Introduction

Zhejiang Province has experienced rapid economic growth since the 1970s, but factors such as urbanization, industrialization, modern agricultural pollution, and changes in the way land is used have had significant negative impacts on ecological environments and patterns of landscape security; as a result, there has been gradual fragmentation of natural habitats as well as declines in biodiversity1,2. Additionally, some ecologically vulnerable animals in nature can serve as warnings to humans through changes in population size and habitat range3. For survival, fireflies (Coleoptera: Fluoridae) need habitats that are ecologically stable and diversified and meet their demanding standards of ecological quality. Since they can periodically reflect local ecological changes and provide early warning of natural environmental changes, fireflies are usually regarded as excellent environmental indicator species4,5. In addition, fireflies are regarded as one of the most beautiful insects in the world because of how they use their distinctive bioluminescence for courting and communication6,7. Currently, Zhejiang Province is home to approximately 10 different species of fireflies, with Pyrocoelia and Luciola being the most common genera. It is concerning that large firefly populations are now uncommon and occur only in a handful of locations and that certain confined, dispersed, and fragmented habitats could vanish at any moment as a result of commercial development, artificial light pollution, and the threat of pesticides. According to pertinent research investigations, habitat loss has emerged as the greatest threat to fireflies globally8. In Zhejiang Province, both the frequency of occurrence and the size of the firefly population have significantly decreased in recent decades9. The loss of firefly habitat and the population decline have also, to some extent, induced changes in the populations of birds, fish, insects, and other creatures, compounding the decline in biodiversity. Fireflies are a crucial link in the food chain of the ecosystem. The restoration of biodiversity, however, can be greatly aided by the protection of species with specific ecological importance, according to research10. Research on the preservation of firefly populations and habitats in Zhejiang is needed, given the significance of fireflies in relation to biodiversity and environmental stability in Zhejiang Province.

From the perspective of biodiversity conservation practices, understanding the spatial distribution patterns of species is essential for species conservation11. Species distribution models (SDMs) are important for species habitat management because they can use existing species distribution information and relevant environmental data to predict suitable distribution areas and assess and filter key environmental factors12. Several SDMs have been successfully developed and implemented, including MaxEnt (maximum entropy model)13, ENFA (ecological niche factor analysis)14, GLM (generalized linear model)15, GAM (generalized additive model)16, GARP (genetic algorithm for rule-set prediction)17, BIOCLIM18 and CLIMEX19. Among them, the MaxEnt model is based on machine learning and maximum entropy theory, using known species distribution records and their associated environmental variables to infer the ecological demand of species and then predict the potential distribution range of species in the study area. Compared with other models, the MaxEnt prediction results are more accurate and stable and have gradually become one of the most widely used SDMs13,20,21. In recent years, many scholars have used MaxEnt to predict and assess the suitable distribution areas of insect species similar to fireflies, and the final model prediction results are better and can play a guiding role in understanding the potential distribution of insect species and formulating habitat conservation measures22,23,24.

Emeia pseudosauteri belongs to the genus Emeia of the family Lampyridae (Coleoptera), a new species of firefly discovered in 2012 in Emei Mountain, Leshan, Sichuan Province25. According to the survey, a large-scale natural colony of E. pseudosauteri has been occurring annually in Kowloon National Wetland Park, Lishui city, Zhejiang Province, since 2014, and its population size and habitat range have been increasing year by year. The annual firefly glow show staged in the local area is as bright and spectacular as a night sky covered with stars, and this light can induce a strong visual impact and emotional response in viewers, attracting a large number of tourists, gradually forming an ecotourism project in the local area with the frefly viewing experience as the main focus, and bringing tangible economic benefits to the people26. However, the excessive interference of human activities, domestic sewage, pesticides, night lights, road hardening, and the proliferation of tourists are causing damage and pollution to the existing habitat of E. pseudosauteri in Zhejiang Province, making the survival and reproduction of the species in crisis27,28,29. Similar to other species of fireflies, E. pseudosauteri requires stable and high-quality habitats for eggs, larvae and adults and is highly sensitive to environmental changes and disturbances by human activities25, making it an excellent environmental indicator species. In addition, it requires habitat conditions that cover the habitat needs of other similar species and fill a predator niche in a typical environment, and thus, it acts as an umbrella species to maintain the balance and stability of the ecosystem30,31. More importantly, E. pseudosauteri fledge as adults in March–April each year, earlier than other species of fireflies, and the high population densities make it easy to create large areas of stunning firefly glitter25. These advantages have made it a local ecotourism icon in just a few years, and it can serve as a promotional and representative flagship species to raise public awareness of ecological conservation32,33. Therefore, as a highly valuable insect resource, research on the conservation and utilization of E. pseudosauteri must be given full attention. In this paper, we used the MaxEnt ecological niche model to study the distribution status and potential suitable distribution areas of E. pseudosauteri in Zhejiang Province, analyze the key environmental factors affecting the natural distribution of this species, and provide a scientific basis for future habitat conservation, population rebreeding, protected area management and industrial development.

Overview of the study area

Zhejiang Province is located in the Yangtze River Delta region on the southeast coast of China, with geographical coordinates of 118° 01ʹ–123° 10ʹ E; 27° 06ʹ–31° 11ʹ N. The straight-line distance from east to west and from north to south in Zhejiang is approximately 450 km, and the land area is approximately 105,500 km2, with the topography sloping from southwest to northeast. The topography is complex, with mountains accounting for 74.6%, surface water accounting for 5.1%, and flat land accounting for 20.3%. Zhejiang Province has a humid subtropical monsoon climate with significant monsoons, four distinct seasons, abundant light, abundant rainfall and humid air. The annual precipitation ranges from 1100 to 2000 mm, and the average annual temperature ranges from 15 to 18 °C34. The forest vegetation in Zhejiang Province is a central subtropical evergreen broad-leaved forest, and the main forest vegetation types are warm coniferous forest, evergreen broad-leaved forest, deciduous broad-leaved forest, mixed evergreen deciduous broad-leaved forest, mixed coniferous broad-leaved forest, bamboo forest, economic forest, and shrub forest35.

Materials and methods

Distribution data for E. pseudosauteri

From 2014 to the present, our research team has been continuously investigating and recording the distribution data of E. pseudosauteri in Lishui city, Zhejiang Province. Based on the observation results over the years, its distribution range has been gradually expanding. From 2014 to 2015, the distribution of E. pseudosauteri in Kowloon National Wetland Park, expanded from 0.17 to 4 km2; from 2016 to 2017, it was found in approximately 80% of the park, covering an area of 13.5 km2. Starting in 2018, its habitat gradually expanded upstream and downstream along the Daxi Creek watershed, where Kowloon National Wetland Park is located, with a total of three larger aggregations occurring. As of April 2022, we obtained a total of 282 E. pseudosauteri distribution sites through field research (Fig. 1). Artificial sampling may bias sampling toward easily accessible areas, and because of the significant spatial clustering of small habitat distribution points, model predictions tend to be overfitted around known occurrences, and model performance values tend to be inflated36. To remove spatial autocorrelation and sampling bias, the obtained distribution data were subjected to 1 km spatial dilution using SDMtoolbox37. The final 36-point information was obtained and imported into an Excel sheet and collated and saved as a CSV format file for MaxEnt modeling analysis.

Figure 1
figure 1

Map of distribution data of E. pseudosauteri in Zhejiang Province as of April 2022 (Bigemap v25.5.0.1: http://www.bigemap.com/index.html).

Full size image

Data on environmental variables

Based on E. pseudosauteri habitat research and related studies, four types of environmental variables were selected: biological environmental (normalized difference vegetation index, distance to rivers), bioclimatic (19 bioclimatic variables including annual mean temperature and isothermality), topographic (elevation, slope, slope direction), and anthropogenic disturbance variables (lighting, land use, distance to roads, distance to settlements). Topographic variables, the normalized difference vegetation index (NDVI), and land use data were obtained from the website of the Resource and Environmental Science Data Center of the Chinese Academy of Sciences (https://www.resdc.cn/Default.aspx). Bioclimatic variables were obtained from the World Meteorological Data website (http://www.worldclim.org/) by downloading a total of 1938. Light data were obtained from the Earth Observation Group (EOG) website (https://eogdata.mines.edu/products/vnl/) for the 2021 annual VNL V2 data39. River, road, and settlement data were obtained from the National Basic Geographic Information Database 2021 1:1 million geographic vector data set, and Euclidean distance calculations were performed through ArcGIS 10.6 to obtain the distance to river, distance to road, and distance to settlement raster data, respectively. Then, all 28 environmental variable raster data were resampled to a 1 km resolution by ArcGIS 10.6 to ensure the same range boundary, unified geographic coordinates as WGS_1984, and projection coordinate system as WGS_1984_UTM_Zone_50N. Finally, the processed environmental variable raster data were converted to ASCII format.

Spatial autocorrelation between environmental variables may affect the analysis of the relationship between species distribution and the environment40. To reduce the spatial autocorrelation between environmental variables and avoid the influence of highly correlated environmental variables on model prediction results. The first 19 bioclimatic variables were subjected to principal component analysis, and the top three with the highest contribution rates were selected as principal components based on the accumulation of eigenvalues. The standardized eigenvectors of each variable were obtained (Table 1), and then Pearson correlation analysis was conducted on 28 environmental variables (Supplementary Table S1). If the correlation coefficient |r|≥ 0.8, the two variables were considered to be highly correlated, and only the environmental variables had larger eigenvector values41. Eleven environmental variables, including elevation, slope direction, slope, seasonal variation in temperature, annual precipitation, lighting, land use, NDVI, distance to roads, distance to rivers, and distance to residential sites, were finally screened for E. pseudosauteri species distribution prediction modeling (Table 2).

Table 1 Principal component analysis results of 19 bioclimatic variables.
Full size table
Table 2 Environmental variables used in predictive modeling of Emeia pseudosauteri species distribution.
Full size table

Maxent model selection and construction

After the E. pseudosauteri distribution data and environmental variable data were processed, modeling operations were performed using MaxEnt (version 3.4.4). Twenty-five percent of the distribution points were randomly selected as the test data set, and the remaining 75% of the distribution points were used as the training data set. The maximum number of iterations of the model was set to 500, the number of repetitions was set to 10, the category of repetition was selected as cross-validation, and the relative contribution of 11 environmental variables to the prediction of the suitable distribution area of E. pseudosauteri was calculated using the knife cut analysis of the software. The threshold value of each environmental factor was responded to by the response curve, other parameters were set by default, and the prediction results were output. ASCII format was chosen for the format42.

Results and analysis

Evaluation of model prediction results

AUC (area under the receiving operator curve) is a statistic widely used to evaluate the prediction performance of species distribution models43,44, but related studies have proven that AUC is insufficient as a predictor in spatial distribution modeling45. Therefore, we introduce TSS (true skill statistics) as a supplement, which inherits all the advantages of Kappa statistics and has a significant correlation with AUC statistics46. We refer to the study by Liu et al.47, select maxSSS (maximum sum of sensitivity and specificity) as the threshold, and calculate the TSS average of the 10 repeated operation results of the MaxEnt model to evaluate the model performance. The AUC value is generally between 0.5 and 1, and the higher the value is, the better the prediction effect of the model48. When the AUC is less than 0.7, it indicates poor model performance, 0.7–0.9 indicates moderate model performance, and 0.9–1.0 indicates good model performance49. In the TSS value range, 0.2–0.5 is poor, 0.6–0.8 is useful, and greater than 0.8 is excellent50.

The mean AUC of 10 repeated runs of the final model was 0.985, the standard deviation was 0.011 (Fig. 2), and the mean TSS was 0.81 (Table 3). It shows that the model has good performance and high prediction accuracy. The species distribution data and environmental variable data used in the modeling can effectively predict the potential suitable distribution area of E. pseudosauteri in Zhejiang Province.

Figure 2
figure 2

ROC curves and AUC averages for MaxEnt with 10 repeated runs.

Full size image
Table 3 Average TSS of the results of 10 iterations of the MaxEnt model.
Full size table

Contribution of environmental variables

We analyzed the importance and contribution of each environmental variable in the model prediction by the jackknife method in the MaxEnt prediction model. The results showed that the contributions of bio_4 (temperature seasonal variation) and Alt (altitude) were 33.5% and 30%, respectively, and the regularized training gain values were both greater than 0.75, which can provide a greater benefit to the prediction model. Thus, bio_4 and Alt had more useful information than the other environmental variables and had the greatest effect on the distribution of E. pseudosauteri in Zhejiang Province. Additionally, riv_dis (distance to rivers), bio_12 (annual precipitation) and LUCC (land use) were the three environmental variables with the next highest contribution and regularized training gain values and had some influence on its distribution in Zhejiang Province. The remaining environmental variables, including the NDVI, slope direction, slope, light, distance from roads, and distance from settlements, had a small or even negligible effect on its distribution (Table 4, Fig. 3).

Table 4 Percent contribution of environmental variables in the MaxEnt model.
Full size table
Figure 3
figure 3

The importance of jackknife analysis of environmental variables.

Full size image

Response curve analysis of environmental variables

To explain the effect of each environmental variable on the distribution of suitable habitat, the MaxEnt model gives response curves between the probability of species presence and environmental variables51. The response curves for the main environmental variables affecting the distribution of E. pseudosauteri are shown below (Fig. 4). The seasonal change in the air temperature response curve showed an overall trend of increasing and then decreasing, and its presence probability reached a maximum of 93% at 7.7 °C. When the amount of seasonal change in air temperature was less than 7.7 °C and more than 8 °C, the presence probability decreased rapidly, while the curve changed slowly between 7.7 and 8 °C. Therefore, it can be indicated that it is more sensitive to seasonal temperature changes. In terms of altitude, the presence probability of E. pseudosauteri increased rapidly with increasing altitude, reaching an extreme value at 100 m, after which the presence probability gradually decreased with increasing altitude. This result suggests that a specific altitude can provide a unique and suitable habitat for it. The presence probability of E. pseudosauteri was negatively correlated with the distance from the river, with the presence probability decreasing as the distance from the river increased. The probability of E. pseudosauteri survival was most suitable when the annual precipitation was approximately 1500 mm, and the distribution probability reached an extreme value of 95%. Between 1000 and 1500 mm, the probability of presence gradually increased and then gradually decreased. The response curves of land use variables indicated that E. pseudosauteri habitat selection preferred mudflats, river canals, and shrubby woodlands. Collectively, this species is highly dependent on wetland environments near rivers with high annual precipitation, and the dense vegetation of shrubby woodlands can provide it with sheltered, safe, and stable habitat environments.

Figure 4
figure 4

Response curve of the main environmental variables affecting the distribution of E. pseudosauteri. (A) Response curves of bio_4 (temperature seasonal variation). (B) Response curves of Alt (altitude). (C) Response curves of riv_dis (distance to rivers). (D) Response curves of bio_12 (annual precipitation). (E) Response curves of LUCC (land use).

Full size image

Prediction of suitable habitat for E. pseudosauteri in Zhejiang Province

According to the fitness index P output from MaxEnt software, the predicted fitness zones were classified into four levels: 0–0.08, nonsuitable zone; 0.08–0.32, low suitability zone; 0.32–0.64, medium suitability zone; and 0.64–1, high suitability zone. The suitable area for E. pseudosauteri was mainly distributed in southern Zhejiang Province, and the highly suitable area was approximately 775.91 km2, accounting for approximately 0.74% of the land area of Zhejiang Province; the moderately suitable area was approximately 1458.09 km2, accounting for approximately 1.38%; and the less suitable area was approximately 3996.96 km2, accounting for approximately 3.79%. The distribution of medium- fitness and high-fitness zones was more scattered, and a continuous larger area of high-fitness zones was found only in Lishui city (Fig. 5).

Figure 5
figure 5

Map of the suitable distribution area of Emeia pseudosauteri in Zhejiang Province (MaxEnt v3.4.4: https://biodiversityinformatics.amnh.org/open_source/maxent/, Bigemap v25.5.0.1: http://www.bigemap.com/index.html).

Full size image

Highly suitable areas were: Daxi watershed in Lishui city, Longquan Creek and Jingshuitan watershed in Yunhe County, Songyin Creek watershed in Songyang County, parts of Longquan City, Xiaoxi and Qianxia Lake watersheds in Qingtian County and parts of the Oujiang River watershed from Qingtian County to Wenzhou City; Dannan Creek watershed in Yongjia County, Wenzhou; around Xianxia Lake watershed in Hunan Town, Quzhou city; around Qiandao Lake watershed in Chun'an County, Hangzhou city; around Changtan Reservoir watershed in Huangyan District, Taizhou city.

The distribution of the middle fitness zone was basically the same as that of the high fitness zone, mainly extending outward to a certain extent with the high fitness zone at the center, including the area around Xianxia Lake waters in Lishui city, Dongdu Township in Jinyun County, Lanju Township in Longquan city, Xiakou Township in Quzhou city, Baita Township in Xianju County in Taizhou city, around Niutoushan Reservoir in Linhai city, and some areas in Wenling city, in addition to those in the high fitness zone mentioned above.

Discussion

The suitable distribution areas predicted by the MaxEnt model essentially reflect a selection made by E. pseudosauteri for the natural environment in the region. Among the 11 environmental variables, the seasonal variation in temperature, elevation and distance from rivers were the main factors affecting the habitat suitability and distribution probability of the species. Temperature is an important condition for firefly habitat reproduction, and temperature changes can directly affect firefly populations52. Combined with the response curves of environmental variables output from the prediction model, when the amount of seasonal variation in temperature was between 7.7 and 8 °C, it can provide an effective accumulation temperature for E. pseudosauteri and ensure the hatch rate of eggs. Elevation is considered to be one of the determinants of species diversity distribution patterns and can influence the climatic characteristics of the habitat, the distribution of water systems, and vegetation community types to some extent53,54,55. From the output of the prediction model, E. pseudosauteri tended to select the area between 50 and 300 m elevation as their habitat, and the area between that elevation was mostly hilly. In southern Zhejiang, there are continuous, gently undulating low hills, with a mild climate, abundant precipitation, and diverse and distinct vegetation types, which successfully shape the unique and suitable habitat environment. As terrestrial fireflies, E. pseudosauteri exhibits extremely high hydrophilicity, and habitat suitability decreases with increasing distance from rivers, decreasing the probability of E. pseudosauteri being present. Based on previous studies, this may be due to egg hatching as well as larval food sources25. Compared to the results of other studies on firefly habitats, the effects of environmental variables such as the NDVI, lights, distance from roads, and distance from residential sites on E. pseudosauteri were not significant in this study, which may be due to the different scales of research and the existence of scale effects. Moreover, because E. pseudosauteri has been a new species of firefly in Zhejiang Province in recent years, it is currently distributed only in part of the Daxi watershed in Lishui city, Zhejiang Province, with a small distribution range and population size. Although it has been demonstrated that the MaxEnt model can also achieve good prediction results with less species distribution data, it may have an impact on the prediction assessment results because the input environmental variables are still relatively coarse at the spatial scale56,57. Studying the effects of ecology on species has always been an important basis for capturing the habitat requirements and spatial distribution of species. Based on the results of this MaxEnt model prediction, this research team will conduct long-term monitoring of E. pseudosauteri in their medium- to high-suitability distribution areas and, in the future, validate the screening of more comprehensive environmental variables to explain the distribution of E. pseudosauteri in their suitability areas through habitat requirements, species behavior, and habitat range studies.

In recent years, increasing numbers of scholars in China have been actively conducting research on firefly habitat restoration, ecological conservation and artificial rebreeding9. The relevant research results provide the basis for the exploitation of firefly resources as well as industrial development. The development and utilization of E. pseudosauteri value resources need to be fully integrated with other industries, e.g., the development of firefly organic agriculture, so that E. pseudosauteri, as an excellent environmental indicator species, can be used to monitor the excellent quality of local organic agriculture and enhance the market competitiveness and added value of local agricultural products9. The viewing of fireflies can also improve the health of the heart to a certain extent, thus allowing the development of a corresponding recreation industry58,59. In addition, the development of firefly science education, firefly cultural exhibitions, and firefly creative products can make E. pseudosauteri a local cultural symbol, bring tangible economic benefits to local people, and stimulate public awareness of conservation. Moreover, a portion of the revenue generated by the development of firefly-related industries can be used for conservation research on habitat ecology and population breeding60. Only in this way can the value of E. pseudosauteri be used rationally and fully to achieve safe and sustainable development.

Conclusion

Potentially suitable distribution areas for E. pseudosauteri are basically along river waters, mainly in the Ou River, Nanxi River and Wuxi River and their tributaries in southern Zhejiang. Although the medium–high suitable distribution area covered a large and extensive area, the overall suitable habitat area was small and fragmented, which led to a poor ecological carrying capacity and resistance of the habitat and was vulnerable to anthropogenic disturbance. Therefore, it is particularly important to develop corresponding conservation mechanisms to strengthen the regulation and protection of E. pseudosauteri habitats. By establishing protected areas, ecological corridors and conservation nodes, we can improve the landscape ecological security pattern of suitable habitats for this species, solving the shortcomings of the current narrow and fragmented habitats and further improving habitat suitability. The E. pseudosauteri habitats are diversified and contain wetlands, farmlands, bamboo forests and shrublands, which overlap with many species' habitats and have similar ecological niches, so the conservation of E. pseudosauteri will, to a certain extent, contribute to the extensive conservation of biodiversity in Zhejiang Province.

Although this paper focuses on the potential suitable distribution areas and environmental factor characteristics of suitable habitats for E. pseudosauteri in Zhejiang Province, there are certain similarities in the habitat environments of different species of fireflies, and the predicted evaluation of the habitat suitability of E. pseudosauteri can provide a reference for conservation studies of other species of fireflies, which will have wider research significance.