An accepted goal of conservation is to build a conservation network that is resilient to fragmentation and environmental change. However, fragmentation is a relative concept as well as connectivity. Effective corridors should provide suitable and reliable connectivity among habitats across scales for species mobile or less mobile for gene exchange under uncertainty and change. However, often "static" background data of landscapes are used to this end (i.e. the cartography of land uses/covers) whereas landscapes (habitats included) are dynamic. Indeed, they do change either under different seasonal conditions, or under multiple driving forces like, for instance, climate change.Thus, how can we rely on connectivity we are going to measure?
Dear Theo, thanks. What you advanced is a modelling on the relative robustness of the landscape and is certainly a way to approach the problem. Another way, as we did, is focusing on time series of relevant state variables and looking for predictability of each segment of the landscape. Indeed, what we are looking for, i.e. fragmentation or effective corridors, can systematically change on the map, and what is fragmented or suitable as corridor under certain conditions could not be suitable or fragmented when season, conditions or the set of focal species are changed. Just because we are not so good in predicting the future and what could be a suitable network sustaining biological diversity and gene exchange, we have to rely on past time series (at a suitable scale) to define the trajectory of every landscape segment to see whether it is predictable or not, that is, if it is persistent or not. Once you get a "predictability" map of invariant structures then you can think of applying different modelling tools to derive, under uncertainty; what possibly could be an effective corridor network and a suitable fragmentation for the future. So you could discover that along with "classical" green and blue ways other elements in the landscape could be crucial for the network based on their predictability. You could also discover which unpredictable landscape pieces are crucial for the maintenance of the overall connectivity in the face of climate change and try to transform them in "persistent" through planning and management efforts. See, for instance, the paper "Highlighting order and disorder in social–ecological landscapes to foster adaptive capacity and sustainability" recently appeared in Landscape Ecology. The same principle should be applied to fragmentation /connectivity for marine systems (see Modeling population connectivity by ocean currents, a graph-theoretic approach for marine conservation appeared in Landscape Ecology). Indeed, for many marine species, population connectivity is determined largely by ocean currents transporting larvae and juveniles between distant patches of suitable habitat. So, connectivity relies on the persistence of ocean currents suggesting areas that might be prioritized for marine conservation efforts and that are working like "stepping stones" in the maintenance of the overall network. On the other hand, you might identify "new" candidate stepping stone areas in case of predicted changes in the oceanic current pattern due to climate change. Unfortunately most of marine biologists and ecologists involved in marine conservation do not consider the importance of ocean currents.
Dear Giovanni,
This is indeed a very valid point that is raised. The tools provided by GIS, landscape indices etc. often lead to much attention being paid to technical aspects, dealing with the (quality of the) tool itself.
Important question is indeed the effectiveness of the corridor (or landscape connectivity, for that matter). A corridor across a major highway through a large natural territory like the Carpathians may not be effective at population level, since the populations on both sides are large enough, and gene flow does occur. At most we can say we avoid traffic collisions, which is good from an ethical point of view (for animals, and people), but the impact at population level may be close to nil.
The point you raise, is the landscape dynamics in relation to landscape connectivity (in fact, the sustainability of the landscape). One way we addressed this is species modelling in which we calculated the chances of arrival at a certain point for different landscape configurations. We did this for the Brown bear in Abruzzo, and my colleague Hans Baveco used the 'smallsteps' model - but there will be more.
This modelling gives the relative 'robustness' of the landscape in different scenarios. Still it depends on the number of landscape configurations, the general connectivity for a species, and the level of fragmentation for 'an' ecosystem, how good this prediction will be in the end, considering the landscape dynamics.
Baseline is that you do not exceed a certain threshold level of fragmentation: species' survival should not depend on one or two corridors but rather a permeable landscape. Generally multifunctional landscapes with structural variety support more biodiversity than monofunctional landscapes.
Article Ecological networks, a challenge for territorial planning - ...
Dear Theo, thanks. What you advanced is a modelling on the relative robustness of the landscape and is certainly a way to approach the problem. Another way, as we did, is focusing on time series of relevant state variables and looking for predictability of each segment of the landscape. Indeed, what we are looking for, i.e. fragmentation or effective corridors, can systematically change on the map, and what is fragmented or suitable as corridor under certain conditions could not be suitable or fragmented when season, conditions or the set of focal species are changed. Just because we are not so good in predicting the future and what could be a suitable network sustaining biological diversity and gene exchange, we have to rely on past time series (at a suitable scale) to define the trajectory of every landscape segment to see whether it is predictable or not, that is, if it is persistent or not. Once you get a "predictability" map of invariant structures then you can think of applying different modelling tools to derive, under uncertainty; what possibly could be an effective corridor network and a suitable fragmentation for the future. So you could discover that along with "classical" green and blue ways other elements in the landscape could be crucial for the network based on their predictability. You could also discover which unpredictable landscape pieces are crucial for the maintenance of the overall connectivity in the face of climate change and try to transform them in "persistent" through planning and management efforts. See, for instance, the paper "Highlighting order and disorder in social–ecological landscapes to foster adaptive capacity and sustainability" recently appeared in Landscape Ecology. The same principle should be applied to fragmentation /connectivity for marine systems (see Modeling population connectivity by ocean currents, a graph-theoretic approach for marine conservation appeared in Landscape Ecology). Indeed, for many marine species, population connectivity is determined largely by ocean currents transporting larvae and juveniles between distant patches of suitable habitat. So, connectivity relies on the persistence of ocean currents suggesting areas that might be prioritized for marine conservation efforts and that are working like "stepping stones" in the maintenance of the overall network. On the other hand, you might identify "new" candidate stepping stone areas in case of predicted changes in the oceanic current pattern due to climate change. Unfortunately most of marine biologists and ecologists involved in marine conservation do not consider the importance of ocean currents.
Monitoring of habitat status becomes more expensive with higher resolution and temporal frequency requirements. I suspect a moderate resolution land cover condition or % cover (aka continuous field land cover) would probably be a good starting point. Examples exist in the National Land Cover Database (NLCD) percent imperious surface and tree cover products (http://pubs.usgs.gov/fs/2012/3020/fs2012-3020.pdf), and percent bare soil, shrub, herbaceous, and litter (http://www.gis.usu.edu/~doug/pubs/JWM(57)1-78.pdf) some of which are being added to NLCD standard products. At coarser resolutions (500m – 250m) percent tree cover (http://ftp.glcf.umd.edu/library/pdf/ei_hansen.pdf) and several other land covers as well as vegetation condition (actual productivity compared to weather and site-based estimated productivity; (http://lca.usgs.gov/blm/information/Mointoring_EPA_in_Rangleands_and_Forests.pdf) should be useful for monitoring corridor condition (currently at an annual or 5 year time step). The USGS is also helping to develop automated Essential Climate Variables (ECVs, http://remotesensing.usgs.gov/ecv/index.php) routinely globally (most, I believe, at 30m resolution). These include Leaf Area Index, biomass, and surface water.
Dear Giovanni,
This is an interesting topic indeed. Landscapes and habitats are very dynamic, In parts of the world where the seasons are shifting dramatically between the summer and the winter I believe it is important that we perhaps look more at winter conditions in the future if possible. Climatic events such as wam periods in the middle of the winter that creates periods with no or less snow cover may change the movements of animals. When the snow-pack dissapeares completely or and ice-crust forms close to the ground it forces e.g. smaller animals to move in order to find food or shelter when otherwise they would remain relatively still. In such cases it may be important for example to map connectivity based on locations of human settlements providing food/shelter or including areas with local snow-conditions that are better for dispersal in winter than e.g. the normally preferred habitats in summer.
I believe this is a challenge for the future and I know of no such winter studies yet in my field working with small mammals and landscape ecology.
Reacting on Magnus, and Giovanni:
What is the function of the specific landscape, or corridor we are focusing on? If it is gene exchange, a passing of one specimen in a generation is enough for exchange with another population.
In the case of metapopulations: if just one pair reaches a patch, a 'vacant' area, the corridor functions well enough because a new population can establish itself.
Magnus is mentioning changing seasons, part of winter in which no migration may occur if I understand him well. At population level this is not so relevant perhaps (In our metapopulation models we consider a population 'connected' or part of a -metapopulation- network if exchange occurs once a year).
The seasonal movements become of importance if you consider the daily or regular foraging trips: then connectivity can be crucial, and construction of a road, a barrier can be very detrimental for not mobile (terrestrial) species.
So landscape connectivity should be considered also in the context of landscape functioning. Different ecological functions result in different demands regarding connectivity.
Hi Giovanni,
FRAGSTATS has not been mentioned yet.
See http://www.umass.edu/landeco/research/fragstats/fragstats.html
FRAGSTATS is a computer software program designed to compute a wide variety of landscape metrics for categorical map patterns (for example a classified remote sensing image).The original software (version 2) was released in the public domain during 1995 in association with the publication of a USDA Forest Service General Technical Report (McGarigal and Marks 1995). Since then, hundreds of professionals have enjoyed the use of FRAGSTATS. Due to its popularity, the program was completely revamped in 2002 (version 3). Recently, the program was upgraded to accommodate ArcGIS10 (version 3.4). The latest release (version 4) reflects a major revamping of the software, with a completely redesigned architecture intended to support the addition of cell-level metrics and surface pattern metrics, among other things. The current release of version 4 (v4.2) has essentially the same functionality as version 3, but with a new user interface that reflects the redesign of the model architecture, support for additional image formats, and a variety of sampling methods for analyzing sub-landscapes.
Version 4 is currently the development version and is being developed and supported by Dr. McGarigal and Dr. Cushman with programming by Eduard Ene.
Dr. McGarigal is the primary contact person for questions and comments regarding all versions of FRAGSTATS.
FRAGSTATS mght be of use with regard to connectivity and especially with categorical datasets obtained at different scales with remote sensing techniques. . Although connectivity can be evaluated using a wide variety of FRAGSTATS metrics that indirectly say something about either the structural or functional connectedness of the landscape, FRAGSTATS computes a few metrics whose sole purpose is to measure connectivity. Patch cohesion (COHESION) was proposed by Schumaker (1996) to quantify the connectivity of habitat as perceived by organisms dispersing in binary landscapes. Patch cohesion is computed from the information contained in patch area and perimeter. Briefly, it is proportional to the area-weighted mean perimeter-area ratio divided by the area-weighted mean patch shape index (i.e., standardized perimeter-area ratio). It is well known that, on random binary maps, patches gradually coalesce as the proportion of habitat cells increases, forming a large, highly connected patch (termed a percolating cluster) that spans that lattice at a critical proportion (pc) that varies with the neighbor rule used to delineate patches (Staufer 1985, Gardner et al. 1987). Patch cohesion has the interesting property of increasing monotonically until an asymptote is reached near the critical proportion. Another index, connectance (CONNECT), can be defined on the number of functional joinings, where each pair of patches is either connected or not based on some criterion. FRAGSTATS computes connectance using a threshold distance specified by the user and reports it as a percentage of the maximum possible connectance given the number of patches. The threshold distance can be based on either Euclidean distance or functional distance, as described elsewhere (see Isolation/Proximity Metrics). Connectedness can also be defined in terms of correlation length for a raster map comprised of patches defined as clusters of connected cells. Correlation length is based on the average extensiveness of connected cells, and is computed as the area-weighted mean radius of gyration across all patches in the class or landscape. Correlation length is not included with the connectivity metrics in the FRAGSTATS graphical user interface because it is already included as a distribution metric for patch radius of gyration (GYRATE_AM) under the Area/Density/Edge metrics. A map's correlation length is interpreted as the average distance one might traverse the map, on average, from a random starting point and moving in a random direction, i.e., it is the expected traversability of the map (Keitt et al. 1997). Hence when connectivity is quantified, categorical datasets preferably at different scales are a strong requirement. Hence accroding to my opnion we should acquire imagery with remote sensing for a certain Region of Interest (ROI), classify these datasets (using post-processing remote sensing methods en finally when the classifications are geographically registrered, they can be used in FRAGSTATS even in combination with a GIS.
GIS is not the essence, the metrics and the data determine the complete exercise of measuring and mapping connectivity (to my opnion at least).
Cheers,
Frank
Hello Frank,
This bypasses the question if we must focus on tools, or on the kind (or quality) of data we use.....
Besides, Fragstats is just a very generic measure. And in fact, it provides many measures, but what they exactly say?
And, lastly, it depends on the mapping process and map quality what the results will be.
The ability to accurately describe habitat will in the end be decisive in results with something like Fragstats. Just processing raster data will always give some figures, but may not be meaningful.
Hello Theo. A short reflection on your answer and to clearify my thoughts on the matter. I was thinking about seasonal movements of small mammals but also think it can be valuable to discuss effects on the gene flow. I think the seasonal movements during winter could be enhanced by a suitable habitat patch where the animals resided in autumn being connected to e.g. a barn or an area with plentiful of dead wood providing good shelter from predators during a warm period in winter. Even though such movements in winter may not result in actual gene flow, since it is movement (usually) outside the normal reproduction season, I think it may be important for the whole population. In winter these type of movements may be what determines life from death and consequently the local extinction of a population. Such local extinction events may be widespread during a bad winter and affect the overall population structure in spring. In spring, the animals normally start to reproduce (timing dependent on the particular species of course) and their distribution and numbers compared with the previous autumn will be determined by the losses during the winter. Hence, if there are no animals or a limited number of survivors in spring due to the connectedness of the landscape being low with regards to "sheltered winter hideouts", this will probably have an effect on the distribution of local populations in spring. Consequently the potential of gene flow in the population during the coming year may be affected if for example local populations in an area surrounded by rather strong barriers are becoming extinct or the local population being separated by far distances, (longer distances than the dispersal capacity of the animals). What do you think of these thoughts and are they considered in your metapopulation models?
I think data is more important than tools. I've not good exerience with specific software. I use a GIS, but first I make thematic maps, with photo but with important field working. A good vegetation shape, made specifically, is basic to make the first proposal of corridors; but you need studying each one, on photo a on site. Never a tool or a software make the work of a professional; only helps you.
I'm not sure why the question is necessary. The better our tools are, the better we can understand our data and - most importantly - the limitations of our data relative to our questions. This cycle leads to better tools, and better understanding of our measurements, and more informed hypotheses. One paper I'd like to write is about the "Curse of high resolution data". Recent and continuous improvements in measurement tools (e.g. high resolution satellites) means that in order to benefit from all that resolution, we need tools and understandings that equip us to deal with the lack of sub-pixel heterogeneity. Now landscape variability is quantifiable as variance and no longer measurement error. These are welcome technological advances. But, they require new/ improved tools to evaluate them.
Dear Raymond,
I agree that "the better our tools are, the better we can understand our data and - most importantly - the limitations of our data relative to our questions". Probably the question has not been put right. The question refers to the fact that many scholars are rather focusing on tools for analyzing networks and corridors, based primarily on static information, while it would be necessary not to lose sight of the dynamics of landscapes at different time scales, and how to use the dynamic information for a better and more useful identification of corridors. Today we begin to have even rather sophisticated tools to study such dynamics (remote sensing, see Bruce Wylie's contribution here), but we still continue to study the effectiveness of corridors on the basis of static information without taking into account their dynamics at different spatial-temporal scales. If we could shift this common static perspective we could possibly develop new and better analytical tools for investigating corridor effectiveness just in the line of what you said.
Giovanni - I understand better now. I agree with you. Ultimately, I think it is a hard question, with expensive solutions. But, some work is being done that advances the question some. if you haven't seen these papers, they might be worth a read.
Fausch, K.D., C.E. Torgersen, C.V. Baxter, and H.W. Li. 2002. Landscapes to Riverscapes: Bridging the Gap between Research and Conservation of Stream Fishes. Bioscience, 52(6):483-498
and a newer one that applies the theory put forth in the first paper.
Brenkman, S.J., J.J. Duda, C.E. Torgersen, E. Welty, G.R. Pess, and M.L Mchenry. 2012. A riverscape perspective of Pacific salmonids and aquatic habitats prior to large-scale dam removal in the Elwah River, Washington, USA. Fisheries Management and Ecology, 19: 36-53
Also - a little shameless self-promotion.
Timm, R.K., and R.C. Wissmar. 2013. Response to disturbance in a highly managed alluvial river: Does it conform to Le Chatelier's general law? Geomoprhology. 182: 116-124
Dear Magnus,
Thanks for clarification.... At population level this habitat connectivity you describe is important, and I guess that climate change may have all kinds of (rather unpredictable) effects on this connectivity. A better understanding of local dynamics may for that reason be important, and specific habitat models can help in acquiring this understanding. I did not use such models myself, but colleagues did (for common vole/ root vole, badger, bittern etcetera), models like Metaphor on population dynamics, or small-steps, on species movement. In some cases these models were combined, like the case I cited for Abruzzo.
If it were specifically for conservation, one may rather choose for metapopulation models, which operate more at a landscape level. With the right choice of species (flagship species, umbrella species) you may increase the sustainability for a larger group of species of conservation interest (see also my publication on European corridors, I'll attach the link).
Species like small mammals (mice?) are very dynamic, showing large fluctuations, and are also very versatile. They may not be the prime target species for conservation. A different story is the lynx, or leopard which may encounter similar dynamics in habitat. Here corridors are at a different scale, and essential for long term survival.
But, as mentioned, other models using life history of species are very valid again for different purposes, a better understanding of local scale dynamics, or possibly to increase the knowledge of events like climate change.
Book European corridors: strategies for corridor development for ...
I have just seen this discussion, and in my view we need landscape ecology tools fit for purpose. It may be that a range of tools must be available to support management decisions about connectivity related issues at the landscape scale, as well as high quality data to demonstrate the utility of different approaches. I can offer 2 citations that illustrate these points.
Firstly, a meta-analysis that demonstrates that connectivity theory provides a useful contribution to landscape ecology -
Eycott AE, Watts K, Brandt G, Buyung-Ali LM, Bowler D, Stewart GB, Pullin AS (2012) Factors affecting the impact of the matrix on species movement: systematic review and meta-analysis. Landscape Ecology
Secondly, species movement data that demonstrate the utility specfic models -
Stevenson CD, Ferryman M, Nevin OT, Ramsey AD, Bailey S, Watts K (2013) Using GPS telemetry to validate least-cost modeling of gray squirrel (Sciurus carolinensis) movement within a fragmented landscape. Ecology and Evolution 3 (7):2350-2361
Howdy folks,
@Theo,
Maybe I was somewhat unclear in explaining the use of FRAGSTATS, but the following reference of a book from Cushman and Landguth, clarifies it somewhat.
Ecological Associations, Dispersal Ability, and Landscape Connectivity in the Northern Rocky Mountains by Samuel A. Cushman and Erin L. Landguth.
http://www.fs.fed.us/rm/pubs/rmrs_rp090.pdf
This is the summary of their work:
Population connectivity is a function of the dispersal ability of certain species,
the influences of landscape elements on the species movement behavior, density and the distribution of the population. Let's not forget the structure of the landscape. Often, researchers do not carefully consider each of these factors when evaluating connectivity and making conservation recommendations. The authors present a general method for efficient evaluation and calculation of functional connectivity for large numbers of native species across vast geographical areas. In their work connectivity is evaluated for 36 groups of species with different ecological associations; within each of these groups, three dispersal abilities were evaluated across the United States northern Rocky Mountains. They quantified the extent and fragmentation of the predicted connected habitat for each of these 108 species and identified those for which the current landscape has the lowest area and the highest fragmentation of habitat.
To clarify the role of "the very generic measures" offered by FRAGSTATS metrics in the work of mentioned authors, the following rationale is followed by the authors mentioned above.
Resistant kernel modeling produces spatial predictions of the areas of a landscape connected by dispersal, given a resistance model, species distribution and density as wellas the dispersal ability of the species. The authors objectives are to quantify the differences in the extent and pattern of connected habitats across resistance models and dispersal abilities. To accomplish this, they calculated a suite of fragmentation metrics uisng FRAGSTATS metrics (McGarigal and others 2002) on each of the 108 maps of predicted connected habitats. They then selected eight landscape metrics reflecting several universally important gradients of landscape structure (Cushman and others 2008). Specifically, these metrics were chosen to robustly reflect two major gradients of landscape structure, particularly important in driving population responses with respect to landscape structure (Cushman and others 2010b). These two gradients are
(1) A gradient from a high extent to a low extent of connected habitat; and
(2) A gradient from low edge density, a small number of isolated patches, and a high habitat aggregation to highly fragmented conditions characterized by a high density of isolated habitat patches of a small size and a high total edge density.
Just to state that typically, some scientitsts like a more mathematical approach to quantify for example migration trajectories, while others don't like that at all. Evidently, validation of a migration moel appraoch should always be validated with migration measurements.
Each approach probably has its merits. Though I know for sure that a quantitative approach can be verified and scrutinized much more rigorously than a qualitative approach. If I am not mistaken quantitative scrutiny is typically one of the most important characteristics which distinguishes science from rubbish.
Cheers,
Frank
I must disagree with a throw away comment from Giovanni which is mis-ldeading...
Unfortunately most of marine biologists and ecologists involved in marine conservation do not consider the importance of ocean currents
This is not true - there is a substantial quantity of literature and research on oceanic connectivity and this issue nearly always comes up in discussions regarding Marine Protected Areas. What is true is that consideration of this issue has not always been taken into account in the actual implementation of MPA designs e.g. the recent issues regarding MPA designation in English waters and whether the sites accepted so far can define an ecologically coherent network - which implies a degree of connectivity..
I am happy that Clive agrees at least on the fact that consideration of ocean currents has not always been taken into account in the actual implementation of MPA designs. As far as I know, very few MPA designs, or none, have been established considering the actual pattern of currents, at least in the Mediterranean sea. If he could provide some examples of MPA design and implementation actually considering the spatiotemporal pattern of currents in an explicit way I would be very grateful to him.
Hi Giovanni et al.
Defra and Natural England did at least consider connectivity in the exploratory work on the English MPA network. The report is available at
http://publications.naturalengland.org.uk/publication/46009
and connectivity is supposed to be taken into account in the Site Selection Guidance issued by Defra
The site selection exercise proposed 127 sites which external peer review considered would fulfill the need for connectivity. So far 31 sites have been accepted by the Minister with a request for further science investigation on the other sites. The general concensus from the external review seems to be that a network of less than the original 127 proposed sites will not deliver the inter-connectivity required.
I can probably dig out more papers and discussions on this if of interest once I get back to the office but at present I am working away while at a meeting.
It might be interesting to make a comparison with how this whole issue is being addressed (or not) in different parts of the European seas.
This is very interesting discussion, i am not GIS, but i learn about modeling, so i can follow the discussion and increased my point of view that ecological connectivity must be considered spatial distribution and its characteristic.
Indonesia has a few of marine protected area, which focused on specific or whole coastal habitats. Most of them been established using socio-ecological-economic approaches, but less of quantitative spatial consideration.
Hope it will further discussed and shared more scientific evidances according to GIS tools utilizations.
Yudi - there is some interesting work being done in Europe right now based on a concept called "Ecological Stepping Stones". I have not seen how the models work. But, in theory, they estimate the virtual connectedness of habitat features in order to create restoration plans. There is some literature on the topic from the European Water Framework Directive.
Dear Raymond, how do i get the references? I am working on the connectivity between seagrass habitat and its neighbor (other coastal and marine good and services) for my dissertation.
Hi Clive,
I agree that It might be interesting to make a comparison with how this whole issue is being addressed (or not) in different parts of the European seas. As far as I know, not much of ocean current patterns are being considered, at least not as sound data from oceanographers. It appears that oceanographers do not often dialogue with whom is charged of MPA design, and viceversa. However, my point was that connectivity among MPAs is determined largely by ocean currents transporting larvae and juveniles between distant patches of suitable habitats, and we do not have often idea of what is the spatiotemporal persistence of this oceanic current pattern. I have looked at the Natural England site, and apparently they look at connectivity and viability as two of the seven network design principles they are using to identify sites to contribute to an ecologically coherent MPA network through the Marine and Coastal Access Act. Very good, it's a good start.
Yudi - I am a river ecologist and not familiar with the marine literature - especially with regard to MPAs. But, there were a couple of folks in my cohort during graduate school who were doing work on the California current and the Berring Sea. There were spatially explicit components to their work as I recall. Start with Google Scholar and search for Vera Agostini, and John Field. I'm sure there were other students in their lab working on similar kinds of things - like Lorenzo Ciannelli.
As far as conectdness goes, there are a whole bunch of questions about your data that I would want to know. All the technical questions notwithstanding, one of the most interesting ecological questions in my opinion is this: How discontiguous can habitat patches be and still be functionally connected? Are there patch composition and configuration covariates to distance?
Raymond, as regards the issue" one of the most interesting ecological questions in my opinion is this: How discontiguous can habitat patches be and still be functionally connected? Are there patch composition and configuration covariates to distance?" that would depend upon the species ability to disperse across gaps, as well as the success a species has to survive in a gap or secondary habitat, ie. the degree of habitat-specificity or how generalist a species is. In the end, a habitat has a community, so then would the minimum connectivity depend upon the least generalist species ?
I think you are exactly right Amartya. Of course not all steelhead can clear a 3m waterfall - but some can. Not all monarchs can traverse the Gulf of Mexico - but many can. What is the nature of the gap and when does it become a barrier? In metapopulation theory, the tails of the distributions are super important to population structures. Unraveling the answers to the questions you posed is really key to conservation biology/ ecology. It would be fun to spend some more time thinking about this and formalizing it.
Prof Giovanni- this questions quite resonates with my research interest and domain.
I will share my perspective with a combined lens of a landscape ecologist cum spatial analyst. Ecological connectivity is quite rooted in the application of landscape indicators for conservation research and management planning of natural resource systems. In my early works, I had modelled spatio-temporal analysis outputs (as surrogate indicators like patchiness, porosity and fragmentation) to comment about significance of ecological connectivity using SPLAM interface (program developed in the lab- reference stated below). The focus was forest ecosystem and habitat conservation. In recent times, I work on agro-ecosystems that typically represent human-nature coupled environment systems. The overarching shift to ecosystems based management (EBA) framework was an embedded element of this discourse, as of existing shift to ecosystems service and benefits as key while maintaining agriculture landscapes (somewhere overlapping with the concept of conservation agriculture). Moreover, I found it very interesting to employ visually explicitly methods (disconnected agriculture strands derived using in high resolution data) to explain to local stakeholders, both resource users (farmers) and resource managers, the concept of ecological connectivity and its impact on delivery of ecosystems services.
My idea of sharing these notes is to present the connectivity concept beyond its conventional application
Related Refernces
P.S.Roy, Hitendra Padalia, Nidhi Chauhan, M.C.Porwal, Sas Biswas and Rajendra Jagdale, 2005; Validation of Geospatial model for Biodiversity Characterization at Landscape Level—a study in Andaman & Nicobar Islands, India, Ecological Modelling [Elsevier], Volume 185, Issues 2-4, 10, July 349-369
Nagabhatla Nidhi and P.S Roy. 2007. Measuring Landscape Parameters: Fragmentation, Disturbance and Biological Richness in Baratang Islands (Andaman) for estimating Landscape Structure, Human and Environment Interlinkages. International Journal of Ecology and Development. Vol. 6, No. S07, 22-36
I had a quick look to the answers before and let me present you my approach.
Lets rearange the given question in two ways:
1. Should we use the science as per given data? or
2. Should we use existing data towards the improvements of science?
When I was working on my three Master Degrees and 2 PhDs, I had the following experience for following areas:
1. Computer Science (CS)
CS is a new science where the data collection is almost not required. May be on the long range, it could be, if now we keep notes on the CS development procedures. Therefore, the answer is pretty clear. We should keep on looking for the best GIS tools.
2. Industrial Engineering and Systems Management (IE&SM)
IE&SM is again relatively new science. If we keep looking on the production and consumption chain, innovation is important in the chosen industry. Again the picture looks the same, as it is presented with the case of CS. New techniques should be developed to manage smooth industrial business.
3. Environmental Management (EM)
EM is the science to manage the environment. The environmental science in terms of biology is old science. Management in terms of economics is a bit newer science. Remote sensing or GIS techniques are new sciences to deal with environmental sciences. In order to deal with EM, the consideration of several sciences are required. Thus, we need to put an aim. Which science is more important to proceed the research?
When the collected data are not enough, then some questions arise: "Are the scientists willing to gather data per defined time period for the sake of science? How much are expenses and time requirements? Can the scientist bear those expenses and time periods? What are the benefits to reconstruct the scientific approach as per existing data?"
4. Economics (EC)
Economics mainly deals with the manipulation of existing data. If we care for the supply and demand case, then we deal with the case of IE&SM. If we care for the valuation techniques or questionnaires inquiries then data collection and information manipulation is required.
5. Biodiversity (BD)
The nature preservation has the main impact on biodiversity issue. BD has several approaches: biology, economics, Remote Sensing etc. BD case requires data collection for some decades by the real researchers. Any target appearance can lead to the collection of BD data per given or life time period.
6. Sustainable Development (SD) as process of Supportive Progress (SP)
SD as SP is a new science as a part of economics containing three dimensions: society, economy and environment. The following staff happened during the years of performing my PhD at NTUA, Greece. The indicators should have been defined to construct 3D of SP in terms of Biodiversity Economics combining political, societal, economic and natural indicators. After the definition of all eighteen (18) branches of SP in 3D, the region should have been defined to perform the gathered info involvement into SP model. This described concept has its roots in the ideology of information pyramid. However, the reality is opposite due to lack of existing data in many defined indicators. Therefore, the researchers doing a complex job in a limited time period with the limited finance come up with totally other picture.
I had the same experience during the years of my 3rd Master Degree on Radar Remote Sensing Science in the direction of vegetation changes at MAICh, Greece.
There was no need in data collection during my years of study at my 1st MSc and PhD at SEUA Armenia on 3D modelling of Interconnects at VLSI. When I was preparing my 2nd MSc at American University of Armenia with the subject of Chemical Industries impacts on Nature, the first purpose was to collect data and then to go for the scientific approach.
Now let me summarize by giving the conclusion to the initiative question. The science is very important to perform the apt decisions. The limitations of existing data, time and finance can bottleneck or redirect the scientific truth to the other direction. As the final result is to compare and to show that the predicted (SP) model is defined in the best way. Thus, we better focus on the kind of data we use to help young researchers to perform the best out of the existing data and finding new scientific techniques according to the existing information. When someone have a budget and no limitation on time then the scientist can keep on looking for the best scientific tools because biodiversity is so huge that there will be always a need in the collection of data per species.
With best regards
Dr. Azniv Petrosyan
Interesting discussion. You may be interested to read a paper we just published in Journal of Applied Ecology that investigates how the data used can result in very difference connectivity metrics. We used dispersal data and compared that to adult males and females. While our landscape was static, our results show that it is critical to utilise data most relevant to connectivity - ie. dispersal data.
Article can be downloaded here:
http://onlinelibrary.wiley.com/doi/10.1111/1365-2664.12282/abstract
Thank you Nicholas for sharing the paper and thanks again Prof Zurlini for initiating this interesting discussion
I'd say we have to focus on the data as so many aspects of "connectivity" are under-explored. To give a personal example from my own research group, we know very little about how narrow connecting landscape features influence pollinator movement and subsequent pollen movement between plant populations - "connectivity" in a subtle, gene flow sense. In fact the only published study I'm aware of that has addressed this experimentally is our own:
Cranmer, L., McCollin, D. & Ollerton, J. (2012) Landscape structure influences pollinator movements and directly affects plant reproductive success. Oikos 121: 562-568
Here's a link to the PDF if anyone is interested:
http://oldweb.northampton.ac.uk/aps/env/lbrg/journals/papers/cranmer2011-landscape-structure-oikos.pdf
hi everyone.
structural versus functional connectivity across ecological communities is an important and difficult issue. We have spent quite a lot of time thinking about it in my group, with particular reference in the first instance to a bird community in complex habitat. The inspiration was the emerging classic work of Jim Radford, Andrew Bennett et al. who did an extraordinary job of evaluating bird population responses to extent and connectivity of habitat. We added a process-based view over the top of that body of work.
We have published a series of papers that tackle various aspects of the original question in this ResearchGate chain.
Our approach is a combination of intensive field ecology, genetic estimates of mobility, GIS approaches to landscape classification, and associated information, all in a hypothesis-testing framework involving Circuitscape and causal modelling. It's too much to summarize here but two papers with Nev Amos as first author make predictions for the system and then test them for up to 10 species:
Amos JN, Bennett AF, Mac Nally R, Newell G, Pavlova A, Radford JQ, Thomson JR, White M, Sunnucks P. (2012) Predicting landscape-genetic consequences of habitat loss, fragmentation and mobility for multiple species of woodland birds. PLoS ONE 7 e30888.
Amos JN, Harrisson KA, Radford JQ, White M, Newell G, Mac Nally RM, Sunnucks P and Pavlova A. (2014) Species- and sex-specific connectivity effects of habitat fragmentation in a suite of woodland birds. Ecology 95, 1556–1568.
The general underlying approaches involving population persistence and genetic connectivity are summarized in:
Sunnucks P (2011) Towards modelling persistence of woodland birds: the role of genetics. Emu- Austral Ornithology, 111, 19-39. http://www.publish.csiro.au/?paper=MU10008
and associated with all this, we have published some more specific and detailed studies:
Pavlova A, Amos JN, Goretskaia MI, Beme IR, Buchanan KL, Takeuchi N, Radford JQ and Sunnucks P (2012) Genes and song: genetic and social connections in fragmented habitat in a woodland bird with limited dispersal. Ecology 93, 1717-1727.
Harrisson KA, Pavlova A, Amos JN, Takeuchi N, Lill A, Radford JQ, Sunnucks P. (2012) Fine-scale effects of habitat loss and fragmentation despite large-scale gene flow for some regionally declining woodland bird species. Landscape Ecology, 27, 813-827.
Harrisson KA, Pavlova A, Amos JN, Takeuchi N, Lill A, Radford JQ, Sunnucks P. (2013) Disrupted fine-scale population processes in fragmented landscapes despite large-scale genetic connectivity for a widespread and common cooperative breeder: the superb fairy-wren (Malurus cyaneus). Journal of Animal Ecology, 82, 322-333. (doi: 10.1111/1365-2656.12007)
Harrisson KA, Pavlova A, Amos JN, Radford JQ, Sunnucks P. (2014) Does reduced mobility through fragmented landscapes explain patch extinction patterns for three honeyeaters? Journal of Animal Ecology. 83: 616–627. doi: 10.1111/1365-2656.12172
At a broader scale, the following produced a stunning case where natural selection apparently overwhelms structural connectivity, and prevents functional connectivity in some senses:
Pavlova A, Amos JN, Joseph L, Loynes K, Austin JJ, Keogh JS, Stone GN, Nicholls JA and Sunnucks P (2013). Perched at the mito-nuclear crossroads: divergent mitochondrial lineages correlate with environment in the face of ongoing nuclear gene flow in an Australian bird. Evolution, 67, 3412–3428. doi: http://dx.doi.org/10.1111/evo.12107.
Ultimately we might also be concerned about how restricted gene flow and natural selection promote local adaptation (or not), and what the implications are for conservation units:
Pavlova A, Selwood P, Harrisson KA, Murray N, Quin B, Menkhorst P, Smales I and Sunnucks P. (2014) Integrating phylogeography and morphometrics to assess conservation merits and inform conservation strategies for an endangered subspecies of a common species. Biological Conservation, 174, 136–146.
hope some of that is of interest
best wishes
Paul
Dear all, just a small remark, downvoting is a quite legitimate attitude, but I would be very interested in knowing the reasons for that, as it is much more productive for the ongoing discussion that just saying "I don't agree with you". I believe that RG is an arena where different and sometime contrasting opinions can have their own space. We should not hide behind a downvote. Therefore, I kindly invite the downvoter to express his point of view on the issue clearly by taking the floor and thus fostering the ongoing discussion .
You are making jokes dear @Gianni by inviting such person to express his point of view!!! Such person(s) need serious help indeed! They even do not appear to be a followers of a thread!!! They are haters!
Again the downvote!!!! Very good, it is really funny isn't? Can you please downvote also this one? Thank you so much my dear downvoter , whoever you are!
The downvoter has changed his mind now. Why? I think that this is a rather chaotic attitude, isn't? Dear downvoter, if you decided to downvote keep on doing it unless you really do not know what to do...... If this is a game let's play then!
With reference to my range of articles on elephant corridors, disjunctive Apennine bear habitat and forest expansion in Majella providing connectivity for forest dwellers (but disconnecting grasslands), my general experience is that real-world design of corridors is often the art of the political possible rather than evidence-based. An example is the Pan-European Ecological Network (PEEN) that I know on the ground in places, without seeing evidence of species using the network for its intended purpose. However, connectivity for animals has been effectively maintained by ecoducts over freeways in several European countries, including Germany.
Connectivity obviously is not positive by definition; for example, I wonder whether we are creating avenues for alien species in invasion-prone landscape.
I am aware that my contribution presumes a time-scale of decades rather then centuries or millennia; the latter would lead to a different discussion.