Result of a year-long animal survey in a state-owned forest farm in Beijing, China

Abstract Background Artificial forest can have great potential in serving as habitat to wildlife, depending on different management methods. As the state-owned forest farms now play a new role in ecological conservation in China, the biological richness of this kind of land-use type is understudied. Once owned by a mining company, a largest state-owned forest farm, Jingxi Forest Farm, has been reformed to be a state-owned forest farm with the purpose of conservation since 2017. Although this 116.4 km2 forest farm holds a near-healthy montaine ecosystem very representative in North China, a large proportion of artificial coniferous forest in the forest farm has been proven to hold less biodiversity than natural vegetation. This situation, however, provides a great opportunity for ecological restoration and biodiversity conservation. Therefore, from November 2019 to December 2020, we conducted a set of biodiversity surveys, whose results will serve as a baseline for further restoration and conservation. New information Here, we report the result of a multi-taxa fauna diversity survey conducted in Jingxi Forest Farm mainly in year 2020 with explicit spatial information. It is the first survey of its kind conducted in this area, revealing a total of 19 species of mammals, 86 birds, four reptiles, two amphibians and one fish species, as well as 101 species of insects. Four species of mammals are identified as data-poor species as they have less than 100 occurrence records with coordination in the GBIF database. One species of insect, representing one new provincial record genus of Beijing, is reported.


Introduction
A large effort on reforestation and afforestation has been made in China, constituting nearly one fourth of the global growth in forest as from year 2000 to 2017 as an example, which has been suggested as a great opportunity for biodiversity conservation (Hua et al. 2016. However, there are very few considerations of biodiversity effects of these projects, with monoculture species, high density stems and low underlayer coverages. Monoculture forests have been proven to support less biodiversity, as well as other critical ecosystem services, such as water yield and soil erosion control, than natural forests, both old-growth and natural-recovered , Ke et al. 2020, Hua et al. 2022. Therefore, it is necessary to set up new plans for transforming the existing artificial forest to more natural and complex ones, as well as implementing reforestation with a better design, including natural and assisted natural regeneration, as a crucial part for further ecological restoration.
State-owned forest farms are government entities that manage state-owned forest and related land in China (Ni et al. 2022). There are 64.7 million hectares of forestlands in China managed by state-owned forest farms up to 2019, consisting of 22.8% of forestlands of the country (National Forestry and Grassland Administration 2021). With the "Natural Forest Protection Policy" and other forest policies, timber logging is strictly limited national wide; therefore, most forest farms now serve with a new role in nature conservation and recreation. Therefore, there is a great potential to estimate and improve biodiversity conservation in the artificial forests of forest farms. However, earlier studies on forest farms mainly focused on single species groups or other environmental indicators (Kwok and Corlett 2000, Liang et al. 2013.
In this study, we provide the first comprehensive inventory of the largest state-owned forest farm in Beijing, the Jingxi Forest Farm, which is a representative of the forest ecoysystem of North China Taihang Mountain Range. The dataset results from our multi-species approach including mammals, birds and insects. The data would be a baseline to be provided for a biodiversity restoration pilot project by the Beijing City government, that restoration intervention and management would improve the biodiversity function of existing artifical forests.

Study site
The study site, Jingxi Forest Farm (39.94N, 115.81E) is located in the west of the Municipality of Beijing, the capital city and one of the mega-cities in the world with ca. 25 million residents. This state-owned forest farm is by far the largest one of its kind in Beijing, with an area of 116.4 km . It used to belong to the Beijing Jingmei Group, a state-owned coal mine company and a large proportion of its artificial forest was planted for industrial use since 1970s. The whole forest farm was transferred under the management of Beijing Municipal Forestry and Parks Bureau in 2017; therefore, all forests were then managed for the ecological purpose of public interest alone. The key responsiblilty of the forest farm management includes tending and thinning of existing forests, ecological restoration on former mining area by reforestation, fire prevention and patrolling against poaching. No other human activity is officially allowed so far, although some level of fungi collecting and hiking may occur. This area also serves an important role in preserving connectivity from the great mountains of of the centre Taihang where leopard (Panthera pardus japonesis) still exists (Luo et al. 2020) to the North-China Plain where tens of millions of people live.
Most part of the Jingxi is mountain area, ranging from 200 m a.s.l. to 1610 m a.s.l. with a temperate continental climate. According to the management office, this site has a mean annual temperature of 7-10°C and mean annual precipitation around 600 mm. It has a mixed vegetation of artificial and natural secondary forest on the shady slopes, as well as scrublands on most of the sunny slopes, except for the artificial forest on less steep sunny surfaces. As of 2017, Jingxi has 36.3 km of forest with canopy coverage above 20%, 63% of the area of which is pure coniferous forest, all artificially originated. The remaining part is a mixture of artificial and secondary natural forests, consisting of broadleaved or mixed forests. With rising elevation, the main tree species in artificial coniferous forest change from Pinus tabuliformis to Larix gmelinii var. principis-rupprechtii, with occasionally naturally-growing Ulmus parvifolia and Fraxinus chinensis subsp. rhynchophylla individuals. Meanwhile, the foundation species in secondary natural forests changes from Populus tremula var. davidiana to Betula pendula subsp. mandshurica, with Quercus mongolica being seen within natural forests at all elevations. This area shared some of its western boundaries with Beijing Baihua Mountain National Reserve, resulting a similar composition in flora community (Zhang et al. 2020).

Sampling design and methods
Our work mainly aimed at collecting fauna biodiversity information for this newlyestablished forest farm, providing a baseline for further management and planning. We undertook a systematic survey on several animal taxa, including mammals (camera traps and mouse traps, thus no bats were sampled), birds (visual observation and vocal identification on transects at dawn, usually within 3 hours after sunrise; for detailed survey time and effort, see Suppl. material 1; no nocturnal species were noticed) and insects (pitfall traps for quantitative surveying Carabidae species, Malaise trap for more general species), while using convenience sampling on other vertebrates and other taxa of insects during daytime (mainly Odonata and Papilionoidea). The project was started in late 2019 and is still in progress; and here we are reporting results mostly up to 31 Dec 2020, before all restoration measurement took place. For different taxa and methods, spatial and temporal coverage may vary.
Details of sampling methods, efforts and date coverage are listed in Table 1, while spatial distribution of sampling points is listed in Table 2 ( Fig. 1). As the whole area had various vegetation types, we recorded the vegetation type at each sampling point. Abbreviation of forest types in Table 2: NB = natural secondary deciduous broad-leaved forest, mainly a mixture community consisting of Quercus mongolica and Populus tremula var. davidiana at our sampling points (Fig. 2); AB = artificial deciduous broad-leaved forest, usually near abandoned (from 1970s) villages, including Robinia pseudoacacia, Juglans regia, Crataegus pinnatifida and other commonly cultivated tree species in Beijing (Fig. 3); AC = artificial coniferous forest, first planted in 1970s, mainly Pinus tabuliformis or Larix gmelinii var. principis-rupprechtii pure forest or a mixture of both species (Fig. 4). We did not sample in any scrub habitats (Fig. 5), which are abundant on most steep sunny slopes, but hard to approach for both human and other medium-large mammals, except Naemorhedus griseus.   Sampling points and their distribution in Jingxi, see Table 2. Contour lines representing elevation, using data from Earth Resources Observation and Science (EROS) Center, USGS (2018).

Data resources
Our survey records 19 species of mammals, 86 birds, four reptiles, two amphibians and one fish, as well as 101 species of insects. By far, we are only reporting occurrence information here and this information does not represent any absence data.
All mammals recorded, but one species (    Other vertebrates are mostly recorded, based on convenience sampling and are listed in Table 6. One sepcies (Elaphe carinata) is listed as Endangered in China's Red List of Biodiversity (Wang et al. 2021, Jiang et al. 2021) with other species listed as Least Concern. There are two main streams in the area and we sampled one for fish and amphibians. Reptiles were recorded on encounter.
A checklist of insects are also reported with our method of survey (  Table 6.
Checklist of other vertebrates and coordinates of their occurrence records.

Results
The total number of 213 species of animals were recorded in Jingxi area by the end of the year 2020. Amongst them, one species was listed as EN and two species were listed as VU in China's Red List of Biodiversity. More species of mammals (natural forest: 14, in total 19; artifical coniferous forest: 11, in total 19), birds (natural forest: 75, in total 86; artifical coniferous forest: 69, in total 86) were recorded in natural forest than in artificial coniferous forests. No insect sampling was done in natural forest, so no comparison was available.

Discussion
Limited by our human power and project design, we did not conduct systemtic sampling all over the site. Additionally, some of the sampling points lie in close vicinity with each other, so occurrence data from these points may not be independent. It is advised that great caution should be taken when accounting our occurrence data into any further modelling. However, comparing to exsiting GBIF data, our occurrence data, as well as environment information still expand the knowledge of several data-poor species in this less-studied area and add evidence to the comparison of biodiversity between artificial forest and natural forests (Carnus et al. 2006, Brockerhoff et al. 2008, Horák et al. 2019. Our data show that there is a clear trend that artificial coniferous forests hold less mammalian and avian biodiversity than natural forest of the same age, altitude and slope, even with more sampling effort, larger coverage (see Table 2) in artificial forests. This justifies our suggestion that artificial coniferous forests should be modified for better biodiversity and other ecosystem services outcomes. Our result also suggest that, despite large areas of artifical forest present, these sites still server as important wildlife habitats, as other artificial forests can do (Harich andTreydte 2016, Tanalgo et al. 2021). Jingxi supports biodiversity similar to one important nature reserve, Baihua Mountain National Nature Reserve, which lies in the close vicinity to the west of Jingxi. The NNR have better natural vegetation (Zhang et al. 2020) and further distance to the urban area, thus less human disturbance, comparing to Jingxi; thus, there is a good opportunity for comparing biodiversity on different spots in the names of different vegetation, human disturbance and management planning. As an example, comparing with latest camera trap report and historical checklist from Baihua Mountain NNR (Fu et al. 1994), we reported a nearly similar composition of mammal species, except for small mammals like bats, rodents and shrews. This similarity in mammal composition shows that Jingxi has a great importance in providing suitable and specific habitats and connections for wildlife.
Our results also call for surveying and monitoring projects to take place in nearby areas and their results to be published with clear spatial information. Although not rarely captured by camera traps in our study, four over 19 total recorded mammal species are data-poor in GBIF, indicating great potential of occurrence data to be published. Furthermore, monitoring of the population trend of vulnerable species N. griseus at larger scale is crucial, as it was once listed as VU in the IUCN Red List, while it is not currently assessed in the Red List as Mori et al. (2019) suggested that N. griseus should be treated as a subspecies of N. goral. The lack of assessment and information could become a barrier in the conservation of this nationally important and vulnerable species/population. Therefore, more information and long-term monitoring on this population is important for further conservation actions.
Finally, there are still knowledge gaps on information about amphibians and reptiles, as well as bats in vertebrates, while no flying squirrels were witnessed within the area. No research on either amphibian or reptile diversity was found for this area or adjoining areas like Baihua Mountain NNR. With a total of five species of amphibians and reptiles beingreported in Jingxi, it shows great potential for a systematic survey to be conducted, comparing with the result of 22 species municipality-wide (Shi et al. 2022). Additionally, we still have a poor level of different taxa of insects covered in our survey as only pitfall traps and malaise traps were used, while it is still more challenging when it comes to other invertebrate taxa.

Conclusions
This dataset shows that an artificial forest farm near a mega city such as Beijing can harbour a considerable biological richness and serve as natural habitat to wildlife. It adds evidence to findings from other forest types with a multi-species approach (Kwok and Corlett 2000, Stephens and Wagner 2007. This is also the first report (both in English and Chinese) on biodiversity in Jingxi. Our results provided a baseline for a next step of ecological restoration, while showing a trend that natural forest held all and more species than in artificial forests on site, justifying our design that transforming artificial forest to natural forest as a means of the restoration.
Considering the large proportion of artificial habitat, especially pure coniferous forest, presenting in Jingxi, it is also worth implementing further management on plant community, based on our knowledge of fauna species' preference, promoting richer biodiversity and more lively environment near the urban community.