Magnolia sinica (Law) Noot. (Magnoliaceae), a tree endemic to China, was first described as Manglietiastrum sinicum Law (Law, Reference Law1979); the type locality is in broadleaved evergreen forest near Fadou village, Xichou County, in south-east Yunnan Province. Most Chinese botanists refer to the species as M. sinicum and also use its common name huagaimu (Sun et al., Reference Sun, Zhou, Li, Yang, Tao, Maschinski and Haskins2012). The species is categorized as Critically Endangered on the IUCN Red List, based on limited data suggesting a population of < 10 mature individuals (Wang & Xie, Reference Wang and Xie2004; Cicuzza et al., Reference Cicuzza, Newton and Oldfield2007; Zheng et al., Reference Zheng, Sun and Zhao2008). The term ‘plant species with extremely small populations’ is used to describe species that are restricted to a narrow geographical distribution, have been subjected to disturbance and other external stress factors over a long period of time, and whose population size has decreased below a critical threshold necessary to prevent extinction (Ren et al., Reference Ren, Zhang, Lu, Liu, Guo and Wang2012; State Forestry Administration of China, 2012; Ma et al., Reference Ma, Chen, Grumbine, Dao, Sun and Guo2013). Magnolia sinica is one such species, and it has been targeted as one of the 20 such species approved by the Yunnan government for urgent rescue action before 2015. Here we (1) update distribution data for M. sinica and reassess the number of wild individuals, (2) revise the species' main floral characteristics, and (3) examine the conservation of the species.

We firstly gathered information covering most of Yunnan Province, with a focus on the south-east (including almost all counties in Wenshan, Honghe: Maguan, Xichou, Malipo, Jinping and Hekou/Pingbian prefectures). We reviewed Flora Reipublicae Popularis Sinicae (Law, Reference Law1996), Flora Yunnanica (Law, Reference Law2006) and Flora of China (Xia et al., Reference Xia, Liu and Nooteboom2008) for information on the species, and examined all M. sinica specimens in the three major herbaria of the Chinese Academy of Sciences (KUN, IBSC, PE) to identify potential sites for field verification. We then surveyed for M. sinica during 2004–2014, using field measurements and participatory rural appraisal (Chambers, Reference Chambers1994, Reference Chambers2007; Upjohn et al., Reference Upjohn, Attwood, Lerotholi, Pfeiffer and Verheyen2013). We interviewed 20 people (five members of staff at each of four reserves), in particular regarding factors affecting the vulnerability of the species. During the field surveys several previously unrecorded individuals of M. sinica were discovered, bringing the total number of known individuals in the wild to 52, scattered throughout the prefectures of Wenshan and Honghe in Yunnan (Fig. 1) at altitudes of 1,339–1,707 m. Thirty-six of the remaining individuals are in Wenshan prefecture (Table 1).

Fig. 1 Known locations of Magnolia sinica in China, with the total number of individuals at each location and, in parentheses, the number of these that are within nature reserves. New records are indicated in blue, and previously known populations in black. The shaded rectangle on the inset shows the location of the main map in China.

Table 1 Records of Magnolia sinica in the wild, with locality, geographical coordinates, altitude, and number of individuals in categories III–IV (DBH 2.5–7.5 cm and 7.5–22.5 cm, respectively) and V–VIII (DBH 22.5–45 cm, 45–67.5 cm, 67.5–100 cm and ≥ 100 cm, respectively).

¹ The individuals in Xichou are all in nature reserves. Two individuals in Dalishu, Maguan are in state-owned forest. The individuals in Hekou are all in non-commercial forest.

² New records

³ Records include seven previously known individuals

⁴ Records include three previously known individuals

⁵ Based on information provided by local people

We collected morphological data for each of the 52 individuals, and grouped them into eight categories, based on height (H) for smaller individuals and diameter at breast height (DBH) for larger individuals: I (H < 33 cm), II (H ≥ 33 cm, and individuals for which DBH can be measured with DBH <2.5 cm), III (DBH 2.5–7.5 cm), IV (DBH 7.5–22.5 cm), V (DBH 22.5–45 cm), VI (DBH 45–67.5 cm), VII (DBH 67.5–100 cm) and VIII (DBH ≥ 100 cm). There was considerable variation in size, implying differences in age. There were no seedlings near those adult trees that could have produced seeds (Fig. 2). The height of adult trees (V–VIII) was 13–42.6 m, and DBH was 25.5–127 cm; the height of immature trees (III–IV) was 4–15 m, and DBH was 5–20 cm. Our revision of the floral traits of M. sinica is provided in Supplementary Material 1.

Fig. 2 Numbers of wild M. sinica individuals in each of eight size categories, defined in the text. Information provided by local people about six individuals is not included.

Magnolia sinica in Kunming Botanical Garden can be traced to a single introduction event in 1983. Of the 10 individuals introduced to the garden in 1983 for ex situ conservation the first flowering (of an individual 13 m in height) occurred in March 2013. During the flowering period (March–April) in 2013 we performed bagged pollination treatments to assess seed-set from self-pollination (n = 6 flowers) and outcrossing (cross-pollination with pollen from a wild tree in Jinping County, n = 2 flowers); as a control 10 flowers were untreated and accessible to natural pollinators. None of the flowers open to pollinators produced any fruit, whereas all hand-pollination experiments showed seed-set of 14.29–60.42% (calculated as the number of successfully fertilized ovules producing seed relative to the total number of ovules per flower). The eight follicles that resulted from the pollination experiments contained a total of 154 seeds. At least some of these seeds were viable, as several germinated after incubation in humid sand for 6 months.

Seeds were collected for ex situ conservation and propagation from four trees at the type locality in Fadou village, Xichou County, in 2004 (Batch 1) and in 2005 (Batch 2; Zhang et al., Reference Zhang, Shen and Wang2014). Both batches of seeds were sown in a nursery affiliated with Kunming Botanical Garden. After 2 years of nursery growth 100 saplings of each batch were transplanted to Shangchang (SC1 and SC2) and another 100 saplings of each batch to Xiaoqiaogou (XQG1 and XQG2). These two reinforcements were monitored by local rangers, who recorded the height, basal diameter and crown breadth of the saplings. Saplings were also transplanted to Kunming Botanical Garden, batch 1 (KBG1) after 3 years of nursery growth and batch 2 (KBG2) after 6 years of nursery growth, and both batches were monitored to record annual growth. Beginning in June 2012 we measured the growth of batches KBG1 (45 individuals) and KBG2 (22) annually. In June 2013 the survival rate of both batches was 100%. In 2013 the numbers of surviving individuals at the reinforcement sites were 20 (SC1) and 18 (XQG1) after 6 years, and 21 (SC2) and 23 (XQG2) after 5 years. We believe that the low survival rate of these two batches of reintroduced seedlings is attributable to inappropriate handling of the saplings during transplanting (c. 48% of the saplings were planted directly without removing the plastic bags in which they had been grown, thus hindering the young plants from taking root). The annual increments of height, basal diameter and crown breadth are in Supplementary Table S1.

Although the species is protected at the national level as well as locally in Yunnan, it is on the brink of extinction in the wild. The low number of saplings and young trees is limiting the potential for regeneration, and the surviving individuals all occur in remnant populations that are fragmented and comprise too few individuals for natural recovery. As M. sinica is unlikely to recover naturally, effective conservation of this species requires intervention.

We therefore propose two groups of measures to protect M. sinica: (1) in situ conservation, population monitoring, and public engagement, and (2) ex situ conservation with reinforcement or reintroduction.

(1) Many of the individuals remaining in the wild (22 of 52, in three populations) are outside protected areas, and therefore expansion of reserves or establishment of new sites for in situ conservation is necessary (Shen et al., Reference Shen, Wang, Zhang, Wu, Jiang and Jiang2013). Local governments and nature reserves could limit any further habitat destruction by reducing cultivation of amomum, rubber and banana, which are grown in former M. sinica habitat, and renaturalization of farmland could create new habitat for reinforcement or reintroduction. To bridge the gap between scientific research and public education, establishing better cooperation between local governments and research institutes, including universities and environmental organizations, is needed for long-term monitoring and research and to develop strategies for the conservation of M. sinica.

(2) The seeds cultivated in Kunming Botanical Garden and used for reinforcement of wild populations were from only one site (Xichou). It is essential to collect seeds from other sites and individuals to increase the genetic diversity of the seed pool for further ex situ conservation and for reinforcement or reintroduction (Ren et al., Reference Ren, Jian, Chen, Liu, Zhang and Liu2014). One of the criteria for judging the success of ex situ conservation is whether the transplanted plants can produce offspring (Law et al., Reference Law, Zhou and Zeng1997), and M. sinica has done so in Kunming Botanical Garden.