River Habitat Survey: Does This Help to Explain the Nature of Water Mite (Acari and Hydrachnidia) Assemblages?

Abstract

In the European Union, assessments of the quality of the aquatic environment based on aquatic invertebrates are mandatory. Biological methods are supplemented with hydromorphological assessments of watercourses. There are many studies analysing the relationships between aquatic invertebrates and the hydromorphological assessment of the environment by the River Habitat Survey (RHS) method, but thus far, there has been no detailed study including water mites (Acari, Hydrachnidia) and the application of this method. In the present study, the following research hypothesis was put forth: a hydromorphological characterization of habitats is a significant element explaining the nature of water mite communities, and the RHS method can be used to predict the characteristics of Hydrachnidia populations in a river. The research was carried out in a small lowland river, the River Krąpiel (north-western Poland). Six locations were selected as representative of some state of habitat modification for the assessment of the hydromorphological conditions of the river and the collection of biological samples. The following conclusions can be drawn from the research: (1) the biology and ecology of water mites make them suitable as bioindicators of the environment, including hydromorphological modifications, and (2) the hydromorphological characteristics of habitats explain the nature of water mite communities in the river at the level of general population parameters (number of specimens and species), while at the species level, general regularities in water mite fauna distribution in river ecosystems, the continuity of the river ecosystem, and characteristics at a smaller spatial scale (habitat scale) better explain water mite community structure than the hydromorphological indices determined for a given site or section of the river.

1. Introduction

Aquatic macroinvertebrates are commonly used to evaluate water quality and, more broadly, the quality of the aquatic environment all over the world [1,2,3,4]. In the EU, the European Union Water Framework Directive requires assessments of the aquatic environment based on aquatic invertebrates [5]. At the same time, the Directive indicates that biological methods based on aquatic macroinvertebrates should be supplemented by the use of physicochemical analyses of water and by hydromorphological assessments of watercourses. The mention of hydromorphological assessments as essential to a comprehensive evaluation of environmental quality is important because hydromorphological changes are among the most significant factors influencing invertebrate communities in European rivers [6,7,8].

The most common method of hydromorphological assessment is the River Habitat Survey (RHS), a method originally developed in the UK [9,10]. It makes use of multiple environmental parameters, which can then be presented in the form of two synthetic indices: the Habitat Quality Assessment (HQA) and the Habitat Modification Score (HMS) [10]. This method is commonly used in many countries in its original form or is sometimes modified by introducing additional hydromorphological parameters [11] or new elements of the assessment, enabling a more complete reflection of the local conditions encountered in the fluvial systems of a given country [12]. The RHS focuses on abiotic factors, such as river bed morphology, the occurrence of natural elements in the river bed, the material of the banks and bottom, and flow types. The only biotic element taken into account is the aquatic and riparian vegetation [10]. The RHS does not include other biotic elements, including invertebrates.

As macroinvertebrates are the most important element of biological assessments of water quality in the EU, and it is recommended that the biological assessment of environment quality should be complemented via an analysis of hydromorphological elements [5], there is a need to examine and understand the links between abiotic environmental parameters and aquatic biota [13]. There was and is a need to combine biological methods based on the responses of aquatic invertebrate communities with the River Habitat Survey in order to carry out a biological validation of the latter method, which focuses on the physical factors of the environment [14]. Many papers have analysed the relationships between aquatic invertebrates and various hydromorphological factors [15,16,17,18,19,20]. Other studies have investigated the relationships linking macroinvertebrate communities to methods of hydromorphological assessment based on the RHS [21] or directly to the RHS method, which provides a comprehensive assessment of the hydromorphological state of a river [8,11,14,22,23,24,25,26,27,28,29,30,31].

Integrated studies of invertebrates in combination with the River Habitat Survey method have shown that the RHS method not only provides a detailed hydromorphological characterization of habitats but is also an important tool, helping to explain patterns of macroinvertebrate distribution in rivers [22,29,30]. According to some authors [29,30], methods of environmental assessment based on invertebrates alone sometimes fail to fully explain the ecological state of a habitat. Only supplementation of the biological assessment with other methods, including the River Habitat Survey, provides a complete picture of the ecological state of a habitat and explains changes in communities of aquatic invertebrates.

Studies dealing with the RHS and macroinvertebrates make use of various benthic invertebrates, most often insects [8,24,29,30]. Analyses combining the RHS with invertebrates completely overlook an important element of the invertebrate communities of lotic environments—water mites. Water mites (Hydrachnidia) are invertebrates commonly found in a variety of aquatic environments, including lotic environments. They are often numerous in flowing water, at high densities, and with high species diversity [32,33,34]. In rivers, they can be the dominant element of multi-taxon macroinvertebrate communities in terms of the number of species and individuals [35]. Water mites play an important role in food webs as predators (adult forms and deutonymphs) and as parasites (the larvae of most species) [34,36]. The main food of water mites in flowing water bodies is Chironomidae larvae [37]. Water mites themselves have a negligible role in the diet of other organisms, including fish [38,39], so their abundance does not change significantly as a result of predation pressure. Water mites are sensitive to water pollution, and throughout their complex life cycle, they are also susceptible to other anthropogenic impacts, such as habitat modification, which affects their food base and host species (see [40] for a review). These characteristics make water mites very useful organisms for assessments of the ecological status of the environment. Despite this, they are overlooked in both biological assessments of lotic ecosystems and analyses combining invertebrates with a comprehensive hydromorphological assessment of rivers using the RHS method. The only study in which water mites are specified as an element of multi-taxon communities of fauna analysed in the context of the RHS method is a paper by Czerniawska-Kusza and Szoszkiewicz [41]. In that paper, however, water mites are treated as the group Hydrachnidia as a whole, with no detailed analysis at the genus or species level. Therefore, it can be said that thus far there has not been a single detailed study presenting an analysis of Hydrachnidia communities in connection with a hydromorphological characterization of habitats using the River Habitat Survey method.

The aim of the present study was to carry out a biological validation of the RHS method with respect to water mites by i) analysing Hydrachnidia communities in connection with a comprehensive hydromorphological characterization of habitats performed by means of the River Habitat Survey, and ii) determining whether the RHS method and hydromorphological parameters explain patterns of water mite species distribution. The following research hypothesis was put forth: a hydromorphological characterization of habitats is a significant element explaining the nature of water mite communities, and the RHS method can be used to predict the characteristics of Hydrachnidia populations in a river.

2. Materials and Methods

2.1. Study Area, Sites, and Field Work

The research was carried out in the small lowland river, the River Krąpiel (north-western Poland). The River Krąpiel is classified as a medium river [5]. The water body codes for this river are PRLW 600009198833 (from the sources to Kania) and PRLW 600011198899 (from Kania to the empty into the Ina River). Other information about the River Krąpiel and a description of the study area can be found in Zawal et al. [42].

Six locations for assessments of the hydromorphological conditions of the river and for the collection of biological samples were established along the course of the river (Figure 1): site 1 (S1) (53°28′10.63″ N 15°21′41.79″ E), S2 (53°27′36.97″ N 15°16′33.2″ E), S3 (53°27′41.47″ N 15°12′22.94″ E), S4 (53°21′6.4″ N 15°11′5.23″ E), S5 (53°20′29.56″ N 15°9′15.04″ E), and S6 (53°19′58.14″ N 15°7′57.54″ E).

Each sampling site was selected as representative of a certain state of habitat degradation and/or modification. Site 1 can be described as 80% natural. The modification of this site involved the placement of large stones in the river bed just beyond the culvert. This resulted in a change in flow from laminar to turbulent and gave the habitats a pseudo-mountainous character (Figure 2a). At site 2, the river had undergone severe anthropogenic alterations and took the form of a straight, deepened channel (Figure 2b). Site 3 is located on a natural stretch of the river, which flows through an alder carr. In some places, it is like a mountain river, with some parts having a laminar flow and others with a turbulent flow (Figure 2c). Site 4 was anthropogenically modified, but long ago, and it is currently in an advanced restoration process (Figure 2d). Site 5 (Figure 2e) is located on a natural stretch of the river. There are anthropogenic modifications at site 6: regulation of part of the channel located near the road and the presence of a bridge that raises the water level (Figure 2f).

A hydrological assessment of the hydromorphological conditions was made for each site (S1–S6) using the standard River Habitat Survey (RHS) method, a technique ensuring that the results are comparable with those of other studies [12]. The RHS methodology was modified somewhat for the purposes of this study: assessments were made for stretches of 100 m rather than the standard 500 m length of river channel. The habitat quality assessment (HQA) and habitat modification score (HMS) were calculated on the basis of the fieldwork [12]. Two other hydromorphological indices were used as well: river habitat quality (RHQ) and river habitat modification (RHM) [11]. On the basis of the values of the two main indices, HQA and HMS, the final classification of the hydromorphological state of each sampling site was performed by assigning the sites to one of the five RHS classes [25].

Water mites were collected in May, July, September, and November 2011. To obtain complete faunistic data and ensure that the biological material was compatible with the methodology of hydromorphological data collection, the biological samples were collected from riffles and pools (hereafter referred to in the text as habitats). In the case of both riffles and pools, all environments present at a given site were included, taking into account the type of bottom sediment and the degree of bottom cover by aquatic plants. A total of 27 habitats (16 riffles and 11 pools) were distinguished at the sites. The number of habitats was not the same at each site, and ranged from 3 (sites 2 and 5) to 6 (sites 1 and 6). A complete list of the habitats with their descriptions is given in

Table 1. List of habitats within sampling sites with their characteristics. S/h—sampling site/habitat; R—riffle; P—pool.

S/h	Type	Depth	Flow [m/s]	River Width	Organic Matter	Mineral Matter	Surroundings	Bottom	Vegetation/Remarks
		[m]		[m]
S1/1	R	0.2	0.16–0.18	1.0	0.55–0.58	99.42–99.45	alder carr	sand
S1/2	R	0.2	0.38–0.47		0.11–0.46	99.54–99.89		stones
S1/3	P	0.5	0.001–0.003		2.49–2.78	97.22–97.51		sand, silt, mud
S1/4	R	0.2	0.48–0.62		0.01–0.10	99.90–99.99		rocks	mosses, algae
S1/5	R	0.2	0.24–0.31		0.35–0.43	99.57–99.65		gravel, sand
S1/6	P	0.4	0.005–0.01		0.87–1.23	98.77–99.13		sand, silt	absence of water mites
S2/1	R	1.2	0.35–0.42	4.0	0.45–0.58	99.42–99.55	alder carr, willow thickets	gravel, sand	absence of water mites
S2/2	P	0.5	0.02–0.04		32.70–36.30	63.70–67.30		silt, mud	Phragmites australis (Cav.) Trin. ex Steud.
S2/3	P	0.5	0.05–0.07		1.41–1.84	98.16–98.59		silt, mud
S3/1	R	0.2	0.2–0.36	7.0	1.21–2.49	97.51–98.79	alder carr, willow thickets	gravel, sand, mud, leaves
S3/2	R	0.5	0.41–0.57		0.15–0.19	99.81–99.85		gravel	Sparganium emersum Rehmann absence of water mites
S3/3	R	0.7	0.31–0.36		0.01	99.99		rocks, stones	Fontinalis antipyretica Hedw.
S3/4	P	0.2	0.05–0.06		1.58–2.23	97.77–98.42		sand, mud
S3/5	P	0.1	0.01–0.02		1.23–1.31	98.69–98.77		sand, mud	Mentha aquatica L., Carex acutiformis Ehrh.
S4/1	R	0.4	0.28–0.38	5.0	1.78–2.89	97.11–98.22	alluvial forests with Alnus glutinosa and Fraxinus excelsior	sand, leaves	grasses
S4/2	P	0.2	0.1–0.12		32.20–35.30	64.70–67.80		mud	Glyceria maxima (Hartm.) Holmb.
S4/3	R	0.5	0.12–0.18		0.82–0.88	99.12–99.18		sand, gravel
S4/4	P	0.2	0.01–0.02		1.13–1.59	98.41–98.87		sand, mud	Sagittaria sagittifolia L.
S4/5	R	0.4	0.2–0.22		0.59–0.78	99.22–99.41		sand, mud	Sagittaria sagittifolia L.
S4/6	R	0.7	0.27–0.31		4.50–4.84	95.16–95.50		sand, mud
S5/1	R	0.7	0.2–0.4	10.0	0.98–9.39	90.61–99.02	oak–hornbeam stands	sand, leaves
S5/2	P	0.2	0.01		21.00–23.40	76.60–79.00		sand, mud
S5/3	R	0.5	0.43–0.57		0.73–0.76	99.24–99.27		gravel, stones
S6/1	R	0.7	0.18–0.35	10.0	0.29–0.35	99.65–99.71	alder carr	sand
S6/2	P	0.2	0.07–0.12		0.88–1.10	98.90–99.12		sand, mud	sedges
S6/3	R	0.7	0.34–0.54		1.35–3.63	96.37–98.65		rocks
S6/4	P	0.4	0.05–0.16		0.28–1.25	98.75–99.72		sand, mud

. A total of 108 biological samples were taken (one sample from each habitat within the sampling sites, 4 times during the year).

In addition to gathering of hydromorphological data (RHS procedure) and biological data (collection of water mites), the following physicochemical properties of the water were measured: temperature (°C), pH, electrolytic conductivity (µS/cm), dissolved oxygen (mg/dm₃) (those parameters were measured with a CX-401 multifunction meter), nitrate nitrogen (NO₃, mg/dm³), ammonia nitrogen (NH4, mg/dm³), phosphates (PO₄, mg/dm³), ferric ions (Fe, mg/dm³), turbidity (mg/dm³), total hardness (mg CaCO₃/dm³) (those parameters were measured with a Slandi LF205 photometer), and BOD₅ (Winkler’s method). Some other environmental parameters were included as well: density of aquatic vegetation (using a scale from 0 to 5, where 0 indicates the absence of plants and 5 means total overgrowth by plants), insolation (%) (measured with a CEM DT-1309 light meter), water flow (measured with a SonTek acoustic FlowTracker flowmeter), proportion of mineral sediment (%), proportion of organic sediment (%), mean sediment grain size (M, mm), and sediment sorting (W). The methodology for determining the last four parameters is given by Zawal et al. [42].

2.2. Laboratory Work

The biological material collected in the field was transported to the laboratory and sorted in white laboratory trays. Water mites were separated out from the samples. The following works were used for the identification of water mites: Davids et al. [36], Di Sabatino et al. [43], and Gerecke et al. [44]. Adult forms were identified by species, and deutonymphs were identified by genus. Species nomenclature and systematics were adopted according to Davids et al. [36], Di Sabatino et al. [43], and Gerecke et al. [44].

On the basis of the literature [36,43,44,45,46,47], the water mite fauna was divided into four synecological groups: rheobionts, rheophiles, crenobionts and crenophiles, and lenitobionts.

2.3. Statistical Analyses

Similarities between parameters were determined using Biodiversity Pro v.2 software [48]. The Euclidian distance formula was used for abiotic parameters (RHS indices) and the Bray–Curtis formula for faunal similarities (water mite fauna). The complete link clustering method was used. Biodiversity Pro v.2 was also used to analyse the co-occurrence of species using quantitative data (Bray–Curtis formula). Analyses of species co-occurrence took into account the sets of species included in the CCAs (see below).

The normality of the data distribution was checked by the Shapiro–Wilk test. The Spearman correlation coefficient (r_S) or Pearson correlation coefficient (r_xy), depending on the data distribution, was used to calculate the correlations. The Mann–Whitney U test (Z) was used to compare two independent samples. The Kruskal–Wallis test (H) was used to compare multiple independent samples. All tests were carried out in Statistica 13.1 software. The statistical significance level was set at p < 0.05.

A Canonical Correspondence Analysis (CCA) was used to determine relationships between the occurrence of water mite species and environmental parameters at the sampling sides. The data for the calculations were not transformed [49]. To test the significance of the environmental variables (p < 0.05), forward selection (FS) was used with Monte Carlo permutation tests (499 unrestricted permutations). FS analyses were run until no more variables remained to improve the fit. CCAs were performed using CANOCO 5 statistical software [50]. Only individuals identified to species were included in the ordination analyses. To avoid skewing of the ordination due to species occurring in small numbers, species represented by fewer than 5 individuals were discarded from the CCAs, so that the analysis was performed on a set constituting 96.1% of the total material and including 17 of the 35 species identified.

3. Results

3.1. Hydromorphological Assessment

The values of the hydromorphological indices indicate that the river was most natural (HQA = 57) and least modified (HMS = 0) at site 5, where it was assigned to the first (best) RHS class (

Table 2. The values of hydromorphological indices and RHS classes at each study site.

Study Site	HQA	HMS	RHQ	RHM	RHS Class
S1	53	24	135,50	14,5	4
S2	22	34	86,31	35,2	5
S3	55	0	131,23	0	2
S4	47	13	137,85	5,8	3
S5	57	0	125,35	0	1
S6	56	15	143,94	25,4	3

). The favourable hydromorphological characteristics of the river at this site are underscored by the high river habitat quality index (RHQ = 125.35) and low river habitat modification index (RHM = 0). The lowest hydromorphological variation (HQA = 22) and highest degree of anthropogenic modification (HMS = 34) were noted at site 2. Based on the values obtained for the HQA and HMS indices, this site was assigned to the fifth (worst) RHS class. The other two indices (RHQ = 86.31, RHM = 35.2) underscore the low quality of this site in terms of its hydromorphological characteristics.

Figure 3 presents a dendrogram grouping the sampling sites on the basis of hydromorphological indices. Study site 2 is clearly separated from the other sites. Within the remaining sampling sites, two groups can be seen. The first comprises sites 1, 4, and 6, belonging to the third and fourth RHS classes (Table 2), while the second group consists of sites 3 and 5, belonging to the first and second RHS classes (Table 2).

3.2. Water Mite Communities

There were 742 water mite individuals belonging to 35 species caught in the River Krąpiel (

Table 3. Number of water mite species and individuals collected from the River Krąpiel. S1–S6—sampling sites; R—riffle habitats within the site; P—pool habitats within the site. SG—synecological group; rb—rheobionts; rh—rheophiles; lb—lenitobionts; cr—crenobionts and crenophiles.

	Species	SG	S1		S2		S3		S4		S5		S6		Total
			R	P	R	P	R	P	R	P	R	P	R	P
1.	Eylais degenerata Koenike, 1897	lb						1							1
2.	Parathyas palustris (Koenike, 1912)	cr	9												9
3.	Hydryphantes hellichi Thon, 1899	lb						1							1
4.	Lebertia fimbriata Thor, 1899	rb									1		6		7
5.	Lebertia inaequalis (Koch, 1837)	rf										1	2		3
6.	Lebertia insignis Neuman, 1880	rf							2			1		4	7
7.	Lebertia oblonga Koenike, 1911	rb									4		2	6	12
8.	Lebertia porosa Thor, 1900	rf							3	4	3	1	5	11	27
9.	Lebertia pusilla Koenike, 1911	rb												2	2
-	Lebertia sp. Neuman, 1880	-							1		2	1		4	8
10.	Sperchon clupeifer Piersig, 1896	rb	62	4			3		3		4				76
11.	Sperchon setiger Thor, 1898	rb											2		2
12.	Sperchon thienemanni Koenike, 1907	cr	2												2
-	Sperchon sp. Kramer, 1877	-								1			1		2
13.	Sperchonopsis verrucosa (Protz, 1896)	rb					1								1
14.	Torrenticola amplexa (Koenike, 1908)	rb					1		35	1	58	8	31	30	164
15.	Torrenticola anomala (Koch, 1837)	rb							2		1		3		6
16.	Torrenticola barsica (Szalay, 1933)	rb									43				43
17.	Torrenticola brevirostris (Halbert, 1911)	rb									1		14	7	22
18.	Torrenticola dudichi (Szalay, 1933)	rb									14		3		17
19.	Torrenticola similis (K. Viets, 1939)	rb											1		1
-	Torrenticola sp. Piersig, 1896	-							2		3		3		8
20.	Albia stationis Thon, 1899	rb										1		1	2
21.	Aturus scaber Kramer, 1875	rb							2						2
22.	Parabrachypoda montii (Maglio, 1924)	lb								2					2
23.	Atractides nodipalpis Thor, 1899	rb							18	1					19
24.	Atractides neumani (Lundblad, 1962)	rb							1						1
25.	Hygrobates fluviatilis (Ström, 1768)	rf							1				1		2
26.	Hygrobates longipalpis (Hermann, 1804)	rf				1			4		2				7
27.	Hygrobates longiporus Thor, 1898	rf												1	1
28.	Hygrobates nigromaculatus Lebert, 1879	lb												5	5
29.	Hygrobates setosus Besseling, 1942	rf					2	12			1	32	1	20	68
-	Hygrobates sp. Koch, 1837	-												5	5
30.	Limnesia maculata (Müller, 1776)	lb				1									1
31.	Forelia variegator (Koch, 1837)	lb						1							1
-	Piona sp. Koch, 1842	-				1		1							2
32.	Neumania callosa (Koenike, 1895)	lb										1			1
33.	Neumania papillosa (Soar, 1902)	rf												2	2
34.	Mideopsis crassipes Soar, 1904	rf							8	12	4	71	6	58	159
35.	Mideopsis orbicularis (Müller, 1776)	lb						1	1	5		21	3	10	41
	In-habitat specimens		73	4	0	3	7	17	83	26	141	138	84	166	742
	species		3	1	0	2	4	5	10	6	12	9	14	13
	At-site specimens		77		3		24		109		279		250		742
	species		3		2		6		12		15		20

). More water mites were caught in riffle habitats (388 individuals) than in pools (354), but the differences were not statistically significant (Z = 0.312, p = 0.754).

Substantial, statistically significant (H (5, N = 93) = 13.67459, p = 0.0178) differences were noted between the sites in terms of the number of individuals caught (Table 3). The most individuals (279) were caught at site 5, and the most species (20) at site 6. The fewest individuals (3) and species (2) were caught at site 2. The number of individuals and species caught at individual sites was negatively correlated with the number of habitats at these sites (r_xy = −0.32, p = 0.52; r_xy = −0.31, p = 0.53, respectively).

The most abundant species in the river were Torrenticola amplexa (164 individuals, 22.1% of the material collected) and Mideopsis crassipes (159 individuals, 21.4%). The group of dominants (dominance >5%) also included Sperchon clupeifer (10.2%), Hygrobates setosus (9.2%), Torrenticola barsica (5.8%), and Mideopsis orbicularis (5.5%). Differences were noted in the dominance structure between the habitat types. In the riffle habitats, the group of dominants was formed by Torrenticola amplexa (32.2%), Sperchon clupeifer (18.7%), and Torrenticola barsica (11.1%). In the pool habitats, the dominant species were Mideopsis crassipes (39.8%), Hygrobates setosus (18.1%), Torrenticola amplexa (11.0%), and Mideopsis orbicularis (10.5%).

Among the water mites collected, rheobionts were represented in the highest numbers (52.6% of the material, 16 species). The second largest synecological group was rheophiles, which accounted for 38.5% of the fauna collected (nine species). The share of lenitobionts was much smaller (7.4%, eight species). The proportion of species associated with springs was negligible—only 1.5% of the material (two species) (Table 3). Differences were noted in the shares of individual ecological groups between habitat types. The largest group in the riffle habitats was rheobionts (84.1% of the fauna collected in these habitats), followed by rheophiles (11.9%). The shares of crenobionts/crenophiles and lenitobionts were negligible (2.9% and 1.1%, respectively). In the pool habitats, rheophiles were clearly dominant (67.7% of the total fauna collected in these habitats). Rheobionts accounted for 17.9% of individuals, and lenitobionts for 14.4%. No species associated with springs were caught in the pool habitats.

Figure 4 shows the faunal similarities between the sampling sites. The sites are arranged from 1 to 6 along the course of the River. Sites 4–6 form a group with higher similarity values. Within this group, the highest similarity of fauna was noted between sites 5 and 6 (69.0%). There were significant differences in the fauna of sites 1 and 2 in comparison with the other sites (low similarity values) as well as between these two sites (no co-occurring species) (Table 3).

On the basis of dominant genera, water mite zones in the River Krąpiel were distinguished (

Table 4. Water mite zones in the River Krąpiel based on dominant genera.

Water Mite Zone	Site	Distance from the Source (km)
Sperchon	1	11.2
-	2	19.7
(Hygrobates)	3	25.8
Torrenticola–Mideopsis	4	46.2
Torrenticola–Mideopsis	5	53.2
Torrenticola–Mideopsis	6	56.4

). Site 1 can be designated as the Sperchon zone. The dominant species defining this zone was Sperchon clupeifer, but S. thienemanni was caught there as well (Table 3). The number of species and individuals caught at site 2 was too small for a water mite zone to be distinguished. At site 3, Hygrobates setosus was caught in the highest numbers, but the low abundance of this species (14 individuals) and the low overall abundance of water mites (24 individuals) do not allow a specific water mite zone to be distinguished. The core of the fauna at sites 4–6 was formed by species of the genus Torrenticola (with Torrenticola amplexa dominant at all sites) and Mideopsis (with Mideopsis crassipes dominant at all sites) (Table 3), so these three sites can be assigned to the Torrenticola–Mideopsis zone (Table 4).

The Hydrachnidia present in the two habitat types (riffles and pools) are grouped in two clusters (Figure 5). Cluster I (the upper part of the dendrogram) comprises habitats at sites in the lower reaches of the river (S4–S5). Within this cluster, there are two sub-clusters—sub-cluster R, consisting of the riffle habitats at sites 4–6, and sub-cluster P, grouping the Hydrachnidia communities inhabiting pools at sites 4–6. Cluster II (the lower part of the dendrogram) includes the habitats located at sites 1 and 3. The faunal similarity was lower here than in cluster I, and the habitats are grouped in R–P pairs within each site. At site 2, water mites were caught only in pool habitats. The Hydrachnidia community at S2P was faunistically very distinct from the communities at the other sites (Figure 5).

In the dendrogram of the co-occurrence of species, two complex clusters are formed (Figure 6). In the first cluster (the lower part of the dendrogram), the highest co-occurrence (92.3%) was obtained for Lebertia insignis and Torrenticola anomala. These species co-occurred at sites 4, 5, and 6 (Table 3). A fairly high rate of co-occurrence (83.3%) was also found for Lebertia fimbriata and Hygrobates nigromaculatus (54.2%), which co-occurred at site 6. In the second cluster (upper part of the dendrogram), the highest rate of co-occurrence was noted between Torrenticola amplexa and Mideopsis crassipes (91.0%). These species co-occurred at sites 4, 5, and 6. There are two pairs of species situated outside the clusters. The first pair consists of Atractides nodipalpis and Hygrobates longipalpis (30.8%), which co-occurred only at site 4, and the second is Parathyas palustris and Sperchon clupeifer (21.2%), which co-occurred only at site 1 (Table 3).

3.3. Water Mites and Environmental Factors

An analysis of the correlations linking faunistic indices with hydromorphological indices and RHS classes revealed the following relationships: (1) the more natural the habitats (HQA), the more individuals and species were caught, and the higher species diversity was recorded; (2) the greater the modification of habitats (HMS), the fewer individuals and species were caught, and the lower species diversity was recorded; (3) the higher the habitat quality (RHQ), the more individuals and species were caught, and the higher species diversity was recorded; (4) the greater the habitat modification (RHM), the fewer individuals and species were caught, and the lower species diversity was recorded; (5) the higher (worse) the RHS class, the fewer individuals and species were caught, and the lower species diversity was recorded (

Table 5. Values of correlations linking numbers of individuals and species and the Shannon–Wiener index with hydromorphological indices and RHS classes.

	HQA	HMS	RHQ	RHM	RHS Class
Number of individuals	0.77 *	−0.45	0.48	−0.21	−0.69
Number of species	0.77	−0.61	0.53	−0.26	−0.69
Shannon–Wiener index	0.43	−0.62	0.44	−0.31	−0.64

Note(s): * All correlation values are statistically non-significant (p > 0.05).

).

Figure 7 presents relationships between water mite distribution and the hydromorphology of the river bed, expressed by hydromorphological indices. Four variables used in the ordination explain 25.2% of the total water mite species variance. Two of them (the HMS and RHQ indices) were statistically significant (p = 0.001, p = 0.002, respectively) and explained 15.6% of the total variance in the occurrence of species (RHQ—8.1%, HMS—7.5%).

Most (13 of 17) of the species included in the analysis, including dominant species, form a dense group situated to the left of axis 2 and above and below axis 1. These species preferred hydromorphologically varied habitats (HQA) of good quality (RHQ) and avoided modified habitats (HMS). Two pairs of species are located outside this group: Parathyas palustris and Sperchon clupeifer, co-occurring in riffle habitats at site 1, and Atractides nodipalpis and Hygrobates longipalpis, co-occurring in riffle habitats at site 4 (Table 3).

Figure 8 presents relationships between the distribution of water mite fauna and the physicochemical parameters of the water. Eleven variables used in the ordination explain 39.4% of the total water mite species variance. Three of them (BOD₅, temperature, and NH₄) were statistically significant (p = 0.005, p = 0.014, p = 0.034, respectively) and explained 18.0% of the total variance in the occurrence of species (BOD₅ 6.2%, temperature 6.0%, and NH₄ 5.8%).

Most (12) of the species form a group situated at the intersection of axes 1 and 2, showing no clear relationship with any of the parameters analysed. Species outside the group are Torrenticola dudichi and Hygrobates setosus, positively correlated with NO₃, Parathyas palustris, positively correlated with O₂, Sperchon clupeifer, positively correlated with BOD₅, and Torrenticola barsica, positively correlated with water temperature (Figure 8). Torrenticola dudichi and Hygrobates setosus were caught in the highest numbers at site 5 (Table 3), where high concentrations of nitrate nitrogen were recorded (5.62 mg/dm³ average for the site during the study period). Parathyas palustris was caught only at site 1, and Sperchon clupeifer was most abundant at this site (Table 3). This site had the best oxygen conditions of all sites (8.66 mgO₂/dm³ average for the site during the study period), as well as high BOD₅ values in comparison with the other sites (7.3 mg/dm³ average for the site during the study period). Torrenticola barsica was caught only at site 5 (Table 3), where the average water temperature was 14.7 °C.

Figure 9 presents relationships between the distribution of water mite fauna and selected environmental parameters. Seven parameters were taken into consideration, but during the CCA, the forward selection procedure excluded the parameter ‘mineral sediment’. The other six parameters used in the ordination explained 26.7% of the total water mite species variance. Two of them (velocity and sediment sorting) were statistically significant (p = 0.001, p = 0.006, respectively) and explained 16.8% of the total variance in the occurrence of species (velocity—10.7%, sediment sorting—6.1%).

The largest group of species (eight) shows a relationship with habitats with organic sediment and abundant aquatic plants (left part of the diagram). The species Lebertia fimbriata and Torrenticola amplexa were associated with habitats with a rapid current. Lebertia fimbriata was caught in riffle habitats at sites 5 and 6, characterized by mineral sediments and a fast water current (Table 1, habitats S5/1, S6/1, and S6/3). Torrenticola amplexa clearly preferred environments with a faster water current, as of the 164 specimens caught in the river, 125 were collected in riffle habitats and only 39 in pool habitats (Table 3). A connection between Parathyas palustris and sediment sorting (W) can be seen. This species was collected only from habitat S1/1, with homogeneous sand sediment (Table 1). Another group (six species) showed a preference for a mineral bottom with large particle sizes and fast water flow (Figure 9, cluster II). Two species, Torrenticola barsica and T. brevirostris, were present in sunlit places, and Parathyas palustris preferred a coarse-grained mineral substrate.

4. Discussion

Aquatic invertebrates are very good biological indicators for assessments of environmental conditions in river ecosystems [1,2,3,4]. Macroinvertebrate communities react to both organic pollution and habitat changes [8,51,52]. The effects of these two factors are often difficult to separate [22,24], but in the case of some taxonomic groups, it is fairly easy to determine the effect of one of them—water pollution. One such group is water mites. Lotic water mites, as a whole group, are highly sensitive to water pollution; a deterioration of water quality results in a marked impoverishment of the fauna, and in heavily polluted water, water mites are not present at all [53,54,55]. In addition to the decrease in numbers, lotic water mite communities also respond to water pollution with a decrease in the abundance of sensitive rheobiontic species and their replacement with more tolerant lentic species [56]. The present study, as well as previous studies on water mites of the River Krąpiel [35,42,57], shows that Hydrachnidia communities in the Krąpiel are rich and diverse in terms of both the number of species and the number of individuals. Moreover, among the various ecological groups, species typical of lotic ecosystems, i.e., rheobionts and rheophiles, are prevalent. These findings are evidence that in the case of the River Krąpiel, the effect of organic pollution can be ruled out as an environmental stressor that decreases the species richness and abundance of water mites. As water pollution has been excluded as a factor influencing the structure of the water mite population in the river, other groups of environmental factors, i.e., hydromorphological parameters, physicochemical parameters of the water, and physical parameters of the environment, should be analysed and discussed.

Assessments of the quality of the aquatic environment in the European Union are based on methods involving the use of aquatic invertebrates, but at the same time, analyses of the physicochemical factors of the water and hydromorphological assessments of habitats are recommended to supplement biological methods [5]. The advantage of this approach, i.e., the integration of biological and habitat components, is that it leads to a holistic view of the ecological status of rivers [13]. Hydromorphological assessments, including the commonly used River Habitat Survey, not only provide a detailed description of habitats but are also helpful in explaining the patterns of macroinvertebrate distribution in rivers [22,29,30]. The RHS method involves assessments of numerous individual elements of the environment [10], but it is convenient to use two indices that combine the assessment of multiple partial parameters into a single numerical value: the habitat quality assessment (HQA) and the habitat modification score (HMS). Research combining both macroinvertebrate fauna and the RHS method has shown that the HQA seems to be the most important hydromorphological index for explaining the distribution of aquatic macroinvertebrates, as there is a strong connection between its values and macroinvertebrate community structure [22,29,30]. The same conclusion can be drawn in the case of water mites and the River Krąpiel—among all the hydromorphological indices, the HQA index had the greatest influence on parameters such as the number of specimens and species caught (Table 4). When we look at the individual study sites included in this research, even without carrying out an RHS assessment, we can see an overall characterization of the habitats, including anthropogenic modifications altering the hydromorphology of the river (Figure 2). That overall impression is confirmed by the values of the hydromorphological indices for individual sites. A comparison of the values of the hydromorphological indices reflecting the quality of habitats with the water mite fauna at the level of general population parameters, i.e., the number of individuals and species caught at each study site, reveals a convergence of the characteristics of abiotic (hydromorphological) and biological (water mite fauna) factors. Thus, the hydromorphological characterization of habitats explains the nature of water mite communities in the river at the level of population parameters (Figure 10, left side of the diagram). It can be concluded that the RHS assessment can be used to predict the nature of water mite communities in a given river in terms of their potential numbers and species richness.

However, the dendrogram grouping the sampling sites on the basis of hydromorphological indices (Figure 3) does not correspond to the dendrogram grouping them on the basis of faunal similarities (Figure 4)—different patterns are seen in the grouping of sites. This discrepancy indicates that in a detailed faunistic analysis, i.e., at the species level, hydromorphological indices do not fully explain the distribution and numbers of particular water mite species. This discrepancy may be explained by some general regularities in the distribution of water mites in rivers. First of all, rivers should be considered in the context of a certain continuum, because the main features characterizing river ecosystems are the continuity of ecological processes and the gradient nature of changes along the course of the river [58]. In the case of Hydrachnidia, a zonal distribution of the fauna along the course of the river can be seen [59,60,61]. A clear zonation can be distinguished in rivers, based on zones of particular species [59] or single-genus or two-genus zones [60]. Another general rule in the distribution of water mite fauna in the longitudinal profile of a river is that in the absence of severe disturbances in the form of water pollution or extreme habitat modification, the number of individuals and species caught increases along the course of the river [42,62,63,64,65]. Finally, water mites migrate along the course of the river or are carried as drift [66], which leads to similarities between water mite populations inhabiting successive sections of the river [42,60,63,64]. All of these regularities and phenomena influenced the species distribution of water mites and their numbers along of the course the River Krąpiel. In the faunal similarity dendrogram, the sites are arranged from 1 to 6, according to the course of the river, and the group with the highest fauna similarity values is formed by sites 4–6 (Figure 4). These sites were located close together, forming a spatial group (Figure 1). They belonged to the first and third RHS classes (Table 2), which means that the hydromorphological modifications assessed at these sites were not an environmental stressor that could significantly affect water mite communities. Thus, the structure of the water mite community, in terms of both species composition and the numbers of particular species, resulted mainly from the natural processes arising from the biology and ecology of particular water mite species, as well as from the functioning of river ecosystems as a continuum. The relatively close proximity of sites 4–6 and the many similarities in their environmental features (Table 1) led to the occurrence of similar species in similar numbers. In terms of both environmental and biological features, this section of the river, which includes these three sites, belongs to a single water mite zone that can be described as the Torrenticola–Mideopsis zone (Table 4).

The continuity of the river ecosystem also explains the relative impoverishment of the fauna at site 3. Although this site belonged to the second RHS class (Table 2), only twenty-four specimens and six species were caught there (Table 3). Site 3 lay about 6 kilometres downstream from site 2, which was the least natural and most altered of all the sites and thus belonged to the fifth RHS class. Due to the factors mentioned above, i.e., zonation in water mite distribution along the course of the river and the migration and drift of species, as well as the spatial arrangement of sites 2 and 3 (Figure 1) and the continuity of the river ecosystem, the water mite community at site 3 was much poorer than the hydromorphological conditions might indicate. The distance between site 3 and site 2, with the poorest water mite fauna, was too short to allow for the restoration of diverse and numerous water mite communities comparable to those noted in further sections of the river.

In the stretch of the river including sites 4–6, where the most water mite species and specimens were caught, the faunistic similarity between habitat types at different sites was greater than the similarity between habitats within the same site (Figure 5). This pattern of similarities indicates that habitat conditions within riffles or pools, i.e., habitats constituting only a segment of the cross-section of the river, were of greater importance for species distribution than the morphological characteristics of the entire cross-section of the river.

The considerations discussed above can be summarized as follows: at the species level, especially in sections of the river that are not greatly modified, general regularities in water mite fauna distribution in river ecosystems, the continuity of the river ecosystem, and characteristics at a smaller spatial scale (the habitat scale) explain the structure of the water mite community to a greater extent than the hydromorphological indices determined for a given site or section of the river (Figure 10, right side of the diagram).

Sites 1 and 2, assigned to the worst RHS classes, require a separate discussion. Site 1 can be described as 80% natural, but a significant modification of the site, involving the placement of stones in the river bed, reduced the HQA index and increased the HMS index, in effect placing this site in the fourth RHS class (Table 2). In terms of the RHS assessment, these modifications constitute significant interference in the environment, but putting aside the anthropogenic nature of the changes, the placement of large stones increased habitat diversity by creating a pseudo-mountainous section in the lowland river (Figure 2a). The stony bottom and rapid, turbulent water flow were reflected in the nature of the water mite population. There were 66 specimens of Sperchon clupeifer caught at this site (of the 76 caught in the river). Within site 1, S. clupeifer was most abundant in habitat S1/2, which confirms this species’ predilection for fast-water-current environments with a stony bottom [62,67,68,69]. This example of an abundant population of S. clupeifer inhabiting an anthropogenically modified section of the river confirms other authors’ observations that certain anthropogenic modifications (e.g., changes in the bottom structure) can have a positive effect on aquatic invertebrate populations, leading to an increase in species richness, species diversity, or the numbers of individual species [24,65,70].

Site 2 was the most hydromorphologically altered of all the sites; the river here took the form of a straight channel (Figure 2b). The HQA and HMS values placed this stretch of the river in the fifth (worst) RHS class (Table 2). A common consequence of river channelization is a decrease in habitat diversity and in the number of niches, which leads to negative changes in invertebrate communities [71]. A modification of the river channel can affect invertebrate communities more than even changes in the physicochemical characteristics of the water [52]. It may impact aquatic invertebrates directly through habitat changes or indirectly by reducing the quantity of available food [72]. All of these mechanisms are present at site 2. Only three water mite specimens were caught at this site (Table 3). Apart from the extreme impoverishment of the fauna, the ecological character of the species caught additionally confirms the negative effect of anthropogenic modification in the form of river channelization. The specimens caught at this site were Hygrobates longipalpis, a rheophile; Limnesia maculata, a lenitobiontic species (caught only at this site); and a deutonymph of Piona sp.—a genus whose representatives are also lenitobionts. The replacement of rheobionts and rheophiles, species characteristic of lotic environments, by lentic species is a common reaction of Hydrachnidia communities to the channelization of riverbeds [56].

The RHS assessment takes into account not only parameters in the river itself but also environmental parameters outside the aquatic environment, such as riparian vegetation and the use of the land immediately surrounding the river [10]. Including parameters from outside the aquatic environment in the RHS method to determine the habitat conditions for aquatic invertebrates is justified because aquatic invertebrates can be influenced by factors acting in the terrestrial environment [73,74,75,76]. This applies to water mites as well. Water mites are absolute hydrobionts at the deutonymph and adult stage, but the life cycle of most species includes a parasitic larva, which spends some time outside the aquatic environment [32,34,77]. Due to the presence of a terrestrial stage in the life cycle of water mites, factors acting outside the aquatic environment can have a direct or indirect impact on water mite communities inhabiting aquatic environments as deutonymphs and adults [35,42]. Measures such as the removal of riparian vegetation may have affected populations of insect hosts of water mite larvae (mainly Chironomidae), which in turn may have influenced the number and species composition of water mite larvae returning to the river from their insect hosts [35]. At site 2, the Krąpiel flows through an open area with no riparian vegetation (Figure 2b). This may be another reason, besides channelization, for the extreme impoverishment of the water mite fauna at this site.

Among the physicochemical parameters of the water, the distribution of water mite species was most influenced by BOD₅. One species—Sperchon clupeifer—was most clearly associated with this parameter (Figure 8). Although S. clupeifer tolerates moderate organic pollution [56], the relatively high numbers of this species at a site with elevated BOD₅ values (site 1) should not be interpreted as a preference of S. clupeifer for environments with a high content of organic matter, but rather as the effect of the species’ greater resistance to organic pollution than in the case of most lotic water mite species [78]. Therefore, the association of S. clupeifer with BOD₅ can be regarded as accidental—as stated above, the main environmental features contributing to the presence of a large population of this species at site 1 were the fast water current and stony bottom.

Among the general physical parameters of the environment (with the addition of aquatic vegetation), the most important factor explaining the distribution of water mite species in the river was the water current. This was to be expected, as in lotic ecosystems, the water current is one of the most important factors determining the distribution of aquatic invertebrates [79,80], including water mites [32]. The important role of the water current for the distribution of water mite species can be seen in the fact that the Hydrachnidia fauna of the Krapiel was dominated by rheobionts and rheophiles, i.e., species for whose occurrence in aquatic ecosystems this environmental factor is fundamental [32]. The species most strongly associated with habitats with a faster water current were Torrenticola amplexa and Lebertia fimbriata. Torrenticola amplexa clearly preferred environments with a faster water current, as of the total 164 specimens caught in the river, 125 were collected in riffle habitats and only 39 in pool habitats (Table 3). Lebertia fimbriata was caught in riffle habitats at sites 5 and 6, characterized by mineral sediment and a fast water current (Table 1, habitats S5/1, S6/1 and S6/3). This distribution of these species confirms the literature data about their environmental preferences—both species prefer lotic conditions [43,81].

The comparison of the CCA diagrams of the distribution of water mites in relation to the hydromorphological indices (Figure 7), the physicochemical parameters of the water (Figure 8), and the physical parameters of the environment (Figure 9) reveals that in each diagram, there is a group of species (the ‘core’ of the fauna) that lies close to the intersection of ordination axes 1 and 2. This group of species present in the CCA diagrams can be considered in relation to the two complex clusters of species seen in the diagram of species co-occurrence (Figure 6). Both the main group of species in the CCA diagrams and the species belonging to complex clusters in the diagram of species co-occurrence consist of the most numerous species collected at sites 4–6 (Table 3). The similar distribution of the main group of species on all three CCA diagrams, as well as their groupings in specific clusters, indicates that the distribution of the most abundant water mite species in the Krąpiel was influenced by a combination of all environmental factors, not by a single group of factors—hydromorphological, physicochemical parameters of the water, or physical factors. However, only in the CCA diagram for the hydromorphological factors are there two pairs of species outside the main group: Parathyas palustris with Sperchon clupeifer and Atractides nodipalpis with Hygrobates longipalpis. The arrangement of those two separated pairs of species in the CCA diagram for the hydromorphological factors coincides with the diagram of species co-occurrence, where the same two pairs of species are also separated from the core of the fauna (Figure 6, upper and lower part of the diagram). The convergence of species distribution in the CCA diagram for the hydromorphological indices and in the diagram of species co-occurrence indicates that of the three groups of environmental factors (hydromorphological, physicochemical parameters of the water, and physical parameters of the environment), hydromorphological factors played the most significant role in explaining water mite distribution in the Krąpiel.

Other authors, however, have not found close connections between hydromorphological assessments and water mite fauna. Czerniawska-Kusza and Szoszkiewicz [41] showed that larger water mite assemblages were present at less natural (lower HQA values) and more modified (higher HMS values) sites. Žeželj Vidoša et al. [65] showed that hydromorphological changes did not reduce either the number of water mites or species richness. According to the authors, these results can be explained by the fact that ‘water mites, because of their size, are not affected by hydromorphological alterations (i.e., do not have significant relationships with scored variables of hydromorphological alteration). Because of their body size, water mite assemblages are not influenced by hydromorphological alteration, but by microhabitat composition’. This conclusion may be correct as analyses on a smaller (mesohabitat or microhabitat) spatial scale appear to be more important for aquatic invertebrate communities than analyses on larger spatial scales [82]. This general rule can also be applied to water mites. Zawal et al. [42] found that of the three levels of organization of the environment (landscape level, macrohabitat level, and mesohabitat level), only an analysis at the mesohabitat level fully explains the structure of Hydrachnidia assemblages in the river.

5. Conclusions

• The biology and ecology of water mites make them no less suitable bioindicators of the environment—including hydromorphological modifications—than aquatic insects, commonly used for this purpose.
• The hydromorphological characteristics of the habitats, expressed as hydromorphological indices, explain the nature of water mite communities in the river at the level of general population indices (the number of individuals and species), whereas at the species level, especially in parts of the river that are not highly modified, general regularities in water mite fauna distribution in river ecosystems, the continuity of the river ecosystem, and characteristics at a smaller spatial scale (habitat scale) better explain water mite community structure than the hydromorphological indices determined for a given site or section of the river (without distinguishing specific habitats). However, these two components complement one another, and ultimately both of them together explain the nature of the water mite fauna in the river (Figure 10).
• More studies should be carried out to provide a better understanding of the connections and dependencies between water mite fauna, which is a very important element of multi-taxa invertebrate communities in lotic ecosystems, and the hydromorphological characteristics of the environment.