249growth rate exists, although not strict, it also corroborates the
observation of a moderate to low growth rates (Simek et al.
2006 ; Livermore et al. 2014 ). In the same way these results
are coherent with the small cell size reported for lacustrine
Actinobacteria (Hahn et al. 2003 ) and their presence in the
LNA (Low Nucleic Acid) fraction of lacustrine environ-
ments. Moreover, freshwater actinobacterial genomes show
small genome sizes (1,2 Mb for acI-B1, 3 Mb for acTH2 –
Livermore et al. 2014 ), a low G + C content as compared to
known Actinobacteria , ranging from 37 to 53 % and small
intergenic distances (Ghai et al. 2014 ). Altogether these
observations suggest that some freshwater Actinobacteria
clades may be subjected to genome streamlining (Garcia
et al. 2012 ), or at least genome reduction. It has been postu-
lated that genome streamlining and small cell size could be
driven by selection for living in nutrient poor environments
(Giovannoni et al. 2014 and references therein). This hypoth-
esis does not hold for freshwater Actinobacteria as they were
reported to be prevalent in a large spectrum of trophic status
(Newton et al. 2006 ; Debroas et al. 2009 ; Humbert et al.
2009 ). Another hypothesis is that small cell size promotes
better escape from bacterivores and phages. Some studies
have indeed reported actinobacterial reduced grazing load
(see references in Newton et al. 2011 ). However, the link
between genome reduction and selection against predation
remains unclear, especially in the designation of which gene
could be lost. Loss of non-essential metabolic pathways and
regulatory machinery is possible for intracellular bacteria
which live in a stable environment (Batut et al. 2014 ), but it
seems diffi cult to consider lakes as stable environments for
free living bacteria. Moreover, for a genome reduction to be
adaptive, “selective gene loss” should correlate with ecologi-
cal niches. Recent literature suggests potential specialization
in substrate preference for freshwater Actinobacteria (Ghylin
et al. 2014 ). However, work is still to be done to draw the
metabolic potential of these organisms and understand the
selective cause, if any, of this genome reduction.
15.3 Metagenetics for Assessing
the Activity of Archaea
15.3.1 Archaea in Freshwater Ecosystems
Archaea have been the focus of intensive research in the last
decades because of their ubiquity and abundance in aquatic
ecosystems. Nevertheless, data available for freshwater eco-
systems are still scarce compared to marine systems mainly
due to the physico-chemical and limnological heterogeneity
of the environments studied (rivers, lakes, inland waters).
The recovery of archaeal genes encoding the ammonia
monooxygenase subunit A ( amoA) , within Thaumarchaeota
Marine Group I, in many marine and terrestrial environments
(Schleper et al. 2005 ) confi rms the global importance of
ammonia oxidizing Archaea (AOA). AOA are also detected
in oxic layers of freshwater ecosystems ( e.g. Hugoni et al.
2013a , b ), however, in lacustrine deeper layers ( i.e. anoxic
zone and above sediments) AOA are rare (Llirós et al. 2010 )
and methanogenesis performed by Euryarchaeota is often
the most important process. Moreover, initially thought to
occur only in anoxic environments, methane production has
been recently described in a well-oxygenated water column
of an oligotrophic lake (Grossart et al. 2011 ). Thus,
Euryarchaeota may also be important players in carbon
cycling in oxic aquatic environments and water to air meth-
ane fl ux.15.3.2 Assessing Community Structure
of Active Archaeal PopulationsWithin freshwater ecosystems, meromictic lakes with per-
manent oxic/anoxic interfaces constitute optimal study sites
to monitor archaeal population changes through physico-
chemical variations in their habitats. However, most studies
investigating lacustrine ecosystems have focused on archaeal
diversity based on gene abundance ( i.e. metagenetics), but
only a few studies have examined active archaeal assem-
blages (La Cono et al. 2013 ). Recent studies have shown the
importance of differentiating the active communities from
the total communities (Hugoni et al. 2013a , b ). One method
to explore an aspect of their activity ( i.e. , the growth rate for
specifi c taxa) is then to investigate microbial communities at
both the 16S rRNA genes (DNA) and 16S rRNA level
(RNA). Nevertheless, the use of 16S rRNA should be
addressed considering the environmental parameters and
specifi c taxa due to some inconsistent relationships between
16S rRNA and activity (Blazewicz et al. 2013 ). Thus, the
changes in the community structure of active archaeal popu-
lations could be evaluated over time by pyrosequencing 16S
rRNA genes and 16S rRNA.
This method allowed correlating the presence of some
taxonomic groups with their potential activity, as in the epi-
limnion of Lake Pavin, where archaeal assemblage was
clearly dominated by Thaumarchaeota Marine Group I
(MGI, 83 % of the reads in the 16S rRNA genes dataset)
(Fig. 15.2a ), which was also the most active group in this
zone (99.6 % of the reads in the 16S rRNA dataset). Some
other groups, such as the Lake Dagow Sediment (LDS) clus-
ter or the Deep-sea Hydrothermal Vent Eukaryotic Group 6
(DHVEG-6) were present in the 16S rRNA genes fraction in
the epilimnion (Fig. 15.2 ) even if they accounted for a negli-
gible proportion of potentially active archaeal fraction. The
LDS cluster is a highly diverse euryarchaeal lineage
(Barberán et al. 2011 ) identifi ed in rivers (Galand et al. 2006 )
where they accounted for a large proportion of the archaeal15 Omic Approaches for Studying Microorganisms and Viruses