Is Metal Leakage from Aluminum Foil without Adverse Effects? A Study on Ants as Models

Citation: Cammaerts MC, Cammaerts R (2018) Is Metal Leakage from Aluminum Foil without Adverse Effects? A Study on Ants as Models. J Nutr Health Sci 5(1): 103. doi: 10.15744/2393-9060.5.103 Volume 5 | Issue 1 Journal of Nutrition and Health Sciences


Liquid Containing Aluminum Given to the Ants
A piece of aluminum foil (3cm×5cm=15cm 2 ) was cut in 10 parts (of 1.5 cm 2 ) which were then set into 100 ml of sugar water, then kept at -15 °C as a stock solution. The pieces of aluminum foil remained in this stock solution all along the entire experimental work. The tap water used had a pH of 7.75 [50]. The sugar water containing aluminum was delivered to the ants in their usual feeder tubes (which had a capacity of 5 ml). One to two pieces of aluminum foil were always present in these feeder tubes. It was checked each day if ants effectively drunk the provided solution, and they did. The cotton plug shutting the tubes was refreshed every 2-3 days, and the entire solution was renewed every 7 days.

Sugar Water and Meat Consumption, General Activity
The ants drinking the sugar water, eating the T. molitor larvae, and being active anywhere in their habitat were counted six times per day during six days, at the same times o' clock each time (Table 1, Daily counts), as done for previous works [49]. The mean of these daily counts were established (Table 1, Daily means) and the six daily means obtained for ants consuming sugar water containing aluminum were compared to those obtained for ants under normal diet using the non-parametric test of Wilcoxon [51]. The means of the daily means were also calculated ( Table 1, Average of daily means).
As previously [42,43,45], the experiments were made on ants freely moving in their tray. The linear and angular speeds were quantified giving no stimulus to the ants; the orientation to an alarm signal was quantified presenting them with a nestmate tied to a piece of white paper ( Figure 1A). Such a tied worker emits its alarm pheromone produced by its mandible glands. As previously [49], for each variable, the trajectory of 20 ants of each colony (n=20 ants×2 colonies=40 trajectories) was recorded on glass and then copied on an appropriate substrate which remained affixed to a PC monitor screen due to its static electricity charge. As in previous works [46,47,49], the recorded trajectories were analyzed using specifically designed software [52]. The linear speed (in mm/s) is the length of a trajectory divided by the time spent for travelling it. The angular speed (in ang. deg./cm) is the sum of the angles, measured at several successive points of the trajectory, between the segment 'point i to point i-1' and the following segment 'point i to point i+1' , divided by the length of the trajectory. The orientation to a given point (in ang. deg.) is the sum of the angles, measured at several successive points of the trajectory, between the segment 'point i of the trajectory-given point' and each segment 'point i-point i+1' , divided by the number of measured angles. If the obtained value is lower than 90 °, the animal has a tendency to orient itself towards the point; if it is larger than 90°, the animal has a tendency to avoid the point. Each distribution of 40 values was characterized by its median and quartiles ( Table 2, lines 1,2,3) and the distributions obtained for ants consuming aluminum were compared to those obtained for ants under normal diet, using the non-parametric χ2 test [51]. Experimental details and statistics are given in the text. Aluminum increased the ants' sinuosity of movement, decreased their orientation ability, their trail following, their audacity, and largely their tactile perception Table 2: Effect of aluminum on six physiological and/or ethological traits

Speeds of Locomotion and Orientation Ability
The method is described in many previous works [46,47,49]. The trail pheromone of Myrmica ants is produced by the workers' poison gland. A solution of 10 of these glands in 500 µl of hexane was made and set for 15 min at -25 °C. To make an experiment, 50 µl of the solution was deposited, thanks to a metallic normograph pen, on a circle (R=5 cm) pencil drawn on white paper and divided into 10 angular degrees arcs. One minute later, this artificial trail was deposited in the ants' tray, and the response of 20 ants of each colony to that trail was assessed by the number of arcs of 10 angular degrees the ants walked along the trail without departing from it ( Figure 1B). The distribution of the 40 values was characterized by its median and quartiles ( Table 2, line 4), and the distribution obtained for ants consuming aluminum was compared to that obtained for ants under normal diet using the non-parametric χ² test.

Trail Following Behavior
The method has already been used in several previous works [46,47,49]. A tower standing on a platform, both being made of strong white paper (Steinbach , height=4 cm; diameter=1.5 cm), was set in the ants' tray, and the ants present at any place of this unknown apparatus were counted 12 times over 12 min ( Figure 1C). The mean and the extremes of the obtained values were established ( Table 2). The values obtained for the two colonies as well as those obtained during each successive time period of two minutes were pulled and the results for ants under aluminum and normal diets were compared using the Wilcoxon test. Figure 1: Some views of the experiments. A: an ant under normal diet having reached a tied nestmate which emits its alarm pheromone. B: an ant under aluminum diet following for a time a circular trail. C: three ants under aluminum diet not inclined in coming onto an unknown and risky apparatus An ingested product may adversely impact the central and the sensitive nervous system. This is why we assessed the ants' locomotion on a rough substrate (and later, on several cognitive abilities), doing so after the ants consumed aluminum for 5 days. Ants correctly perceiving the uncomfortable character of the substrate walk with difficulties, slowly, sinuously, while ants poorly perceiving such a character walk more confidently, more quickly and less sinuously. As previously [46,47,49], a folded piece (length: 3 cm, width: 2+7+2=11 cm) of rough emery n ° 280 paper was tied to the bottom and the borders of a tray (15 cm x 7 cm x 4.5 cm), the tray becoming so divided into a small zone 3 cm long, a zone 3 cm long where ants' moving was difficult ( Figure 1D), and a 9 cm long smooth zone. Each colony had its own apparatus (3+3+9 cm=15 cm, the length of the tray). For each of them, 12 ants were deposited, at a time, in the small zone. When moving away from that zone, the ants walked for a time on the rough paper. During that time, their speed of locomotion and their sinuosity were assessed (n=12 trajectories×2 colonies=24; Table 2, line 6). The values obtained for ants consuming aluminum were compared to those obtained for ants under normal diet using the non-parametric χ² test.

Tactile ('Pain') Perception
Ants of colonies A and B set their brood far from their nest entrance, at an inaccessible place inside the nest. The usual protocol allowing studying the ants' care of brood could no longer be used (removing larvae from the nest and assessing the ants' relocation of them inside the nest) and a novel one was set up. Before the ants consumed aluminum, then after they had consumed that product for 7 days, we observed the brood of each colony, together with the ants located on and in the vicinity of the brood ( Figure 1E). During each observation, we recorded over 15 minutes, at the end of each minute, the number of ants present on and very near the brood (='p'), as well as the number of ants interacting with the brood (='i'), i.e., licking, moving, relocating a larva, regurgitating food. For each kind of diet, the 15 'p' numbers as well as the 15 'i' numbers recorded for each colony were added and the mean proportion of 'p/i' was calculated ( Table 3, line 1). The numbers 'p' and 'i' obtained for ants under aluminum diet were compared to the corresponding numbers previously obtained for ants under normal diet using the non-parametric χ² test for 2×2 contingency tables [51].  Experimental details and statistics can be found in the text. Aluminum did not impact the ants' brood caring and their behavior in front of nestmates and aliens, but largely impacted their cognition and ability in escaping from an enclosure. p: number of ants present on the brood; i: number of ants interacting with the brood; aggressiveness levels: 0: doing nothing, 1: antennal contact, 2: mandibles opening, 3: gripping, 4: stinging; var a =n ° (2+3+4)/n ° (0+1) Table 3: Effects of aluminum on five physiological and/or ethological traits This protocol was set up when studying the effects of nicotine [53]. Two folded pieces of white strong paper (Steinbach , 12 cm×4.5 cm) were inserted in a tray (15cm×7cm×4.5cm) for creating a path with twists and turns between an initial small loggia and a large one. Each colony had its own apparatus. For each of them, 15 ants were set all together in the initial loggia, and just after, the ants present in this initial loggia and in the large one were counted after 30s, 2,4,6,8,10 and 12 min. The numbers obtained for the two colonies were added (Table 3, line 2), and the sums obtained for ants consuming aluminum were compared to those obtained for ants under normal diet using the non-parametric Wilcoxon test.

Cognition
This treat was assessed as previously [46,47,49]. Five dyadic encounters with a nestmate and with an alien were realized for each colony. Each encountering was conducted in a small cylindrical cup (diameter=2 cm, height=1.6 cm), the borders of which being slightly covered with talc. Each time (5×2=10 encounters with nestmates, 5×2=10 encounters with aliens), one ant of colony A or B was observed during 5 min and its encounter with the opponent was characterized, as previously, by the number of times it did nothing (level 0 of aggressiveness), touched the other ant with its antennae (level 1), opened its mandibles (level 2), gripped and/ or pulled the other ant (level 3), tried to sting or stung the other ant (level 4) ( Figure 1F, G). The numbers recorded for the two colonies were added (Table 3, lines 3, 4), and the results obtained for ants under aluminum diet were compared to those obtained for ants under normal diet using the non-parametric χ² test. As in previous works [46], the ants' aggressiveness was also quantified by 'a' , equaling the number of recorded aggressiveness levels 2+3+4 divided by the number of recorded levels 0+1.

Aggressive Behavior against Nestmates and Aliens
As in previous works [46,47,49], for each colony, six ants were enclosed in a reversed polyacetate glass (h=8 cm, bottom diameter=7 cm, ceiling diameter=5 cm) set in the ants' foraging area. The rim of the bottom of the glass was provided with a small notch (3 mm height, 2 mm broad) for giving to the ants the opportunity of escaping from the enclosure ( Figure 1H, I. To quantify such ability, the ants still under the glass and those escaped after 30 s,2,4,6,8,10 and 12 min were counted. The results obtained for the two colonies were added (Table 3, line 5), and the sums obtained for ants consuming aluminum were compared to those obtained for ants under normal diet using the non-parametric Wilcoxon test. As previously [42,43,45], we also evaluated the ants' ability in escaping by the variable "n ° of ants escaped after 12min/12". The protocol of these experiments have been set up many years ago and used several times, among others in [46,47,49]. Here, we used it again for studying the effect of aluminum on learning and memorizing capabilities, experimenting on colonies A and B after they had consumed aluminum for 7 days, and using control results previously obtained on colony C, never provided with that compound, while studying the effects of statins [54]. At a given time, a yellow hollow cube under which the ants could go was set above the entrance of the sugar water tube, the ants undergoing so visual operant conditioning. One week later, after the end of that visual conditioning experiment, pieces of basilica were deposited all around the entrance of the sugar water tube, the ants undergoing then olfactory operant conditioning. Tests were performed over time, while ants were expected to acquire conditioning and after removal of the cue, while they were expected to lose it. The ants were individually tested in a Y-apparatus constructed of strong white paper, and set in a small tray (30cm×15cm×4cm), as explained in previous studies [same references

Visual and Olfactory Conditioning and Memory
as above]. The Y-apparatus was provided with a yellow hollow cube or pieces of basilica in one branch. Half of the tests were conducted with the cue in the left branch and the other half with the cue in the right branch. Moving into the branch containing the cue was considered as giving the correct response ( Figure 1J, K). For each test, 10 ants of each colony were tested, the numbers of ants under aluminum diet being thus 10 ants×2 colonies=20 ants, and of ants under normal diet, n=10. The percentage of correct responses was established for each test (Table 4). The results obtained for ants under one and the other diet were compared thanks to the non-parametric Wilcoxon test. Ants were trained to a visual, then an olfactory cue, and tested over time in a Y-apparatus provided with the cue in one of its branches. Details and statistics are given in the text. Aluminum drastically impacted the ants' ability in acquiring conditioning, and thus their short and middle term visual and olfactory memory. *: control results previously obtained in [54] Table 4: Impact of aluminum on ants' conditioning ability and memory After the ants lived under a diet with aluminum during 10 days, their angular speed was again assessed (Table 5, line 1) in the manner it had been before consuming the food complement and after they had consumed it for one day, in order to examine if ants went on walking sinuously or if they became adapted to the effect of that compound on their locomotion. The distribution of values obtained after 10 days of aluminum consumption was compared to the control one and to that obtained after one day of aluminum consumption using the non-parametric χ² test. Linear speed Angular speed on a rough substrate Angular speed in ang.deg./cm; linear speed in mm/s; normal diet=sugar water diet; Al diet=sugar water+aluminum diet. These traits, i.e., sinuosity and tactile perception, initially impacted by aluminum (see Table 2), were still impacted after 10 or 11 days of Al ingestion. Ants developed thus no adaptation to the adverse effects of aluminum Table 5: Adaptation to the impact of aluminum on sinuosity of movement and on tactile perception (i.e., locomotion on a rough substrate)

Adaptation to Aluminum Consumption
After the ants consumed aluminum since 11 days, their locomotion on a rough substrate (so, their tactile and pain perception, impacted by that compound) was again quantified (Table 5, line 2) in the manner it had been before the ants consumed aluminum and after they had consumed it for 5 days, in order to know if ants adapted themselves to the impact of aluminum on their perception. The numerical results obtained after 11 days of consumption were compared to the control ones and to those obtained after 5 days of consumption using the non-parametric Wilcoxon test.
This trait was not examined, a decision taken after having examined all the here above potential effects of aluminum.

Habituation to Aluminum Consumption
After the ants lived under an aluminum diet during 12 days, we conducted an experiment in order to reveal if they became dependent on that product. The protocol of this experiment was identical to that used for examining potential ants' addiction to different substances [46,47,49]. For colony A and for colony B, 15 ants were removed and set in a small tray (15cm×7cm×5cm) in which two tubes (h=2.5cm, diam.=0.5cm) had been deposited, one containing sugar water, the other containing a sugar solution of aluminum i.e., the solution used throughout the hole study ( Figure 1L). In one tray, the tube containing aluminum was deposited on the right, and in the other tray, it was deposited on the left. After that, the ants seen drinking each liquid were counted 15 times over 15 min. The two sums of these two different counts were compared to those which should have been obtained if ants randomly went drinking each kind of provided liquid, using the non-parametric goodness of fit χ² test [51]. On the basis of these two sums, we also established the proportion of ants which have chosen each kind of liquid. The protocol of such a study is detailed in previous works, for instance in [49]. A fresh solution of aluminum was provided to Loss of the Effects of Aluminum after Its Consumption was stopped the ants 12 hours before the start of the present experiment (and a first assessment was made at t=0) which began just when that solution of aluminum was replaced by an aqueous solution of sugar, free of the compound. Since that weaning time, the ants' angular speed (=sinuosity) was assessed just like it had been before the ants consumed the food complement, after they had consumed it for one day and after they had consumed it for 10 days, except that 20 instead of 40 ant's trajectories were recorded for being able to make the assessments in the course of the experimentation. These assessments were performed after several time periods. The obtained numerical results were compared to those obtained at t=0 (see here above) and to the control ones, taking these two latter values as control groups in a non-parametric Kruskal-Wallis ANOVA for multiple comparisons [51], using Statistica v.10 software. The non-parametric χ² test was also used for comparing some distributions. All tests were two-tailed. The results of the present experiment are given in Table 6, and graphically presented in Figure 2. The experiment ended when the ants' angular speed became similar to that presented under normal diet.  Table 6, details are provided in the text. The effect of aluminum scarcely decreased during 4 hours, then decresed slowly and linearly during 12 more hours, losing 1/12 of its intensity each hour. Such a slow, regular decrease accounted for the absence of addiction to aluminum use After having again assessed the ants' sinuosity while under aluminum diet (note that the effect of aluminum on that trait increased over its consumption time), weaning began when the ants' solution of aluminum was replaced by pure sugar water. Since that time t=0, the ants' sinuosity was assessed until it equaled again the control one. The effect of aluminum decreased slowly, linearly over time, and vanished in 15 to 16 hours. The ants' sinuosity in function of weaning time is plotted in Figure 2. Details are given in the text. Columns 3 and 4: statistical comparison with control values; columns 5 and 6: statistical comparison with values at t=0; K-W test=non-parametric Kruskall-Wallis ANOVA; χ²test=non-parametric χ² test (bilateral tests) Table 6: Loss of the impact of aluminum on the ants' sinuosity of movement after its consumption was stopped

Results
These physiological traits were impacted by aluminum consumption ( . This result was significant (N=6, T=20, P=0.031) and the amount of active ants (consuming aluminum) increased over the six experimental days. A larger activity due to aluminum may explain the slightly larger consumption of sugar water. Such an increase of general activity was very obvious while experimenting, essentially during the late evening (Table 1).
Aluminum affected the ants' sinuosity of movement but not their speed of locomotion ( Table 2, lines 1 and 2). Under that diet, the ants walked at a linear speed nearly similar to that presented while under normal diet (13.7 vs 14.1 mm/s), the difference between the two speeds of locomotion being not significant (χ²=1.94, df=3, 0.50<P<0.70). While observing the ants, it was obvious that the insects' locomotion under aluminum diet differed from that under normal diet. Numerical results revealed that the difference consisted in an increase of sinuosity (148 vs 126 ang.deg./cm), this increase being at the limit of significance (χ²=7.64, df=3, P~0.05). An increase of sinuosity of movement together with an increase of general activity (see above) was spectacular during some evenings (personal observation). Apparently, this impact of aluminum on the ants' physiology continued days after days, in the course of that metal consumption. An experiment was made after 10 days of consumption for checking this presumption (see below ' Adaptation to aluminum consumption').

Linear and Angular Speeds
This physiological and ethological trait was affected by aluminum consumption ( Table 2, line 3). Under normal diet, the ants oriented themselves very well towards a tied worker (38.3 ang. deg.; Figure 1A). When consuming aluminum, these ants did so not so well (59.9 ang. deg.). The difference of orientation capability between ants under one and the other kind of diet was significant (χ²=16.89, df=2, P<0.001). The large sinuosity of ants consuming aluminum may explain this decrease of orientation capability. However, the ants' olfactory perception (or sensorial perception in general) may also be affected, making thus their orientation towards an alarm signal more difficult. Following experiments examined if aluminum impacted the ants' perception (see below 'Trail following behavior' and 'Tactile (pain) perception').

Orientation towards an Alarm Signal
This physiological and ethological trait was impacted by aluminum consumption ( Table 2, line 4), what was obvious while experimenting ( Figure 1B). Ants living under normal diet followed a circular trail along meanly 12.5 angular arcs of 10 °. When consuming aluminum, they followed such a trail along meanly 5.0 angular arcs of 10 °. The difference of trail following ability between ants consuming or not aluminum was significant: χ²=23.26, df=2, P<0.001. Such a decrease of trail following ability resulted from the ants' increase of sinuosity while consuming aluminum, but may also be due to some decrease of their olfactory perception (or sensorial perception in general). The latter cause was submitted to experimentation in a following experiment (see below 'Tactile (pain) perception').

Trail Following Behavior
This trait was slightly affected by aluminum consumption ( Table 2, line 5). When living under normal diet, meanly 1.25 ants were seen on the presented unknown apparatus. When consuming aluminum, only 0.90 were meanly seen on that apparatus. This was obvious while experimenting ( Figure 1C). Even if running continuously on their foraging area, near their nest entrance as well as inside their nest, the ants consuming aluminum were thus less incline in making risky tasks than while consuming no aluminum. However, the difference of straightforwardness between ants under the two kinds of diet was not significant (N=3, NS) probably due to the smallness of the sample.

Audacity
This physiological trait was impacted by aluminum consumption ( Table 2, the two last lines) and this was obvious while experimenting ( Figure 1D). Under normal diet, the ants moved very slowly and very sinuously on a rough substrate (4.7mm/s, 257ang.deg./cm). While consuming aluminum, they moved more frankly on such a substrate, just like if they less perceived its uncomfortable character (8.4mm/s, 192ang.deg./cm). The difference, between ants under the two kinds of diet, of linear speed as well as of sinuosity on a rough substrate, was significant (linear speed: χ²=30.48, df=2, P<0.001; angular speed: χ²=23.12, df=2, P<0.001). Aluminum impacted thus the individuals' sensorial perception, a presumption we had on basis of previous experiments (see above 'Orientation towards an alarm signal' and 'Trail following behavior').

Tactile (Pain) Perception
This trait was not at all impacted by aluminum consumption (Table 3,  We could thus conclude that, in the absence of any contradictory information, aluminum did not affect the individuals' social relationship, a first conclusion again submitted to experimentation in a following handing (see below ' Aggressiveness against nestmates and aliens').
Concerning visual conditioning, ants consuming this compound could never acquire any conditioning. After 72 hours of training, they still presented the null score of 50%, while ants living under normal diet reached the excellent score of 80%. This difference was statistically significant (N=6, T=-21, P=0.016). Aluminum impacted thus the ants' short and probably also middle term visual memory. The ants under aluminum diet having retained nothing of their training, we could not assess the impact of aluminum on their long term memory. We could only state that such ants went on correctly finding their nest entrance and their food sites. Their long term visual memory was thus not, or at least not very largely, impacted by aluminum consumption.
These physiological traits were largely impacted by aluminum consumption (Table 4; Figure 1J, K).

Visual and Olfactory Conditioning and Memory
This trait was largely impacted by aluminum consumption (Table 3, last line). While under normal diet, ants could escape from the enclosure with some delay. After having been enclosed for 6 minutes, 5 ants among 12 could escape and after 12 minutes, 11 ones could do ( Figure 1H). While living under aluminum diet, the ants behaved completely differently. We never saw such a behavior before, even when examining effects of drugs impacting the nervous system [41,48]. When under aluminum diet, the enclosed ants walked sinuously all around and also along the rim of the enclosure. However, when being in front of the exit, they went on walking forwards without going out of the enclosure, or looked for a few seconds through the exit, then moved backwards and went on moving inside the enclosure. Among the 12 enclosed ants, no one could escape after 6 minutes, and only 2 could do so after 12 minutes ( Figure 1I). The variable assessing the ants' ability in escaping equaled 0.91 for those under normal diet and 0.17 for those under aluminum diet. The difference of escaping ability between ants under one and the other kind of diet was significant (N=6, T=21, P=0.016). This experimental result showed that aluminum impacted the ants' cognition and straightforwardness, confirming so previous results (see above: ' Audacity' and 'Cognition').
This trait, reflecting the individuals' physiology and ethology, was impacted by aluminum consumption (Table 3, line 2). Under normal diet, six ants could cross the twists and turns path and reach the area lying beyond, and 10 ones were still moving on the area lying in front of the difficult path, in the course of the 12 experimental minutes. When consuming aluminum, only 4 ants could reach the area lying beyond the twists and turns path, and 16 ones were still in front of it after the 12 experimental minutes. The difference of ants' ability in crossing the twists and turns path between ants under one and the other kind of diet was significant (area in front of the difficult path: N=7, T=28, P=0.008: area beyond: N=5, T=+5.5, -9.5, P=0.359, the latter result may be due to the smallness of the sample). Since the ants' cognitive abilities (present experiment) and their sensorial perception (three previous experiments, see above) were affected by aluminum consumption, this metal may affect their nervous system in general.
A following experiment checked this hypothesis (see below 'Visual and olfactory conditioning and memory').

Cognition
These traits were not at all affected by aluminum consumption (Table 3, lines 3 and 4).

Aggressiveness against Nestmates and Aliens
Concerning the aggressiveness against nestmates, ants consuming aluminum behaved as while living under normal diet ( Figure  1F). They never gripped or stung their opponent; they contacted them with their antennae, and sometimes, slightly opened their mandibles. The latter act was however a little more often exhibited by ants consuming aluminum (this is why the variable assessing the ants' aggressiveness equaled 0.10 and 0.16 for ants under normal and aluminum diets respectively) but, in fine, the difference of behavior between ants under the two kinds of diet was not significant (χ²=2.74, df=2, 0.20<P<0.30).
As for the ants' aggressiveness against aliens ( Figure 1G), it was slightly higher for ants consuming aluminum than for those under normal diet, the variable assessing it equaling respectively 7.08 and 4.50. In fact, ants under aluminum diet less often contacted their opponent with their antennae, and more often gripped it than ants under normal diet. However, this slight difference (due to the high activity and perhaps the lower perception of ants consuming aluminum) did not make the difference of behavior between ants under one and the other kinds of diet significant (χ²=3.76, df=3, 0.20<P<0.30).
Thus, aluminum did not impact the relationship between members of a colony, as well as that with aliens. This result was in agreement with that concerning the ants' care of brood, another non impacted intra-colonial relationship (see above).

Ability in Escaping From an Enclosure
From 4 h to 16 h after weaning, the effect of aluminum (E t ) decreased over the running time (t), (from its initial value at t=0 (E t0 ) until a value identical to the control one) according to the linear function: During the first four hours after weaning, the effect of aluminum on the ants' locomotion only very weakly decreased. After that, it slowly and linearly decreased, vanishing in a total of 15 to 16 h (  20). It was only after 16 hours of weaning that the ants' sinuosity was statistically different from (smaller than) that at t=0 (χ²=13.31, df =2, P~0.001). The effect of aluminum fully vanished thus in a total of 16 hours after its consumption was stopped ( Table 6).

Decrease of the Effect of Aluminum on the Ants' Sinuosity of Movement after Its Consumption was stopped
As for olfactory conditioning, an identical severe impact of aluminum was observed. After 7 hours of training, the ants presented the null score of 50% while the ants under normal diet had reached the high score of 90%. Such a result was of course statistically significant: N=6, T=-21, P=0.016. Since the ants consuming aluminum acquired no olfactory conditioning, the impact of that compound on their olfactory memory could not be assessed. We could only affirm that aluminum affected the ants' short and middle term memory, and that, probably, it did not, or only slightly, affected the long lasting olfactory memory since the ants went on knowing the odor of their alarm pheromone, trail pheromone, nestmates, brood and queens.

Adaptation to Adverse Effects of Aluminum
Aluminum impacted the ants' locomotion, increasing their sinuosity of movement (see above and Table 2, line2). After having consumed this metal for 10 days (let us recall that the ants' feeder tubes always contained one or two pieces of aluminum foil), the ants went on moving with a large sinuosity, presenting thus no adaptation to the impact of aluminum on their locomotion. Their sinuosity was even somewhat larger after 10 days than after one day of consumption ( We can thus definitively conclude that ants did not adapt themselves to the adverse effects of aluminum on their locomotion and tactile perception. Since the effect of aluminum on ants' conditioning and memory was examined after the ants had consumed that compound for 7 days and until they consumed it for 7 more days, and since these ants' traits were found largely impacted by aluminum, we could also conclude that no adaptation occurred as for the impact of aluminum on their nervous system.
We found no favorable effect of that compound on any examined ants' physiological and ethological traits. We could thus not study habituation.

Habituation to Effect of Aluminum
Briefly, when experimented after 12 days of aluminum consumption, the ants presented no dependence on that product ( Figure  1L). Indeed, when confronted to sugar water containing aluminum and to sugar water free of that metal, 12 ants of colony A as well as 13 ants of colony B were counted on the former liquid while 16 ants of colony A and 13 ants of colony B were counted on the latter liquid. In total, 25 ants preferred the liquid containing aluminum and 20 ones the liquid without this metal. This corresponded to 46

Dependence of Aluminum Consumption
Aluminum impacted the ants' tactile perception (see above and Table 2, last line). After having consumed this compound for 11 days, the ants were still less sensitive to the uncomfortable, painful character of a rough substrate. This was obvious while experimenting, and was confirmed by the obtained numerical ( We observed no adaptation to the adverse effects of Al foil, on the contrary: these effects seemed to increase in the course of the metal consumption. We observed no addiction to aluminum, and this was in accordance with a slow decrease of its effect after its consumption was stopped. Since there is no addiction to aluminum for ants, it is highly probable that there is also no addiction for humans (e.g., they do not try having aluminum in their food, drink or cosmetic). It should thus be easy for humans to live without the presence of aluminum in the numerous products they used, and consequently to find palliatives. Unfortunately, for economical reasons, most people cannot stop using aluminum cookware. However, Al migration can be decreased by a factor up to 60 thanks to a preliminary immersion of the utensils in a more than 90° water during 5 hours, a process modifying the passivation layer of the foil by forming a low porosity AlOOH interface [61]. This protection against leaching still exists at higher temperature. Anyway, we recommend the advice of Bassioni et al. [28] to not use Al foil for cooking.
A supplementary point must be related. Due to man-made activities including the acidification of rainwater and freshwater and of that of soils by intensive farming, the inert aluminosilicates of the earth crust are now increasingly transformed into aluminum hydroxide and organoaluminum complexes, enhancing in this way our exposure to aluminum. This metal moreover reacts with phosphates which become then less bio available [2]. For environmental reasons also, it is thus recommended to reduce our use of aluminum.
Another form of aluminum, Al hydroxyde, a compound poorly soluble in water, but injected as a vaccine adjuvant, has nevertheless been found to be neurotoxic. In contrast to control mice injected with a saline physiological serum, those injected with Al hydroxyde suffered from apoptosis of motor neurons and from motor function impairments (such as a lower linear speed and a higher sinuosity) as well as from spatial learning and memory decreases [57]. The present study also showed that ants fed with sugar water containing pieces of Al foil walked with a higher sinuosity, were less able to orient themselves and to follow a trail. They also presented a large deficit of cognition, learning ability and memory. Moreover, they had their audacity and pain perception reduced. Thus, ants clearly revealed the effects of Al on the nervous system and muscular systems, and our results converged with those of Shaw & Petrick [57]. An epidemiological study indicates a correlation between the presence of Al in injected vaccines and the occurrence of autism spectrum disorders [8], revealing once more an impact of Al on the brain functioning. Note that our study on ants, as well as numerous ones of other researchers, never found any impact of Al on social relationship.

Discussion
Using ants as models, we found that sugar water containing pieces of Al foil affected the ants' locomotion, ability in orientating and following a trail, audacity, tactile perception, cognition, ability in escaping from an enclosure, learning, and memory. We now compare our results with those of other researchers on the subject.
Rats receiving, during 6 months, water containing Al dissolved as a salt (Al chloride) showed elevated anxiety: they presented higher ambulatory, rearing activity and defecation index. Their memorization during spatial learning was lower than that of untreated rats, what revealed some cognitive deficit associated with aluminium administration [10]. When routinely orally consumed by rats at human-relevant levels, Al chloride has also been found to impair memory in aging individuals, this cognitive decline being positively correlated with the Al ingested dose. The cognitive score of these middle and old aged rats was inversely correlated with the percentage of Al loaded in brain entorhinal cortex cells [55]. A following experiment on Al-feed rats showed that increasingly more Al-rich microtubule-depleted pyramidal cells became incorporated in hippocampus lesions when Al gradually accumulated in corticolimbic brain regions, a situation possibly leading to dementia [56]. In these experiments, as in the present study in which we showed that ants' locomotion and memory were affected, Al was present in the water given to the animals. This confers to AL a more than ten times greater bioavailability than when present in solid food [4].
A possible association between Al in drinking water and the risk of Alzheimer's disease was shown in 9 among 15 former epidemiological studies on human populations [4] and an evident link between ingested Al and this disease was provided by Tomljenovic [3].
A theory explaining how Al finally injures the central nervous system has been proposed by Shaw and co-authors [58]. Al may disrupt the water-based cellular homeostasis and so cause a cascade of events which begins with inflammatory reactions leading to autoimmune actions impairments and ends by injuring the central nervous system. Indeed, there is a growing evidence that Al ingested through drinking water promotes the rate of brain aging caused by heightened inflammatory activity within brain tissue [7,59]. Furthermore, an histochemical and immunocytochemical research [60] suggests that enhancement of inflammation and pathological changes of cholinergic fibres may be the modes of action through which Al may cause learning and memory deficits.
It lost 165-126=39 ang.deg./cm of its full effect in 16 h-4 h=12 h, and thus 39/12=3.25 every hour. This hourly rate represented 3.25/39=1/12 of the total effect. Thus, four hours after the consumption of aluminum was stopped, its effect decreased of 1/12 each hour, what was a slight (slow) decrease, probably not perceived by the individuals, and this accounted for the absence of dependence to aluminum [Cammaerts, accepted by the Journal 'Frontiers in Physiology']. During the first four hours after weaning, the effect of aluminum nearly did not decrease. Consequently, after weaning, aluminum lost its effect in 12 hours following a 4 hours period with no significant decrease.