Sustainable chicken meat production is enhanced by tangibly reduced crude protein diets


The environmental impact of soyabean meal is estimated to be approximately three times that of grain and successful implementation of reduced crude protein diets for meat-type chickens supports future sustainability for the poultry sector.

P.V. Chrystal1,2, S. Greenhalgh1, P.H. Selle1, F.J. Kleyn3, J.C. De Paula Dorigam4, S.Y. Liu1
1 Poultry Research Foundation, Camden, NSW, Australia
2 Baiada Poultry, Pendle Hill, NSW, Australia
3 Spesfeed Consulting, Broederstroom, South Africa
4 Evonik Nutrition & Care GmbH, Hanau, Germany

In a series of 4 trials using Ross 308 male broilers to 35 days post-hatch, dietary crude protein was, on average, reduced by 39.7 g/kg (from 209.9 to 170.2 g/kg) without compromising bird performance. Notably, in one of the studies broiler performance was maintained with a tangible decrease in dietary crude protein of 57 g/kg and a 66.2% reduction in soyabean meal content, suggesting it is possible to achieve sustainable growth of chicken meat production to meet an estimated 72% growth in demand to 2050. In the context of sustainability, the environmental impact of chicken meat production was assessed using an Australian developed model whereby twelve environmental parameters were included in a single calculated “ecopoint” value. Additionally, as a comparison, six individual environmental parameters utilising European l’Institut National de Recherche Pour l’Agriculture (INRA) data, applied to individual feed ingredients were assessed. The environmental impact of soyabean meal is estimated to be approximately three times that of grain and successful implementation of reduced crude protein diets for meat-type chickens supports future sustainability for the poultry sector.


Whilst meat-type chicken production compares well with other forms of livestock production systems it is nevertheless considered to have a negative environmental impact primarily due to dependence on soyabean meal as the main source of dietary protein. In a review of comparative life cycle assessment (LCA) de Vries and de Boer (2010) concluded that 1 kg of edible chicken meat production requires between 8.1 and 9.9 m2 of land, similar to pork (8.9 to 12.1 m2) and on average 76% less than beef (27-49 m2). Higher reproduction rates, reduced methane emissions and improved feed efficiencies were identified as the major drivers of these differences. However, reported LCA methodologies are not consistent partly due to the complexity in assigning Co2 emissions to deforestation for soyabean production, confounded by dynamics of grazing by livestock and logging. 

More recently, a LCA was conducted by calculating single score “ecopoints” using 12 environmental parameters for Australian chicken meat production on a weighted average annual impact scale of 100 points per person. Authors combined the life cycle impact of global warming, abiotic resource depletion, land transformation and occupation, water consumption, eutrophication, acidification, eco-toxicity, photochemical smog, ozone depletion, ionizing radiation, human toxicity and respiratory effects into a single score. On average, cooked roast chicken was ascribed 11.9 calculated ecopoints/tonne with feed ingredients identified to contribute 49% (5.83 ecopoints/tonne) towards this value and grains, soyabean meal and meat & bonemeal making up 79% of the feed contribution.

Over the past 10 years approximately 5.6 million hectares of additional land has been required annually to accommodate increasing soyabean production. Furthermore, projected increases of 72% in chicken meat production over the next 30 years have been estimated requiring an extra 121 million hectares globally, based on current levels of dietary soyabean meal inclusion. Thus, the purpose of this paper is to quantify the differences in ecopoints as described in Bengtsson and Seddon (2013) per tonne of cooked roast chicken between conventional and tangibly reduced crude protein (CP) diets over four broiler studies. Only feeds that resulted in similar broiler performance were selected and assessed for their conduciveness to sustainability of chicken meat production. Additionally, for comparative purposes, individual environmental impact parameters for feed ingredients; cumulative energy demand, phosphorus consumption, climate change (CO2 emissions) eutrophication and land competition, adapted from INRA data ( were assessed.


A combined total of 168 off-sex male Ross 308 broilers were used in this comparison from 7 to 35 days post-hatch, comprising 6 birds per replicate treatment over 6, 7 or 8 replicates as described in Chrystal et al., 2020. Within each study, a conventional (high protein) diet was compared with the lowest selected dietary protein treatment, where broiler performances were statistically similar to the conventional diet. Dietary AMEn averaged 12.90 MJ/kg and digestible lysine 11.15 g/kg across all treatments whilst average CP ranged from 209.9 to 170.2 g/kg (difference of 39.7 g/kg). All diets were steam-pelleted at a conditioning temperature of 80 °C and comprised of either maize-soyabean meal or wheat-soyabean meal, supplemented with non-bound amino acids (AA) to maintain “ideal protein” ratios as recommended by Aviagen, Ross 308 (2019). Concomitant with a reduction in dietary CP, dietary soyabean meal inclusion declined by 44.6% (from 325 to 180 g/kg) and grain inclusion increased by 28.4% (from 549 to 705 g/kg). All feeding studies fully complied with specific guidelines (2016/973) approved by the Animal Ethics Committee of the University of Sydney.

Soyabean meal was allocated 6.7 ecopoints/tonne whilst wheat varied from 0.4 to 0.8 and barley was estimated at 3.1 ecopoints/tonne; surprisingly, canola seed was highest at 8.5, canola meal 5.0 and by deduction, canola oil 3.5 per tonne. Non-bound, supplemental AA were not individually defined but formed 21% contribution within a normal diet and could thus be proportionately calculated for all diets. Using ecopoints for grain (average 2), soyabean meal and vegetable oil and allocating ecopoints to the balance of the feed, the relative influence of feed differences on ecopoints per tonne of cooked roast chicken was calculated. Thus, the lower the ecopoint value, the lower the total environmental impact. In contrast, applying INRA data and soyabean meal associated with Brazilian deforestation, six environmental impact parameters were individually assessed, emphasising disparity between measures of environmental degradation and depletion of non-renewable resources. Furthermore, INRA assign environmental impact values for only five supplemental AA, requiring assumptions for non-documented dietary AA. Thus, there is a paucity of data on environmental impact of individual AA not commonly used in feeds.


Reducing dietary CP resulted in an average reduction of ecopoints/tonne of feed by 21.2% (5.83 versus 4.59 ecopoints/tonne) whilst dietary soyabean meal inclusion declined by 44.6% and grain content increased by 28.4% (Table 1). Interestingly, the largest decrease in calculated ecopoints/tonne (31.4%) was in maize/soyabean-based diets whereby dietary CP was successfully reduced by 57 g/kg (from 222 to 165 g/kg) and supplemented with non-bound AA. In these diets, soyabean meal inclusion declined by 66.2%, from 334 to 113 g/kg and maize inclusion increased by 41.1% from 511 to 721 g/kg. Average broiler performance across all treatments compared favourably with Aviagen Ross 308 (2019) performance objectives, whereby weight gain exceeded the breed performance objectives by 7.2% (2067 versus 1929 g) cumulative feed intake increased by 3.1% (3113 versus 3018 g) and feed conversion improved by 3.6% (1.510 versus 1.566 g feed/g of liveweight). Assuming standard feed contributes 49% towards ecopoints/tonne of roast chicken, an average reduction in dietary CP by 39.7 g/kg results in a decline of 1.2 ecopoints/tonne, reducing the average environmental impact of roast chicken by 10.1% from 11.9 to 10.7 ecopoints/tonne.

Surprisingly, when applying INRA data to the diets used in the ecopoints exercise, four out of six calculated environmental impact parameters were higher for reduced CP diets (Table 2).

INRA climate change value for L-lysine HCl is 34.7 times higher than maize (10004 versus 288 CO2 eq/kg) whilst land competition is 377 times higher than maize (0.0377 versus 0.0001 m2 yr/kg) and further elucidation of these discrepancies is required. Thus, in reduced CP diets, land competition (m2 yr/kg) was calculated to be 140% higher than standard broiler diets; cumulative energy demand (MJ/kg) 48% higher and similar values of 22% for eutrophication (PO4 eq/kg) and acidification (H+ eq/kg). However, on climate change measured as CO2 eq/kg, reduced CP diets were 11.8% lower and utilised 29.4% less phosphorus (Table 2).


Irrespective of the method used to determine impact of chicken meat production on the environment, data reported varies considerably, primarily due to the complexity of calculating relevant data. For example, published values for liveweight broiler emissions range from 2000  to 5480 kg CO2 equivalent per tonne, with Ingham’s average in Australia reported to be 2613 kg CO2 (Bengtsson and Seddon, 2013), which is in close agreement with 2600 kg CO2 estimated by Williams et al. (2009).

Commonly, feed and the broiler farm are identified as the two main components of the LCA for chicken meat production.

In the Australian scenario, based on ecopoints, Bengtsson and Seddon (2013) attributed 29.9% of the environmental impact to cereal grains compared with an equivalent quantity of soyabean meal. In contrast, a Portuguese case study by González-García et al. (2014) regarded both maize and wheat equal and estimated far higher values of 65.4% of the environmental impact of soyabean meal. Presumably geographical distance between country of origin (South America or USA) and Australia relative to Europe may partly explain these differences. A single ecopoint value is useful in assigning an overall dietary effect on the environment although individual values calculated using INRA data provide absolute measures per LCA category of feed indredients. Importantly, decreasing dietary CP by 30 g/kg is associated with a dietary equivalent CO2 reduction of 306 kg/tonne of chicken meat (Martin, 2020) and a 14% reduction in N excretion emphasising the need to include farm effects in environmental LCA of reduced CP diets. Additionally, INRA estimate yields of cereal grains per hectare to be superior to soyabean meal by a factor of 6 with respect to land competition, in close agreement with Williams et al. (2009). This suggests there is less pressure on deforestation with successful development of tangibly reduced CP broiler diets. These data appear to be inadequately assessed in environmental LCA and warrant further consideration. However, despite data inaccuracy, these tools provide new insights into improving sustainability of chicken meat production.

Projected growth in demand for chicken-meat production will increase its environmental impact. However, based on present studies, adopting reduced CP diets with dietary soyabean meal inclusion equivalent to 310 g/kg, without a deterioration in broiler performance, means this demand can be met without reliance on increased soyabean production. Whilst predicted ongoing improvements in efficiency of broiler performance have not been factored into this future scenario, initial outputs from this study suggest that chicken meat is well placed to meet growing demand over the next 30 years in a sustainable manner.

References are available on request

From the Proceedings of the Australian Poultry Science Symposium 2021