Translator Disclaimer
1 July 2018 Assessment of Animal-Based Methods Used for Estimating and Monitoring Rangeland Herbivore Diet Composition
Author Affiliations +
Abstract

Researchers and managers need effective tools for monitoring the use of forages by large herbivores. Since 2000, the number of herbivore diet studies has nearly doubled. In this review, we determine trends in the field; assess the utility of key techniques against five criteria (cost, accuracy and precision, resolution, utility for long-term monitoring programs, and appropriateness for browsers and grazers); and make recommendations to give managers appropriate tools. Three techniques stand out: microhistology, near infrared reflectance spectroscopy, and deoxyribonucleic acid (DNA) barcoding. Microhistology has a long history of use in rangelands and is often considered the gold standard for understanding diet composition, albeit at a high cost of labor. Near infrared reflectance spectroscopy can resolve the presence of target groups or species more quickly than microhistology, especially for grazers. DNA barcoding provides the greatest resolution of dietary items with less quantitative certainty than microhistology. The costs associated with DNA barcoding come primarily from technology and sequencing, while in microhistology they are associated with labor. Therefore, an improved, streamlined microhistology method could provide rangeland managers a rapid and cost-effective method for diet monitoring. Ultimately, the complex challenges facing rangeland managers today may require the use of more than one method to achieve acceptable resolution within actionable time frames.

Introduction

Herbivorous mammals play a key role in the structure and function of rangelands (Hobbs, 1996), particularly in systems where predators are functionally absent (Vavra et al., 2007) and where native species interact with domestic livestock and exotic or feral herbivores (Schwartz and Ellis, 1981; Bakker et al., 2006; Vavra et al., 2007; Nuñez et al., 2010).The impacts of mammalian herbivores on rangelands depend on both environmental features such as aridity and history of disturbance and animal features including body size, diet type, and evolutionary origin (native/exotic/domestic) (Augustine and McNaughton, 1998; Bakker et al., 2006; Vavra et al., 2007). Diet is particularly important, as it gives insight into ecological and evolutionary processes like habitat selection, competitive interactions (e.g., coevolution of primary producers), and body condition in the typically nutrient poor environment of rangelands (Krebs, 1998). Forage for large herbivores consists primarily of graminoids (monocotyledons such as grasses and sedges), forbs (herbaceous dicotyledons), and browse (woody dicotyledons), with fruits, fungi, and seeds making up smaller components of the diet (du Toit and Olff, 2014).

Understanding and managing the complex relationships between herbivores and plants in rangelands requires effective techniques for monitoring herbivore diet. As all communities experience temporal change, one of the biggest challenges for monitoring programs is to distinguish among factors (e.g., natural climate change vs. over stocking) (Magurran et al., 2010). For this reason, long-term monitoring, ideally across 20- to 30-yr time frames, is desirable for making decisions about landscape management practices such as habitat modification through prescribed burning, managed stocking densities, and harvest regimes.

Methods to assess herbivore diet focus on measuring either the amount of plants that have been removed (plant-based methods) or physical/ chemical aspects of samples collected from animals (animalbased methods). Plant-based methods are of limited use for freeranging animals, so they will not be discussed further here (Holechek et al., 1982; Mayes and Dove, 2000). Animal-based methods to assess diet composition in free-ranging mammalian herbivores range from direct observation of feeding activity to measures of markers and metabolites from the diet in feces (Holechek et al., 1982; Mayes and Dove, 2000; Barboza et al., 2009). Reported diets are affected by the method, material analyzed, and their interaction. Appropriate methodology may differ depending on whether the herbivore is domestic or wild, as the diets of domestic livestock can be more readily sampled by observation and invasive procedures (e.g., fistulas). For long-term monitoring programs, sampling should be noninvasive, as this allows for repeated measurements and is appropriate for common and rare species. Fecal collection is simple, repeatable, inexpensive, and applicable to both wild and domestic species. However, all techniques that rely on postingestive samples are subject to the effects of differential digestion (e.g., rates of flow and extraction of nutrients) of plant species and plant parts (Fig. 1). Increased fragmentation and breakdown of soft components (e.g., leaves), compared with harder components (e.g., stems) all impact the accuracy and precision of the diet estimate over and above any sources of error inherent in particular techniques.

Comprehensive reviews of techniques for estimating herbivore diets were published by Holechek and colleagues in 1982 and again by Mayes and Dove in 2000. Recent advances in technology, particularly the advent of deoxyribonucleic acid (DNA) barcoding, have increased the options available to managers to assess herbivore diets. It is therefore appropriate to revisit techniques for herbivore diet measurement and to make recommendations about timely and effective methods for answering key management questions.

In this review we:

  • Survey the techniques used in recently published literature (since 2000), to determine trends in the field;

  • Assess the utility of these key techniques for current applications against five criteria (cost, accuracy and precision, resolution, utility for long-term monitoring programs, and appropriateness for browsers and grazers); and

  • Make recommendations to give managers appropriate tools for timely diet assessment.

Figure 1.

Noninvasive, animal-based techniques to assess free-ranging mammalian herbivore diet composition. Morphological techniques identify diet components based on structural characteristics of plants. Behavioral observation is the only technique that does not involve digestion of a sample by the herbivore before observation by the researcher or processing by the researcher before analysis. This means that this technique is uniquely unaffected by digestion of diet components. Due to the time involved in behavioral assessment, this technique is best suited to research questions. Microhistology is equally suited to research and monitoring. Associative techniques range in scale from chemically assessing elements (isotope analysis) to chemically assessing sequences (deoxyribonucleic acid [DNA] barcoding). Isotopic analysis gives the coarsest taxonomic scale (grass vs. browse), whereas DNA barcoding can identify plant subspecies in some applications.

f01_449.jpg

Survey of Techniques

We focused our review on commonly used animal-based methods to assess free-ranging herbivore diets: behavioral observation, microhistology, near infrared reflectance spectrometry (NIRS), stable isotopes, cuticular wax alkanes, and DNA barcoding. Using Web of Science (an online database containing information from ≈ 8 500 research journals worldwide), we searched for papers published between 2000 and 2017 with the words “herbivore” and “diet.” We only included records within documents classified as articles, data papers, database reviews, discussions, early access, or proceedings papers. Our search in February 2018 yielded 2 981 results. On the basis of the inspection of the top results, as well as our understanding of the field, we narrowed our search by including one of “bite count OR bite rate,” “microhistology,” “NIRS,” “n-alkane AND wax,” “stable isotope,” or “DNA barcode” We also included either “selection” or “composition” to try to eliminate papers that looked solely at diet quality or digestibility. We excluded “insect,” “reptile,” “bird,” “fish,” “marine,” or “carnivore” because we were interested in the use of techniques based on plant fragments. There were 417 papers included in our final survey.

There were almost twice as many papers published on herbivore diet in 2015 – 2017 (n = 80) as there were in 2000 – 2002 (n = 45). The relative use of behavioral observation and microhistology has declined slightly in the past few years, while NIRS and stable isotope methods have remained relatively consistent (Fig. 2). Use of cuticular wax n-alkanes has noticeably declined since about 2009. In the same time period, the use of DNA barcoding has greatly increased.

Assessment Criteria

Forages are identified by morphological characteristics in behavioral observation and microhistology, whereas the chemical properties of samples are used to reconstruct diets in stable isotope analysis, NIRS, cuticular wax alkanes, and DNA barcoding (see Fig. 1). We separately evaluated each technique against five criteria: cost, precision, resolution, timescale, and diet type. We compare the techniques in the conclusion of our paper.

Cost

The cost of monitoring includes setup costs, longer-term running costs, and personnel time and effort. These costs can be prohibitively high in some cases, especially in developing countries (e.g., Caro, 2016). While some authors explicitly compare costs between techniques (e.g., Newmaster et al., 2013), we focus instead on the relative contribution of labor versus equipment to the expense of particular techniques. This is because the costs associated with each technique depend greatly on the level of replication (i.e., number of samples) and resolution (e.g., conducting microhistology on functional groups vs. individual plant species). Large sample size reduces the cost for techniques with high setup costs and low labor costs for individual samples (e.g., NIRS) but not techniques that have high labor cost per sample (e.g., behavioral observation). Lower resolution can decrease the cost for microhistological analysis because less time is spent identifying individual species, but reduced resolution is not an option for some techniques (e.g., stable isotope analysis). Decisions about the tradeoff between replication and resolution depend critically on the objectives of particular studies.

Accuracy and Precision

Precision refers to the repeatability of a technique when performed on the same sample. Accuracy is the similarity between the estimated diet and actual diet consumed by the herbivore (Figs. S1, S2 [available online at  https://doi.org/10.1016/j.rama.2018.03.003) and is much harder to assess, especially when precision is low or the actual amount of the diet item is low. All techniques will have some tradeoffs among cost, labor input, accuracy, and precision (Amos et al., 2014).

Figure 2.

Relative use of techniques to assess herbivore diet composition between 2000 and 2017, assessed through citations in Web of Science. Numbers above bars indicate number of citations observed in each time period.

f02_449.jpg

Resolution

Resolution refers to the taxonomic level (e.g., family, genus, species) that a technique is capable of detecting and affects the level of change required before a signal can be detected. Some methods are only capable of determining differences in browse versus graze consumption (Mayes and Dove, 2000), while others may detect plant identity down to the subspecies level (Soininen et al., 2009). Variation is higher at lower taxonomic ranks and requires more intensive sampling to detect differences. As diet diversity increases, the number of minor species increases and the ability to detect changes in proportions of minor components declines. While low abundance items can be important (e.g., for nutritional balancing: Raubenheimer and Simpson, 1997), changes in the signals of these items are typically hard to detect (but see Altangerel et al., 2017). It is more likely that monitoring will pick up long-term shifts in the bulk of the diet of graminoids or browse.

Timescale

Ideally, long-term monitoring programs use a single, unchanged technique for data collection (Magurran et al., 2010). Although changes to methodology can be incorporated (e.g., Freeman et al., 2007), analysis of data is simpler when methodological variation is minimized. The stability of methods over time is an important determinant of their utility in long-term monitoring. This criterion should not count against innovation (Magurran et al., 2010), but it urges caution in using new, relatively untried techniques in establishing monitoring programs.

Diet Type

Herbivores tend to specialize as browsers or grazers to best contend with the challenges associated with particular forage types (Hofmann and Stewart, 1972). Browsers have morphological and phyisiological adaptations to avoid or tolerate plant chemical defenses and to digest highly lignified cell walls of dicotyledonous plants. Conversely, grazers are adapted for longer retention of plant material to increase digestion of less lignified cell walls of monocotyledonous plants (Clauss et al., 2008). Grazers are better studied than browsers because domestic livestock are mostly grazers (sheep, cattle, and horses), and only domestic goats are browsers (du Toit and Olff, 2014). Compared with grazers, the diets of browsers tend to contain more species and a wider range of plant secondary compounds, which complicate some analyses.

There is a need to develop rapid and reliable techniques to assess browser diets.

Behavioral Observation

Behavioral observation has long been used to determine diet composition in livestock (Cory, 1927) and wildlife (Dixon, 1934; Wallmo and Neff, 1970). In its simplest form, the technique requires little equipment and is straightforward to implement. Recently, animal-borne video recording systems show promise for capturing behavioral data such as diet selection without the need for direct observation. First used on turtles in 1987 (Moll et al., 2007), these systems have been modified for use in ungulates (Beringer et al., 2007) and are reasonably efficient at capturing data on foraging behavior. For example, Newmaster et al. (2013) found a high correlation (70%) between plant groups in the diet of woodland caribou (Rangifer tarandus caribou) estimated by video recording and DNA barcoding.

Diet selection is directly observed, so there is no processing of the diet sample by the animal or the researcher (see Fig. 1). This eliminates errors such as differential digestibilities that plague techniques based on fecal samples (e.g., Shrestha and Wegge, 2006). Even with extensive training, though, it may be impossible to identify some plant species from a distance. Direct observation of behavior requires little equipment, but the cost of equipment may be offset by the large commitment of labor and limited number of animals visible to each observer. Labor costs are lower for indirect observation of behavior with recorders, but the systems can be expensive to acquire and maintain (e.g., Newmaster et al., 2013). Due to the time-consuming nature of observations, behavioral observation studies typically have limited sample size, particularly those using animal-borne video (e.g., Beringer et al., 2007, n = 3; Thompson et al., 2012, n = 5; Newmaster et al., 2013, n = 15). Functional groups, such as grasses/forbs/shrubs and trees can easily be identified, and with appropriate training, it is possible to identify diet items to species or genus level. Currently, video recording systems are not adequate to distinguish plant species (e.g., Newmaster et al., 2013).

Provided that sampling protocols are well documented, a condition that applies equally to all techniques, direct observation of behavior is well suited to long-term monitoring and has a long history of use. Other advantages include the ability to simultaneously collect data on animal condition, habitat use, behavioral interactions, and activity budgets, all of which can help influence management decisions (Holechek et al., 1982).

The key disadvantage of this technique is its inapplicability to situations where animals cannot be directly observed, such as for rare and cryptic species or visually dissected habitats. At the landscape scale, such as in a savanna, this limitation applies disproportionately to browsing species, as grazers typically forage in open habitats. At the bite scale, it is easier to observe selection of aboveground browse than surface-level grasses. Behavioral observation is therefore more viable for use in livestock, as they are more readily observed (Shrestha and Wegge, 2006; Damiran et al., 2013). One alternative is to use habituated or tame wildlife (Wallmo and Neff, 1970; Timmons et al., 2010). However, because tamed animals are normally captured shortly after birth with limited opportunity to experience the socially facilitated learning component of diet selection, the assumption that these animals are representative of the broader population should be confirmed (Holechek et al., 1982) by comparing diets of habituated and wild animals (Spalinger et al., 1997). Here animal-borne video shows promise as a technique to augment other data capture systems (e.g., Global Positioning System telemetry, Thompson et al., 2012), particularly for species that are challenging to observe (Moll et al., 2007).

Microhistology

First reported by Baumgartner and Martin (1939), microhistology is a well-established and stable technique. Users identify plant cuticle fragments through characters such as cell size, shape, and arrangement using a reference collection of known plant species (Holechek et al., 1982; Norbury, 1988). Recently, some innovative approaches to machine learning and pattern recognition have capitalized on the capacity for computers to recognize and identify images to help solve ecological questions (e.g., Bolger et al., 2012; Yu et al., 2013; Swinnen et al., 2014). While we are unaware of any attempts to do so, such methods could reasonably be applied to reading microhistology slides of herbivore diet to improve the accuracy, precision, and speed and reduce expenses.

Because it uses simple equipment that is widely available, setting up microhistology is relatively inexpensive (Holechek et al., 1982; Mayes and Dove, 2000). However, microhistology is time intensive and it may take users a substantial amount of time to become proficient in the technique (Holechek and Gross, 1982). Automated image recognition processes would assist in reducing labor costs.

A time-consuming aspect of the technique involves creating and becoming familiar with the reference collection. Recent advancements in the ability to digitize and classify reference slide collections (e.g., Shrestha and Wegge, 2006) enhance the ability to transfer reference collections between labs, thereby reducing setup time for this technique but not training time for each technician. Recently, Desbiez and Santos (2014) developed an interactive, online taxonomic key for identifying microhistological characters for 206 species of plants consumed by herbivores in the Brazilian Pantanal. The defining feature of interactive keys is that characters can be used in any order; characters that are not available in a particular epidermal fragment on a slide, or whose interpretation is not clear to the user, can be avoided ( http://delta-intkey.com). Such keys could be developed for other ecosystems and herbivores. Their use would greatly speed up the time for new users to learn microhistological plant identification and improve consistency between users (Desbiez and Santos, 2014).

The greatest disadvantage of microhistology is the potential effect of differential digestibilities of individual plant parts, species, or functional groups (Holechek et al., 1982; Mayes and Dove, 2000). Under such a situation, highly digestible components of the diet are likely to be underestimated (e.g., Shrestha and Wegge, 2006; Wam and Hjeljord, 2010). To improve precision, correction factors for differential digestibilities (e.g., Leslie et al., 1983; Norbury, 1988) and observer training (Holechek and Gross, 1982) have been developed. In some studies, macrohistological examination of rumen or stomach contents has been used to detect stems, seeds, fungi, or flowers (Forsyth and Davis, 2011).

Microhistology can generate high-resolution information about diets, down to individual species at times (Holechek et al., 1982; Mayes and Dove, 2000). However, to minimize concerns about differential digestibilities and the number of fragments that cannot be identified, many studies report diet composition at higher levels, such as family or functional group.

As for all methods using fecal samples, microhistology is well suited to long-term monitoring programs because sample collection is relatively straightforward and repeat samples can be collected from individuals or populations. Due to its long history of use, in many cases there is a wealth of published data using microhistology against which to compare current patterns (e.g., Scasta et al., 2016). Finally, microhistology is equally suited for use in grazers and browsers.

Near Infrared Reflectance Spectroscopy

This technique is based on the principle that reflectance in the near infrared spectrum represents the chemical structure of a sample. It was first applied to understanding herbivore forage quality by Norris et al. (1976). NIRS can be conducted on fecal, ingesta, or forage samples. Applications on fecal samples are not subject to the effects of differential digestibility of diet components because calibrations are based on the relationship between fecal spectra and diets of known composition.

Fecal NIRS has most often been used to determine the botanical composition of diets by estimating the presence of one (e.g., Walker et al., 1998, 2002; Snowder et al., 2001; Waldron et al., 2009; Jean et al., 2014) or two focal species/genera (e.g., Glasser et al., 2008). The greatest weakness of NIRS is that the technique requires independent validation and continual monitoring of calibrations (Dixon and Coates, 2009), which has rarely been done. Despite this, the technique has been independently validated in several studies that showed calibrations were robust for different communities, plant species within a genus, and herbivore species (Walker et al., 1998, 2002). However, Walker et al. (2002) cautioned that fecal NIRS for determining botanical composition should only be considered interval-scale data.

Because NIRS is nondestructive and only requires a small amount of sample, it is potentially useful for long-term monitoring (e.g., Walker et al., 2013), particularly where time and funding are available to develop a detailed reference set for calibrations. Although the technique does not identify individual diet components well, it can be used to track changes in diets over time. Furthermore, multiple constituents can be simultaneously assessed (e.g., nitrogen, fiber, tannins) (Foley et al., 1998), which gives managers additional information about animal nutritional status. For example, NIRS is widely used to monitor the diets of grazers such as domestic sheep (Ovis aries), cattle (Bos taurus) (Walker et al., 2002), and bison (Bison bison) (Craine et al., 2013). In contrast, NIRS is potentially more challenging to use in browsers (but see references by Walker and colleagues), due to the diversity of plant species and chemical compounds typically found in their diets.

Stable Isotope Analysis

Stable isotopes are used as markers in a wide range of organisms. While 13C/12C and 15N/14N are most frequently used (Kelly, 2000; Mayes and Dove, 2000), other isotopes have also been used (Ehleringer and Rundel, 1988). Stable isotopes were first used to assess herbivore diet by DeNiro and Epstein (1978) and Vogel (1978). Tissue (e.g., hair, bone, tooth enamel), ingesta, or fecal samples can be used.

Although isotope analysis requires expensive equipment and considerable expertise, it is provided as a standard laboratory service that can be quick and accessible. In interpreting stable isotope data, it is important to use an appropriate discrimination factor, which may significantly influence estimates of dietary composition (Barboza et al., 2009; Codron et al., 2012; Gustine et al., 2014). Furthermore, fractionation and turnover of isotopes in tissues may obscure the contributions of particular plant types if the diet varies through time (Tieszen et al., 1983; VanSomeren et al., 2017).

The greatest disadvantage of isotopic analysis is that it is unable to give species-level resolution, so it is limited to applications such as understanding diet type (grazer vs. browser) or indicating trophic position (Mayes and Dove, 2000). The ratio of 13C to 12C isotopes can indicate the relative proportions of dicotyledonous (C3) and monocotyledonous (C4) plants in the diet (Kelly, 2000). 15N can indicate the trophic position of a consumer and aridity of particular habitats, and it is more useful than 13C for ecosystems with few C4 plants or a mixture of C3 and C4 monocots (Kelly, 2000; VanSomeren et al., 2017).

The technique has been widely used with tissues from extinct animals. As samples from feces and tissue represent different time frames, it is possible to estimate diet across a range of scales if different samples are used simultaneously (Tieszen et al., 1983). For this reason, stable isotopic analysis is a useful technique for long-term monitoring where time and funding are available to establish a reference library and detailed data on diet composition are not required.

In some situations (e.g., where there are few C4 plants or a mixture of C3 and C4 monocots), multiple isotopes and markers may need to be used to distinguish browsers from grazers (Codron and Codron, 2009). Otherwise, the technique works well for both browsers and grazers.

Plant Cuticular Wax Alkanes

The indigestible waxes and alkanes of plant cuticles contain lipidsoluble compounds that can identify plant species and parts. This method was suggested by Mayes et al. (1986) and substantially reviewed by Mayes and Dove (2000). Because the technique relies on fecal samples, it is affected by differential digestibility of diet components.

Alkane measurement requires equipment for extraction and detection that is not readily available as a standard laboratory service. Furthermore, the technique requires substantial investment to develop appropriate calibrations. The principal detraction of this technique is the finding that the signature of individual plants and plant species changes throughout the year and between locations (Carnahan, 2011). For this reason, the technique is infrequently used to estimate diet composition in free-ranging animals (see Fig. 2).

Only as many diet components can be estimated as there are markers, which depends on the uniqueness of the patterns in each mixture and statistical discrimination of multiple alkane concentrations (Barboza et al., 2009). For this reason, the technique is of limited use in systems where animals consume botanically complex diets (Bugalho et al., 2004). To get around this issue, additional components of cuticular waxes, such as wax esters, long-chain fatty acids, and long-chain fatty alcohols, have been used in addition to n-alkanes with some success (e.g., Bugalho et al., 2004; Carnahan et al., 2013).

If resources are available to develop appropriate calibration data sets, cuticular wax markers may be appropriate for longer-term monitoring. These markers also allow concurrent estimation of dietary intake (Mayes and Dove, 2000), which may provide additional data to influence management decisions. Due to limitations of use in assessing botanically complex diets (Bugalho et al., 2004) like those of browsers, the technique is best suited to grazers.

DNA Barcoding

DNA barcoding involves sequencing target DNA and matching it to a database of known sequences to identify the taxonomic origin (Valentini et al., 2009; Clare, 2014). The potential to use fecal DNA to identify diets in free-ranging animals was first suggested by Hoss et al. (1992). With the advent of next-generation sequencing, DNA analysis has become more accessible (Valentini et al., 2009; Clare, 2014) and has recently become popular for use in determining diet composition in large herbivores (e.g., Pegard et al., 2009; Ait Baamrane et al., 2012; Hibert et al., 2013; Newmaster et al., 2013; Kartzinel et al., 2015).

DNA barcoding offers fast processing and turnaround times and is provided as a standard laboratory service (Valentini et al., 2009; Clare, 2014). The costs therefore depend on the number of samples analyzed. DNA barcoding is particularly useful when the diet cannot be determined morphologically (Kohn and Wayne, 1997). Furthermore, sequences are not prone to subjective interpretation (Valentini et al., 2009; Clare, 2014). The chief advantage of the technique is that it can generate a complete list of species in the diet, down to subspecies level at times (e.g., Soininen et al., 2009), and fewer samples are required to detect a particular or rare species than other methods (Clare, 2014). It is therefore particularly useful for detecting the presence of invasive or threatened plants in herbivore diets to determine the impact of animals as vectors or consumers (Valentini et al., 2009; Clare, 2014).

Sources of error using barcoding to quantify botanical composition of diets include technological errors (e.g., amplification efficiency, extraction efficiency, bioinformatic sorting) and biological errors (e.g., tissue cell density, gene copy number, DNA survival) (Pompanon et al., 2012). DNA barcoding is a single-locus identification system (Valentini et al., 2009), and the choice of primer dictates which species will be identified (Deagle et al., 2010; Clare, 2014). Recent work on herbivore diets has focused on the chloroplast trnL intron (e.g., Taberlet et al., 2007; Pegard et al., 2009; Raye et al., 2011; Ait Baamrane et al., 2012; Hibert et al., 2013; Kartzinel et al., 2015). However, at least four common target regions have been recommended for use in plants (Clare, 2014; Kartzinel et al., 2015) as DNA barcoding of plants to species level often cannot be accomplished using a single region. In most applications, diet items can only be resolved to the family or genus level, even with the use of additional primers (e.g., Pegard et al., 2009; Raye et al., 2011 ; Ait Baamrane et al., 2012; Hibert et al., 2013; Kartzinel et al., 2015).

Table 1

Key limitations and best applications for noninvasive diet monitoring.

t01_449.gif

As fecal DNA is typically degraded, it is difficult to amplify fragments longer than 150 base pairs (Valentini et al., 2009). Sequencing larger regions maximizes the taxonomic information extracted, but sample degradation affects the length that can be recovered (Clare, 2014). There may be a significant bias toward detecting undegraded DNA, which would limit taxonomic identification. Using shorter amplicons may overcome this bias but could overestimate diversity (Clare, 2014). A related concern is that heteroplasmy (the presence of multiple copies of genes), which is common in plants, can further complicate species identification (Valentini et al., 2009). There is also evidence that the mean copy number of shorter DNA fragments is higher than longer fragments, such that shorter sequences are overrepresented in the total number of sequences obtained per fecal sample (Raye et al., 2011). Finally, all DNA barcoding relies on existing databases, such as GenBank, as reference libraries. These libraries are imperfect (Harris, 2003) as a result of sequencing errors, contaminations, sample misidentifications, or taxonomic problems (Valentini et al., 2009).

While the assignment of sequences to species may be relatively straightforward, the quantitative interpretation of barcoding results is not (Valentini et al., 2009). Neither traditional quantitative genetics techniques (e.g., qPCR/rtPCR: McCracken et al., 2012), nor using the number of sequences recovered as a proxy for abundance (e.g., Deagle et al., 2010) have been particularly effective (Clare, 2014). This is because original diet items are differentially digested (fragmentation of DNA) and tissues contain different densities of DNA originally (mean copy number per diet item). These issues affect interpretations of quantity in the original diet, even in very simple systems (Deagle et al., 2010; Clare, 2014). Kartzinel et al. (2015) employed a novel approach to using relative read abundance as a proxy for proportional composition by first validating assessments of grass versus browse in the diet through stable isotope analysis. Further safeguarding against misidentification, Kartzinel et al. (2015) used multiple primers in addition to trnl to identify species and used a reference collection from their study site to compare sequence reads against. While the interpretation of DNA barcoding results is being refined (e.g., Kartzinel et al., 2015) and correction factors are being developed (e.g., Thomas et al., 2016), best practice dictates that both rare and common items should simply be recorded as “present” (Pompanon et al., 2012; Clare, 2014).

Figure 3.

Decision tree for refining monitoring technique. Blue boxes indicate key decision points. Orange boxes on the edges of the figure indicate techniques. Complementary techniques may be required in any one study if multiple questions are important.

f03_449.jpg

DNA barcoding has been successfully used for diet analysis in extinct species (van Geel et al., 2008; Welker et al., 2014). Endogenous DNA in feces can also be used to identify or confirm contributing animal species, sex, or even individual identity (Kohn and Wayne, 1997; Raye et al., 2011). However, because DNA barcoding is a rapidly developing technique, changes in the methodology, particularly in the use of primers, mean that current diet estimates may not be comparable with those generated in the future under different methodologies. The technique is equally suited for use in browsers and grazers.

Management Implications

The best technique to assess herbivore diet depends on the monitoring priority (Table 1, Fig. 3). Therefore, well-executed monitoring programs need to have a clearly defined question (Nielsen et al., 2017). There will often be a tradeoff between cost and sample size for adequate resolution, precision, and accuracy of the system. Morphological methods (behavioral observation and microhistology, see Fig. 1) require less equipment but more labor. Associative methods (NIRS, stable isotopes, wax alkanes, and DNA barcoding) require expensive technology to process samples, but the provision of many of these techniques as standard laboratory services reduces the cost per sample and allows for quick turnarounds (Foley et al., 1998). Morphological methods potentially over-represent resource overlap between consumers by failing to detect subtle differences between plant species; molecular methods potentially over-represent rare species and overemphasize resource partitioning (Clare, 2014).

For studies that require data on diet composition, microhistology and DNA barcoding present the best options (see Fig. 3). Despite the well-known challenges, microhistology is often seen as the “gold standard” for diet assessment and is often used as reference technique against which newer techniques are compared (e.g., Soininen et al., 2009; Carnahan, 2011; Newmaster et al., 2013). If the potential for image recognition is realized for microhistology, it could become the method of choice to influence decision making about vegetation management (e.g., fire fuels) and herbivore management (e.g., harvest quotas) in a time-sensitive manner because of the potential for rapid high-resolution determination of diet composition that could be routinely validated by trained technicians. DNA barcoding is a relatively new yet increasingly popular technique (see Fig. 2). While refinements are being continuously developed (e.g., Kartzinel et al., 2015; Thomas et al., 2016), until some of the more serious challenges have been resolved, the results of DNA barcoding for diet studies in herbivores should be interpreted conservatively. Behavioral observation may be useful for some studies requiring diet composition data but is limited to situations where animals can be readily observed (see Fig. 3). If a coarser scale of measurement is appropriate, and there are resources available to develop calibration data sets, cuticular wax alkanes, NIRS, and stable isotopes are potential options that can provide rapid results (see Table 1). In particular, for studies that require botanical composition of one or two species of interest, N1RS is the most viable alternative. Ultimately, the complex challenges facing rangeland managers today may require the use of complementary techniques to achieve acceptable resolution within actionable time frames.

Acknowledgments

We thank Bruce Davitt and Betty Blatner, Wildlife Habitat and Nutrition Laboratory, Washington State University, for advice about microhistology; Dr. Rola Barhoumi Mouneimne, Image Analysis Laboratory Texas A&M University, for advice about high-quality microscopy; and Dr. Richard Ploeg, Department of Veterinary Pathobiology, Texas A&M University, for advice about image recognition in histology.

References

1.

Ait Baamrane, M.A., Shehzad, W., Ouhammou, A., Abbad, A., Naimi, M., Coissac, E., Taberlet, P., Znariet, M., 2012. Assessment of the food habits of the Moroccan Dorcas gazelle in M'Sabih Talaa, west central Morocco, using the trnL approach. PLoS One 7, e35643. Google Scholar

2.

Altangerel, N., Walker, J.W., González, P.M., Bailey, D.W., Estell, R.E., Scully, M.O., 2017. Comparison of near infrared reflectance spectroscopy and Raman spectroscopy for predicting botanical composition of cattle diets. Rangeland Ecology & Management 70, 781–786. Google Scholar

3.

Amos, M., Baxter, G., Finch, N., Lisle, A., Murray, P., 2014. I just want to count them! Considerations when choosing a deer population monitoring method. Wildlife Biology 20, 362–370. Google Scholar

4.

Augustine, D.J., McNaughton, S.J., 1998. Ungulate effects on the functional species composition of plant communities: herbivore selectivity and plant tolerance. Journal of Wildlife Management 62, 1165–1183. Google Scholar

5.

Bakker, E.S., Ritchie, M.E., Olff, H., Milchunas, D.G., Knops, J.H.M., 2006. Herbivore impact on grassland plant diversity depends on habitat productivity and herbivore size. Ecology Letters 9, 780–788. Google Scholar

6.

Barboza, P.S., Parker, K.L., Hume, I.D., 2009. Integrative wildlife nutrition. Springer, Berlin, Germany, 342 pp. Google Scholar

7.

Baumgartner, L.L., Martin, A.C., 1939. Plant histology as an aid in squirrel food-habit studies. Journal of Wildlife Management 3, 266–268. Google Scholar

8.

Beringer, J., Millspaugh, J.J., Sartwell, J., Woeck, R., 2007. Real-time video recording of food selection by captive white-tailed deer. Wildlife Society Bulletin 32, 648–654. Google Scholar

9.

Bolger, D.T., Morrison, T.A., Vance, B., Lee, D., Farid, H., 2012. A computer-assisted system for photographic mark-recapture analysis. Methods in Ecology and Evolution 3, 813–822. Google Scholar

10.

Bugalho, M.N., Dove, H., Kelman, W., Wood, J.T., Mayes, R.W., 2004. Plant wax alkanes and alcohols as herbivore diet composition markers. Journal of Range Management 57, 259–268. Google Scholar

11.

Carnahan, A.M., 2011. Determining the diets of moose (Alces alces) in Alaska using plant wax compounds [thesis]. University of Alaska, Anchorage, AK, USA. Google Scholar

12.

Carnahan, A.M., Spalinger, D.E., Kennish, J.M., Collins, W.B., 2013. Extraction and analysis of plant alkanes and long-chain alcohols using accelerated solvent extraction (ASE). Wildlife Society Bulletin 37, 220–225. Google Scholar

13.

Caro, T., 2016. Guidelines for wildlife monitoring: savannah herbivores. Tropical Conservation Science 9, 1–15. Google Scholar

14.

Clare, E.L., 2014. Molecular detection of trophic interactions: emerging trends, distinct advantages, significant considerations and conservation applications. Evolutionary Applications 7, 1144–1157. Google Scholar

15.

Clauss, M., Kaiser, T., Hummel, J., 2008. The morphophysiological adaptations of browsing and grazing mammals. In: Gordon, I.J., Prins, H.H.T. (Eds.), The ecology of browsing and grazing. Springer, Berlin, Germany, pp. 47–88. Google Scholar

16.

Codron, D., Codron, J., 2009. Reliability of delta C-13 and delta N-15 in faeces for reconstructing savanna herbivore diet. Mammalian Biology 74, 36–48. Google Scholar

17.

Codron, D., Sponheimer, M., Codron, J., Newton, I., Lanham, J.L., Clauss, M., 2012. The confounding effects of source isotopic heterogeneity on consumer-diet and tissue-tissue stable isotope relationships. Oecologia 169, 939–953. Google Scholar

18.

Cory, V.L., 1927. Activities of livestock on the range. Texas Agricultural Experimental Station Bulletin 367, 1–47. Google Scholar

19.

Craine, J.M., Towne, E.G., Tolleson, D., Nippert, J.B., 2013. Precipitation timing and grazer performance in a tallgrass prairie. Oikos 122, 191–198. Google Scholar

20.

Damiran, D., DelCurto, T., Findholt S.L., Johnson, B.K., Vavra, M., 2013. Comparison of bitecount and rumen evacuation techniques to estimate cattle diet quality. Rangeland Ecology & Management 66, 106–109. Google Scholar

21.

Deagle, B.E., Chiaradia, A., McInnes, J., Jarman, S.N., 2010. Pyrosequencing faecal DNA to determine diet of little penguins: is what goes in what comes out? Conservation Genetics 11, 2039–2048. Google Scholar

22.

DeNiro, M.J., Epstein, S., 1978. Carbon isotopic evidence for different feeding patterns in two hyrax species occupying the same habitat. Science 210, 906–908. Google Scholar

23.

Desbiez, A.L.J., Santos, S.A., 2014. New free tool to assess the use of forage resources in deer: DELTADIET. Deer Specialist Group Newsletter 26, 72–75. Google Scholar

24.

Dixon, J.S., 1934. A study of the life history and food habits of mule deer in California. Part II. Food habits. California Fish and Game 20, 316–354. Google Scholar

25.

Dixon, R., Coates, D., 2009. Review: Near infrared spectroscopy of faeces to evaluate the nutrition and physiology of herbivores. Journal of Near Infrared Spectroscopy 17, 1–31. Google Scholar

26.

Ehleringer, J.R., Rundel, P.W., Rundel, P.W., Ehleringer, J.R., Nagy, K.A., 1988. Stable isotopes: history, units, and instrumentation. Stable isotopes in ecological research. Springer, New York, pp. 1–15. Google Scholar

27.

Foley, W.J., McIlwee, A., Lawler, I., Aragones, L., Woolnough, A.P., Berding, N., 1998. Ecological applications of near infrared reflectance spectroscopy- a tool for rapid, costeffective prediction of the composition of plant and animal tissues and aspects of animal performance. Oecologia 116, 293–305. Google Scholar

28.

Forsyth, D.M., Davis, N.E., 2011. Diets of non-native deer in Australia estimated by macroscopic versus microhistological rumen analysis. Journal of Wildlife Management 75, 1488–1497. Google Scholar

29.

Freeman, S.N., Noble, D.G., Newson, S.E., Baillie, S.R., 2007. Modelling population changes using data from different surveys: The Common Birds Census and the Breeding Bird Survey. Bird Study 54, 61–72. Google Scholar

30.

van Geel, B., Aptroot, A., Baittinger, C., Birks, H.H., Bull, I.D., Cross, H.B., Evershed, R.P., Gravendeel, B., Kompanje, E.J.O., Kuperus, P., Mol, D., Nierop, K.G.J., Pals, J.P., Tikhonov, A.N., van Reenen, G., van Tienderen, P.H., 2008. The ecological implications of a Yakutian mammoth's last meal. Quaternary Research 69, 361–376. Google Scholar

31.

Glasser, T., Landau, S., Ungar, E.D., Perevolotsky, A., Dvash, L., Muklada, H., Kababya, D., Walker, J.W., 2008. A fecal near infrared reflectance spectroscopy-aided methodology to determine goat dietary composition in a Mediterranean shrub land. Journal of Animal Science 86, 1345–1356. Google Scholar

32.

Gustine, D.D., Barboza, P.S., Addison, J., Shively, R., Oliver, L., 2014. Isotopic nitrogen in fecal fiber as an indicator of winter diet in caribou and muskoxen. Rapid Communications in Mass Spectrometry 28, 625–634. Google Scholar

33.

Harris, D.J., 2003. Can you bank on GenBank? TREE 18, 317–319. Google Scholar

34.

Hibert, F., Taberlet, P., Chave, J., Scotti-Saintagne, C., Sabatier, D., Richard-Hansen, C., 2013. Unveiling the diet of elusive rainforest herbivores in next generation sequencing era? The tapir as a case study. PLoS One 8, e60799. Google Scholar

35.

Hobbs, N.T., 1996. Modification of ecosystems by ungulates. Journal of Wildlife Management 60, 695–713. Google Scholar

36.

Hofmann, R., Stewart, D., 1972. Grazer or browser: a classification based on the stomach structure and feeding habits if East Africa ruminants. Mammalia 36, 226–240. Google Scholar

37.

Holechek, J.L., Gross, B., 1982. Training needed for quantifying simulated diets from fragmented range plants. Journal of Range Management 35, 644–647. Google Scholar

38.

Holechek, J.L., Vavra, M., Pieper, R.D., 1982. Botanical composition determination of range herbivore diets: a review. Journal of Range Management 35, 309–315. Google Scholar

39.

Hoss, M., Kohn, M., Paabo, S., Knauer, F., Schroder, W., 1992. Excrement analysis by PCR. Nature 359, 199. Google Scholar

40.

Jean, P.-O., Bradley, R.L., Giroux, M.-A., Tremblay, J.-P., Côte, S.D., 2014. Near infrared spectroscopy and fecal chemistry as predictors of diet composition of white-tailed deer. Rangeland Ecology & Management 67, 154–159. Google Scholar

41.

Kartzinel, T.R., Chen, P.A., Coverdale, T.C., Erickson, D.L., Kress, W.J., Kuzmina, M.L., Rubenstein, D.I., Wang, W., Pringle, R.M., 2015. DNA metabarcoding illuminates dietary niche partitioning by African large herbivores. PNAS 112, 8019–8024. Google Scholar

42.

Kelly, J.F., 2000. Stable isotopes of carbon and nitrogen in the study of avian and mammalian trophic ecology. Canadian Journal of Zoology 78, 1–27. Google Scholar

43.

Kohn, M.H., Wayne, R.K., 1997. Facts from feces revisited. TREE 12, 223–228. Google Scholar

44.

Krebs, C.J., 1998. Niche measures and resource preferences. In: Fogarty, E., McDougal, V., Murray, N. (Eds.), Ecological methodology. Addison-Welsey, San Francisco, CA, USA, pp. 455–495. Google Scholar

45.

Leslie, D.M., Vavra, M., Starkey, E.E., Slater, R.C., 1983. Correcting for differential digestibility in microhistological analysis involving common coastal forages of the Pacific Northwest. Journal of Range Management 36, 730–732. Google Scholar

46.

Magurran, A.E., Baillie, S.R., Buckland, S.T., Dick, J.M., Elston, D.A., Scott, E.M., Smith, R.I., Somerfield, P.J., Watt, A.D., 2010. Long-term datasets in biodiversity research and monitoring: assessing change in ecological communities through time. TREE 25, 574–582. Google Scholar

47.

Mayes, R.W., Dove, H., 2000. Measurement of dietary nutrient intake in free-ranging mammalian herbivores. Nutrition Research Reviews 13, 107–138. Google Scholar

48.

Mayes, R.W., Lamb, C.S., Colgrove, P.M., 1986. The use of dosed and herbage n-alkanes as markers for the determination of herbage intake. The Journal of Agricultural Science 107, 161–170. Google Scholar

49.

McCracken, G.F., Westbrook, J.K., Brown, V.A., Eldridge, M., Federico, P., Kunz, T.H., 2012. Bats track and exploit changes in insect pest populations. PLoS One 7, e43839. Google Scholar

50.

Moll, R.J., Millspaugh, J.J., Beringer, J., Sartwell, J., He, Z., 2007. A new ‘view’ of ecology and conservation through animal-borne video systems. TREE 22, 660’668. Google Scholar

51.

Newmaster, S.G., Thompson, I.D., Steeves, R.A.D., Rodgers, A.R., Fazekas, A.J., Maloles, J.R., McMullin, R.T., Fryxell, J.M., 2013. Examination of two new technologies to assess the diet of woodland caribou: video recorders attached to collars and DNA barcoding. Canadian Journal of Forest Research 43, 897–900. Google Scholar

52.

Nielsen, J.M., Clare, E.L., Hayden, B., Brett, M.T., Kratina, P., 2017. Diet tracing in ecology: method comparison and selection. Methods in Ecology and Evolution  https://doi.org/10.1111/2041-210X.12869Google Scholar

53.

Norbury, G.L., 1988. Microscopic analysis of herbivore diets—a problem and a solution. Australian Wildlife Research 15, 51–57. Google Scholar

54.

Norris, K.H., Barnes, R.F., Moore, J.E., Shenk, J.S., 1976. Predicting forage quality by infrared reflectance spectroscopy. Journal of Animal Science 43, 889–897. Google Scholar

55.

Nuñez, M.A., Bailey, J.K., Schweitzer, J.A., 2010. Population, community and ecosystem effects of exotic herbivores: a growing global concern. Biological Invasions 12, 297–301. Google Scholar

56.

Pegard, A., Miquel, C., Valentini, A., Coissac, E., Bouvier, F., Francois, D., Taberlet, P., Engel, E., Pompanon, F., 2009. Universal DNA based methods for assessing the diet of grazing livestock and wildlife from faeces. Journal of Agricultural and Food Chemistry 57, 5700–5706. Google Scholar

57.

Pompanon, F., Deagle, B.E., Symondson, W.O., Brown, D.S., Jarman, S.N., Taberlet, P., 2012. Who is eating what: diet assessment using next generation sequencing. Molecular Ecology 21, 1931–1950. Google Scholar

58.

Raubenheimer, D., Simpson, S.J., 1997. Integrative models of nutrient balancing: application to insects and vertebrates. Nutrition Research Reviews 10, 151–179. Google Scholar

59.

Raye, G., Miquel, C., Coissac, E., Redjadj, C., Loison, A., Taberlet, P., 2011. New insights on diet variability revealed by DNA barcoding and high-throughput pyrosequencing: chamois diet in autumn as a case study. Ecological Research 26, 265–276. Google Scholar

60.

Scasta, J.D., Beck, J.L., Angwin, C.J., 2016. Meta-analysis of diet composition and potential conflict of wild horses with livestock and wild ungulates on western rangelands of North America. Rangeland Ecology & Management 69, 310–318. Google Scholar

61.

Schwartz, C.C., Ellis, J.E., 1981. Feeding ecology and niche separation in some native and domestic ungulates on the shortgrass prairie. Journal of Applied Ecology 18, 343–353. Google Scholar

62.

Shrestha, R., Wegge, P., 2006. Determining the composition of herbivore diets in the trans-Himalayan rangelands: a comparison of field methods. Rangeland Ecology & Management 59, 512–518. Google Scholar

63.

Snowder, G.D., Walker, J.W., Launchbaugh, K.L., Van Vleck, L.D., 2001. Genetic and phenotypic parameters for dietary selection of mountain big sagebrush (Artemisia tridentata Nutt ssp. vaseyana [Rydb] Beetle) in Rambouillet sheep. Journal of Animal Science 79, 486–492. Google Scholar

64.

Soininen, E.M., Valentini, A., Coissac, E., Miquel, C., Gielly, L., Brochmann, C., Brysting, A.K., Sonstebo, J.H., Ims, R.A., Yoccoz, N.G., Taberlet, P., 2009. Analysing diet of small herbivores: the efficiency of DNA barcoding coupled with high-throughput pyrosequencing for deciphering the composition of complex plant mixtures. Frontiers in Zoology 6, 16–25. Google Scholar

65.

Spalinger, D.E., Cooper, S.M., Martin, D.J., Shipley, L.A., 1997. Is social learning an important influence on foraging behavior in white-tailed deer? Journal of Wildlife Management 61, 611–621. Google Scholar

66.

Swinnen, K.R.R., Reijniers, J., Breno, M., Leirs, H., 2014. A novel method to reduce time investment when processing videos from camera trap studies. PLoS One 9, e98881. Google Scholar

67.

Taberlet, P., Coissac, E., Pompanon, F., Gielly, L., Miquel, C., Valentini, A., Vermat, T., Corthier, G., Brochmann, C., Willerslev, E., 2007. Power and limitations of the chloroplast trn L (UAA) intron for plant DNA barcoding. Nucleic Acids Research 35, e14. Google Scholar

68.

Thomas, A.C., Deagle, B.E., Eveson, J.P., Harsch, C.H., Trites, A.W., 2016. Quantitative DNA metabarcoding: improved estimates of species proportional biomass using correction factors derived from control material. Molecular Ecology Resources 16, 714–726. Google Scholar

69.

Thompson, I.D., Bakhtiari, M., Rodgers, A.R., Baker, A.J., Fryxell, J.M., Iwachewski, E., 2012. Application of a high-resolution animal-borne remote video camera with global positioning for wildlife study: observations on the secret lives of woodland caribou. Wildlife Society Bulletin 36, 365–370. Google Scholar

70.

Tieszen, L.L., Boutton, T.W., Tesdahl, K.G., Slade, N.A., 1983. Fractionation and turnover of stable carbon isotopes in animal tissues: Implications for 13C analysis of diet. Oecologia 57, 32–37. Google Scholar

71.

Timmons, G.R., Hewitt, D.G., DeYoung, C.A., Fulbright T.E., Draeger, D.E., 2010. Does supplemental feed increase selective foraging in a browsing ungulate? Journal of Wildlife Management 74, 995–1002. Google Scholar

72.

du Toit, J.T., Olff, H., 2014. Generalities in grazing and browsing ecology: using acrossguild comparisons to control contingencies. Oecologia 174, 1075–1083. Google Scholar

73.

Valentini, A., Pompanon, F., Taberlet, P., 2009. DNA barcoding for ecologists. TREE 24, 110–117. Google Scholar

74.

VanSomeren, L.L., Barboza, P.S., Gustine, D.D., Bret-Harte, M.S., 2017. Variation in δ15N and δ13C values of forages for Arctic caribou: effects of location, phenology and simulated digestion. Rapid Communication in Mass Spectrometry 31, 813–820. Google Scholar

75.

Vavra, M., Parks, C.G., Wisdom, M.J., 2007. Biodiversity, exotic plant species, and herbivory: the good, the bad, and the ungulate. Forest Ecology and Management 246, 66–72. Google Scholar

76.

Vogel, J.C., 1978. Isotopic assessment of the dietary habits of ungulates. South African Journal of Science 74, 298–301. Google Scholar

77.

Waldron, D.F., Taylor Jr. , C.A., Walker, J.W., Campbell, E.S., Lupton, C.J., Willingham, T.D., Landau, S.Y., 2009. Heritability of juniper consumption in goats. Journal of Animal Science 87, 491–495. Google Scholar

78.

Walker, J.W., Clark, D.H., McCoy, S.D., 1998. Fecal NIRS for predicting percent leafy spurge in diets. Journal of Range Management 51, 450–455. Google Scholar

79.

Walker, J.W., McCoy, S.D., Launchbaugh, K.L., Fralcer, M.J., Powell, J., 2002. Calibrating fecal NIRS equations for predicting botanical composition of diets. Journal of Range Management 55, 374–382. Google Scholar

80.

Walker, J.W., Waldron, D.F., Campbell, E.S., Taylor Jr. , C.A., Lupton, C.J., Landau, S., 2013. Intermediate periodicities in juniper consumption and sampling strategies to accurately estimate the diet of free-grazing goats. Rangeland Ecology & Management 66, 209–215. Google Scholar

81.

Wallmo, O.C., Neff, D.J., 1970. Direct observations of tamed deer to measure their consumption of natural forage. Range and wildlife habitat e valuation—a research symposium. US Department of Agriculture and Forest Service Miscellaneous Publication no. 1147, Washington, DC, USA PP. 105–110. Google Scholar

82.

Wam, K.H., Hjeljord, O., 2010. Moose summer diet from feces and field surveys: a comparative study. Rangeland Ecology & Management 63, 387–395. Google Scholar

83.

Welker, F., Duijma, E., van der Gaag, K.J., van Geel, B., de Knijff, P., van Leeuwen, J., Mol, D., van der Plicht, J., Raes, N., Reu mer, J., Gravendeel, B., 2014. Analysis of coprolites from the extinct mountain goat Myotragus balearicus. Quaternary Research 81, 106–116. Google Scholar

84.

Yu, X., Wang, J., Kays, R., Jansen, P.A., Wang, T., Huang, T., 2013. Automated identification of animal species in camera trap images. Journal on Image and Video Processing 52, 1–10. Google Scholar

Appendices

Appendix A. Supplementary data

Supplementary data to this article can be found online at  https://doi.org/10.1016/j.rama.2018.03.003.

© 2018 The Authors. Published by Elsevier Inc. on behalf of The Society for Range Management. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Sarah Garnick, Perry S. Barboza, and JohnW. Walker "Assessment of Animal-Based Methods Used for Estimating and Monitoring Rangeland Herbivore Diet Composition," Rangeland Ecology and Management 71(4), 449-457, (1 July 2018). https://doi.org/10.1016/j.rama.2018.03.003
Received: 10 November 2017; Accepted: 9 March 2018; Published: 1 July 2018
JOURNAL ARTICLE
9 PAGES


SHARE
ARTICLE IMPACT
Back to Top