Amphibians are highly sensitive to environmental changes such as climate warming. Here, we report unusually early oviposition in two spatially isolated urban subpopulations of the green toad Bufotes viridis Laurenti, 1768, in Poznań, Western Poland. To our knowledge, we report the earliest breeding date for Central and Eastern Europe, for areas of similar latitude. We ascribe the early onset of B. viridis reproduction to an exceptionally warm spring in Western Poland in 2017. B. viridis shows flexibility in the timing of reproductive activity, however, shifts in breeding phenology may have both beneficial and detrimental population consequences.

Amphibia, Anura, climate change, global warming, phenology, Poznań

Global warming affects the phenology of amphibians (e.g. Beebee 1995; Muths et al. 2017), inducing shifts in reproductive periods that may influence amphibian populations both directly (e.g. mortality rates) and indirectly (e.g. impacts on terrestrial and aquatic habitats, changes in food webs or the spread of diseases) (Blaustein et al. 2010; Li et al. 2013). Amphibians, as a group, are highly sensitive to changes in their environment (Wake and Vredenburg 2008; but see: Kerby et al. 2009). Shifts towards earlier reproduction as an effect of climate changes progress twice as quickly amongst amphibians as in any other taxonomic group (Parmesan 2007). Here, we report unusually early reproductive activity and spawning events in the green toad Bufotes viridis Laurenti, 1768 in Western Poland. Bufotes viridis has a prolonged breeding season beginning in April in Central Europe (Table 1). This species readily inhabits human settlements (Stöck et al. 2008).

In 2017, we monitored the occurrence of B. viridis in the two largest known breeding sites in downtown Poznań (area: 262 km², total human population: about 540,000 according to the Polish Central Statistical Office, 2017): Park Cytadela [52°25'26"N, 16°55'56"E], a stone garden/amphitheatre with a permanent shallow concrete pond (area: ca. 5,150 m²) and Park Rataje [52°23'9"N, 16°57'20"E], a post-industrial area containing debris and concrete waste, where, due to a non-permeable clay substratum, water from precipitation forms irregular ephemeral pools and habitats (Fig. 1).

Figure 1. 

The breeding sites of Bufotes viridis in Park Cytadela (upper), and in and in Park Rataje (lower).

Evening counts were conducted twice a week with additional visits after intensive rainfalls, from mid-March to June in Park Cytadela and from the end of March to July 2017 in Park Rataje. We recorded mating calls, numbers of adults, including amplexed individuals and the presence of eggs and tadpoles. We used meteorological data from Weatherbase (http://www.weatherbase.com/, accessed 13 April 2018) and Weather Underground (https://www.wunderground.com, accessed 25 April 2018).

On 20 March 2017 in Park Cytadela, we observed 10 calling males in the pond; the number of mating individuals increased progressively, with the first two females being observed on 24 March 2017. On 31 March 2017, we counted 105 mating individuals (85 males, 20 females); the next day, we found a number of egg strings, we also observed interspecific amplexus between the common toad Bufo bufo (male) and B. viridis (female). However, electrofishing, performed in April 2017, revealed the presence of abundant fish in the pond, mainly Carassius auratus Linnaeus, 1758 and Tinca tinca Linnaeus, 1758; hence, apparently no tadpole survived to metamorphosis. In the second breeding site, Park Rataje, on 31 March 2017, we observed 71 mating individuals of B. viridis and the first three strings of eggs. The next day, there were several dozen strings of eggs, as well as the first developing embryos. However, the tadpoles did not survive, most probably as a result of pool desiccation in mid-May. On 7 June 2017, we observed 57 breeding individuals, including four amplexed pairs, as well as, two days later, abundant spawn; however, by the end of June, the tadpoles had perished once again due to pool desiccation. Nevertheless, on 23 August 2017, we found a number of toadlets at Gosner stage 45 and 46 (Gosner 1960), indicating that toads had reproduced after our last field visit (4 July 2017); thus, the breeding season of B. viridis in Park Rataje spanned about 3.5 months.

We suggest that the early spawning date of B. viridis was related to the exceptionally warm period in late March. According to the Köppen climate classification system, the city of Poznań is characterised by a maritime temperate climate (Cfb), with mean annual precipitation of approximately 513 mm. The mean monthly air temperature for March over the previous 10 years (2007‒16) had been 4.1°C, with 5.4°C in the last decade of March. In March 2017, the mean air temperature was 6.5°C, with recorded daily mean temperatures in the last decade of March varying between 5.0°C and 16.0°C with a mean of 9.3°C. The days 31 March and 1 April 2017 were exceptionally warm: the mean daily temperature reached 17.3°C and 17.8°C, respectively. The effect of the urban heat island, with a mean annual temperature higher by 1°C in Poznań than in the surrounding non-urban areas (Majkowska et al. 2017), may have additionally accelerated the mating period. In addition, in Park Rataje, intense rainfalls exceeding 11.1 mm from 18 to 19 March 2017 created numerous pools providing reproductive sites.

The above observations of breeding B. viridis specimens are, to our knowledge, the earliest in the year amongst published data from Central and Eastern Europe sites with similar climates and at similar latitudes (see Table 1 for details). Reports on similar early breeding events originate from regions further to the south of the continent or simply characterised by warmer climates, such as Polish Jura, southern Poland (Kowalewski 1974), Rhineland-Palatinate, south-western Germany (Sinsch et al. 1999) or Lower Austria (Cabela et al. 2001). Notably, our observations do not concern single individuals, but apparently whole local subpopulations in two spatially isolated parts of the city (separated by an approximate distance of 4.5 km).

As a species adapted to climatic variability, B. viridis may show flexibility in the timing of its reproductive activity (Kyriakopoulou-Sklavounou 2000; Sicilia et al. 2006). Although the reproductive period of B. viridis in Central Europe usually spans three consecutive months (April‒June) (Juszczyk 1987), in Park Rataje, Poznań, in 2017, this period lasted presumably until July, with metamorphosis recorded in the third decade of August. However, according to Zavadil et al. (2011), in ephemeral water bodies in Czechia, metamorphosing individuals can be observed as late as October.

In conclusion, we would like to highlight that shifts in phenology may be both advantageous and unfavourable. Earlier spawning gives juveniles more time to develop and accumulate energy reserves before winter hibernation (Reading and Clarke 1999; Tryjanowski et al. 2003). On the other hand, earlier breeding may imply the development of tadpoles in a colder environment, which may contribute to increased mortality of larvae. B. viridis tadpoles that develop later, in warmer water, have a higher survival rate, but at metamorphosis are smaller than tadpoles developing in colder water (Dastansara et al. 2017). Typically, in anurans, larger metamorphs have higher survival rates (Reques and Tejedo 1997), while smaller individuals are more threatened by predation (John-Alder and Morin 1990) and desiccation (Child et al. 2008). Earlier spawning may contribute to a mismatch between tadpole food demands and the availability of algae, resulting in reduced tadpole growth (Reading and Clarke 1999). In human-dominated landscapes, preterm spawning may also locally prolong the exposure of adult B. viridis to glyphosate-based herbicides during their spring migration to breeding ponds, while reducing the exposure of eggs, larvae and juveniles following their emergence from the water (Lötters et al. 2014). Furthermore, a tendency to early onset of reproduction in B. viridis may increase the risk of hybridisation with B. bufo due to the overlap of breeding periods (Stöck et al. 2008; Canestrelli et al. 2017).

We are grateful to Wiktor Grenas and Łukasz Dylewski for help in the field. All research activities were approved by the Regional Directorate for Environmental Protection (permission no. WPN-II.6401.190.2016.AC.2).