1. Introduction

Wetlands are essential water reservoirs for our ecosystems, which have been affected by human activity in recent decades on a global scale, producing a devastating impact on numerous plant and animal species that use these spaces as refuges (e.g., Scholz, 2015; Valdés-Pineda et al., 2020; Yang et al., 2020; Machuca-Sepúlveda et al., 2021). The advance of agro-industrial activities has exerted increasing pressure on water resources in these systems, especially in recent decades, due to both the use of more efficient water extraction technologies and higher water consumption for agricultural, industrial, residential, and touristic activities (e.g., Fernández, 2006; Valdés-Pineda et al., 2020, 2022; Yang et al., 2020; Machuca-Sepúlveda et al., 2021). In addition, water bodies in wetland systems often present surface area variations due to climatic and/or environmental forcings, leading the system to stressful situations when combined with anthropic drivers (e.g., Bianchi et al., 2015; Fuentealba et al., 2021).

In central-western Argentina (~32° S), the condition of the wetland systems is critical due to the prevailing arid-semiarid environmental conditions (Roig et al.,1991; Rubio et al., 2014; Pereyra et al., 2022). These wetland areas are highly fragile ecosystems given the scarcity of water availability throughout the year, combined with agricultural and urban development pressures. In the study area, one of the main factors influencing regional landscape change has been the growth of the cultivated area at a rate of 427 ha/yr in the middle-lower Mendoza River Basin, representing an increase of 10.4% for the period 1986-2018 when compared to previous estimates (Otta et al., 2022). These changes have induced a greater exploitation of water resources for agricultural irrigation. In fact, four aquifer levels are currently being exploited around the Leyes-Tulumaya system: (1) a phreatic level between 10 and 40 m depth; (2) a first level from 60 to 120 m depth; (3) a second level between 150 and 200 m depth, and (4) a third level from 240 to 350 m depth (Pazos et al., 1993). In the year 1993, a total of 16,000 wells had been registered around the city of Mendoza, with water extraction volume oscillations between dry and wet years of around 560-600 hm³ and 100 hm³, respectively, and an estimated average of water annual pumping at around 380 hm³ (Pazos et al., 1993). Based on historical analyses and satellite imagery, Mirábile et al. (2005) recognized a significant surface reduction of the shallow lakes in the study area between 1997 and 2002, which was then linked to the expansion of the agricultural frontier.

The Leyes-Tulumaya system was historically known as part of the Gran Ciénaga del Bermejo and experienced variations in its extension over the last 400 years (Prieto, 2000; Prieto et al., 2008; Prieto and Rojas, 2012) (Fig. 1A, B). In the study area, there is great interest in understanding the causes behind the wetland changes since the ancient system is presently associated with a series of shallow (<10 m deep) lakes similar to oxbows, well preserved along the Leyes-Tulumaya stream (Vitali, 1940; Zambrano, 1978) (Fig. 1C). Action measures have been implemented since the second half of the 20^th century, including public policies aimed at the conservation of natural resources and ecosystems under the Ramsar Convention. In consequence, current challenges relate to the changes the wetland system could be subjected to, particularly in the face of climate change scenarios (Fernández Cirelli and Abraham, 2002; Sampietro Vatuone and Peña Monné, 2019; Lauro et al., 2022; Otta et al., 2022).

Fig. 1. Satellite image of the study area. A. Subregional figure with main rivers in blue. LTS: Leyes-Tulumaya system, GW: Lagunas de Guanacache wetland, DW: Desaguadero wetlands, and NR 7: National Route 7. B. Maximum surface of the Gran Ciénaga del Bermejo wetland at around year 1802 (light green area; modified from Prieto et al., 2008), with main streams and rivers in blue. C. Interconnected lakes along the Leyes-Tulumaya stream: (1) El Viborón, (2) La Paloma, (3) De Soria, (4) Lauriente, and (5) Montenegro. Historical lake locations taken from Mirábile et al. (2005).

Under the present climate change conditions, the global and regional impact of anthropic activities on natural ecosystems is exacerbated. Understanding the causality between natural and anthropogenic factors affecting wetlands can provide valuable tools for environmental protection. This study therefore analyses the changes in wetland water areas along the Leyes-Tulumaya system over the last two decades. Regional/local climatic and hydrological variables were analyzed together in order to determine their potential role behind the observed changes in lake surface areas, thus shedding light on how to direct wetland management and protection.

2. Study area

The Leyes-Tulumaya system, formed by a group of shallow lakes, wetlands, and the Leyes-Tulumaya stream, is the remnant of a huge historical wetland located in eastern Mendoza Province in west-central Argentina, at 610-680 m a.s.l. (Vitali, 1940; Zambrano, 1978) (Fig. 1B, C). The Leyes-Tulumaya stream corresponds to a ~N-S paleo-stream bed of the Mendoza River, emerging from an upwelling area near 32°59’ S/68°39’ W and extending for about 35 km in a semi-straight line northwards from the town of Villa Tulumaya (Fig. 1C). A series of interconnected oxbows occur along this stream, called ponds and swamps in the literature (Prieto, 2000) and locally known as lakes (Fig. 1C). The El Viborón, La Paloma, and De Soria lakes are preserved at present while other historical lakes (e.g., the Lauriente and Montenegro) disappeared as they were drained for agricultural practices as a consequence of the expansion of the agriculture frontier in the 1970s (Mirábile et al., 2005) (Fig. 1C).

2.1. Physiographic characteristics of the study area

The lakes of the Leyes-Tulumaya system are located in the fluvial-alluvial plain of the Mendoza River at the Andes piedmont (Fig. 1A). This extensive geomorphological unit is limited to the south by the current course of the Mendoza River, to the north by the Lagunas de Guanacache wetland, which is where the Mendoza and San Juan rivers merge, and to the east by the large sandy fluvial-aeolian plains of the Desaguadero River (Sepúlveda, 2001; Devincenzi et al., 2015) (Fig. 1A). The sedimentary infill of this fluvial-alluvial plain is the result of the erosion of different lithologies that crop out in the mountain (western) areas of the upper Mendoza River Basin (González Díaz and Fauqué, 1993; Sepúlveda, 2001). In the Late Cenozoic, tectonic subsidence favored a large accumulation of sediments that reaches a current maximum thickness of up to 600 m (Rodríguez and Barton, 1993; Sepúlveda, 2001). This basin configuration was apparently structurally controlled by a submeridional fault system that would have conditioned the current disposition of the Mendoza riverbed (Zambrano, 1978).

A semi-arid climate characterizes the region with a mean annual precipitation of 210 mm (Mendoza Aero Meteorological station, https://www.smn.gob.ar/observaciones). Precipitation mainly occurs during the warmest months (December to March, 75%) through easterly surface winds and convective storms mainly forced by the South Atlantic Anticyclone (e.g., Viale and Garreaud, 2014; Viale et al., 2019). The Mendoza River is fed by snow melting in the Andes (Fig. 2). There, snowfall occurs between June and August, associated with frontal systems from the west along narrow latitudinal bands known as storm tracks (Garreaud et al., 2008). In the Leyes-Tulumaya area, the mean annual temperature is 17.1 °C, with a mean maximum during January of 32.3 °C and a mean minimum of 2.1 °C in July. Consequently, this region exhibits a water deficit, with an average evapotranspiration rate of 2.13 mm/day in summer (https://goo.su/54gdEq4).

Fig. 2. Satellite image of the Mendoza River Basin. Basin boundaries are shown in aquamarine. Numbers refer to the gauge stations used to measure mean monthly streamflow along the Mendoza River: (1) Punta de Vacas, and (2) Guido. (3) refers to the Mendoza Aero Meteorological Station, used for mean monthly and annual precipitation records.

Native vegetation communities associated with the Leyes-Tulumaya system belong to the Monte Phytogeographic Province (Oyarzabal et al., 2018). The dominant xerophytic shrubland is characterized by plants with small leaves, reduced to thorns or absent, waxy or resinous leaves and thickened cuticles belonging to the Zygophyllaceae family and particularly represented by different species of the genus Larrea (Larrea divaricata, Larrea cuneifolia, Larrea nitida and Larrea ameghinoi). The arboreal stratum includes Parkinsonia aculeata and isolated specimens of Neltuma flexuosa and Geoffroea decorticans, accompanied by cactuses such as Denmoza rhodacantha, Cereus sp. and Echinopsis sp. Open azonal forests of Neltuma sp., Salix humboldtiana and Acacia visco develop associated with soils rich in silt and clay content (Oyarzabal et al., 2018). On the other hand, halophytic communities including Atriplex argentina, Atriplex spegazzinii, Atriplex lampa, Suaeda divaricata, and Plectrocarpa tetracantha, among others, along with abundant species of the Cactaceae family (e.g., Denmoza rhodacantha, Opuntia sp.) are present in well-drained sandy-clayey soils (Roig et al., 1996). The halophytic vegetation has an upper stratum composed of Allenrolfea vaginata, Cyclolepis genistoides, Atriplex vulgatissima, and Neltuma alpataco, and a lower stratum of Neltuma strombulifera in areas with clayey soils where swamps form periodically, but if the flooding period lasts significantly longer, species like Baccharis spartioides, Distichlis spicata, and Tessaria absinthioides dominate instead (Roig et al., 1996).

2.2. The Leyes-Tulumaya stream system

The Leyes-Tulumaya stream is not permanent at present since surface runoff (which depends on groundwater fluctuations and the frequency/magnitude of rainfall) is not enough to allow for surface water to circulate through the system. This system has been affected by human intervention, such as modification on the stream paths, partial damming, and stream bed invasion for growing crops or urban development. These anthropic changes, along with the passive (i.e., low speed) surface water circulation through the system, have led to water stagnation resulting in the El Viborón, La Paloma, and De Soria lakes (Fig. 1C).

As mentioned in the Introduction, the Leyes-Tulumaya system used to be part of a huge swamp known as the Gran Ciénaga del Bermejo, whose first historical records date back to the arrival of the Spaniards in 1561 AD. Based on historical maps, this wetland presented its maximum area in the mid-1800s, with an approximate surface of 45,000 ha (Fig. 1B). At its largest, the Gran Ciénaga del Bermejo was characterized as a network of ‘green islands’ with variable water extension and marshy sectors ideal for cattle breeding. Their surroundings were inhabited by the Huarpe people whose economy was based on fishing and goods exchange (Prieto, 2000; Prieto et al., 2008).

This extensive swamp decreased in area in the following decades in response to the seasonal discharges of the Mendoza River. Thus, any interannual episodic areal increase/decrease of the Gran Ciénaga del Bermejo was related to the increase/decrease of the Mendoza River seasonal streamflow (Prieto, 2000). Based on early 18^th century chronicles of ancient settlers (Prieto, 2000), the extension of the Gran Ciénaga del Bermejo showed a marked increase during the Little Ice Age (1520-1660 AD in the Southern Hemisphere; Villalba, 1994). Historical maps from the 18^th century show how the surplus water from the first agricultural plots and the Mendoza River contributed to maintaining the wetland environment, even with a semi-permanent presence of water bodies. The first important modifications were observed at the beginning of the 20^th century, associated with the drainage of wetland areas by local settlers for the development of activities such as agriculture and cattle raising. This process was dramatically accelerated with the intervention of public policies to improve irrigation systems and the allocation of more arable lands (Prieto, 2000).

Based on historical records, climatic variations and anthropic activities on a regional scale may have been responsible for changes in the Leyes-Tulumaya system, modifying the geographic space drastically (Prieto et al., 2008). While the area is geologically conditioned from a morphotectonic point of view, since it is a depression collecting both surface and subsurface water, the activities directly aimed to drain the wetland were likely one of the main drivers of change during the 20^th century.

Upstream the Leyes-Tulumaya system, in the Mendoza River, the construction of the Potrerillos Dam (1,270 m a.s.l.) was certainly one of the main modifications to the basin during the 21^st century (Fig. 2). The dam was finished in 2003 and provides water yearlong for the population and the productive agricultural sector of northern Mendoza Province (Salomón et al., 2008). Although regulated, the Mendoza River flows downstream the dam with variable flow regimes related to the agricultural cycle, particularly in periods of water deficit.

3. Materials and methods

3.1. Antecedents and remote sensed imagery

Background knowledge on the Leyes-Tulumaya system, with a focus on the Gran Ciénaga del Bermejo, was built upon the compilation of previous studies (from the 1940s to the present), public information, geological information, and published reports (Vitali, 1940; Romanella, 1957; Zambrano, 1978; Prieto, 2000; Rodríguez and Barton, 1993; Sepúlveda, 2001; Mirábile et al., 2005; Prieto et al., 2008; Prieto and Rojas, 2012; Gómez et al., 2014). These literature data were complemented with high-resolution satellite imagery for each lake (Landsat 8 and Maxar Technologies, freely available on Google Earth), spanning the years 2003-2021. The temporal image resolution varied from lake to lake, so direct comparisons were not always possible (Supplementary Table S1). Pre-2003 data were retrieved by using Landsat 7 imagery, which provides good temporal coverage since the 1980s although the spatial resolution was not as detailed (30 m per pixel). In consequence, the pre-2003 data were not used to determine the area of the lakes along the Leyes-Tulumaya stream.

3.2. Streamflow and meteorological data

The mean annual and monthly streamflows of the Mendoza River were obtained from the National Water Institute (INA by its acronym in Spanish). Only gauge stations upstream the Potrerillos Dam were chosen: Punta de Vacas (32°45’ S/69°50’ W, 2,380 m a.s.l.; series 1949-2021), and Guido (32°50’ S/69°16’ W, 1,540 m a.s.l.; series 1956-2021) (see gauge locations in Fig. 2). The selection of upstream gauge stations only was due to the limited data from gauge stations downstream the Potrerillos Dam since the discharges are immediately deviated for hydroelectric and irrigation purposes.

Cumulative annual and monthly precipitation data were obtained from the Mendoza Aero meteorological station (32°50’ S/68°48’ W, 700 m a.s.l.; series 1956-2021) due to its proximity to the study area (see location in Fig. 2). Snow precipitation data (taken from the Argentinian and Chilean Andes Snow Observatory platform, https://observatorioandino.com/nieve/) was not included within the analyzed climatic variables due to a strong correlation with the Mendoza River streamflow data for the period 2000-2022. This correlation is not new, being discussed in Masiokas et al. (2006), among others.

3.3. Topographic and lake surface data

Digital topographic data were retrieved from the ALOS Global Digital Surface Model, with a vertical accuracy of 10 m that can be corrected to 7 m, of the Japan Aerospace Exploration Agency (http://surl.li/zeaafs). A finer spatial resolution of 12.5 m was generated by applying a low-pass filter with a square kernel of 21x21 pixels using the r.neighbors function in a GRASS GIS environment. A morphometric analysis was then performed to highlight landscape features and determine the geomorphological configuration of the area drained by the Leyes-Tulumaya stream (Fig. 3).

 

Fig. 3. Topographic map along the Leyes-Tulumaya stream system. North arrow points to the right. Bottom: Topographic profile along the main course of the stream, depicting the location of the El Viborón (EV), La Paloma (LP), and De Soria (DS) lakes.

The surface extent of the Leyes-Tulumaya system lakes El Viborón, La Paloma, and De Soria were calculated for the period 2003-2021 by loading the satellite images on QGIS (Supplementary Table S1). The temporal surface evolution of each individual lake was analyzed by using 65 images. For each lake, the measured areas were transformed into percentages based on the maximum area recorded during the timespan analyzed and compared across lakes only for data captured at the same time (Figs. 4 and 5). Yearly averages for the years 2003, 2011-15, and 2017-21 were obtained. By May 2022, the El Viborón lake was already fully desiccated and colonized by halophyte plants, as observed during field trip activities.

Fig. 4. Areal extension of the Leyes-Tulumaya system lakes from January 2003 to September 2021. Areas represented only for the months where data were obtained for all lakes

Fig. 5. Area percentages of the Leyes-Tulumaya system lakes El Viborón (EV), La Paloma (LP), and De Soria (DS), from January 2003 to September 2021. Right plot: Integrated temporal series of the three lakes.

The analysis of lake areas to differentiate between natural and anthropogenic forcings through satellite monitoring is abundant in the literature (e.g., Kuppel et al., 2015; Houspanossian et al., 2016; Ferrelli, 2020; Otta et al., 2022; Whitworth-Hulse et al., 2023). Recent studies have also integrated remote sensed imagery and statistical tools to elucidate the natural and/or artificial factors responsible for variations in different water bodies (e.g., Sumiya et al., 2020; Fuentealba et al., 2021; Wang et al., 2021; Soria and Apostolova, 2022).

3.4. Statistical correlations and ancillary information

The hydroclimatic variables considered in this study were (1) the streamflow of the Mendoza River, and (2) the precipitation in the Mendoza Aero meteorological station (Fig. 2), both on an annual and monthly basis. To correlate these two variables with the lake area data for the last two decades, the Pearson correlation coefficient was used.

The Global Lake Evaporation Volume (GLEV) dataset, which provides surface area and evaporation volumes for more than a million lakes and water reservoirs worldwide (Zhao et al., 2022), does not include information about the Leyes-Tulumaya lakes. To remedy this, in the present study an artificial water body associated with a sewage treatment plant (El Paramillo Wastewater Treatment Plant), close to the Leyes-Tulumaya system (32°50’ S/68°33’ W) was analyzed in Google Earth Engine. Data for this artificial reservoir were taken from the GLEV dataset as well, thus providing useful monthly information for the period 1985-2018. As for the vegetation and soil water contents, they were calculated by using the Normalized Differential Moisture Index (NDMI) by means of Sentinel-2 MSI Level-1C imagery (spatial resolution of 10 m in band B8 and 20 m in band B11) with Google Earth Engine for the 2016-2022 period (satellite data only available for this period). The NDMI was calculated by comparing bands 8 and 11 for each pixel during the analyzed period by keeping the coarser resolution (Strashok et al., 2022). NDMI values were also obtained for a specific area and plotted over time, highlighting the inter-annual variations of the index.

4. Results

4.1. Topographic controls on lake formation along the Leyes-Tulumaya system

The topographic profile of the Leyes-Tulumaya stream along ~18 km shows a progressive elevation decrease towards the north from a maximum of 644 m a.s.l. in the southern area to a minimum of 617 m a.s.l. in the northern area (stream slope of 0.086°) (Fig. 3). Therefore, groundwater flows from south to north, propelling that lakes located at lower altitudes have more stable areas (Gomez et al., 2024).

4.2. Lake area dynamics along the Leyes-Tulumaya system

The lake area changes along the Leyes-Tulumaya system reveal a similar evolution although at different rates (Fig. 4). The El Viborón lake, elongated in a north-south direction, reached a maximum extent in July 2011 (~280,000 m²) and a minimum in June 2021 (~70,000 m²) (Figs. 4 and 5; Table 1). This lake showed area variations of 40% at the beginning of the record up to 75% at the end of the period analyzed. After a transient recovery in September 2021, this lake dried up in May 2022. The La Paloma lake, characterized by a meandering shape (Fig. 4), had its maximum area in January 2003 (~180,000 m²) while its minimum was recorded in January 2014 (~12,000 m²) (Figs. 4 and 5; Table 1). This lake showed area variations close to 90% (Fig. 5). The De Soria lake, the northernmost of the three, reached its largest surface area in July 2011 (~135,000 m²) and its smallest in April 2020 (~40,000 m²) (Figs. 4 and 5; Table 1). Water area variations of up to 70% were determined (Fig. 5).

Together, these three lakes reached large areas in July 2011, October 2013, April 2014, and August 2019, and small areas in January 2014 and April 2020. In addition, pronounced episodes of lake contraction-expansion were observed in the southernmost part of each lake (Fig. 4), thus reflecting the south-to-north gradient in topography.

The year 2003 marks in the region a sudden and significant increase in the areas used for agricultural activities and urban development, adding pressure on the stream bed system. It is worth noticing that lake water areas remained with no significant changes at least until the year 2011.

4.3. Hydroclimatic variables and their relationship with lake area data

The streamflow of the Mendoza River shows a clear negative trend since around the year 2006, with average values of ~43 m³s^-1 in the Guido gauge and ~6 m³s^-1 in the Punta de Vacas gauge (Fig. 6). At Guido, maximum and minimum flow values of ~73 m³s^-1 and ~29 m³s^-1 were estimated; whereas at Punta de Vacas these were ~12 m³s^-1 and ~3 m³s^-1 (Fig. 6). The maximum values occurred in the same year (2006) for these two gauges.

Fig. 6. A. Annual precipitation record from the Mendoza Aero Meteorological Station (mm) and streamflow record of the Mendoza River (m3s-1) at the Guido and Punta de Vacas gauges. B. Box and whisker plot showing the monthly precipitation record from the Mendoza Aero Meteorological Station (mm) for the period 2000-2021.

The annual precipitation records show that most of it remained above 200 mm, peaking at between 400 and 500 mm in 2001 and 2016 (Fig. 6). The Pearson correlation coefficients of the hydroclimatic parameters (streamflow and precipitation) and the lake areas indicate weak correlations at both annual and monthly timescales (Table 2). This suggests that the lake area dynamics have been independent of precipitation and streamflow, at least for the last two decades.

4.4. Evaporation (GLEV) and soil moisture (NDMI) data

The artificial water body analyzed shows an overall steady increase in the evaporation volume for the 1985-2018 period, with a sudden increase in the early 1990s possibly due to the addition of a new water for sewage effluent treatment (Fig. 7A). The maximum area was 0.82 km² and the maximum evaporated volume was 0.16 · 10⁶ m³ month^{-1/, the first occurring in the winter of 2017 and the latter occurring in the summer of 1999.}

Fig. 7. A. Area (in km2) and evaporation volume (in 106 m3 per month) for an artificial water body near to the study area based on the Global Lake Evaporation Volume (GLEV) dataset for the 1985-2018 period. B. Normalized Difference Moisture Index (NDMI) time series for the 2016-2022 period. Months reported in the chain M(March)-M(May)-J(July)-S(September)-N(November). C. Single-band, pseudo color image in the blue range showing the NDMI variations for the 2016-2022 period. Note that the El Paramillo wastewater treatment plant remains with water during the whole time, so this is why it shows a dark blue color. EV: El Viborón, LP: La Paloma, and DS: De Soria.

The NDMI values, on the other hand, show an oscillatory pattern for the 2016-2022 period, with higher values (>0.03) during warmer, rainier months (December to March) and minimum values during colder, drier months (June to September) (Fig. 7B). The discrete peaks observed in the data series may be associated with flooding events after rainfall episodes and/or irrigation of agricultural lands in summertime. When the NDMI values are plotted on a map showing the combination of values for the period 2016-2022, the highest values are observed in cultivated areas, whereas the sector surrounding the Leyes-Tulumaya system lakes exhibits very low (negative) values (Fig. 7C).

During autumn (March to June), there are differences in the NDMI time series before and after 2020 (Fig. 7B). In the first case, NDMI values are typically -0.05, possibly associated with water releases for irrigation, while in the second case they are bit higher, between -0.03 and -0.01, and possibly indicate a higher availability of water resources for crops. In 2022, minimum values oscillate around 0, reflecting less intense dry autumnal periods.

5. Discussion

Regional topography plays an active key role in controlling the water flow pattern of the Leyes-Tulumaya stream, forcing the arrangement of the wetlands along this stream and the presence of shallow lakes like oxbows and/or ponds (Fig. 3). Indeed, Gómez et al. (2014) reported that the flow lines of the aquifer follow the northward orientation of the Leyes-Tulumaya stream. The latter implies that those shallow lakes located in upper topographic positions, even only a few meters above other water bodies, will drain to the north. Particularly, it is more feasible to preserve those water bodies located at lower topographic elevations, such as the De Soria lake, than those located higher up, such as the El Viborón lake (Figs. 1 and 3). Therefore, when it comes to surface water areas, the De Soria lake has varied relatively less in magnitude when compared to the El Viborón lake (Figs. 4 and 5). The El Viborón lake dried up in May 2022 due to important environmental transformations, observing the development of a 20-30 cm-thick layer of salt crystals and the colonization of the surface by Oxybasis macrosperma halophytic plants (field observations). The La Paloma lake, on the other hand, located in an intermediate topographical position, showed a water surface reduction of up to 90% in 2014 and 2020 (Fig. 5), while the El Viborón and La Paloma lakes showed decreases between 60 and 70% for the same years. The latter suggests that the topographic position of the lakes is not the unique factor behind the surface variabilities observed.

The joint analysis of lake water area variations and hydroclimatic variables (streamflow and precipitation) show no evident relationships among them over at least the last couple of decades (Table 2). This finding may be related to the fact that the Leyes-Tulumaya system has long been immersed in a productive matrix experiencing multivariable, smaller-scale natural and anthropogenic forcings that hinder the proper role of climatic variables. Alterations in this impacted system include (1) the construction of the Potrerillos Dam in the Mendoza River in 2003, (2) minor blockages of the Leyes-Tulumaya stream due to increasing urbanization and agricultural activities, and (3) the management of water resources for agro-industrial activities and/or the use of the aquifers for agriculture activities.

The evaporation (GLEV) and soil moisture (NMDI) data are not clearly related to the lake areas either (Fig. 7), but the times series of these two variables do contribute to understanding their behavior during the analyzed periods. The evaporation volume responds well to the interannual cycle dynamics of the Leyes-Tulumaya system, showing greater water availability in summer and lesser in winter (Fig. 7A). On the other hand, the soil moisture data evidence hydrological stress periods associated with low lake levels (high radiance; Fig. 7B) with direct negative effects on the studied wetlands.

Contrary to the findings presented here for the Leyes-Tulumaya system, basin-scale studies developed at the Lagunas de Guanacache, a wetland of higher hierarchy located in northeastern Mendoza Province, noticed a direct relation between climatic variables and lake water areas during the 1960s and 1970s (Contreras et al., 2011; Arancibia Abrahan, 2019; Álvarez et al., 2024). This relationship prevailed until the anthropic intervention of the system in the 1980s with the construction of small dams associated with the creation of agricultural oases. Since then, the lacustrine water areas of the Lagunas de Guanacache began to decrease (Contreras et al., 2011; Arancibia Abrahan, 2019; Álvarez et al., 2024).

Although the Leyes-Tulumaya system corresponds to a wetland of lower hierarchy when compared to the Lagunas de Guanacache, the sustained water area variability of the three lakes analyzed is probably having significant impacts on the long-term dynamics of the ecosystems along the middle-lower Mendoza River Basin. The existence of a large, ancient swamp of higher hierarchy during historical times (the Gran Ciénaga del Bermejo; Prieto, 2000; Prieto et al., 2008), followed by drained (or dry) lacustrine bodies along the Leyes-Tulumaya system in present times, implies a drastic environmental change at least during the last 400 years that continues at present, and is even being reinforced, given the development of anthropic activities (Otta et al., 2022).

The Leyes-Tulumaya system and the Lagunas de Guanacache are not anomalies, in fact, on a global scale, research on higher hierarchy wetlands also shows accelerated impacts on these ecosystems in recent decades (Zuquette et al., 2020; Assefa et al., 2022; Zhu et al., 2022). Major changes in wetlands worldwide are associated with the impacts of climate change, resulting in an increased vulnerability of these fragile environments that needs to be investigated further (Stewart et al., 2013; Wardrop et al., 2019).

6. Conclusions

This study reveals that the areas of three lakes along the Leyes-Tulumaya system, in central-western Argentina, have been highly variable over the last 20 years. These area variations occurred in phase, indicating their connectivity through the Leyes-Tulumaya stream, reason as to why it is considered here as a system. Topography is an important but not unique conditioning factor in the persistence of the lacustrine areas. In fact, the lakes located at lower topographic positions (the La Paloma and De Soria) showed smaller area fluctuations, while the El Viborón lake, located at higher altitude, experienced greater fluctuations until its complete drying in 2022.

No evident relationships were found between the hydroclimatic variables (precipitation and streamflow) and the lake area data for the last 20 years. This is because the Leyes-Tulumaya system has been severely altered by different anthropogenic activities since the beginning of the 21^st century. Therefore, further studies are required to investigate and quantify the impacts of urban and agricultural activities on the Leyes-Tulumaya ecosystem, including surface water and groundwater management. Future studies will benefit from finer satellite temporal resolutions so as to track in more detail, and predict with greater accuracy, the changes that these threatened environments are subjected to.

Acknowledgments
The authors would like to thank L. Bonjour (Herbario Ruiz Leal-IADIZA) for her help in identifying the plant species present in the study area. To the staff of the Club de Pesca Cristóbal Colón for their support and collaboration, especially to M.C. De Huin. This work was conducted under a CONICET grant of Rios and funded by the PICT 2019-3426 project led by Moreiras and PIP 2022-0318 funded by CONICET.
We are grateful to J. Soria, an anonymous reviewer, and D. Bertin (Editor of Andean Geology) for their comments that substantially improved the final version of this manuscript.

References

Álvarez, L.M.; Rivera, J.A.; Calderón Archina, A. 2024. Efectos de la variabilidad climática y las políticas hídricas en el Sitio Ramsar Lagunas de Guanacache (San Juan, Mendoza y San Luis, Argentina): cuando la gestión finaliza en los diques. Cuadernos Geográficos 63 (1): 142-157. https://doi.org/10.30827/cuadgeo.v63i1.27952

Arancibia Abrahan, V. 2019. Fluctuación espacio-temporal del complejo de Lagunas de Guanacache, Divisadero y Bebedero durante 1987-2018. Boletín de Estudios Geográficos 112: 161-180.

Assefa, W.W.; Eneyew, B.G.; Wondie, A. 2022. The driving forces of wetland degradation in Bure and Wonberma Woredas, Upper Blue Nile basin, Ethiopia. Environmental Monitoring and Assessment 194 (838): 26 p. https://doi.org/10.1007/s10661-022-10516-8

Bianchi, L.; Rojas, F.; Rivera, J.; Navarro, M.; Villalba, R. 2015. Desarrollo de un índice regional de balance hídrico inferido a partir de imágenes satelitales de la laguna de Llancanelo, Mendoza. In Congreso Argentino de Meteorología, No. 12, Mar del Plata. Argentina.

Contreras, S.; Jobbágy, E.G.; Villagra, P.E.; Nosetto, M.D.; Puigdefábregas, J. 2011. Remote sensing estimates of supplementary water consumption by arid ecosystems of central Argentina. Journal of Hydrology 397 (1-2): 10-22. https://doi.org/10.1016/j.jhydrol.2010.11.014

Devincenzi, S.M.; Moreiras, S.M.; Chiesa, J.; Gómez, D. 2015. Río Desaguadero: una propuesta como lugar de interés geológico (Argentina). Revista de la Sociedad Geológica de España 28 (1): 61-75.

Fernández Cirelli, A.; Abraham, E. 2002. El agua en Iberoamérica. De la escasez a la desertificación. Cooperación Iberoamericana, Programa Iberoamericano de Ciencia y Tecnología para el Desarrollo: 27-44. Buenos Aires.

Fernández, J.M. 2006. Agua y sostenibilidad: algunas claves desde los sistemas áridos. Polis 14: 1-13.

Ferrelli, F. 2020. Efectos de la variabilidad climática sobre las coberturas del suelo de distintas cuencas hidrográficas de Argentina. Geográfica Digital 17 (34): 3-18. http://dx.doi.org/10.30972/geo.17344444

Fuentealba, M.; Bahamondes, C.; Sarricolea, P.; Meseguer-Ruiz, O.; Latorre, C. 2021. The 2010-2020 ῾megadrought̕ drives reduction in lake surface area in the Andes of central Chile (32°-36°S). Journal of Hydrology, Regional Studies 38 (100952): 15 p. https://doi.org/10.1016/j.ejrh.2021.100952

Garreaud, R.D.; Vuille, M.; Compagnucci, R.; Marengo, J. 2008. Present-day South America climate. Palaeogeo-graphy, Palaeoclimatology, Palaeoecology 281 (3-4): 180-195. https://doi.org/10.1016/j.palaeo.2007.10.032

González Díaz, E.F.; Fauqué, L.E. 1993. Geomorfología. In Geología y Recursos Naturales de la provincia de Mendoza (Ramos, V.; editor). Relatorio Congreso Geológico Argentino, No. 12, y Congreso de Exploración de Hidrocarburos, No. 2: 217-234. Mendoza.

Gomez, M.L.; Aranibar, J.N.; Wuilloud, R.; Rubio, C.; Martínez, D.; Soria, D.; Monasterio, R.; Villagra, P.; Goirán, S. 2014. Hydrogeology and hidrogeochemical modeling in phreatic aquifer of NE Mendoza, Argentina. Journal of Iberian Geology 40 (3): 521-538. https://doi.org/10.5209/rev_JIGE.2014.v40.n3.43302

Houspanossian, J.; Giménez, R.; Baldi, G.; Nosetto, M. 2016. Is aridity restricting deforestation and land uses in the South American Dry Chaco? Journal of Land Use Science 11 (4): 369-383. https://doi.org/10.1080/1747423X.2015.1136707

Kuppel, S.; Houspanossian, J.; Nosetto, M.; Jobbágy, E. 2015. What does it take to flood the Pampas?: Lessons from a decade of strong hydrological fluctuations. Water Resources Research 51 (4): 2937-2950. https://doi.org/10.1002/2015WR016966

Lauro, C.; Vich, A.; Otta, S.; Moreiras, S.; Pascuali E.; Mejías, L. 2022. Recursos hídricos superficiales de la vertiente oriental de los Andes centrales (28°-37°S) en contexto de variabilidad hidroclimática. Boletín de Estudios Geográficos 116: 45-71. https://doi.org/10.48162/rev.40.007

Machuca-Sepúlveda, J.; López, M.; Vargas, E. 2021. Equilibrio ambiental precario en humedales áridos de altura en Chile. Ciencia Latina 5 (4): 4345-4369. https://doi.org/10.37811/cl_rcm.v5i4.625

Masiokas, M.; Villalba, R.; Luckman, B.H.; Le Quesne, C.; Aravena, J.C. 2006. Snowpack variations in the Central Andes of Argentina and Chile, 1951-2005: large-scale atmospheric influences and implications for water resources in the region. Journal of Climate 19 (24): 6334-6352. https://doi.org/10.1175/JCLI3969.1

Mirábile, C.; Zuluaga, J.M.; Fusari, M.; Burgos, V.; Núñez, M. 2005. Estudio de la variación temporal del área palustre del río Mendoza a partir de datos históricos, cartográficos y satelitales. In Congreso Nacional del Agua, No. 20. Mendoza.

Otta, S.; Jobbágy, E.; Vich, A.; Viglizzo, E.; Lauro, C.; Vaccarino, E.; Bastidas, L. 2022. Dinámica de las tierras irrigadas en el centro-oeste de Argentina durante el período 1986-2018: análisis a partir de la anomalía del Índice de Vegetación mejorado. Geoacta 44 (1): 35-55.

Oyarzabal, M.; Clavijo, J.; Oakley, L.; Biganzoli, F.; Tognetti, P.; Barberis, I.; Maturo, H.M.; Aragón, R.; Campanello, P.I.; Prado, D.; Oesterheld, M.; León, R.J. 2018. Unidades de vegetación de la Argentina. Ecología Austral 28 (1): 156 p. https://doi.org/10.25260/EA.18.28.1.0.399

Pazos, A.J.; Bessone, J.L.; Vivas, J.A.; Vaca, A.; Wetten, C.; Valero, C.E. 1993. Recursos hídricos subterráneos y aguas termales. In Geología y Recursos Naturales de la provincia de Mendoza (Ramos, V.; editor). Relatorio Congreso Geológico Argentino, No. 12, y Congreso de Exploración de Hidrocarburos, No. 2: 551-559. Mendoza.

Pereyra, F.X.; Ragas, D.B.; Cornacchia, M.C. 2022. Clasificación geológica-geomorfológica de los humedales de Argentina. Revista de la Asociación Geológica Argentina 79 (2): 275-291.

Prieto, M. del R. 2000. Formación y consolidación de una sociedad en un área marginal del Reino de Chile: la provincia de Cuyo en el siglo XVII. Tesis de Doctorado. Anales del Instituto de Arqueología y Etnología, Facultad de Filosofía y Letras, Universidad Nacional de Cuyo 52-53: 17-366 p. Mendoza.

Prieto, M. del R.; Rojas, F. 2012. Documentary evidence for changing climatic and anthropogenic influences on the Bermejo Wetland in Mendoza, Argentina, during the 16^th-20^th century. Climate of the Past 8 (3): 951-961. https://doi.org/10.5194/cp-8-951-2012

Prieto, M. del R.; Abraham, E.; Dussel, P. 2008. Transformaciones de un ecosistema palustre. La Gran Ciénaga del Bermejo-Mendoza, siglos XVIII y XIX. Multiquina 17: 147-164 p.

Rodríguez, E.J.; Barton, M. 1993. El cuaternario de la Llanura. In Geología y Recursos Naturales de la provincia de Mendoza (Ramos, V.; editor). Relatorio Congreso Geológico Argentino, No. 12, y Congreso de Exploración de Hidrocarburos, No. 2: 173-194. Mendoza.

Roig, F.A.; Gónzalez-Loyarte, M.M.; Abraham, E.M.; Méndez, E.; Roig, V.G.; Martínez Carretero, E. 1991. Maps of desertification hazard of central western Argentina (Mendoza province): study case. In World Atlas of Desertification, United Nations Environment Programme: 50-53. London.

Roig, F.; Martínez Carretero, E.; Méndez, E. 1996. Mapa de vegetación de la Provincia de Mendoza. Instituto Argentino de Investigaciones de las Zonas Áridas- IADIZA. 1 mapa escala 1:1.000.000. Mendoza.

Romanella, C.A. 1957. Los suelos de la región del río Mendoza (ensayo de establecimiento de series). Boletín de Estudios Geográficos 14 (4): 1-57.

Rubio, M.C.; Fermani, S.; Victoria Parera, V. 2014. Evolución de la conservación en la provincia de Mendoza. Desafíos en el proceso de ordenamiento territorial en tierras secas. Zonas Áridas 15 (2): 195-210.

Salomón, M.; Abraham, E.; Sánchez, C.; Rosell, M.; Thome, R.; Lopez, J.; Albrieu, H. 2008. Análisis de los impactos ambientales generados por las presas sobre los Sistemas de Riego. Cuenca del Río Mendoza. In Tecnologías de Información y Comunicación (TIC) para la modernización de los sistemas de irrigación y valorización de los sistemas de riego ancestrales: 22 p. Florianópolis.

Sampietro Vattuone, M.M.; Peña Monné, J.L. 2019. Cambios ambientales y geoarqueología en medios áridos/semiáridos:propuesta metodológica. Boletín de la Sociedad Geológica Mexicana 71 (2): 565-584. https://doi.org/10.18268/bsgm2019v71n2a19

Scholz, M. 2015. Wetland Systems to Control Urban Runoff, 2^nd Edition. Elsevier: 556 p. Amsterdam.

Sepúlveda, E. 2001. Hoja Geológica 3369-II, Mendoza, Provincias de Mendoza y San Juan. Instituto de Geología y Recursos Minerales, Servicio Geológico Minero Argentino, Boletín 252: 53 p. Buenos Aires.

Soria, J.; Apostolova, N. 2022. Decrease in the water level of lake Prespa (North Macedonia) studied by remote sensing methodology: relation with hydrology and agriculture. Hydrology 9 (6): 12 p. https://doi.org/10.3390/hydrology9060099

Stewart, R.I.A.; Dossena, M.; Bohan, D.A.; Jeppesen, E.; Kordas, R.L.; Ledger, M.E.; Meerhoff, M.; Moss, B.; Mulder, C.; Shurian, J.B.; Suttle, B.; Thompson, R.; Trimmer, M.; Woodward, G. 2013. Mesocosm experiments as a tool for ecological climate-change research. Advances in Ecological Research 48: 71-181. https://doi.org/10.1016/B978-0-12-417199-2.00002-1

Strashok, O.; Ziemiańska, M.; Strashok, V. 2022. Evaluation and correlation of Normalized Vegetation Index and Moisture Index in Kyiv (2017-2021). Journal of Ecological Engineering 23 (9): 212-218. https://doi.org/10.12911/22998993/151884

Sumiya, E.; Dorjsuren, B.; Yan, D.; Dorligjav, S.; Wang, H.; Enkhbold, A.; Weng, B.; Qin, T.; Wang, K.; Gerelmaa, T.; Dambaravjaa, O.; Bi, W.; Yang, Y.; Ganbold, B.; Gedefaw, M.; Abiyu, A.; Girma, A. 2020. Changes in water surface area of the lake in the steppe region of Mongolia: a case study of Ugii Nuur Lake, Central Mongolia. Water 12 (5): 18 p. https://doi.org/10.3390/w12051470

Valdés-Pineda, R.; García-Chevesich, P.; Valdés, J.; Pizarro-Tapia, R. 2020. The first drying lake in Chile: causes and recovery options. Water 12 (1): 12 p. https://doi.org/10.3390/w12010290

Valdés-Pineda, R.; García-Chevesich, P.; Alaniz, J.; Venegas-Quiñones, H.; Valdés, J.; Pizarro, R. 2022. The Impact of a lack of government strategies for sustainable water management and land use planning on the hydrology of water bodies: lessons learned from the disappearance of the Aculeo Lagoon in Central Chile. Sustainability 14 (1): 33 p. https://doi.org/10.3390/su14010413

Viale, M.; Garreaud, R. 2014. Summer precipitation events over the western slope of the subtropical Andes. Monthly Weather Review 142 (3): 1074-1092. https://doi.org/10.1175/MWR-D-13-00259.1

Viale, M.; Bianchi, E.; Cara, L.; Ruiz, L.E.; Villalba R.; Pitte, P.; Masiokas, M.; Rivera, J.; Zalazar, L. 2019. Contrasting climates at both sides of the Andes in Argentina and Chile. Frontiers in Environmental Science 7: 1-15. https://doi.org/10.3389/fenvs.2019.00069

Villalba, R. 1994. Tree-rings and glacial evidence for the Medieval Warm Epoch and the Little Ice Age in southern South America. Climatic Change 26: 183-197. https://doi.org/10.1007/BF01092413

Vitali, G. 1940. Hidrología Mendocina: Contribución a su conocimiento. Primera edición. Gildo D’Accurzio: 245 p. Mendoza.

Wang, Q.; Chen, X.; Peng, W.; Liu, X.; Dong, F.; Huang, A.; Wang, W. 2021. Changes in runoff volumes of inland terminal lake: a case study of Lake Daihai. Earth and Space Science 8 (11): 1-17.https://doi.org/10.1029/2021EA001954

Wardrop, D.H.; Hamilton, A.T.; Nassry, M.Q.; West, J.M.; Britson, A.J. 2019. Assessing the relative vulnerabilities of Mid‐Atlantic freshwater wetlands to projected hydrologic changes. Ecosphere 10 (2): 1-30. https://doi.org/10.1002/ecs2.2561

Whitworth-Hulse, J.; Jobbágy, E.; Borrás, L.; Alsina, S.; Houspanossian, J.; Nosetto, M. 2023. The expansion of rainfed grain production can generate spontaneous hydrological changes that reduce climate sensitivity. Agriculture, Ecosystems & Environment 349: 108440. https://doi.org/10.1016/j.agee.2023.108440

Yang, X.; Wang, N.; Chen, A.; He, J.; Hua, T.; Qie, Y. 2020. Changes in area and water volume of the Aral Sea in the arid Central Asia over the period of 1960-2018 and their causes. Catena 191: 104566. https://doi.org/10.1016/j.catena.2020.104566

Zambrano, J. 1978. Mapa geotectónico del centro Oeste argentino, escala 1:1.000.000. Centro Regional de Aguas Subterráneas de San Juan. Informe Inédito.

Zhao, G.; Li, Y.; Zhou, L. 2022. Evaporative water loss of 1.42 million global lakes. Nature Communications 13 (3686): 10 p. https://doi.org/10.1038/s41467-022-31125-6

Zhu, L.; Ke, Y.; Hong, J.; Zhang, Y.; Pan, Y. 2022. Assessing degradation of lake wetlands in Bashang Plateau, China based on long-term time series Landsat images using wetland degradation index. Ecological Indicators 139 (108903): 14 p. https://doi.org/10.1016/j.ecolind.2022.108903

Zuquette, L.; Failache, M.; Barbassa, A. 2020. Assessment of depressional wetland degradation, spatial distribution, and geological aspects in Southern Brazil. Geosciences 10 (8): 31 p. https://doi.org/10.3390/geosciences10080296