Using Coleoptera and Odonata as a monitoring tool for improving protected lake habitat characterisation: a case study from Ireland

doi:10.21203/rs.3.rs-3328979/v1

“Oligotrophic waters containing very few minerals of sandy plains” (code 3110) and “Natural dystrophic lakes and ponds” (code 3160) are two protected lake habitats under Annex I of the European Union Habitats Directive, despite this, their conservation condition is considered unfavourable in most European biogeographic regions. The available classifications rely heavily on water chemistry and phytosociology while leaving other taxa under described. This study used Coleoptera and Odonata communities as a monitoring tool to characterise and distinguish both lake habitat types. Samples were collected from 24 sites in western Ireland over two seasons (spring and summer) using standard sweep netting methods. Despite considerable overlap in assemblages, communities were shaped by lake habitat type, sampling region and season, where lake habitat 3110 hosted a higher diversity of taxa in summer dominated by dragonflies associated with larger, warmer oligotrophic waters, while lake habitat 3160 reflected higher beetle fauna in spring resembling small, turbid, stagnant waterbodies. This research highlights the potential use of aquatic invertebrates as an important monitoring tool for these lake habitats, serves to deepen the current knowledge and fill gaps in their classification, hence improving their conservation condition assessment methods for Member States when reporting under the Habitat Directive.

EU Habitats Directive

dystrophic lakes

oligotrophic

conservation condition

aquatic invertebrates

Atlantic blanket bogs

The European Union (EU) Habitats Directive aims to preserve and restore Europe’s natural heritage by creating a network of protected areas across all member states (MS) based on the presence of protected habitats and species as listed in the Annexes (EC, 2023). Habitats need to be characterised, monitored by appropriate methods, and, every six years, their condition must be reported under Article 17 of the Habitats Directive (EC, 2023).

Despite this legal obligation to protect habitat types, two lake habitat types in Ireland “Oligotrophic waters containing very few minerals of sandy plains” (code 3110) and “natural dystrophic lakes and ponds” (code 3160), are in bad (stable) and inadequate (stable) conservation condition respectively (NPWS, 2019), in line with the other four lake habitat types in the country (NPWS, 2019). This is due to several pressures such as agriculture, forestry, and peat erosion, as well as poor land management practices. Incomplete characterisation and limited data are also hindering the implementation of accurate monitoring methods and assessment tools which determine the conservation condition of the lake habitats (Gray et al., 2022; NPWS, 2019; O’Connor, 2015). All habitats are defined by the Interpretation Manual of European Habitats (EC, 2013), which is predominantly based on vegetation community data. Under this classification, lake habitat 3110 (often referred to as oligotrophic and soft water lakes, “lobelia lakes” or “isoetid lakes”) develops on shallow sandy and rock substrates of oligotrophic clear water lakes which are base poor and slightly acidic. The diversity of plants growing in this lake habitat can be low, however, species adapted to low carbon and nutrients such as the isoetids Isoetes sp., Littorella uniflora, Lobelia dortmanna or algae like Nitella sp. and Chrysophyceae can densely colonise water depths up to 3–6 meters with evident zonation (EEA, 2016a). On the other hand, lake habitat 3160 (often described as “humic lakes”, or brown/acidic bog lakes and ponds) are small shallow waterbodies found in boreal and temperate peatlands (EEA, 2016b). Their waters are rich in organic matter resulting in brown coloured waters, acidic pH and muddy, deoxygenated bottoms. They sustain low macrophyte diversity and, due to low transparency, are characterised by floating species of the Utricularia, Sparganium and Juncus genera or generalist bog flora and mosses (EEC, 2013; EEA, 2016a).

Ireland’s working application of the Interpretation Manual of EU Habitats describes both lake habitat types based on a combination of expected water chemistry conditions, occurrence of typical macrophytes species and geology (O’Connor, 2015). However, the information used to develop this classification was limited to those included in national water and vegetation reports, thus failing to provide a complete and detailed characterisation of the two lake habitats which eventually relied heavily on expert opinion (Gray et al., 2022; O’Connor, 2015). Lake habitat 3110 were identified as nutrient and base poor lakes (low phosphorous, nitrogen and alkalinity) with weakly acidic waters (pH > 5.5) and an isoetid type of vegetation growing on the underlying granite or sandstone bottoms. On the other hand, lake habitat 3160 are considered as acidic waterbodies integrated within peatlands and species poor botanically (dominated by floating macrophytes and mosses). Although these descriptions were generally confirmed, recent research shows that the two lake habitats can exist in very similar water chemistry conditions (such as low pH, alkalinity and nutrients), with differences in colour and dissolved organic matter highly influenced by regional and topographical factors, therefore highlighting the need for developing more accurate and specific monitoring tools (Cappelli et al., 2023).

Current assessment methods for structure and function of protected lake habitats could incorporate additional biological monitoring elements to better define and monitor these lake habitats. Invertebrates are well recognised indicators of habitat condition and are already a reliable assessment tool in current European frameworks for environmental protection such as the Water Framework Directive (WFD). Limited information is available on the specific invertebrate communities of lake habitat 3110 as studies usually focus on their macrophyte assemblages and where invertebrate communities are studied, the lake habitat type is typically not determined. The European Red List of habitats highlights the presence of Amphipoda, Mysida and Gastropoda as well as Ephemeroptera, Plecoptera and Trichoptera for lake habitat 3110 (EEA, 2016a). In contrast, invertebrates are considered particularly important for dystrophic lake habitat 3160, with these lakes potentially hosting multiple rare species, therefore invertebrates could support their monitoring especially since these lake habitats have low macrophyte diversity (EC, 2013; Gray et al., 2022; O’Connor, 2015). Previous knowledge of macroinvertebrate assemblages in dystrophic lakes and ponds reveals them to be dominated by Diptera (Chironomidae), Trichoptera, and predaceous insects such as Coleoptera, Hemiptera and Odonata (Beadle et al., 2015; Desrochers & van Duinen, 2006; EEA, 2016b; Ozoliņš et al., 2021, Pakulnicka & Kruk, 2023). The Irish Interpretation of the EU Manual (O’Connor, 2015) compiled a list of more typical species, mostly based on a survey of blanket bog lakes (presumably lake habitat 3160) in western Ireland by Drinan (2012). Among the characteristic species listed are a dominance of Coleoptera, such as Acilius sulcatus (Linnaeus, 1758), Colymbetes fuscus (Linnaeus, 1758), Gyrinus minutus (Fabricius, 1798), Gyrinus substriatus (Stephens, 1828) and Hydroporus erytrocephalus (Linnaeus, 1758), together with the mayfly Leptophlebia vespertina (Linnaeus, 1758), and Odonata such as Pyrrhosoma nymphula (Sulzer, 1776) and Aeshna juncea (Linnaeus, 1758). Due to the harsh environmental nature of dystrophic lakes (low pH and a near lack of calcium carbonate) Amphipoda, Gastropoda and Annelida are present in low numbers or absent (Friday, 1987; Beadle et al., 2015), while their standing waters reduces the presence of Plecoptera and most Ephemeroptera (Beadle et al., 2015). Dystrophic lakes and ponds tend to be smaller in size than other lake types resulting in fewer mesohabitats capable of sustaining complex communities, resulting in fewer species or species belonging to similar functional feeding categories (Heino, 2008; Ranta, 1985). However, increases in organic matter, presence of anoxic conditions and accumulation of organic matter may favour groups such as the family Chironomidae (Diptera) whose larvae employ a detritivore feeding strategy that is more efficient in such conditions (Jones, 2005; Kesti et al., 2022). In turn, the abundance of Chironomidae could provide an essential food source for predacious insects such as Dytiscidae (Coleoptera) (Nilsson et al., 1994; Frelik & Pakulnicka, 2015). This highlights that dystrophic lakes can maintain high diversity and demonstrate shifts from generalist species towards more acidophilous species (adapted to lower pH) resembling typical peatland fauna such as Hydroporus sp., Anacaena sp. or Enochrus sp. (Pakulnicka & Zawal, 2007; Pakulnicka & Zawal, 2018). Moreover, Ozolinš et al. (2021) assessed differences in intact dystrophic (humic) lakes with those of altered water regimes and found that overall taxonomic richness of Coleoptera and Odonata declined and shifted to a Diptera dominated community with metrics such as overall abundance (as individuals/m²) and Coleoptera richness acting as reliable metrics of lake condition assessment. Overall, dystrophic lakes host more specialised invertebrate communities adapted to harsh environmental conditions and represent breeding sites for several rare and protected species, such as Agabus arcticus (Paykull, 1798), Cordulia aenea (Linnaeus, 1758) and Dytiscus lapponicus (Gyllenhal, 1808) as demonstrated by Drinan (2012) in Ireland, and by Coenagrion hastulatum (Charpentier, 1825), Ilybius aenescens (Thomson, 1870) and Gyrinus minutus (Fabricius, 1798) found by Towers (2004) in Scotland.

To date, it is challenging to characterise the reference invertebrate communities of the two lake habitat types due to a lack of studies directly comparing them, however, research has described shifts in invertebrate composition of clearwater lakes under increasingly humic conditions. For instance, Kesti et al. (2021) compared humic and clearwater lakes in Finland and found Chironomidae to be associated with higher organic carbon content while Asellidae (Crustacea), Baetidae (Ephemeroptera) and Plecoptera where more common in clear-water lakes. While in Poland, Pakulnika & Zawal (2007) found that Lake Szare (a lake in the process of dystrophication) had an increased proportion of predatory Dytiscidae and scavenger beetles (Hydrophilidae) and fewer open lacustrine-fluvial like Haliplidae compared to oligotrophic lakes. Changes in the environmental condition of lakes and pools can also determine new invertebrate assemblages. As an example, Baars et al. (2014) found invertebrate communities of upland lakes in Ireland separated along gradients of pH, altitude and nutrients, whereas other authors argue that it is the presence of new niches created by more vegetation or new types of vegetation that drive changes in invertebrate communities (Heino, 2008; Nilsson et al., 1994; Ranta, 1985). Nevertheless, the influence of environmental conditions on the lakes and pools underlines the role that catchment characteristics play in shaping invertebrate community composition, hence, a level of regional scale variability should be considered and described. Indeed, even in small countries like Ireland, local conditions such as the proximity to the sea, altitude and geology could make the characterisation process of lake habitats more complex (O’Connor, 2015).

This research aims to:

Characterise the Odonata and Coleoptera specific community assemblages of lake habitat types 3110 and 3160;

Evaluate the influence of sampling season, sampling region and environmental factors in describing Odonata and Coleoptera community composition of both lake habitats;

Assess species rarity to inform the conservation condition of these protected lake habitats; and

Make recommendations on the use of Odonata and Coleoptera as indicators for assessing conservation condition within both protected lake habitats.

Site selection

Two Special Areas of Conservation (SAC) containing both lake habitats 3110 and 3160 (O’Connor, 2015) with favourable conservation condition were selected, based on the latest national assessment of protected habitats (NPWS, 2020), namely: Connemara Bog Complex SAC (site code IE0002034), Co. Galway and Owenduff/Nephin Complex SAC (site code IE0000534), Co. Mayo (Fig. 1).

The surveyed sites within the Owenduff/Nephin SAC were located in the Owenduff river catchment and characterised by an extensive area of active blanket bog with metasedimentary quartzites and schists (Geological Survey Ireland, 2018; NPWS, 2017a). The average monthly rainfall recorded at this site was 135.9 mm (Mar 2021 – Feb 2022) (Met Éireann, 2022). Connemara Bog Complex SAC is a large site dominated by lowland Atlantic blanket bog with deep peat lying on igneous bedrock (NPWS, 2015a). The eastern section is dominated by granite rock, while in the central and western areas this is alternated with gabbroic and quartz-dioritic layers (Geological Survey Ireland, 2018). The average monthly rainfall in this site was 116.1 mm during the surveyed period (Met Éireann, 2022).

To ensure near pristine lakes habitats were targeted for the study both SAC sites underwent pre-screening through satellite imagery (Google Earth Pro, version 7.3.4.8642). This process involved rejection of any lake habitat 3110 and 3160 located inside conifer forestry, peat cutting sites or within a catchment with evident anthropogenic impacts (e.g. agriculture or roads). Since the effect of past grazing pressure is generally not assessable by satellite imagery, this was evaluated using the Commonage Framework Planning (CFP) reports (NPWS, 2013). Sites were further rejected where the CFP reports indicated a level of damage (low to high). Final selection of lake habitat sites was determined based on nearby river water quality status (2013–2018) (EPA Maps, 2022) with only sites within proximity to good or high river waterbodies (based on WFD River Waterbody Status 2013–2018 national assessment) were retained.

In total, 24 waterbodies were selected for study, ten waterbodies in Owenduff/Nephin Complex SAC comprising one lake habitat 3110 and nine lake habitat 3160, and 14 waterbodies, across two different regions, in Connemara Bog Complex SAC comprising 10 lake habitat 3110 and four lake habitat 3160 (Table 1). Lake habitats are classified as “potential” on SAC conservation objectives maps – in accordance with O’Connor (2015) and published by NPWS, indicating that their status was not supported by macrophyte data and remains to be confirmed with field data (NPWS, 2015b; NPWS, 2017b). In Ireland, normally, blanket bogs are believed to start shifting from lowland vegetation into upland vegetation at around 150 m a.s.l. (Fossitt, 2000). Therefore, six habitats could be considered at the margins of the upland area. Table 1 provides a summary of the site characteristics.

Water chemistry sampling and analysis

All 24 waterbodies were monitored on the months of May 2021 and August 2022. 13 water chemistry parameters were assessed with seven recorded on site and a further six evaluated by an accredited professional laboratory, Complete Laboratory Solutions (CLS) based in Ros Muc, Connemara, Co. Galway.

In situ monitoring of conductivity (µS/cm), optical Dissolved Oxygen (DO) as percentage saturation and concentration, pH and temperature (°C) was conducted using a YSI EXO1 Sonde, equipped with EXO sensors and associated EXO Go (Yellow springs Instruments, Xylem Inc.). At each waterbody, the Sonde was left sitting underwater, approximately two metres from the shore, for ten minutes performing five readings every two minutes. The mean value of these five readings resulted in the final parameter reading (of the month) per site. Turbidity (0.001–9999) as Formazin Nephelometric Unit (FNU) was recorded in-situ using an HACH portable system by floating the sensor in surface water for 60 second.

Water samples were collected from each waterbody using a 1 Litre (L) high-density polyethylene (HDPE), opaque, wide-mouth bottle and a 6-meter extendable pole with a 1 L plastic beaker attached. Bottles were rinsed three times with water prior to collection and the pole was extended to reach the furthest point from the shore, then bottles were filled with surface water (0–30 cm) from one single point at the site. Samples were kept on ice and delivery of water samples to CLS occurred within 24 hours after collection. Analysis of ammonium (NH₄) (Konelab CLS 40 Salicylate method), chloride (Cl) (Konelab CLS 36 based on APHA 4500-Cl-E Colorimetric Method), true colour (CLS 29 based on APHA 2120-C Colorimetric Method), silica (Si) (APHA 4500-SiO₂-D Heteropoly Blue Method), orthophosphate (PO₄) (Konelab CLS 35 based on APHA 4500 P-E Ascorbic Acid Method) and Total Phosphorous (TP) (CLS 205 digestion followed by APHA 4500 P-E) was conducted by CLS (APHA 2012). A 250 ml water sample was also collected from each site and analysed for alkalinity – as mg/L of CaCO₃ – following a low-level Gran titration method (Mackereth et al., 1989).

Table 1

Summary of study site characteristics
Study code name^a	SAC code	Location coordinates (Lat, Long)	Lake habitat type assigned^b	Altitude (m a.s.l.)	Lake area (km²)	Catchment area (km²)	Approx. Distance from the sea (km)
O/N LL 1	IE0000534	53.998153, -9.687078	3110	151	0.0106	0.0349	7.5
O/NLP 1	IE0000534	53.994844, -9.688436	3160	150	0.0016	0.0256	7.3
O/N LP 2	IE0000534	53.995578, -9.689233	3160	150	0.0018	0.0023	7.3
O/N LP 3	IE0000534	53.995231, -9.687850	3160	151	0.0011	0.0121	7.3
O/N LP 4	IE0000534	53.995900, -9.690250	3160	149	0.0028	0.0061	7.2
O/N LP 5	IE0000534	53.997781, -9.685533	3160	153	0.0008	0.0057	7.6
O/N LP 6	IE0000534	53.995317, -9.690158	3160	149	0.0033	0.0076	7.2
O/N LP 7	IE0000534	53.994808, -9.691239	3160	148	0.0005	0.0019	7.1
O/N LP 8	IE0000534	53.992481, -9.690792	3160	147	0.0060	0.0123	7.1
O/N LP 9	IE0000534	53.991833, -9.690694	3160	147	0.0003	0.0004	7.0
C LL 1	IE0002034	53.420597, -9.718833	3110	27	0.0887	1.4507	4.6
C LL 2	IE0002034	53.427989, -9.712903	3110	41	0.0326	0.1538	5.3
C LL 3	IE0002034	53.427628, -9.709103	3110	42	0.0108	0.0303	5.4
C LL 4	IE0002034	53.426244, -9.706181	3110	40	0.0403	0.3457	5.2
C LL 5	IE0002034	53.419497, -9.702053	3110	28	0.0389	0.2517	4.4
C LL 6	IE0002034	53.416375, -9.702503	3160	32	0.0105	0.0747	4.2
C LL 7	IE0002034	53.416917, -9.709889	3110	26	0.0830	3.8928	4.7
C LP 1	IE0002034	53.425897, -9.714419	3160	40	0.0089	0.0339	5.1
C UL 1	IE0002034	53.395814, -9.443747	3110	190	0.0170	0.0875	6.7
C UL 2	IE0002034	53.402194, -9.445675	3110	200	0.0406	0.3311	6.5
C UL 3	IE0002034	53.399700, -9.432550	3110	202	0.0191	0.2724	7.4
C UL 4	IE0002034	53.390725, -9.436958	3110	166	0.0244	0.7959	7.2
C UP 1	IE0002034	53.392850, -9.436911	3160	173	0.0006	0.0080	7.2
C UP 2	IE0002034	53.401578, -9.433133	3160	208	0.0011	0.0134	7.3
^a O/N: Owenduff / Nephin; LL: Lowland Lake; LP: Lowland Pool; C: Connemara; UL: Upland Lake; UP: Upland Pool; CW: Connemara west; CE: Connemara east. ^b Assigned as per O’Connor (2015)

Macroinvertebrates sampling and identification

Samples of Coleoptera and Odonata were collected in May 2021 and August 2021. All 24 habitats were sampled by net sweeping method using a standard (1 mm mesh and 20x25 cm size) net (Biggs et al., 1998). Waterbodies below 0.01 km² in size were surveyed for one minute while those above 0.01 km² were sampled for 3 minutes to allow for representative samples to be captured and making sure mesohabitats (vegetated areas, sheltered/unsheltered shores) were included in the sweeps. Samples were preserved in-situ in plastic bags with 70% IMS. Successively, individuals were sorted according to Order and separated into plastic tubes with new preservative to prepare for identification. Coleoptera and Odonata larvae were identified to species level with the following keys: Cham (2012), Hammond, (1983); Foster et al. (2014), and Foster & Friday, (2011).

Conservation assessment metrics

Using the recorded red lists species of invertebrates and matching their occurrence in the 24 study sites, Species Quality Scores (SQS) were calculated. SQS represent a metric of species rarity based on the number of 10 Km squares (hectad) in which the species has been recorded on the Irish national grid (Foster et al., 1992). Data on the species frequency at a national level (number of hectads) was obtained from the National Biodiversity Data Centre (2023). Ten SQS categories of increasing rarity were identified with a minimum occurrence of 1 hectad assigned an SQS of 10 and an occurrence greater than 512 hectads assigned an SQS of 1 (Table 2). One species, Hydroporus neglectus, was excluded from analysis because of data deficiency in the national database.

Table 2

Species quality score categories and corresponding number of hectads range
Ranges of occurrence in number of hectads	SQS
1	10
2–3	9
4–7	8
8–15	7
16–31	6
32–63	5
64–127	4
128–255	3
256–511	2
> 512	1

Statistical analyses

All statistical analyses were performed with R software for windows (version 4.1.1; R Core Team, 2021). Overall species richness between the two lake habitats was tested using Wilcoxon rank-sum test applied using “rstatix” package (Kassambara, 2023). Multivariate analysis was performed on log(1 + x) transformed species abundance data. A Non-metric Multi-dimensional Scaling (nMDS) was performed with “metaMDS” function in the Vegan package (Oksanen et al., 2022), and based on a Bray-Curtis dissimilarity distance within a two-dimensional space to assess differences in invertebrate assemblages between the two lake habitat types. A stress value of below 0.2 was considered acceptable. To determine differences among pre-defined groups (Lake habitat type: 3110 and 3160), the log(x + 1) transformed species abundance dataset was tested with permutational MANOVA’s (with 599 permutations) using “pairwiseAdonis” package (Martinez Arbizu, 2017).

To examine if invertebrate species were significantly associated with one lake habitat type over another, a multi-level pattern analysis using the function “multipatt” of Indicspecies package (De Caceres & Legendre, 2009) was carried out, with 599 permutations. This function calculates an indicator value, as in Dufrêne & Legendre (1997), where such values represent an overall percentage of association between species and one predefined group, in this case, one of the two lake habitat types, as well as a P-value of the association generated by permutations. Results also provide a statistic IndiVal and two probability measures for each species (as in Dufrene & Legendre, 1997): specificity (A), or the probability that a sample belongs to one group based on species presence and fidelity (B), the probability that the tested species will be found in one or the other group.

Environmental factors including 13 water chemistry parameters (alkalinity, ammonium, chloride, conductivity, DO (percentage saturation and concentration as mg/l), orthophosphate, pH, silica, temperature, TP, true colour and turbidity), averaged for spring (Feb – Apr) and summer (May – Jul) and six physical features of the lake habitats (perimeter, lake area, catchment size and slope, altitude and distance from the sea) was overlaid with the nMDS to evaluate their correlation with invertebrate communities assemblages within the two lake habitat types. Environmental data was standardised with “decostand” function in the Vegan package and the goodness of fit of each parameter tested with “envfit” function (Vegan), with 599 permutations and a max P-value of 0.05.

To evaluate the influence of sampling season, sampling region and environmental factors on Odonata and Coleoptera community composition the following analyses were conducted. To assess the influence of season (spring and summer 2021) on the community composition of lake habitat types, two separate nMDS ordinations for spring and summer were calculated using the log(1 + x) species abundance datasets while PERMANOVA tests (599 permutations) were conducted to compare habitat communities in seasons separately. In addition, analysis of lake habitat indicator species was performed on the spring and summer datasets to assess whether significant taxa were consistent between seasons.

Regional variation in species abundances in Connemara west (CW), Connemara east (CE), and Owenduff/Nephin (O/N) were tested using pairwise PERMANOVAs (599 permutations) using the log(1 + x) transformed species abundance data. Moreover, indicator species analysis was also performed with respect to region as indicated above.

To evaluate the importance of 19 environmental factors in shaping the invertebrate community structure, the log(x + 1) transformed species abundance dataset was tested using redundancy analysis (RDA). Redundancy analysis uses environmental parameters as explanatory variables and recalculates score points that are directly influenced by them, hence, only the variability that can be explained by the designed model is visualised (Legendre & Legendre, 2012). To decide which parameters would be used in the RDA, a model including all 19 (standardised) explanatory variables was conducted with “capscale” function of Vegan package and analysis based on an Euclidean distance index. Parameters were tested with “anova.cca” function with 599 permutations and six significant environmental factors were selected for the final model build: chloride, lake area, DO percentage saturation, pH, temperature and turbidity. The capscale function provides the constrained proportion of the variability explained by the new RDA.

Characterisation of Odonata and Coleoptera Communities in Lake Habitats 3110 and 3160

Overall, the most commonly encountered specimens were Coenagrionidae larvae (1089 individuals in total), followed by adult Gyrinidae (390), Dytiscidae (340) and Libellulidae larvae (178). In lake habitat 3110, 898 Coenagrionidae individuals were recorded, comprising 62.5% of all 1436 specimens (Table 3), followed by 157 Dytiscidae (10.9%), 109 Gyrinidae (7.6%), 106 Libellulidae (7.4%) and 63 Aeshnidae (4.4%). In contrast in lake habitat 3160, 281 Gyrinidae were recorded (Table 3), accounting for 33.3% of the total (844 specimens), followed by 191 Coenagrionidae (22.6%), 183 Dytiscidae (21.7%), 72 Libellulidae (8.5%) and 57 Lestidae (6.8%). Dytiscidae recorded the highest species richness in both lake habitats (18 and 16 in lake habitats 3110 and 3160 respectively) comprising 47.1% of all species found in lake habitat 3160 and 40.9% in lake habitat 3110 (Table 3). In lake habitat 3160 Dytiscidae were followed by Gyrinidae (11.8%) while in lake habitat 3110 by Hydrophilidae (11.4%).

Table 3

Summary of species richness and abundances at family-level for lake habitat types 3110 and 3160
	Lake habitat 3110				Lake habitat 3160
Family	No. of species	% diversity	Total abundance	% abundance	No. of species	% diversity	Total abundance	% abundance
Aeshnidae	3	6.8	63	4.4	2	5.9	30	3.6
Chrysomelidae	1	2.3	1	0.1	1	2.9	1	0.1
Coenagrionidae	3	6.8	898	62.5	3	8.8	191	22.6
Corduliidae	1	2.3	15	1.0	1	2.9	7	0.8
Dryopidae	1	2.3	12	0.8	1	2.9	20	2.4
Dytiscidae	18	40.9	157	10.9	16	47.1	183	21.7
Elmidae	1	2.3	3	0.2	0	0	0	0.0
Gyrinidae	3	6.8	109	7.6	4	11.8	281	33.3
Haliplidae	1	2.3	15	1.0	1	2.9	1	0.1
Helophoridae	2	4.5	23	1.6	0	0	0	0.0
Hydrophilidae	5	11.4	17	1.2	1	2.9	1	0.1
Lestidae	1	2.3	17	1.2	1	2.9	57	6.8
Libellulidae	4	9.1	106	7.4	3	8.8	72	8.5
Total	44	100	1436	100	34	100	844	100

A total of 44 species were found in lake habitat 3110 and 34 in lake habitat 3160 with 18 species found only in lake habitat 3110 while nine were unique to lake habitat 3160 (Table S2). A maximum of 16 species were found in one Owenduff/Nephin lowland lake site (ONLL1) and one Connemara Bog Complex lowland lake site (CLL2), with an overall average of 12 (± 3) species per lake habitat type. A significantly higher median species richness was recorded in lake habitat 3110 (14 ± 3) compared to lake habitat 3160 (12 ± 2, Wilcoxon W = 116, P = 0.01) (Fig. 2).

The most common species found in both lake habitat types was Enallagma cyathigerum, this was followed by Ischnura elegans and Pyrrhosoma nymphula in lake habitat 3110 and Gyrinus minutus and G. aeratus in lake habitat 3160 (Table 4).

Table 4

List of the top 10 most abundant species, as total number of individuals, found in each lake habitat type
Lake habitat 3110			Lake habitat 3160
Species	Family	Total abundance	Species	Family	Total abundance
Enallagma cyathigerum	Coenagrionidae	481	Enallagma cyathigerum	Coenagrionidae	169
Ischnura elegans	Coenagrionidae	239	Gyrinus minutus	Gyrinidae	135
Pyrrhosoma nymphula	Coenagrionidae	178	Gyrinus aeratus	Gyrinidae	112
Gyrinus substriatus	Gyrinidae	95	Hydroporus obscurus	Dytiscidae	65
Hydroporus obscurus	Dytiscidae	93	Lestes sponsa	Lestidae	57
Libellula quadrimaculata	Libellulidae	83	Libellula quadrimaculata	Libellulidae	42
Aeshna juncea	Aeshnidae	54	Gyrinus substriatus	Gyrinidae	33
Helophorus obscurus	Helophoridae	22	Sympetrum danae	Libellulidae	28
Lestes sponsa	Lestidae	17	Hydroporus erytrocephalus	Dytiscidae	27
Cordulia aenea	Corduliidae	15	Rhantus exsoletus	Dytiscidae	23

The nMDS ordination of species abundance per site showed a separation between the two lake habitat types where 10 of the 11 lake habitat 3110 sites grouped to the left of axis 1 while ten of the 13 lake habitat 3160 sites were grouped to the right of axis 1 (Fig. 3). Despite considerable overlap between the two lake habitat types a PERMANOVA found separations to be significantly different (F = 4.46, R² = 0.17, P = < 0.01). A fitting of environmental factors found several parameters to be significantly correlated with the ordination scores. Altitude, distance from the sea and turbidity were associated with lake habitat 3160 sites, while conductivity, lake area and perimeter, pH and temperature were all associated with lake habitat 3110 sites (Table S3). These environmental factors reflect generally larger, warmer waterbodies with higher pH and conductivity in lake habitat 3110 sites, with lake habitat 3160 sites displaying more turbid waters at higher altitudes.

Indicator species analysis revealed seven species were significantly associated with one or other of the two lake habitat types (Table 5) with Acilius sulcatus, Gyrinus aeratus, G. minutus indicators of lake habitat 3160 and Haliplus fulvus, Ischnura elegans, Pyrrhosoma nymphula and Sympetrum striolatum indicators of lake habitat 3110 (Table 5).

Table 5

Indicator Species Analysis showing significant species associated with lake habitat types 3110 and 3160
	Species	Specificity	Fidelity	Indicator value	P value
Lake habitat 3110	Ischnura elegans	0.90	0.64	0.76	**
	Pyrrhosoma nymphula	0.89	0.73	0.80	**
	Haliplus fulvus	0.92	0.54	0.71	*
	Sympetrum striolatum	1.00	0.45	0.67	*
Lake habitat 3160	Acilius sulcatus	0.81	0.69	0.75	**
	Gyrinus aeratus	1.00	0.38	0.62	*
	Gyrinus minutus	0.74	0.92	0.82	**

* < 0.05, ** ≤ 0.01

Evaluating the influence of Region, Season and Environmental factors on community composition

In terms of species richness, CE was the most diverse sampling location with a total of 37 species and an average of 14 ± 1 per lake habitat type, followed by CW (32 species, average: 12 ± 3 per lake habitat) and O/N (30 species, average of 11 ± 3 per lake habitat) although no significant difference in species richness was observed between the three sampling regions (Kruskal-Wallis χ² = 4.7, df = 2, P-value = 0.1). The most abundant species in the Connemara subregion CE were E. cyathigerum P. nymphula followed by Aeshna juncea and H. obscurus (Table 6), whereas I. elegans, E. cyathigerum followed by G. minutus were the most abundant species in CW. On the other hand, G. aeratus, H. obscurus and E. cyathigerum were the most common species found in region O/N (Table 6).

Table 6

List of top 5 most abundant species, as total number of individuals, found in each sampling region. CE: Connemara east, CW: Connemara west and O/N: Owenduff/Nephin
CE		CW		O/N
Species	Total abundance	Species	Total abundance	Species	Total abundance
Enallagma cyathigerum	415	Ischnura elegans	214	Gyrinus aeratus	77
Pyrrhosoma nymphula	111	Enallagma cyathigerum	172	Hydroporus obscurus	68
Aeshna juncea	57	Gyrinus minutus	100	Enallagma cyathigerum	63
Hydroporus obscurus	56	Gyrinus substriatus	88	Sympetrum danae	30
Ischnura elegans	39	Libellula quadrimaculata	78	Gyrinus minutus	29

Odonata and Coleoptera community differences as observed for lake habitat type in the nMDS ordination also separated out the 24 sampling sites based on their region (Fig. 3). All sites sampled in O/N SAC were observed right of axis 1 (roughly aligning with lake habitat 3160) while those of the Connemara Bog Complex SAC were located on the left of axis 1. Sites belonging to CW were associated more with axis 2 while those of CE were associated with both axis 1 and 2. Pairwise comparisons showed that the community composition of CW and CE were indistinguishable from each other (PERMANOVA F = 2.36, R² = 0.17, adjusted P = 0.65). Conversely, the community composition of O/N was significantly different from the Connemara regions, both when compared at an SAC level (F = 7.05, R² = 0.24, P = 0.0017) and at a subregional level (CW: F = 6.51, R² = 0.29, adjusted P = 0.005 and CE: F = 5.1, R² = 0.27, adjusted P = 0.005). Indicator species analysis also revealed species significantly associated with each sampling region (Table 7) with Ischnura elegans an indicator of CW; Aeshna juncea, Pyrrhosoma nymphula and Enallagma cyathigerum indicators of CE and Sympetrum danae and Rhantus exsoletus indicators of O/N.

Table 7

Indicator Species Analysis showing significant species associated with the three sampling regions. CE: Connemara east, CW: Connemara west and O/N: Owenduff/Nephin
		Specificity	Fidelity	Indicator value	P value
CW	Ischnura elegans	0.75	0.75	0.75	*
CE	Aeshna juncea	0.67	1.00	0.82	**
	Pyrrhosoma nymphula	0.60	0.83	0.71	*
	Enallagma cyathigerum	0.47	1.00	0.68	**
O/N	Sympetrum danae	0.64	0.90	0.76	**
O/N	Rhantus exsoletus	0.80	0.60	0.69	*

* < 0.05, ** ≤ 0.01

An overall increase in total species richness was found between spring and summer 2021. During spring, an average of 6 (± 3) species were found in lake habitat 3160 and 4 (± 3) in lake habitat 3110 while in summer lake habitat 3110 showed a higher species richness (average 13 ± 3) compared to lake habitat 3160 (average 9 ± 2) (Fig. 4). The most diverse family was Dytiscidae in both seasons.

Enallagma cyathigerum was common to both lake habitat types in spring, followed by I. elegans and L. quadrimaculata in lake habitat 3110 while for lake habitat 3160, Gyrinidae species Gyrinus minutus and G. aeratus were more common (Table 8). In summer 2021, numbers of E. cyathigerum and I. elegans remained high for lake habitat 3110 but an increase of P. nymphula was also noted. On the other hand, Gyrinus minutus became the most common species in lake habitat 3160 (Table 8).

Table 8

List of top 5 most abundant species, as total number of individuals, found in each lake habitat type during spring and summer 2021
	Habitat type 3110		Habitat type 3160
	Species	Total	Species	Total
Spring	Enallagma cyathigerum	103	Enallagma cyathigerum	84
	Ischnura elegans	68	Gyrinus minutus	34
	Libellula quadrimaculata	23	Gyrinus aeratus	29
	Pyrrhosoma nymphula	8	Libellula quadrimaculata	28
	Cordulia aenea	5	Rhantus exsoletus	17
Summer	Enallagma cyathigerum	378	Gyrinus minutus	101
	Ischnura elegans	171	Enallagma cyathigerum	85
	Pyrrhosoma nymphula	170	Gyrinus aeratus	83
	Gyrinus substriatus	95	Lestes sponsa	56
	Hydroporus obscurus	90	Hydroporus obscurus	49

Sites scores saw little change when comparing NMDS ordinations of the two seasons, with most lake habitat 3110 sites scoring negatively on axis 1 opposed to lake habitat 3160 sites scoring positively (Fig. 5). Comparing the site ordination scores for the two seasons resulted in differences between lake habitat types for spring (PERMANOVA: F = 4.58, R² = 0.18, P = 0.002) and summer (PERMANOVA: F = 4.65, R² = 0.18, P = 0.002).

Indicator species analysis found that in spring 2021, the only species significantly associated with lake habitat 3110 was Ischnura elegans, while four species were significantly associated with lake habitat 3160: Gyrinus minutus, Acilius sulcatus, Hydroporus erytrocephalus, Rhantus exsoletus (Table 9). While in summer 2021 five species were significant indicators of lake habitat 3110: Pyrrhosoma nymphula, Enallagma cyathigerum, Ischnura elegans, Haliplus fulvus, Sympetrum striolatum while only one indicator was identified for lake habitat 3160: Gyrinus aeratus.

Table 9

Indicator species analysis showing significant species associated with lake habitat types 3110 and 3160 during spring and summer 2021
			Specificity	Fidelity	Indicator value	P value
Spring	Habitat type 3110	Ischnura elegans	0.86	0.64	0.74	**
	Habitat type 3160	Gyrinus minutus	0.88	0.69	0.78	*
		Acilius sulcatus	1.00	0.46	0.68	*
		Hydroporus erytrocephalus	1.00	0.38	0.62	*
		Rhantus exsoletus	1.00	0.38	0.62	*
Summer	Habitat type 3110	Pyrrhosoma nymphula	0.93	0.73	0.82	**
		Enallagma cyathigerum	0.65	1.00	0.81	*
		Ischnura elegans	0.94	0.64	0.77	**
		Haliplus fulvus	0.91	0.45	0.64	*
		Sympetrum striolatum	1.00	0.36	0.60	*
	Habitat type 3160	Gyrinus aeratus	1.00	0.38	0.62	*

* < 0.05, ** ≤ 0.01

RDA modelling found that environmental factors chloride, dissolved oxygen (percentage of saturation), pH, temperature and turbidity along with lake area were significant explanatory variables influencing Odonata and Coleoptera assemblages (Table S4) and together explained 41% of the total variability in the species abundance dataset (Fig. 6). Turbidity was associated with lake habitat 3160 and indicator species A. sulcatus while the remaining parameters were aligned with lake habitat 3110 and indicator species I. elegans and P. nymphula.

Four species recorded in this survey are listed in the Irish National Red List: Agabus arcticus (near threatened), Boreonectes multilineatus (syn. Stictotarsus multilineatus near threatened), Cordulia aenea (endangered) and Dytiscus lapponicus (near threatened) (Foster, Nelson and O’Connor, 2009; Nelson, Ronayne and Thompson, 2011). 58% of sites included at least one red list species, with one site (ONLL1) containing all three near threatened Coloeptera species (Table S2). The most represented species of the red list was Cordulia aenea (Order Odonata) found in 63% of lake habitat 3110 sites and 23% of lake habitat 3160 sites (Table S2). Boreonectes multilineatus (Order: Coleoptera) was found in only one site of both lake habitat types (ONLL1 and ONLP6), Agabus arcticus (Order Coleoptera) was recorded in two lake habitat 3160 sites (ONLP2 and CUP1) and one lake habitat 3110 site (ONLL1), with Dytiscus lapponicus (Order Coleoptera) only recorded in one lake habitat 3110 site (ONLL1).

Species with the highest species quality scores (SQS) were: Cordulia aenea (7), Dytiscus lapponicus (7), Agabus arcticus (6), Boreonectes multilineatus (6) and Helochares punctatus (6) (Table S2). Overall, the mean rarity score within the 53 Odonata and Coleoptera species dataset (mean SQS = 3.2 ± 1.8) and for the 24 sites (mean SQS = 2.6 ± 0.42) was low indicating the study sites contained widespread species (Fig. 7). The median SQS for lake habitat 3110 was 2.3 and 2.8 for lake habitat 3160, with no significant difference in rarity scores between the two lake habitat types observed (Wilcoxon W = 49, P = 0.2).

Characterisation of Odonata and Coleoptera Communities in Lake Habitats 3110 and 3160

At a family level, an overall prevalence of Coenagrionidae in lake habitat 3110 was found with Gyrinidae more frequent within lake habitat 3160 (Table 3). Dytiscidae were the most diverse family in both lake habitats, in line with surveys from similar types of waterbodies within peatland catchments (Drinan, 2012; Free et al., 2006; Hannigan & Kelly-Quinn 2014; Soldán et al., 2012; Towers, 2004). Lake habitat 3110 showed significantly higher species richness compared to lake habitat 3160 (Fig. 2) suggesting that the dystrophic water chemistry conditions (Cappelli et al., 2023) of these sites may lead to lower species diversity, due to their limiting environmental conditions. In addition, lake habitat 3160 sites surveyed in this study were also significantly smaller and isolated compared to lake habitat 3110 sites, and which are known to sustain lower species diversity (Beadle et al., 2015; MacArthur & Wilson, 1967). 44 species of Coleoptera and Odonata were found in lake habitat 3110 and 35 in lake habitat 3160, similar to other studies within peatland regions. As an example, Hannigan et al. (2011) found 13 species of Coleoptera and Odonata in intact blanket bog pools in eastern Ireland and Towers (2004) found 15 species in similar pool habitats in Scotland. While Baars et al. (2014) found 22 species of Coleoptera in upland Irish lakes, and Drinan (2012) 54 species of Coleoptera and 12 species of Odonata from 13 peatland lakes in Ireland.

At a species level, the 24 waterbodies segregated according to lake habitat type (Fig. 3) with grouping correlated with eight significant environmental parameters (Fig. 3 and Table S3) resulting in one group dominated by Dytiscidae and Gyrinidae (lake habitat 3160) and another, more diverse, resembling larger lowland lakes (lake habitat 3110) (Fig. 3 and Figure S1). Although it is challenging to delineate clear profiles for the two lake habitats due to considerable overlap in the Odonata and Coleoptera assemblages. This was expected (O’Connor, 2015), because the lake habitats examined are similar in their chemical attributes (Cappelli et al., 2023) and because both belonged to catchments dominated by blanket bogs (NPWS, 2016a; NPWS, 2016b) therefore many species recorded are common and widespread bog fauna with preferences for acidic conditions (Foster et al., 1992, Nelson & Thompson, 2004). Nonetheless, lake habitat 3160 sites were smaller, more acidic and turbid but also more distant from the sea (colder with lower conductivity) which alignes with identified indicator species Acilius sulcuatus, Gyrinus aeratus and Gyrinus minutus (Table 5). A. sulcatus and G. minutus are both typical of stagnant waterbodies with exposed margins and share a full aquatic predatory life for which they developed strong swimming ability (Foster & Friday, 2011; Nilsson, 1996). On the other hand, lake habitat 3110 were associated with generalist and acid-tolerant dragonfly species such as the two Coenagrionidae I. elegans and P. nymphula or the Libellulidae S. striolatum. These species show preferences for lowland (warmer) nutrient poor lakes and are rarely found in colder or enriched sites in Ireland (Nelson & Thompson, 2004). Haliplus fulvus was the only Coleoptera species identified as an indicator species for lake habitat 3110 (Table 5). This species is among the few Haliplidae recorded in these habitat types (Baars et al., 2014; Drinan, 2012), and also a rather tolerant one, able to colonise acidic and upland waterbodies (Foster & Friday 2011, Nilsson, 1996). A similar community pattern was described by Drinan (2012) from upland and lowland peatland lakes, where the first were characterised by a prevalence of larger Coleoptera species, whereas lowland sites included a larger diversity of dragonflies. However, the altitudinal range within this study was limited (max 208 m a.s.l.) with the average altitude for lake habitat 3160 (138 ± 79 m), similar to that of lake habitat 3110 (101 ± 48 m) which is considered low for this lake habitat type (O’Connor, 2015) therefore the influence of altitude on community composition may be limited. In fact, only a few species found in this survey can be described typical of upland waters: Agabus arcticus, Boreonectes multilineatus and Dytiscus lapponicus (Foster et al., 2009); all of which were recorded in a small number of sites across both lake habitat types.

In summary, lake habitat types 3110 and 3160 show subtle differences in their Coleoptera and Odonata communities. However, the high rate of similarity between lake habitat types based on just Odonata and Coleoptera suggests that a wider target of monitoring invertebrates should be considered to address unexplained variability between the lake habitat types. For example, Drinan (2012) found that lowland, less acidic lakes were also characterised by higher abundances of gastropod Lymnaea peregra (Müller), Baars et al. (2014) identified 100 Chironomidae (Diptera) species from peatland lakes, while Kesti et al. (2022) found an increase in Chironomidae and a parallel decrease of Asellidae (Isopoda) in humic lakes compared to clearwater sites. Therefore, in order to build a comprehensive profile of the invertebrate communities of these protected lake habitats, future surveys should extend the characterisation to all available taxa with particular attention to the identification process, ensuring that specimens are identified at species level to facilitate habitats comparison.

Evaluating the influence of Region, Season and Environmental factors on community composition

As depicted by the NMDS ordination, the separation in Odonata and Coleoptera communities is also a reflection of the varying sampling regions (Fig. 3). Most of the lake habitat 3160 sites (nine of 13) were sampled in Owenduff/Nephin SAC and can be described as small acidic and less oxygenated bog pools whereas, those of lake habitat 3110 consist of larger waterbodies with slightly higher pH, alkalinity and colour (Cappelli et al., 2023). Ten sites in O/N were associated with the beetle Rhantus exsoletus (Dytiscidae) found in stagnant acid /mesotrophic waters (Foster & Friday, 2011) and the dragonfly Sympetrum danae (Libellulidae) a specialist of small acidic peat pools and cold environments (Nelson & Thompson, 2004). Based on PERMANOVA tests on the abundance data, the two groups in Connemara Bog Complex SAC (CW and CE) where not significantly different although indicator species analysis did differentiate between the sites with Ischnura elegans associated with CW and Aeshna juncea, Enallagma cyathigerum and Pyrrhosoma nymphula with CE, in line with indicator species found for lake habitat 3110 (Table 7).

Significant differences between regions based on varying physical or chemical conditions (Cappelli et al., 2023) does not exclude a separation based on habitat type. Oligotrophic lake habitat 3110 usually develops on rocky substrates such as granites, like those found in the Connemara SAC (NPWS, 2015a; EEA, 2016a; O’Connor, 2015) while dystrophic lake habitat 3160 develops in direct contact with the surrounding peatland and are less influenced by the underlying bedrock (EEA, 2016b; O’Connor, 2015). Although parameters connected to geology such as pH and alkalinity can have significant effects on invertebrate community structure, the role of geology on influencing the composition of biota within peatland lakes is considered to be limited (Baars et al., 2014; Drinan, 2012; Free et al., 2006). Therefore, in order to complete the classification of lake habitat types, future research should focus on including sites from representative geologies, elevations and surrounding catchment vegetation types.

Both lake habitat types saw an increase in Coleoptera and Odonata richness and abundance from spring to summer 2021 (Fig. 4). Enallagma cyathigerum was the most common species throughout the year, except in lake habitat 3160 during summer, when Gyrinus minutus became more abundant (Table 8). E. cyathigerum larvae are known to remain active during winter and have a long developmental phase therefore they can remain common throughout the year (Nelson & Thompson, 2004). The species composition remained distinct and essentially unchanged between seasons (Fig. 5) although indicator species of lake habitat 3160 were mostly found in spring (Table 9) and included only Dytiscidae and Gyrinidae. On the other hand, indicator species of lake habitat 3110 where mostly Coenagrionidae and dragonflies in general and were detected during summer (Table 9). Indeed, most adult Coleoptera would be expected to be present all year round (Foster & Friday, 2011), whilst dragonfly larvae generally start their emergence from diapause in June in Ireland (Nelson & Thompson, 2004). Increases in diversity between seasons was also noted by Drinan (2012) in Irish peatland lakes however, Baars et al. (2014) reported little seasonal change in Coleoptera richness from upland lakes in Ireland although this was not the case for other insect orders.

Sampling across two seasons (spring and summer) is normal within the survey methods of invertebrates and ensures good representation of the organisms living within the habitat however, frequency, timing and sampling effort should be designed according to the targets of the survey and would need standardisation (especially based on the size of the habitat) in the future. For instance, future monitoring of lake habitat 3160 would benefit from the use of activity traps which have been proven to be an effective method for collecting larger mobile fauna like beetles of the Dytiscidae family and reduce the impact of sampling in delicate protected environments such as shallow bog pools (Towers, 2004). While for lake habitat 3110 monitoring of the different vegetation zones or mesohabitats (O’Connor, 2015) at different depths such as a multi-habitat stratified lake sampling method would ensure representation from these larger sites.

Differences in Odonata and Coleoptera community composition could partially be explained by the chemical and environmental character of the lake habitat type. Redundancy analysis revealed six parameters significantly influenced the community structure: chloride, DO, lake area, pH, temperature and turbidity (Table S4); all parameters known to affect invertebrate composition in peatland regions (Baars et al., 2014; Drinan, 2012; Heino, 2008; Foster et al., 1992; Ranta, 1985). However, the constrained proportion explained by this model (41%) suggests that the physical and chemical attributes of the lake habitats only play a partial role in shaping the two communities and that other parameters, potentially more influential, were not considered. For instance, vegetation structure has been shown to be a significant parameter shaping aquatic Coleoptera assemblages in lakes undergoing succession (Plakulnicka & Zawal, 2018) and is a parameter which requires further consideration. As a matter of fact, indicator species of lake habitat 3110 were all benthic predators such as Pyrrhosoma nymphula and Sympetrum striolatum which prefer waterbodies with wide shallow areas covered with vegetation (Nelson & Thompson, 2004). Vegetation type (e.g. macrophyte or algae) may also be crucial for specialist taxa. Haliplus fulvus, indicator species in lake habitat 3110, is a tolerant beetle but a specialised algivorous species feeding on Chara sp. and Nitella sp. during its larval stages, hence, would be restricted to lake habitats hosting these macroalgae (Nillsson, 1996). Even under similar chemistry conditions, habitats at different successional stages will have different vegetation types and cover which in turn can favour specific feeding/functional groups of insects and therefore change how food webs are structured (Plakulnicka & Kruk, 2023). Acilius sulcatus, indicator species in lake habitat 3160, is known to avoid sites inhabited by fish (Foster & Friday, 2011), a preference shared with the Libellulidae Sympetrum danae (Boudot & Kalkman, 2015) an indicator species of the O/N region (Table 5). Drinan (2012) observed a prevalence of benthic predators like dragonflies in fishless lakes, with an abundance of Gastropoda and Ephemeroptera in lakes with fish in peatland regions, however, little effect on either abundance or diversity was reported. It is clear more research is needed to determine which environmental and ecological factors shape the invertebrate communities of lake habitats 3110 and 3160. Future research should include additional factors such as vegetation density and type (or stage of succession), presence of predators and prey and descriptions of the life history strategies of invertebrates and their habitat preferences in order to determine preferences between protected lake habitat types.

Despite being dominated by some generalist species, 15 of the 24 sites in this study hosted at least one species included in the Irish National Red List. Agabus arcticus, Boreonectes multilineatus, Dytiscus lapponicus and Cordulia aenea are all species previously recorded from similar waterbodies (Baars et al., 2014, Drinan, 2012, Towers, 2004), highlighting that they belong to recurrent communities strongly associated with peatland catchments and regions. However, the most frequently recorded species was the dragonfly Cordulia aenea (downy emerald dragonfly) (10 sites in total). This endangered species has previously been associated with lake habitat 3160 (O’Connnor, 2015), however, this was not found to be the case in the current study and was mostly recorded in sites assigned lake habitat 3110 (seven sites). C. aenea currently has a range in Ireland restricted to just three areas (Drinan, 2011; National Biodiversity Data Centre, 2023) one of which includes the study regions of Connemara Bog Complex SAC. Hence, a real association with either habitat type 3110 or 3160 needs to be reassessed considering the limited distribution of this rare species. The distribution of C. aenea likely affected SQS calculations, since it was the rarest species found in lake habitat 3110, hence habitats sampled outside of its range would have reported lower levels of this score. Furthermore, a confident rarity assessment needs to be estimated on the overall community of invertebrates (including several orders and families), however, in order to be accurate, the sample identification should be at a species-level which in turn requires a higher level of expertise and effort.

This study confirms the hypothesis that protected lake habitats 3110 and 3160 do differ in their overall Coleoptera and Odonata community assemblages although it is limited to a number of taxa and by region, with considerable overlap observed in the community as depicted by multivariate analysis. Despite this, lake habitat 3110 were characterised by indicator species Ischnura elegans, Pyrrhosoma nymphula, Sympetrum striolatum and Haliplus fulvus, whereas three species of beetles were significantly associated with lake habitat 3160: Acilius sulcatus, Gyrinus minutus and G. aeratus. Communities remained overall distinct in both sampling seasons with a higher number of indicator species observed in spring for lake habitat 3160 while most indicator species of lake habitat 3110 were found during summer. Environmental factors influenced the Coleoptera and Odonata communities in that indicators of lake habitat 3110 were associated with higher pH, and warmer larger waterbodies while species of lake habitat 3160 were typical of small turbid and deoxygenated waters, however, the influence of environmental parameters was limited.

This study represents one of the very first attempts to characterise the Odonata and Coleoptera communities of EU Habitat Directive protected lake habitats 3110 and 3160, a process traditionally entrusted to phytosociological methods. Results from this study suggest that while Odonata and Coleoptera communities represent an important monitoring tool which could improve the conservation condition assessment methods of these lake habitat types, inclusion of all invertebrate taxa as well as sites with varying catchment geologies and from across a broader geographical area within and/or outside Ireland would serve to strengthen their characterisation and thus habitat condition assessments. Seasonal sampling is necessary regardless of the habitat type to ensure the representation of different taxa with different spawning and emergent timings, however, future methodologies need to be habitat-specific, designed according to size and taxa, once their dominant fauna is fully characterised. Furthermore, a comprehensive classification including key physical features chemical parameters or ecological drivers of invertebrate community structure, such as vegetation types and structure, their influence on the fauna and how it changes with successional stage, top-down pressure by vertebrate predators such as fish and the generic differences in food web structure within each lake habitat is also required. Many unknowns still remain related to the classification of freshwater lake habitats, especially considering the multivariate nature of their communities structure, which goes beyond presence/absence of phytosociological orders or water chemistry condition. However, understanding the Odonata and Coleoptera community composition of these two lake habitat types will deepen the current knowledge around EU Habitat Directive protected lake habitat types and their future conservation needs.

Statements and Declarations

Fundings This project is funded under the EPA Research Programme 2021-2030 through the research project ‘Framework for characterizing oligotrophic lakes (3110 and 3160) using practical methods and assessment tools’ (2019-W-MS-44). The EPA Research Programme is a Government of Ireland initiative funded by the Department of the Environment, Climate and Communications. Additional research funding was also award by the National Parks and Wildlife Service (NPWS Ref: SPU 511 G06-2021) to cover the costs of water chemistry analysis.

Data availability The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Conflict of interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Author contributions Supervision: Heather Lally; Project administration: Heather Lally; Funding acquisition: Heather Lally; Conceptualization: Heather Lally, Martin Gammell; Methodology: Heather Lally, Martin Gammell; Writing - Review & Editing: Heather Lally, Martin Gammell, Emma Gray; Resources: Heather Lally; Validation: Heather Lally, Martin Gammell, Emma Gray; Visualization: Heather Lally, Martin Gammell, Emma Gray, Giovanni Cappelli; Formal analysis: Giovanni Cappelli; Investigation: Giovanni Cappelli; Data Curation: Giovanni Cappelli: Writing - Original Draft: Giovanni Cappelli.

Baars J-R, Murray DA, Hannigan E, Kelly-Quinn M (2014) Macroinvertebrate assemblages of small upland peatland lakes in Ireland. Biology and Environment: Proceedings of the Royal Irish Academy. JSTOR: 233–248
Beadle JM, Brown LE, Holden J (2015) Biodiversity and ecosystem functioning in natural bog pools and those created by rewetting schemes. WIREs Water 2:65–84
Biggs J, Fox G, Nicolet P (1998) A guide to the methods of the National Pond Survey. Pond Action
Boudot J-P, Kalkman VJ (2015) Atlas of the European dragonflies and damselflies. KNNV publishing
Cappelli G, Gray E, Gammell MP, Lally HT (2023) Physico-chemical characterisation of protected lake habitats: a matter of dystrophy
Cham SA (2012) Field guide to the larvae and exuviae of British dragonflies: damselflies (Zygoptera) and dragonflies (Anisoptera). British Dragonfly Society
Council of the European Union (1992) Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora. Official J Eur Union 206:7–50
De Caceres M, Legendre P (2009) Associations between species and groups of sites: indices and statistical inference. Ecology, http://sites.google.com/site/miqueldecaceres/
Desrochers A, van Duinen G-J (2006) Peatland Fauna. In: Wieder RK, Vitt DH (eds) Boreal Peatland Ecosystems. Springer, Berlin, Heidelberg, pp 67–100. https://doi.org/10.1007/978-3-540-31913-9_5.
Drinan TJ (2012) The impact of conifer plantation forestry on the ecology of peatland lakes. PhD Thesis, University College Cork
Drinan TJ, Nelson B, Tickner M, O’Donnell G, O’Halloran J, Harrison S (2011) First discovery of larvae of the Downy Emerald Cordulia aenea (L.) in Ireland and the species’ use of lakes in treeless blanket bog in Connemara, Co. Galway. J Br Dragonfly Soc 27:1–12
Dufrêne M, Legendre P (1997) Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecol Monogr Wiley Online Libr 67:345–366
European Commission (2013) Interpretation manual of European Union habitats–EUR 28. European Commission DG Environment Brussels
European Commission (2023) The Habitats Directive. EU measures to conserve Europe’s wild flora and fauna. https://environment.ec.europa.eu/topics/nature-and-biodiversity/habitats-directive_en
European Environment, Agency (2020) State of nature in the EU: results from reporting under the nature directives 2013 2018. Publications Office of the European Union, LU. https://data.europa.eu/doi/10.2800/088178
European Environment Agency (2016a) C1.1a Permanent oligotrophic waterbody with very soft-water species. Retrieved from: https://forum.eionet.europa.eu/european-red-list-habitats/library/terrestrial-habitats/c.-freshwater/
European Environment Agency (2016b) C1.4 Permanent dystrophic waterbody. Retrieved from: https://forum.eionet.europa.eu/european-red-list-habitats/library/terrestrial-habitats/c.-freshwater/
Fossitt A (2007) A Guide to Habitats in Ireland, 2007 reprint. The Heritage Council, Kilkenny
Foster GN, Friday L (2011) Keys to the water beetles of Britain and Ireland (part1). Field Studies Council, Shrewsbury
Foster GN, Nelson BH, O’Connor Á (2009) Ireland red list. No. 1, Water beetles. National Parks and Wildlife Service
Foster GN, Nelson BH, Bilton DT, Lott DA, Merritt R, Weyl RS, Eyre MD (1992) A classification and evaluation of Irish water beetle assemblages. Aquatic Conservation: Marine and Freshwater Ecosystems 2: 185–208
Foster G, Bilton D, Friday L (2014) Keys to the Adults of the Water Beetles of Britain and Ireland (Part 2)
Free G, Bowman J, McGarrigle M, Caroni R, Donnelly K, Tierney D, Trodd W, Little R (2009) The identification, characterization and conservation value of isoetid lakes in Ireland. Aquat Conservation: Mar Freshw Ecosyst Wiley Online Libr 19:264–273
Frelik A, Pakulnicka J (2015) Relations Between the Structure of Benthic Macro-Invertebrates and the Composition of Adult Water Beetle Diets from the Dytiscidae Family. Environ Entomol 44:1348–1357
Friday LE (1987) The diversity of macro invertebrate and macrophyte communities in ponds. Freshw Biol 18:87–104
Geological Survey Ireland Spatial Resources (2022) Retrieved from: https://dcenr.maps.arcgis.com
Gray E, Cappelli G, Gammell MP, Roden CM, Lally HT (2022) A review of dystrophic lake and pool habitat in Europe: An Irish perspective. J Nat Conserv 68:126189
Hannigan E, Kelly-Quinn M (2012) Composition and structure of macroinvertebrate communities in contrasting open-water habitats in Irish peatlands: implications for biodiversity conservation. Hydrobiologia 692:19–28
Hannigan E, Kelly-Quinn M (2014) Hydrochemical characteristics of the open-water habitats of selected Irish peatlands. Biology and Environment: Proceedings of the Royal Irish Academy. JSTOR: 109–116
Hannigan E, Mangan R, Kelly-Quinn M (2011) Evaluation of the success of mountain blanket bog pool restoration in terms of aquatic macroinvertebrates. Biology and Environment: Proceedings of the Royal Irish Academy. JSTOR: 95–105
Heino J (2008) Patterns of functional biodiversity and function-environment relationships in lake littoral macroinvertebrates. Limnol Oceanogr 53:1446–1455
Jones RI (2005) Limnology of humic waters: special theme or universal framework? SIL Proceedings, 1922–2010 Taylor & Francis 29: 51–60
Kassambara A (2023) _rstatix: Pipe-Friendly Framework for Basic Statistical Tests_. R package version 0.7.2, https://CRAN.R-project.org/package=rstatix
Kesti P, Hiltunen M, Strandberg U, Vesterinen J, Taipale S, Kankaala P (2022) Lake browning impacts community structure and essential fatty acid content of littoral invertebrates in boreal lakes. Hydrobiologia 849:967–984
Legendre P, Legendre L (2012) Numerical ecology. Elsevier
MacArthur RH, Wilson EO (1967) The Theory of Island Biogeography. Princeton University Press
Martinez Arbizu P (2017) PairwiseAdonis: pairwise multilevel comparison using Adonis. R Package version 0.4.1
McFarland B, Carse F, Sandin L (2010) Littoral macroinvertebrates as indicators of lake acidification within the UK. Aquat Conservation: Mar Freshw Ecosyst John Wiley Sons Ltd 20: https://www.cabdirect.org/cabdirect/abstract/20103162895
Met Éireann (2022) Historical Data. Retrieved from: https://www.met.ie/climate/available-data/historical-data
National Biodiversity Data Centre (2023) Retrieved from: https://maps.biodiversityireland.ie/Species
National Parks and Wildlife Service (2013) Commonage areas dataset. Retrieved from: https://www.npws.ie/maps-and-data/habitat-and-species-data
National Parks and Wildlife Service (2015a) Connemara Bog Complex SAC 002034, Site Synopsis. Retrieved from: https://www.npws.ie/protected-sites/sac/002034
National Parks and Wildlife Service (2015b) Conservation Objectives: Connemara Bog Complex SAC. Retrieved from: https://www.npws.ie/protected-sites/sac/002034
National Parks and Wildlife Service (2017a) Owenduff/Nephin Complex SAC 000534, Site Synopsis. Retrieved from: https://www.npws.ie/protected-sites/sac/000534
National Parks and Wildlife Service (2017b) Conservation Objectives: Owenduff/Nephin Complex SAC 000534. Retrieved from: https://www.npws.ie/protected-sites/sac/000534
National Parks and Wildlife Service (2019) The Status of EU Protected Habitats and Species in Ireland. Volume 2: Habitat Assessments. Unpublished NPWS report
., SAC and National Parks and Wildlife Service, Datasheets SPA (2020) Retrieved from: https://www.npws.ie/maps-and-data/designated-site-data/sac-and-spa-datasheets-downloads
Nelson B, Thompson R (2004) The natural history of Ireland’s dragonflies. National Museums and Galleries of Northern Ireland
Nelson B, Ronayne C, Thompson RS (2011) Damselflies & Dragonflies (Odonata). National Parks and Wildlife Service
Nilsson AN, Elmberg J, Sjoberg K (1994) Abundance and Species Richness Patterns of Predaceous Diving Beetles (Coleoptera, Dytiscidae) in Swedish Lakes. J Biogeogr Wiley 21:197–206
Nilsson A (1997) Aquatic insects of North Europe: A taxonomic handbook. Apollo books
NPWS (2019) The Status of EU Protected Habitats and Species in Ireland, vol 2. Habitat Assessments
O’Connor A (2015) Habitats Directive Annex I lake habitats: a working interpretation for the purposes of site specific conservation objectives and Article 17 reporting. National Parks and Wildlife Service, Department of Arts, Heritage and Gaeltacht, Dublin, Ireland
Ozoliņš D, Skuja A, Jēkabsone J, Kokorite I, Avotins A, Poikane S (2021) How to Assess the Ecological Status of Highly Humic Lakes? Development of a New Method Based on Benthic Invertebrates. Water Multidisciplinary Digital Publishing Institute 13:223
Pakulnicka J, Zawal A (2007) Chrząszcze wodne (Coleoptera) rezerwatu Jezioro Szare i zbiorników usytuowanych w jego sąsiedztwie. Parki Nar Rez Przyr 25:121–133
Pakulnicka J, Zawal A (2018) Model of disharmonic succession of dystrophic lakes based on aquatic beetle fauna (Coleoptera). Mar Freshw Res CSIRO Publishing 70:195–211
Pakulnicka J, Kruk M (2023) Regional differences in water beetle communities networks settling in dystrophic lakes in northern Poland. Sci Rep Nat Publishing Group 13:12699
Core Team R (2016) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria
Ranta E (1985) Communities of Water-Beetles in Different Kinds of Waters in Finland. Proceedings of the Academy of Natural Sciences of Philadelphia Academy of Natural Sciences 137: 33–45
Soldán T, Bojková J, Vrba J, Bitušík P, Chvojka P, Papáček M, Peltanová J, Sychra J, Tátosová J (2012) Aquatic insects of the Bohemian Forest glacial lakes: Diversity, long-term changes, and influence of acidification. Silva Gabreta 18:123–283
Towers NM (2004) Invertebrate community structure along a habitat-patch size gradient within a bog pool complex. Doctoral dissertation, University of Edinburgh; College of Science and Engineering

No competing interests reported.

Cappellietal.Supplementarymaterial.docx

Using Coleoptera and Odonata as a monitoring tool for improving protected lake habitat characterisation: a case study from Ireland

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Results

Discussion

Conclusions and Recommendations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1