SEASONAL AND DEPTH-WISE DYNAMICS OF SOIL CHEMICAL PROPERTIES UNDER ARID-ZONE TREE PLANTATIONS IN SARDARSHAHAR, RAJASTHAN (INDIA): IMPLICATIONS FOR SAMPLING DESIGN IN SOIL CARBON INVENTORIES

Abstract

Soil carbon and nutrient inventories in dryland regions are commonly built on single-occasion sampling, an approach that implicitly assumes temporal stability of the measured properties. This study tested that assumption by monitoring seven soil chemical properties under six tree species — Tecomella undulata, Ailanthus excelsa, Dalbergia sissoo, Prosopis cineraria, Acacia nilotica and Azadirachta indica — across four seasons and two depths (0–15 and 15–30 cm) in Sardarshahar tehsil, Churu district, Rajasthan, India, from November 2023 to October 2024. All nutrient variables peaked in the rainy season and reached a minimum in summer, whereas pH and electrical conductivity followed the opposite phase. Organic carbon varied from 0.301% in summer to 0.420% in the rainy season, a seasonal amplitude of 33.3% of the annual mean, and was 39.6% higher in the rainy season than in summer (t = 20.43, p < 0.001). Corresponding rainy-to-summer increases were 36.6% for total nitrogen, 35.6% for available phosphorus, 26.4% for available nitrogen and 21.7% for available potassium, while electrical conductivity fell by 18.2% and pH by 4.8%. Seasonal means of every nutrient variable were almost perfectly correlated with those of organic carbon (r ≥ 0.998), indicating a common moisture-mediated driver. Surface-to-subsurface ratios were statistically invariant across seasons (organic carbon 1.30–1.33), showing that season and depth acted additively rather than interactively. Partitioning of the explained sum of squares showed that species dominated variation in pH (75.6%) and electrical conductivity (61.5%), depth dominated available phosphorus (45.6%) and available nitrogen (38.8%), and season contributed a remarkably uniform 17.5–25.5% to every property. Estimated soil organic carbon stock for the 0–30 cm profile ranged from 13.5 Mg ha−1 if sampled in summer to 18.9 Mg ha−1 if sampled in the rainy season. Single-season sampling can therefore bias dryland soil carbon inventories by nearly 40%, and standardisation of sampling season is recommended as a prerequisite for comparability.