Is located at 35˚50'0"E to 37˚30'0"E, longitude and 10˚20'0"N to 11˚10'0"N latitude, and 1583 m.a.s.l. and far 502 km northwest of Addis Ababa. The administrative center of this district is Chagni. The monthly mean minimum and maximum temperature is 13.37˚C, and 28.21˚C, respectively. The mean annual rainfall is 1130 mm. The area possesses extensive grassland coverage and perennial river flows, which is suitable for livestock grazing. The major feed resources for livestock in the study area comprise grasses, shrubs, and tree leaves [19]. Herbaceous and woody plants are the major vegetation types of this district. Among the herbaceous species, grasses accounted for 64.3% of the total herbaceous species with 57% of the grass species being highly desirable (palatable) Cynodon dactylon, Digitaria ternata, Digitaria velutina, Brachiaria dictyoneura, Commelina benghalensis, and Panicum maximum were the most palatable grass species [20].

It is located between 36˚35'30"E to 37˚0'30"E longitude and 11˚05'30"N to 11˚25'30"N latitude at an altitude ranging from 1353–2454 m.a.s.l. Most of the landscape of the area is described as a flat surface with a small portion undulating [21]. The monthly mean minimum and maximum temperature is 10.75˚C, and 25.95˚C, respectively. The mean annual rainfall is 1434 mm. Most of the landscape of the district is described as being a flat surface with a small undulating portion. The production system is characterized by mixed farming in which crop and livestock productions are running side by side [22]. The vegetation of the study area is mostly plantation forest of Eucalyptus, and natural tree species Cordia Africana (Wanza), different Acacia species, and Missana are found in the district [23]. A variety of crops, such as annual and perennial crops, grow under rain-fed and irrigation systems. Most of the cereal crops grown in the area are maize, teff, wheat, barley, and finger millet. Besides cereal crops, a variety of vegetables (e.g., onion, tomato, pepper, and cabbage), cash crops (e.g., chat and coffee), and fruit crops (e.g., papaya and banana) are also grown [18].

It is located between 36˚39'09" to 36˚48'25"E longitude and 10˚57'17" to 11˚03'05"N latitude with an elevation 2560 m.a.s.l. It is situated 122 km far from the regional city Bahir Dar to south, and 447 km north to Addis Ababa [24]. The administrative center of this district is Injibara, which is situated in a predominantly mountainous location. The hills and valleys receive high amounts of rain, especially in the rainy season. This high rainfall permits farmers to grow multiple crops in a year [18]. The monthly mean minimum and maximum temperature of the area was 9.12˚C and 24.65˚C, respectively. The mean annual rainfall was 1602 mm with the main wet season from June to September. The vegetation type was dominated by Albizia gummifera, Croton macrostachyus, Prunus Africana, and Apodytes dimidiate [24] (Figure 1).

Pearson Chi-Square (χ²) was used to test the association between the farmers’ perception and changes in temperature and rainfall patterns [25]. Selected cattle farming attributes were analyzed using Statistical Package for Social Scientists [27]. A one-way analysis of variance was applied for quantitative dependent variables using agro-ecological zone as the independent variable. Tukey Cramer test was used for mean separation for one way analysis of variance. Descriptive statistics, including percentages and frequencies, were applied for summary and presenting the findings. Indices were calculated for perception of farmer’s trait preference, and major Fogera cattle production constraints. The following formula was used to compute index as employed by [28]:

I n d e x = R n ∗ C 1 + R n − 1 ∗ C 2 + ⋯ + R 1 ∗ C n ( R n ∗ C 1 + R n − 1 ∗ C 2 + ⋯ + R 1 ∗ C n ) (2)

where, R_n = the last rank (example if the last rank is 8^th, then

R_n = 8, R_n−1 = 7, R₁ = 1).

C_n = the number of respondents in the last rank,

C₁ = the number of respondents ranked first.

The interannual variability of annual and seasonal precipitation pattern was analyzed as follows:

CV ( % ) = σ μ × 100 (3)

where, CV = coefficient of variation σ = standard deviation and µ = mean.

According to [29], the degree of variability of rainfall events can be classified as low (CV < 20), moderate (20 < CV < 30), and high (CV > 30).

Unlike most of the tropics where two seasons are common (wet season and dry season), three seasons are known in Ethiopia, namely Bega (dry season) which extends from October-January, Belg (short rainy season) which extends from (February-May), and Kiremt (long rainy season) which extends from June-September [4]. Thus, the study areas were classified as dry season (October-January), short rainy season (February-May) and long rainy seasons (June-September) following [4]. Then the monthly mean rainfall and temperature data were classified and analyzed through the application of descriptive statistics technique that involved the calculation of maximum and minimum of both sets, calculation of annual averages. Trend analysis was conducted using the Mann-Kendall (MK) trend test and Sens’s slope estimator [30] [31], as implemented in R-package modifiedmk. The nonparametric Mann-Kendall test (MK) was used for trend analysis as it is widely used to evaluate trends in agro-meteorological and hydrological time series [32] [33] [34]. The strengths of the MK are usually associated with its simple concept and with the fact that as a nonparametric procedure that does not assume a specific joint distribution of the data, it is minimally affected by departures from normality [35]. All-time series data were tested for serial autocorrelation using Durbin-Watson Test in R-car package. Where autocorrelation was not significant, a standard set of tests were applied (command mkktest) while when serial autocorrelation was found to be significant, a trend test was performed following a method proposed by [36] that applies a bias-corrected prewhitening (bcpw) tool found in the same R package. The rainfall and temperature trends were analyzed using R Software version 3.5.2. The Mann-Kendall test [30] [31] was applied to the long-term data to detect statistically significant trends. In this test, the null hypothesis (H₀) was that there has been no trend in temperature and precipitation over time; the alternate hypothesis (H₁) was that there has been a trend (increasing or decreasing) over time.

The Mann-Kendall test [30] [31] was applied using the formula:

S = ∑ k = 1 n − 1 ∑ j = i + 1 n sgn ( x j − x k ) (4)

where n = number of data points, x_k and x_j = data values in time series k and j (j > k), and sgn(x_j = x_k) is defined as:

sgn ( x j − x k ) = { + 1       if   ( x j − x k ) > 0 0           if   ( x j − x k ) = 0 − 1       if   ( x j − x k ) < 0 (5)

The variance of S was calculated as:

V A R ( S ) = { 1 18 n ( n − 1 ) ( 2 n + 5 ) − ∑ p = 1 q t p ( t p − 1 ) ( 2 t p + 5 ) } (6)

where q = the number of tied groups and t_p = the number of data points in the p^th group.

Whereas the values of S and VAR(S) are used to compute the test statistic Z as follows:

Z = { s − 1 V A R ( s ) for   s >   0 0 for   s = 0 s + 1 V A R ( s ) for   s < 0 } (7)

where, the positive values of Z indicate upward (increasing) trends in time series, and the negative values show downward (decreasing) trends. Trends are then tested against some critical values (Z_1−α) to show that either they are statistically significant or not. For example, if |Z| > Z_1−α, (e.g., Z_1−α at α = 0.05); the null hypothesis of no-trend is rejected, and alternative hypothesis of significant trend is accepted.

Sen’s nonparametric method [31] was used to estimate the magnitude of trends in the time series data as:

β = m e d i a n ( x j − x k j − k ) ,   j > k (8)

where, x_j and x_k represent data values at time j and k, respectively. β is Sen’s slope estimate. β > 0 indicates an upward trend in a time series. Otherwise, the data series presents a downward trend during the time period.

The majority of Fogera cattle farmers (80.91%) reported experiencing water shortage as negative impacts of climate change and variability in the study areas. The chi-square test values of survey result indicated that frequency of drought, duration of the dry spell, wind, and floods were significantly increasing (p < 0.001) in all the agro-ecological zones. Moreover, about 84.48%, 65.3%, and 60.47% of Fogera cattle farmers in the lowland, midland, and highland, respectively perceived that there was significantly an increasing trend of cattle mortality in the study areas (p < 0.001) as shown in Table 3. Moreover, about 55.17%, 61.22%, and 55.81% of the Fogera cattle owners in lowland, midland, and highland reported that herd size of the Fogera cattle were decreasing in trend whereas 17.24%, 30.61% and 41.86% reported increasing trend, respectively. However, 27.59%, 8.16%, and 2.32% Fogera cattle owners in lowland, midland, and highland were reported that there was no significant change in cattle numbers due to climate change and variability, respectively over the last 30 years. Livestock production is likely to be adversely affected by climate change, competition

***Significant at p < 0.001 level while Ns = non-significant N: number of respondent, χ²: chi-square.

for land and water, and food security at a time when it is most needed [40]. The most important diseases reported by the Fogera cattle owners were ectoparasites (26.78), foot and mouth diseases (FMD) (15.84), lumpy skin disease (15.84), trypanosomiasis (15.32), endoparasites (13.50), and mastitis (12.72%). Group discussion with key informants and extension workers with Fogera cattle owners revealed that majority of these diseases have occurred during the short rainy season when there is no adequate feed and water supply. Participants of group discussion also reported that the emergence of new diseases during a severe drought when there was no adequate cattle feed and water sources. These constraints were one of the major problems of communities residing in lowland areas. Previous studies in Ethiopia also reported an increased incidence of diseases during the long dry season [41]. A similar study by [42] reported that FMD, blackleg, and contagious bovine pleuropneumonia (CBPP) were identified as the major diseases of cattle in Borana area. According to [41], decreasing water levels in rivers and low levels of water accumulation in community ponds have been observed in recent years in Ethiopia. A study conducted by [43], indicated that the negative impacts of heat stress would become more severe in the future, as a consequence of ever-progressing global warming and genetic selection for higher production continues. The same author also reported that even a minute increase in upper critical temperature might severely hamper the cattle production subjecting the farmers to be highly vulnerable to cope with the challenge. This finding is also in line with review work in India which indicated that climate change negatively impacted cattle production both directly and indirectly [44].

The present study revealed three dominant adaptation strategies in the study areas that were reducing the number of livestock, diversification of livestock species, and replacing Fogera cattle with small ruminants. Moreover, mixed farming (crop-livestock), night grazing during hot days, and water harvesting during long rainy season for livestock use were also reported as important adaptation strategies in the study areas (Figure 2). A significant number of Fogera cattle farmers were reported that currently replacing Fogera cattle with small ruminants was becoming an adaptation option because of the shortage of feed resources, especially in the lowland agro-ecological zones. Most of these respondents were reported that they were replacing with small ruminants as small ruminants need relatively small quantities and low-quality feed compared to cattle and equines. Diversification of livestock species is the second important adaptation option reported by the farmers because they believed that if one species fails, they would remain with the other one. In this regard, [17] also reported that due to differences in tolerance to water and feed shortage and resistance to drought among species, multispecies keeping is of vital importance in minimizing

climate-related risks. Similar study in the Borana area reported that recurrent droughts, bush encroachment, increased cattle herd vulnerability, and thus the growing demand for adapted animal species, were the major drivers of livestock diversification [11]. Another study conducted in the Afar region outlined that water harvesting, livestock diversification, cattle restocking, hay production, and purchasing of hay were utilized as among the major adaptation measures [45]. This finding is also in agreement with study conducted in Pakistan which reported that livestock farmers attempted to adopt conventional climate change in their areas using different strategies such as mixed farming, reduction of animals number, provision of more drinking water, use of tree shades, livestock diversification, use of muddy roof, and floor to cope with climate changes [46]. Reducing livestock numbers to match carrying capacity of grazing lands through increased commercial off-take rates and forage development such as elephant grass and fodder trees were potential practices for enhancing the adaptive capacity of communities in pastoral areas [41], which is also consistence with the present result.

The overall mean annual minimum and maximum temperature in the area range from 11.08˚C to 26.39˚C, and a higher mean maximum temperature was observed in lowland (Figure 5(a)). Mean annual maximum temperatures ranged from 24.65˚C to 28.56˚C and generally higher magnitude was recorded in the short rainy season and lowland agro-ecology. The mean annual minimum temperature ranged from 9.12˚C to 13.37˚C (Table 4). The mean annual maximum temperature trend was significantly increasing except for midland. During the short rainy season, it was considerably increasing for all agro-ecological zones. The mean annual maximum temperature trend showed an increasing trend in

NB: Agro = Agro-ecological zones, µ = mean, β = Sen’s Slope, Z = Mann Kendal test, ***, ** and * significant at 0.001, 0.01 and 0.05 p level respectively, while ns = non-significant at p < 0.05.

the study areas which will help the planners to design appropriate strategies to minimize the impact of temperature on the livelihood of the farmers. Moreover, the mean annual minimum temperature trend was significantly increasing for all the agro-ecological zones. The mean minimum temperature trend was also significantly increasing during the long rainy season, short rainy season, and dry season along all agro-ecological zones considered. The mean annual temperature was significantly increasing for all agro-ecological zones as shown in Figures 5(a)-(c), which supports the statistical result. Farmers, key informants, and extension workers participated in the focus group discussion and interview also explained that temperature was increasing every year, which also confirms the statistical results. Another study conducted in the area also showed that most residents perceived that temperature had been increasing while rainfall was declining over the last decades [12] and [13] which were also consistent with the current findings. Most of the previous studies in this region [53] [54] [55] and in Ethiopia in general [14] also confirmed the presence of significant increasing trends in temperature.

The current result revealed that the mean annual rainfall ranges from 1130 to 1602 mm and a higher mean rainfall was observed in highland during long rainy season (Table 5). The mean annual rainfall trend was significantly decreasing in the lowland agro-ecological zone. Moreover, the mean annual rainfall trend was also decreasing in all agro-ecological zones although the level of decreasing trend was varied (Figures 6(a)-(c)). The MK test also shows that the mean annual rainfall was decreasing in midland and highland nevertheless the trend was not significant (Table 5). The long rainy season mean rainfall trend was decreasing significantly in lowland and highland agro-ecological zones. Even though the trend of long rainy season shows decreasing in midland agro-ecological zone but it was not significant (Table 5). Furthermore, the short rainy season mean rainfall trend was also indicated that decreasing in all agro-ecological zones but the

NB: Agro = Agro-ecological zones, µ = Mean, β = Sen’s Slope, Z = Mann Kendal test, µ = Mean, * and *** is significant at 0.05, and 0.001 while ns = non-significant at p < 0.05.

trend was not significant. The dry season mean rainfall showed that decreasing in lowland while increasing in midland and highland however the trend was not significant in all agro-ecological zones. The majority of the farmers, key informants, and extension workers participated in the focus group discussion and interview also indicated that rainfall had been decreasing every year. About 56.89%, 89.79%, and 83.72% of the respondents reported that rainfall was decreasing year after year in lowland, midland, and highland agro-ecological zones, respectively that are in line with the statistical results. The amount, duration, and intensity of rainfall were also decreasing in the study areas. Another survey research result in the area also reported that farmers perceived that the rise of temperature and a decline of rainfall [12], which is also in agreement with the current findings. Study in Pakistan also showed that summer season experienced significant negative trends [56].

The annual rainfall pattern in the area is dominated by the long rainy season. Due to this reason, annual rainfall and long rainy season rainfall showed similar coefficient of variation (CV) across all agro-ecological zones (Figure 7(a), and Figure 7(b)). As rainfall is comparatively abundant in the long rainy season that contributes (80.68%) in the study area, the CV is generally lowest along all

agro-ecological zones. The CV falls below 20% in all cases for both annual and long rainy season. Moreover, CV was the smallest in the highland agro-ecological zone, where mean rainfall was the greatest. The short rainy season which contributes (11.28%) and dry season which contributes (7.97%) had considerably lower total rainfall than the long rainy season and showed high interannual variability at all along all agro-ecological zones (Figure 7(c), and Figure 7(d)). Even though the highland agro-ecological zone tends to show a slightly smaller CV than the lowland agroecology, it had CV greater than 45% in the short rainy season and greater than 64% in the dry season. The current result is in line with the findings of [15]. Similar study conducted in the region also reported Ethiopian rainfall is highly variable [14].