Differential “Tree Attraction”—Epiphytic Growth of Umbilicus rupestris and Other Lithophytic Crassulaceae

Dawson, 1988 [30] and Wardle, 1991 [31] both claimed that, in principle, any terrestrial species can grow epiphytically in nature given (1) successful dispersal and the availability of a (2) suitable microhabitat on a tree. The results of this study suggest that in the case of most European species of Crassulaceae such a situation is extremely rare. This finding comes somewhat as a surprise since the challenges of lithophytic and epiphytic growth are often described as comparable [32,33]: both habitats share a largely impenetrable substrate that inevitably leads to problems of anchorage and procurement of water and nutrients. Both types of challenges may explain why accidental epiphytism of terrestrial species in Europe is typically limited to particular microsites on a tree, e.g., detritus-filled crotches [34]. This should not affect Crassulaceae, however, because lack of humus or establishment on a vertical substrate are usual challenges in their typical growing sites such as rocks or rocky soils. Moreover, many of the species included in this study occur occasionally on human-made structures like roofs or establish on vertical stone walls with little organic material, e.g., [35], which is testimony to successful dispersal to “unusual” growing sites. The only species that seems to stand out in regard to growth on trees among the studied set of Crassulaceae is Umbilicus rupestris. My findings suggest that about 1% of its members grow as epiphytes. Unlike most other accidental epiphytes in Europe, which are typically restricted to special, relatively benign microhabitats on their host tree like crotches with accumulated humus or knotholes [34], a majority of epiphytic Umbilicus plants were found growing directly on bark, in the lower and mid portions of the trunk or even higher up on branches (Table 2). This may reflect the legacy of its typical growth as a lithophyte, which grows on more or less impenetrable substrate. Moreover, the ability of growth at desiccation-prone microsites may also be related to its photosynthetic pathway—the species is one of relatively few species within the European flora that uses crassulacean acid metabolism (CAM), at least facultatively [36]: plants that use CAM can achieve much higher water use efficiencies than C3 plants [37]. However, CAM is also found in many other species in the family like Sedum acre or Sedum album [38], for which I found no observations of epiphytic growth (Table 1). Since it is well-established that variation in water availability is typically much more pronounced in the epiphytic habitat compared to the ground [39], one would expect that epiphytic occurrences of Umbilicus rupestris are mostly found at the wetter end of the species’ geographical distribution, particularly when growing directly on bark. This is clearly not the case: the modelled estimates of MAP of the growing sites of terrestrial and epiphytic individuals are statistically indistinguishable (Table 3); the median of the latter was even substantially lower. However, clear differences were found in regard to temperature. On average, epiphytic individuals grew at relatively warm sites close to the sea, with little or no frost. This is in line with a proposed explanation for a global limitation of epiphytic growth by sub-zero temperatures [2], although there are also many observations that are not in agreement with this notion [40,41]. Hence, the general role of frost as key factor for the distribution of vascular epiphytes remains contentious [9]. Hoeber, Zotz, 2022 [11] emphasized that the study of accidental epiphytes may reveal much about epiphytism in general. Specifically, Lüttge, 2008 [3] conceptualized epiphytism as the opportunistic use of space, and in evolutionary terms this conquest of tree crowns by vascular plants has happened numerous times independently in different families [42]. Thus, identifying what sets Umbilicus rupestris apart from other Crassulaceae may allow us to understand the circumstances associated with the first step in the evolution of a species that may eventually grow primarily or even exclusively epiphytically such as Sedum tortuosum or Echeveria racemosa [1,43,44]. Such an evolutionary trajectory is only possible, however, if a species does not only establish itself in an arboreal habitat, but is also capable of in situ reproduction. This is clearly the case in Umbilicus rupestris as more than 50% of all observations show plants with inflorescences or infructescences. Moreover, this number most likely underestimates the real proportion of reproductive individuals as photographs uploaded to iNaturalist were taken throughout the year, i.e., also show plants in an early phenological phase that may have later developed reproductive organs. However, the major value of a descriptive study like the current one is to set the scene for experimental approaches like comparisons of the germination and establishment success in epiphytic and lithophytic situation or demographic studies of epiphytic and lithophytic populations. Admittedly, the frequencies of epiphytic occurrences documented for Umbilicus rupestris and other Crassulaceae (Table 1) can only be taken as rough estimates since observations were not made systematically. Thus, it is unclear whether iNaturalist contributors were somehow biased when uploading observations. For example, it is conceivable that epiphytic occurrences seemed to be more worthy of reporting than yet another occurrence in the “typical” habitat, but, alternatively, the author knows from own experience that epiphytic occurrences may often be missed. Globally, very few studies have directly quantified the numbers of individuals of a species that grew either epiphytically, lithophytically or terrestrially in a particular study area, e.g., [45,46], or that documented changes in the proportion of epiphytic vs. terrestrial occurrences along an environmental gradient [47]. In the large majority of plant species, available information on habit is only qualitative. Such information can still be used to come up with a rough estimate of habit preference by translating verbal descriptions like “terrestrial, rarely epiphytic” or “both epiphytic and terrestrial” in reasonable percentages, e.g., 95%/5% for the former and 50%/50% for the latter, but even with averages from multiple sources for each species such an approach obviously can only be semi-quantitative [7,8]. To date, this numerical translation is inevitably arbitrary—a calibration with actual data such as the ones reported here is needed to evaluate the realism of the quantitative results of these studies. Irrespective of the question whether the tendency of a given species to grow epiphytically is quantified entirely correctly or not, these recent papers and the present report call attention to the fact that the traditional form of assigning species to the categories “epiphytic”, “lithophytic” or “terrestrial” does not adequately represent the gradual nature of substrate preferences. Although a scheme of distinguishing “accidental (<5% epiphytic individuals)”, “facultative” (5–95% epiphytic individuals), and “true epiphytes” (>95% epiphytic individuals), has been in use for decades [48], there are hardly any data that would allow us to investigate, e.g., how many of the so-called “epiphytic species” collated in Epilist 1.0 [1] are actually only facultatively using trees as substrate. The two studies by [7,8] suggest that these may amount to a substantial number. In summary, this paper quantifies the proportion of epiphytic occurrences among 29 lithophytic species and presents an in-depth analysis for one species over its entire range. It shows that, among European Crassulaceae, the perennial Umbilicus rupestris is the species with the highest incidence of epiphytic occurrences, although overall epiphytic growth is still rather rare. Epiphytic individuals seem to be limited to regions without frost, while the expectation of a similar limitation to the wettest parts of its geographic distribution was not supported by the data. Species like Umbilicus rupestris could be models for the early steps of epiphyte evolution: I suggest to directly study possible physiological and morphological differences of epiphytic vs. terrestrial individuals as well as to study germination success, establishment and general population dynamics of epiphytic vs. terrestrial individuals. The results should be highly instructive and lead to a better understanding of the challenges that terrestrial species face when conquering tree crowns.