Tomas Lämås

Clustering of harvest activities in multi-objective long-term forest planning

Models used in long-term forest planning were generally, until recently, non-spatial. The locations of harvest activities were considered first in lower, more short-term steps of the planning hierarchy. However, now that issues related to biodiversity, recreation and road planning have to be considered, this is no longer a viable option. The spatial arrangement of harvest activities affects parameters such as the proportion of undisturbed interior forest and the sites of new roads. Thus, in long-term planning the spatial location of harvesting operations needs to be taken into consideration. However, including spatiality in long-term planning complicates the planning problems, and requires the development of new methods and approaches. This study presents a new approach for clustering harvest activities in time and space in long-term forest planning. The planning problem essentially consists of maximizing the weighted sum of the net present value of future forest management and the clustered volume of timber to be harvested. This objective is subject to the restriction that a certain volume should be harvested each period. Since the spatial dimension leads to a problem that is difficult to solve with ordinary optimization techniques, the ensuing problem is solved with a heuristic technique called simulated annealing. In a case study the suggested approach is applied to a landscape consisting of 2600 stands in southern Sweden. The results indicate that the model is effective for clustering the harvest and that it is possible to aggregate the harvest with only a small sacrifice of the net present value.

The influence of forest data quality on planning processes in forestry

Abstract Forestry planning relies on decision support systems for predicting and evaluating the outcomes of different management options. The accuracy of data describing the initial state of the forest will inevitably affect the projections of forest development and, hence, the quality of management decisions. As data acquisition is expensive, there is a trade-off between improving management plans and the increasing cost of gathering more data. This review attempts, from a critical perspective, to summarize research on how the quality of forestry data affects the results of forestry planning and decision making. It is structured as follows: first, the attempts to evaluate forestry data for forestry planning purposes are described, including methodology and main conclusions; secondly, some generic questions concerning how important aspects of this research area can be handled are posed and discussed; and finally, some conclusions are drawn, including methodological recommendations. It is concluded that research in this area is scarce, which is in contrast to the fields of forestry planning and forest data acquisition. The inherent complexity of forestry planning is probably the reason why those who seek to evaluate forestry data often oversimplify the problems.

Simulation of harvester productivity in selective and boom-corridor thinning of young forests

Abstract Forest management practices may change in the future, due to increases in the extraction of forest fuel in first thinnings. Simulation models can be used to aid in developing new harvesting systems. We used such an approach to assess the productivity of innovative systems in various thinnings of young stands with wide ranges of mean breast height diameter (1.5–15.6 cm), stems per hectare (1000–19,100), and mean height (2.3–14.6 m). The results show that selective multiple-tree-handling increases productivity by 20–46% compared to single-tree-handling. If the trees are cut in boom-corridors (10×1 or 2 m strips between strip roads), productivity increases up to 41%, compared to selective multiple-tree-handling. Moreover, if the trees are felled using area-based felling systems, productivity increases by 33–199%, compared to selective multiple-tree-handling. For any given harvesting intensity, productivity increased the most in the densest stands with small trees. The results were used to derive time consumption functions. Comparisons with time study results suggest that our simulation model successfully mimicked productivity in real-life forest operations, hence the model and derived functions should be useful for cost calculations and evaluating forest management scenarios in diverse stands.

The Legacy of Logging—Estimating Arboreal Lichen Occurrence in a Boreal Multiple-Use Landscape on a Two Century Scale

In northern Sweden, the availability of arboreal lichens (Bryoria fuscescens, Alectoria sarmentosa) as winter grazing resources is an important element in reindeer husbandry. With the industrialization of forestry, forests rich in arboreal lichens have diminished considerably. Here, we analyze how forestry has impacted lichen availability from the 1920s to the present day and model its future development assuming different forest management scenarios. We recorded the current occurrence of B. fuscescens in 144 sampling plots, stratified by forest age class and dominant tree species in a 26,600 ha boreal forest landscape that is used for both reindeer herding and forestry. Lichen abundance was visually estimated in four classes: none, sparse, moderate and abundant. A binary logistic model using forest age as the independent variable was developed to predict the probability of lichens being present. Using this model, we found that lichens were present in stands that are at least 63 years old. Because of the relative paucity of stands rich in arboreal lichens, it was not possible to reliably determine how age affects the variation in abundance of older forest stands. The historical development of forests where arboreal lichens could potentially occur was studied using historic forestry records dating back 80 years. Between 1926 and the present day, forestry has reduced the cover of forests older than 60 years from 84% to 34%. The likely future spatial coverage of these stands over the next 120 years was estimated for two different management scenarios and an unmanaged reference scenario, using the Heureka strategic planning program. Under both the “business as usual” scenario and that involving more intensive forestry, continued decreases in lichen availability are projected. Our results emphasize the importance of alternative forestry practices, such as prolonged rotation periods, to increase the availability of arboreal lichens as a grazing resource for reindeer.

Studying the use of forest management decision support systems: an initial synthesis of lessons learned from case studies compiled using a semantic wiki

In order to share information on the development and use of forest management decision support systems (FMDSS), a European-initiated network has established a wiki website as part of its activities. Case studies and associated lessons learned were solicited from the network using semantic structures built on the wiki. A total of 31 cases from 10 different countries and 80 associated lessons were entered into the wiki. The resulting lessons were categorized (non-exclusively) using four major themes: (1) decision support system architecture and design (40 lessons), (2) methods and models (26), (3) knowledge management processes (34) and (4) participatory processes (32). The semantic wiki proved useful for gathering case information and relating it to other information objects, such as FMDSS software descriptions; however, it was not as well suited to the task of analysis and synthesis as commercial qualitative analysis software packages. Future development possibilities for the semantic structures are suggested, and more cases are solicited from the FMDSS community.

Tree retention practices in boreal forests: what kind of future landscapes are we creating?

Tree retention practices promoting biodiversity may reshape future boreal forest production landscapes. Using the Heureka system, scenarios of 0%, 5%, and 20% retained patches at the stand level were projected over 200 years in a 533 ha boreal landscape. Visualizations of future forest states at a landscape scale and a more detailed scale were made based on the projections. The no retention results in no forest >120 years old, and no large trees (diameter at breast height >40 cm for conifers and >35 cm for broadleaved trees) 100 years from now. With retention levels of 5% and 20%, the area of old forest will comprise 7% and 19% of the total area, respectively. The average number of large trees per ha will be 4 and 13, respectively. Deadwood volumes will be 2.5 times higher at 5% retention and 4 times higher at 20% retention compared to no retention. Landscape visualizations indicate that retention patches covering 5% will marginally modify the visual impression, compared to clear-cuts, while 20% cover will create a much more varied landscape. We conclude that the retention approach is essential for restoring natural conditions. Landscape transformation will be slow and depend on starting conditions and retention levels.

A cost function estimating the loss due to extended rotation age

A method for estimating the economic loss in timber production resulting from extending the rotation age was studied. Basically, optimal rotation age depends on the stated objectives. If the objective comprises only timber production, the rotation age is derived from maximizing the net present value (PV) of timber production. If, on the other hand, the objective function includes, for example, the maintenance of biodiversity and scenic values, the optimal rotation age is likely to be extended. In the present study, PVs from successively extended rotation ages were estimated for a sample of economically mature stands. The PVs were based on no thinning allowed (case A), and thinning allowed in the first 5‐yr period (case B). By using regression analysis, functions were estimated that predict the inoptimality loss from extending the rotation age for both cases. Stand data collected by subjective (ocular) inventory methods were the regressors. Only the problem of economic loss in timber production from extend...

Estimating stem diameter distributions from airborne laser scanning data and their effects on long term forest management planning

Data obtained from airborne laser scanning (ALS) are frequently used for acquiring forest data. Using a relatively low number of laser pulses per unit area (≤5 pulses per m2), this technique is typically used to estimate stand mean values. In this study stand diameter distributions were also estimated, with the aim of improving the information available for effective forest management and planning. Plot level forest data, such as stem number and mean height, together with diameter distributions in the form of Weibull distributions, were estimated using ALS data. Stand-wise tree lists were then estimated. These estimations were compared to data obtained from a field survey of 124 stands in northern Sweden. In each stand an average of seven sample plots (radius 5–10 m) were systematically sampled. The ALS approach was then compared to a mean value approach where only mean values are estimated and tree lists are simulated using a forest decision support system (DSS). The ALS approach provided a better match to observed diameter distributions: ca. 35% lower error indices used as a measure of accuracy and these results are in line with the previous studies. Moreover – which is unique compared to earlier studies – suboptimal losses were assessed. Using the Heureka DSS the suboptimal losses in terms of net present value due to erroneous decisions were compared. Although no large difference was found, the ALS approach showed smaller suboptimal loss than the mean value approach.

Evaluating citizen science data for forecasting species responses to national forest management

Abstract The extensive spatial and temporal coverage of many citizen science datasets (CSD) makes them appealing for use in species distribution modeling and forecasting. However, a frequent limitation is the inability to validate results. Here, we aim to assess the reliability of CSD for forecasting species occurrence in response to national forest management projections (representing 160,366 km2) by comparison against forecasts from a model based on systematically collected colonization–extinction data. We fitted species distribution models using citizen science observations of an old‐forest indicator fungus Phellinus ferrugineofuscus. We applied five modeling approaches (generalized linear model, Poisson process model, Bayesian occupancy model, and two MaxEnt models). Models were used to forecast changes in occurrence in response to national forest management for 2020‐2110. Forecasts of species occurrence from models based on CSD were congruent with forecasts made using the colonization–extinction model based on systematically collected data, although different modeling methods indicated different levels of change. All models projected increased occurrence in set‐aside forest from 2020 to 2110: the projected increase varied between 125% and 195% among models based on CSD, in comparison with an increase of 129% according to the colonization–extinction model. All but one model based on CSD projected a decline in production forest, which varied between 11% and 49%, compared to a decline of 41% using the colonization–extinction model. All models thus highlighted the importance of protected old forest for P. ferrugineofuscus persistence. We conclude that models based on CSD can reproduce forecasts from models based on systematically collected colonization–extinction data and so lead to the same forest management conclusions. Our results show that the use of a suite of models allows CSD to be reliably applied to land management and conservation decision making, demonstrating that widely available CSD can be a valuable forecasting resource.

Assessment of Coarse Woody Debris

In this paper, different probability sampling methods for assessing coarse woody debris (CWD) are compared. The methods are circular plot sampling, strip surveying, line intersect sampling, and transect relascope sampling. The two latter methods are suited primarily for the assessment of CWD on the ground.