Discussion
We compiled a large dataset that describes mortality due to wildlife-vehicle collision in Ecuador based on both systematic surveys and non-systematic records that came from citizen science and opportunistic observations reported in the scientific literature. Collectively, these data reveal that 13.66% of described vertebrate species from Ecuador are susceptible to die on roads, a number that likely underestimates the true impact of roads as not all areas are well sampled and smaller and cryptic species may be underreported. We see a need for additional systematic surveys, which can provide more comparable estimates, and have been few and limited to some areas. Likewise, citizen science reports are overrepresenting certain areas, with gaps in other regions.
We found that marsupials (D. marsupialis , and D. pernigra ) represented more than a half of mammalian records (systematic and non-systematic) reported in 21 out of the 24 provinces. Their generalist habits (diet and habitat), together with increased abundance and intermediate body masses could be important factors for their high mortality (Medrano-Vizcaíno et al. , 2022). Likewise, species with a scavenger or omnivorous diet such as the American black vultureCoragyps atratus, the smooth-billed ani Crotophaga ani and the groove-billed ani Crotophaga sulcirostris comprised a great part of the avian records. The common green iguana Iguana iguanawas the most common roadkilled reptile. While this is mainly a herbivore, iguanas have been recorded feeding on carcasses of deer and opossums (Anderson & Enge, 2012). Roadkilled animals and invertebrates that get attracted to these carcasses can be a food source for scavengers, which attracted to road then have a higher probability of being hit by a car.
Among the systematic studies we estimated particularly high standardized mortality rates in some cases. Studies in Galápagos (García-Carrascoet al. , 2020), Napo (Filius et al. , 2020), and Manabí (Gaón & Valdez, 2021) where roadkill rates were disproportionally high in comparison with other studies were all conducted in small areas (13.8, 15.8, and 10 km, respectively). These areas could be hotspots of mortality that do not represent rates across wider areas (overestimate risk). On the other hand, roads in these studies were surveyed by bicycle. Slower survey methods like bicycle or walking can result in higher detectability, 8.4 times higher than using a car (Wang et al. , 2022). Therefore, these studies may be better capturing the true mortality, at least for those smaller areas.
How road impacts wildlife in areas of biological importance should be further studied. For example high road mortality for endemics of the Galápagos Islands could threaten population persistence (Wiedenfeld, 2006; Tejera et al. , 2018). In this territory, threatened species comprise 20 out of 43 birds species, 18 out of 42 reptiles species, and 6 out of 9 non-cetacean mammals species (IUCN, 2022), for which the impacts of roads on their populations remain unexplored. But also non-threatened species may be impacted. For example, the yellow warblerSetophaga petechia , that was commonly reported as roadkill, has experienced a dramatic population decline, and although low insect abundance due to intense use of herbicides has been associated with this decline (Dvorak et al. , 2012), road mortality could contribute to further declines (or already be a factor). While there have been two systematic studies in Galápagos these offer limited insight as one focused on a single species focus (Tanner & Perry, 2007) and the other represented a single one-month survey (García-Carrasco et al. , 2020). The real impact of roads on wildlife populations in Galápagos remains unknown and considering the high endemism and quantity of threatened species, we think that this area deserves special attention for research.
We found roadkill records for 15 Threatened and six Data Deficient species, with some of them showing repeatedly in different regions of Ecuador. The Peters’ running snake Coniophanes dromiciformis , is a Vulnerable and little known species recorded in only nine locations of Ecuador (Cisneros-Heredia, 2021). This range-restricted species showed a roadkill rate of 1.19 ind./km/year in Manabí, with an additional four non-systematic records. Another worrying case is the Violet-throated Metaltail Metallura baroni . This bird, catalogued as Endangered and endemic to Azuay and Cañar provinces, is known from five locations (BirdLife International, 2016), and we estimated roadkill mortality of 0.13 ind./km/year in Azuay, with an additional non-systematic record. In addition, we gathered seven non-systematic records of the pacific horned frog Ceratophrys stolzmanni in Manabí, a Vulnerable and rare species whose entire population is distributed in less than eight subpopulations (IUCN SSC Amphibian Specialist Group, 2018). Directing research efforts on roads within the distribution areas of these range-restricted species could be important to determine the impact of roads on their populations.
Additionally, poorly studied organisms such as Caecilians andAtractus sp. snakes were also regularly reported as roadkill. These animals are among the least known vertebrates, with several taxonomic uncertainties and unknown conservation status for many of them (Cisneros-Heredia, 2005; Wilkinson, 2012; IUCN, 2022). Both groups were frequently detected across three studies in Napo (Filius et al., 2020; Medrano-Vizcaíno & Espinosa, 2021; Medrano-Vizcaíno et al. in prep), and non-systematic data included 41 records of Atractus sp. snakes. In a survey conducted in 2014, Medrano-Vizcaíno & Espinosa (2021), found two roadkilled individuals attributed to the genusAtractus in Napo that were latter described as new species (described by Melo-Sampaio et al., 2021; Arteaga et al., 2022). Road ecology research in areas with poorly-known and undescribed species would be valuable, both to quantify impacts but also to potential expand our understanding of these species.
Our study offers an overview of wildlife mortality on Ecuadorian roads, but likely underestimates the impact. We obtained roadkill records from all the 24 provinces in Ecuador, but systematic studies were only available for five, and non-systematic data was disproportionally distributed in provinces like Manabí, and Napo (more than 300 records for each province). While in Bolívar, Cañar, Carchi, Cotopaxi, Galápagos, Los Ríos, Pastaza, Santa Elena, and Zamora Chinchipe, we gathered less than ten records per province. Our results also show that citizen science data can complement systematic studies and identify top roadkill species (like D. marsupialis, D. pernigra, S. petechia , and R. marina ) that were both commonly reported in non-systematic and systematic data from Galápagos, Manabí, Guayas and Napo (González, 2018; Filius et al. , 2020; García-Carrasco et al. , 2020; Zavala, 2020; Gaón & Valdez, 2021; Medrano-Vizcaíno & Espinosa, 2021). We hope by expanding our network of citizen scientists in REMFA we will be able to fill geographical gaps of information, and gain insights into wildlife mortality in areas where systematic studies are lacking.
A limitation in out compiled dataset is the assumption of correct taxonomic identification in systematic studies (and our own correct identification in non-systematic records). For example, some species were reported outside their known distribution ranges in Ecuador:Neacomys amoneus, Mesoclemmys heliostemma, Myrmochanes hemileucus, Pseudocolopteryx acutipennis, Rhogeessa io, Scinax ruber, andXenoxybelis argenteus, and even one species not distributed in Ecuador was reported (Rhinella arenarum ). These may be misidentifications (likely for R. arenarum ), but previous roadkilled specimens have revealed new distribution areas for certain species (Medrano-Vizcaíno & Brito-Zapata, 2021). Roadkill records can provide valuable information about the biology and ecology of species and with correct taxonomic identification contribute to our understanding of biodiversity.
Road ecology research in Ecuador is gaining interest, but is still relatively limited. Promoting and guiding additional research and public engagement is important. Through our citizen science project REMFA we have given a special relevance to science communication, which has been vital to reach citizens to join our work, and we are now engaging with policy makers. The active involvement of government ministries such as the Ministry of Environment, Water and Ecological Transition, together with the Ministry of Public Works is necessary for the inclusion of adequate policies to reduce wildlife mortality across existing roads, and to plan sustainable roads for the future. We hope that this work can be an initial step towards these national aims for wildlife conservation.