DSOF: A Rapid Method to Determine the Abundance of Microalgae and Methanotrophic Bacteria in Coculture Using a Combination of Differential Sedimentation, Optical Density, and Fluorescence

Abstract

Cocultivation of microalgae and aerobic methanotrophs represents an emerging biotechnology platform to produce high-protein biomass, yet quantifying individual species in mixed cultures remains challenging. Here, we present a rapid, low-cost method—differential sedimentation, optical density, and fluorescence (DSOF)—to determine the abundance of coculture members. DSOF exploits differences in cell size and pigment autofluorescence between the thermoacidophilic microalga and methanotrophic species Galdieria sp. RTK37.1 and Methylacidiphilum sp. RTK17.1, respectively, to selectively sediment algal cells and estimate population contributions via OD₆₀₀ and phycocyanin fluorescence. Evaluation with model suspensions across a wide cell density range (0 ≤ [Galdieria]: ≤ 3.23 A.U., and 0 ≤ [Methylacidiphilum] ≤ 1.54 A.U.) showed strong agreement with known values, with most absolute errors < 0.1 A.U. and relative errors < 10% at moderate biomass levels. Application to live batch cocultures under microalga or methanotroph growth-suppressed conditions, and during simultaneous growth, demonstrated accurate tracking of population dynamics and revealed enhanced methanotroph growth in the presence of oxygenic microalgae. While DSOF accuracy decreases at very concentrated biomass (>2.0 A.U. for Galdieria) or under nitrogen-limiting conditions, the model provides a practical, scalable alternative to more complex, invasive or expensive techniques, enabling near real-time monitoring of microalgae–methanotroph cocultures.

1. Introduction

The cocultivation of microalgae and aerobic methane-oxidizing bacteria (methanotrophs) presents a promising strategy for sustainable biotechnology applications, particularly in the development of high-protein biomass suitable for animal feed [1,2,3]. These microbial partnerships leverage complementary metabolisms (e.g., oxygen production and consumption), creating synergistic interactions under specific environmental conditions [1]. Thermoacidophilic microalgae–methanotroph cocultures further offer a robust and resilient platform for single cell protein (SCP) production, as their ability to thrive in extreme conditions limits contamination and supports continuous cultivation [4,5]. Recently, cocultures of the microalga Galdieria sp. RTK37.1 and the methanotroph Methylacidiphilum sp. RTK17.1 produced biomass with nutritional quality comparable to soybean meal and fishmeal, highlighting their potential as a high-protein feed for animals [6]. Beyond protein yield, such cocultures align with circular bioeconomy principles, enabling methane bioconversion into SCP and other products, while mitigating greenhouse gas emissions and integrating into wastewater valorisation systems [7].

Differentiating the relative proportions of each microorganism in coculture is critical for optimizing protein-enriched biomass production and ensuring consistent process control. Galdieria spp. have been reported to exhibit high protein content—ranging between ~26 and 68% (w/w) and contain a significant fraction of essential amino acids (~43%) [6,8,9]. Similarly, methanotrophic or methylotrophic bacteria can exhibit protein levels approximating 50–70% of dry biomass [6,10]. In both algal–bacterial systems and applied settings such as wastewater treatment, inoculum ratios strongly influence biomass yield, protein/carbohydrate/lipid composition, and granule stability, all of which depend on accurate differentiation of algal and bacterial fractions [3,11,12,13,14]. Accurate quantification of coculture members is therefore essential not only for studying microbial interactions but also to maximize productivity, reproducibility, and nutritional quality in industrial-scale SCP production.

Microalgae–methanotroph coculture dynamics are often reported as a total biomass concentration (i.e., optical density, dry biomass) without differentiating the relative proportion of each microorganism [1,6]. When relative abundance is reported, flow cytometry is typically used for direct cell counting [3,15]. Rarely, different techniques are combined to quantify coculture dynamics (e.g., flow cytometry in concert with selective plate counting) [16]. Cell counting via flow cytometry is typically fast, with low instrumental error (<5%) and high sensitivity [17]. However, it requires dedicated, expensive equipment and highly trained technical support [18,19]. In addition, cell fixation is sometimes required to preserve samples or to make cells permeable to dyes [20]. If protocols are not optimized, substantial count errors can result [21]. Other enumeration methods, such as calculating biomass concentrations using direct cell counts, can be problematic as biomass accumulation can occur in the absence of cell division via glycogen [22,23,24] or polyhydroxybutyrate (PHB) accumulation [25]. More recently, non-invasive monitoring tools such as online fluorescence probes and imaging-based sensors have emerged, but they lack taxonomic specificity in mixed cultures [26,27]. Broader bioprocess monitoring technologies, including software sensors, machine learning algorithms, and multi-spectral probes, have advanced in photobioreactor systems, yet their application to complex cocultures remains underdeveloped [26,28,29]. Monitoring mixed microalgae–bacteria systems remains a major challenge, as physical and biological traits (i.e., cell aggregation, sedimentation, and pigment content) influence biomass dynamics but are difficult to disentangle with conventional methods [30]. This highlights the need for rapid, scalable tools to differentiate microbial groups within consortia. Existing techniques, such as flow cytometry and qPCR, though powerful, are often too costly and resource-intensive for routine use. To scale up coculture-based SCP production, simple, rapid, and inexpensive alternatives are required to replace these invasive and high-cost methods.

Computational methods can quantify the proportional biomass of photoautotroph-methanotroph coculture members by combining mass-balance approaches with individual growth yield coefficients; for instance, the E-C method [21] combines total optical density (with wavelength dependent on the coculture pair), headspace gas composition (e.g., CH₄, CO_2, and O₂), and dissolved CO₂ concentrations. Unfortunately, its reliance on monitoring gas concentration dynamics makes it impractical in small-volume reactors, as repeated headspace samplings are likely to impact experimental conditions. Additionally, changes to headspace pressure or gas flow rates could cause significant errors if not properly accounted. Similar methods, which inject an inert gas (e.g., Ar) to compensate for pressure loss or serve as a tracer, have key limitations as they can: (1) alter headspace composition, (2) interfere with GC analyses due to common carrier gases (Ar, N₂, He), and (3) fail when cocultures generate more gas than they consume [31]. Finally, mass balance-based methods for disambiguating coculture member species require precise knowledge of yield coefficients and often assume their constancy, which is rarely the case. For instance, in Scenedesmus obliquus–Methylocystis bryophila cocultures, biomass yields increased from 0.18 to 0.30 mol_C mol_CH4⁻¹ as O₂ increased from 10 to 50% v/v [32]. Similar variability is likely with Methylacidiphilum spp., which can utilize metabolites from partner species (e.g., methanol, H₂/CO₂, formate), altering CH₄-to-biomass yields [23,33,34].

Nevertheless, the physical characteristics of microorganisms (e.g., cell size, shape, density, and fluorescence) can aid in estimating the relative abundance of species within a coculture. Optical density (OD) is commonly used to correlate biomass concentration in axenic cultures [31], with specific wavelengths employed to quantify Galdieria spp. (e.g., 750 and 800 nm) [22,35,36] and Methylacidiphilum spp. (e.g., 600 nm) [33,37]. However, OD at a single wavelength does not produce unique biomass profiles in mixed cultures [31]. Microalgae possess photosynthetic pigments (e.g., chlorophyll, phycocyanin, and carotenoids) that exhibit autofluorescence and are often measured in vivo to estimate pigment concentration [38,39,40]. For example, phycocyanin in Galdieria fluoresces at 670 nm when excited at 590 nm. While pigment content can approximate photosynthetic biomass [41], it fluctuates with cell physiology, age, light quality, and growth conditions, limiting its reliability for tracking algal growth [38,42]. Recent studies have explored integrating spectroscopic approaches, combining optical density and fluorescence to improve algal biomass estimation [43]. These methods demonstrate the value of leveraging multiple optical signals for rapid and low-cost quantification in monocultures, yet they remain largely untested in mixed microbial systems where overlapping signals complicate analysis.

In this study, we aimed to design and evaluate a simple, low-cost method to quantify the relative abundance of thermoacidophilic microalgae and methanotrophic bacteria in cocultures by combining differential sedimentation, optical density, and autofluorescence (hereafter referred to as DSOF). Unlike flow cytometry or mass-balance models, the proposed method relies on easily measurable physical properties, such as cell size and pigment fluorescence, allowing quantification of coculture members with minimal equipment. Using this approach, the individual contributions of Galdieria sp. RTK37.1 and Methylacidiphilum sp. RTK17.1 were obtained across a broad range of biomass concentrations with acceptable error. Collectively, these findings demonstrate that the DSOF method provides a practical, low-cost, and scalable approach for quantifying coculture member dynamics, offering a valuable alternative to more complex or invasive techniques for resolving species abundance in microalgae–methanotroph systems.

2. Materials and Methods

The DSOF method was developed to exploit differences in the settling velocity and autofluorescence of a microalga (Galdieria sp. RTK37.1) and a methanotrophic bacterium (Methylacidiphilum sp. RTK17.1) and to determine the concentration of each microorganism in coculture. As a conceptual overview, a small volume of coculture is harvested and briefly centrifuged under weak centrifugal forces. Supernatant fluorescence and optical density are then compared to those obtained from the uncentrifuged sample. With these four values, the optical density (and by correlation, biomass concentration) of the bacterium (Methylacidiphilum sp. RTK17.1) and microalga (Galdieria sp. RTK37.1) in the original coculture can be calculated (Figure 1).

The basis of the DSOF method is to exploit the cell size difference between species: Galdieria sulphuraria cells range between ~3 and 9 µm in diameter [44], whereas Methylacidiphilum spp. are smaller rods (0.8–2.0 µm length, 0.40–0.65 µm width) [45,46]. According to Stokes’ law, sedimentation velocity scales with the square of particle diameter [47], suggesting Galdieria cells settle ~100× faster than Methylacidiphilum. Therefore, gentle centrifugation (<400× g) could selectively pellet microalgal cells while leaving bacterial cells in suspension. Measuring the change in OD and pigment fluorescence in the resulting supernatant could then be used to estimate the quantity of separated microalgae—and by extension, their abundance in the original sample.

The DSOF method assumes that at 600 nm, the absorbance of Methylacidiphilum sp. RTK17.1 and Galdieria sp. RTK37.1 is additive (Figure A1), so for any given coculture sample:

Supposition 1: OD₆₀₀ values are additive for Methylacidiphilum sp. RTK17.1 and Galdieria sp. RTK37.1, then:

$${\left[Methylacidiphilum\right]}_1+{\left[Galdieria\right]}_1={\left[\mathrm{C}\mathrm{o}\mathrm{c}\mathrm{u}\mathrm{l}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e}\right]}_1$$

(1)

where subscript “1” denotes measurements made to the uncentrifuged coculture sample. Due to their larger cell size, Galdieria sp. RTK37.1 exhibits significantly faster settling velocities than Methylacidiphilum sp. RTK17.1. Therefore, brief centrifugation at low centrifugal force is expected to selectively pellet the microalgal cells while leaving most methanotrophs suspended. As cells settle out of suspension, the optical density of the coculture decreases; thus, the OD of the supernatant after centrifugation reflects the remaining, predominantly bacterial, biomass:

$${\left[Methylacidiphilum\right]}_2+{\left[Galdieria\right]}_2={\left[\mathrm{S}\mathrm{u}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{t}\right]}_2$$

(2)

where subscript “2” refers to measurements made to the supernatant resulting from coculture centrifugation. The second assumption for the DSOF method is that there is no significant sedimentation of methanotrophic cells:

Supposition 2: There is no significant sedimentation of methanotroph cells (Figure A2), then:

$${\left[Methylacidiphilum\right]}_1={\left[Methylacidiphilum\right]}_2$$

(3)

Galdieria sp. RTK37.1 cells are partially removed by centrifugation, so in the supernatant, only a fraction of microalgae cells, “y”, remains, then the following is written:

$${\left[Galdieria\right]}_2=\mathrm{y}·{\left[Galdieria\right]}_1$$

(4)

Phycocyanin, a pigment produced by Galdieria spp., exhibits fluorescence at 670 nm when excited at 590 nm [40]. If we assume Methylacidiphilum sp. RTK17.1 shows no fluorescence (Figure A3), and the fluorescence of phycocyanin is proportional to the microalgal cells remaining in suspension (Figure A4), which in turn is proportional to the optical density. Then, by measuring fluorescence of the original coculture (F₁) and the supernatant (F₂), we can approximate y as follows:

Supposition 3: The fraction of microalgae remaining in the supernatant can be approximated by the proportion of fluorescence of supernatant and original coculture as follows:

$$\mathrm{y}={\displaystyle \frac{{\left[Galdieria\right]}_2}{{\left[Galdieria\right]}_1}}={\displaystyle \raisebox{1ex}{${\mathrm{F}}_2$}\!\left/ \!\raisebox{-1ex}{${\mathrm{F}}_1$}\right.}$$

(5)

Substituting Equations (4) and (3) in (2), to obtain the following system of equations:

$${\left[Methylacidiphilum\right]}_1+{\left[Galdieria\right]}_1={\left[\mathrm{C}\mathrm{o}\mathrm{c}\mathrm{u}\mathrm{l}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e}\right]}_1$$

(1)

$${\left[Methylacidiphilum\right]}_1+\mathrm{y}·{\left[Galdieria\right]}_1={\left[\mathrm{S}\mathrm{u}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{t}\right]}_2$$

(6)

Thus, we can solve for ${\left[Galdieria\right]}_1$:

$${\left[Galdieria\right]}_1={\displaystyle \frac{{\left[\mathrm{C}\mathrm{o}\mathrm{c}\mathrm{u}\mathrm{l}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e}\right]}_1-{\left[\mathrm{S}\mathrm{u}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{t}\right]}_2}{1-\mathrm{y}}}$$

(7)

And for ${\left[Methylacidiphilum\right]}_1$:

$${\left[Methylacidiphilum\right]}_1={\left[\mathrm{C}\mathrm{o}\mathrm{c}\mathrm{u}\mathrm{l}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e}\right]}_1-{\displaystyle \frac{{\left[\mathrm{C}\mathrm{o}\mathrm{c}\mathrm{u}\mathrm{l}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e}\right]}_1-{\left[\mathrm{S}\mathrm{u}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{t}\right]}_2}{1-\mathrm{y}}}$$

(8)

2.1. Technical Description of the DSOF Method

To differentiate the contribution of Galdieria sp. RTK37.1 and Methylacidiphilum sp. RTK17.1 to the coculture OD₆₀₀, a 0.75 mL sample of coculture was harvested, and the optical density, [Coculture]₁, was measured using an Ultrospec 10 cell density meter (Amersham Bioscience, Buckinghamshire, UK). Fresh V4 medium was used as a blank. If the sample OD₆₀₀ was greater than 0.6, it was diluted in the modified V4 nutrient medium and remeasured. Next, 200 μL coculture was transferred into a black, flat-bottom (chimney well) 96-well microplate (Greiner Bio-One, Kremsmunster, Austria), and the fluorescence emission, F₁, at 670 nm (590 nm excitation), was measured using a microplate reader (Varioskan Lux, Thermo Scientific, Waltham, MA, USA). Following this, a 1.5 mL coculture sample was transferred into a 1.5 mL Eppendorf tube and centrifuged at 380× g for 20 s (Biofuge Pico, Heraeus Instruments, Hanau, Germany). The resulting supernatant (0.75 mL) was then collected, and the OD₆₀₀, [Supernatant]₂, and fluorescence (F₂) were measured as described for the coculture above. Finally, the OD₆₀₀ values for Galdieria sp. RTK37.1 and Methylacidiphilum sp. RTK17.1 in coculture was then calculated using Equations (7) and (8), respectively.

2.2. Growth Medium and Culture Maintenance

Complete information on strain isolation and maintenance is provided in our previous work [6,22,33]. Briefly, both Methylacidiphilum sp. RTK17.1 and Galdieria sp. RTK37.1 were isolated from geothermally heated soils at Parariki in the Rotokawa geothermal area, Aotearoa-New Zealand [22,33]. Prior to experimentation, Methylacidiphilum sp. RTK17.1 was routinely maintained in chemostat culture at 50 °C and pH ~2.5 using a 1 L bioreactor (BioFlo 110; New Brunswick Scientific, Edison, NJ, USA, with 600 mL working volume, 0.0069 h⁻¹ dilution rate, and agitation at 800 rpm) supplied with a CO₂/CH₄/O₂/N₂ gas mixture [33]. As previously described [6], Galdieria sp. RTK37.1 was propagated in batch cultures within sealed 1 L bottles containing V4 medium under 80% CO₂/20% N₂, incubated at 45 °C with continuous agitation and light. The bottles were incubated horizontally in a Lab Companion shaking incubator (Cole-Parmer, Chicago, IL, USA) at 110 rpm. Light was provided via three 50 W halogen lightbulbs and adjusted to 60 µmol_photons m⁻² s⁻¹. Galdieria cultures were routinely diluted to an OD₆₀₀ of 1.0 at the beginning of each cycle and maintained until OD₆₀₀ 6.0, at which point the next cycle was initiated.

A modified V4 medium [48] was used for all cultivation experiments. Briefly, the medium contained per liter: 0.4 g NH₄Cl, 0.05 g KH₂PO₄, 0.02 g MgSO₄·7H₂O, 0.01 g CaCl₂·2H₂O, 3 mL FeEDTA solution, 3 mL trace elements solution, 1 mL trace metals solution, 0.2 µM Ce₂(SO₄)₃, and 0.2 µM La₂(SO₄)₃ solution and was adjusted to pH 2.5 with H₂SO₄. The FeEDTA solution was prepared by dissolving 1.54 g of FeSO₄·7H₂O and 2.06 g of Na₂EDTA in one L of deionized water. The trace element solution was prepared by dissolving (per 1 L deionized water): 0.44 g ZnSO₄.7 H₂O, 0.20 g CuSO₄.5 H₂O, 0.19 g MnCl·4H₂O, 0.06 g Na₂MoO₄.2 H₂O, 0.10 g H₃BO₃, and 0.08 g CoCl₂.6 H₂O. To prepare the trace element solution, 1.5 g of Nitrilotriacetic acid was dissolved in 800 mL of deionized water and the pH was adjusted to 6.5 with KOH. Then the following minerals were dissolved in order: 0.2 g Fe(NH₄)₂(SO₄)₂·6H₂O, 0.2 g Na₂SeO₃, 0.1 g CoCl₂·6H₂O, 0.1 g MnSO₄·2H₂O, 0.1 g Na₂MoO₄·2H₂O, 0.1 g Na₂WO₄·2H₂O, 0.1 g ZnSO₄·7H₂O, 0.04 g AlCl₃·6H₂O, 0.025 g NiCl₂·6H₂O, 0.01 g H₃BO₃, and 0.01 g CuSO₄·5H₂O. The pH was then adjusted to 7.0, and the volume brought to 1 L. Unless stated otherwise, all chemicals and reagents were purchased from Sigma-Aldrich (Darmstadt, Germany), and all gases were sourced from BOC (North Ryde, Australia).

2.3. Biomass Determination

Biomass concentrations were estimated from optical density at 600 nm (OD₆₀₀) using previously established calibration curves [6]. For Galdieria sp. RTK37.1, 1.0 OD₆₀₀ corresponded to 0.308 gDW L⁻¹ in the range 0–9 OD₆₀₀. For Methylacidiphilum sp. RTK17.1, 1.0 OD₆₀₀ corresponded to 0.435 gDW L⁻¹ in the range 0–2.5 OD₆₀₀. These calibrations were used to convert OD₆₀₀ measurements into biomass concentrations throughout this study.

2.4. DSOF Evaluation in Model Suspensions

To test the DSOF method, model “coculture” suspensions were prepared from axenic stock suspensions across a range of predefined OD₆₀₀ values. For the methanotroph stock, the exponential phase of Methylacidiphilum sp. RTK17.1 cells were harvested from a 1 L bioreactor (BioFlo 110; New Brunswick Scientific, Edison, NJ, USA) grown in a chemostat (Dilution rate = 0.28 day⁻¹) on a gas mixture of 67% CO₂, 1.0% CH₄, 2.1% O₂ (balance N₂, all v/v) supplied at 16 mL min⁻¹. Temperature was kept at 45 °C, agitation at 400 rpm, and pH 2.5 (not controlled). Cells were stored at 4 °C until required, and then centrifuged (300 mL at a time in six 50-mL Eppendorf tubes) at 5000× g for 15 min in a 5810R Benchtop Centrifuge (Eppendorf, Hamburg, Germany, temperature controlled to 4 °C). Then, they were resuspended with sufficient V4 medium to achieve an OD₆₀₀ of 4.0 A.U. For the microalgae stock, 300 mL V4 medium (pH 2.5) was inoculated with Galdieria sp. RTK37.1 to an initial OD₆₀₀ of 1.0 A.U. within 1 L Duran Pressure Plus Bottles equipped with bromobutyl rubber stoppers. The bottles were then subjected to a vacuum for 3 min, and re-pressurized to 5 psi with an 80% CO₂ and 20% N₂ (v/v) gas mixture. The bottles were incubated in a shaker at 150 rpm and 45 °C. Illumination was supplied by warm white LED strips, providing 40 µmol photons m⁻² s⁻¹, measured at the outer surface of the bottle wall. Cells were harvested at OD₆₀₀ = 3.5 A.U., which corresponds to the point at which NH₄⁺ is depleted in the medium [6], to ensure cultures were never nitrogen-limited.

Several model coculture groups were prepared from the stock suspensions and analyzed with the DSOF to determine the method’s precision and accuracy. For each group, one of the microorganisms’ OD₆₀₀ was kept constant, while the other was varied in defined increments. In total, eight groups were analyzed, five where the microalgae OD₆₀₀ was constant ([Galdieria]_K: 0.28, 0.52, 1.00, 1.81, and 3.23 A.U.), and three where the methanotroph OD₆₀₀ was constant ([Methylacidiphilum]_K: 0.44, 0.90, and 1.54 A.U.). For each suspension within each group, aliquots of both Galdieria sp. RTK37.1 and Methylacidiphilum sp. RTK17.1 was diluted into sterile V4 medium to the desired final optical density. Ten mL of each suspension was prepared. Triplicates of each model mixture were analyzed.

2.5. DSOF Evaluation in Live Batch Cocultures

To evaluate the DSOF method on growing cultures of Galdieria sp. RTK37.1 and Methylacidiphilum sp. RTK17.1, batch cocultures were performed under one of three conditions: (1) light, no CH4 to promote microalgae growth and suppress methanotroph growth, (2) no light, added CH4 to suppress microalgae growth and promote methanotroph growth, or (3) light, added CH4 to promote growth of both microorganisms. For these experiments, the previously described stocks of microalgae and methanotroph were used to inoculate 1 L Duran Pressure Plus Bottles, equipped with bromobutyl rubber stoppers, with 250 mL of broth in V4 medium (pH 2.5). The growth-suppressed cultures had a starting Methylacidiphilum sp. RTK17.1 OD₆₀₀ value of 0.45 A.U., and a starting Galdieria sp. RTK37.1 value of 0.45 A.U. The non-suppressed cultures had initial OD₆₀₀ values of 0.3 A.U. for Methylacidiphilum sp. RTK17.1, and 1.0 A.U. for Galdieria sp. RTK37.1. To promote methanotroph and suppress microalgal growth, bottles were covered with aluminum foil, sealed, and then 100 mL of CH₄ and 60 mL of CO₂ were injected. To promote microalgal growth and suppress methanotroph growth, bottles were injected with 160 mL of CO₂, no CH₄, and supplied with 40 µmol photons m⁻² s⁻¹ warm white LED lighting (i.e., no foil covering). To promote the growth of both methanotrophs and microalgae, bottles were injected with 100 mL of CO₂, 60 mL of CH₄, and illuminated as above.

Table A1. Initial CO₂, CH₄, and O₂ headspace gas concentrations for DSOF evaluation in live coculture batch experiments. Uncertainties represent one standard deviation (n = 3).

Desired Condition	Headspace Concentrations (% v/v)
	O2	CH4	CO2
1. Methylacidiphilum growth was suppressed	14.32 ± 0.01	0	17.29 ± 0.23
2. Galdieria growth suppressed	14.27 ± 0.05	10.00 ± 0.04	6.55 ± 0.07
3. Galdieria and Methylacidiphilum growth allowed	15.06 ± 0.09	6.94 ± 0.15	10.60 ± 0.04

shows the initial headspace gas concentrations for all conditions. The bottles were then cultivated on a shaking incubator (WiseCube WIS-10, Wisd Laboratory Instruments, Wertheim, Germany) at 45 °C and 150 rpm. Each condition was undertaken in triplicate.

Throughout the incubation period, 2 mL liquid samples were harvested daily and analyzed using the DSOF method. To ensure that CO₂/CH₄/O₂ concentrations weren’t limiting growth, 20 mL headspace gas samples were analyzed daily using a 490 micro-GC equipped with a thermal conductivity detector (Agilent Technologies, Santa Clara, CA, USA). To replenish gases to their initial concentrations, after sampling, the bottles were opened on a Laminar Flow cabinet, left to equalize for 30 min, and re-gassed (Table A1). Growth was stopped after six days or when the total OD₆₀₀ surpassed 3.5 A.U. Growth rates were calculated by fitting the appropriate concentration data to the exponential growth equation model in Prism Graphpad 9.4.1 ($X=X_0{e}^{{\mu }_{max}t}$, where X is the concentration in A.U., X₀ is the concentration at time zero in A.U., µ_max is the specific growth rate in h⁻¹, and t is time in hours). Comparison of growth rates (unpaired two-tailed t-tests, α = 0.05) was performed using Prism Graphpad 9.4.1.

3. Results and Discussion

3.1. Evaluation of the DSOF Method in Model Suspensions

Evaluation of the DSOF method in model (non-growing) suspensions confirmed good agreement between derived values (e.g., [Galdieria]_D) from the DSOF method and known OD₆₀₀ values of each microorganism (e.g., [Galdieria]_K) in coculture (Figure 2). To assist with data visualization, known concentrations (e.g., [Methylacidiphilum]_K) were normalized by dividing them by the corresponding concentration of the coculture partner (e.g., [Galdieria]_K). Indicating a strong correlation, slopes of the derived concentration regressions were not significantly different from the slope of known concentration lines (as y = mx, with m = [Galdieria]_K, p-value > 0.05, using a two-tailed unpaired t-test, Figure 2a). Likewise, the intercepts of derived regressions and the known concentration lines were not significantly different (Figure 2b, p-value > 0.05, using a two-tailed unpaired t-test). A similar analysis for the samples where Methylacidiphilum OD₆₀₀ was kept constant along each group (Figure 3), shows no significant difference between the slopes of the derived concentration regressions and the slope of known concentration (p-value > 0.05 using a two-tailed unpaired t-test).

The relative and absolute errors of the derived concentrations were also analyzed (Figure 4). A summary of the findings can be seen in

Table 1. Summary of the relative and absolute errors of the derived concentrations of Methylacidiphilum sp. RTK17.1 and Galdieria sp. RTK 37.1 in artificial suspensions using the DSOF method.

Known Values (A.U.)		Methylacidiphilum sp. RTK17.1 Error		Galdieria sp. RTK37.1 Error
RTK37.1 (A.U.)	RTK17.1 (A.U.)	Absolute (A.U.)	Relative (%)	Absolute (A.U.)	Relative (%)
<0.5	<0.5	±0.10	<10	±0.05	<20
<0.5	0.5–1.5	±0.10	<10	±0.10	<10
0.5–1.0	<1.0	±0.10	<10	±0.10	<10
0.5–1.0	1.0–1.5	±0.10	<10	±0.10	10–15
1.0–2.0	<0.5	±0.10	<10	±0.10	<10
1.0–2.0	0.5–1.0	±0.10	<10	±0.10	<10
1.0–2.0	1.0–1.5	±0.15	<10	±0.15	<10
2.0–3.2	<1.0	±0.10	<20	±0.10	<5
2.0–3.2	1.0–1.5	±0.20	<10	±0.15	<10

. For [Galdieria]_K ≤ 1.00 A.U., relative errors for both [Galdieria]_D and [Methylacidiphilum]_D remained < 10%, with absolute errors within ± 0.1 A.U. However, as [Galdieria]_K increased, so did relative and absolute errors. When [Galdieria]_K > 1.00 A.U., errors and standard deviations became more pronounced, particularly when [Methylacidiphilum]_D was < 0.5 A.U. at [Galdieria]_K OD₆₀₀ = 3.23 A.U. Nevertheless, even at higher concentrations, most relative errors remained below 20%, and absolute errors were within ± 0.2 A.U. These trends align with the behavior predicted by Stokes’ law for ideal suspensions, where particles sediment independently at low concentrations [49]. As concentration increases, cells collide more frequently, bringing methanotrophs and microalgae into close contact. These interactions could promote adhesion between cells. Such interactions resemble bioflocculation, where the presence of bacteria can induce the rapid formation of multicellular aggregates in bacteria–microalgae cocultures [2]. Floc formation occurs when microorganisms come into contact and adhere to one another [50], often facilitated by exopolysaccharides (EPS), which neutralize cell surface charges and promote aggregation [51]. Galdieria sulphuraria can produce up to 115.8 mg L⁻¹ of EPS [52], while EPS production in Methylacidiphilum spp. has not been documented. However, other methanotrophs, such as Methylomicrobium alcaliphilum 20z, can secrete up to 2.64 g L⁻¹ of EPS [53]. Thus, it is plausible that EPS production occurs in Methylacidiphilum sp. RTK17.1 and Galdieria sp. RTK37.1 cocultures, contributing to the observed measurement errors. Floc formation is concentration-dependent [47], and aggregation efficiency is influenced by the bacterial–algal ratios [50]. Consequently, elevated Galdieria sp. RTK37.1 concentrations may, correspondingly, increase measurement errors. A similar phenomenon was reported in flow cytometric quantification of Methylococcus capsulatus and Chlorella sorokiniana cocultures [21], where methanotroph adhesion to microalgae led to measurement errors as high as 88.3% for M. capsulatus and 18.0% for C. sorokiniana. In contrast, this effect was not observed in cocultures involving species with similar cell sizes [21].

The DSOF method tended to underestimate methanotroph concentrations and overestimate microalgal concentrations. This bias likely arises from the underlying assumption that Methylacidiphilum sp. RTK17.1 cells remain entirely in the supernatant after centrifugation. In practice, even at weak centrifugal forces, a fraction of methanotroph cells were removed (Figure A2). At dilute methanotroph concentrations, this loss was minimal and below the threshold of spectrophotometric determination. As concentrations increased, however, the same fractional loss presented as a larger absolute number of cells. Since the DSOF method assumed total OD₆₀₀ as additive, underestimating methanotroph biomass entailed a corresponding inflation of microalgae biomass (Figure 4). The influence of increasing bacterial cell concentrations on the DSOF method was clearly demonstrated: at [Methylacidiphilum]ₖ = 1.54 A.U., deviations from known methanotroph concentrations remained within ± 0.1 A.U. (Figure 3b and Figure 5). These findings suggest that incorporating a concentration-dependent correction factor could improve DSOF accuracy for more concentrated suspensions by accounting for systematic methanotroph sedimentation. In contrast, Galdieria sp. RTK37.1 showed greater error at lower concentrations, primarily due to the unintended removal of methanotroph cells from the supernatant (Figure 5a). While this removal was relatively minor in absolute terms, it disproportionately affected estimates at low microalgal concentrations, resulting in higher relative error. A similar concentration-dependent bias has been reported in flow cytometry. For example, Badr et al. [21] observed deviations beginning at ≥0.015 g_DW L⁻¹ for M. alcaliphilum, ≥0.08 g_DW L⁻¹ for Synechococcus sp. PC7002, ≥0.04 g_DW L⁻¹ for M. capsulatus, and ≥0.06 g_DW L⁻¹ for C. sorokiniana. In comparison, the DSOF method showed deviations only at substantially greater concentrations (Galdieria: 0.308 g_DW L⁻¹, OD₆₀₀ = 1.00 A.U.; Methylacidiphilum: 0.222 g_DW L⁻¹, OD₆₀₀ = 0.50 A.U.). Nevertheless, most relative errors for Galdieria sp. RTK37.1 remained < 20%, and was typically under 10% for [Galdieria]ₖ ≥ 0.50 A.U. For Methylacidiphilum, errors were generally < 10%, with absolute errors falling between −0.2 and 0.05 A.U.

To assess DSOF performance, we conducted Bland–Altman and root mean squared error (RMSE) analyses (Figure 6). For Methylacidiphilum, the mean bias was 0.026 A.U. with 95% limits of agreement between −0.067 and 0.119 A.U.; only one point fell outside these bounds, and no trend was observed, indicating strong agreement across concentrations. The RMSE of 0.05 A.U. confirmed low prediction error relative to the OD range. For Galdieria, the mean bias was −0.036 A.U. with limits between −0.155 and 0.182 A.U. Most values fell within these bounds, especially at OD < 2.5 A.U., where no trend was evident. At higher OD (~3.2 A.U.), deviations increased, and some points exceeded the limits, consistent with greater error under elevated biomass. Still, the RMSE of 0.07 A.U. shows errors remained modest compared with the overall range. Overall, DSOF demonstrated minimal bias and strong agreement with reference values. It performs best at moderate concentrations, with larger deviations only at very high biomass. Even under these conditions, absolute errors remained within ±0.2 A.U., supporting DSOF as a practical, low-cost tool for coculture monitoring.

Direct comparison of DSOF with other coculture quantification methods is limited by scarce error reporting. The E-C protocol [21] quantified biomass in M. alcaliphilum–Synechococcus and M. capsulatus–C. sorokiniana cocultures, showing strong agreement with flow cytometry (R² = 0.90–0.98) and outperforming it at high concentrations. While E-C accuracy was unaffected by cell concentration within the tested ranges, it is unsuitable for Methylacidiphilum-Galdieria cocultures, as both species fix CO₂ autotrophically [54], creating an unaccounted CO₂ sink, and each can utilize alternative carbon sources [22,23,33,34,44,55,56], violating the mass-balance assumptions on which E-C relies.

3.2. Application of the DSOF Method to Monitor Growth Dynamics in Batch Cocultures

Next, the DSOF method was applied to evaluate member dynamics in actively growing batch cocultures. For these experiments, three conditions were tested: (1) methanotroph growth suppression (no CH₄), (2) microalgae growth suppression (no light), and (3) unrestricted growth of both microorganisms (Figure 7). As expected, in the absence of CH₄, Methylacidiphilum sp. RTK17.1 concentrations did not increase significantly from an initial OD₆₀₀ of 0.46 (±0.02 A.U., p-value > 0.05, Figure 7a). In contrast, Galdieria sp. RTK37.1 exhibited a three-day lag phase before growing from an initial OD₆₀₀ of 0.48 (±0.04 A.U.) to a final concentration of 2.77 (±0.12 A.U., 577% increase, µ = 0.017 h⁻¹). Conversely, in the absence of light, Galdieria growth was effectively suppressed, with OD₆₀₀ values remaining stable (0.47 ± 0.06 A.U.). Meanwhile, Methylacidiphilum sp. RTK17.1 cell concentrations increased by 331% (µ = 0.0055 h⁻¹), from an initial OD₆₀₀ of 0.46 (±0.06 A.U.) to 1.56 (±0.04 A.U.).

The DSOF method was also successfully able to quantify the simultaneous growth of both Methylacidiphilum sp. RTK17.1 and Galdieria sp. RTK37.1 in coculture (Figure 7c). Following a two-day lag period consistent with previous reports [52,57,58], Galdieria sp. RTK37.1 biomass increased from 1.01 A.U. to 2.04 (± 0.06 A.U., µ = 0.015 h⁻¹). Growth rates were not significantly different (p > 0.05) from those observed in the methanotroph-suppressed coculture (Figure 7a). Concomitantly, Methylacidiphilum sp. RTK17.1 concentrations increased from 0.31 to 1.46 A.U. (µ = 0.014 h⁻¹), a rate 2.55x faster than in the microalgae-suppressed coculture. This finding aligns with previous observations that the oxygenic activity of Galdieria sp. RTK37.1 is able to enhance CH₄ consumption of Methylacidiphilum sp. RTK17.1 in batch co-cultures [6]. When compared to axenic batch cultures, Galdieria growth rates were not significantly different (p > 0.05) from rates reported previously [6], while Methylacidiphilum growth rates were 12% faster (p < 0.05). This difference was likely due to the DSOF coculture experiments involving daily replenishment of O₂, CO₂, and CH₄ to prevent substrate limitations. Collectively, these results indicate that the DSOF method was able to accurately and precisely quantify relative microalgae and methanotroph cell concentrations, with acceptable error, in actively growing cocultures.

3.3. Outlook and Limitations of the DSOF Method

Despite its speed, simplicity, and reliability, the DSOF method has limitations. It has not been validated under nitrogen-limiting conditions, where both Galdieria spp. and Methylacidiphilum spp. may accumulate glycogen [4,5,22,24], potentially increasing methanotroph cell size or density, accelerating sedimentation, and altering separation efficiency. Nitrogen-starved Galdieria also loses pigments, particularly phycocyanin [59], which may impair fluorescence-based measurements. These effects could substantially increase measurement error, requiring caution and possible adjustment of centrifugation parameters, though further validation is needed. The method was also only tested at total OD₆₀₀ ≤ 5.0 A.U., with errors increasing at higher concentrations; introducing a concentration-dependent correction factor or diluting samples may mitigate this. A correction factor could be introduced into Equation (3) by incorporating a fractional sedimentation term for Methylacidiphilum. This factor could be modeled under different assumptions: (i) simple Stokes’ law behavior [47]; (ii) a Richardson–Zaki type relationship to capture biomass-dependent settling effects [47,49]; or (iii) empirical polynomial or logarithmic regression corrections, as commonly applied to OD-based biomass estimations [60,61,62]. The most appropriate model could then be calibrated using mixtures of Methylacidiphilum and Galdieria across a range of concentrations.

In this study, OD₆₀₀ measurements were recorded relative to sterile cultivation medium as a blank. While OD₆₀₀ provides a convenient and rapid proxy for biomass, it can also be influenced by extracellular metabolites or changes in medium composition over time, not exclusively by cell density [63,64]. In our defined media systems, such contributions appeared minimal; however, in natural or industrially derived feedstocks, suspended solids or metabolic byproducts could introduce additional scattering effects [64,65]. This limitation should be considered when applying DSOF in feedstocks containing suspended solids or with complex media, where background interferences may reduce measurement accuracy.

Because DSOF relies on differences in sedimentation rates and pigment fluorescence, it could potentially apply to other photoautotroph–methanotroph cocultures wherein large cell size differences exist. Thus, we speculate DSOF should be applicable to other microalgae–methanotroph cocultures, provided centrifugation parameters are optimized for maximum separation efficiency. However, minimal size differences, such as between Synechococcus spp. and M. alcaliphilum (

Table A2. Sizes of photoautotrophs and methanotrophs in commonly studied cocultures.

Strain	Guild	Morphology & Size (l × w)	Reference
Synechococcus spp. Methylomicrobium alcaliphilum	phototroph methanotroph	coccus: 0.6–2.1 µm rod: 1.2–3.0 µm	[66,67]
Chlorella sorokiniana Methylococcus capsulatus	phototroph methanotroph	coccus: 2–5 µm coccus: 0.7–1.0 µm	[68,69]
Galdieria spp. Methylacidiphilum spp.	phototroph methanotroph	ovoid: 3.8–5.0 µm rod: 0.8–0.65 µm	[45,70]

), for example, may render DSOF unsuitable for most cyanobacteria–methanotroph pairs.

Our results demonstrate that the DSOF method provides a simple, rapid, and accurate tool for quantifying the relative abundance of thermoacidophilic microalgae and methanotrophs in coculture. Unlike conventional methods such as flow cytometry or qPCR, DSOF requires only standard laboratory equipment and minimal sample preparation, which makes it practical for routine use in experimental and applied settings.

Accurate differentiation of coculture members is particularly important because the performance of microalgae–bacteria systems depends strongly on inoculum proportion, biomass stability, and biochemical composition. This has been demonstrated in applied contexts such as wastewater treatment, where inoculation strategies influence biomass yield, protein/carbohydrate/lipid content, and photo-granule formation [14]. The DSOF method could therefore support optimization of consortia performance in applied biotechnologies where monitoring mixed populations remains a major challenge.

Monitoring cocultures typically relies on high-resource methods such as flow cytometry, molecular tools, or microscopy. However, as noted in recent reviews, physical and biological traits such as aggregation, sedimentation, and pigment variability make accurate quantification difficult with these approaches [30]. By exploiting differences in sedimentation and autofluorescence, DSOF offers a simpler alternative that addresses these challenges without requiring specialized infrastructure.

Our findings also align with recent spectroscopic approaches that combine optical density, autofluorescence, and UV–Vis measurements to improve estimation of algal biomass [43]. While such methods have been validated in monocultures, their application in cocultures remains limited due to overlapping optical signals. DSOF extends this principle by incorporating a physical separation step, thereby enabling accurate quantification in mixed systems where purely spectroscopic methods may fail.

Beyond experimental validation, the scalability of DSOF suggests potential application as a Process Analytical Technology (PAT) tool for real-time bioprocess monitoring. Recent advances in PAT for synthetic cocultures have highlighted the need for robust, low-cost tools to track biomass dynamics and metabolite fluxes in real time [26]. By integrating sedimentation-based separation with optical readouts, DSOF provides a complementary and accessible approach that could be implemented in routine monitoring and control strategies for cocultures, including microalgae–methanotroph systems.

4. Conclusions

In this study, a method based on differential sedimentation, optical density, and autofluorescence was developed and validated in a series of model (non-growing) and actively growing coculture experiments. Findings show the DSOF method can accurately measure Galdieria sp. RTK37.1 and Methylacidiphilum sp. RTK17.1 and abundance with <0.1 A.U. absolute error when [Galdieria]_K ≤ 2.0 A.U., and [Methylacidiphilum]_K ≤ 1.5 A.U. These errors increase to ±0.2 A.U., for 2.0 A.U. < [Galdieria]_K ≤ 3.23 A.U. Overall, the DSOF method tends to underestimate Methylacidiphilum concentrations, likely because the method assumes no methanotroph cells sediment during centrifugation. Despite its speed and simplicity, DSOF has not yet been validated under nitrogen-limiting conditions, where glycogen accumulation and pigment loss in Galdieria could impair accuracy, nor at higher biomass concentrations where scattering effects and concentration-dependent errors become more pronounced. Nevertheless, the method’s reliance on sedimentation and pigment fluorescence makes it potentially adaptable to other microalgae–methanotroph cocultures, provided that sufficient differences in cell size exist and centrifugation parameters are optimized for separation efficiency

Collectively, the results show that the DSOF method is an easy, quick, and accurate method to measure relative concentrations of Methylacidiphilum sp. RTK17.1 and Galdieria sp. RTK37.1, which does not depend on gas concentrations and only requires small volumes of liquid sample. This allows for the rapid and simple measuring of the relative concentration of microalgae and methanotrophs in coculture. Therefore, this method enables near-‘real-time’ monitoring of coculture response to different growth conditions, which should help to expand understanding of microalgae–methanotroph cocultures and interactions, and support their potential application in circular bioeconomy processes. While the method was developed for one algal-methanotroph pair, the conceptual framework, combining differential sedimentation with optical and fluorescence measurements, could be adapted to other systems where cell size or pigment differences exist, thereby providing a practical tool for bioprocess optimization