Trait-based studies have become extremely common in plant ecology. In this work we analysed intraspecific trait variation of Himantoglossum adriaticum, a European endemic orchid species of Community interest, to investigate whether different populations growing on managed and abandoned semi-natural dry grasslands show differences in the CSR strategy. In seven populations occurring in Veneto Region (NE Italy), we measured H. adriaticum maximum vegetative height, leaf traits (LA, LDMC, SLA) and calculated the CSR strategy. Through CCA we investigated the relationship between plant traits and both plant community attributes (cover and height of herbs and shrubs), and geomorphologic features (aspects and slope). PERMANOVA test was used to investigate if the CSR strategy of H. adriaticum varied according to the management regime. Results showed that individuals of H. adriaticum develop different strategies when growing in different habitats. Specifically, individuals growing in managed fully sunny dry grasslands reached higher vegetative height (MH), had lower values of SLA and a higher relative contribution of the C parameter than individuals growing in abandoned dry grasslands, which, on the contrary, were shorter, had higher values of SLA (and correspondingly lower values of LDMC) and a higher relative contribution of the R parameter. Further data on reproductive traits (e.g. fruit and seed-set) may corroborate our results. Although the number of individuals addressed in this study is rather low, and our conclusions may not be considered of general validity for the species, our study demonstrated the applicability of the CSR strategy scheme in detecting functional strategies at intraspecific level.

CSR plant strategy theory, dry grasslands, Himantoglossum adriaticum, intraspecific trait variation, land-use change, plant traits

Trait-based studies have become extremely common in plant ecology (Chelli et al. 2019). This conceptual framework implies describing species through their biological characteristics by measuring suites of traits, i.e. features measurable at the individual level, that are assumed to impact plant fitness and survival (Pierce et al. 2017), either directly or indirectly (Violle et al. 2007).

Although species identity provides major information for ecological studies, the traditional taxonomic description becomes less effective when we are to understand plant reactions to environmental changes, e.g. climate or natural/anthropogenic disturbance, and the processes behind them (Westoby et al. 2002; Garnier et al. 2004). As measurable properties of individuals, functional traits allow understanding the interactions between organisms and their environment (both abiotic and biotic) and bring a functional perspective to the study on biodiversity and how it affects processes at higher levels of organization. Indeed, plant traits have been proved to vary with e.g. environmental conditions (Tautenhahn et al. 2008; Garnier and Navas 2012; Ivanova et al. 2019), land-use (Purschke et al. 2013; Duflot et al. 2014), and species distributions (Violle et al. 2014) suggesting that traits reflect the strategies plants use to cope with the environment and directly influence species interactions and the properties of communities and ecosystems (Grime 2006; Albert et al. 2011; Fantinato and Buffa 2019; Fantinato et al. 2019a, 2019b).

Several studies suggest the existence of sets of plant traits that co-vary, thereby allowing the identification of syndromes that affect survival (e.g. Díaz et al. 2004; Cerabolini et al. 2010; Funk et al. 2016); however, there is no consensus yet on any particular ecological scheme (Pierce et al. 2017). One of the first suggested frameworks is the so-called “leaf economic spectrum” (LES), proposed by Wright et al. (2004); it distinguishes species according to their resource uptake dynamics from quick (acquisitive) to slow return on investment of resources (conservative). Another notable and popular strategy scheme is the CSR (competitor, stress tolerator, ruderal) theory by Grime (1974, 1977), recently reviewed by Grime and Pierce (2012). According to this theory, plant species are assigned to three main strategies which represent trait combinations arising under conditions of competition, abiotic limitation to growth or periodic biomass destruction, respectively (Pierce et al. 2017). These primary strategies represent the extremes, and CSR classification reflects the full spectrum of possible intermediate types. The scheme has been so far adopted at different levels of organization, worldwide (e.g. Yildirim et al. 2012; Astuti et al. 2019; Giupponi and Giorgi 2019; Nyakunga et al. 2018), and values on CSR strategy are currently available for several species (Cerabolini et al. 2010; Pierce et al. 2017). In most trait databases, species are described by mean trait values, regardless of environmental or genetic context (Albert et al. 2010), mostly due to the search for general patterns (Albert et al. 2011). However, plants change their traits not only from species to species (interspecific trait variability) but also among populations of the same species (intraspecific trait variability) when growing in different environment. Recent examples (Lecerf and Chauvet 2008; Messier et al. 2010) showed that intraspecific trait variability plays a crucial role, especially at local scale (Albert et al. 2011). Volis et al. (2004) found that plants of wild barley (Hordeum spontaneum Koch) from favourable environments were better competitors than those from stressful environments. Similarly, Pierce et al. (2013) classified individual plants of the genus Poa, showing within populations trait variation. Intraspecific trait variability thus might help reveal a species’ ability to respond to environmental variations and has been proven to have direct effects on e.g. community assembly (Jung et al. 2010) and nutrient cycles (Lecerf and Chauvet 2008).

In light of the above, this study aims at investigating whether populations of the Adriatic Lizard Orchid (Himantoglossum adriaticum H. Baumann), growing on managed and abandoned semi-natural dry grasslands show differences in the CSR strategy. The evaluation of intraspecific trait variability might be the key to understanding the interactions between H. adriaticum and the extant community and reveal its ability to respond to environmental variations.

We sampled 29 individuals of H. adriaticum. The number of sampled individuals varied from 1 to 15, with an average number of 4.5 ± 5.2 (mean ± SD) individuals per population (Table 1).

The two CCA axes explained 88.81% (p = 0.018) and 11.19% (p = 0.002) of the variance, respectively (Trace = 0.021; P < 0.001). The first CCA axis was strongly positively related to the percentage cover of the herb (C_h; 0.71) and shrub layer (C_s; 0.63), to the average high of the shrub layer (H_s; 0.62) and to the plot slope (SL; 0.53). The largest loadings on the second axis were from plot aspect (AS; -0.27) and slope (SL; 0.25). The CCA revealed a strong positive relationship between SLA, community attributes (i.e., C_h, C_s, H_h and H_s) and plot geomorphological features (i.e., AS and SL; Fig. 1), and a slightly negative relationship between LDMC, community attributes and plot geomorphological features (Fig. 1). The cluster analysis highlighted that sampled plots belonged to two distinct groups; the first corresponding to flat, managed grasslands without shrubs, while the second to steep, abandoned grasslands, highly encroached by shrubs (Table 1).

Figure 1. 

Scatterplot of the canonical correspondence analysis (CCA) utilized to investigate the relationship between plant traits (blue triangles), plant community attributes and plot geomorphological features (red lines). Different symbols indicate individuals of H. adriaticum growing in managed (black dots) or in abandoned grasslands (black stars), as revealed by the cluster analysis.

The overall strategy of H. adriaticum obtained by averaging single individual strategy irrespectively of the habitat type was CR (C:S:R; 60:0:40 %). Significant differences in the CSR strategy of H. adriaticum growing in managed vs. abandoned dry grasslands were revealed by PERMANOVA (F= 28.4; d.f. = 40.71; P < 0.0001). Tukey test revealed that individuals of H. adriaticum growing in managed dry grasslands showed a significantly higher percentage of the C parameter (P < 0.0001; Fig. 2), and a significantly lower percentage of the R parameter (P < 0.0001; Fig. 2) than those growing in abandoned dry grasslands (Table 2).

In other words, our results suggest that individuals of H. adriaticum show different strategies when growing in different habitats (Fig. 3). In our study, individuals growing in managed dry grasslands were more competitive than individuals growing in abandoned dry grasslands, which, on the contrary, were more ruderal.

Figure 2. 

Differences in the relative contribution (%) of C, S and R parameters, among individuals of H. adriaticum growing in managed vs. abandoned grasslands. Different letters indicate significant differences according to Tukey's test.

Figure 3. 

CSR triangular scheme for the individuals of H. adriaticum growing in managed (red dots) and abandoned grasslands (blue dots).

Plant traits are increasingly used to detect plant responses to changes in environmental conditions, since they allow to determine species ecological roles in the environment (Violle et al. 2007) and to detect their ecological strategies (Grime and Pierce 2012; Pierce et al. 2017). Based on the results of this study, leaf traits and Grime’s (1974; 1977) theoretical triangular scheme of competitor, stress tolerator and ruderal plant strategies (CSR theory) proved to be applicable for assessing the functional strategies at intraspecific level.

According to Delforge (2006), H. adriaticum is a species of light or semi-shaded habitats on shallow rocky soils with neutral or basic reaction. It does not show any strong phytocoenological preference. Its primary habitat includes dry grasslands (Festuco-Brometea Br.-Bl. & Tüxen ex Br.-Bl. 1949 class) and open shrubby communities (Rhamno-Prunetea Rivas Goday & Borja ex Tüxen 1962 class) with a mosaic of fully sunny and shaded patches (Bódis et al. 2019), however it can persist in a wide range of habitats including mesic grasslands (Molinio-Arrhenatheretea Tüxen 1937 class), and can also colonise secondary habitats such as roadsides or abandoned vineyards (Fekete et al. 2017).

Overall, H. adriaticum exhibited a Competitive-Ruderal (CR) secondary strategy, incorporating substantial competitive and, to a lesser extent, ruderal features. Interestingly, neither the species nor the two groups of individuals showed the stress-tolerant component. This low inclination for the S- strategy may be explained by the particular phenology of H. adriaticum. Contrary to other dry grassland orchid species, its growing period starts in late spring, and the time of flowering is usually around late May or June. The capsules mature in 4–6 weeks, and seeds are shed rather quickly, in July or August (Bódis et al. 2019). This period corresponds with the rainfall season in the study area, with the average peak of rain in April; at the same time, temperatures, although increasing, do not have reached their peak, which is in July (Fantinato et al. 2017). Thus, H. adriaticum can take advantage of climatic conditions that allow it to complete its life cycle without incurring in stress conditions. Conversely, its characteristic perennial features and persistent rootstock account for the species’ ability to compete and allow a rapid growth and development of a large standing biomass.

Despite these general characteristics, the CCA ordination showed that specific leaf area (SLA) and leaf dry matter content (LDMC) of H. adriaticum varied along with community attributes and plot geomorphological features, reaching, respectively, the highest and the lowest values in partially shaded, sloping grasslands encroached by shrubs. These traits are known to reflect crucial life history trade-offs between the investment of resources in further resource acquisition vs. conservation (Pierce et al. 2017). High values of SLA ensure for high rates of resource acquisition in productive habitats, while high values of LDMC indicate a conservative leaf economics, implying the acquisition of traits responsible for retention of resources in unproductive conditions. Thus, strategies arise as an adaptive response of plants to site ecological condition (Cerabolini et al. 2010), and our results suggest that individuals of H. adriaticum develop different strategies when growing in different habitats. Specifically, individuals growing in managed fully sunny dry grasslands had lower values of SLA and a higher relative contribution of the C parameter than individuals growing in abandoned dry grasslands, which, on the contrary, had higher values of SLA (and correspondingly lower values of LDMC) and a higher relative contribution of the R parameter. The relatively higher ruderal component of individuals of H. adriaticum growing in abandoned dry grasslands, may be explained considering the local geomorphological characteristics. Dry grassland abandonment started from steep slopes, where management practices were more difficult. Steep slopes are generally characterized by high levels of natural disturbance, such as accelerated erosion processes from stormwater runoff and soil loss, with vegetation destruction and fragmentation. Under these unstable environmental conditions, individuals of H. adriaticum may have been selected according to their strategy, with the ruderal component being favoured on the competitive one.

We are aware that the number of individuals addressed in this study is rather low, and that our conclusions may not be considered of general validity for the species. However, as reported by Astuti et al. (2019), in the case of rare and threatened species, the use of small samples can be justified for conservation purposes. Although further research is needed to disentangle to what extent the morphological and ecological variability of the species is due to phenotypic plasticity or genetic variability of the populations, these results are stimulating as they show that plant species present adaptations to changing environmental conditions and extant community attributes that can be detected through the CSR scheme, at a very small spatial scale. Moreover, and consistently to other previous studies (e.g. May et al. 2017; Giupponi and Giorgi 2019), this research confirms that all components (Competitive, Stress tolerant and Ruderal component) of the CSR functional strategy can vary significantly among individuals of the same species. In this regard, a further step towards the comprehension of processes underlying patterns of trait variation should account for regeneration traits of H. adriaticum. Trait-based ecology has so far mainly focused on adult plants despite early stages of a plant life history being very sensitive to environmental factors (Fantinato et al. 2018; Del Vecchio et al. 2020a, 2020b). Specifically, beside differing in leaf traits, the two components (competitive and ruderal) also vary in trade-offs of resource allocation to seed production, with competitors primarily composed of species with relatively low seed production and ruderals allocating resources mainly to seeds. Thus, reproductive traits and precisely fruit and seed-set may corroborate our results.