Table 3 Analysis of papers supporting standpoint 2 (S2)

Londo et al. (2011)

The advantage of a genetic modification depends on the context where plant lives (duration and strength of selective pressure).

When a plant naturally evolves resistance to herbicide, such resistance provokes fitness cost when herbicide selection is not applied (metabolic drain).

Even if one transgene is disadvantageous, its frequency can be increased by the selection of the second transgene (Hitch-hiking effect).

P1: variability of biological process

H1: Stacking transgenes could be a metabolic drain and decrease fitness:

P2: variability of entity behavior

H2: The costs of each transgene might not be additive, therefore might not be foreseeable from analysis of GM single.

Evaluating the fitness of GM stack and GM (single) lines of Canola relative to control, non-transgenic lines in a common garden environment, under different selective pressures.

Mesnage et al. (2013)

Pesticide residues and herbicides co-occur in the plants, synthetized by the plant itself (GM stack) and/or through external pesticide treatment.

Toxicity studies that check the real effects of combination of toxins are missing.

P1: variability of entity behavior

H1: Pesticides and herbicides in combination could have a real effect on human cells that is not predictable

Evaluating in a sensitive model of human cells in vitro toxicity of a mixture of Bt toxins and Glyphosate formulations

Ben Ali et al. (2014)

Stability of the genetic insert is an important aspect to ensure food feed safety. Guidelines say that the insert must not change during the cultivation and propagation.

Studies showed instances of rearrangement and alteration of the genome in GM plants during the post-release monitoring.

New methods for profiling of epigenome, transcriptome, proteome have identified changes in some GM varieties compared to unmodified comparators.

P1: variability of biological process

H1: GM stacks might be more sensitive to instability and alterations of the transgene. Such alterations might not be identifiable with the low-sensitivity methods used in normal risk assessment (Southern Blot).

Identifying DNA alterations in the transgenic DNA fragment of GM plants, particularly in GM stacks.

Evaluating the impact of the identified mutations in the aminoacid sequence of the transgenic proteins

Agapito - Tenfen et al. (2014)

Compositional and nutritional comparison between GM stacks and GM (single) might not be fit to reveal unintended effects

There is no comparison between molecular characterization of GM stacks and GM (single)

P1: variability of biological process

H1: some effects specific to the process of stacking might remain undetected by traditional analysis.

P1: variability of entity behavior

H1: The combined transgene could provoke variations outside the range found in GM (single)

Evaluating changes of protein profiles in stacked events versus single events and control plants. Comparing the level of transgene expression in stacked events compared to single events and control plants