• fluorescent dye;
  • grasshopper dispersal;
  • mark–recapture;
  • resight


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

1. Rangeland grasshopper movement was studied in Wyoming, USA, with respect to the biological and ecological factors (population density, developmental stage and weather) influencing net displacement and directionality.

2. A novel adaptation of the mark–recapture method was developed to monitor grasshopper dispersal. The method used fluorescent powder and resighting marked grasshoppers in the field with ultraviolet light, rather than physical recapturing of individuals.

3. Rangeland grasshoppers exhibited a strong tendency for directional movement. Adult grasshoppers demonstrated a significant tendency for dispersal in a north-westerly direction across a range of population densities (5–8, 10–15 and ≥ 18 grasshoppers m–2). Although not a definitive explanation, weather might have influenced this behaviour, as the grasshoppers consistently moved upwind.

4. The mean displacement of grasshoppers in a 36-h period ranged from 2·3 m in nymphs to 3·7 m in adults, with the distance of displacement being positively correlated with population density.

5. An understanding of grasshopper movement in terms of directionality and displacement has immediate applicability to reduced agent–area treatments for rangeland grasshopper management.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The rangeland grasshopper control method historically and currently used in North America is a ‘blanket’ treatment with broad-spectrum insecticides. However, this traditional method is costly, and available evidence shows that such large-scale insecticide applications have serious effects on beneficial and non-target organisms (Tepedino 1979; USDA 1987; Lockwood, Kemp & Onsager 1988). Unfortunately, other management methods (e.g. biological control and rotational grazing) are not mature enough in terms of application methods, product development and efficacy to be widely used in rangeland grasshopper programmes.

Although qualitatively different strategies are not available, quantitative modifications of chemical control have been sufficiently tested and refined to have immediate applicability (National Grasshopper Management Board 1998). One of the new, alternative control methods involves treating only a portion of the infested area with very low rates of insecticides in reduced agent–area treatments (RAAT; Lockwood & Schell 1997) instead of the traditional or standard blanketing of the whole infestation area at relatively high rates. In large part, this method apparently depends on grasshoppers in untreated areas moving into the treated areas and subsequently ingesting the insecticides. As such, the insecticide applied to the treated swaths actually works on both the grasshoppers in the immediate area and those that immigrate from the adjacent, untreated swaths. Thus, grasshopper movement is one of the key elements determining the efficacy of the RAAT method (Lockwood & Schell 1997), and this phenomenon is the focus of the present project.

A number of authors have studied movement of one or a few particular grasshopper species in experiments across a range of conditions (Riegert, Fuller & Putnam 1954; Clark 1962; Aikman & Hewitt 1972; Mason, Nichols & Hewitt 1995). These studies have indicated that the net displacement of grasshoppers varies as a function of species and development. The movement of grasshoppers was influenced by the microhabitat conditions, with these insects sometimes exhibiting site attachment, which made the movement more complex. Interpretation of these studies is difficult due to a number of methodological limitations. Normal movement was confounded by the effects of initial, agitation-based dispersal, overcrowding dispersal, and human disturbances during release and recapture of the grasshoppers. Furthermore, these works concentrated on only a few individual grasshopper species, some of them flightless (e.g. Podisma pedestris L.; Mason, Nichols & Hewitt 1995). In some cases, the sites of the studies were abnormally simplified, including a golf course (Aikman & Hewitt 1972) and bare ground (Riegert, Fuller & Putnam 1954), so extrapolation to the structural complexity of rangeland is tenuous. Finally, previous studies have not addressed the effect of density on grasshopper movement, and only one study (Riegert, Fuller & Putnam 1954) analysed dispersal as a function of development. Although these studies provide insight into specific aspects of grasshopper behaviour, the results are difficult to apply in a pest management context, where grasshoppers occur at high densities in complex rangeland habitats. Therefore, our goal was to investigate the potential effects of density and development on the distance and directionality of grasshopper movement in naturally occurring species’ complexes found in native, rangeland habitats.

Various mark–release–recapture/resight methods have been used in grasshopper dispersal studies. While some workers have used P-32 marking (Riegert, Fuller & Putnam 1954), others have employed non-radioactive materials in their studies (Aikman & Hewitt 1972; Joern 1982; Mason, Nichols & Hewitt 1995). Although radioactive marking is appropriate for individual and mass marking in long-term studies, non-radioactive marking methods are ideal for mass marking in short-term studies because of low cost and ease of application and detection (Gangwere, Chavin & Evans 1964).

Considerable success has been obtained using fluorescent dye to mark, and ultraviolet (UV) light to detect recaptured (marked) individual insects. The method is also advantageous because the insects do not have to be individually handled, marked or recovered. The first such studies were conducted with mosquitoes (Zukel 1945; Chang 1946; Pal 1947; Reeves, Brookman & Hammon 1948), tsetse flies (Jewell 1958) and aphids (Post & Anderson 1950). More recently, this approach has been employed in movement studies of parasitoid wasps (Corbett & Rosenheim 1996) and a wide range of Coleopteran pests of crops and forests (Gara 1967; Schmitz 1980; Linton et al. 1987; McMullen et al. 1988; Shore & McLean 1988; Salom & McLean 1989; Naranjo 1990; Oloumi-Sadeghi & Levine 1990). These studies have found little effect of marking on the behaviour of the insects. Most recently, Cook & Hain (1992) concluded that the marking had no effect on the flight initiation and semiochemical perception of bark beetles, and the marking remained intact on stored, dried beetles. Although the marking decreased the adult life span in this particular study, it was not of concern in the context of short-term monitoring. Thus, with this method, it is possible to monitor dispersal of highly mobile insects without affecting their normal behaviour.

Fluorescent dyes have not been used extensively to study grasshopper movement. This approach was considered for monitoring the dispersal of insecticides in locust control programmes (G. Bruge, personal communication). A single grasshopper (Melanoplus sp.), accidentally marked with fluorescent dye during a study of sawfly migration was recovered (Post & Anderson 1950), but intentionally marking grasshoppers with fluorescent dye previously has not been attempted. As such, we developed and implemented this method as a tool for tracking the movement of rangeland grasshoppers.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Mark–resight method development

Selection of marking agent

The nature of our dispersal studies required the development of an effective and efficient method for mass marking and recovering of grasshoppers while minimizing both initial, agitation-based dispersal and movement induced by human disturbance during the recapture phase. Based on examination of relevant literature associated with mark–release–recapture methods, we chose fluorescent ink (Gans Ink & Supply Co. Inc., Los Angeles, CA) and fluorescent powder (Magruder Color Co., Elizabeth, NJ) as the candidate marking agents. We tested the feasibility and effect of these substances on grasshoppers collected from native, mixed-grass prairie 8 km south-east of Hawk Springs, Wyoming (41°2′/104°15′) on 20 June 1996. The grasshopper species assemblage (20 grasshoppers m–2) was dominated by Melanoplus spp. [29%, mostly M. sanguinipes (F.) and M. infantilis Scudder], Opeia obscura (Thomas) (18%), Ageneotettix deorum (Scudder) (17%), Amphitornus coloradus (Thomas) (13%) and six other species (< 10% each). The 5th nymphal instar was the dominant developmental stage (48%), followed by the 4th instar (40%). A total of 530 grasshoppers was collected with a sweep net (diameter 35·6 cm) and divided into lots of 30 individuals, which were maintained in transparent plastic tubes (11·4 cm diameter, 48·3 cm long) with wire mesh endcaps. The insects were maintained at 20–26°C and 12 : 12 light : dark (LD) and fed grass and forbs clipped from the habitat.

After 24 h in the laboratory, the grasshoppers in each tube were immobilized by placing them at 4·4 °C for 5 min. For fluorescent powder marking, the grasshoppers were transferred into Ziplock® bags (946 ml) (S.C. Johnson & Son Inc., Racine, Wisconsin 53403, USA) with 0·1 g of fluorescent powder (orange–red, orange–yellow, magenta, green or chartreuse). The bags with the grasshoppers were shaken gently for 5 s to distribute the powder among the grasshoppers. Once the grasshoppers were marked, they were evenly divided into two tubes and provided with food. The tubes were returned to the aforementioned conditions and mortality was assessed after 24 h. For marking with fluorescent ink, the immobilized grasshoppers were gently scattered on a glass plate and the ink (4 g of blue, red, green, yellow or orange ink in 50 ml of petroleum ether) was applied with a pump sprayer. Ten lots of 15 grasshoppers were marked with the fluorescent ink and transferred to tubes as described above. Controls consisted of three lots of 32 grasshoppers processed similarly, except no marking agent was added.

Pooled across all colours, the average mortalities of grasshoppers were 39·3%, 40·7% and 46·9%, respectively, for those marked with fluorescent powder, ink and nothing (Table 1). Chi-square analysis (MSUSTAT; Lund 1986) revealed no significant difference in mortality among the grasshoppers marked with the fluorescent dye, ink and nothing (χ2 = 1·469, d.f. = 2, P > 0·05). During the experiment, six grasshoppers marked with powder or ink moulted successfully. Finally, different colours of fluorescent powder had no significant effect on grasshopper mortality (χ2 = 3·185, d.f. = 4, P > 0·05). Given the relative ease of marking grasshoppers with powder (marking with ink was cumbersome and required the use of a highly flammable and potentially toxic solvent), we decided to use the fluorescent powder as the marking agent.

Table 1.  Mortality of grasshoppers marked with the fluorescent powder and ink (duplicate trials of 15) vs. unmarked controls (triplicate trials of 32)
Marking agentColour% mortalityNumber deadNumber tested
ControlNot applicable474596
Mortality assessment of field protocol

We tested the durability and effect of fluorescent powder on grasshoppers collected on 18 July 1997 from Warm Springs, a native, mixed-grass prairie 15 km west of Guernsey, Wyoming (42°16′/104°50′; 1460 m elevation). Population density at the time was 21 grasshoppers m–2, and the species assemblage was dominated by M. sanguinipes (31%), Ageneotettix deorum (19%), Aulocara elliotti Thomas (16%), and another seven species (< 10% each). The dominant developmental stage was adult (58%), followed by the 5th nymphal instar (16%) and 4th (10%). Grasshoppers were collected with a sweep net and immobilized, counted (250 individuals), marked, and maintained as described previously. Three sets of 50 grasshoppers were marked with 0·1 g of orange–red fluorescent powder. The controls consisted of handling and maintaining 100 grasshoppers (two sets of 50) in the manner previously described, except no dye was added to the bags. After 24 h, the mortality of marked grasshoppers (10·7%) was not significantly different from that of control (10·0%) (χ2 = 0·029, d.f. = 1, P > 0·05) (MSUSTAT; Lund 1986). The lower mortality in the control grasshoppers related to our previous bioassay reflected the shorter duration during which grasshoppers were caged.

Wash-off test

Given the possibility of summer thunderstorms on the prairie, we assessed the efficacy of marking grasshoppers exposed to moisture. We tested the durability of fluorescent powder marking using grasshoppers collected from the previously described Warm Springs site on 19 July 1997. Two lots of 25 grasshoppers were marked with orange–red fluorescent powder as previously described. The insects were left in the bags at 25 °C for 5 min, after which one bag was filled with water, allowed to sit for 1 min and then shaken vigorously for 30 s. The other bag received no water treatment. Visual observation immediately after flushing and subsequent examination of the dried specimens under UV light (2805 DC Collecting Light, 12 V DC operation, powered with a rechargeable battery CFM12 V12.5LPP-S1, EagleEpicher; both from BioQuip Products Inc., Gardena, CA 90248, USA) indicated that all individuals in both bags retained their markings. During observation, protective glasses (Astrospec 3000, UveX Safety Inc., Smithfield, Rhode Island 02917, USA) were used to prevent ocular injury by UV light.

Cross-contamination assessment

As rangeland grasshoppers can have physical contact between individuals, we conducted a cross-contamination assessment to see if unmarked individuals could become marked via body contact with marked individuals. We collected 25 grasshoppers from the Warm Spring site on 19 July 1997, marked them with orange–red fluorescent powder, and placed them into a plastic tube provided with the food. The insects were maintained at 20–26 °C and 12 : 12 LD for 2 h after marking. Then they were transferred into a Ziplock® bag with 25 newly collected, unmarked individuals for 60 min and killed by freezing. This process forced the grasshoppers to have relatively extensive and close physical contact with each other. The numbers of the marked and unmarked grasshoppers were checked under UV light, which showed that the marked grasshoppers were readily distinguishable from those unmarked. Two unmarked grasshoppers had small flecks of dye on the head and thorax, but the visual effect was distinct from the extensive marking on the originally marked individuals. It was unlikely that the cross-contaminated grasshoppers would even have been detected in the field, let alone mistaken for marked individuals. In a subsequent experiment at Pollet Ranch (see below) on 31 July 1997, we observed a marked M. packardii Scudder female mating with an unmarked male without the latter becoming marked. Therefore, it appeared that body contact would not cause a significant amount of false-positive resightings.

Field test of efficacy and efficiency

In this experiment, 500 nymphs and adult grasshoppers were collected with sweep nets, on 12 August 1996, at a site 42 km west of Guernsey, Wyoming, on the Pollet Ranch (42°16′/105°11′; 1585 m elevation). The vegetation was rather uniform and typical of mixed-grass prairie, being mainly buffalograss Buchloe dactyloides, downy brome Bromus tectorum, needle-and-thread Stipa comata, prairie junegrass Keoleria pyramidata, threeawn Aristida spp., western wheatgrass Agropyron smithii, fringed sagebrush Artemisia frigida, and broom snakeweed Gutierrezia sarothrae. For plant taxonomy, see Craighead, Craighead & Davis (1963), Johnson & Nichols (1982) and Whitson et al. (1991). The plant cover was estimated at 50%, and canopy height was 25 cm. The topography of the site was variable but rather flat, with a range of slopes (0–5%). The density of the grasshoppers at the site was 27 grasshoppers m–2 and the species assemblage was dominated by M. infantilis (33%), followed by Phoetaliotes nebrascensis (Thomas) (25%), Opeia obscura (Thomas) (19%) and 12 other species (< 10 each). The grasshoppers were marked in the field with fluorescent powder using the previously described procedure.

Five sets of 100 grasshoppers were marked with each colour (orange–red, orange–yellow, magenta, green or chartreuse). Approximately 50 m from the collection site, an area was divided into four equal quadrants using cotton string (Fig. 1). Virtually all grasshoppers within the centre release circle (radius 3 m) were collected with a net immediately before release of the marked individuals. It took four people 1·5 h to collect, immobilize, count, mark and release the grasshoppers.


Figure 1. The release circle (A), quadrants marked with string (B) and recovery area, with search rings at 1-m increments (C), on Wyoming rangeland.

Download figure to PowerPoint

Marked grasshoppers were uniformly released into the centre circle at 16.30 h to re-establish the original population density. At the time of release, the temperature was 20 °C and the wind was from the west at 4·2 m s–1. Marked grasshoppers were resighted at 22.30 h (6 h after release) using a UV light. The resighting took two people 80 min. The colour and location of each resighted grasshopper was recorded. As it was cool (16 °C) and dark, grasshopper movement due to disturbance was minimal. Fewer than 5% of the resighted subjects moved to escape the observer, and the distance of movement was < 5 cm. The next morning, at 08.40 h (about 16 h after the release) we visually searched for marked grasshoppers. Two people each searched for 80 min, recording colours and locations without handling the grasshoppers directly. Finally, at 09.40 h both people swept the recapture area for 80 min. The collection began at the centre of the circle and progressed at 1-m increments until 12 m from the centre of the release area. Thus, in all cases the area searched and the time of search were constant.

The results showed that 64% (321 of 500) of the marked grasshoppers were resighted at night (Table 2) via the UV light, while only 28% (140 of 500) were relocated with diurnal, visual searching. Just 9% (45 of 500) were recaptured with the sweep net. This experiment demonstrated 2·3- to 7·1-times higher recovery rates of the marked grasshoppers in nocturnal resighting relative to diurnal searching or recapturing. The latter methods were also associated with extensive escape movements by the grasshoppers. Therefore, we decided to adopt the nocturnal, resighting method.

Table 2.  Recovery rates (% resighted or recaptured) of 500 fluorescent powder-marked rangeland grasshoppers via selected methods in the four quadrants surrounding the release site in south-eastern Wyoming

Directional movement was evident, with 82% (265 of 321) of the marked grasshoppers occurring west of the release area and only 18% (56 of 321) occurring east of it, after 6 h. Chi-square analysis (MSUSTAT; Lund 1986) confirmed that the grasshoppers moved into the north-west and south-west quadrants with a frequency significantly greater than chance (χ2 = 136·1, d.f. = 1, P < 0·05).

Dispersal study

Study sites

In 1997, two sites in south-eastern Wyoming (Pollet Ranch and Warm Springs) were chosen because they were accessible and supported a range of grasshopper population densities. The vegetation and topography of Warm Springs were similar to those of the Pollet Ranch. During the study, temperature, relative humidity, and wind speed and direction were recorded 1 m above the ground on site. The wind speed within the canopy was probably reduced by vegetation drag, but our instruments were not designed for such movements.

Grasshoppers were represented by three subfamilies: Melanoplinae, Gomphocerinae and Oedipodinae (respectively accounting for 34%, 38% and 28% of the assemblages). The sites had similar acridids, with 52% of the species occurring at both sites (Table 3). Population densities of 5–50 grasshoppers m–2 were found within the sites.

Table 3.  Number of individuals of grasshopper species and developmental stages (per 100 sweeps) at study sites in south-eastern Wyoming
Nymphal instar
Warm Springs17 JuneAgeneotettix deorum (Scudder) Amphitornus coloradus (Thomas)31 4910 30  
  Aeropedellus clavatus (Thomas) Arphia pseudonietana (Thomas) Aulocara elliotti Thomas Camnula pellucida (Scudder) Cordillacris crenulata (Bruner) Cordillacris occipitalis (Thomas)3 1 34 122
  Eritettix simplex (Scudder) Hesperotettix viridis (Thomas) Melanoplus bivittatus (Say)3 4111
  Melanoplus confusus Scudder Melanoplus occidentalis (Thomas) Melanoplus packardii Scudder Melanoplus sanguinipes (F.) Melanoplus spp. Opeia obscura (Thomas)2 1 2 1 85 201 22 15
  Psoloessa delicatula (Scudder)   1
Pollet Ranch1 JulyAgeneotettix deorum Amphitornus coloradus6 116 313 
  Aeropedellus clavatus Aulocara elliotti Thomas Eritettix simplex (Scudder) Hadrotettix trifasciatus (Say) Hesperotettix viridis5 11 21 2 511
  Melanoplus infantilis Scudder Melanoplus keeleri (Thomas)18 628292
  Melanoplus occidentalis Melanoplus packardii Melanoplus sanguinipes Opeia obscura (Thomas) Phlibostroma quadrimaculatum (Thomas) Phoetaliotes nebrascensis (Thomas) Spharagemon equale (Say) Trachyrhachys kiowa (Thomas) Xanthippus corallipes (Haldeman)14 2 7 8 29 1 1 14 5 1 142
Warm Springs17Ageneotettix deorum Amphitornus coloradus  716
  Aulocara elliotti Thomas   13
  Cordillacris occipitalis Melanoplus bivittatus32 1
  Melanoplus sanguinipes Mermiria bivittata (Serville) Opeia obscura (Thomas) Phlibostroma quadimaculatum Phoetaliotes nebrascensis1 1 8 13 2 15 118
Pollet Ranch31 JulyAgeneotettix deorum 2546
  Amphitornus coloradus  33
  Arphia conspersa Scudder35144
  Aulocara elliotti Hadrotettix trifasciatus  31
  Hesperotettix viridis Melanoplus angustipennis (Dodge)  6 14
  Melanoplus bivittatus36141
  Melanoplus infantilis Scudder Melanoplus occidentalis133210 414
  Melanoplus packardii2623
  Melanoplus sanguinipes54910
  Metator pardalinus (Sausure)   5
  Opeia obscura 23235
  Phlibostroma quadimaculatum 243
  Phoetaliotes nebrascensis Psoloessa delicatula72718 111
  Trachyrhachys kiowa 1121

Forty 0·1-m2 aluminium rings per site were used to assess densities after preliminary visual estimates (Onsager & Henry 1977). Habitats with grasshopper densities of 5–8, 10–15 and ≥ 18 grasshoppers m–2 were located on 17 June (75% of the grasshoppers were 1–3rd instar nymphs at Warm Springs), 1 July (49% of the grasshoppers were 4–5th instar nymphs at Pollet Ranch), and 17 and 31 July (40–58% of the grasshoppers were early and late adults at Pollet Ranch and Warm Springs). These four dates comprised consecutive developmental assemblages representing stages typically treated in rangeland grasshopper control programmes (DeBrey, Brewer & Lockwood 1993). On each of the sampling dates at each site, the mark–release–resight method was used at two, replicated plots within each of the three densities, with the replications being at least 100 m apart. At each plot and date, grasshoppers were collected with a sweep net from the study area (> 50 m from the release circle), and immobilized by 5 min of cooling at 4·4°C. Sets of 50 grasshoppers were marked with fluorescent powder (orange–red), as previously described. The number of marked and released grasshoppers was sufficient to re-establish the original population density in the centre release circle. The release area was similar to that of the field test previously described, but the quadrants were orientated so the centre line of each corresponded to a cardinal direction. The removal of grasshoppers from the release circle and their replacement with marked individuals was completed between 08.00 h and 11.00 h in each trial. After release, 100 sweeps (50 low-slow and 50 high-fast) were taken 50 m from the release site to analyse the grasshopper species and developmental complex. During the 24 trials, 9120 grasshoppers were marked and released.

At 12 and 36 h after the release of grasshoppers, individuals were resighted at night using an UV light. The resighting began within the release circle and then moved outward at 1-m increments around the circle until no grasshoppers were found at a given distance in all four quadrants. The location of each resighted grasshopper was recorded with respect to direction (quadrant) and distance from the centre of the release circle. It took two people 30–45 min to resight the marked grasshoppers in each trial.

Analysis of directionality

The resighted grasshoppers were pooled over all distances from the release centre within a quadrant. The percentage of the total number of resighted grasshoppers found in each quadrant was normalized using an arcsine transformation. We considered sites, plots, quadrants, densities and developmental assemblages as the sources of variation, and conducted completely randomized factorial analysis with a fixed model (anova; SAS Institute 1988; Bennington & Thayne 1994) to assess the differences in the proportions of the grasshoppers in the quadrants as a function of these factors at 36 h after release. Orthogonal contrasts were conducted to make comparisons for particular sets of factors in the analysis. In all analyses, differences were considered to be significant at P < 0·05.

Analysis of distance

Net dispersal of grasshoppers at 12 and 36 h after release was evaluated as a function of developmental assemblages, densities, and sites using completely randomized factorial analysis (anova; SAS Institute 1988) independently. Grasshoppers relocated within the release circle were assumed to have moved 1·5 m (one-half of the radius), and those resighted outside the circle were assumed to have moved from the centre point of the release circle (i.e. 1·5 m plus each 1-m increment). The net dispersal of grasshoppers was estimated by averaging the sum of the distances that grasshoppers moved by the number of grasshoppers resighted in the trial. A square root transformation of the net dispersal was used to control the variance. Fisher's protected least significant difference test (LSD) was used for mean separation. Differences were considered to be significant at P < 0·05.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References


Winds were generally from the north-west at ≤ 3·0 m s–1 (maximum 6·7 m s–1). Diurnal temperatures ranged from 13 to 37 °C, and relative humidity ranged from 18% to 65%. The only rain was a 30-min shower on 18 July (Table 4).

Table 4.  Weather conditions during dispersal experiments on south-eastern Wyoming rangeland
Time (h post-release)Datem s–1DirectionHumidity (%)Temperature (°C)
017 June1·3–3·1NW4922·4
 1 July5·2–6·7NW1824·8
 17 July0·0–2·2NW2137·1
 31 July0·0–2·2NW4227·6
1217 June0·05718·6
 1 July1·5–3·3NW5112·7
 17 July0·0–2·7SW2526·4
 31 July0·0–0·9NW6518·8
3617 June0·0–1·3NE5018·8
 1 July0·0–2·7NE3416·6
 17 July2·0–3·0N5522·1
 31 July0·0–1·9NW6122·7

Recovery rate

Of the marked grasshoppers, 45·9% (4186 of 9120) were resighted 12 h after release and 21·3% (1942 of 9120) were resighted 36 h after release (Table 5). As expected, the marked grasshoppers were readily found in the dark with UV light, and they did not flush from their positions.

Table 5.  The recovery rates [% (number marked)] of marked grasshoppers at different sites and dates as a function of population density
Time after release (h)
SiteDevelopmental stageDensity (no. m–2)1236
Warm Springs1st–3rd instar5–848·4 (500)34·8 (500)
  10–1562·3 (700)33·9 (700)
  > 1853·6 (1200)30·5 (1200)
  All55·0 (2400)32·3 (2400)
Pollet Ranch4th–5th instar5–865·0 (320)30·0 (320)
  10–1543·4 (700)19·1 (700)
  > 1824·8 (1200)11·7 (1200)
  All36·5 (2220)16·7 (2220)
Warm SpringsEarly adults5–833·8 (400)12·0 (400)
  10–1534·1 (700)10·0 (700)
  > 1848·3 (1200)17·2 (1200)
  All41·5 (2300)14·1 (2300)
Pollet RanchLate adults5–839·5 (400)17·3 (400)
  10–1551·2 (600)15·5 (600)
  > 1853·0 (1200)5·8 (1200)
  All50·0 (2200)21·5 (2200)

Directionality of movement

The experimental sites (Pollet Ranch and Warm Spring) had no effect on the pattern of movement and distance (P > 0·05; Table 6 and 7). Because a significant difference was apparent from the anova (the interaction of density and development with respect to the direction of movement was significant), we conducted orthogonal contrasts. These analyses showed that the proportion of early and late adult grasshoppers resighted in the north and west quadrants was significantly greater than that in the south and east quadrants at all population densities. The directionality of early adult grasshoppers at the medium density was the only exception to this pattern. At high density, 4th and 5th instar nymphs also showed significant directional (north-westerly) movement, but all other comparisons of nymphal distribution were not significant (Fig. 2). Thus, it appears that directionality of movement is virtually absent in grasshopper early nymphs, but this behaviour is clearly manifested in late instar nymphs at high density and adults at all densities.

Table 6.  Analysis of directional movement of rangeland grasshoppers (36 h after release) in south-eastern Wyoming, with the sources of variation
Sourced.f.Sum of squaresMean squareF -valueP > F
Development stage30·002380·000790·110·9557
Developmental stage × quadrant90·123120·013681·840·0858
Developmental stage × density60·000420·000070·011·0000
Quadrant × density60·091530·015262·050·0773
Developmental stage × quadrant × density180·309190·017182·310·0113
Corrected total951·05804   
Table 7.  Analysis of distance of rangeland grasshopper movement in south-eastern Wyoming, with the sources of variation
Time (h post-release)Sourced.f.Sum of squaresMean squareF -valueP > F
12Developmental stage30·915010·3050017·710·0003
 Developmental stage × density Error Corrected total6 10 230·06427 0·17224 1·200930·01071 0·017220·620·7101
36Developmental stage31·188210·3960716·170·0004
 Developmental stage × density Error Corrected total6 10 230·10160 0·24490 1·929330·01693 0·024490·690·6624

Figure 2. Proportion of the marked grasshoppers located in quadrants (north, east, □ south and ▪ west) as a function of the population density and development. (a) 1st–3rd instars; (b) 4th–5th; (c) early adults; (d) late adults.

Download figure to PowerPoint

Net distances of movement

Both developmental assemblages and densities significantly affected the net distance of grasshopper movement (Table 7) (density had no effect on dispersal at 12 h after release). Fisher's LSD test showed that adult grasshoppers moved significantly further than nymphs, while late instar nymphs moved further than early instars (Table 8). Across all densities and sites, adults moved an average of 3·0–3·7 m (12 and 36 h after release, respectively), while nymphs moved 1·8–2·3 m. Across all developmental stages and sites, grasshoppers at high densities moved an average of 2·6–3·6 m (12 and 36 h after release, respectively), while those at medium densities moved 2·5–2·8 m, and grasshoppers in low density populations moved 2·3–2·5 m. During our resampling protocol we did not locate any marked grasshoppers at > 30 m from the centre, but we saw a few marked grasshoppers > 50 m away from the centre area while conducting other studies 36 h after release.

Table 8.  The mean distance of movement (m) of rangeland grasshoppers in south-eastern Wyoming, at selected times after release
Time post-release (h)
SiteDevelopmental stageDensity (number m–2)1236
  1. Means within a column followed by different letters differ significantly according to Fishers’ LSD test.

Warm Springs1st–3rd instar5–81·62a1·82a
  > 181·85a2·39b
Pollet Ranch4th–5th instar5–82·04a2·26b
  > 181·93a2·88c
Warm SpringsEarly adults5–82·53b2·84c
  > 183·36b3·84d
Pollet RanchLate adults5–82·91b3·02c
  > 183·13b5·22f


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Assessment of methodology

The primary drawbacks of the traditional mark–release–recapture approach result from disrupting the animals’ behaviour by the four procedures (capture, mark, release, recapture) and from the low rates of recapturing marked individuals. While our method requires initial capture, marking and release, it represents a significant improvement with respect to the recapture phase. Although fluorescent tracer has been used to assess contact of grasshoppers with sprayed biopesticide (Thomas, Langewald & Wood 1996), the use of such dyes in a mark–recapture (or in our case recapture via resighting of marked grasshoppers using UV light in the field) has not been attempted previously. Nocturnal resighting of grasshoppers marked with fluorescent dyes using UV light is possible at 2 m from the marked individuals. Therefore, we did not have to handle or otherwise disturb the grasshoppers during recapture (resighting). Furthermore, because the grasshoppers were quiescent in cool, dark conditions, we were able to reduce greatly the disturbance due to our presence, and thereby overcome many shortcomings of a classical recapture approach using diurnal sighting, sweep-netting or other disruptive methods. Therefore, the method appears to have a wide range of applications for behavioural studies of grasshoppers.

The release of individuals within a central area may be problematical in terms of initial, agitation-based dispersal, overcrowding dispersal, and human disturbance during release (Aikman & Hewitt 1972). However, our approach mitigates these shortcomings because the grasshoppers in our study were immobilized by cooling before release. Thus their initial movements following release in the field were gradual and agitation-based dispersal was avoided. In addition, our initial removal of grasshoppers from the recovery area and the release of marked grasshoppers in numbers to re-establish the original density prevented overcrowding dispersal.

Turchin, Odendaal & Rausher (1991) mentioned the low recovery rate of marked insects during diurnal searching as one of the major pitfalls of the mark–recapture methodology. The resighting method used here significantly increased the recovery rate compared to traditional diurnal resighting. Given that we stopped searching whenever we failed to find a marked grasshopper in four consecutive quadrants, the efficacy of resighting could have been higher than our reported rate. The resighting rate of 46% in our overall large-scale field experiment was nearly twice as great as that of diurnal searching for marked grasshoppers and five times greater than that of recapturing marked grasshoppers with a sweep net. Based on our cage studies, about 10% of marked grasshoppers died within 12 h of the release in the field and were probably removed by scavengers. Presumably, another 45% of marked grasshoppers were not resighted because they were hidden within the area of inspection, moved out of the area, or were eaten by predators.

Turchin, Odendaal & Rausher (1991) also note that the interpretation of mark–recapture results can be difficult, in that the portion of the population that was not detected may be behaving in a different way from that portion that was recaptured. This concern is partially assuaged by recapturing a large portion of the marked population, such that conclusions are more applicable to the entire population, and our method dramatically improves the rate of recapture (resighting). However, the possibility that the ‘missing’ marked individuals differ in their behaviour is a matter that warrants consideration. Perhaps grasshopper movement is a non-linear phenomenon, in which most individuals move a relatively short distance, but a portion of the population moves a very long distance (i.e. outside of the range of our resighting procedure). While our cage studies indicated that the rate of moulting in 24 h was low (2%) and probably did not account for a significant number of unrecovered grasshoppers, the relatively cool laboratory conditions may have slowed development and led to an underestimation of moulting frequency. As such, it seems likely that a portion of the grasshoppers moved out of the search area, so that our measurements of movement probably underestimated the actual displacement. However, the flight behaviour of adult rangeland grasshoppers, except to escape a disturbance, was rare based on our field observations.

Our wash-off test showed that water would not affect the marking of grasshoppers, so the method is robust in case of heavy dew or rain. In addition, a marked M. packardii female was observed mating with an unmarked male, and exuviae of marked grasshoppers were seen occasionally. Thus, it appears the behaviour and biology of marked grasshoppers were not substantially altered.

Marking did not increase grasshopper mortality, although other studies have shown it can slightly increase mortality. Cook & Hain's (1992) study of adult bark beetles [(Dendroctonus frontalis Zimmermann and Ips grandicollis (Eichhof)] indicated that marking decreased the life span of these insects; however, it had no significant effect on flight initiation and semiochemical perception. Other studies specifically addressing the potential impact of marking beetles (Dendroctonus ponderosae Hopkins) with fluorescent dyes have found no adverse effect on adult mortality and flight (Linton et al. 1987; McMullen et al. 1988). All of these studies concluded that the results should not be affected greatly if resighting is done shortly after release.

Directionality of movement

The data suggest that rangeland grasshoppers have a strong tendency for directional movement, as a function of development and density. Adult grasshoppers had a strong tendency to move north-westerly. The prevailing winds might have been an important factor in determining the direction of grasshopper movement, as grasshopper movement was consistently upwind. Although wind direction and velocity varied somewhat, the greatest variation was at night, during which movement was minimal based on our observations. During the field trial associated with development of our method, the grasshoppers showed a strong westerly movement, also corresponding to the prevailing winds. It should be noted that adult flight was rarely observed during our trials, although most of the species were macropterous. The general, upwind movement of adult grasshoppers may have been a search for resources perceived by olfactory cues (e.g. host-plants, mates or oviposition sites). However, stronger winds may have hindered directional movement, as no discernible trend to the north-west was observed when winds reached 6·7 m s–1 (Pollet Ranch, 1 July). A decrease in grasshopper movement during strong winds has been noted previously (Munro & Telford 1942; Riegert, Fuller & Putnam 1954).

Although there is no uniform trend across all grasshopper species and developmental stages, wind direction and velocity have been reported previously to influence grasshopper movement. Late instar (4th and 5th) nymphs of Camnula pellucida and M. sanguinipes moved against wind, while early instar (2nd) nymphs and adults moved with the wind (Riegert, Fuller & Putnam 1954). Their study suggested further that early instar nymphs and adults showed no ability to orientate themselves towards a food supply. However, it should be noted that these results followed release of grasshoppers on bare ground, instead of in their normal habitat. Upwind movement has been documented in other insect groups (Corbett & Rosenheim 1996) so this behaviour may be a common phenomenon.

Our study indicates that nymphal grasshoppers showed little or no tendency for directional movement, but this finding may have been a function of the scale of our resighting survey in the context of the relatively small size, and consequently limited mobility, of nymphs. Because they were not able to move as far as adult grasshoppers, directionality may not have been manifested within the spatiotemporal limits of our experiment. Furthermore, nymphs need less food per unit time, so there is less necessity for them to seek host-plants. This explanation is supported by the observation that nymphs at high densities exhibited directional movement, possibly because at such densities they become resource limited and must seek food over larger distances.

Net distance of movement

The net distance of dispersal was positively related to grasshopper development and population density. As noted previously, the increase of movement with development and population density might have arisen from the greater competition for resources (host-plants, mates and oviposition sites). Lockwood's (1988) study of Aulocara elliotti indicated that intra- and interspecific non-sexual aggression resulted from the defence of territories based on limited, suitable microhabitats. The resulting avoidance movements might be more pronounced in high density grasshopper assemblages.

Density-dependent dispersal has been found in the grasshopper Myrmeleotettix maculatus (Thumb), but the authors questioned whether density accounted for this behaviour (Aikman & Hewitt 1972). Although Joern (1982) recognized that the pattern of microhabitat use by assemblages of grasshoppers was influenced by biotic interactions among species, intraspecific competition was not entirely responsible for microhabitat choice. However, our experiment clearly demonstrated density-dependent dispersal, with grasshoppers at high population densities moving significantly further than those at medium and low densities. This suggests that intra- and interspecific interactions within a grasshopper assemblage (perhaps involving both direct and indirect interactions mediated through a common resource) can influence the movement of these insects.

Early and late adult grasshoppers moved significantly further than nymphs. Furthermore, the late adults moved significantly more than early adults (36 h after release), but early and late instar nymphs moved similar distances (both 12 and 36 h after release). These differences may reflect that situation in which older adults have less suitable food (due to depletion and senescence of host-plants), as well as having the need to find suitable oviposition sites.

As noted in previous studies (Clark 1962; Aikman & Hewitt 1972; Mason, Nichols & Hewitt 1995), extrapolating grasshopper dispersal from short-term observations results in an overestimation of mobility. In our experiments, the net distances at 36 h after the release were only 1·1- to 1·7-times those observed at 12 h. This inability to extrapolate has been variously explained by site attachment and disturbances at the time of release.

Applications to pest management

The pattern and rate of dispersal in grasshoppers are fundamental to the success of RAAT programmes for grasshopper pest management (Lockwood & Schell 1997). Our study of dispersal indicates that the efficacy of a RAAT application may be markedly enhanced compared with the results expected if grasshopper movement was very limited. Moreover, grasshoppers can be expected to move from untreated swaths into treated areas more frequently than if movement was non-directional (random). Therefore, in principle an insecticide with a 14-day residual (e.g. carbaryl) could be applied with swath spacing of approximately 35 m, and we have found that a spacing of 30 m (with 30-m untreated swaths) provides excellent control (Lockwood & Schell 1997).

The RAAT approach may be further enhanced by applying insecticides in a manner consistent with grasshopper behaviour. That is, the spaces between treated swaths can be adjusted to the target population's density and development. Extrapolating from our study, an infestation of late instar nymphs at 10–15 m–2 would move only half as far as > 18 adults m–2, requiring a commensurate adaptation of swath spacing. In addition, assessment of weather conditions during a treatment (specifically wind velocity and direction) might be used to help to orientate the treatment swaths perpendicular to the direction of grasshopper movement and thereby to enhance the potential efficacy of a RAAT programme. In this context, it is particularly fortunate that the prevailing mid-continental winds are from the west, while insecticide swaths are typically orientated north–south as aerial applicators avoid flying into the rising or setting sun.

Our future research will focus on refining field experiments to understand better the factors and parameters describing movement of the pestiferous acridid species so we can confidently apply this information to rangeland grasshopper control programmes. In particular, we are attempting to isolate the mechanisms of directional movement and to predict effectively the net displacement of rangeland grasshoppers, so that RAAT programmes can be tailored to local conditions. This work will provide us with important insights regarding optimal spacing and orientation of treated swaths in the RAAT strategy.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank E. Norelius, R. Rockwell, S. Schell and K. VanDyke for their help with field work. Special thanks go to D. Legg for his help in statistical analysis of the data.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Aikman, D. & Hewitt, G.M. (1972) An experimental investigation of the rate and form of dispersal in grasshoppers. Journal of Applied Ecology, 9, 809817.
  • Bennington, C.C. & Thayne, W.V. (1994) Use and misuse of mixed model analysis of variance in ecological studies. Ecology, 75, 717722.
  • Chang, H.T. (1946) Studies on the use of fluorescent dyes for marking Anopheles quadrimaculatus Say. Mosquito News, 6, 122125.
  • Clark, D.P. (1962) An analysis of dispersal and movement in Phaulacridium vittatum (Sjöst.) (Acrididae). Australian Journal of Zoology, 10, 382399.
  • Cook, S.P. & Hain, F.P. (1992) The influence of self-marking with fluorescent powders on adult bark beetles (Coleoptera: Scolytidae). Journal of Entomological Science, 27, 269279.
  • Corbett, A. & Rosenheim, H.A. (1996) Quantifying movement of a minute parasitoid, Anagrus epos (Hymenoptera: Mymaridae), using fluorescent dust marking and recapture. Biological Control, 6, 3544.
  • Craighead, J.J., Craighead, F.C. & Davis, R.J. (1963) A Field Guide to Rocky Mountain Wildflowers. Houghton Mifflin, Boston, MA.
  • DeBrey, L.D., Brewer, M.J. & Lockwood, J.A. (1993) Rangeland Grasshopper Management. University of Wyoming Agricultural Experiment Station Bulletin B-980, University of Wyoming Agricultural Experiment Station, Laramie, WY.
  • Gangwere, S.K., Chavin, W. & Evans, F.C. (1964) Methods of marking insects, with especial reference to Orthoptera (sens. Lat). Annals of the Entomological Society of America, 57, 662669.
  • Gara, R.I. (1967) Studies on the attack behavior of the southern pine beetle. I. The spreading and collapse of outbreaks. Contributions from Boyce Thompson Institute, 23, 349354.
  • Jewell, G.R. (1958) Detection of tsetse fly at night. Nature, 181, 13541355.
  • Joern, A. (1982) Vegetation structure and microhabitat selection in grasshoppers (Orthoptera, Acrididae). Southwestern Naturalist, 27, 197209.
  • Johnson, J.R. & Nichols, J.T. (1982) Plants of South Dakota Grasslands: A Photographic Study. Bulletin 566. Agricultural Experiment Station, South Dakota State University, Brookings, SD.
  • Linton, D.A., Safranyik, L., McMullen, L.H. & Betts, R. (1987) Field techniques for rearing and marking mountain pine beetle for use in dispersal studies. Journal of the Entomological Society of British Columbia, 84, 5357.
  • Lockwood, J.A. (1988) Nonsexual interactions in the big-headed grasshopper Aulocara elliotti (Thomas) (Orthoptera: Acrididae). Journal of Entomological Science, 23, 383393.
  • Lockwood, J.A. & Schell, S.P. (1997) Decreasing economic and environmental costs through reduced area and agent insecticide treatments (RAATs) for the control of rangeland grasshoppers: empirical results and their implications for pest management. Journal of Orthopterist Research, 6, 1932.
  • Lockwood, J.A., Kemp, W.P. & Onsager, J.A. (1988) Long-term, large-scale effects of insecticidal control of rangeland grasshopper populations (Orthoptera: Acrididae). Journal of Economic Entomology, 81, 12581264.
  • Lund, R.E. (1986) MSUSTAT Statistical Analysis Package, Version 3.2. Bozeman, MT.
  • McMullen, L.H., Safranyik, L., Linton, D.A. & Betts, R. (1988) Survival of self-marked mountain pine beetles emerged from logs dusted with fluorescent powder. Journal of the Entomological Society of British Columbia, 85, 2528.
  • Mason, P.L., Nichols, R.A. & Hewitt, G.M. (1995) Philopatry in the alpine grasshopper, Podisma pedestris: a novel experimental and analytical method. Ecological Entomology, 20, 137145.
  • Munro, J.A. & Telford, H.S. (1942) Grasshopper Control Aided by Tillage Methods. North Dakota Experiment Station Bulletin 309.Agricultural Experiment Station of North Dakota Agricultural College, Fargo, ND.
  • Naranjo, S.E. (1990) Influence of two mass-marking techniques on survival and flight behavior of Diabrotica virgifera virgifera (Coleoptera: Chrysomelidae). Journal of Economic Entomology, 83, 13601364.
  • National Grasshopper Management Board (1998) Proceedings and Recommendations of the 1998 Meeting of the National Grasshopper Management Board, January 10–12. National Grasshopper Management Board, Denver, CO.
  • Oloumi-Sadeghi, H. & Levine, E. (1990) A simple, effective, and low-cost method for mass marking adult western corn rootworms (Coleoptera: Chrysomelidae). Journal of Entomological Science, 25, 170175.
  • Onsager, J.A. & Henry, J.E. (1977) A method for estimating the density of rangeland grasshoppers (Orthoptera: acrididae) in experimental plots. Acrida, 6, 231237.
  • Pal, R. (1947) Marking mosquitoes with fluorescent compounds and watching them by ultra-violet light. Nature, 160, 298299.
  • Post, R.L. & Anderson, J.R. (1950) Devices for marking insects. North Dakota Agricultural Experimental Station Bimonthly Bulletin, 13, 8182.
  • Reeves, W.C., Brookman, B. & Hammon, W.M. (1948) Studies on the flight range of certain Culex mosquitoes, using a fluorescent dye marker, with notes on Culeseta and Anopheles. Mosquito News, 8, 6169.
  • Riegert, P.W., Fuller, R.A. & Putnam, L.G. (1954) Studies on dispersal of grasshoppers (Acrididae) tagged with phosphorus-32. Canadian Entomologist, 86, 223232.
  • Salom, S.M. & McLean, J.A. (1989) Influence of wind on the spring flight of Trypodendron lineatum (Olivier) (Coleoptera: Scolytidae) in a second-growth coniferous forest. Canadian Entomologist, 121, 109119.
  • SAS Institute (1988) SAS Users’ Guide: Statistics, Version 6.03. SAS Institute, Cary, NC.
  • Schmitz, R.F. (1980) Dispersal of pine engraver beetles in second growth ponderosa pine forests. Dispersal of Forest Insects: Evaluation, Theory and Management Implications (eds A.A.Berryman & L.Safranyik), pp. 4150. Cooperative Extension Service, Washington State University, Pullman, Washington State.
  • Shore, T.L. & McLean, J.A. (1988) The use of mark–recapture to evaluate a pheromone-based mass trapping program for ambrosia beetles in a sawmill. Canadian Journal of Forest Research, 18, 11131117.
  • Tepedino, J.L. (1979) The importance of bees and other insect pollinators in maintaining floral species composition. Great Basin Naturalist Mem, 3, 139150.
  • Thomas, M.B., Langewald, J. & Wood, S.N. (1996) Evaluating the effects of a biopesticide on populations of the variegated grasshoppers, Zonocerus variegatus. Journal of Applied Ecology, 33, 15091516.
  • Turchin, P., Odendaal, F.J. & Rausher, M.D. (1991) Quantifying insect movement in the field. Environmental Entomology, 20, 955963.
  • USDA (1987) Final Environmental Impact Statement on the Rangeland Cooperative Management Program. USDA-APHIS FEIS 87–1, Washington, DC.
  • Whitson, T.D., Burrill, L.C., Dewey, S.A., Cudney, D.W., Nelson, B.E., Lee, R.D. & Parker, R. (1991) Weeds of the West. The Western Society of Weed Science in co-operation with the Western United States Land Grant Universities Cooperative Extension Services and the University of Wyoming, Jackson, WY.
  • Zukel, J.W. (1945) Marking Anopheles mosquitoes with fluorescent compounds. Science, 102, 157.

Received 30 May 1998; revision received 21 April 1999