Introduction

Cyanobacteria or blue green algae (BGA) are a group of oxygenic photoautotrophic prokaryotes that were responsible for the first significant increase in atmospheric oxygen (Gould et al. 2008). These diazotrophic photosynthetic microorganisms are distributed in diverse ecological habitats ranging from hot to cold springs, marine to fresh water, pristine to degraded soils, oligotrophic to hypereutrophic environments (Bhatnagar and Bhatnagar 2005, Bagul et al. 2018). Cyanobacteria are a morphologically diverse group comprising both unicellular and colonial (including filamentous) forms. Cyanobacteria are not only abundant in the soil, but also play important roles in soil fertility and soil productivity (Rossi et al. 2017). Filamentous cyanobacteria, due to their ability to add nitrogen and organic carbon to soil and to improve soil physical properties, are undoubtedly one of the important groups of cyanobacteria. Due to their filamentous nature and abundant production of polysaccharides in many species, they can significantly help in soil aggregation (Chakdar et al. 2012). It is well known that filamentous cyanobacteria are associated with biological soil crusts (BSC) the origin of which is due to entrapment of minerals and sand particles by cyanobacterial filaments (Garcia-Pichel et al. 2001). These BSCs are ecologically very important in semi-arid to arid areas as they serve as microhabitats (Zhang 2005). Filamentous cyanobacteria have also been evaluated for their potential role in salinity stress amelioration (Li et al. 2019).

Despite their ecological significance and application for the amelioration of degraded soils, studies on the diversity of cyanobacteria from saline/alkaline/sodic soils are scanty in India. To our understanding, lack of detailed diversity studies from different habitats, routine sub-culturing based isolation and morphometry based techniques leading to the identification of a few predominant genera are some of the reasons that most of the application-based studies are centred around a few genera like Anabaena or Nostoc. In India, the area under salt-affected soils is about 6.73 million (m) ha with the states of Gujarat (2.23 m ha), Uttar Pradesh (1.37 m ha), Maharashtra (0.61 m ha), West Bengal (0.44 m ha) and Rajasthan (0.38 m ha) together accounting for almost 75% of saline and sodic soils in the country (Sharma and Singh 2015). Understanding of the genetic as well as the functional variability of the cyanobacterial population of saline sodic soils can help to establish a very broad base of genetic resources for their further exploitation. Diversity of cyanobacteria is generally studied in terms of the variabilities in cellular features (e.g. biometric characteristics of vegetative cells, heterocysts, hormogonia and akinetes; branching pattern, presence or absence of sheaths etc.) or by the use of molecular techniques. Undoubtedly, molecular analyses of diversity are more accurate and trustworthy, as morphological features are prone to environmental fluctuations (Doers and Parker 1988, Kato 1991, Chakdar and Pabbi 2012). A number of molecular tools PCR-RFLP, RAPD, finger printing based on repetitive elements have been used to study inter- and intra-generic molecular diversity of cyanobacteria (Mazel et al. 1990, Neilan et al. 2003, Prabina et al. 2005, Ezhilarasi and Anand 2010, Akoijam and Singh 2011, Chakdar and Pabbi 2012, Chakdar and Pabbi 2017).

Highly Iterated Palindrome (HIP1), a repetitive eight base sequence (5’-GCGATCGC-3’), was first identified in a cadmium-tolerant strain of Synechococcus PCC 6301 at the borders of a gene deletion (Gupta et al. 1993). It has been speculated that there are no comparable sequences to HIP1 in other organisms, that they are only known to occur in cyanobacteria and can also be used to fingerprint organisms (Smith et al. 1998). These repetitive sequences have been used to study several cyanobacterial taxa (Lyra et al. 2005, Prasanna et al. 2006, Selvakumar and Gopalaswamy 2008, Chakdar and Pabbi 2012, Singh et al. 2014, Shokraei et al. 2019). The conserved nature along with genome-wide distribution of the repeats and reproducibility has made them ideal tools for biodiversity studies (Selvakumar and Gopalaswamy 2008).

Looking at the insufficient information available about the diversity of filamentous cyanobacteria from saline sodic soils of India, in the present study we analyzed the diversity of 47 filamentous cyanobacteria isolated from saline/alkaline soils of eastern Uttar Pradesh, India. Here we have used HIP fingerprinting as well as fuzzy clustering based on cellular features and empirically compared the techniques to determine whether they can really be correlated.

Material and methods

Procurement and maintenance of cultures

Forty-seven strains of cyanobacteria isolated from saline/alkaline habitats of eastern part of Uttar Pradesh, India were procured from the National Agriculturally Important Microbial Culture Collection (NAIMCC), ICAR-National Bureau of Agriculturally Important Microorganisms (ICAR-NBAIM), India (On-line Suppl. Tab. 1). Cultures were maintained in chemically defined BG-11 media with and without nitrogen source for non-heterocystous and heterocystous cyanobacterial isolates respectively at 28 ± 2 °C under a light intensity of 52–55 μmol photon m^-2 s^-1 and L:D cycles of 16:8 h (Stanier et al. 1971).

Morphometric characterization of cyanobacterial isolates

Freshly raised cyanobacterial strains were viewed under 400 and 1000 magnification under a light microscope (Olympus, Japan) (On-line Suppl. Fig. 1) to check the purity and morphological identity according to the keys provided by Desikachary (1959). The length and width of vegetative cells and heterocysts were determined by analysing the captured images using Prog Res Capture Pro 2.6 software (JENOPTIK, Germany) under 400 magnification using a fluorescent microscope (Olympus BX41, Japan). For each cyanobacteria, five vegetative cells and heterocysts (only for heterocystous cyanobacteria) of different filaments were measured. The mean values for length and width of vegetative cells and heterocysts were used for clustering.

Physiological characterization of isolates for tolerance to salinity and alkalinity

The ability of the cyanobacterial isolates to tolerate alkalinity and salinity were checked by growing them on BG-11 N⁺/N^- agar with varying pH (8, 9 and 10) and salinity (5, 10 and 15% (w/v) NaCl concentrations. One mL of freshly grown cyanobacterial culture suspension was taken in sterile microcentrifuge tube (2 mL) and centrifuged at 10000 rpm for 10 min. Pellets were washed with autoclaved distilled water for 2-3 times. Sterile glass beads were added to the culture tube containing cell pellets and 1 mL of autoclaved distilled water was added to it. Cells were vortexed for 30 seconds to homogenise the filamentous cells into suspension. 40 µL of this suspension was spotted on BG-11 medium agar plates of different pH and salt concentrations. The culture plate was incubated for 10 days under conditions as described in the section Procurement and maintenance of cultures. After incubation, plates were observed for growth and results were recorded. Presence of growth was recorded as “1” and absence was recorded as “0”. This binary data was further used for clustering following the procedures detailed in sub-section Genetic diversity analysesin the section Statistical analyses.

Genomic DNA extraction

Genomic DNA was extracted from 1 mL (50–60 mg fresh biomass) of exponentially grown cyanobacterial strains by using Nucleo-pore^® gDNA Fungal/Bacterial kit (Genetix Biotech Asia Pvt. Ltd) following the manufacturer’s protocol with a few modifications. Quantity and purity of DNA was estimated by comparison with known standards in ethidium bromide stained 0.8% agarose (Vivantis) gel.

Multiplex Highly Iterated Palindrome (HIP) fingerprinting

Primers

DNA samples were subjected to amplification using HIP based primers viz. HIP AT, HIP TG and HIP GC in dual combinations. The details of primers sequence used in this study are: (1) HIP AT: 5’-GCGATCGCAT-3’ (2) HIP TG: 5’-GCGATCGCTG-3’ and (3) HIP GC: 5’-GCGATCGCGC-3’ with 60%, 70% and 80% GC content respectively.

PCR amplification

The standard, optimized PCR was performed in a total volume of 50 µL containing 25 µL of 2X Go Taq Green Master mix (Promega), dual combination of HIP primers with 10 pM each of single primer and 90 ng of template DNA (Chakdar and Pabbi 2012). Thermal cycling was achieved in a Master Cycler Gradient (peQSTAR, Germany) as described earlier by Chakdar and Pabbi (2012). PCR products were resolved along with a molecular weight marker (Promega 1 Kb DNA ladder) on 2.0% agarose gel for 6 hours (50 volts) in 1X Tris–Borate–EDTA (TBE) buffer and stained with ethidium bromide solution (1 µg/mL). These were visualized under UV light and gel photographs were captured through gel documentation system (Universal Hood II, BIO-RAD, USA) and the amplification product sizes were determined using the software FluorChem 5500 (Alfa Innotech Corporation, USA). The bands, ranging from 300 to 3000 bp, were scored for diversity analyses. The banding pattern was scored as “1” for presence and “0” for absence of a band. This binary (0, 1) matrix was used for genetic diversity analyses.

Statistical analysis

Morphological diversity analyses

Fuzzy C-Means (FCM) clustering technique is a soft clustering algorithm proposed by Bezdek (1973, 1981). Unlike K-means clustering algorithm in which each data object is the member of only one cluster, a data object is the member of all clusters with varying degrees of fuzzy membership between 0 and 1 in FCM clustering algorithm. Hence, the data objects closer to the centers of clusters have higher degrees of membership than objects scattered in the borders of clusters. For morphological diversity analysis, FCM clustering algorithm was executed. Fitness of the clustering was measured using Dunn's fuzziness coefficientand fuzzy silhouette index. Dunn's fuzzinesscoefficient is a goodness-of-fit criterion for fuzzy clustering that measures how close the fuzzy solution is to the corresponding hard solution. A higher value of Dunn's fuzziness coefficient indicates good clustering. Dunn’s fuzziness coefficient is generally normalized so that it varies from 0 (completely fuzzy) to 1 (hard cluster). The fuzzy silhouette index is another performance evaluation criterion of FCM clustering; it lies between 0 to 1, where 0 indicates completely fuzzy clustering and 1 indicates completely crisp or hard clustering.

Here, FCM clustering was performed for the data generated from morphometry. Before performing the FCM clustering, principal component analysis (PCA) was carried out to reduce the dimension as PCA is a dimension-reduction technique. Besides, by projecting the data into a lower-dimensional space, PCA can help eliminate noise and irrelevant features, making the underlying structure of the data more apparent for clustering. Further, reducing dimensions through PCA can help visualize the data, allowing us to assess the clustering results and the distribution of data points in a more understandable way. Separate clustering was carried out for heterocystous and non-heterocystous cyanobacteria. PCA and FCM cluster analyses were implemented using R version 3.6.3 (2020-02-29), Platform: x86_64-w64-mingw32/x64 (64-bit).

Genetic diversity analyses

Pairwise genetic similarities among the genotypes under study were determined using Jaccard’s coefficient (Jaccard 1908), J = N₁₁/(N₁₁ + N₁₀ + N₀₁), where N₁₁ is the number of bands present in both individuals i and j, N₁₀ is the number of bands present in the individual i and N₀₁ is the number of bands present in the individual j. Cluster analyses were carried out on similarity estimates using the unweighted pair group method with arithmetic mean (UPGMA) using NTSYS pc, version 1.80 (Shalini et al. 2008, 2009). Bootstrap value was determined using Winboot software and the confidence limit of the clustering was also checked. 1000 replicates were used for bootstrap analyses. To test the goodness of fit of a clustering to the set of RAPD data, cophenetic correlation coefficient or cophenetic value was estimated using the COPH and MXCOMP options in NTSYS pc program as described by Chakdar and Pabbi (2012). The degree of fit was interpreted subjectively as 0.9 ≤ r is very good fit, 0.8 ≤ r ≤ 0.9 is good fit, 0.7 ≤ r ≤ 0.8 is poor fit and r < 0.7 is very poor fit.

Results

Measurements of cellular dimensions

The mean length of the vegetative cells of heterocystous cyanobacteria ranged from 3.17 to 10.67 μm while width ranged from 3.08 to 6.05 μm. Hapalosiphon sp. K23 (10.67 μm) and Nostoc sp. K60 (3.17 μm) were recorded as having the highest and lowest cell lengths respectively. Incidentally, Nostoc sp. K60 also showed the lowest (3.17 μm) vegetative cell width while the highest cell width was recorded in Hapalosiphon sp. K76. The average length and width of heterocyst ranged from 3.87-13.12 μm and 3.67-9.69 μm respectively. The highest heterocyst cell length and width were recorded in Hapalosiphonsp. K28 (13.12 μm) and Nostoc sp. K67 (9.69 μm) respectively. On the other hand, the lowest heterocyst cell length and width were recorded in Aulosiralaxa K53 (3.86 μm) and Toxopsis calypsus K49 (3.70 μm) respectively. In the case of non-heterocystous cyanobacteria, the mean cell length and width ranged from 2.87-9.95 μm and 1.9-9.0 μm respectively. Lyngbya hieronymusii K95 was recorded with the highest cell length (9.95 μm) and width (9.0 μm) while Desertifilum sp. K4 (2.87 μm) and Phormidium sp. K6 (2.8 μm) showed the lowest cell length and width, respectively.

Fuzzy clustering based on cell size

For heterocystous and non-heterocystous cyanobacteria, separate clustering was carried out. Analyses of the heterocystous cyanobacteria showed 3 distinct clusters (Fig. 1A) where 78.51% variation was explained by two principal coordinates or Dimensions (Dim 1: -0.386*LV-0.502*WV-0.511*LH-0.581*WH; Dim 2: 0.891*LV+0.011*WV-0.339*LH-0.303*WH). The fitness of the clustering was good as indicated by Dunn's fuzziness coefficient (0.687) and fuzzy silhouette index (0.775) which however also indicated that few of the isolates could be clustered in more than one group.

Figure 1.

Fig. 1. Fuzzy cluster plot of heterocystous cyanobacteria (A) and non-heterocystous cyanobacteria (B). The fuzzy cluster plot was constructed using the cellular dimension of vegetative cells and heterocysts of heterocystous cyanobacteria. Whereas, length (LV) and width of vegetative (WV) cell were used for non-heterocystous cyanobacteria. In case of clustering of heterocystous cyanobacteria, Dimension (Dim) 1 and 2 were the most important principal coordinates explaining 59.17% and 19.33% variation. In Dim1, width of heterocyst contributed maximum variation while in dimension 2, length of vegetative cell had the major contribution.

Isolates of Hapalosiphon spp. viz. Hapalosiphon sp. K23, Hapalosiphon sp. K28, Hapalosiphon sp. K32, Hapalosiphon sp. K50, Hapalosiphon sp. K62, Hapalosiphon sp. K76 and Hapalosiphon sp. K99) were distributed in Cluster 1 and 2. Hapalosiphonsp. K23 showed ~31% probability of being placed in Cluster 2 while Hapalosiphon sp. K50 showed ~26% probability of being placed in Cluster 3 (On-line Suppl. Tab. 2). All isolates of A. laxa, T. calypsus, Scytonema sp. and Nostoc spp. viz. Nostoc sp. K51, Nostoc sp. K56, Nostoc sp. K60 and Nostoc sp. K68 (except Nostoc sp. K67, which was placed in Cluster 2) were placed in Cluster 3. However, two isolates of A. laxa K24, A. laxa K39, Nostoc sp. K68 showed probabilities of being placed in Cluster 2 (On-line Suppl. Tab. 2). Isolates placed in Cluster 1 showed close proximity among themselves while isolates placed in Cluster 2 and 3 showed high variability. The results related to cluster prototype presented in Tab. 1 showed that clusters had significant variations with respect to cellular dimensions among themselves indicating high inter-generic variations as clustering was related to their generic identity.

Tab. 1. Final cluster prototypes for heterocystous and non-heterocystous cyanobacteria. LV – length of vegetative cell; WV – width of vegetative cell; LH – length of heterocyst; WH – width of heterocyst; NA – not applicable.

Table 1.

I. Heterocystous cyanobacteria
	LV	WV	LH	WH	Remarks
Cluster 1	7.167261	5.36375	7.61384	6.494746	Large vegetative cells and medium sized heterocysts
Cluster 2	4.956879	4.76203	9.432821	6.941643	Small to medium sized vegetative cells and large heterocysts
Cluster 3	4.531235	4.297334	4.943985	4.835359	Small vegetative cells and small heterocysts
II. Non-heterocystous cyanobacteria
Cluster 1	3.507888	2.794574	NA	NA	Large vegetative cells
Cluster 2	9.937472	8.983922	NA	NA	Small to medium sized vegetative cells
Cluster 3	5.76292	2.515487	NA	NA	Small vegetative cells

In the case of non-heterocystous cyanobacteria also, three clusters were formed where Cluster 2 contained only one isolate L. hieronymusii K95 (Fig. 1B). Dunn's fuzziness coefficient (0.796) and fuzzy silhouette index (0.684) showed the good fit of the clustering; however, the fitness indices also indicated that a few of the isolates could be clustered in more than one group. Isolates of Desertifilum sp., Lyngbya sp. and Phormidium sp. grouped both in Cluster 1 and Cluster 3 while Alkalinema pantanalense K25 and isolates of Halomicronema sp. were placed exclusively in Cluster 1 and Cluster 3 respectively. Halomicronema sp. K70 had ~32% probability of being placed in Cluster 1 while Desertifilum sp. K18 had ~27% probability of being grouped in Cluster 3. Lyngbya wollei K80 had ~ 26% probability of being placed in both Cluster 2 and Cluster 3. Similarly, Halomicronema sp. K40 had 50% probability of being grouped in both Cluster 1 and Cluster 3 (On-line Suppl. Tab. 3). For non-heterocystous cyanobacteria also, clustering showed a high degree of both inter- and intra-generic variability with respect to cell dimensions (Tab. 1).

Physiological characterization of isolates for tolerance to salinity and alkalinity

All cyanobacterial isolates except Lyngbya aestuarii K97 (tolerated pH 8-9) showed tolerance to pH ranging from 8-10 (Fig. 2A). However, great variation was observed in the case of salinity tolerance (Fig. 2B). Among the non-heterocystous cyanobacteria, isolates K4, K18, K19 of Desertifilum sp. and Phormidium sp. K30 showed tolerance to 15% NaCl (w/v). Only four isolates viz. Halomicronema sp. K79, Desertifilum sp. K20, L. aestuarii K97 and Leptolyngbya sp. K72 showed tolerance to 5% NaCl. Among the heterocystous cyanobacteria, A. laxa K24, Hapalosiphon sp. K28, Hapalosiphon sp. K62 and Scytonema sp. K55 showed tolerance to 15% NaCl. 16 heterocystous isolates could tolerate NaCl only to 5% while 3 isolates showed tolerance to NaCl only up to 10%.

Figure 2.

Fig. 2. Representative image depicting growth of heterocystous/non-heterocystous cyanobacteria on BG11 N⁺/N^- agar plate amended to different alkalinity (i.e., pH 8, 9 and 10) (A) and salinity (i.e., 5, 10 and 15% NaCl concentration) (B).

Cluster analyses of the non-heterocystous isolates showed (Fig. 3A) three distinct clusters viz. a) high salinity & high alkalinity tolerant (04 isolates); b) Low salinity & high alkalinity tolerant (04 isolates) and c) only high alkalinity tolerant (09 isolates). On the other hand, heterocystous cyanobacteria showed (Fig. 3B) four distinct clusters viz. a) high salinity & high alkalinity tolerant (04 isolates); b) moderately high salinity & high alkalinity (03 isolates); c) low salinity & low alkalinity tolerant (09 isolates) and d) only high alkalinity tolerant (14 isolates).

Figure 3.

Fig. 3. Dendrogram representing the clustering of filamentous cyanobacteria based on their ability to tolerate variable levels of alkalinity and salinity. A – clustering of non-heterocystous cyanobacteria; B – clustering of heterocystous cyanobacteria. The heterocystous cyanobacteria were clustered in three distinct groups: high alkalinity tolerant; low salinity and high alkalinity tolerant; and high salinity and high alkalinity tolerant. The non-heterocystous cyanobacteria were clustered in four distinct groups: high alkalinity tolerant; low salinity and high alkalinity tolerant; high salinity and high alkalinity tolerant; and moderately high salinity and high alkalinity tolerant.

HIP fingerprinting

All the three primer combinations HIP AT + HIP TG; HIP AT + HIP GC and HIP TG + HIP GC showed 100% polymorphism (Tab. 2, Fig. 4A-C). HIP AT + HIP TG produced a total of 357 fragments ranging from 170-2156 bp. HIP AT + HIP GC set produced 379 fragments ranging from 157-2129 bp. 269 fragments ranging from 103 bp to 2547 bp were generated by HIP TG + HIP GC. For the purpose of scoring, only the fragments ranging from 300-3000 bp were used.

Tab. 2. Size range of the PCR products generated by three primer combinations.

Table 2.

Sl. No.	Primer combination	Total no. of fragments	% polymorphic bands	Size range (bp)
	HIP AT + HIP TG	357	100	170 - 2156
	HIP AT + HIP GC	379	100	157 - 2129
	HIP TG + HIP GC	269	100	103 - 2547

Figure 4.

Fig. 4. Gel photograph presenting the HIP (Highly Iterated Palindrome) element based fingerprint of the forty-seven different filamentous heterocystous and non-heterocystous cyanobacterial isolates generated using dual primer combinations: A – HIP AT + HIP TG, B – HIP AT + HIP GC, C – HIP TG + HIP GC. L: 1kb DNA ladder (Promega, G571A).

Genetic diversity analyses

Genetic diversity analyses for heterocystous and non-heterocystous cyanobacteria were carried out separately. Genetic similarity among the 17 non-heterocystous cyanobacteria range from ~13-90% with two major clusters. The major cluster 1 comprised all isolates of Halomicronema, Leptolyngbya sp. K72, L. wollei K80, L. hieronymusii K81 and K95, Lyngbya sp. K5, Phormidium sp. K30 and A. pantanalense K25 (Fig. 5A). On the other hand, major cluster 2 contained all isolates of Desertifilum, Phormidium sp. K6, and L. aestuarii K97. Desertifilum sp. K18 and K19 showed ~90% similarity indicating that they may be different isolates of the same species. The fitness of the clustering was found to be good (r = 0.87).

Figure 5.

Fig. 5. Dendrogram depicting genetic diversity based on multiplex HIP fingerprinting: A – non-heterocystous cyanobacteria, B – heterocystous cyanobacteria. The values at the nodes are bootstrap values. Bootstrap values > 50 have been presented in the dendrograms.

Among the 30 heterocystous cyanobacteria, the genetic similarity ranged from ~12-76%. Two major clusters viz. cluster 1 with 29 cyanobacteria and cluster 2 with only T. calypsus K49 were observed (Fig. 5B). In the major cluster 1, all 15 isolates of Aulosira and 7 isolates of Hapalosiphon formed separate sub-clusters. On the other hand, Nostoc spp. showed clustering with Hapalosiphon (Nostoc sp. K51, N. carneum K56, and Nostoc sp. K68) as well as Scytonema sp. K55 (Nostoc sp. K67 and Nostoc sp. K60). Among the A. laxa isolates clustered together, the genetic similarity ranged from 26-76% while the same was found in a range of 31-57% in the case of Hapalosiphon sp. isolates. The degree of fitness of clustering was good (r = 0.89). The result clearly showed a very high intra-generic diversity among the studied filamentous heterocystous cyanobacteria.

In the majority of the cases the clustering of both non-heterocystous and heterocystous cyanobacteria was supported by significant bootstrap values (Fig. 5). More than 70% bootstrap values correspond to > 95% probability that true phylogeny have been found (Hillis and Bull 1993).

Discussion

Filamentous cyanobacteria like Nostoc, Anabaena, Phormidium, Microcoleus, Lyngbya, Calothrix etc. are frequently encountered in BSC (Garcia-Pichel et al. 2001, Zhang 2005). However filamentous cyanobacteria like Nostoc, Anabaena, Calothrix, Microcoleus, Hapalosiphon, Cylindrospermum, Scytonema, Aulosira, Phormidium and Oscillatoria have been reported to be present not only in BSCs from arid and semi-arid areas but also in alkaline and saline soils (Pandey et al. 2005, Srivastava et al. 2009). The predominance of heterocystous or non-heterocystous cyanobacteria is reported to be governed by factors like pH, salinity or nitrogen availability (Srivastava et al. 2009). Many of such filamentous cyanobacteria have been used for the amelioration of salinity stress in various crops, but it is noteworthy that the majority of these cyanobacteria were from four genera viz. Anabaena, Nostoc, Hapalosiphon and Calothrix (Li et al. 2019).

Cluster analyses of the 47 filamentous cyanobacteria based on cellular dimensions and tolerance to salinity and alkalinity showed high variability among the isolates. Among the heterocystous cyanobacteria, significant morphological variability was observed within the members of the genus Hapalosiphon while the other genera like Aulosira or Nostoc had limited intra-generic variability. In case of non-heterocystous cyanobacteria, inter- and intra-generic morphological variability was higher as all the clusters contained isolates belonging to different genera. Among non-heterocystous cyanobacteria, isolates of Lyngbya showed higher morphological variations. Initial classifications of cyanobacteria were solely based on morphology and still an important criterion for identification. However, many times their variation with environmental conditions makes them unsuitable for correct identification. Very few studies, like Mishra et al. (2015), reported that morphological attributes like trichome aggregation, heterocyst shape and akinete shape are stable features and could be used for identification. The results of the present study indicated that dimensions of vegetative cells and heterocysts can be effective for identification and differentiation of cyanobacteria belonging to Hapalosiphon and Lyngbya. This observation is in compliance with the distinction of various species of Hapalosiphon (like H. welwitschii West & G.S.West, H. delicatulus West & G.S.West, H. intricatus West & G.S.West, H. pumilus Kirchner ex Bornet & Flahault, etc.) and Lyngbya (like L. chaetomorphae Iyengar & Desikachary, L. lachneri (W.Zimmermann) Geitler, L. infixa Frémy, L. baculum Gomont, etc.) based on dimensions of filaments or trichomes (Desikachary 1959). Looking at the advances in computational techniques and the development of robust algorithms, it does not seem to be really impossible to use morphological information for understanding actual biological diversity. Heterocystous cyanobacteria showed high intra-generic variability in the case of salinity while all isolates were highly alkali-tolerant. Isolates of Aulosira and Hapalosiphon were found to have variable tolerance to salinity. However, the variability of salinity tolerance was lower in non-heterocystous than in heterocystous cyanobacteria. Growth and colonization of cyanobacteria are known to be greatly influenced by soil pH and salinity (Pandey et al. 2005, Nayak and Prasanna 2007). Cyanobacterial growth is favoured under neutral to alkaline conditions while acidic conditions may limit the growth of many cyanobacteria (Šesták 2001). In the present study, all the isolates of cyanobacteria studied could also tolerate up to pH 10. It was been reported that the metabolic trade-off between ionic balance and heterocyst formation or diazotrophy may limit the proliferation of heterocystous cyanobacteria under highly saline conditions (Vitousek et al. 2002, Berman-Frank et al. 2003). Srivastava et al. (2009) reported that low salinity favoured the presence of heterocystous cyanobacteria while highly saline soils predominantly harboured non-heterocystous cyanobacteria. Kirkwood et al. (2008) reported that heterocystous cyanobacteria could still persist under saline conditions although it might be suboptimal for growth. As observed in the present study, the varying tolerance of heterocystous cyanobacteria to salinity might be an important fitness trait to proliferate under saline/alkaline soils. Consistently with the reports of Srivastava et al. (2009), Aulosira turned out to be highly salinity-adapted (5-15%) heterocystous cyanobacteria. All the isolates of Desertifilum sp. showed higher salinity tolerance than the other non-heterocystous cyanobacteria. The salinity tolerance exhibited by the Desertifilum isolates in the present study are much beyond the tolerance limit (3%) of all the reported species of the genus Desertifilum (Dadheech et al. 2014, Cai et al. 2017).

HIP fingerprinting of heterocystous revealed higher genetic variability than in non-heterocystous cyanobacteria. Clustering of non-heterocystous cyanobacteria based on HIP fingerprinting also showed high intra- and inter-generic variability. Despite being the most commonly used molecular marker for understanding phylogeny and describing novel prokaryotic taxa, the 16S rRNA gene often cannot resolve intra-generic diversity (Woese 1987, Tindall et al. 2010). Although 16S rRNA gene contains informative hypervariable regions, it does not have enough divergence to resolve the differences among the closely related members of a genus (Fox et al. 1992, Drugă et al. 2013). Repetitive elements like HIP can be very effective for such purposes due to their abundance throughout the cyanobacterial genomes (Xu et al. 2018). Earlier studies showed that molecular tools like RAPD and HIP fingerprinting can effectively resolve the intra-generic diversity of cyanobacteria like Nostoc, Anabaena, Hapalosiphon, Calothrix etc. (Chakdar and Pabbi 2012, Shukla et al. 2013, Singh et al. 2014). In the present study, the isolates of A. laxa, Hapalosiphon sp. and Nostoc sp. showed distinct clustering with high intra-generic variability indicating that multiplex HIP fingerprinting could effectively differentiate among closely related members of these genera. In the case of non-heterocystous cyanobacteria like Phormidium and Lyngbya 16S rRNA genes have been reported to be insufficient for intra-generic taxonomic resolution (Marquardt and Palinska 2007, Engene et al. 2010). In the present study, Lyngbya was also found to be genetically highly heterogeneous and HIP fingerprinting could distinguish between two L. hieronymusii isolates. Furthermore, HIP fingerprinting could effectively distinguish the isolates belonging to Halomicronema and Desertifilum.

Conclusion

The results of the present study showed the importance of morphological, physiological and genetic analyses to understand the diversity of filamentous cyanobacteria. All these analyses can supplement each other to provide a better understanding of the population structure and ecology of these cyanobacteria. The findings indicated that the variation in salinity tolerance of filamentous cyanobacteria along with their inherent alkali tolerance helped to proliferate a heterogeneous population of both heterocystous and non-heterocystous cyanobacteria in saline/alkaline soils. With great genetic and physiological diversity, such cyanobacteria can be a potential biological resource for the reclamation of such degraded soils. However, further in-depth studies are required to understand their actual genetic and physiological potential.

Author contribution statement

H.C. contributed in conceptualization of the work. S.V. carried out the experiments and validated the study. R.V. carried out the morphometric characterization. A.B. contributed to the statistical analysis. S.V. and H.C. investigated the study and prepared the original draft. H.C., S.V., S.Y.B., N.S. and V.M. contributed in writing-reviewing the draft. H.C. and A.K.S. supervised the study and edited the final draft and gave final approval for the publication of this version. All authors read and approved the final manuscript.

Availability of data and material

Gene sequences are available in NCBI, Cultures are available in National Agriculturally Important Microbial Culture Collection (NAIMCC) at ICAR-NBAIM, Mau, Uttar Pradesh, India and other information are available with the corresponding author.

References

Akoijam, C., Singh, A. K., 2011: Molecular typing and distribution of filamentous heterocystous cyanobacteria isolated from two distinctly located regions in North-Eastern India. World Journal of Microbiology and Biotechnology 27(9), 2187–2194. https://doi.org/10.1007/s11274-011-0684-8

Bagul, S. Y., Tripathi, S., Chakdar, H., Karthikeyan, N., Pandiyan, K., Singh, A., Kumar, M., 2018: Exploration and characterization of cyanobacteria from different ecological niches of India for phycobilins production. International Journal of Current Microbiology and Applied Sciences 7(12), 2822–2834. https://doi.org/10.20546/ijcmas.2018.712.321

Berman-Frank, I., Lundgren, P., Falkowski, P., 2003: Nitrogen fixation and photosynthetic oxygen evolution in cyanobacteria. Research in Microbiology 154(3), 157–164. https://doi.org/10.1016/S0923-2508(03)00029-9

Bezdek, J. C., 1973: Cluster validity with fuzzy sets. Journal of Cybergenetics 3, 58–73. https://doi.org/10.1080/01969727308546047

Bezdek, J. C., 1981: Partition recognition with fuzzy objective function algorithms. Springer, New York. https://doi.org/10.1007/978-1-4757-0450-1

Bhatnagar, A., Bhatnagar, M., 2005: Microbial diversity in desert ecosystems. Current Science 89, 91–100.

Cai, F. F., Chen, Y. X., Zhu, M. L., Li, X. C., Li, R., 2017: Desertifilum salkalinema sp. nov. (Oscillatoriales, Cyanobacteria) from an alkaline pool in China. Phytotaxa 292(3), 262–270. https://doi.org/10.11646/phytotaxa.292.3.6

Chakdar, H., Jadhav, S. D., Dhar, D. W., Pabbi, S., 2012: Potential applications of blue green algae. Journal of Scientific & Industrial Research 71(1), 13-20.

Chakdar, H., Pabbi, S., 2012: Morphological characterization and molecular fingerprinting of Nostoc strains by multiplex RAPD. Microbiology 81(6), 710–720. https://doi.org/10.1134/S0026261712060070

Chakdar, H., Pabbi, S., 2017: A comparative study reveals the higher resolution of RAPD over ARDRA for analyzing diversity of Nostoc strains. 3 Biotech 7(2), 1–10. https://doi.org/10.1007/s13205-017-0779-5

Dadheech, P. K., Mahmoud, H., Kotut, K., Krienitz, L., 2014: Desertifilum fontinale sp. nov. (Oscillatoriales, Cyanobacteria) from a warm spring in East Africa, based on conventional and molecular studies. Fottea 14(2), 129–140. https://doi.org/10.5507/fot.2014.010

Desikachary, T. V., 1959: Cyanophyta. Indian Council of Agricultural Research, New Delhi.

Doers, M. P., Parker, D. L., 1988: Properties of Microcystis aeruginosa and M. Flos‐aquae (cyanophyta) in culture: taxonomic implications. Journal of Phycology 24(4), 502–508. https://doi.org/10.1111/j.1529-8817.1988.tb04254.x

Drugă, B., Welker, M., Sesărman, A., Hegedus, A., Coman, C., Sicora, C., Dragoş, N., 2013: Molecular characterization of microcystin-producing cyanobacteria from Romanian fresh waters. European Journal of Phycology 48(3), 287–294. https://doi.org/10.1080/09670262.2013.822563

Engene, N., Cameron Coates, R., Gerwick, W. H., 2010: 16S rRNA Gene heterogeneity in the filamentous marine cyanobacterial genus Lyngbya. Journal of Phycology 46(3), 591–601. https://doi.org/10.1111/j.1529-8817.2010.00840.x

Ezhilarasi, A., Anand, N., 2010: Characterization of cyanobacteria within the genus Anabaena based on SDS-PAGE of whole cell protein and RFLP of the 16S rRNA gene. ARPN Journal of Agricultural and Biological Science 5(4), 44–51.

Fox, G. E., Wisotzkey, J. D., Jurtshuk, J. R. P., 1992: How close is close: 16S rRNA sequence identity may not be sufficient to guarantee species identity. International Journal of Systematic and Evolutionary Microbiology 42(1), 166–170. https://doi.org/10.1099/00207713-42-1-166

Garcia-Pichel, F., López-Cortés, A., Nübel, U., 2001: Phylogenetic and morphological diversity of cyanobacteria in soil desert crusts from the Colorado Plateau. Applied and Environmental Microbiology 67(4), 1902–1910. https://doi.org/10.1128/AEM.67.4.1902-1910.2001

Gould, S. B., Waller, R. F., McFadden, G. I., 2008: Plastid evolution. Annual Review in Plant Biology 59, 491–517. https://doi.org/10.1146/annurev.arplant.59.032607.092915

Gupta, A., Morby, A. P., Turner, J. S., Whitton, B. A., Robinson, N. J., 1993: Deletion within the metallothionein locus of cadmium‐tolerant Synechococcus PCC 6301 involving a highly iterated palindrome (HIP1). Molecular Microbiology 7(2) 189–195. https://doi.org/10.1111/j.1365-2958.1993.tb01110.x

Hillis, D. M., Bull, J. J., 1993: An Empirical Test of Boot strapping as a Method for Assessing Confidence in Phylogenetic Analysis. Systematic Biology 42(2), 182–192. https://doi.org/10.2307/2992540

Jaccard, P., 1908: Nouvelles recherches sur la distribution florale. Bulletin de la Société Vaudoise des Sciences Naturelles 44, 223–270. http://dx.doi.org/10.5169/seals-268384

Kato, T., 1991: Allozyme divergence in Microcystis (Cyanophyceae) and its taxonomic inference. Algological Stududies 64, 157–226.

Kirkwood, A. E., Buchheim, J. A., Buchheim, M. A., Henley, W. J., 2008: Cyanobacterial diversity and halotolerance in a variable hypersaline environment. Microbial Ecology 55(3), 453–465. https://doi.org/10.1007/s00248-007-9291-5

Li, H., Zhao, Q., Huang, H., 2019: Current states and challenges of salt-affected soil remediation by cyanobacteria. Science of The Total Environment 669, 258–272. https://doi.org/10.1016/j.scitotenv.2019.03.104

Lyra, C., Laamanen, M., Lehtimäki, J. M., 2005: Benthic cyanobacteria of the genus Nodularia are non-toxic, without gas vacuoles, able to glide and genetically more diverse than planktonic Nodularia. International Journal of Systematic and Evolutionary Microbiology 55(2), 555–568. https://doi.org/10.1099/ijs.0.63288-0

Marquardt, J., Palinska, K. A., 2007: Genotypic and phenotypic diversity of cyanobacteria assigned to the genus Phormidium (Oscillatoriales) from different habitats and geographical sites. Archives of Microbiology 187, 397–413. https://doi.org/10.1007/s00203-006-0204-7

Mazel, D., Houmard, J., Castets, A. M., de Marsac, N. T., 1990: Highly repetitive DNA sequences in cyanobacterial genomes. Journal of Bacteriology 172(5), 2755–2761. https://doi.org/ 10.1128/jb.172.5.2755-2761.1990

Mishra, S., Bhargava, P., Adhikary, S. P., Pradeep, A., Rai, L. C., 2015: Weighted morphology: A new approach towards phylogenetic assessment of Nostocales (Cyanobacteria). Protoplasma 252(1), 145–163. https://doi.org/10.1007/s00709-014-0629-9

Nayak, S., Prasanna, R., 2007: Soil pH and its role in cyanobacterial abundance and diversity in rice field soils. Applied Ecology and Environmental Research 5(2), 103–113.

Neilan, B. A., Saker, M. L., Fastner, J., Törökné, A., Burns, B. P., 2003: Phylogeography of the invasive cyanobacterium Cylindrospermopsis raciborskii. Molecular Ecology 12(1), 133–140. https://doi.org/10.1046/j.1365-294x.2003.01709.x

Pandey, K. D., Shukla, P. N., Giri, D. D., Kashyap, A. K., 2005: Cyanobacteria in alkaline soil and the effect of cyanobacteria inoculation with pyrite amendments on their reclamation. Biology and Fertility of Soils 41(6), 451–457. https://doi.org/10.1007/s00374-005-0846-7

Prabina, B. J., Kumar, K., Kannaiyan, S., 2005: DNA amplification fingerprinting as a tool for checking genetic purity of strains in the cyanobacterial inoculum. World Journal of Microbiology and Biotechnology 21(5), 629–634. https://doi.org/10.1007/s11274-004-3566-5

Prasanna, R., Kumar, R., Sood, A., Prasanna, B. M., Singh, P. K., 2006: Morphological, physiochemical and molecular characterization of Anabaena strains. Microbiological Research 161(3), 187–202. https://doi.org/10.1016/j.micres.2005.08.001

Rossi, F., Li, H., Liu, Y., De Philippis, R., 2017: Cyanobacterial inoculation (cyanobacterisation): perspectives for the development of a standardized multifunctional technology for soil fertilization and desertification reversal. Earth-Science Reviews 171, 28–43. https://doi.org/10.1016/j.earscirev.2017.05.006

Selvakumar, G., Gopalaswamy, G., 2008: PCR based fingerprinting of Westiellopsis cultures with short tandemly repeated repetitive (STRR) and highly iterated palindrome (HIP) sequences. Biologia 63(3), 283–288. https://doi.org/10.2478/s11756-008-0065-4

Šesták, Z., 2001: Whitton, B. A., Potts, M., (ed.): The ecology of cyanobacteria: Their diversity in Time and Space. Photosynthetica 39(3), 466. https://doi.org/10.1023/A:1015127720748

Shalini, Dhar, D. W., Gupta, R. K., 2008: Phylogenetic analysis of cyanobacterial strains of genus-Calothrix by single and multiplex randomly amplified polymorphic DNA-PCR. World Journal of Microbiology and Biotechnology 24(7), 927–935. https://doi.org/10.1007/s11274-007-9569-2

Shalini, S., Dhar, D., Gupta, R., 2009: Morphological and physiochemical characterisation of Calothrix strains. Acta Botanica Hungarica 51(1-2), 195–216. https://doi.org/10.1556/abot.51.2009.1-2.19

Sharma, D. K., Singh, A., 2015: Salinity research in India-achievements, challenges and future prospects. Water and Energy International 58(6), 35–45.

Shokraei, R., Fahimi, H., Blanco, S., Nowruzi, B., 2019: Genomic fingerprinting using highly repetitive sequences to differentiate close cyanobacterial strains. Microbial Bioactives 2(1), 68–75. https://doi.org/10.25163/microbbioacts.21015A2624310119

Shukla, E., Singh, S. S., Mishr, A. K., 2013: Fingerprinting and phylogeny of some heterocystous cyanobacteria using short tandemly repeated repetitive and highly iterated palindrome sequences. Microbiologia 82(6), 801–808. https://doi.org/10.1134/S0026261714010123

Singh, P., Kaushik, M. S., Srivastava, M., Mishra, A.K., 2014: Phylogenetic analysis of heterocystous cyanobacteria (Subsections IV and V) using highly iterated palindromes as molecular markers. Physiology and Molecular Biology of Plants 20(3), 331–342. https://doi.org/10.1007/s12298-014-0244-4

Smith, J. K., Parry, J. D., Day, J. G., Smith, R. J., 1998: A PCR technique based on the Hipl interspersed repetitive sequence distinguishes cyanobacterial species and strains. Microbiologia 144(10), 2791–2801. https://doi.org/10.1099/00221287-144-10-2791

Srivastava, A. K., Bhargava, P., Kumar, A., Rai, L. C., Neilan, B.A., 2009: Molecular characterization and the effect of salinity on cyanobacterial diversity in the rice fields of Eastern Uttar Pradesh, India. Saline Systems 5(4), 1–17. https://doi.org/10.1186/1746-1448-5-4

Stanier, R. Y., Kunisawa, R., Mandel, M. C. B. G., Cohen-Bazire, G., 1971: Purification and properties of unicellular blue-green algae (order Chroococcales). Bacteriological Reviews 35(2), 171–205.

Tindall, B. J., Rosselló-Móra, R., Busse, H. J., Ludwig, W., Kämpfer, P., 2010: Notes on the characterization of prokaryote strains for taxonomic purposes. International Journal of Systematic and Evolutionary Microbiology 60(1), 249–266. https://doi.org/10.1099/ijs.0.016949-0

Vitousek, P. M., Cassman, K. E. N., Cleveland, C., Crews, T., Field, C. B., Grimm, N. B., Howarth, R. W., Marino, R., Martinelli, L., Rastetter, E. B., Sprent, J. I., 2002: Towards an ecological understanding of biological nitrogen fixation. In: Boyer E.W., Howarth R.W. (eds), The nitrogen cycle at regional to global scales, 1–45. Springer, Dordrecht. https://doi.org/10.1007/978-94-017-3405-9_1

Woese, C. R., 1987: Bacterial evolution. Microbiological Reviews 51(2), 221–271.

Xu, M., Lawrence, J. G., Durand, D., 2018: Selection, periodicity and potential function for Highly Iterative Palindrome-1 (HIP1) in cyanobacterial genomes. Nucleic Acids Research 46(5), 2265–2278. https://doi.org/10.1093/nar/gky075

Zhang, Y., 2005: The microstructure and formation of biological soil crusts in their early developmental stage. Chinese Science Bulletin 50(2), 117–121. https://doi.org/10.1007/BF02897513