J.A. Trofymow

Derivation and validation of Canada-wide coarse-resolution leaf area index maps using high-resolution satellite imagery and ground measurements

Leaf area index (LAI) is one of the surface parameters that has importance in climate, weather, and ecological studies, and has been routinely estimated from remote sensing measurements. Canada-wide LAI maps are now being produced using cloud-free Advanced Very High-Resolution Radiometer (AVHRR) imagery every 10 days at 1-km resolution. The archive of these products began in 1993. LAI maps at the same resolution are also being produced with images from the SPOT VEGETATION sensor. To improve the LAI algorithms and validate these products, a group of Canadian scientists acquired LAI measurements during the summer of 1998 in deciduous, conifer, and mixed forests, and in cropland. Common measurement standards using the commercial Tracing Radiation and Architecture of Canopies (TRAC) and LAI-2000 instruments were followed. Eight Landsat Thematic Mapper (TM) scenes at 30-m resolution were used to locate ground sites and to facilitate spatial scaling to 1-km pixels. In this paper, examples of Canada-wide LAI maps are presented after an assessment of their accuracy using ground measurements and the eight Landsat scenes. Methodologies for scaling from high- to coarse-resolution images that consider surface heterogeneity in terms of mixed cover types are evaluated and discussed. Using Landsat LAI images as the standard, it is shown that the accuracy of LAI values of individual AVHRR and VEGETATION pixels was in the range of 50–75%. Random and bias errors were both considerable. Bias was mostly caused by uncertainties in atmospheric correction of the Landsat images, but surface heterogeneity in terms of mixed cover types were also found to cause bias in AVHRR and SPOT VEGETATION LAI calculations. Random errors come from many sources, but pixels with mixed cover types are the main cause of random errors. As radiative signals from different vegetation types were quite different at the same LAI, accurate information about subpixel mixture of the various cover types is identified as the key to improving the accuracy of LAI estimates. D 2002 Elsevier Science Inc. All rights reserved.

A comparison of digital and film fisheye photography for analysis of forest canopy structure and gap light transmission

Due to the scarcity and high cost of conventional film-based hemispherical photographic systems, some forest scientists are now using multi-purpose, consumer-grade digital cameras for the analysis of forest canopy structure and gap light transmission. Although the low cost of digital cameras and direct capture of digital images appear to offer significant advantages over film camera systems, relatively little is known about their technical differences from an applications perspective. In this study, we compared the performance of a popular digital camera (Nikon Coolpix 950 with FC-E8 fisheye) with a conventional film camera (Nikon F with Nikkor 8 mm fisheye) under different stand structures and sky conditions. Our findings show that the Nikon Coolpix 950 digital camera produced hemispherical canopy photos with substantial color blurring towards the periphery of the exposure. We believe that chromatic aberration associated with the camera’s lens optics may be the source of this phenomenon; however, other factors may have also contributed to the diminished image sharpness. Color blur influenced (i) the size, shape, and distribution of canopy gaps; (ii) the accuracy of edge detection and the binary division of pixels into sky and canopy elements, and (iii) the magnitude, range, and replication of canopy openness, leaf area, and transmitted global radiation results. The Nikon Coolpix 950 produced canopy openness measures that were 1.4 times greater than film estimates in 22 of the 36 photo pairs. Cloud cover and sky brightness also influenced the spectral characteristics of the lateral chromatic aberration (halos), and thus had an added and unpredictable effect on canopy openness. Setting the Nikon Coolpix 950 to record in black and white, and shooting only under uniformly overcast skies will help to minimize the unpredictable effects of chromatic aberration. Nevertheless, we recommend a cautious approach when undertaking canopy measurements with the Nikon Coolpix 950, particularly when stands are dense and canopy openness falls below 10%. High-quality (1:4) JPEG compression had no significant influence on mean canopy openness; however, lower XGA and VGA image resolutions

Patterns of Carbon, Nitrogen and Phosphorus Dynamics in Decomposing Foliar Litter in Canadian Forests

We examined the patterns of nitrogen (N) and phosphorus (P) gain, retention or loss in ten foliar tissues in a litterbag experiment over 6 years at 18 upland forest sites in Canada, ranging from subarctic to cool temperate. N was usually retained in the decomposing litter until about 50% of the original C remained. The peak N content in the litter was observed at between 72 and 99% of the original C remaining, with C:N mass quotients between 37 and 71 (mean 55). The rate of N release from the litters was not related to the original N concentration, which may be associated with the generally narrow range (0.59–1.28% N) in the litters. P was immediately lost from all litters, except beech leaves, with critical litter C:P mass quotients for P release being in the range 700–900. The rate of P loss was inversely correlated with the original litter P concentration, which ranged from 0.02 to 0.13%. The soil underlying the litterbags influenced the pattern of N and P dynamics in the litters; there were weak correlations between the N and P remaining at 60% C remaining in the litters and the C:N and C:P quotients of the surface layer of the soil. There was a trend for higher N and P retention in the litter at sites with lower soil C:N and N:P quotients, respectively. Although there was a large variation in C:N, C:P and N:P quotients in the original litters (29–83, 369–2122 and 5–26, respectively), and some variation in the retention or loss of N and P in the early stages of decomposition, litters converged on C:N, C:P and N:P quotients of 30, 450 and 16, when the C remaining fell below 30%. These quotients are similar to that found in the surface organic matter of these ecosystems.

Characterization of organic matter in a forest soil of coastal British Columbia by NMR and pyrolysis-field ionization mass spectrometry

Organic matter in the soil profile under a young Douglas-fir stand in coastal British Columbia was characterized by examining intact samples of fresh litterfall and organic horizons (LF, H), and fractions (floatables, humic acid [HA], fulvic acid [FA], humin [HU]) from the three mineral horizons (Ae, Bm, BC). Some 30–40% of the carbon in the mineral horizons was found in poorly-decomposed plant material floatable in water, a fraction whose characteristics changed little with depth, and which contained over 1% Fe. The proportion of soil C in HA plus FA was approximately 8%, but the ratio of C in FA/HA increased with depth. Solid-state 13C NMR spectra of litterfall, LH and H samples showed effects of decomposition, in particular a decrease in 0-alkyl C from litterfall to LH to H, and degradation of resolution from LF to H. For the mineral soil fractions, both floatables and de-ashed HU (‘HUd’ prepared by HCl/HF treatment) indicated high levels of the original plant biopolymers, including a large alkyl component. Solution 13C spectra of the HAs from mineral horizons showed little difference with depth, except that peaks due to lignin were more pronounced for the Bm HA. The NMR spectra of FAs were high in 0-alkyl and carboxyl C. Pyrolysis-field ionization mass spectrometry confirmed and extended the results from NMR and chemical analyses, in particular demonstrating the accumulation of suberin in some fractions and the leaching and decomposition of lignin components with increasing depth in the mineral horizons. The general features of the HA, FA and HUd from this forest soil, and the effects of decomposition and pedogenesis were similar to those widely found for agricultural and forest soils. However, the accumulation of suberin, and the leaching and decomposition of lignin are particularly associated with forest soils. The low proportion of soil C in HA and FA, and the high proportion in poorly decomposed, iron-rich plant fragments suggest that decomposition is somewhat limited at this site, which is classified as having low fertility. The high accumulations of alkyl C from suberin may also indicate, or contribute to inhibition of decomposition.

Influence of micro- and macro-habitat factors on collembolan communities in Douglas-fir stumps during forest succession

Abstract We studied the relative importance of micro- and macro-habitat factors on collembolan communities in decaying stumps in three Douglas-fir ecosystems. Each ecosystem contained four seres: regeneration (3–8 years old), immature (25–45 years old), mature (65–85 years old), and old-growth stands (over 200 years old). Stumps were classified, depending on stage of wood decay as: sound, moderate decay, and advanced decay. The relationship between collembolan communities and habitat factors was determined by canonical correspondence analysis (CCA) and detrended correspondence analysis (DCA). Macrohabitat factors (stand age and study site) were the most important ones affecting species distribution. When data were pooled for season and stump decay stages, collembolan communities were characteristically grouped by seres, particularly for the regeneration and old growth. However, only a few species were characteristic of a particular sere, notably Vertagopus alpa, Hymenaphorura cocklei, and Folsomia stella in old-growth forests, and Anurophorus septentrionalis and Ballistura libra in regeneration seres. In contrast to collembolan species, microhabitat factors in the stump were influenced more by season than by stand age or site. DCA-ordination indicated that within an individual season, microhabitat factors, especially %C, %P and microbial biomass, were important determinants of collembolan distribution. Common and abundant collembolan species tended to be positively correlated with %C, but negatively correlated with %N, %P, numbers of nematodes and microbial biomass. We suggest that collembolan numbers were not directly related to the first four factors and that the negative correlation with microbial biomass was caused by excessive grazing on fungi by the Collembola.

Abundance, species diversity, and community structure of Collembola in successional coastal temperate forests on Vancouver Island, Canada

Abstract The ecological implications of the conversion of old-growth temperate rainforests to managed forests have generated much discussion worldwide. This paper examines the effects of such a conversion on different aspects of the biodiversity of the soil collembolan fauna, and attempts to determine the time that will be required for the collembolan fauna to approach the abundance and community structure seen in old-growth forests. The study also investigates the potential of using different measures of species diversity and community structure as indicators of old-growth conditions in forest soils. The study was carried out in three chronosequence sites in Douglas-fir dominated stands on the dry leeward eastern side of Vancouver Island, BC, Canada. Each of the three sites contained stands representing four stages of stand development: regeneration (7–9 years), immature (35–46 years), mature (80–102 years) and old-growth (>248 years). The Collembola were extracted from litterbags containing needle litter or wood chips, and from the forest floor (LFH) layer in the late autumn of four successive years. Overall abundance of Collembola was highest in the old-growth and lowest in the regeneration stands. Although population numbers in the immature and mature forests were significantly higher than in regeneration stands, they still had not achieved the levels observed in old-growth forests. In the forest floor, species richness was low in regeneration stands compared to later stages of stand development, but did not differ significantly among immature, mature and old-growth stands. Measures of species diversity based on Shannon’s and Simpson’s indices of diversity did not differ significantly according to the stage of stand development. It was not possible to distinguish individual collembolan species that could be used as indicators of old-growth conditions. The same species occurred in most or all stand ages, with differences being determined by changes in relative and absolute abundance of the species comprising the community. However, principal component analysis of data on the Collembola of needle litterbags and the LFH layer showed that the collembolan community of the regeneration stands could be clearly differentiated from those of the forested stands. In addition, the collembolan communities of 80–102-year-old forests could still be distinguished from those of the old-growth forests. In contrast, the collembolan fauna of decomposing wood chips was very similar in all stand ages, with the exception of the regeneration stands.

Assessing bias from boles in calculating leaf area index in immature Douglas-fir with the LI-COR canopy analyzer

Measurements of leaf area index (LAI) taken with photometers such as the LAI2000 are biased by boles and branches which stop part of the light from reaching the photometer. We estimated how much of the sky would be obscurred by boles in an immature stand of Douglas-fir (Pseudotsuga menziesii (Mirb) Franco) that was stem-mapped and of known diameter, height and height-to-live-crown. These yielded gap fractions which were then converted to bole area index (BAI) via the algorithm used by the LAI2000. The mean BAI, as a fraction of the effective leaf area index (LAIe) calculated by the LAI2000, ranged from <0.01 for Douglas-fir at 24 years of age to about 0.1 at 48 years. Comparing readings on Douglas-fir trees defoliated by root rot with readings on fully foliated trees yielded a proportion of 0.41, which included branches. Measurements were also taken on clumps of maple (Acer macrophyllum Pursh) and alder (Alnus rubra Bong.) both before and after leaf-fall, yielding corresponding proportions, including branches, of 0.41 and 0.25, respectively. The values obtained this way are probably upper limits. The contributions of boles may be appreciable, particularly in mature unthinnned stands with closed canopies.

Functional role of Collembola in successional coastal temperate forests on Vancouver Island, Canada

Abstract The interrelationships among collembolan feeding groups (FG), decomposition rates, microbial biomass, and forest successional stage were studied at three sites in Douglas-fir dominated stands on the dry leeward side of Vancouver Island, BC, Canada. Each site contained a basic suite (or chronosequence) of four stages of stand development: old-growth (>248 years), regeneration (7–9 years), immature (35–46 years), and mature (80–102 years). Principal component analysis (PCA) was used to identify four collembolan feeding groups on the basis of gut contents. Three of the feeding groups (FG 1–3) were fungal-feeders, but differed from one another in the types of fungal material consumed. The last group (FG 4) comprised species that ingested particulate organic matter (POM). Significant positive correlations between microbial biomass and collembolan abundance were obtained for the fungal-feeding Collembola (FG 1–3), but not for the detritivorous FG 4. The proportional representation of collembolan species feeding on darkly pigmented fungi (FG 1) was reduced in the regeneration stands, where species feeding on particulate organic matter predominated. Over the 4 years of this study, the mass loss of decomposing wood chips in litterbags was not significantly influenced by the age of the forest in which the bags were placed. In contrast, the mass loss of needle litter was significantly reduced in regeneration stands. Multiple regressions based on the abundance of the different feeding groups of Collembola provided better estimates of the mass loss of decomposing litter or wood chips than did regressions based on total collembolan numbers.

Fate of urea and ureaformaldehyde nitrogen in a one-year laboratory incubation with Douglas-fir forest floor

Abstract In a 1 y laboratory incubation of a Douglas-fir forest floor (FH) the effects of two different kinds of organic N compounds, fast-release urea (U) and slow-release ureaformaldehyde (UF), on N transformations were studied. Compounds labelled with 15 N were used to follow the mineralization and distribution of added N in the following pools: extractable NH 4 + N, (NO 2 − + NO 3 − )N, soluble organic N, microbial biomass N and total N in the soil residue. The effects of U and UF on microbial activity (CO 2 production), microbial biomass (FE and SIR) and on the numbers of autotrophic nitrifiers (MPN) were also studied. The pattern of transformation of N was quite different. In the U-treated soils the added N contributed mostly to the exchangeable NH 4 + pool, whereas in the UF-treated soils the highest amount of the added N was found in the soil residue. In the U-treated soils the amount of NH 4 + was constant throughout the experiment, but the 15 N in it was diluted by mineralization of native organic N. In teh UF-treated soils the accumulation of exchangeable NH 4 + started slowly and increased steadily. However, the atom% 15 N excess in the NH 4 + pool stayed constant, as it did in the soil residue. This unchanged 15 N enrichment of NH 4 + indicates formation of a UF-humic complex. Higher atom% 15 N excess in the UF-treated soils in the exchangeable NH 4 + pool (2.8%) than in the soil residue (1.5%) suggests also that the UF-N entered the active organic N pool in the soil. The results presented here help to explain earlier field observations, where UF was shown to improve the N status of forest soil, and the applied N was retained in an available N pool.