Floristic Diversity, Indicator and Suitable Species for Andean Livestock in the Sillapata Micro-Watershed, Acopalca

Abstract

Andean grassland ecosystems in Peru are characterized by diverse plant species adapted to environmental factors including weather, soil type, elevation, slope orientation, and soil moisture. This study evaluated the floristic composition, alpha diversity, indicator species, and suitable species for Andean livestock in the Sillapata micro-watershed, Junín region, Peru, across rainy and dry seasons. Data collection involved 100 m linear transects, and analyses included floristic composition and dissimilarity, Shannon-Wiener (H′) and Simpson (D) diversity indices, and the identification of indicator and suitable species using QGIS vs 3.28.14 and R software vs 4.3.2. Results revealed a total of 130 species classified into 74 genera and 23 families, with Asteraceae and Poaceae as the dominant families, exhibiting variations in richness and dissimilarity between control points and seasonal periods. Alpha diversity (H′) ranged from 2.07 to 3.1867, while Simpson’s index (D) ranged from 0.7644 to 0.9234. Six indicator species were identified, along with 11 families containing suitable species, predominantly Poaceae (38–60%), Cyperaceae (11–15%), and Asteraceae (3–9%). The findings indicate that the studied ecosystem exhibits a heterogeneous floristic composition with medium to low and variable diversity, influenced by seasonal climatic changes and the current grassland management regime, which involves rotational grazing with cattle adapted to high-altitude conditions.

1. Introduction

High-altitude ecosystems in the Peruvian Andes support a diverse array of plant species adapted to specific environmental conditions, including weather, soil type, elevation, slope orientation, and soil moisture [1,2,3]. This floristic diversity underpins the ecological balance and the provision of key ecosystem services, such as water regulation, soil conservation, biodiversity maintenance, and climate change mitigation [1,4,5].

Andean grasslands are shaped by a combination of climatic and soil factors, grazing animals, and fire regimes [6,7], with plant species exhibiting varied responses to. grazing intensity [3]. The Sillapata micro-watershed, a part of the Shullcas sub-basin, is representative of these Andean grassland ecosystems and is utilized for grazing various livestock, including cattle, sheep, alpacas, and llamas. Investigating the impact of climate variability (i.e., rainy and dry seasons) and grazing practices on the dynamics of this ecosystem is crucial for understanding its function, productivity, and the well-being of local communities [6].

This study aims to evaluate the floristic composition, diversity, and the presence of indicator species to assess the ecological condition of the grassland ecosystem and to provide a scientific basis for conservation and rangeland management strategies.

The assessment of ecological condition necessitates an understanding of grassland degradation, encompassing its drivers (e.g., overgrazing, land-use change), processes (e.g., declines in plant growth and primary production), and consequences (e.g., reductions in productivity and species richness) [2,7]. In this research, we examine floristic composition and alpha diversity, measured using the Shannon-Wiener index [8,9], with consideration of the influence of rainfall patterns on vegetation development and species diversity [10]. Indicator species, defined as those with distinct morphological traits and high indicative value for ecosystem health [11,12], were identified based on their sensitivity to environmental changes and specific habitat requirements [11,13].

Furthermore, we analyzed the suitability of plant species for livestock, categorizing them as suitable, poorly consumed, or unsuitable [13,14], and highlighted the ecological and economic importance of the Poaceae family [15,16,17,18]. While traditional field methods remain prevalent in grassland ecosystem assessments, the effectiveness of GIS technology and remote sensing for accurate plant species identification warrants further investigation [3,16,19,20]. Sustainable grazing management is essential for the conservation of suitable forage species, as appropriate grazing regimes can maintain biodiversity and ecosystem services [21].

This study focuses on determining the variation in floristic diversity, ecological indicator species, and suitable livestock forage species along an altitudinal gradient within the Sillapata micro-watershed of the Acopalca community in the Junín region of Peru. We employed Shannon-Wiener and Simpson indices and identified indicator and suitable species across nine sites during both rainy and dry periods [22,23,24].

2. Materials and Methods

2.1. Study Location

The study was carried out in 2024 within the Sillapata micro-watershed, a 672-hectare area within the Shullcas sub-basin. The micro-watershed is located in the territory of the Acopalca Peasant Community, in the province of Huancayo, department of Junín, Peru. Nine monitoring points were established across the study area based on differences in vegetation cover, determined by the variation in dominant species observed in the field (Figure 1). The surrounding areas have similar varying vegetation cover and are under a moderately rotational grazing regime with cattle and sheep, in other cases with alpacas and sheep.

The region exhibits two well-defined rainfall patterns: a rainy period, occurring between November and April, with rainfall ranging from 800 to 1150 mm, and a dry period, from May to October, with minimal rainfall (0–20 mm). Temperature fluctuations in the area include daily maxima from 12 to 21 °C and nighttime/early morning minima from −10 to 4 °C. The local population consists of pastoral families who depend on grazing mixed herds of cattle, sheep, alpacas, and llamas as the primary economic driver of their social and economic well-being.

2.2. Data Collection Techniques

Nine control points were established based on a spatial analysis of the micro-watershed using QGIS 3.28.14 software, according to differences in vegetation cover [22], and marked with wooden stakes for evaluations during rainy and dry seasons [22,24]. Transect starting points were georeferenced using a Garmin GPS 64s.

An initial survey identified control points and collected natural grass taxa. These taxa were identified by specialists from the “Andean Ecosystem” research group at the Universidad Nacional del Centro del Perú and compared with certified specimens in the group’s laboratory. Floristic presence was recorded along 100 m linear transects (using a cm-graduated winch), noting species whose leaves or stems contacted the measurement mark, or the leaf projection for tall plants.

Plant botanical identity, soil, mulch, rock, or moss (if no species was present) were recorded on field cards, along with control point, site name, and other details. Suitable species for Andean livestock were identified by compiling a list of suitable and unsuitable species from previous studies in the Peruvian Andes. A matrix was created to show the relationship between suitable species and livestock type [25].

2.3. Data Processing Techniques

Vascular floristic richness was counted, and data were organized into a double-entry matrix for each control point. After cleaning the data, it was integrated into a larger matrix for analysis using Rstudio software vs 4.3.2. Specific richness was determined by counting families, genera, and species. Dominant species were identified at each control point. Multivariate dissimilarity analysis [26] of richness was performed using the INDVAL method used by Dufrene & Legendre, 1997 [27].

Indicator species were identified using statistical analysis of ordered data [22]. An indicator value for each species was determined based on specificity (habitat exclusivity) and fidelity (habitat occurrence frequency), using the vegan package and the vegdist function in R with the Bray-Curtis method [27]. Indicator species had an indicator value >90% and a significance of p < 0.05.

Data matrices were tested for normality using an accumulation curve.

To calculate the dissimilarity, the Bray-Curtis distance was used in the following equation, widely used in ecology for abundance or cover data:

$${\mathrm{BC}}_{\mathrm{ij}}=1-{\displaystyle \frac{{2\mathrm{C}}_{\mathrm{ij}}}{{\mathrm{S}}_{\mathrm{i}}{+\mathrm{S}}_{\mathrm{j}}}}$$

(1)

where: BC_ij is the sum of the lowest values (see example below) only for those species common to both sites, S_i, S_j are the total number of specimens counted at both sites [27,28].

The Shannon-Wiener index (H′) was calculated using the equation:

H′ = −Σ (Pi ∗ ln(Pi))

(2)

where S = number of species, ln = natural logarithm, Pi = proportion of individuals of species i (calculated as ni/N), ni = number of individuals of species i, N = total number of individuals [28].

Simpson’s index (D) was calculated using the equation:

D = Σ (pi²)

(3)

where pi = proportional abundance of species i (ni/total individuals) [29].

Indicator species were also identified using the INDVAL method in R, with the equation:

VIS_ij = Specificity_ij ∗ Fidelity_ij ∗ 100

(4)

where: VIS_ij = indicator value of a species at a site, Specificity_ij = proportion of the j-th site, Fidelity_ij = proportion of individuals of the i-th species [30].

Indicator species were determined by an indicator value > 90% and significance at p < 0.05. Dominant and indicator species were mapped using free GIS software.

3. Results

3.1. Variation Dynamics of Floristic Diversity According to Altitudinal Gradient in the Sillapata Micro-Watershed in the Community of Acopalca

3.1.1. Species Richness and Composition

The study area comprises 130 species classified into 74 genera and 23 families. The most represented families (Figure 2) were Asteraceae (23% of genera, 19% of species) and Poaceae (20% of genera, 31% of species), indicating the dominance of grasses. Cyperaceae was the third most prevalent family (9% of genera and species). Cariophyllaceae had 8% of the genera but only 2% of the species. Gentianaceae and Umbelliferae also showed notable representation, with 6% and 4% of the species, respectively. The remaining families each accounted for less than 2% of both genera and species.

Two species with the highest abundance at each control point for the two seasonal periods were located on the map (Figure 3).

Figure 3 illustrates the two dominant plant species identified at each control point, demonstrating variations in species dominance associated with differences in altitude, soil type, and soil moisture. Where Fehum: Festuca humilior, Lapi: Lachemilla pinnata, Citar: Cinnagrostis tarmensis, Ferigi: Festuca rigidifolia, Nabra: Nassella brachyphylla, Hyta: Hypochoeris taraxacoides, Wenu: Werneria nubigena, Dimu: Distichia muscoide, Platu: Plantago tubulosa, Caec: Carex ecuadorica, Sciri: Sirpus rigidus, Plari: Plantago rigidus, Civic: Cinnagrostis vicunarum.

Figure 3 shows the two dominant species at each control point, which vary according to altitude, soil type, and soil moisture.

During the rainy season (Figure 3A), nine sampling points were identified, each with distinct plant associations determined by soil depth, water saturation, and slope; for instance, P1 was dominated by Festuca humilior and Lachemilla pinnata in oversaturated deep soils, while P7 featured Distichia muscoides and Plantago tubulosa, reflecting wetland characteristics. In the dry season (Figure 3B), points P1–P3, P7, and P8 maintained the same dominant species, but significant changes were observed in P4, P5, P6, and P9, such as the replacement of Cinnagrostis tarmensis and Lachemilla pinnata by Nassella brachyphylla and Festuca humilior in P4 due to heavy grazing, and the dominance of Hypochaeris taraxacoides in P5 and P6, favored by its semi-erect growth habit during the rainy season.

Details on location, elevation, and other characteristics are in Appendix A 

Table A4. Table of attributes of the first and second evaluation.

A	Point	Ord.	E	N	Z	de1	de1n	de2	de2n	Ind_sp
	P1	1	491,069	8,672,207	4156	Fehum	27	Lapi	21	Lapi
	P2	2	491,709	8,672,011	4328	Citar	42		0	Azcre
	P3	1	492,206	8,671,828	4328	Ferigi	28	Lapi	18	Lapi
	P4	1	492,195	8,671,565	4308	Citar	26	Lapi	23	Lapi
	P5	3	492,588	8,671,419	4385	Platu	15	Caec	14
	P6	3	493,206	8,671,106	4453	Caec	15	Wenu	13
	P7	3	493,415	8,672,203	4418	Dimu	29	Platu	17
	P8	2	493,505	8,672,010	4432	Citar	39	Scri	13	Welam
	P9	3	493,964	8,672,728	4543	Plari	34	Poca	13
B	Point	Ord.	E *	N *	Z	de1	de1n	de2	de2n	Ind_sp
	P1	2	491,069	8,672,207	4156	Fehum	23	Lapi	17
	P2	3	491,709	8,672,011	4328	Citar	18		0	Lapi
	P3	2	492,206	8,671,828	4328	Ferigi	27	Lapi	18
	P4	3	492,195	8,671,565	4308	Nabra	14	Fehum	13	Gebe
	P5	1	492,588	8,671,419	4385	Hyta	18		0	Lapi,
	P6	1	493,206	8,671,106	4453	Hyta	10		0	Lapi
	P7	1	493,415	8,672,203	4418	Dimu	29	Platu	14	Lapi
	P8	3	493,505	8,672,010	4432	Citar	18	Wenu	11	Wepy
	P9	3	493,964	8,672,728	4543	Citar	17	Civic	16	Gebe

* E and N correspond to the geographical coordinates, Z correspond to altitude in meters, de1 corresponds to the first dominant species, and de1n to the number of shares; de2, and de2n correspond to the second dominant species, both in the first and second evaluation, Ind_sp are the indicator species.

.

The multivariate analysis of richness dissimilarity between control points (

Table 1. Wealth dissimilarity values between control points, in two seasonal periods.

Control Points: P1 to P9 Sites
Assessment 1
P1	P2	P3	P4	P5	P6	P7	P8	P9
P2	0.7305699
P3	0.7025641	0.9285714
P4	0.4358974	0.5408163	0.7373737
P5	0.7435897	0.8571429	0.8888889	0.8484848
P6	0.7142857	0.7664975	0.7889447	0.7386935	0.6180905
P7	0.7978723	0.978836	0.9581152	0.947644	0.6230366	0.875
P8	0.8367347	0.5025381	0.919598	0.6482412	0.8592965	0.63	0.94791
P9	0.8877551	0.9187817	0.8894472	0.9095477	0.8190955	0.62	0.96875	0.77
Assessment 2
P1	P2	P3	P4	P5	P6	P7	P8	P9
P2	0.6701031
P3	0.6326531	0.744898
P4	0.7860963	0.7860963	0.8941799
P5	0.7040816	0.744898	0.8383838	0.9153439
P6	0.6020942	0.7172775	0.8134715	0.9021739	0.5958549
P7	0.6432432	0.8594595	0.9037433	0.9550562	0.657754	0.61538
P8	0.8162162	0.6216216	0.8823529	0.8764045	0.9037433	0.67032	0.90912
P9	0.6938776	0.6020408	0.7676768	0.6402116	0.8080808	0.76165	0.83957	0.74332

) showed significant differences between sites (dissimilarity > 0.6). In the rainy season, P7 differed notably from P3, P4, and P8, and P9 from P3, P5, and P8 (dissimilarity > 0.9). In the dry season, P7 remained distinct from P4 and P5, and P8 from P6 and P8. Dissimilarity values varied between the two seasons, with some increasing and others decreasing.

3.1.2. Shannon-Wiener (H′) and Simpson Diversity Index (D)

During the rainy season, the Shannon-Wiener index (H′) ranged from 2.0762 at the highest control point (P9) to 2.9497 at P6. Notable variation was observed at P5 (2.7841) and P6, followed by a decline at P7 (2.3137). At the remaining points (P1, P2, P3, P4, and P8), H′ values ranged between 2.1741 and 2.2722, reflecting medium diversity levels (i.e., values between 2 and 3.5).

In the dry season, H′ values increased, ranging from 2.2102 at P7 to 3.1837 at P6. However, this increase was not uniform; H′ at P7 decreased by 0.10 (Figure 4C) in the permanent wetland conditions, while values at other points, such as P2 and P8, increased more substantially. Numerical details are in Appendix A 

Table A3. Shannon-Wiener and Simpson diversity values in two periods, rainy and dry seasons.

Control Points	P1	P2	P3	P4	P5	P6	P7	P8	P9
First assessment
Shanon	2.272	2.177	2.223	2.230	2.784	2.950	2.314	2.174	2.076
Simpson	0.841	0.764	0.845	0.833	0.915	0.923	0.842	0.802	0.825
Second assessment
Shanon	2.519	3.079	2.353	2.576	2.976	3.184	2.210	2.990	2.477
Simpson	0.881	0.931	0.861	0.902	0.928	0.947	0.835	0.920	0.896

.

In the first evaluation, Simpson’s diversity index (D) varied from 0.7644 at P2 to 0.9234 at P6, indicating greater heterogeneity. Lower values (0.7644–0.8417″) were observed at P1, P2, P3, P4, P7, P8, and P9, while higher values occurred at P5 (0.9150) and P6 (0.9234). In the second evaluation, D values increased, except for P7, which decreased by 0.1 (Figure 4C).

Consistent with H′ results, Simpson’s index (D) values closer to 1 indicate lower diversity, aligning with the presence of 2 dominant species across the evaluated transects (Table A3 of the Appendix A).

3.2. Presence of Indicator Species of the Grassland Ecosystem According to Altitudinal Gradient in the Sillapata Micro-Watershed in the Community of Acopalca

Species within the 9 control points (Figure 5) were classified for both evaluations. Both graphs show a similar structure (3, 2, and 4 sites), but control point groupings differ between the rainy and dry seasons. Points in the first group tended to shift to the third or fourth group, reflecting changes in floristic composition between the two periods.

During the rainy season (

Table 2. Presence of indicator species in two seasonal periods: rainy and dry.

First Assessment					Second Assessment
Species	group	indval	p value	freq	Species	group	indval	p value	freq
Lachemilla pinnata	1	0.942	0.017	5	Deschampsia eminens	1	1.000	0.027	2
Azorella crenata	2	1.000	0.024	2	Cinnagrostis densiflora	1	1.000	0.036	2
Phylloscirpus rigidus	2	1.000	0.035	2	Eleocharis albibracteata	1	1.000	0.024	2
Wernera lamprophylla	2	1.000	0.033	2	Lachemilla pinnata	1	0.627	0.043	8
					Gentianella bellidofolia	3	1.000	0.011	3
					Wernera pygmaea	3	1.000	0.006	3

), indicator species were identified. Lachemilla pinnata was the indicator species for the first hierarchical grouping (P4, P3, and P1 in Figure 5 and Figure 6), represented by a dark green buffer on the map. Azorella crenata and Phylloscirpus rigidus were indicator species for the second grouping (P2 and P8), represented by a light green buffer. The third grouping (P5, P6, P7, and P9) had no significant indicator species (p-value > 0.05).

In the dry season, 6 indicator species were identified. Deschampsia eminens, Cinnagrostis densiflora, Eleocharis albibracteata, and Lachemilla pinnata were indicator species for the first hierarchical grouping (P5). Gentianella bellidofolia and Werneria pygmaea were indicator species for the third hierarchical grouping (P2, P4, P8, and P9, represented by green). P1 and P3 had no indicator species

3.3. Variation Dynamics of Suitable Species for Andean Livestock According to Altitudinal Gradient in the Sillapata Micro-Watershed in the Community of Acopalca

Above 3800 m in the Sillapata micro-watershed, the grassland ecosystem supports grazing by 4 domestic livestock species: cattle, sheep, alpacas, and llamas, along with wild vicuñas. Fifty-five suitable species (SS) for grazing have been identified. Less suitable species (LSE) were omitted, as they are considered complementary when SS are depleted or absent. Figure 7 shows that alpacas and sheep benefit from the largest number of forage species, while vicuñas, as wild species, rely on the fewest.

Eleven families with suitable species (SS) were identified (Figure 7). Of these families, 39% are preferred by alpacas, 36% by sheep, followed by llamas and cattle. Vicuñas prefer only 17% of the suitable species. Not all species are suitable for all livestock (

Table 3. List of suitable species specified for each livestock type.

Family	Species	Cattle	Sheep	Alpaca	Vicuña	Llama
Apiaceae	Lilaoepsis andina A.W. Hill		SS	SS	SS
Apiaceae	Lilaoepsis macloviana (Gand) A.W. Hill				SS
Asteraceae	Hypochaeris taraxacum (Meyen & Walp) Ball.		SS	SS
Asteraceae	Wernera nubigena Kunth.		SS	SS
Asteraceae	Werneria pygmaea Gillies ex Hook.		SS	SS
Cariophyllaceae	Cerastium imbricatum Kunth vel aff.					SS
Cyperaceae	Carex ecuadorica Kük.		SS	SS
Cyperaceae	Carex hebetata Boott.					SS
Cyperaceae	Carex mandoniana Boeck.			SS
Cyperaceae	Eleocharis albibracteata Nees & Meyen ex Kunth.		SS	SS	SS
Cyperaceae	Isolepis setacea (L.) R. Br. (Steud.) Boeckeler		SS	SS	SS
Cyperaceae	Scirpus rigidus (Steud.) Boeckeler		SS	SS	SS
Fabaceae	Lupinus chlorolepis C.P. Sm.		SS	SS
Gentianaceae	Halenia umbellata (Ruiz & Pav.) Gilg.	SS	SS	SS
Geraniaceae	Geraniium sessiliflorum Cav.		SS	SS
Juncaceae	Distichia muscoides Nees & Meyen.			SS
Juncaceae	Juncus stipulatus Nees & Meyen.			SS
Juncaceae	Luzula racemosa Desv.		SS
Plantaginaceae	Plantago tubulosa Decne.				SS
Poaceae	Agrostis breviculmis Hitchc.		SS	SS	SS
Poaceae	Agrostis tolucensis Kunth		SS	SS	SS
Poaceae	Anaterostipa hans-meyeri (Pilg.) Peñail	SS	SS	SS		SS
Poaceae	Anthoxanthum odoratum L.		SS	SS
Poaceae	Aristida adscensionis L.		SS	SS		SS
Poaceae	Aristida enodis Hack.	SS	SS	SS		SS
Poaceae	Bromus catharticus Vahl.	SS	SS	SS	SS	SS
Poaceae	Bromus lanatus Kunth.	SS	SS	SS	SS	SS
Poaceae	Cinnagrostis rigida Kunth.	SS				SS
Poaceae	Cinnagrostis densiflora (J. Presl)					SS
Poaceae	Cinnagrostis fuscata (J. Presl)					SS
Poaceae	Cinnagrostis jamesonii (Steud.)	SS				SS
Poaceae	Cinnagrostis rigescens (J. Presl)	SS				SS
Poaceae	Cinnagrostis spicigera (J. Presl)		SS	SS	SS	SS
Poaceae	Cinnagrostis tarmensis (Pilg.	SS			SS	SS
Poaceae	Cinnagrostis vicunarum (Wedd.)		SS	SS
Poaceae	Festuca huamachucensis Infantes	SS				SS
Poaceae	Festuca humilior Nees & Meyen.	SS	SS	SS
Poaceae	Festuca rigescens (J. Presl) Kunth.	SS	SS	SS		SS
Poaceae	Festuca rigidifolia Tovar.	SS	SS	SS		SS
Poaceae	Hordeum muticum J. Presl.	SS	SS	SS	SS	SS
Poaceae	Koeleria spicata (L.) Barberá.	SS	SS			SS
Poaceae	Nassella brachyphylla (Hitchc.) Barkworth		SS	SS		SS
Poaceae	Nassella meyeniana (Trin. & Rupr.) Parodi.		SS	SS		SS
Poaceae	Nassella mucronata (Kunth) R.W.Pohl.	SS				SS
Poaceae	Paspalum pygmaeum Hack.		SS	SS
Poaceae	Paspalum pilgerianum Chase.		SS	SS
Poaceae	Poa aequigluma Tovar	SS
Poaceae	Poa annua L.	SS
Poaceae	Poa gymnantha Pilg.	SS	SS	SS	SS	SS
Poaceae	Poa kurtzii R.E. Fr. vel aff.		SS	SS	SS	SS
Poaceae	Poa marshallii Tovar		SS	SS
Poaceae	Poa perligulata Pilg.		SS	SS
Poaceae	Poa spicigera Tovar vel aff.		SS	SS
Rosaceae	Lachemilla diplophylla (Diels) Rothm	SS		SS	SS	SS
Rosaceae	Lachemilla pinnata (Ruiz & Pav.) Rothm.		SS	SS	SS
	TOTAL SPECIES SS	19	36	39	17	24

); preference varies with plant morphology and the animal’s anatomy and physiology. Llamas and sheep consume the greatest number of unique species, while cattle, sheep, and alpacas share 3–4 pasture species. Poaceae has the highest species representation (38% in the first evaluation, 62% in the second), followed by Cyperaceae (15% and 11%, respectively) and a third family (13% and 5%). The remaining families each represent less than 4%.

Suitable species were ranked according to the preference of the three most abundant domestic livestock types (Figure 8). Rankings were compared for K = 2 and K = 3, and the graph generated by K3 was used to provide a more comprehensive interpretation. During the rainy season, the best grazing areas for cattle were sites P3 and P7 (first group), followed by P5 and P9 (second group), and then P2, P8, P4, P6, and P1 (third group). In the dry season, the ranking changed: P1, P6, P4, and P5 (first group); P2, P8, and P9 (second group); and P3 and P7 (third group).

For sheep, the rainy season assessment identified three groups of suitable species (SS): the first (P6 and P8), the second (P5 and P7), and the third (P1 and P4-P3). The second evaluation grouped the sites into two clusters: the first (P5, P6, P7, and P8) and the second (P1, P3, P4, P2, and P9). For alpacas, two groups were optimal in both evaluations. The first evaluation’s groups were: P2, P5, P1, P4, and P3 (first group) and P5, P6, P7, and P8, and P2, P9, P4, P1 and P3 (second group). The hierarchical grouping differed between evaluations due to an apparent inversion in SS presence between the rainy and dry seasons.

4. Discussion

This study demonstrates that Andean grassland ecosystems support communities with complex species richness and composition (Figure 1 and Figure 2), varying with geographic area, altitude, climatic conditions, grassland management, and soil type.

The identification of 23 families, 74 genera, and 130 species (Figure 2) does not define a unique richness for Andean grasslands. Compared to other studies (51 families, 21 genera, and 78 species in dry grassland in Junín; 81 species, 52 genera, 22 families in Tanta Lima; 22 families, 59 genera, and 83 species in Reserva Paisajística Nor-Yauyos; 32 families with 172 species in Ica; 14 families of vascular plants, 33 genera, and 41 species in Junín’s wetland; and 61 vascular species in Puno) [31,32,33,34,35,36], our findings simply confirm that each site’s species composition is adapted to its specific environmental conditions, climate, and anthropogenic influences. However, the study site’s current species composition appears stable and healthy.

The presence of the sole dominant perennial plant Cinnagrostis tarmensis at P2 in both assessments was likely determined by the inceptisol soil type, the slow decomposition of organic matter, and the steep, rocky slopes at high elevations with intense solar radiation—conditions also found at P8. Additionally, its resilience allows it to thrive in cold environments with well-drained soils, some solar exposure, and moderate levels of disturbance [35,36].

The shift in dominant species at P4 between evaluations (Figure 3) resulted from changes in dominant cover. During the dry season evaluation, this site had been grazed, with lingering effects from the rainy season. Conversely, during the rainy season evaluation, the area was rested, allowing erect perennials like Nassella brachyphylla and Festuca humilior to regain vigor (i.e., growth height and canopy cover). Consequently, the annual Plantago tubulosa and the perennial, prostrate Carex ecuadorica (cushion species) were overshadowed [36]. Dominant species changes at other sites also stemmed from the temporary appearance of species like Hypochaeris taraxacoides due to rainy season effects.

The dissimilarity in species richness among the 9 control points (Table 1) reflects differences in species composition, influenced by factors such as soil conditions and heterogeneous moisture levels across microhabitats defined by topography, elevation, slope orientation, and inclination [1,3,26]. Although distance and elevation differences between control points were relatively small, the observed increases and decreases in dissimilarity result from richness changes driven by the factors mentioned above.

Alpha diversity, measured by the Shannon-Wiener index (H′), ranged from 2.0762 to 3.1837 across both evaluations, indicating medium diversity. The slight increase in H′ during the dry season evaluation (ranging from −0.10 to 0.9) suggests that the rainy season promoted the presence of temporary species [22,27,37]. Simpson’s diversity index (D) ranged from 0 to 1, with values closer to 1 indicating lower diversity [38,39]. This aligns with P2’s single dominant species in both evaluations. Simpson’s index, based on the probability of two randomly selected individuals belonging to the same species, complements the H′ results.

These shifts likely result from rainfall variations between the two distinct Andean seasons, coupled with temperature fluctuations and grazing effects on plant community composition due to livestock feed preferences. Our H′ (2.55–2.90) and D (0.93–0.89) values are similar to those reported for Peruvian grasslands above 4000 m in Junín, suggesting that these grasslands typically have 1 or 2 dominant species, influenced by natural conditions and grazing. Similar results were obtained by Yaranga et al., (2018) in research carried out in other areas of Junín, as well as the D index from 05 to 0.71 in Aguada Blanca in Arequipa [40,41,42].

To analyze indicator species (Figure 5), control points were hierarchically clustered using the Bray-Curtis dissimilarity. The similarity in graphic structure reflects the floristic composition of each control point, which varied between evaluations due to site-specific characteristics and the influence of rainfall and temperature [2,5,13,40]. Consequently, the number of indicator species changed from 4 (typically prostrate and permanent) during the rainy season to 6 during the dry season, including four temporary species: Eleocharis albibracteata, Lachemilla pinnata, Gentianella bellidofolia, and Werneria pygmaea (Table 2). This dynamic indicates that indicator species in these grasslands are not fixed but change over time [26], findings on the varying indicator species status of grazed and ungrazed plants.

Figure 5 and Figure 6 show no clear relationship between indicator species and altitude or control point specificity during the rainy season evaluation. For instance, species in the green group 1 are separated by more than 200 m in elevation, and those in the orange group 3 are similarly dispersed [24,39,41]. This suggests niche sharing as a survival strategy, with species seeking soil nutrients and competing for resources for vegetative growth [24,39]. This interpretation also explains the changes observed in indicator species between evaluations.

Of the 23 plant families recorded in the assessed high-Andean grassland ecosystem, only 11 included species suitable for herbivorous livestock. Figure 7 and Table 3 summarize the forage species identified for the four main native Andean herbivores. The Poaceae family was the most dominant, corroborating previous findings [14], which highlight the ecological prominence of grasses in high-altitude ecosystems due to their physiological and morphological adaptations to cold, radiation-intense environments.

The cluster diagrams in Figure 8 illustrate that each of the three main livestock types in the micro-basin exhibits specific plant species preferences. Consequently, site similarity varies depending on the richness and composition of suitable species, which, in turn, is influenced by habitat conditions. The heterogeneity in suitable species composition results in distinct clusters reflecting livestock-specific preferences and differences between evaluations. These findings are crucial for informing rotational and complementary grazing management strategies to enhance ecosystem sustainability under mixed grazing regimes [42,43,44]. Changes in site clustering between evaluations are related to changes in site richness between the rainy and dry seasons, along with the grazing regime.

5. Conclusions

Andean grassland ecosystems exhibit a diverse and complex floristic community, composed of 23 families, 79 genera and 172 species, which show variations in their spatial distribution among control points. This variation was not strongly affected by altitude, as initially proposed, but was influenced by other conditions, such as humidity, soil type, slope orientation, precipitation, relief and grazing regime, which proved to be the most visible factors in the experience. Regarding alpha diversity, the Andean grasslands showed values within the medium range, which is in agreement with the results of studies carried out under similar conditions. On the other hand, the seasonal change from rainy to dry also showed changes in the richness of each site, with a greater number of temporary species in the rainy season, which also determined the presence of indicator species that were not unique to each site. Regarding the suitable species (D) for livestock, these corresponded to less than 50% of the dominant families, such as Poaceae and Cyperaceae, which shows that very few natural grass species were preferred by each Andean domestic species. This evidences the existence of a specificity in the natural grasses that support the diet of each type of livestock