Prey abundance supporting unusual water mite (Acari: Hydrachnidia) community in a sublacustrine spring and tributary river

Water mites are predators on microcrustations (Copepoda, Cladocera, Ostracoda) and insect larvae (mostly Diptera). Their communities usually have great diversity with high species richness but relatively low abundance. This may have important implications for poorly investigated predator-prey interactions of water mites in natural habitats. A sublacustrine karst spring Torak and the mouth of its tributary river Čikola were examined seasonally from August 2016 to June 2017, with the aim of determining water mite distribution connected to prey availability. Triplicate samples were taken with an Ekman dredge on three distinct depth zones (littoral, sublittoral and profundal zone) at both study sites. Both, the karst spring and the tributary river, had typical lentic water mite communities with no rhithrobiont or crenobiont taxa present. Water mite species richness, abundance and alpha diversity were not governed by depth, but rather by prey availability. Greater copepod, ceratopogonid and chironomid abundances were found to positively correlate with higher abundances of water mites. The population of Limnesia connata, previously reported from lime-poor waters exclusively, was most probably supported by vast prey availability, thus unexpectedly surviving in this karst environment. L. connata, Neumania vernalis and Piona longipalpis were recorded for the first time in Croatia, while L. connata was recorded for the first time in the Dinaric karst. #0#


Introduction
Predation is a biotic interaction that links water mites to different taxonomic groups and life stages of benthic invertebrate communities (Goldschmidt, 2016). Predatory aquatic adult and deutonymph water mites feed on insect larvae and other small invertebrates (Martin, Stur 2006;Stoch et al., 2011) such as microcrustations (Copepoda, Cladocera, Ostracoda) and insect larvae -mostly Diptera (Martin, 2005;Stoch et al., 2011). Their communities usually have great diversity with high species richness but relatively low abundance (Di Sabatino et al., 2008). This may have important implications for poorly investigated predator-prey interactions of water mites, mostly known from observations and ex-situ studies (Martin, 2005), while the distribution patterns and co-occurrence of water mites and their prey in natural habitats are poorly investigated.
Depth has been shown as an important abiotic driver in water mite composition and structure across lentic habitat profiles (Zawal et al., 2013). Concerning both depth and prey availability, the objectives of this study were: 1) to examine the effect of depth on water mite abundance, species richness and diversity and 2) to examine the dispersal of water mites in connection to prey abundance.

Study area
Torak is a sublacustrine karst spring situated on the southeastern bank of the upstream end of a Lakeland formed by the Čikola river near the Skradinski buk waterfall. The altitude is approximately 50 m above sea level. The submerged spring is circular, having a diameter of about 180 m. The eye of the spring is 47 m deep and is a tributary of the Čikola River ( Figure 1). The three sampling depths (littoral, sublittoral and profundal zone) were sampled at approximately 2, 10 and 45 meters.
The second sampling site Čikola is a Dalmatian river with the length of 47 km and is Krka's largest tributary. The Čikola River enters the Krka above Skradinski Buk. The lower reach of the river and the Torak spring lie within the boundaries of Krka National Park (Terzić, Frangen, 2010). At our sampling site, Čikola is at its lowest reach and is very slow flowing, even showing seasonal stratification. The three sampling depths (littoral, sublittoral and profundal zone) were sampled at approximately 2, 10 and 25 meters.
Both sites can be referred to as semi-lentic habitats with some flow present, but with stratification occurring regularly.

Sampling and laboratory methods
Each site was sampled four times (once per season) from August 2016 to June 2017. Triplicate samples were taken with an Ekman dredge (15 × 15 cm surface) on three distinct depth zones (littoral, sublittoral and profundal zone) at both sites. In total, we collected 72 benthic macroinvertebrate subsamples (36 from each site). All benthic samples of macroinvertebrate fauna were preserved in 96% ethyl alcohol. Water mites from the samples were subsequently isolated and fixed in Koenike's solution. All water mite individuals were determined to either species or genus (some deutonymphs) level using general keys by Tuzovskij (1990), Davids et al. (2007), Di  and Gerecke et al. (2016). All water mite specimens are deposited at the Department of Biology, Faculty of Science, Zagreb, Croatia. The number of specimens per water mite species were classified as Male/Female/Nymph.

Data analysis
We used Pearson's correlation coefficient to determine if a significant correlation could be detected between depth and water mite species richness, abundance and diversity. We used five different indices at each site, as they all have slightly different approaches when calculating alpha diversity.
We used Pearson's correlation coefficient to determine significant correlations between diversity indices and water mite abundance with prey availability.
All values were log(x + 1) transformed before testing. For all tests, p < 0.05 indicated statistical significance. All statistical analyses were done using Statistica, 13.0 (TIBCO Software Inc., 2017).

Results and discussion
In total, only thirteen water mite specimens from four taxa were collected during the study period, proving once again the diverse, but not abundant nature of water mite distribution. Both, the karst sublacustrine spring and tributary river, had typical lentic water mite communities with no rhithrobiont or crenobiont taxa present. The most abundant taxon was Neumania vernalis, recorded in both sites with eight specimens found (1/1/6). Arrenurus sp. (0/0/2) and Piona longipalpis (0/1/0) were found only in Čikola river, while Limnesia connata was found solely in Torak (1/1/0).
Water mite specimens were not found in the profundal zone. The thermocline, as found by Zawal et al. (2013) seemed to pose an obstacle for water mite colonisation of the deeper layers of our two study sites. However, water mite species richness, abundance and alpha diversity were found not to be governed by depth. Neither specific abundances nor indices based on water mite assemblages were found to correlate with depth significantly (Table 1). Greater copepod, total microcrustacean, ceratopogonid and chironomid abundances were found to significantly correlate with higher abundances of water mites, indicating that water mite spatial distribution was governed by prey availably, rather than depth (Table 2).

New records for the water mite fauna of Croatia
A checklist of water mites in Croatia was published by Pešić (2002), which was later on broadened within the research on water mites of the Balkan Peninsula (Pešić et al., 2010(Pešić et al., , 2018. These check lists are missing records from Matoničkin and Pavletić (1959) (Woolastookia rotundifrons) and a doctoral thesis from Romana Latiger (1988) in which groundwater water mites were determined by Prof. Schwoerbel (Partnunia angusta, Panisus torrenticolus and Protzia squamosa). These make up a total of 64 water mites recorded in Croatia, and with the current paper three new species are added, making the total number of water mite species found in Croatia 67.
Limnesia (Limnesia) connata Koenike, 1895 Only two specimens found in the sublacustrine spring confirms the records of van Haaren and Tempelman (2009) who stated that this species is widely distributed, but rather scarce in abundance and is not numerous anywhere. A discrepancy was found with other recordings of the habitat preferences of this species such as: lime-poor waters, quagfens and temporary waters (Schwoerbel, 1956;Smit, van der Hammen, 1996). Two adult specimens (1/1/0) were found in the littoral zone in winter. The population of Limnesia connata, previously reported from lime-poor waters exclusively, was most probably supported by vast prey availability, thus unexpectedly surviving in this karst environment.
Neumania (Neumania) vernalis (Müller, 1776) In total eight specimens were found (1/1/6) at both sampling sites. The species was found in spring and autumn, in the littoral and sublittoral zone. The findings of Pešić et al. (2007) that stated: "the species is common in eutrophic to dystrophic standing waters" is in slight discrepancy with the findings of this study where the species was found in both the lower reaches of Čikola river and the sublacustrine karst spring Torak, which we consider to be semi-lentic habitats.
Piona longipalpis (Krendowskij, 1878) This ubiquitous species was found in the sublittoral zone of the slow flowing, lower reaches of Čikola river. Only one specimen was recorded (0/1/0) in autumn. The species is not habitat specific and is widespread in Europe (Gerecke et al., 2016).