Environmental enrichment as an ethical principle in animal research

Animals used in experiments have reduced levels of welfare and are thus vulnerable to pain and suffering. Considering that environmental enrichment increases the quality of life of captive animals, we aimed to assess its scientific application regarding the promotion of animal welfare by recognizing the vulnerability of biological models. Documentary analysis and experimental tests were conducted, stating that despite this technique having been applied mainly in neuroscience studies, endorsing their feasibility and the improvement in learning, the rationale for its use has been limited to the successful development of research, that is, not in recognition of animals’ vulnerability, their need for welfare and quality of life. In contrast, we proposed the application of environmental enrichment for laboratory animals, within a framework of scientific contractualism and the researcher’s responsibility, as a standard to be adopted for the mutual benefit of scientific development and the quality of animal life.


Environmental enrichment as an ethical principle in animal research
http://dx.doi.org/10.1590/1983-80422016243153 Although knowledge of non-human animals has increased significantly throughout history, it has increased in recent decades through the development of technology associated with the field of neuroscience.This process has added to our understanding of the biopsychosocial needs of animals, the moral value of their lives and the legal and ethical need to respect them.This is especially true with regard to captive animals and interferences without valid reasons and for which alternatives do not yet exist, such as in research and education 1 .
The definition of animal welfare (AW) involves a momentary state of harmony between the body and the environment that surrounds it, in which the animal seeks to meet its physiological needs through adaptation, aided by good health and environmental opportunities 2,3 .In order to develop technologies that promote both the diagnosis of conditions of AW and alternatives to mitigate the effects of captivity, the science of AW has developed, supporting the appropriate use of animals, and recognizing the inherent guidelines of the 3Rs: reduction, replacement e refinement 4 .While this behavior is widely encouraged in production systems, it is endorsed especially in the area of science by the justification that the greater the AW, the more reliable the results of research involving non-human animals will be.Recent studies reveal that a space enriched with toys, tunnels and physical activities 5 significantly stimulates the neurogenesis 6 of rodents.This improves the ability to perform cognitive tasks 5 , due to increased exploration activity and new sensory experiences, thus increasing AW.The field of laboratory research with rodents, targeted mainly at experiments aimed at developing biotechnology, has proven conservative in applying techniques of environmental enrichment (EE).
The standardizations of biosecurity and husbandry techniques are important for ensuring that the material introduced into the reproductive system is practical for handling and cleaning, and does not put the animals at risk 7 .Most of the diseases that affect animals in captivity come, or are maximized by laziness 8 , and EE aims to increase quality of life 9 .Thus, the application of EE in rats acts to improve the results of scientific research in experiments in the areas of both health and learning 10 , as this hypothesis is based on the fact that animals display more natural physiological and behavioral characteristics 11 .
Lab exercises with rats are a reproduction of the effects of environmental variables, influencing the functional relationships established by animal behavior 12,13 .One way to impose natural environmental variables would be to use experimental psychology to work with the results of the manipulation of important variables in controlled conditions 12 .Thus, it is the duty of the researcher to take care of feeding and provide a hygienically suitable environment, as well as to avoid the discomfort of animals 12 .It is important to note that all animal research should adopt ethical principles.The animals must be acquired legally and their treatment must comply with local laws and regulations 14 .
For Weinberg 15 , a scientist may be brilliant, imaginative and intelligent, but he or she will not go beyond being a mere scientist if he is not also responsible.Therefore, responsibility is the principle which guarantees the autonomy and freedom of the researcher.This responsibility is social, as it must ensure the quality of the research, and also be ethical to animals, as vulnerable beings are being manipulated in the research process 16 .
As well as the fact that vulnerability is intrinsic to life, it must also be considered that individuals may be directly affected by unfavorable circumstances.Accordingly, being or being in a state of vulnerability refers to a range of meanings that extend from a latent state to a manifested state, or in other words, from the possibility to the probability that the vulnerable being is vulnerable 17 .Not providing welfare to a non-human animal is to leave it in a manifest situation of vulnerability, thus compromising its quality of life and, as a consequence, the results of the research.
This study is justified in bibliographic data 17,18 which shows that EE is effective in promoting AW.The question, however, is why its use remains so poorly promoted.The hypothesis tested theorizes that researchers, despite knowledge of current legislation 19 and the improved quality of life of animals provided by AW -and the consequent increase in the reliability of data -do not consider the vulnerable condition of animals when making their decisions.Therefore, this study aimed to evaluate the applicability of EE through documentary and experimental data, with a view to attributing responsibility for the promotion of AW to researchers, based on the evidence of the effectiveness of EE, as well as its link with the results of their research.

Materials and method
The present article consisted of two methodological approaches: one consisted of the systematic Environmental enrichment as an ethical principle in animal research http://dx.doi.org/10.1590/1983-80422016243153documentary analysis of the use of EE in scientific research, while the other is the experimental evaluation of the effect of environmental and social enrichment on the learning of animals evaluated, to support the discussion about animal vulnerability and the ethical responsibility of the researcher with regard to the evidence of the effectiveness of EE.

Documentary analysis
Documentary analysis was performed using articles recovered from the Capes Periódicos portal, a tool that brings together indexers such as Pubmed, Scopus, ASFA, SciELO, OneFile, Medline, SpringelLink, BioOne and JSTOR.The search term "environmental enrichment" was applied, subject to the presence of the term "rats".The first stage was carried out between September and October 2014, analyzing the total records per year and then categorizing them into specific topics.The first 100 articles published between the years 2013 and 2014 were retrieved, in order to verify that the use of the term "EE" in research was aimed at achieving welfare and recognized the vulnerability of laboratory rats.Review articles that met the exclusion criteria, as well as items that could not be accessed, were rejected, leaving 80 articles.

Research animals and location
The study was performed at the Laboratório de Análise Comportamental da Escola de Saúde e Biociências (the Behavioral Analysis Laboratory of the School of Health and Biosciences) of the Pontifícia Universidade Católica do Paraná (the Pontifical Catholic University of Paraná) (PUCPR), Curitiba campus.The behavioral patterns of 40 male rats of the Rattus norvegicus, aged approximately 30 days, was analyzed.The animals were provided by the PUCPR bioterium and housed in standardized boxes, arranged on shelves.They were provided with daily food and water and bedding was changed twice a week.
The R. norvegicus rodent was used as an experimental model because of its high genetic homogeneity and because it is commonly used in experimental studies and experimental behavior analysis.The fact that they are gregarious animals with a high exploration and locomotor activity was also taken into account 11 .

Experiments
The experimental design aimed to evaluate the effect of EE and social enrichment (SE) comparing the performance of learning among individuals kept in enclosures with and without EE and SE.The animals were therefore separated into four groups, each with 10 animals ¬ Group 1: isolated rats with EE; Group 2: isolated rats without EE; Group 3: rats in pairs with EE; and Group 4: rats in pairs without EE.

Environmental enrichment
The environmental enrichment for both isolated animals and those in pairs consisted of the addition of a PVC tunnel, 7.5 cm in diameter, with an angle of 180°.The tunnel was arranged in a lateral position on bedding in a 20 cm tall by 30 cm wide box.

Social enrichment
The SE consisted of keeping the animals in pairs.A mark on the animals' tails was made with an overhead projector pen and reinforced weekly, so they could be recognized.To avoid the influence of this variable, all animals were marked, even those kept in isolation.Through behavior analysis, the subordinate and dominant animals were determined.

Evaluation of learning performance
To evaluate the learning of the animals in different conditions, tests were conducted by the classic experimental analysis of behavior procedure 12 , using the Skinner box.The operant conditioning chamber is one of the main elements of behavior-analytic methodology 20 .
In this methodology, the assessment of learning occurs by conditioning through positive reinforcement, using the water dispenser.It should be mentioned that for water to function as an effective reinforcement in reinforcement learning trials, it is necessary that the animal is deprived of water for 24 hours before the planned exercise 12 .The first step was to determine the operant level of the animal (OL) before behavior modeling (M).This allows the evaluation of the effect of the positive reinforcement (water) by comparing with the response frequency of response before and after the introduction of the reward 12 .The exercise had an average duration of 30 minutes and behaviors were recorded at one-minute intervals.
After determining the operant level, the animal underwent the conditioning exercise in training with the water dispenser.This process includes the Environmental enrichment as an ethical principle in animal research http://dx.doi.org/10.1590/1983-80422016243153adaptation of the possible emotional responses of the subject when hearing the noise of the water dispenser.The experiment had an average duration of 15 minutes.
The next test was response to the pressure bar after continuous reinforcement (CFR), which consisted of modelling the response to the training.The modeling here is the release of stimulus, containing water, after a previously defined response.The exercise had a duration of ten minutes 12 .Another feature evaluated was the satiation level (SL), which analyzed how long and how many drops of water were needed for the animal to be satiated.The experiment had an average duration of 60 minutes.
After the animal is conditioned to the bar pressure stimulus, the exercise of extinguishing the response to the bar pressure (E) was performed.This test aims to observe the effect on the frequency of the pressing the bar action without the presence of reinforcement.The process was complete after a minimum of fifty responses had occurred and the animal went for ten minutes without pressing the bar 12 .
To again stimulate again the behavior of pressing the bar to receive a reward in the rat, reconditioning of the bar pressure response was carried out.This test had an average duration of one hour, and the animal had an interval of five minutes to press the bar.If this did not occur, the response modeling process was again carried out, continuing in this way until the moment that the subject received ten CRF stimuli.
The intermittent extinction of reinforcement (IE) was also considered in this study.This is an experiment for comparing the bar press response rate with the process of the extinction of the continuous reinforcement.The experiment began with ten stimuli sessions.After the initial phase, the control switch was switched to the manual position.Thus, when the subject pressed the bar no reward was issued 12 .The test had an average duration of 1h 30 min.

Statistical analysis and legal considerations
The test results were compared between the groups based on social and environmental enrichment.The Kolmogorov-Smirnov test was used to verify the normal distribution of variables.Comparisons between means were performed by the non-parametric Kruskal-Wallis and Mann-Whitney tests due to the absence of a normal sample.For comparison between the frequency of motor pattern values displayed by the animals the chi-squared test was used.A significant of 95% was used in all tests.
All procedures, as well as those of the bioterium that provided the animals, were in accordance with Brazilian law.The selection of the animal model was due to its prevalence in the experimental protocols of the institution, as well as being traditionally used by experimental psychology 1 .The experimental design sought to minimize the use of the animals in the performance of statistical tests.
Environmental enrichment was not provided for all animals, as its absence was one of the study variables.The deprivation of water for 24 hours also constituted a procedural variable for behavior modeling.However, to alleviate this process, we attempted to provide intervals of at least one week between the tests.Habituation was carried out prior to testing, so that the animals could get used to the presence of the researcher.Their reactions during the study were also monitored, while they underwent the minimum amount of handling and stayed in the lab for as short a time as possible.

Document analysis
The scientific texts reviewed indicated a development that began with the first article found, which was from the 1950s to the peak of production in 2014.Studies that specifically use environmental enrichment and the association of the technique with neuroscience began to emerge in the scientific field in the 2000s, following the overall increase in scientific production in subsequent years (Figure 1).
Of the 80 articles retrieved, most of the tests involved cognitive or neuroanatomical-pathological procedures.The main interest of the studies was how to meet the research needs in support of the benefit sought by the study, seeking homogeneous results rather than animal welfare.Recognition of animals as vulnerable subjects in the laboratory environment was not mentioned or discussed in any article.
The studies included works on neurology, with 50% being about neurophysiology and the others dealing with neuroanatomy, neuropharmacology and neuroendocrinology.Most studies (75%) aimed to verify the effect of EE.The majority of studies (96%) found that EE had a positive effect on the results of research, while only 3.1% obtained neutral conclusions.A total of 91% of studies were neurological, followed by studies about cognition (6%), drugs (2%) and phenotypes (1%).The tests were mainly based on cognition (37%).The combination of physical and social EE (29%) was the most commonly used, followed by physical, cognitive and social (18.5%).

Environmental enrichment as an ethical principle in animal research
http://dx.doi.org/10.1590/1983-80422016243153 Figure 1.Temporal distribution of scientific texts retrieved from Capes Periódicos Portal using the terms "Environmental enrichment" and "Rats", total, approaches with environmental enrichment, behavioral or neuroscience studies based on cognition and morphological or physiological studies

Environmental enrichment
Group 1, where the animals were kept isolated and exposed to an enriched environment, performed better in learning patterns (t = -3.027;p = 0.007) when the evolution of operant level (OL) after continuous reinforcement was evaluated.The group without EE exhibited a more heterogenous mean difference in the number of responses (bar press), with double the response imprecision, as can be seen in Figure 2. The same response pattern, with greater variation and a higher average, was observed in the level of satiation in group 2 (p = 0.005) (Figure 3).There was no significant difference between the response patterns of groups 1 and 2 for the other cognitive tests.When the test performances of the individuals in pairs (groups 3 and 4) with and without enrichment were compared, the difference was not significant.

Social enrichment
There was no difference in performance in the learning process between the dominant and subordinate individuals in any of the tests, despite the stimulus of environments with and without EE.Thus, it was possible to compare the pairs, as a whole, with the isolated individuals.There was a difference in learning performance between pairs and single individuals in the intermittent extinction of reinforcement test (p = 0.043), with no influence of EE.The isolated individuals performed better and had a more homogeneous bar pressure response in the test (Figure 3).There was no significant difference between the pairs and the isolated individuals for the other tests.

Discussion
The documentary analysis and the experiments showed that the use of EE in research involving behavioral and neurological analysis is established and that its use achieves better results.However, the prioritization of neuroanatomical-pathological tests to measure the effects of EE shows that evaluations are aimed at quality of research, not quality of animal life.
The experimental data in this study indicates that EE favored the learning of rats, certified in modeling tests.These tests evaluated the bar pressure response as a learning determinant and found a significant improvement in the performance of animals kept in enriched environments.These results corroborate other studies that showed the influence of EE on cognition in rats 17,18 , mice 21,22 , rabbits 23 and pig 24 .Schaeffer 6 points out that enriched environments result in an increased proliferation of neural stem cells, the survival of new neurons 25 and increased brain weight 23,26,27 , thus promoting improved performance in cognitive tasks, given the increase in variables in an environment is a stimulating factor.Many researchers have considered that the adoption or otherwise of EE is a possible compromise of experimental results, mainly due to greater variability 27,28 .However, this study found that EE contributed to cognitive ability in the responses to the initial testing, and also that the satiation level of rats was more homogeneous than for individuals without EE.Instead of the complexity of the enriched environment leading to an increased response variation among individuals due to the number of stimuli 29 , what seems to occur is that the inclusion of just one PVC tube provides environmental improvement, making the individuals more able to handle the new challenges posed in testing, thus generating uniform responses, as has been observed in other studies [30][31][32] .Nonhuman animals in EE conditions have shown greater physiological and psychological stability, and are thus able to generate better scientific results 33 .
In this study, EE consisted of the provision of an object that protected from light and aggressions, which is recognized as an increase factor for AW 34 .In the case of rodents, which are gregarious, social animals 35 , it is important to consider that social isolation can stimulate behavioral disorders 36 .As for SE, the results of the present study only showed differences in the most complex phases of the learning tests, with isolated individuals exhibiting an advantage in the re-elaboration of responses to positive stimuli.This result may reflect the generation of a mild social instability generated between the pairs in cages, as literature recommends using three individuals per environment due to the reduction of aggressive behavior and better social stability 37 .

Environmental enrichment as an ethical principle in animal research
http://dx.doi.org/10.1590/1983-80422016243153 The lack of a difference observed in learning performance between pairs and isolated animals in the other tests reinforces the recommendations of Van Loo and Baumans 38 who said that if there is a need to keep animals in social isolation, it is appropriate to use nests, aimed at providing rest and a sense of security, as well as suitable thermal control conditions.
The concern that the application of EE can affect the variability of results caused by environmental conditions is in fact more related to the importance of a wider discussion of the subject in the field of science.Based on the proposal of EE, several studies have been conducted aimed at identifying the variables that affect laboratory experiments.The results point to several important parameters, such as basic housing conditions 28 , sex, species, the age of the animal, cage lighting conditions and temperature 39 40 .
It is clear that factors that influence the results of non-human animal experiments are directly interconnected with the structure in which the individual finds itself.This fact highlights two important aspects, the first being the social responsibility of researchers regarding the quality of their experiments and the importance of ensuring reliable results.The second aspect is the duty to respect the lives of animals exposed to experimental situations in which their existential vulnerability is increased 16 due to pain, discomfort and suffering.

Social responsibility in research with nonhuman animals
According to Jonas 41 , the human being, due to possessing the capability to understand, experiences responsibility as an act of liberty, and such responsibility as essentially ethical.Every researcher has a social responsibility to the community in general 42 , and the primary social responsibility is to prevent the corruption of research data, thus ensuring the integrity of the study.The literature review 43 explained that basic conditions such as the brightness of the environment of the animals can interfere with the response of individuals.Therefore, the non-standardization of the appropriate environmental conditions for the animals, as well as an ethical failing, represents scientific misconduct with regard to research data 43 .
From the principle of ethical and social responsibility linked to research, EE is proposed as a guideline and standard of ethical conduct in research with non-human animals.Giorgini et al. 44 pointed out that the role of the code of ethics in the conduct of researchers in the laboratory practice has been restrained.According to Habermas 45 , in a plural society there is no traditional ethos or a substantial consensus of standards, values or principles to justify answers to practical questions.However, in the absence of such a consensus, there is a form of "rational consent" by independent, free and equal individuals, who identify, through deliberative communication practices, a rational justification for adopting a certain standard.This defines modern contractualism, which is rationally justified through dialogic procedures based on communicative rationality 46 .
In this sense, the best rational argument among the scientific community is the establishment of strategies to promote AW, thus establishing EE in the scientific contract as a standard to be adopted for the mutual benefit of scientific development and quality of animal life.
Another important point is the vulnerability of the animal in experimental conditions.For Hossne 16 , all living beings are subject to vulnerability, which is the result of the inability to an individual to protect his, her or its own interests.The vulnerable individual that suffers unmet needs becomes more prone to being easily affected and victimized 47 .Animals used in laboratory experiments suffer a reduction in the range of possibilities to achieve greater degrees of AW, even taking into account the obligation that their suffering is avoided and their survival ensured, as they are unable to protect their own interests and are thus vulnerable.This takes us back to the principle of social responsibility and the ethics of the researcher, who when handling a vulnerable life incurs the risk of compromising the search results.In addition, there is an ethical responsibility for the lives of the laboratory animals and to provide essential conditions of AW and quality of life.
Given these circumstances, the fact that research has already attested to the benefits of AW cannot be overlooked, leading us to consider whether the omission of this variable in the analysis and placing of results can be considered fraud 48 .However, it is also worth considering the vulnerability of the researcher in the face of the academic pressure to publish, the encouragement and incentives offered for positive results in high impact journals 49 and the necessity of meeting deadlines, as well as the desire for studies that are not properly justified by the results.These factors can lead the professional to only consider the immediate usefulness of the animal rather than the effectiveness of their research and the ethics of their behavior.However, even considering the pressure that researchers may be under, the reliability of the research can be measured by the interpretation of the results.Therefore, it is necessary to improve the quality of animal experiments, especially considering that irresponsible practice includes the ill-treatment of laboratory animals.Studies should establish experimental protocols that comply with the justifications for the experiment, with an appropriate statistical design and detailed methodology, following the guidelines of Concea 50 .This resolution also emphasizes the importance of minimizing the pain and distress of animals so that physiological and behavioral changes are avoided, and do not lead to the misinterpretation of the data 50 .

Final considerations
It was not expected that identifying the threshold between the vulnerable being and being vulnerable is a bioethical issue that would be discussed or addressed directly in a search of scientific articles involving EE, although it was hoped that such a search would find recognition and assessment of EE as an ally of quality of life and AW.The fact that is being tested for the best search result is a positive sign for the improvement of the conditions of captive animals.But it is important to note that so that EE can effectively protect the vulnerable, from an ethical point of view, it can only be considered adequate in a systematic assessment of the context of vulnerability.
In this sense, the concern with neuroanatomical patterns or cognitive performance in tests restricts the focus of research and makes EE an ethical procedure that should be associated with standardized protocols.Promoting AW, as corroborated by the experimental results of this study, in addition to providing appropriate environmental conditions for the animals, will lead the researcher to obtain more reliable and reproducible results.However, for this to happen it is necessary to understand the biological and behavioral needs of the animal and EE in a broad and integrated manner.This is ethical and social responsibility linked to research and to all involved, involving the creation, maintenance, handling and transport of animals.
In summary, it is proposed that EE is one of the guidelines and standards of ethical conduct in research with animals in the context of scientific contractualism.In this context, the researcher adopts the standard as a matter of social responsibility, both for the benefit of scientific development and for vulnerability.

Research articles Environmental enrichment as an ethical principle in animal research
Mean difference between standard responses (bar pressure) of operant level for mice undergoing continuous reinforcement with EE and without EE (WEE).b.Level of response to bar pressure related to the satiation of isolated animals in an environment with EE and without EE (WEE) http://dx.doi.org/10.1590/1983-80422016243153Figure 2. a.Figure 3. Response pattern (bar pressure) in the intermittent extinction of reinforcement test between individuals with social enrichment (doubles -D) and without social enrichment (isolated -I) a. with enrichment (EE) and b. without enrichment (WEE)