Enhancement of euryhaline Asterarcys quadricellulare biomass production for improving biogas generation through anaerobic co-digestion with carbon rich substrate

Abstract

The microalga was isolated from Muktsar, the southwestern zone of Indian Punjab and identified as Asterarcys quadricellulare BGLR5 (MF661929) by 18S rRNA sequence analysis. The optimization of various cultural factors by the Plackett–Burman and central composite (CCD) designs helped in discerning the significant cultural factors for the increased production of biomass and other functional components (chlorophyll, carbohydrate, lipid and protein). The optimal cultural conditions as per the model were pH 9.9, 81 μmol m⁻² s⁻¹ light intensity, 22 °C temperature, growth period of 25 days, NaNO₃ 12 mM, 15 mM NH₄Cl, and 7 mM K₂HPO₄. In comparison to the basal condition biomass (0.886 g L⁻¹), a 0.42-fold increase in biomass yield was attained. Further, the highest yield of biogas (P: 361.81 mL g⁻¹ VS) with enhanced biogas production rate (R_m: 8.19 mL g⁻¹ day⁻¹) was achieved in co-digesting paddy straw with Asterarcys quadricellulare biomass in 1:1 ratio compared to their digestion individually. Further, the co-digestion resulted in the positive synergistic effect which increased the observed biogas yield compared to the estimated yield by 11–58% depending upon the amount of algal biomass and paddy straw used. Hence, the present study signifies that the biomass of Asterarcys quadricellulare BGLR5 can be utilized as a co-substrate with paddy straw to enhance the biogas yield.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13205-021-02792-x.

Introduction

Energy is one of the fundamental requirements for day to day life and the development of human civilization. However, the increased global population at an alarming pace has increased both utilization of fossil fuels and their demand, thereby resulting in the soaring up of energy reserves and unabated emission of greenhouse gases (GHGs). The search for environmentally viable and renewable sources of energy has gained importance (Ganesh Saratale et al. 2018; Saratale et al. 2016). The production of biofuels from food crops and food sources prompt various issues like depletion of resources and decreased biodiversity (Fenton and Ó Huallacháin 2012). Microalgae offer a fascinating alternate substrate for biofuel production as these are high yielding and can also be grown in areas (non-arable land) that are not suitable for crop production using fresh, saline or wastewater (Chen et al. 2015; Kerrison et al. 2015; Magdalena et al. 2018; Milledge et al. 2019). In addition to this, the microalgae utilize the CO₂ emissions (thus, contribute to CO₂ mitigation) and produce the carbohydrates, lipids and proteins which can be processed into various carbon–neutral biofuels (viz., methane, biodiesel, bioethanol, hydrogen) (Formighieri 2015; Klassen et al. 2017; Kruse and Hankamer 2010; Ryan Georgianna and Mayfield 2012).

Biofuels produced from microalgae are regarded to be a feasible alternative energy source (Borowitzka and Moheimani 2013; Klassen et al. 2017). The high water content of microalgae biomass and its less concentration in the culture medium affects the energy balance of the processes viz., gasification, pyrolysis, direct combustion that depend on the dry substrate (Fasaei et al. 2018; Milledge et al. 2019; Milledge and Heaven 2014), thus making it an unattractive substrate for aforementioned processes. However, it can serve as a preferred substrate for processes like anaerobic digestion (AD) that use wet biomass, thus saving energy used for the dewatering and drying processes (Barbot et al. 2016; Milledge et al. 2015). AD involves a sequential action of different groups of bacteria that metabolize organic materials into biogas (Ayala-Parra et al. 2017; Barbot et al. 2016). It can utilise all the organic carbon material (carbohydrates, lipids and proteins) of microalgae for methane production (Magdalena et al. 2018) rather than a specific compound like that of biodiesel from lipids (Allen et al. 2015). However, the low carbon to nitrogen (C:N) ratio of the microalgal biomass hampers the efficiency of the AD process (Fermoso et al. 2016; Ganesh Saratale et al. 2018; Klassen et al. 2017). Low C:N ratio (6:1) of algae elevates ammonia–nitrogen production which reduces the methanogenic activity and if accumulated in higher concentration can lead to the failure of the AD process (Mahdy et al. 2016). The ammonia accretion can be precluded by improving the C:N ratio (Zhen et al. 2016) and the possible strategy to improve it involves co-digesting it with the carbon-rich substrate (Dębowski et al. 2013; Zhen et al. 2016).

Paddy straw having high carbon content has a high C/N ratio and can, therefore, be co-digested with algal biomass. In India, the annual production of rice is about 136.5 million tonnes, and about 136.50–150.00 million tonnes of paddy straw are produced annually (www.indiastat.com). The utilization of paddy straw as a co-feedstock with microalgae for biogas production will also help in managing its on-field burning. This way it will protect the environment from GHGs produced from straw burning which otherwise has deleterious effects on climate change, affecting public health and killing of productive soil microflora and fauna, thereby affecting the organic matter of soil (Bhuvaneshwari et al. 2019). Further, the AD of microalgal demands the high-density cell cultivation of microalgae, which can be achieved through optimization of cultural conditions. Optimization of these cultural conditions for maximum biomass production is paramount for increasing the industrial production of microalgae. The optimization of culture conditions for the maximum growth rate involves repeating experiments and is time-consuming. Response surface methodology (RSM), a statistical approach of optimization, can facilitate it by determining the optimal values in less time with limited experimental runs.

The southwestern zone constituting 34% area of Indian Punjab is unsuitable for crop cultivation due to salinity and waterlogging problems. The livelihood of the small-scale farmers of this region is greatly affected. However, if this region could be used to cultivate the microalgae for biofuel production, it will not only produce a significant amount of bioenergy (biogas) but will also improve the economy of the people living in that area. Apart from bioenergy (biogas) production, the microalgae could serve as a source of other valuable nutraceutical and bioactive compounds. To the best of our know-how, as reflected in the literature not much attempt has gone into testing paddy straw (high C:N ratio) co-digestion with the euryhaline microalgae (having a low C:N ratio) compared to the freshwater microalgae. This has drawn our attention to isolate the microalgae from this area and utilize it with paddy straw for the biogas production to address both the issues (paddy straw management and utilization of salinity affected waterlogged area). The microalga used in this study was isolated from the salinity and waterlogged area of Indian Punjab, which is unexplored in terms of algal diversity. Therefore, the present study focused on the co-digestion of the euryhaline Asterarcys quadricellulare isolated from the Indian Punjab’s waterlogged area, with the paddy straw. The specific objectives of the present study include (i) to isolate and characterize the microalgal culture, (ii) to optimize and assess the cultural conditions and interactions between the operating variables for maximum biomass production of Asteracycs quadricellulare BGLR5, and (iii) to carry out the co-digestion of microalgal biomass (produced under optimized conditions) with paddy straw for biogas production and its kinetic study. We hypothesized that the co-digestion of a carbon-rich substrate with microalgal biomass produced under optimized conditions will enhance the biogas yield.

Materials and methods

Isolation, primary identification and preservation of microalgal isolate

The water sample collected in a sterile bottle from Marar of Muktsar (N 30° 32′ 477″, E 074° 40′ 903″), a waterlogged area of Indian Punjab, were brought to the laboratory. The physicochemical analysis of the water sample in terms of pH, electrical conductivity and salinity (CO₃²⁻, HCO₃⁻, Ca²⁺, Mg²⁺) was carried out. The pH was measured by a pH meter (Mettler Toledo™ SevenEasy 8603), electrical conductivity (EC) using a digital electrical conductivity meter (CC 603–1 Sanco) and the salinity (CO₃²⁻, HCO₃⁻, Ca²⁺, Mg²⁺) was determined through titration methods according to Richards et al. (1954). Microscopic observation (Magnus Icon 528,293 Freedom Model Olympus microscope) was carried to look for the existence of the microalgal species. After this, the algal water sample was enriched in BG-11 medium (Stanier et al. 1971). The composition of the medium is given in supplementary Table 1. For enrichment, 10 mL of each sample was inoculated in the Erlenmeyer conical flasks (250 mL) containing medium (100 mL), separately in triplicates. Then, after 20–30 days of incubation on achieving the visible growth in the flasks kept in the growth chamber maintained at 25 ± 2 °C and 120 rpm under the light intensity of 40.5 µmol m⁻² s⁻¹, the isolation of microalgal isolates was carried out. The conical flasks were continuously examined by an optical microscope for microalgal growth. The flasks which showed the growth was then used for isolation of the axenic algal culture. The microalgal culture was purified using the cocktail of antibiotics (to avoid bacterial and fungal growth) comprising of ampicillin (50 μg mL⁻¹), chloramphenicol (50 μg mL⁻¹), penicillin (70 μg mL⁻¹), amikacin (50 μg mL⁻¹) and amphotericin B (5 μg mL⁻¹), added @ 2 mL/100 mL of culture medium followed by repeated subculturing, serial dilutions and plating on petri plates containing BG-11 medium supplemented with 1.5% (w/v) of agar. The microalgal culture was allowed to grow in the medium containing antibiotics for 2–5 days and then retransferred to the fresh, sterile medium without antibiotics. All these procedures were adopted for every flask in repetition. The antibiotic usage, repeated subculturing, serial dilutions and streaking on the BG-11 agar plate and routine microscopic examination ensured the purity of the culture. The morphological study up to the genus level was done using optical microscopy aided with standard morphological feature keys (Bellinger and Sigee 2010; Wehr et al. 2003). The Olympus microscope (Magnus Icon 528,293 Freedom Model) attached with the camera (Debro 5.1 Megapixel digital) was used to produce the microphotographs (at 40X) of pure culture. Once the pure culture was obtained, it was tentatively labelled as BGLR5. The pure strain was then maintained under low light on a cool shelf, and continuously sub-cultured onto a fresh nutrient medium once a month or cultured in a fresh culture broth to maintain it.

Table 1.

Experimental design and results of Asterarcys quadricellulare BGLR5 in Placket–Burman design

Run	pH	Temp (°C)	Light Intensity (µmol m−2 s−1)	Growth Period (days)	NH4Cl (Mm)	NaNO3 (mM)	K2HPO4 (mM)	Biomass (g L−1)	Chlorophyll (mg L−1)	Carbohydrate (g L−1)	Lipid (g L−1)	Protein (g L−1)	Observed desirability	Predicted desirability
1	10.5	35	40.5	25	20	5	2	0.25 ± 0.03	10.31 ± 1.31	0.09 ± 0.03	0.03 ± 0.01	0.09 ± 0.02	0.34	0.33
2	10.5	25	81	15	20	5	8	0.92 ± 0.09	32.72 ± 6.16	0.36 ± 0.04	0.08 ± 0.01	0.19 ± 0.03	0.87	0.88
3	10.5	35	40.5	25	35	5	8	0.41 ± 0.04	9.72 ± 2.06	0.13 ± 0.03	0.04 ± 0.03	0.05 ± 0.02	0.37	0.38
4	7.5	25	40.5	15	20	5	2	0.52 ± 0.05	23.77 ± 3.75	0.18 ± 0.03	0.05 ± 0.01	0.14 ± 0.04	0.53	0.53
5	7.5	35	40.5	15	20	20	8	0.26 ± 0.06	4.72 ± 1.18	0.10 ± 0.01	0.02 ± 0.01	0.05 ± 0.01	0.25	0.24
6	7.5	25	81	25	35	5	8	0.91 ± 0.03	26.39 ± 4.48	0.34 ± 0.05	0.07 ± 0.03	0.19 ± 0.03	0.63	0.63
7	7.5	35	81	25	20	20	8	0.44 ± 0.04	8.77 ± 1.83	0.18 ± 0.03	0.03 ± 0.01	0.10 ± 0.01	0.37	0.38
8	10.5	25	40.5	15	35	20	8	0.60 ± 0.03	14.75 ± 2.37	0.25 ± 0.03	0.04 ± 0.02	0.16 ± 0.01	0.59	0.60
9	10.5	35	81	15	35	20	2	0.24 ± 0.06	5.35 ± 1.81	0.10 ± 0.01	0.02 ± 0.01	0.09 ± 0.02	0.30	0.29
10	7.5	35	81	15	35	5	2	0.31 ± 0.03	10.88 ± 2.33	0.07 ± 0.01	0.04 ± 0.01	0.06 ± 0.01	0.33	0.34
11	10.5	25	81	25	20	20	2	0.61 ± 0.05	18.96 ± 2.76	0.27 ± 0.03	0.04 ± 0.01	0.25 ± 0.03	0.66	0.67
12	7.5	25	40.5	25	35	20	2	0.412 ± 0.04	9.35 ± 2.23	0.17 ± 0.02	0.03 ± 0.01	0.17 ± 0.02	0.38	0.38

Values are means of triplicate ± SD (standard deviation); biomass is dry cell biomass

Extraction of DNA, PCR amplification and 18S rRNA sequencing

For molecular identification of BGLR5, the DNA was extracted from the culture according to the eukaryotic microalgal nucleic acids extraction method (EMNE), a combination of modified bead–phenol–chloroform (BPC) and modified freeze/thaw (F&T) method (Kim et al. 2012). The microalgal culture cells (1 mL) were harvested by centrifugation (10,000×g), resuspended in deionized water and then mixed with phenol (800 μl, pH 8). The silica beads were then added to the suspension and subjected to three cycles of freezing (liquid nitrogen) and thawing (water bath, 100 °C). After this, Mini-beadbeater (Sigma-Aldrich, India) was used to bead beat at 4000 rpm to break the cell wall and release the DNA from the cells. Thereafter, the sample was centrifuged (10,000×g for 5 min) and the supernatant (700 μl) was transferred to a new microtube; mixed with the same volume of chloroform and centrifuged at 10,000×g for 5 min. The aqueous layer formed (700 μl) was then transferred to the fresh microtube; mixed with the same volume of isopropanol, 140 μl of sodium acetate (3 M) and again centrifuged (10,000×g for 5 min). The pellet obtained was then washed with 70% ice-cold ethanol and air-dried. The air-dried pellet was then resuspended in 50 μl deionized water.

To amplify 18S rRNA gene (~ 1.8–2 kb) of microalgal strain, the 18S forward primer 5′-GTAGTCATAKGCTNGTCTS-3′ and 18S reverse primer 5′-GARACCTDGTTAVGACTY-3′ (18S rRNA specific primers) were used to carry out the polymerase chain reaction (PCR). The reaction mixture (100 μl) for carrying out PCR reactions consisted of 1 μl DNA template (100 ng/μl), 4 μl of dNTPs mixture (10 mM), 400 ng of each primer, 10 μl of 10X Taq DNA Polymerase assay buffer, Taq DNA Polymerase (1 μl of 3U/μl) and rest nuclease free water. The Thermal Cylcer ABI2720 was used to perform the PCR amplification under the conditions as follows: initial denaturation at 94 °C for 5 min and followed by 35 cycles of (denaturation 94 °C for 30 s, primer annealing at 55 °C for 30 s, and chain extension for 2 min at 72 °C), followed by a final extension at 72 °C for 7 min as mentioned by Dar et al. (2019). All PCR reagents were of Chromous Biotech Pvt. Ltd. (Bangalore, India) make. A 500 bp DNA ladder was used to evaluate the amplified PCR product. For sequencing, the PCR product was sent to Chromous Biotech Pvt. Ltd. (Bangalore) where ABI 3500XL Genetic Analyzer was further used to sequence the amplified PCR product. Seq Scape_v 5.2 software was used to analyse the 18S rRNA sequencing data.

Phylogenetic analysis

Homologous sequence search and comparative analyses were performed by the BLAST program at the NCBI server. Aligned sequences were checked manually and edited in BioEdit sequence alignment editor (version 7.2.6.1). Sequences that did not belong to unicellular green microalgal species were removed. The Neighbor Joining method (Saitou and Nei 1987) and bootstrap resampling (100 rounds) were used for phylogenetic analysis and dendrogram construction (Bruno et al. 2000). The dendrogram was constructed through MEGA X (Kumar et al. 2018). The evolutionary distances were computed using the Maximum Composite Likelihood method (Tamura et al. 2011). The 18S rDNA sequence was submitted to the GenBank database.

Optimization of different cultural conditions

Placket–Burman design for factor screening and selection

The significance of different physicochemical variables (light intensity, pH, growth period, temperature, NH₄Cl, NaNO₃, K₂HPO₄) on functional ingredients mainly biomass, carbohydrate, chlorophyll, lipid and protein was evaluated by the Placket–Burman design. It tested every factor at two levels: − 1 for a low level and + 1 for a high level. The experimental design matrix is given in Table 1. The Statgraphics Centurion XVII was used to analyze the data.

Optimization of cultural factors by central composite design (CCD)

To identify the optimum levels of factors significantly affecting various responses as mentioned in the above section, response surface methodology (RSM), based on CCD, was used. The optimization experiment was designed using the Statgraphics Centurian XVI.I software.

The seven significant factors selected by the Placket-Burman approach, used in designing the experiment included light intensity, pH, temperature, growth period, NH₄Cl, NaNO₃ and K₂HPO₄. Each variable was evaluated at three levels (Table 4). A total of 39 experimental runs were performed. The full experimental design with different combinations, under which experiments were carried, is given in Table 5. Biomass, chlorophyll, carbohydrate, lipid and protein production were studied under varied pH (7.5, 9.5, 11.5), temperature (20 °C, 27.5 °C, 35 °C), light intensity (40.5, 60.75, 81.0 µmol m⁻² s⁻¹), growth period (10, 18, 25 days), NH₄Cl (15, 25, 35 mM), NaNO₃ (5, 12, 20 mM) and K₂HPO₄ (2, 5, 8 mM) to find the actual potential of the organism under study.

Table 4.

Different factors used in the CCD study with their levels

Name	Units	Low	High	Levels
A:pH		7.5	11.5	− 1, 0, + 1
B:Temperature	°C	20.0	35.0	− 1, 0, + 1
C:Light intensity	µmol m−2 s−1	40.5	81.0	− 1, 0, + 1
D:Growth period	Days	10.0	25.0	− 1, 0, + 1
E:NH4Cl	mM	15.0	35.0	− 1, 0, + 1
F:NaNO3	mM	5.0	20.0	− 1, 0, + 1
G:K2HPO4	mM	2.0	8.0	− 1, 0, + 1

Table 5.

Central composite design (CCD) and response values for different responses

Run	pH	Temp (°C)	Light intensity (µmol m−2 s−1)	Growth period (days)	NH4Cl (mM)	NaNO3 (mM)	K2HPO4 (mM)	Biomass (g L−1)	Chlorophyll (mg L−1)	Carbohydrates (g L−1)	Lipid (g L−1)	Protein (g L−1)	Observed desirability	Predicted desirability
1	7.5	20	81	10	35	20	2	0.538	10.659	0.282	0.026	0.212	0.325	0.311
2	11.5	20	40.5	10	35	20	2	0.530	10.962	0.286	0.028	0.213	0.304	0.310
3	7.5	20	40.5	10	15	20	2	0.612	14.299	0.290	0.057	0.170	0.362	0.369
4	9.5	27.5	81	18	25	12.5	5	1.078	27.605	0.390	0.130	0.405	0.747	0.745
5	7.5	35	40.5	10	35	20	2	0.519	12.968	0.198	0.040	0.182	0.301	0.312
6	7.5	35	40.5	25	35	20	8	0.488	7.257	0.132	0.018	0.202	0.217	0.205
7	9.5	20	60.75	18	25	12.5	5	0.955	21.841	0.323	0.136	0.379	0.656	0.655
8	11.5	20	81	25	15	20	8	0.598	12.885	0.155	0.038	0.188	0.377	0.375
9	11.5	35	40.5	10	15	20	2	0.512	14.684	0.231	0.037	0.135	0.345	0.338
10	11.5	35	81	10	35	20	2	0.536	9.745	0.273	0.027	0.166	0.309	0.314
11	11.5	35	40.5	10	35	5	2	0.635	13.983	0.251	0.102	0.187	0.447	0.434
12	7.5	35	40.5	10	15	20	8	0.544	10.847	0.132	0.045	0.172	0.267	0.270
13	7.5	35	81	25	35	20	2	0.521	9.195	0.255	0.027	0.201	0.311	0.305
14	7.5	35	40.5	10	15	5	2	0.562	27.674	0.208	0.093	0.153	0.388	0.376
15	9.5	27.5	60.75	18	35	12.5	5	0.889	21.932	0.334	0.129	0.296	0.662	0.661
16	7.5	35	81	10	15	5	8	0.876	22.196	0.183	0.102	0.245	0.471	0.457
17	7.5	20	40.5	25	35	20	2	0.540	10.097	0.290	0.026	0.201	0.326	0.319
18	11.5	35	81	25	15	5	8	0.821	21.773	0.185	0.108	0.253	0.488	0.500
19	11.5	20	81	25	35	5	8	0.837	20.109	0.201	0.105	0.317	0.567	0.553
20	7.5	20	81	25	35	20	8	0.662	12.876	0.231	0.030	0.236	0.364	0.391
21	11.5	35	40.5	25	35	20	2	0.463	8.346	0.148	0.039	0.168	0.276	0.280
22	11.5	27.5	60.75	18	25	12.5	5	0.803	18.614	0.280	0.119	0.311	0.602	0.601
23	7.5	35	40.5	25	35	5	2	0.477	9.390	0.200	0.075	0.158	0.292	0.298
24	11.5	20	81	25	15	5	2	0.657	34.096	0.249	0.093	0.195	0.603	0.602
25	7.5	35	40.5	10	35	5	8	0.770	17.278	0.191	0.107	0.278	0.462	0.475
26	7.5	35	81	10	35	5	2	0.380	6.722	0.161	0.071	0.111	0.221	0.227
27	9.5	27.5	60.75	10	25	12.5	5	0.908	22.977	0.320	0.120	0.362	0.649	0.653
28	11.5	35	40.5	10	35	20	8	0.519	8.583	0.171	0.027	0.196	0.255	0.258
29	11.5	20	81	10	35	5	2	0.520	13.975	0.227	0.071	0.175	0.373	0.378
30	11.5	20	40.5	25	15	5	8	1.098	24.956	0.271	0.062	0.412	0.590	0.600
31	7.5	35	81	10	35	20	8	0.418	6.888	0.129	0.026	0.104	0.176	0.170
32	7.5	35	81	10	15	20	2	0.379	9.587	0.173	0.035	0.107	0.213	0.218
33	9.5	27.5	60.75	18	25	20	5	0.548	11.924	0.206	0.027	0.190	0.289	0.289
34	7.5	20	81	25	15	5	8	1.260	28.983	0.307	0.090	0.396	0.782	0.775
35	7.5	20	40.5	10	35	20	8	0.470	7.982	0.113	0.018	0.180	0.187	0.177
36	11.5	20	81	10	15	5	8	1.120	27.995	0.241	0.083	0.345	0.614	0.623
37	7.5	35	40.5	25	15	20	2	0.366	8.340	0.152	0.033	0.107	0.197	0.200
38	7.5	20	40.5	10	35	5	2	0.455	10.231	0.186	0.061	0.128	0.275	0.276
39	9.5	27.5	60.75	18	25	12.5	8	0.967	24.049	0.260	0.117	0.375	0.655	0.654

Further, a validation experiment was carried out to validate and verify the model using the combination of different optimized variables, which resulted in the maximum production of the response variable.

Analytical procedures

The microalgal biomass weight was determined as per the method described by Ratha et al. (2012) by centrifuging (5000×g) the known volume of homogenous microalgal suspension for ten minutes to obtain the pellet, which was then washed with distilled H₂O. The washed pellet was dried at 60 °C till constant weight was achieved. The modified method of El-baky et al. (2008) was used for the estimation of chlorophyll. According to this method, 20 mL of microalgal culture was centrifuged at 6000×g for 10 min and the pellet obtained was dissolved in acetone (20 mL, 80%) and kept overnight in dark at 4 °C for complete extraction. After this, it was centrifuged at 10,000×g for 5 min and the contents of total chlorophyll (T-Chl) in the supernatant were spectrophotometrically (Hitachi UV–Vis Spectrophotometer U-2800) determined (Lichtenthaler 1987).

The carbohydrates were analyzed according to DuBois et al. (1956) by taking 2 mL of homogenized microalgal suspension, then diluted with 1 mL of distilled water, mixed with 5 mL of 5% phenol and vortexed. Further, concentrated H₂SO₄ (5 mL) was added to the test tube, swirled, cooled and finally absorbance was measured (490 nm).

For protein estimation according to Lowry et al. (1951) using bovine serum albumen standards, 1 N NaOH was mixed with 2 mL of homogenized microalgal sample, heated (100 °C) for 15 min and then cooled. Then, 5 mL of alkaline copper sulphate reagent was added and the contents were thoroughly mixed. Finally, after allowing the contents for 30 min to develop colour, absorbance was measured at 620 nm.

The lipid quantification was done as per the sulfo-phospho-vanillin assay (Mishra et al. 2014), wherein, a known amount of microalgal biomass in 100 μl water obtained from either known volume of liquid microalgal culture or harvested by centrifuging the known volume at 4000 rpm for 5 min, was used. After this, the addition of concentrated (98%) H₂SO₄ (2 mL), heating of contents at 100 °C for 10 min and cooling in an ice bath for 5 min was done. Thereafter, 5 mL of freshly prepared phospho-vanillin reagent was added to the sample and incubated at 37 °C for 15 min in a shaking incubator shaker (200 rpm). Finally, to quantify the lipid within the sample, absorbance at 530 nm was recorded.

To ascertain the mineral composition of the dried biomass, the modified acid digestion method (Scholz 2004) was followed. The biomass was taken in a crucible, weighed and mixed with the concentrated HNO₃ and heated in a muffle furnace for 4 h at 400 °C. After allowing the sample to cool down, the ash formed was mixed with HNO₃ Suprapur^® (Sigma-Aldrich) and diluted with Milli-Q water. Further, filtration was carried using a membrane filter (0.45 μm) (Mitra et al. 2015) and then the mineral composition analysis was done by inductively coupled plasma-atomic emission spectroscopy (ICP-AES).

Co-digestion of paddy straw and microalgal biomass for biogas production

The co-digestion of Asterarcys quadricellulare BGLR5 biomass (microalgal biomass) with paddy straw (PS) in various combinations was performed to assess the influence of microalgal biomass on biogas production from the high C/N ratio substrate i.e., paddy straw. The PS content was decreased in successive digesters and the same amount of algal biomass was added. The PS was obtained from the experimental fields of Punjab Agricultural University, Ludhiana, India (PAU) while the microalgal biomass was cultivated in our laboratory. The cattle dung slurry (CDS) used as inoculum in this study was acquired from the biogas plant of PAU, India. It was degasified before its utilization as inoculum by incubating at 45–50 °C for a week. The PS used for the co-digestion study was soaked for 12 h before incorporating it into the digesters. The paddy straw and algae were co-digested according to Dar et al. (2019) at 35–38 °C in A–F batch digesters (2L), each in triplicates, with 1:1 w/v as substrate: inoculum ratio.

The standard methods of AOAC (2000) were used to determine the volatile solids (VS) of the samples both at the start and end of the anaerobic digestion period (46 days). For the determination of VS, ash (%) is to be calculated for the samples as VS% = 100 − ash%. The ash content was determined by igniting the oven dried sample (1 g) in a muffle furnace at 600 °C for 3 h (or until the sample became carbon-free). Thereafter, the crucible was placed in an oven at 100 °C for one hour and then weighed after cooled in a desiccator. The ash content was calculated using the formula: ash (%) = [(weight of furnace burnt sample and crucible − weight of oven dried empty silica crucible)/(weight of oven dried sample and crucible − weight of oven dried empty silica crucible)] × 100. The volatile solids content before and after digestion was used to calculate the volatile solids reduction (VSR %) as follows:

$$\mathrm{VSR}(\%)=\frac{\left({\mathrm{VS}}_{\mathrm{bd}},-,{\mathrm{VS}}_{\mathrm{ad}}\right)}{{\mathrm{VS}}_{\mathrm{bd}}}\times 100$$ 1

where VS_bd and VS_ad represent volatile solids before and after AD.

The amount of acidic water (pH < 3) replaced by the gas produced is used to determine the daily biogas production.

Kinetic study of biogas production

The kinetics of biogas production was studied using the modified Gompertz equation Eq. (2) adopted from Prajapati et al. (2015). The biogas produced from the various experimental batch digesters was recorded on the daily basis. The daily biogas production was used to determine the cumulative biogas production (mL biogas g⁻¹ VS) which was then fitted with the Gompertz equation for computing the increase in ultimate biogas yield, the maximum rate of biogas production and lag phase in the gas production profile.

$$M=P\times \exp \left[-,\exp ,\left\{\left(\frac{R_{\mathrm{m}}\times e}{P}\right),\times ,\left(\lambda ,-,t\right),+,1\right\}\right]$$ 2

where M represents the cumulative biogas yield (mL biogas g⁻¹ VS) at time t, t = time in days (d), P is the ultimate/maximum biogas potential (mL biogas g⁻¹ VS) at the end of incubation time, R_m signifies the maximum rate of biogas production (mL biogas g⁻¹VS d⁻¹), λ represents the lag phase (d) and e = 2.718. Further, the synergistic impact of anaerobic co-digestion of PS and microalgal biomass was assessed according to Zhang et al. (2014). The obtained biogas yield was from experimental runs, whereas the calculated/estimated biogas production in each experimental set up was computed using the Eq. (3).

$$Y_{\left(\mathrm{estimated}\right)}=Y_{\mathrm{MA}}\times X_1+Y_{\mathrm{PS}}\times X_2$$ 3

where Y_(estimated) is the estimated/calculated yield of biogas (mL g⁻¹ VS) of the mixture, Y_MA is the measured biogas yield of algal biomass, X₁ is the percentage or proportion of algae as co-substrate (%), Y_PS is the measured biogas yield of paddy straw and X₂ is the percentage of paddy straw as co-substrate (%).

Statistical analyses

To study the influence of seven physicochemical factors on the various response variables, the analysis of variance (ANOVA) was carried using the software Statgraphics Centurian XVI.I. The CCD was used to compute the multiple regression analysis of the experiment and the F test was applied to analyse the statistical significance of the quadratic equation. The coefficient of determination (R²) and the adjusted determination coefficient (R²_Adj.) were calculated between predicted and experimental data to analyse the accuracy and variability explained by the regression equation. The quadratic regression equation was used to elucidate the relationship between the response (biomass) and the independent factors and to develop an empirical model correlating the response to the variables, according to the following equation (Cui et al. 2016):

$$y={\beta }_0+\sum {\beta }_i{x}_i+\sum {\beta }_{\mathit{ii}}{x}_i^2+\sum {\beta }_{\mathit{ij}}{x}_i{x}_j$$ 4

where y represents the predicted response, ${\beta }_0$ is the intercept coefficient, β_i is the linear coefficient, β_ii is the quadratic coefficient, β_ij denotes the interaction coefficient and X_i and X_j are the variables.

The significant mean differences of biogas parameters for different experimental sets (A–F) were compared by Fisher’s least significant difference (LSD) multiple comparison procedure (p ≤ 0.05) using SAS software, version 9.4 (SAS Institute, Inc., Cary, NC, USA). Also, root mean square error (RMSE) and coefficient of variation of root mean squared error (CV(RMSE)) were used to evaluate the precision of the model applied.

Results and discussion

Isolation and identification of the microalgal strain

Microalgae are found not only in aquatic but also in terrestrial ecosystems. These represent a huge variety of species dwelling in a broad range of environmental regimes (Sydney et al. 2011). It is estimated that greater than 50,000 species of microalgae exist, out of which 30,000 have been studied (Richmond 2003). The physicochemical characteristics (units in mEq L⁻¹) of the water sample procured and analysed in this study were: HCO₃⁻ (12), CO₃²⁻ (0), Cl⁻ (20), Ca²⁺ + Mg²⁺ (7), and residual sodium concentration (5). The electrical conductivity was found to be 1.59 Scm⁻¹. The pH of the water sample was 7.27. This shows that the water sample taken from this area is saline in nature. Singh (2013) observed similar characteristics for the water procured from Muktsar, the waterlogged part of Indian Punjab. Through repeated standard microbiological methods of isolation (plating and streaking), a few strains of microalgae were isolated in our study. Based on the purity, growth and biomass production in the BG-11 medium, BGLR5 was chosen for further studies. The microscopic studies of its morphological features revealed its spherical shape and medium size with solitary cells (10 ± 0.54 μm) or temporary groups having distinct chloroplast similar to Chlorella (Fig. 1). Using this information as taxonomic keys, it falls under the genus Chlorella, family Chlorellaceae and division Chlorophyta. The morphological heterogeneity of the microalgae makes the microscopic examination very difficult and, therefore, to circumvent any ambiguity emerging out of this, 18S rRNA gene sequence analysis of this isolate was carried to complement morphological observation.

Fig.1.

Microscopic picture (40X) of Asterarcys quadricellulare BGLR5

PCR amplification, 18S rRNA sequencing and phylogenetic analysis

The rRNA genes, present in high copy numbers are considered good candidates for microbial identification and their detection sensitivity can be enhanced by PCR. So, the tentative identification on the morphological basis of the isolate BGLR5 was further established by 18S rRNA sequencing. The PCR amplification, gene sequencing and phylogenetic analyses of the ~ 2 kb 18S-rDNA fragment confirmed the morphological identification of BGLR5. The BGLR5 18S rRNA sequencing and the evolutionary tree construction (Fig. 2) revealed 99% similarity with the Asterarcys quadricellulare, Scenedesmus sp., Grasiella emersonii, Chlorella emersonii and Coelestrella sp., sequences available in the NCBI database. Asterarcys quadricellulare, Grasiella emersonii and Chlorella emersonii are morphologically very much similar. The examination of the phylogram obtained revealed Asterarcys quadricellulare as the most similar and close to the BGLR5 isolate. Thus, the BGLR5 was identified as Asterarcys quadricellulare BGLR5 and its 18S rDNA sequence was submitted with the accession number MF661929 to GenBank database. The BGLR5 genomic DNA and PCR amplification (18S rDNA) gel images are given in supplementary figure (S1).

Fig.2.

Phylogenetic analysis of the 18S rDNA sequences (approximately 1637 bp) of BGLR5. The optimal tree is shown here. The evolutionary history was inferred using the neighbor-joining method (Saitou and Nei 1987). The evolutionary distances were computed using the maximum composite likelihood method (Tamura et al. 2011) and are in the units of the number of base substitutions per site. Evolutionary analyses were conducted in MEGA X (Kumar et al. 2018)

Significant factor determination by Placket–Burman design

To study the influence of each selected factor (pH, light intensity, temperature, growth period, NH₄Cl, NaNO₃ and K₂HPO₄) on the production of biomass, chlorophyll, carbohydrates, lipids and proteins, 12 experimental runs designed by Placket–Burman design with each run screening independent variables, were conducted. The experimental design to determine the significant variables is shown in Table 1. Each run was conducted in triplicate.

The analysis of variance and multiple regression model determined the relationship between independent and dependent variables. The statistical analysis of the response variables is given in Table 2. The variables having p values < 0.05 (confidence levels of > 95%) were considered to significantly affect the response variables. The linear main effect of all the physicochemical variables (temperature, pH, light intensity, growth period, NH₄Cl, NaNO₃ and K₂HPO_4,), significantly affected the responses under study. The p values given in Table 2 determine the significance of each coefficient. The smaller the p value and larger the F ratio, the more significant the variable (p < 0.05) (Montgomery 2005). All the factors were found to have a low p value (< 0.05) (Table 2); hence are statistically significant and imperative for further study. Temperature, with a probability value (p) of < 0.0001, F value of 1753.30 and coefficient of estimate of − 0.3455, was decided to be the highly significant factor for biomass production, followed by K₂HPO₄ (p value < 0.0001, F value 584.58, coefficient of estimate of 0.1995), light intensity (p value < 0.0001, F value 394.24, coefficient of estimate of 0.1638), NaNO₃ (p value 0.0001, F value 236.28, coefficient of estimate of − 0.1268), growth period (p value 0.0188, F value 14.57, coefficient of estimate of 0.0315) and pH (p value 0.0233, F value 12.78, coefficient of estimate of 0.0295) and NH₄Cl (p value 0.0572, F value 7.00, coefficient of estimate of − 0.0218). Similarly, for other responses, the values are given in Table 2. The model p value for all responses was found to be < 0.05 (< 0.0001, 0.0001, 0.0002, < 0.0001 and 0.0003 for biomass, chlorophyll, carbohydrates, lipid and protein, respectively) (Table 3) implying that the model and variables were significant. To establish the optimal range of the significant variables selected, we carried out optimization through the CCD.

Table 2.

Statistical analysis of Placket–Burman design

Source	Levels			Sum of squares	DFa	Mean square	Coefficient of estimate	F ratio	bp value
	− 1 level	+ 1 level
Biomass
A:pH	7.5	10.5		0.0026	1	0.0026	0.0295	12.78	0.0233
B:Temperature	25	35		0.3581	1	0.3581	− 0.3455	1753.30	0.0000
C:Light intensity	40.5	81		0.0805	1	0.0805	0.1638	394.24	0.0000
D:Growth period	15	25		0.0030	1	0.0030	0.0315	14.57	0.0188
E:NH4Cl	20	35		0.0014	1	0.0014	− 0.0218	7.00	0.0572
F:NaNO3	5	20		0.0483	1	0.0483	− 0.1268	236.28	0.0001
G:K2HPO4	2	8		0.1194	1	0.1194	0.1995	584.58	0.0000
Total error				0.0008	4	0.0002
Total (corr.)				0.6141	11
Chlorophyll
A:pH	7.5	10.5		5.2420	1	5.2420	1.3219	8.15	0.0462
B:Temperature	25	35		483.771	1	483.771	− 12.6987	751.87	0.0000
C:Light intensity	40.5	81		77.283	1	77.283	5.0755	120.11	0.0004
D:Growth period	15	25		6.2971	1	6.2971	− 1.4488	9.79	0.0352
E:NH4Cl	20	35		43.4030	1	43.4030	− 3.8036	67.46	0.0012
F:NaNO3	5	20		224.512	1	224.512	− 8.6509	348.93	0.0000
G:K2HPO4	2	8		28.4279	1	28.4279	3.0783	44.18	0.0027
Total error				2.5737	4	0.6434
Total (corr.)				871.509	11
Carbohydrate
A:pH	7.5	10.5		0.0017	1	0.0017	0.0240	13.24	0.0220
B:Temperature	25	35		0.0654	1	0.0654	− 0.1477	501.27	0.0000
C:Light intensity	40.5	81		0.0124	1	0.0124	0.0643	95.14	0.0006
D:Growth period	15	25		0.0011	1	0.0011	0.0190	8.30	0.0450
E:NH4Cl	20	35		0.0009	1	0.0009	− 0.0177	7.17	0.0553
F:NaNO3	5	20		0.0011	1	0.0011	-0.0190	8.30	0.0450
G:K2HPO4	2	8		0.0202	1	0.0202	0.0820	154.57	0.0002
Total error				0.0005	4	0.0001
Total (corr.)				0.1034	11
Lipid
A:pH	7.5	10.5		0.000009	1	0.000009	− 0.0017	8.69	0.0421
B:Temperature	25	35		0.0013	1	0.0013	− 0.0211	1297.73	0.0000
C:Light intensity	40.5	81		0.0003	1	0.0003	0.0100	293.34	0.0001
D:Growth period	15	25		0.00002	1	0.00002	− 0.0024	16.87	0.0148
E:NH4Cl	20	35		0.00001	1	0.00001	− 0.0019	10.65	0.0310
F:NaNO3	5	20		0.0019	1	0.0019	− 0.0249	1809.99	0.0000
G:K2HPO4	2	8		0.0006	1	0.0006	0.0137	549.40	0.0000
Total error				0.000004	4		0.000001
Total (corr.)				0.0041	11
Protein
A:pH	7.5	10.5		0.0012	1	0.0012	0.0204	18.54	0.0126
B:Temperature	25	35		0.0357	1	0.0357	− 0.1091	532.76	0.0000
C:Light intensity	40.5	81		0.0037	1	0.0037	0.0351	55.09	0.0018
D:Growth period	15	25		0.0023	1	0.0023	0.0275	33.89	0.0043
E:NH4Cl	20	35		0.0009	1	0.0009	− 0.0174	13.52	0.0213
F:NaNO3	5	20		0.0005	1	0.0005	0.0133	7.92	0.0481
G:K2HPO4	2	8		0.0005	1	0.0005	− 0.0133	7.95	0.0479
Total error				0.0003	4	0.00007
Total (corr.)				0.0451	11

^aDF is Degrees of freedom

^bSignificance level at a p value ≤ 0.05

Table 3.

Analysis of the experimental results of Placket–Burman design

Model	Biomass	Chlorophyll	Carbohydrate	Lipid	Protein
Model DF	7	7	7	7	7
p value	0.0000	0.0001	0.0002	0.0000	0.0003
Error DF	4	4	4	4	4
Stndard error of estimate	0.0142916	0.802138	0.0114237	0.00101271	0.00818452
Mean absolute error	0.0074167	0.353815	0.006	0.00048260	0.00455604
R-squared (%)	99.87	99.70	99.49	99.90	99.41
Adj. R-squared (%)	99.63	99.19	98.61	99.72	98.37

DF degrees of freedom

The coefficient of determination (R²) checked the model’s goodness of fit and was found to be close to 1 for various responses (Table 3). For biomass, R² is 0.9987, signifying that 99.87% of the total variation in the dependent variable was demonstrated by the independent variables. Similarly, for others, the coefficient of determination (R²) was found to be 99.70, 99.49, 99.90 and 99.41 for chlorophyll, carbohydrate, lipid and protein, respectively. Also, the significance and accuracy of the model used in our study were demonstrated by higher $R_{\text{Adj}}^2$_. values for biomass (0.9963) and other response variables (chlorophyll 0.9919, carbohydrate 0.9861, lipid 0.9972 and protein 0.9837).

Cultural factor optimization of Asterarcys quadricellulare BGLR5 by CCD

The highly significant independent variables obtained from Plackett–Burman screening affecting biomass production and other responses were selected (Table 4) and further investigated using CCD to determine the interactions among the significant factors. In total, 39 experimental runs each with distinct combinations of the seven physicochemical factors were performed; the observed values for all response variables along with the experimental design and desirability (predicted and observed) are given in Table 5. The results exhibited considerable variation in biomass production and for other response variables. The maximum (1.260 g L⁻¹) and minimum (0.366 g L⁻¹) biomass production was achieved in run number 34 and 37, respectively. The highest observed and predicted desirability was obtained for run number 34 (Table 5).

Variance and regression analysis were employed to interpret the data. The model significance and adequacy tested through ANOVA are presented in Table 6. For each effect, the variation in every dependent variable is segregated into distinct sections through ANOVA and then compared the mean square with an estimate of the experimental error to analyze statistically the significance of each effect. The number of effects with p value < 0.05 in biomass were found to be 33, demonstrating that they are significantly (p < 0.05) different from zero (Table 6). The ANOVA tables for other responses have not been given here but it can be mentioned that the number of significant effects in chlorophyll, carbohydrate, lipid and protein was 26, 33, 31 and 33, respectively. The CCD model described well the response variables as illustrated by the ANOVA for the seven variables. The effect of independent physicochemical factors like pH (p < 0.001), temperature (p < 0.001), light intensity (p = 0.0001), growth period (p = 0.001), NaNO₃ (p < 0.0001), NH₄Cl (p < 0.0001) and K₂HPO₄ (p < 0.0001) reported in Table 6 were significant (p < 0.05), whereas, the significant interactions differed from one dependent variable to another. The linear terms in all the response variables were significant. For biomass, interactive (AB, AC, AD, AF, AG, BC, BD, BE, BG, CD, CE, CF, DE, DF, DG, EF, EG and FG) and quadratic (A², B², C², D², E² and F²) were significant (p < 0.05) model terms (Table 6). The AB, AC, AF, AG, BC, BD, BE, CD, CF, DE, EF, EG, FG interactive terms and A², B², C², E² and F² quadratic terms were significant (p < 0.05) model terms in chlorophyll. In lipid, the interactive (AB, AC, AD, AE, AF, AG, BC, BD, BE, BF, BG, CD, CG, DE, DG, EF, EG and FG) and quadratic terms (A², B², D², E², F² and G²) proved to have significant effects (p < 0.05) while the significant model terms for carbohydrate were the interactive (AB, AC, AD, AE, AF, BC, BD, BE, BF, BG, CD, CE, CF, CG, DE, DF, DG, EF, EG and FG) and quadratic (A², B², C², E², F² and G²) terms. Correspondingly, the significant (p < 0.05) model terms for protein were linear (AB, AC, AD, AE, BC, BD, BE, BF, BG, CD, CE, CF, CG, DE, DF, DG, EF, EG, and FG) and quadratic (A², C², D², E², F² and G²). The coefficient of determination (R²) explains the deviation or variability exhibited by the experimental variables and their interactions in the observed response values. The R² values always lie between 0 and 1; the more the R² value is closer to 1, the stronger is the model and the better it will predict the response (Kaushik et al. 2006). The R² of the model for biomass, chlorophyll, carbohydrates, lipids and proteins were 0.9999, 0.9983, 0.9999, 0.9997 and 1.0000, respectively (Table 7). This can be interpreted as, that 99.99%, 99.83%, 99.99%, 99.97% and 100% of the variability in the responses (biomass, chlorophyll, carbohydrates, lipids and proteins), respectively, could be explained by the statistical model used. The R² value of the regression model greater than 0.9 is regarded to be having a very strong correlation (Pham 2019). The R² values found in our case are > 0.9, thus reflect a very good fit between the observed and predicted responses, suggesting that the model is authentic and valid for the production of the various responses under study. Furthermore, the predicted and observed values of various responses in our study were very much close as can be seen in Fig. 3a–e. The values almost fall on the line of fit, therefore, depicting the significance and accuracy of the model. Besides, the $R_{\text{Adj}}^2$_. was also very high (99.96, 99.47, 99.90, 99.84 and 99.99% for biomass, chlorophyll, carbohydrates, lipid and protein, respectively) in our case (Table 7), thus determining the high significance of the model (Pham 2019). Consequently, we regard the analysis of the response trend to be acceptable and satisfactory using the model.

Table 6.

Analysis of variance (ANOVA) and regression coefficients for the quadratic model of various physicochemical factors of Asterarcys quadricellulare BGLR5 for biomass production

Source	Sum of squares	Df	Mean square	F ratio	p value	Regression coeffs
A:pH	0.0049	1	0.0049	224.84	< 0.0001	0.5309
B:Temperature	0.0888	1	0.0888	4086.59	< 0.0001	0.0176
C:Light intensity	0.0027	1	0.0027	123.19	0.0001	− 0.0521
D:Growth period	0.0023	1	0.0023	106.35	0.0001	0.0329
E:NH4Cl	0.0558	1	0.0558	2567.74	< 0.0001	− 0.0398
F:NaNO3	0.2871	1	0.2871	13,215.38	< 0.0001	0.1165
G:K2PHO4	0.2121	1	0.2121	9766.23	< 0.0001	0.1944
AA	0.0126	1	0.0126	578.61	< 0.0001	− 0.0266
AB	0.0051	1	0.0051	235.87	< 0.0001	0.0010
AC	0.0015	1	0.0015	69.81	0.0004	− 0.0002
AD	0.0240	1	0.0240	1107.08	< 0.0001	− 0.0022
AE	0.0026	1	0.0026	121.99	0.0001	0.0005
AF	0.0057	1	0.0057	264.42	< 0.0001	− 0.0011
AG	0.0036	1	0.0036	164.32	0.0001	− 0.0021
BB	0.0014	1	0.0014	66.49	0.0005	− 0.0006
BC	0.0110	1	0.0110	508.66	< 0.0001	− 0.0001
BD	0.0060	1	0.0060	274.75	< 0.0001	− 0.0003
BE	0.0470	1	0.0470	2163.03	< 0.0001	0.0006
BG	0.0007	1	0.0007	32.84	0.0023	− 0.0003
CC	0.0364	1	0.0364	1674.14	< 0.0001	0.0004
CD	0.0068	1	0.0068	311.43	< 0.0001	0.0001
CE	0.0013	1	0.0013	58.07	0.0006	0.0000
CF	0.0026	1	0.0026	120.61	0.0001	0.0001
DD	0.0010	1	0.0010	43.92	0.0012	− 0.0005
DE	0.0065	1	0.0065	298.42	< 0.0001	0.0002
DF	0.0011	1	0.0011	50.6	0.0009	− 0.0001
DG	0.0020	1	0.0020	90.52	0.0002	0.0004
EE	0.0002	1	0.0002	7.52	0.0407	0.0001
EF	0.0309	1	0.0309	1422.22	< 0.0001	0.0005
EG	0.0252	1	0.0252	1158.29	< 0.0001	− 0.0011
FF	0.0771	1	0.0771	3550.26	< 0.0001	− 0.0047
FG	0.1664	1	0.1664	7661.43	< 0.0001	− 0.0039
GG	0.0038	1	0.0038	176.69	< 0.0001	− 0.0065
Total error	0.0001	5	0.0000
Total (corr.)	2.0734	38
Constant						− 1.0033

Table 7.

Analysis of the experimental results of central composite design

Model	Biomass	Chlorophyll	Carbohydrates	Lipid	Protein
Transformation	None	None	None	None	None
Model df	33	26	33	31	33
p value	0.0000	0.0000	0.0000	0.0000	0.0000
Error df	5	12	5	7	5
Stnd. error	0.0047	0.5469	0.0021	0.0015	0.0007
R-squared (%)	99.99	99.83	99.99	99.97	100.00
Adj. R-squared (%)	99.96	99.47	99.90	99.84	99.99

Fig. 3.

Observed vs predicted results of different responses: a biomass, b chlorophyll, c carbohydrate, d lipid and e protein

The final regression equation (Eq. 5) including physicochemical variables under study significantly (p < 0.05) predicted biomass concentration/production in Asterarcys quadricellulare BGLR5, and all the variable terms were significant (p < 0.05).

$$\begin{array}{cc}\hfill \text{Biomass}=& -1.0033+0.5310\ast A+0.0176\ast B-0.0521\ast C+0.0329\ast D\hfill \\ \hfill & -0.0398\ast E+0.1165\ast F+0.1944\ast G-0.0266\ast A^2+0.0010\ast A\ast B\hfill \\ \hfill & -0.0002\ast A\ast C-0.0022\ast A\ast D+0.0006\ast A\ast E-0.0011\ast A\ast F\hfill \\ \hfill & -0.0021\ast A\ast G-0.0006\ast B^2-0.0002\ast B\ast C-0.0003\ast B\ast D\hfill \\ \hfill & +0.0006\ast B\ast E-0.0003\ast B\ast G+0.0004\ast C^2+0.0001\ast C\ast D\hfill \\ \hfill & +0.00004\ast C\ast E+0.0001\ast C\ast F-0.0005\ast D^2+0.0002\ast D\ast E\hfill \\ \hfill & -0.0001\ast D\ast F+0.0004\ast D\ast G+0.0001\ast E^2+0.0005\ast E\ast F\hfill \\ \hfill & -0.0011\ast E\ast G-0.0047\ast F^2-0.0039\ast F\ast G-0.0065\ast G^2\hfill \\ \hfill \end{array}$$ 5

where A, B, C, D, E, F and G represent pH, temperature, light intensity, growth period, NH₄Cl, NaNO₃ and K₂HPO₄, respectively.

Thus, as per the results achieved, pH of 9.9, 22 °C temperature, 81 μmol m⁻² s⁻¹ light intensity, growth period of 25 days, NH₄Cl of 15 mM, 12 mM NaNO₃ and 7 mM K₂HPO₄ were the most desired and optimal conditions for different responses investigated in this study according to the model (Table 8). The desirability index of 94.91% was attained at these values. The closest run to these values is run 34 of CCD, as determined from observed and predicted desirability. Compared to the biomass produced at the basal medium conditions (0.886 g L⁻¹), the per cent (%) increase of 42.21 and 42.44 was observed from the run 34 (1.260 g L⁻¹) and optimal model value (1.262 g L⁻¹), respectively.

Table 8.

Factor settings at optimum and response values at optimum conditions as per model

	Cultural factors
	pH	Temp. (°C)	Light intensity (µmol m−2 s−1)	Growth period (days)	NH4Cl (mM)	NaNO3 (mM)	K2HPO4 (mM)
Setting at optimum	9.92	22.00	81.00	25.00	15.00	12.00	7.00
Response	Optimized	Prediction	Lower 95.0% Limit	Upper 95.0% Limit	Desirability
Biomass (g L−1)	Yes	1.262	1.247	1.277	1.0
Chlorophyll (mg L−1)	Yes	34.096	32.732	35.460	1.0
Carbohydrates (g L−1)	Yes	0.390	0.384	0.396	1.0
Lipid (g L−1)	Yes	0.142	0.138	0.146	1.0
Protein (g L−1)	Yes	0.436	0.433	0.438	1.0
Optimized desirability	0.949137

Interactive effect of different physicochemical factors having a positive impact on biomass production

The three dimensional (3D) response surface plots (with contours at the base) constructed using the CCD model illustrate the interaction effects and optimum levels of the physicochemical variables. The response surface plots with contours for biomass were generated for the interactive effect of various physicochemical factors (pH, temperature, growth period, light intensity, NH₄Cl, NaNO₃, K₂HPO₄) having a positive regression coefficient (Fig. 4a–h). These figures based on Eq. 5 were generated by varying the two variables within the experimental range whereas keeping others fixed at the optimal level. The regression coefficients depicted the nature of the relation of physicochemical factors with the response variable. The positive regression coefficients (linear or interactive coefficients) in the regression equation represented that biomass values (dependent variable) are increased on increasing the independent variables while negative coefficients suggested that the dependent variable (biomass) tends to decrease on increasing the independent variables. The presence of a peak in all the response surface plots of interactions having a positive regression coefficient indicated that the experimental levels selected in this study contain the optimum point (Priyadharshini and Bakthavatsalam 2016).

Fig. 4.

Response surface 3D plots a–h for various interactions of cultural factors for biomass production

Statistical results obtained showed increased biomass production at the pH range 8.5–9.5, lower temperature (20–27 °C) and moderate NH₄Cl concentration (19–27 mM) (Fig. 4a, b). The biomass yield decreased at higher pH and temperature. The higher pH affects biomass production as it causes cell death by inhibiting photosynthesis (due to free ammonia formation) and accumulation of nitrate (Iasimone et al. 2018). Also, the higher ammonium concentration in the algal growth medium results in the elevated concentration of free ammonia; leading to the attainment of stationary and decline phases rapidly in the microalgae (Iasimone et al. 2018). Thus higher NH₄Cl concentrations are not suitable for efficient biomass production. The microalgal growth and metabolic activities are very sensitive to temperature (Kishore et al. 2017). The higher temperature declines metabolic processes like respiration, photosynthesis and thus growth, due to imbalance between ATP production and energy demand, and denaturation of proteins involved in photosynthesis (Ras et al. 2013). The effect of light intensity and growth period on biomass production (Fig. 4c) while keeping other variables constant at optimal level showed a slight increase on increasing the growth period while the increase in light intensity was associated with an abrupt increase in the biomass yield. Although, in our study, a decrease was noticed in the moderate range of light intensity (50–60 µmol m⁻² s⁻¹), after that it increased the biomass on increasing light intensity. Various studies have shown that the different microalgal species have different light requirements and show different patterns of biomass growth (He et al. 2015; Kishore et al. 2017; Nzayisenga et al. 2020), thus signifying it as a vital factor governing microalgal biomass. Since in our study, the upper limit of light intensity was 81 µmol m⁻² s⁻¹, therefore the response surface plot showed an increasing trend beyond 80 µmol m⁻² s⁻¹, demonstrating that the higher light intensity could be favourable for efficient biomass production. The increased light intensity could increase the microalgal biomass by increasing the growth rate through enhanced chlorophyll content and photosynthesis of microalgae (Kishore et al. 2017; Singh and Singh 2015). However, it has been reported that light intensity enhances the microalgae growth up to a certain point (depending on the microalgae species) and beyond this point, photo-inhibition occurs which limits the growth and similarly, low light intensity also limits the microalgae growth (Difusa et al. 2015; Lee et al. 2015; Metsoviti et al. 2020). The effect of interaction between light intensity and NH₄Cl (Fig. 4d) depicted that NH₄Cl was not found to have a prominent impact on biomass increase while the light intensity appeared to be a favourable factor. This ascertains that light intensity had a pronounced effect on biomass production compared to the growth period and NH₄Cl. The interaction between light intensity and NaNO₃ was significant (Table 6). It is illustrated in the response surface plot with contours at the bottom (Fig. 4e) that the high light intensity and moderate NaNO₃ had a positive impact on biomass yield when other variables were kept at the optimum level. The combined effect of the growth period and NH₄Cl was statistically significant (p < 0.0001) but NH4Cl concentration didn’t affect the biomass yield as shown in Fig. 4f whereas the growth period of 15–19 days depicted the maximum biomass yield. It can be seen from Fig. 4g that a significant (p = 0.0002) increase in biomass production was achieved by increasing the concentration of K₂HPO₄ from 2 to 8 mM, with maximum production of biomass achieved at 7–8 mM K₂HPO₄ concentration. However, with the increase in growth period, a slight decrease was noticed. Figure 4h depicted the significant (p = 0.0407) interaction between NH₄Cl and NaNO_3. The surface plot with contours at the base showed that higher biomass could be achieved at low levels (15–19 mM) of NH₄Cl and medium levels (11–17 mM) of NaNO_3. The contour lines on the bottom of surface plots were curvilinear in all the studied interactions, demonstrating that the interactions were quite significant for biomass production.

Validation of the model

Asterarcys quadricellulare BGLR5 was grown in BG-11 medium under most favourable conditions as per the model (pH 9.9, 22 °C temperature, 81 μmol m⁻² s⁻¹ of light intensity, growth period of 25 days, NH₄Cl of 15 mM, 12 mM NaNO₃ and 7 mM of K₂HPO₄). The value obtained for various dependent variables namely biomass, chlorophyll, carbohydrate, lipid and protein were found to be 1.229 ± 0.08 g L⁻¹, 31.019 ± 1.46 mg L⁻¹, 0.329 ± 0.03 g L⁻¹, 0.117 ± 0.01 g L⁻¹ and 0.409 ± 0.02 g L⁻¹, respectively. The response variable values obtained were close to that obtained from the best run of CCD and model predicted values at optimum setting (Table 8). It revealed the confirmation, validity and acceptability of the model for optimizing the various physicochemical variables considered in our study.

Anaerobic co-digestion study

Production of biogas from Asterarcys quadricellulare BGLR5 biomass (MA), PS and their various combinations was studied in digesters (A–F), each in triplicates for a period of 46 days. The biogas yield data were comprehensively assessed by applying the modified Gompertz equation (Eq. 2) to estimate various kinetic model parameters as given in Table 9.

Table 9.

Estimated kinetic constants using non-linear regression model and other characteristics of the co-digestion of Asterarcys quadricellulare BGLR5 biomass (MA) with paddy straw (PS)

Digester	Substrate	P (mL g−1 VS)	Rm (mL g−1 VS d −1)	Λ (d)	R2	RMSE	CV (RMSE) (%)	VSR (%)
A	100% PS	231.52e	5.45e	9.60a	0.910	5.48	6.81	32.48
B	80% PS + 20% MA	245.14d	7.82ab	3.77b	0.982	7.42	5.52	59.17
C	70% PS + 30% MA	349.50b	7.37bc	3.40c	0.992	4.25	2.91	65.72
D	50% PS + 50% MA	361.81a	8.19a	2.81d	0.997	4.19	2.57	68.08
E	30% PS + 70% MA	318.26c	6.62d	2.16e	0.998	2.43	1.96	64.04
F	100% MA	203.12f	7.17c	2.10e	0.996	6.92	5.54	51.45

P ultimate biogas yield, R_m maximum rate of biogas production, λ lag phase in days (d), R² coefficient of determination, RMSE root mean square error, VSR volatile solid reduction, CV(RMSE) coefficient of variation of root mean squared error, values in a column for a given parameter followed by different superscripted letter(s) differ significantly at p < 0.05

The digester D with 1:1 ratio of PS and MA showed the maximum biogas production potential (P) and rate (R_m) of 361.81 mL g⁻¹ VS and 8.19 mL g⁻¹ VS days⁻¹, respectively (Table 9). This is followed by Digester C with P, R_m of 349.50 mL g⁻¹ VS and 7.37 mL g⁻¹ VS days⁻¹, respectively. Digester F with only algal biomass recorded the lowest biogas production. The per cent increase in P of digester D and C to that of A and F each were found to be 56.28, 50.96 and 78.13, 72.07%, respectively. The cumulative biogas produced followed the sigmoidal trend (Fig. 5) as observed by several researchers (Dar et al. 2019; Dar and Phutela 2020; Srivastava et al. 2020). The per cent increase in R_m of digesters D and C compared to that of digester A (only soaked paddy straw) were 50.28 and 35.23%, respectively. With the increase in MA content up to 50%, it was observed that the biogas production rate (R_m) increased but beyond 50%, it declined. The lag phase (λ) also decreased continuously on increasing the MA. Moreover, MA and PS co-digestion improved the volatile solid reduction (VSR) percentage. The digestibility of feedstock can be described more comprehensively by VSR. The highest VSR was found in digester D. Digester A showed a VSR of 32.48%, whereas D and C exhibited 68.08 and 65.72%, respectively. It explained and established that co-digestion enhances substrate digestibility. The increase in P, R_m, VSR and reduction in λ shown by digester D followed by C and E could be due to the improved C:N ratio by co-digesting PS and MA. The digestion of carbon-rich substrate (paddy straw) with microalgal biomass (low C/N ratio) helps in balancing the C:N ratio and decreases the ammonia inhibitory effects. Besides this, the small size (in μm) and cellular composition of microalgae furnishes a desirable substrate to methanogens for fermentation (Srivastava et al. 2020). Various researchers reported that the co-digestion process improves the C/N ratio and has other advantages like higher biogas yield, nutrient balance and reduction of the inhibitory compound concentration and improved buffering capacity (Ganesh Saratale et al. 2018; Jankowska et al. 2017; Milledge et al. 2019; Solé-Bundó et al. 2019). Similarly, both the sub-optimum and more C:N ratio (resulting in toxicity due to ammonia) could be the reason for less biogas yield in digesters A, B and F.

Fig. 5.

Variation and fitting of the cumulative biogas data of Asterarcys quadricellulare BGLR5 co-digested with paddy straw in different ratios with the Gompertz model for the different digesters a–f with time. The vertical line bars represent standard error

The modified Gompertz kinetic model explained adequately and rationally the biogas production from the PS and algal biomass with a goodness of fit (R²) more than 0.90 and less RMSE (< 8) for all the digesters (Fig. 5 and Table 9). Also, the low coefficient of variation of root mean squared error (CV(RMSE)) (Table 9) for biogas production in all the experimental digesters signifies the more precise estimate of the parameters and indicates a good model fit with acceptable and reliable predictive capabilities.

The feedstock fed into the digesters A–F was evaluated by determining the volatile solids before and after anaerobic digestion. A smooth increase in total VSR was noticed on anaerobic digestion (Table 9). The higher per cent change in volatile solids, as well as biogas production upon co-digestion of MA with PS, can be attributed to the synergistic effect during anaerobic co-digestion (Mata-Alvarez et al. 2014). The synergistic impact was analysed by comparing the measured biogas yield and the calculated yield using Eq. (3). In all the co-digestion experiments, the measured biogas profiles were above the calculated/estimated profiles, thus proving the positive synergistic effect upon co-digestion (Fig. 6a–d). The synergistic impact increased the observed biogas yield compared to the calculated yield by 11–58% (Fig. 6e) depending upon the amount of algal biomass and paddy straw taken. Zhen et al. (2016) achieved an increase of 30–54% on co-digestion of microalgal biomass with food waste. It should be mentioned that the strongest synergy occurred at 50% PS + 50% MA and hereafter with the increase in algal biomass content, the synergistic effect of anaerobic co-digestion attenuated with corresponding value dropping to 46% in 30% PS + 70% MA and further to 11.5% in 20% PS + 80% MA. This demonstrated that improvement in biogas production was not just proportional to the amount of microalgal biomass in the mixture. The positive synergistic impact in all the combinations can be attributed to various factors like balanced nutrient composition, increased buffering capacity, and decreased effect of toxic compounds (like ammonia) and volatile fatty acid build-up on the digestion process. It could be also due to the disruption of the cell wall by the introduction of specific microorganisms. These all factors seem to have upgraded conditions for microbial growth and maximized the biodegradability of microalgae and subsequent biogas production potential. However, it needs further investigation to explore the role of these factors and to evaluate in detail the synergistic impact.

Fig. 6.

Obtained and calculated biogas production (a–d), and final biogas yield as well as the increase in obtained biogas yields over calculated ones (e) for codigestion under various paddy straw: microalgal biomass (PS:MA) ratios

The mineral content of the microalgal biomass (mg/100 g) was determined by ICP-OES: potassium (1000), calcium (1100), iron (130), magnesium (300), manganese (10), phosphorous (1000), sulphur (440), nickel (10) and zinc (10). The quantity of nutrients supplied determines the growth of microorganisms carrying out the anaerobic digestion process (Nsair et al. 2020; Vintiloiu et al. 2012). The biogas producing microbes depend substantially on the amount and presence of micronutrients for their growth and favourable performance like other biotechnological approaches (Nsair et al. 2020). It is noteworthy that the overall enhanced biogas production in digesters D, C, E and B to that of A, could also be ascribed to the existence of elements like potassium (K), sulphur (S), calcium (Ca) and iron (Fe) which speed up the biogas production metabolism (Bożym et al. 2015), and the absence of toxic elements viz., cadmium (Cd), copper (Cu), lead (Pb), arsenic (As), and chromium (Cr), which are having a toxic or inhibitory effect on the anaerobic digestion process (Mudhoo and Kumar 2013). Sulphur (S), an important element for the growth of methanogens, positively affects methanogenesis by lowering the redox potential of the medium (Hirano et al. 2013; Paulo et al. 2015). Similarly, P is vital for providing energy carriers like ATP and NADP to methanogens. Trace elements Fe and Ni are vital for the growth of microorganisms. The conversion of acetic acid (CH₃COOH) to methane (CH₄) by methanogens requires enzymes with Fe as a cofactor (Baek et al. 2019). Nickel (Ni) has been reported to enhance the biomethanation rate as it is a structural component of the microbial enzymes involved in biomethanation (Zhang et al. 2015). The impact of the inhibitory effect of the heavy metals depends on their concentration (Habagil et al. 2020). The optimal concentration of micronutrients has been reported to be favourable for biogas production whereas both higher and lower concentrations have been found to inhibit biogas and biomethane production due to volatile fatty acid accumulation (Guo et al. 2019). In our case, the concentration of the metals within the algal biomass was not high enough to inhibit the AD process (Nsair et al. 2020; Soares et al. 2012). However, in our study, the biogas production did not increase on increasing the microalgal biomass content beyond 50%. This could be probably due to unbalanced nutrient composition in the digester and thus the decreased synergistic impact as mentioned above. Thus, it appears that 50% PS + 50% MA provided more stable conditions compared to others. So, it was concluded that co-digestion enhanced the substrate digestibility as well as biogas production compared to their anaerobic digestion individually. However, to further improve the biogas production, pretreatment of the substrate is required as the overall digestibility depends on the complexity of the cell structure and its composition. Given the fact that the complex cellular structure of microalgae with tough cell wall and also the presence of high lignin and silica content in paddy straw hinder the biodigestibility of these substrates. This suppresses the anaerobic digestion and reduce the biogas yield (Ramos-Suárez et al. 2014; Srivastava et al. 2020). Further research is required to explore the effects of different pretreatment approaches and their combinations on the substrate digestibility and biogas production.

Conclusions

This study demonstrated the isolation, identification, optimization of physico-chemical parameters of Asterarcys quadricellulare BGLR5 and assessment of its biogas potential upon co-digestion with paddy straw having high C/N ratio. The conditions: pH of 9.9, temperature of 22 °C, 81 μmol m⁻² s⁻¹ light intensity, growth period of 25 days, 15 mM CaCl₂, NaNO₃ of 12 mM and 7 mM K₂HPO₄ were found to be optimal as per the model. The co-digestion of 50% PS + 50% MA (1:1 w/w) demonstrated the highest biogas production potential (361.81 mL g⁻¹VS). Co-digestion of Asterarcys quadricellulare BGLR5 biomass with paddy straw enhanced both the digestibility and biogas production through positive synergistic effect. 

Accession number

The molecular data accession number for 18S rRNA gene is MF661929.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

The first author RAD and third author UGP designed the experiments. RAD performed the experimental work, analyzed the data and wrote the manuscript. UGP and RKG thoroughly revised and approved the final manuscript.

Declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.