International Journal of Salt Lake Research 8: 113–126, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands.

Benthos of a perennially-astatic, saline, soda lake in Mexico JAVIER ALCOCER1∗, ELVA G. ESCOBAR2, ALFONSO LUGO1 and LUIS A. OSEGUERA1 1 Limnology Lab., Environmental Conservation and Improvement Project, UIICSE, UNAM

Campus Iztacala. Av. de los Barrios s/n, Los Reyes Iztacala, 54090 Tlalnepantla, Estado de México. México; 2 Benthic Ecology Lab., Institute of Marine Sciences and Limnology, UNAM. Apdo. Postal 70-305, 04510 Coyoacán, México, D.F. México (∗ author for correspondence, e-mail: [email protected]; fax: +52 5277 1829)

Abstract. The effects of multiple stressors on the benthic macroinvertebrate community were monitored in Tecuitlapa Norte, a shallow, perennially-astatic, warm, mesosaline, sodaalkaline lake in Mexico. Physico-chemical and biological variables were determined monthly for one year. Tecuitlapa Norte displayed a clear seasonal environmental pattern (dry and rainy seasons). The benthic macroinvertebrate community consisted of five species: Culicoides occidentalis sonorensis Jorgensen, Ephydra hians Say, Stratiomys sp., Eristalis sp., and Limnophora sp. of which the first two were dominant and the rest scarce. C. occidentalis was the most important species numerically (76 percent of the total), while E. hians dominated the biomass (73 percent of the total). Primarily salinity and secondarily pH appear to be the most important environmental factors controlling dominance of benthic organisms in Tecuitlapa Norte. Seasonal abundance dynamics of the dominant organisms was associated with phases in their reproductive cycles: environmentally-triggered (i.e., temperature rise, water-level descent) pupation and emergence periods. We concluded that whereas physical and chemical factors (i.e., salinity, pH) exerted the primary control on benthic macroinvertebrate community composition in Tecuitlapa Norte, another assembly of variables (e.g., water-level, temperature) influenced species distribution and abundance. Key words: benthic macroinvertebrates, crater-lake, physico-chemistry, shallow lake, tropical lake

Introduction Inland waters, which contain a concentration of salts above 3 g.L−1 (W.D. Williams, 1996) are classed as saline. They are of two main types, those dominated by sodium chloride, and alkaline lakes dominated by sodium carbonate and/or bicarbonate. Such lakes can also be defined according to whether or not, and how frequently they dry up. Hartland-Rowe (1972, quoted by Cole, 1979) defined two categories of freshwater lakes, which he called

114 seasonally and perennially astatic lakes. The former dry up annually whereas, in the latter, water-levels rise and fall but the lake does not dry every year. Astatic freshwater lakes and ponds have received some attention (Mackay, 1996; D.D. Williams, 1987, 1996), although most information about them concerns species of particular taxonomic groups or is restricted to relatively local areas (Wiggins et al., 1980). Additionally, information is scarce on astatic saline, soda-alkali lakes. Such lakes are considered among the most stressed of aquatic ecosystems, since their waters combine both high salinity and high pH (Cole, 1979). In shallow soda-alkaline lakes, as in other saline lakes (Hammer, 1986), wide seasonal and inter-annual water-level fluctuations represent an additional stress factor because changes in water-level on an annual cycle strongly affect the available resources for the development of lacustrine communities (Payne, 1986). As suggested by D.D. Williams (1996), the aquatic biota of such waters is strongly influenced by two main groups of constraints, physico-chemical and biological factors, among which the former are believed to have primacy in controlling community composition (Colburn, 1988). This paper discusses the hypothesis that although salinity has a direct or indirect influence on saline lake communities, salinity per se does not explain species distribution, composition, richness or their spatio-temporal fluctuations in saline lakes. We tested this hypothesis by investigating the benthic macroinvertebrate community of Tecuitlapa Norte, a perennially astatic, saline, soda lake in Mexico. Tecuitlapa Norte displays two extreme and very important stressing parameters: salinity (high and fluctuating) and pH (high and stable). Study site Tecuitlapa Norte is one of three water-bodies in the Pleistocene crater of an extinct strato-volcano in the basin of Oriental (19◦ 080 –19◦ 300 N, 97◦ 200 – 97◦ 510 W; 2,300 m.a.s.l.; 4,982 km2 ) (Figure 1). This endorheic basin is on the extreme south-eastern portion of the Mexican Plateau at the conjunction of the states of Puebla, Tlaxcala, and Veracruz. The area is characterised by a subhumid, temperate climate: mean annual temperature and precipitation are 13 ◦ C and 706 mm, respectively. The warm-rainy season extends from May to November, and the cold-dry season from December to April (García, 1988). The lake is shallow (maximum depth ≈ 0.5 m), and small (200×30 m when fully inundated). Tecuitlapa Norte waters have been characterized as saline (K25 mean = 17.8, s.d. = 3 mS.cm), soda type (Na2 CO3 ), and alkaline

115

Figure 1. Geographic location of Tecuitlapa Norte, Puebla, Mexico.

(pH mean = 10.7, s.d. = 0.07) (Vilaclara et al., 1993). Aquatic macrophytes are absent, but a perennial Spirulina bloom turns the water color to a ‘pea soup green’. The Spirulina, with a mean chlorophyll a concentration of about 1,500 mg.m−3 (Vilaclara et al., 1993), continuously generates large quantities of living and dead organic matter (BOD5 = 24–128 mg.L−1 , COD = 1,660– 2,991 mg.L−1 , Garzón, 1990, unpublished) thus providing an abundant food

116 source for herbivores and detritus feeders. The sediment consists of silty sand with a high organic matter content (8 percent). There are no fish or other large aquatic predators.

Methods A monthly sampling program was established from December 1993 to December 1994. Midday measurements of temperature, pH, conductivity (K25 ), dissolved oxygen, and redox potential were measured with a calibrated Hydrolab DS3/SVR3 (multiparameter water quality monitoring and data-logger instrument). On each sampling date, faunal samples were obtained from twenty sediment cores (the top 15 cm of the sediment; 6 cm in diameter) randomly located, covering the whole lake area. These samples were sieved in the laboratory through a 0.25 mm mesh screen. The fauna was sorted, preserved in 70 percent ethanol, identified, counted, weighed wet (roughly equivalent to biomass), and transformed to gC.m−2 (10 percent of wet-weight) following Margalef (1983). Multivariate analysis (Cluster, and Principal Components Analysis – PCA) was used to examine the seasonal variation of both the environmental and biological factors in Tecuitlapa Norte, as well as to identify the most important stressors affecting the benthic macroinvertebrate community. Transformed environmental (log n, except pH) and biological (log n + 1, where n is abundance in org.m−2 or biomass in gC.m−2 ) data were used in the statistical analysis.

Results In spite of a large concentration of organic matter, midday dissolved oxygen concentrations in Tecuitlapa Norte were high (about saturation to supersaturation) reflecting its shallowness, mixing by wind, and high rate of primary productivity. Salinity, dissolved oxygen and Eh showed wide fluctuations during the year (Table 1) with a clear seasonal pattern revealed through multivariate analysis (Figure 2). There are two well-defined seasons: a drycold season from December-93 to March, and a wet-rainy-warm season from June to October. Between these seasons, there were transitional periods, in April–May and in November–December 1994. During the dry period (A), the concentration phase of the lake was characterized by increasing salinity, high pH values, and lower temperature, dissolved oxygen and redox potential (Table 1). The dilution phase occurred during the wet period (C) and was

117 Table 1. Environmental characteristics of Tecuitlapa Norte. Parameter

Dry season (min–max)

Rainy season (min–max)

PC 1 (75.1%) (factor loading)

PC 2 (22.6%) (factor loading)

Temperature (◦ C) pH K25 (mS.cm) D.O. (mg.L−1 ) D.O. (% Sat.) Eh (mV)

21.9–24.9 11.1–11.6 43.9–55.7 5.5–10.0 98.8–156.3 135–176

23.5–25.3 10.4–10.5 21.2–28.4 9.8–>15 177–>200 218–234

−0.034 0.970 0.073 −0.196 −0.194 −0.122

0.028 −0.124 0.966 −0.219 −0.220 −0.060

K25 = conductivity standarized at 25 ◦ C, D.O. = dissolved oxygen, Eh = redox potential, PC = principal component.

characterized by decreasing salinity and comparatively lower pH values, and higher dissolved oxygen and redox potential (Table 1). Intermediate values occurred during the transitional periods (B). The PCA of the physicochemical factors indicates that salinity (PC 1) and pH (PC 2) were the most important parameters in explaining the environmental variance in Tecuitlapa Norte. Only five species of benthic macroinvertebrates were found in Tecuitlapa Norte, all dipterans (Insecta, Diptera). In order of dominance they were Culicoides occidentalis sonorensis Jørgensen (Ceratopogonidae), Ephydra hians Say (Ephydridae), Stratiomys sp. (Stratiomyidae), Eristalis sp. (Syrphidae), and Limnophora sp. (Muscidae). Note that there are no saline, non-alkaline water species in Tecuitlapa Norte. Species richness therefore ranged from one to five species. E. hians and C. occidentalis were present throughout the year, while the other three species appeared only occasionally in samples. The low species richness of Tecuitlapa Norte is similar to other saline lakes worldwide within a similar salinity range (Alcocer et al., 1997; Colburn, 1988; Dejoux, 1993; Herbst, 1988; Timms, 1982, 1983; Williams and Kokkinn, 1988). The abundance and biomass of C. occidentalis (75.99 and 26.89 percent of the total, respectively) and E. hians (23.96 and 72.88 percent, respectively) overshadowed that of the other three species (<0.05 %). Mean annual benthic macroinvertebrate density and biomass (mean ± s.d.) were 84,552±61,939 org.m−2 (n = 260) and 12.2±15.7 gC.m−2 (n = 260), respectively. Higher densities were detected in March (183,350±9,168 org.m−2 , n = 20) and April (182,219±9,111 org.m−2 , n = 20), while the lowest were in May (41,473±4,977 org.m−2 , n = 20) and September (43,420±5,210 org.m−2 , n = 20) (Figure 3). On the other hand, higher biomass values were detected in December 1993 (52.6±3.68 gCm−2 , n = 20), while the lowest was in November (0.5±0.05 gC.m−2 , n = 20) (Figure 3). Clearly, C. occidentalis

118

Figure 2. Environmental dissimilarity dendrogram (1 − Pearson r, Dlink/Dmax∗100) (top) and PCA (bottom) of Tecuitlapa Norte. (A = dry season, B = transitional period, C = rainy season). (PC 1 = K25 , PC 2 = pH).

is, numerically, the most important species in Tecuitlapa Norte, although E. hians dominates the biomass because of its larger body size. Their temporal patterns of abundance and biomass differed, depending on the variation in body-size of the dominant species. Body-size of C. occidentalis was more or less constant, while that of E. hians changed markedly with different life stages.

119

Figure 3. Abundance (org.m−2 ) (top) and biomass (gC.m−2 ) (bottom) of Ephydra hians and Culicoides occidentalis in Tecuitlapa Norte from December 1993 to December 1994.

120 With regard to species composition and abundance, the cluster analysis formed four clusters (Figure 4). Three represent only a single month each (May, October, November), and the fourth clustered the remaining ten months with a high degree of similarity (1 percent dissimilarity). November showed the most striking differences, being over 90 percent dissimilar to the rest of the year. The PCA showed species richness (PC 1) to be the most important parameter in grouping the months, although abundance differences (PC 2) also influenced the cluster formation. The same two species (E. hians and C. occidentalis) were found throughout cluster 1, although abundance in subcluster 1A (December-93 to April) was lower than in subcluster 1B (June– September, December-94). In May (cluster 2), just one species – C. culicoides – was found, while in October (cluster 3) there were three species. The relative similarity (50 percent) between these two clusters (2 and 3) was associated with extremely low abundance (almost absence) of E. hians and Stratiomys in cluster 3. November (cluster 4) was the only month in which all five species were recorded. To avoid a biassed classification derived from the striking difference between cluster 1 and the other three clusters (2–4), a second biological cluster analysis were carried out from which May, October and November were excluded. Although with rather small differences, two main clusters were formed (Figure 5). Cluster I grouped months related to the dry season (cluster A, Figure 2), meanwhile cluster II joined months related to the rainy and transitional periods (clusters B and C, Figure 2). Although a linear correlation analysis (Pearson coefficient on transformed data) showed no significant relationships between any environmental (temperature, pH, K25, O.D., %Sat, Eh) and biological (abundance and biomass) variables (p > 0.1), the last cluster analysis (Figure 5) does suggest that fluctuations in abundance do follow a similar seasonal pattern to the environmental factors (Figure 2). Changes in abundance were related to key moments in the reproductive cycles of the dominant species. Two massive emergence periods of C. occidentalis were observed. The first was in May, at the end of the dry season, when it appeared to be triggered by an increase in temperature. Many C. occidentalis pupae were found in April, explaining the large decrease in numbers of C. occidentalis in the benthos in May. Linley et al. (1970) reported a similar temperature-triggered emergence of C. furens. The second emergence period, at the beginning of the dry season, was probably triggered by a decrease in water-level during September–October, explaining a major reduction of C. occidentalis in November. Linley and Adams (1972) reported an emergence of C. mellus triggered by declining water-level because pupation cannot occur out of water.

121

Figure 4. Biological dissimilarity dendrogram (1 − Pearson r, Dlink/Dmax∗100) (top) and PCA (bottom) of the benthic fauna in Tecuitlapa Norte. [1 = two species present (1A = low abundance/biomass, 1B = high abundance/biomass), 2 = one species present, 3 = three species present, 4 = five species present]. (PC 1 = species richness, PC 2 = abundance).

The marked change in abundance of Ephydra hians was also related to an emergence period. According to Herbst (1990), E. hians starts pupation at 20 ◦ C, and the higher the temperature, the shorter the pupation time. Shallow saline lakes may allow E. hians to achieve up to three generations per year, according to local temperature variations (Herbst, 1990). The temperature

122

Figure 5. Biological dissimilarity dendrogram (1 − Pearson r, Dlink/Dmax∗100) of the benthic fauna in Tecuitlapa Norte, excluding May, October and November. (I = months associated to the dry season, II = months associated to the rainy season and transitional periods).

range (19–36.4 ◦ C) of Tecuitlapa Norte could favor continuous and shortterm formation of pupae, but in spite of this, only one emergence period of E. hians was detected. A trend of increasing temperature from December onwards explained the large quantity of pupae found in April and the absence of this species in May. After May, the abundance of E. hians was rather low for the rest of the year. Probably the constantly high temperature (above 20 ◦ C) favors continuous emergence of this species; this could explain why E. hians adults were seen flying near Tecuitlapa Norte throughout the year. In November, the lowest abundances which followed the massive emergence of C. occidentalis permitted the appearance of Limnophora, Stratiomys and Eristalis. Nonetheless, it is quite clear from their low abundances that these three species are in an environmentally disadvantageous situation. Although Muscidae, Stratiomyidae, and Syrphidae are typically freshwater families, their presence in Tecuitlapa Norte is probably related to the huge quantity of organic matter available as a food supply (McCafferty, 1981). Much of the fauna of temporary, and permanent, ponds displays characteristic successional patterns of occurrence (Bayly and Williams, 1973). In Tecuitlapa Norte, E. hians dominated during the first two months (December-

123 93 and January-94), but was rapidly replaced by C. occidentalis for the rest of the year. Discussion To colonize Tecuitlapa Norte successfully, any species must first be able to tolerate high salinity and pH. Once this tolerance is achieved, another assemblage of variables (environmental and, probably, but not tested in this study, biological) explain its temporal dynamics. On what is this assumption based? To explore this hypothesis, we compare our findings in Tecuitlapa Norte not only with Tecuitlapa crater-lake, but also with Totolcingo (El Carmen) lake. Tecuitlapa is an alkaline (pH = 9.8±0.1) and eutrophic (chlorophyll a = 200 mg.m−3 ) freshwater (K25 = 1.65±0.05 mS.cm) crater-lake, located 100 m from Tecuitlapa Norte (Figure 1) (Vilaclara et al., 1993). Totolcingo lake is a saline (K25 = 10.6–30.4 mS.cm), alkaline (pH = 9.37–10.14), and eutrophic episodically-filled lake, located in the same drainage basin (Oriental). The three lakes have a similar alkaline carbonate (soda) ionic composition (Alcocer et al., 1997; Vilaclara et al., 1993). Twenty-nine species of benthic macroinvertebrates have been recorded in Tecuitlapa crater-lake (Alcocer, 1995) and of these, just one (Stratiomys) lives in Tecuitlapa Norte as well as in Tecuitlapa crater-lake, where it is scarce. Thus, there must be limitations impeding the colonization of more species from Tecuitlapa crater-lake into Tecuitlapa Norte. We consider salinity is of primary importance in this respect since Tecuitlapa crater-lake has similar environmental characteristics to Tecuitlapa Norte but is fresh not saline. On the other hand, the two dominant species of Tecuitlapa Norte (C. occidentalis and E. hians) are characteristic of saline, alkaline carbonate waters (alkali-adapted), and Tecuitlapa crater-lake has alkaline water. Why are C. occidentalis and E. hians not present in Tecuitlapa crater-lake? It is quite probable that again salinity is the answer, but indirectly, not salinity per se. A freshwater lake allows other species to establish that compete and/or predate on C. occidentalis and E. hians. Ephydrids, for example, are successful and abundant only in lakes lacking predators (Herbst, 1988, 1990). This explains why saline, alkaline lakes, in which predators and competitors are lacking, or greatly diminished due to salinity stress, offer an excellent habitat for ephydrids to live in and multiply in large numbers. This is supported by the flourishing population of E. hians in Totolcingo lake (Alcocer et al., 1997), where predators are also absent. Although salinity and pH seem to have limited species composition and richness in Tecuitlapa Norte, temporal fluctuations in abundance and biomass

124 were more associated with other environmental variables (high temperature and water-levels) which trigger key events in the life-cycles of the species (i.e., pupation and emergence). Our findings support the hypothesis proposed by Colburn (1988), Herbst (1988), and Williams et al. (1990) that physical and chemical factors may exert primary control on community composition in saline waters.

Conclusions Tecuitlapa Norte is a shallow, warm, mesosaline, soda-alkaline lake. It is a perennially astatic, and highly productive (Spirulina bloom) lake. Diverse stressors are represented in Tecuitlapa Norte: high and fluctuating salinity, variable water-level, and high alkaline pH, temperature, and organic matter content. Tecuitlapa Norte displayed a clear seasonal environmental pattern: dry season (December 1993 to March), and rainy season (June–October). Transitional periods were also detected, April–May (dry to rainy seasons), and November–December 1994 (rainy to dry seasons). The benthic macroinvertebrate community was composed of five species: Culicoides occidentalis sonorensis Jørgensen (Ceratopogonidae), Ephydra hians Say (Ephydridae), Stratiomys sp. (Stratiomyidae), Eristalis sp. (Syrphidae), and Limnophora sp. (Muscidae). C. occidentalis and E. hians were dominant, the other three species were scarce. C. occidentalis was the most important species numerically (≈76%), while E. hians dominates the biomass (≈73%). Salinity, primarily, and pH secondarily, seem to be the most important environmental constraints in selecting the dominant species inhabiting Tecuitlapa Norte. The saline, alkaline water clearly favored C. occidentalis and E. hians. The changes in seasonal abundance of the dominant organisms were more associated with environmentally-triggered (i.e., temperature rise, water-level descent) key events in their reproductive cycles (i.e., pupation and emergence periods).

Acknowledgments Many people have given time to preparation of this paper; all are gratefully acknowledged, particularly María del Rosario Sánchez, Mario M. Chávez, Laura Peralta, and María de Jesús Montoya for field and laboratory work. Financial support was partially given by CONACYT project T-25430, and

125 DGAPA project IN204597. The authors specially thank Dr David B. Herbst (Ephydridae), and Dr Donald W. Webb (Ceratopogonidae) for taxonomic assistance. This paper was greatly improved by including the comments of Dr Máximo Florín (Utrecht University), Dr S.S.S. Sarma (UNAM Campus Iztacala), Dr Nandini Sarma (UNAM Campus Iztacala), and Dr Mary Burgis (City of London Polytechnic); we are grateful to them all. References Alcocer, J. 1995. Análisis holístico de la comunidad de macroinvertebrados bentónicos litorales de seis lagos-cráter con un gradiente de salinidad. PhD. thesis dissertation. Facultad de Ciencias. Universidad Nacional Autónoma de México. México. 106 pp. Alcocer, J., Lugo, A., Escobar, E. and Sánchez, M. 1997. The macrobenthic fauna of a former perennial and now episodically filled Mexican saline lake. International Journal of Salt Lake Research 5: 1–14. Bayly, I.A.E. and Williams, W.D. 1973. Inland Waters and Their Ecology. Longman, Melbourne. 314 pp. Colburn, E.A. 1988. Factors influencing species diversity in saline waters of Death Valley, USA. Hydrobiologia 158: 215–226. Cole, G.A. 1979. Textbook of Limnology. C.V. Mosby, St. Louis. 426 pp. Dejoux, C. 1993. Benthic macroinvertebrates of some saline lakes of the Sud Lipez region, Bolivia. Hydrobiologia 267: 257–267. García, E. 1988. Modificaciones al sistema de clasificación climática de Köppen. E. García, México. 217 pp. Hartland-Rowe, R. 1972. The limnology of temporary waters and the ecology of Euphyllopoda. In: R.B. Clark and R.J. Wootton (eds.). Essays in Hydrobiology, pp. 15–31. University of Exeter Press, Exeter. Hammer, U.T. 1986. Saline Lake Ecosystems of the World. Junk, Dordrecht. 616 pp. Herbst, D.B. 1988. Comparative population ecology of Ephydra hians Say (Diptera: Ephydridae) at Mono Lake (California) and Abert Lake (Oregon). Hydrobiologia 158: 145–166. Herbst, D.B. 1990. Distribution and abundance of the alkali fly (Ephydra hians) Say at Mono Lake, California (USA) in relation to physical habitat. Hydrobiologia 197: 193–205. Linley, J.R. and Adams, G.M. 1972. Ecology and behaviour of immature Culicoides melleus (Coq.) (Dipt., Ceratopogonidae). Bulletin of Entomological Research 62: 113–127. Linley, J.R., Evans, F.D.S. and Evans, H.T. 1970. Seasonal emergence of Culicoides furens (Diptera: Ceratopogonidae) at Vero Beach, Florida. Annals of the Entomological Society of America 5: 1332–1339. Mackay, R.J. 1996. Temporary aquatic habitats. Journal of the North American Benthological Society 4: 407. Margalef, R. 1983. Limnología. Omega, Barcelona. 1,010 pp. McCafferty, W.P. 1981. Aquatic Entomology. Science Books International, Boston. 448 pp. Payne, A.I. 1986. The Ecology of Tropical Lakes and Rivers. John Wiley & Sons, Chichester. 301 pp. Timms, B.V. 1982. A study of the benthic communities of twenty lakes in the South Island, New Zealand. Freshwater Biology 12: 123–138. Timms, B.V. 1983. A study of benthic communities in some shallow saline lakes of western Victoria, Australia. Hydrobiologia 105: 165–177.

126 Vilaclara, G., Chávez, M., Lugo, A., González, H. and Gaytán, M. 1993. Comparative description of crater-lakes basic chemistry in Puebla state, Mexico. Verhandlungen Internationalis Vereinigung Limnologiae 25: 435–440. Wiggins, G.B., Mackay, R.J. and Smith, I.M. 1980. Evolutionary and ecological strategies of animals in annual temporary pools. Archiv für Hydrobiologie Supplement 58: 97–206. Williams, D.D. 1987. The Ecology of Temporary Waters. Croom Helm, London. 205 pp. Williams, D.D. 1996. Environmental constraints in temporary fresh waters and their consequences for the insect fauna. Journal of the North American Benthological Society 4: 634–650. Williams, W.D. 1996. The largest, highest and lowest lakes of the world: Saline lakes. Verhandlungen Internationalis Vereinigung Limnologiae 26: 61–79. Williams, W.D. and Kokkinn, M.J. 1988. The biogeographical affinities of the fauna in episodically filled salt lakes: A study of Lake Eyre South, Australia. Hydrobiologia 158: 227– 236. Williams, W.D., Boulton, A.J. and Taaffe, R.G. 1990 Salinity as a determinant of salt lake fauna: a question of scale. Hydrobiologia 197: 257–266.