Decline of Pelophylax lessonae in mixed populations of water frogs over the last 50 years

Abstract Two water frog species: the pool frog Pelophylax lessonae (L) and the marsh frog P. ridibundus (R) occur sympatrically in Central Europe and form mixed populations (genetic systems) with their hybrid, the edible frog P. esculentus (E). The aim of the study was to assess the species composition of water frogs in urban and rural populations and compare our current findings with the results of previous studies. We surveyed the same sites that were investigated by Berger et al. in 1962–1970 (Poznań, urban landscape) and 1977–1997 (Dezydery Chłapowski Landscape Park, rural landscape). Because some ponds surveyed in the past were destroyed or dried-up, we also explored others located in the adjacent areas. We captured frogs during breeding seasons 2020 and 2021 and identified them by the nuclear marker gene SAI-1. We found three types of populations in the urban area: R-E, E-E and R-E-L and four in the rural area: R-E, L-E, E-E and R-E-L. Compared to the historical data, we found a drastic decrease in the frequency of P. lessonae in urban and rural landscapes: from 89.1% and 68% to 2.7% and 1.8%, respectively. At the same time, the frequency of P. ridibundus increased from 2.2% and 0.1% to 40% and 29.6%, respectively. A similar pattern was found for P. esculentus whose frequency increased from 8.7% and 31.9% to 57.3% and 68.6%, respectively. Additionally, we confirmed the presence of a cryptogenic Balkan water frog, P. kurtmuelleri, which was recently discovered in south-western Poland. The frequency of SAI-1 allele specific for this taxon reached 7.3%. The patterns found in both types of landscapes are in line with the current situation of both parental species in Europe. Such dynamic changes show the need for long-term monitoring of the population compositions of water frogs, what is crucial for their conservation management.


Introduction
In the mid-twentieth century, Berger (1973) discovered that the most common species of the frog in Europe, the edible frog Pelophylax esculentus (Linnaeus, 1758) (formerly Rana esculenta), is not a species but a hybrid between the pool frog P. lessonae (Camerano, 1882) and the marsh frog P. ridibundus (Pallas, 1771).Hybrids reproduce by hybridogenesis, i.e. by backcrosses with one of the parental species that co-occurs in the same population (Berger 1973; for review see Graf & Polls Pelaz 1989).During gametogenesis, hybrids exclude one of the parental genomes from the germline cells (precisely from gonocytes) prior to meiosis and transmit the other (non-recombined) genome to the gametes (Uzzell & Berger 1975;Chmielewska et al. 2018Chmielewska et al. , 2022)).When the genotype belonging to the species present in the mixed population is eliminated, the hybrid genotype is reproduced and maintained in a population as a constantly renewing F1 generation.Although the parental species may occur sympatrically, they often are ecologically isolated (Pagano et al. 2001;Holenweg Peter et al. 2002).Pelophylax lessonae, the smallest of the water frogs, inhabits mainly shallow water bodies, including those that dry out periodically.In contrast, P. ridibundus prefers larger and deeper water reservoirs, such as lakes, oxbows or rivers, and artificial ones, including fishponds or clay and gravel pits (Rybacki & Berger 1994;Socha & Ogielska 2010;Jośko & Pabijan 2020).Their hybrid, P. esculentus, is a generalist that lives syntopically with one of the parental species and forms the following mixed populations (genetic systems): lessonae-esculentus (L-E), ridibundusesculentus (R-E) or, rarely, ridibundus-esculentus-lessonae (R-E-L, most often formed temporarily during mating seasons) (Rybacki & Berger 2001).Less frequently, pure esculenta-esculenta (E-E) populations emerge due to different combinations of gametes produced by diploid (RL) and triploid (RRL and/or RLL) individuals (Hoffmann et al. 2015;Dedukh et al. 2022).
Although water frogs belonging to the P. esculentus complex are considered one of the most common in Europe, due to their complex relationships and morphological similarity (Kierzkowski et al. 2011), they are often grouped as one taxon in many ecological studies on amphibian assemblages (e.g.Hartel et al. 2010;Kaczmarski et al. 2020;Konowalik et al. 2020).Thus, there is lack of broader knowledge of long-term trends in their local populations.This makes it difficult to determine the correct status of their protection at the regional level (e.g. in Poland, see discussion) and implementation of the proper protection plan for particular taxa.
They, however, vary in protection status according to EU Habitats Directive (Directive 1992).Pelophylax lessonae is listed in Annex IV (i.e.strict protection within the EU), while P. esculentus and P. ridibundus are listed in Annex V ("Member States must ensure that their exploitation and taking in the wild is compatible with maintaining them in a favourable conservation status").Furthermore, all three taxa are not globally threatened: P. lessonae and P. ridibundus are in "least concern" (LC) category according to IUCN, while P. esculentus is not included on the list (IUCN SSC Amphibian Specialist Group 2022, 2023).However, the population trend of P. lessonae is evidently decreasing, while that of P. ridibundus is increasing.Despite the wide distributional range, local population declines of P. lessonae were noted (Meek 2021), and now this species is considered endangered in some European countries (Plötner 2018;Dufresnes et al. 2020).At the same time, the population of P. ridibundus (sensu lato) became invasive in western and southern Europe due to multiple introductions (Holsbeek et al. 2008;Dufresnes et al. 2018Dufresnes et al. , 2020;;Dufresnes & Dubey 2020;Bruni et al. 2020;Jelić et al. 2022).
In this study, we focused on shifts in taxonomic composition of water frogs that occurred over the last 50 years in urban and rural habitats in western Poland.In order to provide reliable comparative data, we conducted our research in Poznań city and Chłapowski Landscape Park -areas which were intensively studied since the middle of the 20 th century (Berger 1988;Berger & Berger 1992;Berger & Rybacki 1998).In the past, L-E populations dominated therein.In this system, P. esculentus produces mainly haploid gametes with non-recombined R genomes, while the L genome is excluded during gametogenesis (Tunner & Heppich 1981).Mating with P. lessonae ensures hybrid survival and thereby the stability of the L-E systems.Initially, this area abounds in surface water due to a young-glacial landscape characteristic (Vistula Glaciation).However, at the turn of the 19th and 20th centuries, significant land use changes occurred in the Greater Poland region (including the study area).The number of small water bodies, which are habitats of P. lessonae, decreased from 11,061 to 2490 between 1894and 1961(77.5% loss, Stasiak 1991after Rybacki & Berger 2003).Later on, between the sixties and the nineties of the 20 th century, there was another decrease in small water reservoirs in the Chłapowski Landscape Park by 19.1% (Rybacki & Berger 2003).Of the remaining ponds, 88% were degraded, mainly due to water pollution caused by chemical fertilizers, pesticides, littering, and sewage (Rybacki & Berger 2003).Detailed research on two small ponds in this area (of which one -Rogaczewo -is also under study) revealed the collapse of ten amphibian species populations between 1977 and 1985 due to chemical pollution (Berger 1989).Moreover, at those times, Pelophylax frogs were intensively harvested in Poland as a source of frog legs.In 1964, 46 000 kg of water frogs (ca.840 000 individuals) were exported to France, however these numbers decreased to 766 kg in 1977 due to overexploitation (Głowaciński et al. 1980).This was the first significant decline in the amphibian population recorded in Poland in the 20 th century.Since legal protection commenced in 1985, this practice is no longer a threat in Poland.
Here, we aimed to assess the current species composition of water frogs in urban and rural landscapes and to compare our findings with previous studies conducted in the same areas.Furthermore, we discuss the threats that affect Pelophylax frogs nowadays, propose changes in their protection status at the national level and suggest the implementation of P. lessonae conservation management in Poland.

Materials and methods
The study was conducted in western Poland, in the same areas that were surveyed in 1962-1970(Poznań, urban landscape) and 1977-1997 (Dezydery Chłapowski Landscape Park, rural landscape) (Berger 1988;Berger & Berger 1992;Berger & Rybacki 1998).Poznań is one of the largest cities in Poland (52.41,16.93)with an area of 261.9 km 2 (Figure 1(a)).The city population is 532 048 and has increased by over 120 000 since 1960.Poznań is situated in lowlands (50-154 m a.s.l.) in the Warta River Valley.Several lakes and numerous small water bodies are within the city.Ca. 15% of the city is covered by forests (Kaczmarek et al. 2014).Chłapowski Landscape Park (52.06, 16.82) is located in lowlands (70-95 m a.s.l.) 40 km south from Poznań (Figure 1(b)).The area of the Park is 173.2 km 2 .This protected area includes a unique agricultural landscape with a network of mid-field shelterbelts and forest patches (Kujawa et al. 2021).More than 200 small water bodies and watercourses occur in the Park, of which some are currently dry.
We selected two sets of ponds studied in the past in both areas (Figure 1(a,b)).However, since some of them were destroyed or dried-up, we also explored all the other ones situated in the adjacent areas (up to 300 m from the original pond).In total, 23 ponds were studied including 13 ponds studied by Berger and 10 new ones, and they were grouped in five study sites in urban and rural areas (Table I, see table S1 for pond characteristics).In 2020 and 2021, we captured at least 20 frogs per study site.Similar to Berger's research, frogs were collected during breeding season, i.e. in May and June.We caught them from the shore and the shallows or from the dinghy, by hand or by hand net, at night after being blinded by a flashlight, while Berger collected frogs using hand net or fishing rod (with barbless hook) during the day.Buccal swabs were taken from each frog and fixed in 80% ethanol.After the procedure, the frogs were released to their habitats.
In their study, Berger identified frogs based on morphology (Berger 1988;Berger & Berger 1992;Berger & Rybacki 1998).In the current study, species identification was carried out based on the size

96
K. Kolenda et al. polymorphism of serum albumin intron-1 (SAI-1) (Hauswaldt et al. 2012).This approach is a simple PCR-based method of species examination on agarose gel: DNA fragment of P. lessonae is approximately 300 bp, while that of P. ridibundus ca.850 bp.Both variants occur in P. esculentus.Recently, a high frequency of an allele of SAI-1 specific to P. kurtmuelleri (ca.700 bp) was found in southwestern Poland; however, its origin (allo-or autochtomus) is under debate (Kolenda et al. 2017a).Thus, for the purpose of this study, any P. kurtmuelleri alleles are considered as P. ridibundus (see Kolenda et al. 2017a and literature cited therein for details about taxonomical status of P. kurtmuelleri).DNA was extracted using the GeneMATRIX TISSUE DNA Purification Kit (EURx Ltd., Poland).DNA amplification of SAI-1 was performed with the primers Pel-SA-F1 and Pel-SA-R2 designed by Hauswaldt et al. (2012).The PCR reaction was done following the protocol of Kolenda et al. (2017a).PCR products were separated electrophoretically on a 1.5% agarose gel and compared with a 100 bp size marker.
A contingency table and Pearson chi 2 test were used to compare changes in frequency of water frog composition between current and historical data.In most cases, Berger provides only a range of years of study on given area.In case of three sites: Łuszkowo, Rogaczewo and Rąbinek, we had data from his two study periods (separated nine or more years).Because population structure was the same or only slightly differ (but the same species dominated), we combined this data into one set (see supplementary Table S2 for details).The analysis was performed in PAST 4.08 software (Hammer et al. 2001).

Population composition in 2020-2021
A total of 354 frogs were collected, including 185 in urban and 169 in rural sites (Table I, S3).Frogs were found in 9 out of the 10 study sites.All were successfully PCR-genotyped by nuclear marker gene SAI-1.The most common were P. esculentus (62.7%), followed by P. ridibundus (35.0%) and P. lessonae (2.3%).Three types of mixed populations were noted in the urban area: R-E (two sites, Naramowice and Sołacz), E-E (one site, Edwardowo) and R-E-L (one site, Junikowo), while four were found in the rural area: R-E (two sites, Zbęchy and Łuszkowo), L-E (one site, Rąbinek), E-E (one site, Rogaczewo) and R-E-L (one site, Turew).

Comparison with previous studies
In comparison to historical data, statistically significant changes in the frequency of water frog compositions in both habitats were found (Pearson chi 2 test: urban areas: chi 2 : 154,5, df: 2, p < 0.001, rural areas: chi 2 : 107.9, df: 2, p < 0.001; Table I, Figure 1 (c,d)).Overall, in the last century, P. lessonae dominated among water frog populations in the urban and rural areas reaching 89.1% and 68%, respectively.Currently, the frequency dramatically decreased to 2.7% and 1.8%, respectively.At the same time, frequency of P. ridibundus increased from 2.2% and 0.1% in urban and rural sites to 40% and 29.6%, respectively.Similar pattern was found in P. esculentus; its frequency increased from 8.7% and 31.9% to 57.3% and 68.6%, respectively.
When considering particular study sites, single individuals of P. lessonae persist in one population in the urban area (Junikowo) and in two populations in the rural area (Turew and Rąbinek) (Figure 2).Importantly, none of the study site was colonized by P. lessonae over time.Pelophylax ridibundus survived at the only site where it occurred in the past (Naramowice) and appeared in two others in the urban area (Sołacz and Junikowo) and three in the rural area (Turew, Zbęchy and Łuszkowo) (Figure 2).Pelophylax esculentus survived at all sites which they had inhabited during the earlier study, except for Żurawiniec in the urban area where all

98
K. Kolenda et al. water frogs disappeared after the desiccation of the pond in 2014 and its reconstruction in 2016 (Figures 2 and 3).

Discussion
Our comparative study clearly shows long-term shifts in population composition of different genetic systems of water frogs in urban and rural landscapes in western Poland: while the population of P. lessonae decreased, the share of P. ridibundus and P. esculentus increased.Previous studies suggested that the environment is a crucial factor establishing the composition of water frog systems (Berger 1988;Berger & Berger 1992;Berger & Rybacki 1998;Holenweg Peter et al. 2002).Currently, we found a common occurrence of P. ridibundus in relatively small water bodies (see Table S1 for details) that were previously considered as typical habitat of the L-E system (Berger & Berger 1992;Berger & Rybacki 1998).
In addition to taxonomic changes, we also found a marked decline in water frog abundance.Although we did not carry out detailed studies on the population sizes, it is worth noting that most of the studied populations were now small, usually not exceeding several dozen individuals.In contrast, in the past, some of these populations numbered hundreds or thousands of individuals (Table I).Furthermore, the presence of P. kurtmuelleri found in the studied areas suggests its broader range in Poland (for details see Kolenda et al. 2017a).

Factors influencing water frog populations
The essential factors influencing water frog populations in Europe, especially the disappearance of P. lessonae, are habitat loss and fragmentation, water pollution, introduction of fish, overharvesting, diseases and competition with non-native Pelophylax frogs (IUCN SSC Amphibian Specialist Group 2022).Most influential seems to be habitat loss, which is a significant factor responsible for global amphibian decline (Collins 2010, see also Introduction).Moreover, over the years, there has also been a significant deterioration of some of the studied habitats.As we have shown in this study, good examples are urban sites Żurawiniec and

Decline of P. lessonae in Poland
Edwardowo that were inhabited by the L-E populations in the past.Fifty years ago, both ponds had shallow and sunlit shores with a low density of reed rush and a variety of submerged vegetation, providing suitable conditions for the maintenance of numerous frog populations.The pond in Żurawiniec completely dried out and after the complete extinction of frogs due to the lack of alternative habitats, has been restored in 2016.So far, there has been no secondary resettlement with migrants from other locations.Now, both sites are heavily shaded with tree canopy, and due to a lack of proper maintenance focused on amphibians, frog breeding habitats have been lost (Figure 3).
Apart from habitat changes, the appearance of fish in small ponds, mainly due to stocking by anglers or in natural way, also threatens amphibians (Kloskowski et al. 2020;Lovas-Kiss et al. 2020;Kloskowski & Nieoczym 2022).Rybacki and Berger (2003) found fish in at least 10% of the ponds in Chłapowski Landscape Park between 1995 and 1997.We also noticed the presence of fishes in almost all of the studied ponds (see Table S1 for details); for example, numerous fishes have been introduced to Żurawiniec site which reduces the possibility of its re-colonization by Pelophylax frogs (Kaczmarski unpublished data).Although all three taxa of water frogs under study are known to breed in fishponds (Van Buskirk 2003;Kloskowski et al. 2020;Skierska et al. 2023), a significantly negative effect of fish on them was confirmed under experimental conditions (Semlitsch 1993;Kaczmarek et al. 2022).For example, in the study by Semlitsch (1993), only 0.33% of tadpoles belonging to P. esculentus and P. lessonae survived in ponds with fish.However, to our knowledge, differences in the effect of fish on particular Pelophylax taxa remain unknown.
Essential factors that currently affect water frogs are diseases.Both Batrachochytrium dendrobatidis (Bd) and ranaviruses were found in Poland, with the highest prevalence in Pelophylax frogs (Palomar et al. 2021).Daum et al. (2012) revealed that P. lessonae secreted a smaller diversity of antimicrobial peptides which are effective against Bd when compared to P. ridibundus and P. esculentus.In SW Poland, we noted a high prevalence of Bd in the L-E system and suggested that P. lessonae may have suffered from the infection (Kolenda et al. 2017b).So far, the occurrence of this pathogen has not been tested in the studied populations, thus its effect requires further studies.
Another interaction that can affect water frog populations is the appearance of P. ridibundus in the L-E systems.This phenomenon emerged in western and southern Europe, where Pelophylax ridibundus sensu lato was introduced multiple times (Holsbeek et al. 2008;Dufresnes et al. 2018Dufresnes et al. , 2020;;Dufresnes & Dubey 2020;Bruni et al. 2020;Jelić et al. 2022).The emergence of P. ridibundus in such populations can cause predation on smaller frogs (Pille et al. 2021).Moreover, shifts in population composition depends also on genetics, i.e. by additional supply of P. ridibundus genome into stable L-E systems.In the L-E systems, hybrids usually produce and transmit the R genome, while the L genome is eliminated from the germline (Berger 1973).Pairs of P. ridibundus and P. esculentus produce P. ridibundus offspring.Consequently, the share of P. ridibundus increases in the population and may completely eliminate the other taxa (Vorburger & Reyer 2003;Bove et al. 2014;Quilodrán et al. 2015).However, in some regions of Switzerland ecological differentiation is a sufficient factor that limits the destabilizing impact of P. ridibundus, and the three taxa have coexisted over the decades (Leuenberger et al. 2014).Nonetheless, in the face of the expansion of P. ridibundus in central Europe and its overcoming of the ecological barrier, its presence in small water bodies can replace the L-E systems into R-E-L, R-E or R-R (Jośko & Pabijan 2020; this study).On the other hand, the described replacement assumes that in the L-E system, the hybrid restores itself in a classic way, i.e. in backcrosses by transmitting the R genome to gametes (Berger 1973).However, there are at least several types of L-E systems, including those with triploids, and thus hybrids can produce simultaneously different genomes (Rybacki & Berger 2001).A recent study from SW Poland showed that more than 30% of P. esculentus females and 80% of P. esculentus males transmit the L genome to gametes (exclusively or with the R genome; Skierska et al. 2023).This shows the potential diversity of combination of gametes in the L-E systems.In sum, the effect of P. ridibundus expansion on the L-E systems is definitely significant.However, deviations from the hybridogenesis model or ecological barriers in some areas may impede the negative impact of P. ridibundus expansion (but see below).
Climate changes can also cause shifts in genetic systems that favour the expansion of P. ridibundus and negatively impact P. lessonae.Recent experimental study showed that P. ridibundus may benefit from climate warming and explore new habitats due to broad thermal tolerance and high thermal preferences (Padilla et al. 2023).Moreover, mild winters shorten the period when small ponds are covered with ice.This may favour 100 K. Kolenda et al. the maintenance of fish populations and colonization by P. ridibundus that overwinters in water and is sensitive to oxygen deficiency (Berger 1982;Plénet et al. 2000).On the other hand, P. lessonae hibernates on land (Berger 1982), where they burrow shallowly (3-7 cm below the surface in soil) and are active on days with mean temperatures below 1°C (Holenweg & Reyer 2000).Hiding in the shallow and dry soil layer due to the ongoing drought in the region (Ziernicka-Wojtaszek 2021) exposed P. lessonae to a greater risk of water loss.Indeed, higher weight loss during warmer winters than in colder ones is predicted (Holenweg & Reyer 2000).
Increase in air temperature and decrease in snow cover depth that has been ongoing since the 60s of the 20th century in Poland (Tomczyk et al. 2021) may explain why P. ridibundus has broken the ecological barrier and inhabits small ponds, while P. lessonae is disappearing.

Trends in water frogs populations in Poland
Some detailed patterns of trends in water frog populations are known from two regions in southern Poland.A comparative study from the Nida Basin showed a significant decline of P. lessonae and P. esculentus from 62% and 25.5% of studied ponds (N = 71), respectively, and colonization of only 4.2% and 14% ponds, respectively, over 25year period (1979-1984and 2006-2008;Bonk & Pabijan 2010).At the same time, the number of ponds inhabited by P. ridibundus increased from 2 (2.8%) to 17 (24%) (Bonk & Pabijan 2010).Taxonomic composition of water frog populations also changed in a fishpond complex located along the Upper Vistula river valley.Before 1989, this area was inhabited by the L-E system with a marginal share of P. ridibundus (Juszczyk et al. 1989).Currently (years 2017-2019), P. ridibundus (47.4%) dominates over P. esculentus (46%) and P. lessonae (6.6%) (Jośko & Pabijan 2020).Similar long-term changes were found in the city of Olsztyn (NW Poland).Between 1997 and 2015, the number of small ponds inhabited by P. lessonae and/or P. esculentus significantly decreased; however, P. ridibundus was not observed in the city during the study period (Knozowski et al. 2022).

Conservation management and further recommendations
Our research showed that the L-E populations in the studied areas in the past were replaced by the R-E populations.Litvinchuk et al. 2020).The genetic approach is needed to detect the potential hybridization with non-native taxa (Weigand et al. 2022).
Because all three frog taxa can occur in the same location creating stable or temporary populations (L-E-R; Rybacki & Berger 2001;Leuenberger et al. 2014;Hermaniuk et al. 2020), active conservation actions should focus on a) improvement of habitat conditions for P. lessonae and b) limitation of the P. ridibundus expansions rather than their complete eradication.First is modification of the existing breeding habitats by creation of shallows with abundant vegetation that are preferred by P. lessonae and P. esculentus (Holenweg Peter et al. 2002;Quilodrán et al. 2015;Jośko & Pabijan 2020).

Decline of P. lessonae in Poland
Simultaneously, it is recommended to construct new ponds that meet habitat requirements of the threatened species (Magnus & Rannap 2019;Moor et al. 2022).Proper management or creation of overwintering shelters is also worth considering.Jośko and Pabijan (2020) suggest the drainage of most ponds for winter in extensively managed fishpond complexes to limit the overwintering sites of P. ridibundus (see also Berger 1982).On the other hand, constructions of terrestrial hibernacula for amphibians could increase the survival of P. lessonae and P. esculentus.Indeed, it turned out that tree stumps located near several water ponds were used as wintering site by at least nine amphibian species, including P. lessonae and P. esculentus (Jędrzejewska 2016).Increasing the possibility of P. lessonae dispersion requires maintenance of ecological corridors such as untransformed stream valleys (e.g. in urban areas; Konowalik et al. 2020).However, due to P. ridibundus broad ecological tolerance for a variety of pond features (Denoël et al. 2022) and various factors affecting P. lessonae, the suggested habitat management actions should be carried out in parallel with long-term monitoring.

Figure 1 .
Figure 1.Study sites and changes in frequency of water frogs in an urban (a,c) and a rural (b,d) habitat.R -P. ridibundus, E -P. eculentus, L -P. lessonae.

Figure 2 .
Figure 2. Changes in frequency of water frogs at particular sites in an urban (a) and a rural (b) habitat.R -P. ridibundus, E -P. eculentus, L -P. lessonae.

Table I .
Number and frequency of frogs collected in previous and current study.
LL -P.lessonae, RL -P.esculentus, RR -P.ridibundus.Decline of P. lessonae in Poland 97 Considering the confirmed decline of P. lessonae in different parts of the country (e.g.Bonk & Pabijan 2010; Jośko & Pabijan 2020), we suggest including this taxon on the list of strictly protected species.This would reflect the higher protection status of P. lessonae in the European Union when compared to P. esculentus and P. ridibundus.Moreover, further long-term monitoring of population compositions of water frogs should be continued to determine whether the observed changes are reversible or progressive and to assess the exact causes of this ongoing process.The monitoring should include different genetic systems and take place in a) localities with confirmed recent shifts in taxonomic composition, b) populations considered stable (e.g.R-E in Barycz river Valley, Socha & Ogielska 2010; L-E and L-E-R in Biebrza river Valley; Hermaniuk et al. 2020; E-E near Wolin island;