Bi et al.

(2007) 1/42

Post-print of Canadian Journal of Fest Research 37 (2007) 1615-1630.

DOI: 10.1139/X07-018

Yield decline in Chinese-fir plantations:

A simulation investigation with implications for model complexity

J. Bi1, J.A. Blanco2,4 , B. Seely 2, J. P. Kimmins2, Y. Ding3, and C. Welham2

1
Hebei Academy of Forestry Science, 75 Xuefu Road, Shijiazhuang, Hebei, 050061, P.R.China; Tel (+86)

311-8768-4962; Fax: (+86) 311-8683-9334

bijun2003@yahoo.com

2
Department of Forest Science, University of British Columbia, 3041-2424 Main Mall, B.C., V6T 1Z4,

Canada; Tel: (+1) 604-822-3549; Fax: (+1) 604-822-9133.

juan.blanco@ubc.ca; hamish.kimmins@ubc.ca; brad.seely@ubc.ca; clive.welham@ubc.ca

3
Faculty of Forest Resources and Environmental Sciences, Nanjing Forestry University, 159 Longpan

Road, Nanjing, Jiangsu, 210037, P. R. China; Tel/fax: (+1) 604 438-0239.

mineralding@yahoo.com

4
Corresponding author

Tel: (+1) 604 822-8876

Fax: (+1) 604 822-9133

E-mail: juan.blanco@ubc.ca

1 / 41
Bi et al. (2007) 2/42

Abstract

A variety of competing hypothesis have been described to explain yield decline in Chinese-fir plantations.

The difficulty in implementing field experiments suggests ecosystem modeling as a viable option for

examining alternative hypotheses. We present a conceptual model of Chinese-fir yield decline and explore

its merits using the ecosystem-based FORECAST model. Model results suggest yield decline is caused

primarily by a decline in soil fertility, largely as a consequence of slash burning in conjunction with short

rotations. However, as tree leaf area declines, there is a transition (over subsequent rotations) from seed-

rain-based competition to bud-bank-based competition, increasing the competitive impact of minor

vegetation on tree growth. Short rotations increase understory survival between rotations and may cause a

gradual shift from tree dominance to shrub/herb dominance over subsequent rotations. These effects are

most evident on nutrient poor sites, but understory competition poses a significant yield decline risk on

good sites as well. We conclude that sustainable production in Chinese-fir plantations requires the

avoidance of activities that compromise soil fertility and increase understory competition. The risk and

severity of yield decline would be reduced by increasing rotation lengths and avoiding plantations on

infertile sites.

Keywords: Cunninghamia lanceolata (Lamb.) Hook; yield decline; site degradation; soil fertility; weed

competition; rotation length; monoculture; slash burning, sustainability, ecosystem modelling, FORECAST,

model complexity.

2 / 41
Bi et al. (2007) 3/42

1. Introduction

Chinese-fir (Cunninghamia lanceolata (Lamb.) Hook.), a typical subtropical coniferous tree species,

has been widely planted in the southeast provinces of China, a region of approximately 22 to 34N and 102

to 123E (Wu 1984; Yu 1997). It is one of the most important timber tree species in China, accounting for

60%~80% of the total area of timber plantations in southeast China, and for 20%~25% of the national

commercial timber output (Sheng 1992). The timber of Chinese-fir is straight and decay-resistant, and has a

long history of being an important construction and furniture material in China. It is a fast growing species

which can reach a level of timber production in monoculture plantations on good sites of 450 m 3/ha at final

harvest age of 25~30 years (Jiang et al. 1982). There has been a steady increase in its use in plantations in

the past few decades.

The natural range of Chinese-fir is in the humid subtropical area in southeast China. This is a region

of low mountains and hills, with very broken topography and complicated geology. Plantations are

generally located on slopes of more than about 20%, gentler lower slopes generally being used for

agriculture. The soil type is usually red-earth, but the soil can have originated from very different parent

materials. The soil conditions vary significantly in terms of texture, depth, fertility and other physical and

chemical characteristics. The natural distribution of Chinese-fir is as a component of mixed subtropical

evergreen broad-leaved forests. It is a species of moderate shade tolerance, but grows best in full sunlight

(Wu 1984; Yu 1997). The species is moderately nutrient demanding (Zhong and Hsuing 1993), and in

unmanaged “natural” forest it normally grows on moist and fertile sites. The response of Chinese-fir to

forest fertilization varies with stand age. Growing on yellow-red earth soils, the greatest response of young

Chinese-fir was to P and then to K, with little response to added N. In contrast, the greatest growth response

in mid and late rotation was to nitrogen (Li et al. 1992a, b, 1993).

The documented history of Chinese-fir cultivation can be traced back more than one thousand years

(Wu 1984). Traditionally, Chinese-fir plantations were established after native evergreen broad-leaved

3 / 41
Bi et al. (2007) 4/42

forests were harvested and slash-burned. Sloping sites were sometimes terraced and intercropped with food

plants before canopy closure. The plantation sites were generally abandoned after one or two rotations and

allowed to regenerate naturally by stump sprouting and natural seeding to mixed species stands which acted

as a fallow period to restore the site (Yu 1997). However, since the 1950's, the plantation area of Chinese-fir

has been enlarged, and this species has been repeatedly planted on the same sites without intercropping or

periods of fallow. Farmers have generally used a 25-year rotation, with variation from 20 to 30 years

depending on site quality (Wu 1984). However, some plantations in Fujian Province are being harvested as

young as 17 years, a trend driven by the increasing demand for timber because of economic development

and population increase. At present, most Chinese-fir plantations are in the second or third rotation on the

same sites; some are thought to be in even later rotations, but this is difficult to confirm due to lack of

documentation.

Farmers have reported yield declines in multi-rotation Chinese-fir plantations since the 1960's, but

there were few scientifically rigorous investigations until the late 1970’s and the 1980’s (Li 1981; Fang

1987; Ma 2001). A variety of competing hypotheses on the causes have been proposed, including soil

nutrient depletion caused by nutrient removals in harvested materials (Fang 1987), and/or nutrient losses by

slash burning (Sheng 1992; Ding et al. 1999, 2000); physical degradation of soil (Fang 1987; Yang et al.

2000); toxic substance accumulation (Yu and Zhang 1989; Ma et al. 2000); slow decomposition of Chinese-

fir litter and consequent slow nutrient cycling (Li 1981; Sheng and Fan 2002) and consequent reduced

biological activities of the soil (Jiao and Yang 1999a,b).

Nutrient depletion caused by harvest removal has traditionally been considered to be one of the

main factors responsible for plantation yield decline (Rennie 1955, 1957). It has been asserted that

successive short rotations of Chinese-fir on a site will deplete the soil nutrient pool (Fang 1987; Yang et al.

2000, 2004) and soil carbon stocks (Xi et al. 2006). Over the same period, short-rotation harvesting

removes more nutrients than long-rotation harvesting because of differences in the nutrient content of

sapwood and heartwood, in the proportion of sapwood and heartwood in the harvested materials, and the

different quantities of nutrients removed in foliage and branches (Ma 2000). Short rotations result in a

4 / 41
Bi et al. (2007) 5/42

higher frequency of site preparation by burning and the associated loss of nutrients and organic matter, and

deterioration in soil physical, chemical and biological properties. Every tree species has a specific influence

on soil development and properties due to its particular rooting systems and litter quality (Kilian 1998), and

it has been asserted that monoculture stands, especially of coniferous species, cause soil deterioration by

soil acidification and allelopathy (Miles 1985; Yu and Zhang 1989). Pot experiments showed that the

phenolics from litterfall affected the germination of Chinese-fir seeds (Ma et al. 2000), but there is no study

of allelopathic effects on tree growth.

Based on experience, farmers assert that Chinese-fir will not survive and grow well unless the site

is burned prior to plantation establishment (Yu 1997; Yu and Yang 1989). Studies of young plantations

have confirmed that seedling survival and early growth are significantly better on slash burned sites than on

sites without burning (Yu and Yang 1989; Shao 1992; Ye 1992). Some of the early studies even claimed

that burning was a necessary procedure for the successful establishment of Chinese-fir stands; that it

removes competition from the non-crop vegetation and significantly improves the soil physical, chemical

and biological properties. The pH, content of soluble nutrients, enzyme activity and the number of

microorganisms in the top soil were all increased on burned relative to unburned treatments (Zheng 1958;

May and Attiwill 2003). However, there is increasing evidence that broadcast site preparation by slash

burning is the ultimate cause of soil degradation and site yield decline because of accelerated loss of

nutrients and topsoil. Sheng (1992) showed that the reported improvements of some soil physical and

chemical properties attributed to slash burning disappear one year after planting, whereas negative effects

on soil organic matter and soil fauna are persistent.

One of the most serious effects of slash burning is an increase of soil erosion. In a three-year study

conducted in Youxi, Fujian Province, Sheng (1992) concluded that erosion on slashburned study sites was

about 37 times greater for mineral material, 10 times greater for organic matter, and 8 times greater for the

total loss of soil nutrients than that for an unburned control site. Such losses via burning are often much

higher than those through harvest removals (Attiwill 1984).

5 / 41
Bi et al. (2007) 6/42

Failure to control competition for soil moisture, nutrients and light caused by non-crop vegetation

was identified as being as important in the yield decline in second rotation radiata pine in Australia as

nutrient losses caused by harvesting (Evans 1996). Slashburning harvested Chinese-fir sites has traditionally

been conducted largely to reduce competition from herbs and shrubs that can cause plantation failure if not

controlled. Competition from minor vegetation has been identified as one of the main factors that contribute

to yield decline in the second and third rotation plantations of Chinese-fir. The weed problem is not only

related to invasion of weed species after harvesting. In fact, the main problem from weeds may be the

vigorous sprouting of rhizomatous shrub and herb species. Sprouts of non-crop species grow much faster

than their seedlings, and therefore exert much greater competition for light, moisture and nutrients. The

problem of weed competition is exacerbated by soil fertility and tree growth declines, both of which reduce

the competitive strength of the Chinese-fir. Yang et al. (1998, 1999) showed that understory biomass

increased by 63.6% and 279.6%, respectively, in second and third rotation Chinese-fir stands compared to

the first rotation, reflecting a stand volume decrease of 22.8% and 46.6%, respectively.

The question of non-crop vegetation is complicated by the fact that minor vegetation can play an

important role in restricting soil erosion and leaching losses of nutrients from harvested sites. As noted

earlier, understory vegetation is known to have beneficial effects on soil fertility and to play a key role in

nutrient cycling processes (Little and Shainsky 1995). A study (Yao et al. 1992) conducted in Fenyi, Jiangxi

Province showed that when the total biomass of the understory in Chinese-fir plantations was more than 5 t

ha-1, the content of available nutrients in the soil was significantly higher than that in stands with less

understory biomass.

Since the yield decline and site degradation problem was recognized, many researchers have

attempted to quantify the extent of the decline over multiple rotations. Table 1 lists some reports of yield

decline and nutrients status in second and third rotations of Chinese-fir compared to the first rotation. In

successive monoculture Chinese-fir plantations in Huitong, Hunan Province, soil humus, total N, available

N and P and exchangeable K were decreased by 45%, 38%, 28%, 38% and 18%, respectively, by year 15

of the first rotation Chinese-fir compared to the original broad-leaved forests (Fang 1987). In the second

6 / 41
Bi et al. (2007) 7/42

and third Chinese-fir rotations, the average height of dominant trees decreased by 7% and 23% and the

stand site index was decreased from 16-18 to 14-16, respectively, compared to the first rotation.

Because of the long time scale involved, most studies of site degradation and yield decline are

based on chronosequences of plantations in different rotations on different sites. The basic assumption that

must be met to make this approach valid is that all members of the chronosequence are the same ecosystem

type, have had the same history of disturbance and, in the case of managed stands, the same stand

treatments. It is often difficult to satisfy this assumption for a chronosequence, even for a single rotation. It

is much harder to satisfy it over multiple rotations. The alternative is to study a single site over time,

something that has rarely been done because of the time factor. Such long term studies have not been

conducted for Chinese-fir.

In the examples of yield decline listed in Table 1, Shao (1992) did not give an adequate description

of the study sites, which were distributed between different rotations and sites. The third rotation site

reported by Yang et al. (1999) was located 800 m from the second rotation site, but was adjacent to the first

rotation site. The first, second and third rotation stands in Fan et al. (2000) were arbitrary arranged on site

indices of 18, 16 and 14, respectively, thereby confounding rotation number with site quality. This design

was based on earlier studies (Fang 1987; Lin et al. 1992) which reported that site index decreases by one

site class for each successive rotation of Chinese-fir plantations; there was no independent measure of the

original site index of all the sites. There are many examples of planting Chinese-fir on inappropriate sites,

and yield decline may be largely restricted to plantations established on sites that are not suitable for

sustained Chinese-fir production. There are no reports from long-term monitoring of a single site with a

clear description of the management history. The same problem can be found in the review by Zhang et al.

(2004), where the authors show data from different sites but they do not give data for site index or site

quality.

The lack of comparable soil and management conditions in the studies listed in Table 1 or in the

review by Zhang et al. (2004) raises the question as to whether Chinese-fir yield decline is real or is an

artifact of inadequate experimental design. As noted above, the soil parent materials and characteristics,

7 / 41
Bi et al. (2007) 8/42

slope, slope position, aspect, and management history often vary significantly in Chinese-fir plantations

over small distances. The lack of adequate experimental design to account for these variations raises

questions about the validity of the chronosequence data sets on which conclusions with respect to yield

decline have been reached. Nevertheless, examples of site and soil damage resulting from repeated

slashburning on short rotations are abundant in southeastern China, and failed Chinese-fir plantations are

common, as are plantations that are growing well and not showing any signs of yield decline. There is an

urgent need to identify the site types, management practices and other factors that result in yield decline on

some sites and not on others, and the sites on which there is a significant risk of yield decline.

2. Conceptual model of yield decline in Chinese-fir

In spite of the above concerns over the empirical evidence for yield decline, we do not deny its

existence in many Chinese-fir plantations in southeastern China. In fact, basic knowledge of the production

ecology of these forests would predict that yield decline is probably inevitable on many sites under current

practices. Our only issue with the literature on this topic is the lack of statistical validity in many Chinese-

fir studies that therefore fail to identify the specific causes and the magnitude of the problem. Ultimately,

the most reliable method to quantify site degradation and yield decline caused by inappropriate

management is to establish long-term field trials that permit explicit identification of the key determinants

of the problem. Such a study would be spatially extensive as well as of long duration since it should

examine individually and in combination the effects of: rotation length, monocultures vs. mixed species

stands, repeated rotations of the same species vs. alternating stand composition in different rotations, slash

burning, various levels of minor vegetation competition, and litter raking and branch harvesting. These

treatments should be examined for their effects on soil organic matter, nutrient availability, site quality,

erosion, growth of Chinese-fir, site hydrology and water quality, wildlife habitat, and various measures of

8 / 41
Bi et al. (2007) 9/42

biological diversity, including medicinal plants and traditional forest food plants, mushrooms, etc. These

studies are needed on poor, medium and good sites.

A major shortcoming of scientific research is that complex issues like Chinese-fir yield decline are

often examined in terms of single factor hypotheses. Without such simplification it is often impossible to

complete research projects within the budget and time available for the individual studies. The

simplification of this complex issue into a series of causal-hypothesis-testing studies has led to the plethora

of hypotheses described above, none of which has proven capable of explaining the phenomenon

satisfactorily (c.f. Kimmins et al. 2005). What is required is a comprehensive, long-term study with an

experimental design that will permit identification of the individual contributions of the many determinants

of yield decline, and their interactions. In the absence of such comprehensive, long-term studies, an

alternative is to combine our experience of Chinese-fir plantation growth with our knowledge of ecosystem

processes to develop predictive models that can provide an interim basis for designing more sustainable

systems, until the results of long term field trials become available.

A suitable conceptual framework for modeling the issue of forest sustainability is that of ecological

rotations (Kimmins 1974). An ecological rotation (ER) is the time taken for an ecosystem, or a particular

ecosystem condition, structure, process or value, to return to its original level or to some new desired level

following disturbance. The length of the ecological rotation is a function of: 1), the degree of change in the

parameter of interest, and 2), the natural or management-assisted rate of recovery in that parameter.

Sustainability of any particular parameter is determined by whether or not the combined disturbance-

recovery is shorter or longer than the ecological rotation. Sustainability cannot be defined by any one of

severity of disturbance, frequency of disturbance or ecosystem resilience. A conceptual model of Chinese-

fir plantation sustainability thus requires a representation of frequency (rotation length), severity (harvest

removals, losses in burning, leaching and erosion), and rate of recovery (ecosystem resilience). Ecosystem

resilience – the slope of the recovery line, is a function of nutrient, moisture and light availability to the

trees, which is in turn influenced by soil fertility, nutrient cycling, the beneficial effects of minor vegetation

in retaining nutrients on site, and the negative effects of competition from the minor vegetation for the

9 / 41
Bi et al. (2007) 10/42

resources the trees need for the development of leaf area. The resilience of the Chinese-fir system is

reduced when this species is planted off-site (dry, nutrient poor sites); when monoculture plantations cause

a change in forest floor type from mull-moder (under the native mixed forest) to moder-mor (because of the

quality of the Chinese-fir litter) which reduces nutrient recirculation rates and increases nutrient and organic

matter losses during slashburning; and by the increased competition from minor vegetation when the leaf

area and vigor of the fir is reduced because of lack of soil nutrients and moisture.

Based on this conceptual model, we think that yield decline in Chinese-fir plantations is the

combined result of: loss of soil fertility caused by short rotations with burning; slower decomposition of

litter and the attendant slowing of the nitrogen cycle that results from the reduction in quality of litter when

mixed species, broadleaved forest is converted to monoculture Chinese-fir forest; increased exposure of

organic matter and nutrients to loss in burning that results from slower decomposition of litter and the

accumulation of a thick forest floor; increased erosion losses following burning of slopes; reduction in tree

leaf area that slows tree growth and favors understory vegetation by reducing shading, which in turn

increases competition for light and soil resources; increased competition from herbs and shrubs early in the

rotation because they are re-growing as sprouts from rhizomes rather than from seed if the tree leaf area is

insufficient to shade them out before the end of the rotation. The interaction of all these factors reduces both

the growth potential of Chinese-fir and its ability to compete for resources with herbs and shrubs. As a

result, the decline in Chinese-fir will be greater than the decline in soil fertility (Figure 1). Any one or more

of longer rotations, cessation of burning, fertilization, growing of Chinese-fir in mixed species stands, and

restricting the species to the appropriate site should reduce or prevent yield decline.

In this study we examined the dynamics of three of the suggested major determinants of yield

decline in Chinese-fir and evaluated the potential of management options to address the problem through

the use of the ecosystem management model FORECAST (Kimmins et al. 1999). The analysis focuses on

the effects of rotation length, slash burning and minor vegetation on soil fertility (nutritional site quality),

site productivity, tree leaf area development and minor vegetation dynamics. The objective was to identify

10 / 41
Bi et al. (2007) 11/42

management strategies that will help to maintain the long-term site productivity of Chinese-fir forests.

Effects of thinning and litter raking were also investigated but are not reported here.

3. FORECAST model description

The ecosystem management simulation model FORECAST (Kimmins et al. 1999) has been used as

a long-term management evaluation tool in several types of forest ecosystems (e.g. Morris et al. 1997; Wei

et al. 2000, 2003; Seely et al. 2002; Welham et al. 2002). In addition, FORECAST has been evaluated

against 29 years of Douglas-fir (Pseudotusga menziesii) thinning and fertilization response data and found

to provide satisfactory predictive accuracy (Blanco et al., in review). The model was specifically designed

to examine the impacts of different management strategies or natural disturbance regimes on long-term site

productivity. The projection of stand growth and ecosystem dynamics is based on a representation of the

rates of key ecological processes regulating the availability of, and competition for, light and nutrient

resources. The rates of these processes are calculated from a combination of historical bioassay data

(biomass accumulation in component pools, stand density, etc.) and measures of certain ecosystem

variables (e.g. decomposition rates, photosynthetic saturation curves) by relating ‘biologically active’

biomass components (foliage and small roots) with calculations of nutrient uptake, the capture of light

energy, and net primary production. Using this ‘internal calibration’ or hybrid approach, the model

generates a suite of growth properties for each tree and plant species to be represented. These growth

properties are subsequently used to model growth as a function of resource availability and competition

(Figure 2). They include (but are not limited to): 1) Photosynthetic efficiency per unit foliage biomass based

on relationships between foliage biomass, simulated self-shading, and net primary productivity after

accounting for litterfall and mortality; 2) Nutrient uptake requirements based on rates of biomass

accumulation and literature- or field-based measures of nutrient concentrations in different biomass

components on different site qualities; 3) Light-related measures of tree and branch mortality derived from

11 / 41
Bi et al. (2007) 12/42

stand density input data in combination with simulated light profiles. Light levels at which foliage and tree

mortality occur are estimated for each species.

Soil fertility in FORECAST is represented based on a bioassay approach in which empirical input

data describing decomposition (mass loss) rates and changes in chemistry as decomposition proceeds allow

for calculation of nutrient release from litter and humus (Figure 3). Carbon allocation in response to soil

fertility and tree/plant nutrition is based on empirical biomass ratios and biomass turnover rates (e.g.

number of years of leaf retention for evergreens) for sites of different fertility, and on literature or locally-

obtained values for variation in fine root turnover along fertility gradients. Moisture limitation on growth is

currently based on moisture-determined maximum foliar biomass and thus maximum foliar N. A water

balance model is being added that will add more detail to moisture limitation, including effects on

photosynthesis and litter decomposition, but this was not used in this analysis. The time step of

FORECAST is currently annual. A daily time step is being added to permit the simulation of climate

change effects and water budgets.

FORECAST performs many of its calculations at the stand level but includes a submodel that

disaggregates stand-level productivity into the growth of individual stems with user-inputted information on

stem size distributions at different stand ages. Top height and diameter at breast height (DBH) are

calculated for each stem and used in a taper function to calculate total and individual gross and

merchantable volumes.

Model calibration and application

FORECAST has four stages in its use: 1) data assembly, input and validation; 2) establishing the

ecosystem condition for the beginning of a simulation run (by simulating the known or assumed history of

the site); 3) defining a management and/or natural disturbance regime; 4) simulating this regime and

analyzing model output. The first two stages represent model calibration. Calibration data are assembled

that describe the accumulation of biomass (above and below-ground components) in trees and minor

vegetation for chronosequences of stands developed on sites that vary in nutritional quality. Tree biomass

12 / 41
Bi et al. (2007) 13/42

and stand self-thinning rate data are often generated from the height, DBH and stand density output of

traditional growth and yield models in conjunction with species-specific component biomass allometric

equations. To calibrate the nutritional aspects of the model, data describing the concentration of nutrients in

the various biomass components are required. FORECAST also requires data on the degree of shading

produced by different quantities of foliage and the response of foliage to different light levels (this

information is derived from literature values, field measurements, or simulation models). A comparable but

simpler set of data on minor vegetation must be provided if the user wishes to represent this important

ecosystem component (c.f Royo and Carson 2006). Data are obtained from the literature or field

measurements. Lastly, data describing the rates of decomposition of various litter types and soil organic

matter are required for the model to simulate nutrient cycling. A detailed description of the input data

requirements can be found in Kimmins et al. (1999), or from the corresponding author.

The second stage of calibration requires running the model in “set-up” mode to establish initial site

conditions. In this stage, the model is run with nutrient feedback turned off to allow it to accumulate

vegetation, litter and soil organic matter representative of the site(s) to be modeled, and which reflects the

historical patterns of accumulation. This is typically achieved by simulating the known or estimated natural

disturbance and/or management history of the site.

Sources of calibration data

The majority of data for model calibration were derived from published studies from sites with

climates similar to that of the central area of Chinese-fir plantations in Fujian and Hunan provinces. Given

the fact that the data come from several studies, care was taken to ensure that selected data were from

comparable sites. The input data were from sites covering the observed range of Chinese-fir plantation

growth, qualitatively described as very poor, medium and very good. The sites were scaled quantitatively as

17, 21 and 27 (from poor to good, based on top height in m at age 60 years) to represent a relative index of

tree growth required for extrapolation within the model. Numeric site qualities in FORECAST can be, but

are not necessarily, based on site index - top height at some index age (Kimmins 1990, 1993). We assume

13 / 41
Bi et al. (2007) 14/42

in these simulations that the range in tree growth data reflects primarily the range in site nutrient availability

within the climatic area from which the data came.

Data describing precipitation nutrient inputs, slope seepage, mineral soil cation and anion exchange

capacities, humus mass, nutrient concentrations in litterfall, litter decomposition rates, etc. were derived

from Liu et al. (1991), Zhou et al. (1991), Tian and Zhao (1989), Tian et al. (1989), Tian (1994), Liao et al.

(1999, 2000), Huang et al. (2000), Ding et al. (1999), Ding and Chen (1995), and Yang et al. (2000). The

tree data on biomass, mortality and stand density, tree height and canopy height, nutrient concentrations of

live tissues and other data were based on values reported by Pan et al. (1983), Wu (1984), Shao (1992),

Yang et al. (1998, 1999), Xiao et al. (1999),, Lin et al. (1996), Liu (1998), Tian et al. (1989a, b), Tian

(1994), Zhong and Hsiung (1993) and Zhou (1994, 1999) for similar Chinese-fir plantations. The minor

vegetation data for herb and shrub biomass, height, tissue nutrient concentrations and other relevant data

came from Fan et al. (2001), Lin et al. (2001), Xiang et al. (2003) and Yan et al. (2003). Detailed field data

on belowground minor vegetation biomass were not available, and were thus estimated from aboveground

biomass. The implications of these estimates are discussed later.

This is a complex data set. FORECAST can be run with much simpler calibration data sets by

disabling one, several or even most of the processes that the model is capable of representing., but the user

must understand the implications of omitting key population, community and/or ecosystem structures and

processes. The benefits of reducing the calibration data demands must be balanced against the costs in

predictive power and realism of omitting important ecosystem components and functions. In its simplest

rendition FORECAST can be used as a simple inter-tree light competition population model, but we do not

advise such use for most applications as other ecosystem structures and processes are known to be

important determinants of stand dynamics, successional development and response to disturbance.

4. Simulation scenarios

14 / 41
Bi et al. (2007) 15/42

Establishment of initial conditions

Prior to running the model to evaluate alternative scenarios it was necessary to establish the initial

conditions for the model runs. To evaluate the effect of initial nutritional site quality, initial conditions files

were prepared to represent both good and poor sites with quality index values of 27 and 17, respectively

(relative to the calibration site quality range described in Section 2.3). This involved simulating seven

successive 50-year-rotations for poor and good sites with nutrient feedback switched off, thereby forcing

the growth of the forest to reflect the historical data and the accumulation of soil humus and nutrient levels.

One final 50-year rotation with nutrient feedback switched on was simulated to allow the model to stabilize.

The output from this final run (values for all ecosystem variables represented in the model) was used as the

starting state for the simulation of the management scenarios. The values of selected state variables for the

starting condition for both site qualities are shown in Table 2.

Management scenarios

A set of management scenarios was developed to cover a range of past and present stand practices

in monoculture Chinese-fir plantations on both good and poor sites, including four rotation lengths (20, 30,

40 and 60 years), with and without slash burning, and with and without minor vegetation (Table 2). Stand

density was set at 3000 stems per hectare in all simulations to represent current practices (e.g Yu 1997).

Light competition, nutrient competition and nutritional control of growth were represented in all runs. The

main treatments were designed and simulated based on the following considerations.

1. Rotation length. Farmers have generally used a 25-year rotation, with variation from 20 to 30 years

depending on site quality (Wu 1984). For the present analysis we compared rotation lengths of 20, 30, 40

and 60 years (for a total simulation period of 120 years) in which 95% of the stemwood and bark were

harvested at the end of each rotation. Residual above-ground and all below-ground tree biomass and above-

15 / 41
Bi et al. (2007) 16/42

ground minor vegetation biomass (assumed to be broken off during the harvesting) was assumed to be left

on site to decompose or to be burned in a slash burning treatment.

2. Slash burning is perhaps the most universal silvicultural practice applied in Chinese-fir plantations

(Zheng 1958). The main purpose of slash burning is to prepare a harvested site for planting by reducing the

amount of slash cover and inhibiting the growth of competing vegetation. Without burning or some kind of

vegetation control, herb and shrub growth in recently harvested stands can be so dense that it is difficult to

plant Chinese-fir seedlings, and the survival and growth of seedlings are greatly reduced (Ye 1992; Yu

1997). On the other hand, slash burning is thought to be one of the key factors leading to the long-term

decline in soil fertility and forest productivity (Boyer 1994; Ding and Chen 1995; Ma et al. 1995; Zheng

and Ding. 1997, 1998, Zhang et al. 2004). Slash burning is usually done in the winter or early spring after

the final harvest, but sometimes the slash and other cut vegetation is left on site to dry for one summer, and

then burned in the following spring (Yu 1997). In this study we simulated slash burning the year after

harvest and before planting.

3. Minor vegetation. As described above, competition from minor vegetation in Chinese-fir plantations can

pose a serious threat to the growth and development of planted trees by competing for light when trees are

young and for belowground resources both when the trees are young and as they mature. To evaluate the

potential impact of minor vegetation competition on Chinese-fir growth, simulations were conducted both

with and without understory vegetation. Growth of minor vegetation in the first rotation of all scenarios was

simulated from seed whereas growth in subsequent rotations was simulated as from rhizome sprouts with

the exception of cases in which the understory vegetation had been completely shaded out by the end of the

previous rotation when it was re-established from seed.

Simulation results and discussion

16 / 41
Bi et al. (2007) 17/42

Effects of rotation length

As expected, model results showed that the longer the rotation length, the less the simulated yield

decline in subsequent rotations (Figure 4). Relative differences between the poor and good sites were small.

With a rotation of 20 years (without slashburning and minor vegetation), yield decline on the good site (as a

percentage of the first rotation) was 3.4%, 5.4%, 7.1%, 9.4% and 10.9% in the second, third, fourth, fifth

and sixth rotations, respectively. On the poor site, the equivalent yield decline was 2.1%, 2.7%, 5.9%, 7.9%

and 8.5%. In contrast, in the 60-year-rotation scenarios, yield in the second rotation increased 10.8% on the

poor site and 8.9% on the good site. Even in the absence of slash burning, rotation lengths of 30 years or

less had increasingly negative impacts on soil properties related to ecosystem productivity relative to

starting conditions for both the poor and good sites (Table 2). In addition, there were interactions between

rotation length, slash burning and minor vegetation competition that amplified the effect of shortening the

rotation length. These results support the notion that frequency and intensity of disturbance are key factors

to consider in developing forest management regimes to sustain ecosystem productivity. Similar

conclusions were drawn by Morris et al (1997) and Seely et al (1999). While lengthening rotations may

significantly reduce yield decline and soil degradation in Chinese-fir plantations, the sustainability of this

practice in a broader context must also be considered in terms of economic viability as well as indicators

associated with biodiversity.

Effects of slash burning

Model results showed that slash burning had the largest impact on yields of the different

silvicultural actions examined in this study (Figure 4). Simulated slash burning accelerated yield decline on

both sites, reducing both stand growth and levels of simulated soil nitrogen pools. Yield declines (not

including the effect of minor vegetation) were most severe in the 20-year-rotation scenarios with drops in

harvested stemwood biomass (relative to the no-burn scenario) as high as 32.9% and 31.1% after the first

17 / 41
Bi et al. (2007) 18/42

rotation and 47.1% and 58.1% after 120 years in the good and poor sites, respectively. Even when the

rotation was extended to 60 years, the effect of slash burning on yield was substantial on both good

(declines up to 29.4%) and poor (declines up to 39.2%) sites.

The simulated declines in site productivity in the slashburning scenarios were directly related to the

substantial losses of litter and humus mass and their associated N contents (Table 2). Stone and Elioff (1998)

observed similar declines in aspen productivity following the removal of 80% of the forest floor relative to

similar stands with no removal. The large losses of organic matter in the simulations presented here were

not surprising considering that slash burning removed almost all the logging slash and forest floor litter, and

some of the root biomass; a severity of burning that corresponds to field observations and measurements of

burned Chinese-fir sites (Minghe and Ritchie 1999). Most of the nitrogen in these materials was lost

through volatilization during the burning. Mineralized N that was not volatilized was released to the soil

and was subject to leaching loss because the removal of forest floor and reduction of understory vegetation

limited the capacity of the ecosystem to immobilize and retain nitrogen (Boyer 1994; Ding and Chen 1995;

Ma et al. 1995; Zheng and Ding 1997, 1998; Zhang et al. 2004; Xi et al. 2006).

Effects of minor vegetation

The presence of minor vegetation, through the competition for limited site resources (water,

nutrients and light), has been found to reduce tree growth in a wide range of forest plantations including

Pinus taeda L. (Knowe et al. 1985), Pinus radiata D. Don (Richardson et al. 1996), and in Chinese-fir (see

above). However, minor vegetation has also been found to contribute positively to ecosystem nutrient

cycling (Fang 1990; Little and Shainsky 1995; Yang et al. 1995) because of its higher nutrient

concentrations, relatively rapid litter decomposition (Grove and Malajczuk 1985), and its ability to retain

nutrients on site after a disturbance until trees are established. Results of the simulations with minor

vegetation in the absence of burning showed that minor vegetation, in general, had a slightly positive long-

term impact on Chinese-fir productivity (Figure 4a, c). This was related to the ability of the vegetation

reduce the loss of site nutrients following harvest. In contrast, when the sites were slashburned, the presence

18 / 41
Bi et al. (2007) 19/42

of minor vegetation consistently caused a pronounced reduction in Chinese-fir growth on the poor site

(Figure 4b) and a similar but smaller pattern of growth reduction on the good site (Figure 4d). The

simulated negative interaction between slashburning and minor vegetation originated from the increased

competition for the relatively lower levels of available nutrients in the slashburning scenarios (Table 2). The

negative effect of minor vegetation on Chinese-fir growth was increased in scenarios where the Chinese-fir

did not shade out the minor vegetation prior to harvest, resulting in a more rapid regrowth of minor

vegetation from rhizome sprouts in the subsequent rotation relative to regrowth from seed (Figure 5). Once

the minor vegetation was able to regenerate from rhizomes it took relatively longer for the Chinese-fir to

recover from the competitive impacts. As a result of this negative feedback loop, minor vegetation became

increasingly competitive in subsequent rotations as nutrient resources declined. In fact, in the case of the

20-year rotation with slashburning and minor vegetation, the non-crop biomass began to exceeded tree

biomass by the fifth rotation on the poor site – essentially resulting in a shrub climax (Figure 6). The

detailed response of the minor vegetation is shown for the aforementioned scenario in Figure 6. Without

burning, the minor vegetation was shaded out prior to the end of the rotations for all rotation lengths and

both sites, forcing regrowth from seed. In contrast, with burning, minor vegetation survived to the end of

the rotation for rotations less than 60 years in length, resulting in re-growth from sprouting. From these

results it appears that the capability of vegetative regrowth in minor vegetation plays a key role in the

observed yield decline. Moreover, these results provide support for the conceptual model (Figure 1) which

suggests that frequent, high intensity disturbance in these Chinese-fir plantations may push the ecosystem

towards a shrub dominated climax. It is clear that there is a need for more field research to better understand

the interactions between minor vegetation and tree growth in Chinese-fir plantations, and our results

strongly support the conclusion from a recent review by Royo and Carson (2006) which clearly shows the

role of understory in forest canopy growth and development has usually been underestimated.

Validation of model results

19 / 41
Bi et al. (2007) 20/42

While the availability of data sets describing the implications of management activities on long-

term site productivity in forest ecosystems is extremely limited, a recent review of the literature on yield

decline in Chinese-fir plantations (Zhang et al. 2004) provided some suitable data for model validation. A

comparison of model predictions of yield decline of harvested stemwood biomass for 20 and 30 year

rotations (with slashburning and minor vegetation) were within the range of those reported by Zhang et al.

(2004) for both poor and good sites (stratified by first rotation biomass production) (Figure 8a, b).

Projections of long-term declines in soil organic matter were also consistent with data reported for Chinese-

fir plantations with 20-year rotations (Figure 8c) (data for 30-year rotations and for stratification by site

productivity were not available). The fact the model showed similar long-term trends of declining C:N

ratios in soil organic matter (SOM) in subsequent rotations to those reported by Zhang et al. (2004)

provided further evidence of satisfactory model behaviour (Figure 8d). The simulated declines in C:N ratios

are the result of relatively larger losses of SOM from the active soil organic matter pool which has lower N

concentration relative to the passive SOM pool represented in the model (see Figure 2). This result supports

the belief that total soil N pools are not always indicative of the quantity of plant available N and site

productivity.

Sustainability and complexity in Chinese-fir plantations

The analysis presented here suggests that tree productivity of Chinese-fir plantations would be

sustainable with rotation lengths of > 30 years if there is no slash burning. If the practice of slash burning is

continued with the same level of severity as represented here, a minimum of 60- year rotations or longer

would be required to sustain levels of production over multiple rotations (Figure 4). Yet, even with 60-year

rotations, slashburning resulted in a loss of productivity (30-40% of stemwood biomass) over the 120 year

simulation period relative to the equivalent no-burning scenario. If the effects of post-harvest soil erosion

were included in the analysis (not included in the present analysis) the ecological rotations may need to be

lengthened, particularly for plantations developed on steep slopes (c.f. Xi et al. 2006).

20 / 41
Bi et al. (2007) 21/42

While FORECAST appears to be satisfactory for this application based on the limited validation

presented here (Figure 8), there are some limitations of the model with respect to the Chinese-fir yield

decline investigation. One reason for this may be that the model does not presently represent moisture

competition. There are several examples of moisture competition from minor vegetation reducing tree

growth in pine plantations (Kozlowski 1949, Riegel et al. 1992, Richardson 1993). These issues suggest the

need for improved data on herb and shrub growth from both seeds and sprouting, better data on nutritional

regulation of growth of these life forms, and of carbon allocation in response to changing resource levels

(light, moisture and nutrients). Yield decline in Chinese-fir is an ecosystem-level rather than a population-

level phenomenon and simulation of this requires ecosystem-level models that explicitly represent all the

major determinants.

Acknowledgements

Funding for this work was provided by The Service Center for Experts and Scholars of Hebei

Province, P.R.China. The authors also would like to thank Ms Gao Hongzhen for her reliable and patient

assistance in collection of research data and material.

References

Attiwill, P.M. 1984. Effects of fire on forest ecosystems. In Proceedings of the Australian Forest Nutrition

Workshop. Edited by Lansdsberg, J.J., and Parsons, W. CSIRO, Melbourne. pp 159-172.

Boyer, W. D. 1994. Effect of burning and brush treatments on nutrient and soil physical properties in young

longleaf pine stands. For. Ecol. Manage. 70: 311-318.

Ding, Y.X., and Chen, J.L. 1995. Effect of continuous plantation of Chinese-fir on soil fertility. Pedosphere.

5: 57-66.

Ding, Y.X., Tian, Y., and Qi, L. 1999. A testing simulation with FORECAST on long-term productivity of

Chinese-fir plantations (in Chinese). Forestry Studies in China. 1: 34-38.

21 / 41
Bi et al. (2007) 22/42

Ding, Y.X., Tian, Y., and Qi, L. 2000. Modeling of Chinese fir plantation productivities in different

rotations (in Chinese). J. Nanjing Forestry Univ. 24: 21-25.

Evans, J. 1996. The sustainability of wood production from plantations: evidence over three rotations in the

Usutu Forest, Swaziland. Commonwealth For. Rev. 75: 234-239.

Fan, S.H., Ma, X.Q, Chen, S.S., and Lin, S.J. 2000. Comparative study on growth and development of

different generation plantations of Chinese fir (in Chinese). Scientia Silvae Sinicae. 36: 9-15.

Fan, S., Ma, X., Fu, R., and Liu, A. 2001. Comparative study on underground vegetation develop of

different generation plantations of Chinese-fir (in Chinese). Forest Research. 14:11-19.

Fang, Q. 1987. Influence of continuous cropping on soil fertility and stand growth in Chinese-fir plantations

(in Chinese). Scientia Silvae Sinicae. 23: 389-397.

Fang, Q. 1990. Effect of soil and minor vegetation on nutrient cycle and biomass of Chinese-fir ecosystem

(in Chinese). Scientia Silvae Sinicae. 26: 201-208.

Grove, T.S., and Malajczuk, N. 1985. Nutrient accumulation by trees and understory shrubs in an age-series

of Eucalyptus diversicolor F. Muell. stands. For. Ecol. Manage. 11: 75-95.

Huang, Z.Q., Liao, L.P., Wang S.L., and Liu, Y.D. 2000. Dynamics of phenolics content of Chinese-fir

stump-roots and the rhizosphere soil and it's allelopathy (in Chinese). Chinese J. Appl. Ecol.. 11: 190-

192.

Jiang, Z.L., Ye, J.Z., and Zhou, B.L. 1982. Tending Felling of Chinese Fir Plantations (in Chinese). China

Forestry Press House. Beijing.

Jiao, R.Z., and Yang, C.D. 1999a. The changes of the soil microorganism in rhizosphere and outside in

different developing stages of the Chinese fir plantation (in Chinese). Scientia Silvae Sinicae. 35: 73-79.

Jiao, R.Z., and Yang, C.D. 1999b. Change of the soil microorganism in different generation of the Chinese

fir plantation (in Chinese). Forest Research. 12: 21-25.

Kilian, W. 1998. Forest site degradation—temporary deviation from the natural site potential. Ecol. Engine.

10: 5-18

22 / 41
Bi et al. (2007) 23/42

Kimmins, J.P. 1974. Sustained yield, timber mining, and the concept of ecological rotation: A British

Columbian view. Forestry chronicle 50:27-31.

Kimmins, J.P. 1990. Modelling the sustainability of production and yield for a changing and uncertain

future. For. Chron. 66: 271-280.

Kimmins, J.P. 1993. Scientific foundations for the simulation of ecosystem function and management in

FORCYTE-11. Forestry Canada. Northern Forestry Centre. Edmonton, Alberta. Info. Rept. NOR-X-

328. 88 pp.

Kimmins, J.P., Mailly, D. and Seely, B. 1999. Modelling forest ecosystem net primary production: the

hybrid simulation approach used in FORECAST. Ecol. Model. 122: 195-224.

Kimmins, J.P., Welham, C., Seely, B., Meitner, M., Rempel, R., and Sullivan, T. 2005. Science in forestry:

why does it sometimes disappoint or even fail us? For. Chron. 81: 723-734.

Knowe, S.A., Nelson, L.R., Gjerstad, D.H., Zutter, B.R., Minogue, P.J., and Dukes, J.H. 1985. Four-year

growth and development of planted loblolly pine on sites with competition control. South. J. Appl. For.

9: 11-15.

Kozlowski, T.T. 1949. Light and Water in Relation to Growth and Competition of Piedmont Forest Tree

Species Ecol. Monogra. 19: 207-231.

Li, C. 1981. The differences on nutrients balance between Chinese fir forest and broad-leaved mixed forest

(in Chinese). Acta Pedologica Sinica. 18: 255-261.

Li, Y.Q., Ji, J.S., and Chen D.D. 1993. Growth responses of middle-aged Chinese fir plantation to fertilizer

application. Forest Research. 6: 390-396.

Li, Y.Q., Chen, D.D., Ji, J.S. 1992a. Growth responses of young Chinese fir to fertilizer application. In

Research on Site Degradation of Timber Plantation (in Chinese). Edited by Sheng, W.T. Beijing,

Science and Technology Press of China. pp 198-211.

Li, Y.Q., Chen D.D., and Shen, G.D. 1992b. Growth responses of pre-matured Chinese fir plantation to

fertilizer application. In Research on Site Degradation of Timber Plantation (in Chinese). Edited by

Sheng, W.T. Science and Technology Press of China, Beijing. pp . 223-239

23 / 41
Bi et al. (2007) 24/42

Liao, L.P., Ma, Y.Q., Wang, S.L., Gao, H., and Yu, X.J. 2000. Decomposition of leaf litter of Chinese-fir in

mixture with major associated broad-leaved plantation species (in Chinese). Acta Phytoecologica Sinica.

24: 27-33.

Liao, L.P., Yang, Y.J., Wang S.L., and Gao, H. 1999. Distribution, decomposition and nutrient return of the

fine root in pure Cunninghamia lanceolata,Michelia macclurei and the mixed plantations (in Chinese).

Acta Ecologica Sinica. 19: 246-250.

Lin, X., Hong, L.X., and Du, G.J. 1992. Preliminary study on soil quality evaluation on continuous

cropping Chinese fir plantation. In Research on Site Degradation of Timber Plantation (in Chinese).

Edited by Sheng, W.T. Science and Technology Press of China, Beijing. pp 267-275.

Lin, K., Yu, X., He, Z., Qiu, E., and Lin, S. 1996. Study on the difference of biomass structure and soil

fertility in Chinese-fir stands of different densities (in Chinese). Scientia Silvae Sinicae. 5: 385-392.

Lin, K.M., Yu, X.T., Hong, W., and Huang, B.L. 2001.Effect of minor vegetation on soil fertility in

Chinese-fir plantation (in Chinese). Scientia Silvae Sinicae. 37: 94-98.

Little, S.N., and Shainsky, L.J. 1995. Biomass and nutrient distribution in Central Oregon second-growth

ponderosa pine ecosystems. U.S. Department of Agriculture, Forest Service, Pacific Northwest

Research Station, Portland, OR. Res. Pap. PNW-RP-481.

Liu, Q. 1998. A study on the chemical composition of Chinese-fir seeds (in Chinese). Scientia Silvae

Sinicae. 2: 92-95.

Liu, F., Luo, R., and Jiang, J. 1991. Quantities and effects of various phosphorus compounds in soil under

Chinese-fir plantation (in Chinese). Journal of Nanjing Forestry University 4: 7-12.

Ma, X.Q. 2000. Effect of rotation on maintaining the site productivity of Chinese-fir plantation (in Chinese).

Scientia Silvae Sinicae. 36: 47-52.

Ma, X.Q. 2001. Advance in researches on productivity decline of replanting Chinese-fir forests (in Chinese).

Journal of Fujian Forestry College. 21: 380-384.

Ma, X.Q., He, Z.Y., and Yu, X.T. 1995. The effects of different clearances on soil fertility of Chinese-fir

timber plantation (in Chinese). Scientia Silvae Sinicae. 31: 485-490.

24 / 41
Bi et al. (2007) 25/42

Ma, X.Q, Liu, A.Q., and Huang, B.L. 2000. A study on self-poisoning effects of Chinese-fir plantation (in

Chinese). Journal of Nanjing Forestry University. 24: 12-16.

May, B.M., and Attiwill, P.M. 2003. Nitrogen-fixation by Acacia dealbata and changes in soil properties 5

years after mechanical disturbance or slash-burning following timber harvest. For. Ecol. Manage. 181:

339-355.

Miles, J., 1985. The pedogenetic effects of different species and vegetation types and the implication of

successions. J. Soil Sci. 36: 571-584.

Minghe, L., Ritchie, G.A. 1999. Eight hundred years of clonal forestry in China: I. traditional afforestation

with Chinese-fir (Cunninghamia lanceolata (Lamb.) Hook.). New Forests 18: 131-142.

Morris, D.M., Kimmins, J.P., and Duckert, D.R. 1997. The use of soil organic matter as a criterion of the

sustainability of forest management alternatives: a modeling approach using FORECAST. For. Ecol.

Manage. 94: 61–78.

Pan, W.C., Tian, D.L., Lei, Z.X., and Kang, W.X. 1983. Studies on the nutrient cycling in the Chinese-fir

plantations: (II) Contents, accumulation rate and biological cycling of nutrient elements in the fast-

growing Chinese-fir forest in the hill regions (in Chinese). Journal of Central South Forestry Institute. 3:

1-12.

Rennie, P.J. 1955. Uptake of nutrients by mature forest growth. Plant and Soil. 7: 49-95.

Rennie, P.J. 1957. The uptake of nutrients by timber forest and its importance to timber production in

Britain. Quart. J. For. 51: 101-115.

Richardson, B. 1993. Vegetation management practices in plantation forests of Australia and New Zealand.

Can. J. For. Res. 23: 1989-2005.

Richardson, B., Vanner, A., Ray, J., Davenhill, N., and Coker, G. 1996. Mechanisms of Pinus radiata

growth suppression by some common forest weed species. N.Z. J. For. Sci. 26: 421-437.

Riegel, G.M., Miller, R.F., Krueger, W.C. 1992. Competition for Resources Between Understory

Vegetation and Overstory Pinus Ponderosa in Northeastern Oregon Ecol. Appl. 2: 71-85.

25 / 41
Bi et al. (2007) 26/42

Royo, A.A., and Carson, W.P. 2006. On the formation of dense understory layers in forest worldwide:

consequences and implications for forest dynamics, biodiversity, and succession. Can. J. For. Res. 36:

1345-1362.

Seely, B., Kimmins, J.P., Welham, C., and Scoullar, K. 1999. Defining stand-level sustainability. Exploring

stand-level stewardship. J. For. 97: 4–10.

Seely, B., Welham, C., and Kimmins, H. 2002. Carbon sequestration in a boreal forest ecosystem: results

from the ecosystem simulation model, FORECAST. For. Ecol. Manage. 169: 123–135.

Shao, J.F. 1992. Influences of continuous cropping on soil fertility and stand growth in Chinese-fir

plantation. In Research on Site Degradation of Timber Plantation (in Chinese). Edited by Sheng, W.T.

Science and Technology Press of China, Beijing. pp 87-92.

Sheng, W.T. 1992. Soil degradation and its control techniques for timber plantations in China. In. Sheng, W.

T., (ed). Research on Site Degradation of Timber Plantation (in Chinese). Science and Technology

Press of China, Beijing. pp 1-7.

Sheng, W.T., and Fan, S.H. 2002. Effect of characteristics of Chinese-fir on long-term site productivity (In

Chinese). Forest Research. 15: 629-636.

Stone, D.M., and Elioff, J.D., 1998. Soil properties and aspen development five years after compaction and

forest floor removal. Can. J. Soil Sci. 78, 51-58.

Tian, D. 1994. Study on bio-geochemistry cycling in Chinese-fir ecosystem. In Studies on Forest

Ecosystems in China (in Chinese). Edited by Forestry Ministry of China Publishing House of Northeast

Forestry University. Harbin. pp. 136-145.

Tian, D.L., and Zhao, K. 1989. Studies on the litter in a Chinese-fir plantation ecosystem: (I) Amounts,

composition and dynamics of litter (in Chinese). J. Central South Forestry Institute. 9(Sup): 38-43.

Tian, D.L., Zhu, X.P., Cai, B.Y., and Gao, Y.M. 1989. Studies on the litter in a Chinese-fir plantation

ecosystem: (II) Nutrient contents and decomposition rate of litter (in Chinese). J. Central South Forestry

Institute. 9(Sup): 45-55.

26 / 41
Bi et al. (2007) 27/42

Wei, X., Liu, W., Waterhouse, M., and Armleder, M. 2000. Simulation on impacts of different management

strategies on long-term site productivity in lodgepole pine forests of the central interior of British

Columbia. For. Ecol. Manage. 133: 217–229.

Wei, X., Kimmins, J.P., and Zhou, G.. 2003. Disturbance and the sustainability of long-term site

productivity in lodgepole pine forests in the central interior of British Columbia – an ecosystem

modeling approach. Ecol. Model. 164: 239-256.

Welham, C., Seely, B., and Kimmins, J.P. 2002. The utility of the two-pass harvesting system: an analysis

using the ecosystem simulation model FORECAST. Can. J. For. Res. 32: 1071–1079.

Wu, Z.L., (ed). 1984. Chinese-fir (in Chinese). China Forestry Publishing House. Beijing. 583 pp.

Xi, F., Dalun, T., and Wenhua, X. 2006. Effects of different management regimes for cutover areas on soil

carbon storage in Chinese-fir plantations. Front. For. China 1: 38-42.

Xiang, W.H., Tian, D.L., Yan, W.D., Kang, W.X., and Fang, X. 2003. Biomass and nutrient content of

vegetation on fallow site after clear-cutting of Chinese-fir plantation (in Chinese). Acta Ecologica

Sinica. 23: 695-702.

Xiao, W., Nie, D., and Zhang, J. 1999. Study on biomass and energy use efficiency of the stands of

Cunninghamia lanceolata in China. Forest Research. 3: 253-258.

Yan, W., Tian, D., and Jiao, X. 2003 A study on biomass dynamics and distribution of undervegetation in

the secondary generation of Chinese-fir plantation in Hui Tong (in Chinese). Forest Research. 16: 82-86.

Yang, C.D., Jiao, R.Z., and Tu, X.N. 1995. Effect of undergrowth in Chinese-fir on soil properties in the 5-

15 cm layer (in Chinese). Forest Research. 5: 514-519.

Yang, Y.S., Chen, G..S., and Huang, B.L., 2000. Variation in the soil water and nutrients between different

rotation stands of Chinese-fir (in Chinese). J. Nanjing For. Univ.. 24: 25-28.

Yang, Y.S., Chen, G.S., Guo, J.F., and Lin, P. 2004. Decomposition dynamics of fine roots in a mixed

forest of Cunninghamia lanceolata and Tsoongiodendron odorum in mid-subtropics. Ann. For. Sci. 61:

65-72.

27 / 41
Bi et al. (2007) 28/42

Yang, Y.S., He, Z.M., Qiu, R.H., Yu, X.T., and Huang, B.L. 1999. Growth pattern of 29-year-old Chinese-

fir grown from seed in different rotations (in Chinese). Scientia Silvae Sinicae. 35: 32-36.

Yang, Y.S., Qiu, R.H., He, Z.M., Yu, X.T., and Huang, B.L. 1998. The stand net productivity and

biological cycle of nutrients in the 29-year-old plantations of Chinese-fir on different rotations (in

Chinese). Scienta Silvae Sinicae. 34: 3-11.

Yao, M.H., Sheng, W.T., and Xiong, Y.Q. 1992. Effect of understory vegetation on soil fertility in Chinese

fir plantation. In Research on site degradation of timber plantation (in Chinese). Edited by Sheng, W.T.

Beijing, Science and Technology Press of China. pp 161-167.

Ye, J.Z. 1992. Effects of controlling burning on physical and chemical properties of soil. In Research on

site degradation of timber plantation (in Chinese). Edited by Sheng, W.T. Science and Technology

Press of China, Beijing. pp 130-134.

Yu, X. T. (ed). 1997. Silviculture of Chinese-fir (in Chinese). Science and Technology Press of Fujian,

Fuzhou, Fujian Province.

Yu, X.T., and Yang,Y.S. 1989. Effect of controlled burning on artificial ecosystem of Chinese fir forest (in

Chinese). Journal of Fujian Forestry College. 9: 343-349.

Yu, X.T., and Zhang, Q.S. 1989. Studies on biochemical properties and fertilities of soils under successive

plantation of Chinese fir (in Chinese). J. Fujian Forestry College. 9: 263-271.

Zhang, X.Q., Kirschbaum, M.K.U., Hou, Z., and Guo, Z. 2004. Carbon stock changes in successive

rotations of Chinese-fir (Cunninghamia lanceolata (Lamb.) Hook) plantations. For. Ecol. Manage. 202:

131-147.

Zheng, W. 1958. Investigating report of the experience for high-yielded forests of Chinese-fir in Xihou

Village of Nanping City, Fujian Province (in Chinese). J. Nanjing For. College. 2: 59-82.

Zheng, Y., and Ding, Y. 1997. Properties of soil under the mixed forests of Tsoong’s tree and Chinese-fir

(in Chinese). Journal of Nanjing Forestry University. 4: 31-36.

Zheng, Y., and Ding, Y. 1998. Effect of mixed forests of Chinese-fir and Tsoong’s tree on soil properties.

Pedosphere. 2: 161-168.

28 / 41
Bi et al. (2007) 29/42

Zhong, A.L., and Hsiung, W.Y. 1993. Evaluation and diagnosis of tree nutritional status in Chinese-fir

(Cunninghamia lanceolata (Lamb) Hook) plantations, Jiangxi, China. For. Ecol. Manage. 62: 245-270.

Zhou, X. (ed). 1994. Long-term research on China’s forest ecosystems (in Chinese). Northeast Forestry

University Press, Harbin.

Zhou, H. 1999. Study on productivity of mixed forest planted on site of continuously planted Chinese-fir (in

Chinese). Journal of Fujian College of Forestry. 2: 165-169.

Zhou, X., Luo, R., and Ye, J. 1991. Effect of continuous cropping with Chinese-fir upon soil nutrients and

its feedback (in Chinese). Journal of Nanjing Forestry University. 3: 47-52.

29 / 41
Bi et al. (2007) 30/42
Table 1. Yield and nutrients reduction in second and third rotations, compared to the first rotation in Chinese fir plantation (-: no available data).

2ndrotation 3rd rotation

Location Xihou, Fujian1 Xihou, Fujian2 Youxi, Fujian Xihou, Fujian1 Xihou, Fujian2 Youxi, Fujian

Trees/ha 1800 - 2250 1845 - 2048 - 1800 - 2250 1845 - 2048 -

Stand age 20 29 20 20 29 20

Height decrease % 22.4 18.4 7.9 - 11.8 40.6 26.4 17.8 - 22.4

DBH decrease % 10.6 16.1 7.3 - 11.6 31.4 20.5 15.6 - 21.5

Volume decrease % 14.8 22.8 21.0 - 24.6 66.3 46.6 38.7 - 44.2

Humus decreased % 15.1 8.5 - 31.8 17.6 -

Total N decease % 10.8 17.6 - 11.8 10.4 -

Available P decrease % 33.1 1.8 - 56.2 23.6 -

Available K decrease % 12.6 0 - 21.3 2.2 -

Reference Shao (1992) Yang et al. (1999) Fan et al. (2000) Shao (1992) Yang et al. (1999) Fan et al. (2000)
1
Soil sampling from a layer of 50 cm.
2
Soil sampling from a layer of 40 cm.

30 / 41
Bi et al. (2007) 31/42
Table 2. Relative change (in % referred to the starting conditions at year 0. Negative numbers mean net decreases in the pool) in selected soil

properties for the poor site from the beginning to the end of 120-year simulation period for each scenario combination.

Site Slash Minor Rotation Litter Litter N Humus Humus N

quality burning vegetation length
Initial a Initial a Initial a Initial a
(y) (Mg ha-1) % (Kg ha-1) % (Mg ha-1) % (Kg ha-1) %
Poor  20 42.2 -55.1 263.4 -47.3 56.9 -37.8 1251.2 -26.3
 30 -36.6 -24.5 -27.0 -17.2
 40 -21.5 -8.7 -15.0 -7.3
 60 -6.0 13.9 5.6 9.5
  20 -79.0 -76.5 -40.4 -28.4
  30 -67.8 -56.8 -35.8 -24.4
  40 -52.5 -29.4 -26.6 -16.5
  60 -9.7 14.9 6.2 10.0
20 -23.7 -13.4 -22.9 -13.6
30 -2.1 8.6 -12.2 -4.7
40 14.5 22.5 -0.2 5.2
60 21.8 41.5 17.9 20.1
 20 -41.3 -21.8 -24.2 -14.7
 30 -6.2 11.8 -11.9 -4.5
 40 16.2 27.3 1.5 6.6
 60 27.9 47.1 19.2 21.1

Rich  20 72.3 -68.4 376.2 -53.9 67.3 -38.6 1491.8 -27.1

 30 -51.7 -31.7 -27.0 -17.5
 40 -35.1 -13.5 -15.4 -8.1
 60 -22.5 4.5 5.9 9.5
  20 -18.8 3.8 -15.1 -7.7
  30 -57.2 -34.5 -26.8 -17.3
  40 -40.1 -16.2 -14.1 -6.9
  60 -23.6 3.8 6.0 9.5
20 -26.0 -6.1 -16.8 -8.7
30 -5.4 16.2 -5.6 0.5
40 10.2 32.1 3.8 8.3
60 12.8 38.6 7.4 11.3
 20 -72.5 -57.1 -38.0 -20.4
 30 -3.1 19.1 -5.3 0.9
 40 12.6 34.7 3.8 8.2
 60 14.6 40.7 8.3 12.1
a
All simulations for the same site quality had the same starting conditions

31 / 41
Bi et al. (2007) 32/42

Caption for figures

Figure 1. The conceptual model of Chinese-fir decline. The results of the simulations support this

interpretation if regrowth of non-crop vegetation is simulated from sprouting as well as seed. The

simulations are more conservative than this conceptual model, probably because we failed to

represent post burning soil erosion.

Figure 2. Illustration of the key processes represented in the ecosystem simulation module within

FORECAST.

Figure 3. Illustration of the nutrient cycling pathways and major pools represented in FORECAST.

Nutrient cycling is based on a mass balance approach in which specific inputs to and outputs from the

ecosystem are tracked and quantified.

Figure 4. Impact of rotation length, minor vegetation development (panels A and C), slash burning

and combined effect of slash burning and minor vegetation development (panels B and D) on

Chinese-fir stemwood biomass harvested at the end of each rotation on poor and good sites over 120-

year simulations. “Control” is simulations of Chinese-fir with no minor vegetation. “Minor

vegetation” is full stocking of herbs and shrubs as given in the setup PLNTDATA file, growing from

seeds or vegetative resprouting, as appropriate. “Slash burning” is broadcast burning after harvesting

and before planting new Chinese-fir seedlings.

Figure 5. Height growth (upper panels) and foliage biomass (lower panels) during the first five years

of the first and second rotations for Chinese-fir, grass and shrubs. In the left panels minor vegetation

was simulated growing always from seeds. In the right panels, minor vegetation was simulated

growing from seeds or from rhizomes, as indicated.

32 / 41
Bi et al. (2007) 33/42

Figure 6. Effect of the presence or absence of herbs and shrubs, with and without slash burning, for

eight successive 20-year rotations on a poor site. Yield decline with slash burning and minor

vegetation is significantly greater than decline in soil fertility. After three rotations with burning, the

plantation has essentially failed (less than 30 Mg ha-1 of stemwood in the 4th rotation), and after four

cycles of the simulated disturbance results in a shrub disclimax. The lack of erosion in these

simulations renders these predictions conservative.

Figure 7. Details of the minor vegetation total biomass with different rotations on good and poor sites,

with and without slash burning over a 120 year simulation period. Slash burning greatly exacerbates

the growth of minor vegetation, leading to a switch from regrowth from seed to regrowth from

sprouts, with the consequences for Chinese-fir growth shown in Figure 4.

Figure 8. Predicted vs. observed (mean ± standard deviation) yield decline in Chinese-fir plantations

in two different types of sites (poor or rich) and two different rotation lengths (A) 20 years and (B) 30

years in percentage relative to the first rotation yield. (C) Observed and predicted organic matter at

the end of the 2nd and 3rd 20-year rotations, relative to the 1st rotation. (D) Observed and predicted

C:N ratio at the end of the 2nd and 3rd 20-year rotations, relative to the 1st rotation Observed data

calculated from Zhang et al. (2004).

33 / 41
Bi et al. (2007) 34/42

Figure 1. (Bi et al.).

34 / 41
Bi et al. (2007) 35/42

FOLIAGE
NITROGEN
CONTENT

Maximum
foliage
biomass set
by available moisture
PHOTOSYNTHETIC
EFFICIENCY

NET
PRIMARY
AVAILABLE PRODUCTION
AVAILABLE
SOIL LIGHT
NUTRIENTS

ALLOCATION

ROOTS BRANCHES STEMS FOLIAGE

LITTER
CO2 CO2

HUMUS
Active Passive

Fig 2. (Bi et al.)

35 / 41
Bi et al. (2007) 36/42

Input

Precipitation
Inputs
Fertilizer
Inputs Input

Nutrient
Input Uptake Upslope
Seepage Input
Plant Available
Biological
N2 Fixation Biomass Internal Soil
Cycling Mineral
Nutrients Weathering
Input

Foliar
Leaching Soil
Leaching

Natural
Herbivory
Mortality Loss
Decomposition Fire Loss
Harvest Litterfall

Loss

Litter and Soil Organic Matter Site Prep

Loss

Fig. 3. (Bi et al.)

36 / 41
Bi et al. (2007) 37/42

Fig. 4. (Bi et al.)

37 / 41
Bi et al. (2007) 38/42

Figure 5. (Bi et al.)

38 / 41
Bi et al. (2007) 39/42

Figure 6. Bi et al.

39 / 41
Bi et al. (2007) 40/42

Fig. 7. (Bi et al.).

40 / 41
Bi et al. (2007) 41/42

Fig. 8 (Bi et al.).

41 / 41