Fanjingshan Mountain is located at Guizhou Province (south-western China), running in a north-east–south-west direction and having an approximate area of 570 km² (GFNNRAB, 2004). This mountain is part of the Three Gorges Region (Fig. 1), which is one of the three main biodiversity hotspots and endemism centres in China (Ying & Zhang, 1984; Ying, 2001; Li et al., 2009; López-Pujol et al., 2011a , b ). The rich biodiversity in this hotspot is mainly a result of glacial refugia that existed in the area during late Neogene and Quaternary periods (Li, 1940; Axelrod et al., 1996; Qian, 2001; López-Pujol et al., 2011a ). Consequently, this biodiversity hotspot is probably China’s region with the highest concentration of relict species; most of these are “living fossils” at present surviving in a small part of it original distribution area due to range contraction as a result of the late Cenozoic global cooling.

Nowadays in Fanjingshan Mountain and its nearby regions, some wild individual or even wild populations of relict species can still be found (Axelrod et al., 1996; Tang et al., 2011, 2012), including Davidia involucrata Baill., Ginkgo biloba L., Liriodendron chinense (Hemsl.) Sarg., and Metasequoia glyptostroboides Hu & W. C. Cheng (GFNNRAB, 2004; Tang et al., 2011, 2012). This fact indicates that Fanjingshan Mountain provided long-term relatively stable habitats for these relict plants. The field surveys also revealed a considerable number of long-lived trees (≥100 years old) of these relict plants around the mountain (Zhang, 2004). The places where the natural long-lived individuals occur may indicate the locations of suitable habitats for relict species (probably as “refugia within refugia”; Wang et al., 2009). The frequencies and distribution patterns of these long-lived individuals can therefore be used to determine growth conditions and the nature of microhabitats for these relict species or other rare and endangered species, which is important for deciding in situ conservation sites and is of instructive significance for ex situ conservation.

Data collection

Three local forestry governments surrounding the Fanjingshan Mountain—i.e. Songtao County, Jiangkou County, and Yinjiang County (Fig. 2)—had respectively carried out field surveys in summers of 2010–2012 on long-lived trees (≥100 years old) with plant scientists from universities and institutes. The reports of these surveys have not published yet but can be obtained from the local forest governments upon request. According to geographic and topographic characters, we divided the surroundings of the mountain into four regions, i.e. north-east (Songtao County), south-east (Jiangkou County), north-west (the north of Yinjiang County), and south-west (the south of Yinjiang County) (Fig. 2). The NW and SW regions were separated roughly by an expressway. The NE and SE were separated roughly by the Zhaiying and Xiaojiang rivers at the boundary of Songtao and Jiangkou counties (Fig. 2). We collected tree age and diameter at breast height (1.3 m) (DBH) for each long-lived tree.

Data analyses

The data of long-lived individuals were analysed separately for the four regions. Based on the national criterion, the long-lived trees can be divided into three classes: Class-I refers to trees of at least 500 years old; Class-II is trees with an age of 300–499 years, Class-III includes long-lived trees less than 300 years old. We focused our study on six widely-recognized relict plants (Figs. 2 and 3): Cyclocarya paliurus (Batalin) Iljinsk., Ginkgo biloba, Liriodendron chinense, Pinus massoniana Lamb., Podocarpus macrophyllus (Thunb.) Sweet, and Taxus chinensis (Pilg.) Rehder.

To detect the growth status of the trees, we estimated the relationship between tree age and DBH for each relict species respectively for the four regions. These data were also used to compare the habitat quality among four regions, e.g. the habitat can be assumed as of good quality if the DBH of long-lived trees increased steadily with age.

To explore the plant communities and habitat similarity among four regions, we draw the floristic similarity dendrogram according to Tang et al. (2011, 2012) using the PAST software (Hammer et al., 2001). The values of relative DBH for long-lived individuals of Ginkgo biloba and Taxus chinensis were applied in the similarity dendrogram analysis using Euclidean and group average clustering. The other relict species were not included into this analyses because they are too few and some are absent in one or two regions.

Distribution patterns

All the information about tree age, DBH, and tree height of long-lived individuals is available in the Appendix. The age classes and spatial distributions of these plants are shown in Fig. 2 and Table 1. Ginkgo biloba and Taxus chinensis were the most common species among the long-lived trees (Figs. 2 and 3, and Table 1), while the other relict species were restricted to a few small areas. Both SW and NE regions appeared to be the accumulation centres of long-lived trees (Fig. 2). Especially the SW region harboured a large proportion of Class-I long-lived trees of G. biloba, some of which were of ≥1000 years old (Fig. 4 and Appendix). For example, six ≥1000-year-old trees and eight Class-I trees of G. biloba were found at a small site (Yangliu Town) located at the south of this region (Fig. 2). Another distribution centre of G. biloba is the NW region, where three 1000-year-old trees and another five Class-I trees were found (Figs. 2 and 4, and Appendix).

Table 1.

The eastern side of Fanjingshan Mountain, especially the SE region, had fewer long-lived trees as compared with the other regions (Fig. 2 and Appendix). Only four Class-I trees of G. biloba occurred at the NE region, while no Class-I trees were found in the SE region (Fig. 2). All the six relict species were found in the NE region and most individuals grew along the east fringe of the mountain and the boundary with the SE region, where several small rivers and valleys occur. Compared to G. biloba, T. chinensis had much fewer Class-I trees and no trees were older than 600 years (Fig. 4). The distribution centres of T. chinensis were quite similar to G. biloba, located both at the NE region and the west side of the mountain.

Correlation of DBH and tree age

The DBH increased with the age more quickly in G. biloba than in T. chinensis (Fig. 4A, B) for long-lived trees below 600-year old. For G. biloba trees with an age of >600 years, DBH increased very slowly (Fig. 4B).

For G. biloba, the long-lived trees from NW region showed a relatively larger DBH than those from the other three regions (Fig. 4B). For T. chinensis, although the DBH values were similar among the four regions, these increased more quickly in trees from NE region compared to the other regions. Consequently, the long-lived trees at the north side to the mountain had relatively larger DBH values for both relict species.

Region similarity

The dendrogram showed that the SW and NE regions were the most similar regarding the DBH of long-lived trees (Fig. 5). The SE region was clustered as the sister group of SW and NE regions, whereas the NW region is the most dissimilar one (Fig. 5).

Grouping of long-lived trees

We also found that some long-lived trees were growing in groups (Fig. 3C, D). For example, many long-lived trees were found in fengshui (“wind and water”) forests surrounding villages in the valley (Fig. 3C), which were well protected by local residents. Isolated long-lived trees usually occurred near villages (Fig. 3A) or farmlands (Fig. 3B), and some were regarded as “God trees” by local residents for good luck and long life (Fig. 3F).

Our data showed that the west side and the NE region around Fanjingshan Mountain are main distribution sites of long-lived trees and can be seen as the real stable natural habitats for relict species such as G. biloba and T. chinensis. Such information has significant implications for conservation of these “living fossils” and local ecosystems.

All the six relict plants can be found at the NE region to the Fanjingshan Mountain (Fig. 2 and Appendix). This pattern is probably the result from the evolutionary history and demographic dynamics of relict plants during glacial periods. Fanjingshan Mountain has an altitude of >2000 m and would have acted as a main refugium for various plants during the last three million years (Qian, 2001; López-Pujol et al., 2011b ). During the glacial periods of the Quaternary, most parts of China were much colder than at present (Li, 1940), and only the mountainous areas with large altitudinal gradients provided opportunities and habitats for the plants to survive. The Fanjingshan Mountain is at the middle point between East and West China (Fig. 1) and probably is among the first stop-off sites when the plants retreated from the eastern lowlands. Therefore, deep valleys of the NE side of Fanjingshan Mountain are probably one of the main survival places of these relict species during their range contraction in the glacial periods (Ying & Zhang, 1984; Ying, 2001; Li et al., 2009; López-Pujol et al., 2006, 2011a , b ). The SE region, however, have much fewer long-lived trees (Fig. 2); this region is mainly consisting of lowlands without deep valleys (GFNNRAB, 2004) having, thus, a very limited role as a refugium (i.e. plant species had less opportunities to track the climatic changes). Furthermore, the SE region is the main place for tourists entering Fanjingshan Mountain and some habitats for relict plants have been destroyed due to road and hotel construction (GFNNRAB, 2004).

Although the NE region is the area with highest species diversity of relict plants, the distribution centre of long-lived trees, especially the very old trees (Class-I and Class-II) is at the west side of the mountain (Fig. 2 and Table 1). A reason might be that the western surrounding regions are well protected by the mountain from the cold winds. and thus the habitats would have remained more stable than in the eastern side. This is probably why the west side of the mountain, including NW and SW regions, has much more Class-I trees, some of which are >1000 years old (Fig. 4 and Appendix).

Most Class-I and other long-lived trees in the SW region were found in a small village named Yangliu (more than 90% Class-I trees of G. biloba and T. chinensis grew at this village; Fig. 2). Recruits of these long-lived trees can be also found near the trees or at very close sites (H.-Y. Liao, pers. obs.). This information indicates that this place probably constitutes a long-term stable habitat for these relict species or other similar relict plants. Two additional scenarios can also explain the high occurrence of long-lived trees in this small village. First, the village is in a deep valley, far away from intense human activities. Second, the local residents of Yangliu Village are largely belonging to ethnic minorities including Tujia and Miao (GFNNRAB, 2004); these communities normally protect big and long-lived trees as “God trees” for long life or good luck (Fig. 3F) but also protect groups of long-lived trees as fengshui forests for environmental safety and good living conditions (Fig. 3C). Therefore, Yangliu Village should be given highest priority for in situ conservation, not only for its high diversity of long-lived trees but also for its stable habitats that are protected by local residents.

Although the ethnic minorities usually protect trees around their villages, they also utilize the mountain plant resources directly, such as fuelwood gathering and burning grass in fields to clear and fertilize soils (GFNNRAB, 2004; Wandersee et al., 2012). Recently, the mountain is experiencing rapid changes in land use and forest cover due to tourism, timber harvesting, and road construction (GFNNRAB, 2004; Wandersee et al., 2012). Some long-lived trees were also be transplanted to cities for their commercial value (Fig. 3F). All these activities have the potential to seriously affect rare and endangered plants, particularly long-lived trees. This is probably why many long-lived trees are now only found in the protected fengshui forests near villages (Fig. 3A, C). Therefore, it is necessary to set up several suitable sites in this village or nearby places for in situ conservation of these relict species or other rare and endangered plants. Fortunately, reforestation programs from national and local governments, such as the Natural Forest Conservation Program (NFCP) and the Grain-to-Green Program (GTGP), were carried out since about one decade ago (GFNNRAB, 2004; Liu et al., 2008; Wandersee et al., 2012). The NFCP is geared to protect forests, water, and soil for environmental safety and biodiversity (Li, 2004; Wandersee et al., 2012). The GTGP aims to plant trees on steep slopes where were once used by farmers as croplands, and the farmers are compensated through cash or foodstuff (Feng et al., 2005).

In conclusion, our data suggest that, although the Fanjingshan Mountain (which is a National Natural Reserve) is the main distribution region of rare and endangered plants including relict species, the surrounding places are also of conservation value due to their large number of long-lived trees of relict plants and the primaeval habitats protected by local residents. We should not only protect the local biodiversity within the limits of the natural reserve, but also pay enough attention to the rare and endangered species that occur out of the range of the reserve.

Appendix