HLI composition variables
When analyzing variables that made up the Heat Load Index (wind speed, relative humidity and black globe temperature) at 20 and 120 cm (Table 1), wind was the greatest impact factor on the inferred thermal comfort of cattle in this study. The wind speed on treeless pasture at 20 cm (1.735m/s) was proximately 300% higher than SPSnu5 (0.55m/s), and 200% than SPSnu10 (0.84 m/s). At 120 cm the wind speed pattern didn’t change much. TLP wind speed (4.00m/s) was proximately 200% higher than SPSnu5 (2.16 m/s), and 250% than SPSnu10 (2.63 m/s).
Another HLI important variable is relative humidity (RH). In this study RH showed no difference between treatments at 20 cm (TLP 61.7%, SPSnu5 61.5%, SPSnu10 61.7%), and at 120 cm (TLP 60.1%, SPSnu5 59.6%, SPSnu10 59.3%). Even with a higher wind speed, relative humidity did not show any distinction between TLP and pasture with shrubs and low height trees under SPSnu. This also was observed by Souza et al (2010), and seems to be the case in insular areas where winds from the ocean can blow high humidity hiding microclimatic site variation of RH due evapotranspiration. However, in a less windy region far from the ocean, RH was greater in SPSnu when compared with TLP (Deniz et al. 2018; Schmitt et al. 2023). Increase in evapotranspiration within shrub and tree vegetation has been observed to increase RH (Lopes et al. 2016). This higher RH observed by Deniz et al. (2018) and Schmitt Filho et al. (2023) under SPSnu was enhanced by the observed decrease in wind speed in-between SPSnu (Deniz et al. 2018; Schmitt et al. 2023) since winds usually dissipates site specific RH (Farhat 2018).
The black globe temperature (BGT) at 20cm was lower in the SPSnu10 (35.7°C) and SPSnu5 (36.1°C), than TLP (37.5°C). However, at 120 cm BGT did no vary among treatments, SPSnu10 36.5°C, SPSnu5 36.5°C, and TLP – 36.6°C. This demonstrates that besides the insufficient shadow protection from low height young trees, SPSnu interfered in 20 cm BGT but not in the 120 cm. Even so, it was noted the great effect of young SPSnu on the wind speed at 20 cm and 120 cm, and therefore on BGT. Higher wind speed at 120 cm caused homogenization of BGT at 120 cm. SPSs under implementation have insufficient radiation blocking by young trees small shadow, however they slow the wind, and this is an essential factor on BGT (Gaughan et al. 2008; Gaughan et al. 2010; Mader et al. 1997).
Soil and air temperature
The overall analyses of the tree treatments, SPSnu5, SPSnu10 and TLP (Fig. 1), showed no differences for air temperature at both heights, 20 and 120 cm. It is important to note these similar temperatures happened under a quite distinct wind speed, since on TLP the wind speed were between 200 and 300% higher than SPSnu. Understanding that wind has a cooling effect, one could assume that with equal wind speed TLP would present higher temperatures.
Furthermore, the analysis of shaded and sunny areas in each treatment (Online Resource 4) showed that the SPSnu, even in the process of implementation, blocked solar radiation causing a reduction in temperatures in areas with shade. The average temperatures of the shaded spots were lower in the two SPSnu treatments in relation to the TLP in the soil and at the two heights (20 and 120 cm).
The greatest temperature differences between shaded and sunny areas decreased from the ground up to 120 cm. Soil surface temperature differences were up to 7.5° C. At 20 cm the differences were up to 1.6° C, and at 120 cm the differences were not more than 1.2° C. This effect of reducing temperatures in shaded areas can positively affect the behavior of cattle. On hot days, for every 1°C increase in the daily mean air temperature, dairy cattle spend 22 minutes less in decubitus per day (Tresoldi et al. 2019). Lying down behavior is negatively correlated by thermal stress, as it interferes with the body area of exposure to wind and radiation, and makes breathing less effective as a cooling function (Tucker et al. 2021).
When comparing SPSnu four areas (within the nucleus-WN, around the nucleus-AN, shaded around the nucleus-ANS, and internuclei-IN) among themselves and with TLP (Fig. 2 and Fig. 3), we found lower soil temperatures in WN and TLP than in AN and IN in SPSnu5. In SPSnu10, the WN area presented lower values than all the others did. The reasons why these two areas were clustered in the same temperature spectrum are quite distinct. The WN area has a dense vegetation of shrubs and young trees, which shade the most part of the inner nucleus blocking radiation and cooling down the temperatures. The TLP area has no young trees to block the wind with the highest wind speeds, 200 and 300% higher than SPSnu. Higher wind speed can explain this phenomenon of cooling down temperatures in TLP (Cleugh & Hughes, 2002). Therefore, WN has lower temperatures because of radiation block (Schmitt Filho et al., 2023), and TLP because of wind cooling effect (Cleugh & Hughes, 2002).
The lowest temperatures between SPSnu5 and SPSnu10 at 20 cm were those inside the nuclei (WN). These temperature were expected since tree height average in the nuclei was 1.74 m with collection points predominantly deep in the shade. This shade pattern and its relationship with air temperature are expected for most consolidated SPSs (Deniz et al. 2020, Schmitt et al. 2023).
When analyzing shaded areas around the nuclei (Table 2), soil temperature on shaded areas were (ANS) 6.8° C lower than in the sunny areas (AN) in SPSnu5, and 6.4° C lower in SPSnu10. In both measurements were done between 08:00h and 16:00h. Contrary to Deniz et al. (2020), it was noted that only the shaded areas around the nuclei had lower soil temperatures. Higher soil temperatures in the areas around the nuclei compared to TLP can be explained by the higher wind cooling those areas far from the nuclei (Mader et al. 1997). The differences between the temperature values were reduced as the collection points height increased. This is related to the low tree heights (1.74m), the distance from the ground, and a more homogeneous wind speed between the areas.
The comparison of shaded points around the nuclei in SPSnu5 and SPSnu10 with TLP (Table 3) showed differences only in soil temperature. These shaded points are around the nuclei where the herd would have access. These are small shadow projections from the young nucleus to their surroundings. Soil temperatures were 4.4°C and 4.6°C lower than the TLP in these shaded spots around the nuclei in SPSnu5 and SPSn10, respectively. Cattle are highly motivated to access shade (Schütz et al. 2008) preferring shade to sprinkling as a cooling strategy (Schütz et al. 2011). Soil temperature also can be an important factor for this decision. Understanding the relationship of those multiple variable with thermal comfort and production needs further attention (Deniz et al. 2020).
Shaded points around the nuclei (ANS) where expected to have best thermal comfort as observed by Deniz et al. (2018) and Schmitt et al. (2023). Therefore, the herd would enjoy the most microclimatic advantages of the SPSnu and perform comfort behaviors at these areas as observed by Deniz et al. (2020) and Craesmeyer et al. (2017). Nevertheless, our results demonstrated that HLI around the nuclei (AN) was similar to pasture with no trees (TLP), where we expected to find the worst conditions of thermal comfort (control treatment) (Muller et al. 1994; Améndola et al. 2019).
This phenomenon is closely linked to two factors. First, a great reduction in wind speed in SPSnu when compared to TLP. The magnitude of this effect got up to 300% decrease in wind speed at 20 cm at the SPSnu (0.55 m/s) than TLP (1.73 m/s), and 250% decrease in wind speed at 120 cm in SPSnu (2.16 m/s) than TLP (4.00 m/s). The nuclei functioned as a windbreak compromising the cooling down effects of the wind during spring and summer months. The second factor is lack of shade in the system due to the early stages of nuclei development with trees only up to 1.74 m.
These two factors defined the microclimatic pattern during this early phase of SPSnu implementation, with special emphasis to the fact that the young trees acted as a windbreak while they were not providing shade yet. These factors can increase HLI during the implementation of SPSs, especially in the tropics and subtropics. We call this phenomena windbreak countereffect (WbCe).
Heat Load Index
The analysis of the general averages of the treatments (Fig. 1) demonstrated that young nuclei did not offer better HLI than treeless pasture in the two heights. Analyzes of this fact should consider that HLI formula has black globe temperature as one component. Black globe temperature is a microclimate variable that involves solar radiation, relative humidity, and wind speed on its value. HLI formula considers the wind as a cooling element, since the higher its speed, the lower the HLI value (Gaughan et al. 2008). As commented before, the average of tree height during the experimental period was only 1.74 meters, with no shade projection to block radiation.
Furthermore, the implementation of the zero functional group (FG0) produced the expected windbreak effect (Schmitt Filho & Farley 2020). The purpose of the zero functional group is creating a better microclimate for plant development. This means reduction of wind speed, increases in humidity and a better temperature within the nuclei, and on the protected and exposed sides (Cleugh & Hughes 2002). This results in a better environment for tree seedlings development in the nuclei. In contrast, these microclimatic condition created by zero functional group would negatively influence the thermal comfort (HLI) around the nuclei during the hot seasons due the interaction of less wind and lack of shade.
The study shows that at this SPSnu initial phase, the greater nuclei per hectare, the higher were the HLI averages in the two heights. SPSnu in an advanced stage with a good tree canopy enhance thermal comfort around the nuclei (Deniz et al. 2020). Otherwise, HLI was impaired during the early stages of SPSnu implementation due windbreak countereffect. This seems to be a tradeoff in order to achieve the late microclimate improvement from well-developed trees (Pent et al. 2021).
Comparing sunny and shaded areas (AN, IN, ANS, WN) of the two SPSnu treatments with the TLP (Table 2), we found that because of windbreak countereffect (WbCe) the SPSnu had worst thermal comfort (lower HLI) than TLP in the two heights. This was caused by the lack of cooling effect of higher wind speed in-between the nuclei due to WbCe. This indicates that the nuclei have influence on the microclimate of the internuclei areas. Previous studies by this research group compared areas surrounding the nuclei with internuclei (Deniz et al. 2020; Schmitt Filho et al. 2023), and surrounding the nuclei with TLP (Deniz et al. 2018; Schmitt Filho et al. 2023). In the present study, we also analyzed inside the nuclei, plus outside areas (surroundings with and without shade, internuclei) and TLP to understand the influence of the nuclei on the microclimate gradients and thermal comfort in SPSnu.
Due to the windbreak countereffect, TLP presented better thermal comfort than in sunny areas of SPSnu treatments. This implies that, regions with high incidence of winds during hot seasons, the windbreak countereffect during the implementation of SPSs must be seem as a tradeoff, to be minimized by management strategies.
Comparing microclimate in the multiples areas of both SPSnu treatments and TLP (Fig. 2 and Fig. 3), the HLI analyzes at 20 cm showed better thermal comfort in TLP, because the windbreak countereffect of the zero functional group reducing the wind speed (Cleugh & Hughes 2002). As a result, the HLI decreased in the areas as the collection points moved away from the nuclei. In SPSnu5, only WN presented HLI in the “extreme” category, while in SPSnu10 WN, AN and IN presented HLI in the “not extreme” category (Gaughan et al. 2010). This is an indicator of the windbreak causing wind blockage, and the mentioned countereffect (Gaughan et al. 2010). The HLI analyzes at 120 cm also indicated better thermal comfort in TLP, but none of the area averages were found that fit into the category of “extreme” HLI. At this height, the windbreak countereffect of functional group zero is not so evident. These results also can elucidate why herbivores prefer to stand during times of thermal stress (Tucker et al. 2021), as the heat dissipation capacity increases, HLI decreases.
When comparing shaded areas around the nuclei (ANS) with other areas (Table 2), HLI results in the SPSnu treatments were classified in the “very hot” category (Gaughan et al. 2010), although during hours of higher temperatures (12:00h) were not used in this analysis (Karvatte Jr et al. 2016). When 12:00 was part of (Fig. 2 and Fig. 3) HLI, it end up being classified as “extreme” (Gaughan et al. 2010). These conditions indicate that mitigation of heat stress is necessary in hot days.
Mitigation of heat stress during the SPSnu implementation process
It was noted that the first phases of SPSs could negatively influence microclimate and thermal comfort due windbreak countereffect. Setting aside areas for tree planting and establishment up to 41 months has been recommended to protect trees and mitigate thermal stress (Porfírio-da-Silva et al. 2012). However, this strategy is not always possible due the costs of setting aside pasture areas, or even lack of paddocks and cultural reasons (Workman et al. 2003).
As sucessional agroforestry nuclei, fast growing pioneering trees will replace functional group zero. As time pass by, windbreak countereffect will be masked by the cool off effect of 30% tree shade from the nuclei, as observed by Deniz et al. (2020), Schmitt Filho and Farley (2020), and Schmitt et al. (2023).
In order to avoid heat stress related losses, farmers must be aware of the windbreak countereffect during the implementation of SPSs. Special attention must be given during the hot months in temperate and subtropical regions, and all year around in the tropics. When necessary, we do recommend strategies to mitigate heat stress during this hot periods. The provision of artificial shade (Blackshaw & Blackshaw 1994; Schutz et al. 2011), paddocks without wind barriers, paddocks with old growth trees, or even paddocks with forest remnants could be considered (Mader et al. 1997; Mader et al. 1999).