Seed supply and the regeneration potential for plantations and shrubland in southern China

Forest Ecology and Management 259 (2010) 2390–2398

Contents lists available at ScienceDirect

Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco

Seed supply and the regeneration potential for plantations and shrubland in southern China Jun Wang, Danyan Li, Hai Ren ∗ , Long Yang Heshan National Field Research Station of Forest Ecosystem, South China Botanical Garden, Chinese Academy of Sciences, Xingke Road 723, Tianhe District, Guangzhou 510650, China

a r t i c l e

i n f o

Article history: Received 10 January 2010 Received in revised form 15 March 2010 Accepted 21 March 2010 Keywords: Indigenous species Plant recruitment Seed bank Seed dispersal Seed limitation Succession

a b s t r a c t Assessing the characteristics of seed supply will be vital to better understand the dynamics of forest regeneration. In this study, we surveyed the aboveground vegetation, the seed rain, the seed bank, and natural seedling emergence in four typical 24-year-old plantations (eucalyptus, mixed-native, mixedlegume, and mixed-conifer) and a naturally successioned shrubland in southern China. The dominant species in the understory were similar among the five plant communities. The seed rain and the seed bank were dominated by shrubs and herbs but indigenous tree species were rare. Species that were common to all five-plant communities represented a great proportion of the seeds in the seed rain and seed bank. The seed rain consisted mostly of seeds derived from the local plant community. Seed abundance was greater in the seed bank than in the seed rain, and species richness was greater in the seed bank and in the corresponding plant community than in the seed rain. Species composition similarity between the seed rain, the seed bank, and the aboveground vegetation was low, because the seed rain contained much fewer species, and the seed bank and aboveground vegetation contained many different species, respectively. These findings indicate that both the seed rain and the seed bank play important roles in providing seeds for plant recruitment in the understory, but the seed bank contributes more than the current seed rain to the diversity of recruited plants. The current plant community has little impact on the qualitative composition of the seed rain and seed bank. Based on these data, it appears that succession to the desired zonal, mature forest community is unlikely to result from seeds in the seed rain or seed bank. Lack of seed availability of desired zonal mature forest species is the main bottleneck currently limiting succession in the plantations. Reintroduction of late-successional species could facilitate the desired succession. © 2010 Elsevier B.V. All rights reserved.

1. Introduction During the last half-century in southern China and Southeast Asia, deforestation and land use change have severely reduced forest cover. Because of economic transformation and conservation initiatives, however, forest cover has steadily increased since the 1990s as a result of plantation establishment and natural restoration (Howlett and Davidson, 2003; Peng, 2003). For example, the forest cover in Guangdong Province in South China increased from 26% in 1979 to 50% in 1998. Still, the understories of established plantations, which are mainly planted with exotic trees species, are dominated by shrubs and herbs and are rarely colonized by indigenous tree species (Ren et al., 2002; Duan et al., 2008). Such plantations remain in the pioneer stages, and succession to more natural communities is inhibited (Ren et al., 2007).

∗ Corresponding author. Tel.: +86 20 37252916; fax: +86 20 37252916. E-mail address: [email protected] (H. Ren). 0378-1127/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.foreco.2010.03.015

Recruitment of indigenous woody plants may be constrained at different life stages. Seedling emergence and establishment are critical in determining the fate of individual plants, and these processes greatly influence forest regeneration (Grubb, 1977; Harper, 1977). For emergence and establishment to occur, however, seeds must be available (Denslow et al., 2006; Shono et al., 2006; Lentink et al., 2009). Therefore, understanding the characteristics of seed supply in the current plantations, which cover a large portion of southern China, is crucial for understanding long-term community dynamics in general and succession to more natural communities in particular. In many ecosystems, forest regeneration relies on the seed rain and soil seed bank (McClanahan, 1983; Urbanska et al., 1998; Bossuyt and Hermy, 2004; Pakeman and Small, 2005). The seed rain of a community is the result of seed production from plants within the community and seed input from adjacent communities (Booth and Larson, 1998). Seed rain plays a key role in the subsequent recruitment of new plants and thus underlies forest community structure, dynamics, and regeneration (Fuller and

J. Wang et al. / Forest Ecology and Management 259 (2010) 2390–2398

Moral, 2003; Pakeman and Small, 2005; Tackenberg and Stocklin, 2008). Previous research has also demonstrated that seed dispersal can facilitate recolonization by indigenous species and thereby accelerate forest succession to more natural stages (Hubbard and McPherson, 1999; Shono et al., 2006). Besides influencing the future plant community, seed rain also greatly affects the seed-bank composition because seed-bank renewal depends on seed rain (Alvarez-Buylla and MartinezRamos, 1990). The soil seed bank, which reflects the current and past plant community, can help prevent local species extinction and can also act as a seed source in forest regeneration (Maranon, 1998; Olano et al., 2002; Auld et al., 2007). Most importantly, the soil seed bank can be a source of colonizing species (e.g., zonal mature forest species) that can accelerate forest succession (Augusto et al., 2001; Luzuriaga et al., 2005). The contributions of the seed rain and seed bank to vegetation regeneration depend on seed quantity and composition (Thompson and Grime, 1979; Fuller and Moral, 2003). To date, however, the effects of the seed rain and seed bank on early stages of regeneration in the established plantations in southern China have not been well studied. In this study, we investigated the quantitative and qualitative composition of the seed rain and soil seed bank and natural seedling emergence in four typical plantations and in a natural shrubland, located in southern China. All five sites had experienced 24 years of succession. The general question asked was whether the seed rain and seed bank can speed up the succession of current plantations to more natural stages. The specific questions were: (1) How does species composition of the seed rain and the seed bank relate to species composition of the aboveground vegetation? (2) What is the pattern of seedling recruitment in the field? (3) How does species composition differ in the seed rain and the soil seed bank? In addition, we discuss the relative importance of the seed rain and seed bank in contributing to plant recruitment.

2. Materials and methods 2.1. Study site The study site is located at the Heshan National Field Research Station of Forest Ecosystem, Chinese Academy of Science (112◦ 54 E, 22◦ 41 N), Heshan City, Guangdong, southern China. This site is characterized by a typical subtropical monsoon climate with a mean annual temperature of 21.7 ◦ C. The mean annual rainfall is 1700 mm, which is concentrated between April and September. The mean annual evaporation is approximately 1600 mm, and the elevation ranges from 0 to 90 m. The soil is an acrisol. The zonal climax vegetation is lower subtropical monsoon evergreen broad-leaved forest (typically comprising Cryptocarya concinna, Cryptocarya chinensis, and Aporosa yunnanensis), but this zonal climax vegetation is not represented at the research station; the closest remnant of the zonal climax vegetation is located at Dinghushan Mountain, about 70 km north of the research station. The closest Fengshui forests (i.e., secondary forests that consist of native tree species and that are usually located adjacent to villages) are approximately 3 km from the research station. In 1984 (24 years before this study was conducted), experimental plantations of native and introduced species were established on degraded hilly-land at the research station to restore the degraded ecosystem. Topography, initial soil properties, and initial vegetation composition were similar across this hilly-land. At the same time, part of the degraded land was left unplanted and without further disturbance and had naturally succeeded to the shrubland stage when the current study was started. The dominant species in the degraded land before plantation establishment were

2391

Ischaemum indicum, Eriachne pallescens, and Baeckea frutescens. The experimental plantations and the shrubland consisted of five sites. The main established species at the EP site (1.79 ha) were Eucalyptus exserta and Eucalyptus citriodora, with a mean basal area of 154 cm2 . The main established species at the NP site (2.68 ha) were native species Schima superba and Cinnamomum burmanii, with a mean basal area of 201 cm2 . The main established species at the LP site (3.99 ha) were legume species Acacia mangium, Acacia auriculaeformis, Acacia confuse, and Acacia holosericea, with a mean basal area of 255 cm2 . The main established species at the CP site (3.17 ha) were coniferous species Pinus massoniana and Cunninghamia lanceolata, with a mean basal area of 227 cm2 . The SL site (3.5 ha) was a shrubland that had undergone 24 years of natural succession from the former degraded land and that was currently dominated by Ilex asprella, Evodia lepta, and Trema tomentosa, with a mean basal area of 28 cm2 . All trees in the plantations were planted at a 2.5 m × 2.5m spacing. All four plantations and the shrubland had been left to develop naturally without anthropogenic disturbance. 2.2. Experiment and sampling design To assess seed rain in each site, we established three transects (10 m × 40 m) at each site; these transects were located on upper slope, middle slope, and basal slope of each site, respectively. The slopes within each site were homogeneous, i.e., with the topsoil and inclination similar. Each transect was divided into four quadrants (10 m × 10 m). One seed trap (0.5 m × 0.5 m = 0.25 m2 trapping area) was established in the middle of each quadrant; thus, adjacent seed traps in a transect were 10 m apart, and there were 12 seed traps per site. Seed traps were positioned in the field in December 2007. A seed trap consisted of plastic mesh (0.5-mm mesh) whose shape was maintained by the attachment of PVC tubes to the four corners. The seed trap was 10 cm deep, and its bottom was positioned 10 cm above the ground (Zou and Yang, 2005; Barbosa and Pizo, 2006). Once each month, from January 2008 through December 2008, all seeds, fruit and seed-bearing fruit fragments were collected from each trap, taken back to the laboratory, sorted by species, and then counted. To assess the seed bank, we collected soil samples from one plot (1 m × 1 m) adjacent to each seed trap. In March 2008, five soil samples (10 × 10 × 10 cm) with litter intact were randomly and carefully excavated from each plot. The soil samples were divided into three depths (0–2, 2–5, and 5–10 cm), and the five soil samples were pooled for each plot and depth class, as suggested by Bossuyt et al. (2002). The total sampling area at each site was 0.6 m2 . Seed abundance and species composition in the seed banks were determined with germination assays, which were performed as described by Ter Herdt et al. (1996). Each soil sample was passed through a 2-mm sieve to remove coarse debris. Seeds with diameter >2 mm were retrieved and returned to the soil samples. Each soil sample was spread on a seed germination tray with perforations on the bottom for drainage. Before the soil samples were spread, the bottom of each germination tray was covered with a 2-cmthick layer of heat-sterilized (120 ◦ C for 10 h) sand to prevent the soil samples from becoming water-saturated. All germination trays were placed in an experimental greenhouse and watered daily to keep the soil moist. The mean temperature in the greenhouse was 28 ◦ C, and the mean relative humidity was 66%. Newly germinated seedlings that were identified at the species level were counted and then removed from the seed trays every 2–5 days. Unidentified seedlings were transplanted into additional germination trays for further growth until the species could be identified. After the newly germinated seedlings were identified and removed, the soil in each tray was thoroughly stirred to stimulate germination of remaining viable seeds (Smith et al., 2002). Twelve seed trays filled with sterilized sand only were kept under the same conditions as a control for

2392

J. Wang et al. / Forest Ecology and Management 259 (2010) 2390–2398

Table 1 Species and total number of seeds captured in the seed rain at five sites over 1 year. EP = eucalyptus plantation. NP = mixed-native species plantation. LP = mixed-legume plantation. CP = mixed-conifer plantation. SL = shrubland. For dispersal mode, Z = zoochory, A = anemochory, and B = ballistic. * indicates the species also exists in the aboveground vegetation. Species

Trees Eucalyptus citriodora Michelia macclurel Litsea glutinosa Schima Superba Cunninghamia lanceolata Shrubs and vines Leucaena leucocephala Mimosa sepiaria Rhus chinensis Litsea cubeba Lindera communis Psychotria rubra Ilex asprella Evodia lepta Trema tomentosa Breynia fruticosa Mallotus apelta Rhodomyrtus tomentosa Brassaia actinophylla Eurya chinensis Ardisia villosa Helicteres angustifolia Herbs Clerodendron fortunatum Polygonum chinense Lophatherum gracile Miscanthus sinensis Ottochloa nodosa Solanum torvum Unidentified 1 Unidentified 2 Unidentified 3 Unidentified 4

Family

Dispersal mode

Total number of seeds EP

NP

LP

CP

SL

Myrtaceae Magnoliaceae Lauraceae Theaceae Taxodiaceae

A Z Z A A

7640* 0 5* 14 0

0 3* 0 401* 0

0 0 0 0 0

0 0 0 0 79*

110* 0 0 0 0

Mimosaceae Mimosaceae Anacardiaceae Lauraceae Lauraceae Rubiaceae Aquifoliaceae Rutaceae Ulmaceae Euphorbiaceae Euphorbiaceae Myrtaceae Araliaceae Theaceae Myrsinaceae Sterculiaceae

B B Z Z Z Z A A A Z A Z Z Z Z B

0 0 7 15* 0 2* 36* 604* 0 0 0 33* 0 0 2* 0

0 0 0 0 0 0 29* 13* 0 0 0 0 0 0 0 0

61* 2* 0 29+ 0 0 2066* 133* 0 1 2* 0 0 0 10* 0

0 0 9* 2* 0 1* 25* 567* 1* 1 0 0 1 0 1* 0

0 0 0 0 1 6* 284* 337* 29* 0 0 0 0 4* 2* 178

Verbenaceae Polygonaceae Gramineae Gramineae Gramineae Solanaceae

A A A A A A

0 2 2 8 972* 0 2 0 0 0

1* 0 127* 14 850* 0 0 0 0 0

0 2* 0 32 2383* 0 0 1 1 1

0 0 58* 4* 4335* 0 18 0 0 0

0 0 118* 29 851* 268 0 0 0 0

seed contamination; no seedlings were found in these control trays. The germination assay continued for 7 months (March to October 2008) and was terminated when new seedlings had not emerged for 4 weeks (Wang et al., 2009). The regeneration potential of natural seedling emergence was determined in two 0.5 × 0.5-m plots near each seed trap. In December 2007, the existing understory vegetation and litter on the soil surface were removed from each of these plots, and the identity and number of the newly emerged seedlings were recorded monthly from January 2008 till December 2008. Each newly emerged seedling was marked with a plastic tag to facilitate recording. Because they are difficult to monitor, herbaceous plants in the Cyperaceae and Gramineae were not recorded. For comparisons of species composition between the aboveground vegetation and the seed rain or the seed bank, the species was identified and the coverage of the aboveground plants was estimated visually in each quadrant for each species of seed plant. These surveys were performed just after the sampling of soil for seed bank assessment. 2.3. Data analysis The number of seeds captured in each seed trap was converted to seed density (seeds per m2 ). Seed bank density (seedlings per m2 ) was calculated from the number of emerged seedlings. Quantitative and qualitative differences between community composition of the seed rain and seed bank were analyzed with one-way ANOVA tests followed by LSD tests when ANOVAs were significant. All data were log10 or arcsine square-root transformed when they did not satisfy normality assumptions. Statistical analyses were performed with

SPSS 11.5 for Windows (Li and Luo, 2005). The relationship between species composition in the aboveground vegetation and in the seed rain of each community was analyzed with the Sorensen similarity index (SI), SI = 2a/(b + c), where a refers to the number of species common to both the aboveground vegetation and the seed rain, and b and c represent total number of species detected in the seed rain and the corresponding aboveground vegetation, respectively (Cox, 1985). The SI index was also used to describe the relationship between the species composition in the aboveground community and in the seed bank. 3. Results 3.1. Seed rain A total of 22 805 seeds belonging to 31 species from 18 families were collected from all five sites during the 12-month period (Table 1). Seed rain density was significantly different among the five sites (df = 4, F = 5.276, P < 0.01), and ranged from 471 to 3111 seeds m−2 (Table 2). The total number of seed taxa collected was highest in EP and lowest in NP (Table 2), but the mean number of species detected per seed trap did not differ significantly among the five sites (df = 4, F = 1.612, P = 0.184). Ilex asprella, Evodia lepta, Miscanthus sinensis, and Ottochloa nodosa were the common species in the seed rain of all the sites. At all sites, seed rain was dominated by a few species. O. nodosa was dominant in the seed rain at the CP, LP, NP, and SL site, representing from 38.4 to 85.0% of the total seeds. E. citriodora was dominant in the seed rain at the EP site, accounting for 81.8% of total seeds. The seed rain contained a much greater proportion of herbaceous seeds than seeds of trees

J. Wang et al. / Forest Ecology and Management 259 (2010) 2390–2398 Table 2 Seed density (seeds m−2 ) in the seed rain, seedling density (seedlings m−2 ) in the seed bank, and total number of species detected at five sites. Values of seed/seedling density are means ± SE per plot, and the results from LSD test after one-way ANOVA. EP = eucalyptus plantation. NP = mixed-native species plantation. LP = mixed-legume plantation. CP = mixed-conifer plantation. SL = shrubland. Site

Seed rain Seed density

EP NP LP CP SL

3111 471 1677 1701 1093

± ± ± ± ±

669 a 137 c 569 bc 342 ab 362 bc

2393

stored seeds in the soil was uneven, and many species occurred only in low density, such as the tree species Litsea glutinosa (5 seeds m−2 ) at the NP site and Evodia meliaefolia (2 seeds m−2 ) at the CP site. Some exotic herbaceous species, like Ageratum conyzoides and Conyza bonariensis, were relatively abundant in the seed bank.

Seed bank Total number of species

Seedling density

15 8 14 14 13

2370 5097 4225 11460 1608

± ± ± ± ±

529 b 1795 b 677 b 2444 a 404 b

Total number of species 35 35 37 35 22

Seed/seedling densities marked with different letters (a, b, and c) indicate that they were significantly different from each site.

and shrubs at all sites except the EP site. The most common seed dispersal modes in our study were zoochory and anemochory, representing 88.9% of the known species detected in the seed rain. For the tree species, however, only Michelia macclurel and Litsea glutinosa had zoorchorous seeds. 3.2. Seed bank The germination assay detected a total of 14 862 seeds belonging to 52 species from 24 families (Table 3). Seed bank density at the CP site was 11 460 seeds m−2 and was much higher than that at other sites (P < 0.01). For all sites, seed abundance gradually decreased with increasing soil depth (Fig. 1). The number of viable seeds was significantly higher in the 0-2 cm layer than in the 5–10 cm layer at all sites except the LP site (Fig. 1). The total number of species in the seed bank of each site ranged between 22 and 37, with the lowest in the SL site (Table 2). The mean number of species per plot was significantly lower at the SL site than at the EP, LP, and CP site (P < 0.05). Thirteen species cooccurred in the seed banks of all sites. Seeds of woody species made up only a minor component of the seed bank; the seed banks were dominated by herbs in term of species richness and seed abundance (Table 3). The most abundant species detected in the seed bank was the herbaceous species O. nodosa, which accounted for 70.5, 93.0, 73.5, 92.1, and 72.0% of the seeds detected in the soil at the EP, NP, LP, CP, and SL site, respectively. The distribution of

3.3. Species composition of the aboveground vegetation vs. that in the seed rain or the seed bank The total number of seed plant species recorded in the aboveground vegetation ranged from 24 to 36 (Table 4). The CP site contained more understory species than the other sites (data not shown). Herb coverage was relatively high in the understory at the CP and NP sites (data not shown). The dominant species in the understory were largely similar among the five sites (data not shown). The total number of species in the seed rain was much less than that in the aboveground vegetation at all sites (Tables 2 and 4). Moreover, from 60 to 88% of the species in the seed rain also existed in the aboveground vegetation. The Sorensen similarity index for the five sites ranged from 0.35 to 0.44, with the lowest and highest value in the EP and NP site, respectively (Table 4). Fewer species were detected in the seed bank than in the aboveground vegetation at sites EP, and SL but not at sites NP, CP or LP (Tables 2 and 4). The Sorensen similarity index ranged from 0.29 to 0.46, indicating that composition similarity between the seed bank and aboveground vegetation was low (Table 4). 3.4. Seedling emergence and survival A total of 928 seedlings belonging to 29 species from 18 families naturally emerged at all sites (Table 5). Among the seedlings recorded in the field, three species were trees, 20 species were shrubs, three species were herbs, and three species were not identified. Emerged seedling density did not differ significantly among the sites (Table 6; df = 4, F = 0.748, P = 0.564). The number of species that emerged in the field also did not differ significantly among the five sites (df = 4, F = 0.251, P = 0.908). Five common species, Mussaenda pubescens, E. lepta, T. tomentosa, I. asprella, and Melastoma dodecandrum, were recorded at all sites. Seedlings of tree species were scarce, and only a total of four seedlings were recorded. Many other species besides the trees had low seedling densities in the field (Table 5). Two shrubs, E. lepta and T. tomentosa, were dominant at the EP, LP, CP and SL sites, accounting for 81.4, 63.9, 79.4 and 83.4% of the total emerged seedlings, respectively. Most seedling emergence occurred between April and May (Fig. 2). In total, 640 seedlings died at all sites (i.e., 69.0% mortality) during the 1-year period (Table 5). The density of dead seedlings did not differ significantly among the five sites (df = 4, F = 0.932, P = 0.452). Consistent with the pattern of dominant species in seedling emergence, E. lepta and T. tomentosa experienced 86.4, 65.1, 82.8 and 85.6% seedling mortality at the EP, LP, CP and SL sites, respectively. The final number of surviving species was smaller than the number of species that emerged because some species, such as S. Superba and Psychotria rubra at the NP and CP site, respectively, were not represented at the end of the 1-year study. Most seedling mortality occurred between May and July (Fig. 2). 4. Discussion 4.1. Seed rain

Fig. 1. Vertical distribution of seeds (mean ± SE) in the soil seed bank of the five sites. Different letters above bars indicate significant differences between soil layers at the same site. Mean separation was based on a one-way ANOVA followed by an LSD test. EP = eucalyptus plantation. NP = mixed-native species plantation. LP = mixedlegume plantation. CP = mixed-conifer plantation. SL = shrubland.

In our study of four plantations and a shrubland in southern China, seed inputs were dominated by herbaceous species. Such results are in line with previous findings that species producing

2394

J. Wang et al. / Forest Ecology and Management 259 (2010) 2390–2398

Table 3 Species and numbers of seedlings that germinated from the seed banks of five sites. Each value indicates the number of seedlings emerged from 0.06 m3 of soil per site. EP = eucalyptus plantation. NP = mixed-native species plantation. LP = mixed-legume plantation. CP = mixed-conifer plantation. SL = shrubland. * represents the species also exists in the aboveground vegetation. Species

Family

Number of seedlings emerged EP

Trees Litsea glutinosa Evodia meliaefolia Eucalyptus citriodora Shrubs Litsea cubeba Rhus chinensis Mallotus apelta Ficus hirta Ficus erecta Baeckea frutescens Rhodomyrtus tomentosa Eurya chinensis Hedyotis hedyotidea Mussaenda pubescens Evodia lepta Trema tomentosa Melastoma candidum Leucaena leucocephala Rubus aculeatiflorus Herbs Hedyotis diffusa Hedyotis acutangula Jussiaea linifolia Oxalis corniculata Elephantopus scaber Eupatorium catarium Ageratum conyzoides Gnaphalium affine Conyza bonariensis Emilia sonchifolia Eupatorium japonicum Melastoma dodecandrum Fimbristylis dichotoma Cyperus haspan Fimbristylis miliacea Cyrtococcum patens Digitaria ciliaris Paspalum conjugatum Eleusine indica Ischaelnum indicum Ottochloa nodosa Lophatherum gracile Torenia biniflora Torenia flava Adenosma glutinosum Lindernia crustacea Haloragis chinensis Halorrhagis micrantha Polygonum chinense Centella asiatica Juncus prismatocarpus Solanum nigrum Dianella ensifolia Mitrasacme pygmaea

Lauraceae Rutaceae Myrtaceae

NP

LP

CP

SL

0 0 3*

3 0 0

0 0 0

0 1 0

0 0 0

Lauraceae Anacardiaceae Euphorbiaceae Moraceae Moraceae Myrtaceae Myrtaceae Theaceae Rubiaceae Rubiaceae Rutaceae Ulmaceae Melastomataceae Mimosaceae Rosaceae

5* 5* 0 2* 0 6 5 1* 81* 30* 32* 733* 29* 0 1

1* 0 0 1 1* 3 6* 0 3* 25* 6* 17* 4* 0 0

0 1 1 1* 0 5 2 5* 183* 63* 16* 205* 18* 6* 2

0 0 0 2* 0 3 2* 2* 72 37* 23* 157* 57* 0 4

2* 1 0 2* 0 13 11* 1* 0 10* 11* 110* 22* 0 0

Rubiaceae Rubiaceae Onagraceae Oxalidaceae Compositae Compositae Compositae Compositae Compositae Compositae Compositae Melastomataceae Cyperaceae Cyperaceae Cyperaceae Gramineae Gramineae Gramineae Gramineae Gramineae Gramineae Gramineae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Haloragidaceae Haloragidaceae Polygonaceae Umbelliferae Juncaceae Solanaceae Liliaceae Loganiaceae

0 0 3 1 0 0 37 1 4 1 3 45* 6 1 0 1 0 3 2 6 1002* 0 0 4 12 5 7 2 0 2 0 1 0 0

1 1 0 0 1 1 5 2 6 1 1 60 6 0 0 4 1 3 0 10 2844* 5* 0 0 23 1 8 1 1 0 1 0 0 1

1 0 0 9 0 0 24 0 5 2 1 14 9 2 0 21 2 0 0 1 1863* 1* 7 27 9 8 3 3 2 13 0 1 2 0

0 0 0 0 1 1 1 0 19 3 4 103* 13 1 1 6* 0 1 2 3 6333* 1* 4 1 0 7 7 3 1* 2 0 0 0 0

0 1* 1 5 0 0 2 0 0 0 0 14* 43 0 0 1 0 5 0 14 695* 1* 0 0 0 0 0 0 0 0 1 0 0 0

Table 4 Species richness and similarity of the aboveground vegetation vs. the seed rain or seed bank at five sites. EP = eucalyptus plantation. NP = mixed-native species plantation. LP = mixed-legume plantation. CP = mixed-conifer plantation. SL = shrubland. Site

EP NP LP CP SL

Number of species in the aboveground vegetation

36 24 31 35 30

Seed rain Number of co-occurring species 9 7 9 10 9

Seed bank Sorensen similarity index

Number of co-occurring species

Sorensen similarity index

0.35 0.44 0.40 0.41 0.42

13 10 10 12 12

0.37 0.34 0.29 0.34 0.46

J. Wang et al. / Forest Ecology and Management 259 (2010) 2390–2398

2395

Table 5 Species and numbers of seedlings that naturally emerged (E) and survived (S) in a 1-year period in the five sites. Values for emergence and survival are totals from twentyfour 0.25-m2 plots per site. EP = eucalyptus plantation. NP = mixed-native species plantation. LP = mixed-legume plantation. CP = mixed-conifer plantation. SL = shrubland. * represents the species is exotic. Species

Family

No. E/No. S EP

Trees Schima superba Syzygium hancei Sapium sebiferum Shrubs Litsea cubeba Leucaena leucocephala* Psychotria rubra Mussaenda pubescens Hedyotis hedyotidea Evodia lepta Trema tomentosa Rhus chinensis Aralia chinensis Jasminum elongatum Eurya chinensis Glochidion eriocarpum Breynia fruticosa Mallotus apelta Clerodendron fortunatum Ilex asprella Wikstroemia indica Rhodomyrtus tomentosa Melastoma candidum Rubus aculeatiflorus Herbs Melastoma dodecandrum Torenia flava Polygonum chinense Unidentified 1 Unidentified 2 Unidentified 3

Theaceae Myrtaceae Euphorbiaceae Lauraceae Mimosaceae Rubiaceae Rubiaceae Rubiaceae Rutaceae Ulmaceae Anacardiaceae Araliaceae Oleaceae Theaceae Euphorbiaceae Euphorbiaceae Euphorbiaceae Verbenaceae Aquifoliaceae Thymelaeaceae Myrtaceae Melastomataceae Rosaceae Melastomataceae Scrophulariaceae Polygonaceae

NP

LP

CP

SL

1/0 2/1 1/1 3/1

3/3

7/5 6/5 117/64 62/7

17/4 4/3 1/0 32/6 6/0

2/1 29/4 4/2 5/1 6/4 37/17 94/6

2/0 17/8 4/2 62/42 88/7 2/2

1/1 11/4 49/26 127/19 0 0 0 4/0

1/1 2/0 4/0

5/2 13/6 1/1 2/2 4/2 2/2

1/0 3/0 1/1 3/0

11/2 6/1 2/1

2/1 1/1 1/0 2/0

1/0

1/0 1/1 1/0 1/1

2/0 1/0 1/1

large quantities of small seeds are likely to reach higher concentrations in the seed rain than are species producing small numbers of large seeds (Willson et al., 1990; Holzel and Otte, 2004). A great proportion of the seeds and species in the seed rain corresponded with the species present in the aboveground vegetation, indicating that seeds of most species in the seed rain were derived from the local plant community. However, the seed rain of these plantations plays a minor role in determining the plant community composition. In our study, some species in the seed rain were absent from the aboveground vegetation. Despite most of the seed rain derived from the local vegetation, seed dispersed by birds or wind from outer plant community is also with potential (Au et al., 2006). However, the immigrated species may not able to incorporated into current plantations. After seeds arrive from other communities, many biotic and abiotic factors such as seed predation, competition and soil moisture deficiency may lead to seed or seedling mortality (Booth and Larson, 1998; Lenz and Facelli, 2005; Dupuy and Chazdon, 2008). Meanwhile, seed dispersal varies from year to year (Grubb, 1977; Tackenberg and Stocklin, 2008), and the migration rate of a species relates to the number of seeds dispersed (Fuller and Moral, 2003).

2/1 3/2 8/0 1/0

1/1 2/0 3/3 1/1 2/0

6/2 1/1 2/1 1/0

1/1

6/2

2/1 1/0 1/0

1/0

Although seed rain density differed among the five sites, species richness and the dominant species were similar among the sites. The seed rain in the current study, however, was represented by a low proportion of indigenous tree species and a lack of typical mature-forest species, indicating ineffective seed dispersal from more natural plantations to the plantations and shrubland. Forest succession has been reported to depend on the distance to remnants of indigenous forests, which act as sources of propagules, and on the characteristics of the seeds of the desired species (McClanahan, 1983; Fuller and Moral, 2003; Zamora and Montagnini, 2007). Both these factors could explain the small numbers of late-successional tree species detected in the seed rain of the current study. 4.2. Seed bank Seed bank composition is the result of the different life strategies of individual species (Olano et al., 2002). In line with seed bank composition of other young regrowth stands (Dalling and Denslow, 1998), the seed bank composition in all five sites of the current study was dominated by herbaceous species, while seeds of woody

Table 6 Total number of species and densities (seedlings m−2 ) of emerged (E) and dead (D) seedlings in a 1-year period in the five sites. Values of seedling densities are means ± SE of 24 0.25-m2 plots per site. EP = eucalyptus plantation. NP = mixed-native species plantation. LP = mixed-legume plantation. CP = mixed-conifer plantation. SL = shrubland. Site

EP NP LP CP SL

E

D

Total number of species

Seedling density

18 17 17 15 13

36.7 17.2 34.2 31.8 35.5

± ± ± ± ±

14.8 2.6 7.8 6.6 9.8

Total number of species

Seedling density

4 5 7 5 3

20.7 12.0 27.7 20.3 25.5

± ± ± ± ±

8.9 1.9 6.8 4.3 7.0

2396

J. Wang et al. / Forest Ecology and Management 259 (2010) 2390–2398

Fig. 2. Seasonal variation in the numbers of naturally emerged and dead seedlings in the five sites. Each value is the mean of 24 0.25-m2 plots per site. EP = eucalyptus plantation. NP = mixed-native species plantation. LP = mixed-legume plantation. CP = mixed-conifer plantation. SL = shrubland.

species were scarce. This result may be due to the relatively long persistence of herbaceous seeds in the soil because seed mass is negatively correlated with seed longevity (Holzel and Otte, 2004). The scarcity of tree and shrub seeds in the seed bank may be due to poor seed dispersal, as indicated by the absence of tree and shrub species in the seed rain. The scarcity of woody species in the seed bank also may be attributed to the rapid deterioration and loss of germinability of tree seeds in soil (Sester et al., 2006). Numerous studies have shown that seed bank composition changes with soil depth (Olano et al., 2002; Luzuriaga et al., 2005). In our study, species richness and seed abundance were greatest in the upper soil layer at all sites. According to Thompson and Grime (1979), the seed banks under the established plantations and the shrubland are transient. Although aboveground vegetation can greatly influence seed bank composition (Olano et al., 2002), the similarity in species composition between the seed bank and the aboveground vegetation was low in the current study, i.e., many species in the aboveground vegetation did not occur in the soil seed bank and vice versa. The absence of some species in the aboveground vegetation may result from unsuitable conditions for seed germination (Russell-Smith and Lucas, 1994). The lack of sufficient light under those plantations would inhibit the germination of shade-intolerant species such as Halorrhagis micrantha and Ischaelnum indicum.

Although the abundance of the seeds stored in soil was much higher in the mixed-conifer plantation than in the four other sites, many species co-occurred in the seed banks of the five sites, and the dominant species were similar in the seed banks of the four plantations and the shrubland. The seed banks contained very few seeds of indigenous tree species. Hence, we conclude that there is no important difference in the seed bank composition among the five plant communities. The seed bank appears to be structured as a consequence of contrasting driving forces such as canopy structure, understory composition, and microhabitat features (Olano et al., 2002). The plantations and the shrubland in our study were established on or naturally succeeded from a similar initial condition of degraded hilly-land. With decades of development, the dominant understory species are largely similar (Duan et al., 2008). Additionally and as discussed earlier, seed dispersal of indigenous tree species into the planted forests is ineffective. These factors may cause the five seed banks to have similar compositions. 4.3. Relationship between the seed rain, seed bank, and seedling emergence The quantitative and qualitative differences between the seed rain and the seed bank were observed in the current study, which is

J. Wang et al. / Forest Ecology and Management 259 (2010) 2390–2398

consistent with previous findings (Pakeman and Small, 2005). The low correlation between seed density of the seed rain and the seed bank is not surprising because the seed bank reflects the accumulation of viable seeds from the seed rain of multiple former fruiting seasons, while the seed rain was observed only for a 1-year period. Long-lived seeds of some species in the soil would lead to the quantitative differences in seed density between the seed bank and seed rain. For example, the number of herbaceous species was much higher in the seed bank than in the seed rain. Annual variation of seed inputs (Arrieta and Suarez, 2005) and the predation of seed after seed input (Hubbard and McPherson, 1999) will also affect the size of the seed bank. Some species, such as the overstory S. superba at the NP site and C. lanceolata at the CP site, were represented in the seed rain but not in the seed bank. Such findings indicate that seeds of these planted species do not enter the soil or they do not persist after entering the soil. For these kinds of plantation species, natural recruitment will depend more on the seed rain than on the seed bank. Many of species that were represented in the seed rain and in the seed bank did not emerge in the field plots during the 1-year period of this study. This indicates that in addition to a seed supply, a suitable site for seedling establishment is also critical for forest regeneration. In our study, most species, e.g., Conyza bonariensis and Lindernia crustacea, that emerged in the greenhouse but not in the field plots had tiny seeds. Small seeds are vulnerable to drought and other extreme weather conditions (Bakker et al., 1996). Another possible explanation for the larger number of seedlings that emerged from the soil samples in the greenhouse than in the field plots is that a relatively thick layer of soil (10-cm depth) was sampled for seed bank detection, and the disturbance and spreading of this soil in the germination assay could have increased the chances of germination in the greenhouse. Seedling emergence depends on many environmental factors including soil moisture availability (McLaren and McDonald, 2003). Soil moisture is directly influenced by rainfall, which in southern China is highly seasonal and occurs mostly from April to September. In agreement with this pattern, many seedlings emerged in the field plots following the spring rain in our study. With the desiccation in the dry season, the ungerminated seeds originating from the seed rain or the seed bank are unlikely to germinate (Veenendaal et al., 1996; Lee et al., 2004). Hence, seedling emergence in the four plantations and the shrubland gradually decreased after the peak in spring. In accordance with our results, Ellsworth et al. (2004) also found that seeds in the seed bank that failed to germinate in the spring did not remain viable. High irradiance during the summer and pathogens attack may also increase seedling mortality (Ellsworth et al., 2004; Mori and Mizumachi, 2005). Our results demonstrate that both the seed bank and the seed rain play important roles in providing seeds for plant recruitment. Some species that emerged in the field (such as I. asprella, the dominant shrub in the understory of the plantations) only occurred in the seed rain. Thus, the current year’s seed inputs are primary sources for seedling recruitment of species that with relatively short-lived seeds in the soil following dispersal. The number of species common to the seed rain and aboveground vegetation was lower than the number common to the seed bank and aboveground vegetation, and the seed rain contained relatively few species. Such results suggest that seeds already present in the seed bank contribute more to the diversity of the recruited plants than seeds provided by the seed rain, which is in contrast to previous findings (Pakeman and Small, 2005). In our study, the contribution of anemochorous species to seedling emergence was much more than zoochorous species. Such emergency pattern may be owing to the high seed abundance of anemochorous species observed in the seed rain and seed bank.

2397

4.4. Restoration implications In the established plantations and shrubland, shrubs and herbs currently dominate the seed supply. It follows that these species are more likely to dominate the understory of the plantations for several years and that this will retard the succession of current plantations to more natural stages. Indigenous tree species rarely disperse from outside communities and do not accumulate in the soil seed bank, indicating that the seed rain and seed bank are poor sources for the indigenous tree species associated with natural succession. In our study, the source of many species in the seed rain was the local plant community, and zoochorous seeds in tree species were scarce. This suggests either the absence of nearby sources of zoochorous seeds in tree species or the absence of animal dispersal agents (Rico-Gray and Garcia-Franco, 1992). Previous research has indicated that forest fragments dominated by anemochorous plants are usually unattractive to vertebrate seed dispersers, thus potentially diminishing the arrival of zoochorous seeds (Barbosa and Pizo, 2006; Zamora and Montagnini, 2007). The established plantation species that we studied (A. mangium, A. auriculaeformis, E. exserta, and S. superba) are anemochorous, and the seeds of these trees may not attract zoochoric dispersal agents. Such speculation is supported by previous findings that these planted forests contain low numbers and low species diversity of understory birds (Zou and Yang, 2005). From the perspective of seed dispersal, the selection of plantation species with fruit or seeds that attract birds should be considered during plantation improvement. As important sources of propagules, remnant mature forests can drive the rate and trajectories of primary succession in nearby degraded plant communities (Fuller and Moral, 2003). With the expansion of agricultural land over the last several decades in southern China, large areas of original forests were deforested, and undisturbed natural forests are scarce and patchily distributed. Under the circumstance of habitat fragmentation, seed dispersal by animals is frequently hampered (Garcia and Banuelos, 2003; Bertoncini and Rodrigues, 2008). In our studied areas, the low occurrence in the seed rain of large-seeded indigenous species characterized by zoochoric dispersal and the slow succession of the plantations and the shrubland. From this perspective, remnant mature forests should be conserved and plantations should be established very close to such mature forests to facilitate succession. In conclusion, the planted forests in this study have developed and now contain many more species than they did at the time of plantation establishment. These plantations, however, do not contain desired mature forest species. Lack of availability of seeds of these desired zonal mature forest species is currently the main bottleneck in the succession of these plantations. Reintroduction of late-successional species by direct seeding or transplanting should be considered to facilitate the development of more natural vegetation. Even when seeds or seedlings are planted, however, they can fail to establish because of numerous abiotic and biotic conditions such as moisture deficit, seed predation, and competition (Holl et al., 2000; Camargo et al., 2002). Therefore, the successful establishment of indigenous tree species in the plantations after artificial seeding or transplanting will require further insights into factors that limit seedling establishment.

Acknowledgements This study was supported by the National Natural Science Foundation of China (30670370, 40871249), and Guangdong Program (07118249, 2008A060207017). The authors are indebted to Prof Bruce Jaffee for English polishing, Mr. Yongbiao Lin and

2398

J. Wang et al. / Forest Ecology and Management 259 (2010) 2390–2398

Mr. Xingquan Rao for field assistance. We also thank anonymous reviewers for their valuable comments on the early version of the manuscript. References Alvarez-Buylla, E.R., Martinez-Ramos, M., 1990. Seed bank versus seed rain in the regeneration of a tropical pioneer tree. Oecologia 84, 314–325. Arrieta, S., Suarez, F., 2005. Spatial dynamics of Ilex aquifolium populations seed dispersal and seed bank: understanding the first steps of regeneration. Plant Ecology 177, 237–248. Au, A.Y.Y., Corlett, R.T., Hau, B.C.H., 2006. Seed rain into upland plant communities in Hong Kong, China. Plant Ecology 186, 13–22. Augusto, L., Dupouey, J.L., Picard, J.F., Ranger, J., 2001. Potential contribution of the seed bank in coniferous plantations to the restoration of native deciduous forest vegetation. Acta Oecologica 22, 87–98. Auld, T.D., Denham, A.J., Turner, K., 2007. Dispersal and recruitment dynamics in the fleshy-fruited Persoonia lanceolata (Proteaceae). Journal of Vegetation Science 18, 903–910. Bakker, J.P., Bakker, E.S., Rosen, E., Verweij, G.L., Bekker, R.M., 1996. Soil seed bank composition along a gradient from dry alvar grassland to Juniperus shrubland. Journal of Vegetation Science 7, 165–176. Barbosa, K.C., Pizo, M.A., 2006. Seed rain and seed limitation in a planted gallery forest in Brazil. Restoration Ecology 14, 504–515. Bertoncini, A.P., Rodrigues, R.R., 2008. Forest restoration in an indigenous land considering a forest remnant influence (Avai, Sao Paulo State, Brazil). Forest Ecology and Management 255, 513–521. Booth, B.D., Larson, D.V., 1998. The role of seed rain in determining the assembly of a cliff community. Journal of Vegetation Science 9, 657–668. Bossuyt, B., Hermy, M., 2004. Seed bank assembly follows vegetation succession in dune slacks. Journal of Vegetation Science 15, 449–456. Bossuyt, B., Heyn, M., Hermy, M., 2002. Seed bank and vegetation composition of forest stands of varying age in central Belgium: consequences for regeneration of ancient forest vegetation. Plant Ecology 162, 33–48. Camargo, J.L.C., Ferraz, I.D.K., Imakawa, A.M., 2002. Rehabilitation of degraded areas of Central Amazonia using direct sowing of forest tree seeds. Restoration Ecology 10, 636–644. Cox, G.W., 1985. Laboratory Manual of General Ecology, 7th ed. McGraw-Hill, New York, p. 248. Dalling, J.W., Denslow, J.S., 1998. Changes in soil seed bank composition along a chronosequence of lowland secondary tropical forest, Panama. Journal of Vegetation Science 9, 669–678. Denslow, J.S., Uowolo, A.L., Hughes, R.F., 2006. Limitations to seedling establishment in a mesic Hawaiian forest. Oecologia 148, 118–128. Duan, W., Ren, H., Fu, S., Wang, J., Zhang, J., Yang, L., Huang, C., 2008. Community comparison and determinant analysis of understory vegetation in six plantations in South China. Restoration Ecology, doi:10.111/j.1526-100X.2008.00444.x. Dupuy, J.M., Chazdon, R.L., 2008. Interacting effects of canopy gap, understory vegetation and leaf litter on tree seedling recruitment and composition in tropical secondary forests. Forest Ecology and Management 255, 3716–3725. Ellsworth, J.W., Harrington, R.A., Fownes, J.H., 2004. Seedling emergence, growth, and allocaion of Oriental bittersweet: effects of seed input, seed bank, and forest floor litter. Forest Ecology and Management 190, 255–264. Fuller, R.N., Moral, R.D., 2003. The role of regugia and dispersal in primary succession on Mount St. Helens, Washington. Journal of Vegetation Science 14, 637–644. Garcia, D., Banuelos, M.J., 2003. Matrix matters for seed dispersal—a comment to Jules and Shahani. Journal of Vegetation Science 14, 931. Grubb, P.J., 1977. The maintenance of species-richness in plant communities: the importance of the regeneration niche. Biological Reviews 52, 107–145. Harper, J.L., 1977. Population Biology of Plants. Academic Press, London, p. 892. Holl, K.D., Loik, M.E., Lin, E.H.V., 2000. Tropical montane forest restoration in Costa Rica: overcoming barriers to dispersal and establishment. Restoration Ecology 8, 339–349. Holzel, N., Otte, A., 2004. Assessing soil seed bank persistence in flood-meadows: the search for reliable traits. Journal of Vegetation Science 15, 93–100. Howlett, B.E., Davidson, D.W., 2003. Effects of seed availability, site conditions, and herbivory on pioneer recruitment after logging in Sabah, Malaysia. Forest Ecology and Management 184, 369–383. Hubbard, J.A., McPherson, G.R., 1999. Do seed predation and dispersal limit downslope movement of a semi-desert grassland/oak woodland restoration. Journal of Vegetation Science 10, 739–744.

Lee, C.S., Kim, J.H., Yi, H., You, Y.H., 2004. Seedling establishment and regeneration of Korean red pine (Pinus densiflora S. et Z.) forests in Korea in relation to soil moisture. Forest Ecology and Management 199, 423–432. Lentink, D., Dickson, W., Leeuwen, J.V., Dickinson, M., 2009. Leading-edge vortices elevate lift of autorotating plant seeds. Science 324, 1438–1440. Lenz, T.I., Facelli, J.M., 2005. The role of seed limitation and resource availability in the recruitment of native perennial grasses and exotics in a South Australian grassland. Austral Ecology 30, 684–694. Li, Z.H., Luo, P., 2005. SPSS for Windows. Publishing House of Electronics Industry, Beijing, p. 683 (in Chinese). Luzuriaga, A.L., Escudero, A., Olano, J.M., Loidi, J., 2005. Regenerative role of seed banks following an intense soil disturbance. Acta Oecologica 27, 57–66. Maranon, T., 1998. Soil seed bank and community dynamics in an annual-dominated Mediterranean salt-marsh. Journal of Vegetation Science 9, 371–378. McClanahan, T.R., 1983. The effect of a seed source on primary succession in a forest ecosystem. Vegetatio 65, 175–178. McLaren, K.P., McDonald, M.A., 2003. The effects of moisture and shade on seed germination and seedling survival in a tropical dry forest in Jamaica. Forest Ecology and Management 183, 61–75. Mori, A., Mizumachi, E., 2005. Season and substrate effects on the first-year establishment of current-year seedlings of major conifer species in an old-growth subalpine forest in central Japan. Forest Ecology and Management 210, 461–467. Olano, J.M., Caballero, I., Laskurain, N.A., Loidi, J., Escudero, A., 2002. Seed bank spatial pattern in a temperate secondary forest. Journal of Vegetation Science 13, 775–784. Pakeman, R.J., Small, J.L., 2005. The role of the seed bank, seed rain and the timing of disturbance in gap regeneration. Journal of Vegetation Science 16, 121–130. Peng, S.L., 2003. Study and Application of Restoration Ecology in Tropical and Subtropical China. Science Press, Beijing, p. 506 (in Chinese). Ren, H., Huang, P., Zhang, Q.M., Hou, C.M., 2002. Forest Resources and its Ecosystem Service in Guangdong. Chinese Environmental Science Press, Beijing, p. 118 (in Chinese). Ren, H., Shen, W.J., Lu, H.F., Wen, X.Y., Jian, S.G., 2007. Degraded ecosystems in China: status, causes, and restoration efforts. Landscape and Ecological Engineering 3, 1–13. Rico-Gray, V., Garcia-Franco, J.G., 1992. Vegetation and soil seed bank of successional stages in tropical lowland deciduous forest. Journal of Vegetation Science 3, 617–624. Russell-Smith, J., Lucas, D.E., 1994. Regeneration of monsoon rainforest in northern Australia: the dormant seed bank. Journal of Vegetation Science 5, 161– 168. Sester, M., Durr, C., Darmency, H., Colbach, N., 2006. Evolution of weed beet (Beta vulgaris L.) seed bank: quantification of seed survival, dormancy, germination and pre-emergence growth. European Journal of Agronomy 24, 19–25. Shono, K., Davies, S.J., Kheng, C.Y., 2006. Regeneration of native plant species in restored forests on degraded lands in Singapore. Forest Ecology and Management 237, 574–582. Smith, R.S., Shiel, R.S., Millward, D., Corkhill, P., Sanderson, R.A., 2002. Soil seed banks and the effects of meadow management on vegetation change in a 10-year meadow field trial. Journal of Applied Ecology 39, 279–293. Tackenberg, O., Stocklin, J., 2008. Wind dispersal of alpine plant species: a comparison with lowland species. Journal of Vegetation Science 19, 109–118. Ter Herdt, G.N.J., Verweij, G.L., Bekker, R.M., Bakker, J.P., 1996. An improved method for seed bank analysis seedling emergence after removing the soil by sieving. Functional Ecology 10, 144–151. Thompson, K., Grime, J.P., 1979. Seasonal variation in the seed bank of herbaceous species in ten contrasting habitats. Journal of Ecology 67, 893–921. Urbanska, K.M., Erdt, S., Fattorini, M., 1998. Seed rain in natural grassland and adjacent ski run in the Swiss Alps: as preliminary report. Restoration Ecology 6, 159–165. Veenendaal, E.M., Ernst, W.H., Modise, G.S., 1996. Effect of seasonal rainfall pattern on seedling emergence and establishment of grasses in a savannah in southeastern Botswana. Journal of Arid Environments 32, 305–317. Wang, J., Zou, C., Ren, H., Duan, W.J., 2009. Absence of tree seeds impedes shrubland succession in southern China. Journal of Tropical Forest Science 21, 210–217. Willson, M.F., Rice, B.L., Westoby, M., 1990. Seed dispersal spectra: a comparison of temperate plant communities. Journal of Vegetation Science 1, 547–562. Zamora, C.O., Montagnini, F., 2007. Seed rain and seed dispersal agents in pure and mixed plantations of native trees and abandoned pastures at La Selva biological station, Costa Rica. Restoration Ecology 15, 453–461. Zou, F.S., Yang, Q.F., 2005. A study of understory bird communities in an artificial forest in Heshan, Guangdong, China. Acta Ecologica Sinica 12, 3323–3328 (in Chinese with English abstract).