Effects of Temperature Changes on the Spatial Distribution and Ecology of Ostracod (Crustacea) Species

To understand the possible effects of changes in ambient temperature on spatial distribution and ecology of ostracods, samples were randomly collected from 70 aquatic sites with 12 different habitat types from Hatay (Turkey) province during the summer season of 2012. 14 of 19 ostracod species were newly reported for the province. The first two axes of CCA explained 79.7% of the cumulative variance of the relationship between the 12 most common species and five environmental variables. Accordingly, water temperature and electrical conductivity were the most effective factors on species occurrences (p<0.05). Estimating ecological optimum and tolerance values of species revealed that Herpetocypris chevreuxi and Cypridopsis vidua displayed the lowest and highest tolerance values for water temperature, respectively. TWINSPAN results illustrated that ostracod species can be used to determine characteristics of habitat conditions. Indeed, the co-occurrence of H. chevreuxi with one or more cosmopolitan species is the indication of an increase in salinity and temperature values. Results suggested that temperature changes can cause critical alteration in shallow water bodies where species with lower ecological tolerances will eventually be negatively affected. Therefore, such species, which may be called “potential candidate species for local extinction” will either be eliminated from the habitats in short term or become extinct in long term.


Introduction
Extinction is not one inevitable result that organisms must face. However, it is a fact that most (if not all) species are under the threat of global and/or local extinction (Black et al. 2001;Eisenhauer et al. 2019) due to climatic and anthropogenic factors. The loss of biodiversity is especially noticeable in freshwater ecosystems. Because, cumulative effects of climate change and human-induced factors (e.g., land-use change, destruction, overexploitation) cause changing in flow regime and chemical composition of freshwaters (Dudgeon et al. 2006;Ertürk 2012;Leigh 2013). Consequently, this situation leads to critical alteration in species composition in a habitat and/or geographic distribution of species Finlayson et al. 2013). Because, when species are faced with changes, most (if not all) of them will not be able to develop fast responses enough to the new environmental conditions. While the species may have a chance to survive as they fit the new conditions by different adaptation abilities, the others, which have restricted ecological ranges, may either become more vulnerable or die out.
Invertebrate animals, playing an important role in the continuous function of ecosystems, are examples of facing threats of extinction or rapid decline in their numbers (Benateau et al. 2019). Indeed, examples of human effect coped with local climatic factors have been illustrated in aquatic habitats along with rapid loss of invertebrates. For example, levels of response to climate change conditions were found various among different benthic invertebrates and four ecoregions in 26 European streams and rivers (Jourdan et al. 2018). Of which, an abundance of sensitive Plecoptera was declined during warmer years while the abundance of Ephemeroptera was increased in northern regions. Besides, Jourdan et al. (2018) found a significant increase in the abundance of invasive species with an increasing number of harsh days induced by climatic changes.
Of invertebrates, Ostracods (Ostracoda: Crustacea) are one of the diverse and abundant taxonomic groups with about 65000 fossil and living species (Kempf 1980(Kempf , 1997. There are about 2300 subjective non-marine species distributed worldwide (Meisch et al. 2019). They are also an important element in the food chain in shallow aquatic bodies (Mesquita-Joanes et al. 2012) and work as key species on production and community metabolism of micro-or mesocosm freshwater beds (Ruiz et al. 2013). Having with wide global distribution in a variety of aquatic habitats, ostracods are considered to provide supportive evidence for environmental changes because of their speciesspecific habitat preferences and different levels of ecological tolerance and optimum ranges.
Combining developing technology with historical and palaeoclimatological data can help to produce future climate change scenarios. Accordingly, different future climate estimation models showed that the climate of the world with an increase in temperature and a shift in precipitation patterns has already changed (IPCC 2007). In Turkey, mean air temperature is foreseen to increase between 1°C-2°C in 2016-2040 and 1.5°C-4°C in 2041-2070, while rainfall trends are less predictable with possible increases and decreases in average precipitation rates (Demircan et al. 2017). Ragab and Prudhomme (2002) projected that amount of precipitation will decrease, and temperature will increase by about 1.75-2.5°C by the year 2050 for the Mediterranean region. In potent climate fluctuations especially the precipitation changes, in some areas such as the Mediterranean region of Anatolia where our study area Hatay province is located, is one of the susceptible areas in Turkey (Karabulut 2009).
Aside from these, Hatay province was chosen as a study area since there has been no spatially extensive and comprehensive study asking the relationship between temperature changes and their effect on ostracod assemblages. Hence, this study aimed to (i) investigate correlation between the effects of ecological changes (e.g., temperature) and distribution of individual ostracod species, and (ii) estimate species ecological tolerance and optimum values along with determining their indicator values.

Materials and Methods
Hatay province with a surface area of 5403 km² is bordered by Syria and the Mediterranean Sea on the south-eastern and western sites, respectively ( Figure 1). Having with İskenderun and Antakya ports, it is one of the heavily industrialized districts on the Mediterranean coast. Besides, organized industrial zones located around the sea, freshwater, and terrestrial ecosystems have negative effects on natural sources in its environment. Because of its location in the southern part of Turkey, the city is under the influence of the Mediterranean climate characterized by very hot, long, and dry summers with cool rainy winters. Samples were randomly collected from 12 different inland water bodies (lake, dam, pond, pool, trough, ditch, irrigation canal, creek, stream, river, waterfall, and spring) in 70 sites located at about sea level (11 m) to 740 m a.s.l of elevations. Since July and August are the hottest months of the region, we thought that sampling between 31 July and 7 August 2012 might be better to show the utmost effect of the air temperature on water bodies. Thus, materials were collected with a plankton hand net (200 μm in mesh size) from each site and stored in 250 ml of a plastic container with 70% ethanol. Then, the material was filtered over four standard sieves (1.5, 1.0, 0.5, and 0.25 mm) and ostracods separated from sediment with fine needles under the Meiji-Techno stereo microscope in the laboratory. Species description was done based on soft body parts and carapace, which are dissected and preserved in Lactophenol -Orange G solution, under the Olympus CX-41 light microscope by using different taxonomic works (Broodbakker and Danielopol 1982;Karanovic 2006Karanovic , 2012Meisch 1984Meisch , 1985Meisch , 2000. All forms of the species are preserved at the Limnology Laboratory of Bolu

Abant
İzzet Baysal University, Bolu, Turkey. During the sampling, we used a GPS 45 XL for recording the coordinate and elevation of each sampling site. While air temperature (Ta, °C) and moisture (Moi, %) were measured by Testo 410-2 model of anemometer, water temperature (Tw, °C), pH, dissolved oxygen (DO, mg L -1 ), saturation (S, %), salinity (Sal, ppt), electrical conductivity (EC, μS cm -1 ) and total dissolved solids (TDS, mg L 1 ) were measured by YSI Professional Plus Series in situ. The physicochemical and geographical data of stations along with distributions of the obtaining species were shown in the Appendix.
Shannon-Wiener (H') was calculated via Species Diversity and Richness, version 4 program (Seaby and Henderson 2006) to examine the species diversity within different habitat types. Canonical Correspondence Analysis (CCA) along with Monte Carlo permutation test (499 permutations) was used to determine the most effective environmental variable(s) on species (ter Braak 1986;ter Braak and Verdonschot 1995). Suitability of CCA was tested with a priori analysis of DCA (Detrended Correspondence Analyses). Length of DCA (>3) suggests possible linear correlation and suitability of the data for CCA. To reduce the influence of multicollinearity and arc-effect, rare species were automatically down-weighted, and the data was log-transformed by the program of Canoco 4. C2 program was used to calculate ecological tolerance and optimum values of individual ostracod species (Juggins 2003). In all statistical analyses, live adults occurred in at least three different samples were used while juveniles, damaged individuals, and sub-fossils were excluded from the analyses. We used Two Way Indicator Species Analysis (TWINSPAN) to estimate the indicator values of individual ostracod species among the habitats in the Community Analysis Package program (CAP 4.1.3) (Seaby and Henderson 2006). This method provides a clustering relationship among the habitats distinguished by species with positive or negative indicator values.  ) is important to extend the known geographical distribution of the species toward the southern parts of Turkey.

Total of 19 ostracod species [Darwinula
The distribution of habitat types along with a total number of species and numbers of individuals among the sampling sites at different elevational ranges were shown in Table 1. According to this, both abundance values and numbers of species were found to show a gradual decrease with increasing elevation. Besides, the majority of the stations (21 stations) with the highest habitat diversity (9 types) were found in the range of 101-200 m a.s.l. The highest species numbers (13 species) were observed at 0-100 m and 201-300 m, even though the numbers of stations were significantly different with 18 and 11 sites, respectively. On the other hand, the highest number of individuals (abundance) was calculated for 301-400 m with seven sites and 1616 individuals. Among the habitat types, creeks displayed the highest species diversity (H'= 2.473) ( Table 1). Following creeks, species diversity was found high in streams (H'= 2.282) and spring (H'= 1.758) waters.
The first two axes of CCA diagram explained about 79.7% of the relationship between cumulative percentage variance of species and environmental variables. Eventually, the most influential factors on species were water temperature (P= 0.002, F= 4.327) and electrical conductivity (P= 0.014, F= 2.562) (Figure 2 a, b). According to the tolerance and optimum estimate values (Table 2) of the species, two species (H. incongruens and H. salina) displayed the highest tolerance values for dissolved oxygen and salinity (referring to electrical conductivity) while C. vidua and P. albicans showed the highest values for water temperature and pH, respectively. H. chevreuxi known as a pure freshwater species showed the lowest tolerance for water temperature (tk= 1.29) and conductivity (tk = 75.85). TWINSPAN results outlined that species (e.g., H. chevreuxi) with certain indicator values can be used to discriminate characteristics of habitats ( Figure 3).  Hartmann (1964) was the first to report five ostracod species [Heterocypris incongruens, Heterocypris.

Discussion
salina, Ilyocypris bradyi, Herpetocypris chevreuxi and Potamocypris zschokkei (Kaufmann, 1900)] from the province. Then after, Gülen (1985Gülen ( , 1988 Özuluğ and Kılıç (2002) listed one more species Costa edwardsii (Roemer, 1838), which is especially known with its preference of marine waters but can also be found in brackish waters (Meisch 2000). Four of these species (H. incongruens, H. salina, I. bradyi and H. chevreuxi) were also found during the present study. Combining the species with our 19 ostracods, total numbers of freshwater ostracods of Hatay province increased to 26.
Finding 15 of 19 species from the creeks corresponds to highest species richness (H'= 2.473) in these habitats. According to Connor and McCoy (1979) more species can be encountered with increasing sampling sites in wider areas. This suggests that there can be a positive correlation between the numbers of species and sampling sites. Thus, the fact that more species were encountered in creeks may be related to the high numbers of sampling efforts (20 out of 70 sites). On the other hand, numbers of samples alone cannot explain relatively high species diversity in different habitat types. For example, streams also had high diversity with 11 species, collected from only 8 different sites (Table 1). In such a case, we assume that habitat suitability may also play an important role in species diversity and distribution.
In general, the highest number of species was found at low elevations when the least number of species was found at high elevations. According to Rapoport's Rule (see e.g., Stevens 1992), numbers of species decrease with increasing elevation. The results of this study do not correspond to the assumption of the rule due to differences in the numbers of stations and habitat types that the greater part of the stations with more different sampling sites were already situated at low elevations. On the other hand, the situation can be explained with the "habitat diversity hypothesis" (Williams 1943) and then with ecological features that individual species showed. Some species (e.g., H. incongruens) with wide tolerance levels to different environmental variables exhibited a wide geographical distribution in a variety of habitats. These species so-called "cosmoecious species" (Külköylüoğlu 2013) usually display elevation free distribution. For example, during the present study, H. salina was reported from almost all elevations. In contrast, stenoecious species (e.g., H. chevreuxi) with a narrow tolerance levels to some of those environmental variables are of limited distributional ranges. This may imply that such these species do have advantages over other species. However, cosmoecious species concept does not include biological factors on species distribution that is probably effective with the arrival of species into a new area. As stated above, water temperature plays an important role for water quality and life in aquatic ecosystems (Morril et al. 2005). Besides, herein as illustrated in CCA diagram showed that water temperature was the most effective factor on species distribution (p< 0.05). In reality, changes in water temperature are relatively slower and narrower interval ranges than the air temperature due to physicochemical properties of water bodies (Odum and Barrett 2005). This is generally (and especially) true for deep waters. However, in the present study, ostracods were collected from shallow aquatic habitats which are more prone to the influence of changes in air temperature. This supports the idea of Preud'homme and Stefan (1992) that predictions can be better for shallow water bodies than a deeper one. In this way trends of water and air temperatures in each site were changed almost parallel to each other. Following such changes, most of the ostracod species must deal with the new conditions, to which that the species with high tolerances will have better adapted. For instance, when a freshwater habitat is altered to hypersaline waters due to increasing temperature gradients, Cyprideis torosa can survive approximately 30 days in dormant form (Meisch 2000), but stenoecious species can suffer from this change.
In this study, estimating ecological optimum and tolerance values of species revealed that H. chevreuxi displayed low tolerance values for water temperature (tk= 1.29) and conductivity (tk= 75.85) ( Table 2). This species usually prefers stagnant waters in Turkey but there is not much ecological data about it. In a recent study, H. chevreuxi was found among those with low tolerance value for temperature and dissolved oxygen in the shallow waters of Bolu (Turkey) . If the current climatic scenarios are correct, one may consider that H. chevreuxi will possibly the first species put under the risk of local extinction in this region due to its low tolerance level for water temperature (tk= 1.29). According to Meisch (2000), the species can be considered as pure freshwater species and its cooccurrence with one or more halophilic ostracods (e.g., Heterocypris salina) indicates an increase in salinity levels of that water body. Our results support this view that H. chevreuxi was found from five different sampling sites where the species was commonly found with one or more of those cosmopolitan species (e.g., H. salina, H. incongruens, I. bradyi, C. vidua) which are known to tolerate different levels of environmental variables, including high levels of salinity changes.
The first two clustering groups in TWINSPAN are separated from group III (Figure 3) with five sites where all sites (numbered as 18, 19, 37, 68, 70) were creeks. The species, H. chevreuxi, was only found in site 18 where the electrical conductivity value was over the freshwater range (EC 789 µS/cm). This may partially reinforce the idea that site 18 can be under the effect of increasing salinity. Six species (except H. incongruens) in group I have rare occurrences in creeks while species in group II have a wide distribution in several different habitat types characterized in freshwater ranges. Supporting evidence can be provided to generalize this result. Considering the dissolved oxygen level, a similar situation is applied to some of the species such as I. bradyi and P. fontinalis with low tolerance levels to dissolved oxygen as tk= 1.26 and tk= 1.4, respectively. One might reasonably think that since the two species have similar tolerance levels, they are in some sense ecologically equivalent. The same idea may be applied to other species with similar occurrences. Also, C. vidua showed a relatively low tolerance (tk= 1.99) level for oxygen tolerance. Kiss (2007) explained that a low density of C. vidua was related to its low tolerance value for poorly oxygenated waters despite its high ecological tolerance levels. Unlike these species, the optimum and tolerance levels of cosmopolitan species such as N. neglecta, H. incongruens, and H. salina were generally higher than the mean values for 12 species (Table 2). Therefore, such species can be considered to tolerate fluctuations in temperature changes and other ecological factors as well.
In conclusion, inland waters, especially the shallow water bodies, are critical for species conservation, biogeochemical cycling, and hydrological management even though they occupy an only a small portion of the Earth Zhang et al. 2020). However, because of current temperature changes elevated by human activities, ecological features of many aquatic habitats have been changed, causing a critical alteration in species composition and their geographic distribution Finlayson et al. 2013). Additionally, it is, however, expected that some ostracod species with low tolerance levels to ecological changes can be affected by such temperature changes earlier than those with high tolerance levels. To make an assumption about which ostracod species can be disappearing in the future (or which will have more chance to survive) is important in terms of the determining of inland waters dynamics due to changes in the structure of biological communities which affect freshwater ecosystem processes negatively (Dudgeon et al. 2006). Nevertheless, longterm studies are required to use ostracods as an early warning sign of changes in shallow-water habitats. and Dr. Özlem Makaroğlu (İstanbul University-Cerrahpaşa) for drawing topographic map of Hatay. Part of this study was presented within an oral presentation at the 17th International Symposium on Ostracoda in Rome, Italy.