Comment on “The extent of forest in dryland biomes”

Griffith, Daniel M.; Lehmann, Caroline E. R.; Strömberg, Caroline A.E.; Parr, Catherine L.; Pennington, R. Toby; Sankaran, Mahesh; Ratnam, Jayashree; Still, Christopher J.; Powell, Rebecca; Hanan, Niall P.; Nippert, Jesse B.; Osborne, Colin P.; Good, Stephen P.; Anderson, T. Michael; Holdø, Ricardo M.; Veldman, Joseph W.; Durigan, Giselda; Tomlinson, Kyle W.; Hoffmann, William A.; Archibald, Sally; Bond, William J.

doi:10.1126/science.aao1309

Cited by 48 publications

(42 citation statements)

References 15 publications

(26 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At the current time, these habitats are managed under a blanket policy of fire exclusion and suppression and fire-setting is a punishable offense according to the Indian Forest Act (Ratnam et al, 2016;Thekaekara et al, 2017). These policies, which stem from a historical misreading of these ecosystems as "forests" are inimical to the conservation and sustainable management of these savannas (Lehmann and Parr, 2016;Griffith et al, 2017). First, the notion that fires are always undesirable disturbances in these ecosystems, widespread amongst both managers and vegetation scholars, prevents nuanced understanding, and appropriate research on the ecology of these systems.…”

Section: Discussionmentioning

confidence: 99%

Functional Traits of Trees From Dry Deciduous “Forests” of Southern India Suggest Seasonal Drought and Fire Are Important Drivers

et al. 2019

Self Cite

View full text Add to dashboard Cite

Two dominant biomes that occur across the southern Indian peninsula are dry deciduous "forests" and evergreen forests, with the former occurring in drier regions and the latter in wetter regions, sometimes in close proximity to each other. Here we compare stem and leaf traits of trees from multiple sites across these biomes to show that dry deciduous "forest" species have, on average, lower height: diameter ratios, lower specific leaf areas, higher wood densities and higher relative bark thickness, than evergreen forest species. These traits are diagnostic of these dry deciduous "forests" as open, well-lit, drought-, and fire-prone habitats where trees are conservative in their growth strategies and invest heavily in protective bark tissue. These tree traits together with the occurrence of a C 4 grass-dominated understory, diverse mammalian grazers, and frequent fires indicate that large tracts of dry deciduous "forests" of southern India are more accurately classified as mesic deciduous "savannas."

show abstract

Section: Discussionmentioning

confidence: 99%

Functional Traits of Trees From Dry Deciduous “Forests” of Southern India Suggest Seasonal Drought and Fire Are Important Drivers

et al. 2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…Global initiatives could greatly benefit from CNNs, such as those recently implemented by the United Nations Food and Agricultural Organization ( [60]) to estimate the overall extension of forests in drylands biomes, where they used the collaborative work of hundreds of people that visually explored hundreds of VHR images available from Google Earth to detect the presence of forests in drylands. The uncertainties in such initiatives ( [61,63,64] could be decreased following our approach to build a CNN-based tree mapper). CNN-based tree and shrub detectors could serve to produce global characterizations of ecosystem structure and population abundance as part of the satellite remote sensing essential biodiversity variables initiative ( [65]).…”

Section: Discussionmentioning

confidence: 99%

Deep-learning Versus OBIA for Scattered Shrub Detection with Google Earth Imagery: Ziziphus lotus as Case Study

et al. 2017

View full text Add to dashboard Cite

There is a growing demand for accurate high-resolution land cover maps in many fields, e.g., in land-use planning and biodiversity conservation. Developing such maps has been traditionally performed using Object-Based Image Analysis (OBIA) methods, which usually reach good accuracies, but require a high human supervision and the best configuration for one image often cannot be extrapolated to a different image. Recently, deep learning Convolutional Neural Networks (CNNs) have shown outstanding results in object recognition in computer vision and are offering promising results in land cover mapping. This paper analyzes the potential of CNN-based methods for detection of plant species of conservation concern using free high-resolution Google Earth TM images and provides an objective comparison with the state-of-the-art OBIA-methods. We consider as case study the detection of Ziziphus lotus shrubs, which are protected as a priority habitat under the European Union Habitats Directive. Compared to the best performing OBIA-method, the best CNN-detector achieved up to 12% better precision, up to 30% better recall and up to 20% better balance between precision and recall. Besides, the knowledge that CNNs acquired in the first image can be re-utilized in other regions, which makes the detection process very fast. A natural conclusion of this work is that including CNN-models as classifiers, e.g., ResNet-classifier, could further improve OBIA methods. The provided methodology can be systematically reproduced for other species detection using our codes available through (https://github.com/EGuirado/CNN-remotesensing).

show abstract

“…The importance of capturing the photosynthetic physiologies of C 3 and C 4 grass types, and their variation with CO 2 and climate, has long been recognized (Collatz, Berry, & Clark, ; Sellers et al, ; Still, Berry, et al, ). However, additional ecological and hydrological factors control C 3 and C 4 distributions and can interact in complex ways, producing offsetting effects driven by interactions with hydrological cycles (Griffith, Lehmann, et al, ) or by driving systems towards deterministically woody states (Bond, ; Moncrieff, Scheiter, Bond, & Higgins, ; Oliveras & Malhi, ). For example, in an experimental study conducted in a semi‐arid, mixed C 3 /C 4 grassland in North America, Morgan et al () showed that C 4 grasses were counterintuitively outperforming C 3 grasses owing to increases in water‐use efficiency driven by interacting effects of higher CO 2 and warming on soil moisture and plant physiological responses.…”

Section: Discussionmentioning

confidence: 99%

Assessing earth system model predictions of C₄ grass cover in North America: From the glacial era to the end of this century

Still

Cotton

Griffith

2018

Global Ecol Biogeogr

Self Cite

View full text Add to dashboard Cite

Aim C4 grasses are distinct from C3 grasses, because C4 grasses respond in a different manner to light, temperature, CO2 and nitrogen and often have higher resource‐use efficiencies. C3 and C4 grasses are typically represented in earth system models (ESMs) by different plant functional types (PFTs). The ability of ESMs to capture C4 grass biogeography and ecology across differing time periods is important to assess, given the crucial role they play in ecosystems and their divergent responses to global change. Location North America. Time periods Last Glacial Maximum (LGM), historical modern period (ca. 1850) and end of this century. Major taxa studied C4 grasses. Methods Proxy data representing relative cover and productivity of C4 grasses were collated, including carbon isotope ratios of soil carbon and animal grazer tissue, and vegetation plot data in undisturbed grasslands. We selected available model predictions of C4 PFT percentage cover. Models were compared against one another and assessed against proxy data at key time points: the LGM, the historical modern period before widespread grassland conversion to agriculture, and the end of this century. Results We highlight large differences among model predictions of percentage C4 grass cover across North America: all pairwise combinations have correlations < .5, and most are < .2. Models also do not capture spatial patterns of the percentage C4 grass cover from proxy data, during either the LGM or the historical modern period. Models generally under‐predict percentage C4 grass cover, particularly during the historical modern period. Main conclusions Earth system models do not accurately represent the biogeography of C4 grasses across a range of time‐scales, and their outputs do not agree with one another. We suggest model improvements to represent this crucial functional type better, including more collection and greater integration of C3 and C4 grass trait data, explicit representations of tree–grass competition for water, and a greater focus on disturbance ecology.

show abstract

Comment on “The extent of forest in dryland biomes”

Cited by 48 publications

References 15 publications

Functional Traits of Trees From Dry Deciduous “Forests” of Southern India Suggest Seasonal Drought and Fire Are Important Drivers

Functional Traits of Trees From Dry Deciduous “Forests” of Southern India Suggest Seasonal Drought and Fire Are Important Drivers

Deep-learning Versus OBIA for Scattered Shrub Detection with Google Earth Imagery: Ziziphus lotus as Case Study

Assessing earth system model predictions of C₄ grass cover in North America: From the glacial era to the end of this century

Contact Info

Product

Resources

About

Comment on “The extent of forest in dryland biomes”

Cited by 48 publications

References 15 publications

Functional Traits of Trees From Dry Deciduous “Forests” of Southern India Suggest Seasonal Drought and Fire Are Important Drivers

Functional Traits of Trees From Dry Deciduous “Forests” of Southern India Suggest Seasonal Drought and Fire Are Important Drivers

Deep-learning Versus OBIA for Scattered Shrub Detection with Google Earth Imagery: Ziziphus lotus as Case Study

Assessing earth system model predictions of C4 grass cover in North America: From the glacial era to the end of this century

Contact Info

Product

Resources

About

Assessing earth system model predictions of C₄ grass cover in North America: From the glacial era to the end of this century