Studies from different continents have proved amphibians to be the most frequently killed vertebrates on roads. In Central-Europe their ratio is between 70 and 88 percent. Local populations are known to become extinct or genetically isolated, and avoidance is also recognised, especially where the road network is dense and the traffi c is intensive. Besides ecological and conservation considerations, amphibian road kills also present a hazard for motorists when amphibians migrate in large numbers. Mitigation measures for amphibians have been applied since the 1960s. In Central-Europe the fi rst amphibianrelated culvert modifi cation occurred at Parassapuszta, Hungary, in 1986. A number of amphibian-oriented mitigation measures have been made in the region under roads and motorways since then, especially after 1995. The aim of this paper is to describe the main features of these constructions, overview the different designs, and make suggestions for their improvement as well as for future amphibian-oriented mitigation measures in general. A total of 31 road sections was monitored. Besides amphibian tunnels, game passages and game bridges were also investigated. Both the detailed characterisation of the technical solutions and the survey of amphibian populations and habitats were included in the methodology. The investigation of the tunnel systems showed a great diversity, e.g., in tunnel and fence material, their position in relation to the road, and connections between them. For economical reasons concrete tunnels were the most common. Both circular and square cross-section tunnels were in place. The accessibility of the entrance was a possible problem, especially in areas where erosion is considerable. Plastic mesh and concrete fences were both applied with a height fl uctuating between 45 and 70cm. Plastic fences are usually fi xed to wooden poles, which need to be checked before the migration period starts. However, the advantage of such fencing is fl exibility, which makes possible, e.g., the turning back of its ends to prevent amphibians from getting on the road. Some systems did not work because certain elements (usually fences) were in bad condition. Elsewhere lack of maintenance reduced the effi ciency of mitigation measures. Missing elements should be replaced immediately, even if the amphibian migration period is over, because other animals, e.g., small mammals, also use these systems. The lowest distance between tunnels in amphibian mitigation systems is 40m in the region (Kudowa Zdrój, Poland). Usually, amphibian tunnels were placed 50-100m from each other, which is an acceptable distance. In case of adequate fencing, game passages and game bridges would also be appropriate for the crossing of amphibians as well as reptiles and small mammals, similar to slightly modifi ed existing culverts under high road mortality sections, and there would be a need for such conservation improvements at several sites. As a result of this work, several recommendations on the maintenance of amphibian tunnels and fences were also developed. Further cooperation among different agencies and organisations was urged, nationally as well as internationally. The improvement of public relation activities on fauna passages also seems to be needed for the effective protection of wildlife on roads.

Transportation infrastructure is a major factor determining land use forms. As global changes in this factor are the most important for biodiversity, roads fundamentally influence wildlife. The effect of roads on wildlife has been categorized in several ways resulting in six to ten categories with road kill as an obvious and important component, and amphibians are greatly affected by this factor. As this animal group has been documented to decline from multiple threats worldwide, the study and mitigation of their deaths on roads has become an important conservation priority. It was also detected as a single cause of decline, and data have accumulated on related population fluctuations, isolation, decline, and extinction in several countries. Genetics studies greatly improve our insight into these processes, e.g., by repeatedly proving significantly low heterozigocy in populations of several species living near roads. Amphibian road kills have been long documented and described due to their spectacular nature, but the overall effect of transportation infrastructure on amphibians was often underestimated due to contrasting research results. The speed of transport and the duration and timing of the surveys in which information was collected turned out to be decisive factors, causing differences of 5.5-16 times the number of road-killed amphibians recorded, mainly in connection with the low visibility and retention time of amphibians on roads. In light of such amphibian-related differences, the often cited national road kill estimates may well be considerably higher in practice, as well. Amphibian road mortality studies have been conducted almost exclusively in developed countries, mostly in Europe and North America, and under temperate zone conditions. In general, all terrestrial and semi-aquatic amphibian species can suffer from road kills where they have populations near roads. However, different amphibian species are threatened to a different extent by traffic because of their specific life history characteristics. Besides amphibian- specific factors (amphibian movement types, length and direction of movement, velocity, temporal movement pattern, behavioural changes on roads), the spatio-temporal pattern of amphibian road kill is also influenced by habitat and transportation characteristics (especially aquatic habitats and vegation, road density, traffic intensity, vehicle speed, position and structure of roads, and awareness of drivers, respectively) and weather conditions (precipitation, temperature, wind). The effect of these factors must be understood before the need for mitigation can be evaluated and measures designed and built. Many mitigation measures have been built since the first amphibian tunnels were created in 1969 near Zürich, Switzerland, and a high diversity of technical solutions successfully reduced amphibian road kills under different conditions. New research results have shown that amphibian tunnels can also be permeable for reptiles, such as snakes and small mammals. However, the lack of maintenance and construction deficiencies are common problems, which lower the efficiency of these measures worldwide. Road kills also have socio-ecological importance. Successful road-kill related projects have the potential to improve the understanding of decision-makers regarding road-related problems, also leading to their support of more complex conservation projects, including, for example, habitat restoration or compensatory developments near roads. Using the media to educate the general public about conservation efforts to reduce road kill, such as setting up frog fences in the USA and toad saving campaigns in Europe, clearly helps to realise this aim by influencing support provided by various authorities.

The effects of roads and railways on animals such as direct mortality caused by these infrastructure elements were recognised as early as the end of the ninetenth century. In the first half of the twentieth century further evidence gathered related to different vertebrate groups. Besides the increasing amount of information available on the environmental impact of roads and railways in the second half of the twentieth century, crossing structures, game bridges, amphibian tunnels and game passages were built as mitigation measures to provide corridors over or under roads and railways, especially in Europe. In most cases, however, they were aimed to help one animal group or species. With the development of an ecosystem-level approach, however, the investigation of the possible involvement of these constructions in helping non-target groups also started together with building green bridges. A further recognition of the special needs of certain species also led to the development of new structural elements, for example tunnels built within green bridges to help burrowing animals to cross.

Mitigation measures representing all animal crossing structures in Hungary were selected to study their use by amphibians, reptiles and small to medium-sized mammals. They included a toad tunnel system with eight tunnels and approximately five-hundred meter concrete fences along road 8518. and six tunnels under the bicycle road running along the same road stretch at Lake Fert?, one wet and two dry passages of one meter in diameter under the M1 motorway with 60 centimetre high concrete fences and two twelve meter wide game bridges with game fences over the same motorway. All sites are located in the same, Arrabonicum fauna district in the western part of Hungary. Due to differences in the studied animal groups a complex sampling methodology was applied. Besides site visits during the day to find the shed skins of reptiles, footprints of mammals on sand beds or their droppings in the passages, the mitigation measure use of amphibians was also investigated in night visits especially during the breeding season while mammals were also caught by baited traps and hair traps were also used. To check the efficiency of the toad tunnel system the frequency of amphibian road kills were also studied.

Amphibians were found both in the tunnel system and the wet passage under the road, but their presence was not proved either in the dry passages or on the game bridges. The tunnel system worked very efficiently, i.e. it lowered road kills by at least 90%, which can even be improved by maintenance. As a consequence, more amphibians died on the bicycle road and a side road nearby than on the main road. The mitigation measure use of reptiles was proved at all investigated sites even if none of the constructions were planned to provide corridors for that animal group. Grass snakes were found in toad tunnels and passages, sand lizards on game bridges. An important difference between them was that snakes moved through the tunnels while lizards lived on them and used game bridges as a habitat. Small mammals used all investigated measures, vole and mice species were trapped in all of them. What is more, they used tunnels as part of their habitats. Besides, shrews were present in toad tunnels as well the presence of foxes and martens was also indicated. However, their road kill was low in the section studied.

During the study period eight species of amphibians as well as mammals and two reptiles were proved to utilise the investigated crossing structures. Besides providing corridors, large constructions, such as game bridges also function as habitats e.g. for lizards. The use of large, mammal-oriented mitigation measures by amphibians and reptiles is needed to study further as well as efforts should be made to construct more passages or alter existing structures in the future to lower habitat fragmentation along transportation infrastructure.