Rigorous science-based planning is important in prioritizing investments of limited conservation resources, especially in the face of high global extinction rates (Myers et al. 2000, Balmford et al. 2003). Islands are renowned hotspots of endemism and extinction (Mulongoy et al. 2006, Kier et al. 2009). Invasive alien species pose the greatest threat to insular biodiversity (Reaser 2007). Though eradication programs can be an efficient and effective means of reducing extinction risk on islands (Howald et al. 2007), limited funding demands careful prioritization among islands to ensure the highest conservation return on investment (Januchowski-Hartley et al. 2011).

In Mexico, 149 islands compose only 0.2% of the country's land surface but host 8% of all Mexican vertebrate and plant species (Aguirre-Muñoz et al. 2008). Approximately 300 species are endemic to Mexican islands, 10% of which are considered vulnerable per the endangered species list of Mexico, the NOM-059-SEMARNAT-2001 (CONABIO 2007). Eighteen percent of all currently threatened birds and mammals are insular species (Aguirre-Muñoz et al. 2008). Invasive alien species have been implicated as the cause of extinction for 16 vertebrate species from Mexican islands (Aguirre-Muñoz et al. 2011a). Islands also support the livelihood of 0.6% of Mexico's population (INEGI 2012), through myriad economic and social values, including lobster, abalone, and tuna fisheries in the rich surrounding waters. Many seabirds and pinnipeds also use the Mexican islands as breeding and resting sites.

Considerable progress has been made in recent years to eradicate invasive mammal species from Mexican islands. As of April 2014, fifty-five invasive mammal populations of 11 species have been eradicated from 35 Mexican islands. These eradications have contributed to the restoration of over 50,815 ha and the protection of approximately 134 endemic plant species, 117 endemic vertebrates, and 220 populations of seabirds (Table 1 Aguirre-Muñoz et al. 2011b). Technological and methodological advances, such as aerial hunting and aerial baiting, have been used on the Mexican islands and have improved efficiency of eradication programs and delivered important conservation benefits (Aguirre-Muñoz et al. 2009). Despite this progress, there are still 36 Mexican islands with one or more known infestations of invasive alien mammal species (Aguirre Muñoz et al. 2011b; Fig. 1). In this paper, we describe a multicriteria decision analysis to prioritize eradication efforts among those islands. Multicriteria decision-making techniques are helpful in conservation planning because they can account for various, sometimes conflicting inputs and can enhance the transparency of decisions (Malczewski 1999, Laskar 2003).

Table 1.

Mexican islands with successful eradication projects.

Fig. 1.

Completed and pending eradications in Mexican islands.

Methods

To prioritize islands for restoration, we first defined the set of attributes used to rank islands. These included island size (surface area), distance from mainland, species richness, pre sence of endemic taxa, presence of threatened species, land use, presence of human population, likelihood of reinvasion, feasibility of successful eradication (e.g., given current technologies), and estimated economic cost (based on past eradication expenditures). We populated a database of those attributes for each of the 36 islands with invasive mammal species. Data were insufficient for 7 islands, so our prioritization analysis was based on 29 islands (Table 2).

Using expert input from conservation practitioners from the Mexican NGO Grupo de Ecología y Conservación de Islas, we developed decision rules (Table 3) and used the rank sum method (Malczewski 1999) to assign weights to attributes. Our schema prioritized islands with the highest presence of endemic taxa, followed by presence of important seabird nesting areas, highest number of species enumerated on the endangered species list, and highest species richness. We analyzed our data using the weighted linear combination procedure (Malczewski 2000) with ArcGIS 10 software (ESRI).

We conducted 3 different multicriteria analyses. We compared (1) outputs based only in biological considerations (e.g., island biodiversity value, including data from endemism, species richness, protected species, and important seabird areas); (2) outputs based only on “strategic” feasibility considerations (e.g., economic cost, feasibility of eradication, and probability of reinvasion); and (3) outputs based on the combination of both biological and strategic values. These analyses provided a comparison of islands where eradication campaigns could be implemented somewhat easily with “hotspot” islands where the biodiversity value may be greater.

Table 2.

Input information for the decision analysis.

Table 3.

Decision rules for prioritization analysis.

Table 4.

Weights set for the final analysis of biodiversity and strategy combined.

Table 5.

Priority matrix for eradication efforts on Mexican islands based on biodiversity scores and strategy. 1 = higher priority (no shading), 4 = lower priority (darkest shading).

Results

We categorized 29 islands into 4 priority groups, higher (1) to lower (4) (Table 4). The highest priority set included Socorro, Espíritu Santo, María Cleofas, María Magdalena, and Cozumel islands. The combined area of these islands is 95,813 ha. Thirteen invasive mammal taxa are found on these islands. Removal of those taxa could benefit 115 endemic taxa, 178 imperiled taxa, and numerous seabird nesting colonies.

The “biological” and “strategic” analyses provided very different results (Table 5), showing that more complex islands with higher species richness pose complex challenges to the success of eradication programs and that a few small, simple islands have the potential for successful, yet simply executed eradication programs. The “strategic” priority set are San Marcos, Cayo Centro, Coronado Sur, San Benito Oeste, Alcatraz, San Diego, Mejía, and Granito islands. Only 2 islands coincided in all priority categories: María Cleofas and Espíritu Santo.

Discussion

With this analysis, we provide a recommendation for strategic investment for the restoration of the Mexican Islands (Table 5; Fig. 2). We also provide a framework that can be used to evaluate archipelagos in other regions that may benefit from conservation action. The multicriteria analysis framework we developed provides many advantages, including that it can be adaptive and dynamic and can help improve the transparency and objectivity of decision-making. This sort of regional prioritization framework can also facilitate planning and implementation of eradications in a programmatic and strategic sequence designed to maximize efficiency and reduce investment risk (e.g., Saunders et al. 2011).

Given the importance we placed on endemism and species richness in our weightings, these characteristics prevailed over economic cost, feasibility of eradication, and reinvasion probability. As a consequence, many small islands in our candidate set, for which eradications may be relatively simple (e.g., Alcatraz, San Diego), rank lower in priority than other more diverse and complex islands. Islands in the tropical Pacific and Caribbean were our highest priority, largely based on species richness and endemism; however, eradications on some of these islands may be more difficult due to the presence of native mammals and large human settlements (e.g., Cozumel). We also note that some important seabird nesting colony islands were disadvantaged in our analysis if they were not on an island with high species richness; such was the case of San Benito Island. In cases where eradication of introduced taxa is not logistically or politically feasible, other approaches may be needed, such as invasive species population control or fencing of nesting areas or other sensitive resources.

This analysis can contribute to both the Estrategia Nacional Sobre Especies Invasoras en México (National Strategy for Invasive Species in Mexico) and the Estrategia nacional para la Conservación y el Desarrollo Sustentable del Territorio Insular Mexicano (National Strategy for the Conservation and Sustainable Development of the Mexican Insular Territory) by generating informa tion that can inform decision-making to prevent, control, and eradicate invasive species. Indeed, we are hopeful that this database will be used and improved into the future. A priority research need is to improve the database with information regarding native insect and plant taxa and to expand the utility of the database for evaluating eradication priorities of nonmammalian invasive alien species.

Fig. 2.

Priorities for eradication efforts on Mexican islands.

UPDATE

In December 2013, the Grupo de Ecología y Conservación de Islas, A.C., in collaboration with the Mexican government (SEGOB, SEMAR, SEMARNAT, CONANP, CONABIO), and with the support of the National Fish and Wildlife Foundation (NFWF), the Packard Foundation, and the Marisla Foundation, successfully conducted the eradication of a population of invasive mouse (Peromyscus eremicus) on San Benito Oeste, one of the islands included in our analysis of priorities.