Biodiversity and disease risk in an algal biofuel system: An experimental test in outdoor ponds using a before-after-control-impact (BACI) design

polycultures) and between time periods (Before vs. After fungal infection). S3 Fig in S1 Appen- dix shows the time-series for average daily water pH, salinity, conductivity, oxygen saturation, dissolved oxygen, and water temperature. These time-series data are represented as treatment and time period means in S4 Fig in S1 Appendix, and a statistical comparison of treatment and period means is given in S1 and S2 Tables in S1 Appendix. S4 Fig in S1 Appendix and the two supplemental data tables in S1 Data show that the daily mean physical and chemical con- ditions of the ponds did change between time periods (e.g. water temperature increased, and oxygen saturation declined through time). However, there was no evidence that the physical/ chemical conditions changed differently through time among treatments. Therefore, raceway ponds were homogenous with respect to each other in their physical/chemical conditions. Areal productivity No significant difference in areal productivity, feedstock stability, or maximum biomass yield between the Control (monoculture) and Impact (polyculture) treatments was found. Areal productivity averaged 12.66 ± 0.44 g m-2 day-1 prior to infection in the Control (monoculture), and a nearly identical 12.50 ± 0.29 g m-2 day-1 in the Impact (polyculture) treatment (Fig 2A). After introduction of the parasitic fungus, areal productivity declined in both treatments to 9.72 ± 0.07 g m-2 day-1g in the Control (monoculture) and 10.39±0.40 g m-2 day-1 in the Impact (polyculture) treatment (Fig 2A). BACI analysis revealed no significant differences in

areal productivity between treatments, and no difference in response of the treatments between the two time periods (Table 1). Thus, the means of the replicate ponds were the same among treatments and declined by the same magnitude after fungal infection. Biomass stability The mean feedstock biomass stability (inverse CV) prior to infection averaged 3.50 ± 1.09 for the Control (monoculture) and 4.34 ± 0.39 for the Impact (polyculture) treatments (Fig 2B). After fungal infection, mean feedstock stability decreased in both treatments to 0.50 ± 0.15 in the Control (monoculture) and 0.52 ± 0.10 in the Impact (polyculture) treatment (Fig 2B). BACI analysis showed no significant differences in either feedstock stability between treat- ments or response of the treatments between periods (Before vs. After infection). Therefore, the mean feedstock stability did not differ between treatments, and decreased by the same magnitude after fungal infection. Maximum biomass yield The maximum biomass yield achieved prior to a harvest for the time-frame before fungal infection was 0.99 ± 0.04 g L-1 for the Control treatment, which was virtually equal to the 0.95 ± 0.04 g L-1 for the Impact treatment (Fig 2C). During the after-infection period, maxi- mum biomass yield declined for both treatments to 0.81 ± 0.03 g L-1 for the Control verses 0.83 ± 0.04 g L-1 for the Impact treatment. The BACI analysis showed no significant difference Table 1. Analysis of variance effects test results. Mean Productivity Mean Stability Mean Maximum Biomass Effect df MS F p MS F p MS F p

Treatment 1 0.194 0.589 0.465 0.508 0.497 0.501 0.0003 0.080 0.785 Period 1 19.165 58.289 <0.000 34.919 34.1793 <0.000 0.069 18.290 0.003 Trmt:Period 1 0.531 1.6159 0.239 0.491 0.481 0.508 0.003 0.716 0.422 https://doi.org/10.1371/journal.pone.0267674.t001 PLOS ONE Biodiversity, biofuel and disease: Ecological engineering in algal feedstocks PLOS ONE | https://doi.org/10.1371/journal.pone.0267674 April 28, 2022 9 / 15