Data from: Validating a practical methodology for thatch - mat - soil distinction in turfgrass soils

We described a practical method for the distinction of thatch, mat and soil layers in turfgrass soils, being a combination of visually and manually observable characteristics. For two widely different turfgrass species, we analyzed total organic matter (TOM) and dry bulk density (ρd) in thin slices of 6 mm (between 0 and 10 cm soil depth), resulting in clear patterns for both soil properties with increasing depth. Statistical analysis of TOM patterns resulted in similar boundary depths between calculated and observed layers, validating our practical method for the distinction of thatch, mat and soil layers as a reliable method. Furthermore, we characterized thatch, mat and soil layer by different TOM fractions. TOM was fractionalized into three distinctive and functional pools of organic matter: (1) visible organic matter (VOM), consisting of mainly non-decomposed plant structures, (2) decomposed organic matter (DOM), consisting of mainly decomposed plant structures with its associated microbial biomass, and (3) soil organic matter (SOM), being the background value or recalcitrant native organic matter in a soil and its local microbial biomass.

We distinguished thatch, mat and soil layer based on visual and manual observable characteristics of the layers and a protocol as described in Evers et al. (2024)  https://doi.org/10.1002/its2.148.  This study was conducted on a well-established turfgrass demonstration field with monoculture plots of turfgrass varieties (turfgrass seed company DLF; Moerstraten, the Netherlands; 51°32'27'' N 4°20'54" E) 3.5 years after sowing, reflecting the result of organic matter accumulation over these initial years of turfgrass establishment. The climate regime was marine with cool to medium summer temperatures and mild winters (Cfb/Cfa according to the Köppen-Geiger climate classification system (Peel, et al., 2007)). The field was built on a sandy soil (Hortic Anthrasol as described in the FAO/UNESCO soil map of the world (2006)). Sampling of the soil took place in June 2016 before the field was sown. We tested our methodology with two turfgrass species, slender creeping red fescue (Festuca rubra trichophylla (Frt), variety Beudin of DLF) as an example of a spreading turfgrass and perennial ryegrass (Lolium perenne (Lp), variety Duparc of DLF) as an example of a bunch-type grass, as these two species were expected to differ widely in thatch and mat depth. The individual plot size for each variety was approximately 1 m2 (0.8 x 1.2 m). Plots of each species were sampled at the end of February 2020. A subplot of 0.25 m2 (0.5 m x 0.5 m) in the center of one plot per species was selected, to avoid contamination with other varieties (at least 90% pure monoculture), and it was marked with a metal frame. From the 25 cells of 5 x 5 cm in this frame, nine evenly dispersed cells were chosen to take a set of soil samples of 10 cm depth, using a core sampler with 2.8 cm diameter, for thatch-mat and mat-deeper soil boundary observation and TOM and ρd analyses. Another set of nine samples was taken next to the previous cells for VOM analyses. Every fresh 10 cm soil core was first photographed (Canon Powershot S5 camera, 24 megapixels; Canon Europe, Amstelveen, the Netherlands), judged on thatch-mat and mat-deeper soil boundaries following the method described in supplementary information, and then sliced into 14 subsamples of 6 mm each plus a 16 mm subsample at the bottom, starting to measure just below the green canopy with an accurate ruler (Sola HK ¼ W12, EU-accuracy class 3), for analyzing TOM and ρd.in every subsample. TOM and ρd were analyzed after samples were dried at 105°C for 24 h. VOM was analyzed after all sediment per slice was carefully washed off with tap water in a fine sieve (approximately 600 µm (27 mesh)), after which the remaining (dead and living) plant biomass, mainly roots and rhizomes, was dried at 65 °C for at least 48  h. SOM was determined separately in the bulk soil of the study site and is 2.6% of the dry matter. DOM was calculated via subtraction of VOM and SOM from TOM

Based on the analyzed TOM content of the 14 slices per nine replicates, the boundaries of distinctive layers were calculated. For this, we rescaled the TOM results per replicate via the normalized function F (χ) = (χ-χmin)/(χmax-χmin) into values between 0 and 1 to overcome scale differences between replicates but keeping distributions the same. Per turfgrass species, the best fitted line, i.e., the smallest rse, and its 95%-confidence interval through all 135 points was iteratively calculated by non-linear least square regression with the nls function of R (version 3.5.2; 2018-12-20). For the fitting of the statistical models, either a logistic function (F (χ) = α/(1+e-(βχ+γ)) or a bell-shaped Gaussian function (G (χ) = αe-((χ-β)^2/2γ^2)) was used, based on the best fit for the respective turfgrass species. The characteristics of these mathematical functions were used to explain the TOM-dynamics in the soil. To this end, the turning points of the curves were calculated by finding where the first derivative of both functions equals zero, i.e., solving F’ (χ) = 0 and G’ (χ) = 0 respectively. The inflection points of each curve were determined by analyzing the second derivative, i.e., identifying the points where the second derivative F’’ (χ) or G’’ (χ) = 0, indicating changes in concavity. The inflection points and turning points indicated changes in TOM content in the turfgrass soil profile as likely boundaries between distinctive soil layers. Differences in parameters between distinctive soil layers and turfgrass species were determined based on the calculated means of parameters per nine replicates of soil slices Normality of residuals and the equality of variances was checked with diagnostic plots and Levene’s test, respectively. Normally distributed means were compared with one-way ANOVA for a 3-layered soil system, followed by either Tukey post hoc tests in case of equality of variance, or by the Games-Howell post hoc test in case of no equality of variance, or with a one sample t-test for a 2-layered soil system.