Links soil, grazing indicators and vegetation repartition
September 2024
Data soil exploration
We collected three soil samples at 2, 5 and 8 meters of the 10-m-length transect. We did so for 62 vegetation transects done during spring 2022 and 2024.
We measured the following soil physico-chemical variables from the soil sampling:
Dataset features explained
wetanddry weightphweightand volume of stones (v_stones) within the cylinders. We have the weight just for the soil sampling of 2024.bulk.density: Soil bulk density (\(\frac{g}{cm^{3}}\))depth: Soil depth (cm)soil compaction: just for the soil sampling of 2024.HCL Test:Limestone : In cold condition, generalized effervescence under HCl
Dolomite : In cold condition, little or non effervescence under HCL
Four categories from 0 to 3 indicating effervescence under HCL
0 : No detectable reaction
1 : Weak, a few bubbles or murmur of effervescence
2 : Medium, bubbles in continuous layer
3 : Vivid, several superimposed layers of bubbles
Munsell soil colour :
hue: basic colorChroma: the strength or weakness of a colourValue: the lightness or the darkness of the colourdescription color: name of the color
om: Percentage (%) of organic matter calculated after being in muffle furnace at 550 °C during 4 hours.N: Percentage (%) of nitrogen in a sample weighing ~ 35 mg. CHN analysis.C: Percentage (%) of carbon (organic + inorganic) in a sample weighing ~ 35 mg. CHN analysis.C:N ratios
For the statistical analysis we kept the variables that are presented in the following section.
Data summary of variable used in the statistics
Here we show the summary of the variables that were already averaged by replicates by transect. In brief, in the dataset we have 62 values of each soil variable.
Also we recorded litter, stones,
rock and moss in the field along the
vegetation transect.
## ph bulk_density soil_depth hcl_category
## Min. :6.480 Min. :0.5000 Min. : 5.80 Min. :0.0000
## 1st Qu.:6.822 1st Qu.:0.9000 1st Qu.: 11.70 1st Qu.:0.0000
## Median :7.015 Median :1.0000 Median : 16.50 Median :0.7000
## Mean :7.054 Mean :0.9887 Mean : 23.99 Mean :0.6806
## 3rd Qu.:7.265 3rd Qu.:1.1000 3rd Qu.: 25.45 3rd Qu.:1.0000
## Max. :7.710 Max. :1.6000 Max. :101.00 Max. :3.0000
## value N C_mg om
## Min. :2.000 Min. :0.1930 Min. :0.840 Min. : 5.470
## 1st Qu.:3.000 1st Qu.:0.3415 1st Qu.:1.480 1st Qu.: 9.153
## Median :3.000 Median :0.4535 Median :2.015 Median :11.521
## Mean :3.319 Mean :0.5059 Mean :2.272 Mean :12.691
## 3rd Qu.:4.000 3rd Qu.:0.7000 3rd Qu.:2.868 3rd Qu.:17.019
## Max. :5.700 Max. :1.0970 Max. :4.800 Max. :26.191
## v_stones C_N lit ston
## Min. : 1.00 Min. : 9.53 Min. : 4.00 Min. : 0.00
## 1st Qu.: 7.30 1st Qu.:10.62 1st Qu.:14.00 1st Qu.:10.25
## Median : 10.85 Median :11.18 Median :17.00 Median :14.50
## Mean : 24.72 Mean :14.27 Mean :18.94 Mean :16.11
## 3rd Qu.: 41.83 3rd Qu.:12.29 3rd Qu.:22.75 3rd Qu.:21.00
## Max. :123.70 Max. :53.33 Max. :37.00 Max. :42.00
## rock moss
## Min. : 0.000 Min. : 0.000
## 1st Qu.: 1.000 1st Qu.: 1.000
## Median : 4.000 Median : 3.500
## Mean : 5.258 Mean : 6.129
## 3rd Qu.: 9.000 3rd Qu.:10.500
## Max. :20.000 Max. :24.000Soil correlation matrix
Spearman correlation
## ph bulk_density soil_depth hcl_category value N C_mg om
## ph 1
## bulk_density 0.42 1
## soil_depth 0.13 -0.04 1
## hcl_category 0.48 0.34 0.01 1
## value 0.27 0.44 0.17 0.1 1
## N -0.26 -0.49 -0.37 -0.01 -0.7 1
## C_mg 0.18 -0.19 -0.25 0.35 -0.24 0.62 1
## om -0.27 -0.53 -0.33 -0.06 -0.71 0.98 0.58 1
## v_stones -0.03 -0.18 -0.32 -0.06 -0.03 0.2 0.22 0.16
## C_N 0.49 0.38 0.21 0.34 0.43 -0.5 0.13 -0.46
## lit -0.04 -0.07 -0.09 -0.21 -0.02 -0.05 -0.02 -0.07
## ston 0.26 0.17 -0.25 0.37 0.02 0.18 0.13 0.17
## rock -0.3 -0.38 0.02 -0.43 -0.02 0.11 -0.07 0.12
## moss -0.32 -0.44 -0.04 -0.39 -0.01 0.04 -0.05 0.08
## v_stones C_N lit ston rock moss
## ph
## bulk_density
## soil_depth
## hcl_category
## value
## N
## C_mg
## om
## v_stones 1
## C_N -0.03 1
## lit 0.03 0.03 1
## ston 0.43 0.01 -0.34 1
## rock 0.24 -0.26 0.05 -0.14 1
## moss 0.2 -0.23 0.07 -0.3 0.26 1Following criterion of |r| > 0.7 for avoid collinearity (Dormann et al. 2013), organic matter is highly correlated to Value and Nitrogen so we removed these two latter variables for the multivariate analysis.
PCA of soil variables
## $importance
## Importance of components:
## PC1 PC2 PC3 PC4 PC5 PC6 PC7
## Eigenvalue 3.1793 2.028 1.4468 1.2062 1.03903 0.87470 0.75712
## Proportion Explained 0.2649 0.169 0.1206 0.1005 0.08659 0.07289 0.06309
## Cumulative Proportion 0.2649 0.434 0.5545 0.6550 0.74162 0.81452 0.87761
## PC8 PC9 PC10 PC11 PC12
## Eigenvalue 0.53058 0.39662 0.27753 0.22912 0.034843
## Proportion Explained 0.04422 0.03305 0.02313 0.01909 0.002904
## Cumulative Proportion 0.92182 0.95488 0.97800 0.99710 1.000000Principal component one explains 26.5%, component two explains 16.9%, and component three explains 12.1% of the variance.
Variables and principal components 1-2
Soil variables accros CSP commons
Commons and principal Components 1-2
PERMANOVA
## Permutation test for adonis under reduced model
## Terms added sequentially (first to last)
## Permutation: free
## Number of permutations: 999
##
## adonis2(formula = euclidean_dist_soil ~ common, data = inv_scores_c, permutations = 999)
## Df SumOfSqs R2 F Pr(>F)
## common 2 8278 0.09785 3.1997 0.006 **
## Residual 59 76317 0.90215
## Total 61 84595 1.00000
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## $parent_call
## [1] "euclidean_dist_soil ~ common , strata = Null , permutations 999"
##
## $C_vs_P
## Df SumOfSqs R2 F Pr(>F)
## common 1 5518 0.12025 4.3741 0.01 **
## Residual 32 40371 0.87975
## Total 33 45890 1.00000
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## $C_vs_S
## Df SumOfSqs R2 F Pr(>F)
## common 1 6501 0.09523 4.6313 0.007 **
## Residual 44 61764 0.90477
## Total 45 68265 1.00000
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## $P_vs_S
## Df SumOfSqs R2 F Pr(>F)
## common 1 603 0.01181 0.5018 0.655
## Residual 42 50500 0.98819
## Total 43 51103 1.00000
##
## attr(,"class")
## [1] "pwadstrata" "list"Bonferroni adjustment of p-values
p_value = c(pairwise.c_2$C_vs_P$`Pr(>F)`[1],pairwise.c_2$C_vs_S$`Pr(>F)`[1],pairwise.c_2$P_vs_S$`Pr(>F)`[1])
p.adjust(p_value, method = "bonferroni")## [1] 0.030 0.021 1.000Significant statistical differences between Castril-Pontones and Castril-Santiago
Statistical differences by variable accros commons
Differences of organic matter (om) and soil depth across commons
##
## Kruskal-Wallis rank sum test
##
## data: om by common
## Kruskal-Wallis chi-squared = 7.5298, df = 2, p-value = 0.02317## # A tibble: 3 × 9
## .y. group1 group2 n1 n2 statistic p p.adj p.adj.signif
## * <chr> <chr> <chr> <int> <int> <dbl> <dbl> <dbl> <chr>
## 1 om C P 18 16 1.71 0.0882 0.265 ns
## 2 om C S 18 28 2.73 0.00636 0.0191 *
## 3 om P S 16 28 0.761 0.447 1 ns
##
## Kruskal-Wallis rank sum test
##
## data: soil_depth by common
## Kruskal-Wallis chi-squared = 14.291, df = 2, p-value = 0.0007883
## # A tibble: 3 × 9
## .y. group1 group2 n1 n2 statistic p p.adj p.adj.signif
## * <chr> <chr> <chr> <int> <int> <dbl> <dbl> <dbl> <chr>
## 1 soil_depth C P 18 16 -3.48 0.000503 0.00151 **
## 2 soil_depth C S 18 28 -3.11 0.00186 0.00559 **
## 3 soil_depth P S 16 28 0.816 0.415 1 nsNo differences in the following variables
##
## Kruskal-Wallis rank sum test
##
## data: v_stones by common
## Kruskal-Wallis chi-squared = 0.86959, df = 2, p-value = 0.6474##
## Kruskal-Wallis rank sum test
##
## data: ston by common
## Kruskal-Wallis chi-squared = 0.51873, df = 2, p-value = 0.7715##
## Kruskal-Wallis rank sum test
##
## data: C by common
## Kruskal-Wallis chi-squared = 3.9798, df = 2, p-value = 0.1367##
## Kruskal-Wallis rank sum test
##
## data: hcl_category by common
## Kruskal-Wallis chi-squared = 0.80646, df = 2, p-value = 0.6682##
## Kruskal-Wallis rank sum test
##
## data: C_N by common
## Kruskal-Wallis chi-squared = 3.3949, df = 2, p-value = 0.1831##
## Kruskal-Wallis rank sum test
##
## data: bulk_density by common
## Kruskal-Wallis chi-squared = 0.28394, df = 2, p-value = 0.8676##
## Kruskal-Wallis rank sum test
##
## data: ph by common
## Kruskal-Wallis chi-squared = 3.4696, df = 2, p-value = 0.1764Soil variables accros transhumance modalities
Transhumance modalities and principal Components 1-2
PERMANOVA
## Permutation test for adonis under reduced model
## Terms added sequentially (first to last)
## Permutation: free
## Number of permutations: 999
##
## adonis2(formula = euclidean_dist_soil ~ thr_cat2, data = inv_scores_c, permutations = 999)
## Df SumOfSqs R2 F Pr(>F)
## thr_cat2 2 2741 0.0324 0.9879 0.397
## Residual 59 81854 0.9676
## Total 61 84595 1.0000There are not significant statistical differences concerning transhumance modalities
Grazing indicators
Materials and methods
Field level grazing indicators
Indicators measured at the field level, i.e.nearby or along the 72 vegetation transects carried out during spring 2022 and 2023.
Plant Utilization Rates (PUR): We followed the method described by Ruiz-Mirazo et al. (2011) which proposes qualitative categories of plant utilization ranging from 0 to 5, with 3 referring to intermediate utilization. PUR have been measured nearby the vegetation transect in 25 individuals of dominant species, i.e. Helianthemum cinereum, Helianthemum oelandicum, Helianthemum appenninum, Thymus serpylloides, Koeleria vallesiana, Seseli montanum and Festuca segimonensis). Also we measured an overall Plant Utilisation Rate (PUR) of the vegetation
pur_RA, which was measured at each meter of the 10m transect. For the statistical analysis we aggregated species with similar PUR. We generated two categories;low palatablespecies (PUR < 2) including Koeleria vallesiana, Seseli montanum and Festuca segimonensis, andhigh palatablespecies (PUR >4) including species of Helianthemum spp. and Thymus serpylloides.pl_density: total number of plant contacts in the transects. Plant density allows evaluating grazing effects (Castañeda et al., 2023).n_flower: total number of plant with flowers recorded in the transect. Flower abundance has been negatively associated with grazing pressure (Tadey et al. 2015).bare: Bare ground recorded in the transect. Bare ground can be an indicator of herbivores’ activity and is positively correlated with runoff and soil erosion (Pyke et al. 2002).dung: Dung measured at the beginning (dung_sp) and at the end (dung_au) of the grazing period. We employed quadrats of 50 x 50 cm to measure the dung amount. The quadrats were located adjacent to the 10m transect every 1 meter so carrying out 10 dung amount measurement. Dung can be a measure of grazing pressure on the transect scale (Jordan et al. 2022).
Individual grazing area level
Stocking Rate,which refers to the number of livestock per hectare over a specified time (Allen et al. 2011), is usually used to asses the effect of livestock on vegetation. Here we used animals*days per hectare (Scarnecchia 1985), which consider the number of days that the comarca is grazed. We gather such informations through the observations in field of Pau and Francisco and our observations during 2022 and 2023 fieldwork. Also, I check the transcriptions of the interviews done by me and Adrià during 2023.
\[\text{Stocking Rate} = \frac{Animals * Days}{Hectares}\]
However, this indicator may not be suitable in case of environmental and vegetation heterogeneity, also it does not take into account herding practices at the local scale (Genin and Hanafi 2010).
1. Which are the relationship between grazing indicators ?
Spearman correlation between indicators.
Correlation matrix:
## stoking_rate_2 pl_density pur_RA pur_high_palatable
## stoking_rate_2 1
## pl_density -0.12 1
## pur_RA 0.27 -0.16 1
## pur_high_palatable 0.25 0.19 0.37 1
## pur_low_palatable 0.06 -0.27 0.26 -0.51
## dung_au 0.21 -0.02 0 0.17
## dung_sp 0.01 -0.09 0.16 0.06
## n_flower -0.1 0.17 -0.08 -0.18
## bare 0 -0.53 0.12 -0.2
## pur_low_palatable dung_au dung_sp n_flower bare
## stoking_rate_2
## pl_density
## pur_RA
## pur_high_palatable
## pur_low_palatable 1
## dung_au -0.25 1
## dung_sp 0.12 0.42 1
## n_flower 0.03 -0.22 -0.05 1
## bare 0.25 -0.13 0.03 -0.12 1The indicators are not highly correlated between them. As they are not highly correlated, we do not have the issue of multicollinearity so we can keep them all to the PCA analysis.
PCA of selected soil variables and grazing indicators
Selected soil variables : bulk density, C:N, organic matter, soil depth, volumes stones, HCL category
# do not subset db to keep rownames
soil_ind_pca_1 <- rda(db_soil_ind, scale = TRUE) #scale=TRUE calls for a standardization of the variables
sum_2 = summary(soil_ind_pca_1) # Default scaling 2
sum_2$cont## $importance
## Importance of components:
## PC1 PC2 PC3 PC4 PC5 PC6 PC7
## Eigenvalue 3.181 1.9810 1.9049 1.4354 1.24229 1.18814 0.92007
## Proportion Explained 0.212 0.1321 0.1270 0.0957 0.08282 0.07921 0.06134
## Cumulative Proportion 0.212 0.3441 0.4711 0.5668 0.64962 0.72883 0.79016
## PC8 PC9 PC10 PC11 PC12 PC13 PC14
## Eigenvalue 0.64369 0.6120 0.54609 0.42428 0.36763 0.2655 0.18688
## Proportion Explained 0.04291 0.0408 0.03641 0.02829 0.02451 0.0177 0.01246
## Cumulative Proportion 0.83308 0.8739 0.91028 0.93857 0.96308 0.9808 0.99324
## PC15
## Eigenvalue 0.101431
## Proportion Explained 0.006762
## Cumulative Proportion 1.000000Principal component one explains 21.2%, component two explains 13.2%, and component three explains 12.7% of the variance.
Variables and principal components 1-2
Soil variables in red. Grazing indicators in blue.
Variables and principal components 1-3
Soil variables in red. Grazing indicators in blue.
distance-based Redundancy Analysis (db-RDA): Modelling of grazing indicators + soil variables and vegetation data
# similar of grazing_indicator rmd
# capscale() with raw (site by species) data (Borcard et al. 2018)
rda_soil_ind <- capscale(community_data_density_22_23.72_sin_cus_filt ~ bulk_density+ soil_depth + om+ C_N + v_stones +hcl_category +
stoking_rate_2 + pl_density + pur_RA + pur_high_palatable + pur_low_palatable + dung_au + dung_sp + n_flower + bare +
Condition(scores(as.matrix(coord_22_23))), data = db_soil_ind, distance = "bray",add = "lingoes")Variability explained by the first three axes.
## CAP1 CAP2 CAP3
## 6.715658 5.953246 3.834516Variance Inflation Factor (VIF)
## scores(as.matrix(coord_22_23))x scores(as.matrix(coord_22_23))y
## 2.011135 1.867628
## bulk_density soil_depth
## 4.191210 1.848894
## om C_N
## 4.136944 2.056068
## v_stones hcl_category
## 1.743875 1.937498
## stoking_rate_2 pl_density
## 1.349158 2.113002
## pur_RA pur_high_palatable
## 3.057139 3.138775
## pur_low_palatable dung_au
## 2.911266 1.767354
## dung_sp n_flower
## 1.723867 1.650024
## bare
## 2.084939Test the significance of the model
## Permutation test for capscale under reduced model
## Forward tests for axes
## Permutation: free
## Number of permutations: 999
##
## Model: capscale(formula = community_data_density_22_23.72_sin_cus_filt ~ bulk_density + soil_depth + om + C_N + v_stones + hcl_category + stoking_rate_2 + pl_density + pur_RA + pur_high_palatable + pur_low_palatable + dung_au + dung_sp + n_flower + bare + Condition(scores(as.matrix(coord_22_23))), data = db_soil_ind, distance = "bray", add = "lingoes")
## Df SumOfSqs F Pr(>F)
## CAP1 1 1.3498 4.8710 0.002 **
## CAP2 1 1.1965 4.3180 0.001 ***
## CAP3 1 0.7707 2.7813 0.069 .
## CAP4 1 0.5571 2.0106 0.566
## CAP5 1 0.5051 1.8229 0.713
## CAP6 1 0.4146 1.4963 0.961
## CAP7 1 0.3448 1.2441 1.000
## CAP8 1 0.3190 1.1512 1.000
## CAP9 1 0.2719 0.9812 1.000
## CAP10 1 0.2259 0.8152 1.000
## CAP11 1 0.2179 0.7863 1.000
## CAP12 1 0.1710 0.6172 1.000
## CAP13 1 0.1653 0.5964 1.000
## CAP14 1 0.1383 0.4990 1.000
## CAP15 1 0.1251 0.4514 0.999
## Residual 44 12.1926
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Figure: Species and variables
rda_soil_ind_sum_b = summary(rda_soil_ind,scaling = 1)
#sites
rda_sites = rda_soil_ind_sum_b$sites[,1:3] %>% as.data.frame
rda_sites = rda_sites %>% rownames_to_column("transect") %>% separate(transect,c("common"),sep = 1,remove = FALSE)
rda_sites_c = left_join(rda_sites,sites_density_coding%>%select(transect,thr_cat2),by="transect")
Figure: Transect repartition across commons
Figure: Transect repartition across transhumance modalities