KNO13: Technologies and techniques to advance biodiversity monitoring and modelling

Date: Wednesday October 10, 2018

Location: Erottaja, ELY

Time: 15:15-17:15

The unprecedented changes being experienced in the Arctic emphasize the importance and urgency of getting information to decision-makers in a timely manner. To do so requires easily accessible, comprehensive data, coordinated and consistent monitoring, up-to-date assessments of trends and informed responses. This session explores case studies in methodologies and technologies to help monitor and track changes in Arctic biodiversity and ecosystems.

Chairs: Catherine Coon, Bureau of Ocean Energy Management (BOEM)

Format: Series of presentations followed by discussion

  • Leveraging drones to quantify the landscape-context of tundra biodiversity change: Jeffrey Kerby, Dartmouth College pdf
  • Rapid decay of palsas monitored using RTK GPS, UAS data and aerial photographs: Timo Kumpula, University of Eastern Finland pdf
  • Geomorphology matrix as a base of Arctic coastal zone monitoring in Global change dynamic: Dmitriy Dobrynin, Working Group on Anseriformes of Northern Eurasia pdf
  • Animal energyscapes: A new dimension for arctic environmental spatial planning: David Grémillet, French National Center for Scientific Research - CNRS 
  • Exploring Arctic ecosystem futures through biodiversity models and using these models for evaluation of global biodiversity models: Fiona Danks, UN Environment World Conservation Monitoring Centre pdf
  • Model evaluation by specialists – using Reindeer Movements for map quality assessment: Sven Adler, Swedish University of Agricultural Sciences pdf

 


Abstracts:

Leveraging drones to quantify the landscape-context of tundra biodiversity change

Jeffrey Kerby, Dartmouth College; Isla Myers-Smith and the HiLDEN Network (arcticdrones.org)

Arctic monitoring efforts reflect the inherent tradeoffs between research priorities and logistical realities.  For decades, ecological research has combined plot- and satellite-informed datasets to yield critical insights about Arctic environmental change. However, often meso-scale observations from 10-1 – 102 km2 are missing. The rapid development and widespread accessibility of small Unmanned Aerial Systems (a.k.a. drones) offers promising solutions to some of the historical constraints Arctic biodiversity monitoring. Improved observations of the landscape context for ecological change will inform predictions of Arctic biodiversity, trophic food webs, carbon cycling and climate feedbacks. Biodiversity data syntheses have been instrumental in documenting widespread changes in plant cover, composition, and phenology throughout the Arctic in recent decades. These findings are commonly invoked to support hypotheses about the drivers of Arctic ‘greening’ (or ‘browning’) - biome-scale patterns of environmental change that have emerged from the remote sensing literature. Ongoing research linking ground- and satellite-based datasets and monitoring philosophies is revealing scale-dependent mismatches between metrics and patterns of vegetation responses. Spatial variability in landscape-level dynamics are increasingly understood to underpin these differences at site-specific scales. The High Latitude Drone Ecology Network (HiLDEN) was established to share common protocols across Arctic research sites to better facilitate meso-scale ecological research, while also providing a framework for building plot- to satellite-based research syntheses. The first summer of data collection in 2017 included participation from over a dozen tundra research teams spread across six Arctic nations. In 2018, we will extend this work to another two sites capturing landscape-level data for long-term ecological monitoring sites in the Yukon, Fennoscandia, on Ellesmere Island and Svalbard. Here, we report on the continued development of this network and present a multi-site analysis of tundra productivity (quantified with various metrics of ‘greenness’) and landscape heterogeneity within and across tundra systems. Our aim is to encourage spatially contextualizing existing and future biodiversity monitoring within their surrounding landscape, and thus to promote future synthesis and cross-scale understanding of vegetation change in the Arctic.

 

Rapid decay of palsas monitored using RTK GPS, UAS data and aerial photographs

Kumpula, T., M. Verdonen P. Korpelainen Department of Geographical and Historical Studies, University of Eastern Finland

Palsa is a form of discontinuous permafrost in the circumpolar zone. Palsa are peat mounts with a core of permanently frozen peat, ice and mineral soil. There are several studies which indicate that palsa’s are melting and decaying as a result of climatic warming. Finland. In two palsa’s high accuracy Real Time Kinematic (RTK) GPS measurements and active layer depth have been conducted 2007-2017. A measurement grid with two meter interval was defined over both palsa, (approximately 200 points). Measurements were carried out yearly in the end of August (2007-2017). Active layer depth of each point was measured with an active layer probe. With ArcGIS we created 3-D models of palsa and yearly active layer surfaces. Weather data used is from the Kilpisjärvi climate station allocated about 15 km north of the study sites (1951-2017). Both palsa’s has experienced significant decay especially along the edges. The palsa in Laassaniemi has large collapsing side towards a thermokarst pond, here the core has collapsed more than one meter during the study period. In the 1959 image there are no signs of a thermokarst pond. With detailed RTK GPS and active layer monitoring of palsa for 10 years we can follow accurately the development of palsa and its correlation to local climatic factors. We have gathered aerial photograph times series from 1960’s to present from 20 palsa mire basins in Enontekiö Lapland. Combining historical aerial photographs the time span of the study can be stretched up to 50-60 years. In some of the studied palsa mire basins more than 60% of the palsas have disappeared during the past 60 years. In 2015 we began to monitor changes in the palsas with UAS (Unmanned Aerial System) drones. Since 2016 we have 12 palsa mires under investigation with twice a year data collection (mid June and late August). From UAS -based remote sensing data we have noticed that palsa decay rate in all sites is surprisingly significant.

 

Geomorphology matrix as a base of Arctic coastal zone monitoring in global change dynamic

Dmitriy Dobrynin, Working Group on Anseriformes of Northern Eurasia; Mariya Dobrynina, Working Group on Anseriformes of Northern Eurasia

Coastal zone ecosystems are the result of terrestrial and marine factors superposition. Both of this groups are affected by Global Change influence. The sublatitudinal orientation of coastal zone in Eastern Europe and Asia increases the response of subshore ecosystem on global dynamic. All these aspects cause difficulties in monitoring arrangement. Choice of method sets and monitoring sites can be implemented on the base of satellite multispectral and radar image processing results. Images from space give us total information about coastal ecosystems structure and dynamic. By analysis of this data we select several groups of coastal landscapes the most sensitive to Global Change influence. They are: - thermokarst lowlands; - thermoabrasive seashores; - estuaries of the Arctic basin rivers; - accumulative shallow bottom formations; - sandbanks and sand bars The most obvious scenarios of ecosystem dynamics under the influence of Global Factors were estimated for each landscape type. Species and ecosystems belonging to the group of the most unstable and threatened were identified. Also the trends of changes in biotopes of most sensitive species are revealed on the base of series of temporary satellite images.

 

Animal energyscapes: A new dimension for Arctic environmental spatial planning

David Grémillet, French National Center for Scientific Research - CNRS; Jerome Fort, French National Center for Scientific Research - CNRS

Arctic landscape and seascape ecology are essential for a better understanding and forecasting of the impact of global changes on arctic biodiversity. This fusion of geography and ecology takes advantage of the geospatial revolution and the rapid development of global data acquisition and mapping tools. It allows displaying and analyzing the impact of environmental conditions upon the distribution and the movements of living organisms. This multi-layered, GIS-based approach integrates a wide range of gradients in geological, physical, chemical, and biological variables. Such forcing factors all act upon the energy balance (the ratio of energy intake to energy expenditure) of wild animals, which conditions their capacity to survive and reproduce, leading to their resilience. Animal energyspaces, defined as the spatial distribution of animal energy balance according to environmental conditions, therefore arises as a fundamental dimension in landscape and seascape ecology. This dimension has so far seldom been considered, by lack of appropriate methodologies to evaluate the energy expenditure of animals on the move. Recent developments of electronic tools, notably GPS and 3D motion recorders, now fill this gap, by providing both the position of animals, and estimations of their metabolic rates. Drawing from recent work on arctic birds and mammals, notably migratory seabirds, we will illustrate why, and how animal energyscapes can become a crucial aid to arctic environmental spatial planning. Specifically, we will draw from our long-term study of little auks (Alle alle), one of the most numerous arctic seabirds, to infer energyscapes and test migration theory. Overall, we will show how energyscapes can be used to evaluate the impact of climate change (vanishing cryosphere) upon arctic animals, to identify hotspots important for conservation and management at the species, community and ecosystem levels, and to define networks of protected areas and test connectivity issues. By integrating these many facets, animal energyscapes will allow evaluating cumulative environmental impacts (including those of pollutants), and mitigating their effects, thereby serving scenario planning for reduced risk.

 

Exploring Arctic ecosystem futures through biodiversity models and using these models for evaluation of global biodiversity models

Michael Harfoot, UN Environment World Conservation Monitoring Centre; Fiona Danks, UN Environment World Conservation Monitoring Centre

Arctic ecosystems, both terrestrial and marine, are amongst the most rapidly changing on the planet. On land, climate change is causing rapid vegetation shifts, both within species and by community turnover, and this process interacts with nutrient availability. Fauna is also likely change substantially. Arctic ecosystems therefore present a challenge for models of biodiversity because of the interaction of global change drivers. The pace of change also makes these system potentially excellent for evaluating model projections – testing how good we are at forecasting ecological futures. Here we present ideas for why we would apply some cutting edge models of biodiversity (e.g. Madingley General Ecosystem Model, PREDICTS model, LPJ-Guess dynamic vegetation model, HYBRID dynamic vegetation model) to Arctic ecosystems, in particular what they might tell us about Arctic ecosystem futures. We also present ideas for how Arctic ecosystems could be used to evaluate biodiversity models through targeted experiments where models make projections for ecosystem changes over a period of several years, and then observed changes are used to evaluate how accurate those model projections were. Applying biodiversity models to the region encourages vital collaboration among scientific, policy, NGO, academic and industry communities in advancing the understanding of key aspects of the Arctic Biodiversity Assessment (with the hope of being able to advise and impact policy recommendations) and increases the visibility of Arctic biodiversity in global fora. Arctic biodiversity model development can aid in safeguarding biodiversity under changing conditions, in particular a rapidly changing climate. Predictive capacity, through advancing models as tools for understanding processes governing change in the Arctic, is improved, and efforts to reduce stressors and implement adaptive measures will be supported through such effort.

 

Model evaluation by specialists – using reindeer movements for map quality assessment

Henrik Hedenås, Swedish University of Agricultural Sciences, Institutionen för skoglig resurshushållning, Skogsmarksgränd, 901 83 UMEÅ, Per Sandström,Swedish University of Agricultural Sciences, Institutionen för skoglig resurshushållning, Skogsmarksgränd, 901 83 UMEÅ, Anna Skarin, Swedish University of Agricultural Sciences, Department of Animal Nutrition and Management, Ulls väg 26, 75007 Uppsala

Mapping of ecosystem services, e.g. tree biomass, is highly demanded as decision basis for e.g. green-infrastructure planning, sustainable management and multiple-use of the landscape. The quality of the maps depends on the model training data and model techniques used. For common phenomena national monitoring schemes like National Forest Inventories (NFI) are supposed to deliver good training data. However, national monitoring data may not function as training data when mapping more uncommon phenomena. Reindeer lichens (Cladonia ssp.) is a pivotal resource for reindeer husbandry through winter time as reindeer lichens is the major winter diet for reindeer. Thus, reindeer lichens can be interpreted as an ecosystem service for reindeer herders. Mapping lichen abundant forests have become quite important since reindeer lichens have decreased strongly over the last 50 years. Maps of important ecosystem services such as reindeer lichens are considered to facilitate communication between forestry, reindeer herders and the administration boards. As the area of lichen abundant forest is very small in comparison to the total forested area of Sweden, only few sample plots within the Swedish NFI describe areas with high values of lichen cover, more than 80% of all has a reindeer lichen abundance <1%. The resulting model based on topographical, satellite and Lidar data using Generalized Additive model technique underestimate the lichen coverage, explains a low proportion of the variance and has a very high cross validation error. According to standard model evaluation tools such kind of model would be discarded. Nevertheless, a map of reindeer lichens was anyhow produced based on the model. But, how usable is such a map? We used, in a first step, GPS positions from reindeer provided by the reindeer herding community Vilhelmina Norra to map the movement of the reindeer and to identify areas preferred by reindeer using Brownian Bridge Moving Models. In a second step, we compared the areas that were preferred and not preferred by reindeer according to the amount of lichen coverage according to the predicted map showing the predicted coverage of reindeer lichens. Surprisingly, the predicted coverage of lichens within the areas preferred by reindeer was significantly higher (p<0.01) compared to the areas not preferred by reindeer. This shows that even models with inadequate performance can provide useful and reliable maps and the information within national monitoring schemes enfold important information even for uncommon phenomena.

 

Efficient monitoring of Swedish sub-arctic habitats with the use of directed balanced sampling

Hans Gardfjell, Swedish University of Agricultural Sciences; Sven Adler Swedish University of Agricultural Sciences

With a changing climate and plausible alterations in land use it is likely we will see substantial environmental changes in the arctic region. That has increased the demand for accurate monitoring data on habitats and species. However, traditional random sampling designs are commonly only efficient in collecting data on the most common habitats and species. The distribution of habitats and vegetation types are highly skewed. Some few dominates whereas the majority of habitats only covers a minor part of the landscape. We will present a method using habitat models and directed balanced sampling to improve the efficiency of probability sampling.

Like us on Facebook
Follow us on Twitter
Subscribe to our YouTube Channel
Join our LinkedIn Group
Check us out on Google+
Follow Us on Instagam
Follow Us on Flickr