Effect of Grazing on Restoration of Endemic Dwarf (Pinus culminicola Andresen et Beaman) Populations in Northeastern Mexico Javier Jime´nez,1,2 Enrique Jurado,1 Oscar Aguirre,1 and Eduardo Estrada1

Abstract lished: (1) seedlings protected from cattle plus small mam- A pilot experiment designed to test the effect of cattle, mals, (2) seedlings protected from cattle, and (3) seedlings small mammals, and elevation on the success of refores- with free access to cattle and small mammals. Seedling sur- tation of an endemic dwarf pine species in northeastern vival was approximately 50% in (1) after 4 years, but there Mexico was implemented. Pinus culminicola (Andresen et were no surviving seedlings with free access to cattle. Ele- Beaman) grows only in four high peaks in the Sierra Ma- vation in general did not account for variation in survival. dre Oriental and is under pressure from grazing, wildfires, Seedling growth was poor during the 4 years, which implies and human activities such as mining, road development for that seedlings remain susceptible to grazing and trampling timber extraction, and telecommunication and aerial navi- by cattle and small mammals. The implications for a large- gation devices. We planted and monitored 2-year-old - scale restoration program are discussed. lings at three elevations within the natural distribution range of this species at Cerro El Potosı´ in Nuevo Leo´n, Key words: cattle exclosure, grazing, Pinus culminicola, Mexico. At each elevation three treatments were estab- rehabilitation, seedling.

Introduction above sea level and is the highest peak in Northern Restoration of endemic species should be done with Mexico (Garcı´a & Gonza´lez 1991). Due to its geographi- knowledge of the species biology as well as an understand- cal isolation and peculiar geological strata, it has many ing of the community organization (Gross et al. 1989; rare, endemic and endangered species (Rzedowski Bowles & Whelan 1994; Teketay 1997). Endangered spe- 1978). The vegetation types at the summit of Cerro El cies recovery is always difficult, and biologists need to use Potosı´ are alpine and subalpine prairie, Pinus culminicola the best tools, skills, and experience available. Hence, the () matorral (shrubland), mixed forest of Pinus knowledge of the development of spatial pattern of the culminicola–P. hartwegii (Pinaceae), mixed forest of Pinus species in the communities under restoration allows for hartwegii–P. culminicola, mixed forest of Abies vejari a better understanding of the interactions between ani- (Pinaceae)–Pseudotsuga menziesii (Pinaceae)–Pinus hart- mals and in the ecosystem (Dale 2000). wegii–P. ayacahuite (Pinaceae), and forests of Pinus hart- Fragmentation and habitat loss are the main threats to wegii. Endemic and rare plant species in the area include the survival of most endangered species; hence, the in- Pinus culminicola, Juniperus sabinoides (Cupressaceae), clusion of dynamic processes of plant communities in res- Lupinus cacuminis (Fabaceae), Hackelia leontis (Boragi- toration plans would result in lower habitat loss and naceae), Erigeron potosinus (Asteraceae), Senecio hinto- fragmentation (Falk et al. 1996; Pickett et al. 1997; Tucker niorum (Asteraceae), and Arracacia schneideri (Apiaceae) & Murphy 1997; Huxel & Hastings 1999; Jime´nez et al. and the endangered plant species Ageratina potosina 1999). Lesica and Allendorf (1999) argue that a primary (Asteraceae), Astranthium beamanii (Asteraceae), Castilleja goal of restoration is the establishment of long-term viable bella (Scrophulariaceae), Stachys vulnerabilis (Lamia- populations that will restore ecosystem functions and pro- ceae), Poa mulleri (Poaceae), Potentilla leonine (Rosa- cesses, prevent erosion, and protect biological diversity. ceae), Trifolium schneideri (Fabaceae), Viola galeanaensis The protected natural area ‘‘Cerro El Potosı´’’ is located (Violaceae), and Arracacia. tolucensis (Apiaceae) (Mc- in the , in the southern part of the Donald 1990; Garcı´a&Gonza´lez 1991). Some animal spe- State of Nuevo Leo´n, Mexico; it has an elevation of 3,670 m cies in the area are also considered endangered (Golden Eagle [Aquila chrysaetos], Maroon-fronted Parrot [Rhyn- chopsitta terrisi], Harris’ (Bay wing) Hawk [Parabuteo uni- 1 Facultad de Ciencias Forestales, Universidad Auto´ noma de Nuevo Leo´ n, cinctus], American Gold Finch [Carduelis tristis], Black bear Apartado Postal 89 67700 Linares, Nuevo Leo´ n, Me´xico [Ursus americanus eremicus]) or endemic (Painted Redstart 2 Address correspondence to J. Jime´nez, email [email protected] [Myioborus pictus], Carmen mountain shrew [Sorex milleri], Ó 2005 Society for Ecological Restoration International Flat headed myotis bat [Myotis planiceps], Coyote [Canis

MARCH 2005 Restoration Ecology Vol. 13, No. 1, pp. 103–107 103 Effect of Grazing on Dwarf Pine Restoration latrans mearnsi]) (Guzma´n 1998; Garcı´a et al. 1999; Jime´nez Experiment et al. 2002). The experiment was laid out in December 1997 at three Dwarf pine (P. culminicola) was described by Andresen elevations and vegetation types: (1) high elevation at and Beaman (1961) as a new species endemic to Cerro El 3,500 m in Pinus culminicola matorral, (2) intermediate Potosı´. Later the same authors described the distribution elevation at 3,400 m in P. culminicola–P. hartwegii for- of the species as a total of 106 ha (Beaman & Andresen est, and (3) lower elevation at 3,300 m in P. hartwegii– 1966). In 1975 the species was also found in three other P. culminicola forest. At each elevation three square plots sites of the same Sierra (Sierra La Marta and Sierra La of 625 m2 were established for planting 2-year-old dwarf Viga in the state of Coahuila, and Cerro Pen˜aNevadain pine seedlings that had been grown in a seedling nursery. Nuevo Leo´n) with similar spatial patterns and altitudinal Three treatments were made for each plot: (1) small mam- distribution ranges to those determined for Cerro El Potosı´ mal plus cattle exclusion (chicken wire), (2) cattle exclu- (Riskind & Patterson 1975). In 1978 wildfires at Cerro El sion (barbed wire), and (3) free range. For each treatment Potosı´ burned 34% of the dwarf pine population (Garcı´a at each elevation, 110 seedlings were planted at a density 1989). During the past four decades, a reduction of the area of approximately 2 seedlings in each 10 m2. At the time formerly covered by dwarf pine has been observed, due to of planting, seedlings were approximately 100 mm tall and human impact involving timber extraction, road building, 5 mm in stem diameter. and land use change for the development of telecom- munication and aerial navigation infrastructure. A manga- nese mine was opened and later abandoned in the region Monitoring (Jime´nez et al. 1996). Recently (1998), several wildfires All planted seedlings were labeled so they could be drastically reduced the area formerly covered by dwarf distinguished from any naturally established seedlings. We pine at Cerro El Potosı´. Currently, only 30 ha of frag- monitored seedlings annually from November 1998 to mented dwarf pine area exist at Cerro El Potosı´, and these November 2001. Plant survival was determined by check- include many old trees with low seed production that are ing the position of each planted seedling. Stem diameter subject to cattle grazing (Jime´nez et al. 1999). The species and total height were measured for all surviving seedlings. is now considered endemic and is subject to special protec- To test whether grazing and elevation affected seedling sur- tion (Diario Oficial de la Federacio´ n 2000). There are no vival, we used a two-way analysis of variance (ANOVA). studies of P. culminicola populations in the other moun- To test whether elevation affected seedling growth in tains where it occurs; unfortunately these areas are not pro- diameter and height, we used one-way ANOVA. tected and are also subjected to grazing and wildfires. In spite of the uniqueness of the species, there are few published studies of the ecology of dwarf pine. Garza et al. (2002) described a high fungal diversity (51 species of mac- Results romycetes from 19 families and 42 genera) associated with P. culminicola. Aguirre et al. (2001) described the vegeta- Seedling Survival tion types of the area and examined the stem density pat- Both elevation and exclosure had an effect on seedling tern of P. culminicola and P. hartwegii and found the survival (p = 0.0016) for the duration of this experiment. former to be more abundant at higher elevations. In this Seedling mortality was high in all treatments, with only study we tested restoration techniques for P. culminicola, about 50% of seedlings surviving 4 years after planting in through replanting and the use of herbivore exclosures in the areas excluded for cattle and small mammals (Fig. 1). the protected natural area Cerro El Potosı´, within its natu- Seedling survival was significantly higher in areas ex- ral distribution range. cluded from both cattle and small vertebrates than in areas with cattle exclosures only. Survival was much lower where no exclosure was used (Fig. 1). Many seedlings in the free-range treatment were grazed and trampled, al- Methods though this did not happen uniformly to all seedlings. This implies that grazing is low and erratic in the area. How- Study Area ever, 2 years after planting, more than 80% of seedlings The study site is located at the top of the protected natural had died at the highest elevation. By the fourth year after area Cerro El Potosı´, in the south of the State of Nuevo planting, no free-range seedlings had survived at any ele- Leo´ n, Mexico (lat 24°320N, long 100°130W). This moun- vation. Where small mammals were not excluded, seed- tain is within the Sierra Madre Oriental range, between lings were found to be nibbled and occasionally had the Gulf Coastal Plain and the Mexican High Plateau. The broken stems. This damage was also present, although mountain top is exposed to high winds for most of the rarely, in the area excluded for cattle plus small mammals, year. Snow may fall several times during winter, persisting where it was attributed to the parrots that are common in for a few days on the north and northeast slopes (Garcı´a the area or to small mammals that may have climbed over 1996). or burrowed under the chicken wire fence.

104 Restoration Ecology MARCH 2005 Effect of Grazing on Dwarf Pine Restoration

Seedling Growth Mean stem diameter (Fig. 2) for all seedlings was low dur- ing 1998 (5.09 mm), 1999 (5.15 mm), and 2000 (5.66 mm), perhaps at the expense of root growth. In 2001 stem diam- eter increased to 8.07 mm. There were no differences in stem diameter that could be attributed to elevation or access to cattle and small mammals (Fig. 2). Seedlings doubled their height since planting in 1997 and 2001 (Fig. 3). The increase in height was smaller dur- ing 1998–2000 than in 2001, perhaps as a result of early

Figure 1. Mean percent seedling survival by treatments (i = small mammals plus cattle exclusion, ii = cattle exclusion, and iii = free range) and monitoring time for three elevations (3,500 (a), 3,400 (b), and 3,300 (c) m).

In general, elevation did not account for differences in seedling survival (Fig. 1); the patterns found for exclosures were consistent at all elevations. The fact that seedling sur- vival was constant between environments suggests that seed- ling mortality during these first years is the result of seedling vigor rather than competition. However, in 1999 seedling Figure 2. Mean stem diameter (±SD) by treatments (i = small mam- mortality was consistently greater at the highest elevation mals plus cattle exclusion, ii = cattle exclusion, and iii = free range) where the surrounding vegetation is Pinus culminicola mator- and monitoring time for three elevations (3,500 (a), 3,400 (b), and ral, perhaps as a result of a very harsh winter. 3,300 (c) m).

MARCH 2005 Restoration Ecology 105 Effect of Grazing on Dwarf Pine Restoration

Discussion In addition to low density of Pinus culminicola due to hab- itat loss and wildfires (Jime´nez et al. 1996), in the present study we have found low levels of natural and planted seedling recruitment to be a further threat to this endemic species. Naturally emerging seedlings were not detected during the 4 years of this study. It is possible that were consumed by rodents, as found for other habitats (Beagan et al. 1996), or that conditions for seedling estab- lishment occurred in events that do not happen annually, such as masting or above-average spring rainfall. Elevation in general did not account for variations in survival, which is in agreement with distribution of the species within Cerro El Potosı´ as described by Aguirre et al. (2001). Seedling growth was poor during the 4 years, which im- plies that seedlings remained susceptible to grazing and trampling by cattle and small mammals. In this study we have shown that free-range cattle severely limit seedling recruitment as suggested by Jime´nez et al. (1999); hence, its special protection status (Diario Oficial de la Federa- cio´n 2000) should be followed by management plans that exclude free-range cattle, and perhaps seed-consuming par- rots (Guzma´n 1998) and rodents (Beagan et al. 1996), from restoration areas. Seedling planting is a must given that natural germination was not observed during the study. Reforestation programs in Mexico do not traditionally evaluate survival of planted seedlings or saplings. How- ever, because the omnipresent cattle are typically not excluded, we believe that plant mortality must be very high (Jaffee 1997), and thus further studies dealing with the impact of herbivory and trampling after reforestation should be encouraged throughout Mexico.

Conclusions Seedling survival was poor after 4 years for seedlings pro- tected from cattle and small mammals, and no seedlings survived after being exposed to grazing and trampling for 3–4 years. Mortality was similar at all sites in spite of dif- ferences in environments and plant communities present Figure 3. Mean length (±SD) by treatments (i = small mammals at different elevations. After 4 years, surviving seedlings plus cattle exclusion, ii = cattle exclusion, and iii = free range) and were still very small and thus susceptible to trampling and monitoring time for three elevations (3,500 (a), 3,400 (b), and grazing by both cattle and small mammals. 3,300 (c) m). There were no naturally occurring seedlings found in this study; hence, it appears that replanting will be neces- sary to ensure the continued survival of this endemic spe- root growth (Fig. 3). The final mean seedling height was cies. In addition to protection of natural populations of approximately 120 mm. After 4 years in the field, plants dwarf pine from wildfires, reforestation programs that were still very short and hence in the layer where tram- include protection for seedlings from cattle and small pling and grazing by cattle and small mammals may occur. mammals should be implemented. There was a tendency for seedlings to be taller at the lowest elevation (3,300 m). Whether this is a result of bet- ter growing conditions or heavier shade remains to be tested. Seedlings excluded from small mammals also had Acknowledgments a tendency to be taller, perhaps as a result of occasional This research was supported by Fondo Mexicano para la nibbling of shoots of unprotected seedlings. Conservacio´ n de la Naturaleza and Programa de Apoyo

106 Restoration Ecology MARCH 2005 Effect of Grazing on Dwarf Pine Restoration a la Investigacio´ n Cientı´fica y Tecnolo´ gica from the Uni- Garza, F., J. Garcı´a, E. Estrada, and H. Villalo´n. 2002. Macromicetos, versidad Auto´ noma de Nuevo Leo´ n. ectomicorrizas y cultivos de Pinus culminicola en Nuevo Leo´ n. Ciencia UANL 5:204–210. Gross, E., I. Steinblums, C. Raltson, and H. Jubas. 1989. Emergency water- shed treatments on burned lands in southwestern Oregon. Pages 109– 114 in N. H. Berg, editor. Proceedings of the symposium on fire and LITERATURE CITED watershed management. General Technical Report GTR-PSW-109. Aguirre, O., J. Jime´nez, H. Kramer, and A. Akca. 2001. Ausscheidung U.S. Forest Service, Berkeley, California. und Strukturanalysen von Bestandestypen in einem unberu¨ hrten Guzma´n, A. 1998. Distribucio´ n altitudinal de la avifauna del Cerro El Naturwald Nordmexikos –als Grundlage fu¨ r die Biotopforschung. Potosı´, Galeana, Nuevo Leo´ n, Me´xico. Facultad de Ciencias Biol- Forstarchiv 72:17–25. o´ gicas, UANL, Me´xico. Andresen, J., and J. Beaman. 1961. A new species of Pinus from Mexico. Huxel, G., and A. Hastings. 1999. Habitat loss, fragmentation and restora- Journal of the Arnold Arboretum 42:437–441. tion. Restoration Ecology 7:309–315. Beagan, M., J. Harper, and C. Townsend. 1996. Ecology. Individuals, Jaffee, D. 1997. Restoration where people matter. Reversing forest degra- populations and communities. 3rd edition. Blackwell Science, dation in Michoaca´n Mexico. Restoration and Management Notes Malden, Massachusetts. 15:147–155. Beaman, J., and J. Andresen. 1966. The vegetation, floristics and phytoge- Jime´nez, J., O. Aguirre, E. Trevin˜ o, E. Jurado, and M. Gonza´lez. 1999. ography of the summit of Cerro Potosı´, Mexico. The American Patrones de desarrollo en un ecosistema de Pinus culminicola y P. Midland Naturalist 75:1–33. hartwegii. Revista Ciencia-UANL II:149–154. Bowles, M., and Ch. Whelan. 1994. Restoration of endangered species. Jime´nez, J., O. Aguirre, E. Trevin˜o, H. Villalo´n, M. Cotera, and E. Estrada. Conceptual issues, planning and implementation. Cambridge Uni- 2002. Area Natural Protegida Cerro El Potosı´, Galeana, Nuevo Leo´n, versity Press, Cambridge, United Kingdom. Me´xico. Gob. Edo de Nuevo Leo´n. Dale, M. 2000. Spatial pattern analysis in plant ecology. Cambridge stud- Jime´nez, J., H. Kramer, and O. Aguirre. 1996. Pinus culminicola.Zur ies in ecology. Cambridge, University Press, Cambridge United Entdeckung und Erhaltung einer mexikanischer Zwergkiefer. Forst Kingdom. und Holz 51:664–667. Diario Oficial de la Federacio´ n. 2000. Norma Oficial Mexicana—NOM- Lesica, P., and F. Allendorf. 1999. Ecological genetics and the restoration 059-ECOL-2000. Proteccio´ n ambiental-de especies de flora y fauna of plant communities: mix or match. Restoration Ecology 7:42–50. silvestres de Me´xico. DLXV 11, Mexico, D. F. McDonald, J. 1990. The alpine–subalpine flora of northeastern Mexico. Falk, D., C. Millar, and M. Olwell. 1996. Restoring diversity. Strategies Sida 14:21–28. for reintroduction of endangered plants. Island Press, Washington, Pickett, S., R. Ostfeld, M. Shachak, and G. Likens. 1997. The ecological D.C. basis of conservation. Heterogeneity, ecosystems and biodiversity. Garcı´a, A. 1989. Ana´lisis de la flora y vegetacio´ n de la cima del Cerro Chapman & Hall, New York. El Potosı´, Mpio. de Galeana, N.L. Me´xico. Fac. Biologı´a, UANL, Riskind, D., and T. Patterson. 1975. Distributional and ecological notes Mexico. on Pinus culminicola. Notes and News. Madron˜ o 23:159–161. Garcı´a, A., and S. Gonza´lez. 1991. Flora y vegetacio´n de la cima del Cerro Rzedowski, J. 1978. Vegetacio´ ndeMe´xico. Limusa, Mexico. Potosı´, Nuevo Leo´ nMe´xico. Acta Bota´nica Mexicana 13:53–74. Teketay, D. 1997. Seedling populations and regeneration of woody spe- Garcı´a, M. 1996. Ana´lisis de la cubierta vegetal y propuesta para la zonifi- cies in dry Afromontane forests of Ethiopia. Forest Ecology and cacio´ n ecolo´ gica del Cerro El Potosı´, Galeana, N.L., Mexico. Facul- Management 98:149–165. tad de Ciencias Forestales, UANL, Mexico. Tucker, N., and T. Murphy. 1997. The effects of ecological rehabilitation Garcı´a, M., E. Trevin˜ o, C. Cantu´, and F. Gonza´lez. 1999. Zonificacio´ n on vegetation recruitment: some observations from the wet ecolo´gica del Cerro El Potosı´, Galeana, Nuevo Leo´ n, Me´xico. Inves- tropics of North Queensland. Forest Ecology and Management tigaciones Geogra´ficas 38:31–40. 99:133–152.

MARCH 2005 Restoration Ecology 107