Seven primer pairs based on microsatellite regions with high polymorphism and Mendelian inheritance were developed to manage the genetic conservation of the endangered Japanese aucha perch, Coreoperca kawamebari. The genetic diversity and presence or absence of genetic disturbance of Japanese aucha perch in Tokushima Prefecture were evaluated. The allelic richness was less than 2.0 among 10 populations, showing the low genetic diversity of C. kawamebari in the prefecture. Genetic structure analysis showed that most of the individuals from the Minami River, a tributary of the Naka River in Tokushima Prefecture, were not from rivers in Tokushima Prefecture, but included individuals from Okayama and Hiroshima Prefectures. We conclude that the Minami River population suffered a serious genetic disturbance with the artificial introduction of individuals from other areas. Based on the results of the genetic diversity analysis, we have made recommendations for the conservation of C. kawamebari in Tokushima Prefecture.

We drew 1-ha mesh maps of the Crown Size Index of globally rare subtropical lucidophyllous forests on Amami Oshima Island (712 km2), Japan, using morphological analysis of aerial photographs taken in the fiscal years 1965, 1984, and 2008. The Crown Size Index corresponded to the vegetation map and the history of human intervention on the island. We roughly categorised forests into three types based on Crown Size Index transitions: mature lucidophyllous forests with consistently large Crown Size Indices since the 1960s; secondary lucidophyllous forests with small Crown Size Indices, which were established after clear-cuts during the 1960s.1980s; and stands of pioneer trees, which grew to form large crowns by the 2000s. Large mature lucidophyllous forest patches are distributed in the designated protected areas of Amami Gunto National Park, and the average Crown Size Index in special protection areas has been consistently maintained at a large value. The species diversity of sub-large trees was higher in areas with larger Crown Size Indices. These results imply that our method may be useful in prioritising forest areas with high conservation values, and that the zoning of Amami Gunto National Park is appropriate for the conservation of subtropical lucidophyllous forests of important biodiversity.

Monitoring and mapping species and their habitats are foundational tasks in ecology and natural resource management. However, the conventional method use to accomplish these tasks, i.e., field surveys, is highly labour intensive. Automated species identification from visual images is one of the most promising avenues to reducing the costs of monitoring and mapping species distributions. Deep learning technologies are becoming increasingly accurate for object detection and image classification. In this paper, we focus on deep learning methods that have shown excellent performance in object detection and image classification in recent years. First, we provide a brief overview of convolutional neural networks, one of the primary algorithms used in deep learning. Second, we review case studies that have applied deep learning to plant species identification and vegetation mapping and discuss future applications. We recommend that future work aim to provide a framework for systematically collecting labelled image data to enable the mapping and monitoring of biodiversity at high temporal resolutions and low cost.

A primer set was developed for detecting environmental DNA (eDNA) of an endangered medium-sized marine mammal, Dugong dugon. We designed a species-specific primer set (target gene: 138 bp of cytochrome b region) for the dugong in silico. We examined the efficiency and specificity of the primer set using DNA derived from hair roots and faeces of the dugong, and water samples collected from an aquarium where a female dugong had been reared. Artificially synthesised DNA and water samples from an aquarium of the most closely related species (African manatee) were also used for reference. By real-time quantitative PCR using SYBR-Green I, DNA was successfully amplified from the dugong samples, whereas manatee DNA was not amplified. These results show the designed primer set to be adequately efficient and specific. These results suggest that eDNA analysis can be an effective tool in surveys of the dugong in combination with conventional methods. The range of distribution of the dugong can be re-examined using eDNA analysis to implement more effective conservation strategies.

Kamagatani Natural Moment and Place of Scenic Beauty in Ikeda, Gifu Prefecture, Japan, contains a nationally significant cherry tree population that is known to contain wild hybrids and varieties. To determine the species composition and probable hybrid parentage within the Kamagatani population, we examined the morphology of 684 individuals. We confirmed the population’s size at 583 individuals, in contrast to a previous inventory that included 157 individuals. Taxa previously known from this area, such as Prunus jamasakura, P. spachiana, P. × mochizukiana, and P. spachiana ‘Pendula,’ were successfully relocated. Newly identified taxa included P. sargentii and P. × yedoensis ‘Somei-yoshino’. Of these two, the latter was the most commonly encountered species (201 individuals) and was likely intentionally planted in the 1950s. Many unidentifiable trees were also encountered; these and individuals of P. sargentii were likely accidentally planted from 1983 to 2004. In total, only 34% of the cherry trees at Kamagatani belong to designated taxa. To increase the abundance of designated taxa, we suggest conservation measures such as improving the vigour of existing trees and establishing a new planting method for new trees.

Many endangered hygrophytes and hydrophytes appeared in habitats created by the tsunami disaster following the 2011 Tohoku Earthquake and subsequently disappeared following the reconstruction work. We surveyed the aquatic flora of a small wetland on Nono-shima Island, Miyagi Prefecture, Japan, in 2015 that had been disturbed by the 2011 Tohoku Earthquake and tsunami. Four brackish water submerged vascular plant species and one charalean alga were recorded. The brackish water submerged macrophyte species diversity of this wetland was higher than that of salt marshes in Japan. As shown by aerial photographs and satellite images, this wetland is behind a beach ridge and was a paddy field or fallow land for at least 50 years before the tsunami. There was insufficient time to consider conservation plans for the rare species that emerged after the disaster because of the rapid reconstruction of infrastructure required in the disaster-stricken areas. To provide information for species conservation rapidly, it may be effective to predict the emergence of species from buried soil seeds using land use history and topographical information.

One of the two major red-crowned crane (Grus japonensis; Japanese, tancho) populations worldwide is located in southeastern Hokkaido, Japan. From 2003 to 2015, aircraft surveys were conducted over northern Hokkaido to monitor the breeding range expansion of this population. Until the late 19th century, red-crowned cranes inhabited northern Hokkaido; however, none were recorded in this region during the 20th century. In 2002, two red-crowned cranes were found, and in 2004, an aircraft survey recorded one pair with a single chick. In 2015, the northern population included three breeding pairs in the Sarobetsu area, two breeding pairs near Lake Kutcharo, and one breeding pair in the Wakkanai Ohnuma area. Expansion of the nesting range continued from 2004 to 2015, with a population increase marked by the fledging of 22 juveniles. Due its high density, the main population in southeastern Hokkaido is at a greater risk of disease. The colonization of northern Hokkaido mitigates this risk, although cranes from the northern population come into contact with the main population in winter. Research on the migration and breeding status of this crane species by aerial survey is essential for maintaining the isolation of the northern population from the southeastern main population.

The water level in Lake Izunuma-Uchinuma in the winter of 2016/17 was 26 cm lower than average. The highest recorded number of whooper swans in the lake during this period came to 6,461 individuals, more than double that of usual years. Analysis of images from monitoring cameras showed that swans were mainly distributed along the shoreline. Where the water depth was shallow enough, the swans dug into the lake bottom in the shallow water, feeding on lotus roots. The following summer, there was no lotus vegetation in areas where the swans had dug. Lotus vegetation disappeared from areas where the water depth was 0.74 m or less, although lotus had occurred in these areas in normal winters, when the water depth was an average of 1.0 m. As a result, 64.4 ha of lotus vegetation were eliminated, leaving open water. The dissolved oxygen of the lake water increased in the new open-water areas created by swans, probably due to enhanced water circulation. Foraging activity of swans thus affected the extent of lotus vegetation, causing a change in the water quality of the lake.

Although the positive contribution of urban green spaces to human health is well known, behavioural preferences for specific types of green spaces within urban landscapes have not been clearly evaluated. Developing empirical relationships between human behaviour (i.e. use) and green spaces may contribute to effective city planning. To address this, we surveyed dog owners by asking them to draw their typical dog walking route on a map. We then calculated the proportion of dog walking routes that passed through built-up, open green, and forest areas relative to a null model of the availability of these land-use types. When the degree of preference for built-up areas was assumed to be 1.0, the preference for open green space was significantly greater at 2.18 (25th percentile = 2.04, 75th percentile = 2.32) and the preference for forest areas was 1.34 (1.14. 1.55). Dog walking is a major activity in urban areas and assessing landscape use by dog owners provides insight into urban green space preferences.

We developed new guidelines for aggregating land-use classifications based on the classification of a 1:25,000 national vegetation map archived by the Ministry of the Environment. We also show the correspondence between the new guidelines and previous guidelines based on a 1:50,000 vegetation map. The guidelines based on the 1:25,000 vegetation map for large (nine classes) and intermediate (16 classes) classifications are the same as the previous guidelines based on the 1:50,000 vegetation map, so the new land-use map can be used as time series data with the previous land-use map. For the fine classification, there are eight new classes and three modified classes related to deer herbivory and alien species.