Introduction

The seasonally dry forest (BES, for its acronym in Spanish)) in the Neotropics are distributed from northern Mexico to southern Brazil and make up 66.7 % of this type of vegetation in the world (^{Miles et al., 2006}). In Perú, BES are threatened and under-studied ecosystems and comprise several subtypes (^{Linares-Palomino, 2006}; Linares-Palomino and Pennington, 2007). The mountain and plain BES of the north coast are very fragile; they extend from the Peninsula Santa Elena, in southern Ecuador, to several regions of northwestern Perú, among which are Tumbes, Piura, Lambayeque and northern La Libertad, as well as the lower layer of the Marañón Valley. These two areas communicate through the Porculla Pass, although they are distinct phytogeographic units (^{Brack and Mendiola, 2004}). It is also known as the Tumbesina endemism region, which is shared by Ecuador and Perú, and it is considered one of the most important areas worldwide for this reason (^{Aguirre et al., 2006}).

Some studies have revealed the large number of endemic species present in the BES of northern and northwestern Perú, with about 26 % of them disseminated in various communities, such as Prosopis limensis Benth. (carob) and Colicodendron scabridum Seem. (zapote) (^{Sagástegui et al., 1999}). But not only the levels of diversity and endemism refer to plants but also to vertebrates such as Tremarctos ornatus Cuvier, 1825 (Spectacled bear), Lagidium Perúanum Meyen, 1833 (Vizcacha), Penelope albipennis Taczanowski, 1878 (Pava aliblanca) and Bothrops barnetti Parker, 1994 (Macanche), among others (^{Aguilar, 1994}).

The works on the BES, although not very abundant, have been significant, among which those carried out by ^{Weberbauer (1922}; ¹⁹³⁰; ¹⁹³⁶; ¹⁹⁴⁵) are particularly interesting, as he contributed floristic and phytogeographic aspects of the flora of Perú and elaborated a map of vegetation, hitherto unsurpassed, with emphasis on BES. In the past decade, the annotated list of woody species and plant formations in the BES of the Tumbesian Endemic Region of Ecuador and Perú (^{Aguirre et al., 2006}) as well as woody plants as a new online tool for taxonomic ecological and biodiversity studies stand out (^{Linares-Palomino and Pennington, 2007}).

Based on data from forest inventories, the structure and diversity patterns of the BES of two conservation areas of the Northwest Biosphere Reserve (Perú), the Cerros de Amotape National Park and the Tumbes Reserved Zone were analyzed and discussed (^{Leal-Pinedo and Linares-Palomino, 2005}). Also addressed were the spatial distribution patterns of tree species such as Eriotheca ruizzii (K. Schum.) A. Robyns (pasayo), Bursera graveolens Triana & Planch. (palo santo), Caesalpinia glabrata Kunth (charán) and Cochlospermum vitifolium (Willd.) Spreng. (polo polo), in a BES of the Cerros de Amotape National Park, in northwestern Perú (^{Linares-Palomino, 2005}). In the surroundings of Jaén (Cajamarca, Perú), in relicts of the seasonally dry tropical forest, eight inventories were made on woody vegetation, endemisms and conservation status (^{Marcelo-Peña, 2008}).

On the other hand, in order to design maps and classify vegetation in seasonal ecosystems of the dry forests of Piura (Perú), with data from floristic inventories, six groups of vegetation were identified, of which five coincided with types of vegetation described 100 years ago by ^{Weberbauer (1922}; ¹⁹³⁰; ¹⁹³⁶; ¹⁹⁴⁵) (^{La Torre-Cuadros and Linares-Palomino, 2008}). In the Ayabaca province (Piura), the edaphic characteristics and floristic composition of La Menta and Timbes BES were evaluated (^{Rasal et al., 2011}).

In the Lambayeque region (Perú), some studies have also been carried out at the BES. In the Chaparrí Ecological Reserve, which is part of the Tumbes-Chocó-Magdalena hot spot of biodiversity, a structural pattern and floristic inventory was developed, from which a fairly impoverished forest became evident, in relation to other BES in the region (^{Linares-Palomino and Ponce-Álvarez, 2009}). In this same Chaparrí Private Ecological Reserve, the phenology (vegetative development, flowering and fruiting) of 17 plant species that feed the white wing turkey (pava aliblanca), an endemic species seriously threatened (^{Martos et al., 2009}), was assessed.

Based on the above, in addition to the fact that the natural forests of Perú have been poorly studied, that the anthropic pressure that leads to their predation is increasing, that climate change continues to alter the patterns of plant distribution, among other factors (^{Steffen et al., 2011}), the present work was carried out with the aim to determine the floristic composition of the BES of Cerro Tres Puntas with the purpose of facilitating the elaboration of reforestation, conservation and utilization plans and, thus, contributing to the knowledge of the natural ecosystems of the country.

Materials and Methods

Study area

The study area is located between 6°14'16.23" S - 79°35'34.81" W and 6°12'46.46" S - 79°33'59.58" W, and includes Cerro Tierra Blanca (1 986 masl), Cerro Gato (1 545 masl), both belonging to the Salas district, and Cerro Tres Puntas (1 590 masl), from the Motupe district, in the province and region of Lambayeque, Perú (Figures 1a -b); these three hills cover around 2 975 ha.

Source: ^{Google Earth (2019)}.

Figure 1 a) Map of the Lambayeque region where Cerro Tres Puntas is located and the village of Pilasca is indicated with a star. b) Area comprising the Tierra Blanca, Cerro Gato and Cerro Tres Puntas of the BES (Salas-Motupe, Lambayeque, Perú). 

The Pilasca village is one of the poorest in Lambayeque (^{Gobierno Regional de Lambayeque, 2013}) where the maize agriculture (Zea mays L.), yucca (Manihot esculenta Crantz) and fruit trees, as well as cattle and goat breeding is for subsistence; it grows in small clearings in the forest, where there is no seed selection, fertilizers or control of pests and diseases. The peasants of Pilasca have used the forest species since ancient times for the manufacture of furniture, construction of rustic houses and pens as well as firewood as fuel, which is why Loxopterygium huasango Spruce ex Engl. (hualtaco), Handroanthus chrysanthus (Jacq.) S. O. Grose (yellow guayacán), Colicodendron scabridum (Kunth) Seem. (zapote), Prosopis limensis Benth. (carob tree) and Cedrela sp. (cedar) practically no longer exist.

In the geological aspect, the study area corresponds to the litho-stratigraphic unit Salas Group (Oi) of the Paleozoic or Salas Formation, (Pi-s) Lower Ordovician (480 to 470 million years) (^{Gobierno Regional de Lambayeque, 2013}; ^{López, 2013}).

The area in which the research was conducted is a seasonally dry semi-dense forest ecosystem of hill and mountain (Bes-cm) (^{Minam, 2019}). The data was collected in a place between 373 and 720 masl; the extreme point A corresponds to the entrance of the Pilasca village (373 m), where the evaluation began, to the extreme point E (720 masl), where it ended, whose topography is rugged with huge boulders and deep ravines, dry and scarce stony clearings. The temperature, relative humidity and precipitation data were taken from the Jayanca CO-Meteorological Station (LA Vic) (78 masl), the closest to Cerro Tres Puntas de Pilasca, from which the following data were recorded: maximum and minimum temperature and relative humidity of March 2018, 34.0 and 19.0 ºC and 78.4 %, respectively, while in June 2018, 26.2 and 15.0 ºC and 79.1 %, respectively. In these four months of evaluation, the precipitation was 0.0 mm day^-1 (^{Senamhi, 2018}).

Sampling

Field work was carried out from March to June 2018 and the method proposed by ^{Gentry (1982}; ^1995a) was used for the sampling of woody plants, with substantial modifications from the difficult access and dangerous geography of the land. A linear sampling was applied taking as a baseline a relatively safe trail that the locals use to occasionally access the top of Cerro Tres Puntas de Pilasca (Figure 2).

Figure 2 Study area showing the Cerro Tres Puntas (Salas-Motupe, Lambayeque, Perú). 

Four sampling units were established in a total area of 3 801 m, designated with points A (373 masl) -B (456 masl), BC (565 masl), CD (677 masl) and DE (720 masl), which covered 958 m, 961 m, 968 m and 914 m, respectively. From the center of the trail were measured 15 m to the right and 15 m to the left, making a total of 30 m which allowed to evaluate about 114 030 m², with a GPS Garmin 60CSX and a measuring tape. All woody individuals whose DBH (1.3 m from the base of the trunk) was ≥ 5.0 cm were recorded, taken as a measure of the basimetric area with a measuring tape.

For individuals with branched trunks, the diameter of each branch was measured at a height of 1.3 m from the base of the trunk, with a diametric tape Forestry Suppliers, Inc. (Modelo 283D/5M). As the species evaluated were known, only botanical samples were taken for registration and deposit in the Herbarium of the Universidad Nacional Pedro Ruiz Gallo de Lambayeque (HPR Herbarium). Likewise, herbaceous species, present in the collection area, were picked up and pressed for subsequent taxonomic identification. For families, the phylogenetic postulates of the Phylogeny Group of Angiosperms were followed (^{APG, 2009}; ²⁰¹⁶).

Results

Discussion

The sampling area is classified as seasonally dry mountain forests on the north coast that covers the BES over 500 masl, and covers the western flanks of the Andes in the Tumbes, Piura and Lambayeque regions to La Libertad, including part from Cajamarca (^{Linares-Palomino and Pennington, 2007}). However, as indicated by ^{La Torre-Cuadros and Linares-Palomino (2008)}, this altitudinal limit is quite variable, debatable and depends on several factors. As sampling began at 373 masl, it engulfed a small area of the plain BES on the north coast that covers the BES below 500 masl along the western flanks of the Andes from Tumbes to La Libertad (Linares- Palomino and Pennington, 2007). This would be the reason why, under the conditions of Pilasca, there was little presence of carob (P. limensis) and only a single specimen of zapote (C. scabridum) and others of vichayo (Beautempsia avicennifolia (Kunth) Gaudich.); the latter outside the sampling area.

These authors also pointed out, that the dry mountain BES at the north coast as those of the interandine system of the Marañón river, are the richest in woody species with 193 and 184, respectively. Such information has been confirmed by the study conducted in the seasonally dry tropical forest (BTES) of the Marañón valley, in northern Perú, in which 440 woody species, and 143 (33 %) endemic, as well as two endemism centers were identified within the valley, with 56 % of species and 78 % of genera found in Peruvian BTES (^{Marcelo-Peña et al., 2016}).

In the analysis of species diversity patterns ─4660 woody plant species─ of BTES from Latin America and the Caribbean, it was determined that the outstanding floristic rotation observed implies the need to establish multiple conservation areas to protect such a great flora diversity (^{DRYFLOR et al., 2016}).

Even though an annotated list of the BES of Ecuador and other related studies (^{Svenson, 1946}; ^{Gentry, 1995b}; ^{Aguirre et al., 2006}) indicated that the BES of Ecuador and Perú, especially those of the coast, formed a single geographical unit, with a high number of endemic species (Tumbesino Endemic Center), the few species (17), genera (17) and families (10), inventoried in the Cerro Tres Puntas de Pilasca, would demonstrate a high degree of depredation of the original flora. This would confirm what has already been observed in other BES such as Chaparrí in Lambayeque (^{Linares-Palomino and Ponce-Álvarez, 2009}; ^{Martos et al., 2009}) and La Menta and Timbes in Piura (^{Rasal et al., 2011}).

On the other hand, other factors such as the diversity and density of species and precipitation gradients, which decrease as the BES extend south, should be considered, including the Lambayeque and La Libertad regions.

Starting from the rugged and dangerous topography of the land, the linear sampling that was used in the present investigation made it possible to evaluate a large area that covered 11.4 ha, a sampling that has already been used in the rapid botanical inventories (IBR) in which it woody vegetation, endemisms and the conservation status of the BES of Jaén (Perú) were addressed (^{Marcelo-Peña, 2008}), based on the methodology of ^{Hawthorne and Abu-Juam (1995)} and ^{Gordon et al. (2004)}.

In spite of the interest in BES in recent years (^{Espinosa et al., 2012}; ^{Aguirre and Geada-López, 2017}), in the few works carried out in Ecuador and Perú it was observed that the sampling methods were very variable between the use of rectangular transects and linear sampling, as well as in the size of the sample units, from 0.25 to 6.5 hectares and in the DAP of the sampled species, from ≥ 1.0 to 10 cm. The difference between the number of individuals, species, genera and families should be added, which is usually higher in protected natural areas such as the Cerros de Amotape National Park (^{Linares-Palomino, 2005}) and very depressed in other historically depredated areas, as the Chaparrí Ecological Reserve (Linares-Palomino and Ponce-Álvarez, 2009).

From all of them, the closest to the described one here is that used in La Menta and Timbes in Piura (^{Rasal et al., 2011}), where, despite the fact that woody species with only DAP ≥ 1.0 cm were registered, the number of inventoried individuals was significantly smaller, although there was quite similarity in the number of species, genera and families.

In their study at the Chaparrí Ecological Reserve, ^{Linares-Palomino and Ponce-Álvarez (2009)} pointed out the complication of establishing comparisons of inventories in the seasonally dry forests of the region, from the great variation between the sampling methodologies used, altitudinal floors, total sampling area, DAP of the sampled specimens, conservation status and depredation of the sampling area at the time of its realization, and even in the prioritization of the information collected, among other factors.

The previous arguments are confirmed by several studies in Perú such as those in the forests of Jaén, Cajamarca (^{Marcelo-Peña, 2008}), La Menta and Timbes in Ayabaca, Piura (^{Rasal et al., 2011}),Biosphere Northeaster Reserve of Perú (Cerros de Amotape National Park and Tumbes Reserved Zone) (^{Leal-Pinedo and Linares-Palomino, 2005}), Piura coastal plains (^{La Torre-Cuadros and Linares-Palomino, 2008}), Cerros de Amotape National Park (^{Linares-Palomino, 2005}) and Chaparrí Ecological Reserve, Lambayeque (Linares-Palomino and Ponce-Álvarez, 2009); however, it has not been possible to establish rigorous patterns of comparison between all this valuable information and the study carried out at the Cerro Tres Puntas de Pilasca.

In the plant formations of the BES of Ecuador and Perú, lists of many woody species were made for the deciduous forest, semi deciduous dry forest mixed with evergreen species, deciduous forest and the dry thorny thicket (^{Aguirre et al., 2006}; ^{Aguirre and Kvist, 2009}); however, in the actual work, only eight species were recorded: P. limensis, C. lutea, P. excelsum, C. glabrata, V. macracantha, E. smithiana, B. graveolens and L. huasango, which somehow, would indicate the strong pressure that Cerro Tres Puntas de Pilasca has suffered in recent years, even though it was possible to find sporadic individuals from Bougainvillea peruviana Nees & Mart., Bauhinia aculeata L., C. vitifolium and C. scabridum although they were outside the linear transect.

The trend of distribution of heights that does not fit exactly with the inverted “J”, described by ^{Lamprecht (1990)} and ^{Günter et al. (2011)}, would exhibit the succession characteristic of the forest from the constant and selective disturbances exerted by the local inhabitants, keeping it in a secondary succession stage (^{Rasal et al., 2012}). Although the referred experience did not include individuals with DAP <5.0 cm, although with a plant height <4.9 m, its results allow a thorough understanding of the regeneration process of the species present in the higher diameter and height classes, as recommended by ^{García et al. (2010)} studying the structure and floristic diversity of two natural forests in Cauca, Colombia.

In BES La Menta and Timbes of Piura (Perú), about 55.0 % of the sampled individuals had a DAP between 0-10.0 cm with some L. huasango and B. graveolens trees with DAP between 50-60 cm for La Menta, and Ceiba trischistandra Bakh. and E. ruizii with DAP between 80-90 cm for Timbes (^{Rasal et al., 2011}). Regarding these five indicated species, at BES Cerro Tres Puntas only E. ruizii was found. Likewise, of the 410 registered individuals, 196 had DAP ≥ 20.0 cm, where three species, V. macracantha, C. iguanaea and E. smithiana contributed 58.8 % of the total.

Very few studies conducted in the dry forests of Perú have collected information or performed analyzes related to AB (Basimetric Area), IVI (Importance Value Index) and VIF (Importance Value for the Family); however, ^{Linares-Palomino and Ponce-Álvarez (2009)} calculated an AB of 11.74 m², in which E. ruizii and L. huasango stood out with 4.40 and 4.12 m², respectively in the Chaparrí Ecological Reserve (Lambayeque). In the dry forests of La Menta and Timbes (Piura), ^{Rasal et al. (2011)} estimated an AB of 128.86 m² in La Menta, in which L. huasango (45.97 m²) and B. graveolens (34.69 m²) stood out, and in Timbes an AB of 196.63 m² was obtained, with special contribution from E. smithiana (47.16 m²) and C. trischistandra (38.88 m²). In both cases, the AB was much lower than the one reported in the current study in which the AB was 343.86 m² ha^-1, which indicated that despite the fact that the Cerro Tres Puntas forest has suffered a strong anthropic pressure, it has not been as dramatic as what happened in Chaparrí and La Menta and Timbes, where even the Chaparrí forest was defined as an impoverished version of the seasonally dry forests of the region (Linares- Palomino and Ponce-Álvarez, 2009).

In this same community, four species recorded IVI numbers over 30.0: E. ruizii (58.22), L. huasango (44.93), C. lutea (33.76) and B. graveolens (30.08), which belong to the Malvaceae, Anacardiaceae, Boraginaceae and Burseraceae families, respectively (^{Linares-Palomino and Ponce-Álvarez, 2009}); while in La Menta, B. graveolens (50.0), L. huasango (43.0) and C. lutea (36.0) had the highest values, and in Timbes, C. lutea (62.0), E. smithiana (40.0) and C. trischistandra (25.0) (^{Rasal et al., 2011}); also, as mentioned above, in the work in the dry forests of the province of Loja, ^{Aguirre et al. (2013)}. defined the five species with the highest IVI: C. trischistandra, Simira ecuadorensis (Standl.) Steyerm., Tabebuia chrysantha G. Nicholson, E. ruizii and Terminalia valverdeae A. H. Gentry, although no quantitative information was provided. When comparing these results with those of the present study, notorious differences are observed since the species that stood out were F. obtusifolia (49.34), V. macracantha (46.75) and B. sulcata (41.57), which were relevant in other seasonal forests. It is possible to hypothesize that the depredation of the species of these forests has not followed a definite pattern, and this would have been subject to the preference for the most valuable timber species and to the same protection that would have given them the local inhabitants, where, certainly, on the Cerro Tres Puntas, predation has been more selective.

The VIF results of the research described for the Cerro Tres Puntas de Pilasca, in which the families Fabaceae (111.86), Moraceae (46.74) and Lauraceae (45.33) stood out, disagree widely with those referred to Chaparrí (^{Linares-Palomino and Ponce- Álvarez, 2009}), La Menta and Timbes (^{Rasal et al., 2011}), since none of the families listed in these works has been recorded in the study carried out in the area of interest.

Conclusions

The study in Cerro Tres Puntas de Pilasca (Salas-Motupe, Lambayeque) contributes to the knowledge of the natural ecosystems of Perú, specifically that of seasonally dry forests, by determining their floristic structure and composition. In the context of biodiversity, the results obtained describe its plant richness, which is a valuable contribution to the conservation and management of these resources, and ecological data as a preliminary approach to the dynamics of the BES in the region.

It is suggested to extend the studies to other areas in the same Tres Puntas Hill, in adjacent geographical areas, in other BES of Lambayeque such as those of Tongorrape, Ñaupe and Tocmoche and even increase the altitude range of sampling. Also, conduct carbon capture assessments and identify the species that would most contribute to mitigate climate change and thus, to promote their large-scale multiplication for reforestation purposes.