Ostracods response to environmental anthropogenic disruption in a neotropical karstic lake in southern Mexico

Ostracods are highly sensitive to environmental changes and represent a powerful tool to assess anthropogenic impact. In this work we present ostracod and ephippium records of Lake San Lorenzo sediments dating from 1956 to 2013. Diatom and geochemical data presented by Caballero et al. (2022) are also considered for the discussion. The ostracod assemblage consists of Darwinula stevensoni, Cypridopsis okeechobei, Cytheridella ilosvayi and two unidenti�ed species. We suggest that one of them could be a variant of C. okeechobei with different ornamentation developed as a response of increasing salinity caused by anthropogenic catchment disruption around 1981. The second unidenti�ed species (Unknown 2) could belong to the Subfamily Cypridopsinae Kaufmann, 1900. This species rapidly adapted to the new environmental conditions of enhanced salinity, high nutrients level and variable oxygen, whereas the other ostracods were absent in the sediments dating from 1982 to 1992. Overall, D. stevensoni, C. okeechobei and C. ilosvayi are benthic species that occur in shallow waters or littoral zones with rich vegetation. However, we suggest C. ilosvayi was the most sensitive to oxygen variations as it is only present in the record in 1978. The lack of ostracod preservation in the sedimentary record, especially during 1995–2013 could be result of anoxic lake bottom conditions or dissolution caused by high organic matter content during periods of anthropogenic catchment disturbance, from 1970–1985 and from 1990–2000. Finally, we suggest analysis of C. okeechobei in natural populations and laboratory studies need to be assessed to determine if this species develops ornamentation changes when salinity conditions increase. Moreover, analysis and description of soft parts and live specimens of Unknown 2 will be necessary to determine this species.


Introduction
Ostracods can easily occur in any kind of aquatic environments and are considered as environmental traces, especially in freshwater ecosystems.They possess the ability to respond fast to physical and chemical changes of their environments such as salinity, water depth, temperature, or pH uctuations.Moreover, in microcosms experiments (e.g., Fennikoh et al. 1978; Ali and Lord 1980, Chial and Persoone 2003) ostracods have showed similar or higher sensibility to herbicides, pesticides, oil spills or heavy metal pollution than other traditional groups such as copepods, amphipods or bacteria, commonly used to assess anthropogenic impacts.Lake San Lorenzo belongs to the Montebello Lakes (ML) system in Chiapas, southern Mexico (Fig. 1).These lakes are considered as a touristic attraction due to their clear waters and spectacular views.
However, during the past decades anthropogenic activity has damaged this area and enhanced input of sewage water, fertilizers and pesticides have turned some of these lakes turbid with unpleasant odours (Alcocer et al. 2016).Recent research in the area based on geochemistry, magnetic susceptibility and diatoms suggest two main periods of basin-wide disturbance, from 1970 to 1985 and from 1990 to 2000 (Caballero et al. 2022).During these periods, Lake San Lorenzo´s turbidity and nutrient concentration increased as a result of long-distance sediment transport through the "Río Grande de Comitán" (RGC) caused by deforestation, intense agriculture, and land use changes.Moreover, in Lake Balamtetik, also part of ML, the sedimentary record exhibits as well these two major disturbance events with a resilience phase.However, the lake capacity to recover seems unlikely for the future if it continues receiving larger discharges as population on the area continue increasing (Caballero et al. 2020).
In this work, we present the ostracod analysis from the sedimentary record of Lake San Lorenzo and compare our results with the diatom record and geochemical data presented by Caballero et al. (2022) with the aim to assess the anthropogenic impact on the ML area.Moreover, our results also include ephippium counts, which are resting eggs that Cladocera produces when environmental stress increases (Ramírez 1981).

Study Area
Lake San Lorenzo (16º09' N, 91º46´ W, 1460 m.a.s.l.) belongs to the protected area "Parque Nacional Lagunas de Montebello" in the State of Chiapas, southern Mexico and close to the Guatemala border (Fig. 1).The ML complex was formed by a combination of karstic and tectonic processes where the outcropping rocks are partially or totally dolomitized Cretaceous limestones (Quezada-Muñetón 1987; Durán-Calderón et al. 2014).The lake exhibits an elongated shape with an area of 181 ha, a maximum depth of 67 m and an average depth of 11.8 m (Alcocer et al. 2016).The nearest meteorological station, La Esperanza (16°09'15" N, 91°52'05" W, 1500 m.a.s.l.; Fig. 1), provided mean monthly temperature and precipitation data from 1981 to 2010 CE (Common Era; Source: Servicio Meteorológico Nacional, Mexico), recording an average annual temperature of 17.2°C.The region experiences a maximum monthly temperature of 19°C during June and a minimum monthly temperature of 15°C in January and February (Fig. 1).Average annual precipitation is ~ 1209 mm and has a unimodal distribution, with most (~ 1011 mm) falling between May and October, and the maximum in September from tropical cyclones at the Paci c Ocean.Precipitation during the rest of the year is 199 mm.Vegetation in the region has been altered due to intense deforestation during the last decades, but originally it was dominated by Pinus-Quercus-Liquidambar and montane forests (March andFlamenco 1996, CONAP 2007).

Materials and Methods
The San Lorenzo sedimentary record was collected in July 2013 from the north part of the basin with a UWITEC gravity corer at 12 m depth, and sediments were sampled every 1 cm.For constructing the age model (Caballero et al. 2022; Fig. 2) 210 Pb results were acquired with relatively constant high activity values along the sequence (60-90 Bq/kg).This suggests that sediments of the San Lorenzo record are too young to allow signi cant 210 Pb to decay and reach the lower supported levels.However, a radiocarbon calibrated date (41 cm, Beta-376719, 102.8 ± 0.3 pMC) determined that sediments at the bottom correspond to the year 1956.Considering that the surface sediments represent the 2013, the year of collection, an age model based on linear interpolation was constructed (Fig. 2).Furthermore, values of 137 Cs activity along the San Lorenzo record range from 0 to 50 Bq/kg with relatively high frequency variation (Fig. 2).These low 137 Cs values suggests that the entire record postdates the beginning of the nuclear testing in 1945, as the radiocarbon age model suggests.However, the high frequency 137 Cs variations prevents the identi cation of individual events and suggest that Lake San Lorenzo has been constantly receiving 137 Cs contaminated soils eroded from the catchment (Poreba 2006; Appleby et al. 2019).For more information about the geochronology see Caballero et al. (2022).Elemental concentrations of Mn were determined by X-Ray uorescence (FRX) using a portable Niton XL3t95 equipment.For ostracod analyses a total of 40 samples from Lake San Lorenzo sediment sequence were processed.Each sample (~ 1 g) was oven-dried to determine dry weight, then soaked in tap water for 1-2 days prior to sieving at 177 and 63 µm.Sieve residues were oven-dried and examined under an Olympus stereomicroscope (SZX12) to determine species abundances, following Furtos (1936), Meisch (2000), Karanovic (2012) and Cohuo et al. (2017).The total ostracod abundance is reported as the number of adult and juvenile valves of all species in 1 g of dry sediment (valves/g).However, the relative abundance (in %) of each species only includes adult counts as juveniles have not fully developed their ornamental features and this can lead to an inaccurate taxonomic identi cation.All the ecological inferences here presented are based on adult relative abundances (in %).Adult ostracod valves and ephippium from cladoceran were extracted and stored in micropaleontology plates, while images were acquired with a Lane-MC500W camera.

Results
Five ostracod species were present as adults in Lake San Lorenzo sediment sequence (Fig. 3).Most of the shells were well-preserved, but some were coated with authigenic carbonate.The San Lorenzo ostracod assemblage consists of: Darwinula stevensoni (Brady & Robertson, 1870), Cypridopsis okeechobei (Furtos, 1936), Cytheridella ilosvayi (Daday 1905), and two unknown ostracod species Unknown 1 and Unknown 2. Figure 4 summarises the total ostracod (adult and juvenile instars) abundances (valves/g) and relative abundances (in %) of the ve species (only adult instars) from Lake San Lorenzo record.Moreover, we present total counts of ephippium from cladoceran per gram of dry sediment (ephippium/g).Total ostracod abundance ranged from 0 to 24 valves/g.Sediments at depths of 40 cm and 23 cm possessed the highest total abundances in the record, with 15 and 24 valves/g, respectively.Sediments from 37 − 31 cm and top 7 cm did not preserve ostracod valves.Adult relative abundance reached a maximum of 42%.Sediments at depths of 39 cm and 14 cm exhibited the highest adult abundances, at 40% and 42%, respectively.Darwinula stevensoni was preserved in sediments at depths of 39 cm, 23 cm and 14 cm, with 40%, 11% and 17%, respectively.Cypridopsis okeechobei was preserved in sediments at depths of 23 cm and 14 cm, with 11% and 25%, respectively.Cytheridella ilosvayi was only preserved in sediment at 25 cm depth with relative abundance of 25%.Unknown 1 was only preserved in sediment at 23 cm depth with relative abundance of 11% and presents very similar characteristics in size and shape of C. okeechobei.However, Unknown 1 exhibits a distinctive ornamentation (Fig. 3).Finally, Unknown 2 was preserved in sediments at depths from 22 − 20 cm and 16 cm, with relative abundances of up to 33%.Valves of this unknown specie are subcircular, with dorsal margin straight, anterior and posterior margins rounded (Fig. 3), and they exhibit sieve porosity.Measurements (Fig. 5) range from 222-267 µm length, 233-278 µm width, and 56-89 µm height.Mean calculated values are 244 µm, 250 µm and 75 µm, respectively.On the other hand, ephippium abundance ranged from 0 to 8 ephippium/g.Sediment at depth of 30 cm possessed the highest abundance in the record (8 ephippium/g), whereas sediments from 38 − 31 cm, 23 − 15 cm and top 13 cm did not preserve ephippium.

Discussion
Ostracod and ephippia records were used to determine the environmental characteristics of Lake San ).On the other hand, ephippium are resting eggs produced by Cladocera when disruption of optimal environmental conditions occurs.Ephippium sediment in the lake bottom where they remain safe until conditions become favourable again for a new parthenogenetic stage (Ramírez, 1981).Although we recognized three different types (Fig. 3), speci c classi cation was not possible as head remains are necessary for this.In the following section we discuss the possible environmental changes of Lake San Lorenzo based on ostracod and ephippia variations (this work), diatom record and geochemical data (Fig. 4) previously published in Caballero et al. (2022).As explain before, we use ephippium results as indicative of environmental stress, and Ti variations as indicative of catchment disturbance due to anthropogenic activities.Moreover, we use Mn variations as indicator of bottom redox conditions as this element is mobile in anoxic environments (e.g.Schaller andWehrli 1996, Caballero et al. 2019).In Fig. 4, we also present the diatom data highlighting the taxa which ecological distribution is indicative of higher turbidity, higher nutrients, and slightly higher salinity environments such as Aulacoseira granulata and its var.angustissima, Stephanocyclus meneghinianus, Stephanodiscus hantzchii and Ulnaria delicatissima.

Ecological implications
During the late 1950s Caballero et al. (2020) reports that changes in land use during the agrarian reform disturbed Lake Balamtetik (Fig. 1) environment with high local and regional erosion, soil organic matter, eutrophication, and anoxic bottom water conditions.Similarly, the Lake San Lorenzo Ti record also re ects higher erosion during this period.The Mn data suggest that anoxic lake bottom conditions prevailed except for a brief episode at the base of the sequence when ostracod valves were also present.Moreover, diatoms that are indicative of disturbance and eutrophication coexist with D. stevensoni and Cladocera ephippium.These results suggest that conditions of Lake San Lorenzo during the late 1950s were already with relatively high environmental stress possibly associated to changes in land use.
During the 1960s, ostracods and ephippium were not preserved in San Lorenzo record, whereas diatom productivity increased probably as a re ection of increased productivity in the lake.This could be indicative of a resilience period where environmental conditions enhanced the productivity of planktonic diatoms and Cladocera reestablished its parthenogenetic reproduction.However, ostracods absence in the sedimentary record correlates with anoxic bottom water conditions.Thus, we suggest ostracods are mainly responding to oxygen lake bottom availability with preference to oxygenated conditions and that they are lost from the record during times when increased lake productivity favoured anoxic bottom waters.
Based on magnetic susceptibility data, C/N values, and Ti and Fe concentration, Caballero et al. (2022) reported two periods of major catchment disturbance that increased the sediment discharge to Lake San Lorenzo through the RGC.The rst one from 1970 to 1985 whereas the second from 1990 to 2000.Overall, during the rst period diatom productivity decreases but proportions of the disturbance taxa remain high.Moreover, Cladocera also re ects increasing water turbidity as ephippium were produced and deposited in the sediment.On the other hand, the ostracod record exhibits presence of juveniles and no adults during the 1970s.This correlates again with oxygenated lake bottom conditions as the Mn record re ects.We suggest, the increasing sediment discharge caused by anthropogenic activity might have led to mixing lake water conditions, bringing some oxygen to the lake bottom.However, this new oxygen level was not optimal for the ostracods to reach their nal adult stage, until the 1978.During this time, the only presence of C. ilosvayi suggests conditions of enhanced dissolved oxygen and slightly higher salinity.However, in 1981 enhanced turbidity and salinity might have caused this species to disappear, while D. stevensoni, C. okeechobei and Unknown 1 are present in the sedimentary record.We suggest, Unknown 1 could be a variant of C. okeechobei with different ornamentation due to enhanced salinity conditions.Slightly higher salinities in these lakes seems to be a response to increased sulphates derived from the use of fertilizers (Mora Palomino et al. 2017) and in the diatom record some salt tolerant species were also recorded (Stephanocyclus meneghinianus) (Caballero et al. 2022).For example, Wrozyna et al. (2017) suggest that differences in chloride and sulphate concentrations, related to lake level uctuations and precipitation, are considered to affect valve development via controlling osmoregulation and carapace calci cation on Neotropical cytheroidean Ostracoda.However, this hypothesis on C. okeechobei needs to be tested.
Later on, from 1982 to 1992 the ostracod record exhibits only the presence of the Unknown 2 species.This period is the transition between the two intervals of major catchment disturbance reported by Caballero et al. (2022).During this period, diatom record shows the lowest productivity, and Mn data suggests uctuating bottom oxygen conditions as sediment discharge and water mixing ceased.Moreover, ephippium absence suggest Cladocera reestablished its parthenogenetic reproduction probably due to water turbidity decreasing.Overall, we suggest less oxygenated water and less turbidity conditions favored Unknown 2 occurrence.However, this unknown species disappeared from the record in 1994, when conditions of catchment disturbance (Ti) and turbidity settled back, also marked by presence of ephippium, D. stevensoni and C. okeechobei in the record.Moreover, we suggest Uknown 2 was a species that adapted faster to the new environmental conditions compared to the other ostracod species.Based on its smaller size (Fig. 5), sieve porosity and ability of rapid adaptation, we suggest Unknown 2 could belong to the Subfamily Cypridopsinae Kaufmann, 1900.Furthermore, no sexual dimorphism was observed on Unknown 2 valves.For example, C. vidua exhibits sieve porosity, it is considered as a cosmopolitan species with wide range of temperature toleration (Keyser 1976;Lorenschat 2009), and sex ratio analyses of C. vidua in natural populations and laboratory studies have indicated that this species is exclusively female (Havel and Hebert 1989).Analysis and description of soft parts and live specimens of Uknown 2 will be necessary to determine this species.
Finally, during the period from 1995 to 2013 the diatom productivity increased with higher proportions of the disturbance indicator taxa.Moreover, the Mn record exhibits the lowest values indicating the most severe anoxic conditions of the record.This correlates again with no ostracod preservation, but juveniles in 2002 when Mn data shifts from higher that the average to lower than the average values.We suggest the lack of oxygenated water at the bottom of the lake is affecting the ostracod productivity in Lake San Lorenzo.Modern sediments collected in 2013 were devoid of ostracods, when lake bottom oxygen levels were anoxic (0.4 mg/L).

Conclusions
Ostracods preserved in Lake San Lorenzo sediments representing from 1956 to 2013 exhibit ornamentation changes as well as species adaption to disturbed environmental conditions caused by anthropogenic activities such as land use changes and deforestation.Moreover, ephippium, resting eggs from Cladocera, also helped to determine periods of environmental stress such as increased water turbidity.Overall, we observed a negative correlation between ostracod and diatom total abundances.Low or absence of ostracods correlates with periods of higher diatom productivity, generally with higher proportions of eutrophic taxa.Moreover, we observed a positive correlation between ostracods and Mn data.Higher ostracod productivity occurred during periods of higher than the average Mn concentration, representing periods of oxygenated lake bottom conditions.Caballero et al. (2022) previously identi ed two periods of catchment disturbance that increased the sediment discharge to Lake San Lorenzo through the RGC as a result of changes in land use and deforestation.The rst one from 1970 to 1985 whereas the second from 1990 to 2000.In general, D. stevensoni and C. okeechobei coexisted during these anthropogenic environmental changes, where sediment discharge led to mixing waters and oxygenated lake bottom conditions.These species re ected tolerance to enhanced salinity conditions and high nutrients waters, whereas C. ilosvayi was more sensitive to oxygen variations.We suggest, C. okeechobei possibly developed ornamental changes (Unknown 1) as a result of increased salinity conditions during 1981.However, during the transition period between these anthropogenic environmental disruption in Lake San Lorenzo (1982 to 1992), where oxygen conditions were very variable (Mn record), the ostracod record re ects only presence of a species (Unknown 2) possibly belonging to the Subfamily Cypridopsinae Kaufmann, 1900.As we observed, valves of this species are relatively small, with sieve porosity and no sexual dimorphism was detected.
We suggest this species adapted rapidly to the new disturbed environmental conditions, and specially to variable dissolved oxygen levels.Analysis and description of soft parts and live specimens will be necessary to determine this species.
No ostracod preservation during 1995 to 2013 was re ected in sediments, but some juveniles in 2002.This period correlates with anoxic lake bottom conditions re ecting the high sensibility of these species to oxygen variations.Comparison through depth and time of the total ostracod abundance as valve/g, relative abundances of total adults, Darwinula stevensoni, Cypridopsis okeechobei, Cytheridella ilosvayi, and two unknown species as percentages (%).Ephippium from Cladocera is also presented as total abundance per gram sediment (ephippium/g).Indicators of catchment disturbance (standardized Ti), anoxic-oxic conditions (standardized Mn) and diatom records (total abundance per gram sediment and relative abundance of eutrophic taxa as %) are previously published by Caballero et al. (2022) Figure 5 3D scatter plot of measurements from unknown ostracod species Uknown 2; length ranges from 222-267 µm (244 µm mean value), width from 233-278 µm (250 µm mean value), and height from 56-89 µm (75 µm mean value)

Declarations Figures
Lorenzo from 1956 to 2013.D. stevensoni, C. okeechobei and C. ilosvayi have been previously reported in the Neotropical-Caribbean region(Pérez et al. 2010;Díaz et al. 2017; Amezcua-Vargas 2022).D. stevensoni is a benthic species with cosmopolitan distribution and tolerates mesohaline environments (≤ 15‰; Meisch 2000).This species is considered as an indicator of shallow waters (up to 15 m) and prefers aquatic environments with slow currents(Pérez et al. 2010).C. okeechobei is a nektobenthonic species that presents a high hydrochemical tolerance and prefers calm waters.Pérez et al. (2010) suggest it could be indicative of stressful environments with waters no deeper than 40 m, as living specimens in the Neotropical-Caribbean region prefer littoral zones with rich vegetation.C. ilosvayi is mainly a neotropical benthic species that prefers waters with salinity < 3.2‰, tolerates relatively high concentrations of sulphate (up to 2300mg/L), temperatures > 20°C and shallow waters (< 40 m)(Pérez et al. 2010 a