Legacy effects of herbivory on treeline dynamics along an elevational gradient.

Treelines are expected to expand into alpine ecosystems with global warming, but herbivory may delay this expansion. This study quantifies long-term effects of temporally varying sheep densities on birch recruitment and growth in the treeline ecotone. We examined treeline ecotone successional trajectories and legacy effects in a replicated experimental setup, where enclosures were present for 14 years with three different sheep densities (0, 25, 80 sheep km-2). Before and after the enclosures were present, the site had an ambient sheep density of 20-25 km-2. We sampled field data 4 years after enclosure removal and compared these to data sampled 8 and 9 years after enclosure erection. We sampled data on birch browsing pressure, birch distribution across life-stages (recruits, saplings, and mature trees), and birch annual radial growth. Fourteen years of increased or decreased sheep density had observable legacy effects depending on birch life-stage. Birch recruit prevalence decreased in areas, where sheep were reintroduced after being absent for 14 years. For the same areas, sapling and mature tree prevalence increased, indicating that these areas have entered alternative successional trajectories compared to areas, where sheep were present the whole time. Birch annual radial growth showed a lag effect of 2 years after enclosure removal, with growth decreasing in areas where sheep had been absent for 14 years and increasing where sheep densities were high. Thus, decadal-scale absences of herbivores can leave legacy effects due to increased numbers of trees that have high resistance to later-introduced herbivore browsing.

Savanna woody encroachment is widespread across three continents.

Tropical savannas are a globally extensive biome prone to rapid vegetation change in response to changing environmental conditions. Via a meta-analysis, we quantified savanna woody vegetation change spanning the last century. We found a global trend of woody encroachment that was established prior the 1980s. However, there is critical regional variation in the magnitude of encroachment. Woody cover is increasing most rapidly in the remaining uncleared savannas of South America, most likely due to fire suppression and land fragmentation. In contrast, Australia has experienced low rates of encroachment. When accounting for land use, African savannas have a mean rate annual woody cover increase two and a half times that of Australian savannas. In Africa, encroachment occurs across multiple land uses and is accelerating over time. In Africa and Australia, rising atmospheric CO2 , changing land management and rainfall are likely causes. We argue that the functional traits of each woody flora, specifically the N-fixing ability and architecture of woody plants, are critical to predicting encroachment over the next century and that African savannas are at high risk of widespread vegetation change.

Climate warming feedback from mountain birch forest expansion: reduced albedo dominates carbon uptake.

Expanding high-elevation and high-latitude forest has contrasting climate feedbacks through carbon sequestration (cooling) and reduced surface reflectance (warming), which are yet poorly quantified. Here, we present an empirically based projection of mountain birch forest expansion in south-central Norway under climate change and absence of land use. Climate effects of carbon sequestration and albedo change are compared using four emission metrics. Forest expansion was modeled for a projected 2.6 °C increase in summer temperature in 2100, with associated reduced snow cover. We find that the current (year 2000) forest line of the region is circa 100 m lower than its climatic potential due to land-use history. In the future scenarios, forest cover increased from 12% to 27% between 2000 and 2100, resulting in a 59% increase in biomass carbon storage and an albedo change from 0.46 to 0.30. Forest expansion in 2100 was behind its climatic potential, forest migration rates being the primary limiting factor. In 2100, the warming caused by lower albedo from expanding forest was 10 to 17 times stronger than the cooling effect from carbon sequestration for all emission metrics considered. Reduced snow cover further exacerbated the net warming feedback. The warming effect is considerably stronger than previously reported for boreal forest cover, because of the typically low biomass density in mountain forests and the large changes in albedo of snow-covered tundra areas. The positive climate feedback of high-latitude and high-elevation expanding forests with seasonal snow cover exceeds those of afforestation at lower elevation, and calls for further attention of both modelers and empiricists. The inclusion and upscaling of these climate feedbacks from mountain forests into global models is warranted to assess the potential global impacts.

Twentieth-century development of floodplain forests in Polish Carpathian valleys: The by-product of transformation of river channels?

Changes in forest extent in the corridors of four rivers of the Polish Carpathians over the last 130 years and their relation to changes in planform river geometry were investigated through the analysis of 1:25000-scale maps from the 1870s and aerial images from the mid-20th century and 2009. Average proportions of river and its geomorphic units as well as floodplain and its land cover features in the total width/area of the analysed river corridors were determined and compared between the three dates. All the analysed rivers narrowed significantly over the study period. This increased considerably the proportion of floodplains in the area of the river corridors, while lateral parts of the former, wide channels became a place of forest development. In the Koszarawa and Raba valleys, forest developed also on parts of the former floodplains following a decline in agricultural and pastoral use of lands with shallow, poor soils. The proportion of forest in the total area of the river corridors increased from 0-7.5% in the 1870s to 28.5-46.5% in 2009, and the forest expansion was mainly driven by the timing and scale of channelization works that reclaimed parts of the former channels from the rivers. A reduction in flow and sediment dynamics of Carpathian rivers over the 20th century enabled development of islands in their active zones. However, channelization works eliminated islands from most river reaches and thus islands persisted only in scarce unmanaged reaches. The expansion of floodplain forests in Carpathian valleys improves functioning of the river ecosystems but the resultant increased delivery of large wood to river channels may generate flood hazard. Optimal river management should avoid removal of riparian trees to maximize the environmental benefits but enable undisturbed transfer of driftwood through bridge cross-sections to minimize the flood hazard resulting from floodplain forest development.

Elevated temperature and CO2 cause differential growth stimulation and drought survival responses in eucalypt species from contrasting habitats.

Climate change scenarios predict increasing atmospheric CO2 concentrations ([CO2]), temperatures and droughts in tropical regions. Individually, the effects of these climate factors on plants are well established, whereas experiments on the interactive effects of a combination of factors are rare. Moreover, how these environmental factors will affect tree species along a wet to dry gradient (e.g., along tropical forest-savanna transitions) remains to be investigated. We hypothesized that under the simulated environmental conditions, plant growth, physiological performance and survivorship would vary in a manner consistent with the species' positions of origin along this gradient. In a glasshouse experiment, we raised seedlings of three Eucalyptus species, each occurring naturally in a wet forest, savanna and forest-savanna ecotone, respectively. We evaluated the effect of drought, elevated temperature (4 °C above ambient glasshouse temperature of 22 °C) and elevated temperature in combination with elevated [CO2] (400 ppm [CO2] above ambient of 400 ppm), on seedling growth, survivorship and physiological responses (photosynthesis, stomatal conductance and water-use efficiency). Elevated temperature under ambient [CO2] had little effect on growth, biomass and plant performance of well-watered seedlings, but hastened mortality in drought-affected seedlings, affecting the forest and ecotone more strongly than the savanna species. In contrast, elevated [CO2] in combination with elevated temperatures delayed the appearance of drought stress symptoms and enhanced survivorship in drought-affected seedlings, with the savanna species surviving the longest, followed by the ecotone and forest species. Elevated [CO2] in combination with elevated temperatures also enhanced growth and biomass and photosynthesis in well-watered seedlings of all species, but modified shoot:root biomass partitioning and stomatal conductance differentially across species. Our study highlights the need for a better understand of the interactive effects of elevated [CO2], temperature and drought on plants and the potential to upscale these insights for understanding biome changes.

Humid tropical rain forest has expanded into eucalypt forest and savanna over the last 50 years.

Tropical rain forest expansion and savanna woody vegetation thickening appear to be a global trend, but there remains uncertainty about whether there is a common set of global drivers. Using geographic information techniques, we analyzed aerial photography of five areas in the humid tropics of northeastern Queensland, Australia, taken in the 1950s and 2008, to determine if changes in rain forest extent match those reported for the Australian monsoon tropics using similar techniques. Mapping of the 1950s aerial photography showed that of the combined study area (64,430 ha), 63% was classified as eucalypt forests/woodland and 37% as rain forest. Our mapping revealed that although most boundaries remained stable, there was a net increase of 732 ha of the original rain forest area over the study period, and negligible conversion of rain forest to eucalypt forest/woodland. Statistical modeling, controlling for spatial autocorrelation, indicated distance from preexisting rain forest as the strongest determinant of rain forest expansion. Margin extension had a mean rate across the five sites of 0.6 m per decade. Expansion was greater in tall open forest types but also occurred in shorter, more flammable woodland vegetation types. No correlations were detected with other local variables (aspect, elevation, geology, topography, drainage). Using a geographically weighted mean rate of rain forest margin extension across the whole region, we predict that over 25% of tall open forest (a forest type of high conservation significance) would still remain after 2000 years of rain forest expansion. This slow replacement is due to the convoluted nature of the rain forest boundary and the irregular shape of the tall open forest patches. Our analyses point to the increased concentration of atmospheric CO(2) as the most likely global driver of indiscriminate rain forest expansion occurring in northeastern Australia, by increasing tree growth and thereby overriding the effects of fire disturbance.