Overgrowth competition or facilitation from cushion plants: Implication for the role of plant-plant interactions

“Your eyes can deceive you, don't trust them.”(Ben Obi-Wan Kenobi, 1977; Star Wars Episode IV: A New Hope) The association of other species with cushion-forming plants has become a classic example of facilitation in arctic and alpine systems (Reid et al., 2010 and reference therein). Cushion plants sometimes create locally suitable conditions for other plants to better establish, grow, or reproduce by ameliorating the thermal environment, improving soil moisture or nutrients, or generally by providing shelter against different local stresses (e.g., wind and light) (Bonanomi et al., 2016; Callaway, 2007; Griggs, 1956). Yet, the general inference of biotic interactions, such as competition or facilitation, from patterns of co-occurring plants has been questioned many times (e.g., see Callaway, 1995; Kershaw, 1973; Leps, 1990; Steinbauer et al., 2016). Growing together may not necessarily reflect a positive effect of the cushion plants (see Griggs, 1956; Michalet, 2006). Alternatively, cushion plants could simply occupy the same safe site or even compete. Our observations may illustrate the behavior of cushion plants with their neighbors that may give us some reasons to re-evaluate the inference of facilitation from co-occurrence with cushion plants (Liancourt & Dolezal, 2021) and its implication for the role of facilitation and competition in “extreme” environments (Liancourt et al., 2017; Michalet et al., 2006; Michalet, Le Bagousse-Pinguet, et al., 2014). During our ascent of the eastern flanks of Korzok Range (Ladakh, NW Himalaya, India; Appendix S1), ~5000 m above sea level (m asl) (Figure 1a), while we were on our way to perform measurements on Caragana versicolor (Fabaceae), a dominant low-stature shrub in the region (Dvorský et al., 2011; Liancourt et al., 2017), we came across individuals of the very common cushion plant Thylacospermum caespitosum (Caryophyllaceae). Although we were not specifically interested in this species for our particular question on that day, individuals were very large at this location, so we stopped and spent some time observing these amazing plants (Figure 1). It seemed evident that C. versicolor was not using T. caespitosum either as a pillow or as a blanket. Instead, T. caespitosum appeared to be overgrowing, to the point that it partially or even entirely engulfed adult C. versicolor (Figure 1b–d). Moreover, it was also overgrowing the other adjacent herbaceous plants (Figure 1e), with few of them managing to persist or survive at the periphery of the cushion. Was this behavior observed for T. caespitosum really facilitation (benefits for species that are overgrown by the cushion plants), in line with the most common interpretation of co-occurrence pattern with cushion plants? Or could this pattern actually reflect an “aggressive” behavior of the cushion species competing with associated species by providing costs for the overgrown species? To test these alternative hypotheses, we sampled twigs of C. versicolor and applied dendrochronological techniques (Doležal et al., 2018) to determine the aboveground growth of shrubs overgrown or free from T. caespitosum (Figure 1f; Appendix S1). Our results demonstrated that T. caespitosum depressed the growth of C. versicolor averaged during the last 10 years (n = 47, F1,20 = 10.67, p < 0.01; Figure 2), after controlling for the positive relationship between age and growth (F1,36 = 21.1, p < 0.001; Figure 2). The twigs of C. versicolor were sampled on individuals overgrown and free from T. caespitosum and did not differ in age (F1,22 = 1.28, p = 0.27) and were on average ~24 years old. The dendrochronological measurements, therefore, revealed that this co-occurrence pattern reflected the competitive behavior, at least aboveground, of T. caespitosum toward C. versicolor, and seems to be a perfect illustration of overgrowth competition (sensu Schoener, 1983). But different components of plant fitness, growth, survival, and reproduction, could provide a different interpretation of the pattern (see Liancourt et al., 2005 and references therein, Schöb et al., 2014). This competition for growth did not lead to rapid competitive exclusion, but it did not seem that being overgrown would extend longevity either (i.e., no evidence for facilitation for survival based on age measurements). As for reproduction, we surmised that being partly or almost entirely covered by T. caespitosum would limit the ability of C. versicolor to reproduce, with a probable negative consequence at the population level. Altogether, these results were consistent with previous observations comparing plant density and species richness within the cushion to the adjacent area without cushions that had shown that the effect of T. caespitosum on associated plant community was competitive in this dry and cold region (de Bello et al., 2011; Dvorský et al., 2013; but see Jiang et al., 2018). It remains, however, uncertain whether T. caespitosum benefited from their behavior toward C. versicolor. If so, C. versicolor (or the adjacent overgrown herbaceous plants) might be the benefactor instead of the beneficiary. It would have established first, and maybe then facilitated the establishment of T. caespitosum, which in turn would compete with C. versicolor, following a sequence of events similar to autogenic succession (Clements, 1916). Facilitation is a broad term that encompasses several types of pairwise interactions such as mutualism (+/+) and commensalism (+/0), but also antagonism (+/−). As such, the association between T. caespitosum and C. versicolor would still qualify as facilitation, and it is not uncommon that facilitative interaction involves a cost for some of the associated species (Callaway, 1995; Losapio et al., 2021; Schöb et al., 2013), as documented for the interaction between T. caespitosum and grasses in Northern China (Michalet et al., 2016). Yet, studies focusing on the effect of the cushion plants on associated species, along stress gradients for instance, may wish to consider the effect as competitive. Our observation have raised several questions that could have far-reaching implications for the role of biotic interactions in harsh environments. In the study of plant–plant association, it is generally assumed that plants establish within cushions and that this association reflects facilitation from the cushion on associated species (e.g., Butterfield et al., 2013; Cavieres et al., 2014). Moreover, the general increase in association with cushion along environmental gradients is taken as support for the prevalence of facilitation in the harsh environment: the stress gradient hypothesis (Bertness & Callaway, 1994). Yet, in some cases such as the one documented here, it can be the opposite, with the association being the consequence of cushion plants overgrowing and competing with their neighbors (see also Griggs, 1956). The behavior of T. caespitosum might be more common than usually thought for cushion species. This would be problematic if the pattern produced by two opposite mechanisms could be similar and if the aggressive behavior of such cushion plants were to be misinterpreted as facilitation. If those competitive cushion plants were to prevail in particularly harsh environments, where plants would tend to be clustered because of the low availability of safe microsites (e.g., very high elevation systems), the association pattern with competitive cushions plants may in fact support the idea that plant–plant interactions switch back to competition in harsh environments (Michalet et al., 2006; Michalet, Le Bagousse-Pinguet, et al., 2014). Cushion plants have intrigued botanists for a long time, they are considered a typical adaptation to temperature-limited systems and are often dominant growth forms in the alpine environment (Aubert et al., 2014). As such, their interactions with other plants are a key driver of diversity (Cavieres et al., 2014; Doležal et al., 2019; Kikvidze et al., 2015; Michalet, Schöb, et al., 2014), and they could play an important role in the context of ongoing climate change in alpine environments (Anthelme et al., 2014; Michalet, Schöb, et al., 2014). Yet, we still know very little about the traits that make a cushion a good or a bad nurse (Anthelme et al., 2017; Michalet et al., 2011; Schöb et al., 2013) and about the circumstances under which their positive effects and more generally plant–plant interactions could wane or even switch back to competition (Michalet, Le Bagousse-Pinguet, et al., 2014; see Liancourt et al., 2017 for a recent example). Our observation calls for more in-depth studies on the complex and dynamic interactions between nurses and beneficiaries, which are essential for a better understanding of the ecological and evolutionary implications of plant–plant interactions (Liancourt et al., 2012; Michalet et al., 2011). We thank Maria Majekova, Michal Gruntman, Udi Segev, Maximiliane Herberich, Stefan Abrahamczyk, Lucien Liancourt, and all the members of the Tielborger Laboratory for their insight, discussion, and feedback on the manuscript and photograph selection. None of this work would have been possible without the contribution of the numerous crew members of the multiple Ladakh expeditions, particularly Thomas Galland and Agnes Albert who accompanied us on this day and helped with field sampling. We are also grateful to Veronika Jandova for the laboratory work. The project was funded by the Czech Science Foundation (project no. 21-26883S). Open Access funding enabled and organized by Projekt DEAL. The authors declare no conflict of interest. Data (Liancourt & Doležal, 2023) are available in Dryad at https://doi.org/10.5061/dryad.qjq2bvqm2. Appendix S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.