover letter accompanying the submission that you wish the manuscript to be considered for publication as part of this special issue. Pre-submission inquiries are not necessary, but any questions can be directed to: manager.molecol@wiley.com
Special issue editors: Nick Fountain-Jones, Megan Smith & Frédéric Austerlitz
Karoo scrub-robin (Cercotrichas coryphaeus) in its typical arid habitat in southern Africa. Photo by Krista N. Oswald.
Written by Ângela M. Ribeiro
Arid environments are ecosystems of energetic stringency. Their typical high temperatures, low primary productivity, and unpredictable water availability prove physiologically challenging for birds. How these vertebrates cope with such harshness remains a conundrum in physiological evolutionary biology. While physiological adaptation likely involves energetic metabolic phenotypes, the underlying mechanisms (plasticity, genetics) are largely uncharacterized. To explore this, we developed a intra-specific level framework (Figure 1) that links environmental conditions, phenotypes and genotypes in a passerine bird whose range spans an aridity gradient. We found variation in energetic physiology phenotypes (a measure of energy expenditure) and gut microbiota composition (involved in energy retrieval from food) to be associated with environmental features and identified a small list of candidate adaptive genes. By working at the interface of physiology and genomics, we suggest that selective pressures on energetic physiology mediated by genes related to energy homeostasis and possibly with contribution of gut microbiota may facilitate adaptation to local conditions. Ultimately, our findings offer a possible explanation to the high avian intra-specific divergence observed in harsh environments, raises awareness that accounting for intra-specific variation is fundamental when modeling physiological responses to climate change, and provides a stepping-stone for further research into the mechanisms of phenotypic adaptation to aridity.
Figure 1. Conceptual framework to infer the mechanisms of physiological adaptation to aridity: linking environment (climate and primary productivity), phenotype (organism-level energetic metabolism: basal metabolic rate and metabolic expansibility; microbiome composition) and genotype (genetic variation in genes underlying the biochemical machinery of energy production).
Ribeiro ÂM, Puetz L, Pattinson NB, Dálen L, Deng Y, Zhang G, da Fonseca RR, Smit B, Gilbert MT. (2019). 31° South: The physiology of adaptation to arid conditions in a passerine bird. Molecular Ecology. 2019. 28-16. 3709-3721.
Individuals within a species vary, and this variation can have important implications for the role a species may play within ecosystems. We compared the relative importance of variation within species due to genetic changes within its own genome versus symbiotic interactions between the focal species and its associated bacteria, also called their microbiome. We focused on Microcystis aeruginosa, a globally distributed photosynthetic cyanobacterium, also known as blue-green algae, that often dominates freshwater harmful algal blooms.
Colony of Microcystis aeruginosa from Gull Lake. Colony photographed by O. Sarnelle of Michigan State University and image prepared by John Megahan of University of Michigan.
These blooms have recently become more common and intense worldwide, causing major economic and ecological damages. We studied Microcystis and their associated microbiomes from lakes in Michigan, USA that vary in phosphorus content, which is the primary limiting nutrient in lakes. We found genomic changes among strains of Microcystis along this phosphorus gradient that indicated increased efficiency in the use of phosphorus and nitrogen. Intriguingly, we found that genotypes adapted to different nutrient environments co-occurred in phosphorus‐rich lakes. This co-occurrence may have critical implications for understanding how Microcystis blooms persist for many months, long after nutrients become depleted within lakes. Similar to previous findings in for example the human microbiome, we uncovered that the bacteria comprising the microbiomes of Microcystis varied in community composition but were more stable at the level of functional contributions to their hosts across the phosphorus gradient. Finally, while our work was mostly focused on unraveling the genomic underpinnings of nutrient adaptation, we also observed consequences of these differences in Microcystis genome and microbiome composition at a physiological level. In particular, when nutrients were provided in abundance, Microcystis (and its microbiome) that had evolved to thrive in low-phosphorus environments could not grow as rapidly as strains from high-phosphorus environments.
– Sara Jackrel, Postdoctoral Fellow, University of Michigan.
Climate change is causing dramatic changes to coral reefs and the eukaryotic life they provide habitat for, but what about the bacterial communities? In this interview with the author, Susana Carvalho from the Red Sea Research Center gives us a behind the scenes take on the paper she and colleagues published using autonomous reef monitoring structures (ARMS) to uncover insights into how bacterial communities respond to environmental stress.
What led to your interest in this topic / what was the motivation
for this study?
Coral reefs face unprecedented decline due to local pressures and
climate change, making assessment of species diversity and responses to
environmental change a priority. Despite the critical roles of bacteria in reef
functioning, their communities remain largely overlooked, partially due to the
lack of standardized tools and protocols. Fostered by the recently developed
Autonomous Reef Monitoring Structures (ARMS), the team – already studying
eukaryotes associated with ARMS – decided to expand the research to bacterial
communities. This step was further motivated based on the knowledge previously
gathered on the eukaryotic reef benthic communities as well as the fact that
the Red Sea can be seen as a natural laboratory for ‘Future Oceans’ due to
clear environmental gradients in sea surface temperature and salinity.
What difficulties did you run into along the way?
The biggest difficulties for this project were logistical and also the
fact that it was an exploratory study. Firstly, the reefs which were studied
were spread across 2000km of the Saudi Arabian coastline. This posed problems
not only in getting permissions from a number of different agencies to visit
the reefs but also in organising the logistics so equipment and people could
undertake the sampling safely over such a wide spatial extent and across
numerous years. Secondly, the nature of the experiment was exploratory, as very
little knowledge is known about reefs in the Red Sea. While this makes the
experiment vital to further knowledge it also complicates the ability to
understand the trends observed as no prior knowledge is available.
What is the biggest or most surprising innovation highlighted
in this study?
While ARMS have previously been used to study the small cryptic eukaryotes (small-sized organisms that are hidden in the reef structure), they had not been utilised to study bacterial communities. For the first time, we employed ARMS to investigate structure and composition of bacterial reef communities across pronounced environmental gradients spanning 16 degrees of latitude. Using this standardized framework, we found that bacterial community structure and diversity aligned with environmental differences. We also found that the decrease in taxonomic diversity was not mirrored by a decrease in functional diversity, suggesting that resilience is not a direct function of taxonomic biodiversity. Importantly, the structure of ARMS devices feature crevices and light-/dark-exposed surfaces as well as exterior and interiors surfaces; in a word it displays many different microhabitats perfectly designed to sample a large diversity of bacteria. This is not possible using other sampling protocols (e.g. water sampling or sediment sampling). The current approach can be expanded to other regions allowing global questions such as the effect of climate change in coral reefs to be addressed and to build a standardized comparable ARMS-based database that allows for meta-analyses beyond the insight from single studies.
Moving forward, what are the next steps in this area of research?
To fully understand how coral reef microbial communities are being affected by local and global stressors and how the functioning of the reef changes, long term time series are required. Thus, repeated sampling over a number of years is critical. Currently, we have 19 reefs along the Red Sea where ARMS are deployed and retrieved every two-years along with traditional reef surveys (photo-transects for benthos and fish visual censuses). To properly understand the function within the reef we also aim to undertake a metatranscriptomic analysis of the sessile community to see which genes are active. This would ideally be expanded to include manipulative experiments to understand how specific stressors affect not only the community composition but also the functional activity of the biological communities.
What would your message be for students about to start developing or using novel techniques in Molecular Ecology?
First of all, we would recommend to
consider fully the question/issue that they would like to address with these
novel techniques. It should be taken in consideration whether the use of
molecular tools will benefit the answering of the desired question in
comparison with more traditional methods. While novel techniques can bring substantial
improvements in understanding questions, such as the current study, sometimes
improvements are just marginal, or can just be in cost and speed.
What have you
learned about methods and resources development over the course of this
project?
The aim of this project was to characterize the taxonomic and functional diversity of coral reef-associated bacteria using ARMS. Whilst the team had previous experience working with bacterial taxonomic characterization, a ‘deep dive’ had to be undertaken to perform the functional analysis. These areas are rapidly developing with a wealth of new algorithms appearing in the literature. While there is no perfect approach, during the course of this project we learnt that one of the most crucial aspects when developing a new method is to understand both the benefits and the limitations of it. Only by gaining this understanding can you confidentially present your results and highlight the areas in which these techniques can be applied.
Describe the significance of this research for your scientific community in one sentence.
ARMS provide a standardized platform to investigate the response of coral reef-associated bacteria to environmental change, with current results suggesting that this research should be conducted from taxonomic and functional perspectives.
Describe the significance of this research for the general scientific community in one sentence.
This study lays the foundation for a holistic understanding of how reef communities respond to environmental changes and proposes a framework that, if applied worldwide, can be vital in providing answers to global questions such as the impacts of climate change on ecosystem diversity and functioning.
Citation: Pearman, JK, Aylagas, E, Voolstra, CR, Anlauf, H, Villalobos, R, Carvalho, S. Disentangling the complex microbial community of coral reefs using standardized Autonomous Reef Monitoring Structures (ARMS). Mol Ecol. 2019; 28: 3496– 3507. https://doi.org/10.1111/mec.15167
Free‐air carbon dioxide enrichment (FACE) experiments have dramatically increased our understanding of how plants may respond to future climate change scenarios. These experiments also provide unique opportunities to better predict how below-ground symbiont communities, crucial to plant health, may respond to climate change. Dr Irena Maček and Prof. Alex Dumbrell give us their behind the scenes insights into their paper that combines Illumina HiSeq sequencing and one of the longest running FACE experiment to find novel insights into mycorrhizal fungi communities and climate change.
The Giessen FACE experimental setup. Photo courtesy of Irena Macek
What led to your interest in this topic / what was the motivation for this study?
Arbuscular mycorrhizal (AM) fungi form mutualistic associations with over two‐thirds of plant species, providing numerous benefits in exchange for carbon. Importantly, different AM fungi provide different plant species with different resources. Thus changes to AM fungal communities can alter plant competition and above-ground productivity. This functional differentiation has motivated both of us to explore the ecological mechanisms regulating AM fungal communities, and it was apparent there was a lack of knowledge on how AM fungal communities respond to elevated atmospheric CO2. This is a major research gap, as AM fungi are entirely dependent on their hosts for carbon and changes in photosynthesis in a high CO2 world may influence this. Thus the chance to sample plant roots from one of the longest running free‐air carbon dioxide enrichment (FACE) experiment in the northern hemisphere in Giessen (Gi-FACE) was an excellent opportunity to address this gap in global climate-change research.
What difficulties did you run into along the way?
Long-term FACE experiments in natural ecosystems are extremely rare, because such set-ups demand stable funding and the persistence of several generations of highly motivated researchers. This would have been a far greater current problem if Prof. Christoph Müller (Justus‐Liebig University Giessen) and his group had not continuously addressed these difficulties since the Gi-FACE experiment’s inception. Another difficulty that we are increasingly finding in metabarcoding research is how to present often extensive and complex data. Dr Dave Clark (University of Essex) was central in working on innovative ways to address this challenge and moving beyond coarse summaries of total community change.
What is the biggest or most surprising innovation highlighted in this study?
For the first time, we have combined a long-term FACE experiment in a natural habitat with high-throughput molecular sequencing (Illumina HiSeq) and new ways of presenting community data. This has allowed us to see subtle population-level responses within broader community-level responses of AM fungi to elevated atmospheric CO2 across the course of a year.
Moving forward, what are the next steps in this area of research?
Understanding the impacts of global-climate change on terrestrial ecosystems requires an integrative approach that explores responses across all levels of biological organisation and spatiotemporal scales, both above- and below- ground. We still lack a comprehensive understanding about how interactions between the above- and below- ground components of biodiversity respond to both acute short-term (e.g. episodic heatwaves, drought etc.) and chronic longer-term climate changes (e.g. warming, elevated CO2). New experiments aimed at addressing these knowledge gaps with robust levels of replication and appropriate experimental durations for capturing longer or shorter-term responses are required, and these must allow for combined sampling of above- and below- ground biota.
Plant root with arbuscular mycorrhizal fungal structures (hyphae and arbuscules). Photo courtesy of Irena Macek
What would your message be for students about to start developing or using novel techniques in Molecular Ecology?
The same as our message for all earlier career researchers – identify your research question, read around your research area, develop your hypotheses and plan an appropriate study to address them, and then choose the correct tools/techniques to conduct the research with. The novel techniques can have a lot of analytical power but can also produce a lot of erroneous data due to the rapid development, a lack of testing and a lack of experience. It is very important to initiate the study with clear questions resulting in hypotheses driven research. Do invest time in skills of data analyses and bioinformatics from the very beginning as there is never enough time to do that the later you are in a career, the more difficult it gets with many other obligations.
What have you learned about methods and resources development over the course of this project?
The field of molecular ecology continuously develops at a rapid pace. It is crucial to have a good network of people with a multidisciplinary range of expertise to collaborate with and to capitalise on all these new and often varied developments. Essentially, good collaboration is crucial. It should also be fun, which is always a good recipe for it to be sustainable and develop into a long-term connection.
Describe the significance of this research for the general scientific community in one sentence.
Elevated levels of atmospheric CO2,reflective of those we will experience in the next ~100yrs, drive changes in symbiotic AM fungal populations with the potential to resonate throughout their associated plant communities, changing above-ground competition dynamics and ecosystem productivity in currently unpredictable ways.
Describe the significance of this research for your scientific community in one sentence.
Predictions regarding future terrestrial ecosystems must consider changes both above-ground and below-ground, but avoid relying on broad‐scale community‐level responses of soil microbes observed on single occasions.
Citation: Maček, I, Clark, DR, Šibanc, N, et al. Impacts of long‐term elevated atmospheric CO2 concentrations on communities of arbuscular mycorrhizal fungi. Mol Ecol. 2019; 28: 3445– 3458. https://doi.org/10.1111/mec.15160Citation:
Estimated to be around 17,000 years old, the Paleolithic paintings in the Lascaux cave of southwestern France give us a rare insight into the history and culture of communities that existed long before modern society. The conservation of caves such as Lascaux is a high priority for historians, scientists, and the general public. The anthropization, or human use, of caves may have dramatic effects on cave-dwelling macro- and micro-organisms, though few studies have been conducted on this topic. By comparing ‘pristine’ caves with anthropized caves frequently visited by humans, Dr. Lise Alonso and colleagues demonstrate that the anthropization of caves is associated with reduced microbial diversity for bacteria and archaea living on cave walls, though microeukaryotes and arthropods were not as strongly affected. In this post, we go behind-the-scenes with Dr. Yvan Moënne-Loccoz on their recent publication in Molecular Ecology and talk about the importance and challenges of working in cave ecosystems.
Great Hall of the Bulls in Lascaux Cave. The cables connect to monitoring probes. Source: DRAC Nouvelle Aquitaine
What led to your interest in this topic / what was the motivation for this study? Cave conservation is an important issue, especially when dealing with caves displaying Paleolithic artwork, as engravings and particularly paintings can be very fragile. There are many of these caves in Dordogne (South-West of France), some of them listed on the UNESCO World Heritage List (https://whc.unesco.org/en/list/85). The most famous Paleolithic cave in Dordogne is the Lascaux Cave, which was closed to the public in the 1960s for conservation reasons. To guide conservation efforts, it is important to understand the ecology and functioning of these caves, especially at the levels of microorganisms and arthropods, which form the main communities present. Against this background, the project was carried out to understand better the biotic communities residing in Lascaux Cave.
Entrance of Lascaux Cave. Source: DRAC Nouvelle Aquitaine
What difficulties did you run into along the way? When dealing with microorganisms and arthropods populating soils, sediments or water, in a majority of cases it is rather straightforward to collect samples and there is no restriction on sample size. In caves, taking samples from walls for microbial analyses, using a scalpel, may leave long-lasting marks. This is an issue in all caves, and particularly so in Paleolithic caves. In the Lascaux Cave, the sample list was prepared after discussions with the cave staff and approved by the cave conservator, and the samples were collected (away from ornate surfaces) by qualified restorers, under the guidance of microbial ecologists, so as to avoid any marks on the wall. It also means that only minute samples were available. Restrictions also apply for the type and location of arthropods traps, as sediments at the bottom of caves might contain historical artefacts.
Sampling of rock wall surface in a pristine cave, using a sterile scalpel. Source: B. Bigaï
What is the biggest or most surprising finding from this study? Caves are oligotrophic environments, so it is always a surprise to find diversified, rather large microbial communities on cave walls. In this study, the Lascaux Cave was compared with eight other caves from the same region, and these caves were quite different from one another in terms of size, architecture, distance from the soil surface, presence/absence of stream underground, human frequentation patterns, etc. Yet, there were clear distinctions in terms of microbial and arthropod communities when comparing anthropized caves versus non-anthropized (almost pristine) caves, which suggests that anthropization was more influential than these cave-specific features. Finally, we were rather surprised to find that prokaryotes (bacteria and archaea) were comparatively more impacted than eukaryotic residents (fungi, other micro-eukaryotes, arthropods) by cave anthropization.
Pristine cave used for sampling. Source: Y. Moënne-Loccoz
Moving forward, what are the next steps for this research? This work was carried out with the Lascaux Cave and eight other caves from Dordogne, which corresponds to a relatively small area. There were at the most 35 km between two caves in this study. Therefore, it remains to be seen whether the results of the current investigation are also relevant elsewhere. At a larger geographic scale, several differences in cave properties can be expected, for instance in geological features (e.g. limestone type) and climatic conditions, which have the potential to influence cave biotic communities. In addition, we evidenced parallel variations in the diversity of microbial and arthropod communities, and it will be important to explore and understand better the ecological interactions between both types of cave inhabitants.
What would your message be for students about to start their first research projects in this topic? First of all, the underground world and the interface between ecology and artwork conservation issues are fascinating, so welcome to the field! More importantly, each cave is different and represents a complex situation of its own, so one can be very busy focusing on a single cave only. This is reflected by the literature on cave microbial ecology, where often a single cave is considered at a time. However, we found that the comparison of different caves, following the path of various groups (e.g. Campbell et al. 2011 J Cave Karst Stud 73:75 ; Hathaway et al. 2014 Geomicrobiol J 31:205 ; De Mandal et al. 2017 BMC Microbiol 17:90 ; Pfendler et al. 2018 Sci Tot Environ 615:1207), brought very interesting insights, so comparative assessments are worth the effort.
What have you learned about science over the course of this project? The majority of participants to this project usually work on soil or aquatic ecosystems, and we found (once again) that concepts and methodology are applicable across different types of ecosystems. More specifically, we realized that underground systems represent interesting models to investigate ecological perturbations, because they are rather confined environments, where community fluctuations in response to mild environmental variations can be documented.
Describe the significance of this research for the general scientific community in one sentence. This research shows that microbiome diversity can be used as a bioindicator of the level of cave anthropization.
Citation Alonso L, Pommier T, Kaufmann B, Dubost A, Chapulliot D, Doré J, Douady CJ, Moënne‐Loccoz Y. Anthropization level of Lascaux Cave microbiome shown by regional‐scale comparisons of pristine and anthropized caves. Molecular Ecology, 28(14), 3383-3394. https://onlinelibrary.wiley.com/doi/10.1111/mec.15144
Picture: Green Anole Lizard (Anolis carolinensis) on railing in Hilo, Hawaii. Author: Paul Hirst. CC-BY-SA-2.5
The green anole (Anolis carolinensis), also called the American chameleon due to its ability to change color, is a common species in South-East USA. It has been studied for decades to understand how reptiles adapt to their environment. Unlike other species of its genus, its range encompasses territories outside tropical climate, reaching the winter-exposed flanks of the Appalachians. The green anole colonized these colder regions from Florida in the last 300,000 years. We used DNA variation covering the whole genome and contrasted populations having recently colonized colder territories with the ones from tropical Florida. We compared multiple approaches to detect which segments in DNA sequences harbored variation compatible with selection. Since these signatures can also be produced by past demography, we took the latter into account to limit the detection of false positives. We then identified the most likely function of genes overlapping with candidate regions for selection, and observed that many of those were involved in exploratory behavior, immunity and response to cold. This suggests that the success of green anoles may have been due to changes in both physiology and behavioral shifts, a hypothesis that could be further tested experimentally.
– Yann Bourgeois and Stephane Boissinot
Bourgeois, Y., & Boissinot, S. (2019). Selection at behavioral, developmental and metabolic genes is associated with the northward expansion of a successful tropical colonizer. Molecular Ecology. 2019. 28-15. 3523-3543
The
spatial representation of species’ data is needed in most areas of biodiversity
related research. In fact, mapping the species’ continuum to guide the
prioritization of areas for conservation was the main driver for PHYLIN
development, but the possible application is far more vast.
Spatial representation of distances between georeferenced samples is challenging. The PHYLIN input are distance matrices and a table of samples classified in groups (lineages, for instance) with locations. PHYLIN relates a matrix measuring a particular distance between samples (for example, a genetic distance) with a matrix representing spatial distance between the same samples. PHYLIN then applies a kriging interpolation: models the relation by means of a variogram and uses that information as weights to interpolate to other locations a probability of belonging to each of the groups
Different applications of PHYLIN with randomly generated data. a) using a simple euclidean distance with 3 dimensions is possible to interpolate over 3d environments; b) using a layer of climate as resistance to movement it is possible to analyse the impact of climate change on connectivity; c) using a Jaccard distance matrix instead of genetic distance to map the contact zone between two species (click on the image for source code).
The latest version of PHYLIN adds the possibility of using multiple spatial distance metrics, opening an exciting avenue with different applications. In our recent paper in Molecular Ecology Resources, we showed how different mechanisms of genetic isolation can be represented in space by PHYLIN. The application of the method is not limited to that and we show here three other possible applications: using 3 dimensional distance (similarly to an ocean environment), climate change connectivity and species distributions/contact zone.
Pedro Tarroso, Guillermo Velo-Antón and Silvia Carvalho
Circadian clocks provide a mechanism that allows organisms to anticipate environmental rhythms, like light-dark cycles. Nematostella vectensis, an estuarine sea anemone, has a surprising degree of overlap in genomic complexity with vertebrates, including circadian clock genes. These genes are predicted to serve a similar role in driving circadian patterns in sea anemones, but we have not worked out the exact mechanism they use.
Photo courtesy of Whitney Leach
In this study, we utilize next-generation sequencing to investigate the time-course transcriptional profiles of animals over 3 days, to dissociate true circadian gene expression vs. photo-responsiveness, by exposing animals to regular light-dark cycles for one month, then abruptly removing the light cue. Hypothesized ‘clock’ genes were rhythmic in the presence of light-dark cycles; however, several of these genes lost their characteristic oscillation after 1 or 2 days in the dark, suggesting lack of endogenous circadian regulation. One would expect a truly circadian gene to continue to cycle in the absence of light, however our results indicate either: 1) the hypothesized ‘clock’ genes simply respond directly to light cues, which implies they are not circadian, or 2) a circadian regulator resides in specific cell types, and the expression signal is too dampened when measuring in the whole animal.
Whitney Leach, Doctoral Candidate, The Reitzel Laboratory, University of North Carolina at Charlotte
Molecular Ecology Spotlight 