Pleistocene and Holocene procurement of the clam Donax serra (Bivalvia: Donacidae) along the southern Cape coast of south Africa

•Earliest Donax serra clams in southern Africa were harvested unselectively c.110 ka.•By c. 91 ka, people at PP13B focussed on mid-intertidal where large clams live.•New 115 ka data show small sizes, but lack of younger data hampers generalising.•Foraging efficiency of D. serra has been relatively higher since the early Holocene.•Different D. serra foraging ranges between two nearby Holocene sites are apparent. Humans began to exploit aquatic resources routinely since the late Middle Pleistocene (c. 162 ka), as evidenced by PP13B Cave in the southern Cape of South Africa. Hard bottom or rocky shore species were among the first to be collected, and although they dominate Pleistocene and Holocene archaeomalacological assemblages locally, the soft-bottom Donax serra clam began to be collected systematically near PP13B by 110 ka. Since then, D. serra appears intermittently in early and Holocene middens. Earlier studies on PP13B D. serra assemblages dated to 110–91 ka concluded that initial collection of D. serra ~ 110 ka was unselective in terms of shell size and was undertaken along the entire tidal gradient. In later visits ~ 91 ka, foragers narrowed their collection to the mid-intertidal where most larger clams live, thus increasing the efficiency of D. serra procurement. In this paper, further metrical data from other Pleistocene and late Holocene assemblages is presented in order to test these conclusions. Pleistocene data from mainly the Klasies River main site also reflect small sizes as in contemporary levels from PP13B; however, it is not possible to generalise that the increase in foraging efficiency observed in PP13B also applies to post 90 ka contexts overall due to a lack of Pleistocene samples dating to this latter period. However, it is clear that foraging efficiency was comparatively higher in the Plettenberg Bay and Mossel Bay areas since the early Holocene. Variability among Holocene D. serra assemblages appear to be explained also in terms of foraging ranges. Settlement patterns, group size and/or its composition may be factors behind the generally intermittent presence of D. serra in the southern Cape middens.