The significance of reproduction-related traits for wet meadow species survival in a fragmented landscape

doi:10.21203/rs.3.rs-4734104/v1

Download PDF

Research Article

The significance of reproduction-related traits for wet meadow species survival in a fragmented landscape

https://doi.org/10.21203/rs.3.rs-4734104/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Habitat fragmentation refers to the breaking of continuous habitat into multiple patches; these patches create less connected vegetation areas than before, which can result in smaller plant species populations due to, among other factors, limited pollinator visitation. Specific plant reproductive success traits related to pollination can filter species success in the remaining vegetation, affect their relative abundance and distribution and further shed light on relevant conservation efforts. The European grasslands comprise the most fragmented biome in the world. We explored whether wet meadow plant species are affected by connectivity degree between patches and if so, which traits related directly or indirectly to reproduction and pollination are responsible for their success degree. More particularly, we constructed a trait suite for each species which consisted of 15 interaction, phenotypic and propagation life-history traits mainly focusing on pollination process. Seven traits were revealed as important with flower colour, a categorical variable, flowering duration and rewards accessibility being the strongest predictors. Partial dependencies further revealed that, generally, the most successful species exhibited non-specialised life-history, phenotypic and interaction traits. These results imply that species with specific specialised traits require greater conservation attention. We further stress the importance of many different types of traits and ecological processes needed to be studied at the same time, to better understand what drives species success in not well-connected fragmented habitats or other stressful environments.

habitat fragmentation

connectivity

pollination

floral phenotype

propagation life-history traits

wet meadows

Fragmentation of natural habitats has long been considered one of the greatest threats to biodiversity (e.g. Saunders et al. 1991; Baur and Erhardt 1995). One critical aspect of fragmentation is the size reduction of habitat fragments, leading to less connectivity between them (Hadley and Betts 2012). One of the most prioritised factors for species conservation in habitat fragments is the viability of the remnant populations (Hobbs and Yates 2003) Less connected and smaller populations are expected to face the greatest threat (MacArthur and Wilson 1967; Fischer and Stöcklin 1997; Kearns et al. 1998), eventually leading species to lower pollinator mate availability (Cunningham 2000; Aguilar et al. 2006). This could eventually lead to the collapse of these subpopulations, particularly at fragments highly disconnected from core areas which hold higher assemblages of pollinators.

Fragmentation can have a lot of negative reproductive success consequences on biotically-pollinated plant species. Indeed this was shown plenty of times with results such as, lower species genetic diversity (Ellstrand and Elam 1993; Hadley and Betts 2012; Morgan et al. 2013; Schlaepfer et al. 2018), similarly lack of gene flow ( Young et al. 1996; Buza et al. 2000; Honnay et al. 2007; Cristóbal-Pérez et al. 2021), high levels of pollen limitation (Nayak and Davidar 2010) and less amount of seed or fruit set (Aguilar and Galetto 2004; Cunningham 2000; Jennersten 1988; Young and Pickup 2010). However, some species may benefit from a low degree of connectivity in isolated patches (i.e. fragments) (Steffan-Dewenter and Tscharntke 2002), but in these cases, plants may experience selection for increased spontaneous autogamy (Aguilar et al. 2008) or higher levels of pollinator-mediated self-pollination in small populations, with adverse consequences for parental inbreeding (Johnson et al. 2009).

Semi-natural European grasslands are among the richest plant communities at a small scale (Janecek et al. 2013; Chytrý et al. 2015). Central European grasslands have been cultivated in different ways for centuries and extensively managed for a long time. However, during industrialisation, especially after the Second World War (Baur and Erhardt 1995), a dramatic decrease in extensively managed grasslands began, due to land use intensification. The continuous habitat eventually reached a very high degree of fragmentation, causing temperate grassland communities to exhibit the highest fragmentation rates of all biomes worldwide with approximately 84% of the core area being lost (Jacobson et al. 2019). Wet meadows, which suffer from high land draining due to intensive management, are especially vulnerable, and the vulnerability of wet meadow biota within European grasslands has been previously emphasised (Lepš 1999; Prach 2008; Krause et al. 2011).

Approximately 87.5 to 90% of all flowering plants worldwide are adapted to animal pollination (Ollerton et al. 2011; Tong et al. 2023); thus, they all require pollination vectors to a certain extent. Pollinators’ visitation rates and attractiveness via floral phenotype together, affect pollination success and reproductive success (Barrett 2003) and selection for floral traits is largely mediated by specific types of pollinators (Fenster et al. 2004; Caruso et al. 2019). This, underscores the major importance of pollination and pollinators to plant reproductive success regarding plants’ ability to persist in less connected fragments. However, some plant species showed slow responses to the negative impact of fragmentation (Helm et al. 2006; Krauss et al. 2010) which would raise questions on whether plant species are less threatened than other taxa.

In interaction and pollination systems it is believed that generalist species would be expected to be less affected by fragmentation than specialist species (Steffan-Dewenter et al. 2006; Aguilar and Galetto 2004; Fischer and Lindenmayer 2007; Ramiadantsoa et al. 2018). The same notion on the vulnerability of specialised species exists regarding fragmented habitats (Bond 1994; Kunin 1997; Steffan-Dewenter and Tscharntke 2002; Xiao et al. 2016), because the distances between patches can make it difficult for high-fidelity species to easily find their mutual partners. However, this long-standing idea has been challenged several times (Jonsen and Fahrig 1997; Aizen et al. 2002; Ashworth et al. 2004), raising questions about the importance of species-specific pollinator spectra, floral phenotype, and life-history strategies. In pollination systems, species can interact more or less asymmetrically (Bascompte et al. 2006; Vázquez and Aizen 2004) and therefore, in fragmented habitats, different foraging strategies, i.e. different pollinator functional groups, can offer different services (Librán-Embid et al. 2021) and they can contribute differently to species-specific pollination success (Tommasi et al. 2023).

Even though, angiosperms exhibit a wide variety of floral phenotypic traits in order to promote outcrossing (e.g. Armbruster 2014), in absence of pollinators, plants use certain mechanisms to compensate for pollinator losses and avoid the risk of extinction (Bond 1994; Aguilar et al. 2006). These mechanisms include longer flowering periods and pollination-compensatory mating systems such as self-compatibility, homogamy (the simultaneous ripening of male and female organs) (Cruden and Hermann-Parker 1977), and/or clonality. In post-fragmented areas, the fate of such species in less connected patches can be severe, since compensation via these mechanisms might increase selfing or geitonogamy and thus, inbreeding depression (Charpentier 2001; Barrett 2003). On the other hand, self-incompatibility, as a way to avoid selfing, could result in the collapse of populations in small patches, as well (Charpentier 2001; Barrett 2003; Wiberg et al. 2016). Moreover, focusing on mating systems is important for plant species conservation practices (Kearns et al. 1998).

Overall, different pollinator spectra, floral phenotype, plant breeding systems and other related to reproduction life-history traits are important predictors determining spatial distribution and plant composition when local extinction risks are high, such as in fragmented habitats (Bond 1994; Aizen et al. 2002; Dupré and Ehrlén 2002; Steffan-Dewenter and Tscharntke 2002; Aizen and Feinsinger 2003; Aguilar et al. 2006), and further help take relevant conservation measures (Kolb and Diekmann 2005).

In this study, we created a species-specific pollination trait suite by combining an ensemble of interaction, floral-phenotypic and propagation life-history traits in order to disentangle and predict their importance on the wet meadow species success related to patch connectivity in fragmented central European grasslands. Aiming to investigate whether wet meadow plant species are affected by habitat fragmentation, we ask: Does habitat connectivity affect plant species presence in wet meadows? If so, do more successful species display distinct traits linked to reproductive and pollination success? Finally, do more susceptible species demonstrate more or less specialised traits?

Study area and design

The study was performed in a 385 km² landscape in the Iron Mountains (Železné hory), Bohemian-Moravian Highlands, Czechia, partially protected by the Železné hory Protected Landscape Area (15.68N, 49.92E − 15.93N, 49.70E, Fig. 1). From a biodiversity perspective, remnants of oligotrophic and mesotrophic semi-natural grasslands and wetlands belong to the most valuable habitats in the region. These grasslands used to be extensively managed before the 1950s and either abandoned or intensively managed thereafter (Janečková et al., 2017; Bartoš et al. 2020b). Recently, many of the grasslands have been extensively mowed, and a part of them has been included into protected areas. From a mosaic of 1,307 patches of 11 mapped semi-natural habitat types (Chytrý et al. 2010; Janečková et al. 2017) we extracted 697 patches of wet meadow vegetation, namely 393 patches of the “Calthion” vegetation type, 107 of “Filipendulenion”, 77 of “Molinion”, 79 fens, and 41 transitional mires (Chytrý et al. 2007). Individual wet meadow patches were determined as being separated from others by distance, or due to a distinctly different vegetation composition. The average patch size was 7,197 m². The plant species composition of each grassland patch was recorded (2008–2012; Janečková et al. 2017) and the cover of all species in the total patch area was estimated using the Braun-Blanquet method (see Mueller-Dombois and Ellenberg 1974).

Studied plant species

We pooled data for 39 plant species that are considered to be associated to the wet-meadow habitat (Table S1). Into this list, no bushes or trees were included. Few species for which we did not collect sufficient data of pollination systems were not included in the study. Furthermore, all dichogamous species were protandrous, with exception of Geum rivale, which was protogynous. Except for one species, Ranunculus auricomus, all species were non-apomictic, so we did not consider this type of mating system.

Species success (CRPc) to habitat fragmentation

We calculated the species-specific connectivity for each sampled meadow patch, adapting the similarity index proposed by McGarigal et al. (2002). Within a 1-km buffer zone around each focal patch, Connectivity (C) was computed using the formula:

$$\:{C}_{j}={\sum\:}_{i=1}^{n}\frac{{A}_{i}*{B}_{i}*\left({V}_{i}+1\right)}{{h}_{ij}}$$

where j represents the focal patch, i is the neighbouring patch, n is the number of patches in the buffer zone, A_i is the area of the neighbouring patch, B_i is the Beals smoothing index for the neighbouring patch (representing the focal species presence probability if higher than 30%), V_i is the cover of the focal species in the neighbouring patch, and h_ij is the edge-to-edge distance between the focal and the neighbouring patch.

To evaluate how plant species respond to patch Connectivity, we initially determined the expected occurrence of each species in the focal patch using the modified Beals smoothing index (Beals 1984, Münzbergová and Herben 2004). We then calculated the Species relative presence as Observed minus Expected occurrence. "Observed" was a binary variable, defined by presence/absence of the species, and "Expected" was the Beals index which indicates the probability of occurrence for each focal species.

As a measure of the plant species’ response to fragmentation, we used the slope between the logarithm of Connectivity and the Species relative presence, named Connectivity Relative Presence correlation (CRPc). We used this variable as the species-specific response to degree of fragmentation. Positive relationships indicate that plant species were negatively affected by lack of connectivity between patches(Table S5).

Species traits

We measured or recorded a set of 15 traits associated with reproduction and pollination for each focal plant species (Table 1). These characteristics reflect the pollinator spectrum, floral phenotype, and propagation life-history of a specific plant species. The individual traits were selected to describe how plants reproduce and thus potentially affect the species’ persistence in fragmented habitats. Table 1 provides a brief description of each trait as well as an overview of the methods used to collect, measure, and scale them. For a few particular traits where values were missing, we searched the literature and filled in the missing data accordingly (see information in Supplementary Material).

Pollinator spectrum

The pollinator spectrum was described by functional-group specialisation (sensu Armbruster 2017), i.e. the number of functional groups of pollinators recorded to visit the plant’s flowers, and primary pollinator, i.e. the functional pollinator group with most recorded flower visits (e.g. Ashworth et al 2015). We identified nine functional pollinator groups: bumblebee, honeybee, other bee, hoverfly, other fly, butterfly, moth, beetle, and other hymenopteran (Table S2). We excluded ants as they are not considered to be efficient pollinators in the temperate grasslands (Wills et al. 2018). For details on the sampling method on the pollinators visitation see in Bartoš et al. (2020a).

Floral phenotype and Life History Traits

Floral phenotype is associated with advertisement, rewards, and mechanical fit of the flower to its pollinator, and describes the level of phenotypic specialisation (Armbruster 2017). We directly measured or quantified most of the phenotypic traits in the field, whilst four traits were taken from Klotz et al. (2002) and Klimešová et al. (2017), (see Table 1 for details). Flower colour was determined using spectrometry, with frequencies ranging from 300 nm to 700 nm (in 0.05 nm step). Floral reflectance measurements were performed using the Ocean optics UV-VIS Spectrometer JAZ and the Ocean Optics DT-mini light source (200–1,100 nm; Dunedin, FL, United States), calibrated with a white reflection standard (WS-1-SL, Ocean Optics). Readings were taken through a fibre-optic reflection probe (UV/VIS 400 mm) held at 45° ca. 5 mm from the flower surface. We used a K-means cluster analysis in the factoextra package (Kassambara and Mundt 2017) to create a categorical variable representing the colour trait. Based on the determination of optimal number of clusters and their visualization in cluster plot (Fig S2), we recognized eight colour clusters. The spectrometry data values represent the reflectance intensity along the complete colour spectrum, which can be translated into the degree of brightness (Athira et al. 2019). We missed spectrometry data for two species (Succisa pratensis and Scorzonera humilis), thus we placed them into the existing clusters using the BIOFLOR (Klotz et al. 2002) and FReD (Chittka et al. 1994) databases. Because the detailed information for S. pratensis was not available, we relied on data from a sister species, Scabiosa lucida (Chittka et al. 1994). As propagation life-history traits, we considered three reproductive strategies namely, dichogamy, self-compatibility, and clonality. All three reproductive strategies can function as compensating mechanisms to the possible inefficiency of pollinators. More particularly, dichogamy can promote outcrossing, as another self-incompatibility mechanism (Freeman et al., 1997) but reversely, homogamous plants can easily promote selfing (Brys et al. 2013; Kalisz et al 2012) or geitonogamy (Barrett 2003). Self-compatible species are well-known to be prone to selfing as well, and lastly, extensive clonal spread of species due to stress in lack of pollinators, can create even large monoclonal isolated populations of plants (Wiberg et al. 2016).

Traits association to Connectivity Relative Presence correlation (CRPc)

Firstly, we tested correlations among traits. We used the Spearman correlations among non-categorical traits (Fig. S1), Kruskal-Wallis Chi-squared test between numerical and categorical variables (Table S3), and Pearson’s Chi-squared between categorical variables (no important correlations were found at a significance level 0.05).

To explore and evaluate the association of traits to CRPc, we built a multivariable model with conditional interference trees using the cforest algorithm with the party package (Hothorn et al. 2015), which belongs to Random Forests algorithms (Breiman 2001). We evaluated the model based on the smallest RMSE (Root-Mean-Square error) using the caret package (Kuhn 2008). To explore individual traits importance, we performed a variable selection based on a permutation-process, using the permimp algorithm (Strobl et al. 2007; Strobl et al. 2008; Debeer et al. 2021) named hereafter, Conditional Permutation Importance (CPI). This variable importance measure determines each variable’s relative importance using decision trees that are constructed after shuffling each variable and comparing the response pattern before and after shuffling. Because there were few replicates in our dataset, we had to manually adjust the hyperparameters before building the cforest model. Firstly, we chose a safe high number of trees, ntree = 1000, i.e. 1000 individual decision trees which built the final random forest. We manually tuned the hyperparameters minsplit (an individual decision tree stopped to further split when reaching less than the given number of observations) and minbucket (a new node of an individual decision tree was not performed when there were already less than the given number of observations) (Debeer et al. 2021) to fit our dataset size and found that the optimal values were 4 and 2, respectively.

We then used all possible mtry hyperparameter values (number of variables − 1) to see which ones better fit our model for the highest accuracy by obtaining the smallest RMSE for 50 different seeds. We used this first model and then the permimp algorithm, to evaluate the variables’ importance. For the permimp algorithm we chose the default threshold of 0.95, based on the moderately high correlations of our variables (Fig. S1), and set the out of bag (OOB) argument to TRUE ( i.e. only the Out Of Bag sample was used to predict the result observation) as an alternative to cross-validation. After obtaining the important variables, we built a second model, by rejecting all possible noisy variables, in order to get clearer permutation importance results. To reveal how traits change with species response to fragmentation, we computed partial dependencies (pdp package) based on our conditional random forest models (cforest) (Fig. 3). All statistical analyses were performed using the software R version 4.1.0 (R Development Core Team 2014).

Table 1

Traits details and sources
Plant trait	Trait description	Data source	Details on methods
Pollinator spectrum
(1) functional-group specialisation	number of functional pollinator groups recorded to visit flowers	own data	obtained by 3 24-hr (covering day and night) recordings in the field in different localities (details in Bartoš et al. 2020a)
(2) primary pollinator	the functional pollinator group with the highest number of visits	own data	obtained by 3 24-hr (covering day and night) recordings in the field in different localities (details in Bartoš et al. 2020a)
Floral phenotype
(3) number of flowers	total number of open flowers in an individual plant	own data	> 20 individuals measured in the field
(4) flower size (mm)	size of a flower or a floret (in species holding inflorescence)	own data	up to 15 specimens measured; not less that 5 specimens for rare species
(5) flower height from ground (cm)	height of median open flower from ground	own data	up to 40 specimens measured; not less than 20 for rare species
(6) flower colour	spectrometry data; eight clusters “UV, yellow/bright”, “yellow/bright”, “white, light blue/bright”, “white, yellow/mid-bright”, “yellow, white/non-bright”, “magentas/bright”, “magentas/non-bright”, “dark red/non-bright”, based on the reflectance and brightness	own data	5 flowers per plant species measured. Clusters created by K-means clustering analysis
(7) flower tube length (mm)	length of a flower tube or spur	own data	15 specimens measures, 5 for rare species, by a digital calliper
(8) rewards accessibility	ordinal trait, ranked: ”3”: easy access to both rewards, “2”: “easy access to one reward”, “1”: “difficult access to both rewards”.	own data	decided based on flower architecture
(9) nectar production per plant (mg)	total amount of sugars produced by all open flowers in individual plant per 24 hours	own data	Nectar production per flower (see below) multiplied by an average number of open flowers per plant individual
(10) nectar production per flower(mg)	total amount of sugars produced by a single flower or floret per 24 hours	own data	nectar sampled from individual flowers covered for 24 hours to avoid visitors; sugars quantified by the HPLC analysis (details in Bartoš et al. 2020a)
(11) nectar guides	presence or absence of nectar guides	own data	observation by human eye
(12) flowering duration	number of months during which the plant is flowering	Klotz et al. 2002
Propagation life-history traits
(13) dichogamy	semi-continuous trait of the species’ position on the homogamy-dichogamy continuum	Klotz et al. 2002	six categories standardised to a scale from 0 to 1.
(14) self-compatibility	ordinal trait of the species’ position on the outcrosser-selfer continuum	Bartoš et al. 2020b, Klotz et al. 2002	four categories
(15) clonality	ordinal trait calculated as the sum of ordinal values of multiplication rate and lateral spread	Klimešová et al. 2017	seven categories

Species response to fragmentation

Initially, we found that 22 plant species (approximately 56%) were significantly affected by lack of patch connectivity, in contrast to 17 species (about 44%) which exhibited a non-significant response to it; in our landscape study area, there were no species found to be positively affected by lack of patch connectivity. (Table S4).

Traits association to CRPc

There were few moderately high intercorrelations between the studied reproductive traits (Fig. S1). When testing all traits against the CRPc, under the variable selection Conditional Permutation Importance seven traits were revealed as important (Fig. 2a). The results of the initial first model, including all the predictor traits, revealed three traits as distinctly strong predictors for CRPc (Fig. 2a) namely, flower colour, flowering duration, rewards accessibility, which were consistently the most important predictors in the smaller second model (Fig. 2b). The results from the smaller second model were quite consistent to the first model also for the weaker predictor traits (Fig. 2a, Fig. 2b). More specifically, as they appear in the second model (Fig. 2b), the four weaker traits which significantly predicted CRPc, were nectar guides, nectar production per plant, functional-group specialisation, and dichogamy. The first in a row non-significant trait after CPI, flower size, did not appear as important in neither model, verifying the robustness of the first model.

The three most important traits exhibited Conditional Permutation Importance > 0.001 and a steeper relation to the response CRPc (Table 2, Fig. 3a,b,c). The best predictor trait was flower colour (explained CRPc range ~ 0.11–0.16). More specifically, the group of “UV, yellow/bright” flowers showed the lowest CRPc (least affected by lack of connectivity), whilst “yellow, white/non-bright” and “dark red/non-bright” flowers exhibited the highest CRPc (Fig. 3a). The second most important trait was flowering duration, with species of a 2-month flowering period showing higher CRPc compared to the species with 3 to 5-month flowering duration (Fig. 3b). Rewards accessibility also quite strongly predicted CRPc with species exhibiting easy access to both rewards species (nectar and pollen) and easy access to one reward, showed lower CRPc than species that had both rewards hidden (Fig. 3c).

The four reproductive traits with a gradually weaker association to CRPc (Conditional Permutation Importance < 0.001) were nectar guides, nectar production per plant, functional group specialisation, and dichogamy. (Table 2, Fig. 2). More particularly, based on the cforest models and the partial dependencies, plants without nectar guides showed lower CRPc than plants with nectar guides (Fig. 3d). Nectar production per plant exhibited a weak relationship to CRPc, with species producing more nectar exhibiting higher CRPc values (Fig. 3e). Functional-group specialisation and dichogamy showed a very weak, still significant, effect on CRPc. Plants visited by less than 3 functional pollinator groups, showed higher CRPc values than species being visited by 4 to 8 functional pollinator groups (Fig. 3f). Dichogamy displayed a weak effect with the general trend of homogamous species showing the lowest CRPc values and thus be more successful (Fig. 3g).

Table 2

Conditional Permutation Importance as a variable selection process after the second *cforest* model (visualised in Fig. 2b) in which only the most important predictor traits, (above and close to zero) were selected to better implement the traits importance. The analysis revealed seven important predictors.
Traits selected	Conditional Permutation Importance
flower colour	0.00505658
flowering duration	0.00431075
rewards accessibility	0.00131456
nectar guides	0.00042577
functional-group specialisation	0.0002341
nectar production per plant	0.0001209
dichogamy	0.00012057
flower size	-0.0005320

Connectivity Relative Presence correlation (CRPc) was used as a proxy to represent the species success based on the connectivity degree between patches. More than half of the selected species in the wet meadow suffered from fragmentation due to lack of connectivity between patches (positive CRPc), i.e. were less successful in more isolated patches (Table S4). The rest of the species were not significantly affected. Interestingly, no species displayed negative CRPc, i.e. was affected positively by less connectivity. Our findings are similar to those of Aguilar et al. (2006) which also showed a negative effect of habitat fragmentation on plant reproduction in grassland biome. Similarly, comparison of recent data with historical ones, in Estonian grasslands showed a loss of 29% of species populations, corresponding to an average loss of 21% of the studied species per site (Saar et al. 2012). Another finding showed that in less connected fragments, in open areas of gypsum habitats, the studied species had significantly less viable seeds and a lower seed set (Matesanz et al. 2015). However, fragmentation has not always consistently yielded adverse outcomes; for example, in fragmented Scandinavian agricultural areas, low degree of connectivity did not influence the seed set of Viscaria vulgaris, which was pollinated by bumblebees (Nielsen and Ims 2000). Other instances where plant species exhibited a favourable reaction to fragmentation, primarily involved tree species (e.g. Ballal et al. 1994; Smith-Ramírez et al. 2007). Consequently, considering different types of pollinator groups and vectors could help to shed more light on species-specific extinction risks of particular plants.

The results of Conditional Permutation Importance were quite consistent between the two models (Fig. 2a, Fig. 2b). Floral colour appeared as the most important predictor trait in both importance models (Fig. 2). The role of corolla colour in plants is connected to the effective attraction of suitable pollen vectors (Fenster et al. 2004; Briscoe and Chittka 2001; Chittka et al. 1994). More particularly, the species less affected by lack of connectivity belonged to the “UV, yellow/bright” flower colour group, displaying a “human yellow” or “UV-green” bee colour (Chittka et al. 1994) (Fig. 3a). This group included all four species from the genus Ranunculus sp. (buttercup) in the studied community assemblage, namely: Ranunculus acris, R. auricomus, R. flammula, R. repens, as well as the species Tephroseris crispa, Hypericum maculatum, Crepis paludosa, and Potentilla erecta (Fig. S2). This result emphasizes the importance of UV floral colour reflectance for pollinator attraction (Briscoe and Chittka 2001; Klomberg at al. 2019) which is highly visible to all pollinators (Briscoe and Chittka 2001), particularly to bees (Chittka et al. 1994). Moreover, especially buttercups’ flowers (Ranunculus sp.) hold cells with special structural patterns which act as strong diffuse reflectors and give them an extra glossy display (van der Kooi et al. 2017). This, in combination to UV reflectance, probably contributes to higher visitation rates and successful reproduction. On the contrary, species with “yellow, white/non-bright” and “dark red /non-bright” flower colour represented the least successful species in isolated patches (Fig. 3a). It seems that brightness, the sum of the flowers reflecting frequencies (e.g. Athira et al. 2019), plays an important role in attracting pollinators, and thus, in the success of plant species in the non-well connected fragments.

Flowering duration was the second most important trait in predicting species success in less connected patches (Fig. 2). More specifically, species with a two-month flowering period showed higher CRPc (lower success) compared to species with longer flowering period, from 3 to 5 flowering months, which exhibited similar success rates (Fig. 3b). Similarly, Janečková et al. (2017) found higher relative frequencies of species displaying longer flowering phenology in isolated meadows compared to more connected ones, studying the flora in the same study area. A longer duration of flowering phenology has previously been associated with higher probabilities for pollination success (Olesen et al. 2008) and increased interaction strengths (Encinas-Viso et al. 2012), though not specifically in fragmented habitats. It seems that wet meadow species invest in long temporal niches in order to overlap with larger assemblages of pollinators and assure higher reproductive success and temporal niche aggregation strategies are preferred whereas, specialised short temporal phenologies, which would target specific pollinators, are likely avoided.

The third most important selected trait was rewards accessibility, which was a quite strong predictor (Fig. 2). More particularly, species holding flowers with easy access to both rewards (nectar and pollen), presented considerably lower CRPc than species with easy access to only one reward, whereas species with both rewards hidden, presented even higher CRPc than the above values (Fig. 3c). Species with nectar guides present, and high amounts of nectar per individual plant presented high CRPc values (Fig. 3d, e). This, in combination with the low success of species with hidden rewards, indicates that phenotypically specialised species are more vulnerable than phenotypically generalised species. Contrary to theories on resource importance for species persistence (Kunin 1997; Minckley and Roulston 2006), resources for pollinators, i.e. nectar production per plant, did not drive plant species success in the wet meadows assembly. Possibly for plants to ensure higher reproductive success, easily accessible rewards seems to be preferable, addressing to a large pollinator spectrum, such as hoverflies and other taxa of flies with short proboscises, which were the most frequent visitors in our study (Table S2).

We found marginally weak but still important evidence for the susceptibility of interaction specialists compared to that of generalists in the fragmented habitats (Fig. 2). This is contrasting to previous results of Aizen et al. (2002) and Aguilar et al. (2006), which showed no difference between specialised and generalised species in fragmented habitats (reviewed in Ashworth et al. 2004). Specifically, species visited by less than three pollinator functional groups presented higher CRPc than species visited by four to eight groups which depicted constantly higher success (Fig. 3f). Our results corroborate other studies, in which specialised species were found to be more easily lost in smaller fragments (Aizen et al. 2012; Burkle and Knight 2012), and where, overall, species interactions were found to be more negatively affected in smaller fragments than in larger ones (Sabatino et al. 2010; Aizen et al. 2012; Burkle and Knight 2012). It is worth noting that, although the number of pollinator functional groups predicted species success, primary pollinator identity did not show any particular trend. A possible explanation is that the majority of plant species’ flowers in the local community of wet meadows are accessible to the majority of pollinator functional groups, i.e. they are relatively generalists, and that visitation rates play a more important role than species trait matching. Dichogamy was one of the weakest predictor traits revealing more homogamous species to hold higher CRPc values (Fig. 2, Fig. 3g). Dichogamy is the temporal separation of pollen presentation and stigma receptivity within and between flowers on a plant (Cruden and Hermann-Parker 1977). The presence of homogamous breeding strategies can have adverse consequences on gene flow and result in worse thriving in the next generations.

All the above considered, the wet meadow successful species select for less specialised life-history, interaction, and floral phenotypic trends. This is in accordance with the general notion that generalised species can survive better in fragmented habitats (Steffan-Dewenter et al. 2006; Aguilar and Galetto 2004; Fischer and Lindenmayer 2007). The specialist species susceptibility is mainly attributed to the narrow range of pollinators addressed by plants, which can be even more constrained in small fragments (e.g. Aizen et al. 2002; Bond 1994). Our study represents a large-scale landscape investigation examining plant species-specific relative success in response to fragmentation via reproductive success and pollination-related traits and their particular importance. We suggest that combinations of carefully chosen trait guilds are required to reveal the underlying mechanisms of a species’ ability to persist under risky conditions such as in small, not well-connected patches of suitable vegetation. As it was stressed before, it is unlikely to unravel species susceptibility regarding reproductive declines withoutcombining ecological data along with a suite of particular traits (Aizen et al. 2002).

This study revealed that lack of connectivity between fragmented patches had a negative impact on approximately 56% of Central European wet meadow plant species. Furthermore, the final analysis identified important traits coming from an ensemble of phenotypic, propagation life-history and interaction traits which determined wet meadow species success in the fragmented grasslands based on the between patches connectivity. Considering every significant trait trend, our findings show that specialists are more vulnerable than species with more widespread or generalised traits. More specifically, it turned out that species that require our attention for conservation have specific corolla colours (mostly non-bright), short flowering duration, flowers with hidden rewards, they hold higher nectar amounts and nectar guides, receive visits by few pollinator functional groups and hold a homogamous mating system. The study emphasizes the significance of characteristics that are frequently disregarded, particularly categorical traits, which can show important trends for species' success under specific environmental conditions.

Author Contribution

Conceptualisation: Š.Ja.; P.J, A.S., M.B.,R.T.; methodology: Š.Ja., P.J., A.S., M.B.,R.T; validation: Š.Ja.,P.J, R.T., A.S., formal analysis, Š.Ja., P.J, A.S.; Investigation, Š.Ja., P.J, M.B., R.T., J.J., J.H.,Y.K., E.C., V.M., S.D. and Š. Ji; resources: M.B.,R.T.; writing-original draft preparation, A.S., P.J.; writing-review and editing, A.S., Š.Ja., P.J, M.B., R.T., J.J., J.H.,Y.K., E.C., V.M., S.D. and Š.Ji; visualisation, P.J., A.S.; supervision, P.J., Š.Ja; project administration, M.B.,R.T., A.S.; funding acquisition, M.B., R.T. All authors reviewed the manuscript.

Acknowledgement

We would like to thank to Jan Filip, Pavel Potocký, Jakub Štenc, Mercy Murkwe, Adéla Jurová, Alena Bartošová, Aneta Hospodková, Barbora Zdvihalová, David Jura, Jan Mertens, Jitka Kocková, Klára Kabátová, Lars Götzenberger, Anna Vojtkó, Laura Mlynárová, Pavel Kratochvíl, Pavla Halamová, Zdenka Herová, and Zuzana Sejfová for their help in the field.

Aguilar R, Galetto L (2004) Effects of forest fragmentation on male and female reproductive success in Cestrum parqui (Solanaceae). Oecologia 138(4):513–520. https://doi.org/10.1007/s00442-003-1468-3
Aguilar R, Ashworth L, Galetto L, Aizen MA (2006) Plant reproductive susceptibility to habitat fragmentation: review and synthesis through a meta-analysis. Ecol Lett 9(8):968–980. https://doi.org/10.1111/j.1461-0248.2006.00927.x
Aguilar R, Quesada M, Ashworth L, Herrerias-Diego YV, Lobo J (2008) Genetic consequences of habitat fragmentation in plant populations: susceptible signals in plant traits and methodological approaches. Mol Ecol 17(24):5177–5188. https://doi.org/10.1111/j.1365-294X.2008.03971.x
Aizen MA, Feinsinger P (2003) Bees not to be? Responses of insect pollinator faunas and flower pollination to habitat fragmentation. How landscapes change: human disturbance and ecosystem fragmentation in the Americas. Springer, Berlin Heidelberg, pp 111–129
Aizen MA, Ashworth L, Galetto L (2002) Reproductive success in fragmented habitats: do compatibility systems and pollination specialisation matter? J Veg Sci 13(6):885–892. https://doi.org/10.1111/j.1654-1103.2002.tb02101.x
Aizen MA, Sabatino M, Tylianakis JM (2012) Specialisation and rarity predict nonrandom loss of interactions from mutualist networks. Science 335(6075):1486–1489. https://doi.org/10.1126/science.1215320
Armbruster WS (2014) Floral specialisation and angiosperm diversity: phenotypic divergence, fitness trade-offs and realised pollination accuracy. AoB Plants 6. https://doi.org/10.1093/aobpla/plu003
Armbruster WS (2017) The specialisation continuum in pollination systems: diversity of concepts and implications for ecology, evolution and conservation. Funct Ecol 31(1):88–100. https://doi.org/10.1111/1365-2435.12790
Ashworth L, Aguilar R, Galetto L (2004) Why do pollination generalist and specialist plant species show similar reproductive susceptibility to habitat fragmentation? J Ecol 92(4):717–719. https://doi.org/10.1111/j.0022-0477.2004.00894.x
Ashworth L, Aguilar R, Martén-Rodríguez S, Lopezaraiza-Mikel M, Avila-Sakar G, Rosas-Guerrero V, Quesada M (2015) Pollination syndromes: a global pattern of convergent evolution driven by the most effective pollinator. In: Evolutionary biology: biodiversification from genotype to phenotype (pp. 203–224)
Athira K, Sooraj NP, Jaishanker R, Saroj Kumar V, Sajeev CR, Pillai MS, Dadhwal VK (2019) Quantitative representation of floral colours. Color Res Appl 44(3):426–432. https://doi.org/10.1002/col.22347
Ballal SR, Foré SA, Guttman SI (1994) Apparent gene flow and genetic structure of Acer saccharum subpopulations in forest fragments. Can J Bot 72(9):1311–1315. https://doi.org/10.1139/b94-160
Barrett SC (2003) Mating strategies in flowering plants: the outcrossing–selfing paradigm and beyond. Philos Trans R Soc Lond B Biol Sci 358(1434):991–1004. https://doi.org/10.1098/rstb.2003.1324
Bartoš M, Janeček Š, Janečková P, Chmelová E, Tropek R, Götzenberger L, Jersáková J (2020a) Are reproductive traits related to pollen limitation in plants? A case study from a central European meadow. Plants 9(5):640. https://doi.org/10.3390/plants9050640
Bartoš M, Janeček Š, Janečková P, Padyšáková E, Tropek R, Götzenberger L, Jersáková J (2020b) Self-compatibility and autonomous selfing of plants in meadow communities. Plant Biol 22(1):120–128. https://doi.org/10.1111/plb.12956
Bascompte J, Jordano P, Olesen JM (2006) Asymmetric coevolutionary networks facilitate biodiversity maintenance. Science 312(5772):431–433. https://doi.org/10.1126/science.1123412
Baur B, Erhardt A (1995) Habitat fragmentation and habitat alterations: principal threats to most animal and plant species. GAIA 4(4):221–226
Beals EW (1984) Bray–Curtis Ordination—an effective strategy for analysis of multivariate ecological data. Adv Ecol Res 14:1–55
Bond WJ (1994) Do mutualisms matter? Assessing the impact of pollinator and disperser disruption on plant extinction. Philos Trans R Soc Lond B Biol Sci 344(1307):83–90. https://doi.org/10.1098/rstb.1994.0056
Breiman L (2001) Random forests. Mach Learn 45:5–32. https://doi.org/10.1023/A:1010933404324
Briscoe AD, Chittka L (2001) The evolution of colour vision in insects. Annu Rev Entomol 46(1):471–510. 10.1146/annurev.ento.46.1.471
Brys R, Geens B, Beeckman T, Jacquemyn H (2013) Differences in dichogamy and herkogamy contribute to higher selfing in contrasting environments in the annual Blackstonia perfoliata (Gentianaceae). Ann Bot 111(4):651–661. 10.1093/aob/mct033
Burkle LA, Knight TM (2012) Shifts in pollinator composition and behavior cause slow interaction accumulation with area in plant–pollinator networks. Ecology 93(11):2329–2335. 10.1890/11-2297.1
Buza L, Young A, Thrall P (2000) Genetic erosion, inbreeding and reduced fitness in fragmented populations of the endangered tetraploid pea Swainsona recta. Biol Conserv 93(2):177–186. 10.1016/S0006-3207(99)00132-6
Caruso CM, Eisen KE, Martin RA, Sletvold N (2019) A meta-analysis of the agents of selection on floral traits. Evolution73(1):4–14. 10.1111/evo.13644
Charpentier A (2001) Consequences of clonal growth for plant mating. Evol Ecol 15:521–530. 10.1023/A:1012140120837
Chittka L, Shmida AVI, Troje N, Menzel R (1994) Ultraviolet as a component of flower reflections, and the colour perception of Hymenoptera. Vis Res 34(11):1489–1508. 10.1016/0042-6989(94)90151-1
Chytrý M, Kočí M, Šumberová K, Sádlo J, Krahulec F, Hájková P, Rozehnal J (2007) Vegetace České republiky 1.Travinná a keříčková společenstva. Academia
Chytrý M, Dražil T, Hájek M, Kalníková V, Šibík J, Ujházy K et al (2015) The most species-rich plant communities in the Czech Republic and Slovakia (with new world records). Preslia 87(3):217–278. 10.23855/preslia.2015.217
Chytrý M, Kučera T, Kočí M (eds) (2010) Katalog biotopů české republiky. Second edition. Praha: agentura ochrany přírody a krajiny ČR. 445 s. ISBN 978-80-87457-03-0
Cristóbal-Pérez EJ, Fuchs EJ, Martén-Rodríguez S, Quesada M (2021) Habitat fragmentation negatively affects effective gene flow via pollen, and male and female fitness in the dioecious tree, Spondias purpurea (Anacardiaceae). Biol Conserv 256:109007. 10.1016/j.biocon.2021.109007
Cruden RW, Hermann-Parker SM (1977) Temporal dioecism: an alternative to dioecism. Evolution 31(4):863–866. 10.1111/j.1558-5646.1977.tb01049.x
Cunningham SA (2000) Depressed pollination in habitat fragments causes low fruit set. Proc R Soc Lond B Biol Sci 267(1448):1149–1152. 10.1098/rspb.2000.1135
Debeer D, Strobl C (2020) Conditional permutation importance revisited. BMC Bioinformatics 21(1):1–30. 10.1186/s12859-019-3312-0
Debeer D, Hothorn T, Strobl C, Debeer MD (2021) Package ‘permimp’. 10.18129/B9.bioc.permimp
Dupré C, Ehrlén J (2002) Habitat configuration, species traits and plant distributions. J Ecol 90(5):796–805. 10.1046/j.1365-2745.2002.00714.x
Ellstrand NC, Elam DR (1993) Population genetic consequences of small population size: implications for plant conservation. Annu Rev Ecol Syst 24(1):217–242. 10.1146/annurev.es.24.110193.001245
Encinas-Viso F, Revilla TA, Etienne RS Phenology drives mutualistic network structure and diversity. Ecol Lett 15(3):198–208. doi:, Fenster CB, Armbruster WS, Wilson P, Dudash MR, Thomson JD (2004) Pollination syndromes and floral specialization. Annu Rev Ecol Evol Syst 35:375–403. doi:10.1146/annurev.ecolsys.34.011802.132347, Fischer J, Lindenmayer DB (2012) (2007) Landscape modification and habitat fragmentation: a synthesis. Glob Ecol Biogeogr 16(3):265–280. doi:10.1111/j.1466-8238.2007.00287.x
Fischer M, Stöcklin J (1997) Local extinctions of plants in remnants of extensively used calcareous grasslands 1950–1985: Extinciónes Locales de Plantas en Remanentes de Pastizales Calcáreos de Uso extensivo entre 1950 y 1985. Conserv Biol 11(3):727–737. 10.1046/j.1523-1739.1997.96089.x
Fischer M, Stöcklin J (1997) Local extinctions of plants in remnants of extensively used calcareous grasslands 1950–1985: Extinciónes locales de plantas en remanentes de pastizales calcáreos de uso extensivo entre 1950 y 1985. Conserv Biol 11(3):727–737. https://doi.org/10.1046/j.1523-1739.1997.96074.x
Freeman DC, Doust JL, El-Keblawy A, Miglia KJ, McArthur ED (1997) Sexual specialization and inbreeding avoidance in the evolution of dioecy. Bot Rev 63(1):65–92. 10.1007/BF02857883
Hadley AS, Betts MG (2012) The effects of landscape fragmentation on pollination dynamics: absence of evidence not evidence of absence. Biol Rev 87(3):526–544. 10.1111/j.1469-185X.2011.00207.x
Helm A, Hanski I, Pärtel M (2006) Slow response of plant species richness to habitat loss and fragmentation. Ecol Lett 9(1):72–77. 10.1111/j.1461-0248.2005.00845.x
Hobbs RJ, Yates CJ (2003) Impacts of ecosystem fragmentation on plant populations: generalising the idiosyncratic. Aust JBot 51(5):471–488. 10.1071/BT02110
Honnay O, Adriaens D, Coart E, Jacquemyn H, Roldan-Ruiz I (2007) Genetic diversity within and between remnant populations of the endangered calcareous grassland plant Globularia bisnagarica L. Conserv Genet 8(2):293–303. 10.1007/s10592-006-9181-6
Hothorn T, Hornik K, Strobl C, Zeileis A, Hothorn MT (2015) Package ‘party’. Package Ref Man Party Version 0(9–998):16:37
Jacobson AP, Riggio J, Tait AM, Baillie JEM (2019) Global areas of low human impact (‘Low Impact Areas’) and fragmentation of the natural world. Sci Rep 9(1):1–13. 10.1038/s41598-018-36525-9
Janeček Š, De Bello F, Horník J, Bartoš M, Černý T, Doležal J, … Klimešová J (2013)Effects of land-use changes on plant functional and taxonomic diversity along a productivity gradient in wet meadows. J Veg Sci 24(5):898–909. doi:10.1111/jvs.12041
Janečková P, Janeček Š, Klimešová J, Götzenberger L, Horník J, Lepš J, de Bello F (2017) The plant functional traits that explain species occurrence across fragmented grasslands differ according to patch management, isolation, and wetness. Landsc Ecol 32(4):791–805. 10.1007/s10980-017-0486-y
Jennersten OLA (1988) Pollination in Dianthus deltoides (Caryophyllaceae): effects of habitat fragmentation on visitation and seed set. Conserv Biol 2:359–366. 10.1111/j.1523-1739.1988.tb00198.x
Johnson SD, Torninger E, Ågren J (2009) Relationships between population size and pollen fates in a moth-pollinated orchid. Biol Lett 5:282–285. 10.1098/rsbl.2008.0718
Jonsen ID, Fahrig L (1997) Response of generalist and specialist insect herbivores to landscape spatial structure. Landsc Ecol 12:185–197. 10.1023/A:1007927910532
Kalisz S, Randle A, Chaiffetz D, Faigeles M, Butera A, Beight C (2012) Dichogamy correlates with outcrossing rate and defines the selfing syndrome in the mixed-mating genus Collinsia. Ann Bot 109:571–582. 10.1093/aob/mcr308
Kassambara A, Mundt F (2017) Package ‘factoextra’. Extract and visualize the results of multivariate data analyses, 76(2). 10.18637/jss.v076.i02
Kearns CA, Inouye DW, Waser NM (1998) Endangered mutualisms: the conservation of plant-pollinator interactions. Annu Rev Ecol Syst 29:83–112. 10.1146/annurev.ecolsys.29.1.83
Klimešová J, Danihelka J, Chrtek J, de Bello F, Herben T (2017) CLO-PLA: a database of clonal and bud-bank traits of the Central European flora
Klomberg Y, Dywou Kouede R, Bartoš M, Mertens JE, Tropek R, Fokam EB, Janeček Š (2019) The role of ultraviolet reflectance and pattern in the pollination system of Hypoxis camerooniana (Hypoxidaceae). AoB Plants 11. 10.1093/aobpla/plz057
Klotz S, Kühn I, Durka W (2002) BIOFLOR—A database on biological and ecological traits of vascular plants in Germany. Schriftenreihe für Vegetationskunde 38:1–334
Kolb A, Diekmann M (2005) Effects of life-history traits on responses of plant species to forest fragmentation. Conserv Biol 19:929–938. 10.1111/j.1523-1739.2005.00068.x
Krause B, Culmsee H, Wesche K, Bergmeier E, Leuschner C (2011) Habitat loss of floodplain meadows in north Germany since the 1950s. Biodivers Conserv 20:2347–2364. 10.1007/s10531-011-0096-0
Krauss J, Bommarco R, Guardiola M, Heikkinen RK, Helm A, Kuussaari M, Steffan-Dewenter I (2010) Habitat fragmentation causes immediate and time-delayed biodiversity loss at different trophic levels. Ecol Lett 13:597–605. 10.1111/j.1461-0248.2010.01457.x
Kuhn M (2008) Building Predictive Models in R Using the caret Package. J Stat Softw 28:1–26. 10.18637/jss.v028.i05
Kunin WE (1997) Population size and density effects in pollination: pollinator foraging and plant reproductive success in experimental arrays of Brassica kaber. J Ecol 225–234. 10.2307/2955544
Lepš J (1999) Nutrient status, disturbance and competition: an experimental test of relationships in a wet meadow. J Veg Sci 10:219–230. 10.2307/3237174
Librán-Embid F, Grass I, Emer C, Ganuza C, Tscharntke T (2021) A plant–pollinator metanetwork along a habitat fragmentation gradient. Ecol Lett 24:2700–2712. 10.1111/ele.13885
MacArthur RH, Wilson EO (1967) The theory of island biogeography. Princeton Univ, Princeton
Matesanz S, Gómez-Fernández A, Alcocer I, Escudero A (2015) Fragment size does not matter when you are well connected: effects of fragmentation on fitness of coexisting gypsophiles. Plant Biol 17:1047–1056. 10.1111/plb.12349
McGarigal K, Cushman SA, Neel MC, Ene E (2002) FRAGSTATS v3: Spatial Pattern Analysis Program for Categorical Maps. University of Massachusetts, Amherst. http://www.umass.edu/landeco/research/fragstats/fragstats.html
Minckley RL, Roulston TH (2006) Incidental mutualisms and pollen specialization among bees. In: Plant-pollinator interactions: from specialization to generalization, pp 69–98
Morgan JW, Meyer MJ, Young AG (2013) Severe habitat fragmentation leads to declines in genetic variation, mate availability, and reproductive success in small populations of a once-common Australian grassland daisy. Int J Plant Sci 174:1209–1218. 10.1086/671288
Mueller-Dombois D, Ellenberg H (1974) Aims and methods of vegetation analysis. Wiley, New York
Münzbergová Z, Herben T (2004) Identification of suitable unoccupied habitats in metapopulation studies using co-occurrence of species. Oikos 105:408–414. 10.1111/j.0030-1299.2004.12746.x
Nielsen A, Ims RA (2000) Bumble bee pollination of the sticky catchfly in a fragmented agricultural landscape. Ecoscience 7:157–165. 10.1080/11956860.2000.11682416
Olesen JM, Bascompte J, Elberling H, Jordano P (2008) Temporal dynamics in a pollination network. Ecology 89:1573–1582. 10.1890/07-0576.1
Ollerton J, Winfree R, Tarrant S (2011) How many flowering plants are pollinated by animals? Oikos 120:321–326. 10.1111/j.1600-0706.2010.18834.x
Prach K (2008) Vegetation changes in a wet meadow complex during the past half-century. Folia Geobot 43:119–130. 10.1007/s12224-008-9014-8
R Development Core Team (2014) R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from http://www.R-project.org/
Ramiadantsoa T, Hanski I, Ovaskainen O (2018) Responses of generalist and specialist species to fragmented landscapes. Theor Popul Biol 124:31–40. 10.1016/j.tpb.2018.01.004
Saar L, Takkis K, Pärtel M, Helm A (2012) Which plant traits predict species loss in calcareous grasslands with extinction debt? Divers Distrib 18:808–817. 10.1111/j.1472-4642.2012.00886.x
Sabatino M, Maceira N, Aizen MA (2010) Direct effects of habitat area on interaction diversity in pollination webs. Ecol Appl 20:1491–1497. 10.1890/09-1344.1
Saunders DA, Hobbs RJ, Margules CR (1991) Biological consequences of ecosystem fragmentation: a review. Conserv Biol 5:18–32. 10.1111/j.1523-1739.1991.tb00384.x
Schlaepfer DR, Braschler B, Rusterholz HP, Baur B (2018) Genetic effects of anthropogenic habitat fragmentation on remnant animal and plant populations: A meta-analysis. Ecosphere 9. 10.1002/ecs2.2488
Smith-Ramírez C, Rovere AE, Núñez-Ávila MC, Armesto JJ (2007) Habitat fragmentation and reproductive ecology of Embothrium coccineum, Eucryphia cordifolia and Aextoxicon punctatum in southern temperate rainforests. Biodivers Conserv 102–119. 10.1007/s10531-007-9131-5
Steffan-Dewenter I, Tscharntke T (2002) Insect communities and biotic interactions on fragmented calcareous grasslands—a mini review. Biol Conserv 104:275–284. 10.1016/S0006-3207(01)00191-9
Steffan-Dewenter I, Klein AM, Gaebele V, Alfert T, Tscharntke T (2006) Bee diversity and plant-pollinator interactions in fragmented landscapes. In: Specialization and generalization in plant-pollinator interactions, pp 387–410
Strobl C, Boulesteix AL, Kneib T, Augustin T, Zeileis A (2008) Conditional variable importance for random forests. BMC Bioinformatics 9:1–11. 10.1186/1471-2105-9-307
Strobl C, Boulesteix AL, Zeileis A, Hothorn T (2007) Bias in random forest variable importance measures: Illustrations, sources and a solution. BMC Bioinformatics 8:1–21. 10.1186/1471-2105-8-25
Tong ZY, Wu LY, Feng HH, Zhang M, Armbruster WS, Renner SS, Huang SQ (2023) New calculations indicate that 90% of flowering plant species are animal-pollinated. Nat Sci Rev 10. 10.1093/nsr/nwad219
Tommasi N, Biella P, Maggioni D, Fallati L, Agostinetto G, Labra M, Galimberti A (2023) DNA metabarcoding unveils the effects of habitat fragmentation on pollinator diversity, plant-pollinator interactions, and pollination efficiency in Maldive islands. Mol Ecol 32:6394–6404. 10.1111/mec.16392
van der Kooi CJ, Elzenga JTM, Dijksterhuis J, Stavenga DG (2017) Functional optics of glossy buttercup flowers. J R Soc Interface 14:20160933. 10.1098/rsif.2016.0933
Vázquez DP, Aizen MA (2004) Asymmetric specialization: a pervasive feature of plant–pollinator interactions. Ecology 85:1251–1257. 10.1890/02-0650
Wiberg RAW, Scobie AR, A'Hara SW, Ennos RA, Cottrell JE (2016) The genetic consequences of long term habitat fragmentation on a self-incompatible clonal plant, Linnaea borealis L. Biol Conserv 201:405–413. 10.1016/j.biocon.2016.07.008
Wills BD, Landis DA (2018) The role of ants in north temperate grasslands: a review. Oecologia 186:323–338. 10.1007/s00442-017-3991-4
Xiao Y, Li X, Cao Y, Dong M (2016) The diverse effects of habitat fragmentation on plant–pollinator interactions. Plant Ecol 217:857–868. 10.1007/s11258-016-0601-5
Young AG, Pickup M (2010) Low S-allele numbers limit mate availability, reduce seed set and skew fitness in small populations of a self‐incompatible plant. J Appl Ecol 47:541–548. 10.1111/j.1365-2664.2010.01809.x
Young A, Boyle T, Brown T (1996) The population genetic consequences of habitat fragmentation for plants. Trends Ecol Evol 11:413–418. 10.1016/0169-5347(96)10045-8

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

The significance of reproduction-related traits for wet meadow species survival in a fragmented landscape

Status:

Version 1

Abstract

Figures

INTRODUCTION

METHODS

Study area and design

Studied plant species

Species success (CRPc) to habitat fragmentation

Species traits

Pollinator spectrum

Floral phenotype and Life History Traits

Traits association to Connectivity Relative Presence correlation (CRPc)

RESULTS

Species response to fragmentation

Traits association to CRPc

DISCUSSION

CONCLUSION

Declarations

Author Contribution

Acknowledgement

References

Additional Declarations

Supplementary Files

Status:

Version 1