Introduction
There are 304 million lakes worldwide, covering approximately 402 million km2 of the continental terrestrial surface. Most of these ecosystems are geologically young; they are mainly of glacial, tectonic, volcanic, or fluvial origin (Hutchinson 1957; Downing et al., 2006). Lakes are not considered highly permanent features of the landscape, and it is estimated that they are destined to disappear due to the accumulation of sediments and organic matter (Thorp & Covich, 2001; Lasso, Gutiérrez & Morales, 2014). Despite most lakes being found in the northern hemisphere, there is a high diversity and variety of lake systems in South America, which are the result of the geomorphology and variety of climates present in the region (Aguilera et al., 2006; Likens, 2010; Benito et al., 2018).
Llames and Zagarese (2009) classified South American tropical lakes according to geomorphology and climatic characteristics. The present study was undertaken in two types of shallow (< 2.5 m): tropical mountain lakes and tropical lowland lakes. Tropical mountain lakes are characterized by being found at heights above 3 500-4 000 m.a.s.l.; they are oligotrophic lake environments with low ionic concentrations, low temperature, and high transparency; they are categorized as cold polymictic (Florez & Rios, 1998; Gutiérrez, Morales & Lasso, 2014; Catalan & Donato, 2016). The chemical conditions of these lakes vary according to their geomorphological origin (Vila & Mühlhauser, 1987; Kappelle & Horn, 2016). Tropical lowland lakes are characterized by being found between 0 and 1 000 m.a.s.l.; they are meso- to eutrophic environments; they present high temperatures and are categorized as warm polymictic, with high concentrations of humic substances and organic matter (Hutchinson & Löffler, 1956; Ramírez & Viña, 1998; Wilhelm & Adrian, 2008; Montoya & Aguirre, 2009).
The aquatic communities present in lakes respond to constant changes in physical and chemical variables (Webb et al., 2002; Li et al., 2019). However, the plankton community responds more quickly than others to these changes and has a direct effect on the productivity of lentic environments because it supports to a large extent the trophic web, affecting the ecosystem structure and functioning (Anderson et al., 2002; Woodward et al., 2010). Phytoplankton is the main energy receptor in aquatic ecosystems; its composition and dynamics in the water column are regulated mainly by the incidence of light, temperature, nutrient availability, competition, hydraulic dynamics, and pression from herbivory (Reynolds, 1984, Reynolds, 2008; Kruk et al., 2011; Esteves & Suzuki, 2011; Muñoz-López et al., 2017). Zooplankton not only regulates primary productivity, can also indicate the environmental state of aquatic systems over a given time period, due to the changes in the community in terms of diversity and abundance (Phan et al., 2015). Zooplankton also plays a fundamental role in the cycle of biogeochemical elements (Allen et al., 2005), increasing CO2 inputs through grazing, and mediating effects on primary productivity (Urabe et al., 2003; Li et al., 2019).
Since tropical mountain and lowland lakes can have different water physicochemical characteristics as well as different plankton communities inhabiting them, it is important to characterize and compare the organisms found in these environments (Barón et al., 2002; Gunkel, 2003; López-Martínez et al., 2017). The present study adds new records of previously described species and to contribute to the knowledge of the planktonic communities present in tropical mountain and lowland lakes of the northeast Colombian Andes.
Some authors proposed that the difference between the community composition in tropical mountain lakes and lowland lakes is due to the biogeographic barrier of the Andes mountains (Coesel, 1992). Morever, from the high mountains to lowlands, a diversity of conditions associated with continuum of ecosystems impose a variety of dispersal pathways in aquatic organisms (Moritz et al., 2013). We hypothesized that regional and local variables influence the plankton richness and occurrence of species in the tropical lakes from the Northeast Colombian Andes. We predicted that species richness of the tropical lowland lakes is greater than tropical mountain lakes.
Materials and methods
Study area: Samples were collected at nine tropical lakes located in the northeastern part of Colombia in the Department of Santander (Fig. 1). The studied systems belong to two types of tropical lakes (mountain and lowland). Five tropical mountain lakes, located within the jurisdiction of the Páramo del Almorzadero, and four tropical lowland lakes located in the middle valley of the Magdalena River Basin were sampled (Junk, 1993; García & Dister, 1990; Morales-Rivas et al., 2007).
The Páramo del Almorzadero belongs to the biome known as Paramo. It belongs to a Tropical Montane Grasslands and Shrublands system; the rain and evaporation pattern is monomodal and reaches 1 379 mm in annual precipitation (Holdridge, 2000; Rivera, 2001; Olson et al., 2001, Hoekstra et al., 2005). The depth of mountain lakes is determined by the topography of the terrain and the retention capacity created by the moss mats and plants found in the system (Buytaert et al., 2006; Morales-Rivas et al., 2007; Aguilera, Lazzaro & Coronel, 2013).
The middle valley of the Magdalena River Basin is surrounded by tropical dry forest and wet tropical forest. It has bimodal precipitation and reaches 3 000 mm in annual precipitation (García & Dister, 1990; Holdridge, 2000; Gutiérrez et al., 2014). The depth of lowland tropical inland lakes is modulated by precipitation patterns (Arias, 1985; Montoya & Aguirre, 2010; Montoya & Aguirre, 2013; Ricaurte et al., 2019).
Sampling and data collection: A total of 27 plankton samples were collected in nine tropical lakes during July and August 2018. In each lake, three samples were collected. The following physical variables were measured in situ at each station: temperature (ºC), dissolved oxygen (mg/L), pH, and electrical conductivity (µS/cm), using a Hach multiparameter probe (Conductivity: 0.5 µS/cm, temperature: ±0.3 °C and mV: 0.1 mV) and a Hanna probe (Dissolved oxygen: ±1.5% F.S., ±0.5°C). Water transparency was measured with a Secchi disk (Tyler, 1968) and maximum depth was measured with a HONDEX ps-7 depth sounder. One liter of water-sample was taken at each station and filtered using Whatman GF/C filters (0.7 μm) to estimate chlorophyll-a concentration ex situ by centrifugation and measuring absorbance with a spectrophotometer (Shimadzu UV-Vis 1 700 Series), following APHA Standard Methods 10200H (2005).
Phytoplankton samples were collected with horizontal tows using a standard 23 μm net. Samples were fixed and preserved in buffered 4% formaldehyde and Lugol, using 0.5 ml for 100 ml of sample (APHA, 2005).
Zooplankton samples were collected with horizontal tows using a standard 63 μm net and straining 80 L of water with a 41 μm net. Samples were fixed and preserved in buffered 4% formaldehyde (APHA, 2005).
Laboratory and stadistical analyses: Organisms were identified using an optical Zeiss microscope with an AxioCam camera. Images of the species were obtained using the software Zen Blue 2.5 lite (Carl Zeiss Microscopy GmbH). Identification was performed to the lowest possible taxonomic level using taxonomic guides by Huber-Pestalozzi (1955), Patrick & Reimer (1966), Komárek (1974), Ruttner-Kolisko (1974), Koste (1978), Barber & Haworth (1981), Sendacz & Kubo (1982), Parra et al., (1982a), Parra et al., (1982b), Parra et al., (1982c), Parra et al., (1983a), Parra et al., (1983b), Ettl (1983), Komárek (1983), Komárek & Foot (1983), Starmach (1983), Kudo (1985), Reid (1985), Tell (1986), Koste & Shield (1987), Shield & Koste (1992), Shield & Koste (1993), Nogrady et al., (1993), Nogrady et al., (1995), Segers (1995), Alonso (1996), Cox (1996), Comas (1996), Suárez et al., (1996), Elmoor-Loureiro (1997), Komárek et al., (1999), Gaviria (2000), Nogrady & Segers (2002), Fernando (2002), Komárek & Anagnostidis (2005), Elías et al., (2008), Haney (2013), González & Inostroza (2017) and Siemensma (2019).
One percent of phytoplankton was identified to variety level; 11.3 % of specimens were identified to species level and 87.3% of specimens were identified to genus level. A total of 70.7 % of zooplankton was identified to species level and 29.3 % was identified to genus level.
Analyzed specimens were deposited at the Natural History Museum of the Universidad Industrial de Santander, in the Hydrobiological collection. Biological records were structured following the Darwin Core standard (DwC) and uploaded through the Integrated Publishing Tool of the Colombian node of the Global Biodiversity Information Facility (GBIF), SiB Colombia (Colombia Biodiversity Information) (2019).
We used the taxonomic classification system of Guiry y Guiry (2019), WoRMS Editorial Board (2019) and Siemensma (2019). These primary occurrence localities were obtained by searching in two electronic sources GBIF (2019) and SiB Colombia (2019), and bibliographic references.
Analysis of variance (ANOVA) was used to compare physical and chemical variables among tropical mountain lakes and tropical lowland lakes. Analysis of variance performed using the PAST software 4.02 (Hammer et al., 2009). Spatial relationships between the species found were established with a Jaccard similarity analysis (presence-absence) of identified organisms in the nine tropical lakes. The Jaccard similarity matrix was checked for significant pruning of groups (group average method) via the SIMPROF routine (999 simulations; p = 0.05). Analysis of similarity (ANOSIM) was used to test if significant differences between the groups derived from the cluster analysis. Community analyses were performed with the statistical package Primer V7 (Clarke & Gorley, 2006).
Results
Physical and chemical variables: The nine studied lakes were located at between 92 and 3 976 m.a.s.l. (Table 1). The highest pH values were found in the tropical lowland lakes (mean = 6.5) and the lowest in the tropical mountain lakes (mean = 5.6). The greatest conductivity was found in the tropical lowland lakes (mean = 90.6 µS/cm) and the lowest in the tropical mountain lakes (mean = 5.1 µS/cm). Dissolved oxygen was higher in the tropical mountain lakes (mean = 8.5 mg/L) and the lowest in the tropical lowland lakes (mean = 2.1 mg/L). The highest temperatures were observed in the tropical lowland lakes (mean = 31.5 °C) and the lowest in the tropical mountain lakes (mean = 11.5 °C). The maximum depth was found in the tropical lowland lakes (mean = 1.6 m) and the lowest in the tropical mountain lakes (mean = 0.8 m). The highest transparency values were observed in the tropical mountain lakes (mean = 67 cm Secchi disk) and the lowest in the tropical lowland lakes (mean = 43.5 cm Secchi disk). The highest chlorophyll-a concentrations were found in in the tropical lowland lakes (10.1 ± 4.34 mg/m3; mean ± SE) and the lowest in the tropical mountain lakes (6.1 ± 3.04 mg/m3; mean ± SE). ANOVA showed significant differences in pH (F=23.4, p=0.003), conductivity (F=192, p<0.001), dissolved oxygen (F=58.2, p=0.003) and temperature (F=291.3, p<0.001) among tropical mountain lakes and tropical lowland lakes.
Coordinates (m.a.s.l.) | Altitude | pH (µS/cm) | C (mg/L) | DO (°C) | T (m) | Depth (cm) | Secchi Disk (mg/m3) | Chl-a | ||
Tropical mountain lakes | Laguna del Sol | 7° 2’ 29,3” N 72° 51’ 39,5” W | 3 907 | 4.5 | 11.1 | 6.4 | 15.5 | 0.24 | 24 | - |
Laguna Frailejones | 7° 3’ 32,0” N 72° 51’ 4,1” W | 3 976 | 6.1 | 2.2 | 10.2 | 10.9 | 0.20 | 20 | 6.0 | |
Laguna los Tutos | 7° 3’ 32,2” N 72° 52’ 16,1” W | 3 963 | 6.1 | 2.6 | 9.1 | 10.6 | 2.60 | 250 | 1.6 | |
Laguna Verde | 7° 3’ 26,8” N 72° 51’ 6,7” W | 3 916 | 5.7 | 5.4 | 6.5 | 8.2 | 0.31 | 31 | 6.4 | |
Pozo Verde | 7° 2’ 57,2” N 72° 52’ 15,9” W | 3 646 | 5.8 | 4.3 | 10.3 | 12.4 | 0.10 | 10 | 10.2 | |
Tropical lowland lakes | Ciénaga la Colorada | 6° 42’ 54,1’’ N 74° 8’ 11,0’’ W | 92 | 6.4 | 82.2 | 2.4 | 33.7 | 1.00 | 35 | 3.0 |
Ciénaga la Duda | 6° 40’ 32,4’’ N 74° 10’ 42,8’’ W | 92 | 6.5 | 79.3 | 1.5 | 29.7 | 1.30 | 49 | 11.0 | |
Ciénaga Río Viejo | 6° 34’ 21,0’’ N 74° 17’ 17,2’’ W | 96 | 6.4 | 94.5 | 2.3 | 30.6 | 2.50 | 43 | 11.6 | |
Ciénaga el Limonal | 6° 40’ 55,1’’ N 74° 10’ 43,0’’ W | 93 | 6.6 | 106.5 | 2.2 | 31.8 | 1.70 | 47 | 14.8 |
A total of 391 planktonic organisms were identified in the nine studied lakes within phytoplankton and zooplankton (Digital Appendix). Species richness among lakes ranged between 41-70 taxa, except Ciénaga Río Viejo which showed a between two and four times more diversity. Tropical lowland lakes showed 17.6% higher richness than tropical mountain lakes.
Phytoplankton: A total of 299 phytoplankton species were identified in the nine studied lakes (Fig. 2); these species were distributed in seven phyla, 13 classes, 31 orders, 57 families, and 96 genera. The phylum with greatest number of species was Charophyta (70), followed by Bacillariphyta (58), and Chlorophyta (58) (Digital Appendix).
A total of 120 phytoplankton species were identified in the five tropical mountain lakes; these species were distributed in six phyla, ten classes, 26 orders, 40 families, and 64 genera. The phylum with greatest number of species was Charophyta (49 species), followed by Bacillariophyta (23 species), Cyanobacteria (20 species), Chlorophyta (19 species), Ochrophyta (5 species), and Miozoa (4 species) (Digital Appendix).
A total of 186 phytoplankton species were identified in the four tropical lowland lakes; these species were distributed in six phyla, 13 classes, 26 orders, 45 families, and 67 genera. Among the species found, four were identified to variety level (Aulacoseira granulata var. angustissima, Trachelomonas armata var. armata, Trachelomonas armata var. litoralensis, and Trachelomonas similis var. spinosa) (Fig. 3). The phylum with greatest number of species was Euglenozoa (47 species), followed by Bacillariophyta (39 species), Chlorophyta (39 species), Charophyta (21 species), Cyanobacteria (19 species), Ochrophyta (17 species) and Miozoa (4 species) (Digital Appendix).
Of the phytoplankton identified in the tropical lakes in this study, 37.1 % of species were unique to tropical mountain lakes, 60.6 % were unique to tropical lowland lakes, and the remaining 2.3 % corresponded to species shared between mountain and lowland tropical lakes (Digital Appendix). The seven phytoplankton species shared between tropical mountain lakes and tropical lowland lakes were Ankistrodesmus fusiformis, Eunotia sp6, Eunotia sp7, Pinnularia cf. major, Pinnularia sp13, Planktolyngbya sp1, and Pseudanabaena sp2.
A total of 60 new phytoplankton species records for the Department of Santander; 19 corresponded to species identified in tropical mountain lakes and 35 corresponded to species identified in tropical lowland lakes (Fig. 4). 15 new records were identified for Colombia (Digital Appendix).
Zooplankton: A total of 92 zooplankton species were identified in the nine studied lakes (Fig. 5); these species were distributed in four phyla, six classes, eight orders, 27 families, and 40 genera. The phyla with greatest number of species were Rotifera (58 species), Arthropoda (20 species), and Amoebozoa (12 species) (Digital Appendix).
A total of 53 zooplankton species were identified in the five tropical mountain lakes; these species were distributed in four phyla, five classes, five orders, 18 families, and 27 genera. The phylum with greatest number of species was Rotifera (58 species), followed by Arthropoda (18 species), Amoebozoa (12 species), and Cercozoa (4 species) (Supporting Information).
A total of 47 zooplankton species were identified in the four tropical lowland lakes; these species were distributed in four phyla, six classes, eight orders, 21 families, and 24 genera. The phylum with greatest number of species in these lakes was Rotifera (32 species), followed by Arthropoda (9 species), Amoebozoa (5 species), and Cercozoa (1 species) (Supporting Information).
Of the zooplankton identified in tropical lakes, 47.9 % of species were unique to tropical mountain lakes, 41.7 % were unique to tropical lowland lakes, and the remaining 10.4 % were shared between tropical mountain and lowland lakes (Digital Appendix). The ten species shared between tropical mountain and lowland lakes were Arcella conica, Arcella sp1, Ascomorpha ecaudis, Asplanchna seiboldii, Brachionus falcatus, Calanoida sp1, Centropyxis sp1, Lecane bulla, and Trichocerca similis (Fig. 6). A total of 41 new records of zooplankton species were found for the Department of Santander; 12 corresponded to species identified in tropical mountain lakes and 22 corresponded to species identified in tropical lowland lakes (Fig. 7).
A total of 27 plankton samples were entered into the Hydrobiological collection of the Natural History Museum UIS. The entered records correspond to UIS-MHB-2299, UIS-MHB-2302, UIS-MHB-2704, UIS-MHB-2166, UIS-MHB-2187, UIS-MHB-2194, UIS-MHB-2195, UIS-MHB-2198, UIS-MHB-2199, UIS-MHB-2201, UIS-MHB-2206, UIS-MHB-2209, UIS-MHB-2223, UIS-MHB-2240, UIS-MHB-2243, UIS-MHB-2251, UIS-MHB-2254, UIS-MHB-2255, UIS-MHB-2257, UIS-MHB-2280, UIS-MHB-2281, UIS-MHB-2287, UIS-MHB-2291, UIS-MHB-2292, UIS-MHB-2296, UIS-MHB-2299, UIS-MHB-2302 y UIS-MHB-2704. The complete list of species is available http://datos.biodiversidad.co
The Jaccard similarity analysis showed differences in the plankton communities found in the tropical mountain and lowland lakes; the cophenetic correlation was 0.87. The five tropical mountain lakes were grouped in the analysis; within this group there was similarity between Laguna del Sol and Laguna Los Tutos, and between Laguna Frailejones and Pozo Verde, whereas Laguna Verde was the least similar. The four tropical lowland lakes were grouped; within this group there was similarity among Ciénaga La Colorada, Ciénaga El Limonal, and Ciénaga Río Viejo, whereas the least similar lake in this group was Ciénaga La Duda (Fig. 8). ANOSIM showed significant differences between the groups derived from the cluster analysis (Global R = 0.89, p = 0.001).
Discussion
The present study contributes to a better limnological understanding of mountain and lowland tropical lakes located in the northeast Colombian Andes. We present the data corresponding to six physicochemical variables and a list of 391 plankton species, and expand the available information on the planktonic organisms of tropical lakes. Fifteen taxa of planktonic organisms are recorded for the first time in the mountainous and lowland areas of the Northeast Colombian Andes.
Physical and chemical variables: The physicochemical variables of the five tropical mountain lakes studied here presented similar characteristics to those reported by Zapata-Anzola et al., (2006) and Roldán & Ramírez (2008). The low conductivity and pH values tending towards acidity observed in this study were probably due to geomorphology and basin substrate, because the lakes located in the northeast Colombian Andes are of glacial origin and present crystalline rocks of slow chemical weathering (Flórez, 2003; Buytaert et al. 2006; Bare & Ashton, 2015; Borsdorf & Stadel, 2015; Catalan & Donato, 2016). The high dissolved oxygen levels reported for these tropical lakes are probably due to location, as they are found in areas of great wind exposure (Donato 1991; Rivera-R et al., 2005; Hernández-Atilano et al., 2012; Herrera-Martínez et al., 2017).
The pH and conductivity values recorded for the tropical lowland lakes in the study area were similar to those reported for other lowland lakes of the Middle Magdalena River, such as Ciénaga de Paredes, Ciénaga de Chucuri, Ciénaga de Aguas Negras, Ciénaga San Silvestre, Ciénaga del Opón, and Ciénaga Miramar (Pedraza et al., 1989; Ramírez & Viña, 1998; Pava et al., 2006; Barón-Rodríguez et al., 2006, Roldán & Ramírez, 2008; García & Dister, 1990; Criales-Hernández & Jerez-Guerrero, 2016; Solís-Parra & Criales-Hernández, 2016). Conductivity values indicated high ionic and dissolved solids content, probably associated with river runoff and anthropogenic activities from agricultural and livestock farming activities in the area (Kalff, 2002; Peña et al., 2005; Gallo-Sánchez et al., 2009; Pinilla et al., 2010; Barón-Rodríguez et al., 2006; Montoya & Aguirre, 2013; Silva et al., 2017).
Plankton analysis: The planktonic organisms identified in this study coincide with information recorded for other tropical mountain (Gaviria, 1989; Donato, 1991; Gaviria, 1993; Donato et al., 1996; Donato, 2010; Aranguren-Riaño et al., 2011; Muñoz-López et al., 2017) and lowland lakes (Duque & Núñez-Avellaneda, 2000; Pinilla, 2005; Barón-Rodríguez et al., 2006; Sala et al., 2008; Álvarez, 2010; Jaramillo-Londoño & Pinto-Coelho, 2010; Rivera et al., 2005; Andrade-Sossa et al., 2011; Jaramillo-Londoño & Aguirre-Ramírez, 2012; Montoya-Moreno et al., 2012; Aguirre-Ramírez, 2014; Sala et al., 2015) of Colombia.
The greatest phytoplankton richness was found for the phyla Charophyta, Bacillariophyta, and Chlorophyta; this was possibly due to these species constituting the greatest percentage of species in lentic systems, despite differences in the environmental characteristics of mountain and lowland lakes (Donato et al., 1996; Donato, 2010; Alba-M et al., 2011; Montoya-Moreno et al., 2013; Osorio-Ávila & Manjarres-García, 2015).
The species richness of phytoplankton in this work is high as compared with the results of other studies in tropical lakes. For example, Castaño et al., (2010) observed 25 phytoplankton species in Laguna Negra, Muñoz-López et al., (2017) reported 49 phytoplankton taxa in Lago de Tota, Jaramillo-Londoño & Aguirre-Ramirez (2012) recorded 74 phytoplankton species in the Ciénaga de Ayapel, Pava et al., (2006) reported 58 in the Ciénaga de San Silvestre. However, Donato, (2010) found a higher species richness than our studie, they identified 142 phytoplankton species in 20 tropical mountain lakes distributed along the three Andean Cordilleras.
The richness of zooplankton species belonging to the Rotifera phylum was probably due to the amount of food available through detritus pathways and the availability of particulate organic matter, small algae, and decomposing organisms such as bacteria and fungi (Nogrady et al., 1993; Pinilla et al., 2007; Andrade-Sossa et al., 2011; Burian et al., 2016). All the tropical lakes studied in this work had associated aquatic vegetation. The high richness of microcrustaceans was probably due to the presence of macrophytes, which act as both shelter against predation and food source for cladocers and copepods (Cazzanelli et al., 2008; Villabona-González et al., 2011; Choedchim et al., 2017).
The proportion of planktonic species unique to tropical lowland lakes (51.2 %) and tropical mountain lakes (42.5 %), as well as the low similarity found in the Jaccard analysis, indicated high heterogeneity in the ecological conditions of the studied tropical lakes. The remaining 6.4 % of specimens corresponded to 17 planktonic species shared between the two types of tropical lake. These shared species were found in lakes with high concentrations of organic matter (Donato, 1987; Tuchman 1996; Klug, 2002; Tuchman et al., 2006; Andrade-Sossa et al., 2011; Brighenti et al., 2018; Mello et al., 2018).
After reviewing databases of biological collections and published studies on plankton species present in tropical lakes of Colombia, the present study reported 15 new records for the country. This high number of new records of planktonic species is probably due to a lack of published descriptions. The present study fills a gap in the knowledge of planktonic communities of the tropical lakes from the Northeast Colombian Andes.
Ethical statement: authors declare that they all agree with this publication and made significant contributions; that there is no conflict of interest of any kind; and that we followed all pertinent ethical and legal procedures and requirements. All financial sources are fully and clearly stated in the acknowledgements section. A signed document has been filed in the journal archives.