Paraphysoderma sedebokerense Infection in Three Economically Valuable Microalgae: Host Preference Correlates with Parasite Fitness

The blastocladialean fungus Paraphysoderma sedebokerense parasitizes three microalgae species of economic interest: Haematococcus pluvialis, Chromochloris zofingiensis and Scenedesmus dimorphus. For the first time, we characterized the developmental stages of isolated fungal propagules in H. pluvialis co-culture, finding a generation time of 16 h. We established a patho-system to compare the infection in the three different host species for 48 h, with two different setups to quantify parameters of the infection and parameters of the parasite fitness. The prevalence of the parasite in H. pluvialis and C. zofingiensis cultures was 100%, but only 20% in S. dimorphus culture. The infection of S. dimorphus not only reached lower prevalence but was also qualitatively different; the infection developed preferentially on senescent cells and more resting cysts were produced, being consistent with a reservoir host. In addition, we carried out cross infection experiments and the inoculation of a mixed algal culture containing the three microalgae, to determine the susceptibility of the host species and to investigate the preference of P. sedebokerense for these microalgae. The three tested microalgae showed different susceptibility to P. sedebokerense, which correlates with blastoclad’s preference to the host in the following order: H. pluvialis > C. zofingiensis > S. dimorphus.


Introduction
The aquatic lower fungi belonging to Blastocladiomycota [1] and their close relatives, the better studied Chytridiomycota, have a high impact on phytoplankton dynamics [2,3], and commonly constitute a plague for microalgae in mass outdoor cultures [4][5][6]. However, little is known about the life cycle, ecology and host range of many algal parasites [7]. Even more, the time needed to complete the life cycle of parasitic Chytridiomycota and Blastocladiomycota species, an essential parameter for understanding epidemics, has not been investigated and is only rarely reported from in situ studies [8][9][10][11][12].
Paraphysoderma sedebokerense is a facultative parasite, which was isolated for the first time from Haematococcus pluvialis collapsed cultures in Israel and can be maintained in laboratory culture growing as a saprobe [13]. The life cycle of P. sedebokerense is complex and includes both the thin cell wall vegetative cysts, and the darker resting cysts with thicker cell wall which acts as a resistant stage, and two types of dispersion propagules: flagellated zoospores (produced only in infected cultures) and amoeboid swarmers [13,14]. In the infective stage, the propagules cause lethal epidemics in H. pluvialis, which is the best natural source of the high value ketocarotenoid astaxanthin [15]. The economic impact of these epidemics is reflected in efforts to control the development of the infection by P. sedebokerense, by selection of resistant H. pluvialis strains phenotypically dominated by motile cells which are not vulnerable to the parasite [16,17], changing culture conditions [18] or using surfactants [19]. In laboratory tests, P. sedebokerense showed the ability to also We used the blastoclad Paraphysoderma sedebokerense isolate AZ_ISR (2019), considering it is the same TJ-2007a strain described in Hoffman et al. [13], since it was isolated from the same host (H. pluvialis) and from the same place at Sede-Boker, Israel. The ITS sequence (GenBank MW336992) of our isolate is 99.7% identical to P. sedebokerense isolates FD61 [21] and JEL0821 [41]. After isolation, we maintained it in a pure culture and continuously subculture it on solid blastoclad growth media (BGM) in our laboratory [40]. Pure cultures of blastoclad were grown in BGM at 30 • C in an incubator shaker (180 rpm) supplemented with 2% CO 2 , under continuous dim white light (15 µmol photons m −2 s −1 ) illumination [40]. Blastoclad propagules were harvested in propagules stimulation medium (PSM), following the protocol developed by Asatryan et al. [40]. The quality of the propagules was visually checked under light microscopy, with attention to propagule movement, absence of encysted propagules and general appearance. Propagules were directly counted in a Neubauer chamber (Reichert Bright-Line 1492 hemacytometer, Hausser Scientific, Horsham, PA, USA), at 400× magnification with phase contrast, using a Zeiss AxioSkop HBO 50 W microscope (Zeiss, Oberkochen, Germany).

Algal Strains and Growth Conditions
Haematococcus pluvialis Flotow 1844 em. Wille K-0084 was obtained from Scandinavia Culture Center for Algae and Protozoa at the University of Copenhagen, Denmark. Scenedesmus dimorphus UTEX B 1237 was obtained from the culture collection of algae at the University of Texas at Austin. Chromochloris zofingiensis SAG-211.14 was obtained from the Culture Collection of Algae of the University of Göttingen (SAG), Göttingen, Germany. Algal monocultures were grown for 7 days in modified BG-11 (mBG11 [42]), in 250 mL Erlenmeyer flasks (100 mL culture), at controlled temperature (25 • C), and constant illumination (80 µmol photon m −2 s −1 ), in an incubator shaker (150 rpm) enriched with 2% CO 2 . For maintenance, H. pluvialis cultures were weekly diluted to approximate densities of 2 × 10 5 cells/mL. Maintenance of C. zofingiensis and S. dimorphus cultures was also done by continuous dilutions of cultures to approximately 5 × 10 6 cells/mL. At these cell densities, cultures reach the stationary stage after one week.

Estimation of Algal Cell Surface Area
Using a LUNA-FL™ Dual Fluorescence Cell Counter automated cell counter (Logos Biosystems, Gyunggi-do, Korea), we measured the cell diameter (long axis for Sd) of each species used in this work. For H. pluvialis and C. zofingiensis, we considered them as perfect spheres to calculate their surface area (4 × π × r 2 ). For S. dimorphus, we measured the two axes, and we simplified the calculation, considering S. dimorphus cells as cylinders without base, to calculate the surface area (2 × r × π × h) (Supplementary Table S1).

Infection of Algal Cultures and Infection Parameters
The infections were carried out in fresh mBG11 medium, in the same conditions used to grow the blastoclad, as described above. We tested different types of inoculum: P. sedebokerense logarithmic monoculture, blastoclad coculture with H. pluvialis (Ps-Hp) and purified propagules routinely isolated from 7-10 days old blastoclad monoculture (referred to in the text as pure propagules) or in specific experiments purified propagules isolated from blastoclad coculture with H. pluvialis 3 days after inoculation (referred to as propagules isolated from Ps-Hp coculture).
Macroscopically, we assessed the virulence of the blastoclad against different algal monocultures as the time needed to cause collapse of the algal culture (flocculation and color change from green to brown), using all different types of inoculums described above.
The quantitative infections were all done only with P. sedebokerense pure propagules, isolated from blastoclad monoculture. To avoid bias, we compared the infection development in the three microalgae (H. pluvialis 4 × 10 5 cells/mL, C. zofingiensis 4 × 10 6 cells/mL, and S. dimorphus 1 × 10 6 cells/mL, cell densities which offer the same host surface area per mL) by inoculating synchronously with the same batch of fresh pure propagules. We assessed five parameters: prevalence (Pr.) as the percentage of host cells carrying encysted parasite, the average number of parasites cysts per cell (considering all cells) as intensity (In.), the average number of parasites cysts per algal surface area as areal density (Ad.), the propagule survival as the sum of host attached cysts and free-swimming propagules at the indicated time (before the first propagule release) out of the total inoculated propagules at zero time of the experiment and propagule production as the measured number of propagules released from the inoculated cultures. The values of Ad. are in areal units; in our case, a unit area is the average surface area of a C. zofingiensis cell (88 µm 2 , Supplementary Table S1).
Since we used different densities of microalgae normalized on their surface area to have the same probability for the parasite to find a suitable host, we needed to use different inoculation ratios. One ratio to quantify the parameters of the infection which are related to the ratio of parasite per host, choosing 1:1 to allow fast development of the infection without excessive effort to harvest propagules. For the parameters of the fitness, we used the same number of propagules per algal culture, to compare survival and production with respect to the inoculated amount. Since we fixed different microalgal cells densities and we needed the same infective ratio for infection parameters and the same number of propagules inoculated for fitness parameters, it ended up with different ratios for each setup.
We considered Pr., In., and Ad. as parameters of infection, and to measure these parameters, we inoculated P. sedebokerense propagules/host cell at the same ratio (1:1). To detect the encysted propagules, we used the combination of phase contrast and fluorescent Nile red staining of the blastoclad [20], at 40× magnifications when inoculated with H. pluvialis and S. dimorphus cultures, and at 100× magnifications when inoculated with C. zofingiensis cultures. For quantification, 50 host cells were counted and considered as a representative subsample.
We considered propagule survival and propagule production as fitness parameters of the parasite, and we inoculated the same number of propagules per infected culture resulting in dissimilar ratios of infection (10:1 in H. pluvialis, 4:1 in S. dimorphus and 1:1 in C. zofingiensis) and determine both parameters by direct cell counting under the microscope. To measure propagule survival, we counted both free-swimming propagules and encysted ones, considering as live propagules those keeping the amoeboid shape and dead propagules the burst or collapsed ones (which were not counted). Since we inoculate a known number of propagules, we calculated the percentage of survival as = (counted number of live propagules and cysts at each time point/number of inoculated propagules at time zero) × 100.
For each parameter, at each sampling point and microalgal species, three to five replicates were measured.
Finally, we considered preference as the successful encystment events of the blastoclad propagules on a specific microalgal species when the other microalgae species are present. We assessed the preference by inoculating purified propagules in a mixed algal culture, where each of the three algae were present at a density that offered equal surface area to the parasite, hereafter referred to as "mixed culture": H. pluvialis at 2.5 × 10 4 cells/mL, S. dimorphus at 7.5 × 10 4 cells/mL and C. zofingiensis at 2.1 × 10 5 cells/mL. The preference was calculated as the number of encystment events divided by the maximum possible encystment events (under assumption that all propagules were alive at the onset of inoculation), multiplied by 100. Preference = (In. microalga × microalga cell number) / (number of inoculated propagules).
For additional details about the inoculum type and conditions of the experiments, see supplementary material (Supplementary Table S2)

Statistical Analyses
Statistical analyses were performed in StatGraphics centurion XVII, considering statistically significant p-values lower than 0.05. The comparison of infection parameters and the fitness parameters of parasite of algal cocultures was done by general linearized model (GLM), separately for each sampling time. To discriminate which means are different, a multiple comparison was performed using the Fisher method (Supplementary Table S3).
The statistical analysis for the preference test was manually done using the chi-square: x 2 = ((Observed-Expected) 2 )/Expected and comparing the value of the results with the chi-square tables.

Infection Cycle of H. pluvialis by P. sedebokerense Isolated Pure Propagules
In order to establish the infection system, the chronology of the infection of H. pluvialis by P. sedebokerense was assessed via inoculating pure amoeboid swarmers of P. sedebokerense (obtained from pure culture in BGM) ( Figure S1) and following the different developmental stages during the first infection cycle, until new propagules are released.
The attachment of the propagules to the algal cells that begins the encystment process is visible 30 min after inoculation, when the shape of the amoeboid swarmers changes into a completely round cyst-like structure ( Figure 1A). One hour after inoculation, the encysted propagule penetrates the algal cell wall with a germ tube ( Figure 1B) and the rhizoidal system grows. After encystment, the blastoclad cyst feeds on the algal cell and consequently grows in size; it initiates cell division 9 h after inoculation ( Figure 1C). The mature sporangium, which loses its spherical shape when propagules increase the pressure on its cell wall, is visible 16 h after inoculation. At this stage, the discharge pore of the sporangia opens to enable the release of new infective propagules, completing the first infection cycle ( Figure 1D).

Infection Cycle of H. pluvialis by P. sedebokerense Isolated Pure Propagules
In order to establish the infection system, the chronology of the infection of H. pluvialis by P. sedebokerense was assessed via inoculating pure amoeboid swarmers of P. sedebokerense (obtained from pure culture in BGM) ( Figure S1) and following the different developmental stages during the first infection cycle, until new propagules are released.
The attachment of the propagules to the algal cells that begins the encystment process is visible 30 min after inoculation, when the shape of the amoeboid swarmers changes into a completely round cyst-like structure ( Figure 1A). One hour after inoculation, the encysted propagule penetrates the algal cell wall with a germ tube ( Figure 1B) and the rhizoidal system grows. After encystment, the blastoclad cyst feeds on the algal cell and consequently grows in size; it initiates cell division 9 h after inoculation ( Figure 1C). The mature sporangium, which loses its spherical shape when propagules increase the pressure on its cell wall, is visible 16 h after inoculation. At this stage, the discharge pore of the sporangia opens to enable the release of new infective propagules, completing the first infection cycle ( Figure 1D).

Virulence of P. sedebokerense against Different Algae
We inoculated H. pluvialis, S. dimorphus and C. zofingiensis with P. sedebokerense, assessing the virulence macroscopically, by observing the flocculation and change of color of the algal cultures from green to brown (Figures 2 and 3); the macroscopic observations were confirmed under a light microscope. For inoculation, P. sedebokerense logarithmic

Virulence of P. sedebokerense against Different Algae
We inoculated H. pluvialis, S. dimorphus and C. zofingiensis with P. sedebokerense, assessing the virulence macroscopically, by observing the flocculation and change of color of the algal cultures from green to brown (Figures 2 and 3); the macroscopic observations were confirmed under a light microscope. For inoculation, P. sedebokerense logarithmic cultures (Figure 2A), previously infected H. pluvialis monocultures ( Figure 2B), or propagules isolated from Ps-Hp coculture ( Figure 2C) or pure isolated propagules ( Figure 3) were all tested; these four inoculum types showed similar virulence, but the susceptibility of the three different microalgae was different. H. pluvialis showed high susceptibility to the infection, collapsing in 24-72 h after inoculation, depending on the amount of P. sedebokerense inoculum. We found a similarly high virulence of P. sedebokerense in C. zofingiensis cultures, which collapsed 24-96 h after inoculation in most of the cases. In contrast, we never observed total collapse of the S. dimorphus culture after inoculation with P. sedebokerense; the culture always maintained its vivid green color, although partial flocculation of the cul-ture was observed ( Figure 2B). These symptoms in S. dimorphus cultures never progressed, even if we used higher P. sedebokerense inoculum and/or if we maintained the inoculated culture for longer time (at least two weeks). Only for S. dimorphus, infection tests were also conducted in the medium used by Letcher et al. [21], and still collapse of the culture was not observed.
were all tested; these four inoculum types showed similar virulence, but the susceptibility of the three different microalgae was different. H. pluvialis showed high susceptibility to the infection, collapsing in 24-72 h after inoculation, depending on the amount of P. sedebokerense inoculum. We found a similarly high virulence of P. sedebokerense in C. zofingiensis cultures, which collapsed 24-96 h after inoculation in most of the cases. In contrast, we never observed total collapse of the S. dimorphus culture after inoculation with P. sedebokerense; the culture always maintained its vivid green color, although partial flocculation of the culture was observed ( Figure 2B). These symptoms in S. dimorphus cultures never progressed, even if we used higher P. sedebokerense inoculum and/or if we maintained the inoculated culture for longer time (at least two weeks). Only for S. dimorphus, infection tests were also conducted in the medium used by Letcher et al. [21], and still collapse of the culture was not observed.  were all tested; these four inoculum types showed similar virulence, but the susceptibility of the three different microalgae was different. H. pluvialis showed high susceptibility to the infection, collapsing in 24-72 h after inoculation, depending on the amount of P. sedebokerense inoculum. We found a similarly high virulence of P. sedebokerense in C. zofingiensis cultures, which collapsed 24-96 h after inoculation in most of the cases. In contrast, we never observed total collapse of the S. dimorphus culture after inoculation with P. sedebokerense; the culture always maintained its vivid green color, although partial flocculation of the culture was observed ( Figure 2B). These symptoms in S. dimorphus cultures never progressed, even if we used higher P. sedebokerense inoculum and/or if we maintained the inoculated culture for longer time (at least two weeks). Only for S. dimorphus, infection tests were also conducted in the medium used by Letcher et al. [21], and still collapse of the culture was not observed.

Infection in Mixed Algal Culture and Cross Infections
A mixed algal culture composed of H. pluvialis (2.5 × 10 4 cell/mL), S. dimorphus (7.5 × 10 4 cell/mL) and C. zofingiensis (2.1 × 10 5 cell/mL) was prepared and inoculated with blastoclad pure propagules (2.3 × 10 5 propagules/mL). These algal densities were chosen in order to expose the same surface area of each algal species to the blastoclad inoculum. We measured the preference as the percentage of inoculated propagules that successfully encysted on live green algae. We expect the same number of encystment events (7.67 × 10 4 ) for each of the three microalgae, if the blastoclad has no preference for a specific algal specie. However, 9 h after inoculation, 52.1% of the propagules encysted on H. pluvialis cells, 1.6% of propagules encysted on C. zofingiensis cells, while no cysts were detected on S. dimorphus ( Table 1). The differences between observed and expected encystment events were statistically significant under chi-square test with a pValue lower than 0.001. We thus conclude that P. sedebokerense shows high preference for H. pluvialis cells, intermediate preference for C. zofingiensis and null preference for S. dimorphus. Moreover, 20 h after inoculation all H. pluvialis cells carried P. sedebokerense cysts, while 42% of C. zofingiensis cells and 95% of S. dimorphus cells remained uninfected and parasite-free 48 h after inoculation (data not shown). A big proportion of the inoculated propagules were not directly attached to any of the three microalgae; it was not quantified, but many of them encysted on empty mother cell walls (especially from H. pluvialis) and on other blastoclad cysts. To test cross infections, fresh algal monocultures were inoculated with the five-daysold infected algal monocultures described in Figure 2C. The results showed different susceptibility to the blastoclad, depending not only on the algal host, but also depending on the type of inoculum (Table 2). H. pluvialis was the most sensitive monoculture, suffering fast culture collapse 48-60 h after inoculation with all four inoculum types; at the same time, Ps-Hp co-culture was the most virulent towards both H. pluvialis and C. zofingiensis. The C. zofingiensis monoculture showed less susceptibility as compared to H. pluvialis monoculture; epidemics developed slower and cultures collapsed 60-168 h after inoculation, depending on the inoculum type Table 2. The inocula of Ps-Cz and Ps-Sd co-cultures showed similar virulence towards C. zofingiensis, causing only flocculation and lower virulence than Ps-Hp and Ps-mixed co-cultures, which caused culture collapse in C. zofingiensis (Table 2) In all tests, S. dimorphus did not collapse nor flocculate.

Quantification of the Infection
The cell size is remarkably different between the tested algal species. To offer the parasite the same probability to encyst on an algal cell, we used cell densities which provide equal surface area per ml in each of the tests; consequently, algal cell density is different for each tested alga. To enable the comparable measurement of the prevalence (% of infected host cells) and intensity (average number of cyst/host) of the infection (parameters that are expressed on the basis of individual algal host cell), the three different microalgae were inoculated with P. sedebokerense pure amoeboid swarmers at a ratio of 1:1 (parasite/host). We found statistical differences in prevalence values (α 0.05); H. pluvialis suffered higher prevalence than C. zofingiensis in the first 8 h and higher than S. dimorphus during the 48 h of the experiment. During the first day, at sampling points 3, 8 and 12 h, H. pluvialis showed high and constant prevalence of about 60%; the prevalence increased, reaching 100% of prevalence 24 h after inoculation ( Figure 4A). In the case of C. zofingiensis, the initial prevalence (3 h and 8 h after inoculation) was low (<12%) and not statistically different from that of S. dimorphus (Supplementary Table S3). However, the prevalence sharply increased later on, to 75% at 12 h after inoculation. In contrast, the prevalence in S. dimorphus increased slowly, reaching a maximum of 20%, 12 h after inoculation, and stayed the same until the end of the experiment (48 h) ( Figure 4A). We also found statistical differences (α 0.05) in the intensity of the infection in the three microalgae: H. pluvialis suffered higher intensity of the infection as compared to the other two microalgae, and only at 12 h and 22 h after inoculation, the intensity in H. pluvialis is not statistically different from C. zofingiensis intensity. The intensity of the infection in S. dimorphus was the lowest, and only in the first sampling point (3 h) it was not statistically lower than the intensity of the infection in C. zofingiensis (Supplementary Table S3 Figure 4A). In both algae, the first increase phase is due to the attachment/encystment of the inoculated propagules, while the other increase phases are suggesting new propagule release. The consequent increases in intensity and areal density are in accordance with the propagule release pulses. In the case of S. dimorphus, a plateau was reached 12 h after inoculation, with no further increase ( Figure 4C) within the two days of the experiment. We normalized the unit areas, considering 1-unit area is the average cell surface of C. zofingiensis (for more details, see Supplementary Table S1). Data presented in this figure are from cultures shown in Figure 3A. Letters indicate statistical significance (p value < 0.05).
Blastoclad fitness tests were conducted as mentioned above with inoculations of algal monocultures based on equal algal cell surface area. However, in this experiment, inoculation with the blastoclad was done with an equal propagule number in each tested alga culture, to enable the comparison of survival and production with respect of the inoculated amount at time zero. Here, we found higher propagule survival in H. pluvialis (p value 0.047) than in C. zofingiensis and S. dimorphus monocultures ( Figure 5A). Propagule production was different among the different algae. The values in H. pluvialis were significantly different from the other algae in each sampling time (p value 0.0054), while the production in C. zofingiensis was different from S. dimorphus only at the measured peak of production (22 h after inoculation). In H. pluvialis, a peak of 3.4 × 10 6 propagules/mL was observed after 24 h, in C. zofingiensis it was observed after 22 h (8.4 × 10 5 propagules/mL), and in S. dimorphus it was observed 24 h after inoculation (2.7 × 10 5 propagules/mL) ( Figure 5B).  Figure 3B. Letters indicate statistical significance (p value < 0.05).
In the three different inoculated monocultures, two types of cysts, the vegetative cyst and the resting cysts, were produced, and also both types of propagules, amoeboid swarmers and flagellated zoospores. Although these observations were not systematically quantified, we observed that resting sporangia ( Figure 6A) were more abundant and appeared faster in S. dimorphus cultures at 48 h after inoculation ( Figure 6B). In contrast, in H. pluvialis infected culture, vegetative cysts developing into mature sporangia releasing amoeboid swarmers ( Figure 6C,D) are more common in the first hours or days after inoculation. In S. dimorphus, P. sedebokerense cysts were frequently attached to senescent cells which did not shows the red typical autofluorescence of the chlorophyll ( Figure 6E-H).

Discussion
We carried out a comprehensive quantitative study of infections of microalgae by an aquatic true fungus, inoculating-for first time-with a known number of pure isolated propagules ( Figure S1). Previous efforts carried out, by homogenization of fungal culture and inoculating with a known density [13], can only be considered as semi-quantitative, since propagules and cysts burst by homogenization, so that there was no real estimation of the infecting units. We also characterized here, for the first time, the chronology of infection stages (Figure 1) of P. sedebokerense, starting with pure propagules in Ps-Hp coculture. We found that the time needed to complete the cycle (new propagules release) is 16 h, which is significantly shorter than the period previously reported (24-30 h) to reach the sporangium stage in H. pluvialis culture [13]. This is also lower than the periods reported to complete the cycle in chytrid parasites of phytoplankton: 2 days [8] or 4-8 [9] in Rhizophydium planktonicum, 3 days in Rhizosiphon crissum [11] or 7-8 days in Rhizophydium scenedesmi [12]. The knowledge on the chronology of infection in H. pluvialis allowed us to design experiments to quantify and compare the kinetics of infection of the three commercially valuable microalgae: H. pluvialis, C. zofingiensis and S. dimorphus, establishing the sampling times in the first 48 h after inoculation.
We found prevalence values similar to the ones reported by Gutman et al. [20], with inoculated H. pluvialis and C. zofingiensis cultures reaching 100% of prevalence, and a Scenedesmus species reaching prevalence of 20%. However, in the current work, the experiment took 48 h with a parasite: host ratio of 1:1 at the onset of inoculation (Figure 4), and we used S. dimorphus (UTEX 1237), while in Gutman et al. [20], the experiment took 7 days; the inoculated ratio was not reported and the infected Scenedesmus species was S. vacuolatus (one specific strain). The average number of cysts, which H. pluvialis cells carried 3 h after inoculation, was 1; at the used inoculation ratio of 1:1, it implies that almost each propagule successfully encysted. For this reason, the prevalence and intensity in H. pluvialis remained constant until the second day, where the prevalence reached a value of 100% after the newly released propagules were attaching. The case of C. zofingiensis was different: the prevalence was moderate (10%) until the 8th hour, but sharply increased (to 75%) afterwards. These results imply a pulse of propagule release between 8 h and 12 h, and with an additional pulse to reach the 100% of prevalence occurring at 31 h after inoculation. The values of prevalence and intensity for S. dimorphus increased slowly during the first 12 h and reached a plateau of 20% prevalence and 0.4 of intensity; we therefore suggest that a propagule pulse did not occur. Since we compare host species with different cell sizes: H. pluvialis is 11 times and S. dimorphus is four times bigger than C. zofingiensis, we calculated the algal areal density to normalize the pressure suffered by the host cells per unit area. In this case, we found the highest cysts number per surface area in C. zofingiensis (5), intermediate in H. pluvialis (2.2) and low in S. dimorphus (0.1). We therefore suggest that the shorter P. sedebokerense life cycle found in C. zofingiensis (12 h) as compared to H. pluvialis (16 h) is due to full consumption of the host constituents in the smaller densely infected C. zofingiensis cells. This also explains why C. zofingiensis inoculated cultures collapse faster than H. pluvialis inoculated at the same ratio (data not shown). Why S. dimorphus reaches a prevalence plateau and promotes the development of P. sedebokerense resting cysts remains a mystery, but we can hypothesize that S. dimorphus is a suboptimal host. In studies of host range, it was found that parasites can infect suboptimal hosts and produce descendants that were only capable of infecting the original host [39]. The results presented in Table 2 further support those findings. In fact, S. dimorphus would be considered as reservoir host for P. sedebokerense, since it can be infected at a moderate density without suffering lethal epidemics, but promotes resting cyst formation ( Figure 6A), which can germinate in the short term ( Figure 6B) or resist unfavorable conditions [13,21]. According to our observations ( Figure 6E-H), S. dimorphus cells targeted by the parasite were mainly senescent, which have yellowish chloroplast autofluorescence, in line with previous study's pictures [21].
We investigated two parameters of the fitness starting with the same number of propagules inoculating 11:1 H. pluvialis, 1:1 C. zofingiensis and 4:1 S. dimorphus in different algal densities offering the same surface area, giving the same probability to the propagules to find a host ( Figure 5). P. sedebokerense propagules infect H. pluvialis (39% survival, 8 h after inoculation) with greater success than C. zofingiensis or S. dimorphus (11% survival, 8 h after inoculation). These successful encystment events in H. pluvialis produced one magnitude order more propagules than C. zofingiensis and S. dimorphus; C. zofingiensis produced more propagules than S. dimorphus, but only in the first sampling time (22 h). Why does infection of H. pluvialis provide higher success to P. sedebokerense than infection of C. zofingiensis? We suggest that it is related to cell size (the lower production in S. dimorphus can be related to cell size but also to being a suboptimal host), since bigger host cells offer higher amounts of nutrients, as it was reported before that host cell size affects sporangium size and the number of spores produced [43].
When we investigated cross infection, the virulence of the different inocula could be explained by the number of propagules which are produced in each inoculated culture. That is, as more propagules are produced in Ps-Hp coculture than in Ps-Sd coculture, more intensive infection occurs resulting in higher virulence of Ps-Hp than Ps-Sd as an inoculum. But the susceptibility of each microalgal species shown in the cross-infection experiment seems to be explained by, or at least be related to, the preference of the P. sedebokerense propagules to each algal species. Since in the mixed culture, P. sedebokerense showed higher preference to the more sensitive species (H. pluvialis), the preference showed by the parasite could be interpreted as evolutionary adaption to the best source of food. For instance, Rhizosiphon akinetum evolved to only infect the akinetums of the phytoplanktonic cyanobacteria Planktothrix, presumably because infecting akinetes provides more energy than infecting regular cells [11].
We inoculated the three different algae with different inoculum types (Figure 2), in different algal densities, and different P. sedebokerense amounts, and we never observed S. dimorphus culture collapse, even if inoculating with a high number of P. sedebokerense cells and maintaining the culture for weeks. We observed partial flocculation of the S. dimorphus cultures and confirmed the partial infection by microscopic observations. These findings are in agreement with the findings of Gutman et al. [20] for another Scenedesmus strain but, contradicting those of Letcher et al. [21] who isolated P. sedebokerense from an outdoor collapsed culture of S. dimorphus strain used in the current study. According to our results, using P. sedebokerense AZ_ISR strain, mainly senescent cells of S. dimorphus were infected and culture collapse was never observed as reported by Letcher et al. [21] using P. sedebokerense strain FD16. The observed differences in virulence of P. sedebokerense against S. dimorphus in our work and Lechter et al.'s work can be attributed to being different parasite strains with possibly different host ranges or host preferences. However, since the culture crush reported by Letcher et al. [21] was only in outdoor culture, we can speculate that this crush of the algal culture was produced due to outdoors abiotic conditions which are different from our tested indoor conditions. In addition, since outdoor cultures are not axenic, we also can consider biotic factors, including another agent causing the culture crush as coinfection with P. sedebokerense. More tests with both strains, under the same conditions are needed before conclusion can be drawn up regarding this parasite virulence towards different algal hosts.

Conclusions
Our system offers a useful tool to compare infections in different microalgae cultures, and the inoculation ratio can be adjusted for the specific purpose (e.g., can be decreased to mimic natural conditions).
Among the three economically valuable microalgae tested, P. sedebokerense showed higher preference to H. pluvialis when the three microalgae are present, correlating with the higher susceptibility of H. pluvialis and the faster infection development.
The life cycle of P. sedebokerense is shorter than previously assumed and this needs to be considered when developing effective strategies to control microalgae fungal pests. Rapid isolation and/or treatment of contaminated cultures is essential.
The infective cycle in C. zofingiensis was shorter than in H. pluvialis; we therefor suggest that this is due to the smaller host cell size.
We consider S. dimorphus as a reservoir host of P. sedebokerense since is not collapsing after infection and promotes resting cyst formation. Reservoir hosts would be a threat to be considered by microalgal industry, since their cultures could show less symptoms and spread the parasitic pest.
Supplementary Materials: The following are available online at https://www.mdpi.com/2309-608 X/7/2/100/s1, Figure S1: isolated propagules of P. sedebokerense, Table S1: Surface areas of microalgae, Table S2: Inoculation details, Table S3 Table S3, the ITS sequence of the strain used in this study is available in GenBank as detailed in the Material and Methods section.