Improved wood species identification based on multi-view imagery of the three anatomical planes.

BACKGROUND: The identification of tropical African wood species based on microscopic imagery is a challenging problem due to the heterogeneous nature of the composition of wood combined with the vast number of candidate species. Image classification methods that rely on machine learning can facilitate this identification, provided that sufficient training material is available. Despite the fact that the three main anatomical sections contain information that is relevant for species identification, current methods only rely on transverse sections. Additionally, commonly used procedures for evaluating the performance of these methods neglect the fact that multiple images often originate from the same tree, leading to an overly optimistic estimate of the performance. RESULTS: We introduce a new image dataset containing microscopic images of the three main anatomical sections of 77 Congolese wood species. A dedicated multi-view image classification method is developed and obtains an accuracy (computed using the naive but common approach) of 95%, outperforming the single-view methods by a large margin. An in-depth analysis shows that naive accuracy estimates can lead to a dramatic over-prediction, of up to 60%, of the accuracy. CONCLUSIONS: Additional images from non-transverse sections can boost the performance of machine-learning-based wood species identification methods. Additionally, care should be taken when evaluating the performance of machine-learning-based wood species identification methods to avoid an overestimation of the performance.

Climatic Signals in Tree Rings of Heritiera fomes Buch.-Ham. in the Sundarbans, Bangladesh.

Mangroves occur along the coastlines throughout the tropics and sub-tropics, supporting a wide variety of resources and services. In order to understand the responses of future climate change on this ecosystem, we need to know how mangrove species have responded to climate changes in the recent past. This study aims at exploring the climatic influences on the radial growth of Heritiera fomes from a local to global scale. A total of 40 stem discs were collected at breast height position from two different zones with contrasting salinity in the Sundarbans, Bangladesh. All specimens showed distinct tree rings and most of the trees (70%) could be visually and statistically crossdated. Successful crossdating enabled the development of two zone-specific chronologies. The mean radial increment was significantly higher at low salinity (eastern) zone compared to higher salinity (western) zone. The two zone-specific chronologies synchronized significantly, allowing for the construction of a regional chronology. The annual and monsoon precipitation mainly influence the tree growth of H. fomes. The growth response to local precipitation is similar in both zones except June and November in the western zone, while the significant influence is lacking. The large-scale climatic drivers such as sea surface temperature (SST) of equatorial Pacific and Indian Ocean as well as the El Niño-Southern Oscillation (ENSO) revealed no teleconnection with tree growth. The tree rings of this species are thus an indicator for monsoon precipitation variations in Bangladesh. The wider distribution of this species from the South to South East Asian coast presents an outstanding opportunity for developing a large-scale tree-ring network of mangroves.

High-resolution proxies for wood density variations in Terminalia superba.

BACKGROUND AND AIMS: Density is a crucial variable in forest and wood science and is evaluated by a multitude of methods. Direct gravimetric methods are mostly destructive and time-consuming. Therefore, faster and semi- to non-destructive indirect methods have been developed. METHODS: Profiles of wood density variations with a resolution of approx. 50 µm were derived from one-dimensional resistance drillings, two-dimensional neutron scans, and three-dimensional neutron and X-ray scans. All methods were applied on Terminalia superba Engl. & Diels, an African pioneer species which sometimes exhibits a brown heart (limba noir). KEY RESULTS: The use of X-ray tomography combined with a reference material permitted direct estimates of wood density. These X-ray-derived densities overestimated gravimetrically determined densities non-significantly and showed high correlation (linear regression, R(2) = 0·995). When comparing X-ray densities with the attenuation coefficients of neutron scans and the amplitude of drilling resistance, a significant linear relation was found with the neutron attenuation coefficient (R(2) = 0·986) yet a weak relation with drilling resistance (R(2) = 0·243). When density patterns are compared, all three methods are capable of revealing the same trends. Differences are mainly due to the orientation of tree rings and the different characteristics of the indirect methods. CONCLUSIONS: High-resolution X-ray computed tomography is a promising technique for research on wood cores and will be explored further on other temperate and tropical species. Further study on limba noir is necessary to reveal the causes of density variations and to determine how resistance drillings can be further refined.