Co-occurrence and salinity
With the additional reports in here, numbers of non-marine ostracods in Bursa province increased up to 33 species. This number is more than many other provinces in similar size in Turkey (Külköylüoğlu et al 2021b), but it is still considered underestimation since previous reports are based on random and/or mostly one-time sampling effort that there is no extensive study on the ostracod fauna of the province. Meanwhile, looking at the abundance of individual species, it is clearly seen that C. torosa is the only dominant species encountered along with nine other recent taxa (Table 1). Of which, four candonids (C. angulata, C. meerfeldiana, C. cf. lindneri, Candona sp.) were the most accompanying taxa with C. torosa followed by others (Cypria sp., Heterocypris salina, Eucypris sp., Potamocypris sp., Plesiocypridopsis sp.). It is already known that some of these taxa can be found from fresh to saline habitats together with C. torosa (Meisch 2000; Scharf et al 2017; Pint and Frenzel 2017; McCormack et al 2019). However, comparison of the abundance values amid taxa portrayed the fact that C. torosa was generally over numbered (> 98%) during our study. Plotnikov et al. (2021) reported that C. torosa was the most common species with long-term tolerance to salinity increase in Aral Sea. These authors found different occurrence frequencies of the species between Large Aral (salinity range 8-13 ppt) and Small Aral (salinity range 0-3 ppt). The species was the last survivor during salinity increase in the lake. However, they argued that increasing salinity can even cause extinction of C. torosa in Large Aral but the species present in Small Aral, implying that even C. torosa has some upper limits of salinity tolerance. During the present study, there are, however, variations of the species occurrence patterns among the stations. For example, one possible explanation of such occurrence may be considered that station 1 receives a water discharge from a small creek (Figure 1) but it is also faced with a seasonal sea water intrusion when sea water level rises over the narrow coastal barrier into the lake area (Dalkıran N., pers. comm). Thus, one may argue that this water back up from the creek may reduce salinity (also referring to electrical conductivity) at that sampling point lower than what C. torosa prefers although the species can tolerate wide ranges of salinity (and temperature) values. In a microcosm study, Frenzel et al. (2012) showed highest numbers of reproductive rates at the salinities ranged from 3 to 8 psu while noded valves being most abundant below 2 psu were also found up to 7 psu. The authors reported smooth valves above the limit 7 psu. Accordingly, their study showed similar trend for both males and females. In our case, however, salinity ranges between the station 1 (0.16 and 0.69 ppt) and the station 6 (0.21 and 0.47 ppt) overlap. Therefore, insignificant level of salinity range does not explain why C. torosa was relatively abundant in station 6 but station 1.
Most recently, Pint and Frenzel (2017) proposed a flowchart for paleoenvironmental interpretation based on the species dominancy. Hence, if dominancy of the species is more than 90%, the habitat can be characterized as hypersaline or with oxygen deficiency. In contrast, dominancy with less than 90% refers to fresh to brackish waters. Although their application is suggested to use fossil occurrences of the species, it seems that it can also be used to determine habitats with present conditions. We collected C. torosa from the stations (but cf. station 1) with more than 90% of dominancy almost all year round. This finding suggests that the delta is of hypersaline conditions, but this does not support oxygen deficiency due to a relatively high mean oxygen value (ca. 7.16 mg/L).
Mg, Ca And Noding
This explanation above may have a value since salinity, Mg and Ca measurements were significantly different between stations (1st and 6th ) and others (2-5, 7th ) where the species exhibited seasonal occurrence patterns with high abundance. No significant difference (P > 0.05) was found for other variables. Station 6 is located on the Çapraz River which flows continuously through Marmara Sea but intrusion from the sea occurs seasonally. Thus, its water is mixed all the time where both smooth and smooth-noded individuals were collected during the study. Both elements are necessary for the carapace formation while Ca is generally higher than Mg in the carapace. However, with a few exceptions, Mg values of the stations were found almost always much higher than Ca during the present study. These differences were apparent between two stations (1 and 6) which were the furthest in distance to the Marmara Sea. Indeed, we found C. torosa from the known ranges of these variables obtained in the literature. What is however imperative is to associate species frequent occurrences amid the stations with or without (or both) noded carapaces. As mentioned, carapace morphology seems to be related to salinity (and temperature) changes in waters that noded individuals tend to be found more commonly in freshwaters than brackish or saline waters. In addition to these variables, however, previous studies (Keyser 2005, Frenzel et al. 2012) pointed out that node formation might also be correlated to deficiency of Ca level, suggesting that numbers of nodes can be increased in the waters with low Ca. Although the correlation was medium and not significant (Table 3), our results support the opposite of this view that the mean Ca level (197.85 mg/L) was the lowest among other groups where there were only individuals with smooth carapaces (Table 2). While working on another species (Limnocythere inopinata) in Lake Van (Turkey) known with Ca limitation (0.105 to 0.087 mmol/L) (Reimer et al 2009), similar finding was outlined by McCormack et al. (2019) that node formation may be influenced by several other factors that Mg may be one of them. Our values are clearly much higher than these values and apparently good enough to build carapace structure. However, again, this does not really explain absence (except one female) of C. torosa at station 1 although its chemical composition is similar to station 6 where the species was relatively higher in numbers and in their occurrences.
Several studies (Meisch 2000) showed that some species and/or genera can be associated to lower salinity ranges. For example, finding members of the genus Candona from station 1 may support this view due to their freshwater habitat preferences with low salinities (Neale 1988; Karanovic 2012) but we are still not able to answer why C. torosa was not found and/or was not common in there. This question is important because some taxa reported in here (e.g., Heterocypris salina, Eucypris sp., Plesiocypridopsis sp., Cypridopsis sp.) are already known to survive in wide ranges of salinity, temperature and/or pH values (Delorme 1991). As indicated in their excellent review, Dettman and Dwyer (2012) clearly underlined that there can be several other factors effective on carapace chemistry and structure. Hence, there is no single explanation about the relationships between the formation of nodes on the carapace and Mg, Ca and/or Mg/Ca in waters. On the other hand, Figure 3 suggests that it is electrical conductivity closely related to species occurrence/abundance more than Mg and Ca alone. Moreover, our results with Mg and node formation tend to support similar explanation used for Ca where individuals without nodes were solely found below the mean (555 mg/L) of Mg level.
Temperature, Seasonality And Noding
Herman et al. (1983) showed that C. torosa has one generation that several factors can be effective on its life cycle and occurrences. For example, salinity increase can be directly intimated with temperature. This is actually the case for C. torosa. Heip (1976a, b), after more than four years of his continuous work, illustrated that abundance and occurrence of the adults were triggered and were closely related to water temperature above 15 ⁰C. Our results are mostly supportive on this approach with a few exceptions in some months (see Table 1) where adults are high in numbers below this proposed temperature level. For example, in total, there were more adults at the station 3 during January and February 2019 where water temperature was 6.03 and 13.2 ⁰C, respectively. In contrast, a medium correlation between water temperature and abundance of the species was not significant. Nevertheless, this does not change the general view proposed by Heip (1976a) that numbers of adults increase with increasing temperature (and salinity) but this should be investigated in detailed studies.
On the other hand, relating the temperature to monthly occurrences of the noding, it is apparently valuable to indicate that occurrence of adults without nodes are mostly beginning within the end of fall season (November) until spring season (April). Similarly, the individuals without nodes (but with a few exceptions) were reported all year around from a eutrophic lake, Lake Küçükçekmece (Turkey) (Külköylüoğlu et al. 1993). In another monthly study, however, Külköylüoğlu et al. (1995) reported a similar pattern of the noded and smooth individuals of C. torosa from a brackish water lake (Lake Büyükçekmece) (now the lake is of freshwater characteristics due to separation from the Marmara Sea in 1985) in summer (June) and winter (December) seasons. In both studies, authors failed to measure salinity values of the lakes but Külköylüoğlu et al. (1995) underlined that node formations might be a necessary issue for the species because it probably helps the species movement on the sediment in freshwater conditions while the species may not need the nodes in saline waters due to lifting force. These authors did not ask a specific question about the correlation between nodding and salinity in the study. Additionally, these explanations may not represent true nature of the correlation between node formations and water chemistry. However, they help to deduce an understanding of it. Nevertheless, as shown in previous studies (see above), node formations are possibly a response to environmental factors.
Ph, Alkalinity And Noding
Alkalinity was suggested as an effective factor on the carapace structure and formation of nodes (Van Harten 2000; McCormack et al 2019); for instance, De Deckker and Lord (2017, p.4) stated that “…It is unfortunate that neither Vesper nor Heip measured alkalinity of the waters during their long investigations of the life cycles of torosa, and this needs to be examined in the future so as to better understand ostracod shell composition. Alkalinity, combined with ionic analysis of the ambient waters will lead to identification of the calcite saturation nature of the waters in which ostracods moult and grow.” We did not measure alkalinity during the present study, but pH values were measured. Moreover, we are aware of that pH and alkalinity of waters are not same, but they are closely related (Boyd et al 2017). Implication of this relationship is that increasing pH values (> 7, referring to alkaline or basic waters) means high alkalinity. In a very comprehensive work of Boyd et al. (2017), this relationship in waters is provided as: pH = 6.6, alkalinity = 1 mg/L; pH = 7.3, alkalinity = 5 mg/L; pH = 7.6, alkalinity = 10 mg /L; pH = 8.3, alkalinity = 50 mg/L. This information may be applied to the studies; for example, C. torosa was reported in the waters of Terschelling Island where pH values were measured between 7.5 and 8.5 (Scharf and Hollwedel 2010). Implication is that alkalinity was at least 10 or more in the waters of the island. During the present study, we have 77 pH measurements. There are only 16 of 77 cases where pH values were below 8.0. Of which, there are only three cases (pH = 7.84, 7.92, and 7.96) where we identified live C. torosa (first two with noded individuals and the last one with smooth individuals, respectively) while we found no ostracods or only valves/carapaces in six and seven cases (mostly smooth and noded-smooth individuals but no single population with solely noded individuals found), respectively. Rest of the cases (61 of 77) includes pH values ≥ 8.0. Adapting the equations of Boyd et al. (2017), we may link the pH values (now referring to alkalinity values above) to the noding on carapaces. The mean pH values (8.14-8.47) among the stations did not show significant difference but it can be inferred that the species may prefer waters with alkalinities above 10 mg/L or even 50 mg/L. This can be useful information provided in here for the first time that such a view may be used in fossil forms for understanding past environmental conditions in paleontological studies.
Fossil Vs Recent (Live) Forms
In Turkey, C. torosa was reported from Early Miocene (Ilgar and Nemec 2005) corresponding to the previous records (cf. Van Harten 2000; Witt 2010; Wouters 2017). When we compare dispersion of the fossil and live species reported so far (Figure 2), numbers of fossil records from about 24 provinces (aka cities) are higher than living specimens in 20 provinces. With a few exceptions (Figure 2), living forms have been mostly coupled with fossil records reported from nearby the coastal zones of west and northwest (around Marmara Sea) of Turkey. Although there are extensive studies in some provinces (e.g., Sinop, Çankırı, Eshişehir, Elazığ, Konya), which includes about 1000 water samples, there are only surface sediemeant samples of (sub/fossils) C. torosa populations reported from them. Last four of these cities (and more others, see the Figure 2) are far away from the seas and are located within Anatolia where fossils were found in several different water bodies. Two other similar proxies can be worth to discuss: First, C. torosa with smooth and noded individuals were reported from Holocene samples of the Lake Sevan (Armenia) (Wilkinson et al 2005). The lake is located at 1900 m asl and has no connection to seas. The authors pinpointed those smooth forms were encountered in a Holocene sequence more than noded forms, implying that the lake salinity had been increased during at least the last 5000 years or late Holocene. Second, similarly, in Germany, Scharf et al. (2017) reported Quaternary fossils of C. torosa from 32 of 45 inland sites far away (more than 200 km) from the Baltic and the North seas. Opposite situation is also true for live populations with a few cases. There can be at least three possible ways to delineate this situation (1) lack of studies, (2) unsuitable habitats for the species, and (3) no time for the species migration yet. On the other hands, we believe that such a map showing overlapping ranges of both fossil and live forms can help us to understand species distribution since the Early Miocene in Turkey.
Overall, in conclusion, as stated above, alkalinity was not directly measured in situ, we cannot provide a good explanation for its correlation with noding of the carapace. However, we agree that a combination of salinity and/or alkalinity with other environmental variables and biotic variables may be a better way to apply in future studies. Indeed, total nitrogen (and phosphorous) portrayed medium correlation (P > 0.05) to the species abundance among the stations. Consulting Figure 1 and site description above, one can recognize agricultural activities or so called “human activities” around the study area. According to Chen et al. (2015), global distribution of TN and TP values in lakes can be found between 0.526 mg/L and 0.014 mg/L. Our mean values are all higher than these values (cf. Table 2). This implies possible sources of nitrogen and phosphate and their compounds reaching to the sampling sites due to human activities. It appears that C. torosa can even overcome all these artificial inputs due to its high tolerance ranges. As indicated by Frenzel et al. (2012), C. torosa can be used as a good indicator species because of the populations inhabiting or preferring a wide range of salinities. For example, individuals of the athalassic populations from stable water bodies can be used to describe continuous and detailed water bodies. Although this finding is of a scientific merit, we cannot make detailed discussion in here about these compounds due to lack of studies.