Presence of Leptospira spp. in a Mosaic of Wetlands Used for Livestock Raising under Differing Hydroclimatic Conditions

ABSTRACT Knowledge about the life cycle and survival mechanisms of leptospires in the environment is scarce, particularly regarding the environmental factors associated with their presence in ecosystems subject to livestock farming, where precipitation, seasonal floods, and river overflows could act as facilitators of leptospire dispersion. This study aimed to identify and study the presence of Leptospira spp. in the Lower Delta of the Paraná River and describe the physical, chemical, and hydrometeorological conditions associated with their presence in wetland ecosystems impaired by livestock raising intensification. Here, we show that the presence of Leptospira was determined mainly by water availability. We detected the species Leptospira kmetyi, L. mayottensis, and L. fainei and successfully cultured the saprophytic species L. meyeri from bottom sediment, suggesting the association of leptospires with microbial communities of the sediment’s biofilm to enhance its survival and persistence in aquatic environments and adapt to changing environmental conditions. Knowledge of Leptospira sp. diversity in wetlands and the impact of climate variability on the transmission of these organisms is crucial for predicting and preventing leptospirosis outbreaks in the context of human health. IMPORTANCE Wetlands are environments that are often conducive to the survival and transmission of Leptospira because they provide a suitable habitat for the bacteria and are often home to many animal species that can act as reservoirs for leptospirosis. Bringing humans and animals into closer contact with contaminated water and soil and increased frequency and intensity of extreme weather events may further exacerbate the risk of leptospirosis outbreaks, which is mostly relevant in the context of climate change and a widespread intensification of productive activities, particularly in the Lower Delta of the Paraná River. The detection of leptospiral species in wetland ecosystems impaired by livestock raising intensification can help to identify propitious environmental factors and potential sources of infection, develop preventive measures, and plan for appropriate responses to outbreaks, ultimately improving public health outcomes.

IMPORTANCE Wetlands are environments that are often conducive to the survival and transmission of Leptospira because they provide a suitable habitat for the bacteria and are often home to many animal species that can act as reservoirs for leptospirosis. Bringing humans and animals into closer contact with contaminated water and soil and increased frequency and intensity of extreme weather events may further exacerbate the risk of leptospirosis outbreaks, which is mostly relevant in the context of climate change and a widespread intensification of productive activities, particularly in the Lower Delta of the Paraná River. The detection of leptospiral species in wetland ecosystems impaired by livestock raising intensification can help to identify propitious environmental factors and potential sources of infection, develop preventive measures, and plan for appropriate responses to outbreaks, ultimately improving public health outcomes.
KEYWORDS Leptospira spp., physical, chemical, and hydrometeorological conditions, wetlands L eptospirosis is one of the most important zoonoses worldwide, affecting both developing and developed countries (1,2). Leptospirosis arises from the infection of spirochaetes of the genus Leptospira. The epidemiology of leptospirosis and the ecology of Leptospira spp. are particularly complex given the genetic diversity of the genus, which involves 82 species segregated into two phylogenetic clades. The saprophytic clade includes species isolated in the natural environment and not responsible for infections, while the pathogenic clade of evapotranspiration, provision of forage for livestock and habitat for wildlife species, and climate change mitigation (27,28). Particularly in the noninsular portion of the Lower Delta, extensive livestock raising has been, traditionally, the most widespread anthropic activity. Nevertheless, the expansion of soybean production in the late 20th century has forced the relocation of cattle toward marginal sites for agriculture, such as the Paraná Delta region (29). In consequence, cattle numbers have significantly increased accompanied by an unrestricted and widespread implementation of water management infrastructure (30). In these wetlands, precipitation and a high water level would facilitate the introduction of Leptospira species eliminated from urine of infected cattle to water bodies. In the context of intensification of livestock-raising activity, this study aimed to identify and study the presence of Leptospira spp. in the Lower Delta of the Paraná River (LDPR) across three periods differing in their hydroclimatic conditions and water quality variables. We hypothesize that the presence of Leptospira in wetland ecosystems is higher during periods of higher water availability.

RESULTS
Hydrometeorological variables. A map of the study area is provided in Fig. 1. In general, no significant El Niño or La Niña events were identified between January 2018 and December 2019. Land surface temperatures exhibited a strong seasonal pattern with no particular anomalies. Nevertheless, multivariate ENSO index (MEI) v2 values were lower and below zero during most of 2018 and higher and above zero during 2019. These values were consistent with lower water stage levels of the Paranacito River in 2018 and higher levels in 2019, including a series of extraordinary overflow events that occurred between January and March 2019 (Fig. 2). With regard to precipitation, the observed values exhibited a mild seasonal pattern. However, several extraordinary precipitation events occurred in December 2018 and January 2019 (Fig. 2).
Thus, three different hydroclimatic scenarios were identified: a drier scenario in November 2018, consisting of mostly lower precipitation values and lower yet increasing water stage levels (August 2018 to November 2018); a more humid period during April 2019, characterized by higher precipitation rates and higher yet decreasing water stage levels (January 2019 to April 2019); and a much drier period in November 2019, characterized by lower and sustained Paranacito River water stage levels and lower precipitation rates, which translated into widespread lack of water availability (August 2019 to November 2019).
Leptospira isolation, detection, and identification. Thirty-four water samples and 12 sediment samples from 15 sampling sites were collected in marshlands and ditches during the three sampling periods. With regard to water samples, the pathogenic species Leptospira kmetyi (S1 and S5a) and L. mayottensis (S2), as well as the intermediate species L. fainei (S5b), were identified in four freshwater marshlands during April 2019 (humid period). In addition, L. kmetyi was detected in a ditch (S12) during November 2018 (dry period) ( Table 1). Regarding culture from water samples, 44.1% (15 samples from a total of 34) showed no development of leptospires, whereas the remaining water samples (55.9%) were contaminated with a higher number of bacteria. In sediment samples, the saprophytic species L. meyeri was successfully isolated from a single sample relative to a freshwater marshland (S1) in November 2019 (much drier period). The remaining sediment cultures were contaminated.
Comparison of water quality variables among sampled periods. Most physical and chemical variables exhibited a strong spatial and temporal variability ( Table 2; Fig. 3). The first two principal components (PC1 and PC2) of the principal-component analysis (PCA) explained 47.4% of the total variability. The score plot of PC1 versus PC2 showed that water samples were grouped in terms of sampling periods. November 2018 (dry period) (Fig. 3, lower left quadrant) was mostly associated with higher pH values, whereas April

DISCUSSION
This is the first study detecting and identifying Leptospira species across a mosaic of heavily altered wetlands of the Paraná River system, in the context of intensified livestock raising. We detected the pathogenic species L. kmetyi and L. mayottensis, the intermediate species L. fainei, and the saprophytic species L. meyeri.
Our results showed that the presence of Leptospira species was strongly associated with the varying hydrometeorological conditions and, thus, mostly determined by water availability. Our findings were complemented by a description and assessment of the physical and chemical properties of water samples in freshwater marshlands and anthropogenic ditches, in the context of different hydrometeorological periods. Leptospira meyeri and L. kmetyi had already been found in other aquatic systems in subtropical and tropical regions around the world (13,24,(31)(32)(33). Our results identify L. mayottensis for the first time in a wetland subjected to cattle use in Argentina. This genomospecies has been isolated from blood of leptospirosis patients in Mayotte (34) and from environmental samples in New Zealand (35).
Our results support the idea that climatic factors such as heavy rainfall and flooding can increase the risk of leptospirosis outbreaks by promoting the growth and spread of Leptospira, particularly during El Niño southern oscillation (ENSO) events (14,36). In accordance with our hypothesis, a greater number of Leptospira species, and particularly pathogenic and intermediate species, were detected in water samples from the humid period (April 2019) than in samples from the drier periods (November 2018 and 2019). The aforementioned findings imply a positive relationship between Leptospira incidence and water availability. In the study area, cattle graze and urinate across the full extent of the topographical gradient, including seasonally flooded freshwater marshlands that usually dominate in the lower topographical settings. These areas constitute an important source of water for cattle as long as the hydrologic regimen and hydrological connectivity are preserved. In this context, river overflows and rainfall events play an important role in water replenishment of topographically lower areas (37). Periods of higher precipitation rates and increased water stage levels would flush Leptospira from topographically higher areas to water bodies, resulting in a higher incidence of these bacteria. Several studies carried out in tropical and subtropical regions have revealed the association between the occurrence of Leptospira in water bodies and periods of heavy rainfall and flooding as well (15,36,(38)(39)(40)(41).
Differing hydroclimatic scenarios in the study area had previously been associated with changes in physicochemical conditions of water (19). Furthermore, the results of this study showed that the highest values of CDOM evaluated as A 440 , TP, TN, chlorophyll a, and  Table 1. depth occurred during the humid period (April 2019). This period exhibited a higher presence of Leptospira in water, despite the fact that pH and dissolved oxygen conditions were unfavorable for its survival. This suggests that organic matter from nonpoint sources, including animal waste, was dragged by rains and increased water stage levels into freshwater marshlands and ditches, along with the microorganisms that it might contain, such as Leptospira. This result agrees with other studies of wetlands of the Paraná River system that associate wet conditions with the mobilization of organic matter from higher topographical areas into aquatic systems (42). In wetland ecosystems, bacterial mineralization of organic matter consumes oxygen and releases carbon dioxide (43), which might explain the generally low values of dissolved oxygen and slightly acidic waters in the study sites. In this regard, our results suggest that hydroclimatic conditions favoring Leptospira transport from terrestrial to aquatic habitats were more important drivers of their presence than abiotic variables able to affect bacterial survival (e.g., pH and dissolved oxygen).
During November 2019 (the driest evaluated period), L. meyeri was isolated from a sediment sample of a freshwater marshland subjected to livestock raising. This result suggests that L. meyeri could survive and persist in unfavorable conditions through its association with microbial communities of the sediment's biofilm, which would increase its ability to adapt to changing environmental conditions (9,15,23,44,45). Biofilms are believed to share a protector effect (20,46), contributing to the persistence and long-term survival of leptospires in adverse aquatic and wetland environments (15,44). It has been observed that leptospiral biofilm has a 5-to 6-fold increase in antibiotic resistance in all the strains used. It is tempting to speculate that biofilms may protect Leptospira against other toxic compounds in the environment (11).
With regard to the difficulty in properly culturing pathogenic species of Leptospira and their low growth rate, previous research found that it takes at least 10 6 UFC/mL to obtain an isolate (22). Saprophytic Leptospira spp. are most frequently isolated from environmental samples because they are common inhabitants of the environment and grow faster (47)(48)(49), whereas pathogenic and intermediate Leptospira spp. are able to survive but not multiply in the environment (6,22). In agreement with this, we successfully isolated the nonpathogenic specie L. meyeri, but most sediment and water samples were contaminated with others microorganisms despite the use of the combination of 5-fluorouracil with sample prefiltration through 0.22-mm-pore-size filters. These results suggest that future strategies might include more antimicrobial agents in the culture medium (6,50).
It is well known that the presence of pathogenic and intermediate Leptospira spp. constitutes a risk to human health and wildlife. In cattle, leptospirosis has been identified as one of the major causes of reproductive failure, including infertility. Increases in the number of services per conception, longer calving intervals, abortion, the occurrence of stillbirths and weak offspring, leading to significant economic hazards, have been observed as well (51). In this context, veterinarians, agriculture workers, abattoir workers, farm workers, fishers, and hunters can become infected during occupational activities that involve contaminated water. In addition, there is also a significant risk of exposure associated with recreational activities, including swimming, kayaking, canoeing, and triathlons, as well as military training exercises (52). Considering that human settlements and livestock numbers will probably increase over the years (30), measures should be considered to prevent the transmission of leptospirosis from its habitat to livestock or humans, and vice versa. Particularly in wetlands of the Paraná River Delta, the aforementioned considerations should be made while taking into account the periodicity of the flood pulse, as well as anomalous seasonal patterns of precipitation during extraordinary meteorological events such as El Niño southern oscillation.
The results presented in this study reinforce the positive association between humid conditions and pathogenic Leptospira in wetland ecosystems in a context of anthropogenic land use intensification. Informing decision makers of the seasonal hydrometeorological drivers of Leptospira presence and abundance should be the first step in preventing, forecasting, and planning control of leptospirosis, which affects not only humans in the contexts of recreational and productive activities but also animal health.
Conclusion. We describe the physical, chemical, and hydrometeorological conditions that could be associated with the presence of Leptospira in wetland ecosystems impaired by intensified livestock raising. The results of our study reinforce the idea that leptospires thrive and survive in both natural and heavily modified wetlands, such as freshwater marshlands and anthropogenic ditches, respectively. In turn, evidence supports the idea that Leptospira adapts to unfavorable and diverse environmental conditions. Detection of pathogenic and intermediate Leptospira in these wetlands during relatively humid periods could imply an increased human health risk, especially for those who come into contact with contaminated water and sediment in seasonally flooded areas. To help expand the current knowledge regarding leptospirosis survival as well as the spatial and temporal distribution of leptospires in the environment and animals, further studies should aim to describe the physicochemical characteristics not only of surface waters but also of bottom sediments of wetlands (including artificial water bodies) and consider the associations between Leptospira and biofilms.
Knowledge of Leptospira diversity in wetlands and the impact of climate variability on their transmission is crucial for predicting and preventing leptospirosis outbreaks in the context of human health. It can help identify potential sources of infection, develop preventive measures, and plan for appropriate responses to outbreaks, ultimately improving public health outcomes.

MATERIALS AND METHODS
Study area. The study area is located in the noninsular portion of the Lower Delta of the Paraná River, which is part of the Delta Region (17,500 km 2 ), in Argentina (South America) (Fig. 1). The Paraná Delta is a vast and highly biodiverse macromosaic of wetlands. It is characterized by a high environmental heterogeneity, mainly due to its geomorphological and hydrological complexity (27,28,53). The noninsular portion of the LDPR (233°459S; 58°519W) exhibits conspicuous geomorphological patterns, which are a result of marine ingressions and regressions that occurred in the Holocene and are still affected by ongoing fluvial processes (29,53). These patterns have been characterized as several landscape units differing in their hydrological regimens as well as in their geomorphological settings and land cover patterns (28,54). Its hydrologic regimen is affected not only by local rains but also by seasonal and extraordinary river overflows from the Paranacito, Paraná, Gualeguay, and Uruguay rivers, whose effects are spatially variable and dependent on hydrogeomorphological and anthropogenic features (37,54).
The climate is subhumid; the average (2001-2020) mean annual temperature and accumulated annual precipitation were 21.9°C and 1,358 mm, respectively (37). Normally, the study area exhibits a mild seasonal pattern with regard to temperature and precipitation. In addition, the area presents an important interannual variability due to ENSO, as well as high seasonal variability in precipitation and river water stage levels based on local hydroclimatic factors (37). As livestock raising has expanded and intensified, most of these patterns have been altered due to the unrestricted development of water management infrastructure, such as polders, embankments, and channelizations (30,55). The study area is entirely affected by livestock raising. Different cattle management practices occur across the full extent of the topographical gradient, which implies direct contact with mostly permanently flooded freshwater marshlands that comprise the most important wetland vegetation physiognomies. In addition, recreational activities such as fishing, swimming, and canoeing are conducted in some areas of these wetlands (54).
Sampling sites. Sampling sites were located in landscape units IIb and IV of the noninsular portions of the Lower Delta (54) (Fig. 1). These units are characterized by gentle undulations due to the presence of sandy ridges separated by depressions. Sampling sites were located either in the lower portion of the topographical gradient (freshwater marshlands) or in ditches within polderized livestock fields. Sampling sites were characterized by sediments rich in organic matter over a dense horizon of clay and fine silt content. Ditches were mainly characterized by free-floating macrophytes, mostly represented by Eichhornia crassipes, Salvinia spp., Azolla filiculoides, Pistia stratiotes, Limnobium laevigatum, and Lemna spp. In contrast, freshwater marshlands were mainly characterized by their most conspicuous species, Schoenoplectus californicus, as well as by accompanying emergent macrophytes such as Ludwigia peploides, Enhydra anagallis, Sagittaria sp., Myriophyllum aquaticum, and Pontederia spp. and the free-floating species mentioned above. In some cases, submerged vegetation, including algae and vascular plants such as Ceratophyllum demersum, were present as well. Samplings were conducted in November 2018 (spring), April 2019 (autumn), and November 2019 (spring) across nine freshwater marshlands (S1, S2, S3, S4, S5a, S5b, S6, S7, S8, and S9) and six ditches (S10, S11, S12, S13, S14, and S15) across livestock fields subject to permanent livestock raising and differing livestock management practices (Fig. 1). Due to the larger extension of freshwater marshland S5, and in order to consider its internal heterogeneity, two subsamples were collected (S5a and S5b).
Hydrometeorological conditions. In order to assess and describe the hydrometeorological conditions for each sampling period, we selected four variables that serve as indicators of the hydrologic regimen and water availability in the study area through the years 2018 and 2019: multivariate ENSO index, monthly mean precipitation, land surface temperature, and hydrometric levels of the Paranacito River. The MEI v2 data (56), provided by the Physical Sciences Laboratory (https://psl.noaa.gov/enso/mei/), identify the occurrence of El Niño (wet)/La Niña (dry) cycles and intensities. Due to the lack of precipitation-measuring instruments on the field, monthly precipitation data were obtained from the Integrated Multi-satellite Retrievals for GPM (IMERG) (57). Monthly values of land surface temperature were estimated from the MOD11C3 product (58). Water level time series data for the Paranacito River between 2018 and 2019 were provided by Prefectura Naval Argentina (https://contenidosweb.prefecturanaval.gob.ar/alturas/). In order to properly assess the hydrometeorological conditions for each sampling period, data covering the 3 months prior to each sampling date were considered.
Sample collection. Five subsurface water samples were collected at each sampling site and period, in sterile 50-mL plastic Falcon tubes. During November 2019, a period of extremely low water level occurred, which might increase the risk of human and animal exposure to sediment microorganisms as a consequence of the reduced water column. Therefore, in addition to water samples, approximately 5 g of sediment was collected during this period by using sterile 50-mL plastic Falcon tubes.
Sampling design was conditioned by the accessibility and flooding conditions at each sampling site. In case of floods due to heavy rain or river overflow events that impeded access to sampling sites, no samples were collected.
Culture of leptospires. Culture of Leptospira was performed immediately upon sample collection. For each water sample, 1 mL was filtered through a sterile membrane (0.22-mm pore size). The filtered water was inoculated into liquid and semisolid Ellinghausen-McCullough-Johnson-Harris (EMJH) medium (Difco Laboratories, Detroit, MI, USA) with the addition of 5-fluorouracil (FLU) (300 mg/mL) as a selective antimicrobial agent. Sediment culture was performed by adding 20 mL sterile distilled water to each plastic tube, shaking vigorously, and allowing sediment to settle for 15 min. Approximately 1 mL of the supernatant was taken and filtered with a sterile membrane with 0.22-mm pores. Two drops of the filtrate was inoculated into the semisolid EMJH medium. Cultures were incubated at 28°C, and leptospiral growth was monitored weekly for up to 4 months using dark-field microscopy. After 4 months without the presence of Leptospira, the culture was considered negative and discarded. In the case of sample contamination, 1 mL of the old culture was extracted and filtered into new EMJH medium with the addition of FLU (300 mg/mL), using a 0.22-mmpore-size membrane syringe filter.
DNA extraction. All samples were maintained and transported to the laboratory on ice for protection against sudden temperature changes and processed within 12 h after collection. Both water and sediment samples were centrifuged at 3,000 rpm (relative centrifugal force [RCF], 100) for 5 min. The supernatant was recovered and centrifuged at 8,000 rpm (RCF, 12,000) for 30 min at 4°C. Pellets of each quintuplicate sample were pooled, and then DNA was extracted using the QIAamp DNA minikit (Qiagen, Valencia, CA) following the manufacturer's instructions.
16S rRNA. Determination of species was performed using 16S rRNA as the amplification target (59). For each reaction, 1 U of GoTaq DNA polymerase (Promega, Madison, WI, USA), 200 mM deoxynucleoside triphosphates (dNTPs), and 1 mM primers were added to a total volume of 50 mL. Amplification was carried out using a Veriti thermal cycler (Applied Biosystems, Foster City, CA, USA), and the PCR products were analyzed on 2% agarose gels.
Sequencing and sequence analysis. PCR amplification products of 16S rRNA were purified using a GeneJET PCR purification kit (Thermo Scientific, Waltham, MA, USA) prior to DNA sequencing. PCR products were then sequenced by Macrogen Inc. (Seoul, South Korea). The sequences were edited using Chromas Lite 2.1.1 (Technelysium Pty., Ltd., Australia). The contigs were assembled using Staden package software (MRC-LMB, United Kingdom), and the alignment was performed using MEGA 5 (60). To obtain the Leptospira species, the assembled sequences of 16S rRNA were analyzed using the Ribosomal Database Project (RDP) (http://rdp.cme.msu.edu/). Physicochemical analysis. Water temperature, dissolved oxygen, conductivity, and pH were measured in situ with Hanna portable checkers. Subsurface water samples were collected in duplicates at each sampling site. Samples were filtered through Whatman GF/C glass fiber filters (pore size, 1.2 mm) within 24 h after sampling. Filters were stored at 220°C up to 3 weeks for spectrophotometric analysis of chlorophyll a, which was extracted with 90% acetone according to Lorenzen's method (61).
Filtered samples were passed through Millipore filters (pore size, 0.45 mm) and kept frozen until spectrophotometric determination of dissolved nutrients following (61). Nitrite (NO 2 2 ) was determined by diazotizing with sulfanilamide and coupling with N-(1-naphthyl)-ethylenediamine dihydrochloride, nitrate plus nitrite (NO 3 2 1 NO 2 2 ) was assessed by reduction of nitrate with hydrazine sulfate and subsequent determination of nitrite, ammonium (NH 4 1 ) was assessed by the indophenol blue method, and soluble reactive phosphorus (SRP) was assessed by the ascorbic acid method. The concentration of nitrate (NO 3 2 ) was calculated from the difference between NO 3 2 1 NO 2 2 and NO 2 2 . Total phosphorus (TP) and total nitrogen (TN) were estimated from unfiltered water samples that were kept frozen until analysis. TP was estimated by digestion with nitric and sulfuric acids followed by SRP determination, and TN was estimated by digestion with potassium persulfate in alkaline medium followed by NO 3 2 1 NO 2 2 determination (61). An aliquot of each water sample filtered through 0.45-mm filters was kept in darkness and refrigerated (4°C) until optical analysis of CDOM. Absorbance at 440 and 700 nm was measured using 1-cm quartz cuvettes and filtered Milli-Q water as a baseline. The absorbance at 700 nm was subtracted from the absorbance at 440 nm to correct offsets (62). The absorption coefficient (m 21 ) at 440 nm (A 440 ) was calculated according to (63) from the corrected absorbance at this wavelength and used as a measure of CDOM concentration. All the spectrophotometric determinations were carried out using a Shimadzu UV-1800 UV/visible-light spectrophotometer.
Data analysis. Patterns of spatiotemporal variability of physical and chemical variables were assessed via PCA, using the software CANOCO version 5 (Microcomputer Power, Ithaca, NY, USA). Variables were log transformed (except pH due to its logarithmic nature), centered, and standardized prior to the analysis. Comparisons among the three sampling periods for each physical and chemical variable were performed using repeated-measures ANOVA. Data were Box-Cox transformed when needed to fit homogeneity of variance assumption.