Research Article |
Corresponding author: Enrico Simonetti ( enrico.simonetti86@gmail.com ) Academic editor: Alessandro Campanaro
© 2019 Stefano Chelli, Enrico Simonetti, Giandiego Campetella, Alessandro Chiarucci, Marco Cervellini, Federico Maria Tardella, Michela Tomasella, Roberto Canullo.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Chelli S, Simonetti E, Campetella G, Chiarucci A, Cervellini M, Tardella FM, Tomasella M, Canullo R (2019) Plant diversity changes in a nature reserve: a probabilistic sampling method for quantitative assessments. In: Mazzocchi MG, Capotondi L, Freppaz M, Lugliè A, Campanaro A (Eds) Italian Long-Term Ecological Research for understanding ecosystem diversity and functioning. Case studies from aquatic, terrestrial and transitional domains. Nature Conservation 34: 145-161. https://doi.org/10.3897/natureconservation.34.30043
|
Species pool conservation is critical for the stability of ecosystem processes. However, climate and land use changes will likely affect biodiversity, and managers of protected areas are under increasing pressure to monitor native species diversity changes by approaches that are scientifically sound and comparable over time. Here we describe a plant diversity monitoring system in use since 2002 in the “Montagna di Torricchio” Nature Reserve (LTER_EU_IT_033), a Central Apennines representative area of 317 ha, most of which is under strict protection. The aim of this paper was to assess changes in plant species richness over time and to deduce the patterns of species assemblage. The monitoring system was based on a probabilistic sampling design representative of the different physiognomic vegetation types occurring in the Reserve. A total of 34 plots (10×10m) were sampled in 2002, 2003 and 2015, and their species presence/absence and relative coverage were estimated. Repeated measure ANOVA was used to test for plot-level and ecosystem-based changes in species richness along the study period. Temporal nestedness and temporal turnover metrics were used to assess patterns of species’ compositional changes. The results showed significantly different levels of species richness depending on the year, with the lowest value in 2003, probably linked to extreme drought events. Forest systems were comparatively stable, demonstrating the capacity to buffer interannual climate variability. Regarding compositional changes along the entire period (2002–2015), we found random patterns of both temporal nestedness and turnover, indicating stability in species composition. However, we also showed the contemporary occurrence of species loss and species replacement processes, considering the dry year 2003, a finding which should be further explored through fine-scale studies to unravel mechanisms of community assembly under drought. The use of a probabilistic sampling design representative of the different physiognomic vegetation types proved to be advantageous in monitoring the Nature Reserve vegetation and collecting reliable quantitative information. This data, in turn, provides the basis for improvements in management practices and proposed adaptation measures.
compositional changes, monitoring, LTER, species richness, temporal nestedness, temporal turnover
Species pool conservation is critical to the stability of ecosystem processes (
Global change trends are undoubtedly producing effects on species diversity in different ecosystems (e.g.
In addition to climate, dynamic processes deriving from land use changes can produce significant changes in species richness and composition both in grasslands (
In addition to species richness, patterns of species assemblages over time are also particularly significant for understanding ecological processes (
Despite their potential, such metrics have been rarely used in diachronic datasets (but see
Temporal analyses of this kind could be affected by imperfect detection of species, such as when species are overlooked, misidentified, or when assessments contain errors, and this can lead to ‘pseudoturnover’ (
The aim of this paper was to assess changes in vascular plant species richness over time and to deduce the patterns of species assemblage through a monitoring system based on a standard probabilistic sampling design. The “Montagna di Torricchio” LTER site (LTER_EU_IT_033), which includes typical Mediterranean-montane systems, was used as the study area. The study site is part of the Italian “Important Plant Areas” identified by
In particular, we hypothesised that:
(a) species richness at plot level significantly changes over time according to interannual climate variability (including extreme climatic events, e.g. drought) and/or dynamic processes deriving from land-use change;
(b) forest ecosystems buffer climate variability and show a certain stability in terms of species richness over time, compared to grasslands. Finally, we explored the patterns of species compositional changes over time (in terms of both temporal nestedness and temporal turnover) along the entire monitoring period and for each couple of years of relevés. Given the scarcity of approaches exploring patterns of compositional changes in different ecosystems, our findings can be useful to support future hypotheses for monitoring programmes in complex landscapes.
The “Montagna di Torricchio” Nature Reserve is located in the Central Apennines, Italy (+130050E, +425740N, WGS84; Fig.
Based on a multi-scalar sampling protocol adopted to assess plant biodiversity (originally including three grain-levels: 1 m2, 100 m2 and 10,000 m2, see
a Location of the “Montagna di Torricchio” Nature Reserve (LTER_EU_IT_033) in the Central Apennines, Marche Region, Italy (From:
Changes in species richness over time
The changes in the mean number of species in the study period, considering the entire dataset and the two main vegetation types separately (grasslands, including grasslands with shrubs, n = 22; forest habitats, n = 12), were tested using repeated-measures ANOVA. The Bonferroni post-hoc test was used to assess differences in species richness among the three study years.
Temporal nestedness analysis
In order to assess the species nestedness that occurred among the three survey periods (2002, 2003, 2015), we used temporal nestedness analysis (TNA). We compared the observed temporal nestedness (TN) between different survey periods (2002 vs 2003, 2003 vs 2015 and 2002 vs 2015) with the distribution of 999 TN values generated by random reshuffling of the survey period between temporally paired samples. More specifically, TNA is based on a comparison of the observed temporal nestedness between subsequent surveys from the same sample and the nestedness of the same subsequent surveys whose species configuration was randomly reshuffled according to species presence/absence collected in the field. In detail, (a) we calculated the Nestedness measure based on Overlap and Decreasing Fill for sites (NODFsites;
Temporal turnover analysis
To quantify the species temporal turnover (TT) that occurred among the three survey periods (2002, 2003, 2015), we used the Simpson dissimilarity index (βsim) based on pairwise comparison (
βsim = min(b,c)/[a+min(b,c)]
where a is the number of species common to both plots, b is the number of species that occur in the first plot but not in the second and c is the number of species that occur in the second plot but not in the first (
All statistical analyses were performed with R, version 2.14.1. In particular, the following R packages were used: the stats package for repeated measures ANOVA; the betapart package (function beta.pair) for the Simpson dissimilarity index calculation; the vegan package (function nestednodf) for temporal nestedness analysis.
Overall, 345 species were sampled during the monitoring period. The lowest number of species was recorded during 2003 (210 species), while 2015 showed the highest number of species (271 species). At plot level, species richness significantly differed among years (Sum of squares = 2811.78, df = 2, F = 24.69, p < 0.001, Fig.
Mean values and variability of species richness per plot a for all the ecosystems, and b according to the two main ecosystems, i.e. grasslands (blue colour) and forests (green colour). Different letters indicate significant differences in mean values across time following post-hoc test results.
The observed TN value along the entire period (2002–2015) did not differ significantly from the randomly generated assemblages (TN = 11.07, SES = -0.95, p > 0.05, Fig.
Temporal nestedness observed for a 2002 vs 2015 surveys b 2002 vs 2003 surveys and c 2003 vs 2015 surveys. The arrow indicates the position of the observed temporal nestedness value (NODFsites) with respect to the scores of the random loop (999 random nestedness values).
Similarly, the observed TT value along the entire period (2002–2015) did not significantly differ from the randomly generated assemblages (TT = 0.37, SES = -1.56, p > 0.05, Fig.
Our results showed significantly different values of species richness at the plot level, confirming our prediction, with the year 2003 having the lowest species richness value. The three study years showed different climatic conditions: the vegetative season of 2003 had 47% less precipitation than that of 2002, and 57% less than that of 2015. Focusing on spring, the year 2003 was particularly dry, registering 42% and 73% less precipitation than that measured for the spring seasons of 2002 and 2015, respectively (Source: ASSAM, Marche Region). Seasonal distribution of precipitation is known to influence processes triggering plant survival and growth, such as tiller production, root-shoot biomass, root depth, canopy leaf area, stomatal conductance and photosynthesis (
Exploring in detail the pattern of species richness variation across the two main systems of the Nature Reserve (grasslands and forests), our results showed that forest ecosystems maintain a certain stability in terms of species richness at plot level over time, according to our predictions. The structure of ecosystems is known to influence local microclimate which in turn can buffer macroclimate variability. This is the case with forest ecosystems, where the canopy cover regulates temperature, light availability, wind speed and soil moisture, affecting ecological processes of the understory layer (
Along the entire monitoring period (2002–2015), the species’ compositional changes were random, i.e. not related to significant patterns of species impoverishment or species turnover, indicating a certain stability in terms of species composition without tendencies due to the legacy of the previous land use. This result highlights that in the study area, the time interval of 13 years was not wide enough to detect successional effects in terms of compositional changes. However, comparisons including the dry year 2003 showed significant patterns of both species’ impoverishment and turnover, with important implications in terms of community assembly. Our data probably illustrates how interannual environmental variation can generate communities that are nested across time: common species are always present, while rare species disappear in bad years such as the dry 2003, largely as a function of their relative abundances (
Thanks to the long-term approaches, LTER sites can provide unique insights potentially useful for managers of protected areas where there is the urgent need of monitoring species diversity over time. Our approach demonstrated the usefulness of long-term plant diversity monitoring programmes based on probabilistic sampling designs for the study of species richness and patterns of species assemblages over time in a protected area. Although we did not study environmental drivers directly, we showed how interannual climate variability could play a key role in shaping plant species richness and assemblages over time. Forest systems seem to buffer the impact of short-term variations, at least in terms of species richness. Nevertheless, the effects of interannual climate variability on species interactions and community structure are only beginning to be evaluated empirically, and more studies are needed. Accordingly, we suggest several considerations for future studies. First, spatial scale should be taken into account, since changes in species richness across vegetation types varies with scale of observation (
We thank Sandro Ballelli and Domenico Lucarini† for species identification and Chiara Peconi and Fulvio Ventrone for assistance during the field sampling. Sheila Beatty kindly corrected the English usage of the final version of our manuscript. This research was supported by the “Montagna di Torricchio” Nature Reserve.