Assessing the Impact of Environmental and Management Variables on Mountain Meadow Yield and Feed Quality Using a Random Forest Model

Seasonal climate variability and agronomic management profoundly influence both the productivity and nutritive value of temperate hay meadows. We analyzed five years of data (2019, 2020, 2022–2024) from 15 meadows in the central Spanish Pyrenees to quantify how environmental variables (January–June...

Full description

Saved in:
Bibliographic Details
Main Authors: Adrián Jarne, Asunción Usón, Ramón Reiné
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Plants
Subjects:
Online Access:https://www.mdpi.com/2223-7747/14/14/2150
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849418356649623552
author Adrián Jarne
Asunción Usón
Ramón Reiné
author_facet Adrián Jarne
Asunción Usón
Ramón Reiné
author_sort Adrián Jarne
collection DOAJ
description Seasonal climate variability and agronomic management profoundly influence both the productivity and nutritive value of temperate hay meadows. We analyzed five years of data (2019, 2020, 2022–2024) from 15 meadows in the central Spanish Pyrenees to quantify how environmental variables (January–June minimum temperatures, rainfall), management variables (fertilization rates (N, P, K), livestock load, cutting date), and vegetation (plant biodiversity (Shannon index)) drive total biomass yield (kg ha<sup>−1</sup>), protein content (%), and Relative Feed Value (RFV). Using Random Forest regression with rigorous cross-validation, our yield model achieved an R<sup>2</sup> of 0.802 (RMSE = 983.8 kg ha<sup>−1</sup>), the protein model an R<sup>2</sup> of 0.786 (RMSE = 1.71%), and the RFV model an R<sup>2</sup> of 0.718 (RMSE = 13.86). Variable importance analyses revealed that March rainfall was the dominant predictor of yield (importance = 0.430), reflecting the critical role of early-spring moisture in tiller establishment and canopy development. In contrast, cutting date exerted the greatest influence on protein (importance = 0.366) and RFV (importance = 0.344), underscoring the sensitivity of forage quality to harvest timing. Lower minimum temperatures—particularly in March and May—and moderate livestock densities (up to 1 LU) were also positively associated with enhanced protein and RFV, whereas higher biodiversity (Shannon ≥ 3) produced modest gains in feed quality without substantial yield penalties. These findings suggest that adaptive management—prioritizing soil moisture conservation in early spring, timely harvesting, balanced grazing intensity, and maintenance of plant diversity—can optimize both the quantity and quality of hay meadow biomass under variable climatic conditions.
format Article
id doaj-art-5cb74e02f3cd46f984ca5e9560b82446
institution Kabale University
issn 2223-7747
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Plants
spelling doaj-art-5cb74e02f3cd46f984ca5e9560b824462025-08-20T03:32:27ZengMDPI AGPlants2223-77472025-07-011414215010.3390/plants14142150Assessing the Impact of Environmental and Management Variables on Mountain Meadow Yield and Feed Quality Using a Random Forest ModelAdrián Jarne0Asunción Usón1Ramón Reiné2Departamento de Ciencias Agrarias y del Medio Natural, Escuela Politécnica Superior, Universidad de Zaragoza, Ctra Cuarte s/n, 22071 Huesca, SpainDepartamento de Ciencias Agrarias y del Medio Natural, Escuela Politécnica Superior, Universidad de Zaragoza, Ctra Cuarte s/n, 22071 Huesca, SpainDepartamento de Ciencias Agrarias y del Medio Natural, Escuela Politécnica Superior, Universidad de Zaragoza, Ctra Cuarte s/n, 22071 Huesca, SpainSeasonal climate variability and agronomic management profoundly influence both the productivity and nutritive value of temperate hay meadows. We analyzed five years of data (2019, 2020, 2022–2024) from 15 meadows in the central Spanish Pyrenees to quantify how environmental variables (January–June minimum temperatures, rainfall), management variables (fertilization rates (N, P, K), livestock load, cutting date), and vegetation (plant biodiversity (Shannon index)) drive total biomass yield (kg ha<sup>−1</sup>), protein content (%), and Relative Feed Value (RFV). Using Random Forest regression with rigorous cross-validation, our yield model achieved an R<sup>2</sup> of 0.802 (RMSE = 983.8 kg ha<sup>−1</sup>), the protein model an R<sup>2</sup> of 0.786 (RMSE = 1.71%), and the RFV model an R<sup>2</sup> of 0.718 (RMSE = 13.86). Variable importance analyses revealed that March rainfall was the dominant predictor of yield (importance = 0.430), reflecting the critical role of early-spring moisture in tiller establishment and canopy development. In contrast, cutting date exerted the greatest influence on protein (importance = 0.366) and RFV (importance = 0.344), underscoring the sensitivity of forage quality to harvest timing. Lower minimum temperatures—particularly in March and May—and moderate livestock densities (up to 1 LU) were also positively associated with enhanced protein and RFV, whereas higher biodiversity (Shannon ≥ 3) produced modest gains in feed quality without substantial yield penalties. These findings suggest that adaptive management—prioritizing soil moisture conservation in early spring, timely harvesting, balanced grazing intensity, and maintenance of plant diversity—can optimize both the quantity and quality of hay meadow biomass under variable climatic conditions.https://www.mdpi.com/2223-7747/14/14/2150grasslandprotein contentrelative feed valuecutting datemodeling
spellingShingle Adrián Jarne
Asunción Usón
Ramón Reiné
Assessing the Impact of Environmental and Management Variables on Mountain Meadow Yield and Feed Quality Using a Random Forest Model
Plants
grassland
protein content
relative feed value
cutting date
modeling
title Assessing the Impact of Environmental and Management Variables on Mountain Meadow Yield and Feed Quality Using a Random Forest Model
title_full Assessing the Impact of Environmental and Management Variables on Mountain Meadow Yield and Feed Quality Using a Random Forest Model
title_fullStr Assessing the Impact of Environmental and Management Variables on Mountain Meadow Yield and Feed Quality Using a Random Forest Model
title_full_unstemmed Assessing the Impact of Environmental and Management Variables on Mountain Meadow Yield and Feed Quality Using a Random Forest Model
title_short Assessing the Impact of Environmental and Management Variables on Mountain Meadow Yield and Feed Quality Using a Random Forest Model
title_sort assessing the impact of environmental and management variables on mountain meadow yield and feed quality using a random forest model
topic grassland
protein content
relative feed value
cutting date
modeling
url https://www.mdpi.com/2223-7747/14/14/2150
work_keys_str_mv AT adrianjarne assessingtheimpactofenvironmentalandmanagementvariablesonmountainmeadowyieldandfeedqualityusingarandomforestmodel
AT asuncionuson assessingtheimpactofenvironmentalandmanagementvariablesonmountainmeadowyieldandfeedqualityusingarandomforestmodel
AT ramonreine assessingtheimpactofenvironmentalandmanagementvariablesonmountainmeadowyieldandfeedqualityusingarandomforestmodel