Red-crowned crane (Grus japonensis) prefers postharvest reed beds during winter period in Yancheng National Nature Reserve

Study area

This study was undertaken in the YNNR (32°48′47″−34°29′28″N, 119°53′45″−121°18′12″E) in Jiangsu Province, eastern China (Fig. 1). Field studies were conducted under the permission from the Administrative Bureau of the YNNR. The YNNR was established in 1983 and was promoted to a national nature reserve in 1992. Currently, the YNNR is the largest coastal zone reserve in China, with a total area of 247,300 ha. As a transition region with a temperate zone, the YNNR has the characteristics of a monsoon marine climate, where the mean temperature is approximately 4 °C in winter, and the annual average precipitation is 980–1,070 mm. Winter precipitation is rare in the area. Vegetation in the core zone of the YNNR shows a clear zonal distribution from the coast to inland with bare tidal flats or supporting communities, such as those dominated by exotic smooth cordgrass, common seepweed, reed and other xeromorphic vegetation (Li et al., 2019). Degree of plant cover is high in reed beds whereas it is relatively low in the common seepweed tidal flats.

Reed management

As a traditional agricultural activity, local people harvested a partial of the reed beds in the core zone of the YNNR in mid-late December annually. After the reed were harvested, reed beds were exposed, and regulation of shallow water level also occurred by introducing water from adjacent rivers and fish ponds. Thus, we distinguished two phases of the wintering period of red-crowned crane: the preharvest period (8 November to 20 December) and the postharvest period (22 December to 6 March).

Study species

Red-crowned crane, which is an endemic species in East Asia and is listed as “endangered” on the IUCN Red List of Threatened Species, has a population of approximately 3,000 individuals (Birdlife International, 2016). Its breeding areas include Japan, eastern Russia, Mongolia, and northeastern China. The wintering areas are in Japan, South Korea and the eastern coastal wetlands of China. Annually, red-crowned cranes usually aggregate in large groups before their autumn migratory from northeast China to the YNNR in late October and relatively remained stable until their spring migratory in early March.

Crane numbers and distribution

Regular crane habitat distribution surveys were performed throughout days with good weather conditions along a fixed route (Fig. 2, approximate nine km) by motorcycle in the core zone from November 2014 to March 2015. Surveys were conducted roughly for every two days, and also for several consecutive days. Because of the crane’s behaviors different though out a day (Shao et al., 2018), the beginning time of our survey was around 9:30 in the morning as wintering cranes were gregarious and shared communal roosting sites at night and usually departed their roosting sites during the morning (06:30–08:00) to forage and returned at night (Lv, 2007). Approximate 2.5 h were spent for each survey. A total of 71 surveys were conducted during the entire winter season. Based on differences in vegetation and micro geomorphic features, the distribution habitat patches of red-crowned crane were divided into four types according to dominant plants, including alien smooth cordgrass tidal flats, common seepweed tidal flats, reed beds and aquaculture fish ponds. Winter fish harvest was common in the buffer and transition zone in the YNNR. However, people in our study area in the core zone did not drain the water off from fish ponds for fish harvesting during the winter. Whenever cranes were encountered, we used a monocular telescope (Nikon, 20 × 60) to scan, count and record the number of birds and habitat types. In order to facilitate the observability, we also get up to scan with greater vision on the mound platforms, which were approximate five m high and built for bird monitoring, along the fixed route. Crane surveys in uncut reedbeds were also conducted before reed harvest. We recorded as “0” when no cranes were observed in uncut reed beds.

Potential animal food resource collection of the cranes

During the wintering period, the cranes tend to consume animal food resources because of the rapid aging of food plants. In addition, due to the high variability in digestibility of different animal food items, a fecal analysis would create a bias (Redpath et al., 2001). Consequently, field sampling provides a simple, minimally invasive manner for estimating the potential animal food resources available in the foraging habitats of this threatened bird species (Bishop & Li, 2002; Dong et al., 2016). According to previous studies about diet of red-crowned crane by stomach anatomy and fecal analysis, we aggregated the species into four items for the analysis: Annelida, Mollusca (snails and shellfish), Arthropoda (crabs) and Chordata (fish) (Shi & Wu, 1987; Ma, Wang & Tang, 1999). We chose two habitats, including common seepweed tidal flats and reed beds, for specimen collection in the core zone before and after reed management. Three 100 × 100 m plots were marked out in common seepweed tidal flats on the sites most occupied by the cranes and another three were marked out in reed beds before reed harvest. Same sample size was applied after reed cut. We sampled these twelve plots for one time during the study period (Fig. 2). Five random quadrats of one m × one m with 20 cm deep were dug up and removed from within each plot (Jia et al., 2013). To sample crabs in these plots, pitfall trap was deployed in each quadrat by burying cylindrical plastic buckets (20 cm in diameter, 30 cm deep) at the soil surface (Chen et al., 2019). The crabs in each trap were checked in the next day. Because of the shallow water level in these habitats, we directly used a customized fish net (one m × one m, four mm mesh) to sample them before digging. The samples were dried at 60 °C for 48 h. We determined animal species and measured the dry weight using an electronic balance (Shengke Instrument Co., Ltd., Shanghai, China; 2,000 g/0.01 g). Weight determinations were then converted to biomass per plot (g/m²).

Data analysis

We assessed the habitat distribution patterns of red-crowned crane based on its occurrence in different habitats. The relative abundance (%) of cranes in each type of habitats were calculated throughout the winter period. The ratio of the number of cranes that occurred in each habitat type to the total number of cranes in all habitats was estimated as the relative abundance (Zheng et al., 2015) using the following equation:

Relative abundance = Ni/N

“Relative abundance” is the relative number of cranes found in habitat type i over all crane individuals in all habitats; Ni is the recorded number of cranes in habitat i; and N is the total number of cranes in all habitat types. Differences in the relative abundance within or between habitats indicate changes in crane habitat distribution patterns over spatial–temporal scales. Linear regression was used to estimate changes in the crane numbers in different habitats throughout the wintering period. To assess differences in total food biomass between the two phases and the major distribution habitats, we performed a two-way ANOVA with phase/habitat and the interactions as fixed effects followed by a post hoc pairwise Tukey’s HSD test (Faraway, 2002). All statistical analyses were performed using the software R (version 3.5.1; R Core Team, 2018).

Crane habitat distribution

From November 2014 to March 2015, we conducted 71 regular red-crowned crane surveys (31 surveys and 40 surveys before and after the reed harvest, respectively). A total of 1,586 individual cranes in the four types of habitat, including smooth cordgrass tidal flats, common seepweed tidal flats, reed beds and aquaculture fish ponds, in the core zone of the YNNR were recorded during our study period. Throughout the wintering period, the number of cranes recorded in the common seepweed tidal flats remained stable (R² = 0.031, F = 3.244, df = 69, and P = 0.076, linear regression), whereas the number increased significantly (R² = 0.625, F = 111.7, df = 69, and P = 0.000) in the reed beds throughout the wintering period (Fig. 3).

During the study period, the cranes occurred mainly in the common seepweed tidal flats (relative abundance = 51.8%) and reed beds (relative abundance = 45.8%) (Table 1). In addition, the relative abundance and number of cranes were 1.6% and 26 individuals in the smooth cordgrass tidal flats and 0.8% and 13 individuals in the aquaculture fish ponds, respectively, throughout the wintering period, which meant that cranes rarely occurred in these two habitats.

Reed management activities noticeably influenced the distribution patterns of red-crowned crane. In the preharvest period, the number of cranes in the common seepweed tidal flats increased significantly (R² = 0.176, F = 6.19, df = 29, and P = 0.019), and cranes occurred more often in this habitat (relative abundance = 95.0%) than in the other three habitats (relative abundance = 5.0%) (Table 1). Moreover, cranes were recorded in the common seepweed tidal flats only during 23 surveys (n = 31, Figs. 3 and 4). During the postharvest period, both the relative abundance (36.4%) and number of cranes in the common seepweed tidal flats decreased significantly (R² = 0.102, F = 4.307, df = 38, and P = 0.045), whereas the relative abundance (61.0%) and number increased and remained stable (R² = 0.083, F = 3.428, df = 38, and P = 0.072) in the reed beds until early March before the spring migration (Figs. 3 and 4).

Variability in potential animal food resources

Food biomass was considered the potential animal food resource availability in the common seepweed tidal flats and reed beds. The species composition and biomass of potential animal food resources in these two habitats throughout the two phases contrasted as the relative abundance of red-crowned crane decreased in the common seepweed tidal flats and increased in the reed beds after reed harvest. A total of 14 species were found among these two habitats. More species were recorded in the common seepweed tidal flats than in the reed beds and throughout the wintering period (Table 2). Before the reed harvest, three crab species and only one fish species were sampled in the red beds. However, crab species were not recorded, and the number of fish species increased to seven after the reed harvest due to inundation (Table 2).

It should be noted that the food biomass differed significantly among the habitat types (n = 12, P = 0.012) due to the interactive effect of phases and habitats (n = 12, P = 0.004) (Table 3). Food biomass was more abundant in the common seepweed tidal flats than that in the reed beds before the reed harvest, and the food biomass decreased after the reed harvest. In contrast, the food biomass increased in the reed beds after the reed harvest (Fig. 5).