Description of the study area

The area comprises 10 municipalities of the state of Durango and 5 from Coahuila, with an area of 4.8 million hectares. The climate is characterized by an annual average temperature of 20.9 °C, warm summers with average maximum temperatures of 30.2 °C (±1.9), minimum 10.5 °C (±3.0) and mean annual rainfall of 287 mm (±98), during the evaluation period⁴. Information obtained by municipalities were: 1) Inventory of dairy cattle, total and lactating animals; 2) Annual milk production; 3) Total harvested area, and forage-harvested area; 4) Annual production of forage crops. The above information was obtained from statistical yearbooks from 1999 to 2011, available in the of SAGARPA Delegation in the Laguna Region⁴. During the study period, the total agricultural harvested area averaged 130,000 ha under irrigation plus 20,000 under rainfed. The rainfed areas are concentrated in six municipalities of Durango: Mapimí, San Pedro del Gallo, San Luis del Cordero, Rodeo, Simón Bolívar and San Juan de Guadalupe. As none of these municipalities provides more than 1 % of dairy cattle inventory, for purposes of the present study were added in “Others” and the N balance was calculated jointly.

N requirement for the crops

To estimate the N requirements, forage crops with more than 100 ha yr^-1 per municipality in the evaluation period were considered. Data of fresh forage production for each crop was transformed to dry matter production (DM), using the DM percentages recorded in Table 1. The requirement of N in forage crops was obtained by N extraction values for ton of DM yield, based on information generated mainly in the Comarca Lagunera region and Mexico^15-19. In the case of alfalfa, it was assumed that 75 % of the total N in the forage comes from biological fixation and the remaining 25 % is taken from the soil; this because these soils are manured on a regular basis and with high levels of available N²⁰.

Table 1 Values of dry matter percentage at harvesting and nitrogen extraction per unit yield according to its harvest stage. 

*Kilograms of Nitrogen per ton of dry matter.

Manure production and nitrogen excretion

Equations generated by Nennich et al²¹ were used to estimate manure production and N excretion (Table 2), similar to other studies(22,23):

E_t = 0.616(DMP) + 46.2

E_DM = 0.0874(DMP) + 5.6

N_E = 2.82(DMP) + 346

Where E_t is the production of total excreta (kg cow^-1 d^-1), E_DM is the production of excreta in DM (kg cow^-1 d^-1), N_E is the total N excreted (g cow^-1 d^-1) and DMP is the daily average milk production (kg cow^-1). As the used equations apply just for lactating cows, the number of cows for calculation purposes was estimated under the following assumptions: 1) From the total inventory of dairy cattle by municipality, 20 % are younger than 1-yr old and were not considered in the balance¹²; 2) Of the number of non-lactating cows, two animals are equivalent to one lactating, as it is considered by Spears et al²⁴.

Table 2 Area, production and forage yield in dry matter, harvested under irrigation in municipalities of the Comarca Lagunera from 1999 to 2011 (mean ± standard deviation). 

NR= Nitrogen requirement.

Cows density

The density (cows ha^-1) was calculated by dividing the number of animals used for the calculation of the balance between the number of harvested hectares of forage crops by municipality.

N balance

N balance by municipality was calculated as the difference between the manure-N incorporated into the soil and N demand for forage crops .i.e., the value of total N excreted was not used in the N balance. It was assumed that 30 % of the total N excreted is lost through volatilization during handling, storage and manure application²⁵.

Statistical analysis

Reviewed data from statistical yearbooks, as well as the calculated data of manure production and N excretion, were analyzed through descriptive statistics (average and deviation standard). Regression analyses were performed between the harvested crop area and milk production, respect to the assessment years; analysis was also ran between the N balance and dairy cattle density per hectare sown with forage, according to the procedure of Saam et al¹³.

Crop pattern

The total irrigated area by crops was 130,000 ha on average for the study period, of which 69 % were harvested with forage; the most important crops by harvested area are: alfalfa, maize, sorghum and oats, as well as some areas sown with rye grass, wheat and triticale (Figure 1). During the study period, the rate of increase in the surface harvested with forages was 4,228 ha yr^-1, from 64,000 ha in 1999 to 115,000 in 2011. The crop contributing more to this increase was forage sorghum, with an increment of 3,561 ha yr^-1 during the period from 2005 to 2011, followed by forage maize, with an increase of 2,584 ha yr^-1 in the period from 1999 to 2007.

Figure 1 Harvested area of forage crops in the Comarca Lagunera region, during the period from 1999 to 2007. 

The regional DM forage production was 1.25 (±0.23) million t yr^-1. The municipality that most contributes to the regional production of forage was Gómez Palacio, with around 25 %, of both the harvested area and forage production. Lerdo, Matamoros, and Francisco I. Madero follow it in importance. Forage yield ranged from 12.8 t ha^-1 of DM in Francisco. I. Madero, 16.1 t in Viesca, with a regional average of 13.9 t (Table 2).

The nitrogen requirement for forage production was 15,070 t yr^-1 on average, which represents a dose of 168 kg N ha^-1 at the regional level. The municipality of Gómez Palacio builds up 27.5 % of the regional requirement, for an average rate of 189 kg ha^-1, while the lower requirements occurred in Nazas, with 2.5 % of the total and a dose of 142 kg.

Inventory of dairy cattle and milk production

The total number of bovines in 1999 was 363,000, with 52.7 % lactating animals; in the year of 2000 the inventory increased to 415,000 and since then, it was between 400 and 450 thousand, with 53.1 % lactating animals. The milk production (MP) in the region increased significantly at a rate of 80.9 million L yr^-1, between 1999 and 2008; the regional production went from 1,535 million L in 1999 to 2.255 million L in 2008; from 2009 the milk production fell down but remained around 2,100 million L yr^-1. The above is the result of an increase in production per lactation by 7,220 L cow^-1 in 2001 to 9,480 L in 2008⁴.

The distribution of dairy cattle by municipality indicates that 79 % of the inventory was concentrated on 4 of the 15 municipalities (Table 3): Gómez Palacio (28 %), Lerdo (19 %), Matamoros (18 %) and Torreón (13 %). With respect to animal density, the lowest value was recorded in “Other” municipalities with 0.9 cows ha^-1, whereas the municipality of Torreón recorded the highest density with 7.72 cows ha^-1. The four municipalities which account for 79 % of the dairy herd in the region (Gómez Palacio, Lerdo, Torreón and Matamoros), had an average density of 5.0 cows ha^-1, whereas the rest of the municipalities averaged 1.7 cows, with a regional average of 3.2 cows.

Table 3 Total inventory, cows in lactation and animal density per hectare of harvested forage from 1999 to 2011 (mean ± standard deviation). 

*Cows per hectare of forage.

Manure production

Total production estimated in fresh weight was 6.624 million t yr^-1, with a 12.71 % of DM, resulting in a manure production of 842,000 t DM yr^-1 (Table 4). From this previous value, 80 % is provided by lactating cows and 20 % for the rest of the herd (dry cows, heifers, and growing calves). As the amount of manure depends mainly on the number of animals, four municipalities which have 79 % of the inventory of dairy cattle (Gómez Palacio, Lerdo, Torreón and Matamoros), accounted for 78 % of the manure produced. Total N excreted was almost 44,200 t yr^-1, with a minimum of 471 t yr^-1 in Nazas, to a maximum of 11,785 t yr^-1 in Gómez Palacio.

Table 4 Manure production and nitrogen in excreta, in municipalities of the Comarca Lagunera from 1999 to 2011 (mean ± standard deviation). 

DM= Dry matter.

N balance

Figure 1 illustrates the N balance of the manure by municipality. The available N to be incorporated into the soil corresponds to 70 % of the values listed in Table 4, assuming that the remaining 30 % is lost by volatilization²⁵; values range from 154 kg ha^-1 at Nazas to 859 kg in Torreon. N requirement by forage crops is less variable among municipalities, with values of 140 kg in Nazas to 190 kg in Gomez Palacio. The N balance illustrated in Figure 3 is the difference between the incorporated manure-N and the N requirement by forage crops. On average, San Pedro, Tlahualilo, and “Others” had a negative balance, while the rest had a positive balance. i.e. in most municipalities manure-N is greater than N requirement for forage crops. The municipalities with the highest concentration of cattle, Gómez Palacio, Lerdo, Matamoros and Torreón, had an average balance of 183, 336, 313 and 691 kg ha^-1, respectively, while the other municipalities had balances of less than 150 kg. The regional average balance is positive, with 187 kg (Figure 2).

(Lines in bars are the standard deviation of the means).

Figure 2 Values of available nitrogen to be incorporated into the soil, nitrogen requirements for forages and of nitrogen balance in municipalities of the Comarca Lagunera. Average from 1999 to 2011. 

Figure 3 N balance in the neighboring municipalities of the Comarca Lagunera, compared with the Regional average and the rest of the municipalities. 

Analyzing the trend of the N balance (Figure 3), it can be seen that the highest average was recorded in 2001, with 291 kg ha^-1; from 2002 the balance value decreases up to 109 kg in 2011 (final year of evaluation). By separating the four municipalities which account for 79 % of heads of cattle (Gómez Palacio, Lerdo, Torreón and Matamoros) from the others, the information shows that those locations reached a maximum balance of 597 kg ha^-1 in 2001 to then decrease to 271 kg in 2011, i.e., they were always above 150 kg. The rest of the municipalities always had a balance of less than 150 kg, reaching a maximum of 88 kg in 2002 and a minimum of -16 kg in 2010. The regional average showed values of 158 kg in 2003 and 2005, and less than 150 kg from 2006 to 2011.

By relating the animal density (CD cows ha^-1) with the N balance (N_B kg ha^-1), the resulting linear regression model is presented in Figure 4:

N_B = 106.6(C_D) - 156

Figure 4 Relationship between the density of cows and the nitrogen balance (average data from 1999 to 2007). 

The previous model was significant (r²= 0.98) and predicts that a system in equilibrium (N_B= 0) is obtained with a value of C_D= 1.6 cows ha^-1. The municipalities with the most positive balance, or greater N excess (Figure 2), are also those who have a higher cow density (Figure 4).

Crop pattern

During the 13 yr analyzed period, it has been a major change in the crops pattern of the Comarca Lagunera. At the beginning of the period (1999), alfalfa held 53.8 % of the forage area; in that same year, maize along with the sorghum occupied 32.5 % of the area, whereas oats forage harvested 7.9 %. At the end of the study (2011), alfalfa decreased to a 34 %, while maize + sorghum increased to 51.2 % and oats 13.3 % of the crops forage area (Figure 1). These changes in the pattern increased the N demand in the region, by decreasing the N biologically fixed by alfalfa and increasing the crops conventionally fertilized²⁶; it increases also the water use efficiency at the regional level without affecting the milk production²⁷.

The regional average yield of 13.9 t DM ha^-1 corresponds to all forage crops, which is greater than the 9.9 t ha^-1 found by Hillhorst et al¹² in an experimental farm in the Netherlands. At farm level, in the Comarca Lagunera Region DM yield per hectare can be greater, as alfalfa has an average yield of 16 t ha^-1. In addition, in irrigated areas with groundwater (64 % of the area) is common to have two or three crops in the same year, with experimental yields of 25 and 18 t DM ha^-1 in maize forage planted in spring and summer, respectively, and 12 t DM ha^-1 with winter cereals²⁸.

Manure production

There are different constants to estimate manure production in animal species^5,29. Fortis et al⁵ estimated a production of 619,000 t DM yr^-1 in the Comarca Lagunera, based on statistics from 2008; the value is less than the average estimated in this study of 842,000 t. In the region of Delicias, Chih., with the use of the Nennich et al equations²¹ a total manure production of 2.9 million t yr^-1 was estimated (freshly excreted weight), with a dairy cattle population of 56,800 heads²². These equations have the advantage that are based on the milk production and can be used at the farm level.

With respect to estimation of the excreted N, dairy cattle has a low efficiency in the use of the ingested N, because 70 % of N intake is excreted¹. Main factors associated with excretion of N are dry matter intake, content of crude protein in the diet, milk production, live weight and days in lactation²¹. An excretion of 44,154 t yr^-1 was estimated in the present study; however, taking into account the manure production in DM (842,000 t yr^-1), with an average total N concentration of 1.61 %³⁰, results a total N content of 13,557 t yr^-1. The difference represents a 69.3 % of N losses during handling, storage and manure application. The process that contributes most to the previous losses is ammonia volatilization, estimated as 25 to 50 % from the total N³¹. In the present study it was assumed a 30 % N loss by volatilization. Part of this high volatilization is because around 38 % of the excreted N, mostly in the urine, is in the form of urea or ammonia⁶. Urea in the presence of urease is hydrolyzed to ammonium (NH₄⁺) and this turns in to ammonia (NH₃) in conditions of alkaline pH³², a common condition in calcareous soils in the North of Mexico. On the other hand, the management of manure by anaerobic digestion (biodigester) can reduce losses by ammonia volatilization to values between 5 and 10 %³³; this technology is now in use and extending in the Comarca Lagunera.

N balance

Balance in Figure 2 indicates an excess of N at the regional level of 168 kg ha^-1, with a variation of -43 kg (other municipalities) to 691 kg ha^-1 (Torreón). These values are comparable with those found in other studies. At farm level, in Australia an average balance of 193 kg ha^-1 was estimated, with a variation of 47 to 601 kg³⁴. A study of 120 farms in Belgium³⁵ found an excess balance of 238 kg. The average balance obtained in this study is higher than the estimated at national level of 22 kg¹⁰, but close to the maximum value of 150 kg suggested to maintain the sustainability of the production system³⁵. In the present study, the municipalities with the highest cows density had a balance greater than 150 kg ha^-1 throughout the period evaluated, although with a clear downward trend, with a maximum of 1,343 kg in 2001, to 508 kg in 2011 (Figure 3). The remaining 11 municipalities recorded a maximum balance of 88 kg in 2002, and -16 kg in 2010.

The decrease of the N balance is due to the increase in the harvested area with forage crops, and that inventory of dairy cattle remained steady from 2000 onwards. A practice to lower the N balance is to move manure surplus to N-deficit municipalities or outside the Comarca Lagunera. This is the approach suggested by FAO³⁶, i.e., integrate livestock areas with neighboring areas of agricultural production where nutrients from manure can be recycled. However, the cost of cattle manure application accounts for 25 to 148 % of the cost for applying fertilizers, depending on the dose required for crops; in addition, the maximum distance to which manure can be moved to match the cost of fertilizer that would apply, can vary from 1 to 92 km³⁷.

Animal density

According to the regression model between animal density (A_D) and N balance (N_B), the AD value for a balanced system of 150 kg ha^-1 is 2.87 cows ha^-1. This value is greater than estimations in Wisconsin by Saam et al¹³, who point out that more than 1.1 cows ha^-1 generate an excesses of N. However, Van Horn et al¹ in Florida, report that at farm level, 1 ha can recycle manure of 10.4 cows; in this balance the manure incorporation is in liquid form and crops used 55.1 % of excreted N in three crops per year, with 29 t DM ha^-1 yield. In the Comarca Lagunera, during the study 66,400 ha were cultivated with irrigation, including 38,000 ha of alfalfa⁴, so the rest of the 28,400 ha are suitable to obtain two or three harvests per year. By including a second crop (62 % of the area with maize and 38 % with sorghum), the DM production increases from 13.9 to 25.0 t ha^-1, increasing the density at 6.7 cows ha^-1, while maintaining a N balance of <150 kg ha^-1.