Introduction

Genetically modified crops (GM) are the result of a breeding process through which features or characteristics such as resistance to certain insects and herbicide tolerance are introduced in commercial crops such as cotton (Gossypium hirsutum L.) (^{Gatehouse, 2008}). Bt commercial cotton shows a soil bacterium (Bacillus thuringiensis) as an insecticidal toxin source, this strain has different effects on their activity towards pests and insects and is a reserve of genes coding for insecticidal proteins (^{Silva, 2005}).

This technology is considered as an alternative instrument to modify and improve crops, especially cotton where losses caused by insects and weeds are highly significant. The implementation of crops (GM) has brought numerous benefits to agriculture advancement; however, its development and commercialization has been the subject of much debate and opposing positions (^{Conner et al., 2003}).

One of the main concerns is the possible loss or reduction of the diversity associated with these crops, especially because of the ecological impact that these plants can cause, the concern is particularly focused on the hybridization of GM crops since they can invade other species through repeated hybridization cycles causing genetic contamination (^{Singh et al., 2006}). Another important ecological risk of the release of transgenic crops is the possible negative effect it may have on non-pest organisms that generate benefits for the crop and agriculture in general so it becomes necessary and important to raise some alternatives to minimize the risk of Bt crops on non-target insects (^{Wisniewsky et al., 2002}; ^{Singh et al., 2006}).

It should be noted that previous papers do not report negative effects on the diversity of arthropod fauna associated with these crop types. A good model for analyzing negative effects of GM crops on non-target species are the papers on the effect of GM maize with tolerance to lepidopteran insects on the monarch butterfly. ^{Angharad et al. (2002}) reported a series of studies that were carried out to rigorously evaluate the impact of GM maize pollen on monarch butterfly larvae to quantify risk.The results showed that large-scale cultivation of Bt maize hybrids does not represent a great risk to monarch butterflies.

In this context, Colombia is not alien to this problem, which is part of a complex biological and social component due to the impact on diversity that would be reflected in the decrease of the beneficial arthropods community associated with the crop. In this way it would be expected to be able to determine the impact on arthropod fauna associated to the transgenic cotton compared to the conventional one.This paper evaluated arthropofauna in agroecosystems of transgenic and conventional cotton to establish whether there are differences in the arthropofauna diversity when this technology is implemented.

Materials and methods

Data collection

Arthropods were collected in commercial batches of Bt and conventional cotton in the municipalities of Cereté and San Pelayo. The evaluated plant material was: (1) isogenic line (DP 174) tolerant to glyphosate and (2) a genotype with Bt technology (FM 9162 B2F). In each locality, four lots were available (Table 1).

Table 1 List of commercial batches for growing Bt and conventional cotton in the municipalities of Cereté and San Pelayo. 

Sampling

On each batch five sampling methods were implemented according to the microhabitat (Figure 1), for soil arthropofauna the Berlesse and Pitfall method was used. In the first method the sample was taken at four sampling sites for each batch with a “cup cutter” type borehole capable of extracting 1 500 g of soil and 20 cm deepening, then carried to the Berlesse funnels for 24 h, Pitfall traps were located at ten points on the furrow and between plants, this method consisted of installing a perforated 12-ounce plastic cup at ground level, which was fixed throughout the crop cycle and corresponded to the Pitfall trap base. Inside this base a glass of 3.5 ounces was placed; the collections started seven days after the emergence of the crop (DEC), at 24 h after the traps were installed, for 15 consecutive days (^{Secretaría del Convenio sobre la Diversidad Biológica, 2000}).

Figure 1 Distribution scheme of sampling methods. 

In order to capture aerial arthropods and in herbaceous strata, entomological networks, adhesive traps and light traps were used. For sampling with entomological networks, 10 points were randomly selected in each lot, in which ten double passes of jama (pdj) were made; this was repeated weekly until the crop reached an average height of 1.5 m. Yellow plastic plates (PPA) of 20*10 cm dimensions and 2 cm² squares fixed in a wooden rod of 2.2 m long were used on sticky traps, as adhesive the plate was smeared with mechanical fat. These adhesive traps were located at ten points in each lot at the height of the crop from seven days after the emergence of the crop. Finally, a light trap was installed in each location, the readings were made during seven consecutive days per month during the crop development.

Results and discussion

A total of 23 787 individuals were collected during the sampling phase, with a total of 240 taxa grouped in 25 species and 215 morphospecies (4 identified untill order, 151 to family and 60 to genus). For the transgenic crop, the greatest number of individuals was recorded, with 12 286 (51.7%), registering 237 taxa grouped in 25 species and 212 morphospecies, for the conventional crop 11 501 individuals (48.3%) registering 240 taxa grouped in 25 species and 215 taxa morphospecies. Among the most abundant orders is Hemiptera representing the largest number with 6 241 individuals, being 50.8% for transgenic (TC) crop and for conventional crop (CC) 5513 individuals with 47.9%, followed by Coleoptera Order with 1 660 individuals for 13.8% CT and 1591 being 13.5% in CC, followed by the Hymenoptera Order with 1 224 individuals, which equals 10.2% for (CT) and 1 174 individuals; that is, 10% in CC, finally the Diptera Order recorded 9.5% with 1168 individuals in CT and 9.5% with 1 093 individuals in the CC.

As for wealth, the Coleoptera Order recorded 22 families (20%), being the group showing the greatest richness. In this way the artropofauna was represented mostly by the groups considered more important that were coleoptera, hemiptera, diptera, hymenoptera, arachnida, and lepidoptera. The diversity between the conventional and transgenic crop was not different, the abundance and species richness showed no significant differences between the two crop types, which coincides with that reported by ^{Almada et al. (2011}) by comparing the spider community in Bt and conventional cotton in northern Santa Fe (Argentina), no significant differences between crops were observed, only that the abundance of spiders increased in Bt crops, which could indicate that these crops did not affect the spider population.

The high abundance and richness reported in this study could be due to the fact that cotton cultivation, both conventional and transgenic, provides conditions that favor the establishment of arthropods, taking into account that the variations are influenced by the phenological development stage of the crop, small differences were evidenced between the two crops types, being the transgenic crop the one that had greater abundance, this can be attributed to the lower use of insecticides and herbicides compared with conventional cultivation, a fact that favors the arthropods diversity on its development and natural behavior in the transgenic crop, which coincides with those reported by several authors (^{Novillo et al., 1999}; ^{Durán et al., 2000}; ^{Pérez-Guerrero et al., 2009}; ^{Benamú, 2010}), who argue that the application of insecticides in the case of conventional crop exerts a negative effect on the decrease of diversity in arthropods populations.

According to the behavior and dietary habits of the arthropods present during the phenological development of the crop, taking into account the number of families present in the two crop types, only the most abundant ones were plotted and it was shown that in both crops they expressed in the same way, although their values were different, that is why a grouping was made taking into account the alimentary habit (phytophagous, predator and parasitoid) limited to a few families, being the most common: Aleyrodidae, Thripidae, Cicadellidae, Aphididae, Chrysomelidae, and Agromizidae for the phytophagous group, a marked difference was shown between the families represented in terms of abundance, being the Aleyrodidae family the most representative in both agroecosystems. The Mann-Whitney test showed no significant difference between the diversity of families and types of genetic material (U= 188, p= 0.7453).

As for the group of predators the abundance was higher in the transgenic crop, and in both agroecosystem the families Formicidae, Coccinellidae, Anthocoridae, Carabidae, Staphylinidae and Syrphidae were observed, being the most representative the Formicidae family accounting for 881 individuals; i. e., 7.7% for conventional crop and for the transgenic crop 969 individuals which is equivalent to 7.9%. This can be related to the type of action that predators exert on their prey (phytophagous), and to the symbiosis they can share, in the case of many species that excrete exudates that attract ants and that they in turn provide protection from other predatory species, from these results it can be inferred that the transgenic crop, depending on the implemented management strategy offers greater variety and availability of resources, which coincides with that reported by ^{Almada et al. (2011}). The Mann-Whitney test showed no significant differences between the diversity of families and the type of crop (U= 156, p= 0.8494).

For the parasitoids group the Eulophidae, Braconidae and Scelionidae families showed the highest abundance in both crops, with higher values in the conventional crop compared to the transgenic. The Student-t test showed no significant difference between the diversity of families and the type of crop (t-value= 1.6394; p= 0.1153), it can be said that this behavior is due to the number of hosts that are associated with the crop and for prey availability in each of them, the presence of parasitoids may be linked to lepidoptera complex, grouped within the main phytophagous for which Bt technology is directed to, noting that predators were similar for both crops; this is consistent with studies conducted by ^{Lobos et al. (2003}), where it is argued that this trend may be influenced by nutritional ecology, biological attributes such as specificity and seasonality, characteristics of the agroecosystem, particularities of its hosts and prey of preference.

Which is also backed by several authors in their research suggesting that transgenic cotton varieties that produce toxins of B. thuringiensis, have insecticidal attributes that may as well affect populations of target pests of this technology, and influence the dynamics of the beneficial fauna that is associated, such is the case of many that are beneficial such as predators and parasitoids who has been attributed with importance in controlling the crop associated phytophagous considered pests (^{Novillo et al., 1999}; ^{Durán et al., 2000}; ^{Pérez-Guerrero et al., 2009}; ^{Benamú, 2010}).

The marked difference in the number of individuals belonging to groups considered to be beneficial in the transgenic crop compared to the conventional one allows to believe that genetically modified crops favor the arthopofauna diversity, promoting a great decrease of applications of insecticides and avoiding the appearance of new pests that develop under some present conditions which matches with ^{Whitehouse et al. (2005}) when comparing communities between Bt and conventional crops, which shows that Bt crops generate indirect effects due to the reduction of broad spectrum insecticide applications that allow the presence of natural enemies that would control the potential pests that may appear. Proposed studies claim that the reduction of insecticides favors the arthropods diversity and the maintenance of a more uniform distribution of species, also report that the greater abundance and richness of species associated to the crop depends to a great extent on the vegetation structure and the one around it (^{Romeis et al., 2004}).

As for the richness of families present in the two types of crop, the Formicidae family was the most representative, followed by Chrysomelidae, Cicadellidae and Coccinellidae with greater number of genera, in contrast to the Agromyzidae, Aleyrodidae and Anthocoridae families being the least rich.

When ordering species by its abundance in each crop and comparing the phytophagous group, the Bemisia and Frankliniella genera were predominant for both crops, while Aphis gender showed more individuals in the transgenic crop, being lower for conventional, which in part might be explained by the lower use of insecticides in the crop with transgenic technology. The Mann-Whitney test showed no significant differences between gender diversity and crop type (U= 638, p= 0.9103).

For the predators group, must genera showed activity on the crop with abundance values representative for both (CC and CT) being the case of Sthetorus, Solenopsis, Brachimirmex, Paratrechina, Dorymyrmex and Orius. The Mann-Whitney test, showed no significant differences between gender diversity and crop type (U= 442.5, p= 0.5926). For the parasitoids group, Telenomus and Architas genres showed higher abundance in the conventional culture and otherwise to Euplectrus and Casinaria whose abundance was reflected with higher values in the transgenic crop. The Student-t test showed no significant differences between gender diversity and crop type (t-value= 2.4019, p= 0.5315).

When analyzing the arthropods community present for the two crop types;conventional and transgenic by range-abundance curves (Figure 2), it is generally observed that the slope of the graphs indicate a trend among species, curves were dominated by the most common for cotton agroecosystem such as Bemisia tabaci, Frankliniella occidentalis, Sthetorus sp., Aphis gossypi, Brachimirmex sp. and Solenopsis spp., which usually are in aggregations characteristics of their groups; which were reported as abundant in a paper by ^{Sosa et al. (2015}). In their paper on the diversity of arthropod community in transgenic cotton varieties, whereby it is concluded that the Bt and conventional cotton crops host arthropods without expressing differences in their community.

Figure 2 Abundance range curves of arthropods caught in cotton crops the two types of conventional and transgenic crop. Córdoba-Colombia, 2013-2014. 

In the case of Bemisia tabaci and Aphis gossypi for being gregarious insects and also for the growth of their populations with exponential trend tend to be captured in large numbers of individuals, these results go hand in hand to those reported by ^{Carapia et al. (2013}). A paper in which it is affirmed that these aggregations can be found by many factors like rains, high or low temperatures, relative humidity, and also to effects related to the ecology of the insect. Moreover, the greater abundance A. gossypi can be influenced by the presence of caterpillars that feed on plant structures which induces the production of terpenoids (^{Sosa et al., 2015}).

Meanwhile Brachimirmex sp. and Solenopsis sp. ants were very abundant since they are generally found in colonies and are closely related to the local vegetation, in this case with the cotton crop for food or shelter, results that resemble those described by Casanova (2014), who indicates that ants can be beneficial since they consume weed seeds that may affect the crop and they are also considered as bioindicators of environmental quality of ecological status of ecosystems assessment (^{Sosa et al., 2015}).

Rare or unusual species showed in Figure 2, reported in conventional breeding two species Diabrotica speciosa and Tasogodes orizicolus and one for the transgenic crop Linogeraus capillatus, which might have been caused by the impact of human activities affecting the population dynamics of the communities and the viability of populations. ^{Tapia et al. (2010}) refers that on their paper is highlighted the biology of rare species that can be shown through the low population size and the impact of human activities.

The analysis of completeness (Figure 3) by the batch sample coverage method yielded for the conventional crop a wealth value of S= 218 with a coverage deficit (DC) of 9.76 for lot 1, for lots 2, 5 and 6 were S= 181, D = 10.37; S= 185, DC= 6.72; S= 194, DC= 9.09 respectively. In relation to the lots for the transgenic crop, a wealth estimate of S= 185 was obtained with a DC of 8.72 for lot 3, and S= 183 with DC of 10.2 for lot 4, for lots 7 and 8 the wealth values obtained were S= 198, DC= 7.58; S= 174, DC= 8.47 respectively.

Figure 3 Completeness analysis for the types of crops studied in batches. (estimated wealth 95%, IC- sample coverage= 0.95). 

Species diversity for conventional and transgenic crop (Figure 4) represents the same trend when wealth is analyzed through alpha diversity true type ¹D, represented by the species richness (⁰D), with a larger number of effective species in the conventional crop (¹D= 36.39) compared to transgenic crop (¹D= 33.39). Considering the values obtained in the different orders of diversity, for the order ²D values of (6.81) in conventional breeding and (6.31) for the transgenic crop were obtained, based on these results no differences related to the three diversity profiles between conventional and transgenic crops were observed.

Figure 4 Profiles of true alpha diversity for arthropods of the conventional and transgenic cotton crops in Córdoba-Colombia. 

Considering the degree of change or replacement in the species composition between the two types of agroecosystems (Figure 5), a value of 1.013 with q= 0 was shown. For the beta diversity of order q= 1 and q= 2 of 1.001 effective communities. These results showed that there are no differences between the effective species of the two types of agroecosystems because the values are very close to 1, showing a similarity among communities.

Figure 5 True beta diversity profile for arthropods of the conventional and transgenic crops in Córdoba- Colombia. 

Regarding the analysis of true alpha and beta diversity these indicators showed values that do not differ significantly between the two crop types, indicating that for this research the diversity of arthropofauna in conventional and transgenic cotton crops is not different.Results that when compared with other researches such as that conducted by (^{Pilcher et al., 1997}; ^{Riddick et al., 1998}; ^{Armer et al., 2000}) show that modified species containing the Bt protein are highly selective and have no effect on the survival, development and feeding of non-target arthropod species. This contrasts with what is proposed in studies by ^{Hilbeck et al. (1999}); ^{Romeis et al. (2004}); ^{Andow and Zwahlen (2006}) suggest that long-term exposure of these organisms could affect soil biodiversity and in general the ecological functionality of soil as the toxin reaches the arthropods by means of root exudates.

In studies focused on the risk that may result from the use of transgenic plants on the diversity of arthropod species associated with crops by ^{Armer et al. (2000}), no negative effects were reported. In the same way, in this research, the results obtained coincide with those mentioned above. Many of the species are not affected, as they do not have a direct association with the plant, but to generalize that this type of plants does not have any effect on the associated arthropods groups is very risky.

Conclusions

In this research, the arthropofauna diversity associated with the cotton agroecosystem in conventional and transgenic crops was not different, there was no evidence of differences between arthropod populations in the transgenic crop compared to the conventional one. The small differences that were shown are not attributed to the use of transgenic, it is possible that there are external factors that may be influencing the dynamics of arthropod groups such as the chemical control implemented by the farmers that takes place during the development of the crop, In order to bring the incidence of arthropod pests and of arthropods to levels of low economic importance.