**2. Materials and methods**

*Land Use Change and Sustainability*

cropland.

War I, farmers optimized their profits by plowing and planting grasslands with annual crops. The period of high yields was followed by drought during the 1930s which resulted in the dust bowl and bust. A recent boom occurred between 2006 and 2012 due to increase in maize and soybean price. During this timeperiod, Reitsma et al. [4] reported that 730,000 ha of grassland was converted to

During rapid land conversion periods, grasslands are often converted to cropland.

This conversion can strengthen the financial resources of individual farms while simultaneously reducing wildlife habitat [5]. Thus, land conversion from grasslands to croplands creates the classical dilemma of balancing economic development with environmental impacts. Between 2008 and 2011, all across the USA, 23.7 million acres of grassland, shrub, and wetland were converted to agricultural land, and 3.2 million acres of wildlife habitat disappeared in North and South Dakotas alone [6]. Grasslands

Worldwide, the NGP ecoregion in North America is considered one of best remaining opportunities for grassland maintenance [7]. Similarly, other adverse side effects of land-use change are increased greenhouse gas emissions [8, 9], reduced water quality [10], and higher soil erosion [11, 12]. In the NGP, the adoption of management practices that improve soil health and minimize soil degrada-

Technology improvements, such as the development of new planting equipment and the wide-scale adoption of transgenic crops, have provided the opportunity to seed annual crops in areas that previously were considered unsuitable for crop production [14]. Moreover, complex interaction of various factors like climatic variability, soil quality, topography, and socioeconomic factors may influence individual decisions [22, 23]. In the NGP, higher rainfall and temperatures linked to

From soil erosion perspective, the conversion of grasslands to cropland may be sustainable if conversion occurs on suitable land type [4]. One approach to assess suitability is the land capability classification (LCC) approach. In this approach, soils with LCC values ≤ 4 are generally considered sustainable for annual crops if appropriate management practices are followed. Soils with LCC values ≥ 6 are not considered suitable for annual crops. Soils with a LCC value of 5 may be prone to flooding. The number of restrictions increases as the LCC value increases from 1 to 4 and from 6 to 8. However, Rashford et al. [25] found that between 1978 and 2008, 0.4 million hectare of cropland increased and most conversions occurred on land are considered suitable for crop production (LCC ≤s 1–4). Rashford et al. [25] also reported that grassland with LCC ≤ 2 has a 30–50% greater probability of being converted to cropland than grassland with

In light of current pressure on land and various forces driving land-use change, it is essential to examine the dynamics of land changes. The objectives of this study were to calculate the rate of land-use change from 2006 to 2012 and from 2012 to 2014 in South Dakota and Nebraska and assess if land-use changes were sustainable. This region was selected as a model system because it is located in a climate transition zone and it has a humid continental climate on the eastern border and semiarid

tion is critical to insure long-term sustainability [13–19]. We believe that increasing the adoption of sustainable management practices requires a clear understanding of factors driving the land-use change. Reitsma et al. [4] reported that land-use change most likely resulted from many factors including recent technological improvements, land ownership structure changes, climatevariability, various governmental policies, crop prices, and aging workforce [4, 14, 20, 21].

are one of the most threatened and least protected ecosystems.

climate change were important [7, 24].

climate on the western border [26, 27].

LCC values of 3 and 4.

**4**

South Dakota and Nebraska were selected as model systems because these states contain a wide range of soil, crops, and climate which are representative of other larger areas; both states have a large production capacity for livestock and annual crops; most of the soils were developed in tall and mixed grass prairies; they are located in climate transition zone; and the two states have different access to irrigation water. This region receives most of its precipitation in the spring and fall [14].

The most common annual crops in South Dakota include maize (*Zea mays* L.), soybean [(*Glycine max* (L.) Merr.], and wheat (*Triticum aestivum* L.). In South Dakota, rainfall decreases from east to west, and temperatures decrease from South to North. Additional information on characteristics of South Dakota soils is available in Reitsma et al. [4] and Clay et al. [13]. Farmers in this region use crop rotations that include maize, soybean, wheat, sunflower, canola (*Brassica napus* L.), barley (*Hordeum vulgare* L.), lentil (*Lens culinaris* Medik.), flax (*Linum usitatissimum* L.), and pea (*Pisum sativum* L.).

Eastern Nebraska has a humid continental climate, whereas the western region has a semiarid climate [26, 27]. Eastern part of Nebraska has fertile, moist, and warm soil making it well suited for maize and soybean production. It consists of

#### **Figure 1.**

*South Dakota and Nebraska states in the US map along with USDA's National Agricultural Statistics Service (NASS) reporting regions. (Source of Data, USDA-NASS).*

loess and glaciated till soils. The Nebraska Sand Hills are contained almost entirely within the Nebraska North NASS region, and it represents one of the most unique and homogenous ecoregions in North America. The Sand Hills are one of the largest areas of semiarid grass-stabilized sand dunes in the world [28].

## **2.1 Assessing land-use change**

The method to assess land use was previously discussed in Reitsma et al. [4] and summarized below. South Dakota has nine National Agricultural Statistics Service (NASS) regions (USDA-NASS, 2015) that include the northeast (NE), south east (SE), north central (NC), east central (EC), central (C), south central (SC), northwest (NW), west central (WC), and south west (SW). Similarly Nebraska has eight NASS regions that include the northwest (NW), north (N), northeast (NE), central (C), east (E), south west (SW), south (S), and south east (SE) (**Figure 1**). Stratified random sampling approach was used for sampling and within each of 17 USDA-NASS reporting districts. In each NASS region, 1600 sampling points were randomly identified using ESRI® ArcMap 10.2.2. These points were laid over high-resolution imagery, obtained from the US Department of Agriculture (USDA), Farm Service Agency (FSA), National Agricultural Imaging Program (NIAP) (USDA-FSA, 2013). The NAIP imagery for 2006 had a 2 m resolution, and the 2012 and 2014 imagery had a 1 m resolution. At each point (8 by 8 m), the dominant land use (cropland, grassland, habitat, non-Ag, and water) was visually identified (for 2006, 2012, and 2014). In South Dakota, 43,200 points in total were visually classified (14,400 points each year), whereas in Nebraska, 38,400 points were classified (12,800 points each year). For validation of our visual assessment and classification system, we randomly selected 100 sampling points from 17 different counties in South Dakota. The predicted management based on the remote sensing data (visual classification) was identical to the known management at these points 100% of the time.

### **2.2 Assessing changes in soil quality**

Land capability class (LCC) and dominant subclass were obtained from the Soil Survey Geographic (SSURGO) data set by superimposing the sampling points over SSURGO [29]. At these points, the LCC value was determined [24]. LCC subclasses are used to help define the limitation. The most common subclass limitations are erosion hazard (e), wetness (w), rooting-zone limitations (s), and climate (c). However this was different than Reitsma et al. [4] where the LCC value was the sum of the component soils within a mapping percent multiplied by its numeric LCC value. This change in classification approach may result in slightly different percentages of soils within a LCC category as reported by Reistma et al. [4].
