Grass páramo as hunter-gatherer landscape

Abstract

Tussock grass páramo constitutes the dominant vegetation of the high tropical Andes and cordilleras of Costa Rica. Its distribution, composition and the location of the upper forest line are ascribed, by broad consensus, to climate. The zonal argument finds support in the altitudinal movements of páramo during the Pleistocene, clearly responding to changes in precipitation and temperature. I ask here, however, if the principal ecological variables driving post-Pleistocene páramos are circumscribed solely by climate. The combined pollen, charcoal and archaeological evidence generated in recent decades suggests a distinct Holocene etiology. Five principal conclusions emerge: (1) the sedimentary charcoal record establishes that grass páramo is a fired landscape, (2) natural sources of fire, specifically volcanoes and lightning, are incapable of generating the fire regime apparent in the sedimentary charcoal record, (3) burning at most sites intensified significantly between 13,000 and 11,000 cal. yr BP, and maintained heightened levels during the Holocene, (4) archaeological findings suggest that the original settlement of the Andes coincided with the sedimentary charcoal rise, and (5) the subsistence logic of hunter-gatherers argues strongly for firing the vegetation to increase resource productivity and reliability. From the pioneering mixed vegetation after deglaciation, anthropic fire selected in favor of tussocks and against woody species, generating a novel plant association. I propose, therefore, that Holocene grass páramo is not zonal vegetation, but rather a hunter-gatherer landscape.

Keywords

grass páramo Holocene fire hunter-gatherer pollen sedimentary tropical Andes

Introduction

Neotropical páramos form an archipelago of elongated herbaceous islands covering high mountain terrain between Costa Rica and northern Peru. In sum they occupy over 35,000 km² and extend from 11°N to 8°S latitude (Hofstede et al., 2003; Luteyn, 1999). Botanists and ecologists, struck by the beauty of these landscapes and by the resilience and diversity of their flora, have studied páramos since at least the times of Alexander von Humboldt, 200 years ago (von Humboldt and Bonpland, 2009 [1807]). Humboldt viewed páramo as a response to elevation, in the manner of alpine vegetation, and elegantly schematized the altitudinal distribution of tropical Andean flora from lowlands to perennial snow. Páramos occupied the highest of these vegetation zones, between the upper montane forest and the snowline. The zonal paradigm employed by Humboldt became the measure by which páramos were understood and scientifically explored, and today remains the consensus view.

Twentieth century landscape-scale ecologists universally viewed grass páramo as natural vegetation (Hartshorn, 1983; Troll, 1968; Walter, 1985; Weberbauer, 1945), and most contemporary páramo researchers concur (Bakker et al., 2008; Cleef, 1981; Hooghiemstra, 1984; Hooghiemstra and Cleef, 1995; Horn, 1993; Horn and League, 2005; Körner, 1998; Kuhry et al., 1993; Moscol Olivera and Cleef, 2009a; Moscol Olivera and Hooghiemstra, 2010; Van der Hammen, 1974). In the zonal paradigm the transition from forest to open vegetation and the altitude of the generalized upper forest limit (UFL) primarily reflect temperature constraints, as do the characteristic forms and physiological adaptations of páramo plants. Localized microclimates above the UFL permit the establishment of patch forests, defining an upper tree line. Precipitation regimes control atmospheric and soil humidity and govern, for a given mean temperature, the characteristic life forms and floristics, and influence the altitudinal limits of páramo communities (Sklenář and Lægaard, 2003). Physiological aridity associated with high transpiration rates, wind and cold provides an explanation for the curious combination of a wet climate and the xeric adaptation of many páramo plants (Leuschner, 2000; Weberbauer, 1945). Secondary to climate, edaphic factors including soil pH, water deficit or excess and nutrient status determine most azonal formations.

In the zonal view three broadly horizontal altitudinal belts are emplaced in ascending elevation on the mountainside (Cleef, 1981; Cuatrecasas, 1968; Hooghiemstra, 1984; Luteyn, 1999): (1) shrub- or sub-páramo, interpreted to be a transitional habitat directly above montane forest; (2) tussock-dominated grass páramo, also called ‘páramo proper’ or ‘true páramo’ to emphasize its zonal centricity (Lauer, 1981); and (3) desert páramo, or superpáramo, a formation poor in long-stem tussock grasses and dominated by isolated cushion plants, sclerophyllous dwarf shrubs and acaulescent rosettes, mosses, lichens and open ground (Monasterio, 1979; Sklenář, 2006; Sklenář and Balslev, 2005). Locally specific plant associations can be identified within each altitudinal belt, but the architecture and growth forms of these associations are similar across the broad expanse of northern Andean páramos.

Pleistocene pollen records from the high plain of Bogotá, Colombia, document the altitudinal displacements of páramo and contiguous forest. These records detail how herbaceous vegetation descended and ascended the mountain profile in consort with the waxing and waning of Pleistocene glacial cycles (Groot et al., 2011; Hooghiemstra, 1984; Hooghiemstra and Cleef, 1995; Hooghiemstra and Van der Hammen, 2004; Marchant et al., 2002; Van der Hammen, 1979; Van der Hammen and Cleef, 1986). Contemporary researchers consider the early-Holocene emplacement to be the latest manifestation of prior cycles, on the assumption that the present reflects Pleistocene interglacial conditions (Hooghiemstra et al., 2006). This assumption, however, requires examination. To this end, evidence generated only in recent decades, including new pollen analyses, sedimentary charcoal studies, and the emerging history of early peopling of the Andes, will be appraised.

The term ‘grass páramo’, of central interest here, refers to the exuberant tussock-dominated grasslands found between the UFL and the desert páramo above (Figure 1a, b). This band occupies elevations between a lower limit adjoining the UFL and an upper limit at 4000–4300 m. Mature tussocks at moderate grass páramo elevations establish a closed canopy typically of 70–90% coverage, and reaching 60 cm or more in height. Most shrub páramos retain a tussock coverage of 30–40%. Grass páramo may also include isolated forest patches, gallery forests, and wetlands of variable size within the grassland matrix. The objective of the present discussion is to examine the etiology of grass and brush páramo specifically. Desert páramo at higher elevations and puna grasslands at higher latitudes are not objectives, although they are referenced in the discussion and tables where they provide context for the analysis of grass páramo.

Figure 1.

(a) A grass páramo at 3600–3800 m, Eastern Cordillera, Ecuador. (b) A pure stand of tussock grass at 3600 m, Eastern Cordillera, Ecuador. Plants are 50–70 cm high and form a closed canopy.

Data sets

Pollen record

At local last glacial maximum (LLGM), 26,000–20,000 cal. yr BP, ice covered the northern Andes down to 3300–3600 m, and for some wetter locations to 2800 m, where ice tongues may locally have been in proximity to montane forest (Clapperton, 1993a, 1993b; Clark et al., 2009; Heine, 2000; Lachniet and Vázquez-Selem, 2005; Thouret et al., 1996; Van der Hammen and Cleef, 1986). At this time the UFL appears to have been situated at about 2000–2400 m (Hooghiemstra and Van der Hammen, 2004; Van der Hammen and Cleef, 1986; Van der Hammen and Hooghiemstra, 2003). By 19,000–18,000 cal. yr BP deglaciation was underway. It proceeded unevenly but in general terms sites at 3700 m were free of ice by 17,000–15,000 cal. yr BP, and at 4000 m by 15,000–12,000 cal. yr BP (Smith et al., 2008). Essentially modern temperatures and circulation patterns were established by 11,000–10,000 cal. yr BP. The observable grass páramo, above the band of UFL displacements and across its entire latitudinal range, achieved its placement during this period.

The Pleistocene movements of forest and herbaceous belts can be traced because the pollen signature of forest and páramo differ notably (Grabandt, 1980, 1985; Grabandt and Nieuwland, 1985; Hansen and Rodbell, 1995; Hansen et al., 2003; Horn, 1993; Horn and League, 2005; Islebe and Hooghiemstra, 1997; Melief, 1985; Moscol Olivera et al., 2009; Niemann et al., 2010; Rodgers and Horn, 1996; Van der Hammen et al., 1981; van ’t Veer et al., 1995). Simple percentages of (1) Poaceae, (2) indicator páramo taxa, and (3) arboreal pollen can be employed to distinguish forest from páramo over its broad latitudinal range (e.g. Groot, et al., 2011; Hooghiemstra, 1984). This approach is applied in Table 1 to 50 high elevation pollen cores, in which defining percentages of the three assemblages are considered in consort.¹

Table 1.

Pollen studies >2500 m elevation, montane tropical America. All páramo sites with adequate dating controls from the northern Andes and Costa Rica are included. Studies are listed by geographic proximity ordered from lower to higher altitude. To facilitate identification, contemporary forest (or agriculture from forest) and near-UFL sites are shaded.

Site name	Altitude (m a.s.l.)	Latitude^a	Current vegetation	Basal date cal. yr BP^b [interpolated date]	Date site became forested (cal. yr BP ^c) or Páramo to indicate that site remained as páramo grassland since establishment	Reference
Costa Rica
La Trinidad I, II	2700	9°50′N	Bog in montane forest	[9800]	9500	Islebe et al. (1996)
La Trinidad III	2700	9°50′N	Bog in montane forest	13,919	11,000	Islebe and Hooghiemstra (1997)
Lago de las Morrenas	3480	9°29′N	Bamboo- and shrub- páramo	11,773	Páramo	Horn (1993)
Venezuela
Páramo de Mucabají	3540	8°47′N	Páramo	9263	Páramo	Salgado-Labouriau et al. (1992)
Páramo de La Culata	3600	8°45′N	Páramo	8349	Páramo	Salgado-Labouriau and Schubert (1976)
Páramo de Mucabají	3650	8°47′N	Páramo	14,715	Páramo	Salgado-Labouriau (1979, 1986); Salgado-Labouriau et al. (1977)
Páramo Miranda	3920	8°55′N	Superpáramo	13,324	Páramo	Salgado-Labouriau et al. (1988)
Laguna Verde Alta	4215	8°51′N	Superpáramo	14,240	Páramo	Rull et al. (2005)
Colombia
Eastern Cordillera (Sierra Nevada de Cocuy, High Plain of Bogotá, Páramo de Sumapaz)
Duitama Plain (Pantano de Vargas)	2510	5°47′N	Agriculture	9109	c. 9800	Gómez et al. (2007)
Ciudad Universitaria	2560	4°67′N	Agriculture	28,576	c. 11,400	Van der Hammen and Hooghiemstra (1995)
El Abra III	2570	5°00′N	Agriculture and secondary forest	14,299	c. 11,400	Schreve-Brinkman (1978); Van der Hammen and Hooghiemstra (1995)
El Abra II B3	2570	5°00′N	Agriculture and secondary forest	13,052	c. 11,200	Van ‘t Veer et al. (2000)
Laguna Fúquene I, II	2580	5°27′N	Agriculture	24,520	10,700–11,000	Van der Hammen and Cleef (1986); Van der Hammen and Hooghiemstra (1995); Van Geel and Van der Hammen (1973)
Laguna Fuquene III	2580	5°27′N	Agriculture	48,263	10,057	Van der Hammen and Hooghiemstra (2003)
Laguna Fuquene VII	2580	5°27′N	Agriculture	23,479	15,372	Van ‘t Veer et al. (2000)
Lake Alsacia	3100	4°00′N	Montane forest	17,643	12,100	Melief (1985)
Agua Blanca I	3250	5°00′N	UFL/Páramo w/forest patches	40,856	Páramo ^d	Helmens and Kuhry (1986)
Agua Blanca (PAB II, III)	3250	5°00′N	UFL/Páramo w/forest patches	9296	c. 4900^e	Kuhry (1988b)
Ciénaga del Visitador	3300	6°08′N	Páramo	14,870	Páramo	Van der Hammen (1974); Van der Hammen and González (1965); Van der Hammen and Hooghiemstra (1995)
La Guitarra	3450	4°00′N	UFL/Shrub páramo	18,323	Páramo ^f	Melief (1985)
Laguna Ciega	3510	6°30′N	Páramo	24,899	Páramo	Van der Hammen et al. (1981)
La Primavera	3525	4°00′N	Páramo	13,482	Páramo	Melief (1985); Van der Hammen and Cleef (1986); Van der Hammen and Hooghiemstra (1995)
Páramo de Palacio	3550	5°16′N	Páramo	9132	Páramo	Van der Hammen (1974); Van der Hammen and Cleef (1986); Van der Hammen and González (1960a)
Peña Negra (PPN I)	3625	5°05′N	Mire w/ grass páramo	14,402 [16,300]	Páramo	Kuhry (1988b)
Laguna Verde (PLV I, II)	3625	5°15′N	Mire w/ páramo	5972 (I); 5942 (II)	Páramo	Kuhry (1988b)
Andabobos	3750	4°08′N	Páramo	16,901	Páramo	Melief (1985); Van der Hammen and Hooghiemstra (1995)
Laguna de los Bobos	3800	6°15′N	Páramo	c. 5000	Páramo	Marchant et al. (2001); Van der Hammen (1974)
El Gobernador	3815	4°00′N	Páramo	8400 [10,500]	Páramo	Melief (1985)
Valle de Lagunillas	3880–4220	6°25′N	Páramo	11,238; 14,149	Páramo	González et al. (1965)
Cordillera Central (Nevado de Ruiz)
La Laguna	2900	4°55′N	Mixed Weinmannia forest and pastures	22,511	c. 10,100	Helmens et al. (1996)
El Billar I and II	3600	4°54′N	Páramo	15,544	Páramo	Melief (1985); Van der Hammen and Hooghiemstra (1995)
El Bosque Mire	3650	5°45′N	UFL/Montane forest	6553	Before 6553	Kuhry (1988a)
Parque Los Nevados Transect (TPN 26)	3900	4°45′N	Páramo	[6500]	Páramo ^g	Salomons (1986)
Laguna Otún (TPN 21B)	3950	4°47′N	Páramo	8266 [10,500]	Páramo ^h	Salomons (1986)
La Cachucha	4050	5°40′N	Páramo	7100 [10,000]	Páramo ⁱ	Bakker and Salomons (1989); Salomons (1986)
Northern and Central Ecuador
Surucucho	3180	2°45′S	UFL/Montane forest w páramo above	13,704	11,000 ^j	Colinvaux et al. (1997)
Guandera (G15-II)	3400	0°36′N	Montane forest	6758 [7150]	~6000	Bakker et al. (2008)
Guandera 1^k	3540	0°36′N	UFL/Montane forest	12,489	4000–5000	Di Pasquale et al. (2008)
Chorreras	3700	2°45′S	Páramo	c.17,000	Páramo	Hansen et al. (2003)
Quimsacocha	3780	3°02′S	Páramo	7672	Páramo	Jantz and Behling (2012)
Guandera 2^k	3810	0°36′N	Páramo	13,006	Páramo	Di Pasquale et al. (2008)
Guandera 3^k	3890	0°36′N	Páramo	12,910	Páramo	Di Pasquale et al. (2008)
Pallcacocha	4060	2°45′S	Páramo	15,500	Páramo	Hansen et al. (2003)
Southern Ecuador
Cocha Caranga	2710	4°02′S	UFL/ shrub páramo above	12,569 [14,500]	14,000–9700^l	Niemann and Behling (2009)
El Tiro Pass	2810	3°50′S	Shrub páramo	19,836	Páramo	Niemann and Behling (2008)
Cerro Toledo (CT core)	3150	4°22′S	Shrub páramo	19,999	Páramo ^m	Brunschön and Behling (2009)
Laguna Rabadilla de Vaca	3312	4°15′S	Shrub and grass páramo	10,380 [11,700]	Páramo	Niemann et al. (2009)
Peru
Laguna Chochos	3285	7°38′S	UFL/Forest close to lake in páramo matrix	14,370 [17,150]	11,500ⁿ	Bush et al. (2005)
Laguna Baja	3575	7°42′S	Páramo	13,915	Páramo ^o	Hansen (1995); Hansen and Rodbell (1995)
Laguna La Compuerta	3950	7°30′S	Páramo (termed ‘Wet puna’)	30,041 [33,000]	Páramo	Weng et al. (2006)

Notes: ^aLatitude taken from cited study, and Marchant et al. (2001).

All ¹⁴C dates have been calibrated using the Fairbanks0107 curve (Fairbanks et al., 2005). Dates interpolated from dated segments or from sedimentary patterns shared with other cores are given in brackets. Where calendar year dates are given as a range, a mean date is calculated.

To determine if a site remained herbaceous or became forested, the criteria discussed in Note 1 were applied to the pollen diagrams of each study. Páramo thus means that Poaceae percentage either never descended below 20%, or that during the period in which they were <20%, indicator species totals were greater than 20%. High Poaceae percentages associated with human disturbance (mostly late in the Holocene) were discounted as representative of grass páramo. Sites near UFL and which may have been briefly forested are discussed in table footnotes. In all cases the dates of forest establishment are estimated from the published pollen sum diagrams.

Interpreted by Helmens and Kuhry (1986) to be a ‘timberline forest’ after 8300 cal. yr BP. But because páramo taxa continue to account for 30–40% of pollen, and subpáramo taxa for another 20–30%, by the criteria applied here the site was more likely a shrub páramo with proximal forest. The current vegetation surrounding the coring site is open shrub páramo with forest patches.

Forests were dominated by Weinmannia, but the pollen evidence leaves unclear if the forest cover was continuous or, as now, patchy. The persistence of sub-páramo elements (Helmens and Kuhry, 1986) suggests a shrub páramo matrix with Weinmannia patch forests.

Interpreted by Melief (1985) to have been forested by 8800 cal. yr BP, becoming shrub páramo at c. 3200 cal. yr BP. However, the persistence of Poaceae (20–30%) during the ‘forest’ phase, and isolated tree pollen peaks reaching only 35–50%, suggest a shrub páramo, albeit with forest nearby.

Shrub páramo c. 6250 cal. yr BP, then grass páramo to present.

Shrub páramo c. 6000–3000 cal. yr BP, then grass páramo to present.

Shrub páramo between c. 7500 and 6250 cal. yr BP, then grass páramo to present.

Polylepis forest early; then after 10,000 cal. yr BP dominated by Weinmannia.

Guandera (Di Pasquale et al., 2008) consists of three sites: One (3540) currently in montane forest and two (3810 m and 3890 m) in páramo, noted here as Guandera 1, 2 and 3. Macroscopic charcoal was used to identify the charred remains of characteristic páramo and forest species, and thus the plant community.

Site at modern UFL. Authors argue that forest was replaced by páramo 9700–4200 cal. yr BP, and then reverted to forest. However, the persistence of Poaceae (≥20%) in mid- to late-Holocene forests suggests a shrub páramo rather than montane forest. Continued burning throughout the ‘forest’ phases, as evidenced by persistent sedimentary charcoal (except for the last 1200 years), is inconsistent with montane forest.

But UFL close after 8900 cal. yr BP

Lakeside forest was stable since the initial woody colonization of the site, but the páramo matrix above and beyond the lake appears to have increased coverage after 6000 cal. yr BP, as suggested by an increase in Poaceae to 25–35%.

But UMF close after 10,700 cal. yr BP.

For 18 forested sites (or sites deforested for agriculture) in the northern Andes at elevations of 2500–3200 m, replacements of Pleistocene herbaceous vegetation by forest occurred principally between 12,000 and 9500 cal. yr BP. Early-Holocene upslope advance of forest halted at approximately current local UFLs. In general these are found at 3200–3400 m, but range from 2800 m in southern Ecuador (Niemann and Behling, 2009) to 3650 m in northern Ecuador (Bakker et al., 2008) and the wetter páramos of Colombia (Kuhry, 1988a). Holocene UFLs never reached higher than 3650 m at any location yet examined, and remained below 3300–3400 m for most sites.

Cores taken near the modern local UFLs function as sensitive indicators of forest advances and retreats, and these sites have constrained the postglacial variation of UFL elevations. As evident in Table 1, all 32 current shrub and grass páramo sites in the northern Andes and Costa Rica remained in páramo during the entire Holocene, as often concluded by study authors (Bakker et al., 2008; Di Pasquale et al., 2008; Horn, 1993; Horn and League, 2005; Melief, 1985; Moscol Olivera and Hooghiemstra, 2010; Niemann et al., 2010; Salgado-Labouriau and Schubert, 1976; Salgado-Labouriau et al., 1992; Wille et al., 2002). Evidence is absent for a generalized tree cover at current grass páramo elevations as proposed by Lægaard (1992), and supported by others (Hansen and Rodbell, 1995; Hansen et al., 1994; Holtmeier, 2003; Keating, 2000, 2007; Luteyn, 1999; Sarmiento, 2000, 2002; Sarmiento and Frolich, 2002; Young and Keating, 2001). Hence, with the exception of variation around the present UFL, grass páramo maintained its approximate current distribution since its original emplacement at the Pleistocene/Holocene transition.

Sedimentary charcoal record

For the zonal school tussock grass dominance is achieved by climatic and edaphic controls alone, and thus fire is of minor import for grass páramo etiology (Baruch, 1984; Marchant et al., 2001; Pérez, 1998; Thouret et al., 1996). Anthropic fire may alter formerly pristine páramo or extend its range, as when brush páramo is absent and presumably eliminated by burning (Bader and Ruijten, 2008; Bakker et al., 2008; Hofstede et al., 1995; Moscol Olivera and Cleef, 2009a; Verweij and Beukema, 1992). Environmental regulations of northern Andean countries largely subscribe to the zonal school of páramo etiology, with its distinction of natural versus human fires and broad condemnation of the latter (Keating, 2007). Urban public sentiment considers burning to be inimical to the survival of grass páramo (Horn, 1998), and associates its occurrence with páramo ‘destruction’.

A subgroup of researchers holding the zonal view, however, allows that fire is part of páramo ecology within the dominant forcing of climate and soils. They note broad plant tolerance to fire and argue or assume that lightning and volcanoes have provided natural sources of ignition (Horn, 1993; Horn and League, 2005; Luteyn, 1999; Moscol Olivera and Cleef, 2009b). Whether this tolerance represents an evolutionary adaptation to fire or the fortuitous availability of traits that allow persistence in a burning regime remains to be determined (Lægaard, 1992). From the zonal perspective of this subgroup, human fire merely augments a natural fire regime adequate to páramo creation and maintenance. Anthropic fire, and human disturbances in general, may explain grass páramo expansion or magnify the dominance of tussocks but is not considered pertinent to the etiology of the flora (Brunschön and Behling, 2009; Niemann and Behling, 2009; Sklenář and Ramsay, 2001; Thouret et al., 1996). Although the evidence is clear that páramo plant community floristics evolved with each glacial cycle (Hooghiemstra and Van der Hammen, 2004), the strong dominance of tussocks in contemporary grass páramo is believed to have its analogue in Pleistocene páramos. Thus, a contemporary ‘undisturbed’ grass páramo (sensu Hofstede et al., 1995) not only represents a zonal ‘true páramo’, but it also represents a grass páramo that is assumed to pre-date the arrival of humans.

Because fire burdens required by the zonal hypothesis are negligible, natural sources of ignition will be largely absent and low concentrations of sedimentary charcoal are predicted. Nonetheless, for zonal advocates who acknowledge fire as an ecological variable secondary to climatic controls, natural fire should be of sufficient intensity to explain the observed fire tolerance of páramo plants, and be represented by moderate concentrations of sedimentary charcoal from these natural sources.

Mid 20th century pollen analyses for Andean climatic reconstructions did not report charcoal concentrations until this invaluable line of inquiry was undertaken by Horn (1986, 1989a). Now the inclusion of charcoal data has become standard practice (Horn and Kappelle, 2009). Novel pedoanthracological methods are being introduced (Di Pasquale et al., 2008, 2010), as are higher time-resolution analyses (League and Horn, 2000). Investigators commonly use charcoal occurrence as a proxy for climate, where an increase in charcoal values over the previous period is taken to represent drier, more seasonal or warmer conditions, because these facilitate burning. Cooler conditions, even when dry, are associated with less burning (Power et al., 2010). In addition to its use as a proxy for climate, the abundance of charcoal informs the vegetation’s fire ecology. Table 2 summarizes charcoal data from 22 páramo and upper montane forest/UFL sediment cores and soil profiles, plus three cores from puna.

Table 2.

Páramo, puna and UFL Pleistocene/Holocene charcoal records.

Site (core or profile name)/ country/ latitude	Altitude (m a.s.l.)	Current vegetation	Particle size (µm)	Basal date in cal. yr BP^a Oldest date reported(mean of date range)[Interpolated date]	Date of initial charcoal rise (cal. yr BP^b) (bold) (subsequent peaks in parentheses; asterisk indicates major rise)	Reference
Lago de las Morrenas (1, 1989-II)/ CR/9°29′N	3477	Páramo	>500	12,360 (11,795)	9500–9100 (4750, 2600, 500)*	Horn and League (2005); League and Horn (2000)
Lago Chirripó (1985-II)/ CR/9°29′N	3520	Páramo	>700	4850 (4630)	NA^c (4450, 2200, 1100)	Horn (1989a, 1993); Horn and League (2005)
Guandera^d (Gua1, Gua2, Gua3)/EC/ 0°36′N	1: 3890 2: 3810 3: 3540	1: Páramo 2: Páramo 3: Forest	400–2000	1: 13,033 (12,489) 2: 13,158 (13,006) 3: 12,994 (12,910)	1: NA^c (12,500) 2: NA^c (13,000) 3: NA^c (12,900)	Di Pasquale et al. (2008)
Guandera^e (Gua1; Gua2a; Gua2b; Gua3; Gua5; Gua6) /EC/ 0°36′N	1: 3890 2a: 3810 2b: 3810 3: 3540 5: 3530 6: 3790	1: Páramo 2a: Páramo 2b: Páramo 3: Forest 5: Forest 6: Páramo	>400	1: 13,250 (12,250) 2a: 13,200 (12,875) 2b: 13,250 (12,950) 3: 12,895 (12,848) 5: 11,200 (11,130) 6: 11,420 (11,310)	1: NA^c (12,500) 2a: NA^c(12,900)* 2b: NA^c (12,900) 3: NA^c (12,800) 5: NA^c (11,200) 6: NA^c (11,400)	Di Pasquale et al. (2010)
Laguna Pallcacocha/ EC/2°45′S	4060	Páramo	>100	15,000	12,900* (12,100, 11,100, 9500)	Hansen et al. (2003)
Laguna Chorreras/ EC/2°45′S	3700	Páramo	>100	17,500	11,100 (9600, 7000, 5300)*	Hansen et al. (2003)
Laguna Surucucho/ EC/2°55′S	3180	Forest near UFL	>25 & <25	13,704	12,000 (9000, 3500)*	Colinvaux et al. (1997)
Laguna Quimsacocha/ EC/3°02′S	3780	Páramo	10–25, 25–50, 50–100, >100	7672	NA^c (7500, 3900)*	Jantz and Behling (2012)
El Tiro Pass/EC/3°50′S	2810	Shrub páramo	2–150	19,836	8000 (5000, 1900, 350)	Niemann and Behling (2008)
Cocha Caranga/ EC/4°02′S	2710	Shrub Páramo near UFL	10–150	[14,500]	9700 (7700, 6000, 4200)	Niemann and Behling (2009)
Laguna Rabadilla de Vaca/ EC/4°15′S	3312	Páramo	10–150	10,380 [11,700]	NA^c (11,500, 8200, 3400)*	Niemann et al. (2009)
Cerro Toledo/ EC/4°22′S	3110	Shrub páramo	5–150	20,000	19,000* (16,000, 15,000, 6000)	Brunschön and Behling (2009)
Laguna La Compuerta/ PE/7°30′S	3950	Páramo	<180	30,041	9500* (8400, 3500*)	Weng et al. (2006)
Laguna de Chochos/ PE/7°38′S	3285	Forest and páramo	>100 & <100	14,3709 [17,150]	15,000 (12,000, 8000, 2000)	Bush et al. (2005)
Laguna Baja/ PE/7°42′S	3575	Páramo	>100	13,915	13,000-11,600 (6500, 5600)	Hansen and Rodbell (1995)
Lake Pacucha/ PE/13°36′S	3095	Agriculture/ near UFL	>180	24,000	15,000 (9800, 5500)	Valencia et al. (2010)
Lake Titicaca/ PE/16–17°S	3810	Agriculture/puna	>180	370,000	Interglacials (major rise): 320 kya; 210–190 kya; 135–120 kya. Glacials (limited rise): 240 kya; 60 kya.	Hanselman et al. (2011)
Lake Titicaca/ PE/16–17°S	3810	Agriculture/puna	>180 & 179–65	27,500	13,200* (11,000, 10,000, 7800, 5200)	Paduano et al. (2003)

Notes: ^a¹⁴C ages have been calibrated to calendar years BP using the Fairbanks 0107 curve (Fairbanks et al., 2005). Where authors report date ranges, the oldest date is cited first, followed in parentheses by the mean age for the range. Interpolated dates as provided by authors are given in brackets.

Dates are interpreted from graphic presentations in the reference, or as stated in textual descriptions. Asterisks indicate the higher absolute charcoal concentration(s) of the series.

Sites showing high charcoal concentrations at the basal date indicate a significant fire regime but cannot be used to identify a rise.

Because this pedoanthracology study used macro-charcoal fragments to identify forest versus páramo indicator species, and did not examine charcoal concentrations, reported dates show the earliest evidence of burning but do not indicate fire regime intensity. The authors however report that all soil profiles contained charcoal at all depths, except the top organic layer of the contemporary forest (Di Pasquale et al., 2008).

Three of the six soil profiles reported in Di Pasquale et al. (2010), located at 3890, 3810 and 3540 m a.s.l., may be those described in Di Pasquale et al. (2008). Earliest charcoal dates between the two studies are quite similar but not identical, suggesting a separate dating procedure for the 2010 results.

A number of limitations apply to these data. Although sedimentary charcoal is used to infer fire regimes, concentrations may reflect other factors including hydrology, wind directions, soil residence time and the nature of the vegetation burned (Horn, 1989a). The studies reported in Table 2 resist direct comparison because abundances are calculated in a variety of ways and for different size classes of charcoal particles. The diagnostic value of micro- (<100 µm) and macroscopic (>100 µm) particles is particularly problematic: Although most microscopic charcoal is generated in the local watershed, wind-borne deposition may carry particles from distant sites (Horn and Kappelle, 2009). Nonetheless, with regard to the studies reported in Table 2, the microscopic concentration curves generally parallel macroscopic curves, suggesting that both respond to a local fire regime. An additional difficulty is that intervals between samples vary widely, representing time periods of half a century to half a millennium. As the resolution of the record lessens, the absolute fire frequency becomes more difficult to discern.

But a fundamental observation follows from the data of Table 2: All studies report persistent macroscopic charcoal in páramo soils and sediments from the terminal Pleistocene and throughout the Holocene. No study that looks for charcoal fails to find it in the core or profile, except during periods when the site was under forest (Di Pasquale et al., 2008), and during glacial intervals, when vegetation cover was negligible or absent (Hanselman et al., 2011; Paduano et al., 2003). These data thus suggest that grass páramo fire is ubiquitous and that Holocene fire regimes began before the Pleistocene/Holocene boundary (Di Pasquale et al., 2010). For example, for a Costa Rican páramo core covering the last 11,000 cal. yr, 29 of 30 sediment samples contained at least one fragment greater than 700 µm in size (Horn and League, 2005). Venezuela and Colombia are not represented because to date their pollen analyses of páramo sites have not reported charcoal concentrations. In the absence of future studies that show little or no charcoal for extended periods, the evidence strongly suggests that Holocene grass páramo is a fired landscape.

Fire sources

Were grass páramo zonal vegetation, but also to some degree fired, natural ignition sources must exist. In this search a recorded incident of ignition, or its mere likelihood, does not establish a sufficient incidence of natural fire on a very large and fragmented landscape. Fire must occur at a spatial density and with a return frequency adequate to generate the sedimentary charcoal abundances recorded. Volcanoes and lightning are considered the two principal sources of natural fire in Neotropical mountains. Ignitions caused by sparks from falling rocks, crystal lensing and spontaneous combustion have been documented in other habitats, but there is no evidence to suggest that they occur at the densities and return intervals required for páramo creation and maintenance (Trabaud, 1987; Vogl, 1974). The absence of confirmed reports in the páramo literature of natural fires (Di Pasquale et al., 2008; Horn, 1989b; Luteyn, 1999) underscores the need to critically assess volcanism and lightning as sources for the ubiquity of charcoal in páramo sediments.

Volcanism

Researchers often assume that volcanoes generate fire (Horn and Kappelle, 2009; Keating, 2007; Moscol Olivera and Cleef, 2009a), even though direct observation of fire resulting from volcanic eruptions has not been documented in páramo. Indirect evidence is occasionally cited as, for example, charcoal in a páramo bog core ascribed to the activity of the neighboring Nevado Santa Isabel volcano in Colombia (Kuhry, 1988a). Likewise, charcoal in a páramo profile was linked to volcanism because the core examined showed interbedded Holocene tephra layers (Horn and Kappelle, 2009). But associations of this sort cannot establish causality for a fire regime at the landscape scale. I argue that eruptions are an inadequate source of natural fire for three principal reasons: (1) the localized and non-incendiary nature of most volcanic phenomena, (2) the spatial distribution of northern Andean volcanoes active during the Holocene, and (3) the episodic nature of eruptions.

Volcanoes can cause vegetation fires (Mueller-Dombois, 1981), but their eruptive behavior would generate limited ignition foci. Although pyroclastic flows are hot, their leading edges travel fast and immediately bury what they burn. When a pyroclastic flow stops, it may have cooled to below the ignition temperature, but in any case the potential ignition area is limited to the border of the flow’s final emplacement on the original surface. Because pyroclastic flows follow the pre-existing hydrology, they concentrate hot material and carry it quickly downstream and away from headwater páramos. When an eruption includes molten lapilli, deposition is most intense on the flanks and base of the volcano (as is the lightning generated by the pyrocumulus clouds of violent eruptions), where normally no plant matter is available to burn or to carry fire. The principal eruptive products in most cases in the northern Andes are debris flows and ash fallout. These are not associated with fire, and thus a distinct tephra bed reminds us that the páramo in question was not burned but rather buried (Hall and Mothes, 2008).

Second, volcanoes as natural fire sources are limited in their distribution. Were volcanoes a principal explanatory source of natural fire, the location of the major páramo islands should correspond to that of volcanoes active now and throughout the Holocene. This is not the case, as illustrated by the páramos of the Andean depression in southern Ecuador and northern Peru, which have considerable sedimentary charcoal but have not seen volcanism for 2 million years (Beate et al., 2001). Instead charcoal is found everywhere that the mountain terrain achieves páramo elevations of 3200 to 3600 m or higher. In areas of relatively dense volcanic activity, such as central and northern Ecuador, volcanoes active in the Holocene were separated by 20–60 km (Hall and Mothes, 2008; Hall et al., 2008). Isolated páramo fires initiated by volcanic activity would have great difficulty spreading across the intervening distances, because natural obstacles to fire passage make most grass páramos highly fragmented surfaces, even without contemporary human infrastructure (Horn and Kappelle, 2009).

A third reason to doubt volcanoes as sources of páramo charcoal is the sporadic nature of volcanic activity. Intervals between eruptions along Ecuador’s volcanic arc, for example, ranged from hundreds to thousands of years during the Holocene (Hall and Mothes, 2008). In the nearly 500 years since the Spanish Conquest, only six of the arc’s 68 volcanoes have experienced eruption (Barba et al., 2008). Perhaps within a given eruptive episode fire frequency could be sufficiently high to burn a local páramo repeatedly. But if volcanoes were a singular source of fire, charcoal production should diminish or cease during the prolonged quiescent period between eruptions, a reduction not supported by sedimentary charcoal data.

Lightning

Lightning is more evenly distributed in space and in time than volcanic activity, and would seem at first glance to be a suitable natural source of páramo fire. Páramo students frequently invoke lightning (Cleef, 1981; Horn, 1989a, 1989b; Horn and League, 2005; League and Horn, 2000; Luteyn, 1999; Moscol Olivera and Cleef, 2009b; Sarmiento and Frolich, 2002). The Tropics have the highest global flash rate, which would appear supportive of lightning as a fire source (Horn and Kappelle, 2009; Horn and Sanford, 1992). This, however, ignores important distinctions at regional scales as, for example, the relatively low flash rate in the tropical Andes compared with the much higher Amazonian forest rate. But the principal difficulty of lightning as a source of páramo fire is its low ignition ratio, which compares the number of lightning strikes with the number of ignitions caused by those strikes (Christensen, 1993). Mid-latitude coniferous forests and prairies support the highest ignition ratios. Andean montane forest and undisturbed lowland forest, in contrast, are practically immune to ignition (Mueller-Dombois, 1981). For this reason no suggestion is made in the literature that páramo was created from undisturbed forest by the direct action of lightning.

But it may be argued that even in a wet mountain climate grass can dry out enough to burn, and thus the lightning ignition ratio will be considerably higher for páramo vegetation than for closed forest. Nonetheless, for grass fires to spread over large areas, two conditions must be met:

Strikes must coincide with periods when vegetation is dry. Since the likelihood of lightning striking any particular point is very small, the occurrence of fire at any point not directly hit by lightning depends on the ability of fire to spread to that point. Thus, if lightning is to spread widely it must occur during a dry period, of which two kinds occur in páramos: The dry season per se (of <1 to 5 months), and a rainless and sunny period of a few days to a week or two that may occur within a rainy season. The dry season is an unlikely period for lightning to occur because the dominant air masses are stable and not conducive to thunderstorms. Lightning is more likely when the air is unstable during the wet season or in the transition between seasons. But tropical mountain thunderstorms form close to the surface in humid air masses that restrict evaporation, and thus virtually all precipitation produced by the cloud reaches the ground (Mueller-Dombois, 1981; Young, 1993). Exceptionally, positive lightning emanating from a cloud top may travel a sufficient horizontal distance to strike outside the precipitation footprint of the cloud, but this phenomenon has not been reported for the high tropical Andes. Instead, when lightning is most likely to occur, vegetation is simultaneously most likely to be wet.

Fire must transcend fire obstacles. The spread of fire is not only impeded by rain but by a variety of obstacles on the ground. Because these obstacles delimit a discrete fire surface, the process of ignition must be repeated for each surface, and thus the number of ignition points required would be very large. This contradicts our impression when observing an expanse of grass páramo, which appears to be an integrated whole because the vegetation is uniformly dominated by tussock grasses. But our observation in fact subsumes many fire surfaces, each burned in a separate ignition event. Especially where the topography is broken, or interrupted by watercourses or rock fields, the number of fire surfaces per unit area will be high. Modern residents in these kinds of páramo who burn their holdings to improve forage for cattle or sheep testify to the need to plant fire in many places. To burn an entire observed expanse of grass páramo requires, therefore, a number of ignitions equal to the number of discrete fire surfaces.

The probability of a given páramo area being burned due to lightning can be summarized: A low probability of lightning striking any particular point would be multiplied by a low ignition ratio, and this very small number by the inverse of the total number of fire surfaces that comprise an observed páramo. But if grass páramo becomes woodier in the absence of fire (Brandbyge, 1992; Cierjacks et al., 2008; Gade, 1999; Jokisch and Lair, 2003; Lægaard, 1992; Sarmiento and Frolich, 2002) this grand improbability must be further reduced by two elements of the temporal dimension: Mean return interval and deviation around the mean. To be held responsible for burning páramo, lightning must initiate a fire for all discrete fire surfaces with a sufficiently short return frequency, and a low deviation around this value. The lightning return frequency and initiation of localized fires cannot delay much beyond the mean because the facility of a pioneering forest to ignite is considerably less than the grass it replaced, and thus with each failure of fire to return with sufficient frequency, the dominance of woody species increases and the likelihood of fire decreases. In sum, lightning ignitions (as opposed to strikes) would have to be spatially dense, temporally frequent and invariable to a degree that would appear extremely unlikely. These observations suggest that although natural ignitions are possible, no natural sources of fire can explain the terminal Pleistocene and Holocene sedimentary charcoal records.

Humans as a source of páramo fire

It appears, by default, that human burning generated virtually all Late Pleistocene and Holocene sedimentary charcoal. If this is the case, then charcoal concentrations for páramo during human occupation should be elevated as compared with those of a pre-human period (Di Pasquale et al., 2010; Hanselman et al., 2011; Pinter et al., 2011; Piperno et al., 1990; Rull, 2009). When associated in a tropical forest setting with disturbance pollen or that of plant domesticates, or with burned phytoliths of disturbance indicators, charcoal reliably represents human-set fires (Behling, 2000; Gómez et al., 2007; Piperno, 2006; Piperno and Jones, 2003; Piperno et al., 1990). This confidence derives from the charred cultural artifacts themselves, but also from the established difficulty of causing ignitions in undisturbed wet forest by natural sources, and the alien nature of fire to wet forest ecology.

Páramo is a more difficult environment in which to segregate the human fire contribution because the vegetation is combustible and the plant community’s tolerance of burning may suggest a history of fire exposure. But if a pre-human burning regime can be identified, an initial and sudden rise should represent the first arrival of humans to the sampled site. To qualify undeniably as ‘pre-human’ would require dates prior to 15,000 cal. yr BP (and perhaps 20,000 cal. yr BP, as discussed below) for elevations currently in páramo. Ideally the concentration curves would include a long pre-human time series, to reliably identify the background frequency and variability of natural fire.

Of the charcoal sites summarized in Table 2, 19 have basal dates older than 12,500 cal. yr BP, and eight of these are ≥15,000 cal. yr BP. They universally reveal a low level of fire in páramo and puna immediately following the LGM. Fire regimes at most sites intensify suddenly during the period c. 13,000–11,000 cal. BP, and three intensify at 15,000 or earlier. Because of their relatively long pre-rise periods and moderate elevations (allowing for early vegetation development and thus an early capacity to be burned), the studies at Chorreras, El Tiro, La Compuerta and Los Chochos are particularly valuable in detecting presumable anthropic inputs to the fire regime (Bush et al., 2005; Hansen et al., 2003; Niemann and Behling, 2008; Weng et al., 2006).²

Parallel to this finding, the sudden rise in the percentage of Poaceae pollen in the final millennia of the Pleistocene may reveal not only a warming climate but also human occupation and burning, given the competitive advantage gained by bunchgrasses in a fire regime (Brunschön and Behling, 2009; Niemann and Behling, 2009). Accompanying the intensified fire regime and Poaceae increase, the pollen data often signal a reduction of shrubs characteristic of brush páramo, or of Polylepis, in these cases reflecting woody intolerance to burning (Bush et al., 2005; Cierjacks et al., 2008; Weng et al., 2006; Williamson et al., 1986).

Low fire incidence in the earliest core dates certainly reflects limited biomass, especially for higher elevations sites (Mayle et al., 2009). To illustrate, a core covering 370,000 years and three glacial cycles recovered from Lake Titicaca, at 3810 m in the Peruvian Altiplano, establishes that fire was largely absent during glacial periods of up to 50,000 years, but that during interglacials of 15,000 to 20,000 years’ duration a vegetation was being burned that resembled modern puna (páramo-like grassland) and puna brava (super páramo-like) (Hanselman et al., 2011). The macroscopic charcoal concentrations for these interglacials, all presumably resulting from natural ignitions, generated sedimentary charcoal concentrations of 0.1–0.9 mm²/cm³ (with a single outlier of 1.9 mm²/cm³ at 210 kya), and a mean of ~0.3 mm²/cm³.

But how intense a fire regime do these values suggest? Two additional puna studies address this question. From a separate Lake Titicaca core covering the last c. 27,500 years, pollen first appears c. 21,000 cal. yr BP and charcoal at 17,700 cal. yr BP, with very low abundances (Paduano et al., 2003). Then, at c. 13,200 cal. yr BP concentrations suddenly rise to 15 mm²/cm³ and continue at about 10 mm²/cm³ before declining to almost zero in the last 4000 years, perhaps reflecting overgrazing or the agricultural transformation of the zone around Lake Titicaca. These two complementary studies establish that fire was virtually absent during the last three glacial maxima, when plant cover was minimal. But they also indicate that in the presence of puna and puna brava vegetation during interglacials, burning was very limited as compared with levels reached after 13,200 cal. yr BP, when they increased by orders of magnitude.

This sequence and timing is echoed by the charcoal abundances at Lake Pacucha in southern Peru, now situated in upper montane forest at 3095 m but dominated during the terminal Pleistocene by Poaceae, with abundant Asteraceae and Polylepis. Macroscopic charcoal concentrations remained below 4 mm²/cm³ between 24,000 cal. yr BP and 15,000 cal. yr BP, with a mean of <1 mm²/cm³. They rose suddenly after 15,000 cal. yr BP to 10–15 mm²/cm³ and remained elevated but variable until lowering dramatically 500 cal. yr BP (Valencia et al., 2010).³ From these charcoal studies with basal dates older than ~12,500 cal. yr BP, combined with the severe limitations on natural ignition as argued earlier, the data indicate that the charcoal rises beginning after 15,000 cal. yr BP at páramo and puna sites initiated a distinct Holocene fire regime, and likely marks the original period of human colonization of these regions.

A low fire incidence during the LGM in the Andes followed by heightened levels throughout most of the Holocene finds parallels across a variety of Neotropical habitats, as well as austral South America (Haberle and Ledru, 2001; Mayle et al., 2009; Moreno et al., 2010). Human intervention after 14,000–13,000 cal. yr BP is increasingly suggested as the cause for this increase (Di Pasquale et al., 2010; Pinter et al., 2011). Greater biomass, periodically drier climates, higher atmospheric levels of carbon dioxide and rising temperatures beginning in the early Holocene would certainly have facilitated burning (Power et al., 2008, 2010), but in the absence of significant natural ignition sources in Neotropical páramos, the necessary variable in manifesting this potential was Homo sapiens (Figure 2).

Figure 2.

Cubillín Lake and grass páramo, Ecuador, at 3700 m elevation, with column of smoke from a fire started by sheep herders.

To explain an early and then persistent administration of fire, we can envision a transient first wave of humans, limited in numbers and currently invisible to us except for their charcoal, moving through the foreland vegetation and using fire as a universal conditioner of their landscape, in the manner of their forbears. By the early Holocene, with the establishment of well-developed and familiar-looking grass páramo, articulated subsistence systems appeared. Territorial bands in larger numbers gained the capacity to sustain their landscape, in the process seeding the soil with charcoal. If this scenario describes the terminal Pleistocene fire regime, the archaeological record should contain evidence for human settlement coincident with the period of fire intensification. This data set is now examined.

Archaeology of early peopling

Archaeologists presently concur that the earliest peopling of South America occurred before 13,000 cal. yr BP (Bruhns, 1994; Cooke, 1998; Faught, 2008; Fiedel, 1999; Gruhn and Bryan, 1991), and many remain open to dates as early as 15,000–20,000 cal. yr BP (Bryan and Gruhn, 2003; Guidon et al., 1996; Lavallée, 2000). Archaeological and recent genetic evidence appear to converge on an original colonization of the Americas by 18,000–15,000 cal. yr BP (Goebel et al., 2008; Pitblado, 2011). At least ten sites in South America have been identified that were first occupied before 12,000 cal. yr BP, and over 30 sites between 12,000 and 10,000 cal. yr BP. A large number of additional sites from this period are now being investigated, many in tropical forest (Borrero, 1996; Bruhns, 1994; Cardale, 2005; López et al., 2003; Miotti and Salemme, 2003). These pre-agricultural populations inhabited a vast array of coastal, riverine, grassland and forest niches by means of locally adapted subsistence systems (Aceituno, 2000, 2002; Aldenderfer, 2008; Bryan, 1973; Dillehay, 2008; Gnecco, 1999; Stothart et al., 2003). Among the earliest sites is Monte Verde, Chile, reliably dated to 14,600 cal. yr BP (Dillehay, 1997; Dillehay and Collins, 1988; Goebel et al., 2008; Meltzer et al., 1997). Although a broad occupation of the Americas by the terminal Pleistocene is now well established, controversy will continue regarding sites with dates prior to 15,000 cal. yr BP, and the interpretation of genetic evidence used to develop a chronology of early diffusion (Bryan and Gruhn, 2003; Gruhn and Bryan, 1991; Pitblado, 2011; Roosevelt et al., 2002).

The earliest peopling of the high Andes provides a beginning point for anthropic inputs to páramo ecology. Were these inputs to occur after páramo had become established, humans would be regarded as disturbance factors. This is the consensus view among zonal advocates. But if humans were present at or before the period of deglaciation then páramo vegetation and human populations in some manner developed together, and the nature of this relationship needs to be elucidated. To this end Table 3 summarizes northern Andean archaeological sites at the Pleistocene/Holocene boundary from Perú, Ecuador and Colombia, and a sampling of sites of this period from the central Andes. I include puna sites because they form part of the same process of occupation as páramos of the northern Andes, and provide insights into hunter-gatherer subsistence (Bruhns, 1994; Church, 1996; Cooke, 1998; Hurt et al., 1972; Ijzereef, 1978; Lavallée, 2000; Miller and Gill, 1990; Salazar, 1984; Van der Hammen and Correal, 1978). All except two sites are rock shelters, reflecting their utility to hunter-gatherers and the generally dry conditions that permitted the preservation of cultural remains.

Table 3.

Terminal Pleistocene/early Holocene hunter-gatherer sites in or proximal to páramo and puna.^a

Site name	Altitude (m)	Date(s) of occupation (cal. yr BP)	Site type	Associated faunal and plant remains
Sabana de Bogotá/Colombia
Tibitó	2600	11,700	Rockshelter	Extinct horse and mastodon, fox, deer, birds
El Abra	2570	12,400	Rockshelter	White tailed deer and cavia; also armadillo, rabbit, peccary, agouti, birds
Sueva	2600	11,000– 10,000	Rockshelter	Modern fauna
Nemocón	2600	11,000	Rockshelter	Modern fauna
Tequendama	2600	10,920	Rockshelter	White tailed and brocket deer; also cavia, armadillo and canids
Highlands/Ecuador
El Inga (Ilaló, northern EC)	2549	11,000–9000	Open-air camp and obsidian workshop	Mastodons; white tailed and brocket deer, pudu, bear, tapir and small rodents
Chobshi (central EC)	2400	10,010–7535	Rockshelter and tool workshop	White tailed deer, pudu, brocket deer, rabbit, paca, guinea pig, opossum, tapir, Andean bear, porcupine and tinamou; no plant remains recovered due to damp conditions
Cubilán (southern EC)	2800–3100	10,300–9100	Open-air camp and flint workshop	No associated faunal or plant remains recovered
Marañón-Huallaga Divide/Perú
Manachaqui Cave	3625	12,033	Rockshelter	No remains reported from early horizon
Callejón de Huaylas-Cordillera Negra/Perú
Guitarrero Cave	2580	12,500– 11,000	Rockshelter	Camelids, cervids, viscacha, rabbits and skunk; tubers and rhizomes, beans, Capsicum, Cucurbita, Solanum hispidum and Pouteria lucuma; remains suggest vertical mobility (access to >4000)
Huánuco Plain/Perú
Lauricocha	4100	9500	Rockshelter	Mainly deer (various species) and camelids; seasonal exploitation
Junin Plain/Perú
Pachamachay Panaulauca Uchcumachay	4050–4420	10,500–9000	Rockshelter	Extinct horses and deer for the earliest periods at Uchcumachay, then hunting of a more diversified fauna; specialized vicuña hunting; gathering focused on wetland sites
Ayacucho/Perú
Pikimachay	2350	14,000– 13,000	Rockshelter	Armadillo, sloth, horse, deer

Notes: ^aSources for Lateglacial and early-Holocene sites include summary archaeological reviews (Bruhns, 1994; Cooke, 1998; Dillehay, 2008; Dillehay et al., 2004; Lavallée, 2000; Roosevelt et al., 2002; Scheinsohn, 2003) and specific site studies: Sabana de Bogotá (Correal, 1990; Hurt et al., 1972; Ijzereef, 1978; Van der Hammen and Correal, 1978); Ecuador (Burger et al., 1994; Lynch, 1989; Lynch and Pollock, 1981; Mayer-Oakes, 1986; Salazar, 1984); Manachaqui (Church, 1996); Guitarrero (Lynch, 1980, Smith, 1980, Wing, 1980); Lauricocha (Cardich, 1976); Panaulauca, Pachamachay, Uchcumachay (Rick, 1980; Wheeler Pires-Ferreira et al., 1976); Pikimachay (MacNeish and Vierra, 1983).

Current discussions of the human colonization of páramo routinely assume a temporary and limited use of páramo resources, often combined with an original occupation in the middle to late Holocene (Bakker et al., 2008; Brunschön and Behling, 2009; Ellenberg, 1979; Lægaard, 1992; Moscol Olivera and Hooghiemstra, 2010; Niemann and Behling, 2010). Often this view is based on the assumption that the earliest ceramics or other manifestations of Andean agricultural societies (such as pollen of disturbance indicators or plant domesticates) represent the earliest human occupation (Hansen and Rodbell, 1995; Niemann et al., 2009). But Table 3 reveals that humans had become settled in puna and at the lower edge of páramo by the terminal Pleistocene. These data show that the high Andes participated fully in the earliest peopling of South America, suggesting that páramos were productive habitats (Bruhns, 2003; Lavallée, 2000; Salazar, 1985; Weng et al., 2006). The intensity of this participation remains to be clarified through the discovery of more sites and other evidence of occupation, but also by calculating the likely human carrying capacity of páramo, and thus its subsistence attractiveness. High-altitude sites would have seen less use in the terminal Pleistocene than in the early Holocene because of the incipient nature of higher-elevation vegetation after glacial retreat, and thus its limited herbivore population and plant food resources (Butzer, 1988; Dillehay et al., 2004; Lanata and Borrero, 1999; Lynch, 1971; Smith, 1980). A similar situation unfolded at the same time along South American coasts as consumable ocean resources developed along rising shorelines. Given that sites at 4000 m in the central Andes and those above 2500 m in the northern Andes were utilized by 12,000 cal. yr BP, it appears that the vegetation between UFLs and retreating ice developed relatively quickly, and was of immediate interest as a subsistence resource.

The limited number of early sites currently described at páramo and near-páramo elevations reflects incipient human populations, but it may also reflect a variety of biases of the record. One is the restricted material culture of hunter-gatherers and the absence of pottery, ceremonial sites, terraces, a formal road network, and settlements to mark their presence in the manner of later agricultural societies. A second bias is the limited number of rock shelters in many páramos, especially as compared with puna. Hunter-gatherer open-air shelters constructed of degradable materials would only leave identifiable remains if fortuitously preserved, as in the bog conditions of Monte Verde (Bocek and Rick, 1984; Dillehay, 1989, 1997; MacNeish and Vierra, 1983). A third bias may be communicational: Some northern Andean hunter-gatherer sites are reported but have not been formally excavated, or are described in national anthropology publications and conference proceedings but have not yet found a wider audience and scrutiny through international peer-reviewed journals (e.g. Salazar, 1985; Weng et al., 2006). Finally, the geographic focus of some active researchers, as would be the case of Correal and Van der Hammen in the Sabana de Bogotá, may produce a cluster of sites. This concentration, combined with an archaeological paradigm disinclined to admit significant early human populations, may generate an erroneous sense of absence elsewhere (Gnecco and Salgado, 1989).

There is no doubt, as well, that early occupants of the northern Andes were mobile and thus neighboring páramos could be exploited for their food resources using base camps at the forest edge or below (Bruhns, 1994; Salazar, 1984, 1985). Vertical mobility is the rule where a dissected topography allows direct access to a variety of elevations (Figure 3). In the pre-agricultural Holocene, as now, rounds that involve walking from forested elevations to páramo and back, even in the course of a day, may have been the norm. Landscape-scale effects of human subsistence activities, especially when fire is a tool, do not require a large or stable resident population.

Figure 3.

Because they are islands, many grass páramos adjoin upper montane forests and would have permitted hunter-gatherer access to resources from both habitats.

If these observations are grounded, the eventual identification of additional early hunter-gatherer sites in the northern Andes is likely, some in rock shelters unexamined until now and others in open páramo tied to fortuitously encountered hearths, kill sites, tool-making workshops, garbage pits, and bones and teeth from burials. The first step in this endeavor is to accept the prospect of finding additional signs of early hunter-gatherer occupation and evidence of their impacts. By the very nature of their subsistence and the size of their populations, our inquiry should complement the identification and study of sites and artifacts with that of cultural landscapes, using an expanding array of proxies for vegetation, disturbance, diet, material culture and land use change. By placing páramo hunter-gatherers in the larger context of the Late Pleistocene peopling of essentially all productive environments in South America, a growing array of biological and geochemical proxies that do not require cave settings can be applied to environmental reconstruction.

Pollen analysis will remain of great importance, especially as it becomes more rigorous (Seppä and Bennett, 2003). But the analysis of pollen is now being complemented by the use of fungal spores, diatoms, starch grains, phytoliths, coprolites, soil biomarkers, and other geochemistry indicators (Eckmeier and Wiesenberg, 2009; Jansen et al., 2008, 2010; Montoya et al., 2010; Nierop and Jansen, 2009; Piperno, 2006). These new techniques will supplement and also perhaps challenge the vegetation and land-use reconstructions based to date on pollen and a limited array of non-pollen polymorphs. Many new techniques still need to be calibrated and data bases created of reference specimens and geochemical values. Soil charcoal is one of these landscape markers already in use, and we have seen its value in illuminating páramo origins.

The logic of hunter-gatherer fire

Humans were witnesses of deglaciation in the northern Andes, and likely utilized the foreland habitat below the ice front. The substrate remaining after ice retreated consisted of rock, regolith and sorted materials such as gravel and sand. A mixture of hardy woody and herbaceous species, whose growth was limited initially by unstable inorganic substrates, cold temperatures and freeze-thaw cycles, colonized the newly exposed lands. The species that constituted the incipient vegetation would have come from both the forest below and from the herbaceous zone between the forest edge and the glacial ice. Over the final millennia of the Pleistocene the mixed vegetation established after ice retreat achieved growing complexity and mass, in tandem with developing soils. Were the early inhabitants passive in their hunting and gathering, accepting the animal and plant foods made available by the climatic shift? Or were they instead inclined to intervene in their environment to make it more diverse and productive, and to enlarge it?

Given the collective ability of hunter-gatherers throughout the world to both adapt to and transform landscapes, it seems likely that páramo hunter-gatherers would have actively manipulated their environment, especially through the use of fire. Such is the case for other open-vegetation hunter-gatherer groups in Asia, Africa and Australia, as in Neotropical lowland settings (Aceituno and Loaiza, 2007; Bird et al., 2008; Gnecco, 1998; Gould, 1985; Kirkpatrick, 1994; O’Connell and Allen, 1995; Piperno et al., 1990; Pyne, 1993; Rull, 2009; Wharton, 1968). Prescribed, moderate-interval burning in páramo would have generated higher plant productivity and herbaceous diversity, resulting in increased graze and thus grazers. Within a fire regime and warming climates, Late-Pleistocene foreland vegetation was progressively transformed to grass páramo. Fire would also have increased the predictability of resources because successional development could be anticipated. Availability is synchronized, for example, when a burn stimulates bunchgrasses to flower and seed as a cohort (Mills, 1986).

As fire initiators, humans may choose the moment and the place to burn that will provide the greatest spread for each ignition point. If for a wet páramo only 20 or 30 days in the year are dry enough for grass to carry fire, humans can target those days with great precision. This is the case now for contemporary páramo cattle and sheep ranchers, as it certainly would have been for hunter-gatherers whose survival depended upon adequate grass and successional stage management. Hunter-gatherers would have known to wait until the morning dew dries, and the daily winds pick up, before starting fires. They would have known to choose adequate points of ignition upwind and at the bottom of slopes, and in the taller grasses. They only needed to carry a smoldering fire stick that would break into flames with a bit of blowing. Easily in a day, and with an expense of energy only slightly greater than that required to walk, a single hunter-gatherer could set fires at multiple points and burn hundreds of hectares over many discrete fire surfaces.

Terminal Pleistocene inhabitants of páramo forelands appear to have hunted species of megafauna until they became locally extinct, by about 11,000 cal. yr BP (Bruhns, 1994; Cooke, 1998; Correal, 1990; Wheeler Pires-Ferreira et al., 1976). The mega-herbivores occupied primarily open habitats, to which hunters were drawn. Nonetheless, the plant and animal subsistence base likely was broad from the beginning of occupation. Once vegetation became established after ice retreated, hunting centered on camelids and cervids in the central Andes, while in the páramos of the northern Andes the white tailed deer, Odocoileus peruvianus, provided most of the animal protein (Ijzereef, 1978; Lynch and Pollock, 1981; Miller and Gill, 1990; Van der Hammen and Correal, 1978; Wing, 1986). White-tailed deer combine browsing and grazing habits and are unusual in their acceptance of a largely grass diet. The universal dependence of early páramo hunter-gatherers on white-tailed deer confirms the herbaceous nature of the environments in which hunters engaged their game. Other important hunted species include the rabbit Sylvilagus brasiliensis and the wild guinea pig Cavia aperea (Stahl, 2008), both requiring grasslands with forest edges. Montane forest deer (Mazama sp. and Pudu mephistophiles), paca (Agouti sp.), armadillo (Dasypus novemcinctus), birds and snails offered additional animal protein (Ijzereef, 1978; Lynch and Pollock, 1981; Miller and Gill, 1990; Wing, 1986). A variety of lithics, including chert, basalt, quartz, and obsidian, plus wood, bone and horn provided material for projectile points and butchering tools. Obsidian was exploited at specific locations and sometimes exchanged over considerable distances in the northern Andes, especially after the mid Holocene (Burger et al., 1994; Van der Hammen and Correal, 1978).

Three changes in the local environment would have permitted the herbivore population, and thus the human carrying capacity, to expand: An increase in the grass component of the open vegetation, a periodic renewal of tussock grasses to maintain their nutritional value, and an enlarged area of grassland vegetation. It is apparent that the periodic application of fire to the vegetation could achieve these three aims simultaneously. Anthropic fire in páramo was neither gratuitous nor fortuitous (Niemann and Behling, 2010), but rather supremely purposeful (Mills, 1986; Price and Bowman, 1994).

The strategy of burning, however systematically applied, would nonetheless have left certain areas free of fire because of irregular terrain and natural firebreaks. These protected sites could develop into forest. Early páramo inhabitants would have welcomed forest patches because they provided browse and protection for guinea pig, deer and other species that prosper in edge habitats, while also providing humans a source of firewood, materials for shelter, medicines and plant foods (Ojasti, 1983).

Although fire helped support a robust herbivore population, perhaps more importantly it allowed the maintenance of a diverse and open community of gatherable foods (Bonzani, 1997). For this reason Andean hunter-gatherers are considered primarily broad-spectrum foragers (Bruhns, 1994; Bryan, 1986; Dillehay, 2008). By burning páramo they created a mosaic of different ages of plant succession, and each successional stage provided a singular array of food resources. The first stage, beginning 6 to 8 months after a fire, produced vast quantities of Calamagrostis and Festuca seeds, similar to wheat and barley in nutritional content and perhaps utilized as food (Pearsall, 1996). Subsequent stages would have included rhizomes and roots, seeds, berries, leaves, flowers and fruits (Pearsall, 1980; Rick, 1980). To illustrate, the excellent preservation conditions of Monte Verde and an acute interest by researchers in reconstructing the diet of its inhabitants permitted the identification of 60 utilized plant species (Dillehay and Rossen, 2002).

The inventory of plant foods in the early páramo diet is poorly known, in part because the archaeological yield of plant remains is marginal in damp environments. But contemporary herders raised in páramo still have considerable knowledge of edible plants (Cardich, 1976; Pérez, 1998). The species they report are virtually all sun-loving. This finding would be consistent with other heliophytes that dominate the human inventory of cereal and vegetable domesticates. It parallels as well the very early hunter-gatherer strategy in Neotropical forests, where clearings occupied by pioneering species generated gatherable foods, some of which eventually were domesticated (Gnecco, 1998; Piperno et al., 1990). Although as yet little studied, I do not share the view that grass páramo is an autosuccession (Llambí et al., 2003; Sarmiento et al., 2003). Rather, sustained observation of a burned páramo shows site-specific development involving major transformations in plant species composition and structure, over timescales of decades. In sum the mosaic of different-aged successional stages holds great species diversity, which hunter-gatherers could then exploit through mobility. Thus, availability drove mobility rather than scarcity or seasonality as in higher latitudes (Cardich, 1976; Mills, 1986; Salazar, 1984). The successional mosaic would have also created subsistence stability, as has been suggested for hunter-gatherers in the puna (Pearsall, 1983).

Conclusion

Three fundamental observations follow from this discussion:

Pleistocene glacial cycles and vegetation movements inform our understanding of the natural constraints on contemporary páramos, but of themselves are inadequate in explaining the origin of tussock-dominated grass páramo. Although Pleistocene climate oscillations drove the movement of the UFL and the character of the herbaceous vegetation, with human settlement we can no longer argue that the current UFL and páramo plant associations represent simply another interglacial. Because the earliest humans occupying the tropical Andes applied fire to the pioneering vegetation, even in very wet locations, the view must be abandoned that hunter-gatherers were absent or of minor ecological significance, and therefore that the pre-human and human ecologies were similar. Tussock-dominated grass páramos could not exist at their current elevations without periodic fire, and this fire was and remains anthropic.

Humans as ecological factors require precise historical grounding. Food procurement and production link humans to their environment and until modern times generated most anthropic impacts. Because subsistence systems in the high northern Andes have evolved significantly since the first peopling ≥ 13,000 yr BP, the nature and degree of impact will be specific to each stage in this chronology. The primary distinction is between hunter-gatherer and agricultural societies. The former occupied páramo elevations during the period of grass páramo establishment, 8000–10,000 years before the local advent of agriculture. The land use chronology and the archaeological record do not therefore support scenarios of páramo origins that posit a role for agriculturalists, neither as the original páramo inhabitants (Bakker et al., 2008; Brunschön and Behling, 2009; Moscol Olivera and Hooghiemstra, 2010; Niemann and Behling, 2010; Niemann et al., 2009), nor as agents of deforestation and replacement (Ellenberg, 1979; Lægaard, 1992; Sarmiento, 2002). Sedimentary charcoal evidence indicates, as well, that the first instance of anthropic fire in páramo does not date from the advent of agriculture (Hansen and Rodbell, 1995; Paduano et al., 2003; Wille et al., 2002), but rather began in the Late Pleistocene. Linking grass páramo development to subsistence systems also brings into focus the danger of applying contemporary observations of human impacts to those made by hunter-gatherers. Intense current páramo fire regimes and deforestation at the UFL, which have been used to suggest that a similar process caused deforestation and replacement earlier in the Holocene, ignores the uniqueness of modern residence patterns, population densities, tools, command of energy, road network and market links. In effect, both the early agriculturalist and contemporary human relationships to grass páramo fail as proxies for the hunter-gatherer relationship.

Rather than being framed as either primary or secondary vegetation, grass páramo is best understood as a plant community that co-evolved with humans. Vegetation reconstructions commonly frame the community in terms of primary or secondary, the former being original and pre-human, and the latter a replacement, often following human disturbance. This standard choice is a valuable instrument, but it also deeply conditions our thinking and has certainly defined the debate on páramo etiology over the last half century. Following Ellenberg (1979) and Lægaard (1992), this debate centered upon the proposal that grass páramo was largely a secondary formation following human removal of a zonal forest. In response, advocates of a natural grass páramo generated ample pollen evidence that almost all existing páramo was never forested, allowing only that grass páramo is a secondary form near a depressed current UFL, where humans are assumed to have eliminated montane forest (Moscol Olivera and Cleef, 2009b; Moscol Olivera and Hooghiemstra, 2010). This explanation attempts to reconcile the zonal and replacement hypotheses by assigning each a location on the altitudinal gradient. In doing so the primary/secondary paradigm is once again brought to bear. I argue that this paradigm has inhibited our ability to adequately reconstruct the vegetation history of grass páramo. If hunter-gatherer fire created grassland from a pioneering mix of vegetation beginning in the terminal Pleistocene, it would be more fruitful to think of humans as biotic factors – part of the original páramo ecology and necessary to its function. From this perspective a hunter-gatherer fire regime ceases to be an external disturbance factor. Specifically, although the variable fire is abiotic when generated by lighting or volcanism it becomes biotic when initiated by early humans, just as we might classify shade as abiotic when caused by tall canyon walls but biotic when caused by the leaves of competing plants. It follows that a tussock grass páramo should not exclude but rather embrace its hunter-gatherers.

The evidence indicates that hunter-gatherer activities had a decisive bearing on grass páramo origins. In effect, long before the advent of domesticated plants and animals, hunter-gatherers had domesticated a landscape. The pollen, sedimentary charcoal and archaeological data sets are culturally and temporally coherent. I suggest that humans be regarded as a keystone species in grass páramo ecology, and anthropic fire as a fundamental ecological factor.

Footnotes

Acknowledgements

I gratefully acknowledge my affiliations with the Departments of Geography at the University of Vermont (Burlington, Vermont) and the University of New Mexico (Albuquerque, New Mexico) during the library research for this topic and preparation of the manuscript. At the Mazar Wildlife Reserve in Ecuador, colleagues and students from Round River Conservation Studies and the Fundación Cordillera Tropical helped to hone the arguments presented here. Particular appreciation, extending over a decade and more, goes to William Denevan, Daniel Gade, Matt Himley, Catherine Schloegel, Alfredo Martínez and Sven Wunder.

Funding

This work was partially supported with a writing grant from the Center for International Forestry Research-CIFOR (Bogor, Indonesia) .

Notes

References

Aceituno

(2000) Una propuesta para el estudio de las sociedades arcáicas de la cordillera centro-occidental colombiana. Boletín de Antropología 14: 154–182 (Medellín, Colombia: Universidad de Antioquia).

Aceituno

(2002) Interacciones fitoculturales en el Cauca medio durante el Holoceno temprano y medio. Arqueología del Área Intermedia 4: 89–114.

Aceituno

Loaiza

(2007) Domesticación del bosque en el Cauca medio Colombiano entre el Pleistoceno final y el Holoceno medio. BAR International Series 1654, 139 pp.

Aldenderfer

(2008) High elevation foraging societies. In: Silverman

Isbell

(eds) Handbook of South American Archaeology. New York: Springer, pp. 131–143.

Bader

Ruijten

(2008) A topography-based model of forest cover at the alpine tree line in the tropical Andes. Journal of Biogeography 35(4): 711–723.

Bakker

Salomons

(1989) A palaeoecological record of a volcanic soil sequence in the Nevado del Ruiz Area, Colombia. Review of Palaeobotany and Palynology 60(1–2): 149–163.

Bakker

Moscol Olivera

Hooghiemstra

(2008) Holocene environmental change at the upper forest line in northern Ecuador. The Holocene 18(6): 877–893.

Barba

Robin

Samaniego

. (2008) Holocene recurrent explosive activity at Chimborazo volcano (Ecuador). Journal of Volcanology and Geothermal Research 176(1): 27–35.

Baruch

(1984) Ordination and classification of vegetation along an altitudinal gradient in the Venezuelan páramos. Vegetatio 55(2): 115–126.

10.

Beate

Monzier

Spikings

. (2001) Mio–Pliocene adakite generation related to flat subduction in southern Ecuador: The Quimsacocha volcanic center. Earth and Planetary Science Letters 192(4): 561–570.

11.

Behling

(2000) A 2860-year high-resolution pollen and charcoal record from the Cordillera de Talamanca in Panama: A history of human and volcanic forest disturbance. The Holocene 10(3): 387–393.

12.

Bird

Codding

. (2008) The ‘fire stick farming’ hypothesis: Australian Aboriginal foraging strategies, biodiversity, and anthropogenic fire mosaics. Proceedings of the National Academy of Sciences of the United States 105: 14,796–14,801.

13.

Bocek

Rick

(1984) La época precerámica en la puna de Junín: Investigaciones en la zona de Panaulauca. Revista Chungará 13: 109–127. Tarapacá, Chile: Universidad de Tarapacá.

14.

Bonzani

(1997) Plant diversity in the archaeological record: A means toward defining hunter-gatherer mobility. Journal of Archaeological Science 24(12): 1129–1139.

15.

Borrero

(1996) The Pleistocene–Holocene transition in southern South America. In: Straus

Eriksen

Erlandson

. (eds) Humans at the End of the Ice Age. New York: Plenum Press, pp. 339–354.

16.

Brandbyge

(1992) Planting of local woody species in the páramo. In: Balslev

Luteyn

(eds) Páramo: An Andean Ecosystem Under Human Influence. London: Academic Press, pp. 265–274.

17.

Bruhns

(1994) The first peoples: 12 000–6000 BC. In: Bruhns

(ed.) Ancient South America. Cambridge: Cambridge University Press, pp. 43–64.

18.

Bruhns

(2003) Social and cultural development in the Ecuadorian highlands and eastern lowlands during the Formative. In: Raymond

Burger

(eds) Archaeology of Formative Ecuador. Washington DC: Dumbarton Oaks Research Library and Collection, pp. 125–174.

19.

Brunschön

Behling

(2009) Late Quaternary vegetation, fire and climate history reconstructed from two cores at Cerro Toledo, Podocarpus National Park, southeastern Ecuadorian Andes. Quaternary Research 72(3): 388–399.

20.

Bryan

(1973) Paleoenvironments and cultural diversity in Late Pleistocene South America. Quaternary Research 3(2): 237–256.

21.

Bryan

(1986) Paleoamerican prehistory as seen from South America. In: Bryan

(ed.) New Evidence for the Pleistocene Peopling of the Americas. Orono Maine: Center for the Study of Early Man, University of Maine, pp. 1–14.

22.

Bryan

Gruhn

(2003) Some difficulties in modeling the original peopling of the Americas. Quaternary International 109–110: 175–179.

23.

Burger

Asaro

Michel

. (1994) An initial consideration of obsidian procurement and exchange in prehispanic Ecuador. Latin American Antiquity 5(3): 228–255.

24.

Bush

Hansen

Rodbell

. (2005) A 17 000-year history of Andean climate and vegetation change from Laguna de Chochos, Peru. Journal of Quaternary Science 20(7–8): 703–714.

25.

Butzer

(1988) A ‘marginality’ model to explain major spatial and temporal gaps in the Old and New World Pleistocene settlement records. Geoarchaeology 3(3): 193–203.

26.

Cardale

(2005) Calima and Malagana: Art and Archaeology in Southwestern Colombia. Bogota: Pro Calima Foundation, 294 pp.

27.

Cardich

(1976) Vegetales y recolecta en Lauricocha: Algunas inferencias sobre asentamientos y subsistencia preagrícolas en los Andes centrales. Relaciones de la Sociedad Argentina de Antropología 10: 27–41.

28.

Christensen

(1993) Fire regimes and ecosystem dynamics. In: Crutzen

Goldammer

(eds) Fire in the Environment: The Ecological, Atmospheric, and Climatic Importance of Vegetation Fires. New York: John Wiley and Sons, pp. 233–244.

29.

Church

(1996) Prehistoric cultural development and interregional interaction in the tropical montane forests of Peru, Volume 1. PhD Thesis, Yale University.

30.

Cierjacks

Salgado

Wesche

. (2008) Post-fire population dynamics of two tree species in high-altitude Polylepis forests of central Ecuador. Biotropica 40(2): 176–182.

31.

Clapperton

(1993a) Nature of environmental changes in South America at the Last Glacial Maximum. Palaeogeography, Palaeoclimatolgy, Palaeoecology 101(3–4): 189–208.

32.

Clapperton

(1993b) Late Cenozoic glacial history of the Andes, Parts III and IV. In: Clapperton

(ed.) Quaternary Geology and Geomorphology of South America. Amsterdam: Elsevier, pp. 395–471.

33.

Clark

Dyke

Shakun

. (2009) The Last Glacial Maximum. Science 325(5941): 710–714.

34.

Cleef

(1981) The vegetation of the páramos of the Colombian Cordillera Oriental. PhD Thesis, Universiteit van Amsterdam.

35.

Colinvaux

Bush

Steinitz-Kannan

. (1997) Glacial and postglacial pollen records from the Ecuadorian Andes and Amazon. Quaternary Research 48(1): 69–78.

36.

Cooke

(1998) Human settlement of Central America and northernmost South America (14 000–8000 BP). Quaternary International 49–50: 177–190.

37.

Correal

(1990) Evidencias culturales durante el Pleistoceno y Holoceno de Colombia. Revista de Arqueología Americana 1: 69–89.

38.

Cuatrecasas

(1968) Páramo vegetation and its life forms. In: Troll

(ed.) Geo-ecology of the Mountainous Regions of the Tropical Americas. Bonn: Ferd. Dümmlers Verlag, pp. 163–186.

39.

Denevan

(ed.) (1992) The Native Population of the Americas in 1492. Madison WI: University of Wisconsin Press, 353 pp.

40.

Dillehay

(ed.) (1989) Monte Verde: A Late Pleistocene settlement in Chile, Vol. 1. Washington DC: Smithsonian Institution.

41.

Dillehay

(ed.) (1997) Monte Verde: A Late Pleistocene settlement in Chile, Vol. 2: The Archaeological Context and Interpretation. Washington DC: Smithsonian Institution.

42.

Dillehay

(2008) Profiles in Pleistocene history. In: Silverman

Isbell

(eds) Handbook of South American Archaeology. New York: Springer, pp. 29–43.

43.

Dillehay

Collins

(1988) Early cultural evidence from Monte Verde in Chile. Nature 332(6160): 150–152.

44.

Dillehay

Rossen

(2002) Plant food and the peopling of the New World. In: Jablonksi

(ed.) The First Americans: The Pleistocene Colonization of the New World. Memoirs of the California Academy of Sciences 27, pp. 237–253.

45.

Dillehay

Bonavia

Kaulicke

(2004) The first settlers. In: Silverman

(ed.) Andean Archaeology. Malden MA: Blackwell, pp. 16–34.

46.

Di Pasquale

Impagliazzo

Lubritto

. (2010) Soil charcoal analysis as a climato-stratigraphical tool: The key case of Cordillera Real, northern Andes. Nuclear Instruments and Methods in Physics Research B 268(7–8): 1088–1090.

47.

Di Pasquale

Marziano

Impagliazzo

. (2008) The Holocene treeline in the northern Andes (Ecuador): First evidence from soil charcoal. Palaeogeography, Palaeoclimatology, Palaeoecology 259(1): 17–34.

48.

Dull

Nevle

Woods

. (2010) The Columbian encounter and the Little Ice Age: Abrupt land use change, fire, and greenhouse forcing. Annals of the Association of American Geographers 100(4): 755–771.

49.

Eckmeier

Wiesenberg

(2009) Short-chain n-alkanes (C_16–20) in ancient soil are useful molecular markers for prehistoric biomass burning. Journal of Archaeological Science 36(7): 1590–1596.

50.

Ellenberg

(1979) Man’s influence on tropical mountain ecosystems in South America. Journal of Ecology 67(2): 401–416. Originally published as: Ellenberg

(1958) Wald oder Steppe? Die natürliche Pflanzendecke der Anden Perus. Umschau in Wissenschaft und Technik 21: 645–681.

51.

Fairbanks

Mortlock

Chiu

T-C

. (2005) Marine radiocarbon calibration curve spanning 0 to 50 000 years BP based on paired ²³⁰Th/²³⁴U/²³⁸U and ¹⁴C dates on pristine corals. Quaternary Science Reviews 24(16–17): 1781–1796.

52.

Faught

(2008) Archaeological roots of human diversity in the New World: A compilation of accurate and precise radiocarbon ages from earliest sites. American Antiquity 73(4): 670–698.

53.

Fiedel

(1999) Older than we thought: Implications of corrected dates for paleoindians. American Antiquity 64(1): 95–115.

54.

Gade

(1999) Nature and Culture in the Andes. Madison: University of Wisconsin Press, 287 pp.

55.

Gnecco

(1998) Paisajes antropogénicos en el Pleistoceno final y Holoceno temprano en Colombia. Revista de Antropología y Arqueología 10: 45–61.

56.

Gnecco

(1999) An archaeological perspective of the Pleistocene–Holocene boundary in northern South America. Quaternary International 53/54: 3–9.

57.

Gnecco

Salgado

(1989) Adaptaciones precerámicas en el suroccidente de Colombia. Boletín del Museo de Oro 24. Bogotá: Banco de la República. Available at: http://banrepcultural.org/blaavirtual/publicacionesbanrep/bolmuseo/1989/bol24/bod1.htm (accessed 13 October 2011).

58.

Goebel

Waters

O’Rourke

(2008) The late Pleistocene dispersal of modern humans in the Americas. Science 319(5869): 1497–1502.

59.

Gómez

Berrío

Hooghiemstra

. (2007) A Holocene pollen record of vegetation change and human impact from Pantano de Vargas, an intra-Andean basin of Duitama, Colombia. Review of Palaeobotany and Palynology 145(1–2): 143–157.

60.

González

Van der Hammen

Flint

(1965) Late Quaternary glacial and vegetational sequence in Valle de Lagunillas, Sierra Nevada del Cocuy, Colombia. Leidse Geologische Mededelingen 32: 157–182.

61.

Gould

(1985) ‘Now let’s invent agriculture…’: A critical review of concepts of complexity among hunter-gatherers. In: Price

Brown

(eds) Prehistoric Hunter-gatherers: The Emergence of Cultural Complexity. Orlando FL: Academic Press, pp. 427–442.

62.

Grabandt

(1980) Pollen rain in relation to arboreal vegetation in the Colombian Cordillera Oriental. Review of Palaeobotany and Palynology 29: 65–147.

63.

Grabandt

(1985) The use of data concerning the present relation between pollen rain and vegetation in the Colombian Cordillera Oriental in the interpretation of Andean pollen diagrams. PhD Thesis, part 2, Faculteit der Wiskunde en Natuurwetenschappen, Universiteit van Amsterdam.

64.

Grabandt

Nieuwland

(1985) Pollen rain in relation to páramo vegetation in the Colombian Cordillera Real. PhD Thesis, part 1, Faculteit der Wiskunde en Natuurwetenschappen, Universiteit van Amsterdam.

65.

Groot

Bogotá

Lourens

. (2011) Ultra-high resolution pollen record from the northern Andes reveals rapid shifts in montane climates within the last two glacial cycles. Climate of the Past 7(1): 299–316.

66.

Gruhn

Bryan

(1991) A review of Lynch’s descriptions of South American Pleistocene sites. American Antiquity 56(2): 342–348.

67.

Guidon

Pressis

A-M

Parenti

. (1996) Nature and age of the deposits in Pedra Furada, Brazil: Reply to Meltzer, Adovasio and Dillehay. Antiquity 70(268): 408–421.

68.

Haberle

Ledru

M-P

(2001) Correlations among charcoal records of fires from the past 1000 years in Indonesia, Papua New Guinea, and Central and South America. Quaternary Research 55(1): 97–104.

69.

Hall

Mothes

(2008) Volcanic impediments in the progressive development of pre-Columbian civilizations in the Ecuadorian Andes. Journal of Volcanology and Geothermal Research 176(3): 344–355.

70.

Hall

Samaniego

Le Pennec

. (2008) Ecuadorian Andes volcanism: A review of late Pliocene to present activity. Journal of Volcanology and Geothermal Research 176(1): 1–6.

71.

Hanselman

Bush

Gosling

. (2011) A 370 000-year record of vegetation and fire history around Lake Titicaca (Bolivia/Peru). Palaeogeography, Palaeoclimatolgy, Palaeoecology 305(1–4): 201–214.

72.

Hansen

(1995) A review of late-glacial pollen records from Ecuador and Peru with reference to the Younger Dryas event. Quaternary Science Reviews 14(9): 853–865.

73.

Hansen

Rodbell

(1995) A Late-glacial/Holocene pollen record from the eastern Andes of northern Peru. Quaternary Research 44(2): 216–227.

74.

Hansen

Rodbell

Seltzer

. (2003) Late-glacial and Holocene vegetational history from two sites in the western Cordillera of southwestern Ecuador. Palaeogeography, Palaeoclimatolgy, Palaeoecology 194(1–3): 79–108.

75.

Hansen

Seltzer

Wright

(1994) Late Quaternary vegetational change in the central Peruvian Andes. Palaeogeography, Palaeoclimatology, Palaeoecology 109(2–4): 263–285.

76.

Hartshorn

(1983) Plants: Introduction. In: Janzen

(ed.) Costa Rican Natural History. Chicago: University of Chicago, pp. 118–157.

77.

Heine

(2000) Tropical South America during the Last Glacial Maximum: Evidence from glacial, periglacial and fluvial records. Quaternary International 72: 7–21.

78.

Helmens

Kuhry

(1986) Middle and late Quaternary vegetational history and climatic history of the Páramo de Agua Blanca (Eastern Cordillera Colombia). Palaeogeography, Palaeoclimatolgy, Palaeoecology 56(3–4): 291–335.

79.

Helmens

Kuhry

Rutter

. (1996) Warming at 18,000 yr BP in the tropical Andes. Quaternary Research 45(3): 289–299.

80.

Hofstede

Chilito

Sandovals

(1995) Vegetative structure, microclimate, and leaf growth of a páramo tussock grass species, in undisturbed, burned and grazed conditions. Vegetatio 119(1): 53–65.

81.

Hofstede

Segarra

Mena

(2003) Los Páramos del Mundo. Quito: Proyecto Atlas Mundial de los Páramos. Global Peatland Initiative/NC-IUCN/Ecociencia, 299 pp.

82.

Holtmeier

F-K

(2003) Timberlines: Ecology, Patchiness and Dynamics. Advances in Global Change Research 36. Berlin: Springer, 369 pp.

83.

Hooghiemstra

(1984) Vegetational and climatic history of the high plain of Bogota, Colombia: A continuous record of the last 3.5 million years. PhD Thesis, Universiteit van Amsterdam.

84.

Hooghiemstra

Cleef

(1995) Pleistocene climatic change and environmental and generic dynamics in the north Andean montane forest and páramo. In: Churchill

(ed.) Biodiversity and Conservation of Neotropical Montane Forests. New York: NY Botanical Garden, pp. 35–49.

85.

Hooghiemstra

Van der Hammen

(2004) Quaternary ice-age dynamics in the Colombian Andes: Developing an understanding of our legacy. Philosophical Transactions of the Royal Society B 359(1442): 173–181.

86.

Hooghiemstra

Wijninga

Cleef

(2006) The paleobotanical record of Colombia: Implications for biogeography and biodiversity. Annals of the Missouri Botanical Garden 93(2): 297–324.

87.

Horn

(1986) Fire and páramo vegetation in the Cordillera de Talamanca, Costa Rica. PhD Thesis, University of California-Berkeley.

88.

Horn

(1989a) Prehistoric fires in the Chirripó highlands of Costa Rica: Sedimentary charcoal evidence. Revista de Biología Tropical 37(2): 139–148.

89.

Horn

(1989b) Postfire vegetation development in the Costa Rican páramos. Madroño 36(2): 93–114.

90.

Horn

(1993) Postglacial vegetation and fire history in the Chirripó Páramo of Costa Rica. Quaternary Research 40(1): 107–116.

91.

Horn

(1998) Fire management and natural landscapes. In: Zimmerer

Young

(eds) Nature’s Geography: New Lessons for Conservation in Developing Countries. Madison: University of Wisconsin Press, pp. 125–146.

92.

Horn

Kappelle

(2009) Fire in the páramo ecosystems of Central and South America. In: Cochrane

(ed.) Tropical Fire Ecology. Chichester: Springer Praxis Books, pp. 505–539.

93.

Horn

League

(2005) Registros de sedimentos lacustres de la vegetación del Holoceno e historia del fuego en el páramo de Costa Rica. In: Kappelle

Horn

(eds) Páramos de Costa Rica. Heredia: Instituto Nacional de Biodiversidad [INBio], pp. 253–273.

94.

Horn

Sanford

(1992) Holocene fires in Costa Rica. Biotropica 24(3): 354–361.

95.

Hurt

Van der Hammen

Correal

(1972) Preceramic sequences in the El Abra rock-shelters, Colombia. Science 175(4026): 1106–1108.

96.

Ijzereef

(1978) Faunal remains from the El Abra rock shelters (Colombia). Palaeogeography, Palaeoclimatology, Palaeoecology 25(1–2): 163–177.

97.

Islebe

Hooghiemstra

(1997) Vegetation and climate history of montane Costa Rica since the last Glacial. Quaternary Science Reviews 16(6): 589–604.

98.

Islebe

Hooghiemstra

van ‘t Veer

(1996) Holocene vegetation and water level history in two bogs of the Cordillera de Talamanca, Costa Rica. Vegetatio 124(2): 155–171.

99.

Jansen

Huassmann

Tonneijck

. (2008) Characteristic straight-chain lipid ratios as a quick method to assess past forest-páramo transitions in the Ecuadorian Andes. Palaeogeography, Palaeoclimatology, Palaeoecology 262(3–4): 129–139.

100.

Jansen

Van Loon

Hooghiemstra

. (2010) Improved reconstruction of palaeo-environments through unraveling of preserved vegetation biomarker patterns. Palaeogeography, Palaeoclimatology, Palaeoecology 285(1–2): 119–130.

101.

Jantz

Behling

(2012) A Holocene environmental record reflecting vegetation, climate, and fire variability at the Páramo of Quimsacocha, southwestern Ecuadorian Andes. Vegetation History and Archaeobotany 21(3): 169–185.

102.

Jokisch

Lair

(2003) One last stand? Forests and change on Ecuador’s eastern cordillera. Geographical Review 92(2): 235–256.

103.

Keating

(2000) Chronically disturbed páramo vegetation at a site in southern Ecuador. Journal of the Torrey Botanical Society 127(2): 162–171.

104.

Keating

(2007) Fire ecology and conservation in the high tropical Andes: Observations from northern Ecuador. Journal of Latin American Geography 6(1): 43–62.

105.

Kirkpatrick

(1994) A Continent Transformed: Human Impact on the Natural Vegetation of Australia. Melbourne: Oxford University Press.

106.

Körner

(1998) A re-assessment of high elevation treeline positions and their explanation. Oecologia 115(4): 445–459.

107.

Kuhry

(1988a) A paleobotanical and palynological study of Holocene peat from the El Bosque mire, located in a volcanic area of the Cordillera Central of Colombia. Review of Palaeobotany and Palynology 55(1–3): 19–72.

108.

Kuhry

(1988b) Palaeobotanical-palaeoecological studies of tropical high Andean peatbog sections (Cordillera Oriental, Colombia). Dissertationes Botananicae 116: 1–241.

109.

Kuhry

Hooghiemstra

Van Geel

. (1993) The El Abra stadial in the eastern cordillera of Colombia (South America). Quaternary Science Reviews 12(5): 333–343.

110.

Lachniet

Vazquez-Selem

(2005) Last glacial maximum equilibrium line altitudes in the circum-Caribbean (Mexico, Guatemala, Costa Rica, Colombia, and Venezuela). Quaternary International 138–139: 129–144.

111.

Lægaard

(1992) Influence of fire in the grass páramo vegetation of Ecuador. In: Balslev

Luteyn

(eds) Páramo: An Andean Ecosystem Under Human Influence. London: Academic Press, pp. 151–170.

112.

Lanata

Borrero

(1999) The archaeology of hunter-gatherers in South America: Recent history and new directions. In: Politis

Alberti

(eds) Archaeology in Latin America. London: Routledge, pp. 76–89.

113.

Lauer

(1981) Ecoclimatological conditions of the páramo belt in the tropical high mountains. Mountain Research and Development 1(3–4): 209–221.

114.

Lavallée

(2000) The First South Americans. Salt Lake City: University of Utah Press, 260 pp.

115.

League

Horn

(2000) A 10,000 year record of páramo fires in Costa Rica. Journal of Tropical Ecology 16(5): 747–752.

116.

Leuschner

(2000) Are high elevations in tropical mountains arid environments for plants? Ecology 81(5): 1425–1436.

117.

Llambí

Fontaine

Rada

. (2003) Ecophysiology of dominant plant species during old-field succession in a high tropical Andean ecosystem. Arctic, Antarctic and Alpine Research 35(4): 447–453.

118.

López

Pino

Realpe

(2003) Stratigraphic, chronometric and lithic technological evidence for Late Pleistocene and early Holocene occupations in the Middle Magdalena river, Colombia. Current Research in the Pleistocene 20: 46–49.

119.

Luteyn

(1999) Introduction to the páramo ecosystem. In: Luteyn

(ed.) Páramos: A Checklist of Plant Diversity, Geographical Distribution, and Botanical Literature. Brooklyn: New York Botanical Garden, pp. 1–39.

120.

Lynch

(1971) Preceramic transhumance in the Callejón de Huaylas, Peru. American Antiquity 36(2): 139–148.

121.

Lynch

(ed.) (1980) Guitarrero Cave: Early Man in the Andes. New York: Academic Press, 328 pp.

122.

Lynch

(1989) Chobshi cave in retrospect. Andean Past 2: 1–32.

123.

Lynch

Pollock

(1981) La arqueología de la Cueva Negra de Chobshi. Miscelánea Antropológica Ecuatoriana 1: 92–119.

124.

MacNeish

Vierra

(1983) The Preceramic way of life in the puna ecozone. In: MacNeish

Vierra

Nelken-Terner

. (eds) Prehistory of the Ayacucho Basin Volume IV: The Preceramic Way of Life. Ann Arbor: University of Michigan Press, pp. 225–235.

125.

Marchant

Berrio

Cleef

. (2001) A reconstruction of Colombian biomes derived from modern pollen data along an altitude gradient. Review of Palaeobotany and Palynology 117(1–3): 79–92.

126.

Marchant

Boom

Hooghiemstra

(2002) Pollen-based biome reconstructions for the past 450,000 yr from the Funza-2 core, Colombia: Comparisons with model-based vegetation reconstructions. Palaeogeography, Palaeoclimatology, Palaeoecology 177(1–2): 29–45.

127.

Mayer-Oakes

(1986) Early man projectile points and lithic technology in the Ecuadorian Sierra. In: Bryan

(ed.) New Evidence for the Pleistocene Peopling of the Americas. Orono: Center for the Study of Early Man, University of Maine, 133–156.

128.

Mayle

Burn

Power

Urrego

(2009) Vegetation and fire at the Last Glacial Maximum in tropical South America. In: Vimeux

Sylvestre

Khodri

(eds) Past climate variability in South America and surrounding regions from the Last Glacial Maximum to the Holocene. Series: Developments in Paleoenvironmental Research 14: 89–112.

129.

Melief

ABM

(1985) Late Quaternary paleoecology of the Parque Nacional Natural Los Nevados (Cordillera Central) and Sumapaz (Cordillera Oriental) areas, Colombia. PhD Thesis, Universiteit van Amsterdam.

130.

Meltzer

Grayson

Ardila

Barker

Dincauze

. (1997) On the Pleistocene antiquity of Monte Verde, southern Chile. American Antiquity 62(4): 659–663.

131.

Miller

Gill

(1990) Zooarchaeology at Pirincay, a Formative Period site in highland Ecuador. Journal of Field Archaeology 17(1): 49–68.

132.

Mills

(1986) Prescribed burning and hunter-gatherer subsistence systems. Haliksai 5: 1–26. Albuquerque, New Mexico, USA: University of New Mexico.

133.

Miotti

Salemme

(2003) When Patagonia was colonized: People mobility at high latitudes during Pleistocene/Holocene transition. Quaternary International 109–110: 95–111.

134.

Monasterio

(1979) El páramo desértico en el altiandino de Venezuela. In: Salgado-Labouriau

(ed) El medio ambiente páramo. Mérida, Venezuela: Actas del Seminario de Mérida, Centro de Estudios Avanzados, 117–146.

135.

Montoya

Rull

Van Geel

(2010) Non-pollen palynomorphs from surface sediments along an altitudinal transect of the Venezuelan Andes. Palaeogeography, Palaeoclimatology, Palaeoecology 297(1): 169–183.

136.

Moreno

Kitzberger

Iglesias

Holz

(2010) Paleofires in southern South America since the Last Glacial Maximum. PAGES News 18(2): 75–77.

137.

Moscol Olivera

Cleef

(2009a) A phytosociological study of the páramo along two altitudinal transects in El Carchi province, northern Ecuador. Phytocoenologia 39(1): 79–107.

138.

Moscol Olivera

Cleef

(2009b) Vegetation composition and altitudinal distribution of Andean rain forests in El Angel and Guandera reserves, northern Ecuador. Phytocoenologia 39(2): 175–204.

139.

Moscol Olivera

Duivenvoorden

Hooghiemstra

(2009) Pollen rain and pollen representation across a forest-páramo ecotone in northern Ecuador. Review of Palaeobotany and Palynology 157(3–4): 285–300.

140.

Moscol Olivera

Hooghiemstra

(2010) Three millennia upper forest line changes in northern Ecuador: Pollen records and altitudinal vegetation distributions. Review of Palaeobotany and Palynology 163(1–2): 113–126.

141.

Mueller-Dombois

(1981) Fire in tropical ecosystems. In: Proceedings of the Conference: Fire Regimes and Ecosystem Properties. Washington: U.S. Department of Agriculture, General Technical Report WO-26, 137–176.

142.

Nevle

Bird

(2008) Effects of syn-pandemic fire reduction and reforestation in the tropical Americas on atmospheric CO² during European conquest. Palaeogeography, Palaeoclimatology, Palaeoecology 264(1–2): 25–38.

143.

Nevle

Bird

Ruddiman

Dull

(2010) Neotropical human-landscape interactions, fire, and atmospheric CO² during European conquest. The Holocene 21(5): 853–864.

144.

Niemann

Behling

(2008) Late Quaternary vegetation, climate and fire dynamics inferred from the El Tiro record in the southeastern Ecuadorian Andes. Journal of Quaternary Science 23(3): 203–212.

145.

Niemann

Behling

(2009) Late Pleistocene and Holocene environmental change inferred from the Cocha Caranga sediment and soil records in the southeastern Ecuadorian Andes. Palaeogeography, Palaeoclimatology, Palaeoecology 276(1–4): 1–14.

146.

Niemann

Behling

(2010) Late Holocene environmental change and human impact inferred from three soil monoliths and the Laguna Zurita multi-proxi record in the southeastern Ecuadorian Andes. Vegetation History and Archaeobotany 19(1): 1–15.

147.

Niemann

Brunschön

Behling

(2010) Vegetation/modern pollen rain relationship along an altitudinal transect between 1920 and 3185 m a.s.l. in the Podocarpus National Park region, southeastern Ecuadorian Andes. Review of Palaeobotany and Palynology 159(1–2): 69–80.

148.

Niemann

Haberzettl

Behling

(2009) Holocene climate variability and vegetation dynamics inferred from the (11 700 cal yr B.P.) Laguna Rabadilla de Vaca sediment record, southeastern Ecuadorian Andes. The Holocene 19(2): 307–316.

149.

Nierop

Jansen

(2009) Extensive transformation of organic matter and excellent lipid preservation at the upper, superhumid Guandera páramo. Geoderma 151(3–4): 357–369.

150.

O’Connell

Allen

(1995) Human reactions to the Pleistocene-Holocene transition in greater Australia, a summary. Transitions: Pleistocene to Holocene in Australia and Papua New Guinea. Antiquity 69(265): 855–862.

151.

Ojasti

(1983) Ungulates and large rodents of South America. In: Bourlière

(ed) Tropical Savannas. Ecosystems of the World 13, Amsterdam: Elsevier, 427–439.

152.

Paduano

Bush

Baker

Fritz

Seltzer

(2003) A vegetation and fire history of Lake Titicaca since the Last Glacial Maximum. Palaeogeography, Palaeoclimatology, Palaeoecology 194(1–3): 259–279.

153.

Pearsall

(1980) Pachamachay ethnobotanical report: Plant utilization at a hunting base camp. In: Rick

(ed) Prehistoric hunters of the high Andes. New York: Academic Press, 191–231.

154.

Pearsall

(1983) Evaluating the stability of subsistence strategies by use of paleoethnobotanical data. Journal of Ethnobiology 3(2): 121–137.

155.

Pearsall

(1996) Appendix F: Analysis of charred botanical remains from Manachaqui cave, Peru. In: Church

Prehistoric cultural development and interregional interaction in the tropical montane forests of Peru. PhD Thesis, Yale University, 869–895.

156.

Pérez

(1998) Human impact on the high páramo landscape of the Venezuelan Andes. In: Zimmerer

Young

(eds) Nature’s geography: New lessons for conservation in developing countries. Madison, Wisconsin, USA: University of Wisconsin Press, 147–183.

157.

Pinter

Fiedel

Keeley

(2011) Fire and vegetation shifts in the Americas at the vanguard of Paleoindian migration. Quaternary Science Reviews 30(3–4): 269–272.

158.

Piperno

(2006) The role of phytoliths in archaeological reconstruction. In: Phytoliths: A comprehensive guide for archaeologists and paleoecologists. Lanham, Maryland, USA: Altamira Press, 139–164.

159.

Piperno

Jones

(2003) Paleoecological and archaeological implications of a late Pleistocene/early Holocene record of vegetation and climate from the Pacific coastal plain of Panama. Quaternary Research 59(1): 79–87.

160.

Piperno

Bush

Colinvaux

(1990) Paleoenvironments and human occupation in Late-glacial Panama. Quaternary Research 33(1):108–116.

161.

Pitblado

(2011) A tale of two migrations: Recording recent biological and archaeological evidence for the Pleistocene peopling of the Americas. Journal of Archaeological Research 19(4): 327–375.

162.

Power

Bush

Behling

Horn

Mayle

Urrego

(2010) Paleofire activity in tropical America during the last 21 ka: A regional synthesis based on sedimentary charcoal. PAGES News 18(2): 73–75.

163.

Power

Marlon

Ortiz

Bartlein

Harrison

Mayle

. (2008) Changes in fire regimes since the Last Glacial Maximum: An assessment based on a global synthesis and analysis of charcoal data. Climate Dynamics 30(7–8): 887–907.

164.

Price

Bowman

(1994) Fire-stick forestry: A matrix model in support of skillful fire management of Callitris intratropica R.T. Baker by north Australian Aborigines. Journal of Biogeography 21(6): 573–580.

165.

Pyne

(1993) Keeper of the flame: A survey of anthropogenic fire. In: Crutzen

Goldammer

(eds) Fire in the environment: The ecological, atmospheric, and climatic importance of vegetation fires. Chichester, UK: John Wiley and Sons, 245–266.

166.

Rick

(1980) Prehistoric hunters of the high Andes. New York: Academic Press, 360 pp.

167.

Rodgers

Horn

(1996) Modern pollen spectra from Costa Rica. Palaeogeography, Palaeoclimatology, Palaeoecology 124(1–2): 53–71.

168.

Roosevelt

Douglas

Brown

(2002) The migrations and adaptations of the first Americans: Clovis and pre-Clovis viewed from South America. In: Jablonksi

(ed) The first Americans: The Pleistocene colonization of the New World. Series: Memoirs of the California Academy of Sciences 27, 159–235.

169.

Rull

(2009) New palaeoecological evidence for the potential role of fire in the Gran Sabana, Venezuelan Guayana, and implications for early human occupation. Vegetation History and Archaeobotany 18(3): 219–224.

170.

Rull

Abbott

Polissar

Wolfe

Bezada

Bradley

(2005) 15 000-yr pollen record of vegetation change in the high altitude tropical Andes at Laguna Verde Alta, Venezuela. Quaternary Research 64(3): 308–317.

171.

Salazar

(1984) Cazadores recolectores del antiguo Ecuador. Quito: Banco Central del Ecuador, Serie Nuestro Pasado 1, 101 pp.

172.

Salazar

(1985) Investigaciones arqueológicas en Mullamica (Provincia de Pichincha). Miscelánea Antropológica Ecuatoriana 5: 129–160.

173.

Salgado-Labouriau

(1979) Cambios climáticos durante el Cuaternario Tardío paramero y su correlación con las tierras tropicales calientes. In: Salgado-Labouriau

(ed) El medio ambiente páramo. Mérida, Venezuela: Actas del Seminario de Mérida, Centro de Estudios Avanzados, 67–78.

174.

Salgado-Labouriau

(1986) Late-Quaternary paleoecology of Venezuelan high mountains. In: Vuilleumier

Monasterio

(eds) High altitude tropical biogeography. New York and Oxford: Oxford University Press, 202–217.

175.

Salgado-Labouriau

Schubert

(1976) Palynology of Holocene peat bogs from the central Venezuelan Andes. Palaeogeography, Palaeoclimatology, Palaeoecology 19(2): 147–156.

176.

Salgado-Labouriau

Bradley

Yuretich

Weingarten

(1992) Paleoecological analysis of the sediments of Lake Mucubají, Venezuelan Andes. Journal of Biogeography 19(3): 317–327.

177.

Salgado-Labouriau

Rull

Schubert

Valastro

(1988) The establishment of vegetation after late Pleistocene deglaciation in the Páramo de Miranda, Venezuelan Andes. Review of Palaeobotany and Palynology 55(1): 5–17.

178.

Salgado-Labouriau

Schubert

Valastro

(1977) Paleoecologic analysis of a Late-Quaternary terrace from Mucubají, Venezuelan Andes. Journal of Biogeography 4(4): 313–325.

179.

Salomons

(1986) Paleoecology of volcanic soils in the Colombian Central Cordillera (Parque Nacional Natural de los Nevados). In: Van der Hammen

Diaz-Piedrahita

Alvarez

(eds) La Cordillera Central Colombiana Transecto Parque Los Nevados (segunda parte). Series: Studies on Tropical Andean Ecosystems 3, 15–215 . Berlin: J Cramer.

180.

Sarmiento

(2000) Breaking mountain paradigms: Ecological effects on human impacts in man-aged tropandean landscapes. Ambio 29(7): 423–431.

181.

Sarmiento

(2002) Anthropogenic change in the landscapes of highland Ecuador. Geographical Review 92(2): 213–234.

182.

Sarmiento

Frolich

(2002) Andean cloud forest tree lines. Mountain Research and Development 22(3): 278–287.

183.

Sarmiento

Llambí

Escalona

Marquez

(2003) Vegetation patterns, regeneration rates and divergence in an old-field succession of the high tropical Andes. Plant Ecology 166(1): 145–156.

184.

Scheinsohn

(2003) Hunter-gatherer archaeology in South America. Annual Review of Anthropology 32: 339–361.

185.

Schreve-Brinkman

(1978) A palynological study of the upper Quaternary sequence in the El Abra corridor and rock shelters (Colombia). Palaeogeography, Palaeoclimatology, Palaeoecology 25(1–2): 1–109.

186.

Seppä

Bennett

(2003) Quaternary pollen analysis: Recent progress in palaeoecology and palaeoclimatolgy. Progress in Physical Geography 27(4): 548–579.

187.

Sklenář

(2006) Searching for altitudinal zonation: Species distribution and vegetation composition in the superpáramo of Volcán Iliniza, Ecuador. Plant Ecology 184(2): 337–350.

188.

Sklenář

Balslev

(2005) Superpáramo plant species diversity and phytogeography in Ecuador. Flora 200(5): 416–433.

189.

Sklenář

Lægaard

(2003) Rain-shadow in the high Andes of Ecuador evidenced by páramo vegetation. Arctic, Antarctic and Alpine Research 35(1): 8–17.

190.

Sklenář

Ramsay

(2001) Diversity of zonal páramo plant communities in Ecuador. Diversity and Distributions 7(3): 113–124.

191.

Smith

(1980) Plant remains from Guitarrero Cave. In: Lynch

(ed) Guitarrero cave: Early man in the Andes. New York: Academic Press, 87–119.

192.

Smith

Mark

Rodbell

(2008) The timing and magnitude of mountain glaciation in the tropical Andes. Journal of Quaternary Science 23(6–7): 609–634.

193.

Stahl

(2008) Animal domestication in South America. In: Silverman

Isbell

(eds) Handbook of South American archaeology. New York: Springer, 121–130.

194.

Stothart

Piperno

Andres

(2003) Terminal Pleistocene/early Holocene human adaptation in coastal Ecuador: The Las Vegas evidence. Quaternary International 109–110: 23–43.

195.

Thouret

J-C

Van der Hammen

Salomons

Juvigné

(1996) Paleoenvironmental changes and glacial stades of the last 50 000 years in the Cordillera Central, Colombia. Quaternary Research 46(1): 1–18.

196.

Trabaud

(1987) Fire and survival traits of plants. In: Trabaud

(ed) The role of fire in ecological systems. The Hague: SPB Academic Publishing, 65–89.

197.

Troll

(1968) The cordilleras of the tropical Americas. Series: Colloquium Geographicum 9: 15–56.

198.

Valencia

Urrego

Silman

Bush

(2010) From ice age to modern: A record of landscape change in an Andean cloud forest. Journal of Biogeography 37(9): 1637–1647.

199.

Van der Hammen

(1974) The Pleistocene changes of vegetation and climate in tropical South America. Journal of Biogeography 1(1): 3–26.

200.

Van der Hammen

(1979) History of flora, vegetation and climate in the Colombian Cordillera Oriental during the last five million years. In: Larsen

Holm-Nielsen

(eds) Tropical Botany. London: Academic Press, 25–32.

201.

Van der Hammen

Cleef

(1986) Development of the high Andean páramo flora and vegetation. In: Vuilleumier

Monasterio

(eds) High altitude tropical biogeography. Oxford and New York: Oxford University Press and American Museum of Natural History, 153–201.

202.

Van der Hammen

Correal

(1978) Prehistoric man on the Sabana de Bogotá: Data for an ecological prehistory. Palaeogeography, Palaeoclimatology, Palaeoecology 25(1–2): 179–190.

203.

Van der Hammen

González

(1960a) Holocene and Late Glacial climate and vegetation of Páramo de Palacio (Eastern Cordillera, Colombia, South America). Geologie en Mijnbouw 39: 737–746.

204.

Van der Hammen

González

(1960b) Upper Pleistocene and Holocene climate and vegetation of the “Sabana de Bogotá” (Colombia, South America). Leidse Geologische Mededelingen 25: 261–315 + 55 plates.

205.

Van der Hammen

González

(1965) A Late-glacial and Holocene pollen diagram from Ciénaga del Visitador (Dept. Boyacá, Colombia). Leidse Geologische Mededelingen 32: 193–201.

206.

Van der Hammen

Hooghiemstra

(1995) El Abra stadial, a Younger Dryas equivalent in Colombia. Quaternary Science Reviews 14(9): 841–851.

207.

Van der Hammen

Hooghiemstra

(2003) Interglacial-glacial Fuquene-3 pollen record from Colombia: An Eemian to Holocene climate record. Global Planet Change 36(3): 181–199.

208.

Van der Hammen

Barelds

De Jong

De Veer

(1981) Glacial sequence and environmental history in the Sierra Nevada del Cocuy (Colombia). Palaeogeography, Palaeoclimatology, Palaeoecology 32: 247–340.

209.

Van Geel

Van der Hammen

(1973) Upper Quaternary vegetational and climatic sequence of the Fuquene area (Eastern Cordillera, Colombia). Palaeogeography, Palaeoclimatology, Palaeoecology 14(1): 9–92.

210.

Van ‘t Veer

Tan

Mommersteeg

Hooghiemstra

(1995) Multivariate analysis of the Middle and Late Pleistocene Funza pollen records of Colombia. Mededelingen Rijks Geologische Dienst 52: 195–212.

211.

Van ‘t Veer

Islebe

Hooghiemstra

(2000) Climate change during the Younger Dryas chron in northern South America: A test of the evidence. Quaternary Science Reviews 19(17–18): 1821–1835.

212.

Verweij

Beukema

(1992) Aspects of human influence on upper-Andean forest line vegetation. In: Balslev

Luteyn

(eds) Páramo: An Andean ecosystem under human influence. London: Academic Press, 171–175.

213.

Vogl

(1974) Effects of fire on grasslands. In: Kozlowski

Ahlgren

(eds) Fire and ecosystems. New York: Academic Press, 139–194.

214.

von Humboldt

Bonpland

(2009 [1807]) Essay on the geography of plants with a physical tableau of the equinoctial regions. In: Jackson

(ed) Essay on the geography of plants: Alexander Von Humboldt and Aimé Bonpland. Chicago: University of Chicago Press, 55–155.

215.

Walter

(1985) Vegetation of the Earth and ecological systems of the geo-biosphere. Berlin: Springer-Verlag, 318 pp.

216.

Weberbauer

(1945) El mundo vegetal de los Andes peruanos [Revised edition (1911) of Die Pflanzenwelt der peruanischen Anden. Leipzing: Wilhelm Engelmann]. Lima, Peru: Ministerio de Agricultura, 776 pp.

217.

Weng

Bush

Curtis

Kolata

Dillehay

Binford

(2006) Deglaciation and Holocene climate change in the western Peruvian Andes. Quaternary Research 66(1): 87–96.

218.

Wharton

(1968) Man, fire and wild cattle in Southeast Asia. Proceedings of the Annual Tall Timbers Fire Ecology Conference 8: 107–167.

219.

Wheeler Pires-Ferreira

Pires-Ferreira

Kaulicke

(1976) Preceramic animal utilization in the central Peruvian Andes. Science 194(4264): 483–490.

220.

Wille

Hooghiemstra

Hofstede

Fehse

Sevink

(2002) Upper forest line reconstruction in a deforested area in northern Ecuador based on pollen and vegetation analysis. Journal of Tropical Ecology 18(3): 409–440.

221.

Williamson

Schatz

Alvarado

Redhead

Stam

Sterner

(1986) Effects of repeated fires on tropical páramo vegetation. Tropical Ecology 27: 62–69.

222.

Wing

(1980) Faunal remains. In: Lynch

(ed) Guitarrero cave: Early man in the Andes. New York: Academic Press, 149–172.

223.

Wing

(1986) Domestication of Andean mammals. In: Vuilleumier

Monasterio

(eds) High altitude tropical biogeography. Oxford and New York: Oxford University Press, 246–264.

224.

Young

(1993) Woody and scandent plants on the edges of an Andean timberline. Bulletin of the Torrey Botanical Club 120(1): 1–18.

225.

Young

Keating

(2001) Remnant forests of Volcán Cotacachi, northern Ecuador. Arctic, Antarctic and Alpine Research 33(2): 165–172.