Arthropod communities within the olive canopy as bioindicators of different management systems

Arthropods, within the olive tree canopy, which can act as bioindicators of different management systems (conventional, integrated and organic), were collected in costal Croatia during 2007-2009. The aim of the research was to determine the arthropod orders present in each of the observed olive groves, to identify the number of individuals in each order, and to set potential bioindicators of different management systems. In order to do that, arthropods were collected from the olive canopies at three locations (Supetar, Kaštel Stari and Duilovo). After the samples were sorted according to orders, and determined according to arthropods determination keys. Olive groves were characterised by the same growth conditions and environment, with differences in agricultural management systems. In this investigation, 7882 arthropods were collected, belonging to 17 orders. It was found that Diptera, Heteroptera and other Hemiptera individuals, excluded Cicadidae and Psyllidae family individuals, could be defined as potential bioindicators of olive groves under different management systems. The number of Araneae and Coleoptera individuals significantly differed in all researched olive groves. So, these orders can be considered as bioindicators of different management systems. The knowledge obtained in this research can potentially be exploited in olive production in the future. Obtained results represent an important contribution to the inventory of arthropods in olive canopy. Also, this research significantly improves understanding of both harmful and beneficial entomofauna in olive groves and consequently contributes to better understanding of the total entomofauna in Croatia. Additional keywords: conventional olive grove; integrated olive grove; Olea europaea; orders abundance; organic olive grove. Abbreviations used: COG (conventional olive grove); GAP (good agricultural practice); IOG (integrated olive grove); OOG (organic olive grove). Authors’ contributions: Conceived and designed the experiments, interpretation of data, and performed the experiments: EV, SK, MK, PD and JIB. Wrote the paper: EV. Statistical analyses: MK and EV. Critical revision of the manuscript for important intellectual content, supervising the work: JIB and MI. Coordinating the research project: MK. Obtaining funding: SK. Citation: Vitanović, E.; Ivezić, M.; Kačić, S.; Katalinić, M.; Durbešić, P.; Igrc Barčić, J. (2018). Arthropod communities within the olive canopy as bioindicators of different management systems. Spanish Journal of Agricultural Research, Volume 16, Issue 2, e0301. https://doi.org/10.5424/sjar/2018162-12385 Received: 04 Oct 2017. Accepted: 22 May 2018. Copyright © 2018 INIA. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International (CC-by 4.0) License. Funding: Croatian Ministry of Science and Education (project 'Research & Development of Environmentally-Friendly Methods in Olive Pest Control', 091-0910468-0364). Competing interests: The authors have declared that no competing interests exist. Correspondence should be addressed to Elda Vitanović: elda@krs.hr


Introduction
Olive (Olea europaea L.) is the most widespread agricultural plant in the whole Mediterranean, including Croatia.For many years the pest management has been based only on the use of pesticides, of which the result was negative affect on the environment (Heim, 1985;Cirio, 1997;Civantos, 1999).The preservation of biological balance and the entire agroecosystem is the biggest challenge of today.Intensive agriculture is responsible for imbalance of olive grove biocoenosis by reducing diversity between the harmful and beneficial species, and by destroying their natural habitat (Brnetić, 1978;Lacroix & Abbadie, 1998).The excessive use of agrochemicals has triggered the need for development of different plant protection models with minimum negative effect to biodiversity.For that reason integrated and organic management is widely applied nowadays since they limit the use of pesticides.Olive is a host to many pests, the most numerous among which are insects with 110 species, followed by 90 species of pathogenic fungi, 13 species of mites, 11 nematodes, 5 bacteria and 5 species of birds (Cirio, 1997).Numerous investigations by international scientists show that the most numerous arthropods in the olive groves are spiders (Araneae) with about 50 species, followed by beetles (Coleoptera) with 30 species, ants (Hymenoptera, family Formicidae) with 23 species, Hemiptera with 11 species, Neuroptera with 13 species, and Diptera with one species (Morris, 1997;Morris et al., 1999;Campos & Civantos, 2000).Many of these arthropod species belong to the group of phytophagous insects, but also many belong to the group of predators and parasitoids, which are the natural enemies for the most important olive pests.This interaction of phytophagous species, predators and parasitoids is important for the sustainability and preservation of biological balance of an ecosystem (Andrewartha & Birch, 1984;Rodriguez et al., 2003;Garratt et al., 2011).
Integrated or organic plant protection is recommended in order to preserve diversity of arthropod community in olive grove.The analysis of soil, plant material and oil can determine the levels and types of used pesticides, and potentially confirm their overuse.Overuse of pesticides can also be determined by using arthropods, as members of the olive tree canopy.Such bioindication is a valuable tool, as it permits the evaluation of the overall state of conservation for a particular ecosystem, based on the living organisms that it contains (Burel et al., 2004;Jerez-Valle et al., 2014).For many arthropods, survival in agrosystems depends on the suitability of the habitat, which is strongly influenced by agricultural management systems (Jenneret et al., 2003;Jerez-Valle et al., 2014).The use of bioindicators as the reference point for different agricultural management systems does not have a long tradition (Van Straalen & Verhoef, 1997;Markert & Wunschmann, 2011), especially of arthropod communities, which have been used since the beginning of 20 th century (Cairns & Pratt, 1993;Brown, 1997).Bioindicators are defined as organisms or groups of organisms that react in a measurable way to environmental stress, especially to the use of pesticides (Meffe & Carroll, 1994;McGeoch, 1998;Igrc Barčić & Maceljski, 2001).Olive tree canopy is abundant in various species (Arambourg, 1986;De Andres, 1991;Varela & Gonzales, 1999;Campos & Civantos, 2000;Ruiz & Montiel, 2000) that can act as bioindicators of different management system (Cairns & Pratt, 1993;Brown, 1997;Cotes et al., 2011;Jerez-Valle et al., 2015;Gkisakis et al., 2016).These are numerous phytophagous species, among which a large number are olive pests, and many parasitoids and predators.Parasitoids are the most numerous, with about 300-400 different species of Hymenoptera (Arambourg, 1986).Predators comprise mostly the orders of Araneae, Coleoptera and Hymenoptera (family of Formicidae), which are the most diverse (Churchill, 1997;Morris, 1997;Peck et al., 1998;Morris et al., 1999;Marc et al., 1999;Holland & Luff, 2000;New, 2000).Excessive use of pesticides disrupts the balance of the olive tree canopy, which reflects on the number of species in Arthropoda phylum.The use of olive canopy arthropods at the level of order proved to be useful in discriminating between management systems, especially among organic and non-organic practice (Biaggini et al., 2007;Jerez-Valle et al., 2015).Depending on the management systems (conventional, integrated or organic) applied in olive groves, the presence of orders of arthropod and the number of their families and species is different in the canopy, so they can be used as bioindicators of changes caused by the applied management systems.
Negative effect of the use of pesticides on arthropod communities has been investigated by many authors (Heim, 1985;Cirio, 1997;Civantos, 1999;Ruano et al., 2001Ruano et al., , 2004Ruano et al., , 2010;;Scalercio et al., 2009;Santos et al., 2010;Cardenas et al., 2015), who confirmed that the lowest number of arthropods is found in the conventional olive groves (COG).According to research by Ruano et al. (2004), the arthropods collected in the olive canopy are more reliable bioindicators of different management systems than those collected in the soil using pitfall soil traps.These authors point out that the obtained results are likely to be directly affected by pesticide spraying of olive tree canopies.Orders of Coleoptera, Diptera, Heteroptera, Lepidoptera and Thysanoptera can be used as bioindicators of different pesticide treatments in olive groves.These authors especially mention Coleoptera and Lepidoptera as bioindicators of organic plant protection in olives.The claims of these authors are aligned with the selection criteria of bioindicators according to Cilgi (1994).Mentioned insect orders are also determined as sustainability bioindicators of many other agricultural crops (Fauvel, 1999;Kevan, 1999).
In Croatia, diversity of the arthropod fauna in the olive grove has not been investigated until this research, neither were arthropods so far investigated as bioindicators of different management systems.Having in mind the lack of research on this matter in Croatia, and the importance of arthropod communities in olive growing, their role as bioindicators has been selected as the topic of this research.The aim of the research was to determine the orders of arthropod present in each of the observed olive groves and to identify the number of individuals in each order.Based on the obtained results, the final goal was to determine the bioindicators of different management systems.This was the first project of this kind in Croatia, so the obtained results have brought important contribution to the inventory of arthropod communities in olive tree canopies.On top of that the results are setting a whole new direction in using arthropod communities as bioindicators of various agricultural management systems in a particular olive grove.It is important to mention that these results are also contributing to better understanding of harmful and beneficial entomofauna in olive orchards.

Study area and agricultural management systems
The research was done in three experimental olive groves in costal Croatia.First  The selected olive groves have been characterised by similar growth and environmental conditions, as well as the variety; the only difference was the applied agricultural management system in each of them.Each experimental grove covers the area of approx.0.5 ha, comprising about 125 trees of the autochthonous variety ˈOblicaˈ.COG is ten years old, while IOG and OOG are twenty five years old.Good agricultural practices (GAP) were applied in all olive groves.According to Köppen Climate Classification System, observed olive groves are located in the region with Mediterranean or dry summer climate -Csa (Koppen & Geiger, 1954).It is usually characterized by rainy winters and dry, warm to hot summers.
The soil tillage was done twice each year of the research in the COG.Mechanical tillage has been done only in the ground plan projection of canopy at the end of April and in mid-May.Vegetation in the grove between the rows was maintained by mechanical mowing.In March or April, spring fertilisation was applied during shallow mechanical soil tillage in all three years of the research.Each olive tree was fertilised with 1 kg of nitrogen fertiliser KAN N 27 (INA,Kutina,Croatia).During autumn fertilisation in November 1 kg of complex fertiliser N-P-K (7:14:21; INA, Kutina, Croatia) per tree was applied.Treatments against Bactrocera oleae Gmel.and the anthophagous generation of Prays oleae Bern.were applied every year during research.In July and September, treatment with systemic organophosphate insecticide (dimethoate) at the concentration of 0.15 % was applied against B. oleae, while for the anthophagous generation of P. oleae, a synthetic pyrethroid (deltamethrin) at the concentration of 0.03 % was used.It was applied in each year of the research at the beginning of flowering (phenophase F) in May.
In the IOG soil was managed the same as in the COG.During shallow mechanical treatment in the spring, fertilisation with 1.5 kg of inorganic fertiliser Hortyflor (Cedar agro d.o.o., Novi Marof, Croatia) per olive tree was applied.For autumn fertilisation of olive trees, 1 kg per tree of complex fertiliser N-P-K (7:14:21; INA, Kutina, Croatia) was used.In this grove, B. oleae and P. oleae were also treated each year of the research.Poisoned baits were used for B. oleae.In the first two years, a poisoned protein bait consisting protein hydrolysates as a food attractant and insecticide (spinosad) at the concentration of 3 % was used, while in the third year a bait containing 0.07 % of sexual attractant Polycore™ SKl (AgriSense BCS Ltd., Pontypridd, UK) and synthetic pyrethroid (deltamethrin), was used.Each year 1 dL of poisoned baits was used, applied five times according to B. oleae control programme (Katalinić, 1998).Against the anthophagous generation of P. oleae, a growth inhibitor based on the active substance diflubenzuron at 0.02 % concentration was used.
The soil in the OOG was managed in the same way as in the other two groves.Spring and autumn fertilisations were done during shallow mechanical tillage in three years of the research using 1 kg per tree of Hortyflor fertiliser (Cedar agro d.o.o., Novi Marof, Croatia).Against B. oleae, attract-and-kill method was applied using Eco-trap (Vioryl S.A., Afidnes, Greece).A microbiological insecticide containing Bacillus thuringiensis Berliner spores at the concentration of 0.1 % was used against the anthophagous generation of P. oleae.It was applied at the beginning of flowering (phenophase F).

Experimental design and arthropod collection
The samples were taken monthly from April to November during the years 2007-2009, as the time when the largest differences in composition of the arthropod communities are observed between applied management systems in olive groves and which has the highest arthropod abundance (Ruano et al., 2004;Jerez-Valle et al., 2014;2015).One sample contained fauna collected from the canopies of three olive trees of a smaller size or three branches of larger trees was randomly selected and put into linen bags of 4×2 m dimensions.Each sample was collected in three repetitions in three olive groves under different management systems (conventional, integrated and organic), in daytime between 8 am and 2 pm.The bags were firmly tied with a rope to keep the tree fauna inside.The fauna were collected by fogging through an opening in the bag with the solution of 9 mL of dichlorvos in 5 L of water using IGEBA TF 35 thermal fogger (IGEBA, Weitnau, Germany).The pipe was inserted through an opening in the bag and after 4 considering each year separately or grouping them, and also grouping samples in different sizes.

Presence and abundance of arthropods in the olive groves under different management systems
All arthropod fauna collected from the olive groves under different management systems were representatives of two classes, Arachnida and Insecta.In the COG, 15 different orders were found, among which two belong to the Arachnida class, and 13 to Insecta class.In the IOG, 14 orders were found, two of Arachnida and 12 of Insecta class, while in the OOG 14 orders were found, three belonging to Arachnida and 11 to Insecta class (Fig. 1).In the same Fig. 1 the percentages of each order collected in the olive groves with different management systems during 2007-2009, are shown.
In the research a total of 7882 arthropods were collected.The highest number (3433) was collected in COG, followed by IOG where 2569 arthropods were collected.The lowest number (1880) was collected in OOG (Table 1).In the same Table the total number of arthropods collected in all researched olive groves during three years of the research, is shown.The majority of collected arthropods in the COG belong to order of Hemiptera, followed consecutively by Hymenoptera and Araneae.The majority of individuals in Hemiptera order belonged to Cicadidae and Psyllidae families, and the majority of individuals in Hymenoptera order belonged to Formicidae family.In IOG and OOG, the majority of the collected arthropods belong to the order of Hemiptera, largely influenced by the high number of Cicadidae and Psyllidae individuals, followed by orders of Araneae and Coleoptera.
These finding indicates great diversity of arthropods in olive groves.According to the results of Shannon-Weiner (H') and Simpson (1-lambda) diversity indexes (Table 2), great diversity of arthropods between the olive groves under different applied management systems was detected.The greatest diversity of arthropods was found in the IOG, followed by OOG.The lowest diversity of arthropods was found in the COG.
The total number of each order of arthropods collected in the three years of research according to the observed olive groves is shown in Table 3.In the same Table also the arithmetic mean values and standard deviations of each order of arthropods in the fogging pulled out and the opening tightly closed with a rope.Ten minutes after the treatment, the bags were shaken well so that as many insects as possible would fall from the branches into the bag.Then the bags were carefully removed and taken to the laboratory for further examination.The contents of the bags were emptied on a large plastic foil to separate the fauna samples from the leftovers of the trees and other impurities.Each sample was examined with a binocular microscope and those of the Arthropoda phylum were separated and placed into 200 mL plastic bottles containing the solution of ethanol/water 70% (v/v) for further examination.The arthropods were further sorted according to orders, and each order was placed in a 1.5 mL Eppendorf tube containing the solution of ethanol/water 70% (v/v) and determined according to arthropods determination keys (Reitter, 1908;1909;1910;1911;Kuhnt, 1911;Winkler, 1932;Freude et al., 1964Freude et al., -1974;;Schmidt, 1970;Durbešić, 1988).The identification was confirmed at the University of Zagreb Faculty of Science and Faculty of Agriculture, Croatia.
The family Formicidae (Hymentoptera) was separated from the other Hymenoptera, and was considered as an independent group due to its abundance.Further, suborder Heteroptera was separated from the other Hemiptera individuals, as well as individuals of Cicadidae and Psyllidae families.They were also considered as three independent groups due to their abundance.

Data analysis
The Shannon-Weiner (H') and Simpson (1-lambda) diversity indexes were calculated at the order level for each olive grove (COG, IOG and OOG) using PRIMER 6 (Plymouth Routines In Multivariate Ecological Research) software.
For statistical analysis, SPSS 11.0.0 software (SPSS Inc., Chicago, IL, USA) was used.Descriptive parameters (arithmetic mean and standard deviation) were used to describe continuous variables.Shapiro-Wilk test was used to determine normality.The differences among the olive grove locations were examined with one-way variance analysis.Variance homogeneity was assessed by using Levene's test and Bonferroni post hoc test.In the case of significant differences, further tests were applied that are usually used when homogeneity or equality of variances are not found.
A discriminant analysis was performed using the groups that showed significant differences in abundant to establish whether differences among the management systems would arise.The discriminant analysis was performed in different ways, either

Arthropods as bioindicators of different management systems
Discriminant analysis (Fig. 2) was used to determine differentiating functions of each olive grove under different management systems during the three years of research.Comparison among different management systems showed no significant differences except two orders and one family.Only the significant variables (orders of Araneae and Coleoptera, and family of Psyllidae) were used as potential ones, which can discriminate the researched groves.The other groups did not differ statistically or there were lack of some specimens depending on the management system considered.The dependent variable was each of the researched olive groves.
Significance of the individual functions is visible in the eigenvalues shown in  2007-2009, are shown.The differences between the years were investigated using the analysis of variance (ANOVA), and when the number of individual samples was not sufficient, ANOVA was not used.Table 3 also show that there were no statistically significant differences in number of Araneae individuals between 2008 and 2009 (p>0.05).Significantly more samples were collected in the first than in the second year of research (p=0.012).Statistically significant difference can be seen in the number of Psyllidae individuals (Hemiptera), between years 2007 and 2008.Significantly lower number of Psyllidae family individuals was determined in the second year of the research.Also, there were no significant differences in the number of Coleoptera individuals between second and third year of the research in IOG.It can be seen that relatively higher number of Coleoptera individuals was collected in 2007 comparing to 2008 (p=0.05).There were no statistically significant differences in the number of arthropods in OOG among the observed years (Table 3).According to the results, the orders of Araneae, Coleoptera and family of Psyllidae (Hemiptera) showed statistically significant differences; therefore, their mean values were used in the discriminant analysis.Mantodea 2 0 1 3 N COG = number of arthropods in conventional olive grove; N IOG = number of arthropods in integrated olive grove; N OOG = number of arthropods in organic olive grove; N tot = total number of arthropods in all researched olive groves discriminant points were significantly different among the three groups, and the most outstanding separated olive grove was COG.It was characterised by a higher number of Araneae individuals, and a lower number of Coleoptera individuals than the other two researched olive groves.In 2009, the number of Coleoptera individuals contributed the most to the discriminant function.The percentage of correct classifications was high (73%).Arithmetic mean values of discriminant points were significantly different among the three groups, and the most outstanding separated olive grove was IOG.It had higher number of Coleoptera individuals than olive groves with conventional and organic management systems.

Discussion
Olive tree canopy is abundant in various orders of arthropods (Arambourg, 1986;De Andres, 1991;Varela & Gonzales, 1999;Campos & Civantos, 2000;Ruiz & Montiel, 2000), which can be bioindicators of different management systems (Cairns & Pratt, 1993;Brown, 1997;Jerez-Valle et al., 2015).Intensive use of pesticides disturbs the balance of the biocoenosis, which reflects in the number and diversity of arthropods.Therefore, they can serve as bioindicators of changes in the olive grove caused by different management systems.The less abundance of arthropods in the COG than in the groves with other management systems was expected as a consequence of negative effects of pesticide use.Namely, the negative effect of pesticide use on the arthropods has been researched by many authors (Heim, 1985;Cirio, 1997;Civantos, 1999;Ruano et al., 2004Ruano et al., , 2010;;Scalercio et al., 2009;Santos et al., 2010;Porcel et al., 2013;Cardenas et al., 2015;Jerez-Valle et al., 2015), who confirmed that the least number of arthropods was found in the COGs.Ruano et al. (2004) pointed out that the highest number of arthropods after two years of research was collected in the IOG.The second highest number of arthropods was collected in the COG.On the contrary, the results of this research show that the highest number or arthropods was collected in the COG, followed by IOG.The lowest number or arthropods was collected in OOG.In all researched olive groves, and especially in conventional, a large number of Cicadidae, Psyllidae and Formicidae was collected.One of the reasons for their high abundance is their ability to adapt to changes, which are consequence of the pesticide use.Ruano et al. (2004) stated that the family of Formicidae is less sensitive to the pesticide use.One of the reasons that explain high numbers in some of the samples is that they are eusocial insects (Morris et al., 2002).Their 1.00 2.24 0.00 0.00 0.00 0.00 --54 10.80 17.80 0.00 0.00 0.00 0.00 --Balttodea 0 0.00 0.00 0.00 0.00 0.00 0.00 --1 0.00 0.00 0.00 0.00 0.20 0.45 --Mantodea 2 0.40 0.55 0.00 0.00 0.00 0.00 --0 0.00 0.00 0.00 0.00 0.00 0.00 --number can be related to the presence of numerous species that excrete honeydew, on which they feed.If the number of arthropods in the researched olive tree canopies is compared excluding the individuals of Cicadidae, Psyllidae and Formicidae families, different results are obtained.In that case, the most abundant of arthropods was IOG (1422), followed by the OOG (857).COG had the least number of arthropods (723).
In this work, the highest number of arthropods was collected in the first year of the research.The reason that may explain this is the extremely high temperature recorded during the sampling period in 2007.High temperatures, especially in the beginning of vegetation, have probably caused the sudden appearance of a large number of arthropods.
In the first two years of research, the highest number of Araneae individuals was collected in the COG, while in the third year of the research it was collected in the IOG.The lowest number of Araneae individuals, on average, was collected in the OOG.The results of this research are aligned with those of Mansour (1984) and Van Den Berg et al. (1990), but are contrary to those of Cardenas et al. (2006) and Morris et al. (2002).The same authors reported the highest number of Araneae in OOG, where 114.6% more samples were collected than in conventional grove.The results of this research can be affected by diversity of Araneae, as Cardenas et al. (2006) mentioned in the research.Also, Marc et al. (1999) stated that the species within order of Araneae behave differently depending on their exposure to pesticides used in the agriculture with conventional management system.It is well known that Araneae is more resistant to organophosphorus pesticides, carbamates (Mansour, 1984) and pyrethroids (Van Den Berg et al., 1990) than to chlorinated carbohydrates.Mansour (1984) confirmed that some of the species of Araneae developed resistance.
In 2007 the highest number of Formicidae individuals was collected in the COG, which can be explained with proportionally high abundance of Cicadidae and Psyllidae individuals, as it is mention before.In other years of the research, the highest number of Formicidae individuals was collected from the IOG.Santos et al. (2007) investigated the arthropods in olive groves with integrated and organic management systems.The number of Formicidae individuals was the highest  (Finnegan, 1974;El Haidri, 1981;Arambourg, 1986;Gotwald, 1986;Sato & Higashi, 1987;Paulson & Akre, 1992;Way & Khoo, 1992;Majer & Delabie, 1993;Weseloh, 1993).In this work, the highest number of order of Hemiptera was collected in the COG.The lowest number of same arthropods was collected in OOG.Individually, the most numerous families of Hemiptera were Cicadidae and Psyllidae.On the contrary, the highest number of all other individuals from order of Hemiptera was in IOG and the lowest in COG.It is important to emphasise that all collected samples of Psyllidae belong to the same species, Euphyllura olivina Costa.Ruano et al. (2004) also mentioned this species as the most numerous within the family.The same authors pointed out that E. olivina was the most numerous species in the canopies of the olive groves with integrated (95%) and organic (79.3%) management systems.Suborder of Heteroptera is the most numerous in IOG and OOG.In the first two years of the research, the highest number of Heteroptera individuals was collected in the IOG, while in 2009 it was collected in the OOG.Many authors emphasise the sensitivity of Heteroptera to insecticides used in olive groves (Zeletzki & Rinnhofer, 1966;Paternotte et al., 1986).Severin et al. (1984) and Ruano et al. (2004) claimed that the species of this suborder can serve as bioindicators of different management systems.
In all years of research, the highest number of Coleoptera was collected from the IOG, followed by COG.Coleoptera is one of the most diverse orders in IOG, with the greatest diversity observed in the olive groves with integrated management systems.Many   researchers highlight its presence in the IOG and their importance, because it belongs to the group of predators (Churchill, 1997;Morris, 1997;Peck et al., 1998;Morris et al., 1999;Marc et al., 1999;Holland & Luff, 2000;New, 2000;Ruano et al., 2004).
The highest number of Diptera individuals was collected in the IOG, followed by organic grove.In COG, the number of Diptera individuals was the lowest, which was also indicated by other authors.Sol & Sanders (1959) and Sommaggio (1999) pointed out that the use of pesticides in agriculture significantly reduces the number of Diptera.Apart from them, Speight & Castella (2001), Ruano et al. (2004) and many others investigated the order of Diptera as potential bioindicator.According to them, the presence of Diptera in the olive grove and other agricultural crops indicates different management systems, as we can also conclude according to our results.
According to values of Shannon-Weiner and Simpson diversity indexes IOG was the most diverse that was due to fact that integrated management positively affects the abundance and diversity of arthropods, while conventional management negatively affects it.The results of arthropod diversity obtained in this research are aligned with those of Cardenas et al. (2006).
The olive biocenosis is stable and characteristic for a particular growth area and the environmental conditions typical for the olive grove.It responds in a measureable way to stress changes in the environment, especially to those caused by the pesticide use (Meffe & Carroll, 1994;McGeoch, 1998;Igrc Barčić & Maceljski, 2001).Diptera and suborder Heteroptera as part of olive tree canopy are potential bioindicators of different management systems.The number of Araneae and Coleoptera individuals differentiates significantly the olive groves under different management systems, so they can be considered as their bioindicators.Cardenas et al. (2006) also investigated the order of Araneae as bioindicators of different management systems.Similar results as ours were obtained by Ruano et al. (2004), according to whom the orders of Coleoptera, Diptera, Lepidoptera, Thysanoptera, and suborder of Heteroptera are potential bioindicators of different management systems.Orders of Coleoptera and Diptera are pointed out as bioindicators of the olive grove with organic management system.The findings of the above authors, as well as of the present research, match almost completely the criteria for the selection of bioindicators; they are extremely widespread, constant inhabitants of olive biocoenosis, always in great abundance, easy to collect and to determine, and are sensitive to pesticides.They are very important part of olive tree canopy because many of them are predators and parasitoids.The mentioned orders of insects have also been determined as bioindicators of sustainability in other agricultural crops (Fauvel, 1999;Iperti, 1999;Kevan, 1999).
In conclusion, the use of arthropod communities within the olive canopy could be useful in distinguishing between agricultural management systems, especially integrated and conventional.Diptera, Heteroptera and other Hemiptera individuals, excluded Cicadidae and Psyllidae family individuals, are potential bioindicators of olive groves under different management systems.The number of Araneae and Coleoptera individuals significantly differed all researched olive groves, so they can be considered as bioindicators of different agricultural management systems.According to the number of their individuals collected from the olive canopy, it is possible to determine, with great certainty, which management system was applied in particular olive grove.In the future, these two orders could potentially be used as bioindicators to determine the particular management system applied in a particular olive grove.In order to prove this we will expand investigation on more locations and on other Mediterranean crops.

Figure 2 .
Figure 2. Distribution of the total number of arthropods in 2007, 2008 and 2009 (discriminant points of individual functions) and the corresponding centroids.L1= conventional olive grove; L2= integrated olive grove; L3= organic olive grove

Table 4 ,
while Table 5 contains standardised coefficients that show which variables contribute most to discrimination among the groups.In 2008, none of the canonical functions was statistically significant (Table 4).In 2007 and 2009, there was only one discriminant function that statistically significantly differentiated the dependent variables.Orders of Araneae and Coleoptera contributed most to the discriminant function in 2007.The percentage of correctly classified cases was high (80%).In 2009, Coleoptera contributed olive groves under different management systems during

Table 1 .
Total number of arthropods in the olive groves with different agricultural management systems during three years of the research

Table 2 .
diversity indexes between the olive groves with different management systems in the three years of research

Table 3 .
Total number of order of arthropods in olive groves under different management systems (COG= conventional olive grove; IOG= integrated olive grove; OOG= organic olive grove) during2007-2009.

Table 4 .
Values of eigenvalue, variance percentages, canonical correlation, significance function indicator and the percentage of correctly

Table 5 .
Standardised coefficients for the variables of the arthropod orders included in the discriminant analysis (two canonical functions) depending on the year of investigation.