\makeatletter
\@ifundefined{HCode}
{\documentclass[screen,CRGEOS,Unicode,biblatex]{cedram}
\addbibresource{crgeos20250738.bib}
\newenvironment{noXML}{}{}
\let\citep\parencite
\let\citet\textcite
\def\xcitealp#1#2{\citeauthor{#1}, \citelink{#1}{#2}}
\def\citealt#1#2{\citeauthor{#1}, \citelink{#1}{#2}}
\newcommand*{\citelink}[2]{\hyperlink{cite.\therefsection @#1}{#2}}
\def\defcitealias#1#2{}
\let\citetalias\textcite
\let\citepalias\parencite
\def\thead{\noalign{\relax}\hline}
\def\endthead{\noalign{\relax}\hline}
\def\tbody{\noalign{\relax}}
\def\tabnote#1{\vskip4pt\parbox{.98\linewidth}{#1}}
\def\tsup#1{$^{{#1}}$}
\def\tsub#1{$_{{#1}}$}
\def\ndash{\text{--}}
\def\hyphen{\text{-}}
\def\og{\guillemotleft}
\def\fg{\guillemotright}
\def\0{\phantom{0}}
\def\rmDelta{\Delta}
\makeatletter
\g@addto@macro{\UrlBreaks}{\UrlOrds}
\gappto{\UrlBreaks}{\UrlOrds}
\RequirePackage{etoolbox}
\usepackage{upgreek}
\def\inlinefig#1{\includegraphics{#1}}
\def\jobid{crgeos20250738}
%\graphicspath{{/tmp/\jobid_figs/web/}}
\graphicspath{{./figures/}}
\def\xmorerows#1#2{#2}
\newcounter{runlevel}
\usepackage{multirow}
\def\nrow#1{\@tempcnta #1\relax%
\advance\@tempcnta by 1\relax%
\xdef\lenrow{\the\@tempcnta}}
\def\morerows#1#2{\nrow{#1}\multirow{\lenrow}{*}{#2}}
\let\MakeYrStrItalic\relax
\def\refinput#1{}
\let\ubreak\break
\csdef{Seqnsplit}{\\}
\def\botline{\\\hline}
\def\back#1{}
\newcommand*{\hyperlinkcite}[1]{\hyper@link{cite}{cite.#1}}
\DOI{10.5802/crgeos.319}
\datereceived{2025-09-12}
\daterevised{2025-11-23}
\dateaccepted{2025-11-24}
\ItHasTeXPublished
}
{
\PassOptionsToPackage{authoryear}{natbib}
\documentclass[crgeos]{article}
\usepackage[T1]{fontenc}
\def\CDRdoi{10.5802/crgeos.319}
\def\citelink#1#2{\citeyear{#1}}
\def\xcitealp#1#2{\citealp{#1}}
\let\ubreak\relax
\makeatletter
\def\CDRsupplementaryTwotypes[#1]#2#3{}
\def\hyperlinkcite#1#2{\citeyear{#1}}
}
\makeatother

\usepackage{upgreek}

\begin{document}

\begin{noXML}

\CDRsetmeta{articletype}{research-article}

\TopicFR{G\'eochimie, cosmochimie}
\TopicEN{Geochemistry, cosmochemistry}

\title{Paleosalinity reconstruction of the Francevillian Basin (2.1
Ga): geochemical insights into the depositional environment of early
macroscopic life}

\alttitle{Reconstitution de la pal\'{e}osalinit\'{e} du bassin de
Francevillian (2,1 Ga) : apports g\'{e}ochimiques \`{a} la
compr\'{e}hension de l'environnement de d\'{e}p\^{o}t des premi\`{e}res
formes de vie macroscopiques}

\author{\firstname{Anna} \lastname{El Khoury}\CDRorcid{0009-0003-7323-6049}}
\address{Universit\'{e} de Poitiers, IC2MP, UMR 7285, CNRS, 86073
Poitiers, France}
\address{NANOSCOPIUM Beamline, Synchrotron SOLEIL, 91190, Saint-Aubin,
France}
%\email{annackhoury@hotmail.com}

\author{\firstname{Ibtissam} \lastname{Chraiki}\CDRorcid{0000-0001-5940-2192}}
\addressSameAs{1}{Universit\'{e} de Poitiers, IC2MP, UMR 7285, CNRS,
86073 Poitiers, France}
%\email{ibtissamchraiki@gmail.com}

\author{\firstname{Claude} \lastname{Fontaine}\CDRorcid{0000-0001-5686-6725}}
\addressSameAs{1}{Universit\'{e} de Poitiers, IC2MP, UMR 7285, CNRS,
86073 Poitiers, France}
%\email{claude.fontaine@univ-poitiers.fr}

\author{\firstname{Andrea} \lastname{Somogyi}\CDRorcid{0000-0002-3814-4152}}
\addressSameAs{2}{NANOSCOPIUM Beamline, Synchrotron SOLEIL, 91190,
Saint-Aubin, France}
%\email{andrea.somogyi@synchrotron-soleil.fr}

\author{\firstname{Abderrazak} \lastname{El Albani}\CDRorcid{0000-0001-9598-955X}\IsCorresp}
\addressSameAs{1}{Universit\'{e} de Poitiers, IC2MP, UMR 7285, CNRS,
86073 Poitiers, France}
\email[A. El Albani]{abder.albani@univ-poitiers.fr}

\shortrunauthors

\keywords{\kwd{Francevillian}\kwd{Biota}\kwd{Salinity}\kwd{Proterozoic}}

\altkeywords{\kwd{Francevillien}\kwd{Biota}\kwd{Salinit\'e}\kwd{Prot\'{e}rozoique}}

\thanks{La R\'{e}gion Nouvelle Aquitaine (Project Darwin, grant
E458CR25), French government program ``Investissements d'Avenir'' (EUR
INTREE, reference ANR-18-EURE-0010), Synchrotron SOLEIL,  ANR-BIOGEN,
grant, 22-CE49-0010-02)}

\begin{abstract}
The Francevillian Basin of Gabon (2.1 Ga) hosts one of the earliest
known macroscopic organisms preserved within the fossiliferous FB2b
subunit. Understanding the depositional conditions of this interval is
critical for reconstructing the paleoenvironmental context of early
complex life. In this study, multiple geochemical proxies (S/TOC, B/Ga,
Sr/Ba, and Y/Ho ratios) are applied to evaluate paleosalinity across
key stratigraphic units of the basin. Results indicate pronounced
values variability. The FB2b interval is characterized by consistently
low salinity, reflecting a freshwater-influenced depositional setting,
while the FC and FB1c units show more marine-like signatures, in
agreement with evidence of hydrothermal inputs. The FB2a and FB1b
subunits display intermediate values, pointing to fluctuating
freshwater--brackish conditions. These paleosalinity trends are
consistent with previously recognized sea-level fluctuations, facies,
and oxygenation patterns in the basin. The findings highlight the
heterogeneity of hydrological conditions in the Paleoproterozoic
Francevillian Basin and identify FB2b as a unique freshwater influenced
setting that may have favored the development and preservation of early
macroscopic life.
\end{abstract}

\begin{altabstract}
Le bassin de Francevillian au Gabon (2,1 Ga) renferme l'un des plus
anciens organismes macroscopiques connus, pr\'{e}serv\'{e} au sein de
la sous-unit\'{e} fossilif\`{e}re FB2b. La compr\'{e}hension des
conditions de d\'{e}p\^{o}t de cet intervalle est cruciale pour la
reconstitution du contexte pal\'{e}oenvironnemental des premi\`{e}res
formes de vie complexes. Dans cette \'{e}tude, plusieurs indicateurs
g\'{e}ochimiques (rapports S/COT, B/Ga, Sr/Ba et Y/Ho) sont
utilis\'{e}s pour \'{e}valuer la pal\'{e}osalinit\'{e} des principales
unit\'{e}s stratigraphiques du bassin. Les r\'{e}sultats indiquent une
variabilit\'{e} marqu\'{e}e des valeurs. L'intervalle FB2b est
caract\'{e}ris\'{e} par une salinit\'{e} constamment faible,
refl\'{e}tant un milieu de d\'{e}p\^{o}t influenc\'{e} par l'eau douce,
tandis que les unit\'{e}s FC et FB1c pr\'{e}sentent des signatures plus
marines, en accord avec des apports hydrothermaux. Les sous-unit\'{e}s
FB2a et FB1b affichent des valeurs interm\'{e}diaires, t\'{e}moignant
de conditions fluctuantes entre eau douce et eau saum\^{a}tre. Ces
tendances de pal\'{e}osalinit\'{e} sont coh\'{e}rentes avec les
fluctuations du niveau marin, les faci\`{e}s et les profils
d'oxyg\'{e}nation pr\'{e}c\'{e}demment identifi\'{e}s dans le bassin.
Ces r\'{e}sultats mettent en \'{e}vidence
l'h\'{e}t\'{e}rog\'{e}n\'{e}it\'{e} des conditions hydrologiques dans
le bassin pal\'{e}oprot\'{e}rozo\"{i}que de Francevillian et
identifient FB2b comme un environnement unique influenc\'{e} par l'eau
douce qui a pu favoriser le d\'{e}veloppement et la pr\'{e}servation de
la vie macroscopique primitive.
\end{altabstract}

\maketitle
{\vspace*{.8pc}}

\twocolumngrid

\end{noXML}

\defcitealias{GauthierLafayeWeber1989}{ibid.}
\defcitealias{Thieblemontetal2014}{ibid.}

\section{Introduction}

Reconstructing paleoclimate and paleo-salinity remains a key challenge
in Earth sciences, providing crucial insights into past environmental
conditions and their role in shaping the evolution and distribution of
life \citep{Chraikietal2023, Liuetal2025, Sunetal2022, WeiAlgeo2020,
Yeetal2020}. Although paleosalinity proxies are well established in
younger systems \citep{Chenetal2025, Gilleaudeauetal2021, Sunetal2022,
Yeetal2020}, reconstructing salinity in ancient settings such as the
Proterozoic is far more challenging because original geochemical
signals may be overprinted by late diagenesis and because secular
variations in trace-element concentrations and clay-mineral assemblages
can obscure the primary salinity imprint \citep{Liuetal2025}. The
2.1~Ga Francevillian Paleoproterozoic Series of Gabon
\mbox{\citep{Ikouangaetal2024}} is particularly significant in this regard, as
it hosts some of the oldest macroscopic fossils
\citep{ElAlbanietal2010} interpreted as one of the first evidences for
early multicellular life on Earth \citep{ChiFruetal2024,
ElAlbanietal2014, ElAlbanietal2019, ElAlbanietal2023,
ElKhouryetal2025b, ElKhouryetal2025a, OssaOssaetal2023}. Understanding
the depositional environment of these successions is therefore crucial
for accurately placing the biogenic record into its proper
paleoenvironmental context. This will allow new insights into how life
emerged and evolved on Earth.

Within the Francevillian basin, the FB2 unit has attracted sustained
attention because it preserves fossil-bearing black shales.\ The lower
part, the Poubara sandstones (FB2a) \citep{Weber1968}, has been
variably interpreted as offshore bars \citep{AzzileyAzzibrouck1986}, a
shoreface system \citep{Pamboetal2006}, deltaic deposits
\citep{Duboisetal2017, GauthierLafaye1986, OssaOssa2010}, or
high-density turbidites deposited below storm-wave base
\citep{Parizeetal2013}.\ Sequence stratigraphic studies and redox
studies suggest that these sandstones record regressive conditions,
with interbedded shales deposited in the same general setting
\citep{Canfieldetal2013, Reynaudetal2018}.

The overlying FB2b black shales, which \mbox{contain} the macrofossil
assemblages, have given rise to equally diverse interpretations.\ They
have been described as recording an anoxic lagoon
\citep{AzzileyAzzibrouck1986}, a prodelta environment
\citep{GauthierLafaye1986, OssaOssa2010}, a storm-dominated shelf
\citep{Pamboetal2006}, or \mbox{deep-water} turbidites
\citep{Parizeetal2013}. More recently, the discovery of load casts,
liquefaction features and microbial mat textures has supported an
alternative view: rapid fluid--mud deposition in shallow-water
conditions (${<}$80--100~m), consistent with euphotic setting
\citep{Aubineauetal2018, Reynaudetal2018}, but all with a priori marine
environment. 

Taken together, these contrasting models highlight the complexity of
reconstructing the depositional history of the FB2 unit. While facies
analysis and sequence stratigraphy have been central to this debate,
one environmental parameter has remained largely overlooked:
paleosalinity.\ Salinity reflects basin-scale hydrological conditions
and freshwater--marine mixing dynamics. Although it exerts only limited
direct control on sedimentary processes \citep{Mulderetal2003}, it
remains a key parameter for constraining depositional environments and
water--mass connectivity. In this work, we examine paleosalinity across
several sections of the Francevillian Series, with a particular focus
on the fossiliferous FB2b interval. By applying geochemical proxies,
including S/TOC, B/Ga, Sr/Ba and Y/Ho, we aim to refine interpretations
of depositional environments and to better constrain the
paleoenvironmental conditions under which the Francevillian macrobiota
were \mbox{preserved}.

\section{Geological setting} \label{sec2}

The Francevillian basin belongs to a well-known Paleoproterozoic
lithostratigraphic succession that outcrops across 35~000~km$^2$ in
southeastern Gabon. It is only slightly deformed and show no evidence
of metamorphic transformation \citep{Bankoleetal2015,
GauthierLafayeetal1996, Ikouangaetal2023, NgombiPembaetal2014}. It
consists of several lithostratigraphic formations, designated as
Francevillian (F) FA to FE (Figure~\ref{fig1}) \citep{Weber1968}.

\begin{figure}
\includegraphics{fig01}
\caption{\label{fig1}Synthetic lithostratigraphic column of
Francevillian Series \citep[modified form][]{Aubineau2019,
NgwalghoubouIkouanga2022}.\ The fossil symbol indicates the
fossiliferous subunit.}
\end{figure}

The FA Formation primarily comprises fluvial and deltaic coarse-grained
sandstones but also contains red radioactive U--Th rich conglomerates
and some interlayered siltstones indicating a progressive marine
invasion \citep{Bankoleetal2020, Bankoleetal2015}. Its thickness ranges
from 100 to ${>}$1000~m, from the proximal to the distal part of the
basin. This Formation rests unconformably on Archean basement rocks
\citep{Weber1968}. At the top, it contains uranium enrichments
associated with bitumen and the notable Oklo natural nuclear reactors
\citep{GauthierLafayeWeber1989}.

The FB Formation is composed mainly of marine sediments deposited below
the storm wave base that are discordantly overlyfluvial to deltaic FA
Formation. Its thickness ranges from 400 to 1000~m
\citepalias{GauthierLafayeWeber1989} and contains up to 15~wt\% organic
intervals \citep{Canfieldetal2013}. This Formation is further
subdivided lithologically into FB1 and FB2 members with the FB1 Member
divided into the FB1a--c units and the FB2 Member into FB2a--b units
\citep{AzzileyAzzibrouck1986}. In this scheme, the FB1a--b units consist
of interlayered shales, sandstones, and conglomerates that transition
into predominantly shales at the top while the FB1c unit consists
primarily of shales, with a thin iron formation overlain by black
shales and a manganese-rich interval, formed by the weathering of
Mn-rich carbonate layers known as protore \citep{AzzileyAzzibrouck1986,
Reynaudetal2018, Weber1968}. The FB2a unit includes dominantly
``Poubara'' coarse-grained sandstone beds deposited during sea level
fall \citep{Reynaudetal2018}. This unit is sharply overlain by the FB2b
unit, characterized by finely laminated black shales interbedded with
thin siltstone layers deposited by waning storm surges
\citep{Weber1968}. It contains remarkable specimens that are partially
to completely pyritized and non-pyritized macrofossils found abundantly
in the black shales \citep{ElAlbanietal2010, ElAlbanietal2014,
ElAlbanietal2019}. Fossil abundance decreases towards the top of the
section, where silty and sandy beds are more common. In the
fossiliferous quarry, the FB2b black shales are 5~m thick and were
deposited in a quiet, low-energy marine environment from a fully
oxygenated water column \citep{Canfieldetal2013}. While macrofossils
are only found in the FB2b unit, microbial mat structures can be found
in all the FB2 Member \citep{Aubineauetal2018, Duboisetal2017}.\looseness=1

The FC Formation is dominated by massive dolomites, indicative of
shallow-water evaporitic depositional conditions \citep{Preatetal2011,
Weber1968}. It also contains stromatolitic cherts with interlayered
black shales at its base \citep{BertrandSarfatiPotin1994} hosting
Gunflint-type assemblage of microfossils
\citep{LekeleBaghekemaetal2017}. This Formation varies between 5 and
40~m in thickness and is perceived as a lithostratigraphic marker since
it is ubiquitous in the Franceville basin \citep{Thieblemontetal2014}.

The FD Formation corresponds to black shales deposits interbedded with
volcanic tuffs and fine- to medium-grained sandstones
\citepalias{Thieblemontetal2014}. It does not really differ from the FB
Formation, however, the occurrence of silica-rich FC rocks (i.e.,
cherts) allows distinguishing between the FB and the FD Formations.
These sediments were settled in a near-continental margin, during a
transgressive phase \citep{Canfieldetal2013, ChiFruetal2019}. They are
highly enriched in organic content (up to 10~wt\%) and very depleted in
${}^{13}$C, suggesting the incorporation of methanotroph-derived
organic matter~\citep{WeberGauthierLafaye2013}.

The FE Formation has been predominantly eroded across the basin.
However, it can be identified in some restricted geographical
extensions and consists of medium-grained arkose with interlayered
shales \citep{GauthierLafayeWeber1989, Thieblemontetal2014}. It is
considered a localized filling of depressions through weathering of the
Ogoou\'{e} orogenic belt \citep{Thieblemontetal2014}.

\section{Materials and methods} \label{sec3}

Samples were collected from the Moulende quarry, as well as several
core drills in the Francevillian Basin. Selected 218 samples (39 FB1b,
78 FB1c, 21 FB1a, 47 FB2b, 20 FC and 13 FD) were powdered in agate
mortar, and analyzed for whole-rock major, minor and trace element
compositions including CaO, B, Ga, Sr, Ba, Y and Ho at the Service
d'Analyse des Roches et des Min\'{e}raux (SARM; CRPG-CNRS, Nancy,
France). Detection limits were 0.03\% for CaO, 2~ppm for~B, 0.2~ppm for
Ga, 1.5~ppm for Sr, 8~ppm for Ba, 0.1~ppm for Y and 0.003~ppm for Ho as
reported by SARM analytical standards.\ The whole-rock powder of
${\sim}$100~mg of each sample was fused with a lithium metaborate
(LiBO$_2$) and dissolved in nitric acid for elemental analysis. Major
and trace element data were obtained using inductively coupled plasma
atomic emission spectrometry (ICP-AES) and inductively coupled plasma
mass spectrometry (ICP-MS), respectively, following the techniques
described in \citet{Carignanetal2001}. 
Total sulfur (S) and organic carbon (TOC) content were measured by
infrared absorption with carbon--sulfur EMIA 320-V2 (HORIBA) at
SARM-CRPG. Detection limits were approximately 0.02~wt\% and 0.01~wt\%
respectively.

\section{Results and discussion} \label{sec4}

Reconstructing paleosalinity in deep time systems requires a multiproxy
strategy because no single geochemical ratio provides a fully reliable
signature, especially in Proterozoic settings where post-depositional
alteration and secular changes in seawater chemistry can complicate
interpretations. Elemental proxies such as B/Ga, Sr/Ba, S/TOC and Y/Ho
have been used effectively in previous work \citep{Caoetal2023,
Lawrenceetal2006, Liuetal2025, Remirezetal2024, RemirezAlgeo2020,
WeiAlgeo2020}, but each responds to different environmental and
diagenetic processes. For example, \citet{Algeoetal2025} demonstrated
that clay--mineral assemblages can influence B uptake, potentially
biasing B/Ga ratios in some sedimentary contexts. \citet{Liuetal2025}
showed that in Proterozoic shales, Sr/Ba tends to underestimate
salinity relative to B/Ga, likely due to reduced Sr concentrations in
ancient seawater. By combining four complementary proxies in this
study, it becomes possible to cross-validate salinity estimates, reduce
ambiguity, and produce a more robust reconstruction than that with
fewer indicators.

\subsection{S/TOC proxy}

A first proxy would be the S/TOC ratios \citep{Liuetal2025,
WeiAlgeo2020}. The S/TOC ratio in the FB1b and FB1c subunits yield
values between 0.009 and 0.5 (average~0.11) and 0.007 and 0.48
(average~0.23) respectively (Figure~\ref{fig2}a; Supplementary data).
In the FB2a samples, the values are between 0.08 and 1.08
(average~0.55) while in the FB2b subunit, some TOC contents are below
detection limits, with S/TOC ratio that can go up to 2.34
(average~0.18). Values are more variable in the FC unit, ranging from
0.01 to 15.14 (average~4.09). Finally, S/TOC values range from 0.03 to
0.48 (average~0.36) in the FD unit. Overall, the dataset reveals
substantial stratigraphic variability; FB2 interval shows intermediate
values with a clear distinction between FB2a and the fossil-bearing
FB2b subunit while higher and more heterogeneous values occur in FC.
Relatively low and stable values characterize FD.

\begin{figure*}
\includegraphics{fig02}
\caption{\label{fig2}Paleosalinity proxies including (a)~S/TOC,
(b)~B/Ga, (c)~Sr/Ba and (d)~Y/Ho  \citep[according
to][]{LawrenceKamber2006,WeiAlgeo2020}. The vertical axis represents a
normalized composite~stratigraphic depth, reflecting the relative
stratigraphic position of each sample and not the exact depth.}
\end{figure*}

Applied to the Francevillian succession, these thresholds suggest a
progressive shift in salinity conditions through the sequence. At the
base, the FB1b subunit yields S/TOC values close to the freshwater
threshold (${<}$0.1 \citep{WeiAlgeo2020}), indicating limited sulfate
availability. Upsection, FB1c displays more variable but generally
higher values, consistent with brackish to marine influence. The
overlying FB2a subunit shows a further increase (average~0.55),
reflecting more sustained brackish--marine conditions. In contrast, the
fossiliferous FB2b interval is marked by a pronounced decrease in S/TOC
(average~0.18, with several samples below detection), pointing to
reduced sulfate availability and enhanced freshwater input
\citep{Chenetal2025}. Above this interval, the FC unit exhibits the
highest and most heterogeneous values (average 4.09, with several
samples ${>}$1), indicating strong marine influence and efficient sulfur
incorporation during early diagenesis. The succession culminates with
the FD unit, where moderately low and stable S/TOC values
(average~0.36) again reflect brackish to marine depositional
\mbox{conditions}.

When considered collectively, the S/TOC proxy suggests significant
heterogeneity in salinity across the Francevillian succession
(Figure~\ref{fig2}a). Importantly, the fossil-bearing FB2b interval
records amongst the lowest average ratios, implying that macro-biota
developed and were preserved under conditions of reduced salinity
relative to the other units. This provides a first-order constraint on
the depositional environment of FB2b, suggesting a setting influenced
by freshwater input, such as a shallow restricted \mbox{lagoon} or
estuarine system.

\subsection{B/Ga proxy}

The second considered proxy is the Boron/Gallium (B/Ga) ratio
\citep{Liuetal2025, Wangetal2021, WeiAlgeo2020}.\ At the base of the
succession, FB1b exhibits moderate B/Ga values ranging from 1.29 to
4.54 (average~2.83) (Figure~\ref{fig2}b; Supplementary data).
Upsection, FB1c displays markedly higher variability, with values
ranging from below detection to 47.43 (average~7.51). The overlying
FB2a subunit shows a narrower range of values (1.12--2.43;
average~1.94), followed by the fossiliferous FB2b interval, where
ratios remain similarly restricted (1.34--2.13; average~1.85). Above
these units, the FC \mbox{interval} yields values between 0.10 and 3.06
(average~1.65), while the uppermost FD unit records ratios ranging from
2.13 to 3.43 (average~2.83). Overall, the FB2 interval is characterized
by consistently low B/Ga ratios relative to other parts of the
succession, with particularly restricted values in the fossil-bearing
FB2b.

Boron uptake in sediments is enhanced under marine conditions due to
the more alkaline chemistry of seawater, whereas gallium concentrations
remain relatively low owing to strong particulate scavenging
\citep{OriansBruland1988}.\ Consequently, B/Ga is a widely used
paleosalinity proxy, with thresholds of ${<}$3 for freshwater, 3--6 for
brackish, and ${>}$6 for marine facies \citep{WeiAlgeo2020}. Applied to
the Francevillian Series, most FB2 samples plot within the freshwater
domain, with mean values between 1.1 and 2.6. The FB2b subunit
(average~1.85) thus indicates a strong freshwater influence, consistent
with its low S/TOC ratios. Similarly, FB2a (average~1.94) also lies in
the freshwater field, though with slightly higher values. The FD unit
averages 2.83, straddling the upper limit of freshwater, while the FC
unit averages 1.65, again within the freshwater domain despite
occasional higher values. By \mbox{contrast}, FB1c shows highly
variable values, including some ${>}$6 that fall within the marine range,
while FB1b mostly remains below the brackish threshold.

Taken together, the B/Ga proxy suggests that the fossiliferous FB2b
interval was deposited under very low-salinity conditions, in contrast
to marine incursions recorded in FB1c (Figure~\ref{fig2}b). This
reinforces the view that the depositional setting of FB2b was
restricted and strongly influenced by continental runoff, a scenario
compatible with shallow-water or estuarine environments.

Moreover, B shows a close geochemical association with potassium (K)
\citep{Retallack2020}. In the fossiliferous sediments of FB2b, boron is
primarily hosted in clay minerals such as illite, given the absence of
K-feldspar \citep{NgombiPembaetal2014}. The lack of tourmaline, a
boron-rich heavy mineral, further excludes significant detrital
contributions of boron that could bias salinity reconstructions
\citep{Yeetal2020}, leaving only minor inputs from detrital clays
(e.g., chlorite) and micas. If inherited boron from these phases were
accounted for, the corrected boron content (B$^*$), as proposed by
\citet{Yeetal2016}, defined as the difference between measured and
inherited boron, would yield even lower effective values. Boron can
also be expelled from clay structures during advanced diagenesis
(200--300~{\textdegree}C) or metamorphism \citep{Retallack2020}.
However, the FB2b sediments are exceptionally well preserved and show
minimal signs of alteration or thermal overprint
\citep{Bankoleetal2015, ElKhouryetal2025b, ElKhouryetal2025a,
GauthierLafaye1986, Ikouangaetal2023, NgombiPembaetal2014}.

\subsection{Sr/Ba proxy}

Another proxy for salinity is the Sr/Ba ratio \citep{Liuetal2025,
Remirezetal2024, WeiAlgeo2020}. This proxy requires careful
consideration of carbonate content because Sr concentrations can be
artificially elevated in carbonate-rich sediments. In the studied
dataset, no significant correlation is observed between Sr, Ba, and
CaO, (Figure~\ref{fig3}). In the FB1b subunit, Sr/Ba values range from
0.06 to 24.62 (average~1.80) (Figure~\ref{fig2}c; Supplementary data),
while FB1c exhibits similarly broad variability, with ratios between
0.04 and 10.25 (average~0.87). Upsection, the FB2a interval shows
markedly lower and more restricted values, ranging from 0.05 to 0.34
(average~0.10), and the fossiliferous FB2b subunit displays an even
broader range but similarly low average values (0.02--6.13; average
0.07). Above these units, the FC interval yields values from
concentrations below detection to 2.64 (average~0.43). The FD samples
(excluding the carbonate-rich sediments) displays values from 0.03 to
0.11 (average~0.06). Overall, the FC unit shows the highest average
values within the FB--FC interval, whereas the fossiliferous FB2b and
FB2a subunits display consistently low averages.

\begin{figure*}
\includegraphics{fig03}
{\vspace*{-.15pc}}
\caption{\label{fig3}CaO versus Sr/Ba correlations for the different
units and subunits of the Francevillian Series.}
\end{figure*}

Strontium is relatively enriched in seawater, with concentrations
showing little variation with depth, whereas barium is scarce due to
its low solubility, especially when bound in carbonate, sulfate, or
phosphate minerals. For this reason, Sr/Ba is widely applied as a
paleosalinity proxy, with thresholds of ${<}$0.2 indicating freshwater,
0.2--0.5 brackish, and ${>}$0.5 marine conditions \citep{Omaretal2023,
WeiAlgeo2020}. Applied to the Francevillian succession, Sr/Ba ratios in
FB2b (average~0.07) and FB2a (average~0.10) fall within the freshwater
domain, consistent with results from the B/Ga and S/TOC proxies. By
contrast, the FC unit, with an average of~0.43, plots within the
brackish range, suggesting a more sustained marine influence. FB1c and
FB1b show highly variable values, with averages in the brackish to
marine domains but individual samples reaching very high marine-like
ratios (${>}$10), likely reflecting localized carbonate or diagenetic
effects \citep{Onanga2016}.

In sum, Sr/Ba results reinforce the view that the fossiliferous FB2b
interval was deposited under low-salinity, freshwater-influenced
conditions, in contrast to the more marine character of FC
(Figure~\ref{fig2}c). These findings align with the independent
evidence from S/TOC and B/Ga, pointing to significant paleosalinity
heterogeneity across the Francevillian basin (Figure~\ref{fig2}).

\subsection{Y/Ho proxy}

Y/Ho ratios were also evaluated as a paleosalinity proxy
\citep{Bolharetal2004, LawrenceKamber2006}. In the FB1b subunit, Y/Ho
values range from 26.49 to 40.54 (average~30.97) (Figure~\ref{fig2}d;
Supplementary data), whereas FB1c exhibits a broader range, from 28.01
to 47.60 (average~36.35). Upsection, the FB2a interval yields values
between 26.88 and 40.78 (average~32.28), followed by the fossiliferous
FB2b subunit, where values range from 28.49 to 34.45 (average~31.05).
Above these units, the FC interval spans from 10.83 to 51.84
(average~35.96), and the FD unit displays values between 30.07 and 35.9
(\mbox{average}~31.25).

The Y/Ho ratio is a robust indicator of seawater influence, as yttrium
is preferentially stabilized by complexation with anionic salts,
resulting in elevated Y/Ho values in saline waters
\citep{Bolharetal2004, Lawrenceetal2006, LawrenceKamber2006}. Average
shale compositions typically yield ratios of ${\sim}$27, whereas modern
seawater is characterized by higher, superchondritic values (${>}$44).
All average values for the Francevillian succession fall between 30
and~36, which are consistently above the Post-Archean Average Shale
(PAAS) normalized value of~27, placing them within the marine range.
The fossiliferous FB2b subunit (average~31.05) lies within the lower
end of this range, consistent with some freshwater influence that
moderated the seawater signature. In contrast, the FC unit
(average~35.96) and FB1c (average~36.35) display the highest averages,
pointing to more pronounced marine{\break} conditions.

Taken together, the Y/Ho proxy suggests that although all intervals
record a significant seawater component, relative differences exist
across the succession. The lower values of FB2b align with other
proxies (S/TOC, B/Ga, Sr/Ba), reinforcing the interpretation of reduced
salinity in the fossil-bearing \mbox{interval}, while FC and FB1c
reflect stronger marine influence (Figure~\ref{fig2}).

\begin{table*}[t!]%tab1
\caption{\label{tab1}Summary of paleosalinity proxies and inferred
depositional environments for each stratigraphic unit of the
Francevillian FB--FD succession}
\begin{tabular}{cccccl}
\thead
\parbox[t]{1.9cm}{\centering Stratigraphic level} &
\parbox[t]{1cm}{\centering  S/TOC (avg)} &
\parbox[t]{.9cm}{\centering  B/Ga (avg)} &
\parbox[t]{.6cm}{\centering  Sr/Ba (avg)} &
\parbox[t]{1.1cm}{\centering   Y/Ho (avg)}\vspace*{2pt} &
Inferred environment\\
\endthead
FB1b & 0.10 & 2.83 & 1.80 & 30.97 & \parbox[t]{7.9cm}{\raggedright
\textbf{Brackish to slightly marine}, fluctuating salinity, moderate
terrestrial influence}{\vspace*{2pt}}\\
FB1c & 0.23 & 7.51 & 0.87 & 36.35 & \parbox[t]{7.9cm}{\raggedright
\textbf{Highly variable}, episodic freshwater--marine mixing,
hydrothermal contribution}{\vspace*{2pt}}\\
FB2a & 0.55 & 1.94 & 0.10 & 32.28 &
\parbox[t]{7.9cm}{\raggedright\textbf{Brackish}, transitional
conditions, partial freshwater influence}{\vspace*{2pt}}\\
FB2b & 0.18 & 1.85 & 0.07 & 31.05 & \textbf{Freshwater-influenced},
low-salinity\\
FC & 4.09 & 1.65 & 0.43 & 35.96 & \textbf{Marine influenced},
hydrothermal contribution\\
FD & 0.36 & 2.83 & \00.06$^*$ & 31.25 & \textbf{Brackish to marine}, low
variability
\botline
\end{tabular}
\tabnote{$^*$Refers to the value excluding carbonate rich samples.}
\end{table*}

Although the Francevillian basin experienced early diagenesis and
localized late-stage fluid circulation \citep{Bankoleetal2015,
GauthierLafayeWeber2003}, it has been subject to only minimal tectonic
deformation and show no evidence of metamorphic alteration
\citep{Bankoleetal2015, GauthierLafayeetal1996, Ikouangaetal2023,
NgombiPembaetal2014}. In addition, several lines of evidence indicate
that the geochemical ratios used here retain a largely primary
environmental signal. S and total organic carbon (TOC) can be altered
during early diagenesis; however, employing their ratio (S/TOC) rather
than absolute concentrations reduces much of the associated
variability. Moreover, previous work has shown that diagenetic
reworking of organic matter in the FB2 unit is limited
\citep{Aubineauetal2021}. B/Ga, Sr/Ba, and Y/Ho are comparatively more
resistant to low-temperature diagenetic overprinting. Boron remains
bound within clay mineral lattices unless exposed to temperatures
${>}$200--300~{\textdegree}C \citep{Retallack2020}, far exceeding
conditions documented in the Francevillian Basin. Sr/Ba ratios display
limited mobility under early diagenesis once carbonate influence is
accounted for, and rare earth element ratios such as Y/Ho are among the
most stable geochemical indicators and generally preserve their primary
marine-continental \mbox{signatures} \citep{Lawrenceetal2006}. Taken
together, these considerations confirm that diagenesis may contribute
minorly to the mobility of the used elements but does not obscure the
broader paleosalinity trends identified in this study.

\subsection{Implications}

The paleosalinity patterns identified in the Francevillian succession
show strong coherence with previously established paleoenvironmental
frameworks. Low-salinity conditions in the FB2b fossiliferous interval
are consistent with reconstructions of sea-level fluctuations and
oxygenation events in the basin \citep{Canfieldetal2013}. The
variability in salinity across the succession also agrees with facies
models that distinguish between shallow- and deeper-water depositional
environments \citep{Reynaudetal2018}. In this context, the low-salinity
signature of FB2b is compatible with rapid accumulation in shallow
settings (${<}$80~m), and the presence of bacterial mats
\citep{Aubineauetal2018}, whereas more marine-like values in the FC and
FB1c units reflect open-water deposition. The marine incursions
indicated by Sr/Ba and Y/Ho ratios in FB1c are further consistent with
the hydrothermal influence described for this interval
\citep{ChiFruetal2024}. A similar correspondence is observed in the FC
unit, where relatively marine geochemical signatures match independent
evidence of hydrothermal inputs \citep{ElKhouryetal2025b}. These
associations suggest that both hydrothermal \mbox{activity} and episodic
marine inflows played an important role in shaping basin hydrology\break
(Table~\ref{tab1}).

Overall, the integration of paleosalinity proxies with existing
stratigraphic and geochemical data indicates that the Francevillian
Basin was marked by pronounced spatial and temporal heterogeneity in
water chemistry. Within this framework, the FB2b subunit stands out as
a freshwater-influenced environment, providing critical
paleoenvironmental conditions for not only the development but also the
preservation of some of the earliest known macroscopic life forms. In
fact, freshwater environment commonly enhances the delivery of
fine-grained sediments and promotes rapid burial of organic remains, a
key factor for soft-tissue preservation \citep{KidwellHolland2002}.
Moreover, a low salinity environment can decrease the degradation rate
of proteins \citep{Corthesyetal2025} and suppress bioturbators,
minimizing sediment mixing and allowing delicate biological fabrics to
persist \citep{Przeslawskietal2009, ZipperleReise2005}.

\section{Conclusion} \label{sec5}

Geochemical proxies including S/TOC, B/Ga, Sr/Ba, and Y/Ho provide new
constraints on the paleosalinity of the Francevillian Basin. The
results reveal systematic differences among the studied units, with the
fossiliferous FB2b interval characterized by consistently low salinity,
in contrast to more marine-influenced signatures in the FC and FB1c
subunits. These variations demonstrate that the basin experienced
alternating phases of freshwater dominance and marine incursions,
shaped in part by sea-level fluctuations and localized hydrothermal
activity. \looseness=1

The freshwater influence on FB2b is particularly significant, as it
defines the paleoenvironmental context in which some of the earliest
macroscopic organisms evolved and were preserved. The paleosalinity
framework established here thus refines the understanding of
depositional environments in the Paleoproterozoic and provides a
critical foundation for interpreting the ecological and taphonomic
conditions associated with early complex life. Although these patterns
are fundamentally local, they contribute to a broader understanding of
environmental heterogeneity in the early Paleoproterozoic ocean, a
period following the Great Oxidation Event, when surface environments
were becoming increasingly \mbox{diversified}.

\section*{Acknowledgements}

We thank La R\'{e}gion Nouvelle Aquitaine (Project Darwin, grant
E458CR25), the French government program ``Investissements d'Avenir''
(EUR INTREE, reference ANR-18-EURE-0010) and the Synchrotron SOLEIL,
ANR-BIOGEN, grant, 22-CE49-0010-02) for their financial support. We
also thank the CENAREST Gabonais, the COMILOG and SOCOBA Companies, the
French Embassy in Libreville and the Institut Fran\c{c}ais du Gabon,
for their support. We appreciate the technical and logistical help of
C.~Lebailly, A.~\mbox{Ngomanda}, C.~Laforest and R.~Oslisly.

\CDRGrant[NA]{E458CR25}
\CDRGrant[ANR]{ANR-18-EURE-0010}
\CDRGrant[ANR]{22-CE49-0010-02}

\section*{Declaration of interests}

The authors do not work for, advise, own shares in, or receive funds
from any organization that could benefit from this article, and have
declared no affiliations other than their research organizations.

\section*{Supplementary materials}

Supporting information for this article is available on the journal's
website under \printDOI\ or from the author.

\CDRsupplementaryTwotypes[application/zip]{supplementary-material}{\cdrattach{crgeos-319-suppl.xlsx}}

\back{}

\printbibliography
\refinput{crgeos20250738-reference.tex}

\end{document}