Khao Luang National Park (Nakhon Si Thammarat province) (Fig. 1), is an important watershed region for the surrounding farms and villages, located between the following geographic coordinates: 08°30′ N, 99°44′ E. Covering an area of approximately 570 km², it is mountainous with many tall peaks and has forests, wildlife, rivers and waterfalls (National Park, Wildlife and Plant Conservation Department 2006). Due to its unique location on the peninsula, it receives heavy monsoon rains throughout the year from both the east and the west coasts. The climate is tropical, with well-demarcated rainy and dry seasons. We selected 22 stream segments accessible by car around community areas of seven different districts of Nakhon Si Thammarat province in Khao Luang National Park (designated K1 to K22) (Table 1).

Physical and chemical stream variables were measured and two replicates of each alga were collected at each site in each of three seasons over a period of 12 months (May 2014 to April 2015): the early rainy season (July–August 2014, SW monsoon), the heavy rainy (November–December 2014, NE monsoon), and the dry season (March–April 2015). Species exploring were conducted along a transect line marked by permanent plots. A long-transect length surveyed is 100 m per sampling site. In each of the 22 sites, a permanent plot 10 m in length along the stream margin was used as the sampling area were chosen at random (modified from Necchi 1993, Necchi et al. 1999). Along this 10 m long transect, divided into 10 at equal intervals, red algal occurrence (presence and percent cover) was estimated visually on the stream bottom and abundance was averaged over the entire stream segment according to the following scale (modified from Necchi 1993): 0, absent; 1, <10%; 2, 11–20%; 3, 21–30%; 4, 31–40%; 5, >40%. This procedure was followed in two replicates in each site and data were computed as the average for the sampling site.

In each study site, water samples were collected for measurement according to American Public Health Association, American Water Works Association, and Water Pollution Control Pact standards (Eaton et al. 1995). Temperature, water depth, current velocity, conductivity, total dissolved solid, dissolved oxygen (DO), and pH values were measured in the field using standard portable instruments (Portable Multi Parameter Photometer, MD600; Lovibond, Dortmund, Germany). Nutrients, ammonium-nitrogen (NH₃-N), nitrate-nitrogen (NO₃-N), and orthophosphate (PO₄-P), acidity, alkalinity, and hardness were analyzed in the laboratory by standard methods (Eaton et al. 1995) from water samples kept at 4°C until analysis. The trophic status was evaluated using the Applied Algal Research Laboratory Physical and Chemical score (AARL-PC score) (Peerapornpisal et al. 2004). This composite index defines water quality categories using dissolved oxygen, conductivity, nitrate-nitrogen, ammonia-nitrogen, and soluble reactive phosphorus (SRP).

Algal specimens were preserved in 2% glutaraldehyde in the field. Species were identified using morphological characters according to Jose and Patel (1990), Kumano (2002), Traichaiyaporn et al. (2007, 2012), Necchi and Vis (2012), and Necchi (2016). The community structure was evaluated using the Shannon-Wiener diversity index (H′) and Pielou’s evenness index (J′) (Brower et al. 1997). These indices were calculated from the annual average scores of red algal species records. Shannon-Wiener diversity -∑n=inpi log2 pi with i proportion of each algal species (percent cover) to the total abundance, Pielou’s evenness = H′ / H_max, where H′ = Shannon-Wiener diversity index, pi = the proportion of individual average scores of the species i^th species expressed as a proportion of total abundance, H_max = log₂S, where S = the total number of species.

Univariate analysis, using the mean and standard deviation, was applied to describe and evaluate the water quality among sampling sites and seasons. Multivariate indices analysis, cluster and ordination multi-dimensional scaling (MDS) analyses were used to classify similarity of water quality parameters among seasons and sites. The cluster and MDS analyses aimed to find grouping of similarities of samples among different sites and season that samples within a group are more similar to each other than samples in different groups. The resulting cluster was shown in tree-like diagram called a dendrogram and MDS was shown as biplot. The correspondence of the ordination to the distances shown as a stress value, where a value less than 0.05 is an excellent fit, a value less than 0.10 is a good fit, a value less than 0.20 is a potential fit and a value greater than 0.30 is an arbitrary fit. Both the clustering dendrogram and map were constructed using a standardized Euclidean distances matrix for the water variables. Analysis of similarity (ANOSIM) was also performed to determine significant differences in the different groups of water parameters. The similarity percentages (SIMPER) was used to test the parameters within groups. All analysis procedures used were based on Clark and Warwick (1994) using PRIMER program (ver. 5 for Windows; Bourne Press Limited, Bournemouth, UK). The relationships between red algae assemblages and environment variables were explored by Canonical Correspondence Analysis (CCA) using MVSP (MultiVariate Statistical Package) for Windows ver. 3.22. The result was shown by CCA plot.

In addition, the trophic status of water was evaluated from the main parameters such as dissolved oxygen, conductivity, ammonia-nitrogen, nitrate-nitrogen and SRP by AARL-PC score, based on Peerapornpisal et al. (2004).

Most stream segments (20 of 22) were slightly acidic (pH 6.3 ± 0.2 to 6.7 ± 0.6) while two stream segment (K5 and K9) were slightly basic (pH 7.4 ± 0.5 and 7.1 ± 1.2), with high DO content (3.51 ± 1.5 to 7.8 ± 0.7 mg L⁻¹) and conductivity levels (12.4 ± 10.1 to 128.3 ± 24.4 μs cm⁻¹). All stream sites had clear water with low turbidity (except K6 site) and relatively high water temperature (varied from 24.5 ± 0.4 to 26.5 ± 0.6°C). The current velocity was highly variable, ranging from 2.0 ± 0.1 to 116.5 ± 22.2 cm s⁻¹. The turbidity ranged from 1.0 ± 0.1 to 116.5 ± 22.2 nephelometric turbidity unit. Total dissolved solids (TDS) ranged from 6.4 ± 1.1 to 58.4 ± 20.2 mg L⁻¹. In acid samples CaCO₃ varied from 2.3 ± 0.3 to 19.8 ± 4.4 mg L⁻¹. In alkaline samples alkalinity is accurately expressed as CaCO₃, which was 4.3 ± 1.5 to 81.3 ± 0.8 mg L⁻¹. In a wide range of samples hardness as CaCO₃ varied from 5.1 ± 0.8 to 130.8 ± 59.6 mg L⁻¹. Ammonia-nitrogen values varied from <0.02 ± 0.0 to 0.22 ± 0.1 mg L⁻¹ (Table 2).

Most species occurred in slightly acidic water, which is consistent with data from Sheath and Hambrook (1990), who reported freshwater red algae occurring mostly in mildly acidic streams ranging between pH 6.0 and 7.0. The AARL-PC showed most sites within the range of ultraoligotrophic to oligotrophic status, and few in the oligo-mesotrophic (Table 3). Most of the sampling sites showed similar trophic status throughout the year and were classified to clean to very clean. This finding is probably due to geographical proximity of the preserved area of Khao Luang National Park. However, the sampling sites K2, K4, K5, K14, and K22 located in the downstream area showed moderate contamination levels by the agriculture activities, evidenced by oligo-mesotrophic status (clean to moderate clean). These characteristics were similar to those described in eight National parks of five provinces in northern Thailand (Traichaiyaporn et al. 2005).

Cluster and MDS analyses of water quality parameter variations among seasons and sites showed three groups. The three groups (70% similarity and ANOSIM test results of R = 0.977, p < 0.01) were group 1—K6W1 and K2W2, group 2—K3D, K3W1, K5D, K5W1, K5W2, K6D, K7W1, and K7W2, and group 3 composed of the rest of the sampling dates (56 variables) (Fig. 2A & B). One parameter indicated similarity within each groups: turbidity contributed greater than 15% similarity in the first group; TDS, hardness and velocity in the second group; TDS and conductivity in the third group. The status of Khao Luang National Park water quality is similar to the results of Thiemdoa et al. (2007) in other areas of northern and southern Thailand with mostly clean waters.

A total of 16 red algal species were identified (Table 4), belonging to eight genera Audouinella, Balliopsis, Batrachospermum, Caloglossa, Compsopogon, Kumanoa, Sirodotia, and Thorea) that are distributed among the families Acrochaetiaceae, Batrachospermaceae, Compsopogonaceae, Delesseriaceae, and Thoreaceae, and occurred in all collecting sites sampled (Table 3). In 16 sites only one species was found, whereas site K5 had the highest species richness with four species (Table 3). The most species-rich genera recorded was Kumanoa (four species), followed by Sirodotia (three species), and Thorea (two species). Audouinella, Balliopsis, Caloglossa, and Compsopogon has only one species. Thorea clavata Seto & Rathanasabapathy occurred most frequently in eight sites in both seasons, while most appeared rarely (only 1 site). During this study, maximum algal abundance was observed in April 2015 at the Klonggun site (K15), in which the dominant species was Caloglossa beccarii (Zanardini) De Toni sensu lato. The lowest algal abundance was found in Klongtow site (K7), in which Audouninella subtilis Jao occurred in both seasons. The highest species richness was at the Klongwangpai site (K6) with four species, while at 11 sites only one species was found. Among the 16 species sampled, T. clavata and C. beccarii sensu lato were dominant with greatest abundance and maximum density, respectively. In contrast, some species were found only in particular sites, for instance, A. subtilis occurred only at Klong tow (K7), Thorea siamensis was observed only at the Huaycheesuk (K5) and Kumanoa gibberosa occurred only at the Huaytrankaew (K16) (Table 3). Our results show the streams around Khao Luang National Park have a high diversity of red algae compared to studies of other regions of the world (Necchi et al. 1999, Eloranta et al. 2016). This area is well-preserved and has relatively rich red algal communities, when compared with some neighbouring countries like Philippines (Liao 2010), Malaysia and Indonesia (Johnston et al. 2014).

It is also important to point out that the alga we have identified tentatively as C. beccarii sensu lato as occurring in Thailand, well in the morphological characteristics of Caloglossa fluviatilis, a species described from Panama and also reported from India (Kamiya and West 2014). However, molecular characterization of the Thailand specimens are under study to confirm our tentative identification.

Table 3 shows some differences in diversity of study sites and seasons. Of the 22 sites surveyed, eight species were found in all seasons. The highest values (16) of species richness were observed in the period of dry and wet season of early rainy season SW monsoon (16 species), whereas the lowest numbers occurred in wet season of heavy rainy season (8 species). The highest diversity values revealed by Shannon-Wiener, Simpson and Pielou indices were observed in dry season (when the dominant species was C. beccarii sensu lato) (Table 3). In the dry season other common species were Compsopogon caeruleus, Audouinella subtilis, Kumanoa tabagatenensis, Sirodotia huillensis, S. suecica, Thorea clavata, T. siamensis, and Balliopsis prieurii, which were found in all seasons with variable abundances. The highest values of species richness and abundance of red algae observed in the dry season could be due to the higher sunlight irradiance and light penetration in water with lower turbidity increasing photosynthesis, as well as slow current velocity allowing increased biomass, as reported in temperate regions (Sheath 1984) and tropical regions (Necchi 1993). In addition, nutrient run off during the rainy season, especially in NE monsoon, might influence algal growth and abundance. Of all the species surveyed, Caloglosssa beccarii sensu lato, is of special interest, since it had the highest abundance in all seasons. This might be due to their small thallus size (about 2 cm high) coupled with production of new blades, adventitious branches from blade margins and firm attachment by rhizoids. At present in Thailand freshwater Caloglossa species identification is not clear for lack of critical comparative molecular and morphological characterizations. Two species of Caloglossa, C. beccarii (Sato and Akiyama 2002) and C. leprieurii (Montagne) G. Martens (Peerapornpisal et al. 2006, Naulcharoen et al. 2007), have been reported from southern Thailand, the latter species from a peat swamp in Krabi province. The former species has a wide distribution in freshwater streams from Southeast Asia being reported recently from Indonesia and Malaysia (Johnston et al. 2014). As already stated, following recent papers (Kamiya and West 2014, West et al. 2015, Kamiya et al. 2016) we tentatively identified, but based on only on morphological studies, the Thailand specimens as C. beccarii sensu lato, but they seems to belong to a different species (C. fluvaitilis).

Many environmental variables can influence the growth of freshwater red algae. The substrate characteristics have a direct impact on the distribution of these algae (Sheath 1984), whereas other environmental factors affect stream condition and freshwater algal seasonality. In this study, water quality and trophic status were not particularly different among study sites and seasons. In a period of high turbidity, especially during the North-East monsoon, the nutrient concentration increased, water transparency decreased and species richness was lower (8 species). In this aspect, it is similar to a study made in Brazil (Necchi 1993), where algal abundance decreased with a higher nutrient levels, solar radiation, temperature and turbidity.

The relationship of red algae and water quality parameters provided by CCA plot (Fig. 3), indicated that the abundance of Batrachospermum sp. (Bsp) was positively correlated with current velocity (Vel) but it was negatively correlated with TDS, turbidity, and alkalinity. Thus, this species is more abundant in higher water velocity and lower TDS, turbidity, and alkalinity. Caloglossa beccarii sensu lato was more abundant in higher conductivity and temperature and lower pH. Kumanoa tabagatenensis was more abundant in high ammonia concentration and slightly acidic condition. Kumanoa hirosei was more abundant in higher alkalinity, turbidity and TDS with lower DO and SRP. Although the dominant taxon, C. beccarii sensu lato occurred in three sites at all seasons under oligotrophic-mesotrophic status based on AARLPC score. The presence of rare species of Batrachospermum and Kumanoa at most sites could also be relate to oligotrophic condition. These findings corroborate previous studies in Thailand (Suphan and Peerapornpisal 2003, 2007, Suphan et al. 2004, Peerapornpisal et al. 2006). The present trophic water data can provide a good basis for water management in Khao Luang National Park. Also, this research provides more information on freshwater red algal ecology in tropical streams.