Abstract
The present study assessed the phytoremediation efficiency of water fern (Azolla pinnata R.Br.) in the treatment of dairy wastewater (DWW). Batch mode experimentation was done using different dilutions (0 to 100%) of DWW transplanted with five healthy leaflets of A. pinnata. Besides this, the A. pinnata growth was captured using camera vision-based image recognition and further modeled using logistic and modified Gompertz models. The findings showed that after 14 days of phytoremediation experiments, the maximum significant (p < 0.05) reduction efficiency of selected pollutant parameters of DWW i.e. pH (9.41%), electrical conductivity (61.42%), total dissolved solids (71.56%), total Kjeldahl's nitrogen (73.25%), and total phosphorus (65.37%) observed using 75% DWW dilution. Moreover, the periodically taken surface images were useful to recognize the position and number of leaflets which further helped to simulate A. pinnata growth patterns. The maximum number of leaflets (n = 20), fresh biomass (16.18 ± 0.42 g), dry biomass (1.47 ± 0.04 g), and chlorophyll content (3.14 ± 0.03 mg/g fwt.) was also observed using 75% DWW treatment, respectively. The logistic model was found more robust as compared to modified Gompertz to predict leaflet production (y) as revealed from model validation results i.e. coefficient of determination (R2 > 0.9533) and minimum difference between experimental and model-predicted results. Thus, the combined application of phytoremediation and image processing techniques can be used to monitor and maximize plant growth performance.
[Display omitted]
•The phytoremediation potential of Azolla pinnata in the treatment of dairy wastewater has been assessed.•Maximum reduction efficiency was observed using 75% DWW dilution.•Growth simulation was done using image recognition techniques.•The logistic model was more appropriate as compared to the modified Gompertz model.