Next Article in Journal
Genome-Wide Association Study (GWAS) for Examining the Genomics Controlling Prickle Production in Red Raspberry (Rubus idaeus L.)
Next Article in Special Issue
Evaluation of the Contribution of Pastures on the Economic Sustainability of Small Ruminant Farms in a Typical Greek Area
Previous Article in Journal
Genotypic Variation in Nutrient Uptake Requirements of Rice Using the QUEFTS Model
Previous Article in Special Issue
Spatial Variability of Soil Phosphorus Indices under Two Contrasting Grassland Fields in Eastern Canada
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Agronomy
Volume 11
Issue 1
10.3390/agronomy11010025
agronomy-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Floristic Composition Mediates Change in Forage Nutritive Quality of Atlantic Mountain Grasslands after Experimental Grazing Exclusion

by
Gonzalo García-Baquero
1,2,
Iñaki Odriozola
1 and
Arantza Aldezabal
1,*
1
Department of Plant Biology and Ecology, University of the Basque Country, 48940 Leioa, Spain
2
Group of Environmental Epidemiology and Child Development, Biodonostia Health Research Institute, 20014 San Sebastian, Spain
*
Author to whom correspondence should be addressed.
Agronomy 2021, 11(1), 25; https://doi.org/10.3390/agronomy11010025
Submission received: 14 November 2020 / Revised: 18 December 2020 / Accepted: 22 December 2020 / Published: 24 December 2020
(This article belongs to the Special Issue Sustainable Management and Utilization of Permanent Grassland)
Browse Figures
Review Reports Versions Notes

Abstract

:
Simultaneous reduction of biodiversity and forage nutritive value after grazing abandonment represents a critical agroecological problem observed in temperate mountain grasslands. However, how both processes affect each other after the exclusion of large grazers is not well understood. To address this knowledge gap, we used four Atlantic grassland sites in the Aralar masif (northern Iberian Peninsula) to ask whether floristic composition mediates change in forage nutritive quality after grazing exclusion, and, if so, how much of the effect on forage quality is mediated. To examine the effects of grazing exclusion on forage quality and floristic composition a repeated-measures (2005–10) randomized complete block experiment was used. Then, the direct effects of grazing exclusion on forage quality were disentangled from the indirect effects mediated by concurrent change in floristic composition. Grazing exclusion deteriorated forage mineral content, phosphorus content, neutral detergent fiber and, especially, crude protein and enzymatic digestibility. Concurrent floristic change mediated change in crude protein (80%), enzymatic digestibility (55%) and forage content in calcium (31%). Our study shows that plant diversity and forage quality are intimately linked features of Atlantic mountain grasslands and highlights the importance of preserving leguminous and dicotyledonous species to maintain the nutritive value of these grasslands.

1. Introduction

Atlantic and Pyrenean mountain grasslands located below 2000 m.a.s.l. (meters above sea level), preserved by extensive multispecific (sheep, cattle, horses) livestock grazing in local transhumance, are among the most biodiverse habitats in Europe [1]. Decreasing plant diversity in these grasslands might cause negative consequences on ecosystem functioning, stability, and services [2,3]. For instance, plant diversity enhanced quality-adjusted yield and revenues, independently on management intensity, in a large-scale biodiversity experiment in semi-natural grasslands [4]. Many studies have reported that grazing contributes the maintenance of several ecosystem functions and services, especially in historically grazed temperate grasslands; among others, landscape heterogeneity [5], suitable habitat for wild vertebrates [6], plant diversity [7], forage quality [8,9,10], soil microbial activity [11], nutrient cycling [12,13,14], pollinator networks [15] and invertebrate diversity [16,17].
In addition to being a reservoir of biodiversity, semi-natural managed grasslands are also an essential food resource for herbivores and prevent the consumption of supplementary feed (and the consequent economic cost) during the growing season (which could last from four to six months), contributing to the sustainability of the grazing system [18] and GHG mitigation [19]. However, the future of the traditional management of these ecosystems is under constant threat due to socio-economic and market changes [20]. Similar to most extensively managed European grassland systems [21], the use of mountain grasslands by domestic livestock has declined slowly in most of the Pyrenean-Basque-Cantabrian grasslands [22,23].
Grazing can influence forage quality (i.e., palatability and nutritional value) directly, for instance, by promoting the growth of younger plant tissues [24]. This is expected in historically grazed grasslands [10] with several livestock species exerting a non-selective grazing pressure, where plants respond with a tolerance strategy [25]. Grazing may also affect forage quality indirectly, through changes in floristic composition (e.g., [10,26,27]). Non-selective shoot herbivory exerted by mixed livestock equalizes the ability of plant species to compete for light [28], thus allowing the co-existence of short-statured highly palatable legumes and nutrient-rich forbs [29,30], together with tall competitive grasses [31]. After grazing abandonment tall grasses tend to out compete remaining short-statured species [7,32], reducing forage quality and plant diversity altogether. These effects vary both in time and in space, but long-term manipulative studies that explore them are rare, and the relative contribution of direct and indirect effects of grazing abandonment on forage nutritive value are hardly known.
This study aims at addressing this knowledge gap by using the results of a repeated-measurements (2005–2010) randomized complete block design experiment (RCBD) implemented to test for the effects of ungulate grazing exclusion, via experimental fencing, on both forage nutritive quality and species composition. After testing for the effects of grazing exclusion over time on forage nutritive quality and floristic composition, we asked whether simultaneous change in floristic composition mediates change in forage quality and, if so, how much of the effect of grazing exclusion on forage quality is direct and how much is mediated by concurrent change in floristic composition. For that purpose, we used mediation analysis [33] with the aim to disentangle the direct effects of grazing exclusion (X) on forage quality (Y) from those (indirect) effects mediated by change in floristic composition (M), as outlined in Figure 1. Thus, the direct effects of grazing exclusion on forage quality (i.e., the path XY), as they are here framed, include any effect that is not mediated by floristic composition but that may, in fact, be either truly direct or rather mediated by factors other than floristic composition, such as the accumulation of an increasingly thick organic layer that provides insulation to the soil [8]. Mediation analysis, though, must consider the possible existence and influence of variable(s) Z (Figure 1), sometimes called moderators, acting as potential confounders of the relationship between the mediator M and the outcome Y [34]. In this work, Z is represented by the physical and chemical properties of the topsoil. The importance of soil characteristics in determining species composition has been highlighted in previous studies [35,36,37], and although the influence of such characteristics on floristic composition is more pronounced in arid environments compared to temperate grasslands, where species richness and diversity strongly respond to grazing [38], it can also affect the floristic composition and should not be ignored.

2. Materials and Methods

2.1. Study Area

The study area is located in the Aralar Natural Park (42°59′48″ N and 2°06′51″ W), an 11,000-ha Special Area of Conservation (part of the European Natura 2000 network) located in the Atlantic Basque Country (Northern Iberian Peninsula, see Supplementary Figure S1). The area has oceanic climate, with mean annual temperature equal to 7 °C and annual precipitation of about 1330 mm. A mixture of calcareous and acid substrates is present in the area, but its oceanic climate with abundant precipitation throughout the growing season has led to the development of acidic soils. In these soils, the average of Kjeldahl total nitrogen is 6534 ppm, available phosphorus 8.3 ppm, mean soil water content at 10-cm depth 46.3%, and pH 4.8 ppm [7]. The vegetation in the park comprises a mosaic of gorse-heather shrublands and semi-natural grasslands, which are located between 800 m.a.s.l and 1400 m.a.s.l. This area is managed as a communal land, called Enirio-Aralar, and native grasslands included in the Habitat Directive are dominant: the most relevant is the Jasiono-Danthonietum grassland (code 6230, subtype a, 92/43/EEC, European Commission, 2013), located in relatively flat terrain (slope range: 0–20°) and primarily comprising perennial graminoids such as Festuca nigrescens subsp. microphylla (St-Yves) Markgr.-Dannenb. (chewings fescue), Agrostis capillaris L. (common bent), Luzula campestris (L.) DC. in Lam. & DC. (field woodrush) in Lam. & DC., and herbaceous dicotyledons such as Galium saxatile L. (heath bedstraw), Trifolium repens L. (white clover) and Cerastium fontanum Baumg. (starweed) [7,11]. The mean aboveground net primary production (ANPP) of Jasiono-Danthonietum grassland is 1.97 t dry mass ha−1 year−1 with standard error of 0.39 [39]. Livestock herds (sheep, cattle and horses) are managed in short transhumance, using the lowland farms (below 700 m.a.s.l.) in winter, and extensively using the upland grasslands (up to 1400 m.a.s.l.) from May to October–November. In Enirio-Aralar communal land, the number of grazing animals has remained almost stable, with 18,000 dairy ewes, 800 beef cattle and 700 mares [11]. The stocking rate varies from 2.3 to 4.5 LU ha−1d−1 along the growing season [8]. The most common size of the transhumance flocks is between 200 and 500 lactating ewes. Lactating ewes’ milk is destined mainly to the production of the local Idiazabal Protected Denomination of Origin (PDO) cheese, awarded worldwide, and sold mainly in local market or directly on farms.

2.2. Experimental Design and Sampling Procedure

Experimental sites were selected using a GIS application, intersecting slope and vegetation maps and obtaining a new map with suitable areas for establishing exclusion fences. Four field sites were selected randomly into this map to conduct the experiment in the Jasiono-Danthonietum temperate grassland (Supplementary Figure S1): Oidui (42°59′16.6″ N, 2°5′59.4″ W; 876 m.a.s.l), Uzkuiti (43°0′50″ N, 2°4′3″ W; 1300 m.a.s.l); Igaratza (42°59′9.25″ N, 2°2′9.7″ W; 1247 m.a.s.l); and Alotza (43°0′10.6″ N, 2°5′22″ W; 1223 m. a.s.l). The minimum and maximum distance between any of these sites was 2175 m and 4720 m respectively and all sites were located on relatively flat terrain. In order to simulate grazing cessation, one permanent fenced plots (50 m × 50 m each) was installed in May 2005 at each experimental sites. Next to each fenced plot (‘exclosure’ level), we delineated a grazed plot (‘grazing’ level), where dairy sheep, beef cattle and horses were allowed to graze continuously during the vegetative period (from May–November). Additionally, 3–4 temporary moveable cages of 1.5 × 1.5 m were randomly allocated within the grazing plot at the beginning of each growing season. Each month, quadrats outside permanent fences were clipped inside temporary moveable cages that were randomly relocated in grazed plot after biomass measurements. This procedure was carried out every month. The sequential use of such temporary cages prevented that the plant biomass inside the cage was eaten by grazing animals, and allowed us to estimate accurately the nutritive quality of forage consumed by herbivores. Each month, three 1 m2-quadrats were clipped per site and treatment at ground level. Unfortunately, in some cases it was not possible to sample all quadrats due to problems with temporary cages and meteorological conditions, so the total number of samples per growing season varied from 46 to 71 quadrats.
The experiment was implemented according to a repeated-measurements randomized complete block (RCB) design [40]. The field sites acted as blocks (categorical random factor ‘Site’ with four levels). Grazing exclusion represented the experimental manipulation (fixed factor ‘Treatment’ with three levels: grazing, cage, and exclosure). The factor ‘Time’, understood as a fixed but otherwise quantitative variable, constitutes a second fixed factor. Time is indexed in months, starting in June 2005 (month number 1) and ending in October 2010 (month number 65). Thereby, the basic RCB design converts into a repeated-measurements RCB design. Therefore, the statistical model corresponding to this experimental design is (using ‘effects’ model parameterization):
yijk = μ + τi + bij + δk + (τδ)ik + εijk
where yijk corresponds to the response measured in treatment i in site j at time k; μ is the level of reference (grazing, in this application); τi corresponds to the effect of treatment i (with respect to the level grazing); bij is the term for the random factor ‘site’, i.e., for site j under treatment i; δk corresponds to the effect of time k; (τδ)ik is the interaction between the fixed effects; finally, εijk corresponds to the error. bij is independently and identically distributed as normal (0, σb2); likewise, εijk is independently and identically distributed as normal (0, σ2).
Vegetation sampling took place from June 2005 to October 2010, using the quadrat cover method [38,41]: once per month from May/June–October, floristic composition was estimated visually within three randomly placed, non-overlapping 1-m2 quadrats per plot. For estimating species cover, we used the following semi-quantitative indices: + (<1%), 1 (1–10%), 2 (11–20%), 3 (21–30%), 4 (31–40%), 5 (41–50%), 6 (51–60%), 7 (61–70%), 8 (71–80%), 9 (81–90%), 10 (91–100%). All vascular plants present within each quadrat were identified at the species level. These semi-quantitative indices were transformed to quantitative ones, replacing each index by the middle value of the cover range (for instance: index 1 was replace by 5%). The index ‘+’ was converted to 0.5%. After composition determination, standing plant biomass of the same 1-m2 quadrat was clipped in order to measure forage quality variables. Each sample was stored in plastic bags, labelled, and processed in the Eskalmendi Agricultural Laboratory, drying them in an oven at 60 °C for 48 h. The nutritional variables thus measured were the following: macrominerals (calcium, Ca; phosphorus, P; potassium, K; and magnesium, Mg), crude protein (CP, calculated as nitrogen × 6.25), neutral detergent fiber (NDF), enzymatic digestibility (NDF cellulase, and NDFcel). All analytical procedures follow standard protocols [42].
Topsoil properties were measured at the end of the experiment. Soil samples were collected from a depth of 0–10 cm, using a 3-cm-diameter soil core sampler. Each soil sample comprised a pool of 10 cores (subsamples) per 1-m2 quadrat. The measured soil variables were as follows: pH, effective cation-exchange capacity (cmolc kg−1), effective aluminium cation-exchange capacity (%), oxidable organic matter (%), and macronutrients, including Kjeldahl total nitrogen (mg L−1), available phosphorus (mg L−1), available potassium (mg L−1), available sodium (mg L−1), available calcium (mg L−1), and available magnesium (mg L−1), and texture (percentages of gross and fine sand, silt, clay). All analytical procedures were conducted in Fraisoro Agricultural Laboratory following ADAS standard procedures [43].

2.3. Mixed Modelling

After exploratory data analysis [44], we applied mixed effects modeling [45] to test for the effects of the treatment and time factors. Mixed models were fitted with function lme() of package nlme [46] in R software [47]. Restricted maximum likelihood was employed for parameter estimation. Including a correlation structure was needed in some cases, albeit a simple structure of the type autoregressive process of order 1 (corAR1) sufficed. The packages MuMIn [48] and sjPlot [49] were used to obtain R2 statistics and graphs, respectively.

2.4. Mediation Analysis

To disentangle the direct effects of grazing exclusion on forage quality from the effects mediated by concurrent change in floristic composition, mediation analysis was applied using the model-based approach implemented in the R package mediation [50]. We used data from only the end of the experiment (year 2010) to fit sets of linear models according to our experimental design; then mediation effects were estimated from these fitted models [51,52], via function mediation. Three variables derived from species composition (legume-to-grass ratio, proportion of dicots, and Simpson’s index of diversity) were tried as mediators, simultaneously controlling for potential confounding due to topsoil properties (variables Z in Figure 1). Simpson’s index of diversity [53], which indicates the probability that any two individuals in the sample will belong to different species, was computed as follows:
Simpson ’ s   index   of   diversity = 1 − ∑ p i 2
where pi represents the proportional abundance of species i in the sample.
The value of the Simpson’s index ranges between 0 (minimum diversity) and 1 (maximum diversity). Observed Simpson’s index of diversity and Shannon index (H’) are often correlated, even though the latter is upper-unbounded. Both direct and mediation effects are defined according to [54]’s concept of potential outcomes (See also [55]). Significance of mediation effects was assessed via quasi-Bayesian Monte Carlo tests [51].

2.5. Data Availability

The dataset is available from figshare: ForageQuality.Dataset.xls. To access the dataset, go to https://doi.org/10.6084/m9.figshare.13336877.

3. Results

3.1. Forage Quality

Exploratory data analysis (Supplementary Tables S1 and S2 and Figure S2) showed that most forage quality variables were sufficiently symmetrical for linear modeling purposes; besides, they were significantly correlated among them, though, in general terms, not strongly correlated (the exceptions are the pairs Mg-Ca and NFD-NDFcel, whose correlations are close to or above |0.5|).
Except for ash and Ca, the interaction between treatment and time was significant in all cases (Table 1 and Table 2), indicating that the effect of treatment on Mg, K, P, CP, NDF, and NDFcel depends on Time (Figure 2 and Supplementary Figure S3); main tests for time are not significant in many cases, which is due to the observed crossover interactions (Figure 2a and Supplementary Figure S3). Tukey’s post-hoc tests (Supplementary Table S3) indicate that the levels grazing-exclosure and cage-exclosure are significantly different in all cases but K, and also that the levels grazing-cage are significantly different only in the cases of Mg, K, and CP.
In general, the indicators of forage nutritive quality Mg, K, P, CP, NDF, and NDFcel under the levels grazing and cage of the treatment factor tend to remain stable or improve over time, whereas they tend to deteriorate under the exclosure level of treatment. In the case of Ca (Figure 2b), only both main effects are significant; in the case of ash, only treatment is significant (Supplementary Figure S3). In summary, all quality descriptors except ash deteriorated quickly after grazing exclusion (Table 2, Supplementary Figure S3), with no generalized significant differences between grazing and cage levels of the treatment factor.

3.2. Floristic Composition

Simpson’s index of plant species diversity remained stable over time within the experimental exclusions, whereas it improved outside the exclusions (Figure 3a; Table 1 and Table 2), i.e., under grazing and cage levels of the treatment factor. By contrast, whereas the legume-to-grass ratio and the proportion of dicots deteriorated over time within the experimental exclusions, they also tended to improve outside the exclusions (Figure 3b and Figure S4; Table 1 and Table 2). The evidence for statistical difference between grazing and cage levels of the treatment factor, though significant for the legume-to-grass ratio, was, in general terms, weak (Supplementary Table S3).

3.3. Mediation Analysis

As shown in Figure 4, using only the data corresponding to the end of the experiment (i.e., after six years of grazing exclusion), either the legume-to-grass ratio or the proportion of dicots mediated change in crude protein (p = 0.02), enzymatic digestibility (p = 0.02) and forage content in cation calcium (p = 0.05). Further results (not detailed here) also hint at a possible mediation effect of species diversity on potassium forage content (p = 0.09), but no evidence of indirect effects was found regarding magnesium and phosphorus forage content. Since ash did not change over time, mediation analysis was not applied to this outcome variable.

4. Discussion

Discovering causal mechanisms is an essential objective of science [55], for which randomized experiments constitute a primary tool [56]. Very often, though, even randomized experiments may merely ascertain whether a given factor causes change in another (the response), but without unveiling how exactly this relationship takes place [57]. By means of linear mixed modelling [40,45], we found that experimental fencing to recreate exclusion of ungulate grazers in an Atlantic grassland system reduced forage nutrient quality indicators after just six years of experimental manipulation. These quality indicators were mineral content, crude protein, phosphorus content, neutral detergent fiber and enzymatic digestibility. Additionally, by means of mediation analysis [33,34] based on Robins and Greenlands’ [54] concept of potential outcomes, we found that concurrent floristic change at the end of our experiment mediates 80% of change in crude protein. Concurrent floristic change also mediates 55% of change in enzymatic digestibility and 31% of change in forage content in calcium. Thus, a substantial proportion of the decline in the said three forage quality descriptors is due to simultaneous change in floristic composition.
Cessation of grazing in productive long-history grasslands, where plant species are adapted to tolerate grazing, was expected to lead to reduction in forage nutrient content and palatability [24,58]. Similar results have been, in fact, reported in humid and sub-humid grasslands [10,26,59,60]. However, the positive temporal trends in forage quality found in both grazed treatments were unexpected. In a previous study in these field sites, we reported an increasing trend of ANPP between 2005–2010 for both grazed and fenced treatments [39], concurrent with an increase in stocking rate (cattle and, especially, horses) [61]. Increased stocking rate increases defoliation intensity and fertilization by urine and feces [62], both of which are known to contribute to improve forage quality. Defoliation promotes the growth of younger, more palatable plant tissue [24] and fertilization favors a resource acquisition strategy and may also improve forage quality [63,64]. Nevertheless, we believe that the increasing trend in forage quality will probably be short-lived, since the observed increases in stocking rate and ANPP are likely local trends within longer-term oscillations.
We found significant, albeit small in size, differences between grazing and cage treatments for a few forage quality indicators. These differences might be due to the fact that plants collected from cages were complete fully-grown individuals with higher leaf:stem ratio, whereas the plant samples collected in the grazed plots were necessarily constituted by basal parts with lower leaf:stem ratio. In grasslands, foliar nitrogen allocation increases from the bottom to the top of the plant canopy [65]. Grasses contain the leaf sheath and other lignin-rich components in the basal area and dicots have the youngest (with less lignified structures) tissues in the aerial part, due to the apical growth meristems. Lignin, in turn, is the principal factor limiting forage digestibility [66]. Moreover, plant parts also differ in carbon and nitrogen allocation: leaves are often of a higher quality than stems in mature stage, due to lower C/N ratio and richer non-cell wall components (i.e., protein and soluble carbohydrates) [58], whereas the temporal cages represent the forage that was consumed by livestock in the previous time interval, the grazed plots represent what remained after the grazing activity. For this reason, we believe that the best measure of animal-consumed forage quality is the one measured inside the temporal cages.
That tall competitive grasses dominate after grazing exclusion, at the expense of smaller species (i.e., legumes and forbs), is something known since Tansley and Adamson [32] conducted the first-known grazing exclusion experiment. Our results also agree with the predictions of the general model of grazing for productive grasslands with a long evolutionary history of grazing [67,68]. However, plant diversity, as measured by Simpson’s index, remained stable after grazing exclusion. The short time elapsed (6 years.) since the beginning of the experiment might explain the absence of diversity loss in the excluded plots. This explanation is based in that a previous study carried out in the same plots showed that the observed reduction of dicots proportion resulted in plant diversity loss only after 13 years of grazing exclusion [7]. Thus, the observed steady reduction in legumes and forbs after grazing exclusion may be an indication of posterior biodiversity loss.
Dicots proportion and plant species diversity also increased, unexpectedly, under grazing treatments. The aforementioned increase in stocking rate at the end of the experimental period might also explain floristic changes in grazed plots. Creation of colonization opportunities is an important mechanism by which herbivory promotes diversity [69]. Enhanced defoliation and trampling by herbivores opens new gaps in the vegetation and brings about opportunities for ruderal species to become established [70]. Our results support this hypothesis, since increased diversity is mainly explained by the appearance in grazed plots of several annual species of ruderal-nitrophile nature such as Trifolium dubium Sibth., Taraxacum gr. officinale Weber., Sherardia arvensis L., Poa annua L., Malva neglecta Wallr., Medicago lupulina L., and Stellaria media (L.) Vill.).
Changes in floristic composition mediated the effect of grazing on several forage quality indicators. In particular, the legume-to-grass ratio explained most of the variation in crude protein (80%) and digestibility (55%). These results indicate that preserving nitrogen-fixing legumes linked to extensive grazing is of paramount importance to preserve the nutritional quality of Atlantic mountain grasslands. This is so because reducing the proportion of fiber and increasing the proportion of protein of forage are key factors to increase livestock production [71,72]. Moreover, dicots proportion mediated a significant part of the grazing effect on forage content in calcium (31%), but the significance of the effect of Simpson diversity index in potassium content was marginal. Our results agree with previous findings indicating that dicots are richer than grasses in minerals [29,30], which are essential to meet the nutritional requirements of grazing animals. The abundance of legumes and forbs also contributes to enhance the nutritive quality of milk produced by lactating ewes: it correlates with higher fat content, richer in α-linolenic acid and its biohydrogenation intermediates, as well as long-chain saturated branched-chain and cis-monounsaturated fatty acids [73,74].
Our results show that preserving plant diversity is required in order to maintain high forage nutritive value in Atlantic mountain grasslands. In a recent experiment on biodiversity and ecosystem function, carried out in semi-natural grasslands, increasing plant diversity promoted quality-adjusted yield [4]. However, opposite to our findings, found that only biomass yield increased, whereas forage quality remained stable [4]. However, whereas in our experiment plant diversity is correlated with the abundance of nutrient rich legumes and forbs, in Schaub et al.’s experiment, plant diversity was uncorrelated with the presence of specific functional groups. This was because diversity was experimentally augmented by randomly adding plant species. Thus, the link observed in our study between conservation of biodiversity and forage nutritive quality is likely applicable to only natural grassland systems where plant diversity correlates positively with the presence of nutrient-rich plant functional groups.

5. Conclusions

Grazing is a pivotal process for mountain grasslands that links economic use with emphasis on forage productivity and quality with conservation and management for maintaining biodiversity. By unveiling and quantifying mediation mechanisms by which changing plant diversity influences forage quality in Atlantic mountain grasslands, our findings shed light on the precise nature of that link. Thus, in terms of grassland management within the context of global change, our results underline the practical importance of preserving, by means of extensive grazing, high leguminous-to-grass ratio and proportion of dicotyledonous species to maintain the high nutritive value of these grasslands. Our findings indicate that the traditional management of Atlantic mountain grasslands, with livestock grazing in local transhumance regimes, promotes the sustainability of the grazing system (by reducing the need of supplemental feed and consequent reduction on GHG emissions), the production of high-quality local products with much-needed added economic value and the conservation of biodiversity with all the ecosystem services linked to this diversity. For Atlantic grasslands, and possibly for almost every grassland type, sustainable economic use and biodiversity conservation are not opposed management objectives.

Supplementary Materials

The following are available online at https://www.mdpi.com/2073-4395/11/1/25/s1, Figure S1: Study area and experimental plots; Figure S2: Bivariate plots, histograms and Kendall correlations for forage nutritive quality descriptors; Figure S3: Change in descriptors of forage nutritive quality, Figure S4: change in dicots proportion, Table S1: Definition of forage nutritive quality and floristic composition descriptors; Table S2: descriptive statistics of forage quality properties and floristic composition; Table S3: Post-hoc (Tukey) tests; Table S4: Restricted maximum likelihood estimates (ReML) for cation calcium content (% of dry matter, DM), crude protein (% of dry matter, DM), Simpson’s index and legume-to-grass ratio at the end of each season (October) over the period 2005–2010.

Author Contributions

Conceptualization: G.G.-B. and A.A.; methodology: G.G.-B. and A.A.; formal analysis: G.G.-B.; investigation: A.A.; resources: A.A.; data curation: G.G.-B. and I.O.; writing—original draft preparation: G.G.-B., I.O., and A.A.; writing—review and editing: G.G.-B., I.O., and A.A.; visualization: G.G.-B.; supervision: A.A.; project administration: A.A.; funding acquisition: A.A. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Basque Government—Eusko Jaurlaritza—(grant numbers: ETORTEK10/34, IT1022-16), by the UNESCO Cathedra of the University of the Basque Country—UPV/EHU—(grant number: UNESCO07/07), and by the Ministry of Economy and Competitiveness of the Spanish Government (grant number: AGL2013-48361-C2-1-R).

Acknowledgments

We are indebted to all those who helped us with fieldwork and provided logistical support. In particular, we are grateful to Ana Etxeberria (technical assistant), Jose Antonio Irastortza (guard of the Aralar Natural Park) and to the botanists of ARANZADI Scientific Association (especially to Mari Azpiroz and Lorena Uriarte).

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

References

  1. Dengler, J.; Janišová, M.; Török, P.; Wellstein, C. Biodiversity of Palaearctic grasslands: A synthesis. Agric. Ecosyst. Environ. 2014, 182, 1–14. [Google Scholar] [CrossRef] [Green Version]
  2. Cardinale, B.J.; Srivastava, D.S.; Duffy, J.E.; Wright, J.P.; Downing, A.L.; Sankaran, M.; Jouseau, C. Effects of biodiversity on the functioning of trophic groups and ecosystems. Nature 2006, 443, 989–992. [Google Scholar] [CrossRef] [PubMed]
  3. Wang, L.; Delgado-Baquerizo, M.; Wang, D.; Isbell, F.; Liu, J.; Feng, C.; Liu, J.; Zhong, Z.; Zhu, H.; Yuan, X.; et al. Diversifying livestock promotes multidiversity and multifunctionality in managed grasslands. Proc. Natl. Acad. Sci. USA 2019, 116, 6187–6192. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  4. Schaub, S.; Finger, R.; Leiber, F.; Probst, S.; Kreuzer, M.; Weigelt, A.; Buchmann, N.; Scherer-Lorenzen, M. Plant diversity effects on forage quality, yield and revenues of semi-natural grasslands. Nat. Commun. 2020, 11, 1–11. [Google Scholar] [CrossRef] [Green Version]
  5. Augustine, D.J.; Frank, D.A. Effects of Migratory Grazers on Spatial Heterogeneity of Soil Nitrogen Properties in a Grassland Ecosystem. Ecology 2001, 82, 3149–3162. [Google Scholar] [CrossRef]
  6. Pärtel, M.; Bruun, H.; Sammul, M. Biodiversity in temperate European grasslands: Origin and conservation. In Proceedings of the 13th International Occasional Symposium of the European Grassland Federation, Integrating Efficient Grassland Farming and Biodiversity; Lillak, R., Viiralt, R., Linke, A., Geherman, V., Eds.; Organising Committee of the 13th International Occasional Symposium of the European Grassland Federation and Estonian Grassland Society (EGS): Tartu, Estonia, 2005; pp. 1–14. [Google Scholar]
  7. Odriozola, I.; García-Baquero, G.; Fortin, M.-J.; Laskurain, N.A.; Aldezabal, A. Grazing exclusion unleashes competitive plant responses in Iberian Atlantic mountain grasslands. Appl. Veg. Sci. 2017, 20, 50–61. [Google Scholar] [CrossRef]
  8. Odriozola, I.; García-Baquero, G.; Laskurain, N.A.; Aldezabal, A. Livestock grazing modifies the effect of environmental factors on soil temperature and water content in a temperate grassland. Geoderma 2014, 235–236, 347–354. [Google Scholar] [CrossRef]
  9. Peco, B.; Navarro, E.; Carmona, C.P.; Medina, N.G.; Marques, M.J. Effects of grazing abandonment on soil multifunctionality: The role of plant functional traits. Agric. Ecosyst. Environ. 2017, 249, 215–225. [Google Scholar] [CrossRef]
  10. Cingolani, A.M.; Posse, G.; Collantes, M.B. Plant functional traits, herbivore selectivity and response to sheep grazing in Patagonian steppe grasslands. J. Appl. Ecol. 2005, 42, 50–59. [Google Scholar] [CrossRef]
  11. Aldezabal, A.; Moragues, L.; Odriozola, I.; Mijangos, I. Impact of grazing abandonment on plant and soil microbial communities in an Atlantic mountain grassland. Appl. Soil Ecol. 2015, 96, 251–260. [Google Scholar] [CrossRef]
  12. Semmartin, M.; Aguiar, M.R.; Distel, R.A.; Moretto, A.S.; Ghersa, C.M. Litter quality and nutrient cycling affected by grazing-induced species replacements along a precipitation gradient. Oikos 2004, 107, 148–160. [Google Scholar] [CrossRef]
  13. Frank, D.A.; Evans, R.D. Effects of native grazers on grassland N cycling in Yellowstone National Park. Ecology 1997, 78, 2238–2248. [Google Scholar] [CrossRef]
  14. McNaughton, S.J.; Zuniga, G.; McNaughton, M.M.; Banyikwa, F.F. Ecosystem catalysis: Soil urease activity and grazing in the ecosystem Serengeti. Oikos 1997, 80, 467–469. [Google Scholar] [CrossRef]
  15. Carvalheiro, L.G.; Kunin, W.E.; Keil, P.; Aguirre-Gutiérrez, J.; Ellis, W.N.; Fox, R.; Groom, Q.; Hennekens, S.; Van Landuyt, W.; Maes, D.; et al. Species richness declines and biotic homogenisation have slowed down for NW-European pollinators and plants. Ecol. Lett. 2013, 16, 870–878. [Google Scholar] [CrossRef] [Green Version]
  16. Rosa García, R.; Fraser, M.D.; Celaya, R.; Ferreira, L.M.M.; García, U.; Osoro, K. Grazing land management and biodiversity in the Atlantic European heathlands: A review. Agrofor. Syst. 2013, 87, 19–43. [Google Scholar] [CrossRef]
  17. Scohier, A.; Dumont, B. How do sheep affect plant communities and arthropod populations in temperate grasslands? Animal 2012, 6, 1129–1138. [Google Scholar] [CrossRef] [Green Version]
  18. Batalla, I.; Knudsen, M.T.; Mogensen, L.; Hierro, Ó.; Del Pinto, M.; Hermansen, J.E. Carbon footprint of milk from sheep farming systems in Northern Spain including soil carbon sequestration in grasslands. J. Clean. Prod. 2015, 104, 121–129. [Google Scholar] [CrossRef]
  19. Gerber, P.J.; Steinfeld, H.; Henderson, B.; Mottet, A.; Opio, C.; Dijkman, J.; Falcucci, A.; Tempio, G. Tackling Climate Change Through Livestock—A Global Assessment of Emissions and Mitigation Opportunities; Food and Agriculture Organization of the United Nations (FAO): Rome, Italy, 2013; ISBN 978-92-5-107920-1. [Google Scholar]
  20. Bernués, A.; Ruiz, R.; Olaizola, A.; Villalba, D.; Casasús, I. Sustainability of pasture-based livestock farming systems in the European Mediterranean context: Synergies and trade-offs. Livest. Sci. 2011, 139, 44–57. [Google Scholar] [CrossRef]
  21. Pe’er, G.; Dicks, L.V.; Visconti, P.; Arlettaz, R.; Báldi, A.; Benton, T.G.; Collins, S.; Dieterich, M.; Gregory, R.D.; Hartig, F.; et al. EU agricultural reform fails on biodiversity. Science 2014, 344, 1090–1092. [Google Scholar] [CrossRef]
  22. MacDonald, D.; Crabtree, J.R.; Wiesinger, G.; Dax, T.; Stamou, N.; Fleury, P.; Gutierrez Lazpita, J.; Gibon, A. Agricultural abandonment in mountain areas of Europe: Environmental consequences and policy response. J. Environ. Manag. 2000, 59, 47–69. [Google Scholar] [CrossRef]
  23. Lasanta, T.; Nadal-Romero, E.; Arnáez, J. Managing abandoned farmland to control the impact of re-vegetation on the environment. The state of the art in Europe. Environ. Sci. Policy 2015, 52, 99–109. [Google Scholar] [CrossRef] [Green Version]
  24. Díaz, S.; Noy-Meir, I.; Cabido, M. Can grazing response of herbaceous plants be predicted from simple vegetative traits? J. Appl. Ecol. 2001, 38, 497–508. [Google Scholar] [CrossRef] [Green Version]
  25. Strauss, S.Y.; Agrawal, A.A. The ecology and evolution of plant tolerance to herbivory. Trends Ecol. Evol. 1999, 14, 179–185. [Google Scholar] [CrossRef]
  26. Deléglise, C.; Meisser, M.; Mosimann, E.; Spiegelberger, T.; Signarbieux, C.; Jeangros, B.; Buttler, A. Drought-induced shifts in plants traits, yields and nutritive value under realistic grazing and mowing managements in a mountain grassland. Agric. Ecosyst. Environ. 2015, 213, 94–104. [Google Scholar] [CrossRef]
  27. Evju, M.; Austrheim, G.; Halvorsen, R.; Mysterud, A. Grazing responses in herbs in relation to herbivore selectivity and plant traits in an alpine ecosystem. Oecologia 2009, 161, 77–85. [Google Scholar] [CrossRef]
  28. Wilson, J.B. The twelve theories of co-existence in plant communities: The doubtful, the important and the unexplored. J. Veg. Sci. 2011, 22, 184–195. [Google Scholar] [CrossRef]
  29. Koutsoukis, C.; Roukos, C.; Demertzis, P.G.; Kandrelis, S.; Akrida-Demertzi, K. The variation of the chemical composition of the main plant species in a subalpine grassland in northwestern Greece. Legum. Sci. 2019, 1. [Google Scholar] [CrossRef]
  30. Darch, T.; McGrath, S.P.; Lee, M.R.F.; Beaumont, D.A.; Blackwell, M.S.A.; Horrocks, C.A.; Evans, J.; Storkey, J. The Mineral Composition of Wild-Type and Cultivated Varieties of Pasture Species. Agronomy 2020, 10, 1463. [Google Scholar] [CrossRef]
  31. Odriozola, I.; García-Baquero, G.; Etxeberria, A.; Aldezabal, A. Patterns of species relatedness created by competitive exclusion depend on species niche differences: Evidence from Iberian Atlantic grasslands. Perspect. Plant Ecol. Evol. Syst. 2017, 28, 36–46. [Google Scholar] [CrossRef]
  32. Tansley, A.G.; Adamson, R.S. Studies of the vegetation of the English chalk: III. The chalk grasslands of Hampshire-Sussex border. J. Ecol. 1925, 13, 177–223. [Google Scholar] [CrossRef] [Green Version]
  33. Pearl, J. Causal inference in statistics: An overview. Stat. Surv. 2009, 3, 96–146. [Google Scholar] [CrossRef]
  34. Pearl, J. Interpretation and identification of causal mediation. Psychol. Methods 2014, 19, 459–481. [Google Scholar] [CrossRef] [PubMed]
  35. Tahmasebi Kohyani, P.; Bossuyt, B.; Bonte, D.; Hoffmann, M. Grazing as a management tool in dune grasslands: Evidence of soil and scale dependence of the effect of large herbivores on plant diversity. Biol. Conserv. 2008, 141, 1687–1694. [Google Scholar] [CrossRef]
  36. Stohlgren, T.J.; Schell, L.D.; Vanden Heuvel, B. How grazing and soil quality affect native and exotic plant diversity in Rocky Mountain grasslands. Ecol. Appl. 1999, 9, 45–64. [Google Scholar] [CrossRef]
  37. Hulshof, C.M.; Spasojevic, M.J. The edaphic control of plant diversity. Glob. Ecol. Biogeogr. 2020, 1–16. [Google Scholar] [CrossRef]
  38. Frank, D.A. The interactive effects of grazing ungulates and aboveground production on grassland diversity. Oecologia 2005, 143, 629–634. [Google Scholar] [CrossRef]
  39. Aldezabal, A.; Odriozola, I.; García-Baquero, G. Grazing Abandonment Delays the Effect of Temperature on Aboveground Net Primary Production in Atlantic Grasslands. Rangel. Ecol. Manag. 2019, 72, 822–831. [Google Scholar] [CrossRef]
  40. Casella, G. Statistical Design; Casella, G., Fienberg, S., Olkin, I., Eds.; Springer: Gainesville, FL, USA, 2008; ISBN 9780387759647. [Google Scholar]
  41. Higgins, K.F.; Oldemeyer, J.L.; Jenkins, K.J.; Clambey, G.K.; Harlow, R.F. Vegetation sampling and measurement. In Research and Management Techniques for Wildlife and Habitats; Bookhout, T.A., Ed.; The Wildlife Society: Bethesda, MD, USA, 1994; pp. 567–591. ISBN 9780935868814. [Google Scholar]
  42. Riveros, E.; Argamentería, A. Métodos encimáticos de la predicción de la digestibilidad in vivo de la materia orgánica de forrajes. 1. Forrajes verdes. 2. Ensilados y pajas. Av. En Prod. Anim. 1987, 12, 49–75. [Google Scholar]
  43. Jackson, E.; Farrington, D.S.; Henderson, K. The Analysis of Agricultural Materials: A Manual of the Analytical Methods Used by the Agricultural Development and Advisory Service; Ministry of Agriculture, Fisheries and Food: London, UK, 1986; ISBN 0-11-242762-6.
  44. Tukey, J.W. Exploratory Data Analysis; Addison-Wesley Publishing Company: Reading, MA, USA, 1977. [Google Scholar]
  45. Pinheiro, J.C.; Bates, D.M. Mixed-Effects Models in S and S-PLUS; Springer: Berlin, Germany, 2000. [Google Scholar]
  46. Pinheiro, J.; Bates, D.; DebRoy, S.; Sarkar, D.; Team, R.C. Nlme: Linear and Nonlinear Mixed Effects Models. R Package Version 3.1-147. 2020. Available online: http://CRAN.R-project.org/package=nlme (accessed on 10 December 2020).
  47. R Core Team. R: A Language and Environment for Statistical Computing; R Core Team: Vienna, Austria, 2017. [Google Scholar]
  48. Barton, K. MuMIn: Multi-Model Inference. R Package Version 1.43.17. 2020. Available online: http://CRAN.R-project.org/package=MuMIn (accessed on 15 April 2020).
  49. Lüdecke, D. sjPlot: Data Visualization for Statistics in Social Science. R package version 2.8.4. 2018. Available online: http://CRAN.R-project.org/package=sjPlot (accessed on 28 October 2020).
  50. Tingley, D.; Yamamoto, T.; Hirose, K.; Keele, L.; Imai, K. Mediation: R Package for Causal Mediation Analysis. J. Stat. Softw. 2014, 59, 1–38. [Google Scholar] [CrossRef] [Green Version]
  51. Imai, K.; Keele, L.; Tingley, D. A General Approach to Causal Mediation Analysis. Psychol. Methods 2010, 15, 309–334. [Google Scholar] [CrossRef] [Green Version]
  52. Imai, K.; Yamamoto, T. Identification and Sensitivity Analysis for Multiple Causal Mechanisms: Revisiting Evidence from Framing Experiments. Polit. Anal. 2013, 21, 141–171. [Google Scholar] [CrossRef]
  53. Simpson, E.H. Measurment of Diversity. Nature 1949, 163, 688. [Google Scholar] [CrossRef]
  54. Robins, J.M.; Greenland, S. Identifiability and exchangeability for direct and indirect effects. Epidemiology 1992, 3, 143–155. [Google Scholar] [CrossRef] [Green Version]
  55. Pearl, J. Direct and indirect effects. In Proceedings of the Seventeenth Conference on Uncertainty in Artificial Intelligence; Morgan Kaufmann Publishers Inc.: Seattle, WA, USA, 2001; pp. 411–420. [Google Scholar]
  56. Fisher, R.A. Statistical Methods, Experimental Design and Scientific Inference. A Re-Issue of Statistical Methods for Research Workers, The Design of Experiments, and Statistical Methods and Scientific Inference; Oxford University Press: Oxford, UK, 1990. [Google Scholar]
  57. Imai, K.; Keele, L.; Tingley, D.; Yamamoto, T. Unpacking the Black Box of Causality: Learning about Causal Mechanisms from Experimental and Observational Studies. Am. Polit. Sci. Rev. 2011, 105, 765–789. [Google Scholar] [CrossRef]
  58. Chapman, D.F.; Lemaire, G. Tissues flows in grazed plant communities. In The Ecology and Management of Grazing Systems; Hodgson, J.G., Illius, A.W., Eds.; CAB International: Guilford, UK, 1996; pp. 3–36. [Google Scholar]
  59. Pavlů, V.; Hejcman, M.; Pavlů, L.; Gaisler, J.; Nežerková, P. Effect of continuous grazing on forage quality, quantity and animal performance. Agric. Ecosyst. Environ. 2006, 113, 349–355. [Google Scholar] [CrossRef]
  60. Cruz, P.; De Quadros, F.L.F.; Theau, J.P.; Frizzo, A.; Jouany, C.; Duru, M.; Carvalho, P.C.F. Leaf Traits as Functional Descriptors of the Intensity of Continuous Grazing in Native Grasslands in the South of Brazil. Rangel. Ecol. Manag. 2010, 63, 350–358. [Google Scholar] [CrossRef]
  61. Aldezabal, A.; Zubeldia, I.; Mandaluniz, N.; Laskurain, N.A. Estudio de la evolución temporal y los cambios paisajísticos en los pastos de montaña de interés comunitario del País Vasco. In Pastos y PAC 2014-2020; Busqué, J., Salcedo, G., Serrano, E., Mora, M.J., Fernández, B., Eds.; S.E.E.P.: Potes, Spain, 2014; pp. 81–88. ISBN 978-84-697-0561-2. [Google Scholar]
  62. Lezama, F.; Paruelo, J.M. Disentangling grazing effects: Trampling, defoliation and urine deposition. Appl. Veg. Sci. 2016, 19, 557–566. [Google Scholar] [CrossRef]
  63. Pontes, L.D.S.; Louault, F.; Carrère, P.; Maire, V.; Andueza, D.; Soussana, J.F. The role of plant traits and their plasticity in the response of pasture grasses to nutrients and cutting frequency. Ann. Bot. 2010, 105, 957–965. [Google Scholar] [CrossRef] [Green Version]
  64. Pontes, L.D.S.; Soussana, J.F.; Louault, F.; Andueza, D.; Carrère, P. Leaf traits affect the above-ground productivity and quality of pasture grasses. Funct. Ecol. 2007, 21, 844–853. [Google Scholar] [CrossRef]
  65. Milcu, A.; Gessler, A.; Roscher, C.; Rose, L.; Kayler, Z.; Bachmann, D.; Pirhofer-Walzl, K.; Zavadlav, S.; Galiano, L.; Buchmann, T.; et al. Top canopy nitrogen allocation linked to increased grassland carbon uptake in stands of varying species richness. Sci. Rep. 2017, 7, 1–11. [Google Scholar] [CrossRef]
  66. Van Soest, P.J. Plant, Animal and Environment. In Nutritional Ecology of the Ruminant; Cornell University Press: London, UK, 1994; pp. 77–92. [Google Scholar]
  67. Cingolani, A.M.; Noy-Meir, I.; Díaz, S. Grazing Effects on Rangeland Diversity: A Synthesis of Contemporary Models. Ecol. Appl. 2005, 15, 757–773. [Google Scholar] [CrossRef]
  68. Milchunas, D.G.; Sala, O.E.; Lauenroth, W.K. A generalized model of the effects of grazing by large herbivores on grassland community structure. Am. Nat. 1988, 132, 87–106. [Google Scholar] [CrossRef]
  69. Olff, H.; Ritchie, M.E. Effects of herbivores on grassland plant diversity. Trends Ecol. Evol. 1998, 13, 261–265. [Google Scholar] [CrossRef] [Green Version]
  70. Bullock, J.M.; Hill, B.C.; Silvertown, J.; Sutton, M. Gap Colonization as a Source of Grassland Community Change: Effects of Gap Size and Grazing on the Rate and Mode of Colonization by Different Species. Oikos 1995, 72, 273–282. [Google Scholar] [CrossRef]
  71. Lee, M.A.; Davis, A.P.; Chagunda, M.G.G.; Manning, P. Forage quality declines with rising temperatures, with implications for livestock production and methane emissions. Biogeosciences 2017, 14, 1403–1417. [Google Scholar] [CrossRef] [Green Version]
  72. Waghorn, G.C.; Clark, D.A. Feeding value of pastures for ruminants. N. Z. Vet. J. 2004, 52, 320–331. [Google Scholar] [CrossRef]
  73. Dubljević, R.; Marković, B.; Radonjić, D.; Stešević, D.; Marković, M. Influence of changes in botanical diversity and quality of wet grasslands through phenological phases on cow milk fatty acid composition. Sustainability 2020, 12, 6320. [Google Scholar] [CrossRef]
  74. Bravo-Lamas, L.; Aldai, N.; Kramer, J.K.G.; Barron, L.J.R. Case study using commercial dairy sheep flocks: Comparison of the fat nutritional quality of milk produced in mountain and valley farms. LWT Food Sci. Technol. 2018, 89, 374–380. [Google Scholar] [CrossRef]
Agronomy 11 00025 g001 550
Figure 1. Directed acyclic graph representing direct and indirect (via concurrent change in floristic composition) effects of grazing exclusion (X) on forage nutritive quality (Y). Floristic composition acts as a mediating variable (M). The variable(s) Z denotes potential confounders of the mediator M and the outcome Y. Grazing exclusion (X) is a randomized experimental factor, and so stands as a parent node with no incoming arrows. The path from X to Y (X → Y) represents the direct effects and the path from X to Y through M (X → M → Y) represents the indirect effects, once that the potential confounding effects of the Z variables have been controlled for.
Figure 1. Directed acyclic graph representing direct and indirect (via concurrent change in floristic composition) effects of grazing exclusion (X) on forage nutritive quality (Y). Floristic composition acts as a mediating variable (M). The variable(s) Z denotes potential confounders of the mediator M and the outcome Y. Grazing exclusion (X) is a randomized experimental factor, and so stands as a parent node with no incoming arrows. The path from X to Y (X → Y) represents the direct effects and the path from X to Y through M (X → M → Y) represents the indirect effects, once that the potential confounding effects of the Z variables have been controlled for.
Agronomy 11 00025 g001
Agronomy 11 00025 g002 550
Figure 2. (a) Modeled change in crude protein; (b) Modeled change in cation calcium content. The patterns for the change in Mg, K, P and enzymatic digestibility are similar to that for crude protein, whereas the pattern for neutral detergent fiber is just the inverse of the pattern for enzymatic digestibility. Ash does not change over time (see Supplementary Figure S3). Time is indexed in months, starting in June 2005 (month number 1) and ending in October 2010 (month number 65). Restricted maximum likelihood estimates (ReML) for cation calcium content and crude protein at the end of each season (October) over the period 2005–2010 are presented in Supplementary Table S4.
Figure 2. (a) Modeled change in crude protein; (b) Modeled change in cation calcium content. The patterns for the change in Mg, K, P and enzymatic digestibility are similar to that for crude protein, whereas the pattern for neutral detergent fiber is just the inverse of the pattern for enzymatic digestibility. Ash does not change over time (see Supplementary Figure S3). Time is indexed in months, starting in June 2005 (month number 1) and ending in October 2010 (month number 65). Restricted maximum likelihood estimates (ReML) for cation calcium content and crude protein at the end of each season (October) over the period 2005–2010 are presented in Supplementary Table S4.
Agronomy 11 00025 g002
Agronomy 11 00025 g003 550
Figure 3. (a) Change in Simpson’s (plant) diversity index; (b) change in legume-to-grass ratio content. The observed patterns for the change in the proportion of dicots is similar to that for legume-to-grass ratio (see Supplementary Figure S4). Time is indexed in months, starting in June 2005 (month number 1) and ending in October 2010 (month number 65). Restricted maximum likelihood estimates (ReML) for Simpson’s index and legume-to-grass ratio at the end of each season (October) over the period 2005–2010 are presented in Supplementary Table S4.
Figure 3. (a) Change in Simpson’s (plant) diversity index; (b) change in legume-to-grass ratio content. The observed patterns for the change in the proportion of dicots is similar to that for legume-to-grass ratio (see Supplementary Figure S4). Time is indexed in months, starting in June 2005 (month number 1) and ending in October 2010 (month number 65). Restricted maximum likelihood estimates (ReML) for Simpson’s index and legume-to-grass ratio at the end of each season (October) over the period 2005–2010 are presented in Supplementary Table S4.
Agronomy 11 00025 g003
Agronomy 11 00025 g004 550
Figure 4. Direct (paths from X to Y) and indirect effects (paths from X to Y passing through M) of experimental grazing exclusion (X) on descriptors of forage nutritive quality (Y), once that topsoil properties (Z) have been taken into account. (a) Simultaneous change in the legume-to-grass ratio mediates 80% of change in crude protein content and, (b) 55% of change in enzymatic digestibility; (c) concurrent change in the proportion of dicots mediates 31% of change in cation calcium forage content.
Figure 4. Direct (paths from X to Y) and indirect effects (paths from X to Y passing through M) of experimental grazing exclusion (X) on descriptors of forage nutritive quality (Y), once that topsoil properties (Z) have been taken into account. (a) Simultaneous change in the legume-to-grass ratio mediates 80% of change in crude protein content and, (b) 55% of change in enzymatic digestibility; (c) concurrent change in the proportion of dicots mediates 31% of change in cation calcium forage content.
Agronomy 11 00025 g004
Table
Table 1. Hypothesis testing results, after linear mixed modelling, for eight forage quality properties and three derived floristic composition variables measured in a repeated-measurements (2005–2010) RCB experiment, implemented to simulate the effects of grazing exclusion on forage quality. Columns 2–4 report F-values, together with degrees of freedom (among parentheses); p-values are indicated with asterisks. Degrees of freedom differ among variables because, in some cases, there are missing values. Parameter estimation is reported in Table 2 and illustrated in Figure 2 and Figure 3, Figures S3 and S4. FF = Fixed Factors; RF = Random Factor.
Table 1. Hypothesis testing results, after linear mixed modelling, for eight forage quality properties and three derived floristic composition variables measured in a repeated-measurements (2005–2010) RCB experiment, implemented to simulate the effects of grazing exclusion on forage quality. Columns 2–4 report F-values, together with degrees of freedom (among parentheses); p-values are indicated with asterisks. Degrees of freedom differ among variables because, in some cases, there are missing values. Parameter estimation is reported in Table 2 and illustrated in Figure 2 and Figure 3, Figures S3 and S4. FF = Fixed Factors; RF = Random Factor.
TreatmentTimeTreat. × TimeFF R2RF R2
Ash26.40 (2, 333) ***0.39 (1, 333)n.s.0.1070.211
Ca18.44 (2, 331) ***28.14 (1, 331) ***n.s.0.1610.000
Mg39.14 (2, 331) ***0.54 (2, 331)14.26 (2, 331) ***0.2100.130
K18.77 (2, 331) ***2.01 (2, 331)16.49 (2, 331) ***0.1570.113
P26.71 (2, 331) ***9.45 (1, 331) **8.64 (2, 331) ***0.1370.285
CP88.62 (2, 331) ***2.90 (1, 331)17.83 (2, 331) ***0.3810.022
NDF16.95 (2, 331) ***0.35 (1, 331)8.10 (2, 331) ***0.1280.014
NDFcel18.44 (2, 329) ***1.59 (1, 329)13.08 (2, 329) ***0.1560.030
Simpson’s index0.69 (2, 329)22.84 (1, 329) ***7.86 (2, 329) ***0.0910.140
Legume-to-grass ratio20.78 (2, 329) ***7.69 (1, 329) **40.20 (2, 329) ***0.2280.182
Dicots proportion21.60 (2, 329) ***1.50 (1, 329)27.81 (2, 329) ***0.1490.353
n.s., p > 0.05; **, p < 0.01; ***, p < 0.001.
Table
Table 2. Restricted maximum likelihood parameter estimation for the forage quality properties and floristic composition measured in four field sites during 2005–2010. The parameter σb corresponds to the variability due to the random factor Site; the parameter σ corresponds to the residual variability. Cage vs. grazing has been omitted, since, in most cases, these effects are not significant (Supplementary Table S3). Hypothesis testing is reported in Table 1.
Table 2. Restricted maximum likelihood parameter estimation for the forage quality properties and floristic composition measured in four field sites during 2005–2010. The parameter σb corresponds to the variability due to the random factor Site; the parameter σ corresponds to the residual variability. Cage vs. grazing has been omitted, since, in most cases, these effects are not significant (Supplementary Table S3). Hypothesis testing is reported in Table 1.
VariableLevel of Reference: Grazing
(<intercept>)		Exclosure vs. GrazingTimeExclosure × Timeσbσ
Ash7.96−1.65--1.162.08
Ca0.62−0.09−0.003-0.000.20
Mg0.14−0.010.000−0.0010.020.03
K0.940.180.001−0.0080.110.29
P0.24−0.01−0.0003−0.00050.030.04
CP13.12−0.820.01−0.060.382.02
NDF67.550.24−0.040.090.594.62
NDFcel42.640.510.02−0.110.924.77
Simpson’s index0.710.050.001−0.0010.030.07
Legume-to-grass ratio0.140.090.002−0.0050.060.11
Dicots proportion0.310.040.001−0.0030.060.08
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

García-Baquero, G.; Odriozola, I.; Aldezabal, A. Floristic Composition Mediates Change in Forage Nutritive Quality of Atlantic Mountain Grasslands after Experimental Grazing Exclusion. Agronomy 2021, 11, 25. https://doi.org/10.3390/agronomy11010025

AMA Style

García-Baquero G, Odriozola I, Aldezabal A. Floristic Composition Mediates Change in Forage Nutritive Quality of Atlantic Mountain Grasslands after Experimental Grazing Exclusion. Agronomy. 2021; 11(1):25. https://doi.org/10.3390/agronomy11010025

Chicago/Turabian Style

García-Baquero, Gonzalo, Iñaki Odriozola, and Arantza Aldezabal. 2021. "Floristic Composition Mediates Change in Forage Nutritive Quality of Atlantic Mountain Grasslands after Experimental Grazing Exclusion" Agronomy 11, no. 1: 25. https://doi.org/10.3390/agronomy11010025

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop