Breeding soundness examination and herd proficiency of local genetic groups of bulls in tropical environment conditions in Veracruz, Mexico

Abstract With the aim of assessing bull fertility before or during its use for reproduction under tropical weather conditions where heat stress is present, and correlating it to herd parameters, BSE was carried out to determine which genetic groups of bulls have better reproductive performance. 223 bulls from different genetic groups, Zebu bulls (N = 72), European bulls (N = 58), and Crossbred Bulls (N = 93), were evaluated in situ over a two-year period, 2018–2019. Bull (age, corporal conditions, libido, and scrotal circumference), semen (concentration, motility and volume), herd (pregnancy rate and calving intervals) and environmental variables (THI, season, sampling month and spermatogenesis month) were analysed with One-Way ANOVA, as well as with factorial, multiple regression, and multivariate analysis. Zebu and European bulls have a higher libido than crossbreed bulls (p < .05), and it decreases with age (p < .05). High-libido bulls (>7) show high correlation (R = 0.70 p < .05) with herd parameters; BCS affects libido (p < .05). Sperm concentration is correlated to herd parameters (R = 0.82 p < .05). European bull sperm concentration and motility showed a decrease in autumn and winter (p < .05) in comparison to other genetic groups. Libido and semen variables are more associated with herd parameters than with other bull characteristics. Environmental variables are not associated with herd parameters (p > .05). Heat stress did not directly affect bull reproductive performance; Zebu bulls and crossbreeds showed better herd parameters (p < .05). Highlights Bull breeding soundness examination and herd proficiency. Herd proficiency of local genetic groups of bulls in tropical conditions. Breeding soundness examination in tropical environment conditions.


Introduction
Bulls have a decisive influence on the reproductive efficiency of herds, regardless of the reproductive programme used: natural, controlled mating or artificial insemination. Therefore, absence, or decrease, of bull fertility causes a reduction in productive parameters, with a possible decline in farm profitability, particularly in extensive livestock production systems. In some specific conditions, fertility does not only mean the characteristics of the seed but also the ability of the bull to obtain, even in natural mounts, a satisfactory number of pregnancies even in adverse environmental conditions. In tropical countries, breeding programmes for beef and dual-purpose cattle are characterised by extensive systems, where the aim is to increase calve numbers and selling weight (Mart ınez-Gonz alez et al. 2017). Livestock production in Veracruz began since the first age of Spanish colonisation in XV century with the introduction of southern Spain's animals (Bos taurus) that originated the Creole cattle population (Cañ on et al. 2011); since the 50's, Bos Indicus cattle (Indo-Brazil, Brahman) from Brazil, as well as international cattle (Brown Swiss, Holstein, Charolais and Simmental), were introduced with the aim of introducing superior production performance through gradual adaptation to environmental conditions (Dom ınguez-Mancera et al. 2017; Ginja et al. 2019). In the full extensive farming system of the region, bulls are required to be extremely fertile in board sense to preserve farm profitability. Highly fertile bulls can achieve a pregnancy rate higher than 60% in females that show adequate reproductive activity; to achieve this, Bull Breeding Soundness Examination (BBSE) is necessary (Nichi et al. 2006; P aez-Bar on and Corredor-Camargo 2014). BBSE is a procedure that reduces risk and improves strategic bull usage and herd fertility (Chenoweth 2002). Breeding Soundness is affected by several factors, such as age, nutrition (body condition), environmental conditions (seasons), management, breed (genetic group) and their interactions (Roberts et al. 2010;Menon et al. 2011;Lemma and Shemsu 2015). BBSE includes biometric measurements, mainly scrotal circumference (SC), semen analysis (sperm concentration, motility and morphology), but behaviour (libido) is not considered in it, although this parameter can be used to predict potential bull fertility (Lessard et al. 2011). SC is an indirect but trusty indicator of testicular weight and sperm production capacity (Chacon et al. 1999;Alexander 2008;Kumar et al. 2016). It has been extensively studied because of its favourable genetic correlations with age at puberty in males and females (Toelle and Robison 1985;Notter et al. 1993), and reproductive characteristics in females, such as heifer pregnancy (Van Melis et al. 2010;Santana et al. 2012). Sperm concentration, motility and morphology are key factors in BBSE, and, consequently, affect herd reproductive parameters in extensive systems (Burnett et al. 2018). Bull behaviour, especially libido, is another very important trait related to BBSE; bulls with high libido (>7 in scale 1-10) produce more ejaculates, have lesser reaction time and refractory period, and better semen quality (Ellis et al. 2005;Bury et al. 2011). In tropical conditions, where high ambient temperature and/or humidity are present, cattle reproduction can be affected by heat stress (Brito et al. 2002); this condition can have important effects on most aspects of reproduction (Hansen 2009). Semen characteristics are not immediately affected by changes in testicular temperature, because damaged spermatogenic cells do not enter ejaculates for some time after suffering from heat stress but other factors such as behaviour can be influenced early and indirectly affect global fertility (Lees et al. 2019). Regarding bulls, whose spermatogenesis takes about 61 days, alterations in semen occur about two weeks after heat stress, and it does not return to normal until up to eight weeks following the end of heat stress conditions (Meyerhoeffer et al. 1985;Paul et al. 2008), given the recent interest in the impact of climate change on the profitability and sustainability of cattle breeding, we present the following research with the aim of verifying the effect of environmental variables on global fertility parameters (BBSE) in bulls of different genetic types under the tropical conditions in Veracruz state, Mexico and its relationship with herd reproductive parameters.

Ethical statement
All the procedures for handling, immobilisation and semen collection carried out on bulls within the farms by official veterinary services and were evaluated and approved with registration number COBIBA010/2017 by Bioethical Committee (School of Veterinary and Animal Sciences of Veracruz University). The welfare of the animal is paramount and stimulation was discontinued if either undue stress was being caused or physical injury to the bull might occur.

Animals
Bull breeding soundness examination (N ¼ 223), carried out by the Animal Biology Reproduction Laboratory, and the Cell Biology Laboratory of Veracruz University, were used. The genetic groups to be evaluated were assigned as proposed by (Menegassi et al. 2015). Zebu bulls 'Bos indicus' (Gyr and Brahman), European bulls 'Bos taurus' (Brown Swiss, Holstein, Charolais and Simmental), Crossbreed bulls (Holstein Â Zebu, Brown Swiss Â Zebu, synthetic breed bulls Â Ze; Beef Master, Brahman Â Hereford Â Shorthorn). All bulls were fed under an extensive grazing system grazed Cynodon nlemfuensis and Brachiaria humidicola grasses (Sollenberger 2008;Cruz et al. 2017), with 1-8 years of age, without any apparent health impairment at the time of evaluation. Animals and semen samples were examined in the farms where they were being kept during the 2018-2019 period.
The cows used to measure the fertility of bulls were clinically healthy and fertile at time when the Breeding Soundness Examination were performed, cows with reproductive abnormalities were excluded of the analysis. Palpation of the reproductive system by rectal examination or ultrasound using a Minitube Probe 6.5 MHz linear (Minitube, Verona, WI, USA) was used for rectal pregnancy and abnormalities detection. The reproductive records and clinical evaluation of the cow reproductive system (N ¼ 6513) from different genetic groups. Zebu cows Bos indicus N ¼ 2160 (Gyr n ¼ 1258 and Brahman n ¼ 902), European cows Bos taurus N ¼ 1749 (Brown Swiss n ¼ 612, Holstein n ¼ 192, Charolais n ¼ 334 and Simmental n ¼ 612), Crossbreed cows N ¼ 2603 (Holstein Â Zebu n ¼ 1093, Brown Swiss Â Zebu n ¼ 1511) with an age range of 4-9 years, mean calving interval of 914 ± 22 days (2.5 ± 0.07 years) and mean pregnancy percentage 35.6 ± 2.9%. The pregnancy rate was calculated as the total number of pregnant cows divide by the numbers of mated cows by bull examined. Calving interval, was calculated as the amount of time (days) between the birth of a calf and the birth of a subsequent calf, both from the same cow. This information was obtained from herd productive and reproductive records, considering the time each bull was in the herd (Perea-Ganchou et al. 2005;Galina et al. 2007). All cows were fed under an extensive grazing system grazed Cynodon nlemfuensis and Brachiaria humidicola grasses (Sollenberger 2008;Cruz et al. 2017).

Morphological and behavioural evaluation
BBSE was carried out following the indications of the Society for Theriogenology's Manual for Breeding Soundness Examination in Bulls (Kennedy et al. 2010). Scrotal circumference was measured in each bull, using the technique described by Beggs et al. (2013); body condition scores (scale 1-5) were measured using the technique described by Kunkle et al. (1994). To evaluate bull libido and breeding ability, a test was performed in a small yard where bull and cow could be easily observed; the bull was allowed to be in contact a cow exhibiting behavioural oestrus during a 5minute period. Libido was rated (0-10 scale) with the system proposed by Chenoweth (1997, Chenoweth et al. 2010).

Semen quality evaluation
The semen samples were collected on a monthly basis, starting in January 2018 until December 2019, each of these were evaluated in the field immediately after being obtained ($5 minutes). Semen was collected in a test tube with a 1-15 mL graduation, and volume, colour and density were evaluated macroscopically (Chenoweth 1983;Chenoweth 2002).
Electro-ejaculation was attempted to procure a semen sample from each bull, using a three-electrode probe (Minitube, Verona, WI, USA; Ø: 2 0 '/5.08 cm; long: 33 cm). At each examination session, no more than one electro-ejaculation attempt was made to collect a semen sample from each bull (Furman et al. 1975). The semen sample was evaluated immediately for sperm motility, the materials that contact the sperm was at the same temperature as the sperm (to avoid temperature shock), clean, dry, and non-toxic. Individual motility was assessed in a sample diluted with warmed saline solution. A drop of diluted sperm was placed on a thermoplatin slide at 37 C, covered with a coverslip and examined at 40X. The proportion of sperm that are moving progressively across the field of view was estimated by finding multiple groups of $10 sperm and estimating how many sperm are progressive versus how many are not (Moskovtsev and Librach 2013). Sperm concentration (Â10 6 /mL), spectrophotometric method was used to measure sperm concentration. Once sample was obtained, a drop of undiluted semen was taken and placed in the Microcuvette for SDM-1 (Minitube) with a capacity of 2 ml, then inserted into SDM-1 photometer (Minitube) calibrated for bovine, finally, we proceed to take the reading provided by screen equipment (Atiq et al. 2011;Minitube 2020). Semen density was categorised as: very good (creamy, 4), grainy consistency with 750-1000 Â 10 6 /mL, Good (milk-like, 3) with 450-750 Â 10 6 /mL, fair (skim milk-like, 2) with 250-450 Â 10 6 /mL, poor (translucent, 1) < 250 Â 106/mL (Barth 2000).

Weather data collection and analysis
Weather data from 2018 to 2019, provided by the Gulf of Mexico forecast centre of the National Meteorological Service, were used (weather stations in closest proximity to evaluated farms $30kM). The Centre's main office is located in the city of Veracruz. Livestock Weather Security Index (LWSI), commonly known as temperature humidity index (THI), was established by the following equation: where: 'T' is the average daily temperature in C, and RH is the relative humidity percentage (%) (National Research Council 1971). Hahn et al. (1999), considers four categories of THI to evaluate thermal environmental conditions, and their impact associated with respirations per minute. In this sense, consider the increase in respiratory frequency as a proportional compensatory response to heat stress (Nienaber and Hahn 2007). Values of THI 74 were considered as the Comfort stage; 75-78 were considered as the Alert stage; 79-83 were considered as the Dangerous stage, and !84 were considered as the Emergency stage (Saizi et al. 2019). With the weather data, a brief description of the climatic conditions of the region where bulls grassing throughout the year was made, thereby showing the environmental effects in spermatozoa of bulls from different genetics groups.

Statistical analysis
STATISTICA V10 (2010) (StatSoft 2011) was used for all statistical analysis, and figures were edited in the Sigma Plot V11 (SigmaPlot 2008) software. First of all, normality was evaluated with the Shapiro-Wilk Test, and homoscedasticity was evaluated through Bartlett's test. One-Way analysis of variance (ANOVA), generalised linear model (GLM) were performed to evaluate main effects and combined effects; in addition linear mixed models (LMMs) as a method for analysing data that are non-independent of the following effects: 1. Intrinsic effects: age, body condition score, libido, genetic group, semen 2. Extrinsic effects: climatic factors (THI level at sampling day '0 day' and day that the spermatogenesis cycle begins 'À60 day', year season, month sampling, month spermatogenesis), reproductive parameters (pregnancy rate and calving interval).
The fixed effects included in the models were genetic group, age, body condition score and random effects were month sampling, month spermatogenesis and year season.
Simple linear relation was made with the following model: where 'Y' is the dependent variable and 'X' is the independent variable. In addition, were obtained the correlations (R), and determination values (R 2 ).
Multiple correlation analysis was used to indicate the variables that were most associated with bull sexual behaviour (libido) and pregnancy rate. Multivariate analyses; Cluster analyses and Principal components & Classifications (Options for Multivariate and Exploratory Analysis from the Statistics Analysis Module, STATISTICA V10), Principal Components was the data analysis tool that was used to reduce the number of variables from a large number of interrelated variables while retaining as much of the information (variation) as possible and calculates an uncorrelated set of variables known as factors or principal components. For this reason, was used to obtain variables and the total variation of the model that best describes the bull and semen variables with climatic variables on pregnancy rate. Table 1 shows least squared means values and mean standard errors for each BBSE variable analysed, as well as differences (p < .05) of post-hoc comparisons through Tukey method for each genetic group.

Regional climatologic analysis
In a first preliminary analysis of the data from meteorological stations using daily data records was done. The regional data are shown in Figure 1 plotted according a monthly distribution. Climatic variations can be appreciated throughout the year, with notable differences in accumulated monthly rainfall, dominant wind speed, relative humidity and average temperature for each season in central Veracruz State, Mexico.

Effects of heat stress on sperm concentration, individual motility and libido
Together with climatological analysis, temperature and humidity values were used to obtain the Livestock Weather Security Index (Temperature Humidity Index,  THI), and, thereby, determine the effect of heat stress on bull sperm concentration and motility ( Figure 2) and Libido (Table 2). Heat stress affects sperm concentration (p < .05); it was observed that sperm concentration decreases when semen samples are obtained on 2 or 3 months after high THI levels (alert and danger stages) day À60 (Table 2). It must be taken into account that the sperm cycle takes 61 days, and, if this period elapses when THI is high (summer), it can affect sperm production and quality, as seen in Figure  2 and Table 2, where an analysis was performed at À60 days before sampling. Bull libido was affected by year month, season and THI (p < .05), showing lower parameters when heat stress is present (Table 2) represented by THI levels (danger and emergency stages). Another variable that must be taken into account is the quantity and quality of forage available for bull feeding in extensive systems. Forage decreases in the period between January to May; this period is known as dry season. Sperm concentration, Individual motility and Libido is low on dry season, and an increase is observed (p < .05) when bulls are sampled in rainy season. Sperm motility does not show changes (p > .05) related to heat stress effects; there are oscillatory changes in both the dry season, when bulls are in the Comfort stage, and the rainy season (june to december), when heat stress is high.   It was decided that one aspect to evaluate would be whether sperm concentration and individual sperm motility of different bull genetic groups present in the tropical regions of Veracruz were affected by heat stress condition (elevated THI); results are shown in Figure 3. European bulls show low sperm    concentrations and individual motility in winter, when effects of heat stress are low (comfort), but also nutrients are low. Sperm concentration and motility increases in summer, when THI is high over the comfort threshold (alert and danger stages), and nutrients are also high due to rainfall, to descend in autumn, when THI is in the comfort stage, and nutrients decrease. Zebu bulls also show variations throughout the year, with increases in seasons when THI is at comfort stage. In spring and summer, when THI is at alert and danger level, sperm concentration values decrease, to increase in summer, when THI and nutrient intake are high. Crossbreed bulls show similar values to those found in Zebu bulls, with increases in seasons when THI is at comfort stage, and low sperm concentration with high THI. Sustained increases can be observed in rainy season (summer).
To determinate the THI value in which semen (sperm concentration and individual motility) and behaviour (libido) parameters change in each genetic group, a linear regression was performed with THI records at day sampling (day 0), to evaluate the effect of heat stress at the beginning of spermatogenesis; which was considered 60 days before sampling day, climatological records were obtained to determine the THI at the beginning of the spermatic cycle (À60 day), the results are showing in Figure 4 and Table 3.
Zebu bulls do not show changes in semen variables and libido when these variables were analysed with THI at the sampling day (day 0) or at the beginning of spermatogenesis (À60 day), Zebu bulls were not affected by heat stress. European bulls, show increments in these variables when THI is high (day 0) and the beginning of spermatogenesis (À60 day) probably due to increments in nutrients (Biomass) present in rainy season, when THI is high. Crossbreed bulls show decrease in sperm concentrations and libido at sampling day (day 0), probably just libido can be affected by heat stress in this genetic group, and not changes where found when THI is analysed at the beginning of spermatogenesis (À60 day). Others factors can be related in heat stress effect of bulls as: age, body condition, intensity and duration of THI at the beginning of spermatogenesis.

BBSE correlation analysis with herd parameters
After having carried out climatic variable analyses on some sperm variables and behaviour, BBSE correlations on herd parameters were obtained: Pregnancy Rate and Calving intervals. Results are shown in Table   Table 3. Temperature Humidity Index (THI) correlation analysis on semen: sperm concentration and sperm individual motility and Behaviour: libido, of bull genetic groups in Veracruz.  4. Variables with high correlations on Pregnancy rate were: age (R ¼ 0.630) and libido (R ¼ 0.704); while low correlation variables were BCS (R ¼ 0.383) and scrotal circumference (R ¼ 0.207). Likewise, variables with medium to high associations for calving interval were: age (R ¼ 0.506) and libido (R ¼ 0.539), and those with low associations were BCS (R ¼ 0.155) and scrotal circumference. When variables that characterise semen were used in association with pregnancy rate and calving interval, high correlations were observed with sperm concentration (R ¼ 0.822) and density (R ¼ 0.811); and medium to high correlations were observed with volume (R ¼ 0.619) and motility (R ¼ 0.686) on pregnancy rate. In the same sense, when the calving interval variable was correlated, medium to high correlations were observed with sperm concentration (R ¼ 0.634) and density (R ¼ 0.659), and medium correlations were observed with volume (R ¼ 0.567) and motility (R ¼ 0.600).

BBSE effect on herd reproductive parameter
The main effects of ANOVA, and its' interactions with pregnancy rate and calving interval were determined with the variables that best explain BBSE (age and BCS) and bull behaviour (libido) on herd reproductive performance ( Figure 5). The libido of bulls in the Veracruz tropical region decreases (p < .05) with age ( Figure 5(a)); likewise, bulls with low BCS 3, and high BCS > 5 (1 À 5 scale), show decreased libido (p < .05). When sperm concentration and motility were analysed in relation to bull age ( Figure 5(c)) and libido ( Figure 5(d)), a decrease (p < .05) in sperm concentration and motility is observed in older bulls (age 6-8). Libido is associated with traits of concentration and motility, showing increases (p < .05) in the presence of higher libido values. It should be taken into account that young bulls (ages 1-2) show low sperm concentrations. Herd reproductive parameters, pregnancy rate and calving interval are affected (p < .05) by bull age (Figure 5(e)) and libido ( Figure 5(f)). Herds with old bulls show low pregnancy rate and long calving intervals, herds with high libido bulls show high pregnancy rates and low calving intervals, in comparison to the average. When different bull genetic groups were analysed (Figure 6), it was observed that Zebu bulls have high libido values (Figure 6(a)) and BCS (p < .05). Testicular growth over the years (scrotal circumference) is similar (p > .05) in all genetic groups analysed (Figure 6(b)), with sustained growth to 5 À 6 years. Crossbreed bulls show lower values (p < .05) of sperm concentration and sperm individual motility than other genetic groups (Figure 6(c)). With respect to herd reproductive parameters values (pregnancy rate and calving interval), Zebu bulls showed the best results, Figure  6(d) (p < .05).

BBSE multivariate analysis on herd proficiency
Finally, multiple linear regression and multivariate analysis (cluster, principal components & classification) were performed to determine mathematical models to describe variables associated with herd proficiency (reproductive parameters). First, Libido was analysed as a response variable, since this characteristic is associated with almost all BBSE variables, obtaining the model: where 'Y' is Libido, a is the intercept, X 1 is the age of bull, X 2 is Body Condition Score, X 3 is the genetic group and c is the standard error of estimate Where age, BCS and genetic group were variables that forward methodology determined. Concerning herd reproductive variables, pregnancy rate was the one that was best associated with BBSE variables; the model obtained from this analysis was: where 'Y' is pregnancy rate, a is the intercept, X 1 is the scrotal circumference, X 2 is the individual sperm motility, X 3 is the sperm concentration and c is the standard error of estimate.where scrotal circumference, sperm individual motility, and sperm concentration were variables that forward methodology determined. Cluster analyses, whose different variables studied are shown in Figure 7(a). Libido is linked to clusters of bull characteristics, while pregnancy rate is linked to clusters of semen variables. Calving intervals are not directly associated with any study variable; its linkage distance is wide.
Principal components and classification analyses were performed to determine total variation from two main components (Factors, Eigenvalue 1 and 2) with bull and semen variables (Figure 7(b)). The total variation determined by analysis was 64.69% (Factor 1 ¼ 52.06% and Factor 2 ¼ 12.63%). Pregnancy rate clusters with libido, concentration and sperm individual motility; calving interval shows association with bull age, as seen in previous analyses, where bull age negatively affected calving interval.
In Figure 8, Principal component & classification analyses were performed to determine total variation of two principal components (factors, eigenvalue 1 and 2) (e, f) Herd reproductive parameters, pregnancy rate (left) and calving interval (right) in bulls of different age (e) and libido value (f). Numbers above the error bar show the number of analysed bulls. ( Ã ) indicates significant statistical differences (p < .05). Bulls 1 and 2 years old, no data on pregnancy rate and calving interval were reported. from bull, climatic variables and semen characteristics; the total variation determined by analysis was 55.59% (Factor 1 ¼ 38.83% and Factor 2 ¼ 16.76%). Pregnancy rate clusters with libido, concentration and sperm motility, volume, scrotal circumference and BCS; calving interval shows association with bull age. Climatic variables (season, month of sampling, month of spermatogenesis, season of spermatogenesis) do not show association with herd reproductive parameters.

Discussion
Livestock from the central region of Veracruz State shows genetic group heterogeneity, and is used for meat and milk production, with significant differences in their reproductive parameters. Several researchers (Perea-Ganchou et al. 2005;Galina et al. 2007), confirm similar results as those obtained in this research, since typical region weather is coupled with pasture seasonality and herd management to make each genetic group behave differently. The tropical region climatology of Veracruz shows significant variations throughout the year, which can influence the reproductive performance of domestic species. Two constants that predominate are high temperature and humidity, which, when combined, cause animals to lose ability to dissipate the heat they produce, causing heat stress (Hahn et al. 1999;Dom ınguez-Mancera et al. 2017).
The highest values of temperature and humidity are observed in the summer season, but monthly  accumulated rain values are also high and, with them, there is an increase of biomass available for bulls in extensive systems (Casagrande et al. 2018). Bull reproductive and genetic potential is influenced by several factors such as management, age, nutrition, environmental conditions and behaviour. A fundamental aspect in bull's behaviour is the environment in which they perform; in tropical conditions, where temperatures exceed the ruminant thermo-neutral zone, research reports show that cows are more affected than bulls (Hansen 2009). Heat stress affects oestrus, as well as conception, and, consequently, low pregnancy rates are reported (Polsky et al. 2017). THI has been used as a tool to indicate the extent of the effect of climate on cattle (Hahn et al. 1999;Nienaber and Hahn 2007).
THI effects, has been demonstrated in relation to sperm quality semen, the increase can cause variations in the lipid conformation, reduces normal morphology and expression of the receptor lipoprotein mRNA very low density (VLDLr), thereby sperm motility and progressive speed can be affected (Sab es-Alsina et al. 2017, 2019). Bulls exposed to THI high levels (danger or emergency stages) semen samples shows alterations 3 weeks after; although alterations are not permanent, semen quality may take up to 3 months to return to normal (Rahman et al. 2018;Sab es-Alsina et al. 2019). In addition, sperm from bulls exposed to high temperatures do not show significant differences in morphology and DNA fragmentation compared to semen obtained from bulls without heat stress effect (Llamas Luceño et al. 2020). In this research, sperm concentration and motility show a slight decrease when THI is high, but it is not significant because it is found in months where the present biomass amount is lower (dry season). It should be noted that sperm cycle lasts 61 days, and semen samples taken at that time do not fully reflect the bull's reproductive potential (Valeanu et al. 2015;Felton-Taylor et al. 2020). When these variables were analysed differently between genetic groups, it was observed that bulls from European origin are more affected at the end of the rainy season than Zebu bulls, thus showing that nutritional intake can be a key factor in sperm concentration and motility, instead of heat stress (Brito et al. 2002;Felton-Taylor et al. 2020). A positive relationship between food quality and sperm concentration has been described by other researchers (Dance et al. 2015;Bourgon et al. 2018), who demonstrated relationship between an adequate amount of nutrients from grass and spermatogenesis. In this research, Zebu bulls and their crossbreeds were not affected by THI, which shows that Zebu bulls in tropical conditions are more thermotolerant than European bulls (Jim enez-Severiano 2002; Deb et al. 2014;Khan et al. 2018;Rahman et al. 2018).
Bulls of European origin show lower reproductive performance than Zebu bulls and their crossbreeds (Polsky et al. 2017;Casagrande et al. 2018), in tropical conditions, as can be seen in the obtained results. Has been reported different results in genetic groups of bulls of European origin than Zebu bulls and their crossbreeds (Polsky et al. 2017), with emphasis in bull behaviour (libido), indicating that Zebu bulls have less libido than European bulls. In this research, there were no differences in libido between Zebu bulls and European bulls, just a decrease was observed in crossbred bulls. Bull behavioural characteristics involve two components: (1) their ability to identify females in oestrus, and (2) their capacity to mount them. Apparently, libido is not related to semen quality, nor is it related to scrotal circumference; thus, it is possible to obtain an excellent semen sample in bulls with low libido (Chenoweth 1997;Galina et al. 2007). Age plays a decisive role in BBSE; libido, sperm concentration and motility show their highest values in young and mature animals, as can be seen in the results. Others researchers (Jim enez-Severiano 2002; Khan et al. 2018) have reported similar results to those found in this research, and decreasing levels as the animal ages. Regardless of the genetic group analysed, once sexual maturity is reached, these indicators stabilise for a short period time (2 years), and when the bull ages, there is a continuous decrease of the indicators towards the last years of bull reproductive performance, causing decrease in herd proficiency (Brito et al. 2002). Bulls of 3-5 years of age show better performances in herd reproductive parameters than bulls more than 5 years old, which shows that the bulls have a useful life in extensive systems, regardless of their genetic group (Rawlings et al. 2008). The best associations of herd reproductive parameters (pregnancy rate) with BBSE variables were found in sperm concentration, motility and bull libido; calving interval was best associated with bull age. Although libido evaluation and mating ability are not sufficient to predict reproductive success, it is prudent to use bulls that passed all BBSE stages, including libido (Menegassi et al. 2015). Lastly, multivariate analysis has been used in the evaluation of the reproductive soundness examination of bulls in order to determine their inclusion and permanence in the herd, in the same way, has been analysed the effects of breed, age, year season, regions on sperm variables (Hancock et al. 2016;Felton-Taylor et al. 2020), this kind of analysis allows associating intrinsic and extrinsic factor that can affect the physiological performance of bull and with it, be able to explain effects as a whole that in simple analyzes could not be determined. The findings of this study will act as a guide for veterinary practitioners and dual-purpose cattle breeders in find bulls that can be expected to pass the BBSE test and libido based on a robust data set.

Conclusions
Environmental conditions do not have a direct effect on sperm concentration and motility in local bull genetic groups. Bulls that show the best performance on herd reproductive parameters are from Indian origin (Zebu, 'Bos indicus') and crossbreeds (Bos Tauris Â Bos indicus); bulls more than 6 years old regardless of the genetic group should not be allowed to be in the herd, since there could be a decrease in herd proficiency. Libido is a characteristic that must be considered in BBSE, since it is highly correlated with herd reproductive parameters. Bull breeding soundness evaluation performed before or during breeding season reduces the risk of sub-fertile bulls in the herd.

Disclosure statement
The authors declare no conflicts of interest. The funders had no role in the design of the study, in the collection, analyses, or interpretation of data, in the writing of the manuscript, or in the decision to publish the results.