\makeatletter
\@ifundefined{HCode}
{\documentclass[screen,CRGEOS,Unicode,biblatex,published]{cedram}
\addbibresource{crgeos20250980.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}\hline}
\def\tabnote#1{\vskip4pt\parbox{.96\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}}
\let\xsection\relax
\def\ubreak{\break}
\skip\footins13pt
\makeatletter
\g@addto@macro{\UrlBreaks}{\UrlOrds}
\gappto{\UrlBreaks}{\UrlOrds}
\RequirePackage{etoolbox}
\usepackage[figuresright]{rotating}
\def\inlinefig#1{\includegraphics{#1}}
\def\jobid{crgeos20250980}
%\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{}
\csdef{Seqnsplit}{\\}
\def\botline{\\\hline}
\def\back#1{}
\newcommand*{\hyperlinkcite}[1]{\hyper@link{cite}{cite.#1}}
\DOI{10.5802/crgeos.336}
\datereceived{2025-12-06}
\daterevised{2026-03-09}
\dateaccepted{2026-04-07}
\ItHasTeXPublished
\def\vspone{\vspace*{18pc}}
%%
\gdef\tabonesplittabular{\hline
\multicolumn{13}{r@{}}{(continued on next page)}
\end{tabular}
\end{sidewaystable*}
\setcounter{table}{0}
\begin{sidewaystable*}
\vspone
\caption{(continued)}
\tabcolsep=5.25pt
\begin{tabular}{lcccccccccccc}
\hline
Rock type & \multicolumn{2}{c}{Basanite} & \multicolumn{2}{c}{Basalt} &
Hawaiite & \multicolumn{6}{c}{Trachyte} & Rhyolite\\
\hline}
%%
\gdef\tabtwosplittabular{\hline
\multicolumn{13}{r@{}}{(continued on next page)}
\end{tabular}
\end{sidewaystable*}
\setcounter{table}{0}
\begin{sidewaystable*}
\vspone
\caption{(continued)}
\tabcolsep=1.2pt
\begin{tabular}{lcccccccccccc}
\hline
Rock type\hspace*{3pc} & \multicolumn{2}{c}{\hspace*{2pc}Basanite\hspace*{2.5pc}} & \multicolumn{2}{c}{Basalt} &
Hawaiite & \multicolumn{6}{c}{Trachyte} & Rhyolite\\
\hline}
}
{
\PassOptionsToPackage{authoryear}{natbib}
\documentclass[crgeos]{article}
\usepackage[T1]{fontenc}
\def\CDRdoi{10.5802/crgeos.336}
\def\citelink#1#2{\citeyear{#1}}
\def\xcitealp#1#2{\citealp{#1}}
\def\href#1#2{\url[#1]{#2}}
\newenvironment{sidewaystable*}{\begin{table*}}{\end{table*}}
\let\tabonesplittabular\relax
\let\tabtwosplittabular\relax
\let\vspone\relax
\let\ubreak\relax
\newcommand\@coi{}
\newcommand\COI[1]{\gdef\@coi{#1}}
\newcommand\printCOI{\ifx\@coi\@empty\else%
\section*{Declaration of interests}
\@coi\fi}
\makeatletter
\def\CDRsupplementaryTwotypes#1#2{}
\def\hyperlinkcite#1#2{\citeyear{#1}}
}
\makeatother

\usepackage{upgreek}

\COI{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.}

\dateposted{2026-06-01}
\begin{document}

%\dateposted{2026-02-16}

\begin{noXML}

\CDRsetmeta{articletype}{research-article}

\TopicFR{P\'etrologie, p\'etrophysique}
\TopicEN{Petrology, petrophysics}

\title{Origin and petrogenesis of volcanic rocks from Ziver in Northern
Cameroon (Cameroon--Chad Volcanic Line, Central Africa): insights from
mineralogical, geochemical and isotopic data}

\alttitle{Origine et p\'{e}trogen\`{e}se des roches volcaniques de
Ziver dans le Nord-Cameroun (Ligne Volcanique Cameroun--Tchad, Afrique
centrale) : analyses tir\'{e}es des donn\'{e}es min\'{e}ralogiques,
g\'{e}ochimiques et isotopiques}

\author{\firstname{Jacques} \lastname{Dili-Rake}}
\address{Institute of Geological and Mining Research, PO Box 4110,
Yaound\'{e}, Cameroon}
\address{Department of Earth Sciences, Faculty of Sciences, University
of Maroua, PO Box 814 Maroua, Cameroon}
%\email{dilirakejacques@yahoo.fr}

\author{\firstname{Daouda} \lastname{Dawa{\"i}}\CDRorcid{0000-0003-0354-1876}}
\addressSameAs{2}{Department of Earth Sciences, Faculty of Sciences,
University of Maroua, PO Box 814 Maroua, Cameroon}
%\email{daoudadawai@gmail.com}

\author{\firstname{Beno\^{i}t Joseph} \lastname{Mbassa}\CDRorcid{0000-0003-4161-1177}\IsCorresp}
\addressSameAs{1}{Institute of Geological and Mining Research, PO Box
4110, Yaound\'{e}, Cameroon}
\email[B. J. Mbassa]{benjo\_mbassa@yahoo.fr}

\author{\firstname{Olivier}\nobreakauthor\lastname{Vanderhaeghe}\CDRorcid{0000-0001-9361-4631}}
\address{GET-OMP, Universit\'{e} de Toulouse, UPS, CNRS, IRD, CNES, 14
avenue E. Belin, 31400 Toulouse, France}
%\email{olivier.vanderhaeghe@get.omp.eu}

\author{\firstname{Michel} \lastname{Gr\'egoire}\CDRorcid{0000-0001-9196-876X}}
\addressSameAs{3}{GET-OMP, Universit\'{e} de Toulouse, UPS, CNRS, IRD,
CNES, 14 avenue E. Belin, 31400 Toulouse, France}
%\email{michel.gregoire@get.omp.eu}

\author{\firstname{Mathieu} \lastname{Benoit}\CDRorcid{0000-0002-0134-4863}}
\addressSameAs{3}{GET-OMP, Universit\'{e} de Toulouse, UPS, CNRS, IRD,
CNES, 14 avenue E. Belin, 31400 Toulouse, France}
%\email{mathieu.benoit@get.omp.eu}

\author{\firstname{Rachid} \lastname{Zayane}}
\address{Geology Department, Cadi Ayyad University, BP 2390, 40000
Marrakech, Morocco}
%\email{zayane@gmail.com}

\shortrunauthors

\keywords{\kwd{Cameroon--Chad Volcanic Line}\kwd{Ziver volcanic
rocks}\kwd{Bimodal alkaline suite}\kwd{Mantle origin}\kwd{Fractional
crystallization}}

\altkeywords{\kwd{Ligne volcanique Cameroun--Tchad}\kwd{Roches
volcaniques de Ziver}\kwd{Suite alcaline bimodale}\kwd{Origine
mantellique}\kwd{Cristallisation fractionn\'{e}e}}

\thanks{CNRS--IRD--LithoCOAC project}

\begin{abstract}
The Ziver volcanism, located in northern Cameroon within the Central
African Rift, forms an integral part of the Cameroon--Chad Volcanic
Line (CCVL). The new mineralogical, geochemical, and isotopic data
presented here provide fresh insights into the sources and petrogenesis
of lavas from this little-studied area. The primary mineral assemblage
consists of olivine, clinopyroxene, Fe--Ti oxides, and feldspars.
Clinopyroxenes are predominantly calcic, with compositions ranging from
diopside to clinoenstatite. Feldspars occur as andesine in mafic rocks,
whereas in felsic lavas they are represented by K-albite and
Na-sanidine. The volcanic suite defines a bimodal alkaline series
composed of mafic (basanite, basalt, hawaiite) and felsic (trachyte,
rhyolite) lavas, characterized by moderate to high alkali contents
(K\tsub{2}O + Na\tsub{2}O = 4.10--12.25 wt\%). These lavas display
moderately enriched radiogenic isotope signatures, with
\tsup{87}Sr/\tsup{86}Sr ratios of 0.70311--0.71856 and
\tsup{143}Nd/\tsup{144}Nd ratios of 0.51276--0.51295. Geochemical and
isotopic data (0.7036 < (\tsup{87}Sr/\tsup{86}Sr)\textsubscript{initial}
< 0.7203; --10.65 < $\varepsilon$Ndi < 6.44) indicate an intraplate
OIB affinity, derived from low-degree (1--3\%) partial melting of an
enriched garnet lherzolite mantle plume source. Magmatic
differentiation is dominated by fractional crystallization with minimal
crustal contamination, consistent with the CCVL as a whole.
\vspace*{-4pt}
\end{abstract}

\begin{altabstract}
Le volcanisme de la localit\'{e} de Ziver, situ\'{e}e au Nord-Cameroun
dans le rift centre-africain, fait partie int\'{e}grante de la Ligne
Volcanique Cameroun--Tchad (LVCT). Les nouvelles donn\'{e}es
min\'{e}ralogiques, g\'{e}ochimiques et isotopiques ici
pr\'{e}sent\'{e}es, permettent de mieux comprendre l'origine et la
p\'{e}trogen\`{e}se des laves de cette zone encore peu \'{e}tudi\'{e}e.
L'assemblage min\'{e}ral primaire est compos\'{e} d'olivine, de
clinopyrox\`{e}ne, d'oxydes ferro-titan\'{e}s et de feldspaths. Les
clinopyrox\`{e}nes sont majoritairement calciques, leur composition
variant du diopside \`{a} la clinoenstatite. Les feldspaths ont des
compositions d'and\'{e}sine dans les laves mafiques, K-albite et
Na-sanidine dans les laves felsiques. La suite volcanique d\'{e}finit
une s\'{e}rie alcaline bimodale constitu\'{e}e de laves mafiques
(basanite + basalte + hawaiite) et felsiques (trachyte + rhyolite) caract\'{e}ris\'{e}es par des teneurs en alcalins mod\'{e}r\'{e}es \`{a} \'{e}lev\'{e}es (K\tsub{2}O + Na\tsub{2}O = 4,10--12,25 \%). Ces laves pr\'{e}sentent des signatures d'isotopes radiog\'{e}niques mod\'{e}r\'{e}ment enrichies, avec des rapports \tsup{86}Sr/\tsup{86}Sr de 0,70311 \`{a} 0,71856 et \tsup{143}Nd/\tsup{144}Nd de 0,51276 \`{a} 0,51295. Les donn\'{e}es g\'{e}ochimiques et isotopiques (0,7036 < (\tsup{87}Sr/\tsup{86}Sr)\tsub{initial} < 0,7203; --10,65 < $\varepsilon$Ndi < 6,44) indiquent une affinit\'{e} d'OIB intraplaque, issue de la fusion partielle de faible degr\'{e} (1 \`{a} 3 \%) d'un panache mantellique de lherzolite \`{a} grenat. La diff\'{e}renciation magmatique est domin\'{e}e par la cristallisation fractionn\'{e}e avec une faible contamination crustale, en coh\'{e}rence avec l'ensemble de la Ligne Volcanique Cameroun.
\end{altabstract}

%\input{CR-pagedemetas}

\maketitle

\vspace*{-4pt}

\twocolumngrid

\end{noXML}

\section{Introduction}\label{sec1}

\vspace*{-2pt}

Alkaline magmatism is a key indicator of various tectonic settings,
including oceanic hotspots and seamounts \citep{Kogarko1998,
Hartetal1992}, continental rifts, and other intraplate environments
\citep{Corfuetal1991}.\ Its genesis occurs at different depths within
the mantle, ranging from the asthenosphere to the core-mantle boundary.\
This type of magmatism significantly contributes to continental crust
evolution and reflects intricate interactions between Earth's
geospheres. The generation of alkaline magmas is commonly attributed to
several mantle processes. One widely invoked mechanism is the
low-degree partial melting of peridotites in the presence of
CO\tsub{2}, which lowers the solidus and favors the production of
silica-undersaturated melts. Alternatively, alkaline magmas may derived
from the melting of recycled oceanic crust as proposed by
\citet{Hofmannetal1986}, where subducted basaltic-gabbroic lithologies
are returned to the mantle and later re-melted. A third possibility
involves the melting of metasomatized lithospheric mantle enriched by
fluids or melts from earlier subduction or plume-related events,
producing alkaline signatures even at relatively shallow depths
\citep{Dasguptaetal2007}.

The Cameroon--Chad Volcanic Line (CCVL) is a prominent intraplate
volcano-tectonic megastructure located in the heart of Central Africa.
The lavas along the CCVL are predominantly alkaline
\citep{Marzolietal2000, Kamgangetal2013, Poucletetal2014} despite the
occurrences of a few transitional lavas in the West Cameroon Highlands
\citep{Fossoetal2005, Kuepouoetal2006, ZiemaBidiasetal2018,
Lemdjouetal2020}, and rare of tholeiitic rocks in Kapsiki 
\citep{Ngounounoetal2001}.
In general, the evolution of felsic lavas is driven by the fractional
crystallization of mafic rocks coupled with crustal contamination
\citep{Kamgangetal2013, TchuimegnieNgongangetal2015}.
\looseness=1

The northern portion of the CCVL, which includes our study area,
remains relatively under-investigated (Figure~\ref{fig1}b). However,
some geochemical studies have been conducted in some neighboring
localities such as the Kapsiki Mountains \citep{NgounounoDeruelle1997,
Tamenetal2015}, Mandara Mountains \citep{NgounounoDeruelle1997},
Biu Plateau \citep{Rankenburgetal2005}, Gawar and
Zama\"{i} \citep{GountieDedzoetal2019}, Mokolo Hossehone
\citep{Tchouhlaetal2022}, and Iriba in the Oudda\"{i} massif Chad
\citep{Djerossemetal2024}.

\begin{figure*}
\includegraphics{fig01}
{\vspace*{-.25pc}}
\caption{\label{fig1}The Cameroon Chad Volcanic Line (CCVL) and its
extension. (a)~Map illustrating the location of CCVL and its northern
extension into Chad, highlighted in pink. The map shows key geological
features of the African continent. (b)~Detailed map showing the
distribution of the major volcanic centers and alkaline complexes along
the CCVL. The Chad branch is defined according to
\citet{Djerossemetal2024}. The red stars indicates the studied area.
(c)~Simplified geological sketch map of the Ziver area, showing the
locations of collected samples (red stars).}
{\vspace*{-.3pc}}
\end{figure*}

The volcanic lavas of the Ziver area exhibit a wide range of
lithologies, from basanites to rhyolites partially covering the
Pan-African granitic basement (Figure~\ref{fig1}c). A comprehensive
study of these rocks could provide valuable insights into the magmatic
source and differentiation processes of this segment of the CCVL. In
this study, we present bulk rock major and trace element compositions,
along with \mbox{mineralogical} and Sr--Nd isotopic data~of the Ziver
lavas. Our primary objective is to decipher their magmatic source,
discuss their petrogenetic and differentiation processes, and compare
them with surrounding lavas to enhance a better understanding of the
magmatic evolution in the northern CCVL.

\vspace*{-3pt}

\section{Geological setting}\label{sec2}

\vspace*{-1pt}

The CCVL is structurally localized along the Central Africa Rift System
and subdivided into a complex network of faults oriented in three major
directions: N30{\textdegree}E, N70{\textdegree}E, and
N120{\textdegree}--N130{\textdegree}E \citep{Moreauetal1987}.\ This
feature is characterized by an N30{\textdegree}E alignment of oceanic
and continental volcanic massifs, tracing a path along the Central
African Rift \citep{NjomeDeWit2014}. Extending from Pagalu Island in
the Atlantic Ocean, onto the continent \citep{GountieDedzoetal2019},
the CCVL is linked to the reactivation of large intracontinental
structures \citep{CornacchiaDars1983}. Some researchers propose its
extension to the Tibesti Massif via Lake Chad (Figure~\ref{fig1}a)
\citep{Deruelleetal2000, Djerossemetal2024}, an extension that has
recently led to the alternative designation ``Cameroon--Chad Volcanic
Line'' by \citet{Djerossemetal2024}. Magmatic activity along the CCVL
began approximately $67 \pm 2$~Ma ago \citep[][and references
therein]{Njonfangetal2011} and continues to the present day, as
evidenced by the 1999 and 2000 eruptions of Mount Cameroon
\citep{Suhetal2003}. The origin of magma remains a longstanding debate,
and several hypotheses have been proposed, including: 
(i)~\citet{Marzolietal1999}
concluded that the CCVL as a whole cannot be interpreted as the surface
expression of simple hot-spot magmatism, confirming the earlier
conclusions of \citep{FittonDunlop1985}, drawn from a more restricted
database. However, some recent geological data also show that the CCVL
lavas do not have the same mantle sources as the St Helena plume,
suggesting that the plume is not the source of the{\break} CCVL
\citep{Marzolietal2000, Rankenburgetal2005, Yokoyamaetal2007,
Lemdjouetal2020}; (ii)~the development of volcanism has been linked to
several hotspots \citep{Ngounounoetal2003, Ngakoetal2006,
Deruelleetal2007} or to small-scale tectonics and convection in the
upper mantle at the base of the lithosphere
\citep{KingRitsema2000, Reuschetal2011, DePleanetal2014,
Adamsetal2015}; (iii)~a lateral flow of asthenospheric
plumes below the continental lithosphere, which was significantly
thinned during Mesozoic rifting, and which could currently
\mbox{contribute} to young volcanism along the LVC 
\citep{EbingerSleep1998};
(iv)~\citet{Nkonoetal2014} and \citet{NoudiedieKamgangetal2020}
propose two successive senestrial geodynamic models to explain the
distribution of magmatic activity from the Cenozoic to the present. The
first, during the Palaeogene, developed around the N70{\textdegree}E
direction, while the second (Neogene) is oriented around the
N130{\textdegree}E direction. A short transition separates the two
periods. The location follows the local reactivation of pre-existing
faults (Pan-African) in relation to the collision between the
Afro--Arabic and Eurasian plates, during Alpine history.

The lithospheric mantle beneath the CCVL may have undergone both
thermochemically and mechanically erosion during the break-up of
Gondwana and the opening of the Central Atlantic Ocean (from
${\sim}$126--100~Ma) \citep{Deruelleetal2007}. Major volcanic activity
along the CCVL has been ongoing since cretaceous (68.8 ${+}$ 1.7~Ma in
the Bamoun plateau, \citep{Njonfangetal2011, Ngongeetal2014} and
continues to the Present with the eruption of Mount Cameroon
\citep{Suhetal2003, Deruelleetal2007}. The evolution of silicic lavas
through fractional crystallization of mafic terms is overall
accompanied by crustal contamination \citep{Kamgangetal2013,
TchuimegnieNgongangetal2015}. The lavas are mainly basaltic at Mount
Cameroon \citep{Deruelleetal1987} and remain the dominant composition
in the other massifs except in the northern part of the CCVL
\citep{Tamen1998,Ngounounoetal2000}. Mafic lavas are generally present
in all volcanic centers of the CCVL, with the exception of Mount
Etind\'{e}, which consists of nephelinites, leucitites and
ha\"{u}ynophyres \citep{Nkoumbouetal1995}. The evolved lavas consist of
trachyte, trachyphonolite, rhyolite, and phonolite. The compositional
gaps observed in most series reflect the strongly bimodal (mafic, and
felsic) character of volcanic centres such as Mount Bambouto
\citep{Marzolietal1999,Youmenetal2005,KagouDongmoetal2010,
GountieDedzoetal2011}, Bamenda Oku \citep{Fitton1987},
and the Kapsiki plateau \citep{NgounounoDeruelle1997,Ngounounoetal2000}.
The geochemical data reviewed by
\citet{Deruelleetal1991} and \citet{Njonfangetal1992}
supplemented by numerous detailed studies, show the alkaline nature of
the lavas from CCVL, despite the presence of some transitional lavas in
the Bamoun Plateau, Mounts Bangou, Bana and Mboutou anorogenic complex.

Geophysical studies \citep{BrowneFairhead1983, PoudjomDjomanietal1995}
have revealed that the thinning of the crust associated with the uplift
of the Adamawa region, with an abnormally hot body in the upper mantle
\citep{Gassetal1978, Dorbathetal1984, Stuartetal1985}, could be
interpreted as a continuous rise of asthenospheric material from the
thinned lithosphere from the Tertiary period to the present day
\citep{BrowneFairhead1983, NoutchogweTatchumetal2006}.
The lithospheric seismic structure of the Central African shear zone
(CASZ) has been studied in detail over the last ten years
\citep{PasyanosNyblade2007, Priestleyetal2008, Fishwick2010,
GallacherBastow2012, Kochetal2012, DePleanetal2014}. These studies
indicate that the mantle beneath the CCVL is characterized by low
seismic wave velocities and that the lithosphere--asthenosphere
boundary is about 100~km deep (${\sim}$60~km).

Seismically, the central CCVL is active, particularly around Mount
Cameroon and Bioko Island. Recorded events reach magnitudes up to 5 and
the epicentral intensities of VII--VIII \citep{Atebaetal2009,
Medzaetal2023}. This sector is therefore considered at risk, with
earthquakes of low to moderate magnitude (0.5--5) but potentially high
intensity, up to VIII on the Modified Mercalli scale \citep{Thierryetal2008}.

The northern segment of the CCVL, extending from the Adamawa Plateau to
the Tibesti massif, is characterized by large volcanic massifs
dominated by alkaline to peralkaline lavas derived from a metasomatized
mantle source \citep{Vicatetal2002, GourgaudVincent2003,
Mbowouetal2012, Shellnuttetal2016, GountieDedzoetal2019,
Tchouhlaetal2022}.

This study examines volcanic formations from the Ziver area
(Figure~\ref{fig1}c), composed predominantly of mafic lavas (basanite,
basalt, and hawaiite) and felsic lavas (trachyte and rhyolite). These
lavas occur as blocky outcrops forming dome-shaped hills, as flows, or
as small cones ${\sim}$50~m high and 100--150~m in \mbox{diameter}.

\section{Analytical methods}\label{sec3}

Powders and thin sections of selected rock samples were prepared at the
\textit{Geosciences Environnement Toulouse} laboratory (GET),
CNRS-IRD-CNES-University Paul Sabatier (Toulouse~3, France), for
geochemical and mineralogical analyses.

Major elements minerals analyses were conducted at the \textit{Centre
de microcaract\'{e}risation \mbox{Raimond} Castaing}, CNRS-University
Paul Sabatier (Toulouse~3, France), using a Cameca SX Five electronic
microprobe. Samples were carbon-coated (15~nm thick layer, density
2.25~g/cm$^{3}$) before analysis. The conditions were accelerating
voltage of 15~kV and a probe current of 10 or 20~nA (depending on
mineral resistance to the electron beam).

Whole-rock major and trace element concentrations were determined at
the \textit{Service d'Analyses des Roches et des Min\'{e}raux} (SARM,
CRPG, France) using an ICP-OES for major elements and an ICP-MS for
trace elements analyses. New Rb/Sr and Sm/Nd isotopic analyses were
performed for four (04) lava samples from Ziver including two
basanites, one trachytes and one rhyolite. The Sr/Nd isotopic data were
performed at GET, using the Thermo Scientific TRITON${+}$ solid source
mass spectrometer, following \citet{Labouetal2020}
and \citet{Lietal2011, Lietal2012} procedures. Before measurement,
about 100~mg of whole rock powder was weighed in a teflon beaker and
dissolved in a mixture HF/HNO\tsub{3} 1:1. After dissolution, samples
were diluted in 1~ml, 2\% HNO\tsub{3} and Nd/Sr were extracted from the
matrix (2N HNO\tsub{3}) using a combination of Sr--Spec and Thru--spec
Eichrom resins. Mixed Sr and REE were loaded on a Re filament and were
run sequentially (first Sr then Nd) using a double Re filament
protocol. Monitoring of the interferences of \tsup{87}Rb and
\tsup{144}Sm was proceeded according to the protocol of
\citet{Lietal2012} and the quality and reproducibility of the
measurements were controlled using a sequential measurement of isotopic
standards (SRM 987 and JNdi), doped isotopic standards (NBS 987 ${+}$ Rb
and JNdi ${+}$ Sm) and laboratory--dedicated Sr ${+}$ REE artificial
\mbox{solutions}.\looseness=-1
\vspace*{-.4pc}

\section{Results}\label{sec4}
\vspace*{-.2pc}
\subsection{Nomenclature and petrography} \label{sec4.1}

The volcanic rocks studied include both mafic (basanites, basalts,
hawaiites) and felsic lavas (trachytes and rhyolites), classified
according to the scheme of \citet{LeBasetal1986} (Figure~\ref{fig2}).
Mafic volcanic rocks have a microlitic to porphyritic texture. The
phenocryst phases are principally plagioclase, olivine, pyroxene and
opaque minerals are the main phenocrysts. Olivine occurs in some
samples and is frequently altered to iddingsite. The groundmass of
these lavas consists of microlites of plagioclases alongside
clinopyroxene, olivine, and oxides (Figure~\ref{fig3}c--f). Felsic
volcanic rocks are light to dark--grey, locally brecciated with
fragments of the granitic basement. They have a porphyritic texture
with a groundmass composed mainly of fine crystallized alkali feldspar,
clinopyroxene and opaque minerals. Phenocrysts consist mainly of alkali
feldspars: 10--15~vol\%, clinopyroxene: 1--5~vol\% and opaque
minerals: 3--10~vol\%. Microcrystals of opaque minerals are rounded or
angular, and, for some, embedded in clinopyroxene crystals. Microliths
show, in some samples, a preferred orientation underlining a magmatic
fabric.\vspace*{-.15pc}

\begin{figure}
\includegraphics{fig02}
{\vspace*{-.4pc}}
\caption{\label{fig2}TAS classification diagram for Ziver lavas after
\citet{LeBasetal1986}. The discrimination dotted line is after
\citet{IrvineBarragar1971}. Domains of the different sectors of the
CCVL, as defined by \citet{Deruelleetal2007}: CS (Continental sector);
OS (Oceanic Sector), ET (Mount Etinde).}
{\vspace*{-.5pc}}
\end{figure}

\begin{figure*}
\includegraphics{fig03}
{\vspace*{-.1pc}}
\caption{\label{fig3}Field photographs and representative
photomicrographs for the magmatic rocks from Ziver. (a)~Basalt outcrop
in the Ziver area; (b)~vacuolar structure observed in basalt outcrops;
(c)~microlitic porphyritic texture; (d)~olivine and clinopyroxene
crystals exhibiting destabilization at their edges; (e)~presence of an
opaque mineral phase; (f)~microlitic aphyric texture in rhyolite, with
plagioclase microlites displaying a preferential orientation. Mineral
abbreviations follow \citet{WhitneyEvans2010}.}
{\vspace*{-.3pc}}
\end{figure*}

\subsection{Minerals major elements composition} \label{sec4.2}
\vspace*{-.15pc}
\subsubsection{Olivine} \label{sec4.2.1}

\looseness=-1
The olivine crystals analyzed from the Ziver lavas are consistently
magnesian, with compositions ranging from Fo$_{70}$ to Fo$_{92}$.\ The
corresponding Mg\# values (100 ${\times}$ Mg/(Mg ${+}$ Fe$^{{2+}}$)) range
from 69 to 92.\ Electron microprobe analyses reveal a broad range of
chemical compositions, with minor element concentrations, which include
0.07--2.19~wt\%~CaO, \mbox{0.05--50.25}~wt\% MgO, and 7.9--27.19~wt\% FeO.
Trace element compositions show 0.03--0.47~wt\% NiO, 0.04--0.54~wt\%
MnO, and Cr\tsub{2}O\tsub{3} concentrations ${\leq}$0.22~wt\%. According
to the classification scheme of \citet{Dick1989}, as presented in 
Figure~\ref{fig4}a, the composition of the analyzed olivines in the
basanite lavas ranges from forsterite to hyalosiderite.

\begin{figure*}
{\vspace*{-.2pc}}
\includegraphics{fig04}
{\vspace*{-.2pc}}
\caption{\label{fig4}Mineralogical compositions of the studied lavas:
(a)~composition of olivines in the classification diagram of
\citet{Dick1989}; (b)~composition of clinopyroxenes in the Wo--En--Fs
ternary diagram of \citet{Morimotoetal1988}; (c)~position of
clinopyroxenes in the Al$^{\mathrm{VI}}$ versus Al$^{\mathrm{IV}}$
variation diagram of \citet{CaldeiraMunha2002}. RMC: Refractory mantle
clinopyroxene fields from \citet{Jagoutzetal1979}. HP ${=}$ High
pressure; LP ${=}$ Low pressure; (d)~composition of analyzed feldspars in
the An--Ab--Or diagram of \citet{SmithBrown1988}; (e)~projection of the
analyzed feldspars iin Al$^{\mathrm{IV}}$ versus Si diagram;
(f)~projection of the Fe--Ti oxides in the classification diagram of
\citet{HaggertyTompkins1984}.}
{\vspace*{-.2pc}}
\end{figure*}

\subsubsection{Clinopyroxene} \label{sec4.2.2}

Clinopyroxene crystals from Ziver lavas are characterized by
TiO\tsub{2} ${\leq}$ 3.41~wt\%, Al\tsub{2}O\tsub{3} ${\leq}$ 8.13~wt\%,
Na\tsub{2}O ${\leq}$ 0.45~wt\%, FeO ${=}$ 5.04--8.58~wt\%, MgO ${=}$
12.26--49.73~wt\%, CaO ${=}$ 0.02--23.91~wt\%, Cr\tsub{2}O\tsub{3} ${\leq}$
0.92~wt\%, and MnO ${=}$ 0.04--0.14~wt\%.\ The
Al$^{\mathrm{VI}}$/Al$^{\mathrm{IV}}$ ratio increases with
differentiation, \mbox{ranging} from 0.11--0.13 in basanite, 0.24--1.0 in
basalt, and up to 2.18 in hawaiite. According to the classification of
\citet{Morimotoetal1988} (Figure~\ref{fig4}b), the analyzed pyroxenes
correspond to clinoenstatite (Wo\tsub{{0.02\ndash
1.13}}En\tsub{{89.97\ndash 92.10}}{\ubreak}Fs\tsub{{0.00\ndash
0.38}}Ac\tsub{{0\ndash 0.3}}) in basalts and hawaiites, and to diopside
(Wo\tsub{{48.90\ndash 50.02}}En\tsub{{36.78\ndash
38.61}}Fs\tsub{{9.76\ndash 11.50}}Ac\tsub{{1.4\ndash 1.7}}) in
basanites. In the Al$^{\mathrm{VI}}$ versus AlI$^{\mathrm{V}}$
discriminant diagram of \citet{CaldeiraMunha2002}, diopside plots in
the low-pressure domain (Al$^{\mathrm{VI}}$/Al$^{\mathrm{IV}} < 1$),
whereas clinoenstatite falls within the high-pressure RMC field defined
by \citet{Jagoutzetal1979} for refractory mantle pyroxenes
(Figure~\ref{fig4}c).

\subsubsection{Feldspars} \label{sec4.2.3}

Representative feldspars from Ziver lavas are
chemically characterized by CaO ${\leq}$
9.24~wt\%,{\break} Al\tsub{2}O\tsub{3} ${=}$ 18.12--26.76~wt\%, FeO ${=}$
0.05--0.25~wt\%, Na\tsub{2}O ${=}$ 1.14--7.98~wt\%, and K\tsub{2}O ${=}$
0.25--15.01~wt\%. In the An--Ab--Or classification diagram of
\citet{SmithBrown1988} (Figure~\ref{fig4}d), feldspars plot as andesine
in mafic lavas, and as potassic albite, sanidine, and anorthose in
felsic lavas. All feldspars show Si ${+}$ Al$^{\mathrm{IV}} \approx 4$
(Figure~\ref{fig4}e). Structural formula calculations yield Si ${=}$
2.98--2.99 and Al ${=}$ 0.99--1.02 a.p.f.u.\ in mafic lavas, and Si ${=}$
2.88--3.01 and Al ${=}$ 0.97--1.10 a.p.f.u.\ in felsic
lavas.

\subsubsection{Opaque minerals} \label{sec4.2.4}

Opaque minerals in the Ziver lavas are represented by ulv\"{o}spinel in
the mafic rocks and magnetite in the felsic rocks (Figure~\ref{fig4}f),
following the classification of \citet{HaggertyTompkins1984}. In felsic
lavas, FeO contents (49.89--88.60~wt\%) are consistently higher than
TiO\tsub{2} (2.89--50.08~wt\%), whereas in mafic lavas TiO\tsub{2}
(36.88--41.70~wt\%) exceeds FeO (34.64--36.61~wt\%). The Cr\#
[100${\cdot}$Cr/(Cr ${+}$ Al)] varies from 0 to 26.46 in the mafic lavas
and from 0 to 7.73 in the felsic lavas.

\subsection{Whole-rock geochemistry} \label{sec4.3}

The geochemical study is based on bulk-rock major and trace elements
analysis of twelve (12) representative lava samples and four (04) Rb/Sr
and Sm/Nd isotopic ratios (Table~\ref{tab1}). The laboratory bulk data
were recalculated on an anhydrous basis because some samples displayed
loss on ignition greater than the critical value of 2~wt\%. The
differentiation index [DI ${=}$ quartz ${+}$ albite ${+}$ orthoclase or
(Nepheline ${+}$ Leucite ${+}$ Albite ${+}$ orthoclase)] of these samples
ranges from 24.45 (basanite Zi10) to 90.13 (rhyolite M29). A
compositional gap is noticed between 47.04 and 64.65~wt\% SiO\tsub{2}
contents.

\begin{sidewaystable*}[p!]%tab1
\vspone
\caption{\label{tab1}Whole rock chemical composition of lavas from Ziver}
\tabcolsep=8.75pt
\begin{tabular}{lcccccccccccc}
\thead
Rock type & \multicolumn{2}{c}{Basanite} & \multicolumn{2}{c}{Basalt} &
Hawaiite & \multicolumn{6}{c}{Trachyte} & Rhyolite\\
\endthead
Sample ID & M33 & Zi10 & Zi16 & M20 & Zi17 & Zi14 & Zi7 & Zi12 & Zi5 & Zi11 & Zi8 & M29\\
SiO\tsub{2} & 42.10 & 43.55 & 45.63 & 45.78 & 46.01 & 63.16 & 63.52 & 63.58 & 63.66 & 63.92 & 64.28 & 70.10\\
Al\tsub{2}O\tsub{3} & 13.38 & 12.07 & 12.95 & 13.90 & 15.19 & 15.63 & 15.72 & 15.53 & 15.78 & 16.14 & 15.76 & 13.79\\
Fe\tsub{2}O\tsub{3} & 13.90 & 12.29 & 11.13 & 11.20 & 11.07 & \05.24 & \05.23 & \05.57 & \05.41 & \05.54 & \05.19 & \05.53\\
MnO & \00.18 & \00.19 & \00.15 & \00.16 & \00.15 & \00.17 & \00.19 & \00.17 & \00.19 & \00.10 & \00.15 & \00.22\\
MgO & \09.17 & 13.90 & 13.10 & 10.55 & \08.04 & \00.06 & \00.08 & \00.08 & \00.04 & \00.13 & \00.09 & \00.00\\
CaO & 10.46 & 10.19 & \08.34 & \08.94 & \08.70 & \01.16 & \01.12 & \01.22 & \01.30 & \00.14 & \00.98 & \00.04\\
Na\tsub{2}O & \03.56 & \03.05 & \02.62 & \02.94 & \03.34 & \06.52 & \06.57 & \06.09 & \06.71 & \05.46 & \07.10 & \05.57\\
K\tsub{2}O & \01.58 & \01.40 & \01.42 & \01.45 & \01.71 & \05.18 & \05.16 & \05.12 & \05.17 & \05.41 & \04.99 & \04.22\\
TiO\tsub{2} & \04.11 & \02.38 & \02.52 & \02.72 & \02.85 & \00.20 & \00.20 & \00.20 & \00.20 & \00.20 & \00.21 & \00.26\\
P\tsub{2}O\tsub{5} & \01.32 & \00.59 & \00.57 & \00.63 & \00.74 & \00.00 & \00.00 & \00.00 & \00.00 & \00.00 & 0.00 & \00.00\\
Sum & 100.08\0 & 100.22\0 & 99.01 & 99.57 & 98.98 & 98.93 & 99.41 & 99.37 & 99.86 & 98.90 & 99.78 & 100.71\0\\
Mg\# & 59.74 & 71.77 & 72.57 & 67.91 & 62.01 & \02.63 & \03.41 & \03.20 & \01.52 & \04.94 & \03.79 & \00.00\\
ID & 29.42 & 24.45 & 28.90 & 30.95 & 35.26 & 84.15 & 84.67 & 85.38 & 84.51 & 85.76 & 85.25 & 90.13{\vspace*{5pt}}\\

\multicolumn{13}{l}{\textit{CIPW norm}}\\
Quartz & \00.00 & \00.00 & \00.00 & \00.00 & \00.00 & \02.05 & \02.09 & \03.85 & \01.62 & \07.90 & \02.52 & 17.07\\
Orthoclase & \08.40 & \09.46 & \08.61 & \08.79 & 10.41 & 31.62 & 31.32 & 31.14 & 31.20 & 33.10 & 30.02 & 24.67\\
Albite & \04.81 & \08.37 & 19.05 & 20.06 & 22.25 & 53.22 & 53.59 & 52.97 & 53.44 & 47.83 & 54.19 & 48.95\\
Anorthite & 15.33 & 16.10 & 19.89 & 21.05 & 22.11 & \00.00 & \00.00 & \00.00 & \00.00 & \00.73 & \00.00 & \00.15\\
Nepheline & 11.60 & 12.03 & \02.00 & \02.96 & \03.76 & \00.00 & \00.00 & \00.00 & \00.00 & \00.00 & \00.00 & \00.00\\
Diopside & 25.77 & 22.55 & 15.04 & 16.37 & 14.16 & \05.26 & \05.06 & \05.54 & \05.88 & \00.00 & \04.39 & \00.00\\
Hypersthene & \00.00 & \00.00 & \00.00 & \00.00 & \00.00 & \05.23 & \05.36 & \05.08 & \05.04 & \07.97 & \05.53 & \07.67\\
Olivine & 25.97 & 18.03 & 27.17 & 21.96 & 17.97 & \00.00 & \00.00 & \00.00 & \00.00 & \00.00 & \00.00 & \00.00\\
Magnetite & \02.15 & \02.43 & \01.97 & \01.99 & \01.97 & \00.00 & \00.00 & \00.94 & \00.00 & \00.99 & \00.00 & \00.97\\
Ilmenite & \04.59 & \07.92 & \04.92 & \05.32 & \05.60 & \00.39 & \00.39 & \00.40 & \00.38 & \00.39 & \00.40 & \00.46\\
\tabonesplittabular%
Apatite & \01.39 & \03.10 & \01.36 & \01.50 & \01.77 & \00.00 & \00.00 & \00.00 & \00.00 & \00.00 & \00.00 & \00.00\\
Sum & 100.00\0 & 100.00\0 & 100.00\0 & 100.00\0 & 100.00\0 & 100.00\0 & 100.00\0 & 100.00\0 & 100.00\0 & 100.00\0 & 100.00\0 & 100.00\0{\vspace*{5pt}}\\

\multicolumn{13}{l}{\textit{Trace elements (ppm)}}\\
Co & 51.14 & 59.04 & 54.08 & 52.38 & 42.97 & \01.30 & \00.58 & \02.49 & \00.72 & \00.71 & \01.10 & \00.34\\
Cr & 534.55\0 & 717.16\0 & 454.66\0 & 429.23\0 & 240.18\0 & 185.59\0 & 92.83 & 351.17\0 & 117.99\0 & 96.60 & 147.24\0 & 62.10\\
Cu & 61.97 & 50.66 & 46.41 & 47.87 & 31.32 & \06.98 & \03.90 & 11.86 & \03.97 & \04.94 & \06.42 & \04.20\\
Ni & 265.72\0 & 422.58\0 & 435.58\0 & 348.53\0 & 142.44\0 & 85.16 & 25.91 & 179.26\0 & 34.72 & 47.19 & 54.25 & \08.30\\
V & 241.03\0 & 201.40\0 & 176.73\0 & 191.00\0 & 185.85\0 & \01.69 & \01.45 & \02.10 & \01.45 & \00.00 & \01.71 & \05.80\\
Cs & \00.17 & \00.36 & \00.27 & \00.29 & \00.32 & \00.40 & \00.28 & \00.26 & \00.50 & \00.37 & \00.46 & \00.34\\
Rb & 19.69 & 34.63 & 22.40 & 28.49 & 28.00 & 115.46\0 & 108.27\0 & 106.25\0 & 115.63\0 & 132.08\0 & 130.39\0 & 130.00\0\\
Ba & 491.75\0 & 403.43\0 & 343.75\0 & 409.75\0 & 446.10\0 & 93.16 & 111.54\0 & 104.04\0 & 161.79\0 & 99.26 & 34.64 & 152.00\0\\
Th & \02.75 & \04.30 & \03.21 & \04.04 & \04.48 & 13.78 & 14.25 & 14.18 & 14.19 & 14.57 & 22.03 & 28.90\\
U & \00.77 & \01.11 & \00.93 & \01.15 & \01.32 & \03.51 & \02.12 & \01.66 & \03.55 & \01.74 & \07.28 & \04.36\\
Nb & 56.10 & 51.98 & 42.42 & 48.74 & 56.54 & 121.64\0 & 129.99\0 & 119.57\0 & 125.97\0 & 125.65\0 & 182.21\0 & 239.00\0\\
Ta & \03.92 & \03.66 & \02.94 & \03.51 & \04.05 & \08.76 & \08.98 & \08.81 & \08.78 & \09.36 & 13.51 & 17.90\\
K & 46.52 & 52.37 & 47.15 & 48.05 & 56.62 & 172.07\0 & 171.28\0 & 169.85\0 & 171.74\0 & 179.54\0 & 165.80\0 & 140.13\0\\
La & 44.29 & 35.91 & 31.79 & 37.73 & 43.61 & 127.39\0 & 128.59\0 & 127.53\0 & 126.93\0 & 92.10 & 156.46\0 & 149.00\0\\
Ce & 96.94 & 70.54 & 65.23 & 76.54 & 87.40 & 254.60\0 & 258.66\0 & 256.22\0 & 255.27\0 & 272.80\0 & 303.63\0 & 310.00\0\\
Pb & \01.77 & \02.38 & \01.94 & \02.40 & \02.31 & 10.04 & 10.06 & 10.11 & 10.09 & 10.33 & \09.64 & 20.80\\
Pr & 12.18 & \08.34 & \07.84 & \09.28 & 10.33 & 29.13 & 29.71 & 29.29 & 29.19 & 24.44 & 33.56 & 34.40\\
Sr & 1678.29\0\0 & 767.84\0 & 911.30\0 & 1191.63\0\0 & 1291.19\0\0 & 24.76 & 16.35 & 53.36 & 27.81 & 14.58 & 15.08 & 23.80\\
Nd & 51.45 & 33.40 & 32.17 & 37.94 & 41.12 & 109.17\0 & 112.15\0 & 108.48\0 & 109.94\0 & 96.47 & 118.30\0 & 128.00\0\\
P & 26.48 & 59.25 & 25.59 & 28.28 & 33.22 & 46.91 & 47.89 & 46.81 & 47.47 & 49.06 & 48.50 & 61.49\\
Sm & 10.19 & \06.92 & \06.73 & \07.65 & \08.23 & 20.83 & 21.26 & 20.78 & 21.07 & 21.78 & 21.53 & 27.30\\
Hf & \04.55 & \04.08 & \04.58 & \05.35 & \05.92 & 22.25 & 22.04 & 22.51 & 22.22 & 20.58 & 32.51 & 37.90\\
\tabtwosplittabular
Zr & 197.95\0 & 175.18\0 & 208.69\0 & 243.28\0 & 282.42\0 & 1123.58\0\0 & 1048.73\0\0 & 1102.96\0\0 & 1073.28\0\0 & 932.28\0 & 1667.70\0\0 & 1619.00\0\0\\
Eu & \03.55 & \02.27 & \02.21 & \02.55 & \02.67 & \03.71 & \03.80 & \03.72 & \03.78 & \03.83 & \01.42 & \04.49\\
Gd & \08.07 & \05.87 & \05.46 & \06.24 & \06.52 & 17.07 & 17.77 & 17.22 & 17.19 & 18.95 & 17.30 & 25.70\\
Tb & \01.07 & \00.83 & \00.74 & \00.85 & \00.90 & \02.56 & \02.64 & \02.60 & \02.57 & \02.81 & \02.65 & \04.10\\
Dy & \05.55 & \04.53 & \04.04 & \04.53 & \04.79 & 14.50 & 14.93 & 14.76 & 14.57 & 15.94 & 15.37 & 25.00\\
Ho & \00.94 & \00.83 & \00.71 & \00.81 & \00.85 & \02.74 & \02.84 & \02.82 & \02.78 & \03.04 & \02.98 & \05.08\\
Y & 22.97 & 20.57 & 17.90 & 20.75 & 21.24 & 69.89 & 70.38 & 70.89 & 69.79 & 75.44 & 75.20 & 130.00\0\\
Er & \02.13 & \01.98 & \01.66 & \01.89 & \01.97 & \06.98 & \07.13 & \07.21 & \07.03 & \07.59 & \07.75 & \013.70\\
Tm & \00.26 & \00.26 & \00.22 & \00.25 & \00.26 & \00.99 & \01.02 & \01.02 & \01.01 & \01.09 & \01.15 & \02.08\\
Yb & \01.50 & \01.52 & \01.27 & \01.46 & \01.50 & \06.15 & \06.33 & \06.35 & \06.16 & \06.65 & \07.33 & 12.70\\
Lu & \00.21 & \00.21 & \00.18 & \00.21 & \00.21 & \00.90 & \00.93 & \00.92 & \00.91 & \00.97 & \01.08 & \01.70{\vspace*{5pt}}\\

\multicolumn{13}{l}{Isotopes}\\
\tsup{87}Sr/\tsup{86}Sr mes & 0.70362\0 & 0.70311\0 &&&&&&& 0.71856\0 &&& 0.71798\0\\
$2\sigma \times 10^{-6}$ & 4 & 5 &&&&&&& 5 &&& 7\\
\tsup{143}Nd/\tsup{144}Nd mes & 0.512913 & 0.512949 &&&&&&& 0.512763 &&& 0.512833\\
$2\sigma \times 10^{-6}$ & 0.000003 & 0.000005 &&&&&&& 0.000006 &&& 0.000005\\
\tsup{145}Sm/\tsup{144}Nd & 0.34841\0 & 0.34841\0 &&&&&&& 0.34841\0 &&& 0.34842\0\\
$2\sigma \times 10^{-6}$ & 1 & 4 &&&&&&& 4 &&& 3\\
\tsup{87}Rb/\tsup{86}Sr & 0.03294\0 & 0.12667\0 &&&&&&& 11.6964\0\0\0 &&& 15.4293\0\0\0\\
\tsup{147}Sm/\tsup{144}Nd & 0.11562\0 & 0.120932 &&&&&&& 0.111854 &&& 0.132268\\
(\tsup{87}Sr/\tsup{86}Sr)$_{\mathrm{i}}$ & 0.703614 & 0.709718 &&&&&&& 0.720302 &&& 0.712068\\
(\tsup{143}Nd/\tsup{144}Nd)$_{\mathrm{i}}$ & 0.512893 & 0.512017 &&&&&&& 0.512816 &&& 0.512810
\botline
\end{tabular}
\end{sidewaystable*}

\subsubsection{Major elements} \label{sec4.3.1}

The major-element compositions of the Ziver lavas are characterized by
TiO\tsub{2} ranging from 0.02 to 4.11~wt\%, Al\tsub{2}O\tsub{3} from
12.70 to 16.13~wt\%, Fe\tsub{2}O\tsub{3} from 4.96 to 13.89~wt\%,
Na\tsub{2}O from 2.61 to 7.09~wt\%, and K\tsub{2}O from 1.40 to
5.18~wt\%, with Na\tsub{2}O/K\tsub{2}O ratios between 1.00 and 2.25,
and Mg\# values [(100${\cdot}$MgO/(MgO ${+}$ FeOt)] varying from 1.52 to
72.57. Mafic lavas contain normative apatite ranging from 1.39 to
3.10~wt\% and nepheline from 2.00 to 12.03~wt\%, whereas felsic lavas
are characterized by normative quartz ranging from 1.62 to 18.20~wt\%
and hypersthene from 5.04 to 7.97~wt\%.

In Harker variation diagrams (Figure~\ref{fig5}), MgO,
Fe\tsub{2}O\tsub{3}, CaO, P\tsub{2}O\tsub{5}, and TiO\tsub{2} contents
of the studied lavas, taken together with published data of northern
CCVL lavas, decrease systematically with increasing SiO\tsub{2},
contrary to Na\tsub{2}O and K\tsub{2}O, which increase from mafic to
felsic compositions. Al\tsub{2}O\tsub{3} contents increase in mafic
lavas and decrease in felsic lavas while MnO shows no significant
variation.

\begin{figure*}[t!]
\includegraphics{fig05}
{\vspace*{-.25pc}}
\caption{\label{fig5}Variation of major elements (Fe\tsub{2}O\tsub{3},
Al\tsub{2}O\tsub{3}, CaO, MgO Na\tsub{2}O, K\tsub{2}O,
P\tsub{2}O\tsub{5} and TiO\tsub{2}) as a function of the SiO\tsub{2} in
the studied lavas, together with other lavas from northern CCVL. Data
sources of the lavas from the northern part of the CCVL are from
\citet{Ngounounoetal2000}, \citet{Rankenburgetal2005},
\citet{GountieDedzoetal2019}, \citet{Tchouhlaetal2022} and
\citet{Djerossemetal2024}. The red dots represent the lava closest to
our study area.}
{\vspace*{-.45pc}}
\end{figure*}

\subsubsection{Trace elements} \label{sec4.3.2}

The concentrations of several trace elements, including Ni, Cu, Cr, V,
and Sr, decrease markedly from mafic to felsic lavas. Nickel ranges
from 16.93 to 435.58~ppm, with the highest contents recorded in mafic
lavas. Co varies from 2.56 to 61.96~ppm, Cr from 92.82 to 717.16~ppm,
and V from 1.45 to 241.03~ppm, all showing a progressive decrease from
basalts to trachytes. Sr contents range from 767.8 to 1678.3~ppm in
mafic lavas and from 14.6 to 27.81~ppm in felsic lavas. In contrast,
the concentrations of Y, Pb, Ga, Nb, Rb, and La generally increase from
mafic to felsic lavas. Y increases from 17.9~ppm in basalt sample Zi16
to 124.2~ppm in rhyolite M29; Pb ranges from 1.77~ppm in sample M33 to
19.98 ppm in rhyolite M29; Hf increases from 4.08~ppm in sample Zi10 to
34.36~ppm in sample M29; and U ranges from 0.77~ppm in sample M33 to
7.28~ppm in trachyte Zi8 (Table~\ref{tab1}). The distribution of trace
elements (V, Cu, Cr, Ni, Sr, Pb, Nb, and Ta) as a function of
SiO\tsub{2} content (Figure~\ref{fig6}) shows that the Ziver lavas plot
in the same fields with previously studied lavas from the northern
segment of the CCVL, and therefore exhibit the similar geochemical
characteristics. For example, all compatible elements (Cr, Ni and V)
show a systematic decrease with increasing SiO\tsub{2}. Conversely,
incompatible elements such as Nb, Ta, Pb, and Sr show a general trend
of increasing or remaining relatively stable with silica.

\begin{figure*}
\includegraphics{fig06}
{\vspace*{-.45pc}}
\caption{\label{fig6}Distribution of traces elements (V, Cu, Cr, Ni,
Sr, Pb, Nb and Ta) versus SiO\tsub{2} content in lavas from Ziver are.
Data sources of the northern part of the CCVL, and Symbols are the same
in Figure~\ref{fig5}.}
{\vspace*{-.45pc}}
\end{figure*}

The primitive mantle normalized multi-element patterns show that Ziver
lavas exhibit geochemical signatures remarkably consistent with those
described for volcanoes in the northern segment of the CCVL
\citep{Ngounounoetal2000, Rankenburgetal2005, GountieDedzoetal2019,
Tchouhlaetal2022, Djerossemetal2024}. Mafic lavas (Figure~\ref{fig7}a)
show a marked enrichment in highly incompatible elements (LILE and
HFSE), notably Cs, Rb, Ba, Th, U, Nb, Ta, K, and La, as well as a
gradual decrease in heavy rare earth element (HREE) contents. This
distribution, characterised by a generally convex curve and an absence
of significant anomalies in Nb--Ta or Ti, closely corresponds to the
spectrum of oceanic intraplate basalts (OIB) described by
\citet{SunMcDonough1989}.

\begin{figure}
\includegraphics{fig07}
{\vspace*{-.15pc}}
\caption{\label{fig7}Primive mantle-normalized multi-element spider
diagrams for Ziver lavas: (a)~mafic lavas and (b)~felsic lavas.
Primitive mantle composition is after \citet{McDonoughSun1995}. Data
for OIB, N-MORB, and E-MORB are from \citet{SunMcDonough1989}. Data
sources of the northern part of the CCVL are the same used in
Figure~\ref{fig6}.}
{\vspace*{-.45pc}}
\end{figure}

The felsic lavas of Ziver (Figure~\ref{fig7}b), represented by
trachytes and rhyolites, retain the enriched general structure observed
in mafic lavas, but are \mbox{accentuated} by higher normalisation
factors, a direct consequence of their advanced differentiation. These
lavas show: (i)~a marked enrichment in \mbox{incompatible} elements
(LILE and LREE) indicating strong magmatic differentiation, (ii)~a
clear negative anomaly in Eu, linked to plagioclase fractionation,
typical of evolved magmas, and (iii)~a relative depletion in HREE, but
less marked than in basalts, \mbox{indicating} melting at shallower
depths, without major influence on garnet.

The chondrite-normalised rare earth element spectra show that the mafic
lavas of Ziver (Figure~\ref{fig8}a) have moderately sloping profiles,
characterised by a strong enrichment in LREE and a gradual decrease
towards HREE. This geometry is almost parallel to the OIB spectra and
clearly distinct from the N-MORB or E-MORB signatures. There are no
significant anomalies in Eu or Ce. This enrichment in light rare earth
elements is marked by the high (La/Yb)$_{\mathrm{N}}$ ratio
(15.19--20.91) and the gradual decrease in MREE with a ratio (Gd/Yb)
(3.19--4.44). Depletion of medium and heavy rare earth elements is
evidenced by (La/Sm)$_{\mathrm{N}}$ ranging from 2.4 to 4.6 and
(Dy/Yb)$_{\mathrm{N}}$ from 1.32 to 1.61, respectively. The Ziver
mafic lavas show a slight positive Eu anomaly. Felsic lavas retain an
overall enriched pattern similar to that of mafic lavas, but with
higher LREE contents, reflecting their more advanced degree of
evolution. The presence of moderate negative anomalies in Eu can be
explained by the fractional crystallisation of plagioclase during the
late stages of magmatic evolution (Figure~\ref{fig8}b).
{\vspace*{-.2pc}}

\begin{figure}
\includegraphics{fig08}
{\vspace*{-.15pc}}
\caption{\label{fig8}Chondrite-normalized REE patterns for Ziver lavas:
(a)~mafic lavas and (b)~felsic lavas. Chondrite values are from
\citet{McDonoughSun1995}. Data for OIB, N-MORB, and E-MORB are from
\citet{SunMcDonough1989}.}
{\vspace*{-.3pc}}
\end{figure}

\subsection{Sr--Nd isotopes} \label{sec4.4}

Four new Sr and Nd isotopic ratios were obtained both in mafic
(basanite) and felsic (trachyte, rhyolite) lavas. Calculations for
major elements were performed on an anhydrous basis. Isotopic data were
corrected to 40~Ma, an age derived from nearby \mbox{rhyolites} on the
Kapsiki plateau \citep{Ngounounoetal2000}, due to their geospatial
proximity to the Ziver area.

The \tsup{87}Sr/\tsup{86}Sr isotopic ratios range from 0.70311 in the
basanite Zi10 to 0.71856 in the trachyte Zi5, while the
\tsup{143}Nd/\tsup{144}Nd ratios vary between 0.51295 and 0.51276,
respectively. The (\tsup{87}Sr/\tsup{86}Sr)$_{\mathrm{initial}}$
varies from 0.70361 (basanite Zi10) to 0.71207 (trachyte Zi5), and
(\tsup{143}Nd/\tsup{144}Nd)$_{\mathrm{initial}}$ ranges from 0.51282
(trachyte Zi5) to 0.51202 (basanite Zi10), with initial $\varepsilon$Nd
values varying from 3.86 (trachyte Zi5) to 7.44 (basanite Zi10).

\section{Discussion}\label{sec5}
\subsection{Petrogenesis of lavas} \label{sec5.1}

The coherence of trace-element patterns across the studied lavas
suggests a genetic relationship, \mbox{indicating} that they belong to
the same magmatic series despite the presence of a compositional gap.
This gap, which is characteristic of the Ziver lavas, emphasizes their
bimodal nature-a feature also documented in several other continental
volcanic massifs along the CCVL, including Nganha \citep{Nonoetal1994},
Bambouto, Oku, and Ngaound\'{e}r\'{e} \citep{Marzolietal1999}, the
Kapsiki Mountains \citep{Ngounounoetal2000}, the Benue Trough
\citep{Ngounounoetal2003}, the Bamenda Mountains
\citep{Kamgangetal2010}, Mbengwi \citep{Mbassaetal2012}, and Tchabal
Gangdaba \citep{Itigaetal2013}.\looseness=-1

Several geochemical criteria support a mantle-derived origin and
subsequent evolutionary differentiation of these magmas. The basanites,
interpreted as the most primitive lavas in the suite, are marked by
high Ni contents (265.72--422.58~ppm) and elevated Mg\# values
(59.74--71.77), consistent with crystallization from an
undifferentiated, mantle-derived mafic magma. 

The major variation diagrams show that the lavas from the Ziver area
follow the same evolutionary trends as those from the northern part of
the CCVL, characterised by an alkaline series controlled by fractional
crystallisation of mafic phases and plagioclase. The basic compositions
overlap closely, indicating a common mantle source. However, Ziver
exhibits more dispersed intermediate to differentiated compositions, as
well as more evolved terms (trachytes--rhyolites), suggesting more
advanced differentiation and locally distinct magmatic evolution
conditions within the CCVL.

Overall, the Ziver lavas closely overlap with the geochemical fields
defined for the volcanoes of the northern CCVL
\citep{Ngounounoetal2000,Rankenburgetal2005,GountieDedzoetal2019,
Tchouhlaetal2022,Djerossemetal2024,Djamilatouetal2025}.
Their composition indicates a mantle source analogous to that of other
edifices in the area and some evolution dominated by fractional
crystallisation, with no significant crustal input. These results
indicate that the Ziver volcanism represents an additional coherent
segment of the regional magmatic field, extending the petrogenetic
processes defined for the northern part of the\break CCVL.

Normative mineralogy further constrains their silica saturation state:
the mafic lavas are silica-undersaturated, as indicated by the presence
of \mbox{normative} nepheline, reaching up to 12.03~vol\% in sample
M33. This high nepheline content underscores their pronounced alkaline
affinity. In contrast, the felsic lavas are silica-oversaturated, with
SiO\tsub{2} contents ranging from 64.66 to 70.30~wt\% and containing
normative quartz (1.62--18.20~vol\%) and hypersthene
(5.04--7.97~vol\%). Their peralkaline index [(Na\tsub{2}O ${+}$
K\tsub{2}O)/Al\tsub{2}O\tsub{3}] varies from 0.92 to 1.08, reflecting
the peralkaline nature of the felsic lavas. The combination of high
SiO\tsub{2} contents and peralkaline character is a diagnostic feature
of evolved alkaline magmas.

\subsection{Nature of the source(s) of the Ziver volcanic rocks}
\label{sec5.2}

The chemical composition of the mantle source can be constrained using
incompatible element ratios (e.g., Nb/Ta, Nb/U, Ta/U), which remain
relatively stable during magmatic differentiation. The Nb/Ta, Nb/U, and
Ta/U ratios of the mafic lavas (0.79--0.82, 1.24--2.14, and 0.71--0.81,
respectively) closely match those of the felsic lavas (0.77--0.82,
1.02--2.75, and 0.76--1.10, respectively), supporting a cogenetic
relationship between the two series. Both groups are enriched in
incompatible elements, consistent with derivation from relatively low
degrees of partial mantle melting. Their (Ce/Yb)$_{\mathrm{N}}$
ratios (5.56--17.94) partly overlap the range defined by
\citet{Deruelleetal2007} for the continental sector of the CCVL
(12--22; average ${\approx}$ 15) and are similar to values reported for
the Kapsiki Plateau \citep[(Ce/Yb)$_{\mathrm{N}} \approx
14.3$;][]{Ngounounoetal2000}. These comparatively low ratios, combined
with strong enrichment in light rare earth elements (LREE), point
toward low to moderate degrees of mantle\break melting.

The almost complete overlap of the Ziver lavas with the geochemical
fields established for neighbouring edifices north of the CCVL (Gawar:
Gounti\'{e} Dedzo et~al., \citeyear{GountieDedzoetal2019}; Mokolo
Hossehone: Tchoulha et~al., \citeyear{Tchouhlaetal2022}; Rankenburg
et~al., \citeyear{Rankenburgetal2005}, Kapsiki: Ngounouno et~al.,
\citeyear{Ngounounoetal2000}; \citep{Djamilatouetal2025}, Iriba: Djerossem
et~al., \citeyear{Djerossemetal2024}) highlights regional homogeneity
in the nature of the mantle source and in the differentiation
processes.\ The data confirm that the Ziver magmas derive from an
OIB-type enriched mantle source, having undergone \mbox{progressive}
differentiation controlled mainly by fractional crystallisation,
without a major crustal contribution. This~\mbox{pattern} reinforces
the idea that Ziver volcanism is a further expression of the intraplate
alkaline magmatism characteristic of the\break CCVL.

Additional constraints on mantle source characteristics are provided by
the Gd/Yb versus La/Yb and Sm/Yb versus La/Sm diagrams
(Figure~\ref{fig9}), combined with partial melting models for
lherzolite with homogeneous compositions \citep{Yokoyamaetal2007}. In
these diagrams, the Ziver samples fall along the melting trajectories
of garnet-bearing peridotite, implying magma generation at depths
consistent with the stability of the garnet in the residue
(${\sim}$70--80~km). This is further confirmed by
very high (La/Yb)$_{\mathrm{N}}$ ratios (15.19--20.91) and Gd/Yb
(3.19--4.44), typical of magmas resulting from the melting of a
garnet-bearing mantle. Elevated Sm/Yb, Gd/Yb, La/Yb, and La/Sm ratios
indicate very low degrees of partial melting (0.5--3\%) of a
garnet-bearing enriched mantle source (Figure~\ref{fig9}), involving
metasomatized lithospheric mantle and/or plume-derived enriched
domains. This limited degree of melting explains the pronounced
enrichment of incompatible elements, particularly LREE, which is a
hallmark of OIB-type magmas. The Ziver lavas therefore share
geochemical affinities with other volcanic centers of the CCVL,
including Mt.~Cameroon \citep{Asaahetal2015}, and Mokolo-Kossehone
\citep{Tchouhlaetal2022}, Mbengwi \citep{Mbassaetal2012}, Gawar and
Zama\"{i} \citep{GountieDedzoetal2019}, Iriba
\citep{Djerossemetal2024}, and the Kapsiki Plateau
\citep{Ngounounoetal2000}. Slight variations in melting degree and
depth of generation likely reflect small-scale heterogeneities in the
mantle source but overall confirm the existence of a common deep mantle
reservoir beneath the CCVL.

\begin{figure}
\includegraphics{fig09}
{\vspace*{-.3pc}}
\caption{\label{fig9}Gd/Yb versus La/Yb diagram illustrating the
partial melting of the Ziver and Mount Cameroon basaltic lavas. The
breakdown of the Mount Cameroon zone is from \citet{Yokoyamaetal2007}.
The two curves marked Grt 4\% and 8\% represent the garnet content of
the source. Calculations of garnet grades in the source were made using
the partition coefficients of \citet{Hallidayetal1990}; (b)~plot of the
Ziver lavas in the Sm/Yb versus La/Sm diagram after
\citet{Gurenkoetal2006}. The red dotted line field is that of basaltic
lavas from northern Cameroon. Values of DMM (Depleted MORB Mantle) are
from \citet{WorkmanHart2005}. Values of PM are from
\citet{SunMcDonough1989}. The different curves represent the partial
melting of garnet and spinel peridotites. The gradations represent the
different melting rates.\looseness=-1}
{\vspace*{-.25pc}}
\end{figure}

The Nb/Y versus Zr/Y diagram of \citet{Condie2005} (Figure~\ref{fig10})
further supports this interpretation: the Ziver lavas plot above the
$\Delta$Nb line, within the plume-derived enriched mantle domain
defined by \citet{Fittonetal1997}.
Their geochemical signature is characteristic of intraplate alkaline
basalts, comparable to OIBs, and consistent with derivation from an
enriched asthenospheric mantle plume source. Isotopic data
\citep[\tsup{143}Nd/\tsup{144}Nd versus
\tsup{87}Sr/Sr\tsup{86};][]{Foucarde1998}
show that the basanites fall along the mantle array within the depleted
mantle (DM) field, suggesting minimal crustal contamination. By
contrast, the more evolved trachytes and rhyolites display enrichment
in radiogenic Sr, indicating the assimilation of the upper continental
crust during their evolution.

\begin{figure}
\includegraphics{fig10}
\caption{\label{fig10}Tectonic setting of Ziver mafic lavas in relation
to various mantle components and basalt fields, as defined by
\citet{Weaver1991} and \citet{Condie2005}. DEP ${=}$ Deeply Depleted
Mantle; REC ${=}$ Recycled Component; PM ${=}$ Primitive Mantle, DM ${{=}{:}}$
depleted mantle, HIMU ${=}$ high-$\upmu$, EM~${=}$ Enriched mantle.}
{\vspace*{-.45pc}}
\end{figure}

Taken together, these data indicate that the Ziver lavas were generated
by the upwelling of a mantle plume in a post-orogenic intraplate
setting, possibly linked to late-stage rifting. This model is
consistent with both the OIB-like isotopic signatures observed in other
CCVL volcanic centers and geodynamic models that invoke deep mantle
upwelling beneath Central Africa.
\vspace*{-.25pc}

\subsection{Magma differentiation processes and crustal contamination}
\label{sec5.3}

\vspace*{-.25pc}

To better constrain the geochemical characteristics of the Ziver
volcanic rocks we investigated the \mbox{potential} interactions
between molten rocks and crustal rocks during the ascent of magma to
the surface, since crustal contamination is an effective means of
modifying the composition of trace elements \citep{McDermottetal2005,
Zellmeretal2005}. Although not consistently observed, previous studies
have already suggested that contamination of primary magmas
by continental crust is significant along the CCVL
\citep{Hallidayetal1988,Marzolietal1999, Rankenburgetal2005,
Kamgangetal2010, Kamgangetal2013, Njonfangetal2013,
TchuimegnieNgongangetal2015}. The influence of the
basement on the composition of magmas can be confirmed by trace element
ratios. Crustal rocks and melts derived from them are characterized by
higher Rb/Nb ratios, whereas alkaline basaltic magmas are relatively
enriched in Nb and therefore have a relatively low Rb/Nb ratio
\citep{WeaverTarney1984}.
Consequently, the low Rb/Nb ratios (0.49--1.05) observed in the
basaltic rocks of Ziver are compatible with the absence or very low
crustal contamination.

Furthermore, the Ziver basanites display primitive mantle--normalized
trace element patterns similar to those of uncontaminated OIBs
(Figure~\ref{fig7}a), consistent with observations reported for other
lavas of the CCVL \citep{Marzolietal2000, Rankenburgetal2005,
Yokoyamaetal2007, Akaetal2009, Kamgangetal2013, Asaahetal2015}. The
positive Nb--Ta anomaly in the Ziver basanite lavas is consistent with
the presence of peridotite xenoliths \mbox{observed} in most basanites
of the study area, suggesting rapid magma ascent to the surface and
ruling out significant crustal contamination. In contrast, the
differentiated Ziver lavas (trachytes and rhyolites) display a negative
Nb--Ta anomaly (Figure~\ref{fig7}b), indicating limited magma--crust
interaction.

The Nb/Y versus Rb/Y diagram of \citet{CoxHawkesworth1985} and 
\citet{LeemanHawkesworth1986}
is indeed generally used to assess the influence of crustal
contamination on the chemical composition of the magma shown in
Figure~\ref{fig11}. The lavas studied are grouped into a domain
characteristic of poorly differentiated alkaline magmas, with moderate
Nb/Y ratios (1--3) and low to moderate Rb/Y ratios (1--4). Their
position is close to that of the basanites, basalts, and hawaiites of
the CCVL, indicating a mantle signature similar to that of Pan-African
alkaline volcanism.

\begin{figure}
\includegraphics{fig11}
\caption{\label{fig11}Position of the studied lavas in the Nb/Y versus
Rb/Y diagram of \citet{CoxHawkesworth1985} and 
\citet{LeemanHawkesworth1986}. Data for Precambrian basement are from
\citet{Tchamenietal2016}. Data sources from the northern part of the
CCVL are similar to Figure~\ref{fig6}.}
\end{figure}

The black cloud representing regional data shows an upward trend
(increase in Nb/Y) consistent with fractional crystallisation from a
parental basanitic magma. The Ziver lavas follow this trend, suggesting
that they derive from a parent magma common to the northern CCVL. The
red dot (the lava \mbox{closest} to Ziver) overlaps with the field of
undersaturated alkaline magmas of the CCVL, supporting a direct
petrogenetic link. The Precambrian basement samples (Mokong and Ziver)
show low Nb/Y values (${\leq}$0.5) and higher Rb/Y values,
characteristic of a crustal substrate enriched in incompatible
elements. None of the lava samples show a clear trend towards these
compositions, indicating limited crustal assimilation. Overall, their
trace element distributions are consistent with mantle-derived alkaline
magmatism that evolved in a closed system through fractional
crystallization. This supports an intraplate tectonic setting involving
an enriched deep-mantle source with limited crustal interaction,
similar to that inferred for other volcanic centers of the CCVL, such
as the Adamaoua Plateau \citep{Nkouandouetal2010, Fagnyetal2020,
Mebaraetal2022, DiliRakeetal2022}; the Gawer and Zama\"{i} lavas
\citep{GountieDedzoetal2019}; the Kapsiki plateau lavas
\citep{Ngounounoetal2000, Djamilatouetal2025}; the Iriba lavas in Chad
\citep{Djerossemetal2024}.

\section{Conclusions}\label{sec6}

The Ziver volcanic area includes both mafic (basanite, basalt, and
hawaiite) and felsic lavas (trachyte and rhyolite), defining a bimodal
series with a Daly gap between 47 and 64.7~wt\% SiO\tsub{2}. The
geochemical data obtained on Ziver lavas show that this volcanic center
is fully integrated into the magmatic dynamics of the northern segment
of the CCVL. The trends in trace elements, the
distribution of multi-element diagrams and rare earth spectra
normalised to the primitive mantle and chondrites respectively reveal
signatures typical of intraplate alkaline magmatism derived from an
OIB-type enriched mantle source. The absence of negative anomalies in
Nb--Ta or Ti, combined with systematic enrichment in LILE, HFSE, and
LREE, rules out significant crustal input and indicates that the magmas
originated from a moderately heterogeneous mantle, similar to that
feeding the other volcanic edifices in the northern part of the CCVL.
These studies reveal a deep, enriched mantle origin for the Ziver
lavas. Diagrams of rare earth element ratios (La/Yb, Gd/Yb, Sm/Yb,
La/Sm) indicate that the magmas originate from the partial melting
(0.5--5\%) of a garnet-lherzolite mantle, suggesting a melting depth
greater than 80~km. The low initial \tsup{87}Sr/\tsup{86}Sr ratios:
0.7016--0.7043 and positive $\varepsilon$Nd isotope of the basaltic
lavas confirm their mantle origin. The isotopic data from the Ziver
lavas (\tsup{143}Nd/\tsup{144}Nd versus \tsup{87}Sr/\tsup{86}Sr)
position the samples on the mantel array, with no significant influence
on the continental crust. In addition, the Nb/Y versus Zr/Y diagram
places the compositions within the field of mantle feather-type sources
(OIB), notably HIMU-EM1, confirming an origin linked to intraplate
activity or a mantle thermal anomaly. The relationships between the
ratios of the different trace elements (Nb/Y and Rb/Y) demonstrate a
magmatic differentiation process mainly controlled by fractional
crystallization with a negligible contribution from crustal
assimilation, as confirmed by the strong chemical distinction from the
Precambrian basement around Ziver.

\section*{Acknowledgements}

This article is part of a Doctoral thesis in progress by the first
author. The current work was financially supported by the
CNRS--IRD--LithoCOAC project. The authors would like to thank Fabienne
De Parseval for thin sections preparation, Philippe De Parseval, Sophie
Gouy, and Sophie Mandrou for their technical assistance, respectively
during electron microprobe and TIMS analyses. We also thank two
anonymous reviewers for their significant and constructive comments,
and the editorial assistance of Adenise Lopes.

\printCOI

\back{}

\printbibliography
\refinput{crgeos20250980-reference.tex}

\end{document}