Study site

The Monserrat island (25^o 41’ 00” N and 111^o 03’ 00” W) has an area of 19.9 km². It is located 13.7 km from the Baja California peninsula with an elevation of 223 masl (^{Álvarez-Castañeda and Cortés-Calva 2002}, Figure 1). A mountain chain runs along the island, which forms two slopes: the western one is more pronounced and the eastern slope is steeper. This tendency is best observed in the southern part of the island, where the highest hill is located (223 masl). The island presents stony beaches; the only sandy beach is located in the north, with an extension of approximately 1 km in length.

Figure 1 Location of Monserrat Island in the Gulf of California, Mexico. 

The soil of the island is poorly developed and only a few centimeters thick; only in the areas where streams flow to the sea, there is sandy soil with an organic substrate of approximately 10 cm thick, due to the organic matter deposited in these areas. The streams are intermittent contingent upon flows occurring during the rainy season.

Marine terraces composed of marine calcareous sediments are present on the island. These are flat areas with slopes of no more than 4 degrees and rocks ranging from 0.3 to 2 m in size, these are located in high elevations of the island that cut abruptly, forming abyss. The average annual rainfall is 185 mm occurring in summer and to a lesser degree in winter. The average annual temperature is 33.02^o C. The vegetation of the island corresponds to that of the Sonoran Desert, dominated by thorny scrub and columnar cactaceae (^{Case et al. 2002}).

Vegetation and terrain survey

Based on aerial photographs (scale 1: 70000), which were georeferenced in the field and 15 excursions on foot made on the island, the geomorphic structural elements were identified according to physical and structural differences of soil, slope characteristics and terrain profiles. The vegetation sampling sites were selected based on aerial photographs installing randomly 66 transects of 5 × 40 m (13 200 m²) taking into account their geographic orientation, slope, altitude, slope exposure, and geoforms identified (Figure 2).

Figure 2 Map of the Monserrat Island with routes traced on the island (yellow lines), where the vegetation transects were established (squares); the quadrats indicate the places where climatic sensors were placed for recording of precipitation, temperature and humidity (HOBBO sensors). 

In each transect, the vegetation present was recorded registering the height of plants and vegetation cover. The height was taken from the ground (base of the trunk) to the highest part of the vegetation cover. The measurement of the vegetal cover (canopy) was made considering its greater diameter and the smaller diameter; then, with the formula of the ellipse = [(π) (Larger diameter) (smaller diameter) (0.25)], the area covered by the vegetation was calculated. The form of growth was recorded grouping the plants in trees (T), shrubs (S), perennial forbs (PF), succulents (SU), and climbers (C). With the use of Geographic Information Systems (GIS) and the software ArcView ver. 3.2 the analyzed information of the aerial photos, geoforms, and the types of vegetation were delimited.

Soil analysis

Soil samples (0-20 cm) were collected at the beginning of each point of transects for chemical and physical analysis. Soil texture was determined by the hydrometer method (^{Bouyoucos 1962}), and the results were used to calculate the percentages of sand, silt, and clay. Organic matter was determined based on loss-on-ignition at 450^o C. Soil-water extracts (1:5) were set for determination of electrical conductivity (EC) and pH.

Environmental and climate analysis

For the environmental and climatic analysis, the slope (measured in degrees), orientation, solar exposure, altitude, percentage of rockiness, and geolocation (at the beginning and the end of the transect) were taken into account for each transect. To record the climatic variables, climatological sensors were established, which registered rainfall (rain gauge) and temperature and humidity of the selected sites (HOBBO® sensors, MicroDaq, Contoocook, NH, USA).

Data analysis

Species richness on the island was calculated as the average number of species per stand. The Shannon index H'=-∑pilogpi for the relative species evenness, and Simpson index C=∑pi2 for the relative concentration of species dominance were calculated for the vegetation present in each geoform based on the relative cover (pi) of species (^{Pielou 1975}). The Importance Value Index (IVI) for plant species was determined as the sum of the relative frequency, relative density, and relative dominance (^{Cottam and Curtis 1956}). For the canonical correspondence analysis (CANOCO software, version 4), a total of 56 plant species were analyzed (Importance Value Index: IVI), 16 environmental variables in 66 sampling sites. An exploratory analysis was carried out with all biotic, abiotic, and environmental variables, to determine the association between them.

Overall vegetation distribution

The geoforms that were recorded on the island were: beaches, plateaus, ravines, hillsides, marine terraces, abrasion terraces, and steep slopes. The vegetation of the Monserrat Island covers most of the terrain, except the rocky beaches. The vegetation is predominantly a xerophilous scrub, in some places, the herbaceous species of the genus Marina, Atriplex, and some grasses prevail, while in other areas the predominant shrub genus are: Gossypium, Jatropha, Encelia, Simmondsia, Fouquieria, and Bursera trees, which are more abundant in areas with steep slopes.

In places with very steep slopes, the vegetation is scarcer, due to the exposure, since the exposure of the northern slope has more abundant vegetation and greater cover. The plant community is dominated by perennial and shrubby herbaceous plants, where Jatropha cuneata is the dominant species throughout the island (Table 1).

A sarcocaule xerophilous scrub develops mainly in ravines and bottoms of streams, where Lysiloma candida, Bursera microphylla, Jatropha cuneata, and climbers (Antigonon leptopus, Passiflora arida, Janusia californica, and Matelea pringlei) grow. In the bottom of ravines, the vegetation is more abundant, presents a greater number of species, of which some are unique to the study area; the arboreal form of growth is the most abundant with Lysiloma candida and Olneya tesota attaining heights of about 5 m with aerial cover greater than 5 m².

Chamaephytes were the predominant life form and constituted 32% of the recorded flora, followed by therophytes (31%) and hemicryptophytes (17%). In this context, chamaephytes, hemicryptophytes, and cryptophytes altogether constituted the main bulk of the floristic structure of each of the recognized geoform unit.

We recorded 114 species, belonging to 92 genera and 38 families of vascular plants. The most abundant families were Asteraceae, Euphorbiaceae, Fabaceae, Poaceae, and Cactaceae (Table 1). The genera with the greatest diversity of species were: Euphorbia, Porophyllum, Opuntia, and Perityle. Herbaceous perennials and shrubs were the more representative forms of growth.

The desert scrub is the dominant vegetation on the island, in some places prevailing herbaceous plants such as Marina, Atriplex, and some grasses, while in other areas, the dominant plants presented are shrubs such as Gossypium, Jatropha, Encelia, Simmondsia, and Fouquieria. The Bursera trees are more abundant in areas with little steep slopes (less than 25^o). On areas with steep slopes (greater than 45^o), the development of vegetation is scarce. Northern exposure slopes have denser vegetation, with higher canopy cover values. Overall, the island is dominated by herbaceous perennial and shrubby species, where Jatropha cuneata is the dominant one.

The dominant species in ravines are Lysiloma candida, Bursera microphylla, Jatropha cuneata, and lianas such as Antigonon leptopus, Passiflora arida, Janusia californica and Matelea pringlei. In streams, vegetation is even more abundant, they have a greater number of species, some of which are unique to these areas; the dominant species are Lysiloma candida reaching about 5 m in height and Olneya tesota with a height of 4-5 m. In alluvial fans with a layer of higher matter contents, the vegetation is predominantly constituted by shrubs, and dominated by Jatropha cinerea.

The coastal scrub develops on sandy soils of the dunes at the north of the island, mainly composed of Jouvea pilosa and Atriplex barclayana, and by several annual grasses growing in the rainy season. The scrub takes varying composition and growth forms composed mainly by desert scrub and sarcocaule thickets. Marine terraces have a community of undeveloped desert scrub, averaging 50 cm tall and 1.60 m² canopy cover values, dominated by Fouquieria diguetii and Jatropha cuneata. Individuals of Ferocactus digueti were recorded only in these areas. Rhizophora mangle was recorded on the north coast of this island.

Vegetation transect study

A total of 5607 individuals belonging to 51 species were recorded in the 66 transects, covering an area of 13 200 m². Shrubs and herbaceous perennials were the most abundant species (18 and 15, respectively), followed by succulent and trees, with nine and seven species respectively, and climbing plants (two species). The most abundant species in all geoforms was Jatropha cuneata, with 503 individuals recorded; Encelia farinosa develops in plateaus with little slope (less than 10^o), being more abundant in the north part of the island; Stenocereus gumosus develops in patches, covering large areas, mainly on the slopes with little inclination (15-20^o) to the east and north side of the island; Lysiloma candida was typical of streams and ravines, being one of the few tree species in the island.

The Shannon-Wiener index showed high diversity in the island (H = 2.95, Table 2); the distribution of vegetation is heterogeneous as well as the growth form, it varies depending on environmental conditions prevailing in each microenvironmental geoform. The highest important values recorded include the arboreal species: Fouquieria diguetii, Olneya tesota, Jatropha cinerea, Bursera microphylla, Cercidium microphyllum, Stenocereus thurberi, Bursera hindsiana, Colubrina viridis and Maytenus phyllanthoides, developed above the middle vegetation, and characterizing the desert scrub on the island. Pseudonodricism was observed in some of these species (J. cinerea, B. microphylla, C. microphyllum, and B. hindsiana). Pachycereus pringlei, Stenocereus thurberi, Fouquieria diguetii, Olneya tesota and Jatropha cinerea are the dominant species and protrude above the rest of them in the vertical stratum, however, in the glens, Lysiloma candida reaches heights up to 5 m tall, with high rates of regeneration.

Table 3 shows results of the canonical correspondence analysis, cumulative variance, and Monte Carlo permutation of the biotic variables of the plant species on Monserrat Island. Analysis of the structural composition of ravines showed more abundant vegetation, where the height average was 1.6 m, and the canopy cover was 4.19 m² and is dominated by Lysiloma candida, J. cuneata, Colubrina viridis, Cerciduim praecox, and C. microphyllum; during the rainy season, climbing species such as Antigonon leptopus and Cardiospermum corindum grow abundantly on the canopies of other species in these sites. The other geoforms (beaches, plateaus, marine terraces, and slopes), the vegetation reaches an average height of 93.1 cm and a canopy of 2.6 m².

Figure 3 shows the canonical map of correspondence. The total inertia obtained was 9.32, indicating that the data present a high heterogeneity in their values (both biotic and environmental variables) and that they are suitable for a canonical correspondence analysis, the results in the Monte Carlo permutation test (0.30) indicate a high correlation between the variables analyzed (biotic vs environmental).

Figure 3 Analysis of canonical correspondence with plant species, importance value index (IVI) and the most significant environmental variables in 66 vegetation transects on the Monserrat Island, Gulf of California, Mexico. 

The canonical correspondence graph shows a differential gradient of the environmental variables of slope, soil, exposure, and altitude; three large groups were detected, the slope is the most influential in some plant species; while soil and exposure variables have less effect on plant species. However, more species are associated with these; the height has a low influence among all the environmental variables, but it has an influence on plant species of ecological relevance such as Olneya tesota, which grows in high places of the island, mainly in geoforms of plateaus and places with gentle slopes.

The most evident association was between the bare soil and hillside exposure, the latter characterized by the north and south exposures, given that the island has an orographic composition in which most of the slopes are oriented towards these courses, being the south-facing exposures with poor coverage, due to their orientation characteristics (southern exposure slopes are more “dry” than those on the north side).

Overall vegetation distribution

Out of the 114 plant species registered on the Island, 34 (28.9%) are endemic to the Baja California peninsula and nearby islands and two species to the islands; five genera are monospecific and 16 are reported as having only one species for the island and the Baja California peninsula (^{Case et al. 2002}). Two species found on the island and the peninsula are listed in the NOM-059-SEMARNAT-2001, while the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) listed 12 species for the island, 11 of them are cacti. Due to the abrupt terrain in most of the island, the ravines have an important role in the development of the vegetation because it is in these sites where most of the vegetation grow.

Jatropha cuneata appears to be generalists because it was found in almost all geoforms. This means that a threshold for extreme exposure, nutrient impoverishment of soil, did not exist on the island for this shrub. J. cuneata presents a rapid response to humidity since they produce their leaves after the first rainfall, regardless of its seasonal occurrence (^{Maya and Arriaga 1996}). Thus, this plant appears to depend on the combination of water availability and high temperature to complete its phenological cycle and produce sexual structures.

Vegetation transect study

A moderate level of species richness was found in the island. The level of diversity on this island surpasses the number of species found in smaller neighboring xeric islands of the Gulf of California receiving less than 200 mm of rainfall, such as Alcatraz (55 plant species), Dátil (102 plant species), Patos (13 plant species), and Cholludo (28 plant species), but it presents a comparable richness to the San Esteban Island (41 km², 122 plant species) (^{Wilder et al. 2008}).

Different from most other descriptions of desert vegetation in islands in the Gulf of California, we have endeavored here to analyze the vegetation patterns on Monserat Island taking into account the geoforms detected in this ecosystem. Exposition, altitude, soil characteristics, and slope seemed to increase the environmental variation across the island, which resulted in a rather distinctive species’ segregation. Also the causal relations between site characteristics and vegetation distribution were determined by standard multivariate techniques. For instance, slope was far more reliably measure for describing vegetation distribution than altitude, although, the latter variable influences Olneya tesota, a plant species ecologically relevant in this landscape because it creates microhabitats that benefit many species of flora and fauna (^{Carrillo-García et al. 1999}), and therefore is a valuable nurse species (^{Suzán et al. 1996}). Although the highest point in Monserrat Island was only 223 m, lower temperatures at higher altitudes imply a greater adaptation of this tree to cooler temperatures.

Euphorbia magdalenae was strongly associated with slope, which agrees with description of this shrub in islands of the Gulf of California where this plant grows on rocky hillsides or ridge crest of these islands (^{Felger et al. 2012}). The distribution of both Jouvea pilosa and Marina parryi was influenced by soil characteristics. The grass Jouvea pilosa grows on dunes, which means that it prosper on deeper soil (^{Romero-Lopez et al. 2006}) than the rocky spots on the island. Rhizophora mangle inhabits the north coast of this island and its growth is due to the very high salinity of its stands, shallow water table, and poor aeration and drainage.

The high species diversity on the island seems to be related to the low salinity of its soil and/or heterogeneity of the topography, which supports clumps of grasses, trees, succulent life forms, and ephemerals during the rainy season. The vegetation and environment of Monserrat Island are alike to those described for other desert islands of the Gulf of Baja California (^{Felger et al. 2012}). All are dry and dominated by maritime exposure and subjected to isolation which has fostered biological endemism. Salvia platychelia and Ferocactus diguetii are of particular note as they are narrow-endemics, so far having only been recorded in the sea of Cortez Islands.

The Monserrat Island presents a topography (geoforms) that give a particular character to the development of its vegetation, which is not different from that which is developed in most of the peninsula of Baja California Sur, Mexico, being considered, together with most of the other islands of the Gulf, as short-term and localized disturbances (^{Tershy et al. 1997}). The geographic isolation of this island has caused, to a greater or lesser extent, biodiversity to be genetically isolated (^{Salm and Clark 1994}), which has favored high endemism of plants. Since the abiotic and biotic factors of the islands of the Gulf of California are unique to each island, it was not surprising to find a great diversity of flora on the island under study, as has occurred on other islands in this region (^{Wilder et al. 2008}).

In this context, the plant species present on the island are sensitive to man damage and natural events, such as hurricanes. Although the literature that relates geomorphology and flora is still scarce, their relationship is evident, which implies a greater understanding of ecosystems and the development processes of plant communities. In this sense, ^{Radhakrishnan et al. (1998)} and ^{Zinck (2016)} point out the importance of geoforms in the study of vegetation and its adaptation and development processes as a new strategy for research and conservation of island ecosystems. On the other hand, ^{McAuliffe et al. (1991}, ¹⁹⁹⁴⁾ showed the importance of the relationship between vegetation and geoforms in continental regions.

The results showed that the ravines and streams of the island present the highest diversity indexes, abundance of species and morphometric averages, as well as unique species, it is worth mentioning that these geoforms are the ones that present greater anthropic activity, mostly by fishermen, who use the resource, mainly for fuel. The damage that these activities have on these geoforms and the vegetation they sustain is not known, so these can be considered as critical areas. This study serves as a basis for understanding the processes of adaptation and distribution of plant communities under particular geomorphic conditions, as is the case of the islands of the Gulf of Cortez. Thus, abiotic and in particular geological features are important to consider in planning conservation projects (^{Groves et al. 2002}, ^{Strasberg et al. 2005}).