Thanks to continued paleontological surveillance of the digging activity during the construction of a dump between 2002 and 2014, a diverse primate assemblage (Alba, 2012; Marigó et al., 2014; Alba et al., 2017; DeMiguel et al., 2021; Fortuny et al., 2021) has been recovered from the Middle to Late Miocene stratigraphic sequence of Abocador de Can Mata (ACM; Alba et al., 2006, 2011a, 2017) in the fossiliferous area of els Hostalets de Pierola (Catalonia, Spain). Both small-bodied catarrhines (Alba et al., 2010; 2012a, 2015) and large-bodied apes—Pierolapithecus (Moyà-Solà et al., 2004; Almécija et al., 2009; Pérez de los Ríos et al., 2012; Hammond et al., 2013; Pina et al., 2014), Dryopithecus (Moyà-Solà et al., 2009a; Alba and Moyà-Solà, 2012; Pina et al., 2019, 2020), and Anoiapithecus (Moyà-Solà et al., 2009b; Alba et al., 2013)—have been recovered there, spanning from ∼12.4–12.3 to ∼11.6 Ma (Alba et al., 2017). Although litho-, bio-, and magnetostratigraphic correlation has enabled an accurate dating of the fossil primate remains from ACM (Moyà-Solà et al., 2009a; Casanovas-Vilar et al., 2011, 2016a, 2016b; Alba et al., 2017), this does not apply to all the previous hominoid finds from the same area.

A mandibular fragment with M2–M3 (IPS1826+1827) found in 1941 at Can Vila (Villalta Comella and Crusafont Pairó, 1941a, 1944) is the holotype of ‘Sivapithecus’ occidentalis, which is considered a species inquirenda (Alba et al., 2020; Fortuny et al., 2021)—i.e., of uncertain taxonomic validity because of the lack of comparable elements in other species. This specimen comes from minimally older levels than ACM/BCV1 (Barranc de Can Vila 1, the type locality of Pierolapithecus catalaunicus) and thus has an age of ∼12.0 Ma (Alba et al., 2013, 2017, 2020). However, the age of the two other previous finds is more uncertain. Thus, the stratigraphic provenance of an M2 (MGSB48486) from Can Mata s.l. (Van der Made and Ribot, 1999), currently assigned to Dryopithecus fontani (Alba et al., 2013, 2017, 2020; Fortuny et al., 2021), is unknown. Therefore, it can only be considered broadly younger than the hominoid finds from ACM (Alba et al., 2013), but it is impossible to determine whether it is latest Aragonian (Mammal Neogene [MN] zone MN7+8) or earliest Vallesian (MN9) in age (i.e., ∼11.3–11.0 Ma). The third previous hominoid find from the area is a C1 of a female individual (IPS1766; Fig. 1) from the locality of Can Mata 1 (CM1) or Bretxa de Can Mata. It was originally assigned to Hispanopithecus laietanus (Crusafont-Pairó and Golpe-Posse, 1973; Golpe-Posse, 1993) or Dryopithecus laietanus (Begun et al., 1990; Harrison, 1991; Andrews et al., 1996), but most recently it was left unassigned to genus as Hominidae indet. (Casanovas-Vilar et al., 2011; Alba, 2012) or Dryopithecinae indet. (Alba et al., 2013, 2020; Marigó et al., 2014). This canine has been excluded from the hypodigm of H. laietanus because this species is not recorded until much later (∼10.0 Ma; Alba et al., 2018) and also because the specimen cannot be directly compared with any of the ACM dryopithecine taxa. Although the provenance of IPS1766 from CM1 is well documented (Crusafont-Pairó and Golpe-Posse, 1973), its age is uncertain owing to the lack of magnetostratigraphic data and poor stratigraphic control of historical finds of hipparionin equids from the area of els Hostalets de Pierola.

The localities of Can Mata are situated nearby the farmhouse of Can Mata de la Garriga within the fossiliferous area of els Hostalets de Pierola, in the Penedès Sector of the Vallès-Penedès Basin (Fig. 2; Casanovas-Vilar et al., 2016a; for details on the geological background, see Casanovas-Vilar et al., 2008; Alba et al., 2009, 2017; Moyà-Solà et al., 2009a). Three Can Mata localities (CM1, CM2, and CM3) have been reported in the literature (Supplementary Online Material [SOM] S1.1), although only CM1 has yielded a rich fossil assemblage. The locality of CM1 (Fig. 3) was discovered in 1941 close to the farmhouse of Can Mata de la Garriga (Fig. 4), being considered pre-Vallesian given the absence of hipparionins and giraffids (Villalta Comella and Crusafont Pairó, 1941b, 1943; Crusafont and Truyols, 1954). In turn, CM2 was initially considered Vallesian in age because of the record of giraffids and the presence of Hippotherium molars in the same area (Crusafont Pairó, 1952; Crusafont and Truyols, 1954). Finally, CM3 corresponds to an isolated find of Hippotherium ∼200 m in NE direction from CM1 (Fig. 4c; Alba et al., 2006; SOM S1.1), which represents the oldest hipparionin find of known provenance from the area of Can Mata. Over the years, and only with a few exceptions (Mein, 1990; Andrews et al., 1996; Pickford, 2013), most authors have correlated CM1 to the latest Aragonian on biostratigraphic grounds (Crusafont-Pairó and Golpe-Posse, 1973; Golpe-Posse, 1974, 1982, 1993; Agustí et al., 1985; de Bruijn et al., 1992; Alba et al., 2006, 2011a, 2017; Moyà-Solà et al., 2009b; Casanovas-Vilar et al., 2011, 2016a, 2016b; Robles et al., 2013), and the same applies to CM2 (Crusafont Pairó and Golpe Posse, 1971; Golpe-Posse, 1974; Agustí et al., 1985, 1997, 2001) with the exception of de Bruijn et al. (1992).

The published faunal list of large mammals from CM1 needs to be updated and revised to confirm the identity of multiple taxa that is confusing according to published sources (see SOM S1.2 for further details). However, it is generally indicative of an age close to the Aragonian/Vallesian boundary. The frequent correlation of CM1 to MN7+8 instead of MN9 almost entirely relies on the lack of Hippotherium and giraffids. These large mammals have even been formally considered in the local biozonation of the Vallès-Penedès Basin (Casanovas-Vilar et al., 2016b). Even the claim that CM2 was located above CM1 (e.g., Agustí et al., 1985) might be exclusively based on the fact that Golpe-Posse (1974) roughly correlated CM1 with La Grive-Saint-Alban in France (MN7+8) and CM2 with Castell de Barberà (then considered latest MN7+8, but currently considered earliest MN9; Alba et al., 2019), based on the presence of giraffids in the latter locality. Along the same line of reasoning, Agustí et al. (1985) remarked that not only Hippotherium but also giraffids—traditionally considered to predate the arrival of hipparionins in the Vallès-Penedès Basin (e.g., Golpe-Posse, 1974; Crusafont Pairó, 1975; Agustí et al., 1985, 1997, 2001)—were missing from CM1.

The important role played by hipparionins and giraffids in the local biozonation of the Vallès-Penedès Basin across the Aragonian/Vallesian transition might seem surprising, given that most of the local biozones are based on rodents (e.g., Agustí et al., 1997; Casanovas-Vilar et al., 2016b). However, besides the role played by hipparionins in the definition of the Vallesian, there are other historical reasons that explain this situation. In particular, CM1 only yielded very scarce micromammal remains, contrasting with the hundreds of micromammal specimens recovered from the so-called ‘bloc de marga’ (marl block)—originally attributed to nearby Mas d'Ocata (Schaub, 1944, 1947), but actually coming from Can Flequer (Agustí, 1981; Agustí et al., 1985), which is Vallesian based on the presence of Hippotherium. Agustí and colleagues (Agustí, 1981, 1999; Agustí and Gibert, 1982; Agustí et al., 1985, 1997, 2001; Agustí and Moyà-Solà, 1991) relied on a composite micromammal faunal list for all the purportedly pre-Vallesian outcrops from the area, collectively denoted as Hostalets Inferior. These authors noted similarities with micromammals from the unambiguously Vallesian localities (Hostalets Superior) and concluded that their composition remained virtually unchanged through the Aragonian/Vallesian boundary (see also Casanovas-Vilar et al., 2011, 2016a, 2016b). Needless to say, this practice might lead to circular reasoning (particularly if Vallesian levels are mistaken as pre-Vallesian due to the absence of Hippotherium). Nevertheless, ever since then, it has been customarily accepted that the latest Aragonian and earliest Vallesian rodent assemblages from the Vallès-Penedès Basin are virtually indistinguishable (Agustí et al., 1997; Alba et al., 2006; Casanovas-Vilar et al., 2011, 2016a, 2016b).

Alba et al. (2006) remained skeptical about whether CM2 was latest Aragonian or earliest Vallesian, but given that CM1 is a well-sampled locality, they assumed that the absence of Hippotherium was not a sampling bias and accepted its customary correlation to MN7+8. Under such an assumption, Moyà-Solà et al. (2009b: Fig. 5) and Casanovas-Vilar et al. (2011) correlated CM1 to the upper part of the reverse polarity subchron C5r.2r (postdating the short-duration normal polarity cryptochron C5r.2r-1), implying an age comprised between 11.263 and 11.188 Ma (Ogg, 2020). Casanovas-Vilar et al. (2016a) further correlated CM1 to the Democricetodon crusafonti–Hippotherium interval subzone of the Vallès-Penedès Basin (Casanovas-Vilar et al., 2016b), whose top boundary in this basin is defined by the first local occurrence (FLO) of Hippotherium at 11.18 Ma (Garcés et al., 1996, 1997; Agustí et al., 1997; Casanovas-Vilar et al., 2016a, 2016b; Alba et al., 2019). In turn, based on the assumption that CM3 was located ∼40 m above CM1 (Alba et al., 2006), Moyà-Solà et al. (2009a) tentatively correlated the former to C5r.1r, although such a correlation was revised to C5n.2n by Casanovas-Vilar et al. (2016b: Fig. 3), following some corrections on the then available composite magnetostratigraphic sequence for ACM and Ecoparc de Can Mata (ECM; Alba et al., 2012b).

The tentative magnetostratigraphic correlation of CM1 by Moyà-Solà et al. (2009b) was based on its geographic location and its lithostratigraphic correlation with paleomagnetic samplings in the nearby Riera de Claret—as CM1 was then located outside the area of the Can Mata landfill. Nevertheless, Robles et al. (2013: 999) noted that “paleomagnetic sampling would be required in order to confirm correlation [of CM1] with C5r.2r […] rather than C5r.1n.” The resumption of the digging activity at ACM in 2018, to enlarge further the landfill, provided the opportunity to extend the sequence upward toward the Vallesian—including the stratigraphic section where CM1 is located, as well as to measure in situ the stratigraphic distance between CM1 and CM3. New paleomagnetic samples were therefore taken in 2020 to update the ACM composite stratigraphic sequence (so as to accurately date the fossil remains recovered between 2018 and 2020) and provide a revised magnetostratigraphic correlation and interpolated ages for CM1 and CM3. Given their biochronological implications, the oldest well-dated hipparionin and giraffid remains from the area of els Hostalets de Pierola are also reported. Finally, our results are discussed in relation to phylogenetic and paleobiogeographic uncertainties regarding the evolutionary origin of Vallesian dryopithecines from Western Europe.