The influence of activity pattern on primate biology has been a major focus of research in evolutionary anthropology because of its perceived importance for understanding the clade's early evolutionary history and the origin of crown Anthropoidea (Allman, 1977; Cartmill, 1992; Ross, 1996, 2000; Heesy and Ross, 2001, 2004; Kirk and Kay, 2004; Ravosa and Savakova, 2004). Numerous studies have demonstrated correlations between activity pattern and aspects of the visual system across primates (Kay and Cartmill, 1977; Kay and Kirk, 2000; Ross, 2004; Kirk, 2006a, 2006b; Ross and Kirk, 2007; Heesy, 2008, 2009), but the ecological consequences of diurnality and nocturnality are likely to be pervasive (Charles-Dominique, 1975; Clutton-Brock and Harvey, 1977; van Schaik and Kappeler, 1996; Kronfeld-Schor and Dayan, 2003).

In addition to provoking adaptive responses at the population and species levels, activity pattern may play a role in structuring phylogenetic diversity. Studies of trait-dependent diversification indicate that diurnal primates have diversified more rapidly than nocturnal lineages, either because they have higher rates of speciation (Magnuson-Ford and Otto, 2012; Scott, 2018, 2022) or lower rates of extinction (Santini et al., 2015). This relationship has also been detected across mammals (Upham et al., 2019a) and in other vertebrate clades (Anderson and Wiens, 2017), suggesting that the effect is a general feature of tetrapod macroevolution. The reasons for the disparity between diurnal and nocturnal lineages are unknown. In the case of a primarily nocturnal radiation such as Mammalia (Heesy and Hall, 2010; Gerkema et al., 2013; Bennie et al., 2014; Maor et al., 2017; Wu et al., 2017; Cox et al., 2021), one possibility is that lineages that establish themselves in the diurnal environment may encounter ecological opportunity (sensu Yoder et al., 2010) more frequently than nocturnal lineages owing to weaker interspecific competition (Santini et al., 2015; Upham et al., 2019a).

Most of the studies that have investigated the relationship between activity pattern and diversification have focused on differences between diurnal and nocturnal lineages without direct consideration of species that are active both diurnally and nocturnally (cathemerality) or mainly at dawn and dusk (crepuscularity). A notable exception is the analysis conducted by Santini et al. (2015), who found that cathemeral primates have high and nearly equal rates of speciation and extinction, resulting in a low net diversification rate (speciation rate minus extinction rate), a high turnover rate (speciation rate plus extinction rate), and a high extinction fraction (extinction rate divided by speciation rate) in comparison to diurnal and nocturnal lineages. This finding may offer insight into why cathemerality is uncommon among primates: clades composed of lineages with this combination of macroevolutionary rates are especially vulnerable to extinction early in their histories (Gilinsky, 1994; Chevin, 2016). Santini et al. (2015) speculated that the diversification dynamics of cathemeral lineages may reflect a macroevolutionary advantage (higher speciation rates) over nocturnal and diurnal lineages in novel or unstable environments, followed by disadvantage (higher extinction rates) as conditions stabilize and lineages that are more specialized for nocturnality or diurnality become more competitive.

In extant primates, cathemerality is found primarily in the lemurid genera Eulemur, Hapalemur, Lemur, and Prolemur and in a few populations of the otherwise nocturnal platyrrhine genus Aotus (Tattersall, 1987; van Schaik and Kappeler, 1996; Wright, 1999; Curtis and Rasmussen, 2006; Donati and Borgognini-Tarli, 2006; Tattersall, 2006; Donati et al., 2013; Santini et al., 2015). Recent observations suggest that cathemerality may be more common among lemurs than previously thought (Bray et al., 2017; Campera et al., 2019). Because cathemerality occurs most frequently among the lemurs of Madagascar, discussion of its evolutionary significance has focused on the island's unique biogeographic history and distinctive environmental and ecological challenges, especially its extreme seasonality and unpredictable resource availability (van Schaik and Kappeler, 1996; Wright, 1999). Cathemerality has a number of potential adaptive benefits for lemurs: thermoregulatory buffering, predator avoidance, reduced interspecific competition, and greater time devoted to digestive processing (reviewed by Curtis and Rasmussen, 2006; Donati and Borgognini-Tarli, 2006; Griffin et al., 2012). The possibility that cathemerality is not adaptive has also been considered: van Schaik and Kappeler (1996) argued that cathemerality may be an ephemeral stage in the ongoing shift from nocturnality to diurnality in some lineages following the late Holocene extinction of the island's diurnal megafauna (Burney et al., 2004; Crowley, 2010). However, evidence from eye morphology (Kirk, 2006b) and the phylogenetic distribution of activity patterns in lemurs (Griffin et al., 2012; Santini et al., 2015) argue against lemur cathemerality being recent and ephemeral, pointing instead to adaptationist explanations.

The notion that a trait can increase an organism's fitness while having negative macroevolutionary consequences has a long history in evolutionary biology (Cope, 1896; Simpson, 1944; Maynard Smith, 1978; Schwander and Crespi, 2009; Bromham et al., 2016). A widely discussed example from plant biology is the evolution of reproductive self-compatibility and self-fertilization in angiosperms. Self-compatibility can be adaptive when outcrossing opportunities are limited but may ultimately increase a lineage's probability of extinction through inbreeding depression or inability to respond rapidly to changes in selection pressures (Stebbins, 1957; Lande and Schemske, 1985; Takebayashi and Morrell, 2001; Igic and Busch, 2013). Phylogenetic estimates of macroevolutionary rates have provided empirical support for this hypothesis (Goldberg et al., 2010; Freyman and Höhna, 2019; Zenil-Ferguson et al., 2019), showing that self-compatible lineages in some clades exhibit the same volatile combination of rates detected by Santini et al. (2015) in cathemeral primates—i.e., high turnover rates and extinction fractions, and low net diversification rates. Thus, cathemerality, in primates at least, may be an example of conflict between selection at the organismal level and selection at the species level (Lewontin, 1970; Vrba, 1989; Grantham, 1995; Jablonski, 2008).

There is, however, reason to be cautious about the conclusion that cathemerality increases a lineage's extinction risk. Although subsequent analyses have confirmed an association between activity pattern and diversification in primates, greater macroevolutionary volatility in cathemeral primates relative to diurnal and nocturnal lineages has not been corroborated (Scott, 2018, 2022). As Santini et al. (2015) noted, the low frequency of cathemeral primate species poses a challenge for estimating rates of speciation, extinction, and transition into other character states and for disentangling their separate effects (Davis et al., 2013). A particular state may be uncommon in a clade because it influences speciation and extinction probabilities or because transitions out of that state are much more likely to occur than transitions into it (Maddison, 2006; Bromham et al., 2016). The inherent flexibility of cathemerality may make it an unstable, short-lived adaptive strategy, with lineages shifting rapidly to nocturnality or diurnality when the ecological conditions that favor cathemerality no longer prevail (Santini et al., 2015). In other words, cathemerality may be less persistent than nocturnality and diurnality (Maor et al., 2017; Baker and Venditti, 2019). These rate asymmetries are not mutually exclusive: differential diversification and differential state persistence may operate synergistically to prevent cathemerality from becoming more common. Given the small number of cathemeral primate species, the power to distinguish the separate effects of speciation, extinction, and transition rates on the distribution of cathemerality in primates is low. The conclusion that cathemeral lineages are more prone to extinction than diurnal or nocturnal lineages is, therefore, open to question.

The present study examines the hypothesis that cathemerality has negative consequences for lineage diversification by examining the macroevolutionary dynamics of activity pattern in Mammalia. This hypothesis predicts that cathemeral mammals will have a combination of high speciation and extinction rates such that turnover rates and extinction fractions are high and net diversification rates are low. The alternative hypothesis—that cathemerality is simply less persistent than diurnality and nocturnality—predicts that transition rates out of cathemerality will be much higher than the transition rates out of the other two activity patterns. Cathemerality has evolved numerous times in mammals (Curtis and Rasmussen, 2006; Bennie et al., 2014; Maor et al., 2017; Cox et al., 2021), providing the opportunity to test for consistent associations with macroevolutionary rates across several independent origins in clades with divergent adaptations and ecologies. Finding diversification dynamics at this broader phylogenetic scale that match those detected by Santini et al. (2015) in primates would strengthen the case for concluding that cathemerality is uncommon in primates because of its effect on extinction probabilities. On the other hand, failure to detect these dynamics across Mammalia would weaken the case, but it would not be sufficient to confidently reject the hypothesis in primates, given the possibility that it is a unique feature of primate macroevolution. In any case, understanding how activity pattern evolves and influences diversification in mammals has the potential to provide insights into the evolutionary history of primates that are not evident from primate-specific analyses.