Using optimization for fitting individual-tree growth models for uneven-aged stands

A convenient model type for simulating the dynamics of uneven-aged and uneven-sized stands of Finland is individual-tree model. This is because the stand structures are complex due to the presence of several tree species and irregular size distributions of trees. The required minimum set of models in this approach consists of species-specific individual-tree diameter increment models, individual-tree survival models, and ingrowth models. The development of these models needs data in which the diameter and survival of each tree of the sample plots is known for at least two time points. For this, the trees need to be numbered, which is tedious in uneven-aged forests due to the great number of small trees and the continuous ingrowth process. This study proposes a modelling approach that fits the above models but requires only the diameter distributions of the plots in the beginning and at the end of the measurement interval. The method uses non-linear optimization to derive such values for model parameters that, when the models are applied to the initial diameter distribution, the simulated stand development results in a diameter distribution which agrees with the measured ending distribution. The study showed that the method produces similar models and model parameters as regression analysis. Since the method is less demanding in terms of modelling data, it brings new data sets available for modelling the dynamics of uneven-aged stands and reduces the cost of collecting new data. The models fitted by the proposed optimization method were rather similar to the models developed earlier for Finnish uneven-aged forests.     

浙江省毛竹直径与年龄的二元Weibull分布模型

利用2004年复查的浙江省森林资源连续清查体系4250个固定样地中的245个毛竹纯林样地数据,用具有严格概率模型性质的二元Weibull分布模型描述浙江省毛竹直径与年龄的二维结构,并讨论二元Weibull分布模型的生成函数,给出几个不同形式的表达式,指出模型良好的概率模型性质。一元Weibull分布模型的直径、年龄模型分别是二元模型的边际分布,而且用边际分布求得的参数与用联合分布求得的参数也很接近。本研究是省级地域的毛竹宏观模型,该模型也可尝试应用到林分水平,以及其他类型森林。     

西藏天然林直径分布模型及其应用研究

以西藏自治区2006年森林资源连续清查的479个天然林实测样地资料为基础,利用相对直径和累积概率的概念,借鉴相对树高曲线模型的建模方法,研究建立了冷杉、云杉、柏木、松类、栎类、桦类、阔叶类、针阔类等8个树种（组）的直径分布累积概率模型。通过与常用于天然异龄林直径分布结构的Weibull分布函数的拟合结果进行对比分析,表明所建模型效果良好。所建的直径分布累积概率模型为预估林分直径分布结构提供了参考依据,尤其为2006年西藏森林资源连清遥感样地和目测样地的样木模拟提供了定量依据。     

A comparative analysis of spatial, temporal, and ecological characteristics of forest fires in seasonally dry tropical ecosystems in the Western Ghats, India

The Western Ghats in India is one of the 25 global hotspots of biodiversity, and it is the hotspot with the highest human density. This study considers variations in the regional fire regime that are related to vegetation type and past human disturbances in a landscape. Using a combination of remote sensing data and GIS techniques, burnt areas were delineated in three different vegetation types and various metrics of fire size were estimated. Belt transects were enumerated to assess the vegetation characteristics and fire effects in the landscape. Temporal trends suggest increasingly short fire-return intervals in the landscape. In the tropical dry deciduous forest, the mean fire-return interval is 6 years, in the tropical dry thorn forest mean fire-return interval is 10 years, and in the tropical moist deciduous forest mean fire-return interval is 20 years. Tropical dry deciduous forests burned more frequently and had the largest number of fires in any given year as well as the single largest fire (9900 ha). Seventy percent, 56%, and 30% of the tropical moist deciduous forests, tropical dry thorn forests, and tropical dry deciduous forests, respectively have not burned during the 7-year period of study. The model of fire-return interval as a function of distance from park boundary explained 63% of the spatial variation of fire-return interval in the landscape. Forest fires had significant impacts on species diversity and regeneration in the tropical dry deciduous forests. Species diversity declined by 50% and 60% in the moderate and high frequency classes, respectively compared to the low fire frequency class. Sapling density declined by ca. 30% in both moderate and high frequency classes compared to low frequency class. In tropical moist deciduous ecosystems, there were substantial declines in species diversity, tree density, seedling and sapling densities in burned forests compared to the unburned forests. In contrast forest fires in tropical dry thorn forests had a marginal positive effect on ecosystem diversity, structure, and regeneration.     

Height–diameter models for tropical forests on Hainan Island in southern China

Measurements on 8352 felled trees from 31 plots in tropical forests on Hainan Island provided information to investigate the relationship between total tree height and diameter at 1.3 m (DBH). Out of 33 model forms either published or derivable from published forms, a modification of the exponential model published by Meyer (Meyer, H.A. A mathematical expression for height curves. J. For. 38 (1940) 415–420) was selected based on low bias and relatively good precision of its predictions on individual plots. Variations from plot-to-plot were significant. Therefore, a parameter prediction method based on predictor variables derived from all DBH values on a plot was developed in order to provide a method to localize the model. In a test with independent data from an additional 36 plots, the localized height-prediction model produced predictions with a pooled mean-squared error of 3.8 m², a pooled R² of 0.72, and an average bias of −0.2 m. Even though tremendous variability exists in tropical forests and modeling height–DBH relationships never provides high levels of precision, this model will provide a useful and economical alternative to attempting height measurements during inventories of these forests.     

Characteristics of typhoon disturbed gaps in an old-growth tropical montane rainforest in Hainan Island,China

Disturbances that create gaps can shape the structure and function of forests. However, such disturbance regimes in Asian tropical montane rainforests remain largely unquantified. Least studied are typhoon disturbances that are attributable to climate change. We investigated gap characteristics in terms of size, age, and gapmaker to quantify the gap disturbance regimes in an intact old-growth tropical montane rainforest on Hainan Island,China. The intensity of typhoons has increased since 1949,and typhoon winds blow mostly(45.5%) from the northeast corner of Hainan Island, resulting in a higher frequency of gaps in the northeast. A total of 221 gap-makers(trees that fell to create canopy gaps) and 53 gaps were observed in a 3.16 ha old-growth rainforest. Most canopy gaps(85%)were 200 m~2. The average size of canopy gaps was smaller in the rainforest than in other tropical forests, while the average size of expanded gaps was similar to those in other tropical forests. The maximum age of gaps was 23.5 years indicating that gaps had more rapid turnover than other parts of tropical forests. The frequency distribution of gap-makers followed a lognormal distribution with a distinctive peak at three gap-makers, which was different from the inverse J-shaped curve typical of other tropical forests. Gaps were recorded mainly on slopes between 20° and 35° and wood density of gap-makers was between 0.6 and 0.7 g cm~(-3). Our results suggest that small-scale disturbance was the dominant agent of gap formation in this old-growth rainforest that is subject to increasing typhoon disturbances.     

长白山过伐云冷杉恢复林分主要树种径阶生长与枯损模拟

以长白山过伐林区金沟岭林场的云冷杉林4个局级固定样地连续12年的观测数据为研究对象,利用固定样地内主要针叶树种红松、冷杉和云杉,从1978年到1984年6年内的胸径与定期平均生长量对应值数据,建立林木径阶生长转移概率模型,预估林木径阶平均生长量,并利用1990年观测数据进行检验,结果表明:所建概率模型实际应用误差较小,精度较高;同时还分析了1978年至1990年12年间云冷杉林的枯损林木株数分布特征,通过模型模拟和检验,表明Weibull分布函数适用于异龄混交林的枯损株数分布模拟。     

广州城区公园风景游憩林径级和高度级结构分布

【目的】探明广州风景游憩林群落结构现状以及存在的问题,进一步为城市风景游憩林群落结构调整和配置优化提供理论依据。【方法】以900 m^2样方为单元分别在广州发展公园风景游憩林群落样地设置样方15个,华南农业大学风景游憩林群落样地设置样方25个,火炉山森林公园风景游憩林群落样地设置样方20个,采用群落生态学的方法,研究了不同生态位风景游憩林径级和高度级结构分布。【结果】1)3个样地风景游憩林径级结构呈现一定差异:广州发展公园以大、中径级乔木占绝对优势,华南农业大学样地风景游憩林小径级乔木相对优势明显,火炉山样地风景游憩林群落以中、小径级乔木占绝对优势。2)根据群落中个体的胸径和树高呈现不同的相关关系表明:广州发展公园样地风景游憩林大径级树木开始出现衰退现象;华南农业大学样地风景游憩林由于存在大量小径级的速生树种,导致胸径—树高曲线在初期迅速增高;火炉山样地风景游憩林结构配置拟自然化,胸径—树高曲线增长较为平稳。【结论】根据高度级比重、相对频度和分布均匀度分析结果得出:广州发展公园样地风景游憩林中、大高度级树木比重较大,且分布更为集中,华南农业大学样地风景游憩林的小、中高度级的树木所占比重较高,小高度级树木分布更为集中;火炉山样地风景游憩林小高度级比重,且分布更为集中。建议通过人工择伐、补植特定树种的方式,调控城市风景游憩林群落的径级和高度级结构分布,促进风景游憩林的逐步稳定。     

Fitting irregular diameter distributions of forest stands by Weibull, modified Weibull, and mixture Weibull models

Irregular diameter frequency distributions of forest stands include multimodal structure of mixed-species stands, highly skewed and highly irregular shapes of uneven-aged stands, and rotated sigmoid form of old-growth stands. In this study, a traditional two-parameter Weibull model, a modified two-parameter Weibull model, and a finite mixture of two-parameter Weibull models were used to fit four artificial example plots. The model fitting and comparison results indicate that the mixture Weibull model is more flexible to fit various irregular diameter distributions, while the traditional Weibull model fails in every case to adequately describe these frequency distributions. The modified Weibull model is a good choice for fitting the “rotated-sigmoid” diameter distribution of an uneven-aged old-growth stand. However, it may not be sufficient when a diameter frequency distribution is multimodal or highly irregular in shape.     

Modeling the spatial structure of topical forests

The spatial structure of tropical forest stands under different management conditions was modeled as a series of different spatial point processes. Spatial patterns were first assessed by K-function analyses to help choose a point process appropriate for observed patterns. The homogenous Neyman–Scott process accurately described live tree distribution in clear cut areas, where tree patterns tended to be aggregated. Parameters were estimated by minimizing Diggle's modified least squares criterion, and goodness-of-fit was assessed by comparison to confidence envelopes constructed by Monte Carlo simulation. Parameter estimates can be interpreted to help understand the ecological processes influencing re-colonization of disturbed areas. The inhomogeneous Poisson process was investigated for simulating the spatial pattern of ingrowth trees in lower canopy strata. The intensity function of this process was inversely proportional to variables representing canopy density. As assessed by Monte Carlo generation of confidence envelopes, the inhomogeneous Poisson process successfully portrayed the influence of canopy structure on understory plant distribution in most stands. Tree mortality was modeled as a thinning process in which the probability of individual tree mortality was conditional on subject tree attributes and competitive environment. The thinning function took the form of a generalized linear model with a binomial error distribution and logit link function. In most stands, tree neighborhood variables were powerful predictors of mortality, but they were not important predictors in all plots. This suggests that the surrounding forest structure of a subject tree has considerable influence on its morality, but competition is not the sole cause of tree morality in tropical forests.     

Relationships between diameter and height of trees in natural tropical forest in Tanzania

The relationship between tree height (h) and tree diameter at breast height (dbh) is an important element describing forest stands. In addition, h often is a required variable in volume and biomass models. Measurements of h are, however, more time consuming compared to those of dbh, and visual obstructions, rounded crown forms, leaning trees and terrain slopes represent additional error sources for h measurements. The aim of this study was therefore to develop h–dbh relationship models for natural tropical forest in Tanzania. Both general forest type specific models and models for tree species groups were developed. A comprehensive data set with 2 623 trees from 410 different tree species collected from a total of 1 191 plots and 38 sites covering the four main forest types of miombo woodland, acacia savanna, montane forest and lowland forests was applied. Tree species groups were constructed by using a k-means clustering procedure based on the h–dbh allometry, and a number of different non-linear model forms were tested. When considering the complexity of natural tropical forests in general and in particular variations of h–dbh relationships due to high species diversity in such forests, the model fit and performance were considered to be appropriate. Results also indicate that tree species group models perform better than forest type models. Despite the fact that the residual errors level associated with the models were relatively high, the models are still considered to be applicable for large parts of Tanzanian forests with an appropriate level of reliability.     

Liana abundance in a Puerto Rican forest

Liana (woody vine) abundance varies among tropical forests and is often high in disturbed forests. In two areas of subtropical wet forest in Puerto Rico, El Verde and Bisley, we recorded the density of liana stems ≥1 cm dbh, and the percent of tree crowns (trees ≥10 cm dbh) that lianas infested. Both study areas have been disturbed by hurricanes several times in the past century; however, sample plots in each area were divided between plots that were less disturbed and those that were more disturbed, by both hurricanes and humans. The mean density and basal area of liana stems at El Verde were significantly higher in the less disturbed plots than in the more disturbed plots. The percent tree crown infested by lianas was higher on certain tree species and on larger trees, both of which characterized the less disturbed forest. Results at Bisley were similar to those at El Verde. Liana density and tree crown infestation in these Puerto Rican forests were low compared with most other tropical forests, contrasting especially with high values in other disturbed forests. Liana abundance varies among forests for complex reasons, including differences in disturbance, biogeography, seasonality, and tree host features.     

Forest dynamics after selective timber harvesting in Papua New Guinea

Forest dynamics after timber harvesting is a major issue for tropical forest managers and communities. Timber harvesting provides income to communities and governments and resources to industry but it has also been identified as a potential contributor to deforestation and degradation of tropical forests. In Papua New Guinea (PNG) harvesting is primarily occurring in accessible primary forests however, the fate of these forests under current harvesting practices is poorly understood.In this study we investigated the impacts of selective harvesting on stand structure, growth and dynamics, recovery and degradation, and species diversity. We also assessed the impacts of forest fire after the 1997-98 El Nino on basal area (BA) growth and mortality rates of natural tropical forests in PNG. For this study we used data from 118 (105 in selectively harvested and 13 in un-harvested forest), one-hectare permanent sample plots distributed across the country and measured for over 15 years by the PNG Forest Research Institute (PNGFRI). We analysed data from 84 of these plots in harvested forest to examine temporal trends in stand condition following harvesting. Mortality rates were investigated in 10 of the 21 plots in harvested forest that were burned during the 1997-98 El Nino drought with sufficient data for analyses. We tested a model developed in Queensland tropical forests to determine whether or not a critical threshold residual BA existed for the recovery of harvested tropical forests in PNG. Results from a logarithmic regression analysis of the relationship between starting BA (BA at first census) and stand BA increment after selective harvesting showed a positive increase in BA growth (r² = 0.74, p < 0.05). However, there was no critical threshold in residual BA that determined whether a harvested forest was likely to degrade or recover BA growth after harvesting. Our analyses suggested that the response to harvesting was variable, with the majority of un-burned plots (75%) showing an increase in BA and remainder a decrease. Average BA of selectively-harvested tropical forests was about 17 m² ha⁻¹ ± 4.17 (SD). Average annual increment in BA across the 84 un-burned plots was 0.17 m² ha⁻¹ year⁻¹ ± 0.62 (SD). Thus these forests generally show capacity to recover after selective harvesting even when the residual BA is low. A proportion of the BA increment is made up of non-commercial pioneer species that originate in significant gaps after harvesting. On burned plots, BA is affected by high mortality rates. The fate of these forests will depend on the degree of future harvesting, potential conversion to agriculture and the impact of fire and other disturbances.     

A multi-criterion index for the evaluation of local tropical forest conditions in Mexico

Despite the ecological and economical importance of tropical forests they are currently affected by human activities, mainly through deforestation and selective extraction. With the aim of making an opportune diagnosis of the condition of forests, we developed an ecological index based on qualitative and semi-quantitative data, allowing a quick diagnosis in order to manage and conserve tropical forests. We evaluated 44 plots of tree vegetation, measuring canopy height, number of strata, tree cover, dominant trees, number of tree species, as well as the management of and damage to the forest. The data of each parameter was classified in categories of 3, 4 or 5, which were normalized between 0 and 1 for the worst and best characteristics, respectively. For the purpose of analysis, the average, a set of IF–THEN rules, and fuzzy logic were applied and as a result we obtained a model that measures the ecological condition of the tropical forests. The model has the advantages of having an ecological basis, allows data to be gathered quickly and is clear and easy to manage and interpret. When running the model, the value of each intermediate variable is displayed, thus allowing the detection of where necessary action is required to improve the ecological condition of the forest. We expect this index to contribute in evaluating the effectiveness of forest management and possibly offer advice for the short-term management and conservation of the remnants of tropical forests.     

Modelling and mapping the suitability of European forest formations at 1-km resolution

Proactive forest conservation planning requires spatially accurate information about the potential distribution of tree species. The most cost-efficient way to obtain this information is habitat suitability modelling i.e. predicting the potential distribution of biota as a function of environmental factors. Here, we used the bootstrap-aggregating machine-learning ensemble classifier Random Forest (RF) to derive a 1-km resolution European forest formation suitability map. The statistical model use as inputs more than 6,000 field data forest inventory plots and a large set of environmental variables. The field data plots were classified into different forest formations using the forest category classification scheme of the European Environmental Agency. The ten most dominant forest categories excluding plantations were chosen for the analysis. Model results have an overall accuracy of 76%. Between categories scores were unbalanced and Mesophitic deciduous forests were found to be the least correctly classified forest category. The model’s variable ranking scores are used to discuss relationship between forest category/environmental factors and to gain insight into the model’s limits and strengths for map applicability. The European forest suitability map is now available for further applications in forest conservation and climate change issues.     

Characterising the diameter distribution of Sal plantations by comparing normal,lognormal and Weibull distributions at Tilagarh Eco-park,Bangladesh

For many years foresters have been using statistical probability density functions to describe and characterise stand structure. Predicting the current and future yields of a stand is essential for successful stand and timber management. Implicit prediction of current yield is accomplished by using diameter distribution methods. All diameter distribution yield systems predict the number of trees per unit area by diameter class. In this study, the normal, lognormal and the three-parameter Weibull probability density function were compared to characterise the diameter distributions of Sal (Shorea robusta) plantations grown at Tilagarh Eco-park, Bangladesh. Data from 70 plots, established in three plantations, were used for this study. The Weibull parameters were estimated by the maximum likelihood and moments estimator methods. A one-sample Kolmogorov–Smirnov test was used for the goodness of fit for all models. The Kolmogorov–Smirnov test results showed that both lognormal and Weibull distributions were suitable to characterise the diameter distributions of Sal plantations in the study area and may be applicable for other Sal forests in Bangladesh.     

Modelling the Growth and Management of Mixed Uneven-aged Maritime Pine - Broadleaved Species Forests in Galicia,North-western Spain

The state of mixed forests of maritime pine ( Pinus pinaster Ait.) and broadleaved species in the coastal region of Galicia (north-western Spain) was described using data from 213 circular sample plots selected among the available 4700 plots of the Spanish National Inventory. A matrix model was developed for this forest type to obtain information about the productivity and potential for sustainable management. The broadleaved species had a diameter distribution close to the inverted J-shape typical of the uneven-aged forests but for maritime pine there were many medium-sized trees and a lack of trees in the first diameter class of 15 cm. The matrix growth model was used to predict the development of mixed forests for three different management options: no harvesting in a well-stocked stand, a regime with a 5 yr harvest cycle and a residual basal area of 15 m 2 ha -1 in a well-stocked stand, and a rehabilitation management applied to an understocked stand. The results showed the possibility of applying uneven-aged silvicultural systems to these forests, leading to the production of high-value timber and to the maintenance of continuous cover and a biodiverse forest.     

不同密度落叶松人工林直径结构研究

文章利用汪清林业局金钩岭林场落叶松人工林不同密度固定样地的长期观测数据,采用偏度、峰度、直径变动系数对林分直径结构动态变化规律进行了探讨,结果表明:①25a时偏度值随密度增加而增大,其后随着年龄增加,各样地的偏度值均逐渐减小,直径分布曲线右偏型逐渐减弱,期初密度大的样地偏度值变化明显;②随着年龄的增加,不同密度样地的峰度值逐渐减小,直径分布曲线逐渐变得平缓,期初密度大的样地尤其明显;③31a后,各样地直径变动系数稳定在0.36附近。     

Structure and composition of the understory treelets in a non-dipterocarp forest of tropical Asia

Species richness, floristic composition, and structure of understory treelets were investigated in three 1-ha plots, distributed in the undisturbed tropical seasonal rain forest of Xishuangbanna, SW China. We investigated all the woody species except lianas in these plots, compared the differences of species diversity of trees and treelets in these plots and the differences of understory structure between this forest and the typical dipterocarp forests in SE Asia. We found 5089 individuals belonging to 356 species, 189 genera, and 63 families in the three plots. The five most important families were Rubiaceae, Euphorbiaceae, Lauraceae, Meliaceae, and Annonacae. The treelet layer was much more diverse than the tree layer, indicating that result based only on trees may be not an appropriate representation of the diversity status of a particular tropical forest type. The three plots were more similar when the treelet layer was considered than when tree layer was considered. In contrast with the dipterocarp forests in SE Asia, the understory of tropical seasonal rain forest was mainly composed of the trees with small and middle stature at maturity classes. These results indicate that the structure of tropical forest understory may differ not only among continents but also within continent.