Connectivity of green and blue infrastructures: living veins for biodiverse and healthy cities


One of the ecological concepts that BIOVEINS investigates is fragmentation. What is fragmentation, and why is it interesting?

Fragmentation refers to the breaking up of a formerly continuous habitat into smaller pieces that are no longer continuous. So, for example, a forest that was once vast and continuous is logged and we only have small patches of standing forest left. Or, a grassland is cut into pieces by paved roads crossing it. Or, areas within a forest-shrub mosaic are converted into agricultural fields of various sizes and shapes, leaving small, isolated patches of habitat mosaic. What is a continuous habitat? We could define it (a bit vaguely) as one that can be moved across by all the species living in it. Fragments, by contrast, are (1) reduced in size and absolutely small, (2) have a high edge-to-interior-area ratio, (3) are embedded in what is usually called a “matrix” of un-natural or degraded landcover, which cannot be crossed.

From a conservation ecology perspective, each of these things is bad news. Fragments, like national parks, are often thought of as islands in a sea of uninhabitable matrix. Thus, drawing on Rober MacArthur’s island biogeography theory, we can expect that habitat fragments, to the extent that they are truly isolated (we will think about this below), should have fewer species in them. They will both lose species through random local extinction, and not gain new ones, due to isolation. A high edge-to-interior ratio is also bad since “edge effects” are generally negative: predation success is higher along edges (this is bad for the prey), edges can reduce the microhabitat benefits of closed canopy forests several meters or more into the fragment, some species that are adapted to interior habitats cannot survive anywhere near an edge, and some animals will not cross edges for perceptual/ behavioural reasons (e.g. a tendency to follow, not cross, linear features), and are thus trapped inside fragments. Edge habitats are essentially the same as habitat transitions, and thus favour generalist species and those adapted to early stages of succession, such as pioneer, ruderal or disturbance-adapted species. The properties of the “matrix” in which the fragment is embedded will determine what species are able to move in and out of the fragment. Roads are dangerous for many species that may be run over by traffic or overheat on tarmac in the sun. Farmland may have poisons, traps, insecticides, herbicides, and simply no suitable food sources or habitat structures for other daily needs (resting spots, nests, etc.). A logged and bulldozed forest is not a suitable habitat meeting the needs of most of the species living in the surrounding forests. And so on.

However, ecosystems are dynamic: ecology is a set of processes of constant change. Fragmentation, in the sense of the formation of patches of degradation or disturbance, is a “natural” ecological process. Habitat mosaics, in which two kinds of “fragments,” e.g. grassland patches and shrub patches, are mixed together and interconvert in a process of transformation, are very normal ecological dynamics. You will notice that we use different language for good processes and bad processes in ecology: fragments are bad, patches are good, mosaics are good, matrices are bad. Its not entirely clear that these are really different processes or phenomena, however. That is, a snail’s matrix is a land-crab’s habitat; a southern lapwing will make a nest in the middle of a degraded field where a colocolo oppossum would never go. Some trees only germinate in full sun, other only in deep shade. How, then, to decide when a fragmentation process is a problem?

Conservation ecology works by some version of utilitarian ethics (the greatest good for the greatest number). Thus a matrix/mosaic or fragment/patch becomes bad when biodiversity declines. How you make this calculation can make a difference. If you compare the matrix to the original habitat, now a fragment, the matrix could have more species in it. This can be for several reasons: (1) the fragment is so small that it randomly happens to contain mainly mice, and no shrews or opossoms. This is due to chance, to what happened to be inside the fragment when it was cut off from the continuous habitat. (2) There were originally mice, shrews, and opossums, but the mice and shrews moved into the matrix and for one reason or another never came back. Now there are only opossums. (3) There are still mice, shrews and opossums in the fragment, but in the matrix there are mice, shrews, rats, and voles. Rats and voles like the matrix habitat, and mice and shrews may be transient populations leaving fragments, or they may do OK in the matrix too. There are thus more species in the matrix.

Scale will make a difference. If we zoom out and look not just at one fragment and its nearby matrix, but all the fragments, then in scenario (1) we will find that there are shrews and opposums in other fragments, so across all fragments combined we find mice, shrews and opposums; in the matrix there are rats and voles so the fragments are more biodiverse (in the aggregate). If we zoom out in scenario (2) we may find other fragments that didn’t lose all their mice and shrews (e.g. bigger fragments) or fragments that received emigrating mice and shrews. Thus, across all fragments we still have mice and shrews as well as opossums, so the fragmented habitat is still more biodiverse. In scenario (3) we have to conclude that the matrix is not really a “bad” habitat and might be more of a mosaic scenario. It also depends when you make this measurement, as fragments are expected to lose species slowly over time.

A second consideration, in addition to scale, is whether the “matrix” can spontaneously experience ecological succession or whether it is being kept in a particular state through human maintenance. A road, parking lot, or agricultural field, for example, will not convert into some other habitat type until it is abandoned (see the blog post on Ruins). This gives us a sense of the time period over which certain species are being favored or disfavored: how long can we maintain a habitat mosaic that favours vole population growth and reduces opossum population growth before opossums go extinct and voles become a source of conflict? The maintenance of stasis in the distribution of habitats is more problematic that habitat change per se.

Finally, how fragments are arranged in the “matrix” also makes a difference since many animals are capable of moving certain distances or making difficult journeys at particular moments: for example, during a “dispersal” stage at the beginning of maturity for animals, or for plants at the seed stage. Other species migrate to find better habitats as the seasons change. Seeds and small animals like invertebrates can be carried in the wind, on passing animals, in vehicles, and so on. Birds can fly, but of course not all birds fly long distances, and they follow certain kinds of landmarks, habitats, and landforms. Many vertebrate animals will tend to follow linear structures and areas with cover when moving from place to place: fallen tree trunks, banks of shrubs, forest edges, rivers. But all species have different, particular, movement patterns and preferences. Habitat corridors and road under- and overpasses can be built to try to help animals cross matrix habitats and roads. These do not always work, as they may not make sense to all the animals concerned, but they are better than nothing for many species.

Urban habitats are “matrix” habitat for most species. We can address the problems that city infrastructures cause for animal and plant movements in two ways: (1) we can make corridors to better link habitat fragments; (2) we can make urban habitats less a “matrix” and more a “mosaic” for more species, through managing urban infrastructures and parks to provide nesting, resting, and foraging sites for a variety of animals, and leaving areas for wild plants to establish.

--Meredith Root-Bernstein