Over the past 25 years, numerous authors have linked climate change and hominin evolution based on the premise that resources essential to hominin development (e.g., brain size, cognition and behavior), such as water and food availability, have fluctuated in tandem with climate in the East African Rift (deMenocal, 1995, 2004, 2011; Vrba et al., 1995; Potts, 1998, 2013; Bobe and Behrensmeyer, 2004; Wynn, 2000, 2004; Trauth et al., 2005, 2007, 2010, 2015; Behrensmeyer, 2006; Bobe, 2006; Bonnefille, 2010; Levin et al., 2011; Cerling et al., 2011; Maslin et al., 2014, 2015; Levin, 2015; Potts and Faith, 2015; Cuthbert et al., 2017; Blumenthal et al., 2017; Timmermann et al., 2022). An important related debate about the relationship between climate change, hominin diet and behavior concerns landscape aridification in the Turkana Basin (Kenya, Ethiopia) (Blumenthal et al., 2017; Fortelius et al., 2016; Lepre and Quinn, 2022; Quinn and Lepre, 2020, 2021; Wynn, 2004). During the Neogene, topographic uplift altered atmospheric moisture transport and distribution in the East African Rift, and the initiation of Northern Hemisphere glaciations at the end of the Pliocene intensified hydrological deficits in this region (Blumenthal et al., 2017; Laporte and Zihlman, 1983; deMenocal, 2004; Sepulchre et al., 2006; Trauth et al., 2021). Along with hypotheses linking hominin adaptation to landscape aridification, the availability of water and food would also have been critical resources for their evolution (e.g., Cuthbert et al., 2017). The debate on aridification and drinking water and food availability is especially relevant to hominin evolution, as some previous hypotheses, such as the savanna and the turnover pulse hypotheses (Vrba et al., 1995; deMenocal, 2004) or the pulsed climate variability hypothesis (Maslin et al., 2015; Maslin and Trauth, 2009), considered climate instability to be a key driver of natural selection, adaptive divergence, and speciation in Plio–Pleistocene hominin evolution. Although the global climate changed during the Plio–Pleistocene transition, the responses of these changes on landscape, fauna and hominins are still debated in the East African Rift and in the different regions of the Turkana Basin in particular (Blumenthal et al., 2017; Blumenthal et al., 2017, 2017; Fortelius et al., 2016; Lepre and Quinn, 2022; Quinn and Lepre, 2020, 2021; Villaseñor et al., 2020, 2023). The aims of this review are to examine the archeological and hominin records from the western region of the Turkana Basin (Kenya; Fig. 1) and the evidence for the existence of freshwater conditions and resources in the western region of Turkana Basin between 4.20 and 0.70 Ma. This specific region of the Turkana Basin provides evidence of hominin and archeological sites dating back to the Late Pliocene–Early Pleistocene, raising new questions about the role of climate and environmental change on water and food resources and hominin evolution.

The western region of the Turkana Basin is unique, as the archeological site of Lomekwi 3 has yielded the first lithic assemblage associated with Late Pliocene sediments, climate, and environment (Harmand et al., 2015; Lewis and Harmand, 2016; Quinn et al., 2021). Lomekwi 3 is at least 300 kyr older than the earliest known Oldowan assemblages discovered at other localities (i.e., Gona, Ledi-Geraru, and Nyayanga; Semaw, 2006; Braun et al., 2019; Plummer et al., 2023). Other important archeological sites have been discovered in the same region of the basin over the past 30 years. For example, the Lokalalei 1A and 2C sites are among the oldest Oldowan assemblages known in Kenya and older than 2.2 Ma (Delagnes and Roche, 2005; Harmand, 2009; Kibunjia et al., 1992; Roche et al., 1999). The artifacts from Kokiselei 6 are the earliest known evidence for the emergence of bifacial shaping (Duke et al., 2021), and the archeological site of Kokiselei 4 has yielded the oldest Acheulean assemblage presently known (Lepre et al., 2011; Texier, 2018). These archeological sites provide some of the earliest evidence of hominin cognitive abilities in eastern Africa. As for the hominin record, the western region of the Turkana Basin has yielded some of the most important hominin fossils known, including Australopithecus anamensis (Ward et al., 1999, 2001, 2013, 2020; Leakey et al., 1995, 1998), Kenyanthropus platyops (Leakey et al., 2001; Spoor et al., 2016), Paranthropus aethiopicus (Walker et al., 1986) and Paranthropus boisei (Leakey and Walker, 1988; Brown et al., 2001; Prat et al., 2003; Wood and Constantino, 2007), as well as specimens assigned to early Homo (Prat et al., 2005), Homo ergaster or Homo erectus, referred to here as Homo ergaster/erectus (Wood and Leakey, 2011; Prat et al., 2003; Walker and Leakey, 1993; Wood and Boyle, 2016), including the nearly complete skeleton of KNM-WT 15000 (‘Nariokotome Boy'; Walker and Leakey, 1993). This fossil record highlights the importance of the western region of the Turkana Basin in terms of hominin taxonomy and phylogeny (e.g., Leakey et al., 2001; Walker and Leakey, 1993; Ward et al., 2020). The western region of the Turkana Basin is also very rich in freshwater micro- and macrofauna demonstrating that freshwater conditions expanded into this arid region of the East African Rift during the Plio–Pleistocene (e.g., Harris et al., 1988; Martin and Trautwein, 2003; Meylan et al., 1990; Stewart, 2003; Stewart and Rufolo, 2020; Storrs, 2003; Van Bocxlaer, 2011, 2020; Williamson, 1981).

Similar, coeval records exist to the northern regions of the Turkana Basin in the Shungura, Mursi and Usno Formations (Ethiopia) along the Omo River and to the eastern regions in the Koobi Fora Formation (Kenya) and at Fejej and Konso (Ethiopia) (de Heinzelin, 1983; Williamson, 1981; Howell et al., 1987; Stern et al., 1993; Isaac, 1997; Nagaoka et al., 2005; WoldeGabriel et al., 2005; Chapon et al., 2005, 2011; Van Bocxlaer et al., 2008; Braun et al., 2010a; Barsky et al., 2011; Delagnes et al., 2011; Leakey et al., 2012; Maurin et al., 2017, Fig. 1). Reviews of the Turkana Basin generally combine data from these different fluvial and lacustrine formations (i.e., located in the northern, western, eastern and southern regions of the Turkana Basin). However, this review covers only the western region of the Turkana Basin (Nachukui and Kanapoi Formations) and does not include data from the northern and northeastern regions of the basin (Shungura and Koobi Fora Formations). In these latter regions, freshwater conditions are more permanent, due to the presence of the Omo River delta. By contrast, in the western and southern regions of the Turkana Basin, conditions have been more alkaline–saline since the closure of the outlet to the Indian Ocean at around 2.2 Ma (Boës et al., 2019a). Given the reorganization of the hydrography, it is important to consider the mosaic of landscapes with their own rhythms, hydoconnectivity and hydroclimatic influences from the northern, southern, eastern and western regions of the Turkana Basin. Here, we focus on understanding how landscapes changed regionally along the western region of the Turkana Basin, where the hydrology and limnology is substantially different from that of the Omo delta region. Moreover, in terms of hydroconnectivity, the upland river-lake systems in the western region of the Turkana Basin are associated with the hydroclimate of the Kenyan Highlands, while in the eastern region of the basin they are associated with the hydroclimate of the Ethiopian Highlands. Finally, volcano-tectonic activity also affected the western and eastern regions of the Turkana Basin differently during the Plio-Pleistocene.

The various sources of data from the western region of the Turkana Basin (see Table 1, Table 2, Table 3, Table 4, Table 5) can be compiled in a reliable chronostratigraphic framework derived from 40Ar/39Ar ages of volcanic tuffs and/or lateral correlations by geochemical signatures specific to each tuff (e.g., Brown and McDougall, 2011; Feibel et al., 1989; Harris et al., 1988; McDougall et al., 2012). Plio–Pleistocene sediments and tuffs are outcropping in the Nachukui Formation, including Lothagam, Loruth Kaado, South Turkwel and in the Kanapoi Formation (Fig. 1; Table 2). The Nachukui and Kanapoi Formations offer ∼700 m of strata with regularly intercalated tuffs that cover the period from 4.20 to 0.70 Ma without major chronostratigraphic gaps, although the interval between 3.1 and 2.7 Ma is represented by fewer outcrops. In this review, we consider that it is not possible at this time to provide accurate ages within the interval between 3.1 and 2.7 Ma because there are no dated tuffs or paleomagnetic inversions, which considerably limits the reliability of sediment chronologies (and/or results in excessively large errors; see section 3.6 below on methods for controlling chronology). In what follows, ages in Ma with two decimal places are argon–argon dated tuffs or ages that were directly interpolated from these tuffs, whereas ages in Ma with one decimal place either were derived from paleomagnetism or were estimated.

We begin by introducing the impact of global climate change on eastern Africa during the Plio–Pleistocene and present our current understanding of the influence of regional climate on the Turkana Basin. We then provide an overview of the former hydroconnectivity of the Turkana Basin and highlight the topoclimatic effects on the regional hydroclimate system. Subsequently, we present the datasets and criteria we used to reconstruct the paleohydrological evolution of the western region of the Turkana Basin and the development of freshwater systems accompanied by freshwater flora and micro- and macrofauna in this region. A synthesis is provided based on available datasets between 4.20 and 0.70 Ma and on field data verifications that were conducted between 2010 and 2023 in the western region of the Turkana Basin (Mission Préhistorique au Kenya; West Turkana Archaeological Project). We then relate our findings from this synthesis to the stratigraphic and spatiotemporal contexts of the archeological and paleoanthropological sites of the western region of the Turkana Basin. Along with the hypotheses linking increasing aridity to hominin evolution, we discuss the availability of drinking water and freshwater food resources for hominins occupying this arid region of the rift.