A four-generation feeding study with genetically modified (Bt) maize in laying hens

A four-generation study with at least 60 laying hens (LSL) and 10 cockerels (LSL) was carried out to investigate the influence of genetically modified maize (Bt 176) on animal health and feed intake, laying performance, feed efficiency, and hatchability of chickens and to compare GM-maize with its near isogenic counterpart. The chickens were divided into two groups (one pen/group) of at least 30 hens each and 3 cockerels. The diets contained 400 (chickens and pullets) or 500 g . kg-1(laying hens) isogenic or genetically modified maize (Bt 176), respectively. Feed and water were provided ad libitum. Eggs for hatching were collected when the laying hen was aged 31 weeks. In the 31st week of life, brooding eggs were collected and brooded for every group. One-day-old chickens from each group were sex sorted and allocated to one pen per group. There were no significant differences in composition between the two maize varieties. For every generation, as well as the average of all four generations, there was no significant influence on the feed intake of chickens (32.2 and 32.0 g per day), pullets (68.4 and 70.4 g per day) and layers (114.9 and 112.9 g per hen per day for control and Bt-maize), body weight of chickens (652 and 636 g), pullets after 18 weeks (1316 and 1305 g), and laying hens after 31 weeks (1656 and 1626 for control and Bt-maize), laying intensity (83.5% and 83.3%), fertility of eggs (96.6% and 97.5%), or hatchability of living chicks (86.8% and 88.9% for control and Bt-maize). In conclusion, feeding of 400 (grower) or 500 g . kg-1(layer period) Bt-maize to chickens, pullets and laying hens for four generations did not significantly influence feed intake, growth, laying or breeding performance compared with an isogenic counterpart.


Introduction
Since 1996 the cultivation of genetically modified plants (GMP) has increased from 1.6 million ha to about 170 million ha in 2012 (James, 2013) or around 12% of arable land on a global scale.Soyabean, maize, cotton seed and rapeseed (canola) are the dominating cultures (see James 2013 for more details).
Safety and nutritional assessment of feed/food from GMP is one of the starting points for public acceptance (Chassy, 2010;Kleter and Kok, 2010).International agreed documents (e.g., OECD, 1993;EFSA, 2008) recommend compositional analysis and comparison of isogenic with transgenic products.Feeding studies with laboratory and/or target animal species are only recommended if the compositional analyses, as well in vitro, in sacco, in silico, or 'omics' measurements are not able to answer all questions concerning feed/food safety (ILSI, 2003(ILSI, , 2007;;FDA, 2007;EFSA, 2008EFSA, , 2011)).

ABSTRACT.
A four-generation study with at least 60 laying hens (LSL) and 10 cockerels (LSL) was carried out to investigate the influence of genetically modified maize (Bt 176) on animal health and feed intake, laying performance, feed efficiency, and hatchability of chickens and to compare GM-maize with its near isogenic counterpart.The chickens were divided into two groups (one pen/group) of at least 30 hens each and 3 cockerels.The diets contained 400 (chickens and pullets) or 500 g • kg -1 (laying hens) isogenic or genetically modified maize (Bt 176), respectively.Feed and water were provided ad libitum.Eggs for hatching were collected when the laying hen was aged 31 weeks.In the 31 st week of life, brooding eggs were collected and brooded for every group.One-day-old chickens from each group were sex sorted and allocated to one pen per group.There were no significant differences in composition between the two maize varieties.For every generation, as well as the average of all four generations, there was no significant influence on the feed intake of chickens (32.2 and 32.0 g per day), pullets (68.4 and 70.4 g per day) and layers (114.9 and 112.9 g per hen per day for control and Bt-maize), body weight of chickens (652 and 636 g), pullets after 18 weeks (1316 and 1305 g), and laying hens after 31 weeks (1656 and 1626 g for control and Bt-maize), laying intensity (83.5% and 83.3%), fertility of eggs (96.6% and 97.5%), or hatchability of living chicks (86.8% and 88.0% for control and Bt-maize).In conclusion, feeding of 400 (grower) or 500 g • kg -1 (layer period) Bt-maize to chickens, pullets and laying hens for four generations did not significantly influence feed intake, growth, laying or breeding performance compared with an isogenic counterpart.
Recently, we (Flachowsky et al., 2012) described objectives of various types of feeding studies with laboratory and target animal species for safety and nutritional assessment of feeds from genetically modified plants, as shown in Table 1.In general, the costs of the studies mentioned in Table 1 increase from laboratory to target animals and from the top to the bottom of Table 1, but should be the highest for long-term and multi-generational studies with large target animal species.In the meantime, many feeding studies with laboratory animals, but also with food-producing animals (about 150) according the recommendations by EFSA (2008, 2011) and ILSI (2003, 2007) have been carried out (see summary by Flachowsky, 2013).
Long-term feeding studies and multigenerational experiments with target animals, especially pigs and ruminants, are very rare, as recently summarized by Snell et al. (2012) and Ricroch et al. (2013).High costs and limited feed amounts in earlier breeding stages may also restrict animal numbers and duration of such studies with large target animals.Recently, Sartowska et al. (2012) and Korwin-Kossakowska (2013) initiated an eight-generation project with Japanese quails and included soyabean meal (39.0% in grower; 29.5% in layer feed) and maize (25% in grower and layer feed) from GM-plants.In the publications cited above they reported the first results (animal performance, meat and egg quality after two generations, Sartowska et al., 2012; transfer of transgenic DNA-fragments in food of animal origin after four generations, Korwin-Kossakowska et al., 2013).To date, they did not find any significant effect on animal yields and no measurable transgenic DNA fragments in animal bodies or eggs.
Nevertheless, there are critical comments about the duration of feeding studies (e.g., Séralini et al. 2011Séralini et al. , 2012)).Although some of those studies and comments were not published in peer reviewed journals, they are nonetheless being considered in the public discussion (e.g., Velimirow et al. 2008;Antoniou et al., 2012).
Therefore, all long-term and multi-generational studies should be published in peer reviewed journals.In addition to our 10-generation study with Bt-maize in quails (Flachowsky et al., 2005), we carried out a four-generation study with Bt-maize in laying hens, which was only published as an abstract (Halle et al., 2006).Therefore, the objective of this paper is to examine the effects of genetically modified (Bt) maize on some traits in laying hens and to describe them in more detail.

Animals and experimental design
In the four-generation trial, a total of 64 laying hens (4 hens were reserves; Lohmann LSL) and 10 cocks (LSL) were purchased at the age of 18 weeks and randomly distributed to two groups (Control: Caesar maize; treatment: Bt 176 maize) and held in two pens (about 10 m 2 per pen, 0.3 m 2 per hen) with litter (at least 30 hens and 3 cocks per group).The laying hen diet was formulated to contain 500 g • kg -1 maize (isogen maize or Bt maize).
Feed (Table 3) and water were provided ad libitum.The lighting programme and climate corresponded to the specifications of the management guides.The number of laid eggs was recorded daily and the offered feed was weighed back weekly on a pen-basis.In the 28 th and 29 th weeks of life, the collected eggs were weighed for four days per group.In the 31 st week of life, brooding eggs were collected and brooded for every group.In generation 1, more eggs were used for hatching because of the unknown fertility of the eggs.The one-dayold chicks from every group were sex sorted and allocated to one pen per group.Failure in sorting was eliminated after 8 or 18 weeks.Different animal numbers during the chicken and pullet periods result from the ignorance of breeding results of the hens, failures in sex sorting, and culling birds with low and nonstandard development of body weight.
The hatched female chicks were used as the next generation.Chicken and pullet diets were fed in the 2 nd to 4 th generations to 31-74 female chickens of each group (weeks 1 to 8 and 9 to 18; see Table 6), layer diets were fed to 31-37 hens and 3-5 cocks, each from weeks 19 to 31.

Composition of maize
For the whole study, isogenic maize of the conventional variety 'Cesar' and transgenic maize (Bt 176; Bacillus thuringiensis maize), which is able to express the Cry 1AB-protein to protect maize against the European corn borer, were used.The conventional maize and Bt maize were cultivated under similar agronomical conditions.Analytical data of both maize-varieties is shown in Table 2.

Composition of complete poultry feed
Three various complete feeds (for chickens, pullets and layers) were mixed according to the nutritive requirements (GfE, 1999) of the animals (Table 3).The feed for chickens/pullets contained 400 g • kg -1 of maize and the laying hen feed, 500 g • kg -1 .The complete feeds were produced in the feed mill of the Institute of Animal Nutrition.
The diets for chickens (0-8 weeks), pullets (9-18 weeks) and laying hens (after 18 weeks) were newly mixed for each generation and stored in containers.

Analytical methods
Feed samples were taken for each generation and for each age group.Before analyses the mixed feed was stored in glass at 4°C.
The following were determined in the feed: dry matter (DM), crude protein, ether extract, crude fibre, further fibre fractions (neutral detergent fibre, NDF, acid detergent fibre, ADF), and ash content according to the methods of VDLUFA (2012).Amino acids and fatty acids were analysed using an amino acid analyzer (Beckmann 6300 amino acid analyzer) and gas chromatography.The mycotoxins zearalenone and dioxynivalenol were determined according to Valenta and Oldenburg (1995) and Valenta et al. (2002).Transgenic DNA was determined as described by Reuter and Aulrich (2003).

Statistical methods
Data from the growing, laying and hatching trials were analysed by one-way analysis of variance (GLM procedure): y i = μ + a i + e i (y i = growing, laying, hatching parameters of chickens or hens per generation (1 to 4) and growing, laying, hatching parameters of chickens or hens per treatment (isogen maize, Bt maize), μ -mean, a i -treatment (isogen maize, Bt maize), e i -error term).All statistics were carried out using SAS software (2002/03).

Results and discussion
The composition of isogenic maize and Btmaize is shown in Table 2.There were no significant differences between the maize hybrids in crude nutrients, fibre fractions, starch, some minerals, amino acids and fatty acids.The content of zearalenon and deoxynivalenol of transgenic maize was below the detection limit; the isogenic hybrid contained small  (Valenta and Oldenburg, 1995;Valenta et al., 2002) amounts of both mycotoxins (Table 2).All of the results agree with previous findings as summarized by Flachowsky (2013).
The content of some nutrients and metabolizable energy of complete feed for chickens, pullets and layers are given in Table 4.Only small differences were found or calculated between mixtures based on control or Bt-maize.The protein and amino acid contents, as well as the P-content, decreased from chicken to pullets and laying hen mixtures, but the Ca content of the feed for laying hens increased markedly.9.0 9.0 7.9 7.9 6.8 6.8 Methionine + cystine 3 6.5 6.5 6.3 6.3 6.0 6.0 Ca 2 11.0 11.0 9.0 9.0 38.4 39.8 P 2 7.0 7.0 6.5 7.0 4.5 4.5 ME, MJ • kg -1 4 11.1 11.1 10.8 10.8 10.7 10.7 1 number of analysed samples; 2 analysed values; 3 calculated values; 4 N-corrected metabolizable energy, calculated according to WPSAformula (1984) The mortality of animals was very low in all age groups and was not treatment-or generation-related (Tables 5 and 6).
The daily feed intake of chickens was between 29.7 and 33.7 g, of pullets -between 64.6 and 76.8 g (Table 5).No significant differences in feed intake or body weight of chickens and pullets after 8 and 18 weeks were registered for any generation fed with control or Bt-maize (Table 5).
The hens of generation 1 (purchased hens) consumed less feed and were lighter at the beginning of the study and after 31 weeks of life than hens of all other generations (Table 6).Animal losses in the laying period (weeks 19-31) were very low (1.5%).No significant differences in feed intake (weeks 19-22 and 23-31) were found among hens of any generation fed with control or Bt-maize.Feed intake of hens fed with control maize or Bt-maize varied in some generations.Laying intensity in hens of generation 1 was lower, and feed intake per kilogram egg mass (feed efficacy) was higher than those of all other generations.
No significant differences in fertility and hatches of living chickens were found for any of the four generations (Table 6).
Table 5. Daily feed intake (g per bird) of chickens and pullets and animal numbers per treatment and age group and body weight (g per bird) of chickens and pullets of generations 1-4 (body weight: P = 0.3-0.7)(mortality: 1 st wk 0-5 per group, wk 2 nd to 18 th 0-2 per group)

Age
Generation  The results of the four-generation laying hen study agree with the previous ten generation study with laying quails fed with control and Bt-maize (Flachowsky et al. 2005).
Presently, we do not know of other generation studies with poultry.Multigeneration studies with Bt-maize in mice (e.g., Brake et al., 2004;Haryu et al., 2009) and rats (e.g., Kiliç and Akay, 2008), but also in pigs (Buzoianu et al., 2012a,b) showed no relevant biological effects or influence on animal performance or reproduction parameters.Some studies have been published on other animal species.For 44 months, Trabalza-Marinucci et al. (2008) fed rations with Bt 176 maize to sheep and observed some differences in cytosolic activities in liver and pancreas cell nuclei and in the immune response to Salmonella vaccination.According to the authors, the significance and the reproducibility of these phenomena is unclear.

Conclusions
Feeding of Bt 176 maize to growing (400 g/ kg diet) and laying hens (500 g • kg -1 diet) over four generations did not significantly influence the growth, laying or breeding performance of hens compared with the isogenic counterpart.Our results are in agreement with other generation studies using various GM-feeds in different animal species as recently summarized by Snell et al. (2012) and Ricroch et al. (2013).

Table 1 .
Important types of feeding studies with animals for nutritional and safety assessment of feed from GMP and animals recommended

Table 6 .
Body weight, laying performance and hatchability, %, of hens per generation (P = 0.09-0.9) 1 hatched living chicks of all hatched eggs Table7summarizes the most important results over all four generations.No significant effects (P > 0.05) were registered between control and Bt-maize for any of the parameters under study.

Table 7 .
Means over four generations