TROPICS  Vol. 23 (2)  83︲90 Issued September 1, 2014
ORIGINAL ARTICLE
Secondary succession of mixed plantations established to rehabilitate 
abandoned pasture in the Peruvian Amazon
Shigeo Kobayashi1*, Manuel Soudre2 and Arberto Ricse2
1  Graduate school of Asian and African Area Studies, Kyoto University
2  Instituto Nacional de Innovación Agraria (INIA), Peru
* Corresponding author : skobayashi@asafas.kyoto-u.ac.jp
ABSTRACT　　Secondary succession or facilitation processes carried out at sites established for rehabilitating abandoned 
pastures and degraded forests (prurmas) are instrumental in their return to original forest status. An understanding of these 
secondary succession processes contributes to the rehabilitation of degraded forest ecosystems and to the livelihoods of local 
communities, and aids in conserving biodiversity. We studied secondary succession in mixed species plantations that were 
established to rehabilitate abandoned land. The initial vegetation in these abandoned pastures and croplands was grassland 
composed of three dominant species: Rottboellia exaltata, Imperata brasiliensis, and Brachyaria decumbens. After tree planting 
and weeding had been carried out, the site was first invaded by R. exaltata and Baccharis floribunda. These two species, which 
depend solely on sexual and not vegetative reproduction, facilitated secondary succession and elevated species diversity by 
enabling subsequent invasion by several species. By contrast, B. decumbens, I. brasiliensis, and Hyparrhenia rufa depend mainly 
on vegetative reproduction involving rhizomes and tillers, and subsequent invasion by other species was relatively less in stand 
types dominated by these three species. We found that further adequate rehabilitation techniques were necessary for the respective 
vegetation types.
Key words: abandoned pasture, secondary succession, grasslands, rehabilitation, Peru
INTRODUCTION 2004). Among the groups of species that appear in 
secondary succession, pioneer species play key roles in the 
　　Anthropogenic impacts and land-use conflicts between land dynamics and recovery process of degraded forests 
forests and agricultural land, caused by population increase, (Kobayashi et al., 1995). In this study, we seek to answer 
are accelerating the degradation of forest ecosystems. Every the following questions. (1) Do rehabilitated forests provide 
year, 16.1 million hectares (ha) of the world＇s primary timber and forest ecosystem services? (2) Can we expect 
forests are disappearing (FAO, 2001). Specifically, 15.2 newly established rehabilitated forests to return to their 
million ha of this annual loss of tropical forests is resulted original forest status? (3) Should an original forest stand be 
from 14.2 ha of land use changes and 1.0 million ha of the final goal for rehabilitating degraded forest ecosystems? 
deforestation. These degraded forests generally fail to To answer these questions, we surveyed initial vegetation 
regenerate naturally, because adequate rehabilitation recovery and classified undergrowth vegetation types to 
techniques are not applied, especially in tropical forests predict changes in several types of abandoned pasture and 
(Nihon Ringyo Chousakai 1989, ITTO, 1991), necessitating secondary forest. We investigated the influences and roles 
silvicultural treatment for their recovery. This is an issue of vegetation types on secondary succession affected by 
requiring attention, because rehabilitated forestland could anthropogenic activities, i.e., facilitation or competition 
potentially contribute to the sustainable use of forest processes relating to environmental changes (Holmgren et 
resources, conservation of primary tropical forests, and al., 1997; Li and Wilson, 1998), their interactions (Callaway 
ecosystem services improvement. The process of secondary and Walker, 1997), the allocation pattern of photosynthesis 
succession provides important information for the products of each indicated species (Kawano 1974), and 
rehabilitation of degraded tropical forests and land. In biodiversity of each forest (Kobayashi 1984a). The aims of 
particular, the availability of information on natural this study were to: (1) investigate dynamic processes of 
secondary succession processes contributes to the vegetation recovery in the context of soils found in 
development of rehabilitation techniques (Kobayashi, abandoned agricultural and pasture lands, (2) determine 
84 TROPICS  Vol. 23 (2) Shigeo Kobayashi, Manuel Soudre et al.
changes in biomass composition and species diversity, and During the last decade, the average monthly temperature 
(3) define vegetation types in terms of facilitators or oscillated between 23℃ and 26℃. The average evapotrans-
competitors in processes of secondary secession. Instead of piration potential is 1200 mm and the average relative 
only focusing attention on the decrease in forests in Peru, it humidity is 77 %. 
is necessary to attend to previously neglected degraded 
forests in each country located along a tributary of the 
Amazon River. In this study, we focused on the region Study methods
around the Ucayali River, an Amazon tributary in Peru, and 
the area of Pucallpa which is emerging as a new front for 　　Three experimental sites, each being 40 m×40 m×3 
natural forest logging. replications in area, were established to rehabilitate 
degraded abandoned pastures and agriculture lands through 
mixed plantations of Tabebuia serratifolia (tahuari), 
SITE DESCRIPTION AND METHODS Calycophyllum spruceanum  (capirona), Amburana 
cearensis (ishipingo), Terminalia oblonga (yacushapana), 
Study site Cedrelinga catenaeformis (tornillo), and Schizolobium 
amazonicum (pashaco). Three sites were scattered in the 
　　The Ucayali region in the Peruvian Amazon has a area of about 31 km2 in Campo Verde. Each tree was 
population of 391,000 and occupies an area of 102,500 km2. planted in a 3 m×3 m space within these plantations. Three 
It covers seven different ecological zones and four 40 m×40 m plots, each containing five quadrats (measuring 
transitional ecological zones (ONERN, 1976 &1981; 2 m×2 m), were demarcated in each site in August 1997. 
Soudre et al., 2001), with tropical rain forest accounting for The plots were originally dominated by Imperata 
the largest area in this territory, followed by sub-montane brasiliensis (cashupsha), Rottboellia exaltata (arrocillo), 
rain forest. Seventy-five percent of the Ucayali forest is and Baccharis floribunda (sachahuaca). Another two grass 
deforested and degraded, and requires rehabilitation. Our species, Brachiaria decumbens (pasture weed) and 
study sites were located at Campo Verde in Pucallpa, along Hyparrhenia rufa (yarahua) were frequently observed at 
the eastern tributaries of the Ucayali River, including the near the study sites. We had also selected three 800-m2 area 
Aguaytia River. Deforestation in this region stands at each representing the area where initially dominated by one 
30,787 ha annually, and is the most intensive in Federico of the grass species, Imperata brasiliensis (cashupsha), 
Basadre, Nueva Requena (along the Aguayia River), Rottboellia exaltata (arrocillo), or Baccharis floribunda 
Tournavista (along the Pachitea River), and in areas near (sachahuaca), for monitoring the distribution area change of 
the Neshuya River (Fujisaka et al., 2000). The land was these grass species and estimating secondary succession. 
formed from residual material and old alluvial deposits The 800-m2 was consisted of former three sites (2 m×2 m
during the Tertiary period of the Cenozoic era. Geo- ×5 replications×3 replications＝60 m2) and added one 
morphologically, it is defined as dissected alluvium controlled plot (no treatment: 2 m×2 m×5 replications＝
(Rasanen, 1993). Topographically, little elevation exists. 20 m2). Species composition in the 800-m2 monitoring areas 
The relief is flat with slightly undulating slopes varying and sampling points set in each area was recorded using 
between 1 % and 8 %. Drainage is generally good, although Braun-Blanquet＇s (1964) method at the following times: in 
some areas in the region near Nueva Requena experience June 1998 (one month after tree planting) and in June 1999 
poor drainage. The study area, Campo Verde, does not have (one year after planting). A species composition table was 
a meteorological station. However, it is located midway drawn up to examine the relationships between vegetation 
between the Pucallpa (8°23′S, 74°34′W) and San Jorge types and litter, grass biomass, soil condition, and landform.
(8°30′S, 74°52′W) meteorological stations. Therefore, 　　Above ground biomass, accumulated litter, soil 
climatic information for the study site was interpolated condition, and landform were annually surveyed for grass 
using data from these two stations. The average annual species occurring in the vicinity or inside of each quadrat, 
precipitation in the area is 1800 mm, with the majority of plot, and site. The above ground biomass of grass was 
the precipitation occurring during the wet season between surveyed in the vicinity of each quadrat (sample area＝1 m
November and April, and less precipitation occurring ×1 m) or in 10 temporary quadrats (1 m×1 m) in the plot. 
during the dry season between May and October. The The litter amount was surveyed at each quadrat (sample 
average annual temperature is 25.2℃, with an average area＝0.5 m×0.5 m). Ratios of dry weight (obtained 
maximum of 30.9℃ and an average minimum of 19.6℃. through heating at 80℃ over 72 hours) to fresh weight were 
Secondary succession of mixed plantations established to rehabilitate abandoned pasture in the Peruvian Amazon 85
determined for each grass and litter type, and fresh weights assess their allocation patterns: Imperata brasiliensis, 
were converted to dry weights. The soil profile at 30 cm Rottboellia exaltata, Baccharis floribunda, Brachyaria 
depth was measured at a thickness of A0 and A horizons, decumbens, and Hyparrheria rufa. Of these five species, R. 
and a hardness of A and B horizons, using a Yamanaka soil exaltata and B. floribunda were individually harvested, 
penetrometer. Slope direction, slope incline, and relief while each of I. brasiliensis, B. decumbens, and H. rufa 
index were measured using a clinometer. We commenced were harvested in one 1 m×1 m sampling site because 
this study in August 1997, and repeated these observations these species developed rhizome and it is difficult to 
in June 1998 and June 1999. distinguish  individual. We dissected each plant into its 
　　We evaluated species abundance in each 2 m×2 m photosynthetic organ (leave/blade), intake organ (root), 
quadrat and in each observation sites using Braun- support organ (stem), sexual reproductive organ (flower and 
Blanquet＇s cover-abundance scale (Braun-Blanquet, 1964). fruit), and vegetative reproductive organ (rhizome), and 
The scale was converted to dominance values which are obtained their fresh weights. We then calculated the 
necessary for the calculation of diversity indices, as shown reproductive effort (sexual and vegetative reproductive 
below. organ/total individual weight) for each species using the 
data set.
Braun-Blanquet＇s cover-abundance scale: dominance value 
of i th species di
　　r: 0.01 RESULTS
　　＋: 0.001
　　1: 0.05 Vegetation recovery and undergrowth vegetation 
　　2: 0.175 types related to soil condition
　　3: 0.375
　　4: 0.625 　　The initial vegetation before tree planting was 
　　5: 0.875 classified into three vegetation types dominated by I. 
　We employed the following diversity indices: brasiliensis, R. exaltata, and B. floribunda. One year after 
　　M ean diversity, H ＇＝3.3219 (log10 D ＇­1/ D ＇∑ di log10 the trees were planted, the soil condition had changed in 
di), (Lloyd and Ghelardi, 1964) terms of the hardness of the A and B horizons (Table 1). 
　　Total diversity, H ＇D ＇＝H ＇×D ＇ (Pielou, 1975) The A and B horizons were compacted from 6.04 mm to 
　　Evenness, J ＇＝H ＇ / log2 S (Pielou, 1975) 9.25 mm and from 14.5 mm to 16.3 mm, respectively. 
where S denotes the total number of species recorded, and 　　The total number of species increased from 8.23 to 
D ＇ denotes the sum of dominance values ( ∑ di). 10.73, and the total dominance value (D＇) also increased 
from 85.5 to 128.4, resulting in increased mean diversity, 
　　We harvested the following five dominant grasses to total diversity, and evenness. In areas where the original 
Table 1. Vegetation changes observed one month and twelve months after planting in Campo Verde, Pucallpa
1 month after planting 12 months after planting
Slope degree 0.95 0.95
Thickness of A0 horizon (cm) 3.38 2.82
Thickness of A horizon (cm) 4.77 6.36
Hardness of A horizon (mm) 6.04 9.25
Hardness of B horizon (mm) 14.5 16.27
Undergrowth biomass (dry t/ha) 4.63 8.19
Litter (dry t/ha) 8.36 5.59
Total species number (/4 m2) 8.23 10.73
Total dominance value 85.52 128.35
Mean diversity 1.09 1.43
Total diversity 100.92 202.1
Evenness 0.36 0.43
Note: All data are averages and are presented without conducting statistic tests. April/98: clearing; May/98: 
planting; every two months: weeding at an average of 20 points.
86 TROPICS  Vol. 23 (2) Shigeo Kobayashi, Manuel Soudre et al.
dominant vegetation type was Imperata, we found the soil surface, whereas the Imperata type was found to be 
largest amount of undergrowth biomass (9397.9 kg/ha) the associated with hard compacted soil with a hardness＞10-
lowest total dominance value (115.66), and the lowest 15 mm at the soil surface. 
diversity indices in 1999 (Table 2). In areas where the initial 
dominant vegetation type was Baccharis, we found the 
smallest amount of undergrowth biomass (6139.8 kg/ha) Energy allocation patterns and reproductive efforts 
and the highest litter, mean diversity, total diversity, and 
evenness. 　　We sampled five dominant grass species in each area 
　　Lastly, in areas where the initial dominant vegetation to compare their allocation patterns of photosynthetic 
type was Rottboellia, we observed the highest total products (Table 3). The estimated biomass including their 
dominance value. These results indicate a relationship roots for I. brasiliensis, B. decumabens, H. rufa, R. exaltata, 
between mean diversity and soil hardness of the A horizon and B. floribunda. The estimated biomasses, including 
for these three grass species (Figure 1). In 1998, the roots, for these species were: 42.2, 39.2, 34.5, 6.9 and 24.8 
Baccharis and Rottboellia vegetation types were found to t/ha, respectively. The highest proportionate allocation of 
be associated with soft soil with a hardness of＜6 mm at the photosynthetic products to the photosynthetic organ 
Table 2. Vegetation changes by types one month (in 1998) and twelve months (in 1999) after planting
1998 1999
Control Imperata Rottboellia Baccharis Control Imperata Rottboellia Baccharis
Slope degree 1 2 0 1 0.75 2 0 1
Thickness of A0 horizon (cm) 2.25 4.3 2.15 5.7 3.59 3.2 2.33 4.3
Thickness of A horizon (cm) 3.75 8.8 3.2 4 6.25 9.4 5.8 5.8
Hardness of A horizon (mm) 4.1 13.4 3.12 4.44 6.88 11.24 11.06 9.08
Hardness of B horizon (mm) 14.92 14.88 13.58 13.62 12.07 14.94 0 0
Undergrowth biomass (dry t/ha) 13.36 0.9 0.83 0.87 8.47 9.4 9.38 6.14
Litter (dry t/ha) 4.51 9.65 6.14 14.95 4.63 4.31 5.55 8.12
Total species number (/4m2) 9.5 6.8 6.2 10.6 9.08 9.2 11.2 13.9
Total dominance value 139.77 70.04 63.8 53.66 145.7 115.66 128.96 116.61
Mean diversity 1.14 0.22 1.25 1.81 1.46 0.93 1.48 1.79
Total diversity 196.14 16.59 78.68 98.82 234.47 117.94 203.05 233.44
Evenness 0.34 0.09 0.48 0.54 0.47 0.29 0.43 0.48
Fig. 1. Relationship between A horizon soil hardness and mean diversity of R. exaltata, B. decumabens and I. brasiliensis tuypes in 1998.
Secondary succession of mixed plantations established to rehabilitate abandoned pasture in the Peruvian Amazon 87
occurred in I. brasiliensis. In H. rufa, 50 % of these highest reproductive effort with 58.9 % associated with the 
products were allocated to the intake organ, and in both R. vegetative organ, whereas only 4 % was associated with the 
exaltata and B. floribunda, high proportions were found in sexual organ in B. floribunda.
the support organs at 58.1 % and 49.8 %, respectively. At 
the time of sampling, I. brasiliensis had no flowering organs 
and R. exaltata and B. floribunda had no vegetative organs. Secondary succession processes of abandoned 
The reproductive effort [(sexual organ weight＋vegetative pastures
organ weight)/individual weight×100 %] of each grass 
species differed significantly. I. brasiliensis showed the 　　Changes in vegetation between 1997 and 1999 were 
Table 3. Energy allocation patterns and reproductive efforts of five grass species
Species name Stem Photosynthetic Intake Support Sexual VegetativeBiomass Reproductive Reproductive Reproductive(common name) Number Organ Organ Organ Organ Organ Effort
(fw t/ha) (/m2) (%) (%) (%) (%) (%) (%)
Imperata brasiliensis 42.2 375 25.7 10.5 5.5 0 58.7 58.7
(Cashupsha)
Brachyaria decumabens 39.2 n 17.8 38.6 31.7 0.001 11.9 11.9
(pasture weed)
Hyparrheria rufa 34.5 n 14.5 53.4 14.2 3.1 14.8 17.9
(Yarahua)
Rottbellia exaltata 6.9 83 12.8 11.3 58.1 17.9 0 17.9
(Arrocillo)
Baccharis floribunda 24.8 1.4 20 26 49.8 4.2 0 4.2
(Sachahuaca)
Note: Photosynthetic organ＝leaves/blades, intake organ＝roots, support organ＝stem, reproductive organ＝seeds and rhizomes, n: no 
main stem.
Fig. 2. Secondary succession process based on changes in dominance area and mean diversity.
88 TROPICS  Vol. 23 (2) Shigeo Kobayashi, Manuel Soudre et al.
recorded in three observation areas (Fig. 2). Portions of Grime (2001) classified these factors into three types in 
these areas occupied by each dominant species were relation to the evolution of plant strategies: the S-type 
separately estimated by observing changes in the number of (stress-tolerant: physical and nutrition environments), the 
sampling points dominated by the species and estimating R-type (ruderal: environmental disturbance), and the C-type 
mean diversity. At the initial stage of treatment, the (competitive: competition for resources among organisms). 
Imperata, Rottboellia, and Baccharis vegetation types were The idea is related to r/K selection (Pianka, 1970). 
dominant. A further five vegetation types appeared one year Baccharis and Rottboellia types showed a preference for 
after planting, which included Brachiaria, Hyparrhenia, soft soil of＜6 mm hardness at the soil surface that could be 
other weeds, and bush types (these dominated by the categorized as a middle C-type, according to Grime (2001). 
planted trees). Imperata occupied 100 % of its original area Imperata type, by contrast, was distributed in hard 
and its distribution range remained unchanged. The mean compacted areas of＞10-15 mm soil hardness that could be 
diversity of the Imperata stand decreased. The Rottboellia classified as a S-type. Therefore, both environment 
stand decreased from 800 m2 to 160 m2 with the remaining conditions and competition between plants must be 
area of 640 m2 showing an expansion into four vegetation considered in relation to secondary succession. The 
types, including Brachiaria, weeds, and bush types. The reproductive effort of a species significantly affected the 
area dominated by Baccharis remained at 1200 m2 and areas it occupied and its mean diversity, and this correlated 
expanded into four vegetation types, including other with either a facilitation or competition process (Tilman et 
Baccharis types, Hyparrhenia, Rottboellia, bushes, and al., 1996; Callaway and Walker, 1997; Holmgren et al., 
planted trees. This area originally contained 50 % of bushes 1997; Tilman, 1997; Li and Wilson, 1998). I. brasiliensis 
and planted trees and showed the highest mean diversity has a tillering structure associated with its predominantly 
one year after treatment. The Rottboellia and Baccharis vegetative reproduction method, with sexual reproduction 
types were seen to progress to bushes and planted trees occurring only after exposure to fire (FAO-RAP, 2003). R. 
(40 % and 50 %, respectively). exaltata and B. floribunda populations flower frequently at 
the initial stages, and these species were found to have 
invaded all populations except for I. brasiliensis. From our 
DISCUSSION observations during the study, we concluded that B. 
floribunda was a typical succession facilitator for 
Vegetation recovery related to energy allocation controlling the direction of secondary succession in 
patterns and reproductive efforts of five grass species abandoned pasture, as well as a catalyst for increasing 
in Campo Verde, Pucallpa biodiversity. By contrast, I. brasiliensis was found to be a 
secondary successional competitor. 
　　Significant differences were noted between the 
allocation patterns and reproductive efforts of the five grass 
species, especially in their supporting organs. Their Secondary succession processes based on changes in 
reproductive efforts indicated that some of their strategic dominance and mean diversity
aspects were being used to facilitate succession (Kawano 
1974, Tilman 1997). I. brasiliensis showed the highest 　　We examined the early phase of vegetation recovery in 
reproductive effort at 59 % of the vegetative organ, while B. abandoned pastures one year after planting. We found that 
floribunda only showed a 4 % sexual organ allocation. The intensive forest clearing to create pasture affected 
reproductive effort of each species contributes to vegetation recovery during the initial phase. As previous 
maintaining and/or expanding its population (Kawano, studies have also shown, we observed relationships between 
1984). Baccharis type decreased its area size and increased grass types and environmental gradients such as soil 
its mean diversity, whereas Imperata remained stable and hardness (Ito, 1979; Wilson and Keddy, 1988; Currie, 
had the lowest mean diversity. Rottboellia type showed a 1991). This is a similar situation to that seen in newly 
medium changing ratio of mean diversity and area size. created primary non-vegetated sites where seed sources are 
These results are similar to those obtained for a tropical the most important agents for vegetation recovery, and wind 
seasonal forest in Thailand where banana was a facilitator dispersal is an important factor (Nakashizuka et al., 1993). 
and bamboo a competitor (Kobayashi et al., 1995). At the initial phase of treatment, Imperata, Rottboellia, and 
Environmental factors significantly affect the plant Baccharis vegetation types were observed. Subsequently, 
strategies (Kobayashi, 1984b; Wilson and Keddy, 1988). seven other vegetation types appeared such as Brachiaria, 
Secondary succession of mixed plantations established to rehabilitate abandoned pasture in the Peruvian Amazon 89
Hyparrhenia, weeds, bushes, and planted tree types 
ACKNOWLEDGEMENTS　　This study is part of a 
(Cecropia, ishipingo, tabebuya, and tahuari). Over a period 
of one year of mixed plantation, Pueraria phaseoloides and cooperative research program jointly organized and 
other climbers, combined with soil condition, further supported by INIA, the Center for International Forestry 
influenced secondary succession. Imperata type was found Research (CIFOR), E-3 research on rehabilitation in the 
to be associated with the competition process because it landscape of degraded tropical forests (supported by the 
showed the smallest mean diversity and no change in area Global Environment Research Fund), and the Ministry of 
size (Kobayashi, 2004). Vegetation suppressed by this the Foreign Affairs, Japan.
process  would  requi re  t rea tment  for  successfu l 
rehabilitation (enrichment and planting). Baccharis, by 
contrast, was part of the facilitation process because of its REFERENCE
higher mean diversity and recruitment of woody shrubs. In 
the facilitation process, vegetation is expected to follow its Braun-Blanquet J. 1964. Pflanzensoziologie. Springer Verlag, 
Wien, Germany.
natural course of secondary succession but may sometimes 
Callaway RM, Walker LR. 1997. Competition and facilitation: a 
require treatment to accelerate natural regeneration. Similar synthetic approach to interactions in plant communities. 
processes have been observed in several other forest types Ecology 78: 1958−1965.
in different regions. These include: mixed Dipterocarp Currie DJ. 1991. Energy and large-scale patterns of animal-and 
forests in Samarinda, East Kalimantan, Indonesia, and plant-species richness. American Naturalist 137: 27−49. 
Pasoh in Malaysia; a mixed deciduous forest in Maeklong, FAO. 2001. State of The World ＇s Forest. Rome. (Japanese version, 
Thailand; a peat swamp forest  in Belait ,  Brunei, by Noubunkyo, 2002, pp.311)
Darussalam; a subtropical moist forest in Misiones, FAO-RAP (Food and Agriculture Organization of the United 
Nations Regional Office for Asia and the Pacific). 2003. 
Argentina; and a teak (Tectona grandis) plantation in Thom 
Advancing assisted natural regeneration (ANR) in Asia and 
Pha Phun, Thailand (Kobayashi, 2004 & 2007). The the Pacific. RAP Publication 2003/19. 44 pp.
facilitation process is necessary to accelerate natural Fujisaka S, Escbar G, Veneklaas EJ. 2000. Weedy fields and 
regeneration and enrichment planting (Kobayashi, 2004). forests: interactions between land use and the composition of 
The competition process can be introduced in large-scale plant communities in the Peruvian Amazon. Agriculture, 
and catalytic plantations (Nambir and Brown, 1997). This Ecosystems and Environment 78: 175−186. 
facilitation-competition process supports the diversity– Grime JP. 2001. Plant strategies, vegetation processes, and 
productivity hypothesis (Tilman et al., 1996), but does not ecosystem properties. Wiley. 417 pp.
＇ Holmgren M, Scheffer M, Huston MA. 1997. The interplay of support Tilman s statement that“ species-rich sites are more 
facilitation and competition in plant communities. Ecology 
resistant to invasion” (Tilman, 1997). Accelerated natural 78: 1966−1975.
regeneration methods, such as“ umbrella” natural Ito S. 1979. Concepts of diversity in plant sociology. Biological 
regeneration,“ side-effect” natural regeneration, and forest Science 31: 200−206. (in Japanese)
patch improvement, are closely related to assisted natural [ITTO] International Tropical Timber Organization. Forestry 
regeneration (FAO-RAP, 2003). Enrichment planting Agency. 1991. Beyond the guidelines – an action program for 
techniques, such as line planting and gap planting, have sustainable management of tropical forests. ITTO Technical 
also been studied (Kobayashi, 2004). Our study suggests Series 7, Yokohama. 
Kawano S. 1974. Speciation and adaptation. Evolutionary biology 
that large-scale planting using site matching, direct sowing, 
in plants II. Sanseido. 407 pp. (in Japanese)
and mixed planting can act as a catalyst. We have addressed Kawano S. 1984. Life history and evolution of forest floor 
both the development of techniques as well as local vegetation. In: Kawano S (ed) Plant history and evolution-2. 
community incentives for the rehabilitation of abandoned Baifukan, Tokyo, 1−19. (in Japanese) 
pastures. Forest recovery takes a long time, and fires and Kobayashi S. 1984a. Biodiversity of forests. Research of Species 
illegal logging are likely to occur during this period. Local Biology 8: 35−45. (in Japanese)
community incentives should focus on ecological forestry Kobayashi S. 1984b. Life history of Disporum smillacinum and its 
resources, such as fallow products and non-wood forest population dynamics. In: Kawano S (ed) Plant history and 
evolution-2. Baifukan, Tokyo, 87−112. (in Japanese)
products, on which forest dwellers depend. Local 
Kobayashi S. 2004. Landscape rehabilitation of degraded tropical 
communities can then use these resources during the forest ecosystems: case study of the CIFOR/Japan project in 
facilitation process in secondary succession. Indonesia and Peru. Forest Ecology and Management 201: 
13−22. 
Kobayashi S. 2007. An overview of techniques for the 
90 TROPICS  Vol. 23 (2) Shigeo Kobayashi, Manuel Soudre et al.
rehabilitation of degraded tropical forests and biodiversity [ONERN] 1981. Peruvian major land use map, scale 1:1,000,000. 
conservation. Current Science 93: 59−66 Lima, Peru.
Kobayashi S, Yarwudhi C, Punyaratana P. 1995. Population Pianka ER. 1970. On r- and K-selection. American Naturalist 104: 
dynamics of wild banana (Musa acuminata) related with the 592−597.
succession of the logged-over forest. Proceedings of the Pielou EC. 1975. Ecological diversity. Wiley-Interscience. 
International Workshop on the Changes of Tropical Forest Rasanen, M. 1993. The geohistory and geology of the Peruvian 
Ecosystems by El Niño and Others. JSTA, NRCT, & JISTEC, Amazon. pp. 46−67. In: Kalliola R, Puhakka (eds) Peruvian 
pp. 265−270. Amazon. Tropical humid vegetation in the sub-Andean plain. 
Li X, Wilson SD. 1998. Facilitation among woody plants PAUT, and ONERN, Jyvaskyla, Finland. 265 pp.
establishing in an old field. Ecology 79: 2964–2705. Soudre M, Ricse A, Carbajal Y, Kobayashi S, Sabogal C, Alegre J. 
Lloyd M, Ghelardi RJ. 1964. A table for calculating the 2001. Adaptability of six native forest tree species to degraded 
“equitability” component of species diversity. Journal of lands in Pucallpa, Peruvian Amazon. In: Kobayashi S et al. 
Animal Ecology 33: 217−225. (eds) Rehabilitation of degraded tropical forest ecosystems. 
Nakashizuka T, Iida S, Suzuki W, Tanimoto T. 1993. Seed dispersal CIFOR. pp. 123−128.
and vegetation development on a debris avalanche on the Tilman D. 1997. Community invasibility, recruitment limitation, 
Ontake volcano, central Japan. Journal of Vegetation Science and grassland biodiversity. Ecology 78: 81−92.
4: 537−542. Tilman D, Wedin D, Knops J. 1996. Productivity and sustainability 
Nambir EKS, Brown AG. 1997. Management of soil, nutrients and influenced by biodiversity in grassland ecosystems. Nature 
water in tropical plantation forests. CSIRO. Canberra, 379: 718−720. 
Australia. 571 pp. Wilson SD, Keddy PA. 1988. Species richness, survivorship, and 
Nihon Ringyo Chousakai. 1989. Midori no booklet No. 1. 84 pp. biomass accumulation along an environmental gradient. Oikos 
(in Japanese) 53: 375−380.
[ONERN] Oficina Nacional de Evaluacion de Recursos Naturales 
(Office of National Assessment of Natural Resources; Peru). Received: 22 July, 2013
1976. Ecological map of Peru (explanatory guide). Lima, Accepted: 29 April, 2014
Peru, 146 pp.