www.mobot.org Research Home | Search | Contact | Site Map  

North America
South America
General Taxonomy
Photo Essays
Training in Latin

Wm. L. Brown Center
Graduate Studies
Research Experiences
  for Undergraduates

Imaging Lab
MBG Press
Climate Change
Catalog Fossil Plants
Image Index
Rare Books

Res Botanica
All Databases
The Unseen Garden
What's New?
People at MO
Visitor's Guide
Jobs & Fellowships
Research Links
Site Map


Páramo Ecosystem

Introduction Checklist of Vascular Plants Gazetteer References

Back to Páramo Flora Home

Text: James L. Luteyn
Photographs: Carmen Ulloa Ulloa and Peter Jørgensen, unless specified.

This "Páramo ecosystem" website is based primarily upon information published in the book Páramos: A Checklist of Plant Diversity, Geographical Distribution, and Botanical Literature (Luteyn, 1999, Memoirs of the New York Botanical Garden Volume 84).


U-shaped valley
U-shaped valley
Lake Patoquinoas
Lake Patoquinoas

"The páramo zone is the most interesting of the life zones of the Andes, since it shows to the highest degree, the struggle of plant and animal life against conditions of extreme cold temperature..." 

(Chardón, 1933)

General Definition

      Within the tropical regions of Mexico, Central and South America, Africa, Malesia including New Guinea, and Hawaii, there is a vegetation type that occurs between the upper limit of continuous, closed-canopy forest (i.e., forest line or timberline) and the upper limit of plant life (i.e., snow line) that is characterized by tussock grasses, large rosette plants, shrubs with evergreen, coriaceous and sclerophyllous leaves, and cushion plants.  This vegetation type is scattered along the crests of the highest mountain ranges or on isolated mountaintops between about 3000 meters (m) and 5000 m, like islands in a sea of forest.  Locally these areas are known as "zacatonales" (the Mexican and Guatemalan volcanic highlands), "páramo" (Central and northern South America), "jalca" (northern Peru), "puna" (drier areas of the altiplano of the central Andes), "afroalpine" and "moorland" (East Africa), and "tropical-alpine" (Malesia).
      Although neither Beard (1944), Troll (1958), nor Lauer (1981) felt the term "alpine" was appropriate for inter-tropical high altitude landscapes, these vegetation types have been more recently referred to as "tropical alpine" (Ramsay & Oxley, 1997;  Rundel et al., 1994;  Smith & Young, 1987) or "tropicalpine" (Smith & Cleef, 1988).  I do not like to apply the general term "alpine" to vegetation in the tropics, because it is a term derived from temperate regions.  Furthermore, high elevation tropical climates differ sharply from those of temperate alpine regions, particularly with respect to seasonality and diurnal patterns of temperature change (Rundel, 1994).  Some authors, including Walter (1973), Lauer (1981), and Monasterio and Vuilleumier (1986), advocate using the term páramo in the broadest sense possible, on a worldwide basis, for all high tropical montane vegetation above the continuous timberline for the sake of nomenclatural simplicity, if for no other reason.  Monasterio and Vuilleumier would simply add a geographic adjective to characterize the particular area being discussed, for example, Andean páramo, African páramo, etc.  Andean páramo is often compared with other alpine and arctic ecosystems (Baruch, 1979;  Billings, 1973, 1974, 1979;  Billings & Mooney, 1968;  Smith & Young, 1987).  For the purposes of this book, the term "páramo" is used in its regional sense, being restricted to the northern Andes of South America and adjacent southern Central America.  It is here used as a collective term for the entire landscape unit (or ecosystem) of the high altitude above continuous forest line and below the perpetual snow line.  Many different plant communities can be found in páramo, and will be discussed below, but the most widespread are dominated by tussock-forming grasses.
      The word páramo comes from the Latin word "paramus", according to the Diccionario de la Lengua Española (Real Academia Española, 1992).  Corominas (1973) states that this Hispanic-Latin word seemingly arose from the mid-western portion of the northern Iberian Peninsula, but that its exact origin is uncertain.  He further suggests that it may have been adopted by the Romans as a celticism, or instead is neither Iberian nor Celtic but originates from another Indoeuropean language in pre-Roman times.  In Spain, up until the epoch of the Conquest, the desertic plateau of arid Castile, which contrasts with the fertile regions lower down, was called "paramera."  Font Quer (1977) states that in Castilian Spanish the general significance of the word páramo is a flat plain.  The early Spanish explorers applied the word páramo to north Andean areas that were high, cold, inhospitable, wind- and rainswept, perhaps reminding them of the plains in their native Iberian Peninsula.  In Colombia, atmospheric moisture in the form of drizzle is often referred to as "paramitos," while in Ecuador the term "parameando" has come to mean "it is raining," and in Venezuela one might say "estoy emparamado" when one is getting wet because of rain and cold (Monasterio, 1980b;  Nuñes & Pérez, 1994;  Ramsay, 1992;  Vareschi, 1970;  Weber, 1958;  pers. comm. with local inhabitants).  For definitions and insight into conceptualizations of the páramo ecosystem by local "parameros" (inhabitants of the páramos), see López-Zent (1995) and Zambrano (1993).
      The páramo landscape has been influenced by glaciation. It is irregular and uneven, from jagged and very rough with erratic rocks to rolling or flat, often times with many small glacial lakes and tributaries.  It is the source of many of the large rivers of northern South America (e.g., Río Magdalena and Río Cauca of Colombia, Río Napo and Río Coca of Ecuador, and Río Orinoco of Colombia and Venezuela).  As will be seen from the discussions in this Introduction, there is no single definition of páramo, because it is characterized by a variety of geographic, geologic, climatic, physiognomic, and floristic features, all of which will be briefly touched upon.

Geographical Distribution

Geographical Distribution

      In the tropics of the Americas, the páramo ecosystem is discontinuously distributed between 11°N and 8°S latitudes.  It is concentrated in the northwest corner of South America, mostly in Venezuela, Colombia, and Ecuador, but there are outliers in Costa Rica, Panama, and northern Peru (see front and back endpapers, which show in black the area above 3000 m elevation that is potentially páramo).  The northernmost páramos are located in the Sierra Nevada de Santa Marta of Colombia, located at about 11°N latitude.  The westernmost páramos are located in Costa Rica, in the Cerro Buena Vista (Cerro de La Muerte) region of the Cordillera de Talamanca, at about 83°W longitude.  The easternmost páramos are located in north-central Venezuela, in the state of Lara, at about 70°W longitude.  The southernmost páramos (locally called "jalca") are found in northern Peru in the department of La Libertad, at about 8°S latitude, just north of the Cordillera Blanca.  Other neotropical areas that have páramo-like vegetation or that correspond ecologically and have sometimes been referred to as páramo, but that fall outside the geographical range or definition used in this book, include Pico Duarte in the Dominican Republic, the "zacatonales" of Mexico, Pico Naiguatá (Avila and the Silla de Caracas areas) in the Cordillera Costal of north-central Venezuela, Pico de la Neblina along the Venezuela/Brazil border, scattered "humid puna" areas of the eastern slopes of the Peruvian Andes, the "yungas" region of northeastern Bolivia, scattered areas in Chile and Argentina, and the Itatiaia area of eastern Brazil.
      For other opinions about the geographical distribution of páramo, see Brack Egg (1986b), Braun (1956), Cleef (1978, 1981b), Cuatrecasas (1968), Mann (1968), Monasterio (1980b), ONERN (1976), Ribera et al. (1994), Tosi (1960), Unzueta Q. (1975), Vareschi (1955), and Vuilleumier and Monasterio (1986).


Cloud-free, cold, crisp
Cloud-free, cold, crisp
Fog, and drizzle
Fog, and drizzle

      The páramos of Colombia and northern Ecuador are influenced by the intertropical convergence of air masses (a low pressure trough), because of their geographical location near the Equator.  They are generally humid throughout all or most months of the year with continuous moisture in the form of rain, clouds, and fog, mostly due to orographic uplift caused by the Andes.  Many páramos receive
Diagramas de Walter para localidades en los Andes de Ecuador
more than 2000 mm of rain on their exposed slopes (absolute range 500-ca. 3000 mm/year).  They have a high relative humidity averaging 70-85% (absolute range 25-100%).  This contrasts with páramos in the northernmost Andes of Venezuela, the Sierra Nevada de Santa Marta in northern Colombia, and in Costa Rica, where there is a marked dry season due to the influence of the northeast Trade winds (Herrmann, 1970, 1971;  Lauer, 1979a,b).  Páramo becomes driest near its southern limits in southern Ecuador and northern Peru, where they are influenced by two air masses, one from the Amazon Basin, which has already released its moisture on the eastern slopes, and another dry cool air mass from the West under the influence of the Humboldt Current.  In addition to the large scale climatic parameters given above, local microclimates may strongly influence regional weather patterns (G. Sarmeinto, 1986).  In northern Peru, the páramo environment grades almost imperceptibly into the "puna" ecosystem, which is more characteristic of the central and southern Andean altiplano highlands of central Peru, Bolivia, and Argentina.  Puna, in contrast to páramo, is typically xeric with lower humidity, a shorter wet season, and six months or more of almost no rain.  Figure 5 illustrates the overall climate of páramo by the use of climate diagrams from scattered localities.  For additional climate diagrams, see Cañadas Cruz (1983), Cleef (1981b), Guhl (1982), Jørgensen and Ulloa U. (1994), Monasterio and Reyes (1980), Van der Hammen et al. (1995), and Witte (1994).  For comparisons of páramo and puna, see reviews by Cabrera (1957, 1968, 1976), Lauer (1952), Quintanilla P. (1983), G. Sarmiento (1986), Troll (1968b), and Young et al. (1997). 
      Páramos have a generally cold and humid climate with sudden changes in the weather and a diurnal fluctuation in temperature from below freezing to as much as 30°C, which often results in a daily freeze-and-thaw cycle that has been referred to as "summer every day and winter every night" (O. Hedberg, 1964).  During the dry season, for example, Páramo Piedras Blancas (Venezuela), at 3700-4700 m elevation, shows extremes in air temperatures ranging from -5°C to -11°C at night to 25-30°C during the day, with freezing temperatures occurring 325-350 nights of the year (Pérez, 1987a, 1996b;  Pfitsch, 1994).  Although overall mean annual temperatures of páramo range from 2°C to 10°C, there is much greater contrast in the climate of higher elevation areas than is found in lower zones of the same mountain ranges.  Therefore, the environment becomes harsher and more severe for plant life as altitude increases (Javellas & Thouret, 1995).
      A typical day often begins cloud-free, cold, crisp, and occasionally windy until mid morning; then increased cloud cover from lower elevations, caused by convectional and orographic uplift, brings rain, sleet, fog, and drizzle for much of the afternoon;  clearing often occurs in the late afternoon or early evening.  Nights are always cold and usually clear with stars filling the skies;  however, frost is frequent in the high páramos and snow is common at the highest altitudes. During any given day of the year rain, ice, snow, and fog may alternate abruptly with clear sunny skies and elevated temperatures;  in one moment the wet cold necessitates heavy clothing, raincoats, and gloves, while in the next moment, lotion is needed to protect against sunburn.
      For other general references that discuss páramo climate, see Cuatrecasas (1968), Eidt (1968), Guhl (1974), Lauer (1981), Monasterio (1980a), Rundel (1994), Schnetter et al. (1976), J. Snow (1976), Sturm (1978), Sturm and Rangel Ch. (1985), Troll (1968b), Van der Hammen and Ruiz (1984), Van der Hammen and Santos (1995), Van der Hammen et al. (1983, 1989, 1995, In press), Weber (1958), and Witte (1994).


      The geology of the Andes is extremely variable and consequently so are the soils.  Most páramo soils are relatively young and only slightly developed, and are broadly classified into the orders Andosols, Inceptisols, Histosols, Entisols, and Mollisols (Buol, et al., 1980;  Buringh, 1979;  FAO, 1975;  Soil Survey Staff, 1975).
      Andosols and Inceptisols include older soil names such as Ando soils, brown tropical soils, black soils, Onji soils, humic allophane soils, hydro humic latosols, volcanic soils, and volcanic ash soils.  They are soils formed from and associated with volcanic ash, have a low supply of (or have lost) bases or iron and aluminum, and show moderate weathering.  In páramo, these may include Andepts and some Aquepts, as well as Tropepts and Umbrepts.  The Andosols most likely to form in páramos are Aquands, Cryands, and Udands.
      Histosols include older soil names such as bog soils, muck soils, organic soils, and peat soils.  They are soils that are highly organic and are found in very wet places such as bogs and swamps.  These may include Fibrists, Folists, Hemists, and Saprists.
      Entisols are soils that have little or no evidence of development (i.e., absence of horizons) and have a highly mineral nature.  These are often found near snow line and may include Aquents, Fluvents, Orthents, and Psamments.
      Mollisols are a less common soil group in the páramo, but are very dark colored and base-rich.  The most likely group to develop are the Aquolls.
      At its highest elevations (i.e., superpáramo), páramo soils are very shallow and coarse with a high percentage of rock and sand, there is little to no production of organic matter, and consequently, low water retention. Superpáramo soils are extremely infertile since without organic matter or fine grains they have practically no ability to hold exchangeable cations (Pérez, 1992c).  Furthermore, in the superpáramo, the soil surface is recurrently disturbed by needle-ice activity (a type of ground frost), and soil-moving phenomena such as frost-heaving and thawing and sorting of material is common (Pérez, 1987c).  In this part of the high páramo, the mean annual air temperatures are always low (4-6°C), but the cold does not penetrate very deeply into the soil.  Soil temperatures, at about 30 cm, more or less reflect those of the mean annual temperatures of the air (Lauer, 1979).  Soil temperature does, however, have a profound effect on nutrient and water availability, root growth, seed survival and germination, and vegetation zonation (Diemer, 1996;  Lauer, 1981;  Pérez, 1987b;  A. P. Smith, 1976;  Walter & Medina, 1969).
      At its middle elevations (i.e., grass páramo), páramo soils are relatively deep, humic, black or dark brownish, and acidic with pH ranging from about 3.7 to 5.5.  They are continuously moist or even saturated with water due to the daily formation of dew or frost and the water-retaining capabilities of the highly organic, peat-like content.
 In the lowermost part of the páramo (i.e., subpáramo), near Bogotá (Colombia) and elsewhere in the northern Andes, Sturm (1978) has stated that soils have in common:  dark color ("black coloured"), moderate to high pH and correspondingly low Ca levels, low free P content, relatively high K and N content and reduced uptake of these elements by plants, higher than 10% organic content in the top layer, little or no "podzolic" features, and high water capacity.
      For a discussion of the factors that help to form páramo soils, see Cortés Lombana (1982, 1995).  For soil types in local páramo studies, see Baruch (1979), Botero (1985), Hofstede (1995c), Pérez (1991c, 1992c, 1996b), Quintanilla P. (1983b), Rangel Ch. (1989), Salomons (1986), Sánchez M. et al. (1989), Sevink (1984), Sturm and Rangel Ch. (1985), Thouret and Faivre (1989), and Vis (1995).  For other general references about páramo soils, see also Del Llano (1990), Jenny (1948), Jenny et al. (1948), Simonson (1979), Sturm (1994a), Van der Hammen and Ruiz (1984), Van der Hammen and Santos (1995), Van der Hammen et al. (1983, 1989, In press), Vareschi (1970), Wright and Bennema (1965), and Zöttle (1970).

Paleohistory and Paleoecology

      Reconstruction of the paleoecology of páramo and high elevation montane forest has been the subject of study by Thomas Van der Hammen and his associates since the 1960s.  The following is a brief summary of an article by Van der Hammen and Cleef (1986) that emphasizes the paleohistoric events that gave rise to and further influenced development of the páramo ecosystem we see today in the high plain of Bogotá (Colombia).  The sequences may be similar or different in other parts of the northern Andes.  Unfortunately, detailed accounts from other areas are few or not yet available.
      The Andes began to arise during the Paleocene, and during the Miocene they were probably only ridges to low mountains up to ca. 1000 m elevation.  It was not until the beginning of the Pliocene or slightly later that the northern Andean region uplifted to its present altitudes.  During the Plio-Pleistocene, ca. 4-5 Ma (million years ago) there was abundant volcanic activity, at which time elevations above the present treeline came into being.  There may not have been forest yet at the high elevations, since considerable time was needed for the forest line to have risen from around 2500 m (the level at that time) up to around 3500 m.  The upper Andean forest and páramo belts evolved more or less simultaneously during the Late Pliocene or Early Pleistocene (2-4 Ma).  There is evidence that an early páramo vegetation, what Van der Hammen and Cleef call "protopáramo" vegetation, was present and consisted of Poaceae, Cyperaceae, Asteraceae, Ericaceae, a Polylepis-Acaena (Rosaceae) type of pollen, and Symplocos (Symplocaceae), Myrica (Myricaceae), Aragoa (Scrophulariaceae), Hypericum (Clusiaceae), Miconia (Melastomataceae), Ilex (Aquifoliaceae), Hydrocotyle (Apiaceae), Borreria (Rubiaceae), Ludwigia (Onagraceae), Polygonum (Polygonaceae), Valeriana (Valerianaceae), Plantago (Plantaginaceae), Ranunculus (Ranunculaceae), Myriophyllum (Haloragaceae), Jamesonia (Pteridaceae), and Hymenophyllum (Hymenophyllaceae).  By 1 Ma (for the high plain of Bogotá, Colombia, at about 2600 m), there is evidence for about 15 to 20 repeated alternations of forest and páramo (i.e., fluctuations of climatic zones) in interglacial and glacial periods. During this time genera of north temperate origin, such as Alnus (Betulaceae) ca. 1 Ma and Quercus (Fagaceae) ca. 0.3 Ma, appeared in the pollen record and must have passed over a Panamanian landbridge.
      During the later part of the Quaternary, ca. 44,000-21,000 yr BP (before present) of the Last Glacial stadial, glacials and interglacials continued to alternate with some short but severe cold periods.  At that time there were numerous changes in the proportions between páramo and forest elements, although the páramo flora became well established and dominated the scene.  Between ca. 45,000 yr BP and 25,000 yr BP there was a cold and wet period during which time glaciation reached its maximum advance.  Van der Hammen and Cleef further explain that during this time the glaciers and forest may even have been in contact at elevations between 2200 m and 2700 m, and the páramo belt must have been relatively narrow and wet.  On the contrary, between 21,000 yr BP and 14,000 yr BP there was a very cold but dry period during which time the mean annual temperature may have been 6-7°C lower than today.  Glaciation was not so extensive, but the páramo belt was broad and dry and covered most of the area above 2000 m (i.e., the area where present-day forest occurs).  This means that páramo vegetation covered a much greater area than it does today, and that many of the currently isolated páramos were then united.  It also means that the upper forest line was lowered by 1300-1500 m.
      At the beginning of the Holocene (ca. 10,000 yr BP) the climate became much warmer;  forest limits rose to elevations even higher than today and páramo vegetation was restricted to above 3300-3500 m.  The lower elevation forest included Dodonaea viscosa (Sapindaceae), Myrica (Myricaceae), Myrsine (=Rapanea) (Myrsinaceae), and Miconia (Melastomataceae);  it continued upwards with Alnus (Betulaceae), and ended at the highest elevations with Weinmannia (Cunoniaceae) and Quercus (Fagaceae).  During the period from 7500 yr BP to ca. 3000 yr BP temperatures rose about 2°C more, causing another upward shift in the forest line of about 300-400 m higher than today, and thereby reducing the area occupied by páramo.  Finally, at about 2900 yr BP, there was a noticeable lowering of the temperature that marked the last downward movement of the forest and páramo belts to their present-day positions.  Van der Hammen and Cleef state that the most important changes in the Holocene period have been the temperature changes mentioned above, the development of soils, and the development of peat bogs and soils with increased humidity.  They summarize by saying that the present-day páramo flora and vegetation is the result of an amalgamation of approximately 4-5 million years of complicated paleohistoric events.
      Some recent publications suggest that the situation may be more complex or may differ in other parts of the Andes than that described by Van der Hammen and Cleef (1986).  Colinvaux et al. (1997), for example, refute Van der Hammen's idea that Andean vegetation zones were compressed and moved in belts during Quaternary times.  Instead they suggest that during times of glacial cooling or Holocene warming plant associations showed different spatial diversity and were reformed according to the temperature tolerance of individual species, with heat intolerant species showing larger displacements.
      For additional general references about geology, glaciation, and paleohistory in the páramo regions of Central and South America, see Horn (1990b) and Weber (1958) for Costa Rica;  Graham (1973, 1989) and Markgraf (1989) for general Central and South America;  Salgado-Labouriau (1986) and Schubert and Vivas (1993) for Venezuela;  Helmens (1990), Hooghiemstra and Ran (1994), Van der Hammen (1981b, 1989), and Van der Hammen et al. (1973) for Colombia;  and Colinvaux et al. (1997), Hastenrath (1981), and Wolf (1892) for Ecuador.

Vegetation Zonation

Arenal del Cerro Amarillo
Arenal del Cerro
Laguna Patoquinoas
Laguna Patoquinoas
Páramo arbustivo
Páramo arbustivo

      Over the years numerous authors have given various names to the different vegetation zones and plant associations within the high Andes.  At times the usage of these names can become confusing and it is difficult to know exactly what is being discussed and/or compared, but see Huber and Riina (1997) and Jørgensen and Ulloa U. (1994) for summaries of this nomenclature.  When thinking of or talking about páramo, for example, reference is generally being made to the open, treeless grasslands with scattered espeletias and shrubs.  Much of the present day páramo vegetation of treeless grasslands, however, probably has anthropogenic origins, being maintained by cutting, periodic burning, and grazing;  practices mostly intensified within the last 300 years.  Therefore, the natural forest line at which forest ends and undisturbed páramo begins was probably higher than what is seen today and above 4000 m elevation in some places.  The evidence for this is that in an undisturbed system, there is usually not an abrupt end to the forest, not a sharp edge or border, but instead more of a transition from the tall forest trees to gradually shorter trees with increased elevation, then small trees and shrubs in a more or less thicket formation, and finally, above the forest line to the grasses, herbs, and scattered small shrubs of páramo. As one reaches the natural limits of one zone with the next, many of the plant species characteristic of these zones intermix.  It must also be remembered that the boundaries of the vegetation zones and the elevations at which they begin and end are not fixed.  Depending on environmental features, such as topography, exposure, soils, and general climatic conditions, and human intervention, timberline may occur anywhere between about 2800 m and 4000(-4800) m in the northern Andes.  Furthermore, on the wetter or windward side of a mountain or cordillera, the boundary between forest and páramo (i.e., the contiguous forest line) is higher than on the dry or leeward side (Lægaard, 1992;  Lauer, 1981, 1993;  A. P. Smith, 1975b,c, 1994;  Troll, 1958, 1959, 1968b, 1973;  Verweij, 1995).
      This book follows Cuatrecasas (1934, 1954, 1958, 1968) in dividing páramo ideally into three broad zones based on overall altitude and vegetation structure, with varying degrees of intergradation.  From the highest elevations to the lowest, these three zones are called superpáramo, páramo (here referred to as grass páramo), and subpáramo.  Superpáramo may be looked upon as the transition belt or ecotone between the permanent snow region above and the grass páramo below.  Subpáramo may be seen as the transition belt or ecotone between the grass páramo above and the montane forest below.  A brief summary of these three zones follows.  Other ideas of páramo zonation or modifications of Cuatrecasas' ideas may be found in papers by Monasterio (1980c) and Vareschi (1970) for Venezuela, by Cleef (1981b), Espinal and Montenegro (1963), Fosberg (1944), and Guhl (1982) for Colombia, by Acosta-Solís (1984), Cañadas Cruz (1983), Harling (1979), Jørgensen and Ulloa U. (1994), and Ramsay (1992) for Ecuador, by Brack Egg (1986b) for Peru, or by Cabrera (1957), Fosberg (1967), and Holdridge (1967) for more general systems of overall vegetation classification that include páramo.

      Superpáramo.  This is a narrow zone of vegetation growing on rocky scree and coarse, sandy soils below the snow-line from about (4000-)4500-4800(-5000) m altitude. Amongst the three páramo zones it is characterized by the lowest air temperature, precipitation, soil water-holding capacity, and nutrient content, and the highest solar radiation and night-frost frequency (Baruch, 1984).  Plants in this zone must be capable of enduring the daily extreme conditions of coldness and strong radiation, and regular or frequent snowfalls.  Superpáramo is the zone of least disturbance by man.  It is very localized because of its scattered occurrence only on the highest mountains and has very high endemism.  Some superpáramos occur on mountaintops that are high enough in elevation to have glaciers in their uppermost regions, while others are without permanent snow.  Vegetation referred to as desert páramo, or sometimes locally called "arenales," may have so much sand that they look like beaches.
      At first glance, superpáramo often looks from a distance to be bare ground, but is in reality home to tiny, clumped or scattered plants, such as Azorella pedunculata (Apiaceae), Hypochaeris sessiliflora, Senecio canescens, S. nivalis, S. adglacialis, S. supremus, S. comosus, Pentacalia gelida and Xenophyllum rigidum (=Werneria rigida) (Asteraceae), Draba pachythyrsa, D. depressa and Eudema nubigena (Brassicaceae), Arenaria spp. and Cerastium floccosum (Caryophyllaceae), Pernettya prostrata and Disterigma empetrifolium (Ericaceae), Astragalus geminiflorus, Lupinus alopecuroides, L. microphyllus and L. smithianus (Fabaceae), Geranium multipartitum (Geraniaceae), Luzula racemosa (Juncaceae), Nototriche jamesonii and N. chimborazoensis (Malvaceae), Aciachne pulvinata, Agrostis nigritella, Bromus oliganthus, Poa cucullata, P. trachyphylla, P. orthophylla and Stipa ichu (Poaceae), Plantago sericea ssp. nubigena (Plantaginaceae), Valeriana alpifolia (Valerianaceae), Viola pygmaea (Violaceae), the only páramo gymnosperms Ephedra americana and E. rupestris (Ephedraceae), and various species of cryptogams including vagrant ball-forming mosses and unattached lichens.

Viola pygmaea
Viola pygmaea
Ourisia chamaedrifolia
Ourisia chamaedrifolia
Ephedra americana
Ephedra americana

      Grass Páramo.  [Cuatrecasas referred to this zone as grass páramo, "páramo propiamente dicho", or páramo proper.]  From about 3500-4100(-4400) m vegetation of the grass páramo is continuous and plant cover is generally 100%. It is composed mainly of tussock- or bunch-grasslands dominated by species of Calamagrostis and/or Festuca.  There is a high proportion of dead shoots among the living that give a yellowish- or olive-brown to grayish look to the grasslands as a whole (Penland, 1941).   This is the classic area of the genus Espeletia and its relatives (Asteraceae: Espeletiinae), in local communities called "frailejonales", which for so many have come to symbolize páramo vegetation with their columnar, woolly, rosette-plant growth form. During October and November, in the higher Venezuelan páramos, there is no greater display of brilliant floral colors anywhere than when Espeletia schultzii (deep yellow) and Senecio formosus (violet-maroon) (Asteraceae), Castilleja fissifolia (bright red, yellow, and green) (Scrophulariaceae), and Chaetolepis lindeniana (intense magenta) (Melastomataceae) are in full bloom.
      The grass páramo is the most broadly circumscribed of the three páramo zones.  It includes not only the dominant grassland communities, but also the greatest number of azonal communities, which are determined by specific factors such as soil moisture and topography, and páramo growth forms.  The most frequently encountered azonal communities are described below under the grass páramo, but several are also found to a limited extent in the superpáramo and subpáramo zones as well.  The grassland communities of the grass páramo zone have suffered greatly from burning and grazing (see Impact of Burning and Grazing below).
      Grass páramo may consist of tall- and short- grass communities ("pajonales", "pastizales", or "prados") that include both herbaceous and woody vegetation, but they are dominated by tussock- or bunchgrasses. In the tall grass communities, with grass up to 1 m tall, Calamagrostis recta usually dominates on drier sites, while C. effusa dominates on moister.  Short grass communities usually indicate grazing and burning pressures and are often dominated by species of Agrostis, Festuca, and Paspalum.  The grass Aciachne pulvinata often forms low or flat cushions with very sharp (to the touch) leaves in drier-site short-grass páramos.  In Colombia, dwarf bamboos (Chusquea spp.) dominate on slopes with very wet climates in communities known as bamboo-brakes or "chuscales".
      If one takes the time to search between the bunchgrasses of the open grass páramo, there is an astonishing array of species of small herbs, e.g., Bomarea spp. (Alstroemeriaceae), Eryngium humile (Apiaceae), Perezia spp. (Asteraceae), Lysipomia spp. (Campanulaceae-Lobelioideae), Paepalanthus (Eriocaulaceae), Lupinus spp. (Fabaceae), Gentiana sedifolia, Gentianella spp. and Halenia spp. (Gentianaceae), Lachemilla spp. (Rosaceae), Gunnera magellanica (Haloragaceae), Sisyrinchium spp. and Orthrosanthus chimboracensis (Iridaceae), Oenothera epilobiifolia (Onagraceae), Ranunculus spp. (Ranunculaceae), Castilleja fissifolia and Bartsia spp. (Scrophulariaceae), the lycopods Huperzia spp., and the fern genus Jamesonia (Pteridacae). Common species rooted within the tussock formations include Cerastium spp. (Caryophyllaceae), Vicia andicola (Fabaceae), Geranium spp. (Geraniaceae), and Bromus lanatus (Poaceae).  Rumex acetosella (Polygonaceae) is a weedy adventive species that often forms a red landscape over large disturbed areas, usually after potato or wheat cultivation. Senecio niveo-aureus (Asteraceae) and related species in Senecio sect. Culcitium  add a stark beauty with their dense, woolly, whitish to pale yellowish pubescence, which covers practically the entire plant.  Acaulescent rosette-plants in the genera Oreomyrrhis (Apiaceae), Hypochaeris (Asteraceae), Acaulimalva (Malvaceae), and Acaena (Rosaceae), and cushion-plants in genera such as Werneria and Xenophyllum (Asteraceae), Draba (Brassicaceae), Arenaria (Caryophyllaceae), and Paepalanthus (Eriocaulaceae) are also common in the grass páramo.  Grass páramo is also rich in small and large shrubs, such as those genera mentioned in the subpáramo (below), but they are more scattered in the grass páramo. They include many species of Baccharis, Diplostephium, Gynoxys, Loricaria and Pentacalia (Asteraceae), Hypericum(Clusiaceae), Gaylussacia, Gaultheria, Pernettya and Vaccinium (Ericaceae), Arcytophyllum (Rubiaceae), and Valeriana (Valerianaceae).
      Swampy or boggy azonal sites, called cushion mires or "turberas", are common, especially in the uppermost grass páramo (Bosman et al., 1993;  Cleef, 1980a. Here species of the spectacular cushion plant growth form attain their best development, e.g., Azorella aretioides, A. multifida and A. pedunculata (Apiaceae), Oreobolus obtusangulus (Cyperaceae), the moss-like Distichia muscoides (Juncaceae), and Plantago rigida (Plantaginaceae).  The long-lived cushions often form the substrate for other smaller plants such as Hypochaeris spp. (Asteraceae), Carex spp. (Cyperaceae), Disterigma empetrifolium and Pernettya prostrata (Ericaceae), Gentiana sedifolia and Gentianella spp. (Gentianaceae), Agrostis spp. and Poa spp. (Poaceae), and Lachemilla spp. (Rosaceae).  These cushion "epiphytes" derive their water and nutrients from the process of litter decomposition and nurient release taking place within the cushion itself (Sklenár, 1998).  Other common species of swampy or boggy sites include Juncus spp. (Juncaceae), the páramo endemic Castratella piloselliodes (Melastomataceae), and Valeriana spp. (Valerianacee).  The lichen genera Cladia (Cladoniaceae) and Usnea (Parmeliaceae), mosses Campylopus (Dicranaceae), Breutelia (Bartramiaceae), Sphagnum (Sphagnaceae), and liverworts Riccardia (Aneuraceae), Frullania (Jubulaceae), and Lophozia (Jungermanniaceae) are also sometimes abundant.
      Other wet or flooded azonal communities such as marshes ("pantanos" or "ciénagas"), seeps, and springs, may also include bunchgrasses (especially Festuca dolichophylla), but with more species of sedges (Carex, Eleocharis, etc.) and mosses such as Drepanocladus (Amblystegiaceae) and Sphagnum (Sphagnaceae).  Also found are the various species of Lilaeopsis and Hydrocotyle (Apiaceae), Oritrophium peruvianum (Asteraceae), Plagiobothrys linifolius (Boraginaceae), Draba lindenii (Brassicaceae), Stellaria media (Caryophyllaceae), Halenia spp. (Gentianaceae), Myriophyllum quitense (Haloragaceae), Juncus spp. (Juncaceae), Huperzia spp. (Lycopodiacee),  Rumex tolimensis (Polygonaceae) to 4-5 m tall, Caltha sagittata and Ranunculus praemorsus (Ranunculaceae), and Mimulus glabratus and Pedicularis incurva (Scrophulariaceae).  Shallow pool, lake, and river communities (Figs. 15C-D) include Callitriche spp. (Callitrichaceae), Elatine spp. (Elatinaceae), Myriophyllum spp. (Haloragaceae), Isoëtes spp. (Isoetaceae), Cortaderia spp. (Poaceae), Potamogeton spp. (Potamogetonaceae), and Ranunculus spp.
      Rock ledge and cliff communities harbor another distinct group of interesting plants including Draba spp. (Brassicaceae), Luzula racemosa (Cyperaceae), Escallonia myrtilloides and Ribes hirtum (Grossulariaceae), Calceolaria spp. (Scrophulariaceae) and numerous fern species in the genera Asplenium and Cystopteris (Aspleniaceae) and Elaphoglossum and Woodsia (Dryopteridaceae).
      Trees of the genus Polylepis (Rosaceae), with their characteristic reddish, exfoliating bark and strangely contorted trunks and branches, may form localized, isolated woodlands or forest communities to over 4000 m elevation within otherwise grass páramo. They are often found on scree slopes, near the shelter of rock cliffs, or in river valleys. It is interesting that within the Polylepis forest there is a noticeable drop in plant diversity compared to the surrounding grass páramo (páramo species are probably shade intolerant), and furthermore that certain plants grow only under Polylepis. Polylepis forest is still relatively underexplored and an overall study of its flora would be very interesting (but see Arnal, 1983;  Fjeldså and Kessler, 1996;  and Hueck, 1960a).

L. microphyllus
Eryngium humile
Eryngium humile
Halenia spp.
Halenia spp.

      Subpáramo.  The lowest zone, called subpáramo, is also the most diverse and floristically showy.  It is a shrub-dominated transition zone at (2800-)3000-3500 m elevation made up of elements from forest below and the grass páramo above. It is often a mosaic of shrubs and small scattered trees, gradually reduced in size, giving way to scrub and low vegetation of dwarf shrubs, grasses, and herbs of the grass páramo above. Sometimes local changes in topography and soils may give way to different microclimatic conditions and so small patches of forest may be present. Where the ecotone between forest and páramo is abrupt, or where isolated shrub-tree islands are found within páramo, it is usually brought about by anthropogenic means, especially cutting, burning, and grazing.  Therefore, it is possible that subpáramo consists of nearly all secondary communities.
      The subpáramo communities, often known as scrub, thickets, "chaparrales"  or "mattorales", are composed predominantly of shrubby or woody vegetation that is sometimes lacking or rare in the Andean forest below.  These include species of the genera Ilex (Aquifoliaceae), Ageratina, Baccharis, Chuquiraga, Diplostephium, Gynoxys, Loricaria, Senecio and Stevia (Asteraceae), Berberis (Berberidaceae), Siphocampylus (Campanulaceae-Lobelioideae), Hypericum (Clusiaceae), Coriaria (Coriariaceae), Desfontainia (Desfontainiaceae), Bejaria, Cavendishia, Gaultheria, Macleania, Pernettya, Semiramisia, Themistoclesia and Vaccinium (Ericaceae), Brachyotum, Chaetolepis, Miconia and Monochaetum (Melastomataceae), Myrsine (Myrsinaceae), Monnina (Polygalaceae), Rubus (Rosaceae), Arcytophyllum (Rubiaceae), Aragoa and Calceolaria (Scrophulariaceae), Symplocos (Symplocaceae), and Ternstroemia (Theaceae).  Woody epiphytes in the Ericaceae (e.g., Disterigma, Plutarchia, Sphyrospermum, Thibaudia) and Loranthaceae (e.g., Aetanthus) (Figs. 10D and 14B) are also common.
      Genera forming low forests of small trees up to eight meters tall, often in tree islands known as "bosques achaparrados" include:  Oreopanax and Schefflera (Araliaceae), Gynoxys, Diplostephium and Senecio (Asteraceae), Buddleja (Buddlejaceae), Weinmannia (Cunoniaceae), Bejaria (Ericaceae), Escallonia (Grossulariaceae), Miconia (Melastomataceae), Myrsine (Myrsinaceae), and Hesperomeles (Rosaceae).
      Subpáramo has many common names, of which the most frequently used are: páramo forest, páramo thicket, shrubby páramo, subpáramo woodland, subpáramo chaparral, subpáramo elfin forest, and tropical subalpine forest (in English);  paramillo, paramito, bosquecillo de páramo, matorral de páramo, matorral de subpáramo, bosque paramero, bosque subparamero, bosque enano, bosque musgoso de subpáramo, bosquete andino, and chirivital (in Venezuela) (in Spanish);  and paramillo thicket and paramillo scrub (mixed Spanish and English).
      In general terms, subpáramo is the most difficult of the three páramo zones to define, because it has been greatly extended and expanded both horizontally and vertically (downwards) by human disturbance and habitat destruction over hundreds, or perhaps thousands of years.  This is especially due to cutting and burning at the upper forest line for herding and agricultural purposes.  Correspondingly, forest lines appear to have been lowered by as much as several hundred meters by human interference (Lægaard, 1992), and many species that were more restricted to relatively high elevations appear to have colonized lower life zones (Budowski, 1968).  Much of both subpáramo (non-controversial) and grass páramo (somewhat controversial) now occur in areas that were probably covered with upper montane forests in the past (Lægaard, 1992).

      Detailed classifications of páramo vegetation have been constructed utilizing relevés in the Zürich-Montpellier approach and described following the international "Code of Phytosociological Nomenclature" (Barkman et al., 1976).  This system divides vegetation into units or communities on the basis of floristic, physiognomic, and ecologic similarities, and usually employs a detailed classification of the vegetation consisting of a syntaxonomical hierarchy (subassociation, association, alliance, order, class, etc.).  For further explanation of this methodology, see R. Becking (1957), Braun-Blanquet (1964), Mueller-Dombois and Ellenberg (1974), and Westhoff and Van der Maarel (1973).  For applied examples of this methodology, with lists of species characterizing the various plant communities, see Aguirre and Rangel Ch. (1976), Cleef (1981b, including figs. 11-79), Franco R. et al. (1986), Lozano C. and Schnetter (1976), Sturm and Rangel Ch. (1985, including figs. 4-7), and Vargas Ríos and Zuluaga (1980).  Páramo studies using other quantitative methods may be seen in Baruch (1984) and Fariñas and Monasterio (1980) for Venezuela, and in Balslev and de Vries (1982), Grubb et al. (Unpublished), Muñoz et al. (1985), and Ramsay (1992) for Ecuador.

Gynoxys miniphylla
Gynoxys miniphylla
Monnina crassifolia
Monnina crassifolia
Rubus coriaceus
Rubus coriaceus

Morphological and Physiological Adaptations

      Since páramo is a high elevation tropical ecosystem, certain characteristic physical, chemical, and climatic features affect the biological functioning of the organisms that live there.  Therefore, the plants that grow in these areas must be adapted to:   1) high elevation air (with less water content and lower partial pressures of gasses such as O2 and CO2),  2) low temperatures (in the shade the annual mean temperature decreases about 0.6°C for each 100 meters increase in elevation, whereas in full sunlight, it increases with elevation because the atmosphere neither absorbs nor disperses as much radiation energy as in the lower regions),  3) intense ultra-violet radiation (highest in equatorial high mountains, but controlled by frequent fog and cloud cover), 4) rapid changes in insolation resulting in quick absorption or loss of heat,  5) drying effect of winds (by increasing transpiration to the point of desiccation), 6) physiological dryness (due to the combination of low temperatures, intense transpiration during sunny periods, and drying effects of winds, along with high soil acidity and high osmotic pressure of soil water making root water-absorption difficult), and 7)  physical damage from hail and possibly snow (Acosta-Solís, 1984;  Cleef, 1981b;  Cuatrecasas, 1968;  Little, 1981;  Mani, 1980;  G. Sarmiento, 1986;  A. P. Smith, 1981;  Vareschi, 1970;  Young, pers. comm.).  Consequently, growth and decomposition are slow, primary productivity is low, and natural succession of the vegetation takes a long time, especially when woody species are involved (Ferweda, 1987;  Hofstede, 1995c;  Horn, 1989, 1997;  Janzen, 1973;  Ramsay & Oxley, 1996;  Salamanca V., 1991;  L. Sarmiento M. et al., 1990;  A. P. Smith, 1981;  Sturm, 1978;  Williamson et al., 1986).  It is important to remember that there is no strong temperature seasonality, no marked change from summer to winter as in the temperate regions, but that in páramo growth is continuous throughout the year and that great changes in temperature (and to a certain extent precipitation) occur every day (diurnally).  Smith and Young (1987) noted that "Many aspects of morphology and physiology seem to provide escape from, or tolerance to, extreme diurnal climatic fluctuations."
      Recent studies have shown how giant Andean rosettes (e.g., Espeletia and Puya), have evolved adaptations that favor temperature insulation and the maintenance of a positive water balance under the severe conditions of the páramo environment.  Adaptations to low temperatures include freezing avoidance mechanisms such as supercooling of adult leaves (Goldstein, Rada & Azócar, 1985;  Larcher, 1975;  Rada R. et al. 1985a);  insulation by retention of dead leaves (marcescent leaves), which protects rosette stems from freezing (Goldstein & Meinzer, 1983;  Rada R. et al., 1985a;  A. P. Smith, 1979);  parabolic leaf geometry and nyctinastic movements of the leaves, which result in protection of the apical leaf bud from freezing (O. Hedberg, 1964;  Larcher, 1975;  A. P. Smith, 1974b);  thermal buffering by mucilaginous fluids secreted by the leaf bases, which protect apical buds (Smith & Young, 1987);  dense leaf pubescence, which reduces transpiration (Baruch, 1972) and increases leaf temperature (Baruch, 1975;  Meinzer & Goldstein, 1985;  Meinzer, Goldstein & Rada, 1994;  Miller, 1986, 1994);  tall aerial stems, which protect buds against the low minimum nighttime temperatures at ground level (Meinzer, Goldstein & Rada, 1994;  A. P. Smith, 1980);  and contractile roots in juvenile plants, which draw the developing stem into the ground (Smith in Smith & Young 1987).
      Rosette adaptations to low moisture levels include changes in the method of CO2 assimilation, such as a switch to the CAM photosynthetic pathway (Baruch & Smith, 1979;  Medina, 1974);  and a well-developed water-storing pith tissue, which can be used particularly during early morning hours when cold or frozen soils limit water uptake during that period of high transpiration (Goldstein & Meinzer, 1983;  Goldstein et al. 1984;  Meinzer & Goldstein, 1986;  Meinzer, Goldstein & Rundel, 1994;  Meinzer et al. 1985).  Some rosette plants, like Draba chionophila (Brassicaceae), which grows up to ca. 4800 m in the Venezuelan Andes, are freeze-tolerant (i.e., freezing injury occurs only when temperatures drop below the temperature at which extracellular ice formation begins) (Azócar et al., 1988;  Goldstein et al., 1994;  Pfitsch, 1994).  For reviews of general adaptive radiation in Espeletia and other plants of the high Andes, see also Beck (1994), Goldstein et al. (1994), Hedberg and Hedberg (1979), Monasterio (1986b), Monasterio and Sarmiento (1991), Ramsay (1992), Rundel et al. (1994), and Smith and Young (1987).
      Some of the physiological and morphological adaptations discussed above work in combination, that is, not only as avoidance/tolerance mechanisms against a cold environment, but also as useful adaptations against a hot environment resulting from frequent fires. Adaptations to high elevation by páramo plants result in the characteristic growth forms discussed below.

Growth Forms


      Most high elevation tropical plant communities have a characteristic physiognomy, which repeats in geographically disjunct areas of the world where they occur, for example, South America, East Africa, and Hawaii (Cuatrecasas, 1968;  O. Hedberg, 1964, 1992;  Hedberg & Hedberg, 1979;  Raunkiaer, 1934;  Troll, 1958;  Vareschi, 1970).  The growth forms (sensu lato) that characterize this physiognomy are examples of convergent evolution, the forms having evolved independently in several different plant families on distant continents in response to the unique high altitude tropical environments.  These growth forms often result from the ecological and morphological adaptations mentioned above.  Many are also apparantly adapted to survive fire (Lægaard, 1992;  Young & León, 1991). 
      While the tree is the dominant growth form in the forest, it is essentially absent from páramo (except the genus Polylepis and a few associated species).  Apparently it cannot survive at such altitudes.  In fact, tree growth ceases when soil tempertures drop to 6-10°C (Larcher, 1975;  Lauer, 1979a, 1981;  Walter & Medina, 1969a,b).  Páramo, however, has its own important and conspicuous growth forms (not all of which are strictly found in páramo), such as bunchgrasses, rosettes plants, cushion plants, microphyllous and dwarf shrubs, vagrant plants, and geophytes.  The presence of rosette plants (some of giant size), for example, is probably a good general indicator of páramo and seems to be one of the clearest distinctions between high elevation areas of tropical and temperate latitudes (A. P. Smith, 1994).  The most characteristic growth forms of the high elevation páramos are summarized below.  See also Balslev and de Vries (1991), O. Hedberg (1964, for comparisons with Afroalpine plants), Hedberg and Hedberg (1979), and Ramsay and Oxley (1997).

      Rosette Plants.  This growth form gives páramo vegetation its distinctive character. Two kinds of rosette plants have been described:
      Stem rosette.--  The most typical and well known growth form of páramo is the columnar woolly rosette plant.  Members of the genus Espeletia (Asteraceae), the so-called "frailejón" (or literally translated "big friar") because of the grayish woolly coat of pubescence, are the classical example.  These plants produce an erect, normally unbranched, thick-woody stem tightly encased by the dense bases of old leaves (Cuatrecasas, 1968).  The erect stems may be as tall as 15 meters in undisturbed páramo. Lateral inflorescences are produced from the single aerial meristem. Other examples of this growth form are the fern Blechnum schomburkii (Blechnaceae) and Plantago sericea ssp. perrymondii (Plantaginaceae). It has been shown that woolliness is a response to ultra-violet light and is associated with thermoregulation, the retarding of evaporation, and general protection from UV light (Miller, 1994).
      Acaulescent rosette.--  These plants develop thick, perennial, tap-roots and a dense rosette of leaves at the ground level.  The flowering stems may be very short with the flowers hidden in and amongst the leaf bases, or longer thereby lifting the flowers above the ground surface.  Some plants, such as Puya spp. (Bromeliaceae), produce giant, bulky inflorescences several meters tall. In all cases, the buds that form the flowers originate in the axils of the rosette leaves, which are sunken a few centimeters below the ground.  Acaulescent plants thereby protect their buds from fire and frost.  Experiments have shown that the mean temperature within the rosette is higher (by about 6°C) than in the surrounding air (Hedberg & Hedberg, 1979), and that these plants seem to buffer temperature variations to such an extent that they are able to avoid both positive and negative extremes (Goebel, 1891;  O. Hedberg, 1964).  Examples are shown by the genera Hypochaeris and Werneria (Asteraceae), Lysipomia (Campanulaceae-Lobelioideae), Paepalanthus (Eriocaulaceae), Lupinus alopecuroides (Fabaceae), Acaulimalva and Nototriche (Malvaceae), Rumex tolimensis (Polygonaceae), Ranunculus gusmanni (Ranunculaceae), Acaena cylindristachya (Rosaceae), Valeriana plantaginea (Valerianaceae), and Viola (Violaceae).

      Cushion Plants.  These plants form a flat, convex, or hemispherical cushion as the result of the regular outward branching of dense radially oriented buds. Each branch has a small rosette of leaves at the tip and only the outer and upper leaves are green and living;  the interior of the cushion consists of a peaty mass, the remains of dried leaves, accumulated humus, dust, soil, and rain water, all of which protect the buds and stems from wind, desiccation and predation, and provide a reservoir of water and nutrients (Sklenár, 1998).  Like in rosette plants, the mean temperature is higher within the cushion than at the cushion surface, thereby protecting the buds from cold temperatures.  The plants are often very prickly and hard to the touch, yet firm enough that one is able to walk on top of certain species, such as in Azorella (Apiaceae), Plantago (Plantaginaceae) (Figs. 9C and 19), and Distichia (Juncaceae). Many different species, in different families, form cushions of different sizes, from flat or only a few centimeters tall to cushions over one meter tall and several meters in diameter, for example Azorella pedunculata.  In moist or humid sites, where cushions are more frequently found, Distichia muscoides and Plantago rigida are dominant species, while in drier places Azorella pedunculata and A. aretioides are common. Other genera that produce cushions include Werneria pygmaea and Xenophyllum spp. (Asteraceae), Draba aretioides (Brassicaceae), Arenaria spp. (Caryophyllaceae), Oreobolus spp. (Cyperaceae), Disterigma empetrifolium and Pernettya prostrata (Ericaceae), Paepalanthus lodiculoides (Eriocaulaceae), Geranium spp. (Geraniaceae), Aciachne pulvinata (Poaceae), Calandrinia acaulis (Portulacaceae), Valeriana rigida (Valerianaceae), and Xyris subulata (Xyridaceae). See O. Hedberg (1964, 1992), Heilborn (1925), Sklená_ (1998), and Rauh (1939) for extensive discussions of morphology and adaptation of cushion plants to high elevation environments.

      Bunchgrasses (or Tussock Grasses).  This growth form is the most widespread in the páramo.  In undisturbed areas, grasses may average 1-1.5 m tall with a coverage of up to 100%.  Members of the grass and sedge families frequently form tufts or dense bunches of stems (culms) with rigid, pointed, tubular or inrolled leaves. These dense tufts in which the dead leaves are maintained and decay on the plant, along with the culms, provide good insulation for the buds and young leaves from cold temperatures, high radiation, evaporation, and high heat of fires to 500°C (Ramsay, 1992;  Ramsay & Oxley, 1996).  Much of their regeneration takes place through the production of vegetative buds near the ground.  Here the tufts are very dense and living shoots are found along with dead culms and leaves.  These tufts protect the vegetative buds.  The most common species are Calamagrostis recta and C. effusa (Poaceae); other important genera and species are Carex and Uncinia (Cyperaceae), Cortaderia spp., Festuca dolichophylla, F. tolucensis, Stipa spp., and Lorenzochloa erectifolia (Poaceae).  See Hofstede (1995c), Ramsay (1992), and Ramsay and Oxley (1997) for additional discussion of how the bunchgrass growth form is an adaptation to fire in the páramo environment.

      Microphyllous Shrubs.  These shrubs are characterized by dense foliage of small, xeromorphic leaves, sometimes with many of the following combinations of adaptations in the same species, all acting as protection from ultraviolet light and/or the reduction of transpiration (O. Hedberg, 1964;  Larcher, 1975).  Examples of genera and/or species with hard or sclerophyllous leaves include Gaultheria anastomosans and Gaylussacia buxifolia (Ericaceae), and Miconia summa (Melastomataceae);  squamous or rolled leaves Baccharis revoluta and Diplostephium revolutum (Asteraceae), and Miconia salicifolia (Melastomataceae);  imbricate leaves Loricaria (Asteraceae) and Aragoa cupressina (Scrophulariaceae);  aciculate or spine-tipped leaves Chuquiraga (Asteraceae), Hypericum laricifolium (Clusiaceae), and Valeriana microphylla (Valerianaceae);  and densely tomentose-pubescent leaves Diplostephium eriophorum and Pentacalia guicanensis (Asteraceae), and Gaultheria lanigera (Ericaceae).
      The so-called bouquet plants ("plantas en ramilletes de florones") are a subset of microphyllous shrubs with a special habit.  These plants feature an increase in the size of an individual flower in relation to the total appearance of the plant (e.g., Bidens humilis), or the dense aggregation of many small flowers into a bouquet of dense flowering stalks (Vareschi, 1970).  This is considered an adaptation to make the flowers more attractive to pollinators by forming a large and noticeable splash of color that attracts insects from a distance, as in species of Draba (Brassicaceae), Gentianella (Gentianaceae), and Chaetolepis (Melastomataceae).

      Prostrate Dwarf Shrubs.  These are small, woody plants that rarely produce shoots over 0.75 m tall.  The special feature separating them from microphyllous shrubs is that they have a larger part of their branch system protected below or upon the soil surface.  They are often prostrate, growing laterally along the ground.  Sometimes entire branching systems occur underground and only the current year's growth is seen above ground.  This growth form often has its regenerative buds below ground where they are protected from fire and frost.  Examples are found in Bidens and Senecio (s.l.) (Asteraceae), Lupinus and Astragalus (Fabaceae), Pernettya and Disterigma (Ericaceae), and Arcytophyllum (Rubiaceae).

      Geophytes.  These are herbs that survive the unfavorable periods of the year (including times of fire) by means of subterranean organs, such as succulent roots, rhizomes, stolons, tubercules, or bulbs (Lægaard, 1992;  Raunkiaer, 1934;  Vareschi, 1970).  Examples of geophytes include Orthrosanthus chimboracensis (Iridaceae), the genus Stenomesson (Amaryllidaceae), Altensteinia and Gomphychis (Orchidaceae), and the fern Ophioglossum crotalophoroides (Ophioglossaceae).

      Vagrant Plants.  These plants grow free, unattached to the substrate, and are found in many biomes throughout the world (Pérez, 1994a, 1997b).  They are found in the superpáramo zone, where frost-heaving is a common phenomenon, and only in cryptogamic plants such as the fruticose lichen Thamnolia vermicularis (family uncertain) and an acrocarpous moss Grimmia longirostris (Grimmiaceae).  These plants have been variously referred to as erratic, vagant, vagrant, solifluction floaters, errant cryptogams, and sometimes globular mosses or moss balls when their growth shape becomes more spherical (see Pérez, 1997b for more details).  Cryptogams as a whole (i.e., bryophytes and lichens) have been considered a true growth form by Cuatrecasas (1968), Griffin (1979), and Ramsay (1992), or as composed of several growth forms by Ramsay and Oxley (1997).


      On a geological timescale the páramo flora is young, the so-called protopáramo vegetation of Van der Hammen and Cleef (1986) having evolved during the Late Pliocene or Early Pleistocene, some 2-4 Ma.  Páramo environments suitable for plant colonization on an extensive scale, however, have only been available since the Quaternary (Simpson, 1975;  Van der Hammen, 1974;  Van der Hammen & Cleef, 1986).  The youth of the páramo flora is also evidenced by the presence of relatively few endemic or near endemic genera (23 genera, ca. 5% based on this work) and the absence of any endemic families in the vascular flora.  Although the páramo ecosystem occupies no more than 2% of the land area of the countries in which it is found, the flora is extremely diverse.  In fact, the páramo flora is the richest high mountain flora of the world (Smith & Cleef, 1988).  The páramo flora has evolved in various ways:  by adaptation of lower elevation plants (i.e., tropical elements) to high elevation environments, by immigration (i.e., dispersal) of cool-adapted plants from north and south temperate regions, and by speciation through isolation from within (i.e., autochthonous element).  For discussion of these and other ideas, see Chapman (1917), Chardon (1938), Simpson (1975), Simpson and Todzia (1990), Van der Hammen (1972a,b), and F. Vuilleumier (1970).

Table I. Numbers of families, genera, and species of major plant groups in the páramo.

Taxonomic Group

No. Families

No. Genera

No. Species

Non-Vascular Plants 130  365 1298
  Lichens* 45 114 465
  Mosses 51 163 544
  Hepatics 34 88 291
Vascular Plants  124 500 3399
  Ferns/Fern Allies 22 52 352
  Gymnosperms  1 1 2
  Angiosperms  101 447 3045
    Monocots 16 101 634








*Lichenicolous fungi are not included in this table.

      General Floristic Diversity.  For the non-vascular plants, the lists herein presented include 114 genera in 45 families of lichens (excluding lichenicolous fungi), 163 genera in 51 families of mosses, and 88 genera in 34 families of hepatics, for a total of 365 genera and 1298 species (see Table I).  For the vascular plants, the lists include 52 genera in 22 families of ferns and fern allies, 1 genus in 1 family of gymnosperms, 101 genera in 16 families of monocotyledons, and 346 genera in 85 families of dicotyledons, for a total of 500 genera and 3399 species (see Table I).  Tables II-VI show the largest families and genera of páramo lichens, mosses, hepatics, ferns and fern allies, and flowering plants, respectively.  Table VII lists the genera endemic to páramo and notes those that are monotypic.  Luteyn (1992) provided a preliminary estimate of specific endemism of páramo vascular plants as high as 60%.  However, it is now realized that species numbers, limits, and distribution patterns are too poorly known and that a realistic approximation of species endemism in the páramo is not yet possible.  As a result, detailed estimates of overall specific endemism and the geographical origins and relationships of the páramo flora have not been calculated from the data herein presented.

      Non-Vascular Plants. --  In this study, 1298 species of non-vascular plants have been found throughout the geographical and elevational range of páramo as here defined (Table 1), of which 36% are lichens (lichenicolous fungi are not included), 42% mosses, and 22% hepatics.
      Within the lichens, Table II shows the 10 largest families and genera of páramo lichenized fungi.  At the family level, Parmeliaceae (25 genera and 159 species) are by far the most diverse, with over three times as many genera and species as the next closest families Physciaceae (8 gen.) and Cladoniaceae (45 spp.), respectively.  Four of the ten most speciose páramo genera are also Parmeliaceae -- Hypotrachyna (50 spp.), Oropogon (21 spp.), Xanthoparmelia (18 spp.), and Parmotrema (13 spp.) -- accounting for just over 100 species.  The second most speciose family is Cladoniaceae, with Cladonia having 38 species.  Additional general comments about lichens are given by Ahti (1992) and Sipman (1992, and below in the introduction to his checklist).
      Table III shows the 10 largest families and genera of páramo mosses.  At the family level, Dicranaceae (17 gen. and 67 spp.), Bryaceae (10 gen. and 65 spp.), and Pottiaceae (19 gen. and 63 spp.) are the most diverse.  The most speciose genera are Campylopus (Dicranaceae, 37 spp.), Sphagnum (Sphagnaceae, 27 spp.), and Zygodon (Orthotrichaceae, 21 spp.).  Additional general comments about mosses are given below by Churchill and Griffin in the introduction to their checklist.  Table IV shows the 10 largest families and genera of páramo hepatics.  Lejeuneaceae (16 gen. and 38 spp.) and Jungermanniaceae (11 gen. and 31 spp.) are the most diverse families in terms of both genera and species.  The most speciose genera are Riccardia (Aneuraceae, 20 spp.), Metzgeria (Metzgeriaceae, 20 spp.), Plagiochila (Plagiochilaceae, 18 spp.), Frullania (Jubulaceae, 13 spp.), and Bazzania (Lepidoziaceae, 13 spp.).  Additional general comments about liverworts are given below by Gradstein in the introduction to his checklist.

Table II Ten largest families and genera of páramo lichenized fungi*
(prepared by H. Sipman).

Family (no. gen./spp.)

Genus (and family) (no. spp.)

Parmeliaceae (25/159) Hypotrachyna (Parmeliaceae) (50)
Cladoniaceae (3/45) Cladonia (Cladoniaceae) (38)
Physciaceae(8/27) Leptogium (Collemataceae) (25)
Collemataceae (2/26) Oropogon (Parmeliaceae) (21)
Lobariaceae (3/24) Stereocaulon (Stereocaulaceae) (19)
Stereocaulaceae (1/19) Xanthoparmelia (Parmeliaceae) (18)
Peltigeraceae (3/13) Heterodermia (Physciaceae) (17)
Ramalinaceae (1/12) Sticta (Lobariaceae) (13)
Lecanoraceae (4/11) Parmotrema (Parmeliaceae) (13)
Pannariaceae (4/10)

Ramalina (Ramalinaceae) (12)

*Lichenicolous fungi are not included in this table.

Table III. Ten largest families and genera of páramo mosses
(prepared by S. P. Churchill and D. Griffin III).

Family (no. gen./spp.)

Genus (and family) (no. spp.)

Dicranaceae (17/67) Campylopus (Dicranaceae) (37)
Bryaceae (10/65) Sphagnum (Sphagnaceae) (27)
Pottiaceae (19/63) Zygodon (Orthotrichaceae) (21)
Bartramiaceae (7/40) Bryum (Bryaceae) (18)
Orthotrichaceae (3/36) Leptodontium (Pottiaceae) (16)
Sphagnaceae (1/27) Orthotrichum (Orthotrichaceae) (14)
Amblystegiaceae (9/19) Breutelia (Bartramiaceae) (13)
Brachytheciaceae (7/18) Daltonia (Daltoniaceae) (13)
Polytrichaceae (6/16) Macromitrium (Macromitriaceae) (13)
Grimmiaceae (4/17) Schizymenium (Bryaceae) (11)

Table IV. Ten largest families and genera of páramo hepatics
(prepared by S. R. Gradstein).

Family (no. gen./spp.)

Genus (and family) (no. spp.)

Lejeuneaceae (16/38) Riccardia (Aneuraceae) (20)
Jungermanniaceae (11/31) Metzgeria (Metzgeriaceae) (20)
Lepidoziaceae (6/29) Plagiochila (Plagiochilaceae) (18)
Aneuraceae (2/21) Frullania (Jubulaceae) (13)
Metzgeriaceae (1/20) Bazzania (Lepidoziaceae) (13)
Plagiochilaceae (2/19) Anastrophyllum (Jungermanniaceae) (8)
Geocalycaceae (7/18) Lepidozia (Lepidoziaceae) (8)
Gymnomitriaceae (5/14) Leptoscyphus (Geocalycaceae) (7)
Jubulaceae (1/13) Isotachis (Balantiopsidaceae) (6)
Balantiopsaceae (2/7) Cephaloziella (Cephaloziellaceae) (6)
  Marsupella (Gymnomitriaceae) (6)
  Radula (Radulaceae) (6)

      Vascular Plants. -- In this study, 3399 species of vascular plants have been found from throughout the geographical and elevational range of páramo (Table 1), of which 10.4% are ferns and fern allies, 0.06% gymnosperms, and 89.6% angiosperms.  Of the angiosperms (flowering plants), 21% are monocots and 79% are dicots.
      Table V shows the largest families and genera within the páramo ferns and fern allies.  Dryopteridaceae, Lycopodiaceae, Polypodiaceae, and Pteridaceae are the largest and most diverse families at the generic and specific levels.  At the generic level, Polypodiaceae (14 gen.) are the most diverse, with twice as many genera as the next closest family Pteridaceae (7 gen.).  At the specific level, Dryopteridaceae are the largest due to the numerous species of Elaphoglossum (65 spp.), this genus is also the largest of the páramo pteridophytes. Huperzia (Lycopodiaceae) with 60 spp. is the second largest genus (or the largest with 69 if considered as Lycopodium in the broad sense), followed by Hymenophyllum (20 spp.), Isoëtes and Jamesonia (18 spp. each).
      Table VI shows the composition of the páramo flowering plants in terms of the 15 largest families and genera.  In this study, the Asteraceae are the largest family by far in both numbers of genera and species.  The data from this study show that Asteraceae are two and a half times larger than the Poaceae in numbers of genera (101 gen. vs. 41 gen.) and nearly four times larger in numbers of species (858 spp. vs. 227 spp.).  Four of the five most speciose páramo genera are Asteraceae -- Pentacalia (89 spp.), Senecio s.s. (69 spp.), Diplostephium (70 spp.), and Espeletia s.s. (61 spp.) -- accounting for nearly 300 species.  Asteraceae also have the highest number of endemic genera of any páramo vascular plant family (16 genera, or 70% of the endemic páramo genera).  The largest genera of Poaceae are Festuca (38 spp.), Calamagrostis (36 spp.), Agrostis (24 spp.), and Poa (20 spp.).  The Orchidaceae are surprisingly diverse in the páramo with 25 genera and 152 species herein recorded, although (as discussed below) their tabulation has proven difficult and their numbers may be questioned.  It is interesting to note that in virtually all páramo studies that have been published, the Asteraceae are always the largest family in numbers of species and genera, followed closely by Poaceae.  Other families (in alphabetical order) that rank consistently high in overall importance in the páramo ecosystem include Apiaceae, Brassicaceae, Bromeliaceae, Cyperaceae, Ericaceae, Gentianaceae, Melastomataceae, Orchidaceae, Rosaceae, and Scrophulariaceae.  There is also a trend in the upper páramo (i.e., superpáramo at ca. 4000+ m) for certain families, often with a more north-temperate element, to become increasingly important.  For example, the Apiaceae, Brassicaceae, Caryophyllaceae, Fabaceae, Gentianaceae, Malvaceae, Poaceae, and Valerianaceae become more conspicuous, while the Bromeliaceae, Melastomataceae, Orchidaceae, Rubiaceae, and Solanaceae decline (see also table 9 in Jørgensen & Ulloa U., 1994).

Table V. Ten largest families and genera of páramo ferns and fern allies.

Family (no. gen./spp.)

Genus (and family) (no. spp.)

Dryopteridaceae (5/77) Elaphoglossum (Dryopteridaceae) (65)
Lycopodiaceae (3/69) Huperzia s.s. (Lycopodiaceae) (60)
Polypodiaceae (14/60)   (Lycopodium s.l. =69)
Pteridaceae (7/43) Hymenophyllum (Hymenophyllaceae) (20)
Hymenophyllaceae (1/20) Isoëtes (Isoetaceae) (18)
Isoetaceae (1/18) Jamesonia (Pteridaceae) (18)
Thelypteridaceae (1/13) Eriosorus (Pteridaceae) (14)
Aspleniaceae (2/11) Thelypteris (Thelypteridaceae) (13)
Blechnaceae (1/9) Polypodium (Polypodiaceae) (12)
Cyatheaceae (1/6) Melpomene (Polypodiaceae) (11)
  Asplenium (Aspleniaceae) (10)

Table VI. Fifteen largest families and genera of páramo flowering plants.

Family (no. gen./spp.)

Genus (and family) (no. spp.)

Asteraceae (101/858) Pentacalia (Asteraceae) (89)
Poaceae (41/227) Senecio s.s. (Asteraceae) (69)
Orchidaceae (25/152)   (Senecio s.l. =172)
Scrophulariaceae (14/144) Diplostephium (Asteraceae) (70)
Melastomataceae (9/107) Calceolaria (Scrophulariaceae) (65)
Gentianaceae (4/93) Espeletia s.s. (Asteraceae) (61)
Ericaceae (16/79)   (Espeletia s.l. =123)
Bromeliaceae (6/78) Lupinus (Fabaceae) (56)*
Rosaceae (10/77) Valeriana (Valerianaceae) (54)
Fabaceaea (9/76) Hypericum (Clusiaceae) (54)
Brassicaceae (13/71) Miconia (Melastomataceae) (54)
Cyperaceae (8/70) Gentianella (Gentianaceae) (48)
Apiaceae (15/61) Puya (Bromeliaceae) (48)
Solanaceae (8/58) Gynoxys (Asteraceae) (46)
Clusiaceae (2/56) Baccharis (Asteraceae) (45)
  Draba (Brassicaceae) (45)
  Geranium (Geraniaceae) (43)

Solanum (Solanaceae) (43)

*Rupert Barneby, who studied Lupinus, felt there are only about 15 species in the genus.  If this is true the overall family number drops to ca. 35 species.

Table VII. Vascular plant genera endemic (or nearly so) to the páramo.
Those marked with an asterisk (*) are monotypic.



Apiaceae   Laestadia
  Cotopaxia   Paramiflos*
  Perissicoelum   Raouliopsis
Asteraceae   Westoniella
  Aphanactus Campanulaceae
  Ascidiogyne   Lysipomia
  Blakiella* Melastomataceae
  Chrysactinium   Castratella
  Coespeletia Scrophulariaceae
  Espeletia   Aragoa
  Floscaldasia* FERNS
  Flosmutisia* Pteridaceae
  Freya*   Jamesonia
  Hinterhubera   Nephopteris*

      General Phytogeography and Origins.  Frequently, the overall floristic and phytogeographical relationships of the páramo flora have been discussed and compared with the Mexican and Guatemalan alpine floras (Beaman, 1965;  González, 1986;  Islebe & Cleef, 1995;  Rzedowski, 1978), the tepuis of the Guayana Highlands (Cleef et al., 1993;  Riina, 1996), the lowland savannas of South America in general (Cleef et al., 1993), the puna flora (Baumann, 1988;  Quintanilla P., 1983b), the subantarctic flora (Cleef, 1978, 1980a), and the world's other high elevation tropical floras such as in Africa and Malesia (Smith & Cleef, 1988).  In common with all other high elevation tropical floras, the páramo flora is predominantly of temperate zone origin at the generic level;  but in contrast, the páramo flora is the richest overall and has the largest actual number of genera and endemic elements (Smith & Cleef, 1988).  Future research efforts are needed to see if the previous comparisons, based mostly on a restricted data-set from the páramos of the Colombian Cordillera Oriental, are representative for the páramo flora as a whole.
      In an effort to understand the origins of the páramo flora, recent phytogeographical studies have followed Cleef (1979a) in assigning each genus to one of seven geographical floristic elements:  páramo, other neotropical, widespread tropical, Holarctic, Austral-Antarctic, widespread temperate, and cosmopolitan (see Table VIII).  In the Colombian Cordillera Oriental, the most comprehensively studied páramo region to date, Cleef (1979a, 1980c) and Van der Hammen and Cleef (1986) found that the páramo genera of vascular plants are about 50% tropical in origin and 50% temperate.  Within the tropical element, strictly neotropical genera are the most strongly represented.  Within the temperate element, the widespread genera are most strongly represented (about 20% of the genera), while the Austral-Antarctic and Holarctic elements are each represented by about 10%.  See also Sturm and Rangel Ch. (1985), who studied the phytogeography of the 130 most important species of the Colombian páramo flora as a whole.
      In the páramo flora of the Cordillera de Talamanca (Costa Rica), Cleef and Chaverri P. (1992) found that of the páramo genera of vascular plants about 36% are tropical in origin and 64% temperate.  Within the tropical and temperate elements of Costa Rica, strictly neotropical genera and widespread genera, respectively, are the most strongly represented, as they are in the Colombian Cordillera Oriental.  Although the Costa Rican páramo flora shares about 95% of its vascular genera with the Andes, the larger proportion of the temperate component in Costa Rica, with genera such as Garrya (Garryaceae), Arctostaphylos (Ericaceae), Mahonia (Berberidacee), and Romanschulzia (Brassicaceae), was attributed to the more northern geographical position of the country.  [Note:  The genera Helianthemum (Cistaceae) and Smilacina (Liliaceae) have also been reported from the páramos of Costa Rica by Cleef and Chaverri P. (1992), but I have not seen any herbarium specimens unequivocally from páramo and, therefore, have not included them in this book.]
      Ricardi S. et al. (1997) studied the superpáramo (above 4000 m) in Venezuela and compared it with other high elevation floras along the entire length of the Andes from Costa Rica to Patagonia.  They found that 34% of the genera of vascular plants are tropical in origin and 66% temperate, with the greatest similarity to that of Colombia and next to Ecuador.
 In northern Ecuador, León Yánez (1993) found that of the genera of vascular plants 45% are of tropical origin and 55% of temperate.  In the southern páramos of Ecuador, Ramsay (1992) found fewer families and genera than in Colombia, with about 33% of the genera of vascular plants of tropical origin and 67% temperate.  He attributed this to the lower humidity and more extreme cold temperatures of Ecuadorean páramos.  Ramsay (1992) also compared the Ecuadorean páramo flora to the puna of Peru and the mountains of East Africa and New Guinea (see also Balslev, 1988;  Mena V., 1984 and Mena V. & Balslev, 1986).  There are, unfortunately, no similar phytogeographical studies from the páramos (jalca) of northern Peru.
      Monasterio (1980b) looked at the entire páramo region and found the following trend:  the jalca and Ecuadorean páramos consist mainly of tussock grasses with genera from extra-tropical regions, the Venezuelan páramos are dominated by rosettes, and the páramos of Colombia show equal importance of grasses and rosettes.
      For additional studies that include phytogeographical analyses, see Ahti (1992, for the lichen family Cladoniaceae), Becker (1988), Cabrera (1957), Cleef (1981b), Duque N. (1987), Jørgensen and Ulloa U. (1994), Keating (1995), Lozano C. and Rangel Ch. (1989), Rangel Ch. (1991b, 1995a), Rangel Ch. and Garzón (1995, 1997), Sipman (1992, for Colombian lichens), Tirado M. and Ricardi S. (1997), Vargas Ulate & Sánchez G. (Unpublished).

Table VIII.   Percentage of vascular plant genera in each phytogeographical element in the Neotropical páramos.

Geographical Element 
Percentage of Vascular Plant Genera
  CRa VEb CO-1c CO-2d EC-1e EC-2f

Páramo 4 7 7 8 4 9
Other Neotropical 25 24 34 30 32 21
Wide Tropical 7 3 10 28 10 3
Holarctic 15 13 11 12 10 14
Austral-Antarctic 14 7 9 5 10 10
Wide Temperate 24 39 20 7 26 26
Cosmopolitan 11 7 8 10 9 17

aCordillera de Talamanca, Costa Rica (Cleef & Chaverri P., 1992); bbased on data from the superpáramo of Páramo de Piedras Blancas, Venezuela (Ricardi S. et al., 1997); cbased on ca. 600 relevés from all parts of the páramo throughout the Cordillera Oriental, Colombia (Cleef, 1979a); dbased on the 130 most important species of the Colombian páramo flora as a whole (Sturm & Rangel Ch., 1985); ePáramo de Guamaní, Ecuador (León Y., 1993); fbased on 192 quadrats in zonal vegetation from 12 scattered páramos in Ecuador (Ramsay, 1992).


Spectacled bear
Spectacled bear
Andean condor
Andean condor

      Páramo grass- and shrublands support a number of native animal species.  It is not the purpose of this book to discuss the general fauna of the páramos, but some of the larger, more common, and conspicuous animals are mentioned here.
      Mammals in the páramo include the puma (Felis concolor), spectacled bear or "oso de anteojos" (Tremarctos ornatus), white-tailed deer (Odocoileus virginianus), mountain tapir or "danta" (Tapirus pinchaque), rabbit (Sylvilagus brasiliensis), guinea pig (Cavia porcellus), Andean fox (Duscicyon culpaes), mountain coatí (Nasuella olivaceae), long-tailed weasel (Mustela frenata), shrew (Cryptotis spp.), rat opossum (Caenolestes spp.), and various small rodents.  Rabbit feces are extremely common in the grass páramo, especially in open areas between tussocks.
      Stotz et al. (1996) list 69 species of birds as "total users" of the páramo habitat, with 41 species making it their primary habitat and 16 as indicator species.  Some of the better known páramo birds include:  vultures such as the Andean condor (Vultur gryphus) and the turkey buzzard (Cathartes aura), eagle, hawk, and falcon (Phalcoboenus carunculatus, Buteo spp., Geranoaetus spp., Falco sparverius), owl (Bubo virginianus and Asio spp.), hummingbird  or "colibrí" (family Trochilidae: e.g., Oxypogon guerinii and Patagona gigas), duck (Anas spp.), and the rufous-fronted parakeet (Bolborhynchus ferrugineifrons).  See F. Vuilleumier (1986) for a discussion of the origins of the high Andean bird fauna.
      Native páramo fish species are few, but rainbow trout or "trucha" (Oncorhynchus mykiss) have been introduced to many lakes and streams for food and sport.
      The herpetofauna has been relatively well studied (see especially Duellman, 1979b) and includes amphibians such as the salamander (Bolitoglossa spp.), frog and toad (Eleutherodactylus spp., Hyla spp., Atelopus spp.), and reptiles such as the lizard and cameleon (Stenocercus spp., Phenacosaurus spp., Proctoporus spp.).
      Invertebrates are less conspicuous and found mostly in the subpáramo.  These include lepidopterans (Descimon, 1986) and other insects such as grasshoppers, cockroaches, beetles, and flies;  molluscs such as a few snails and slugs;  and earthworms.  Arthropods and other microfauna (e.g., mites and springtails) are rarely seen, but are abundant (Sturm, 1978, 1983, 1990, 1994a; van Velez, 1992). 
      A number of the larger páramo mammals, as well as the Andean condor, have been extensively hunted by man and their numbers are now very low. Hummingbirds, bees, and flies seem to be important pollinators in the páramo (see, for example, Berry & Calvo, 1994; Brand Prada, 1994b). Birds, rabbits, and guinea pigs (as well as wind, water, and gravitation) are important agents of dispersal (Frantzen & Bouman, 1989;  Graf Bock, 1984; Simpson & Todzia, 1990).
      Additional references about the páramo fauna include Aagaard (1982), Adams (1973), Amat García (1987, 1991a,b), Arias Lemos (1989), Aristide U. (1969), Barnett and Gordon (1985), Barrientos and Monge-Nájera (1955), Bernal C. (1985), Bernal C. & Figueroa (1980), Brand Prada (1994a), Cadena G. and Malagón (1994), Chapman (1917, 1926), Davis et al. (1997), Del Llano (1990), Duellman (1979a, 1988), Fjeldså and Krabbe (1990), Garay (1981), E. Gaviria (1991), S. Gaviria (1989), Grabandt (1983), Hoffstetter (1986), Hoyos (1991), INDERENA (1984), Janzen et al. (1976), Lynch (1986), Mani (1962), Mora O. and Sturm (1994a), Phelps and Phelps (1958, 1963), Rangel Ch. and Bernal (1980), Reig (1986), Righi (1995), Rivero (1979), Sturm (1984a), Sturm and Rangel Ch. (1978), F. Vuilleumier (1980), Vuilleumier & Ewert, (1978), Wolf and Gill (1986), and Yerena (1994).

Human Influence

Polylepis forest
Polylepis forest

      There have been settlements in the highlands of the Andes for perhaps 15,000 years or more (Eckholm, 1975;  Little, 1981).  Man's influence there has been profound, with the result that nearly 90-95% of the forests of the northern Andes have been cleared (Henderson et al., 1991).  Exploitation of the high elevation puna ecosystem of the central and southern Andes of Peru, Bolivia, and northern Argentina occurred in the pre-hispanic period and has been relatively well documented (see for example, Baker & Little, 1976;  Brush, 1976, 1982;  Ellenberg, 1979).  Intensive land use in the páramos of the northern Andes has been a more recent phenomenon and needs further study (Hess, 1990).  Did similar changes of habitat destruction and alteration take place in the essentially unpopulated high elevation páramo areas as it did in the puna or in the lower montane regions of the Andes?  One of the most frequently discussed questions about the historical development of highland regions of the northern Andes is whether the páramo that we see today, specifically the grass páramo, is a natural ecosystem or to what extent did man create it following his need to provide more pasture and agricultural land?
      Ellenberg (1979) championed the theory that the climax vegetation of the tropical Andes was forest and that man, largely by means of fire, has been responsible for nearly all the large open areas (Becker, 1988;  Fjeldså, 1992;  Lægaard, 1986, 1992).  Other authors feel that both Polylepis forest and some form of páramo grassland have always existed as independent formations, and have expanded and contracted through time in response to climatic changes (Cleef, pers. comm.;  Lauer, 1981;  Troll, 1959;  Van der Hammen & Cleef, 1986;  Walter & Medina, 1969a,b).
      Charcoal fragments in core sediment samples reveal that fires have occurred in páramo since Holocene times (Horn, 1989c;  Horn & Sanford, 1992;  Salomons, 1986).  Those fires may have been ignited by natural sources (i.e., lightning or volcanic activity) or by man (deliberate or accidental), but usually their exact origins are unknown.  Verweij (1995) has stated that fires caused by natural sources occur less than once in 1000 years.  Published records of lightning-set páramo fires have not been found, although Horn (1989c) and Young (pers. comm.) have mentioned lightning strikes in Chirripó páramo (Costa Rica) and the Río Abiseo area (northern Peru), respectively.  In studies of the Costa Rican páramos, Horn (1989c, 1993, In press;  Horn & Sanford, 1992) has correlated modern and prehistoric fires together with pollen analyses of sediment cores.  She has concluded that in the Chirripó highlands páramo communities are the "natural" vegetation and that modern pollen assemblages of individual taxa have not differed greatly from those developed since deglaciation some 10,000 years ago.  She has also found that fires have been a part of Chirripó páramos throughout the Holocene, but have not "carved páramo from forest."
      Without doubt man has had major impact upon the origin and subsequent spread of grasslands throughout the Andes, and perhaps he has been the single most important reason why grass páramo exists today, where shrub/tree woodlands of Polylepis, Buddleja, and Gynoxys may have once dominated.  It is unlikely, however, that we will ever be able to say with confidence what percentage of today's páramo has anthropogenic origins.  Whatever the outcome of this discussion, the facts remain that grass páramo currently exists, covers large expanses of the high elevation Andes, and has great ecological and economical importance.
      The following discussion summarizes what we do know about the history of man's presence in the páramos of the high Andes and the effect he has had on them.  Although many of the references and quotes given below are specific to one country, I am certain that similar events have occurred in all of the north Andean countries including Venezuela, Colombia, Ecuador, and probably by extension to Costa Rica and Panama.  The situation in southern Ecuador and northern Peru is less clear, however, and may be quite different due to its overall drier climate and proximity to the larger pre-Columbian civilizations that occupied high elevations to a much greater extent in the central Andes than in the northern Andes.  For example, the intense grazing pressure may have occurred earlier in Peru because of the presence of native camelids and their herding.

      The Pre-Columbian Period.  It is assumed that the Venezuelan páramos have been exploited for more than 1400 years (López del Pozo, 1992).  They have been occupied for at least 500 years, although there were no permanent settlements until colonial times (Wagner, 1978, 1979, 1988).  The páramos were used, however, as corridors, in rituals, or as hunting areas (Clarac, 1981, 1985;  Rojas, 1985).  According to Salgado-Labouriau (1976) and Monasterio (1980d), pre-Columbian agriculture was present in Venezuela up to 3000 m, the lower edge of the páramo, but it was concentrated below 2000 m and the cultures that developed in or near páramo were dedicated to agriculture and hunting/gathering (Molinillo & Monasterio, 1997).  In Colombia, the pollen record shows that human activity in the páramos around Bogotá began about 800 years ago (Van der Hammen, 1968).  The Colombian páramos were mostly uninhabited, however, although they too were used for the cultivation of some tuber crops and as a crop-storage area (Langebaeck, 1988;  López del Pozo, 1992).  The highest mountains and páramos were mostly considered sacred, often with mystical qualities, and their main use was for religious purposes and to bury the prominent dead (G. Correal U., cited in Cleef, 1981b;  Reichel-Dolmatoff, 1982;  Zambrano, 1993).

      The Colonial and Independence Periods.  Europeans settled the New World in the 16th century.  They introduced exotic animals that were totally new to the northern Andes, viz., cows, sheep, goats, horses, and donkeys, which added to the herds of native camelids (llama, alpaca, and vicuña) in Peru and further south.  [Llamas and alpacas have only recently been introduced into the Ecuadorean páramo (Hervas Ordoñez, 1994).]  Exotic plants were also introduced by Europeans for cultivation at the cooler high elevations, especially the grains wheat and barley, but also vegetable crops such as broad or faba beans, lentils, peas, carrots, radishes, onions, and garlic. These new plants added to the native grain quinoa and tuber crops such as potato.
      Early colonists (including pre-Columbian man) founded their villages and towns primarily in the cooler inter-andean valleys along rivers, where their crops were grown in the fertile floodplains (corn) and along the adjacent upland slopes (potatoes and grains).  Cattle reached the Andes of Venezuela and Colombia in the late 1500s, but probably were not introduced into the páramos until the early 1700s, and then only to ca. 3000 m (Vila, 1962;  Wagner, 1967).  The cattle foraged in the páramo grasslands above the forest, several hours or days walk from their homes.  As herds of cattle grew in numbers, fires were periodically set to burn-off the dead material from the páramo and provide fresh new grass for the cattle.  With time the páramo became overgrazed and excessively burned.  Wood was also cut from nearby forests, mostly for fuel and as building material (including road-building).  Gradually, the need arose to cut and burn the forest immediately below the páramo to provide additional new grasslands for pasturing and arable land for crops.  Sediment cores from Laguna Victoria, Venezuela (3250 m), show that at about this same time large amounts of pollen of Rumex acetosella, an herb introduced with wheat and potatoes, as well as a drastic drop in tree pollen, suggesting a massive deforestation of the upper Andean forest (Salgado-Labouriau & Schubert, 1977).
      The numbers of cattle continued to rise steadily during the 1700s.  In the 1800s, during the wars of independence and civil strife in Venezuela (and possibly throughout Colombia, Ecuador, and Peru), the numbers of cattle oscillated, sometimes decreasing by almost 95%, although they again reached the pre-independence levels by ca. 1847 (Brito, 1972;  Pérez, 1992a, In press;  Rouse, 1977;  Vila, 1962, 1978).
      Thus it would seem that in the early 1700s the páramos of the northern Andes were still not utilized to the extent of the altiplano or puna to the south (Brush, 1982).  However, during the late Colonial period and into the Independence period, agriculture and cattle raising became common (Velázquez, 1986) and the range of páramo-like habitats probably expanded.

      The Modern Period.  During the 20th century, man's impact on páramo has been a story driven by demographics, i.e., the numbers, density, and distribution of populations. As population densities in the interandean valleys increased, people began to colonize new areas in order to satisfy their needs (Bernsen, 1991;  Brush, 1976, 1982;  Ellenberg, 1979;  Hess, 1990;  Hofstede, 1995c;  López del Pozo, 1992;  Molinillo & Monasterio, 1997;  Monasterio, 1980d;  Parsons, 1982;  Pérez, 1992a).  The annual population growth rates, previously over 3% in Venezuela, Colombia, Ecuador, and Peru (Little, 1981) and currently 2.6-2.9% (World Population Data Sheet 1997 from the Population Reference Bureau), are still the highest in South America.  The Colombian population, for example, increased from 3 million to 12 million between 1900 and 1950, rose to 26 million in 1975, to 33 million in 1990, and is estimated that it will reach 49 million in 2025;  the Venezuelan population increased from 5 million to 20 million between 1950 and 1990, rose to 22 million in 1995, and is estimated that it will reach 35 million in 2025;  while the Ecuadorean population increased from 3 million to 11 million between 1950 and 1990, rose to 12 million in 1995, and is estimated that it will reach 18 million in 2025 (World Resources 1996-97). 
      Because of land and agrarian reform policies during the 1960s and 1970s, most parts of Colombia, Venezuela, Ecuador, and northern Peru have experienced two recent tendencies:  the upward movement of agriculture into the páramo belt between 3000 m and 4000 m, and the intensification of animal production in the lower páramo belts (Baruch, 1979;  Becker, 1988;  Bernsen, 1991;  Ferwerda, 1987;  Hess, 1990;  López-Zent, 1993;  Monasterio, 1980d;  Pérez, 1992b;  Verweij, 1995;  Verweij & Beekman, 1995).  In Venezuela, Monasterio (1980d) reported that since the 1960s, the highest limit of cultivation (in Páramo Piedras Blancas, ca. 4200 m) has increased due to the introduction of frost-resistent varieties of potato, such as the "papa negra."
      The last 20-25 years have also seen, however, a regional growth in urbanization, with farmers abandoning the countryside to cities such as Mérida, Bogotá, Nariño, Quito, Cajamarca, etc.  This has resulted in greater demand and increased pressure upon fewer farmers throughout the Andes to cultivate new land to produce more food.  As natural soil fertility decreased in the páramo, usage of chemical fertilizers, insecticides, pesticides, and fungicides increased (Bernsen, 1991;  Ferwerda, 1987;  López-Zent, 1993), followed by soil contamination.  For example, in the crop belt of the Venezuelan páramos (3000-4000 m), the traditional land-use system of short periods of crop production (1-4 years) alternating with fallows of 7-20 years, which allows long-term recuperation of nutrients with negligible erosion rates and maintenance of high levels of soil organic matter, has been interrupted by massive use of chemical fertilizers and a market-oriented agricultural system.  Some of the consequences have been "greater incidence of pest outbreaks, soil exhaustion, a possible decrease in soil organic matter, and intensification of erosion" (L. Sarmiento M. et al., 1990, 1993).  According to Ferwerda (1987), more than 70 years are needed for regeneration of páramo bunch-grasslands after potato cultivation.  Along with the upward movement of agriculture and increased livestock production have come the continued lowering of the forest line due to wood cutting to make more room for crop production and pastureland (Kok et al., 1995).  The timber is used (locally and regionally) for building, fence-posts, fuel, etc.  Fuelwood collection, mainly for cooking purposes, also removes shrubby vegetation from the páramo proper (Becker, 1988).  Verweij (1995) estimated that an average family of seven persons uses about 20 kg of fuelwood every day, which corresponds to a cleared area of one hectare per year.
      Other recent developments that have led to additional páramo disturbance include construction of aqueducts, drainage systems, and roads (Young, 1994), off-road vehicles become more common (Pérez, 1991b), mining (especially of sand, gravel, and peat), and afforestation projects (Rangel Ch., 1989;  pers. observ.).  The impact of plantations of exotic tree species (especially Pinus radiata and P. patula) on the development of soils and vegetation placed in or near (extensively grazed) páramo has recently been described by Hofstede (1997). Even tourists are now becoming a problem in some areas (Horn, 1986;  Urgilés Sánchez, 1990).

Impact of Burning and Grazing

Lachemilla orbiculata
Lachemilla orbiculata

      Fire is used in the páramo for primarily two reasons.  First, it is a means to clear the upper limits of forest, thus providing open grasslands for grazing of livestock and new land for cultivation of crops.  Second, because forage quality is poor and up to 80% of the above ground biomass of the tussock- or bunchgrass growth form is dead material (Cardozo & Schnetter, 1976), farmers burn-off the old or dead non-palatable parts to provide new, more nutrient-rich growth for cattle.  Ramsay and Oxley (1997) say that it is quite difficult to estimate the ultimate effects of fire upon páramo vegetation and its species diversity, since fire has been so much a part of páramo history.  Nevertheless, burning and grazing go hand-in-hand in most high elevation areas of the northern Andes;  seemingly they are the only applied "management practice."  However, is fire necessarily bad for páramo in view of the fact that it has been part of páramo history since at least Holocene times?  And what are the actual effects of fire and grazing upon páramo?  How do the continuous and repetitive practices of burning and grazing combine to affect species composition, nutrient and soil-water content, biomass and productivity, decomposition, and short-term and long-term stability of the páramo ecosystem?  The following discussion outlines what is currently known about the role and impact of fire on the páramo ecosystem.
      Four recent studies about the combined effects of burning and grazing on the páramo ecosystem contribute new insight and provide valuable basic information that is necessary for future management programs (Hofstede, 1995c;  Keating, 1995;  Ramsay, 1992;  Verweij, 1995).  In the following section, I have drawn heavily from the excellent dissertation of Robert Hofstede (1995c), who studied the ecological impacts of burning and grazing on biomass, nutrient status of vegetation and soil, and hydrology, in the Los Nevados National Park (Cordillera Central, Colombia).  In a complimentary study, Pita Verweij (1995) focused on the impacts of management on vegetation development and landscape ecology in the same national park.  In addition, Paul Ramsay (1992) and Philip Keating (1995) studied páramo vegetation in Ecuador, including community ecology, dynamics, and productivity.  Some of their main conclusions are summarized in the following paragraphs.
      Hofstede (1995c), Lægaard (1992), Ramsay (1992), and Verweij (1995) all conclude that natural páramo vegetation is to a certain extent tolerant to the managament practices of burning and grazing, and that fire does offer some short-term advantages to the cattle-farmer (over-all palatability, initial productivity).  It is not, however, a wise long-term management tool, because when the burning frequency is more than once per decade (Verweij, 1995) and grazing intensity increase, stress on the vegetation is too great, effects are mostly negative, and there is degradation of the ecosystem, including a loss in species diversity.  Unfortunately, it has been shown that farmers will choose the short-term perspective, i.e., rapid post-fire regeneration of palatable forage for maximum livestock production at the expense of long-term degradation of the vegetation and soil (Hofstede, 1995;  Ramsay, 1992;  Verweij, 1995).
      Burning and grazing affect the páramo ecosystem most importantly in the following three ways:

      Vegetation Structure, Composition, and Species Diversity.  Despite their many adaptations, rosette plants (especially members of the Espeletiinae) and tussocks or bunchgrasses are some of the most susceptible species to burning and grazing in the páramo.  In Venezuela, A. P. Smith (1981) found that 55% of the adult Espeletia schultzii (Asteraceae: Espeletiinae) plants could be killed by fire in a dry season, while in Colombia, Verweij and Kok (1992, 1995) found over 80% mortality in juveniles of E. hartwegiana immediately after fire.  Verweij and Kok (1992) also found that grazing had a major impact on adult mortality of E. hartwegiana in Colombia, and Pérez (1992a) found that cattle browsing could cause a decrease of 37% in rosette cover of Coespeletia timotensis (Asteraceae: Espeletiinae) in Venezuela.  In southern Ecuador, Keating (1995) found that burning and cutting reduced the number of rare species and the relative frequency of shrub species.  In the Buenavista subpáramos of Costa Rica, Horn (1997, In press) found that 90-99% of the dominant bamboo, Chusquea subtessellata and the ericaceous shrub Vaccinium consanguineum resprouted following fire, while the second most common woody species, the shrub Hypericum irazuense (Clusiaceae), suffered 94% mortality.  This display of fire sensitivity in woody species certainly affects species composition and may be the reason for the present dominance of the bamboo.  For tussocks and bunchgrasses, burning and subsequent grazing also lead to fragmentation because of trampling by livestock after the fire.  Continued grazing or overgrazing leads to replacement of the tussocks and bunchgrasses by short-grass communities (Agrostis, Festuca, Paspalum), rosaceous species (Acaena and Lachemilla orbiculata), and species of low or matted herbs such as Eryngium humile (Apiaceae), Bidens and Hypochaeris (Asteraceae). These plants are better adapted to cattle grazing and to more exposure of bare ground.  Alien species, which are frequently found in these communities, are indicators of human and grazing (with some fire) disturbance;  these include (among others) Achillea millefolium, Gamochaeta americanum (formerly Gnaphalium americanum), Stevia lucida, and Taraxacum officinale (Asteraceae), Arenaria serpens, Cerastium arvense and Stellaria media (Caryophyllaceae), Trifolium repens and Lupinus spp. (Fabaceae), Geranium multipartitum (Geraniaceae), Juncus effusus (Juncaceae), Oxalis spp. (Oxalidaceae), Anthoxanthum odoratum, Dactylis glomerata, Digitalis purpurea, Poa annua and Holcus lanatus (Poaceae), Rumex acetosella (Polygonaceae), and the fern Pteridium aquilinum (Dennstaedtiaceae) (Cleef, 1981b;  Cleef & Rangel Ch., 1984;  Ewel et al., 1976;  Hofstede, 1997;  Hofstede, Chilito P. & Sandoval S., 1995;  Hofstede et al., 1995;  Keating, 1995;  Ramsay, 1992;  Rangel Ch., 1989;  Salamanca, 1991;  A. P. Smith, 1974a;  Verweij & Budde, 1992;  Verweij & Kok, 1995).
      Based on studies of small-scale dynamics within several páramos in Ecuador, Ramsay (1992) generalized that:  a) it is mainly vegetation structure that determines fire temperature: the highest temperatures (up to 600°C) are produced in the tussock canopy while the lowest (ca. 65°C) are found within the tussock base and just beneath the soil surface;  b) the main form of recovery from fire is regeneration by vegetative reproduction of below ground plant parts;  c) the degree of survival of intertussock species depends upon the severity of the fire, which is largely a function of the time interval since the last burning;  d) survival following fire does not guarantee a plant's persistence in the community, since both tussock grasses and intertussock species show significant mortality rates in the months following fire;  e) recovery at higher altitudes is slower;  f) burning may cause the patterns of community development to become cyclical;  g) some species that quickly colonize bare ground by seed may persist, but as the vegetation matures most are killed by competition;  and h) establishment within tussocks after burning is the preferred method of some species, such as Lupinus cf. pubescens.
      Minimum fire recovery periods, i.e., the period of post-fire growth required to generate enough fuel to carry a subsequent fire (Verweij, 1995), of 3-10 years have been reported depending on the vegetation structure (Horn, 1990, 1991;  Janzen, 1973;  Ramsay, 1992;  Ramsay & Oxley, 1996;  van Groen, 1987;  Verweij, 1995).  However, the frequency of burning is typically every 2-4 years (Ramsay & Oxley, 1996), and in practice, fire is applied too frequently to allow full bunchgrass recovery (Verweij, 1995).  Minimum fire recovery periods do not mean that the vegetation will have recovered completely, however, because species composition and distribution will change and more bare spots will be present (Hofstede, 1995c;  Horn, 1989;  Janzen, 1973;  Miller & Silander, 1991;  Verweij, 1995;  Williamson et al., 1986).  Furthermore, since grasses have high relative growth rates, and burning and grazing stimulate tillering, they are favored over woody species by high fire frequencies (Horn, 1991;  Janzen, 1973;  Ramsay, 1992;  Ramsay & Oxley, 1996;  Verweij, 1995;  Williamson et al., 1986).  Verweij (1995) concluded that more than 10 years are needed before the average number of vascular plant taxa stabilizes back to pre-fire levels, implying also that fire frequency may influence species richness in the vegetation.
      In summary, it would appear that occasional disturbances over short intervals may allow for relatively rapid recovery of community structure and composition, but that repeated burning, combined burning and grazing, or severe disturbances at intervals of even 2-4 years is too great to allow complete recovery of páramo vegetation and some species are lost.

      Soil Structure and Water-storage Capacity.  The páramo belt is an indispensable source of water for drinking and irrigation for a majority of the people in Venezuela, Colombia, and Ecuador.  In its natural, undisturbed condition, páramo acts as a sponge, where the excess water is slowly but constantly returned to the ecosystems of lower elevations (Gómez Molina & Little, 1981).  Natural páramo shrub vegetation holds up to 12 times more water than a disturbed grass community (Schnetter et al., 1976), while water retention in soils of the páramo surpasses 200% of its own dry weight (Cañadas Cruz, 1983).  Hofstede and Sevink (1995) showed that undisturbed páramo has a larger water-storage capacity than burned and grazed páramo, and that a continuously wet plant layer is important for maintaining the large soil storage for dryer periods.  With grazing and burning, however, vegetation cover disappears and is replaced by a low growing, ground cover that cannot hold as much moisture since the páramo soils are more compressed, dryer, less acidic, and somewhat less organic (Hofstede, 1995c).  Verweij and Budde (1992) and Verweij (1995) also found that during the first 1.5-5 years following fire the percentage of bare soil increased over 10%, depending on the vegetation structure and grazing intensity.  Meanwhile vegetation is trampled and compacted, and the turf is broken-up by cattle (Grubb, 1970;  Pérez, 1993a), thus favoring erosion, loss of soil, and micro-terracing, especially on slopes (Flores Ochoa, 1979;  Pérez, 1992b, In press;  Troll, 1973).

      Decomposition, Nutrients and Productivity.  Hofstede (1995) and Verweij (1995) have shown that grazing and especially burning practices result in faster decomposition and mineralization due to dryer soils and higher maximum soil temperatures, but this does not necessarily lead to higher nutrient contents of soils and plant tissues.  On a short-term basis, mineralized nutrients may become available for immediate plant regrowth, but generally they are very quickly unavailable in the soil due to immobilization especially of phosphorus, volatilization especially of nitrogen and sulphur, or less importantly to erosion or leaching.  Therefore, the vegetation remains nutrient poor.  The change in vegetation structure and composition eventually leads to less tussocks, which may have more tillers per unit area and a higher live material content, but the overall lower stature and leaf elongation rates yield less productivity (Hofstede, 1995).  This low productivity combined with nutrient poor vegetation results in poor quality cattle (Hofstede, 1995;  Pels & Verweij, 1992;  Schmidt & Verweij, 1992;  Verweij, 1995).

      In summary, the general impact of increased burning and grazing was summarized by Verweij (1995) as follows:  "the structure of the bunch grasslands becomes more open.  Fire events cause temporal changes in vegetation structure over the relatively short time span of a few years.  Without the influence of grazing, the vegetation is able to recover and returns to the initial stucture of dense bunch grassland, the floristic composition remaining largely unchanged.  Under prolonged grazing pressure, however, short matted grasses and forbs increase and a gradual transformation into other communities occurs.  Changes in structure are of vital importance in relation to the water holding capacity of both vegetation and soil.  Hofstede & Sevink (1995) concluded that water retention capacity is reduced in grazed and burned situations, with an increased risk of drying especially in the drier seasons."

The Future of the Páramo Ecosystem

      The páramo is rapidly becoming a threatened ecosystem.  With population size at its greatest in history and steadily rising, pressure for greater food production and land use has intensified and deterioration of the environment has inevitably followed.  Activities such as deforestation, road building, burning, agriculture, and herding of animals have significantly altered this fragile ecosystem.  They have also caused native páramo plants and timberline forests to disappear at an alarming rate, thereby allowing weedy species to dominate some landscapes.  Such activities also lead to accelerated erosion on the slopes and flooding as the soils are not able to retain moisture within their drainage basins.  Furthermore, with the public's increased desire for leisure time and a gradual awakening to the potential of the páramo for recreation and tourism, we need to increase quickly our efforts to study this ecosystem, not only to know what is there by means of inventory and ecological analyses, but also to monitor the effects of disturbance upon the vegetation and aquatic resources so that rational management programs can be established to protect and preserve the páramo.
      Why is the páramo important and what needs to be done to maintain this ecosystem?  Can we still turn the tide of alteration and destruction?  Must it be totally preserved and protected?  As has been discussed above, páramos are important for scientific, ecological, and economic reasons -- all of which are inter-connected and dependent upon each other.  Some of these reasons are summarized below and opinions are offered about what should be done.

      Scientific Reasons.  Páramos occupy no more than 2% of the land area of the northern Andes in Venezuela, Colombia, and Ecuador, and yet they are highly diverse biologically and contain many plants that grow nowhere else in the world.  Their geographical isolation along the mountainous chain from Costa Rica to northern Peru, along with their high diversity and endemism, combine to make them biogeographically unique.  Paleoecological studies indicate that páramo habitats have undergone repeated expansion and contraction during the late Pliocene and Pleistocene.  These high elevation, cold-climate ecological islands of endemism are surrounded by a sea of low elevation, warm tropical forest.  Therefore, the insular dispersal, age, origin, affinities, and evolutionary patterns of the plant taxa, are open to detailed systematic and evolutionary biogeographic studies.
      Man's activities, however, are severely limiting biodiversity in the páramo ecosystem.  Basic facts about the taxonomy, relationships, distribution, and ecological preferences of most high elevation groups is very incomplete.  We must know the names of the taxa we are dealing with before we are able to compare them or effectively communicate data about their ecology, evolution, etc. (Vuilleumier & Monasterio, 1986).  Therefore, it is appropriate and timely that páramo plants be studied and a complete flora with keys, species descriptions, and indications of geographical distribution be written.

      Ecological Reasons.   Perhaps the single most important ecological function of páramo is as a regulator of Andean hydrology.  However, deterioration of the páramo due to loss of vegetation cover, soil compaction, and weakened water storage capacity, may increase soil loss, erosion, and flooding and contribute to droughts during dry periods.  Therefore, ecological studies must also begin at the basic level, including additional short- and long-term studies of the dynamics of páramo vegetation as it affects landscape heterogeneity and plant species diversity, keeping in mind the use of better management techniques that could lead to stabilization of water supplies, increased productivity, and a subsequent lesser need to cut more forest and exploit new areas of land (Ramsay, 1992;  Verweij, 1995).  Also needed is a better understanding of the dynamics of fire, repeated fires, and recovery after fire in order to properly assess its role as a tool in the management and conservation of páramo ecosystems.
      It has also been noted that the páramo is a very sensitive ecosystem and could provide an excellent area for additional environmental studies of global and regional climatic changes, such as the effects of global climate change (the greenhouse effect) and drought or increased rain (brought on by El Niño) on plant species distributions and adaptations.

      Economic Reasons.   The major economic use of the páramo today is for agriculture and livestock.  But, as has been discussed above, the management practices of burning, grazing, and increased use of chemicals emphasize only short- not long-term usage.  If the páramo ecosystem is going to remain an important ecological and economic center, it must continue to be the natural source of clean water, as well as a resource for agriculture and grazing, new foods, ornamentals and medicines, and for recreation and tourism.
      Páramo plants are contributors of important genetic factors to high elevation or cool-habitat crops such as the potato (Solanum tuberosum, Solanaceae) and the grain quinoa (Chenopodium quinoa, Chenopodiaceae).  Páramos are also a source of potential new foods ("new" to non-Andean peoples) derived from pre-Columbian tuber crops such as "oca" or "íbias" (Oxalis tuberosa, Oxalidaceae), "ulluco" or "olluco" (Ullucus tuberosus, Basellaceae), and "añu", "mashua" or "cúbios" (Tropaeolum tuberosum, Tropaeolaceae) (King, 1988;  National Research Council, 1989;  Schjellerup, 1989).  New ornamental plants such as Senecio niveo-aureus (Asteraceae), and new medicines such as those derived from Oritrophium spp. and Chuquiraga jussieui (Asteraceae), and Thamnolia vermicularis (a lichen of uncertain affinity) have as yet unrealized potential (Aranguren B. & Márquez, 1995;  Aranguren B. et al. 1996;  López-Zent, 1993;  Paredes, 1962;  Pérez, In press).  Historically, the páramo has been a source of traditional medicines for local communities, and even today 56% (43 out of 77) of the medicinal plants used by the Cogui, an indigenous people living in the Sierra Nevada de Santa Marta (Colombia), come from the páramo (Carbonó de la Hoz, 1987).
      Although there are limitations, the potential for ecotourism or for recreation in páramo regions, such as fishing, hiking, and camping, has gone almost unnoticed (Black, 1983, 1988;  Ramsay, 1992;  Rangel Ch., 1989;  Urgilés Sánchez, 1990).  The so-called "wilderness experience" can draw many visitors, but as Hofstede (1995) has reminded us these same visitors are "disappointed when meeting unnatural elements such as cattle, fences and meadows.  Both national and international tourists potentially visiting the páramo are not directly interested in a high species richness, a high water storage or a good soil protection, but their demand is that the landscape is as original as possible:  they simply enjoy touching the snow and seeing an Espeletia.  Nonetheless, the ecological and hydrological sustainability benefits directly from the conservation of an original landscape and vegetation structure...  In the future, it should be evaluated which yields more money:  a productive cow or a majestic Frailejón." 

      What can be done?  Virtually any extended use of the high elevation páramos, be it for science, economics (including both resource production as well as extraction), or conservation, will to some extent have an impact on species composition, vegetation structure, biodiversity, and soil hydrology.  It is only if all human activities in páramos are ceased, that ecological and hydrological sustainability will be assured.  With the current socio-economic situation in the Andes, however, the exclusion of all human activities from the páramo is not realistic and agricultural production has to be accepted and respected (Hofstede, 1995;  Lauer, 1993;  Verweij, 1995).  Given the scientific evidence at hand, it is relatively easy to identify the major problems that have changed the páramo ecosystem, but what are the practical alternatives?  At this time, three areas in particular need to looked at -- management practices, conservation efforts, and environmental education.

      Management.  It seems generally agreed upon that there is not any single management practice that addresses all páramo functions;  therefore, perhaps the challenge to managers is to vary their programs in order to meet the demands of one function in one area and those of another somewhere else (Hofstede, 1995;  Horn, In press;  Pérez, In press;  Ramsay, 1992;  Verweij, 1995;  Young, 1997).  While banning the use of burning is advocated by some (e.g., Hofstede, 1995;  Verweij, 1995), Horn (In press) feels that fire management efforts need to be changed from the prevention of all fires to the control of fire frequency thus recognizing its place in the páramo landscape.  Ramsay (1992) advocated shifting agriculture and grazing to lower elevation zones, since it appeared from the results of his productivity studies that lower altitude pastures were "substantially under-productive."  Rangel Ch. (1989) suggested declaring all páramos as nationally protected areas, primarily to protect the water resources;  restricting use of all páramo bodies of water, especially in the production of energy;  avoiding the exploitation of páramo biological resources and especially avoiding grazing by sheep;  the establishment of research centers to study páramo;  and using particular páramo areas for recreation (with controls).  Numerous smaller, but nonetheless important, changes should also be considered for implementation wherever possible, such as:  prohibition of the cutting of natural vegetation and the collection of fallen stems for fuel, use of fencing to exclude cattle or allow rotation of grazing, and the exclusion of motorcycles and off-road vehicles from the páramo (Pérez, In press).
      If new management strategies are to be considered, there needs to be strong coordination and cooperation between scientists, local people, non-governmental institutions, and government departments in the entire planning, design, and implementation process.  Becker (1988) has stressed that within indigenous communities there are needs for alternative sources of income to alleviate the pressures of habitat destruction for agricultural purposes (see also Verweij et al., In press), better social organization with the promotion of traditional technologies, such as terracing for soil conservation (Lauer, 1993;  Nieto Cabrera et al., 1997;  L. Sarmiento M. et al., 1993), reforestation with useful and native species (Brandbyge, 1984;  Brandbyge & Holm-Nielsen, 1986;  Lauer, 1993;  L. Sarmiento M. et al., 1993), and integrated agroforestry systems and controlled grazing and burning to allow regeneration of native plants (Hofstede, 1995c;  Ramsay & Oxley, 1996;  Verweij, 1995).

      Conservation.  Because of their high biological diversity and their biogeographical, evolutionary, ecological and economic significance, páramos are unique and the species that have adapted to them will not grow elsewhere.  These facts, along with the realization that the páramo, like so many other tropical ecosystems, is being altered by human activities, qualify them as high priority areas for conservation.  "Their preservation in the face of these ever-expanding human activities, however, will depend on their rational use" (Vuilleumier & Monasterio, 1986).  The farmer's aim in burning is to increase livestock productivity and this practice may work in the short-term;  however, the process of recovery in páramo takes a long time and subsequent repeated burning and grazing may cause long-term damage in complete contrast to the original aim of the burning (Ramsay, 1992).  Therefore conservation of the páramo ecosystem, including watersheds and natural vegetation, is imperative.
      Until about 40 years ago there was little direct action taken toward preservation of the páramo ecosystem.  Recently, however, páramos have been included within the boundaries of many national parks, frequently with recognition as regulators of Andean hydrology.  National parks that include important stretches of páramo are Chirripó (Costa Rica), Sierra Nevada de Mérida and Tamá (Venezuela), Sierra Nevada de Santa Marta, Sierra Nevada de Cocuy, Sumapaz, Chingaza, Los Nevados, and Puracé (Colombia), Cotopaxi, Sangay, and Podocarpus (Ecuador), and the Río Abiseo watershed in northern Peru.  More representative páramo areas still need to be set aside, however, as well as more efforts to increase the numbers of rangers to enforce existing boundaries and park regulations.

      Environmental Education.  Scientists have the responsibility to see that factual data from páramo studies be presented to the general public, local authorities, and indigenous communities, because if conservation and/or management goals are to be achieved, public awareness and public support are essential.  People from large cities like Bogotá, Quito, Cuenca, and Mérida flock to the páramos on weekends for the natural beauty, the "wilderness experience" and for recreation, but at the same time they often use branches from shrubs for their campfires and leave litter behind (Urgilés Sánchez, 1990). We must make the general public aware of the value of páramo, which can also serve as an educational model in student field trips and ecological tours.  For example, placing signs along roads through the parks that explain different aspects of the environment are an excellent means of public education.  If, as Horn (In press) advocates, fire is recognized as having a place in the páramo landscape, then park managers will also need to educate the general public as to the role of fire in maintaining diversity of habitats and species.
      The training of local students is also critical for long-term maintenance of diversity and identification of critical areas for protection and management (Balslev & Paz y Miño C., 1991;  Verweij, 1995).  For logistic reasons, local scientists can more easily undertake long-term surveys and their follow-ups.  Furthermore, many young scientists will become key government officials or leaders in conservation groups in their countries, thus improving the quality of decision-making about conservation and management issues.

Web page designed by Myriam Fica.
Contact Carmen Ulloa.

© 1995-2014 Missouri Botanical Garden, All Rights Reserved
P.O. Box 299, St. Louis, MO 63166-0299
(314) 577-5100

Technical Support