Predicting reproductive-phase nutrient replenishment in Cannabis sativa L. via a transpiration-driven mass-balance...

Fixed nutrient recipes often overlook developmental shifts in uptake and the coupling between transpiration and ion delivery. We tested a transpiration-driven mass-balance framework to predict nutrient replenishment during the reproductive phase in two cultivars of Cannabis sativa L. (CJ2, First Light). Plants received a solution prepared from a constant recipe; water use was tracked gravimetrically, leachate was periodically sampled, and organs were harvested for ICP-OES/flow injection analysis and biomass composite partitioning. Water-use efficiency (WUE) peaked at 4.5 g L^-1 in CJ2 and 3.4 g L^-1 in First Light, while solution drawdown of N, P, and K declined from early to late flowering. Leachate chemistry resolved cultivar- and stage-dependent transitions between net nutrient removal (net removal) from solution and net accumulation above input. In First Light, N shifted from 130 mg L^-1 net removal to +109 mg L^-1 enrichment and K from 192 mg L^-1 net removal to +195 mg L^-1 enrichment; in CJ2, N ranged from 90 mg L^-1 net removal to +101 mg L^-1 enrichment and K from 102 mg L^-1 net removal to +195 mg L^-1 enrichment. Phosphorus differences were smaller (First Light: 58 to 13 mg L^-1; CJ2: 42 mg L^-1 removal to +2 mg L^-1 enrichment). Mg net removal was greater in First Light than CJ2 (37 vs. 16 mg L^-1). Significant cultivar main effects were detected for P, K, Ca, and Fe (p ≤ 0.05), whereas significant date x cultivar interactions were detected for Mg and Mn (p ≤ 0.05). Across the four harvest intervals retained for mass-balance comparison, biomass-weighted predictors frequently produced the lowest absolute percent deviation (APD), although performance was element- and cultivar-dependent. Single-organ tissue element concentrations misrepresented whole-plant replenishment estimates: leaf Ca and Mg increased while leaf N declined, flowers and roots accumulated P, and stems and roots diverged in K. Predicted input concentrations derived from organ biomass-weighted tissue elemental content improved the alignment between replenishment estimates and measured net solution removal. Biomass-weighted means excluding floral tissue predicted most element replenishment amounts well and performed most consistently in CJ2, whereas biomass-weighted means including all organs more frequently minimized APD in First Light, with absolute percent deviations<10-13%. Collectively, these results indicate that reproductive-phase nutrition in Cannabis sativa L. is well represented by a flux-based mass-balance replenishment framework informed by organ biomass-weighted nutrient composition.