 Up until now we've been talking about ecosystems as networks through which energy and resources flow, but in this module we will be defining more clearly what we mean by this as we talk about ecological networks and active area of research within contemporary ecology. Since the 1970s when networks were imported from physics and the social sciences into ecology, they have grown increasingly popular among ecologists and today offer a potential for advancing our understanding of ecological processes. From the metabolic networks in our cells to the food webs within an ecosystem to the global networks of animal migration, network models are a powerful tool for quantifying how ecosystems work. Network analysis can provide precise metrics that quantify community structure and models for analyzing ecosystem resilience and stability in a formal and rigorous fashion. A network contains nodes and links between those nodes. An ecological node may represent an individual plant or animal, a whole population or species. Interactions between them take many forms, but are generally divided into antagonistic trophic interactions, what are called food webs with interactions such as that between predator and prey or mutualistic symbiotic interactions such as that between pollinator insect and flowering plant. As an example we can take a look at this trophic network with data taken from 12 Galapagos islands indicating the interaction between the different creatures in the food web. This network here considers 19 land bird species and 106 plant species. Network structure describes patterns and structures within the overall ecological network. Here we're asking such questions as how many nodes are there in the network, how connected is the overall network? The overall level of connectivity to a network is a primary factor determining the nature of any network. In an ecosystem this is called connectance. Connectance is a measure of the number of links or connections between species expressed as a proportion of maximum connectance. This will also tell us something about the linkage density which is the average number of links per species. For any given creature their degree can define how much of a generalist or specialist the species is within the food web. Both in its role as a consumer and as a resource. Previous network analysis studies have revealed that ecological networks exhibit the important characteristic of clustering that is found in many different networks. Clustering or modularity describes to what extent the network can be divided into non-overlapping groups of highly interacting species within their local neighborhood in the network. A module or cluster then is a set of species that have a disproportionate number of connections with their module as opposed to connecting to other modules. This illustration of the food web with a number of different ecosystems illustrates prominent clustering. Here the food web structure and sub web frequency distribution for a number of different ecosystems is shown including a park, estuary, lake and coral reef all exhibiting a high clustering coefficient. More recently attention has shifted from network connectance to the idea of degree distribution. Degree distribution is a measurement to the variability in link density. It is asking about how equally or unequally the connections are distributed out within the system. For example koala are very specialized feeders feeding on a very few species of eucalyptus whereas raccoon feed on a wide variety of species, 17 birds or eggs, crayfish, plants and various other invertebrates. Thus they would have many more connections within the food web as consumers. Degree distribution goes a long way to tell us how centralized or distributed the network is. These networks are those with a very high degree distribution having few central hubs with very many connections or many other nodes have very few links. This may also give us what is called a scale free network which are surprisingly common in our world. For example the metabolic network of a biological cell follows this centralized scale free model where the essential molecules of ADP and ATP that provide the energy to fuel the cell play a central role in interacting with a very many different other molecules thus forming hubs in the metabolic network. But equally we can find distributed networks that have a low degree distribution where the connections are distributed out relatively evenly across the network. Keystone species is a measurement of how central a node is within a network and thus how significant it is within the system. A species with a high centrality measure within an ecosystem would be called a keystone species. Keystone species are generally understood as those species that play a role disproportionate to their number in the dynamics of their ecosystem. A classical keystone species is a predator that prevents a particular herbivorous species from eliminating dominant plant species. Since the prey numbers are low the keystone predator's numbers can be even lower and still be effective. Yet without the predators the herbivorous prey would explode in numbers removing all the dominant plants and dramatically alter the dynamics of the ecosystem. Another more concrete example of a keystone species would be a beaver. Beavers can engineer whole wetland ecosystems in their capacity to build and maintain dams which have a major effect on the environment in that they can transform the territory from a stream to a pond or swamp which define a different set of biotic and abiotic elements and interactions. As such these keystone species are described as playing a critical role in maintaining the structure of an ecological community and we could say that they have a high centrality measure within that network. The centrality of a node within an ecological network is quite a complex metric to define as it is a product of a number of different factors such as how connected that node is that is to say the immediate connections it has with other nodes either as a predator or consumer of some resource. Equally we need to think about how critical the creature is to the network. Do they serve some role that no other node in the network could perform and thus they form what we call a bridging link. Without other redundant components to fill their role they can be critical to the ecological network not because of the number of links they have or the size of the population but instead because of their uniqueness and irreplaceability. We can also note that critical ecosystem processes will not always be under the control of individual species but may be mediated nonetheless by a small set of species that thereby form a keystone functional group. For example the groups of microbial species that fix nitrogen can control processes that play a fundamental role in the preservation of ecosystems. With their high influence on the system keystone species can trigger nonlinear responses that lead to cascades of local or global change in the formation of the ecosystem. More generally we may also have abiotic critical elements in the system such as the level of precipitation, pH level of the soil or other geological factors that form critically sensitive values within the network. Network analysis is a key tool for modeling the resilience of an ecosystem in terms of the integrity of its network of connections. Connectivity within networks can both enable robustness but also represent pathways for disaster spreading. It works both ways. On the beneficial side the resiliency and robustness of the network can be correlated to the flow through the system as described by the theory of ascendancy where ascendancy is defined as the level of functionality to the ecosystem's trophic network. One way of interpreting ascendancy is to regard it as the organizational structure of connections that enables resources to flow through the network. The magnitude of the power that is flowing within the system towards particular ends. As such ascendancy is a key index in determining the ability of an ecosystem to prevail against disturbance by virtue of its combined organization connectivity in size. This can be illustrated by analogy to the difference between a living and dead organism which may be interpreted as simply the volume of resources flowing through the system. The greater the flow the greater the vitality meaning more ascendancy. Thus we can use network theory to analyze how integrated the network carrying the resources is and this will give us some interpretation to the robustness of the ecosystem on aggregate. This network integrity is traditionally understood in terms of habitat fragmentation. Habitat fragmentation describes the emergence of discontinuities or fragmentations in an organism's native environment causing population fragmentation and ecosystem decay. It is a process during which a large expanse of habitat is transformed into a number of patches of smaller total size isolated from each other by a set of habitats unlike the original one. It includes discontinuities in the spatial patterning of resource availability affecting the conditions for species occupancy and ultimately individual fitness. Migration can arise via both natural and anthropogenic processes in terrestrial and aquatic systems. This graphic of the river network in part of Denmark illustrates anthropogenic fragmentation along the country's waterways where every red dot indicates physical barriers to fish migration such as dams. Connectivity among fragments, the characteristics of the matrix, the availability of corridors for movement between fragments and the permeability and structure of habitat edges are all important in this context and affect the structure, persistence and strength of the ecological network. In the same way that connectivity can enable ascendancy and robustness, it can also enable large food web disturbances in the form of cascades. As food webs become more interconnected, this creates more pathways for disaster spreading and cascading effects. An ecological cascade effect is a series of secondary extinctions that is triggered by the primary extinction of a key species in the ecosystem. Secondary extinctions are likely to occur when the threatened species are dependent upon a few specific food sources or some other mutualistic interaction. One example of the cascade effect caused by the loss of a top predator is related to the sea otter. Starting before the 17th century, sea otters were hunted extensively for their pelts. Their decline caused a cascade effect through the kelp forest ecosystem along the Pacific coast of North America. One of the sea otters' primary food sources is the sea urchin. When hunters caused sea otter populations to decline, an ecological release of sea urchin populations occurred. The sea urchins then over-exploited their main food source, kelp, creating ecological collapse in those areas. No longer having food to eat, the sea urchin population became locally extinct as well. Also, since kelp forest ecosystems are home to many other species, the loss of the kelp ultimately caused their extinction as well. Thus, the loss of sea otters caused a cascade effect of secondary extinctions due to their interconnectivity within the system. In this video, we've been talking about ecological networks, the application of network theory to modeling and analyzing ecosystems in terms of their network of connections. We talked about some of the overall features to these networks including the overall level of connectivity, linkage density and degree distribution. We discussed network modularity as describing to what extent the network can be divided into highly interacting local clusters, how centrality can be used as a measurement to how central a node is within a network and thus how significant it is within the overall system. Finally, we looked at ecosystem resilience in terms of the integrity of its network, noting how this can both enable robustness by enabling a greater flow of resources through the system, but also how it can add to vulnerabilities as it gives rise to the possibility of cascading food web disturbances.