Sky Island Grassland Assessment: Identifying and Evaluating Priority Grassland Landscapes for Conservation and Restoration in the Borderlands
Total Page:16
File Type:pdf, Size:1020Kb
Sky Island Grassland Assessment: Identifying and Evaluating Priority Grassland Landscapes for Conservation and Restoration in the Borderlands David Gori, Gitanjali S. Bodner, Karla Sartor, Peter Warren and Steven Bassett September 2012 Animas Valley, New Mexico Photo: TNC Preferred Citation: Gori, D., G. S. Bodner, K. Sartor, P. Warren, and S. Bassett. 2012. Sky Island Grassland Assessment: Identifying and Evaluating Priority Grassland Landscapes for Conservation and Restoration in the Borderlands. Report prepared by The Nature Conservancy in New Mexico and Arizona. 85 p. i Executive Summary Sky Island grasslands of central and southern Arizona, southern New Mexico and northern Mexico form the “grassland seas” that surround small forested mountain ranges in the borderlands. Their unique biogeographical setting and the ecological gradients associated with “Sky Island mountains” add tremendous floral and faunal diversity to these grasslands and the region as a whole. Sky Island grasslands have undergone dramatic vegetation changes over the last 130 years including encroachment by shrubs, loss of perennial grass cover and spread of non-native species. Changes in grassland composition and structure have not occurred uniformly across the region and they are dynamic and ongoing. In 2009, The National Fish and Wildlife Foundation (NFWF) launched its Sky Island Grassland Initiative, a 10-year plan to protect and restore grasslands and embedded wetland and riparian habitats in the Sky Island region. The objective of this assessment is to identify a network of priority grassland landscapes where investment by the Foundation and others will yield the greatest returns in terms of restoring grassland health and recovering target wildlife species across the region. As a first step in identifying priority landscapes, we integrated two recent assessments of grassland condition to characterize the historical extent and current condition of grasslands across the Sky Island region. These assessments—the Apache Highlands Grassland Assessment (AHGA) and the New Mexico Rangeland Assessment (REA)—used similar condition classes and an expert-based approach to map the spatial extent of grasslands by condition class. For this analysis, the two assessments were combined into one spatial dataset by grouping finer-scale REA condition classes up to AHGA classes. Five condition types or classes resulted from the integration: (1) no to low woody increase, native grassland; (2) medium woody increase, native grassland; (3) no to low woody increase, non-native grassland; (4) medium woody increase, non- native grassland; and (5) high wood increase, former grassland. The combined spatial data set does not have sufficient information to identify grasslands in Mexico outside of the Apache Highlands ecoregion. To represent these grasslands, we used the Instituto Nacional de Estadística y Geographía land cover map which contains no information on current condition or spatial extent of former grasslands; these grasslands were classified as “undetermined”. Experts in the U.S. and Mexico identified 13,902,000 acres as current or former grassland which corresponds to almost half of the area we analyzed. Furthermore, grasslands were highly connected historically, allowing wide-ranging species, like bison, pronghorn and grassland birds, as well as species with more limited dispersal capabilities, like black-tailed prairie dogs, to move freely within and between these habitats. Only 2.6 million acres, or 19% of these grasslands, are still dominated by native grasses and remain relatively open and shrub free. Approximately 4.9 million acres, or 35%, have experienced more shrub encroachment but still fall into classes that are considered restorable with prescribed fire and other relatively cost-effective methods. However, shrub cover and associated soil erosion has exceeded a threshold on over 3.2 million acres, producing an irreversible type conversion from grassland to shrubland on almost one quarter of the historic grasslands and savannas in this region. The spread of non-native perennial grasses within grasslands has also been substantial. Boer lovegrass and, to a greater extent, Lehmann lovegrass are common on at least 1.5 million acres such that non-native grasslands with little to moderate woody increase comprise 11% of the area’s current and former grasslands. Implications of this ii spread are mixed, with some wildlife species more impacted than others. Native grasses remain present in many invaded areas, albeit at lower density. To assist in the identification and evaluation of priority Sky Island grasslands, we summarized existing information from a variety of sources on the distribution of sensitive grassland species and riparian-aquatic species that occur in wetland habitats embedded within grasslands which were identified as targets and priority species by NFWF and others. In addition, we summarized occurrence information for natural communities that have declined significantly over the last 100+ years due primarily to human impacts: ciénega wetlands, sacaton riparian grasslands, and sandy black grama grasslands. Occurrence records for grassland and riparian-aquatic sensitive species were obtained from the Natural Heritage Programs in Arizona and New Mexico. As a second step, we assembled several experts to both expand our understanding of grasslands across the region and to explicitly delineate priority grassland landscapes. Participants were asked to map areas they viewed as particularly promising for sustaining the region’s grasslands over the long term, drawing on their own knowledge and a variety of supporting information provided at the workshop. Spatial data sets provided to participants included the integrated grassland condition assessment described above, soil classifications, species occurrence locations, land management responsibility and protection status, and results of various groups’ efforts to identify priority conservation areas for other purposes. We seeded the discussion with suggested criteria for identifying, delineating, and evaluating these priority landscapes but participants were free to map as they saw fit. Suggested criteria included size of grassland blocks, ecological condition of those blocks, presence of embedded streams and wetlands and target wildlife species, fragmentation versus connectivity of grassland habitat, presence of intractable threats, and enabling conditions for long-term conservation success such as local community partnerships. Once polygons were drawn and met with group agreement, participants filled out a matrix of conservation value and feasibility for each. Criteria evaluated included many of the same ones used to inform the drawing of the polygons, but this exercise required participants to rate criteria for each site as low, medium, or high and, where necessary, explain those ratings. This group effort provided additional sources of information about distributions of target species and sensitive communities as well as enabling conditions for conservation. Participants identified twelve priority grassland landscapes (polygons), almost all containing substantial blocks of “high-quality grassland” that the combined grassland assessment had identified within the region (83% of overall acreage in native no to low woody increase classes), as well as substantial acreage of surrounding grassland patches deemed “restorable” (moderate woody increase). Landscapes differed in the amount and proportion of open intact grasslands within them. Four landscapes stand out as particularly noteworthy with over 150,000 acres of open intact grassland each: Upper San Pedro Mexico, Animas Valley-Sierra San Luis, Playas Valley-Janos Plain and Burro Cienega-Hachita landscapes. The 12 priority polygons also succeeded in capturing many of the target biological features associated with grasslands in the region. All but two patches of riparian sacaton grassland identified were included in the final polygons and ciénegas were well distributed in all but four of the priority landscapes. Looking at the occurrence of species of interest across these polygons iii revealed the value of the network as a whole. Some species like pronghorn and Chiricahua leopard frog (in embedded wetland habitats) are widespread, occurring in 10 out of 12 priority landscapes, while other targets occur in just one or two polygons. Of the 20 rare fish identified, 17 are found in three polygons or fewer, with most of these species living in just one or two stream systems each. Four landscapes are particularly notable in regard to native fish: San Bernardino Valley and Animas-Sierra San Luis with 7 and 9 species, respectively, of Yaqui drainage fish, and Empire-Cienega-Sonoita and Aravaipa-Muleshoe-Willcox with 10 and 11 species, respectively, of Gila drainage fish. A broader look at rare species in the region further demonstrated how well this network of polygons captured known locations for species dependent on grasslands or their embedded riparian habitats. Half of the species evaluated had >75% of their recorded locations falling inside these polygons. Less than a third of the species had the majority of their locations outside our network of polygons and only 1 species (with one occurrence) did not occur in any grassland polygon. Landscapes vary dramatically in land ownership/management patterns, as well as in protected status. Overall, private lands dominate the priority grassland landscapes. The degree to which these lands are managed for the benefit