Non-Destructive Assessment of Soil Nutrient Variability in Mbeya Catena Using Dielectric Properties via Low-Cost Antenna Systems

Soil nutrient depletion and inefficient fertilizer management remain major challenges for sustainable agriculture in Tanzania’s Southern Highlands. This study evaluates a low-cost, non-invasive dielectric sensing approach to monitor soil nutrient and moisture variability using a log-periodic dipole antenna (LPDA) system integrated with a nano vector network analyzer (NanoVNA). Soil samples from the middle and lower catena positions in Mbeya City were treated with varying concentrations (0–12.5%) of UREA and calcium ammonium nitrate (CAN), and dielectric properties permittivity and conductivity were measured under controlled moisture levels (10–40%). The results showed strong positive correlations between UREA concentration, dielectric constant (r = 0.905), and conductivity (r = 0.858), particularly in the lower catena. CAN showed a reliable response in the middle catena (r = 0.913 for εr) but inconsistent trends in the lower catena. Moisture content had a significant non-linear effect on dielectric behavior, with a peak response at 40% moisture. A two-way ANOVA confirmed statistically significant main and interaction effects (p < 0.05) for fertilizer type, concentration, and catena position. These findings validate the LPDA-NanoVNA system as an effective soil nutrient and moisture monitoring tool. The dielectric method is especially promising for supporting site-specific fertilizer applications and precision farming. UREA is more suitable for dielectric-based assessments in variable terrain, while CAN is more effective in stable slope environments. The approach advances low-cost, scalable technologies for climate-smart agriculture and aligns with SDG 15 goals on sustainable land use and soil health.