FUNDACIÓN MAPFRESeguridad y Medio Ambiente

Year 31 Nº 122 2011

Implementation of a monitoring system for gauging the effects of global change on the ecosystem functioning of protected areas in Spain and South America ENVIRONMENT

This article presents a system for evaluating and monitoring the responses of the ecosystem to global change in six protected areas of Spain, Argentina and Uruguay, setting alert thresholds and offering the results of the pilot scheme carried out in one of these protected sites, the Nature Park Cabo de Gata-Níjar in Almería Spain. The system is based on the analysis of a time series of low-cost satellite images for calculating the Absorbed Photosynthetically Active Radiation (APAR) by the vegetation as a proxy indicator of the net primary production (NPP) of the ecosystems. As a basic application of the system an analysis is made of the 2001-2008 trends in four of the descriptors related to primary productivity,  phenology and seasonal carbon gain . This project offers protected-site managers a monitoring and alert system to keep track of changes in their ecosystem functioning. This scientific support tool for the management of those areas offers a twofold advantage: firstly, it is adaptable to other areas and, secondly, it can be consulted on internet by other scientists, site managers and the public at large.


The project was carried out in six protected areas, including the Nature Park Cabo de Gata-Nïjar in Spain, (left), the Protected Landscape of Quebrada de los Cuervos in Uruguay (top), the National Park El Palmar in Argentina (below left), and the National Park of Doñana in Spain (below).

Dr. of Environmental Sciences and research fellow in the Regional and Remote sensing Laboratory (Laboratorio de Análisis Regional y Teledetección) of the Agronomy School and IFEVA (Facultad de Agronomía e IFEVA), Buenos Aires University and CONICET (Argentina), and in the Plant Biology and Ecology Department (Dpto. de Biología Vegetal y Ecología) of the Andalusian Centre for Assessing and Monitoring Global Change (Centro Andaluz para la Evaluación y Seguimiento del Cambio Global), Almería University (Spain). *Correspondence:
Ctra. Sacramento s/n, La Cañada de San Urbano, Almería (España).
Tel: (+34)950015932.
Fax: (+34)950015069.
email: dalcaraz@ual.es

J. Cabello. Dr. of Biology. Professor in the Dpto. de Biología Vegetal y Ecología, Centro Andaluz para la Evaluación y Seguimiento del Cambio Global, Almería University (Spain).

C. Bagnato. Technician in the Laboratorio de Análisis Regional y Teledetección, Facultad de Agronomía e IFEVA, Buenos Aires University and CONICET (Argentina).

A. Altesor. Dr. of Biology. Professor of the Ecology Department (Departamento de Ecología) , Universidad de la República (Uruguay).

C. Oyonarte. Dr. of Biología. Professor in the Department of Soil Science and Agricultural Chemistry (Departamento de Edafología y Química Agrícola), Centro Andaluz para la Evaluación y Seguimiento del Cambio Global, Almería University (Spain).

M. Oyarzabal. Dr. of Farming Science. Coordinator of the Laboratorio de Análisis Regional y Teledetección, Facultad de Agronomía e IFEVA, Buenos Aires University and CONICET (Argentina).

J.M. Paruelo. Dr. of Ecology. Professor of the Agronomy School (Facultad de Agronomía) and y Director of the Environmental Sciences Degree Course, of Buenos Aires university, Chief Researcher of CONICET and Director of the Laboratorio de Análisis Regional y Teledetección, Facultad de Agronomía e IFEVA, Buenos Aires University and CONICET (Argentina).

Global change, in terms of the combined effect of climate change, biogeochemical cycles, land uses and biotic interchanges, poses a stern challenge for biodiversity conservation [1]. On a world level it has been proven that climate change is altering the duration and phenology of the growing season, the fire regime and food chain, i.e. multiple aspects of the ecological system[2]. Many of these effects have become evident in Spain and the Southern Cone of South America [3-9]. These changes also affect protected areas, where the greatest efforts are being made to preserve biodiversity [10]. It is therefore necessary to find out whether or not the structure and functioning of these areas is changing over time and quantify the changes [11]. This would help to develop adaptive management practices to minimise any biodiversity loss and prioritise conservation efforts in light of these changes. At the same time quantifying the value that the assets and services of protected areas afford mankind and ascertaining the risk they run would boost social acceptance of the conservation measures [12].

It behoves us scientists and site managers to develop and implement monitoring instruments for gauging the health of ecosystems and the changes occurring in them.

In light of all the above it behoves us scientists and site managers to develop and implement monitoring instruments that keep a constant check on the health of ecosystems (ecological integrity) and the changes occurring in them [13,14]. The variables to be factored into these instruments have to abide by the following basic principles [14-17]:

  1. they can be recorded at ecosystem level throughout large areas and in real time
  2. they give an early warning of impacts to serve as a guide to effective adaptive management
  3. they can be measured in an easy and direct way
  4. they pick up the time-space variability caused both by natural perturbations and anthropic impacts
  5. they establish benchmark control values
  6. they can be compared at both local and regional level
  7. they facilitate the establishment of relations between them by means of different space scales.

For this purpose the functional attributes of the ecosystems (i.e., those related to exchanges of energy and material between the biota and the environment) offer several advantages. They show a quicker perturbation response than vegetation structure, preventing structural inertia from delaying perception of the effects of this perturbation on the ecosystems[18]. Furthermore, functional attributes favour monitoring by remote-sensing of different regions under a common observation protocol, allowing quantitative and qualitative characterisation of the ecosystem services [19]. Diverse spectral indices deriving from satellite images are linked to functional variables of the ecosystems, such as primary production, evapotranspiration, surface temperature, land surface albedo and precipitation use efficiency [20-22]. The vegetation index is one of the most important variables deriving from spectral data. Vegetation indices have been widely used in ecology for studying time trends and they play a key role in global change research [23]. They can be calculated from images with very diverse space and time resolution, so they have been incorporated into many monitoring experiments [24, 25] carried out on a wide range of scales (from regional to worldwide). They have proven to be useful for detecting long term changes even at the scale of a protected area [6-8]. The use of ecology remote-sensing tools based on vegetation indices [26, 27] has improved the assessment of ecosystem functioning, particularly of net primary productivity (NPP) on a regional and even worldwide scale. The incorporation of these functional variables is vitally important for ecosystem management, systematic conservation planning [28] and to factor the effects of global change into conservation strategies [10].

Despite all the knowledge built up on a global and continental scale the management of each particular protected site calls for an assessment in its own right, since a given area might not follow the pattern observed at regional level [6]. There are now many assessments based on large-scale remote sensing of ecosystem functioning of the protected areas of Spain and South America [6-8, 29-32], and also of the trends over time [6-8, 32]. Nonetheless, conservation practice and theory call for assessments with a high space resolution, paving the way for specific management actions to suit the particular effects of the environmental changes on the biodiversity-maintaining ecological processes of a protected area.

Objectives

One of our lines of research in the last ten years has been to gauge the effects of global change on ecosystem functioning in protected areas. This objective has both a theoretical and practical side to it. From a theoretical point of view it has enabled us to understand better the environmental controls of the mean ecosystem functioning and its time trends over different space and time scales. From a practical point of view it has enabled us to develop monitoring and alert systems to support the conservation of protected areas. In this research work we have developed a conceptual framework and working tool that provides solid scientific information on the current situation and time trends in the ecosystem functioning of six protected areas of Iberoamerica. This tool is based on the study of spectral vegetation indices deriving from high resolution, low cost satellite images and it aims to furnish protected areas with an online monitoring and alert tool for tracking the trend of net primary productivity , the fundamental service provided by the ecosystem when incorporating solar energy into the food chain. The methodology employed will enable the system to be easily extended to other ecosystem-functioning descriptors such as evapotranspiration or phenology, of great interest for assessing the effects of global change on biodiversity.

Materials and Methods 

Protected Areas Studied

This research was conducted simultaneously in six protected areas of Spain, Argentina and Uruguay. Nonetheless, by way of illustration, this article shows the results only for the pilot scheme carried out in the Biosphere Reserve of the Nature Park Cabo de Gata-Níjar [32]. The six selected areas take in a vast variety of environments and land-use histories, representing an excellent framework for evaluating changes in ecosystem functioning through the different ecosystems and geopolitical circumstances (figure 1):


Figure 1. Maps of the ecosystems studied in each one of the six protected areas studied, showing the 250 m MODIS pixels of the ecosystems (white squares) that were sampled in the analysis

  • In Spain: 1) Biosphere Reserve of the Nature Park Cabo de Gata-Níjar, whose vegetation is predominantly semi-arid scrubland sharing species with arid zones of Africa and Asia; 2) National and Nature Park of Doñana, with Mediterranean scrubland and woodland and saltmarsh, a system of coastal wetlands.
  • In Argentina: 1) National Park Iguazú and its continuation in Brazil with National Park do Iguaçu, harbouring the last representation of the tropical rainforest of Paraná; 2) National Park El Palmar and its continuation in the Wildlife Refuge of La Aurora del Palmar, last redoubt of the temperate savanna dominated by espinal (thorny deciduous shrubland forest and palm groves).
  • In Uruguay a National System of Protected Areas (SNAP in Spanish initials) is being set up. Within this the following areas were chosen: 1) the first area brought into the system, the protected landscape Quebrada de los Cuervos, and 2) the catchment area of the Laureles and Cañas streams, candidate area to be incorporated in SNAP. Both areas are a patchwork of meso-xerophytic and meso-hydrophytic grassland with native woodland around the rivers.

Design of the ecosystem-functioning monitoring system 

The information system for monitoring the vegetation and supporting the decision-taking of the protected areas is based on that proposed by Grigera et al. [33] for monitoring forage production and it uses the Monteith model [34] for measuring ecosystem productivity on the basis of radiometric information. Oyarzabal et al [35] began to apply it to the P.N. Cabo de Gata-Níjar.

In a nutshell, the core of the information system is made up by a complete collection of biophysical databases, pride of place here going to satellite images (figure 2). The first step was to acquire the different information sources. These were then put through a quality control before being integrated into the system. Finally, by means of statistical models and analysis, the mean dynamic was identified and time trends and spatial anomalies of key biophysical variables were detected together with indicators of the ecosystem functioning to help site managers draw up their reports. One advantage of the proposed scheme is that it adapts to the level of available information. The maintenance and development of the system thus involves not only the systematic updating of the satellite information but also the calibration of the biophysical variables and incorporation of new additional information on the protected area.

The proposed information system combines the Grigera ‘et al’ method for monitoring forage production and the Monteith method for assessing ecosystem productivity from radiometric information

Figure 2 (Available long description)
Figure 2. Information system scheme

Spectral vegetation indices and descriptive attributes of the ecosystem functioning

The study of ecosystem functioning was based on analysis of spectral data provided by satellite images. This involved the use of two vegetation indices: the Normalised Difference Vegetation Index  (NDVI) and the Enhanced Vegetation Index EVI).Both enable us to monitor photosynthetic activity, productivity or the leaf area index of ecosystems for wide-ranging territories by means of time series of the satellite images. They are based on the spectral property of green vegetation to absorb differentially the photosynthetically active radiation. The NDVI calculates  the standardised difference of reflectance between two wavelengths in relation to the photosynthesis process (red and near infra-red bands), while the EVI incorporates a third wavelength (blue) to cancel out the influence of the soil and atmosphere.

The study is based on an analysis of spectral data from satellite images of two vegetation indices: the Normalised Difference Vegetation Index (NDVI) and the Enhanced Vegetation Index (EVI) 

The EVI images were obtained from the MODIS sensor in 16-day composites with a spatial resolution of 250 metres. Thanks to their high quality and spatial resolution, these images form the core of the monitoring system. Nonetheless, they have been available only since 2000, so NDVI images from the NOAA-AVHRRLTDR database at 15 day composites were used to go further back in time. Despite having a 5 k pixel size these images are available for the 1981 to 2000 time period. The information system downloads the images, filters out values with a low spectral quality and systematically records the «vegetation indices» product from a NASA server (ftp://e4ftl01u.ecs.nasa.gov/MOLT/ MOD13Q1.005/); for this application the process was automated with Kepler and IDL+ENVI 4.6. This article shows only the results based on the MODIS images for the P.N. Cabo de Gata-Níjar.

As main indicator of ecosystem functioning a calculation was made of the absorbed photosynthetically active radiation (APAR) by the vegetation, an estimator of primary productivity, by means of the following equation:APAR = FAPAR x IPAR; where FAPAR is the fraction of absorbed photosynthetically active radiation by the vegetation and IPAR is the incident photosynthetically active radiation. The FAPAR was calculated from a linear relation with the EVI (also calculated for the NDVI) using the method put forward by Ruimy et al [36], which requires fixing minimum (0%) and maximum (95 %) FAPAR values. In the case of the P.N. Cabo de Gata, the other term of the equation, the IPAR, was taken from daily readings of the meteorological station of Alme airport. For more details of the general APAR calculating algorithm see Grigera et al [33] and Oyarzabal et al [35].

The advantage of using APAR, instead of a vegetation index directly, stems from a certain timelag in the interception of energy (FAPAR, calculated from the EVI) and incident radiation. The mean FAPAR peaked in February and bottomed out in September, while the IPAR peaked in June and bottomed out in September. When this occurs the APAR (calculated from the above equation as the product of both readings) is a better indicator of the vegetation’s photosynthetic activity than vegetation indices.

From the APAR, following the Monteith model, it is possible to calculate vegetation productivity by multiplying APAR by its Radiation or Light Use Efficiency (RUE or LUE,). In the method proposed herein this last step was not taken since there is not yet enough information to hand on the RUE and its determination is a very complex process, though it is likely to be phased in at some moment in the future [37, 38].
The system allows representation of the APAR values of each pixel on each date, thus obtaining the annual APAR curve (figure 3), which then serves as the basis for calculating the following descriptive attributes of the annual behaviour of the APAR, which are then used as indicators of ecosystem functioning: 1) the annual cumulative APAR (C-APAR), calculated as the product of the annual mean APAR and the yearly number of composite images (23 in the case of MODIS), and considered as a linear estimator of primary productivity; 2) the intra-annual coefficient of variation of APAR (APAR_CV), calculated as annual standard deviation divided by the annual mean and taken to be a descriptor of seasonal carbon gain; 3) the maximum APAR values (APAR_Max) and 4) minimum APAR values (APAR_Min) of the year, as indicators of the maximum and minimum photosynthetic activity of the ecosystems; and, lastly, 5) the moment of maximum APAR (MMAX) and 6) the moment of minimum APAR (MMIN). In sum, these six attributes provide us with information on the productivity, seasonality and phenology of the ecosystems [23] (table 1) (figure 3).


Figure 3. Annual curve of the Absorbed Photosynthetically Active Radiation (APAR) by the vegetation and derived descriptive attributes of ecosystem functioning. Adapted from Baldi et al. [39].

Table 1. Descriptive indicators of ecosystem functioning serving as the basis of this research.
Index Type of measure Definition Biological significance
APAR Measurements of quantity, structure and condition of the vegetation Absorbed photosynthetically active radiation (APAR) by the vegetation Amount of energy absorbed by the plants during photosynthesis
C-APAR Total productivity and biomass Total amount of energy absorbed by the plants during a year of photosynthesis Annual vegetation productivity
intra CV Intra-annual coefficient of variability Standard deviation / mean Seasonal carbon gain
Max Total productivity maximum and biomass Maximum value throughout the year Maximum photosynthetic capacity of the ecosystem
Min Minimum total productivity and biomass Minimum value throughout the year Minimum photosynthetic capacity of the system
MMAX Moment of the year Date on which the maximum value is reached throughout the year Phenology of the ecosystem’s maximum photosynthetic activity
MMIN Moment of the year Date on which the minimum value is reached throughout the year Phenology of the ecosystem’s minimum photosynthetic activity

Detection of trends in ecosystem functioning

Trends were sought in the size, seasonality and phenology of the APAR using the Mann-Kendall non-parametric test [6-9, 32], which has the advantage of robustness to non-normal data distribution, data gaps and temporal self-correlation. This rank-order non-parametric test calculates the monotonic trend in terms of the number of times that a year in particular shows a value greater or lower than any previous year. This method has proven to be powerful in preliminary gross-scale analyses of the Iberian Peninsula [6-9, 32] and South America [5, 7, 29, 39]. This analysis was carried out in the six protected areas and their immediate vicinity on two spatial scales (MODIS 250 metres and LTDR 5 kilometres), although in this study the results are shown only for the P.N. Cabo de Gata-Níjar for MODIS.

Detection of spatial anomalies in ecosystem functioning

As an example of the system’s capacity to support decision taking, spatial anomalies were detected in the APAR within the vegetation types «scattered scrub» and «dense scrub» in the P.N. Cabo de Gata-Níjar, to pinpoint areas in need of a restructuring of the plant cover, thus enhancing the ecosystem’s productivity and resilience. For each vegetation type, using the Map of Land Uses and Plant Cover of Andalusia, a random selection was made of 40 pixels fulfilling two conditions: complete inclusion within a type of vegetation and location far from the borderline with other vegetation types.

The aim of these samples is to establish the characteristic variability of the functional variable under consideration (APAR in the example) for a given ecosystem, establishing the normal range of variation. The values falling outside this range are deemed to be spatially anomalous [40]. In this example the mean interval ± 1 standard deviation has been used. If we increase the number of standard deviations around the mean, the system will be more tolerant or less sensitive.

The system phases in new satellite images every 16 days, allowing us to evaluate the «new state» of the protected ecosystems against the benchmark behaviour of previous years

Detection of alerts and evaluation of their relevance

The system automatically phases in new images as they become available (every 16 days in the case of MODIS’s vegetation indices); this allows it to evaluate the «new» state of the protected ecosystems, taking as benchmark their behaviour of previous years. In graphical form the idea is to display the mean year curve of the available time series, with its inter-annual variation interval (± 1 standard deviation) with superimposition on the same graph of the most recent up-to-date values for the same variable. This will trigger an alert when the new value of the variable strays far from the average behaviour
(mean ± 1 standard deviation). The relevance of the alert is evaluated on the basis of its strength, deviation and duration. The strength represents the differences between the new values and those expected on the basis of historic behaviour: the bigger the difference the greater the strength. Deviation tries to determine whether the alert may be simply a passing or tardy event, such as an early start to the rainy season or late start to the fire season. Duration measures the time over which the alert has been recurring and is estimated from the number of previous dates on which the value reached by a functional attribute is lower (or higher) than the mean ± 1 standard deviation. In assessing the relevance of the alert consideration is also given to the conservation value of the ecosystems affected and whether or not they had already been recording long term trends.

Results and Discussion

Ecosystem functioning trends

The results obtained for Cabo de Gata-Níjar (figure 4a) show that, in the time series considered, there was predominantly an increase in the APAR and, ipso facto, of net primary productivity. This chimes in with the global trend observed in the period 1981-2000 [41]. The grouped spatial pattern shown by the positive and negative trends serves as guidance for the management actions of P.N. Cabo de Gata Níjar and also favours research into the causes and consequences of these trends.

Absence of negative trends within the protected area indicates that no reduction in vegetation productivity is currently underway. This contrasts sharply with the situation outside the protected area with a predominance of pixels with negative trends, in many cases significant (figure 4.b). Both the climate conditions and their trend are similar inside and outside the park; this suggests that the discovered differences in ecosystem functioning stem from the different use and management of the land in each case and the recovery and degradation of the plant cover.


Figure 4. Trend maps of absorbed photosynthetically active radiation (APAR) by the vegetation of the protected area Biosphere Reserve of Cabo de Gata-Níjar (Almería-Spain) and bordering zones in the period 2001-2008. The maps show: (a) the trend slope of the annual cumulative APAR (C-APAR) (left) and its significance level or p-valor (right); and the significant trends; (b) the maximum values (APAR_Max), (c) the minimum values (APAR_Min) and d) the coefficient of variation (APAR))_CV). The slopes were calculated with Sen’s method. Their significance was calculated with the Mann-Kendall trend test. The line shows the outline of the protected area. Adapted from Oyonarte et al [32].

The annual dynamic of the APAR also gives us insights into other aspects of ecosystem functioning, such as phenology or seasonal carbon gain. In this study a calculation was also made of the trends of the maximum and minimum annual APAR, of particular interest for the management of ecosystems with borderline environmental conditions, such as the arid ecosystems of P.N. Cabo de Gata-Níjar, where changes in minimum values may be bound up with degradation processes. Analysis of the trends in maximum APAR (APAR_Max) and minimum APAR (APAR_Min) shows that the observed increase in cumulative absorbed radiation C-APAR over the years under study is due above all to an increase in maximum values (figure 4.c) rather than an increase in minimum values (figure 4.b). The presence of significant positive trends in APAR_Min values but not in APAR_Max values brought about significant falls in the intra-annual coefficient of variation, an indicator of seasonality (figure 4d). This means that the difference in absorbed radiation between the growing season and non-growing seasons has fallen, i.e., a reduction in seasonality.

The results obtained for Cabo de Gata-Níjar show an increase in APAR and, ipso facto, net primary productivity during the time series under study. This chimes in with the overall trend observed for the period 1981-2000

Spatial Anomalies

The results (figure 5) show, as was only to be expected, that the C-APAR is significantly higher in areas with «dense scrub» (604.3 MJ.m-2.year-1) than in «scattered scrub» (515.9 MJ.m-2.year-1); at the same time, its spatial variability is lower (coefficient variation of 13.1% and 15.6%, respectively).


These results show «scattered scrub» to be the area in most need of restoration, and we use the benchmark levels to identify those pixels showing behaviour well below the average C-APAR for «scattered scrub». The final result is the map of spatial anomalies (figure 6) showing those areas with a behaviour below (15.9%), above (22.1%) or within the benchmark range (61.9% of the land surface) for the C-APAR of «scattered scrub».


Figure 5. Spatial heterogeneity of the annual absorbed radiation for two land uses in the protected area. The points correspond to the mean of each one of the forty pixels sampled in the period 2000-2008; the lines show the variability intervals determined by the mean functioning of each class (the central line corresponds to the mean, the upper and lower lines to the +/- of standard deviation). Different letters indicate significant differences between means using a t test (p<0.0001 ). Adapted from Oyonarte et al. [32].


Figure 6. Map of spatial anomalies in the annually absorbed photosynthetically active radiation (APAR) by the vegetation in the land surface covered by «scattered scrubland». The classes are established in terms of the benchmark levels «accepted variability» for the type of land use. The interval is defined by the mean ±, a standard deviation of the selected sample (n= 40). All the pixels of the protected area belong to the type of land use area classified according to whether their value lies within this interval (grey), above (green) or below the lower interval limit (brown). As with the trends, this map shows that the pixels with anomalies are spatially grouped. This spatial continuity shows that the results are scientifically consistent and useful for site management purposes, marking as they do sufficiently extensive areas for carrying out activities that would have been non-viable if the result had been a series of small areas (pixels) spread randomly throughout the territory. Adapted from Oyonarte et al. [32].

Decision-taking support system

The above analysis of spatial anomalies has already provided us with a short list of areas where action is most necessary. Given that available resources are limited, however, finer priorities need to be set in terms of which particular areas are to be acted on. To this we propose a decision matrix based on a combination of the information generated by the system on time trends and spatial anomalies. The decision matrix shown in figure 7 enables us to prioritise pixels in terms of their spatial anomaly in relative C-APAR (columns) and relative time trend (rows).

The matrix rates priority from least (1) to most (9): low (yellow cells), medium (orange cells) and high (red cells). The bracketed figure shows the percentage of land surface occupied by each class. In this case the top priority would correspond to areas with a low relative C-APAR and negative trend, while areas with a high C-APAR and positive trend are considered to be low priority areas.

  Mean APAR 1
Lower Normal Higher
C-APAR 2  time trend  Positive 5
(5,3%)
2
(15,4%)
1
(5,9%)
Moderate 7
(9,5%)
4
(42,9%)
3
(15,0%)
Negative 9
(1,2%)
8
(3,7%)
6
(1,2%)

Figure 7. Decision matrix to support management. This helps to evaluate management needs and priorities of a given area in terms of annual absorbed radiation and the trend over time. To interpret this matrix it is necessary to take into account the management objectives. This case, applied to zones of disperse scrubland, shows an example for the selection of reafforestation areas.

(1) Mean of the absorbed photosynthetically active radiation (C-APAR) in the time series 2000-2008. The pixels are classed in terms of their value against benchmark values for the type of land use: Normal: within the variability interval (mean ± standard deviation); Lower: below the lower interval limit; and higher: above the upper limit.
(2) Trend classes for the time series. The pixels are classified in terms of the Mann-Kendall test slope  (see Trends section). Classes: Negative, values below 0 (negative trend);Moderate, values from 0 to 15 (positive trend); and Positive, values over 15 (positive trend). In this example no consideration is given to the degree of significance offered by the Mann-Kendall test.

Implementation of an alert monitoring system similar to the one developed here is feasible and recommendable in all types of protected areas as a backup to their adaptive management

Early alert system

Management of a protected area calls for up-to-date information on the ecosystem’s state of health and the existence of any changes as a guide to management actions, for example, tailoring the livestock load each year to the particular conditions of the ecosystem. Figure 8 gives a quick, intuitive check of an ecosystem against average behaviour in the past, showing, for example, the class of «scattered scrubland» in P.N. Cabo de Gata.

Preliminary versions of this tool already up and running include the «System for estimating and monitoring canopy productivity in real time» (SegF, http://larfile.agro.uba.ar/labsw/sw/gui/Inicial.page) and the Land Ecosystem Change Utility for South America (LechuSA, http://lechusa.unsl.edu.ar/, set up, respectively, by the Regional and Remote-sensing Analysis Laboratory of Buenos Aires and the Environmental Studies Group of the University of San Luis.


Figura 8. Gráfico para el sistema de alerta temprana. Las líneas continuas indican el rango promedio + una desviación estándar (en este caso 2001-2007), y la línea gruesa con puntos muestra la dinámica del año más reciente. Un dato cada 16 días. Adaptada de Oyonarte et al. [32].

Ejemplos preliminares ya en curso de esta herramienta lo constituyen el «Sistema de estimación y seguimiento de la productividad forrajera en tiempo real» (SegF, http://larfile.agro.uba.ar/labsw/sw/gui/Inicial.page) y el «Land Ecosystem Change Utility for South America» (LechuSA, http://lechusa.unsl.edu.ar/, desarrollados por el Laboratorio de Análisis Regional y Teledetección de la Universidad de Buenos Aires y por el Grupo de Estudios Ambientales de la Universidad de San Luis, respectivamente. 

Conclusions

Implementation of a monitoring and alert system similar to the one developed herein is feasible and recommendable in all types of protected areas as support to their adaptive management. The analyses offered by this system enable changes in ecosystem functioning to be pinpointed exactly in space and detected early so that they can be acted upon before they cause structural alterations that might then be very difficult to reverse afterwards. This improves the husbanding of natural resources, heads off threats and increases confidence in the decision-making process. Furthermore, the system is based on the monitoring of an intermediate [42] or backup [43] ecosystem service, namely net primary production (NPP); it therefore directly plums the «state of health» of the protected areas and brings home the advantages they provide for mankind. This enables site managers to operate more effectively in the legal and political spheres to promote public appreciation of the protected resources.

Finally, the satellite images used are free, so the system is low cost and fairly simple to set up. It then builds up information systematically and is written in open source software so the system is versatile and can be customised and its analysis capacity can be increased to phase in new monitoring needs of site managers and scientists.

ACKNOWLEDGEMENTS

Our thanks go to several collaborators who participated in the project: Agustín Giorno, Siham Tabik, Elisa Liras, David Vinazza, José M. Molero, Lucas Sevilla
García, Antonio Castro. To Jeff Burkey (King County) for developing the basic code for the trend test. The authors have been subsidised by the following institutions and programmes: FUNDACIÓN MAPFRE (R&D projects 2008), Organismo Autónomo de Parques Nacionales (project 066), Junta de Andalucía and ERDF (excellence projects RNM1280 and P09-RNM5048), Inter-American Institute for Global Change Research (IAI, CRN II 2031 and 2094), and  the Agreement «Desarrollo rural y sostenibilidad ambiental: diseño y ejecución de programas de seguimiento». The satellite images were downloaded from the MODIS Land website and the website of the Land Long-Term Data Record.


TO FIND OUT MORE

  1. Sala, O. E.; Chapin, F. S.; Armesto, J. J.; Berlow, E.; Bloomfield, J.; Dirzo, R.; Huber-Sanwald, E.; Huenneke, L. F.; Jackson, R. B.; Kinzig, A.; Leemans, R.; Lodge, D. M.; Mooney, H. A.; Oesterheld, M.; Poff, N. L.; Sykes, M. T.; Walker, B. H.; Walker, M.; Wall, D. H., Global biodiversity scenarios for the year 2100. Science 2000, 287, (5459), 1770-1774.
  2. van der Putten, W. H.; de Ruiter, P. C.; Bezemer, T. M.; Harvey, J. A.; Wassen, M.; Wolters, V., Trophic interactions in a changing world. Basic and Applied Ecology 2004, 5, (6), 487-494.
  3. Valladares, F.; Peñuelas, J.; de Luis Calabuig, E.; Moreno-Rodríguez, J. M., Impactos sobre los ecosistemas terrestres. In Evaluación preliminar de los impactos en España por efecto del cambio climático, Ministerio de Medio Ambiente: Madrid, Spain, 2005; pp 65-112.
  4. Fernández-González, F.; Loidi, J.; Moreno-Saiz, J. C.; Moreno-Rodríguez, J. M., Impactos sobre la biodiversidad vegetal. In Evaluación preliminar de los impactos en España por efecto del cambio climático, Ministerio de Medio Ambiente: Madrid, 2005; pp 183-248.
  5. Paruelo, J. M.; Garbulsky, M. F.; Guerschman, J. P.; Jobbagy, E. G., Two decades of Normalized Difference Vegetation Index changes in South America: identifying the imprint of global change. Int. J. Remote Sens. 2004, 25, (14), 2793-2806.
  6. Alcaraz-Segura, D.; Cabello, J.; Paruelo, J. M.; Delibes, M., Trends in the surface vegetation dynamics of the National Parks of Spain as observed by satellite sensors. Appl. Veg. Sci. 2008, 11, 431-440.
  7. Alcaraz-Segura, D.; Baldi, G.; Durante, P.; Garbulsky, M. F., Análisis de la dinámica temporal del NDVI en áreas protegidas: tres casos de estudio a distintas escalas espaciales, temporales y de gestión. Ecosistemas 2008, 17, (3), 108-117.
  8. Alcaraz-Segura, D.; Cabello, J.; Paruelo, J. M.; Delibes, M., Assessing protected areas to face environmental change through satellitederived vegetation greenness: The case of the Spanish National Parks. Environ. Manage. 2009, 43, (1), 38-48.
  9. Alcaraz-Segura, D.; Liras, E.; Tabik, S.; Paruelo, J.; Cabello, J., Evaluating the consistency of the 1982- 1999 NDVI trends in the Iberian Peninsula across four time-series derived from the AVHRR Sensor: LTDR, GIMMS, FASIR, and PAL-II. Sensors 2010, 10, (2), 1291-1314.
  10. Hannah, L.; Visconti, G.; Beniston, M.; Iannorelli, E. D.; Barba, D., The role of a global protected areas system in conserving biodiversity in the face of climate change. In Global Change and Protected Areas, Beniston, M., Ed. Kluwer Academic Publishers: Dordrecht, 2001; pp 413-422.
  11. Barber, C. V.; Miller, K. R.; Boness, M., Securing protected areas in the face of global change: Issues and strategies. IUCN: Gland, Switzerland and Cambridge, UK, 2004; p 236.
  12. Pyke, C. R., The implications of global priorities for biodiversity and ecosystem services associated with protected areas. Ecology and Society 2007, 12, (1).
  13. Yaffee, S. L., Three faces of ecosystem management. Conservation Biology 1999, 13, (4), 713-725.
  14. Ludwig, J. A.; Tongway, D. J.; Bastin, G. N.; James, C. D., Monitoring ecological indicators of rangeland functional integrity and their relation to biodiversity at local to regional scales. Austral Ecology 2004, 29, (1), 108-120.
  15. Costanza, R.; Norton, B. G.; Haskell, B. D., Ecosystem Health: New goals for environmental management. Island Press: Washington D.C., 1992; p 279
  16. Grumbine, R. E., What is ecosystem management. Conservation Biology 1994, 8, (1), 27-38.
  17. Zorn, P.; Stephenson, W.; Grigoriev, P., An ecosystem management program and assessment process for Ontario National Parks. Conservation Biology 2001, 15, (2), 353-362.
  18. Milchunas, D. G.; Lauenroth, W. K., Inertia in plant community structure: State changes after cessation of nutrient enrichment stress. Ecological Applications 1995, 5, 1195-2005.
  19. Costanza, R.; d'Arge, R.; de Groot, R.; Farber, S.; Grasso, M.; Hannon, B.; Limburg, K.; Naeem, S.; O'Neill, R. V.; Paruelo, J. M., The value of the world's ecosystem services and natural capital. Nature 1997, 387, 253-260.
  20. di Bella, C. M.; Rebella, C. M.; Paruelo, J. M., Evapotranspiration estimates using NOAA AVHRR imagery in the Pampa region of Argentina. International Journal of Remote Sensing 2000, 21(4), 791- 797.
  21. Liang, S., Narrowband to broadband conversions of land surface albedo I Algorithms. Remote Sensing of Environment 2000, 76, 213- 238.
  22. Running, S. W.; Thornton, P. E.; Nemani, R. R.; Glassy, J. M.; Sala, O. E.; Jackson, R. B.; Mooney, H. A.; Howarth, R. W., Global terrestrial gross and net primary productivity from the Earth observing system. In Methods in Ecosystem Science, Springer-Verlag: New York, USA, 2000; pp 44-57.
  23. Pettorelli, N.; Vik, J. O.; Mysterud, A.; Gaillard, J. M.; Tucker, C. J.; Stenseth, N. C., Using the satellitederived NDVI to assess ecological responses to environmental change. Trends in Ecology & Evolution 2005, 20, (9), 503-510.
  24. Nemani, R.; Hashimoto, H.; Votava, P.; Melton, F.; Wang, W.; Michaelis, A.; Mutch, L.; Milesi, C.; Hiatt, S.; White, M., Monitoring and forecasting ecosystem dynamics using the Terrestrial Observation and Prediction System (TOPS). Remote Sens. Environ. 2009, 113, (7), 1497-1509.
  25. Crabtree, R.; Potter, C.; Mullen, R.; Sheldon, J.; Huang, S.; Harmsen, J.; Rodman, A.; Jean, C., A modeling and spatio-temporal analysis framework for monitoring environmental change using NPP as an ecosystem indicator. Remote Sens. Environ. 2009, 113, (7), 1486-1496.
  26. Alcaraz-Segura, D.; Paruelo, J. M.; Cabello, J., Baseline characterization of major Iberian vegetation types based on the NDVI dynamics. Plant Ecolog. 2009, 202, 13-29.
  27. Paruelo, J., La caracterización funcional de ecosistemas mediante sensores remotos. Ecosistemas 2008, 17, (3), 4-22.
  28. Margules, C. R.; Pressey, R. L., Systematic conservation planning. Nature 2000, 405, (6783), 243-253.
  29. Garbulsky, M. F.; Paruelo, J. M., Remote sensing of protected areas to derive baseline vegetation functioning characteristics. Journal of Vegetation Science 2004, 15, 711-720.
  30. Paruelo, J. M.; Piñeiro, G.; Oyonarte, C.; Alcaraz-Segura, D.; Cabello, J.; Escribano, P., Temporal and spatial patterns of ecosystem functioning in protected areas in southeastern Spain. Appl. Veg. Sci. 2005, 8, (1), 93-102.
  31. Paruelo, J.; Fernández, N.; Delibes, M., Conservation, land-use changes and ecosystem functioning: The case of the Doñana area in southern Spain. Environmental Conservation En revisión.
  32. Oyonarte, C.; Alcaraz-Segura, D.; Oyarzabal, M.; Paruelo, J.; Cabello, J., Sistema de apoyo a la gestión de reservas de la biosfera basado en el monitoreo de la productividad primaria: ensayo en Cabo de Gata-Níjar (Almería-España). In Reservas de la Biosfera. Su contribución a la provisión de servicios ecosistémicos. Experiencias exitosas en Iberoamérica, Araya Rosas, P.; Clüsener Godt, M., Eds. UNESCO: Paris, 2010; pp 119-154.
  33. Grigera, G.; Oesterheld, M.; Pacín, F., Monitoring forage production with MODIS data for farmers' decision making. Agricultural Systems 2007, 94, 637-648.
  34. Monteith, J. L., Solar radiation and productivity in tropical ecosystems. Journal of Applied Ecology 1972, 9, 747-766.
  35. Oyarzabal, M.; Paruelo, J. M.; Federico, P.; Oesterheld, M.; Lauenroth, W. K., Trait differences between grass species along a climatic gradient in South and North America. Journal of Vegetation Science 2008, 19, (2), 183-1U1.
  36. Ruimy, A.; Saugier, B.; Dedieu, G., Methodology for the estimation of terrestrial net primary production from remotely sensed data. Journal of Geographycal Research 1994, 99, (3), 5263-5283.
  37. Piñeiro, G.; Oesterheld, M.; Paruelo, J. M., Seasonal variation in aboveground production and radiation- use efficiency of temperate rangelands estimated through remote sensing. Ecosystems 2006, 9, (3), 357-373.
  38. Garbulsky, M. F.; Peñuelas, J.; Papale, D.; Ardö, J.; Goulden, M. L.; Kiely, G.; Richardson, A. D.; Rotenberg, E.; Veenendaal, E. M.; Filella, I., Patterns and controls of the variability of radiation use efficiency and primary productivity across terrestrial ecosystems. Global Ecology and Biogeography 2010, 19, 253-267.
  39. Baldi, G.; Nosetto, M. D.; Aragón, M. R.; Aversa, F.; Paruelo, J. M.; Jobbagy, E. G., Long-term satellite NDVI datasets: Evaluating their ability to detect ecosystem functional changes in South America. Sensors 2008, 8, 5397-5425.
  40. Larson, W. E.; Pierce, F. J., The dynamics of soil quality as a measure of sustainable management. Soil Science 1994, 35, 37-51.
  41. Cao, M.; Prince, S. D.; Small, J.; Goetz, S. J., Remotely sensed interannual variations and trends in terrestrial net primary productivity 1981–2000. Ecosystems 2004, 7, (3), 233-242.
  42. Fisher, B.; Turner, R. K.; Morling, P., Defining and classifying ecosystem services for decision making. Ecological Economics 2009, 68, (3), 643-653.
  43. Millenium Ecosystem Assessment, M. E. A., Ecosystems and human well-being. synthesis report. Island Press: Washington D.C., 2005; p 155.

go top


The catchment area of the Laureles stream, in Uruguay, made up by pastureland and woodland, is one of the six areas selected for this project.

The National Parks of Iguazú in Argentina and Brazil, chosen for the project, are an excellent framework for gauging changes in ecosystem functioning.

The vegetation of the Nature Park Cabo de Gata-Níjar in Almería (Spain), is dominated by semi-arid scrubland sharing many species with the arid zones of Africa and Asia

Doñana Nature Park takes in Mediterranean woodland and scrubland and saltmarsh