There is an increasing consensus on the need to recognize the phenomenon of so-called ‘microrefugia’ (Rull et al., 1988) or ‘cryptic refugia’ (Cruzan & Templeton, 2000) to explain the biotic colonization of temperate continents after the Last Glacial Maximum (LGM) because otherwise the required colonization rates exclusively from southern refugia would be unrealistically high (Clark et al., 2003). Sommer & Zachos (2009) point out that this type of refugia could only be considered important if present during the LGM. A microrefugium has been defined as ‘a small area with local favourable environmental features, in which small populations can survive outside their main distribution area (the macrorefugium), protected from the unfavourable regional environmental conditions’ (Rull, 2009, pp. 482–483), whereas cryptic refugia are considered to be ‘refugia situated at different latitudes or longitudes than would normally be expected, and often resemble climatic islands in which conditions differ favourably from the surrounding areas’ (Stewart et al., 2010, p. 662). The idea behind these concepts is the same, but both their definitions and their common usage in the literature may create confusion. For example, some authors use either one or the other of these two terms (e.g. Lebarbenchon et al., 2010; Nevill et al., 2010), others use them interchangeably (e.g. Opgenoorth et al., 2010), and a third group combine them in expressions such as, for example, ‘cryptic microrefugia’ (e.g. Fløjgaard et al., 2009). Moreover, Rull (2009) includes the concept of cryptic refugia within that of microrefugia, whereas Stewart et al. (2010) disagree, arguing that small size is not integral to the definition of cryptic refugia. In this paper the concept of microrefugia is defended and advocated, but the main aim is to provide a framework for discussion that will eventually lead to a terminological and conceptual agreement, in order to avoid duplications and misconceptions. It could be argued that terminological discussions of this nature are unnecessary, and that calling refugia either cryptic refugia or microrefugia does not contribute to advancing the understanding of their nature. However, these terms are widespread in the literature and cannot be ignored. Since their recent emergence as a conceptual necessity to account for geographic and genetic patterns of modern populations, all authors in this field use either one or the other (or both) in this context. As they are largely overlapping concepts, a semantic clarification to avoid confusion seems opportune.
In my recent revision of the concept of microrefugia (Rull, 2009), my aim was to provide a tentative definition and classification that would be useful to deal with a phenomenon of increasing significance for the understanding of such ecological matters as, for example, the shaping of a species’ genetic structure or the assembly of extant communities. In this context, the concept of microrefugia included – either as synonyms or as particular cases – other terms widespread in the literature, such as cryptic refugia, northern refugia, isolate microhabitats, relict isolates, sparse stands, humid microsites, intraglacial refugia, etc. I also classified the microrefugia into remote (or distal), widespread (or diffuse) and ecotonal (or proximal), according to their clustering type and geographical patterns with respect to the macrorefugium. I believe that a good strategy for terminological unification is to provide a term and a definition as general as possible and then to try to subdivide it into different categories. If, eventually, some new type does not fit entirely within this frame, then the original term is not general enough and should be revised. This seems to be the case here, as Stewart et al. (2010) believe that the idea of cryptic refugia does not fit the definition of microrefugia. The arguments provided by these authors are that: (1) ‘microrefugia covers a broader range of phenomena, including widespread but low-density populations, and hypothesized large numbers of small, isolated populations, than we consider here’ (p. 662) and therefore ‘would include any area with a small, isolated population, whereas we limit the refugial concept to the contraction phase of a species’ expansion–contraction cycle’ (p. 662); and (2) ‘cryptic refugia will often be smaller than conventional refugia’ but ‘small size is not integral to their definition’ (p. 662).
Concerning the first argument of Stewart et al. (2010), they are right in that my concept of microrefugia includes widespread and low-density populations and other types of small isolated populations, as my intention was to find a term as general as possible, for the reasons stated above. However, the concept is indeed restricted to the species’ contraction phase, either during a glacial or an interglacial, and does not include ‘any area with a small, isolated population’, as Stewart et al. (2010) suggest. I was very careful, for example, to distinguish the concept of macrorefugium(-ia), linked to glacial–interglacial alternation, from that of microrefuge(s), commonly employed for other types of refuges, such as for example shelter microenvironments that some animal populations use seasonally (Rull, 2009). My mistake was not to include this differentiation in the definition, but this is easily solved by including it or by adding the terms ‘glacial’ or ‘interglacial’ before the word.
Regarding size – the second argument – the prefix ‘micro-’ records what is widespread in the literature, as almost all papers use words such as microsites, microenvironments, microclimates, microhabitats, etc., when referring to the concept under discussion (Willis & van Andel, 2004; Bennett & Provan, 2008; Bhagwat & Willis, 2008; Birks & Willis, 2008; Provan & Bennett, 2008; and literature therein). Even Stewart and co-workers admit that small size is a common feature of cryptic refugia in sentences such as ‘the northern refugia would have been in areas of sheltered topography that provided suitable stable microclimates’ (Stewart & Lister, 2001, p. 608) or ‘Cryptic refugia are generally expected to be smaller than the more traditional southern and polar refugia because they are peripherals and are surrounded by unsuitable habitats’ (Stewart et al., 2010, p. 665). As exceptions to this rule, these authors mention extensive high-elevation plateaus and mountain ranges, which would have acted as cryptic refugia for the northern biota during interglacials. If these refugia were fully (or almost) occupied by the species, then Stewart et al. (2010) are right in that this is not a microrefugium; however, if the species is widespread in small populations isolated in favourable microhabitats, then this would be one of my widespread or diffuse microrefugia (Rull, 2009), which is really a microrefugial complex or network. A similar concept would be the so-called ‘mosaic of refugia’, suggested for African eland antelopes during the LGM (Lorenzen et al., 2010).
The term ‘cryptic refugia’ was used by Cruzan & Templeton (2000) to refer to hypothetical refugia deduced from phylogeographic analyses of present populations, which cannot be addressed using traditional palaeoecological methods. For example, the existence of cryptic refugia during the LGM in central and eastern Europe seems necessary to understand the geographical distribution of genetic variability of Fagus sylvatica and other significant forest trees across the continent (Magri, 2008; Petit et al., 2008). These cryptic refugia, however, are ‘too localized or small to be detected by pollen analysis or other conventional palaeoecological techniques’ (Birks & Willis, 2008, p. 148). Therefore, what makes these refuges different from others is their cryptic character but, once properly located, they become simply refugia. For example, a number of formerly widespread cold-adapted Pleistocene taxa (Dryas octopetala, Betula nana, etc.) are presently restricted to refugia on central and south European mountains (Alps, Pyrenees), where their ranges are well known. These areas are not cryptic at all but, according to the criteria of Stewart et al. (2010), they should be considered as cryptic refugia, which is contradictory. The same would be said for past refugial areas, once localized. For example, Willis & van Andel (2004) and Provan & Bennett (2008) propose several specific areas from Europe and North America as potential cryptic refugia, based on palaeoecological and phylogeographical evidence. In other words, the term ‘cryptic’ only refers to our own (transitory) inability of knowing the settlement of a given refugial area and is not an inherent feature of the refugium itself.
Since its first appearance, the expression ‘cryptic refugia’ has been used in different ways and with different meanings. The term was introduced by Cruzan & Templeton (2000), but they did not in fact provide a definition. Then, Stewart & Lister (2001), although they did not provide a formal definition, specified that the cryptic refugia ‘would have been in areas of sheltered topography that provided suitable stable microclimates’. Subsequently, Willis & van Andel (2004) pointed out that cryptic refugia would have been located in ‘small microenvironmentally favourable habitats’ or would have had ‘a more diffuse distribution’ over larger areas. Birks & Willis (2008) follow the original description of Stewart & Lister (2001), whereas Provan & Bennett (2008) consider two complementary types of refugia during the LGM in Europe: ‘small and localized refugia’ in southern mountains, and ‘broader, less well-defined, cryptic refugia of scattered populations’ elsewhere. Stewart et al. (2010) defined cryptic refugia in general terms (see above), emphasizing their unexpected location (considering latitude or longitude) and avoiding any reference to size, although they state that they are isolated and surrounded by areas of unfavourable climate. The term ‘unexpected’ is also subjective and imprecise. In the context under discussion, it probably refers to what would be expected from the initial propositions of the refuge theory, according to which temperate forest trees were restricted to southern refugia. However, as it is stated, it might be misleading. For example, according to this definition, the present occurrence of Dryas octopetala or Betula nana in the southern European mountains would be unexpected but, considering their environmental envelope, it is not.
Summarizing, the concept of cryptic refugia is subjective (i.e. anthropomorphic), perishable and can be misleading. The term microrefugia and its various categories (Rull, 2009) is preferable for those cryptic refugia restricted to particular microhabitats (or clusters of them), which are the majority. Those special cryptic refugia of relatively large size fully occupied by a given species would be called simply refugia or macrorefugia, if this is the case. The temporary and subjective ‘hidden’ or ‘unexpected’ nature of refugia has no biogeographical interest and should not be emphasized. In contrast, size, clustering type and geographical patterns – as included in the concept of microrefugia – are inherent features with significant ecological and evolutionary implications. A Solomonic alternative is to use simply ‘refugia’– either glacial or interglacial – for any situation, thus avoiding terminological controversies. Bennett & Provan (2008) go further and propose that the term ‘refugia’– which places too much focus on geographic location and on the origin of modern populations, thus neglecting longer-term processes – should be replaced by ‘bottleneck’, which describes better what happens in terms of population and genetic dynamics. Further advances in the knowledge of glacial and interglacial expansion–contraction phases will provide a clearer picture concerning the use and classification of the concept of refugia (see also Ashcroft, 2010). In the meantime, the term ‘microrefugia’ is proposed here as a useful biogeographical term and concept.