Along with the spread of open landscapes and radiation of grasses during the Cenozoic (probably best documented for the Miocene), a striking morphological characteristic of dentitions evolved in different herbivore lineages [1
]: hypsodonty, or high-crowned teeth. While the phenomenon apparently started to develop nearly 20 Ma, differences in crown height are also very obvious among extant grazers (hypsodont) and browsers (brachydont = low-crowned) [4
It is generally agreed that the ultimate explanation for hypsodonty is the maintenance of functionality of teeth under conditions of increased wear [7
]. The most accepted cause of increased wear is a rise of dietary silica content as a consequence of a higher proportion of grass in diets and/or foraging in open landscapes, respectively. Silica is harder than tooth enamel, and therefore critical for tooth wear [8
]. There are several plant groups that are known for particularly high silica contents, like liver mosses or horsetails [9
]. However, among angiosperms, grasses are best known to be silica accumulators, while dicots are generally characterized by lower silica contents. Surprisingly little data are available from direct comparisons, but the difference between grasses and browse (trees, shrubs, herbs) can generally be considered substantial: for example, in a study on East African vegetation, silica contents have been quantified to be 4.95 per cent dry matter (DM) in grasses compared with only 0.56–1.46% DM in browse [11
] or in a sample of alpine plants to be 2.66 ± 1.60 (grasses) versus 0.20 ± 0.23% DM (dicots) [12
grasses generally seem to have higher values than C3
Principally, dietary silica can occur as characteristic crystals in plant cell walls (phytoliths), or can be ingested as dust or contaminations of soil [5
]. But while much of the discussion on the causes of hypsodonty focuses on whether phytoliths or grit should be considered the major abrasive agent (e.g. [17
]), it should not be forgotten that even for a scenario disregarding this distinction and simply considering total silica, several inconsistencies and alternative explanations appear to exist: for example, if the rise of grasses is considered as the dominant trigger of hypsodonty, it is surprising that in the prime example of evolution of hypsodonty (Early to Middle Miocene of North America), the major rise of grasses appears to happen much earlier (4 Ma) than the onset of hypsodonty [2
], described as ‘adaptive lag’ by Janis [3
]. Increased tooth wear was also hypothesized to be caused not only by ingestion of abrasive silica, but also by higher general occlusal stress in combination with large quantities of low-quality food [18
] or potentially also higher occlusal stress loads owing to a longer lifespan [19
], the latter hypothesis being both rejected [18
] and supported [20
] later on. In addition, looking at the data of silica contents at the level of individual plant species, it appears that at least some dicots can reach fairly high silica levels [11
], like Cucurbitaceae and Urticales [21
] potentially rendering the ranking of grass and browse concerning their silica contents less unequivocal as often perceived. In fact, silica has been discussed as causing abrasion in dicot diets, too [22
], and among hypsodont notoungulates, microwear indicated a browsing feeding style [24
]. Once evolved, hypsodonty appears not to be decreased irrespective of a later shift to a less abrasive diet [18
], which could imply a less tight connection of grass diets and hypsodonty and a generally high benefit/cost ratio of this dental characteristic.
Several studies have shown hypsodonty to be positively correlated to grass content of diet [5
]. By contrast, it can be stated that while the focus of discussions is already on the distinction of the significance of different silica sources (exogenous dust versus endogenous plant phytoliths) for abrasiveness of herbivore diets, not even the relation of total ingested silica (sum of exogenous and endogenous silica) and hypsodonty has been tested yet in an empirical, quantitative assay.
A potential approach to tackle this data gap makes use of the fact that besides its mechanical resistance, a striking property of silica is its chemical stability and inertness. It is known to pass through the digestive tract without any significant degradation or absorption [26
], characteristics qualifying silica as one of the standard markers in animal digestibility trials. This also opens the door for an estimate of tooth wear constraints faced by individual species owing to ingested silica: faecal silica should reflect ingested silica (as the sum of phytoliths and exogenous silica), integrating both diet (e.g. browse or grass) and habitat choice (e.g. open versus closed), and offering a way to approach the relation of ingested silica and hypsodonty.
Based on a sample of African herbivores, we tested how faecal silica levels reflect the degree of hypsodonty of a species, and to what extent faecal silica levels change between the wet and dry season.