Index

The problem
We live in a fragmented world. The restoration of large stretches of contiguous habitat is impossible in most parts of the globe. Land managers and conservation biologists are thus faced with the task of enhancing and retaining the conservation value of existing fragmented landscapes. This work takes place in a financial landscape with clear limits and opportunity costs; land purchases and restoration efforts now must be viewed as strategic conservation decisions. As metapopulation theory, reserve networks, and regional reserve planning replace single-site reserve designs, the facilitation of connectivity between core areas is increasingly seen as a critical aspect of conservation in fragmented landscapes. Habitat corridors are thought to provide that connectivity. The intended function of corridors is simple: to facilitate movement and gene flow between connected patches (through this function, corridors can increase connectivity, facilitate patch recolonization, reduce inbreeding, and increase the stability of metapopulations and metacommunities). Yet as land availability and quality goes down, and purchase prices go up, validation of these basic functions is critical.

How we validate the importance of corridors
Our goal is to validate corridor function directly – to answer the question: Does this corridor facilitate movement, increase gene flow, and promote diversity? To answer this question, we use a wide variety of techniques, depending on the temporal and spatial scales and the target organisms.

Molecular validation: migration and gene flow using molecular fingerprinting and assignment tests
Molecular fingerpring, assignment tests, and estimation of contemporary gene flow have now advanced to a level where they can be used for direct assessment of landscape function for target species, particularly where molecular markers have already been developed (Cushman et al. 2006). From experimental test to direct landscape evaluation, our group is using molecular approaches to validate landscape function. In our current work, graduate student Dan Evans (Tewksbury Lab) is currenty using previously validated behavioral models (Levey et al. 2005) combined with molecular fingerprinting to determine gene flow between connected and unconnected patches in multiple plants species, and Sam Cushman, Kevin McKelvey, and Michael Schwartz have recently validated the use of molecular markers in evaluating landscape connectivity from the perspective of the animals under study – in their case, bears (Cushman et al. 2006).

N15 validation: affordable, direct assessment of dispersal services and plant response to landscape features
In a project led by Tomas Carlo (a post-doctoral associate in Tewksbury’s lab) we are currently field-validating the use of 15N- labeled urea to isotopically label maternal plants and recover the label in dispersed seeds and established seedlings, giving us a flexible, powerful tool to measure the impact of landscape features on the movement of plants and the animals that disperse the seeds of plants (NSF-DEB 0636630) . Here is how it works: We apply a foliar spray of 15N labeled urea to plants during the flowering stage. The urea is incorporated in the plant, broken down, and the 15N used in anabolic and metabolic functions. We have already shown that the labeled urea is reliably passed from the plant to the seed via the endosperm and is retained in the seedling, allowing us to directly associate seeds and seedlings with the sprayed plant. In fact, seeds and seedlings from plants sprayed with different concentrations of 15N urea retain distinctive 15N signatures. This 15N labeling method can be applied to many plant species (particularly small plants and shrubs who bear small seeds) and can be applied regardless of seed size or the availability of molecular genetic tools. Importantly, because basic mixing models allow bulk sampling of all propagules collected in a location to determine the proportion of a sample that came form labeled plants, this method is ideally suited to the study of plant movement, as rare events (the long distance movement of a seed) can be examined at a fraction of the cost of of molecular approaches. Currently, we are using this system to examine context-specific seed movement, and validating it through direct comparison with molecular markers, but this technique is designed to be used in a landsca pe context, and is particularly efficient when the landscape itself is the focus.

Direct validation of the movement process
In our previous work, we have used direct fluorescent tagging of fruits to measure seed dispersal (Levey and Sargent 2000, Tewksbury et al. 2002, Levey et al. 2005) passive foliar tagging to measure pollen flow (Townsend and Levey 2005), and manipulated spatial structure of males and females in dioecious plants (Tewksbury et. al. 2002), all of which provide direct measurements of the effectiveness of habitat corridors.

Historical records and meta-analytical techniques
Large-scale data collected over substantial time scales are valuable tools for assessing the role of connectivity in mediating ecological change. Such data exist in the form of annual North American Breeding Bird Surveys, Christmas Bird Counts, the Fourth of July Butterfly count, herbarium collections, and museum collections. By comparing changes in these large-scale data, we can directly estimate rates of biodiversity change and determine if rates are concomitant with landscape change. Importantly, these collections provide an unprecedented means for testing if conservation designs are effective, because they provide baseline data (i.e., controls) for comparing regions before and after conservation action (e.g., the construction of protected areas or reserves). Museum collections offer another way to measure connectivity over relevant ecological and evolutionary time scales. Molecular analyses of museum collections can be used to estimate rates of outcrossing, a proxy for gene flow among populations. When compared with estimates from contemporary populations, this provides a powerful tool for measuring changes in gene flow as a function of landscape connectivity. The scientific literature is replete with studies of local population characteristics. Much of this literature is within theses, dissertations, biological monitoring, and impact assessment documents. By combining thousands of observations from these separate studies with landscape data and modern meta-analysis techniques, we are able to assess how important population parameters (e.g., survival rates, population size) are a function of connectivity.