The first-order importance of tectonic and environmental controls for terrigenous sediment supply has rarely been questioned, but the role of vegetation in the modification of ancient alluvial signatures has been observed since the mid-20th century (Vogt 1941). Studies of sparsely vegetated rivers (Schumm 1968) and alluvial stratigraphic variation (Cotter 1978; Davies & Gibling 2010) led to observations of (1) plant modulation of alluvial signatures and (2) Palaeozoic facies shifts (PFS): unidirectional changes to facies diversity and frequency, in stratigraphic alliance with the plant fossil record. One PFS is the Siluro-Devonian appearance of mud-rich, architecturally complex alluvium, traditionally ascribed to meandering rivers, and sedimentologically distinct from pre-vegetation strata (Davies & Gibling 2010; Long 2011). Using selected secondary data, Santos et al. (2017) dispute the correlation of these observations using three key points, as follows. (1) The mid-Palaeozoic was typified by orogenic assembly of low-gradient equatorial continents and elevated sea-level, which led to tropical weathering (abundant fine sediment) and extensive alluvial plains. This drove the PFS by promoting river meandering independently of vegetation. (2) Meandering does not require vegetation; this is shown by examples in Precambrian deposits, on other planets, and in ‘non-vegetated’ deserts. Meandering rivers were more abundant than the pre-vegetation rock record suggests, owing to selective bypass and deflation of fine material. (3) Early Siluro-Devonian (meaning Ludlow–Early Devonian) land plants were too small, their biomass and cover too limited, and their wetland habitat too narrow to have stabilized meandering channels, influencing landscape little more than earlier microbial communities. We contest the conclusions and method of the paper, and deal with each point in turn.