Close Home
iForest - Biogeosciences and Forestry
vol. 8, pp. 471-479
Copyright © 2014 by the Italian Society of Silviculture and Forest Ecology
doi: 10.3832/ifor1261-007

Collection: IUFRO 7.01.00 - Ilhéus (Brazil, 2013) & Beijing (China, 2014)
“Forest Response to Climate Change and Air Pollution”
Guest Editors: Paoletti E, Kozovitz A, Feng Z, Bytnerowicz A

Research Articles

Advantages of the point-intercept method for assessing functional diversity in semi-arid areas

Alice Nunes (1-2), Susana Tápia (1), Pedro Pinho (1-3), Otília Correia (1), Cristina Branquinho (1)Corresponding author

Introduction 

Semi-arid ecosystems are characterized by water scarcity and often by low soil productivity. Due to global change, they are highly vulnerable to losses of biodiversity, which underpins many critical ecosystem services ([42]). The more dramatic effects on these ecosystems are often preceded by subtle changes in relative species’ abundance and/or in the dominance of specific functional traits ([7], [43]). Even in less abundant functional groups (e.g., nitrogen-fixers), the loss of critical traits that ensure ecosystem functioning and resilience may have important consequences ([23], [52]). The detection of such changes may be a hint of significant ecosystem transitions.

Species richness has traditionally been used to assess an ecosystem’s response to environmental factors, and has been related to ecosystem multi-functionality ([31]). These relationships are largely modulated by other community attributes, such as species evenness and functional identity and divergence, which often respond more rapidly to environmental constraints than richness, and may have a strong impact on ecosystem processes ([7], [35], [32]). Functional diversity, defined as the value, range, and relative abundance of the functional traits of biological communities in a given ecosystem, was shown to be a better and more universal predictor of ecosystem vulnerability than species diversity, which does not reflect the uneven role played by species in the maintenance of ecosystem processes ([48], [12]). Functional diversity is usually assessed by the use of several metrics (e.g., community-weighted mean and functional richness, evenness, and divergence). Recent investigations have demonstrated the better predictive ability of indexes that consider species abundances rather than richness alone ([44], [35]). However, there is no consensus on the best field method for functional diversity assessment. Biomass ([40]), frequency ([9]), and most commonly cover ([18], [28]) are ordinarily used in the estimation of functional diversity. Nonetheless, cover estimates obtained by different methods may vary considerably ([1]) and may therefore affect estimates of functional diversity.

The desirable characteristics of a cover-sampling method to monitor functional diversity are primarily efficiency, precision, and reproducibility. Efficiency expresses the amount of information collected in relation to the resources devoted to achieve that information. Precision, i.e., the bias between two measurements of the same object by the same observer, is essential to detect changes in the dominance of traits and to ensure measurement repeatability. The reproducibility of a method to be used by different observers largely depends on its objectivity, which is assumed to be higher for methods less vulnerable to the observer bias. The methods most commonly used for sampling plant cover are the area-based, modified-Whittaker’s method (MW), and the point-intercept method (PT), based on transects ([22], [46], [14]). The PT method was originally proposed for grasslands ([22]), and is based on the interception of species at predefined points along a transect. It is thus less biased than area-based methods, which rely on the visual assessment of plant cover ([14]). The line-intercept method is mainly used in patchy shrublands ([10]), but it is not suitable for species (e.g., grasses, some forbs, shrubs) with narrow or lacy canopies, whose extension is hard to delineate when plant density is high ([14]). Dengler’s plot (DE - [11]) is an additional area-based, MW-derived method recently proposed, but without the MW shortcomings of non-uniform plot sizes or shapes, nestedness, and spatial arrangement of smaller subplots. Although considerable attention has been devoted to comparisons of plant-sampling approaches over the last 20 years ([16]), few studies have assessed the performance of different cover-sampling methodologies with respect to functional diversity ([1]). For instance, the use of different methods may lead to a different number of species included in the 80% “dominance” threshold - thus affecting the “amount of trait diversity” analyzed - as proposed by Garnier et al. ([19]) and Pakeman & Quested ([37]) based on the mass-ratio hypothesis ([23]).

With the aim of determining the best cover-sampling method for non-destructive fine-scale monitoring of functional diversity in semi-arid areas, we compared the above-mentioned PT, MW, and DE methods in a vegetation survey carried out in Mediterranean Holm oak (Quercus ilex L.) woodland. Specifically, wet and dry locations were selected along a regional precipitation gradient to determine potential differences in methods’ performance in relation to water scarcity. In the study area water deficit, often combined with low soil productivity, is the major limiting factor for plant establishment, thus lowering vegetation density and/or diversity. The following questions were addressed: (i) Are there differences in the efficiency of these methods in quantifying the relative abundances of species and traits? (ii) Does the efficiency of these methods vary depending on the amount of precipitation and, consequently, with changing vegetation density and/or diversity? (iii) Do the differences lead to different estimates of functional diversity metrics? The above methods were compared in terms of time-efficiency and precision of cover estimates for individual species, and main genera and families as well. Also, we assessed the effect of their use on several functional structure and diversity metrics, namely, “community-weighted mean” and functional richness, evenness, and divergence ([19], [50], [26]). We hypothesized the PT method (less biased and more reproducible) would provide more precise cover estimates than the other above-mentioned methods, allowing a better quantification of functional diversity over time by different observers and across different environmental conditions.

Material and methods 

Study sites

The study was performed in a Mediterranean Holm oak (Quercus ilex L.) woodland (montado) in southwestern Portugal. This semi-natural savanna-like ecosystem has been shaped by human use since long time. Dominant soils in the study area are poor and shallow lithosols, grazing is moderate to low, and agricultural activity has ceased in recent years. The climate ranges from dry sub-humid to semi-arid, with large inter-annual variation. The plant community consists of scattered Holm oak trees and an herbaceous understory dominated by annual grasses and forbs, with shrubland patches dominated by Cistus ladanifer L.

Two sampling sites were selected approximately 100 km apart at the extremes of a regional precipitation gradient (Tab. 1). To ensure within-plot homogeneity, sampling was carried out in highly homogeneous grassland areas with no drainage lines or flooding surfaces, and included herbaceous species and sub-shrubs (chamaephytes).

Tab. 1 - Main characteristics of the sites sampled in this study. Sources: Atlas Digital do Ambiente ([4]), Nicolau ([36]).

Sampling methods

In the spring 2011, an area of approximately 1000 m2 was randomly selected at each sampling location (Tab. 1). The sampling designs depicted in Fig. 1 was superimposed as much as possible over the selected area. Data on the herbaceous community and bare soil cover were collected by two experienced botanists working together, and the survey time recorded, using the following methods (Tab. 2):

Fig. 1 - Sampling design and field overlaying schemes. (MW): modified Whittaker’s plots (light gray); (DE): Dengler’s plot (dark gray); (PT): point-intercept method (black).
Tab. 2 - Description of the sampling design adopted for each of the three survey methods analyzed.
  1. the modified-Whittaker (MW) method ([46]), with consistent rectangular proportions and independent and non-overlapping subplots nested within the largest plot.
  2. The Dengler’s (DE) method ([11]), based on the MW method but with fully nested, square sampling units (each plot nested within the parent larger plot) and replicates of equally-distributed smaller subplots.
  3. The point-intercept (PT) method ([14]), using six 20-m linear transects systematically located (41 points each, spaced every 50 cm). At each point, a rod of 5mm in diameter was stuck in the ground with a 90° angle. All plant species, naked soil, lichens, litter, etc., touching the rod were recorded, though only plant data were considered in the subsequent analysis. The same species was recorded only once at each point. Species and group cover were calculated as the proportion of points intercepted per transect.

Data analysis

Overall richness, number of botanical families, and Shannon’s diversity index ([25]) were calculated for each method at each location (mean of 1-m2 subplots or transects). Pielou’s evenness index was also determined ([38]).

Precision of cover estimates

The mean and precision of the cover estimates obtained with the three methods at the two locations were compared (data from 1-m2 subplots or transects, N≥6) for species cumulatively attaining a relative cover ≥ 80%, as well as for genera cumulatively attaining a relative cover ≥ 50% and for the main botanical families, namely, Asteraceae, Fabaceae, and Poaceae (cumulatively attaining a relative cover ≥ 90%).

Functional traits

To compare functional structure and diversity estimates among methods, we selected eight traits reflecting the strategies used by species to cope with the main environmental constraints, i.e., water and soil-nutrient limitations. Binary, semi-quantitative, and quantitative traits were considered, including growth form, N-fixing ability, dispersal mode, life cycle, flowering onset and duration, vegetative height at maturity, and specific leaf area (leaf area/dry weight - Tab. 3). These traits are related to stress and disturbance avoidance/tolerance, resource acquisition and retention, and reproductive and dispersal abilities ([8]). Traits were assigned to each species (Tab. S1 in Appendix 1) based on either direct observations or literature reports ([17], [6], [39], [5]). Growth form (graminoid or other) and dispersal mode (anemochorous or other) were coded as binary traits to reflect the most relevant characteristics in the grassy plant community (Tab. 3).

Tab. 3 - Description of the functional traits considered in the study. For species trait assignments, see Tab. S1 (Appendix 1). Source: (1) direct observation/measurement; (2) Franco ([17]); (3) Porto et al. ([39]); (3) Castro ([6]); (4) Bernard-Verdier et al. ([5]).

Functional diversity metrics

To describe functional diversity, we used the indexes proposed by Villéger et al. ([50]), who considered species within a community distributed in a multidimensional functional space. This approach has the advantage of taking into account species abundance and considering simultaneously several traits. Using these indexes, functional richness, functional evenness, and functional divergence were computed for the combined eight traits listed in Tab. 3. A detailed description of the computational method is reported by Villéger et al. ([50]). Functional richness is the amount or range of functional multidimensional space occupied by a community and is calculated based on the convex-hull volume method. Functional evenness reflects the regularity of the distribution of abundance in a trait space. A higher functional evenness is expected to correspond to a fuller occupation of a niche space by coexisting species ([33]). Functional divergence quantifies the functional dissimilarity of trait values within a community. For instance, divergence is high when the functional trait values of the most abundant species are far outside the center of the functional trait range ([50]). High functional divergence can be used as an indicator of a high degree of niche differentiation and low competition for resources ([33]). Additionally, the functional dispersion ([26]), which is closely related to Rao’s quadratic entropy, was calculated for all the traits combined. Functional dispersion is defined as the weighted mean distance in multidimensional trait space of individual species from the weighted centroid of all species, using as weight the species’ relative abundance.

We also calculated the functional dispersion and functional evenness for each trait individually, as well as the “community-weighted mean” (CWM), proposed by Garnier et al. ([19]), for the three methods at each location. CWM is defined as the average trait value in a community weighted by the relative abundances of the species carrying each value, and reflects the dominant traits in a community. All calculations were carried out using the “dbFD” function implemented in the FD package ([27]) and running under the R© environment ([41]).

Statistical analysis

Separate tests were performed for each location, since site comparison was out of the scopes of this analysis (see above). Departure from normal distribution of data was tested by Kolmogorov-Smirnov tests. Kruskall-Wallis nonparametric tests were applied to test for differences among methods in individual species cover, cover of the main genera and families, and functional structure and diversity estimates. Multiple comparisons were carried out by pairwise Wilcoxon rank sum tests using Bonferroni’s adjustments at α=0.05. To remove variation due to mean effect size ([29]), Levene’s test for homogeneity of variances was applied on log-transformed species’ cover values to detect significant differences in the precision of estimates. To compare the precision of cover estimates among methods, the coefficient of variation (CV: ratio of standard deviation to mean) was analyzed with likelihood ratio tests ([49]), thereby removing variation due to differences between mean cover values. Student’s t-test was used to compare species diversity indexes. Nonparametric and Levene’s tests were conducted using R© version 3.0.2 ([41]).

Results 

The average time needed for two people to complete the vegetation survey at dry and wet locations was 125 and 170 min, respectively, ranging from 93 to 207 min, with no considerable differences among the MW, DE, and PT methods (data not shown).

The overall richness recorded with the MW method was higher than that determined using the DE method within 1000-m2 plots. By contrast, in 1-m2 plots or 20-m transects similar cumulative richness were obtained by the MW and PT methods, whereas fewer species were detected using the DE method (Tab. 4). At the wet location, the PT method revealed a higher diversity index and a higher evenness than either MW or DE methods (Tab. 4).

Tab. 4 - Overall cumulative richness, number of botanical families, Shannon’s diversity index (mean; N≥6), and evenness, recorded at dry and wet locations, for each sampling method. (MW): modified-Whittaker’s method; (DE): Dengler’s method; (PT): point-intercept method. Values with different letters are significantly different across columns after Bonferroni’s test adjusted at P<0.017 (N≥6).

At the species level, whenever differences were noted, the PT method generally yielded absolute cover estimates that were significantly higher than those obtained with the other two methods at both locations (Tab. 5). The PT method also revealed a lower CV both for more and for less abundant species at the wet location (e.g., Vulpia myurus, Chamaemelum mixtum, Cerastium glomeratum) and for less abundant species at the dry location (Lolium rigidum) The number of species necessary to attain a relative cover of 80% (relative to the sum of all species cover) differed among the three methods, with consistently higher values using the PT method. At the dry location, the 80% threshold was reached with 8, 7 and 3 species using the PT, MW and DE methods, respectively. At the wet location, these differences were even higher: 17, 10, and 7 species for the PT, DE, and MW methods, respectively (Tab. 5).

Tab. 5 - Absolute mean cover (%) and coefficient of variation (CV) recorded for the most common species (attaining ≥80% relative cover) by each method (MW: modified-Whittaker; DE: Dengler; PT: point-intercept) at dry and wet locations. Values with different letters are significantly different across columns after Bonferroni’s test adjusted P<0.017 (N≥6). Lowercase letters refer to cover comparison and uppercase letters to CV comparison. Poaceae species: Agrostis pourretii, Chaetopogon fasciculatus, Gaudinia fragilis, Holcus annuus, Vulpia myuros,Lolium rigidum, Bromus lanceolatus; Asteraceae species: Chamaemelum mixtum, Carlina racemosa, Crepis vesicaria, Tolpis barbata, Leontodon taraxacoides; Fabaceae species: Ornithopus compressus, Trifolium campestre, Trifolium cernuum, Trifolium glomeratum, Trifolium striatum; Caryophyllaceae species: Cerastium glomeratum. (§): Species cumulatively attaining ≥80% of relative cover with each method are indicated.

Whenever there were differences in the cover of the main genera and botanical families, significantly higher estimates and lower CVs were obtained with the PT method than with the other methods, particularly at the dry location (Tab. 6). Cover estimates for Poaceae species were highest using the PT method, at both locations. At the dry location, the PT method yielded cover values for Asteraceae family species and for the genus Vulpia higher than those obtained with the DE and MW methods, respectively, while at the wet location estimates for Fabaceae species with PT were higher than with MW (Tab. 6). The precision of the cover estimates was significantly higher for estimates obtained with PT (lower CV) than with MW for the genera Agrostis and Vulpia at the dry location, and for Asteraceae species at both locations (Tab. 6).

Tab. 6 - Absolute mean cover (%) and coefficient of variation (CV) recorded for the main genera (attaining >50% relative cover) and families (attaining >90% relative cover) for each method (MW: modified-Whittaker; DE: Dengler; PT: point-intercept) at dry and wet locations. Values with different letters are significantly different across columns after Bonferroni’s test adjusted at P<0.017 (N≥6). Lowercase letters refer to cover comparison and uppercase letters to CV comparison.

Of the functional metrics performed for the combined eight traits, only functional evenness differed significantly among the methods used, with higher estimates for the PT method (Fig. 2).

Fig. 2 - Overall functional diversity indices (8 traits) at dry and wet locations. Functional richness, functional evenness, functional divergence ([50]), and functional dispersion ([26]). Methods: modified-Whittaker (MW, diamonds); Dengler (DE, triangles); point-intercept (PT, circles). Different letters indicate significant differences between methods after Bonferroni’s test adjusted P<0.017 (N≥6).

Functional metrics computed individually for the four binary, semi-quantitative, and quantitative traits are shown in Fig. 3. The results for the other four traits are provided in Fig. S1 (Appendix 1). The DE method resulted in lower CWM values for N-fixing ability at both locations and in higher estimates of functional dispersion than those provided by either PT or MW at the dry location (Fig. 3). The CWM of flowering onset was lower using PT than DE at both locations, whereas functional evenness estimates with DE were higher than those obtained with either other methods at the wet location. Height functional evenness was higher with PT than with DE at the wet location, whereas the functional dispersion of specific leaf area differed with each of the three methods at both dry (DE>PT) and wet (DE>MW) locations (Fig. 3).

Fig. 3 - Community weighted mean (CWM), functional evenness, and functional dispersion in dry and wet locations. Traits: N-fixing ability (binary), onset of flowering (semi-quantitative), and height and specific leaf area (SLA - quantitative). Methods: modified-Whittaker (MW, diamonds); Dengler (DE, triangles); point-intercept (PT, circles). Different letters indicate significant differences between methods after Bonferroni’s test adjusted at P<0.017 (N≥6).

Discussion 

In this study, three methods commonly used in the field to quantify species/traits showed remarkable differences of efficiency, leading to divergent estimates of the “community-weighted” functional diversity metrics, regardless of the trait values themselves. This “field sampling-method effect” on functional diversity estimates has relevance in plant functional ecology, especially when the aim of the study implies a fine-scale survey. Indeed, a number of studies relied on plant abundance measured in the field to weight trait importance in the computation of functional metrics, while collecting at least some of the species trait values from bibliographic sources or databases ([15], [20]). Our results showed that the field method chosen in the survey affects community weighted means, functional evenness, and functional divergence. All such parameters are required to achieve a reliable assessment of the functional structure and diversity of plant communities.

Although no differences among the three methods were found in the time needed for the survey, the observed differences in their efficiency in quantifying the species/traits abundance point out important advantages of the PT method in non-destructive fine-scale monitoring of plant functional diversity in semi-arid areas. Firstly, this method allowed the detection of as many species as the MW method and of a higher number than the DE method. It also provided a higher precision in cover estimates and, because of higher community evenness, more species/traits could be analyzed within the advocated 80% “dominance” threshold (species representing 80% of the relative cover of the community). Therefore, the PT method offers a cost-effective way to reduce “sampling error”, by reducing the subjectivity in species cover estimates and improving functional diversity estimates. Secondly, this method is less vulnerable to the operator’s bias and thus likely more reproducible when used by different operators ([14]). It is worth to stress that we did not aim at testing the effect of different observers, e.g., through ring tests to assess the “observer error” ([21]). Instead, we assumed a priori that the reproducibility of the PT method is higher (lower observer-bias) when used by experienced botanists familiar with the local flora.

The PT method has proven to be more efficient across contrasting environments (different precipitation regimes) and showed consistency even across different plant density or diversity. These features are critical for a precise and reproducible assessment of changes in functional diversity of plant community in response to environmental changes. In this study, the PT method outperformed other methods commonly used for fine-scale monitoring of plant functional diversity in the understory of Holm-oak woodlands. Similarly, we hypothesized its superior performances also in vegetation surveys of semi-arid areas characterized by low tree density and/or dominated by grasslands and/or shrubland patches. However, further analyses are needed before extending our results to other ecosystems like boreal or temperate forests.

Species quantification and richness

With the PT and MW methods, a higher number of species was quantified than with the DE method, i.e., PT and MW performed better in species/traits quantification. This is a critical step in functional diversity assessments. Even though our study was not aimed at testing methods to assess species richness per se, we found that using the two area-based methods considered (MW and DE) a higher overall number of species was detected, as a consequence of the inclusion of an extra 1000-m2 plot for the detection of “new” species presence. However, if species richness is an index of interest in the survey, the PT method could be easily complemented with a search for new species in a predefined surrounding area with little extra time.

Precision of cover estimates

The PT method adopted in this study provided plant cover estimates more precise than those obtained with the other methods tested. At the dry location, major differences were found for main genera and families. Cover estimates by the PT method were shown to be particularly sensitive to plant architecture and leaf morphology of understory plants ([1]). However, such method performed similarly or even better than others in monitoring groups of species with distinct morphology (e.g., Asteraceae), ensuring at the same time similar or higher precision in the assessment of cover by species group as well. At the wet location, precision differences concerned mainly individual species’ estimates, either dominant or not, with a consistent advantage of the PT method in the case of varying plant density or diversity. In accordance with our results, other authors using the PT method in herbaceous communities, have reported a precision similar or higher than that obtained by cover visual estimates ([51]). Precision in plant abundance quantification is an essential prerequisite of a reproducible method suitable to fine-scale monitoring of functional diversity of herbaceous communities, where changes in species abundance may be hard to detect (e.g., dense/rich communities).

Evenness and quantification of less abundant species

In the wet location analyzed, community evenness estimates obtained with the PT method were higher than those resulted from the application of the other methods, with small differences in relative cover from more to less abundant species. Likely, such differences were underlying the divergence of functional evenness estimates among the three methods considered. Community evenness plays an important role in ecosystem processes and multifunctionality ([24], [31]). Functional evenness reflects the distribution of traits in a community and thus the degree of niche space occupation by coexisting species ([33]). It often responds more rapidly to environmental changes than species richness and may have a rapid and strong impact on ecosystem functions ([7], [35]). Area-based methods implying visual cover estimates (MW and DE) tend to overestimate the dominant and more conspicuous species and underestimate those less abundant ([51]), with accordingly lower evenness. This was the case at the wet location, where seven species attained 80% relative cover according to the MW method, while seventeen species were needed to achieve the same threshold with the PT method. It is important to take into account the role of less common species in functional diversity assessments. A threshold of 80% (in terms of species number or their cover) has been advocated as representative in the description of functional composition ([19], [37]). Such threshold is based on the assumption that dominant species are functionally more important because of their higher abundance, following the mass-ratio hypothesis ([23]). Thus, since the adoption of the PT method included more species (and more functional traits) within the above threshold, it follows that PT performed better for purposes of assessing functional diversity. Moreover, the role of less common species in drought-prone ecosystems may be critical, both over the short- and the long-term. In dry-lands, inter-annual climatic fluctuations are high, forcing species to cope with extreme values of the environmental factors. Consequently, they often exhibit a dynamic turnover involving shifts in the abundance of response groups when a rainy year is followed by a severely dry one, especially in communities dominated by annual species ([3], [2], [13]). Therefore, in contrast to more mesic areas, less abundant species in dry-lands are likely to play a major role in the ecosystem resilience, as a consequence of their capability of exploiting outstanding environmental conditions, as proposed by the complementarity hypothesis ([23], [52], [30]). This argument supports the relative importance of less abundant species and thus their inclusion in functional diversity assessments. In a study on grasslands by McIntyre & Lavorel ([34]), the range of different traits exhibited by forbs and smaller grasses contributed to a varying environmental response, in contrast to local dominant grasses. In our study, this seemed to be the case of N-fixing species. Despite the fact that their relative cover did not exceed 3.0 % and 17.6 % at the dry and wet locations, respectively, they represented a highly relevant functional feature of drylands, usually characterized by soil N shortage ([45]). Furthermore, dry-lands are highly susceptible to land degradation and desertification ([42]). These processes most likely depend on a critical threshold beyond which drastic alterations occur, preceded by more subtle functional changes in communities ([42], [43]). It is therefore important to monitor such changes, as they represent early-warning indicators and allow a timely adoption of counteracting prevention activities.

Cover estimates

Absolute cover estimates were generally higher with the PT method than with either other methods, as previously reported in the literature ([1]). This effect has been attributed to the error due to the diameter of the rod used in sampling, which should be as thin as possible ([14]). However, this is not a problem when the aim is to monitor changes through time and it should not affect the computation of functional metrics using the relative cover of species/traits.

Conclusions 

In this study, we demonstrated that the PT, MW, and DE methods differed in terms of efficiency in the assessment of species/traits relative abundances, thus affecting estimates of functional diversity. The PT method had important advantages over the others with respect to fine-scale monitoring of plant functional diversity in the mainly grassy understory of Holm oak woodlands, and likely in semi-arid areas in general as well. The adoption of the PT method allowed the detection of as many species as the MW method and of more species than the DE method, with a higher precision of cover estimates both for groups of species and at the single-species level. Moreover, due to higher community evenness, it allowed the analysis of a greater number of species/traits within the advocated 80% “dominance” threshold. Precision of the estimates is a prerequisite of functional diversity surveys, in that they must include not only dominant species and traits, because of their larger contribution to ecosystem functionality, but also less common ones, given their decisive role in the resilience and function of semi-arid ecosystems, thereby integrating the mass ratio and complementarity hypotheses ([30]). The advantages of the PT method were proven across contrasting environmental conditions. In addition, it should be less biased than those based on visual estimation and thus more reproducible when used by different operators.

Acknowledgements 

Authors’ contributions: AN, ST, OC, and CB conceived of and designed the experiments, AN and ST performed the experiments, AN and CB analyzed the data, AN, PP, OC, and CB wrote the manuscript. This research was funded through the FCT-MEC DesertWarning project (PTDC/AAC-CLI/10 4913/2008) and through SFRH/BPD/754 25/2010 and SFRH/BD/51407/2011 fellowships.

References

(1)
Abrahamson IL, Nelson CR, Affleck DL (2011). Assessing the performance of sampling designs for measuring the abundance of understory plants. Ecological Applications 21 (2): 452-464.
::CrossRef::Google Scholar::
(2)
Adler PB, Levine JM (2007). Contrasting relationships between precipitation and species richness in space and time. Oikos 116: 221-232.
::CrossRef::Google Scholar::
(3)
Aronson J, Shmida A (1992). Plant species diversity along a Mediterranean-desert gradient and its correlation with interannual rainfall fluctuations. Journal of Arid Environments 23: 235-247.
::Online::Google Scholar::
(4)
Atlas Digital do Ambiente (2011). Climate Atlas of Portugal. Agência Portuguesa do Ambiente (APA), Ministerio do Ambiente, Ordenamento do Territorio e Energia, Governo de Portugal, web site. [in Portuguese]
::Online::Google Scholar::
(5)
Bernard-Verdier M, Navas ML, Vellend M, Violle C, Fayolle A, Garnier E (2012). Community assembly along a soil depth gradient: contrasting patterns of plant trait convergence and divergence in a Mediterranean rangeland. Journal of Ecology 100: 1422-1433.
::CrossRef::Google Scholar::
(6)
Castro H (2008). Effects of land use change on plant composition and ecosystem functioning in an extensive agro-pastoral system: plant functional traits and ecosystem processes. PhD thesis, University of Coimbra, Portugal, pp. 144.
::Google Scholar::
(7)
Chapin FS, Zavaleta ES, Eviner VT, Naylor RL, Vitousek PM, Reynolds HL, Hooper DU, Lavorel S, Sala OE, Hobbie SE, Mack MC, Díaz S (2000). Consequences of changing biodiversity. Nature 405: 234-242.
::CrossRef::Google Scholar::
(8)
Cornelissen JHC, Lavorel S, Garnier E, Díaz S, Buchmann N, Gurvich DE, Reich PB, ter Steege H, Morgan HD, van der Heijden MGA, Pausas JG, Poorter H (2003). Handbook of protocols for standardized and easy measurement of plant functional traits worldwide. Australian Journal of Botany 51: 335-380.
::CrossRef::Google Scholar::
(9)
De Bello F, Leps J, Sebastia MT (2005). Predictive value of plant traits to grazing along a climatic gradient in the Mediterranean. Journal of Applied Ecology 42: 824-833.
::CrossRef::Google Scholar::
(10)
De las Heras MM, Saco PM, Willgoose GR, Tongway DJ (2011). Assessing landscape structure and pattern fragmentation in semiarid ecosystems using patch-size distributions. Ecological Applications 21: 2793-2805.
::CrossRef::Google Scholar::
(11)
Dengler J (2009). A flexible multi-scale approach for standardized recording of plant species richness patterns. Ecological Indicators 9: 1169-1178.
::CrossRef::Google Scholar::
(12)
Díaz S, Lavorel S, de Bello F, Quétier F, Grigulis K, Robson TM (2007). Incorporating plant functional diversity effects in ecosystem service assessments. Proceedings of the National Academy of Sciences USA 104: 20684-20689.
::CrossRef::Google Scholar::
(13)
Elmendorf SC, Harrison SP (2009). Temporal variability and nestedness in California grassland species composition. Ecology 90 (6): 1492-1497.
::CrossRef::Google Scholar::
(14)
Elzinga CL, Salzer DW, Willoughby JW (2001). Measuring and monitoring plant populations. Bureau of Land Management National Business Center, Denver, CO, USA, pp. 477.
::Google Scholar::
(15)
Fischer LK, von der Lippe M, Kowarik I (2013). Urban grassland restoration: which plant traits make desired species successful colonizers? Applied Vegetation Science 16: 272-285.
::CrossRef::Google Scholar::
(16)
Floyd DA, Anderson JE (1987). A comparison of three methods for estimating plant cover. Journal of Ecology 75: 221-228.
::CrossRef::Google Scholar::
(17)
Franco JA (1971). Nova flora de Portugal (continente e Açores) [New flora of Portugal (mainland and Açores)]. Sociedade Astória Lda, Lisbon, Portugal, vol. 1-2, pp. 555+659.
::Google Scholar::
(18)
Frenette-Dussault C, Shipley B, Léger J, Meziane D, Hingrat Y (2012). Functional structure of an arid steppe plant community reveals similarities with Grime’s C-S-R theory. Journal of Vegetation Science 23 (2): 208-222.
::CrossRef::Google Scholar::
(19)
Garnier E, Cortez J, Billès G, Navas ML, Roumet C, Debussche M, Laurent G, Blanchard A, Aubry D, Bellmann A, Neill C, Toussaint JP (2004). Plant functional markers capture ecosystem properties during secondary succession. Ecology 85: 2630-2637.
::CrossRef::Google Scholar::
(20)
Gerhold P, Price JN, Pussa K, Kalamees R, Aher K, Kaasik A, Partel M (2013). Functional and phylogenetic community assembly linked to changes in species diversity in a long-term resource manipulation experiment. Journal of Vegetation Science 24: 843-864.
::CrossRef::Google Scholar::
(21)
Giordani P, Brunialti G, Benesperi R, Rizzi G, Frati L, Modenesi P (2009). Rapid biodiversity assessment in lichen diversity surveys: implications for quality assurance. Journal of Environmental Monitoring 11 (4): 730-735.
::CrossRef::Google Scholar::
(22)
Goodall DW (1953). Point quadrat methods for the analysis of vegetation - the treatment of data for tussock grasses. Australian Journal of Botany 1 (3): 457-461.
::CrossRef::Google Scholar::
(23)
Grime JP (1998). Benefits of plant diversity to ecosystems: immediate, filter and founder effects. Journal of Ecology 86: 902-910.
::CrossRef::Google Scholar::
(24)
Hillebrand H, Bennett DM, Cadotte MW (2008). Consequences of dominance: a review of evenness effects on local and regional ecosystem processes. Ecology 89: 1510-1520.
::CrossRef::Google Scholar::
(25)
Kent M, Coker P (1992). Vegetation description and analysis - a practical approach. John Wiley & Sons, Chichester, UK, pp. 363.
::Google Scholar::
(26)
Laliberte E, Legendre P (2010). A distance-based framework for measuring functional diversity from multiple traits. Ecology 91: 299-305.
::CrossRef::Google Scholar::
(27)
Laliberte E, Shipley B (2011). FD: measuring functional diversity from multiple traits, and other tools for functional ecology. R package version 10-11.
::Online::Google Scholar::
(28)
Lavorel S, Grigulis K, McIntyre S, Garden D, Williams N, Dorrough J, Berman S, Quétier F, Thébault A, Bonis A (2008). Assessing functional diversity in the field - methodology matters! Functional Ecology 22: 134-147.
::CrossRef::Google Scholar::
(29)
Lewontin R (1966). On the measurement or relative variability. Systematic Zoology 15: 141-142.
::CrossRef::Google Scholar::
(30)
Loreau M (2000). Biodiversity and ecosystem functioning: recent theoretical developments. Oikos 91: 3-17.
::CrossRef::Google Scholar::
(31)
Maestre FT, Quero JL, Gotelli NJ, Escudero A, Ochoa V, Delgado-Baquerizo M, García-Gómez M, Bowker MA, Soliveres S, Escolar C, García-Palacios P, Berdugo M, Valencia E, Gozalo B, Gallardo A, Aguilera L, Arredondo T, Blones J, Boeken B, Bran D, Conceicao A, Cabrera O, Chaieb M, Derak M, Eldridge D, Espinosa CI, Florentino A, Gaitán J, Gatica MG, Ghiloufi W, Gómez-González S, Gutiérrez JR, Hernández RM, Huang X, Huber-Sannwald E, Jankju M, Miriti M, Monerris J, Mau RL, Morici E, Naseri K, Ospina A, Polo V, Prina A, Pucheta E, Ramírez-Collantes DA, Romão R, Tighe M, Torres-Díaz C, Val J, Veiga JP, Wang D, Zaady E (2012a). Plant species richness and ecosystem multifunctionality in global drylands. Science 335: 214-218.
::CrossRef::Google Scholar::
(32)
Maestre FT, Castillo-Monroy AP, Bowker MA, Ochoa-Hueso R (2012b). Species richness effects on ecosystem multifunctionality depend on evenness, composition and spatial pattern. Journal of Ecology 100: 317-330.
::CrossRef::Google Scholar::
(33)
Mason N, Mouillot D, Lee W, Wilson J (2005). Functional richness, functional evenness and functional divergence: the primary components of functional diversity. Oikos 111: 112-118.
::CrossRef::Google Scholar::
(34)
McIntyre S, Lavorel S (2001). Livestock grazing in sub-tropical pastures: steps in the analysis of attribute response and plant functional types. Journal of Ecology 89: 209-226.
::CrossRef::Google Scholar::
(35)
Mouillot D, Villéger S, Scherer-Lorenzen M, Mason NWH (2011). Functional structure of biological communities predicts ecosystem multifunctionality. PLoS ONE 6 (3): e17476.
::CrossRef::Google Scholar::
(36)
Nicolau M (2002). Modeling and mapping the spatial distribution of precipitation. An application to Portugal mainland. PhD thesis, Universidade de Lisboa, Lisbon, Portugal.
::Google Scholar::
(37)
Pakeman RJ, Quested HM (2007). Sampling plant functional traits: what proportion of the species need to be measured? Applied Vegetation Science 10: 93-8.
::CrossRef::Google Scholar::
(38)
Pielou, EC (1975). Indices of diversity and evenness. In: “Ecological Diversity” (Pielou EC ed). John Wiley & Sons, New York, pp. 5-18.
::Google Scholar::
(39)
Porto M, Correia O, Beja P (2011). Long-term consequences of mechanical fuel management for the conservation of Mediterranean forest herb communities. Biodiversity Conservation 20: 2669-2691.
::CrossRef::Google Scholar::
(40)
Prieur-Richard AH, Lavorel S, Dos Santos A, Grigulis K (2002). Mechanisms of resistance of Mediterranean annual communities to invasion by Conyza bonariensis: effects of native functional composition. Oikos 99: 338-346.
::CrossRef::Google Scholar::
(41)
R Core Team (2013). R: A language and environment for statistical computing R Foundation for Statistical Computing, Vienna, Austria.
::Online::Google Scholar::
(42)
Reynolds JF, Stafford Smith DM, Lambin EF, Turner II BL, Mortimore M, Batterbury SP, Downing TE, Dowlatabadi H, Fernandéz RJ, Herrick JE, Huber-Sannwald E, Jiang H, Leemans R, Lynam T, Maestre FT, Ayarza M, Walker B (2007). Global desertification: building a science for dryland development. Science 316: 847-851.
::CrossRef::Google Scholar::
(43)
Scheffer M, Carpenter S, Foley JA, Folke C, Walker B (2001). Catastrophic shifts in ecosystems. Nature 413: 591-596.
::CrossRef::Google Scholar::
(44)
Schleuter D, Daufresne M, Massol F, Argillier C (2010). A user’s guide to functional diversity indices. Ecological Monographs 80 (3): 469-484.
::CrossRef::Google Scholar::
(45)
Sprent JI, Gehlot HS (2010). Nodulated legumes in arid and semi-arid environments: are they important? Plant Ecology and Diversity 3 (3): 211-219.
::CrossRef::Google Scholar::
(46)
Stohlgren TJ, Falkner MB, Schell LD (1995). A modified-Whittaker nested vegetation sampling method. Plant Ecology 117 (2): 113-121.
::CrossRef::Google Scholar::
(47)
Talavera S, Aedo C, Castroviejo S, Romero Zarco C, Sáez L, Salgueiro FJ, Velayos M (1999). Flora iberica. Plantas vasculares de la Península Ibérica e Islas Baleares [Flora iberica. Vascular plants of the Iberian Peninsula and Balearic Islands]. Real Jardín Botánico, CSIC, Madrid, Spain, pp. 34.[in Spanish]
::Online::Google Scholar::
(48)
Tilman D, Knops J, Wedin D, Reich P, Ritchie M, Siemann E (1997). The influence of functional diversity and composition on ecosystem processes. Science 277: 1300-1302.
::CrossRef::Google Scholar::
(49)
Verrill S, Johnson RA (2007). Confidence bounds and hypothesis tests for normal distribution coefficients of variation. Communications in Statistics, Theory and Methods 36: 2187-2206.
::CrossRef::Google Scholar::
(50)
Villéger S, Mason N, Mouillot D (2008). New multidimensional functional diversity indices for a multifaceted framework in functional ecology. Ecology 89: 2290-2301.
::CrossRef::Google Scholar::
(51)
Vittoz P, Guisan A (2007). How reliable is the monitoring of permanent vegetation plots? A test with multiple observers. Journal of Vegetation Science 18: 413-422.
::CrossRef::Google Scholar::
(52)
Walker BH, Kinzig A, Langridge J (1999). Plant attribute diversity, resilience, and ecosystem function: the nature and significance of dominant and minor species. Ecosystems 2: 95-113.
::CrossRef::Google Scholar::

Download

Paper Contents

Paper Sections

Paper Figures

Paper Tables

Supplementary Material

 
 
Close

 

Nunes A, Tápia S, Pinho P, Correia O, Branquinho C (2014).
Advantages of the point-intercept method for assessing functional diversity in semi-arid areas
iForest - Biogeosciences and Forestry 8: 471-479. - doi: 10.3832/ifor1261-007
Close
First Previous Next Last
 
Close
© iForest

Download Reference

Paper ID# ifor1261-007
Title Advantages of the point-intercept method for assessing functional diversity in semi-arid areas
Authors Nunes A, Tápia S, Pinho P, Correia O, Branquinho C
Format
Close Download