Skip to main content
NCBI home page
Current Zoology logoLink to Current Zoology
. 2018 Jan 9;64(6):671–679. doi: 10.1093/cz/zoy001

Changes in potential mammal diversity in national parks and their implications for conservation

Alba Estrada 1,2,✉, A Márcia Barbosa 3, Raimundo Real 1
PMCID: PMC6280100  PMID: 30538726

Abstract

Observed species richness (OSR) is a widely used and well-studied biodiversity metric. However, non-observed species in favorable ecosystems are also relevant. Two metrics that include observed and potential species were recently defined: potential biodiversity (hereafter potential species richness—PSR) and geometric mean of favorabilities (GMF). We used these metrics to evaluate the national park network of mainland Spain at two time periods (2002 and 2015), using terrestrial mammals on a UTM 100-km2 grid. PSR and GMF are based on the favorability function, a species distribution model that assesses how favorable an area is for the presence of a species, over and above its prevalence in the study area. For each park and for the whole network, we calculated the mean and sum of OSR, PSR, and GMF in each time period, as well as changes between periods. OSR and PSR were higher inside than outside the park network in both time periods. Thus, although the network covers a very small proportion of the country, it performs well for the representation of mammal species and their favorable areas. However, mean PSR was lower in 2015 than in 2002 inside the national park network, whereas the opposite was the case outside the network. Mountainous Parks generally not only concentrated highly favorable areas for mammals, but they also showed less favorable areas in 2015 compared to 2002, although the reduction was moderate to low. This is a result to consider for future analyses because if the tendency increases, it may have consequences for the conservation of mammals and for the adequacy of the national park network.

Keywords: conservation planning, dark diversity, environmental favorability, potential biodiversity, protected areas

There are several different forms of biodiversity metrics, which are essential tools in conservation ecology (Whittaker 1972; Purvis and Hector 2000). These metrics are needed, among other reasons, to gauge the state of an ecosystem, population trends if they are applied at different time points (Harrison et al. 2014), and the adequacy of protected areas (Estrada et al. 2008). Although the term “biodiversity” encompasses all biotic variation, from genes to ecosystems (Purvis and Hector 2000), more simplistic metrics that are focused on a particular aspect of biodiversity are generally applied. Observed species richness is a very common and well-studied metric, which is widely applied from local to global scales (Griffiths 1999; Orme et al. 2005). However, some authors have pointed out that the set of non-observed species in a favorable ecosystem is as valuable as the set of known species (Pärtel et al. 2011a). From this perspective, they defined the concept of “dark diversity” as being composed of those species that theoretically can inhabit a particular site, but that are absent from it for some reason. Mokany and Paini (2011) offered a probabilistic approach to quantifying dark diversity that incorporates the continuous nature of species distributions, and that can be calculated, for instance, from occurrence probability values given by species distribution models (SDMs).

With this background, Real et al. (2017) proposed the term “potential biodiversity”. In this approach, instead of separating observed diversity and dark diversity, all species can be understood in terms of the probability that they will be found in each region, whether or not they have been observed there. Thus, a species would not be assumed to be inside a region or otherwise; rather, it would be treated as having some relative likelihood of occurring within the region. An important point is that presence probability values are not directly comparable between species differing in prevalence (see below); thus, to estimate potential biodiversity, favorability (a function of probability and prevalence, also taking values between 0 and 1) should be used instead (Real et al. 2006). The summed favorability values may then be considered the potential biodiversity of the region (Real et al. 2017).

Presence probability values are not comparable between species because the probability of a species occurrence in a particular location is affected by both the overall prevalence of the species, and the degree to which the environmental conditions at that location make the occurrence of the species more or less likely than its general prevalence (see Cramer 1999; Hosmer and Lemeshow 2000; Real et al. 2006). Favorability is precisely the latter, i.e., it reflects only the response of the species to the environment, regardless of its prevalence across the study region. Therefore, favorability values are directly comparable between species, even if these have different numbers of presences in the study area (Acevedo and Real 2012). Thus, it is the interaction between favorabilities that enables the combination of SDMs when several species are involved (Estrada et al. 2011). Favorability can also be directly treated with fuzzy logic (Real et al. 2006), which is in line with the views of Pärtel et al. (2011b) who considered species pools as fuzzy sets.

Another biodiversity metric recently defined is the geometric mean of favorabilities (Real et al. 2017), which is inspired by the geometric mean of species abundances (Buckland et al. 2011), and gives more weight to favorability values closer to one. The geometric mean of species abundances has proved to be an efficient measure compared with other biodiversity metrics (Buckland et al. 2005). However, it cannot be used when abundance is 0, and it is too sensitive to species with marked spatial changes in abundance. These limitations are overcome by the geometric mean of favorabilities (Buckland et al. 2011; Real et al. 2017). Moreover, Real et al. (2017) defined the increment in the geometric mean of favorabilities as a measure for assessing changes of this quantity in two periods of time.

Potential biodiversity and the geometric mean of favorabilities are novel metrics that can be used together with more traditional metrics to assess the adequacy of protected area networks, such as Spanish National Parks. Previous studies have evaluated the adequacy of Spanish protected areas in conserving different forms of biodiversity. These comprise a variety of taxonomic groups, including different groups of vertebrates (e.g. Rey Benayas and de la Montaña 2003; López-López et al. 2011; Lisón et al. 2015), invertebrates (Romo et al. 2007; Sánchez-Fernández et al. 2013), plants (Castro et al. 1996; Araújo et al. 2007, 2011), and lichens (Martínez et al. 2006); and a variety of conservation criteria, such as species richness (Araújo et al. 2007; Romo et al. 2007), rarity (Castro et al. 1996; Rey Benayas and de la Montaña 2003), vulnerability (Rey Benayas and de la Montaña 2003; Traba et al. 2007), and endemicity (López-López et al. 2011). As a result, some studies have found that protected areas effectively include the most relevant areas for biodiversity (Martínez et al. 2006; Lisón et al. 2015), whereas others have found that the current configuration of protected area networks is insufficient to cover important areas for biodiversity or hotspots (Rey Benayas and de la Montaña 2003; Traba et al. 2007; López-López et al. 2011). In these previous works, National Parks were generally included as part of the protected reserves (Castro et al. 1996; Rey Benayas and de la Montaña 2003; Romo et al. 2007; Sánchez-Fernández et al. 2013) and were not evaluated independently. However, we consider it necessary to perform a separate evaluation of the National Park network, because conservation policy in Spain is carried out by the Environmental Ministry in the case of National Parks, and by the autonomous regions in the case of protected natural areas (Estrada et al. 2008). Additionally, the National Park network in Spain was established a century ago (in 1916), and thus it is an optimal time to evaluate the contribution of these Parks to the conservation of biodiversity.

Our specific aims in this study are to assess: 1) the contribution of each National Park in mainland Spain to the overall capacity of the National Park network to cover observed species richness (OSR), potential biodiversity (hereafter potential species richness—PSR), and the geometric mean of favorabilities (GMF) of terrestrial mammals, and 2) the current spatio-temporal trends in this capacity.

Materials and Methods

Biodiversity metrics

Our study species were 63 terrestrial non-flying mammals inhabiting mainland Spain. Real et al. (2017) analyzsed their distributions on a grid of 10 km × 10 km UTM (Universal Transverse Mercator) cells (n = 5, 167) at two time periods (2002 and 2015). Data from 2002 came from the mammal atlas of Spain at that time (Palomo and Gisbert 2002), whereas data from 2015 came from the database of the Spanish Ministry of Agriculture, Food and the Environment (downloaded 26 July 2015 from http://www.mapama.gob.es/es/biodiversidad/temas/inventarios-nacionales/inventario-especies-terrestres/inventario-nacional-de-biodiversidad/bdn-ieet-default.aspx). Biodiversity values in each UTM cell of mainland Spain were calculated by Real et al. (2017), who performed species distribution modelling for each species at the two time periods with methods and variables described previously (Barbosa et al. 2009; Real et al. 2009; Barbosa and Real 2010). Environmental favorability for each species was obtained after applying the favorability function (Real et al. 2006). Analyses were performed with package fuzzySim (Barbosa 2015), in the R statistical environment (R Core Team 2014).

From favorability values, potential species richness was obtained in each UTM cell of mainland Spain as the sum of raw favorabilities (i.e., without binary transformations), considering favorable areas for all species, whether or not they have been observed in each region. The analysis was performed separately for the two time periods. The geometric mean of favorabilities was also calculated at both time periods, as well as its increment from 2002 to 2015. These analyses were performed using the R functions provided in Real et al. (2017). For comparison, observed species richness was also calculated at the two time periods, as well as its change from 2002 to 2015. Note that the change in GMF is calculated through the increment (an exponential equation with values always above zero, and higher than one if the increment is positive) (Real et al. 2017), which is not simply a subtraction of the values at two periods of time (as occurs with the change in OSR and PSR).

Maps with values of the different biodiversity metrics across mainland Spain can be seen in Real et al. (2017). In summary, the highest values for observed and potential species richness, although not equally distributed, were located in northern and central-western Spain. The GMF was highest in north-eastern Spain. Overall patterns were essentially similar for 2002 and 2015. The changes in OSR and PSR from 2002 to 2015 were mainly concentrated in the Spanish region of Extremadura (central-western Spain), while the increase in the GMF was more spread throughout the southern half of Spain.

Contribution of each national park

The National Park network of Spain was established a century ago (in 1916) and is currently formed by 15 National Parks, 9 of which are located in mainland Spain. National Parks are the areas with the highest level of protection in Spain. They are defined as “natural areas with high ecological and cultural value, little transformed by exploitation or human activity which, due to the beauty of their landscapes, the representativeness of their ecosystems or the singularity of their flora, fauna, geology or geomorphological formations, possess outstanding ecological, aesthetic, cultural, educational and scientific values whose conservation deserves a preferential attention and are declared of general interest of the State” (BOE 2014). We downloaded a map of the National Parks from the webpage of the Spanish Ministry of the Environment (http://www.mapama.gob.es/es/red-parques-nacionales/sig/; Figure 1) and intersected it with the UTM grid of mainland Spain at a resolution of 10 km × 10 km.

Figure 1.

Figure 1.

Open in a new tab

National Park network of mainland Spain. A: Aigüestortes i Estany de Sant Maurici, C: Cabañeros, D: Doñana, G: Sierra de Guadarrama, M: Monfragüe, O: Ordesa y Monte Perdido, P: Picos de Europa, S: Sierra Nevada, T: Tablas de Daimiel.

For each National Park and for the whole Park network in each time period, we calculated, in the protected cells, the mean and sum of each of the three biodiversity metrics calculated by Real et al. (2017), i.e., OSR, PSR, and GMF, as well as the change in these indices from 2002 to 2015. We performed these analyses in two ways: considering all UTM cells overlapping with a particular Park as equally protected, and considering the degree of protection of each cell according to the proportion of the cell included in the Park. We also estimated mean values of the biodiversity metrics outside the network, i.e., in non-protected cells across mainland Spain. We used the functions group_by and summarise available in the R package dplyr (Wickham and Francois 2016) to calculate these statistics. We tested for significant differences in mean values inside and outside the National Park network with the Mann–Whitney U-test. Significant differences in mean values among National Parks were tested with the Tukey honest significant difference method (Tukey HSD).

Results

The values of the different biodiversity metrics in the cells covered by National Parks can be seen in Figure 2. Highest values for observed and potential species richness were concentrated in northern Parks. The GMF was highest in north-eastern Parks. The changes in OSR and PSR from 2002 to 2015 were concentrated in Monfragüe and Cabañeros National Parks (see location in Figure 1), while the increase in the GMF was highest in the Parks located in the southern half of Spain.

Figure 2.

Figure 2.

Open in a new tab

Observed species richness (A, B, C), potential species richness (D, E, F), and geometric mean of favorabilities (G, H, I), for Spanish terrestrial mammals in 2002, 2015 and the change from 2002 to 2015 in the grid cells intersecting National Parks in mainland Spain. North is up, and grid squares measure 100 km2.

Mean values of the biodiversity metrics were higher within than outside the National Park network for all metrics except the change of PSR and the increment of the GMF (Table 1). However, the differences in GMF between the National Parks and the remaining territory were not significant at any period of time (Table 1).

Table 1.

Mean values for the three biodiversity metrics in the national park network (NP1) and outside the network (NP0) in 2002 and 2015, and changes from 2002 to 2015

Period Metric NP0 NP1 U Sig
2002 OSR 12.55 19.09 173,627 ***
PSR 19.30 24.03 196,485 ***
GMF 0.0252 0.0503 240,162.5 n.s.
2015 OSR 14.63 21.63 154,550 ***
PSR 19.72 23.98 182,713 ***
GMF 0.0256 0.0383 246,214 n.s.
Change from 2002 to 2015 OSR 2.08 2.55 267,696 n.s.
PSR 0.422 −0.0458 329,171 ***
GMF 1.31 1.21 284,730 n.s.
Open in a new tab

OSR: observed species richness, PSR: potential species richness, GMF: geometric mean of favorabilities. The statistic (U) and significance (sig) of the Mann–Whitney tests are also shown.

***

: P < 0.001, ns: P > 0.05.

The National Park with the highest mean and sum values for most indices was Picos de Europa (see location in Figure 1), both when considering all overlapping cells as protected (Figure 3, Table 3) and when values were corrected for the proportion of each cell covered by a National Park (Tables 2 and 4, Figure 4). The Park with the lowest mean values was Tablas de Daimiel (central-southern Spain), although Doñana (SW Spain) also had relatively low values of the mammal biodiversity metrics, and sometimes there were no significant differences between these two Parks (Figure 3). On the other hand, the changes in OSR and PSR between the two time periods were sharpest in Monfragüe (W Spain), whereas the increment in the GMF was largest in Doñana and Sierra Nevada (S Spain; Figures 3, 4, Tables 2–4). We represent in Figure 4 the mean values of the biodiversity metrics in 2015 corrected for the protected proportion of each UTM cell (Table 2). Although sum values could theoretically be affected by the area of the National Park, the Park with the highest values was not the largest (Tables 3 and 4).

Figure 3.

Figure 3.

Open in a new tab

Mean values in each National Park for observed species richness (OSR), potential species richness (PSR) and geometric mean of favorabilities (GMF) in 2002, 2015, and the change from 2002 to 2015. National Park names as in Figure 1. Parks without significant differences according to the TukeyHSD test share the same lower-case letter.

Table 3.

Sum values for the three biodiversity metrics in each national park (NP) in 2002 and 2015, and changes from 2002 to 2015

NP Area (ha) 2002
 2015
 Change from 2002 to 2015
OSR PSR GMF OSR PSR GMF OSR PSR GMF
A 13,927 220 246.03 1.03 230 236.96 0.523 10 −9.07 3.64
C 40,907 138 216.36 0.0497 173 255.09 0.0778 35 38.73 22.67
D 53,416 232 173.74 0.00172 233 177.24 0.00273 1 3.50 22.59
G 33,960 239 374.59 1.05 295 349.29 0.851 56 −25.31 10.67
M 18,010 56 108.73 0.0248 171 159.99 0.0407 115 51.26 13.29
O 15,692 181 335.78 1.69 198 318.41 0.922 17 −17.36 5.38
P 63,601 518 511.10 1.33 516 496.28 1.50 −2 −14.82 14.65
S 85,883 392 501.16 0.0445 423 464.98 0.0556 31 -36.18 29.29
T 3011 9 31.22 0.00365 11 35.71 0.00504 2 4.48 4.15
Open in a new tab

OSR: observed species richness, PSR: potential species richness, GMF: geometric mean of favorabilities. Maximum and minimum values are grey shaded, with maximum values in bold. National park names as in Figure 1.

Table 2.

Mean values in each national park (NP) for the three biodiversity metrics in 2002 and 2015, and changes from 2002 to 2015

NP 2002
 2015
 Change from 2002 to 2015
OSR PSR GMF OSR PSR GMF OSR PSR GMF
A 6.24 6.89 0.0258 6.44 6.57 0.0134 0.192 −0.322 0.104
C 2.95 4.54 0.000993 3.45 5.34 0.00157 0.492 0.804 0.468
D 8.42 6.52 0.0000627 8.49 6.49 0.0000987 0.0714 −0.0352 0.794
G 5.52 7.83 0.0247 6.09 7.31 0.0193 0.565 −0.526 0.205
M 2.13 4.03 0.000952 6.17 5.81 0.00152 4.04 1.78 0.448
O 5.87 11.94 0.0591 6.59 11.06 0.0289 0.721 −0.879 0.169
P 19.50 19.53 0.0527 19.44 18.99 0.0604 −0.0539 −0.538 0.564
S 7.02 9.65 0.00108 7.52 8.85 0.00129 0.500 −0.796 0.487
T 0.472 1.07 0.000135 0.535 1.22 0.000184 0.0629 0.152 0.137
Open in a new tab

Values are corrected for the proportion of the grid cell covered by a national park. OSR: observed species richness, PSR: potential species richness, GMF: geometric mean of favorabilities. Maximum and minimum values are grey shaded, with maximum values in bold. National park names as in Figure 1.

Table 4.

Sum values in each national park (NP) for the three biodiversity metrics in 2002 and 2015, and changes from 2002 to 2015

NP Area (ha) 2002
 2015
 Change from 2002 to 2015
OSR PSR GMF OSR PSR GMF OSR PSR GMF
A 13,927 43.70 48.21 0.180 45.05 45.96 0.0938 1.34 −2.26 0.728
C 40,907 41.34 63.56 0.0139 48.23 74.82 0.0219 6.89 11.26 6.55
D 53,416 117.82 91.30 0.000877 118.82 90.80 0.00138 1 −0.492 11.11
G 33,960 71.76 101.81 0.321 79.11 94.97 0.251 7.35 −6.84 2.67
M 18,010 17.06 32.21 0.00762 49.36 46.45 0.0122 32.30 14.25 3.59
O 15,692 58.71 119.37 0.591 65.93 110.58 0.289 7.21 −8.79 1.69
P 63,601 253.45 253.93 0.685 252.74 246.93 0.785 −0.701 −7.00 7.33
S 85,883 154.40 212.22 0.0238 165.41 194.71 0.0284 11.01 −17.51 10.70
T 3011 1.42 3.20 0.000404 1.61 3.66 0.000551 0.189 0.457 0.410
Open in a new tab

Values are corrected for the proportion of the grid cell covered by a national park. OSR: observed species richness, PSR: potential species richness, GMF: geometric mean of favorabilities. Maximum and minimum values are grey shaded, with maximum values in bold. National park names as in Figure 1.

Figure 4.

Figure 4.

Open in a new tab

Mean values of the biodiversity metrics in 2015 corrected for the proportion of each cell covered by a National Park. (A) Observed (OSR) and potential species richness (PSR), (B) geometric mean of favorabilities (GMF), (C) change in observed and potential species richness from 2002 to 2015, (D) increment of the geometric mean of favorabilities from 2002 to 2015. In each map, bars are scaled by the maximum value. Exact values can be seen in Table 2.

There was a strong positive correlation between OSR and PSR in National Parks (values corrected for the protected proportion) (Table 5). However, the correlation of both biodiversity metrics with the geometric mean of favorabilities, although significant, was weaker. None of the biodiversity metrics correlated significantly with the change in OSR from 2002 to 2015. The correlation of OSR, PSR, and GMF with the change in PSR was negative, but weak. Intermediate correlation values appeared between the increment in the GMF and both OSR and PSR (Table 5).

Table 5.

Pearson’s correlations between biodiversity metrics corrected for the proportion of the grid cell covered by a national park

PSR02 GMF02 OSR15 PSR15 GMF15 OSRChg PSRChg GMFChg
OSR02 0.898 0.619 0.983 0.899 0.766 −0.0722 ns −0.384 0.651
PSR02 1 0.755 0.911 0.994 0.788 0.0897 ns −0.485 0.632
GMF02 1 0.620 0.741 0.900 0.0157 ns −0.455 0.109 ns
OSR15 1 0.924 0.753 0.111 ns −0.293** 0.694
PSR15 1 0.786 0.154 ns −0.391 0.669
GMF15 1 −0.0567 ns −0.362 0.211*
OSRChg 1 0.494 0.245*
PSRChg 1 0.0311 ns
Open in a new tab

Correlations were performed with only the cells intersecting national parks. OSR: observed species richness, PSR: potential species richness, GMF: geometric mean of favorabilities, 02: 2002, 15: 2015, chg: change from 2002 to 2015. All correlations have P < 0.001, except those marked with asterisks or ns.

**

P < 0.01

*

P < 0.05, ns: not significant.

Discussion

In this work, we evaluated the spatial trends in observed and potential mammal diversity in Spanish National Parks at two periods of time. We considered two metrics of potential diversity: potential species richness and geometric mean of favorabilities. The inclusion of these metrics to assess the contribution of each National Park gives importance not only to observed species but also to dark diversity, i.e., to absent species for which the area is favorable (Pärtel et al. 2011a). Different reasons may arise for the absence of these species in favorable areas, with dispersal limitation, habitat specialization, and type of reproduction being recently identified as primary causes related to species life-history traits (Estrada et al. 2015, 2017; Riibak et al. 2017). Additionally, artifactual or anthropic reasons are also possible, such as insufficient sampling effort (Real et al. 2017) or excessive human pressure (Nores 2007; Lucas et al. 2016).

The evaluation of the whole National Park network of mainland Spain showed that mean values of OSR and PSR were higher inside the network than outside it, at both analyzed periods of time (Table 1). Thus, although the network covers a very small proportion of the country, it performs well for the representation of mammal species and their favorable areas. This result is partially in agreement with previous studies that evaluated the level of congruence between Spanish protected areas and hotspots for mammals. For instance, Araújo et al. (2007) found that Iberian protected areas represented more mammal species than expected by chance, and thus protected areas provided effective samples of species among mammals, a result that was not generalizable to other vertebrate or plant groups. On the other hand, Rey Benayas and de la Montaña (2003) found a positive relationship between protected areas and rare mammals, although the relationship was not significant for other conservation criteria, such as richness or vulnerability. This is in accordance with López-López et al. (2011), who found that protected areas were not particularly concentrated in areas of high mammal species richness.

Our results also showed that the change in PSR from 2002 to 2015 was smaller (and negative) inside the network than outside it (Table 1). This means that mean potential species richness was lower in 2015 than in 2002 inside the National Park network, whereas the opposite was the case outside the network. This could indicate that while National Parks have highly favorable areas for mammals, for some reason favorable areas decreased slightly in the studied time period. Hence, some changes occurred outside the network from 2002 to 2015 that implied distributional changes of mammals and their favorable areas in a positive way. This is a result to consider for future analyses. It will be important to assess whether the tendency is reversed or increased, in which case this could have consequences for the conservation of mammals and for the adequacy of the National Park network.

Picos de Europa had the highest values of the biodiversity metrics for mammals. This Park represents the ecosystems linked to the Atlantic forest and is a mountainous system characterized by the presence of several different species of mammals and birds. National Parks with the lowest biodiversity metrics were Tablas de Daimiel and Doñana. This was not an unexpected result, as those National Parks were not characterized by or declared because of the presence of mammals. Tablas de Daimiel is a unique European wetland. It is an exceptional habitat for all fauna linked to the aquatic environment, especially birds and particularly waterfowl. The most important ecosystem in Doñana is the marsh, which is of extraordinary importance as a place of passage, breeding and wintering of thousands of European and African birds (http://www.mapama.gob.es/es/red-parques-nacionales/). Although we evaluated the network at two periods of time (2002 and 2015), it is worth noting that not all the National Parks had been declared before these periods: Monfragüe and Sierra de Guadarrama were declared in 2007 and 2013, respectively. Thus, although natural values are not strictly dependent on the declaration of the Park, a higher level of protection is expected for the oldest National Parks.

Other studies also found that hotspots for mammals were located in northern Spain (Rey Benayas and de la Montaña 2003; Araújo et al. 2007; Levinsky et al. 2007; López-López et al. 2011). This reinforces the relevance of the northern National Parks (Picos de Europa, Ordesa y Monte Perdido, and Aigüestortes i Estany de Sant Maurici) regarding their contribution to preserving mammal species (see Figures 2–4, Tables 2–4). The abovementioned studies also highlighted hotspots for mammals around Sierra de Guadarrama (Rey Benayas and de la Montaña 2003; Araújo et al. 2007; Levinsky et al. 2007; López-López et al. 2011), although, as stated previously, the declaration of this National Park was not made until 2013. All of these previous results are in agreement with ours, and reinforce the importance of mountainous systems as favorable areas for several mammal species (Levinsky et al. 2007).

Regarding changes in the biodiversity metrics between 2002 and 2015, observed species richness increased in all National Parks except for Picos de Europa (Figures 3–4, Tables 2–4). Thus, it is worth remarking that the Park with the highest OSR is the same that simultaneously decreased slightly in its mammal richness, and the reduction is a bit highest for PSR. As stated above, mountainous Parks concentrate highly favorable areas for mammals, but these Parks (Picos de Europa, Ordesa y Monte Perdido, Aigüestortes i Estany de Sant Maurici, Sierra de Guadarrama, and Sierra Nevada) also had less favorable areas based on the data from 2015 compared to 2002. In any case, these negative changes in PSR were moderate to low, and even the Park with the lowest mean value, i.e., Sierra de Guadarrama, was less favorable in 2015 for only two species (PSR Change = −1.9, see Figure 3). The pattern of changes in the geometric mean of favorabilities was less clear, and the results differ if we consider mean or sum values in the National Parks or whether or not the protected proportion was included. For this metric, a negative increment from 2002 to 2015 occurs when values are below one. Although this pattern is not maintained for all statistics, mountainous Parks were again highlighted as having a negative increment on the GMF (Figure 3). Therefore, special attention should be paid to mountainous Spanish National Parks if the intention is for them to maintain high levels of mammal diversity.

Methodological aspects

We performed different summary statistics. However, we consider the mean values of the metrics corrected by the protected proportion (Table 2) to more realistically reflect the contribution of each National Park. Mean values are not affected by the size of the Park, whereas sum values are. The correction for the protected proportion of the cell adjusts the values and balances their importance (Díaz-Gómez et al. 2013; Sánchez-Fernández and Abellán 2015). It is not realistic to assign, for instance, all species recorded in a 100 km2 cell to a National Park if their overlap is small. In our case, mean OSR in Picos de Europa, for instance, was around 40 species (Figure 3) but this figure dropped to ∼20 when the protected proportion of the cells was considered (Table 2). Similar patterns appeared for other biodiversity metrics and National Parks.

One way to evaluate a reserve network with small numbers of Parks and small geographical coverage, such as the Spanish National Parks, is to focus attention on the contribution of each Park, rather than on the biodiversity metric itself. In other words, the intention is not to determine if the most favorable areas for a species or the highest species richness are well covered by protected areas in general (Scott et al. 1993; Estrada et al. 2008), but to assess how each Park contributes to this protection. Thus, the importance of each National Park is relative to the contribution of the other Parks in the network. Another possibility is to focus attention on how well species, or any biodiversity metric, are represented in the Parks. In this case, an overall value of protection is given to each species (Díaz-Gómez et al. 2013), and a species is considered adequately protected by the National Park network when the proportion of the overall value of protection covered by the network is larger than the proportion of the country covered by the network (Estrada and Real submitted). The use of one of these evaluation methods depends on the objective of the study. In this case, our aim was to gauge the contribution of each National Park to the protection of potential mammal diversity, rather than to assess how well each species was covered by the National Park network.

Concluding remarks

Mammal diversity was higher in the Spanish National Park network than outside the network, and the Parks with highest diversity values were mountainous Parks located in northern Spain. However, these high-diversity Parks had less favorable areas in 2015 compared to 2002. Mammal diversity in the Spanish National Parks should be evaluated again in the mid-long term, when updated distribution data become available, to assess whether or not the tendency of these Parks to have less favorable areas has increased. This will have important conservation implications for mountainous Parks to maintain high levels of mammal diversity.

Acknowledgment

We thank Sarah Young for the English revision of the article.

Funding

This study was funded by project 1098/2014 (Organismo Autónomo Parques Nacionales, Spain). A.M.B. is supported by FCT (Portugal) and FEDER/COMPETE 2020 through contract IF/00266/2013, exploratory project CP1168/CT0001, and funds POCI-01-0145-FEDER-006821 to research unit UID/BIA/50027.

References

  1. Acevedo P, Real R, 2012. Favourability: concept, distinctive characteristics and potential usefulness. Naturwissenschaften 99:515–522. [DOI] [PubMed] [Google Scholar]
  2. Araújo MB, Alagador D, Cabeza M, Nogué S, Bravo D. et al. , 2011. Climate change threatens European conservation areas. Ecol Lett 14:484–492. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Araújo MB, Lobo JM, Moreno JC, 2007. The effectiveness of Iberian protected areas in conserving terrestrial biodiversity. Conserv Biol 21:1423–1432. [DOI] [PubMed] [Google Scholar]
  4. Barbosa AM, 2015. fuzzySim: applying fuzzy logic to binary similarity indices in ecology. Methods Ecol Evol 6:853–858. [Google Scholar]
  5. Barbosa AM, Real R, 2010. Favourable areas for expansion and reintroduction of Iberian lynx accounting for distribution trends and genetic variation of the wild rabbit. Wildlife Biol Practice 6:34–47. [Google Scholar]
  6. Barbosa AM, Real R, Vargas JM, 2009. Transferability of environmental favourability models in geographic space: the case of the Iberian desman Galemys pyrenaicus in Portugal and Spain. Ecol Model 220:747–754. [Google Scholar]
  7. BOE, 2014. Ley 30/2014, de 3 de diciembre, de Parques Nacionales.
  8. Buckland ST, Magurran AE, Green RE, Fewster RM, 2005. Monitoring change in biodiversity through composite indices. Philos Trans R Soc B-Biol Sci 360:243–254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Buckland ST, Studeny AC, Magurran AE, Illian JB, Newson SE, 2011. The geometric mean of relative abundance indices: a biodiversity measure with a difference. Ecosphere 2:1–15. [Google Scholar]
  10. Castro I, Moreno JC, Humphries CJ, Williams PH, 1996. Strengthening the Natural and National Park system of Iberia to conserve vascular plants. Bot J Linn Soc 121:189–206. [Google Scholar]
  11. Cramer JS, 1999. Predictive performance of the binary logit model in unbalanced samples. J R Stat Soc Ser D-Stat 48:85–94. [Google Scholar]
  12. Díaz-Gómez DL, Toxopeus AG, Groen TA, Muñoz AR, Skidmore AK. et al. , 2013. Measuring the Insecurity Index of species in networks of protected areas using species distribution modeling and fuzzy logic: the case of raptors in Andalusia. Ecol Indic 26:174–182. [Google Scholar]
  13. Estrada A, Meireles C, Morales-Castilla I, Poschlod P, Vieites D. et al. , 2015. Species’ intrinsic traits inform their range limitations and vulnerability under environmental change. Global Ecol Biogeogr 24:849–858. [Google Scholar]
  14. Estrada A, Morales-Castilla I, Meireles C, Caplat P, Early R, 2017. Equipped to cope with climate change: traits associated with range filling across European taxa. Ecography. doi:10.1111/ecog.02968. [Google Scholar]
  15. Estrada A, Real R. Forthcoming. Assessment of the National Park network of mainland Spain by the Insecurity Index of vertebrate species. PLoS ONE. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Estrada A, Real R, Vargas JM, 2008. Using crisp and fuzzy modelling to identify favourability hotspots useful to perform gap analysis. Biodivers Conserv 17:857–871. [Google Scholar]
  17. Estrada A, Real R, Vargas JM, 2011. Assessing coincidence between priority conservation areas for vertebrate groups in a Mediterranean hotspot. Biol Conserv 144:1120–1129. [Google Scholar]
  18. Griffiths D, 1999. On investigating local-regional species richness relationships. J Anim Ecol 68:1051–1055. [Google Scholar]
  19. Harrison PJ, Buckland ST, Yuan Y, Elston DA, Brewer MJ. et al. , 2014. Assessing trends in biodiversity over space and time using the example of British breeding birds. J Appl Ecol 51:1650–1660. [Google Scholar]
  20. Hosmer DW, Lemeshow S, 2000. Applied Logistic Regression. 2nd edn. New York: John Wiley and Sons, Inc. [Google Scholar]
  21. Levinsky I, Skov F, Svenning JC, Rahbek C, 2007. Potential impacts of climate change on the distributions and diversity patterns of European mammals. Biodivers Conserv 16:3803–3816. [Google Scholar]
  22. Lisón F, Sánchez-Fernández D, Calvo JF, 2015. Are species listed in the Annex II of the Habitats Directive better represented in Natura 2000 network than the remaining species? A test using Spanish bats [journal article]. Biodivers Conserv 24:2459–2473. [Google Scholar]
  23. López-López P, Maiorano L, Falcucci A, Barba E, Boitani L, 2011. Hotspots of species richness, threat and endemism for terrestrial vertebrates in SW Europe. Acta Oecol 37:399–412. [Google Scholar]
  24. Lucas PM, González-Suárez M, Revilla E, 2016. Toward multifactorial null models of range contraction in terrestrial vertebrates. Ecography 39:1100–1108. [Google Scholar]
  25. Martínez I, Carreño F, Escudero A, Rubio A, 2006. Are threatened lichen species well-protected in Spain? Effectiveness of a protected areas network. Biol Conserv 133:500–511. [Google Scholar]
  26. Mokany K, Paini DR, 2011. Dark diversity: adding the grey. Trends Ecol Evol 26:264–265. [DOI] [PubMed] [Google Scholar]
  27. Nores C, 2007. Galemys pyrenaicus (E. Geoffroy Saint-Hilaire, 1811). Ficha Libro Rojo In: Palomo LJ, Gisbert J, Blanco JC, editors. Atlas y Libro Rojo De Los Mamíferos Terrestres De España. Madrid: Dirección General para la Biodiversidad-SECEM-SECEMU, 96–98. [Google Scholar]
  28. Orme CDL, Davies RG, Burgess M, Eigenbrod F, Pickup N. et al. , 2005. Global hotspots of species richness are not congruent with endemism or threat. Nature 436:1016–1019. [DOI] [PubMed] [Google Scholar]
  29. Palomo LJ, Gisbert J, 2002. Atlas De Los Mamíferos Terrestres De España. Madrid: Dirección General de Conservación de la Naturaleza-SECEM-SECEMU. [Google Scholar]
  30. Pärtel M, Szava-Kovats R, Zobel M, 2011a. Dark diversity: shedding light on absent species. Trends Ecol Evol 26:124–128. [DOI] [PubMed] [Google Scholar]
  31. Pärtel M, Szava-Kovats R, Zobel M, 2011b. Discerning the niche of dark diversity. Trends Ecol Evol 26:265–266. [DOI] [PubMed] [Google Scholar]
  32. Purvis A, Hector A, 2000. Getting the measure of biodiversity. Nature 405:212–219. [DOI] [PubMed] [Google Scholar]
  33. R Core Team, 2014. R: A Language and Environment for Statistical Computing Vienna: R Foundation for Statistical Computing. http://www.R-project.org/.
  34. Real R, Barbosa AM, Bull JW, 2017. Species distributions, quantum theory, and the enhancement of biodiversity measures. Syst Biol 66:453–462. [DOI] [PubMed] [Google Scholar]
  35. Real R, Barbosa AM, Rodriguez A, Garcia FJ, Vargas JM. et al. , 2009. Conservation biogeography of ecologically interacting species: the case of the Iberian lynx and the European rabbit. Divers Distrib 15:390–400. [Google Scholar]
  36. Real R, Barbosa AM, Vargas JM, 2006. Obtaining environmental favourability functions from logistic regression. Environ Ecol Stat 13:237–245. [Google Scholar]
  37. Rey Benayas JM, de la Montaña E, 2003. Identifying areas of high-value vertebrate diversity for strengthening conservation. Biol Conserv 114:357–370. [Google Scholar]
  38. Riibak K, Ronk A, Kattge J, Pärtel M, 2017. Dispersal limitation determines large-scale dark diversity in Central and Northern Europe. J Biogeogr 44:1770–1780. [Google Scholar]
  39. Romo H, Munguira ML, García-Barros E, 2007. Area selection for the conservation of butterflies in the Iberian Peninsula and Balearic Islands. Anim Biodivers Conserv 30:7–27. [Google Scholar]
  40. Sánchez-Fernández D, Abellán P, 2015. Using null models to identify under-represented species in protected areas: a case study using European amphibians and reptiles. Biol Conserv 184:290–299. [Google Scholar]
  41. Sánchez-Fernández D, Abellán P, Picazo F, Millán A, Ribera I. et al. , 2013. Do protected areas represent species' optimal climatic conditions? A test using Iberian water beetles. Divers Distrib 19:1407–1417. [Google Scholar]
  42. Scott JM, Davis F, Csuti B, Noss R, Butterfield B. et al. , 1993. Gap analysis: a geographic approach to protection of biological diversity. Wildl Monogr 123:1–41. [Google Scholar]
  43. Traba J, García de la Morena EL, Morales MB, Suárez F, 2007. Determining high value areas for steppe birds in Spain: hot spots, complementarity and the efficiency of protected areas. Biodivers Conserv 16:3255–3275. [Google Scholar]
  44. Whittaker RH, 1972. Evolution and measurement of species diversity. Taxon 21:213–251. [Google Scholar]
  45. Wickham H, Francois R, 2016. dplyr: A grammar of data manipulation. R package version 0.5.0. https://CRAN.R–project.org/package=dplyr.

Articles from Current Zoology are provided here courtesy of Oxford University Press

RESOURCES