Introduction
Environmental conditions are ever-changing, temporally and spatially. Trying to survive amid a change in conditions within a local habitat requires species to adapt and develop specialised mechanisms. In many cases, however, the changes are to such an extent that it is better to leave altogether, be it for one or more generations, or altogether. In these cases, many species have adapted to disperse. Dispersal can be defined as the movement of an individual from site of birth to site of reproduction. If better sites can be reached via dispersal, this drives evolution as organisms escape unfavourable conditions (Matthysen, 2012). Dispersal can be divided into three phases: emigration, inter-patch movement and immigration. During emigration, the individual can obtain clues from both biotic and abiotic factors in the local sub-habitat. Using these clues, the individual can then either disperse or not. If they do, they enter the transfer phase or movement through the habitat, where many species require clues from the environment in order to choose a new location to settle in (Clobert et al. , 2001, 2012). The successful settlement into a new sub-habitat is the immigrant phase. Many species disperse passively, and so are unable to make informed choices about the final sub-habitat in which they find themselves.
Dispersal and emigration can reduce the likelihood of competition with kin, and mitigate against drift and inbreeding (Bengtsson, 1978; Wolff, Lundy and Baccus, 1988; Perrin and Mazalov, 1999; Ronce, 2007; Hidalgo, De Casas and Muñoz, 2016). However, environmental variability is arguably the most important driver for dispersal evolution. Many species have offspring that either can or cannot disperse. Experimental research has demonstrated that the ratio of dispersing to non-dispersing offspring changes in response to the environment; For example, Sinervo et al. demonstrated the interaction between maternal environmental conditions, and the resulting offspring dispersal ratio in lizards (Sinervo et al. , 2006). Dispersal plasticity is taxonomically widespread and similar results are seen in animals, insects and plants (Harrison, 1980; Fox and Mousseau, 1998; Dingle, 2006; King and Roff, 2010; Steiner et al. , 2012; Arendt, 2015; Duckworth, Belloni and Anderson, 2015). The ability to alter offspring dispersal ratio is a selective advantage when persisting in highly variable environments (Arendt, 2015).
Within a habitat, there are often several sub-habitats. These sub-habitats can differ in biotic and abiotic factors. Environmental variability is determined by how much and how frequently these factors fluctuate. When environmental conditions within sub-habitats are constant, but the sub-habitats differ in quality, a non-dispersal strategy is optimal. This is because offspring dispersing from the native sub-habitat will encounter lower quality sub-habitats more often than higher quality sub-habitats (Hastings, 1983). However, in bet-hedging scenarios, if the environment fluctuates, with sites of differing quality across time, but statistically the same overall quality, then producing some dispersing offspring is optimal (Harper, 1977; Den Boer, 1981; Venable and Brown, 1993; Baskin and Baskin, 1998; Starrfelt and Kokko, 2012). If the environmental conditions in sub-habitats fluctuate and their quality is overall statistically different, it is possible that dispersal rates are sub-habitat specific (McPeek and Holt, 1992). But when sub-habitats are of more or less equal quality, yet differ in variability, what dispersal strategies will evolve? Will dispersal be adaptive, and if so, will the dispersal rates be habitat specific (Seale and Nakayama, 2019)?
In mountainous habitats, the sub-habitat at the top of the mountain can be considerably different from the sub-habitat at the bottom of the mountain. If neither sub-habitat is of substantially higher quality, and environmental variability is different within each, is sensing and site-specific dispersal adaptive? Here, we use the term “site-specific dispersal” to describe altering the ratio of dispersing to non-dispersing offspring produced by an individual, in response to the environment that they experience during their lifetime. An example of such site-specific dispersal is given by the plant Aethionema arabicum. Ae. arabicum is an annual which grows along the steep stony slopes of the Anatolian Mountains, at a range of 0-3000 meters above sea level (Bhattacharya et al. , 2019). Ae. arabicum exhibits fruit and seed dispersal dimorphism and is able to undergo site-specific dispersal between sub-habitats of similar quality but different variability by sensing its location through temperature (Fig. 1) (Lenser et al. , 2016; Mohammadin et al. , 2017; Arshad et al. , 2019; Seale and Nakayama, 2019).
Mountainous habitats, including the Anatolian Mountains, can be roughly divided into two sub-habitats: high elevation, and low elevation (Velchev, 1984; Mohammadin et al. , 2017). Neither sub-habitat is favourable, neither are optimal. The higher elevation is dry, exposed, and rocky, making the abiotic conditions unfavourable, however, there is little to no competition. At low elevation, the environment is overcrowded, shaded and highly competitive, providing many biotic stresses. However, the constant availability of water supply and nutrients (Atalay, 2006). One sub-habitat is environmentally variable in terms of abiotic stresses but with few competition stresses, while the other is environmentally constant but much more over-crowded, making them of both low quality and differing in variability.