Phytoremediation results

Data Synthesis through Multivariate Analyses

Data from Experiments 1 (shrimp waste) and 2 (fish solids) were assimilated and combined for meta-analysis using data visualization with multivariate statistical techniques. For each metric reported, the mean observation per flat (n=8) was calculated for each of the three treatments. These data were then transformed to Z-scores (obs-mean/stdev) for each metric to standardize the range of values (i.e., normalizing), and then analyzed using non-parametric multivariate tests in Primer v. 6.1.5 (PRIMER-E Ltd) to visualize among treatment differences. First, the resemblance matrix (i.e., distance or similarity matrix) was calculated using Euclidean distance, where highly similar samples are grouped together. Second, the resemblance pattern in multivariate space was visualized using non-metric multidimensional scaling (MDS) with 50 runs, where points that are close together represent samples that are very similar. Third, the resemblance matrix was analyzed using analysis of similarity (ANOSIM) with 999 permutations to test the null hypothesis of no significant difference among treatments, followed by pairwise (posthoc) tests based on the difference of average rank dissimilarities to determine like groups. Results obtained in the MDS plots (Fig. 4) provide a visual representation of the multivariate responses measured in S. alterniflora and J. roemerianus to the two nutrient enrichment treatments and the control at two time points, at the end of the 6 week dosing period and the end of the 20 week response period. Similarly, hierarchical agglomerative clustering using group means (UPGMA) on the treatment centroids strongly emphasizes the order of fertilization responses: Miracle-Grow was better than Fish Solids or Shrimp Waste (similar to each other), which both were better than no fertilizer (Control plants).

Figure 4: Summary results obtained using multivariate methods. Panels on the left are MDS plots with the four experimental treatments (C = control, MG = Miracle-Grow, FS = fish solids, SW = shrimp biofloc waste). Panels on the right are clustering dendrograms showing the group centroids and relations of the four treatment effects.

Overall, phytoremediation of salty nutrient-rich marine aquaculture wastes can be achieved using salt-tolerant coastal marsh plants. These plants can further be re-purposed for coastal habitat restoration projects, yielding a second source of income for marine aquaculture businesses, while simultaneously reducing the environmental costs associated with waste-water disposal and expensive salt additions to new water.

Results of this research were published here:

Joesting, H., R. Blaylock, P.D. Biber and A. Ray. 2016. The use of marine aquaculture solid waste for nursery production of the salt marsh plants Spartina alterniflora and Juncus roemerianus. Aquaculture Reports 3: 108-114 

https://doi.org/10.1016/j.aqrep.2016.01.004