Attractivity of various artificial light sources to caddisfly (Trichoptera) species and its importance in their sampling and conservation

The artificial light sources are useful tools for sampling night active insects, however, they also possess potential environmental risks in their habitats. To test their applicability and evaluate environmental risk for caddisflies (Trichoptera), the attractivity of different portable light traps working with LED, UV and mixed-white light sources was studied and compared with attractivity of traditionally used mercury-vapour lamp (in Jermy-type light trap), which is tool of sampling and street-light. Analysing 1135 caught individuals of 19 species light sources emitting different wavelength spectra showed different attractivity and selectivity on caddisflies both on species and family levels. Attractivity of mercury-vapour lamp was generally lower than the other tested light sources. We found that the most attractive wavelength range for caddisflies is between 360 and 407 nm. One of the tested LED and mixed-white lamps together could cover this spectrum and a high and wide spectral peak of mixed-white light source between 375 and 391 nm resulted additional catches considering both species and number of individuals. Lamps emitting between 360 and 407 nm may be both a useful tool for sampling caddisflies and dangerous source of light pollution along lowland water courses where the sampled species are common and widespread.


Introduction
Traps using different light sources are traditionally used in both basic and applied entomological surveys since they can operating-and colour temperature and spectral distribution. There are quite a lot of data on the effects of weather (temperature, wind strength, humidity, air pressure, etc.), moon phases and other environmental factors on the efficiency of light traps (Kiss 2002;Nowinszky 2003;Nowinszky et al. 2014), while spectral distribution is one of the most important parameters affecting the attractivity and selectivity of light sources (Nowinszky 2003). Light sources emitting relatively large amount of UV light generally perform better in catching of night-active insects than others (Blomberg et al. 1976;Ashfaq et al. 2005;Cowan and Gries 2009), since beyond the visibility and the spectral distribution of the lamp, at least a minimal amounts of UV radiation are a crucial factor of the attractiveness (Barghini 2008, Barghini andMedeiros 2012).
The most widely used mercury-vapour lamp is an appropriate tool for attracting species of Lepidoptera (van Langevelde et al. 2011;Bates et al. 2013;Brehm 2017;Infusino et al. 2017), Coleoptera (Kádár and Szél 1995;Band et al. 2019) and Trichoptera (Ward et al. 1996;Kiss 2002;Nowinszky 2003;Blahnik and Holzenthal 2004;Calor and Mariano 2012) and even Hemiptera, Hymenoptera, Diptera and Neuroptera (Ramamurthy et al. 2010;van Grunsven et al. 2014). Mercury-vapour lamps attract more species and individuals of moths than the UV light alone probably since they have high peaks both in the UV and in the visible spectrum (Jonason et al. 2014;Tikoca et al. 2016;Brehm 2017). In case of Lepidoptera blacklight (BL) tube (Belton and Kempster 1963;Mészáros 1966;Bhandari et al. 2017) and LED lights with wide spectra including UV radiation (Brehm 2017;Infusino et al. 2017) can be also effective. For attracting mosquitos, the incandescent lights, for agricultural pests the green lights, and for some aquatic insects the cyalume lightsticks are the most commonly used light sources (Epsky et al. 2008).
The high attractivity of the mercury-vapour lamp for caddisflies (Mikkola 1972;Blomberg 1976;Walker and Galbreath 1979) and the role of its UV spectra were also proven (e.g., Chernyshev 1961;Calor and Mariano 2012;Pohe et al. 2017), thus this trap type and black light lamps became the most widely used standard light source types in caddisfly sampling (Blahnik and Holzenthal 2004;Calor and Mariano 2012). According to Mikkola (1972) the eyes of Trichoptera are similar to Lepidoptera and sensitive for 350-600 nm wavelength, with a maximum around 550 nm, but the effect of spectral distribution of light sources on caddisflies is poorly known.
Although it is one of the most effective tools for sampling many taxa mercury-vapour lamps have been phased out in Europe since April 2005 as per EU Regulation 245/2009, which favours more cost-effective and energy-efficient LED technology (van Grunsven et al. 2014;Donatello et al. 2019;Haddock et al. 2019). Furthermore, the manufacturing and trade of mercury lamps have been gradually prohibited worldwide since 2020 by Minamata Convention Agreement (www.mercuryconvention.org), thus it should be replaced also in entomological studies.
Here we compared the efficiency and selectivity of different light sources to find the most appropriate method and tool for the caddisfly sampling and establish the most attractive wavelength range for them using mercury vapour lamp as kind of "traditional control method". Beyond that we provided recommendations on use of different artificial lights considering conservation of caddisfly fauna. Finally large amount of distribution data of caddisfly species was also collected in a mainly unknown underinvestigated region of the Carpathian Lowland (Transcarpathia, W Ukraine).

Sampling site
The samplings were carried out in the Velyka Dobron' Game Reserve, in the eastern part of the Bereg Plain (West Ukraine) (Fig. 1). The Bereg Plain belongs to the Upper Tisza Region on the northeastern part of the Pannonian Lowland. This area has cool, humid continental climate strongly affected by the nearby Carpathians (Simon 1953).
The study area was the margin of the former Szernye Marsh thus it is still water-logged and wet despite the drainage (Szanyi et al. 2015). The Szernye Canal constructed during the drainage of the marsh and the Latorca River flow through the study area. On the floodplain, these watercourses provide a permanent water supply with a maximum in spring. Beyond that, the reserve is surrounded by many artificial canals containing water periodically, and ephemeral ponds and puddles.
The light traps were placed on the banks of the Szernye Canal, near to the Latorca River in order that samplings represents both artificial and natural waterflows of the region.

Sampling method
Between 11th August and 9th September 2018 14 samplings were made with six portable light traps and one Jermy-type light-trap. The Jermy-type light trap worked with a mercuryvapour bulb (80 W Hg), while the portable light traps contained different light sources: one mixed-white light tube, 2 types of LED and 3 types of UV light tubes with different spectral distribution. The spectral distribution of the tested light sources were measured with a Red Tide spectrophotometer working with a linear silicon CCD (Charge Coupled Device) detector. It gives the wavelengths in nanometres and describes the intensity with counts (Fig. 2).
The mercury-vapour lamp (MV) had many peak wavelengths in a wide range (around 366, 405, 408, 436, 492, 547, 578, 594, 621, 697 nm-s). The mixed-white light source (W) had high peaks both in the visible (around 405, 436, 546 nm-s) and in the UV spectrum (around 370 nm). The UV1 and UV2 light sources were similar, and they showed only minor peaks in the visible compared to the UV spectrum. Their main peaks were similar with those of W light source (around 370 nm), but narrower. The peaks of the UV1 were higher in the UV range, while the peaks of the UV2 were higher in the visible spectrum. UV3 covered almost the whole UV-A spectrum, but only with low intensity without high peaks. The two tested LED light sources had similar peaks in the blue spectrum range (around 395 nm), although the peak of the L2 light source was slightly wider (Fig. 2).
Location of the mercury-vapour lamps are limited by available grid power, or they can work with a generator. The small portable light traps can be easily located at different habitats, since they work with 12 V batteries. These portable light traps were placed about 100 m distance from each other along the banks of the Szernye Canal. To eliminate the effect of sampling sites they were relocated site by site in each sampling time. Portable light traps were smaller versions of the Robinson-type light trap (Upton and Mantle 2010). The light tubes were mounted vertically with the top of a large funnel and have three radiating baffles around it. The funnel fitted on a plastic bucket, and had a transparent cover mounted above it to keep the rain out. The structure of the Jermy-type light trap is described in Szentkirályi (2002).
The collected insects were stored in 70 V/V% ethanol till identification carried out in laboratory.

Data analyses
The caught caddisflies were identified based on the keys of Malicky (2004). The nomenclature of Nógrádi and Uherkovich (2002) was followed. In case of Hydropsyche spp. males were identified at species level, while females were identified only at genera level since the extreme similarity of the female genitalia of different species.
The fauna was characterised with checklist and relative frequencies of the caught species and families. To characterise efficiency and selectivity of the tested light sources the

Results
Totally 1135 individuals of 19 caddisfly species belonging to six families were caught by the tested light traps. The most species rich family were the Hydropsychidae with 6 species. Leptoceridae family showed similarly large species richness, however the abundances of species were especially low. The most abundant family was also the Hydropsychidae followed by Limnephilidae. Based on male catches the most abundant Hydropsychidae species was Hydropsyche modesta followed Hydropsyche contubernalis and Hydropsyche bulgaromanorum. Among Limnephilidae Glyphotaelius pellucidus was dominant. Only one or two species belonging to Psychomyiidae, Ecnomidae and Phryganeidae families was caught only with low abundances (Table 1).
Traps caught both sexes with near 1:1 ratio, since the mean percentage of males was 45.5 (± 18.3) % (Table 1). Species showed different sex ratios. Ecnomus tenellus and total and mean number of species and individuals (species/ and individuals/sample) were calculated, and the number of unique species (species attracted only by a given light source) also was provided by light sources. The mean number of caught individuals (individual/samples) was also calculated both for caddisfly families and species. To evaluate how different light sources sample diversity Shannon-Wiener and Hill's diversity indices were calculated and compared by different methods. For calculation of indices an online diversity calculator developed by Young T. M. was used (alyoung.com/labs/biodiversity_calculator.html).
Light sources were compared using the above variables. For comparisons Kruskal-Wallis test was used since data didn't fulfil the assumptions of parametric tests. If the test showed significant differences the pairwise comparison was made with Mann-Whitney U-test using Statistica 7 program.  Table 3).
The number of individuals attracted by tested light sources differed between a smaller range (127-183) and the mean number of caught individuals also did not show significant differences (KW: H(6;98) = 2.6317, p = 0.8534) ( Table 2).
During the study only three species showed light source specific presence ( Table 2). Two especially rare Hydropsyche species (H. ornatula and H. pellucidula) were attracted only with the mixed-white light source (W), while the also rare O. furva was attracted only with L1 light. Beyond that, specimens of Mystacides azureus were caught only with mixedwhite (W) and UV1 traps while individuals of S. punctatus Hydropsyche spp. showed 1:4 and 1:3, contrary Agrypnia varia and Limnephilus flavicornis showed 3:1 and 3:2 ratios. Ratios of other species were average (Table 1).
Although mixed-white light (W) attracted almost all the species registered during the study (17/19, except Setodes punctatus and Oecetis furva) and the number of caught species varied between 9 and 17, but the efficiency provided by mean number of caught species per sample did not differ (KW: H(6;98) = 4.8261, p = 0.5663). Considering species number, the less efficient light sources were the UV1 and the mercury-vapour lamps (Table 2). Using diversity indices different pattern was found. Based on mean value of Shannon-Wiener and Hill's diversity indexes of samples there were no significant differences between the tested sampling methods (KW: H(6;98) = 2.8543-4.489, p = 0.6108-0.8911), individuals than others. Contrary, in case of Oecetis notata mercury-vapour light bulb attracted significantly more individuals than other light sources from which L1, UV1 and UV3 bulbs did not attract any individuals ( Table 4). In case of the other seven species obvious specificity could not be detected.

Discussion
In the present investigation the mercury-vapour lamp (MV) seems to be less effective than the other studied light sources. It attracted the least number of individuals and less species than the most UV, LED and mixed-white light sources. Additionally, 9 species: Cheumatopsyche lepida, Hydropsyche ornatula, Psychomyia pusilla, Grammotaulius nigropunctatus, Mystacides azureus, Hydropsyche pellucidula, Lype phaeopa, Trichostegia minor and Oecetis furva were collected only with alternative light sources and the last 4 of them have not been caught formerly, during the 3-year samplings with mercury-vapour lamp. Szanyi 2018, 2019). Researchers generally use mercuryvapour lamps and black light lamps for collecting caddisflies and also moths (Blahnik and Holzenthal 2004;Calor and Mariano 2012) since the attraction of caddisflies to near-ultraviolet radiation has been known for a long time (Chernyshev 1961;Mikkola 1972;Blomberg 1976;Walker and Galbreath 1979;Pohe et al. 2017). Although between them mercury-vapour lamps are seen more efficient (Mikkola 1972;Blomberg 1976;Walker and Galbreath 1979), but shorter wavelength light sources generally attract higher were caught only with mercury-vapour (MV) and UV2 light sources. Other species were attracted with at least three types of tested light sources. Considering the number of caught individuals the selectivity of light sources could be detected both on family and species level. The mercury-vapour light source attracted significantly more individuals of Leptoceridae than other light sources except mixed-white light bulb, but the difference was conspicuous also in this case. Contrary, in case of Limnephilidae, Phryganeidae and Psychomyiidae families mercury-vapour light source attracted less individuals than the other light sources. In case of Limnephilidae the differences were significant in all comparisons. In case of Phryganeidae mercury-vapour lamp attracted significantly less individuals than mixed-white and UV (UV1-3) lights, while the differences were not significant in case of L1 and L2 lights. Considering Psychomyiidae UV3 light source attracted the most individuals, and the differences were significant in case of UV1, UV2 and mercury-vapour lights. In case of Ecnomidae and Hydropsychidae there were no remarkable differences between the selectivity of the tested light sources (Fig. 3).
Among 19 caught species, 10 were caught with more than 10 individuals. Considering mean number of caught individuals three of them were differently attracted by different light sources. UV1 light attracted the most T. minor individuals. Traps with UV1, UV2 and mixed-white lights caught significantly more individuals than traps with mercury-vapour bulb that did not catch any individuals. Individuals of G. pellucidus attracted by all tested light sources except mercury-vapour bulb, that attracted significantly less 12 135 2.6 (0.4) 9.6 (2.0) 0 it was the less effective one. Individuals of Psychomyiidae were highly attracted by UV3 light while UV1, UV2 and mercury-vapour lamps were totally inefficient. Chernyshev (1961) reported different attractivity of light sources.
In their experiment the black light blue lamp was highly attractive for specimens of Hydroptilidae, Limnephilidae and Psychomyiidae families and in contrast to our results mercury-vapour lamp showed higher attractivity for family of Hydropsychidae. Mikkola (1972) also observed, that the yellow radiation was highly attractive for species of Leptoceridae family. In our investigation, only two light sources (mercury-vapour and mixed-white lights) emitted in the range of yellow radiation and the individuals of Leptoceridae family attracted with higher number exclusively to these light sources. The importance of the green-yellow light was also proved for moths e.g. Pandemis dumetanai Tr. and Pandemis heparana Den. et Schiff. (Járfás 1977;Nowinszky 2013). Laboratory measurements showed two number of species and higher abundance (van Langevelde et al. 2011;Barghini and Medeiros 2012;Kadlec et al. 2016;Brehm 2017). The tested alternative light sources emitted in this shorter wavelength spectra which can explain their higher efficiency and lower selectivity for different caddisfly species. Regarding diversity indexes the samples taken with mercury-vapour lamp showed the highest diversity, that can explain the wide use of this type of light source. The differences might be caused by the attractivity of the other light sources for Limnephilidae species that appeared in their samples with high abundances decreasing the value of diversity indices despite of the higher species diversity of samples. Considering number of caught individuals belong to Ecnomidae and Hydropsychidae families there were no differences between attractivity of different light sources. In case of Leptoceridae mercury-vapour lamp was the most efficient, while in case of Limnephilidae and Phryganeidae The wavelength range of the L1 and mixed-white (W) light sources together cover the whole spectrum that attracted all the collected caddisfly species. This spectrum and light sources provide the most appropriate wavelength range and tool for caddisfly sampling and parallelly cause largest risk of light pollution for caddisflies when they are used in public lighting.
Since the habitat loss and fragmentation and water pollution endanger caddisflies and other aquatic insects, we should reveal all other threatening factors and should provide solution for them. Caddisfly species trapped in the study area were all common and widespread in small watercourses of the Carpathian lowlands. Along these wetlands use of the tested light sources and any light sources emit between 360 and 407 nm should avoid and lamps work in different wavelength range are suggested. On the other hand, we can collect valuable data for monitoring and conservation planning with use the most effective light sources.
Acknowledgements The work of Kálmán Szanyi was supported by International Visegrad Fund (Scholarship number: 52010905) 2020.
Authors' contributions SK, NA and SS designed the experiment and analysed the data, all authors contributed to writing of the manuscript.
Funding The study was supported by International Visegrad Fund: Visegrad Scholarship Program (2020) (Serial Number: 52010905).

Open access funding provided by University of Debrecen.
Data availability Not applicable.
Code availability During the analysis SynTax 2000, and IndVal 2.0 major peaks in the spectral sensitivity of moth species, one UV range between 360 and 365 nm and another in the green-yellow range between 515 and 575 nm (Agee 1973;McFarlane and Eaton 1973;Teel et al. 1976;Pappas and Eaton 1977;Eguchi et al. 1982).
The UV1, UV2 and L1 lights together caught all species (10) attracted with mercury-vapour lamp. The wavelength range of these light sources is between 360 and 407 nm (Fig. 1), thus in case of mercury-vapour lamp these wavelength range can be seen attractive to caddisflies while light above 407 nm is inefficient. The mixed-white light source (W) attracted almost all collected species (17 from 19). The spectral distribution of this light source extends from 360 to 577 nm. Above the wavelength of 391 nm, its spectral peaks are smaller, narrower, and similar to those that the mercury-vapour lamp showed (Fig. 1). It suggests that the high efficiency of the W light source due to its peaks below 391 nm. Differences between UV1, UV2 and L1 lights, which together perform as mercury-vapour lamps, and mixed-white light could be found between 375 and 391 nm. In this range mixed-white light have wider and higher spectral peak, that might result in its higher attractivity. Summarily the 360-391 nm wavelength range attracted most of the collected caddisfly species. It supports result of Mikkola (1972), who found that the eyes of caddisflies are sensitive in the spectrum of 350-600 nm.

Conflicts of interest
The authors declare that they have no conflict of interest.

Ethics approval Compliance with Ethical Standards
Informed consent Informed consent was obtained from all individual participants included in the study.
Research involving human participants and/or animals This article does not contain any studies with human participants performed by any of the authors. All applicable international, national, and/or institutional guidelines for the care and use of animals were followed.
Consent to participate Not applicable.

Consent for publication Not applicable.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/.