Remote sensing for environmental control of Dociostaurus maroccanus (Orthoptera, Acrididae) in Apulia (Italy)
Belviso Luciano(1) and Pizza Maria(2)
(1) University of Bari, via Amendola 165/A, 70100 Bari, Italy
(2) University of Bari, Via Amendola 165/A, 70100 Bari, Italy
Dociostaurus maroccanus (Orthoptera, Acrididae) is a locust widely spread all over the Mediterranean region. This insect is linked to particular oviposition sites (pseudosteppa or garigue). Females oviposit their egg-pods undergound, only in compact soils, with patch vegetation, like wild pastures.
Young locusts larvae feed on wild grasses. When they become adults, the spontaneous vegetation dries and locusts feed on cultivated crops, expanding on wide areas of vegetations and farmlands, threatening economical operators of agriculture and food processing.
In the past years, observation attempts have been made, however, the localization of concentrations of locusts as well as their egg-pods is extremely difficult. Moreover, most of the affected areas are under the protection of National Parks Authorities that makes unacceptable any chemical pesticide treatment.
The objective of this study, performed in cooperation with the University of Bari, is to obtain a behavioural model of locusts propagation using data from remote sensing, in order to predict locusts concentration and promptly act with appropriate biological countermeasures, like enthomopathogenous fungi.
Although insects are affected by climate changes and meteo factors, it is extremely difficult to assess their precise behaviour and the relations among different environmental factors and the global insect population.
The infestation of the past years is probably due to the increase of annual temperature mean of just a couple of degrees but no explicit relation is currently known.
Considering this phenomenon affects wide areas, remote sensing represent the only applicable solution to obtain precise data concerning land coverage, surface temperature, incidence of solar radiation as well as atmospheric contribution.
After data collection, clusters are selected using neural networks methods in order to classify globally big territories.
At present, experiments (using MODIS, ASTER, ENVISAT) show that wide areas can be grouped in just a few clusters with some of them suitable for locusts life.
In the next months further investigation will be performed to refine the classification procedure.
Metamodels are extensively used to simulate complex non-linear systems. Since there is no explicit evidence of a relation betweeen locust life cycle and a particular environmental factor (such as temperature, humidity, etc.) this technique allows to consider all variables and obtain their specific weight in term of importance for insect life.
Different metamodeling techniques have been compared in order to underline these relationships and to explain how to use them for improving population description and modeling.
The use of metamodels can be helpful in order to make consideration on critical factors identification, those variables whose variance is the most responsible for the overall variance of the result.
Since all meteo and environmental factors are know thanks to satellite images taken during the year, simulation techniques will be investigated during the next months to assess relationships between observed variables and insect life-cycle.
Keywords: ESA European
Space Agency - Agence spatiale europeenne,
observation de la terre, earth observation,
satellite remote sensing,
teledetection, geophysique, altimetrie, radar,
chimique atmospherique, geophysics, altimetry, radar,