Effect of South American grazing camelids on soil fertility and vegetation at the Bolivian Andean grasslands

Agriculture, Ecosystems and Environment 207 (2015) 203–210

Contents lists available at ScienceDirect

Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee

Effect of South American grazing camelids on soil fertility and vegetation at the Bolivian Andean grasslands Maria Angeles Muñoz * , Angel Faz, Jose Alberto Acosta, Silvia Martínez-Martínez, Raul Zornoza Sustainable Use, Management, and Reclamation of Soil and Water Research Group, Department of Agrarian Science and Technology, Universidad Politécnica de Cartagena, Paseo Alfonso XIII, 48, 30203 Cartagena, Murcia, Spain

A R T I C L E I N F O

A B S T R A C T

Article history: Received 1 December 2014 Received in revised form 12 February 2015 Accepted 3 April 2015 Available online xxx

The high grasslands at Bolivian Andes provide a natural habitat for a high number of wild and domestic South American camelids such as vicuna (Vicugna vicugna) and alpaca (Lama pacos). Because of the importance of the camelid raising for the Andean inhabitants economy and the sustainable biodiversity, it is fundamental to determine the natural resources condition and their availability to camelids. The objectives of this research were to: (i) evaluate the soil fertility; (ii) characterize the plant communities and its relationship with the landscape; and (iii) analyze the effect of the camelid populations on soil properties, vegetation cover and plant species richness. Soil and vegetation samplings were carried out in eight areas with different vicuna densities. Results pointed out no effect of grazing camelids on soil properties, including soil fertility in all areas. The most plentiful plant community was Pycnophyllum sp. grassland, although it was highly disturbed due to domestic camelid grazing. The studied areas presented medium (30–50%) and high plant cover (>50%). The substitution of palatable by non palatable plant species in those zones with high alpaca concentration highlighted the negative domestic camelid effect on vegetation composition. No effect of vicuna grazing on vegetation was observed. However, some protection actions should be undertaken to prevent biodiversity decline by bringing resource over exploitation by domestic camelids in the Bolivian Andes. ã 2015 Elsevier B.V. All rights reserved.

Keywords: Soil fertility Plant communities Highland grasslands Camelid grazing Biodiversity Bolivian Andes

1. Introduction The Apolobamba Integrated Management National Area is located in the Northern Bolivian Andes and provides a natural habitat for wild South American camelids such as vicuna (Vicugna vicugna) and domestic ones such as alpaca (Lama pacos) or llama (Lama glama). This one is scarcely present in some mixed flocks of alpaca and llama and little flocks of alpaca (SERNAP, 2006). Although no data on llama census are available the total llama is supposed fewer than 5% of the total alpaca (SERNAP, 2006). The presence of guanaco has not been described in Apolobamba. The main economical activity for Aymara and Quechua indigenous communities from Apolobamba is the domestic camelid raising. Vicuna has been a threatened species recognized by The World Conservation Union as vulnerable (IUCN, 1990) and classified as least concern in 2008 (Lischtenstein et al., 2008) pointing out the recovery of populations. Since their fleece is one of the finest fibres in the world, its exploitation is an example of

* Corresponding author. Tel.: +34 968 32 57 52. E-mail address: [email protected] (M.A. Muñoz). http://dx.doi.org/10.1016/j.agee.2015.04.005 0167-8809/ ã 2015 Elsevier B.V. All rights reserved.

sustainable management of the biodiversity by indigenous communities (Agencia Española de Cooperación Internacional, 2004). Furthermore, Apolobamba area is one of the poorest zones in Latin American where Aymara and Quechua ethnic groups from three municipalities (Charazani, Curva and Pelechuco) present a human development index of 0.472 (Instituto Nacional de Estadística, 2005). Ecosystems in the Puna or high altitude grasslands in the Andean region, are degraded as a consequence of anthropogenic activities such as the excessive grazing (Rocha and Sáez, 2003). The wild and domestic camelid populations have increased in the last decades with more than 140,000 animals in the Apolobamba area (130,000 alpacas and more than 10,000 vicunas) distributed in 1200 km2 (SERNAP, 2006). The protection actions of vicuna have reduced noticeably the poaching increasing the populations mainly from 80s to the end of 90s (Agencia Española de Cooperación Internacional, 2004). Likewise, some owners of medium and big alpaca flocks have increased the number of animals coupled with a scarce genetic improvement (Agencia Española de Cooperación Internacional, 2004; SERNAP, 2006). The landscape has a communal use in Apolobamba (Rocha and Sáez,

204

M.A. Muñoz et al. / Agriculture, Ecosystems and Environment 207 (2015) 203–210

2003); therefore, wild and domestic camelids share natural resources such as soil and vegetation, as in others Andean grasslands (Borgnia et al., 2010). The Andean ecosystem is a fragile environment with rigorous climatological conditions and low and variable quantity and quality of food for camelid grazing (Borgnia et al., 2010). Because of the importance of wild and domestic camelids population for the Apolobamba’s inhabitants economy, it is fundamental to assess the condition of the natural resources in the region and their availability to camelids. Soil property determination can contribute to the understanding of the soil fertility and the grazing impact (Brady and Weil, 2008; Dorrough et al., 2006; Li et al., 2008). In addition, the high complexity and heterogeneity of the Andean habitats are assumed to be responsible for the variety in altitudinal distribution and compositional changes of the vegetation (Brunschön and Behling, 2010). There is a need for greater understanding of physico-chemical soil properties and the impact of management in the Andean region (Fonte et al., 2012). The botanical identification of the native species, the landscape description, the plant cover and the species richness provide significant information to understand the foraging ecology of the vicuna and the plant community evolution related to the camelid grazing impact (Borgnia et al., 2010; Seibert, 1993). Limited information is available from previous studies conducted in the Northern Bolivian Andean regarding soil properties, vegetation and impact of camelid grazing (Beck et al., 2002; Seibert, 1993). In order contribute to one of the least well understood ecosystems around the world, the objectives of this research were to: (i) evaluate the soil fertility; (ii) characterize the plant communities and its relationship with the landscape; and (iii) analyze the effect of the camelids populations on soil

properties, vegetation cover and plant species richness in the Apolobamba area. The information obtained could improve the understanding of the proper management and the sustainable camelid exploitation in Andean grassland. 2. Materials and methods 2.1. Study area Apolobamba is an Integrated Management National Area located in the Northwest of La Paz Bolivia, bordering with the Republic of Peru (Fig. 1). This area is composed by three ecological zones: Puna, valley and tropic with high, medium and low altitude range, respectively, covering an area of 4837 km2 (SERNAP, 2014). The research was carried out in the Puna of the Apolobamba area, the highest altitude ecological zone ranging from 4300 to 4900 m.a.s.l., with 1200 km2 of extension (SERNAP, 2006). The study area is characterized by udic and frigid soil moisture and temperature regimes (USDA, 2014), with an annual average temperature of 4.5  C and total precipitation of 505 mm concentrated in five months (November–March) (SERNAP, 2014). The zone exhibits a mountain landscape influenced by glacial processes and parent materials such as metamorphic pelite with slate, metalimonite and sandstone (SERNAP, 2006). Considering the bioclimatic model proposed by Rivas-Martínez (2004) and the bioclimatic map developed by Navarro and Maldonado (2002), the area studied is classified as orotropical. The Puna of the Apolobamba area is included in the North Puna eco-region. Some authors assessed 148 species at the altoandine vegetation subeco-region in the Apolobamba area (Beck et al., 2002; García et al., 2002a).

Fig. 1. Location of the Apolobamba area and sampling zones.

M.A. Muñoz et al. / Agriculture, Ecosystems and Environment 207 (2015) 203–210

2.2. Soil sampling and analytical methods Due to the vicuna management program significance to the Apolobamba inhabitants development and the biodiversity protection, vicuna population density was the main reason to select the zones studied (Fig. 1). With regards to the last vicuna censuses carried out in the Apolobamba area (SERNAP, 2006), eight zones with dissimilar vicuna densities were selected. The selected zones presented similar densities in pairs: 1–2, 3–4, 5–6 and 7–8 although these pairs were not replicates (Table 1). Census results showed that each vicuna population inhabited in a particular zone maintaining stable populations due to the natural geographical barriers of the zones and the territorial and sedentary behavior of the vicuna (Borgnia et al., 2010). Unfortunately, the vicuna censuses were merely conducted from 1996 to 2005 due to the lack of funding for the management program (SERNAP, 2006). Data from the last 5 years were compared selecting those zones with stable populations. The selected zones were: zones 1 and 2 (low vicuna density, from 2.1 to 9.4 individuals km 2), zones 3 and 4 (medium vicuna density, from 9.4 to 16.5 individuals km 2), zones 5 and 6 (high vicuna density, from 16.5 to 23.2 individuals km 2) and zones 7 and 8 (very high vicuna density, from 23.1 to 58.1 individuals km 2) (Table 1). In order to characterize the soils, three replicate plots of 5 m  5 m were chosen in each studied zone located approximately 50 m among them. The zones studied were homogeneous enough regarding landscape, soil and vegetation characteristics to select this size of sampling plots. Since the Puna region presents a marked seasonality in rainfall which affects plant community composition and the seasonal variation in the camelid diet (Marshal et al., 2004), soil and vegetation samplings were carried out in the same season, corresponding to the dry season (from May to October). Since most of the roots were located in the surface layer, three replicates of soil samples were randomly collected from 0 to 5 cm depth. Samples were placed in plastics bags, then sealed and transported to the laboratory. Soil samples were air-dried and passed through a 2 mm sieve to remove gravels, plant remains and root fragments. The following soil analyses were carried out: total organic carbon (TOC) was determined using a TOC Analyzer (Shimadzu 5000, Japan); total nitrogen (TN) was measured according to the Kjeldahl method (Duchaufour, 1970); available phosphorous was determined using the method of Watanabe and Olsen (1965); cation exchange capacity (CEC) and exchangeable cations according to Chapman (1965); pH was measured in 1:1 (w/v) aqueous solution (Peech, 1965) and texture through the pipette Robinson method (FAO-ISRIC-ISSS, 2006). Landscape characterization of each zone was carried out based on the soil description system (FAO, 2006), considering landforms, slope position, gradient classes, land-use classification, exposition, altitude and parent material. 2.3. Vegetation sampling and methodology The same three plots per zone were considered to carry out the vegetation sampling. Vegetation samples were collected and they

205

were pressed and dried in an oven at 80  C for 24 h. The botanical identification was carried out in the Herbario Nacional de Bolivia using its own botanical data base and the Missouri Botanical Garden’s one (W3 Tropicos database). The phytosociological characterization of the plant communities was established following Braun-Blanquet (1964) methodology described by Seibert (1993). The term grassland was used to signify the herbaceous type of vegetation dominated by graminoid growth forms (UNESCO, 1973). The plant cover percentage was determined following Huss et al. (1986) methodology, modified using a sample grid or transect of 50  50 cm and needles inserted with 90 in each sampling plot. The richness was calculated as the average of all plant species found in each plot for each zone. 2.4. Statistical analyses The normality of the distribution of soil parameters and plant coverage was studied by Shapiro–Wilk’s Test (Shapiro and Wilk, 1965) through the residual analysis with Normal Probability Plot Test. When needed, Ln transformations were carried out to achieve normality. Bartlett Test’s confirmed the homegenity of the variance. One-way ANOVA followed by a post-hoc Tukey’s test at P < 0.05 was completed to identify significant differences among zones through the comparisons of all possible pairs of means (Fisher, 1935). Relationships between soil properties, plant coverage, species richness and alpaca and vicuna densities were studied using Pearson correlations. Statistical analyses were performed with the software Statistix 9.0. 3. Results 3.1. Soil properties The highest TOC, TN and CEC were found in zone 4, without significant differences with zone 7 (Table 2). Zones 5, 1 and 6 had the lowest TOC and TN contents The lowest value of CEC was registered in zones 6, 1 and 5 zone 1 exhibited the highest content of available P, with no significant differences with zones 4, 5, 7 and 8; whereas the lowest content was found in zone 2. Exchangeable Na+ showed the highest concentration in zones 6, 2 and 8. Conversely, K+, Mg2+ and Ca2+ showed the lowest contents in this zone. The lowest Na+ concentration was observed in zones 5, 1 and 4. The highest K+ content was found in zone 4 and no significant differences were found with the other studied zones except for zone 6. The maximum Ca2+ content was registered in zones 3, 2 and 4. The highest Mg2+ concentration was observed in zones 2, 3 and 4. The lowest pH mean value was observed in zone 8 while the highest was in zones 2 and 3. According to the particle size distribution, the highest sand content was observed in zones 1, 5 and 6. The highest clay content was found in zones 2 and 3, (18.0  2%) with no significant differences between them (Table 2). Correlation analysis pinpointed a high correlation factor between TOC and TN, TOC and CEC and CEC and clay contents (Table 3). A statistically significant correlation was found between exchangeable Ca2+ and Mg2+.

Table 1 Vicuna and alpaca densities in the studied zones. Zones

N vicunas

N vicunas km

1 2 3 4 5 6 7 8

139–412 139–412 413–691 413–692 692–1383 692–1384 1384–2119 1384–2120

2.1–9.4 2.1–9.4 9.4–16.5 9.4–16.6 16.5–23.1 16.5–23.2 23.1–58.1 23.1–58.2

2

Vicuna density

N alpacas

N alpacas km

Low Low Medium Medium High High Very high Very high

4005–6641 2543–4004 6642–15300 2543–4004 2543–4004 2543–4004 2543–4004 2543–4004

113.0–171.0 57.0–112.0 172.0–331.0 57.0–112.0 57.0–112.0 57.0–112.0 57.0–112.0 57.0–112.0

2

Alpaca density High Medium Very high Medium Medium Medium Medium Medium

206

M.A. Muñoz et al. / Agriculture, Ecosystems and Environment 207 (2015) 203–210

Table 2 Physicochemical soil properties in all studied zones. Values are mean  standard error (n = 9). Zones 1 2 3 4 5 6 7 8

TOC (g kg

1

TN (g kg

)

1

)

Available P (mg kg 1)

CEC (cmo(+) kg

Na+ (cmo(+) kg

1

)

1

)

K+ (cmo(+) kg

Ca+2 (cmo(+) kg

1

)

1

)

Mg+2 (cmo(+) kg

1

pH

Sand (%)

Clay (%)

)

51.5  6.4cd 58.9  5.2bc 55.0  3.1bc 91.7  5.6a 37.3  2.0d 45.9  2.3cd 72.7  3.8ab 61.1  5.9bc

4.3  0.4bcd 5.0  0.2b 4.6  0.3bc 7.2  0.4a 3.3  0.2d 3.9  0.1cd 5.2  0.3b 5.2  0.4b

266.4  45.6a 122.9  14.7c 165.4  9.4abc 209.1  7.0ab 156.6  12.1abc 150.3  17.3bc 172.0  12.8abc 176.7  9.8abc

13.9  0.6de 20.9  0.4ab 16.3  0.8cd 23.5  1.5a 14.2  1.1de 12.3  0.6e 18.3  0.9abc 17.2  1.4bcd

0.2  0.0cd 0.7  0.2ab 0.3  0.1bc 0.2  0.0cd 0.1  0.0d 1.0  0.2a 0.2  0.1bc 0.6  0.2ab

0.3  0.1bc 0.4  0.1abc 0.5  0.1abc 0.6  0.1ab 0.5  0.1abc 0.3  0.0c 0.6  0.1ab 0.5  0.1abc

0.8  0.2c 4.0  0.5a 4.1  0.1a 3.4  0.2ab 2.2  0.3b 0.5  0.1c 2.2  0.4b 0.6  0.1c

1.2  0.0b 1.6  0.2a 1.5  0.1a 1.0  0.1a 0.3  0.0b 0.2  0.0b 1.1  0.2a 0.4  0.1b

5.0  0.1c 5.9  0.1a 5.6  0.2a 5.1  0.1c 5.4  0.1b 4.9  0.1c 4.9  0.0bc 4.2  0.1d

69.0  2a 40.0  3d 60.6  3ab 44.9  1cd 63.4  4ab 66.0  1a 52.0  2bc 49.9  3bc

11.1  0cd 18.0  2a 16.0  1ab 14.0  1b 11.1  2bcd 11.0  1d 13.8  1b 14.0  1b

p(W):0.29 p(B):0.16 F:14.1***

p(W):0.46 p(B):0.30 F:17.0**

p(W):0.53 p(B):0.08 F:10.2**

p(W):0.83 p(B):0.06 F:14.3***

p(W):0.72 p(B):0.06 F:14.4***

p(W):0.06 p(B):0.23 F:3.56**

p(W):0.07 p(B):0.07 F:45.4***

p(W):0.14 p(B):0.56 F:41.5***

p(W):0.34 p(B):0.06 F:34.9***

p(W):0.07 p(B):0.06 F:18.4***

p(W):0.41 p(B):0.09 F:12.3***

TOC: total organic carbon; TN: total nitrogen; available P: available phosphorous; CEC: cation exchange capacity. Different letters indicate significant differences among zones (P < 0.05) according to Tukey’s test. p(W): Shapiro–Wilk’s test; p(B): Bartlett’s test; (F): Statistic ANOVA test. n.s. = not significant. ** P < 0.01. *** P < 0.001.

3.2. Botanical taxonomy The botanical identification of the samples pointed out similar families, genera and species in the different studied zones (Table 4). The most common are the next ones. The Poaceae family species were found in all studied zones; Aciachne acicularis was found in the eight areas. Festuca rigescens was identified in all studied zones except for zones 6 and 7, while Deyeuxia rigescens was identified in zones 1, 3, 4, 5 and 7. The Caryophillaceae family was present in all zones and was represented mainly by Pycnophyllum kobalantum. Different species of Asteraceae family were identified in zones 1, 2, 3, 5 and 6 (Table 4). One Cyperaceae family species (Scirpus rigidus) was found in zones 2, 4, 5, 6, 7 and 8, while Alchemilla pinnata, which belongs to Rosaceae family, was observed in zones 3, 4, 5 and 7. Plantagineaceae family was indentified in in zones 1, 3, and 8 while umbeliferae was observed in zone 3. Additionally, two Pteridophyta species were collected in zones 1, 6 and 8. 3.3. Landscape description and vegetation Landscape parameters for each zone are shown in Table 4. Most zones studied were located in medium gradient hill

with a gently and strongly sloping. The slope position was ranging between toe slope (zones 1 and 3) and crest (zones 6 and 8). Zones 1 and 3 were located in a plateau with a nearly level. Diverse exposition (N, NW, W, S, SW and NE) and altitudes were observed (the lowest in zone 3 and the highest in zone 8). The most common parent materials were consisted of schist, quartzite and although alluvial terraces, alluvial deposits or fans of mixed geological origin were also observed (zones 1, 3 and 5, respectively) (Table 4). The Poaceae species such as F. rigescens and A. acicularis were collected from different landforms and parent materials. However, Deyeuxia risgescens was not found in the crest slope position (zones 6 and 8), in slopes from 0.5% to 15%, between nearly level to strongly sloping from 4342 m.a.s.l. and 4564 m.a.s.l. Caryophyllaceae family were found in all studied zones and the most abundant species (S. rigidus) was observed in different landforms (zones 2, 4, 5, 6, 7 and 8) except for plain shape with slopes above 2%. Plantaginaceae family was observed in the plain shape and the crest hill (zones 1, 3 and 8) including the entire slope range and two expositions (W and SW). Pteridophita species was observed in zones 1, 4, 5, 6, 7 and 8. These zones ranged from lower

Table 3 Pearson correlation analysis among soil properties, plant coverage, species richness and vicuna and alpaca density. TOC TN P CEC Na+ K+ Ca2+ Mg2+ pH Sand Clay Plant coverage Richness Vicuna density Alpaca density

TN

P

CEC

Na+

K+

Ca2+

Mg2+

pH

Sand

Clay

Plant coverage

Richness Vicuna density

0.88*** 0.03 0.72*** 0.03 0.56*** 0.39*** 0.59*** 0.15 0.55*** 0.27* 0.46*** 0.03 0.11

0.03 0.76*** 0.01 0.41*** 0.37** 0.57*** 0.13 0.55*** 0.33** 0.40*** 0.12 0.06

0.20 0.08 0.08 0.25* 0.30** 0.25* 0.39*** 0.28* 0.11 0.05 0.06

0.00 0.37** 0.56*** 0.68*** 0.05 0.70*** 0.53*** 0.31** 0.07 0.04

0.17 0.35** 0.05 0.02 0.02 0.03 0.25* 0.25* 0.17

0.33** 0.40*** 0.14 0.29* 0.13 0.49*** 0.22 0.27*

0.84*** 0.66*** 0.53*** 0.55*** 0.16 0.29* 0.36**

0.40*** 0.67*** 0.64*** 0.29* 0.16 0.42***

0.12 0.29* 0.30* 0.21 0.70***

0.73*** 0.37** 0.29* 0.01

0.09 0.04 0.08

0.06 0.35**

0.02

0.07

0.06

0.05

0.00

0.04

0.08

0.40***

0.40***

0.44***

0.08

0.28*

0.17

0.017

TOC: total organic carbon; TN: total nitrogen; P: available phosphorous; CEC: cation exchange capacity. * Correlation is significant at P < 0.05. ** Correlation is significant at P < 0.01. *** Correlation is significant at P < 0.001.

0.43***

M.A. Muñoz et al. / Agriculture, Ecosystems and Environment 207 (2015) 203–210

207

Table 4 Landscape characteristics and identified species in each studied zone. Zone

Landform

Slope Slope gradient class Exposition Altitude position (%) (m.a.s.l.)

Parent material

Species identified

1

Plateau

Toe slope

Nearly level (0.5– 1%)

North

2

Medium gradient hill

Mid hill

Strongly sloping (10–15%)

Northwest 4675

3

Plateau

Toe slope

Nearly level (0.5– 1%)

West

Alluvial terraces, schist, basalt, quartzite and sandstone Glacial deposits, esquistes, quartzite and slate Alluvial deposits, esquistes and quartzite

4

Medium gradient hill

Lower slope

Gently sloping (2– 5%)

Southwest 4564

5

Medium gradient hill

Lower slope

Gently sloping (2– 5%)

Northeast

4502

6

Medium gradient hill Medium gradient hill

Crest

Strongly sloping (10–15%) Strongly sloping (10–15%)

North

4552

South

4560

Medium gradient hill

Crest

Strongly sloping (10–15%)

Southwest 4890

Festuca rigescens,Deyeuxia rigescens, Nasella brachyphylla, Aciachne acicularis, Pycnophyllum kobalantum, Senecio spinosus, Plantago sericea, Selaginella peruviana, Parmelia sp. Festuca rigescens, Aciachne acicularis, Pycnophyllum kobalantum, Paronychia andina, Belloa sp., Senecio spinosus, Scirpus rigidus Festuca rigescens, Deyeuxia rigescens, Deyeuxia minima, Aciachne acicularis, Pycnophyllum kobalantum, Perezya pygmaea, Alchemilla pinnata, Plantago sericea, Azorella diapensioides Festuca rigescens, Deyeuxia rigescens, Aciachne acicularis, Pycnophyllum kobalantum, Paronychia andina, Scirpus rigidus, Alchemilla pinnata, Selaginella peruviana Festuca rigescens, Deyeuxia rigescens, Stipa ichu, Aciachne acicularis, Pycnophyllum molle, Pycnophyllum kobalantum, Senecio spinosus, Belloa sp., Scirpus rigidus, Alchemilla pinnata, Selaginella peruviana Aciachne acicularis, Nasella brachyphylla, Pycnophyllum kobalantum, Belloa sp., Scirpus rigidus, Selaginella peruviana Deyeuxia rigescens, Stipa sp., Aciachne sp., Pycnophyllum kobalantum, Paronychia andina, Scirpus rigidus, Alchemilla pinnata, Selaginella peruviana Festuca rigescens, Pycnophyllum kobalantum, Paronychia andina, Scirpus rigidus, Plantago sericea, Parmelia sp., Selaginella peruviana

7

8

Middle slope

4367

4342

slope to middle slope, toe slope and crest, different expositions and the altitude varied from 4367 to 4890 m.a.s.l.

Glacial deposits, esquistes, quartzite and slate Alluvial fans, esquistes, quartzite and slate

Glacial deposits, esquistes and quartzite Glacial deposits, esquistes, quartzite and slate Glacial deposits, esquistes, quartzite and slate

4. Discussion 4.1. Soil fertility

3.4. Plant coverage and species richness Regarding plant coverage (PC), significant differences were found among the zones studied (Fig. 2A). Zone 7 exhibited the highest PC percentage with no significant differences with the other zones except for zone 6 which showed the lowest PC. Significant positive correlations were observed between plant cover and TOC, TN, CEC and K+ (Table 3). A positive correlation was found between vicuna density and plant cover whereas no significant correlation was observed between alpaca density and PC. Similar values of plant species richness were found in all zones, with no significant differences among them (Fig. 2B). The highest richness was identified in zones 3 and 5 while the other areas presented 4 different species. Table 4 shows that the highest number of the total different identified species was observed in zone 5 and the lowest in zone 6. No statistical correlations were observed between species richness, soil properties, plant coverage and camelid densities.

Vagen et al. (2006) established a soil fertility classification in highlands located in Madagascar taking into account a natural forest soil as reference and labeling the mean values observed as good fertility. According to this classification, zones 1 and 3 presented TOC concentration close to the mean value for good soil fertility (55.72 g kg 1) while zones 5 and 6 have exhibited TOC contents under this mean value and similar to the average value (43.93 mg kg 1). However, zones 4 and 7 presented the highest TOC contents which were much higher than the good fertility mean value. In fact, zones 4 and 7 exhibited high contents of total organic carbon regarding the Tibet Plateau (Genxu et al., 2007). Whereas Muñoz et al. (2013) showed high recalcitrant carbon stocks in these zones. This fact highlights the valuable role of these soils and environments in the C sequestration (Lal, 2014). Conversely, degraded organic matter was observed in those zones with highest alpaca concentration (Muñoz and Faz, 2012).

Fig. 2. Plant coverage percentages (A) and plant species richness (B) in the different studied areas from Apolobamba. Values are mean. Error bars denote standard error (n = 3).

208

M.A. Muñoz et al. / Agriculture, Ecosystems and Environment 207 (2015) 203–210

With respect to the TN contents, all the studied zones showed concentrations above the mean value for good fertility (3.90 g kg 1), as the same for CEC (11.76 cmol(+) kg 1) (Vagen et al., 2006). Our CEC values were similar to those reported by García-Pausas et al. (2008) in acid soils in the Pyrenean Mountains in Europe. High CEC values in the upper horizons are attributed to high quantities of organic matter or poorly crystalline clays derived from volcanic ash or both in the Puna of Peru (Wilcox et al., 1988). In addition, the relationships between TOC, clay content and CEC are widely described in the soil science literature (Brady and Weil, 2008; Urbano, 2001). We observed high exchangeable magnesium and low exchangeable calcium concentrations in zones 2 and 3. The high positive correlation observed between Ca2+ and Mg2+ was described by Dercon et al. (2003) in acid soils located in the Ecuadorian Andean region. All studied areas were above the mean value of exchangeable Mg2+ for good fertility (0.37 cmol(+) kg 1), except for zones 5 and 6. Owens et al. (2007) obtained similar available phosphorous values that those found in zone 2. However, higher available P contents were registered in the other zones, mainly in zone 1. Some researches established that the presence of high contents of available phosphorous and low clay contents could be associated to the phosphorous mobilization as a result of the erosive processes in grasslands (Quinton et al., 2001). However, the high contents of TOC observed could have associated high contents of available P, both in active and in slow and passive organic matter (Brady and Weil, 2008). Additionally, regarding the landscape, position or slope in zone 1, we can suggest that the landform would not influence the erosion as it was located in a plateau with toe slope. Zone 8 was extremely acid, while the other zones were very strongly acid and strongly acid (Soil Survey Division Staff, 1993). Texture analyses showed sandy-loam and loam soils (FAO-ISRICISSS, 2006). Li et al. (2008) described sand content in Mongolia grazing steppes while García-Pausas et al. (2008) reported similar results to our loam zones from acid soils in high altitude grasslands. Clay contents were similar to good fertility mean (14.32%) except for zones 2 and 3 where the clay contents were similar to the average value established by Vagen et al. (2006). In general, the soil parameters studied pointed out good soil fertility in the area. 4.2. Plant communities characterization Since the most representative plant species found was Pycnophyllum koballantum (identified in the eight studied zones), the most plentiful plant community observed was Pycnophyllum grassland (Seibert, 1993). Taking into account the plant composition of the high altitude grassland in the Andean region, some researches established the Festuca dolichophylla as dominant species (Beck et al., 2002; Seibert, 1993). These authors suggested that these grasslands usually present Deyeuxia, Stipa and Poa genera (Poaceae family). Accordingly, some of these genera such as Festuca and Deyeuxia were identified in most of the studied zones. Particularly, Seibert (1993) assessed the Poaceae species abundance in the Pycnophyllum grasslands and species observed in the most of the zones studied evidenced the general dominance of the Pycnophyllum grasslands. With regards to the plant community variants, the typical one was with Selaginella described by Seibert (1993) involved genera such as Deyeuxia, Aciachne and Alchemilla, which were found in almost all the area studied. Different species could indicate disturbance in zones 1 and 2. In this way, Senecio spinosus is a thorny species described by some authors as indicator of disturbance in the Apolobamba grasslands, due to the camelid grazing over-exploitation, as well as A. acicularis (García et al., 2002a; García and Beck, 2006). Species such as S. rigidus or

Plantago sericea observed in zones 1 and 2 could also indicate a relative disturbance degree (García and Beck, 2006). Therefore, we suggest that zones 1 and 2 represent disturbed Pycnophyllum grasslands characterized as the variant typical with Selaginella. According to Seibert (1993), zone 3 could exhibit a plant community characterized as moisture vegetation of Plantago tubulosa based on the Plantago genera species observed. This author established that species such as Deyeuxia minima, Azorella diapensioides or A. pinnata are abundant in the D. minima grasslands. According to this, we suggest that the zone 3 has the transition vegetation between the P. tubulosa and the D. minima grasslands. 4.3. Landscape description and vegetation The entire zones could be classified as nature protection (P) particularly as wildlife management (PN3), based on the sustainable vicuna management programme of FAO (2006). Regarding the landscape and the species identified in each zone, Poaceae, Caryophyllaceae, Asteraceae and Cyperaceae families appeared in diverse scenarios with different landscape characteristics. Seibert (1993) described the variant typical with Selaginella of the Pycnophyllum grassland between 4300 and 4500 m.a.s.l., with South slope aspect and 0-5%. We identified this kind of grassland in zone 8 where the altitude was around 4900 m.a.s.l. with Southwest aspect and 10–15% of slope. Zone 2 could present the disturbed Pycnophyllum grassland (variant typical with Selaginella) located at 4675 m.a.s.l., Northwest exposition and 10–15% of slope. Therefore, we suggest that the variant typical with Selaginella was distributed in a higher altitudinal area (until 4900 m.a.s.l.) than the Seibert’s (1993) description with dissimilar landscape characteristics (SW and N exposition and 2–15% of slope). Seibert (1993) found the variant Stipa with Selaginella in the same North and South exposition than in ours zones 5 and 7. Zone 3 with the P. tubulosa grassland presented similar landform, slope, aspect and altitude according to this author’s description. 4.4. Plant cover and species richness Researchers such as Beck et al. (2002) assessed that the PC in the altoandine unit of the Apolobamba area were between 5 and 30% in dry soils, whereas García et al., 2002b found the PC between 50% and 80% in moist soils. On the other hand, Genxu et al. (2007) established high PC when it was above 50% moisture in the Tibet Plateau. Following these researchers, the studied zones presented high PC (above 50%) except for zones 3 and 6, with medium PC (between 50% and 30%). The statistical analyses pointed out positive correlation between soil properties and the PC, which suggests that soils with a higher fertility promote higher plant coverage. On the other hand, considering the landscape description, different characteristics were found in those zones with the lowest PC (zones 3 and 6) with toe slope and crest and West and North aspect, respectively. According to this, we can conclude that the landscape influences the amount of PC. Nonetheless, we cannot ascertain that the highest plant cover provided the highest species richness, since no correlation was found between these parameters. With respect to species preferably consumed by camelids, a high number of palatable species were observed in all the zones studied. They were mainly grouped into the Poaceae family. Following to Borgnia et al. (2010), similar palatable genera for the vicuna were described in the Argentinean puna such as Stipa, Festuca or Deyeuxia. Other authors described that the Stipa and Festuca genera were highly consumed by the vicuna in Pampa Galeras, located in the Peruvian Andean steppes (Franklin, 1983).

M.A. Muñoz et al. / Agriculture, Ecosystems and Environment 207 (2015) 203–210

Two unpalatable species described by García and Beck (2006) were observed in zones 1 and 2, S. spinosus and A. acicularis. Also the unpalatable pteridophitas Selaginella peruviana and Parmelia sp. were identified in zone 1.

209

Acknowledgements We thank the Agencia Española de Cooperación Internacional para el Desarrollo (AECID) and the Herbario Nacional de Bolivia for their cooperation.

4.5. Effect of camelid population on soil properties and vegetation

References

Considering the relationships found between the vicuna and the alpaca densities and the soil parameters studied (merely exchangeable Ca2+, Mg2+ and pH) and taking into account the good fertility of the soils, we can conclude that, the soil properties and, therefore its fertility, were not affected by the camelid population. However, Rocha and Sáez (2003) assessed that the alpaca concentration was much higher than the ecosystems can support in some zones of the Apolobamba area, such as zone 1. Zones with no disturbed Pycnophyllum grassland presented the highest vicuna density while the alpaca concentration was medium (zones 4–8). Conversely, zones 1 and 2 presented disturbed vegetation with high alpaca concentration and low vicuna density. So, it could indicate the vicuna population preference for not disturbed Pycnophyllum grasslands whereas alpaca flocks were mainly located in disturbed grasslands. The grassland disturbance could be originated by the domestic camelid grazing regarding the high alpaca and the low vicuna densities in these zones. The positive relationship between vicuna density and plant cover may indicate the vicuna preference for those grasslands with higher plant cover. Since no negative correlation was identified between the alpaca density and the plant cover or species richness, we could not establish the domestic camelid grazing effect on these parameters in contrast to Carilla et al. (2011) observations in the Andean grasslands located in Argentina. However, some researchers established that the excessive grazing results in the increase of the unpalatable species (Carilla et al., 2011; Collins and Smith, 2006). We observed a high number of unpalatable species were observed in zone 1, with high alpaca and low vicuna densities. This point emphasizes the negative alpaca effect in the Pycnophyllum grasslands composition. Since no negative influence on the soil or the vegetation could be assessed due to the vicuna grazing and based on the negative alpaca influence in the grasslands, the alpaca flocks should be keeping under control.

Agencia Española de Cooperación Internacional, 2004. Manejo Sostenible de la Vicuña en Apolobamba, Agencia Española de Cooperación InternacionalPrograma Araucaria, La Paz. Beck, S., Garcia, E., Zenteno, F., 2002. Flora del Parque Nacional y Área Natural de Manejo Integrado Madidi, Proyecto de Apoyo al Parque Nacional Madidi, SERNAP-CARE-HNBCE-WCS, La Paz. Borgnia, M., Vilá, B.L., Cassini, M.H., 2010. Foraging ecology of Vicuáa, Vicugna vicugna, in dry puna of Argentina. Small Ruminant Res. 88, 44–53. Braun-Blanquet, J., 1964. Pflanzensoziologie, Grundzüge der Vegetationskunde, 3rd Edition Springer-verlag, Wien. Brady, N., Weil, R., 2008. The Nature and Properties of Soil. Pearson International Edition, New Jersey. Brunschön, C., Behling, H., 2010. Reconstruction and visualization of upper forest line and vegetation changes in the Andean depression region of southeastern Ecuador since the last glacial maximum – A multi-site synthesis. Rev. Palaeobot. Palynol. 163, 139–152. Carilla, J., Aragón, R., Gurvich, D., 2011. Fire and grazing differentially affect aerial biomass and species composition in Andean grasslands. Acta Oecol. 37, 337–345. Chapman, H.D., 1965. Cation exchange capacity. In: Black, C.A. (Ed.), Methods of Soil analysis. American Society of Agronomy, Madison, Wisconsin, pp. 891–900. Collins, S.L., Smith, M.D., 2006. Scale-dependent interaction of fire and grazing on community heterogeneity in tallgrass prairie. Ecology 87, 2058–2067. Dercon, G., Deckers, J., Govers, G., Poesen, J., Sánchez, H., Vanegas, R., Ramírez, M., Loaiza, G., 2003. Spatial variability in soil properties on slow-forming terraces in the Andes region of Ecuador. Soil Till. Res. 72, 31–41. Dorrough, J., Moxham, C., Turner, V., Sutter, G., 2006. Soil phosphorus and tree cover modify the effects of livestock grazing on plant species richness in Australian grassy woodland. Biol. Conserv. 130 (3), 394–405. Duchaufour, P., 1970. Precis de Pedologie. Masson, Paris. FAO, 2006. Guidelines for soil description, Food and Agriculture Organization of the United Nations, 4th Edition, Rome. FAO-ISRIC-ISSS, 2006. World Reference Base for Soil Resources 2006, A framework for international classification, correlation and communication 103, Rome. Fisher, R.A., 1935. The Design of Experiments. Oliver & Boyd, Edinburg. Fonte, S.J., Vanek, S.J., Ayarzun, P., Parsa, S., Quintero, C., Rao, I.M., Lavelle, P., 2012. Chapter four – pathways to agroecological intensification of soil fertility management by smallholder farmers in the andean highlands. Adv. Agron. 116, 125–184. Franklin, W., 1983. Contrasting sociecologies of south America’s wild camelids: the vicuña and the guano. In: Eisemberg, J.F., Kleiman, D.G. (Eds.), Advances in the Study of Mammalian Behaviour, , pp. 573–629. García, E., Beck, S., Zenteno-Ruiz, F.S., 2002. Plan de Manejo Madidi, Capítulo Botánica, Herbario Nacional de Bolivia, Informe interno. García, E., Beck, S., Zenteno-Ruiz, F.S., 2002. Plan de Manejo Madidi. Capítulo Botánica, Herbario Nacional de Bolivia. La Paz, Unpublished report.. García, E., Beck, S., 2006. Puna, In: Morales, M., Ollgaard, B., Kvist, L., Borchsenius, F., Baslev, H (Eds.), Botánica Económica de los Andes Centrales, La Paz, 51–76. García-Pausas, J., Casals, P., Camarero, L., Huguet, C., Thompson, R., Sebastiá, M.T., Romanýa, J., 2008. Factors regulating carbon mineralization in the surface and subsurface soils of Pyrenean mountain grasslands. Soil Biol. Biochem. 40 (11), 2803–2810. Genxu, W., Yuanshou, L., Yibo, W., Qingbo, W., 2007. Effects of permafrost thawing on vegetation and soil carbon pool losses on the Qinghai-Tibet Plateau, China. Geoderma 143 (1–2), 143–152. Huss D., Bernardon A., Anderson D., Brun J., 1986. Principios de manejo de praderas naturales, Instituto Nacional de Tecnología Agropecuaria y Oficina Regional de la FAO para América Latina y el Caribe, Santiago. Instituto Nacional de Estadística, 2005. Available at: . (accessed 10.05.14.). IUCN, 1990. Red List of Threatened Animals, International Union for Conservation Nature. Available at: . (accessed 7.06.14.). Lal, R., 2014. Soil carbon sequestration. ReykjavikIn: Halldórsson, G., Bampa, F., Posteinsdóttir, A.B., Sigurdsson, B., Montanarella, L., Arnalds, A. (Eds.), Soil Carbon Sequestration for Climate Food Security and Ecosystem Services. Proceedings of the International Conference, 27–29. , pp. 11–16 May 2013. Li, Ch., Hao, X., Zhao, M., Han, G., Willms, W.D., 2008. Influence of historic sheep grazing on vegetation and soil properties of a Desert Steppe in Inner Mongolia. Agric. Ecosyst. Environ. 128 (1–2), 109–116. Marshal, J.P., Bleich, V.C., Andrew, N.G., Krausman, P.R., 2004. Seasonal forage use by desert mule deer in southearstern California. SouthWest. Nat. 49, 501–505. Muñoz, M.A., Faz, A., 2012. Soil organic matter stocks and quality at high altitude grasslands of Apolobamba, Bolivia. Catena 94, 26–35. Muñoz, M.A., Faz, A., Zornoza, R., 2013. Carbon stocks and dynamics in grazing highlands from the Andean Plateau. Catena 104, 136–143. Navarro, G., Maldonado, M., 2002. Geografía Ecológica de Bolivia: Vegetación y ambientes acuáticos. Fundación Simón Patiño, Cochabamba.

5. Conclusions The physicochemical soil characterization indicated a good soil fertility status in general. The most plentiful plant community was Pycnophyllum grassland although it was highly disturbed in two of the zones studied likely due to the alpaca grazing. According to the landscape description, the typical Pycnophyllum grassland with Selaginella was found in dissimilar landscape description than currently is assessed by other authors in the Apolobamba area. The zones studied showed medium and high plant coverage which presented a positive relationship with the soil fertility. We suggest the vicuna preference for undisturbed Pycnophyllum grasslands with high plant coverage. No vicuna affection in the soil fertility and the vegetation was observed. However, the substitution of palatable species by no palatable ones in disturbed Pycnophyllum grasslands with high alpaca concentration could indicate the negative effect of the domestic camelid grazing in the original grassland composition. Therefore, some protection actions focused on the control of the over exploitation by domestic camelids should be undertaken. These actions would prevent the changes in the grassland composition and the vicuna environment affection in the Bolivian Andean grasslands.

210

M.A. Muñoz et al. / Agriculture, Ecosystems and Environment 207 (2015) 203–210

Owens, P., Deeks, L., Wood, G., Betson, M., Lord, E., Davison, P., 2007. Variations in the depth distribution of phosphorus in soil profiles and implications for modelbased catchment-scale predictions of phosphorus delivery to surface waters. J. Hydrol. 350 (3–4), 317–328. Peech, M., 1965. Hidrogen-ion activity. In: Black, C.A. (Ed.), Methods of Soil Analisis. American Society of Agronomy, Madison, Wisconsin, pp. 914–916. Quinton, J., Catt, J., Hess, T., 2001. The selective removal of phosphorus from soil: is event size important. J. Environ. Qual. 30, 538–545. Rivas-Martínez, S., 2004. Clasificación bioclimática de la Tierra. Available at: . (accesed 8.01.14.). Rocha, O., Sáez, C., 2003. Uso pastoril en humedales altoandinos. Ministerio de Desarrollo Sostenible y Planificación (Eds.), La Paz. SERNAP., 2006. Plan de Manejo, Área Natural de Manejo Integrado Nacional Apolobamba, La Paz, Unpublished results. SERNAP, 2014. Área Natural de Manejo Integrado Nacional Apolobamba, Características del Área. Available at: . (accessed 15.03.14.). Seibert, P., 1993. Ecología en Bolivia. Instituto de Ecología-Universidad Mayor de San Andrés, Revista de Ecología N 20.

Shapiro, S.S., Wilk, M.B., 1965. An analysis of variance test for normality (complete samples). Biometrika (3–4), 591–611. Soil Survey Division Staff, 1993. Soil survey manual. USDA Handbook 18. U.S. Government Printing Office, Washington, DC. UNESCO, 1973. Clasificación Internacional y Cartografía de la Vegetación. Ecology and Conservation. UNESCO. París. Urbano, P., 2001. Tratado de Fitotecnia General. Mundi-Prensa, Madrid. USDA, 2014. Keys to Soil Taxonomy, 12th Edition USDA-Natural Resources Conservation Service, Washington, DC. Vagen, T.G., Shepherd, K.D., Walsh, M.G., 2006. Sensing landscape level change in soil fertility following deforestation and conversion in the highlands of Madagascar using Vis-NIR spectroscopy. Geoderma 133, 281–294. Watanabe, F.S., Olsen, S.R., 1965. Test of ascorbic acid method for determining phosphorous in water and NaHCO3 extracts from soil. Soil Sci. Soc. Am. Proc. 677–678. Wilcox, B.P., Allen, B.L., Bryant, F.C., 1988. Description and classification of soils of the high-elevation grasslands of central Peru. Geoderma 42 (1), 79–94.