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

in maximum biomass yield between treatments or a differential response to fungal infection (Table 1). Thus, both maximum biomass yield between treatments was the same for each period (Before vs. After infection) and decreased by the same magnitude after fungal infection. Fungal infection Through microscopic examination (as described in Section Data Analysis) D. armatus successfully introduced the fungal pathogen, without becoming established in the cultures. Initial morphologi- cal characteristics identified the fungus as an Aphelida (class Aphelidea) [23]. Subsequent samples of infected algae from AzCATI, were cultured in the lab and later identified as Ameobophelidium occidentale using rDNA analysis. Since this fungi were from separate cultures outside of the exper- iment it is not certain that this is the same species, however the morphology and life cycle observed during the experiment are very similar to those identified for various Aphelida species [22, 23, 43]. Additionally, the D. armatus inoculum was from an open-air pond and could have contained other pathogenic agents. However, the dominant signs of disease within the cultures presented morphologically as Aphelid fungus in each of the algae species. This led to the belief that an Aphelid fungus was the main pathogenic stressor in the feedstock cultures. The lack of any differences in response variables between treatments (Control—vs. Impact) was most likely due to S. capricornutum being both a competitively dominant species and being

least susceptible to the fungal parasite. Prior to the intentional introduction of the fungal para- site on day 23, no significant signs of infection were observed in any algal species (infection met- ric were calculate as the proportion of total cells showing signs infection, 2.5 Data Analysis). Prior to fungal infection, S. capricornutum maintained the highest proportion of total cell den- sity in the Impact-polyculture treatments and C. sorokiniana and S. obliquus maintained a lower, but relatively similar proportional cell density to one another (Fig 3A). However, within five days of fungal infection, the average Impact (polyculture) treatment pond became domi- nated by S. capricornutum (>90% of total cell density), and these ponds were almost entirely composed of S. capricornutum by the end of the experiment (Fig 3A). S. capricornutum domi- nating the polyculture treatments after infection is likely attributed to S. capricornutum being least susceptible to the fungal pathogen and C. sorokiniana and S. obliquus both being highly susceptible. The average proportion of S. capricornutum becoming infected in either monocul- ture or polyculture was maintained at less than 0.03 (Fig 3B). In contrast, both C. sorokiniana and S. obliquus demonstrated high susceptibility to infection, and within 2 days after introduc- ing fungus the proportion of cells infected of each species increased rapidly (Fig 3B). Through- out the remainder of the experiment C. sorokiniana and S. obliquus maintained high average

infection proportions (~0.49 and ~0.32, respectively). The high susceptibility to infection of C. sorokiniana: and S. obliquus in conjunction with very low susceptibility of S. capricornutum likely explains why the Impact-polyculture treatments quickly became dominated by S. sorokini- ana, in effect transitioning to a monoculture. Thus, this rapid shift in community composition within the Impact-polyculture treatment after infection reasonably accounts for Treatments (Control-monoculture vs. Impact-polyculture) not differing with respect to productivity, stabil- ity, and maximum biomass yield within the period After infection. Discussion Many prior studies have shown that diverse algal cultures can enhance certain properties that are desirable in algal biofuel feedstock production. For example, polycultures relative to mono- culture feedstocks have been shown improve temporal stability in variable temperature envi- ronments, delay invasion from unwanted algae species, and improve nitrogen and phosphorus use efficiencies [12, 16, 17]. Additionally, polycultures have demonstrated the ability to PLOS ONE Biodiversity, biofuel and disease: Ecological engineering in algal feedstocks PLOS ONE | https://doi.org/10.1371/journal.pone.0267674 April 28, 2022 10 / 15