Research Article

Journal of Orthoptera Research 2023, 32(1): 33-42

Probability of a Central American locust Schistocerca piceifrons piceifrons
upsurge in the Yucatan Peninsula, Mexico

Mario A. Poot-PECH!

1 Yucatan State Plant Health, Yucatan, Mexico.
2 Grasshopper-Locust Management in Central America, Yucatan, Mexico.

Corresponding author: Mario A. Poot-Pech (mpootpech@gmail.com)

Academic editor: Hojun Song | Received 1 September 2021 | Accepted 10 July 2022 | Published 21 February 2023

https://zoobank. org/ODD815F9-F98A-4826-A5F4-F2311 CA4E964

Citation: Poot-Pech MA (2023) Probability of a Central American locust Schistocerca piceifrons piceifrons upsurge in the Yucatan Peninsula, Mexico.
Journal of Orthoptera Research 32(1): 33-42. https://doi.org/10.3897/jor.32.73824

Abstract

From ancient times to the present, infestations of the Central Ameri-
can locust (CAL) [Schistocerca piceifrons piceifrons (Walker, 1870)]| have oc-
curred periodically and with varying intensities in the Yucatan Peninsula
(YP), Mexico. Despite efforts to survey the recession zone, an upsurge is
still difficult to predict and prevent, and high economic costs are incurred
in controlling this pest. For this study, two models were developed to de-
termine the probability of an upsurge in the YP. The first was the Markov
chain (MC) with transition probability matrix, which estimates probabil-
ity by determining the proportion of times that the system moved from
one state to another (n,) over 71, 33, and 24 years in Yucatan, Campeche,
and the Quintana Roo States, respectively, divided into different periods;
a correlation of the matrix and probability (n,) of the next period was
performed to evaluate the accuracy of the estimation. The other method
is the classic probabilistic (CP) model, which uses the number of times
the upsurge could happen and the number of possible events. In the MC
model, great variation was found in CAL upsurge probabilities between
periods, with a similar number of upsurges from the past to the present
but with varying intensity. In recent years, the treated area with insecticides
has been less than that of the past. The CP model revealed that the locust
population reached its maximum peak every four years, with the migra-
tion of swarms to neighboring states at the end/start of the year. Valida-
tion of the MC and CP models was performed considering information
on areas treated in 2019 and 2020, and good accuracy was obtained. Both
models provide information on the probability of an upsurge in the YP.
This information can be incorporated into economic models to improve
management decisions, such as when to announce early warnings, and to
implement preventive control strategies.

Keywords

early warning, forecast, preventive management, recession period

Introduction

Locusts are among the most devastating pests of human agri-
culture (Lockwood 2015), making them a major threat to global
food security and the livelihoods of farmers in numerous coun-

tries (Simpson and Sword 2008). Schistocerca piceifrons (Walker,
1870) (Orthoptera: Acrididae) is the true locust of tropical America
(Harvey 1983). This species has two subspecies: the Central Ameri-
can locust (CAL; S. piceifrons piceifrons), known in Mexico and Cen-
tral America, and the Peruvian locust (S. piceifrons peruviana), dis-
tributed in Peru and Ecuador (Harvey 1983, Barrientos et al. 1992).

CAL damage has been documented from the Mayan culture
and colonial period, which reported drought and hunger as results
of this pest (Flores 2011, Garcia 2012), to the present (SENASICA-
DGSV 2016). According to Trujillo (1975), a CAL swarm can feed
on 30 tons of vegetation per day and, based on its food preference
and distribution in 10 Mexican states, the species could affect 5.9
million ha (SENASICA-DGSV 2016).

An important CAL breeding zone is located in the Yucatan
Peninsula (YP), Mexico (SENASICA-DGSV 2016), which possesses
ecological conditions-such as mainly grass vegetation-for high-
density locust development (Poot-Pech et al. 2018). According to
SENASICA (2019), the YP receives 50% of the national budget for
locust control because of the importance of breeding zones. The YP
is located in the southeast of Mexico. It is made up of three states—
Yucatan, Campeche, and Quintana Roo-that comprise 181,000 km?
(9.2% of the national territory). The overall climate is tropical, with
strong seasonality and dry (November-May) and rainy (June-Oc-
tober) seasons (Barber et al. 2001). Tropical forest is the dominant
vegetation, but grassland is increasing (Duran and Garcia 2010).
The soil is heterogeneous and mostly stony (Bautista et al. 2011).

Currently, CAL in Mexico are controlled through government
regulations and advice from the International Regional Organiza-
tion for Plant and Animal Health (OIRSA), an intergovernmental
organization founded in 1953. OIRSA was created as a cooperative
effort of locust control between Central American countries that
are part of the insect’s migration path (Barrientos et al. 1992).

Locust plagues occur after a series of events that increase the
number of locusts. The series normally begins with a calm period
of recession followed by localized outbreaks and upsurges from
which a plague may develop and eventually decline (Symmons
and Cressman 2001). Locusts survive the dry season as adults and

Copyright Mario A. Poot-Pech. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original author and source are credited.

JOURNAL OF ORTHOPTERA RESEARCH 2023, 32(1)

34

oviposit in bare areas. Young nymphs hatch and feed on adjacent
short ephemeral grasses, while older nymphs and young adults
survive on taller vegetation (Hunter and Spurgin 1999). An out-
break occurs when multiplication causes a marked increase in
locust numbers and densities so that individuals become gregari-
ous and, unless checked, hopper bands and/or swarms are formed
(Van Huis et al. 2007). The gregarization process takes several
months (Symmons and Cressman 2001). Upsurges are periods af-
ter an outbreak that are followed by two or more successive gener-
ations of transient-to-gregarious breeding, with populations occu-
pying an expanding breeding area (Van Huis et al. 2007). Periods
of one or more years during which upsurges are widespread are
considered plagued years (Symmons and Cressman 2001).

The CAL is an old pest in Mexico (Contreras and Galindo 2014)
that is still present in breeding areas today. To analyze and forecast
the distribution of CAL, Hernandez-Zul et al. (2013) built a dynam-
ic simulation to identify suitable conditions for high CAL reproduc-
tion rates. Galindo et al. (2014) proposed an early warning system
that used multicriteria models and satellite images. Aldama et al.
(2014) suggested that outbreak characterization of endemic chan-
nels could help in forecasting short-term locust thresholds in zones
where outbreaks can occur. The climate, El-Nino Southern Oscil-
lation (ENSO) (Retana 2000, Contreras and Galindo 2014), veg-
etation, locust density, and predators are additional indicators that
could be valuable in predicting a locust outbreak (Henschel 2015).

The strategies for locust control have has recently changed,
with attempts focusing on prevention with the use of forecasting
tools. Methods for producing Short- and medium-term forecasts
have been made indicating potential locust migration and breed-
ing areas (Cressman 1996). Zhang et al. (2019) proposed includ-
ing monitoring and forecasting as part of current preventive locust
management. Presently, biomodels and geographic information
systems (GIS) are used to support the monitoring and forecasting
of CAL outbreaks (Barrientos et al. 2021), but efficient statistical
tools, with or without historical information, are important for the
early detection of rapid changes in the breeding and migration of
pest species that lead to outbreaks (Naumova and MacNeill 2005).
Bayesian statistics and Markov models offer an effective method for
long-term population forecasting of pests and can provide locust-
control agencies and organizations with the information neces-
sary to implement appropriate management strategies (Xiang-Hui
et al. 2010). Knowledge of indirect effects is also a valuable tool
for predicting the probability of CAL upsurge, as it is important to
be aware of the relationship between areas treated with any pesti-
cide and locust evolution (outbreak, upsurge, or pest) (Prospatin
2012). Van Huis et al. (2007) described the treated areas with any
control agent as an answer to the function of locust area and time.

The objective of this research was to develop and compare two
models that can produce probability estimates for future upsurges
in the YP by analyzing non-weather-related historical CAL data.

Materials and methods

Data acquisition.—To document the occurrence of CAL upsurges,
the following sources of historical information on locust were
used: Marquez (1963), Trujillo (1975), Pereyra (1991), Chi (2000),
newspapers, interviews with locust field officers, and information
provided by the National Service of Health, Safety and Agrifood
Quality (SENASICA) on the locust control campaign in the YP,
Mexico, through its informatics system called the Phytosanitary
Campaign System (SICAFI; www.sicafi.gob.mx/DGSV/).

M.A. POOT-PECH

For Yucatan state, CAL data for 71 years (1948-2018) were
available (every year had a value). Data for 33 years (1986-2018)
were available for Campeche, and 24 years of data (1995-2018)
were available for Quintana Roo.

Locust upsurge definition.— Upsurges were used instead of out-
breaks because there were two high-density generations-—tran-
sients-to-gregarious or gregarious-to-gregarious—and migration
to other regions (Van Huis et al. 2007); when swarms move be-
yond breeding areas to other regions, the overall situation exceeds
a hazily demarcated threshold from “outbreak” to “upsurge”
(Showler 2019). The CAL was present in Yucatan State in a gre-
garious form in almost years which is reflected in larger treated
surface with insecticides (Fig. 1).Upsurges, therefore, allow con-
sideration of the large number of migrant swarms that formed
and reached the state of Campeche. As a result of such upsurges,
the treated area required for CAL control was high in one year as
in 2004, 2006, 2009, 2010, 2014, and 2018 or consecutive years
1952-1955, 1979-1981 and 1986-1989 on gregarious popula-
tions (Fig. 1) that reached Campeche during migration. The reac-
tion in this state was the development of highly controlled surface
in 2007, 2011, 2015, and 2019 because subsequent swarms (451
in 2006-2007, 245 in 2010-2011, 243 in 2014-2015, and 120 in
2018-2019) migrated from Yucatan (Fig. 2). In contrast, in years
without an upsurge, the number of swarms detected in Yucatan
was at a minimum (0-60 swarms) and failed to reach Campeche.

Markov chain probabilistic model.—The data were organized by as-
signing “O” to years with none or low intensity of CAL swarms
and “1” to those in which an upsurge occurred. Second, the data
were organized into periods of upsurge based on number of years,
resulting in 5 periods of 12 years and one of 11 years in Yucatan
state; 3 periods of 12 years, except for the first period of 9 years, in
Campeche; and 2 periods of 12 years in Quintana Roo.

Classic probabilistic model.—The maximum peak years were ob-
tained using the treated area (ha) in Yucatan state from 2003-
2018. Subsequently, a consecutive number was assigned to each
of the following years, finalizing in the next peak. A probabilistic
analysis of occurrence was performed for each consecutive year.
The probability of occurrence in the most recent years (2003-
2018) was compared with that of previous periods [1977-1989
(Pereyra 1991) and 1952-1955 (Marquez 1963)].

Model validation.—The probabilistic results of the two models
were compared to the results of the treated area in the YP in 2019
and 2020.

Data analysis

Markov chain probabilistic model.—The probability of matrix transi-
tion for every period was estimated by determining the proportion
of times that the CAL upsurge situation, 0 and 1, moved from one
state to another (Kemp 1987). In this context, there were four pos-
sible transitions: no upsurge — no upsurge (0-0), no upsurge >
upsurge (0O— 1), upsurge — no upsurge (1-0), and upsurge — up-
surge (1—1) (modified from Zimmerman et al. 2004). These esti-
mates were represented by a p matrix developed for each year, where

ar PoosPor
PagsPa4

JOURNAL OF ORTHOPTERA RESEARCH 2023, 32(1)

M.A. POOT-PECH

35

Table 1. Matrix (one-step transition) and n,, for CAL upsurges in a 71-year period.

Number Years Yucatan Number Upsurge Upsurge Probability n,
Matrix (One-step transition) of years years average

0—0 0-1 1-0 1-1 0-0 0-1 1—0 1-1
1 1948-1959 0.5 0.5 0.43 0.57 M2 if ral 0.46 0.54 0.46 0.53
ps 1960-1971 0.5 Ons 0.6 0.4 12 i, 2.40 0.55 0.45 0.54 0.46
3 1972-1983 0.71 0.29 Ov75 0.25 12 4 3.00 0:72. 0.28 0.72 0.28
4 1984-1995 0.2 0.8 0.67 0.33 12 6 2.00 0.57 0.43 0.35 0.65
5 1996-2007 0 a. 0.72 0.28 12 7 1.71 On Fa 0.28 @22 0.8
6 2008-2018 0.58 0.42 0.66 0.34 11 4 2.75 0.61 0.39 0.60 0.4

71 (total) 33 (total) 2.15 (mean)

The elements P,,, and P,, refer to the probability that in X year
(present), the following year’s CAL population will be classified as
low (P,,) or high (P,,) if X year has low densities. Alternately, if X
year has a high locust density, P,, is the probability that the follow-
ing year's populations will be classified as having a low density,
and P,, is the probability that the state will continue to be classi-
fied as having a high density of CAL the next year (Kemp 1987).

The probability of correctly determining the locust density in
the next period was obtained using the recursive properties of a
two-state Markov chain (Bhat 1972). The matrix P(n-1) was mul-
tiplied by the matrix P to generate the matrix p(n). The element-
by-element calculations for P™ were as follows:

rE al

Ol ~~ 00 01 01 U1;
lige Fe ye by = Ness a
ie) + Bgae an + Pe

The correlation between the probability n, of the two-state
Markov chain and the probability of the next period (one-step
transition)-for example, the probability n, of period 1 with the
matrix of period 2—was determined using Pearson correlation
(p < 0.05) in R software version 3.6.0 (R Core Team 2019).

Classic probabilistic model.—We obtained the classic probability us-
ing P(1) = #1/M, where P(1) is the probability of an upsurge, #1 is
the number of times the upsurge can happen, and #M is the num-
ber of possible events (Infante and Zarate 1990). Upsurge years
were designated as 1, with years of no upsurge designated as 0.

Results

Markov chain probabilistic model

Yucatan state.—A great deal of variation was found in CAL upsurge
probabilities between periods, except in transitions 00 and 0-1
of periods 1 and 2. In the first period, transition 1—1 (0.57) was
highest, indicating years with contiguous upsurges. The second

period was characterized by a reduction in upsurge frequency of
1—1 (0.4). In the third period, this transition had its biggest re-
duction (0.25), indicating an increase in “recession years,” and
the transition 0—0 was increased to 0.5 to 0.71 . During periods 4
and 5, the values of the transition matrix were very similar (0-0:
0-0.2, O01: 0.8-1, 1-0: 0.67-0.72, and 1—1: 0.28-0.33). Thus,
an upsurge appeared in one year, and in the next, it was reduced.
The last period was very similar to the second period, where the
recession years were increasing 0-0 (0.58) and 1-0 (0.66) and
remained high. From 1948 to 2018, there were 33 upsurges at an
average of 2.15 per period and with a range of 4 to 7. Periods 1
and 5 had a major upsurge, and periods 3 and 6 had a minor lo-
cust presence.

In 4 of 5 cases, the correlation between the matrix and prob-
ability n, in the next period was negative, and the P-value was
not statistically significant (P > 0.05). In the period 1972-1983,
the correlation was high, positive, and statistically significant
(P < 0.01).

Table 2. Pearson correlation of the matrix and probability n, in
Yucatan state (P < 0.05).

Period Correlation P-value
1960-1971 -0.65 0.34
1972-1983 0.98 0.01
1984-1995 -0.26 0.73
1996-2007 -0.75 0.24
2008-2018 -0.64 0.35

Campeche State.—In the matrix, the three periods of CAL upsurge
in Campeche State were different. In period 1, which was the
shortest, O01 (1) and 1—1 (0.72) stood out. Periods 2 and 3
showed identical values-1—0 (1) and 1—1 (0)-and had similar
0—0 (0.57-0.75) and 0—1 values (0.43-0.25).

The number of upsurges per year decreased; the first period
had the highest number of upsurges (7), followed by periods 2
(4) and 3 (3).

Table 3. Matrix (one-step transition) and n,, for CAL upsurges in a 33-year period.

Number Period Campeche Years Upsurge Upsurge Probability n,
Matrix (one-step transition) years overage
0-0 0-1 1-0 1-1 0-0 0-1 1-0 1-1
il 1986-1994 0 i 0.28 0.72 9 ‘f 129 0.28 0.72: 0.2 0.8
4 1995-2006 0.57 0.43 1 0 12 f 3 0.75 0.25 0.57 0.43
3 2007-2018 0.75 0.25 1 0 aly? 3 4 0.81 0.19 O75 0.25

33 (total) 14 (total) 2.3 (mean)

JOURNAL OF ORTHOPTERA RESEARCH 2023, 32(1)

36

Table 4. Pearson correlation of the matrix and probability n, in
Campeche (P > 0.05).

Period Correlation P-value
1986-1994 -0.88 0.11
1995-2006 0.67 0.32

The Pearson correlation was different in the two periods. In
1986-1994, it was negative, and in 1995-2006, there was a posi-
tive association. However, the differences were not statistically sig-
nificant (P > 0.05).

Quintana Roo State.—In the 24 years of data (1995-2018), only
one year had a CAL upsurge: 2006. Therefore, the probability for
1-0 and 1-1, equal to 1 and n, (Table 5), could not be obtained.
According to data from the Quintana Roo Plant Health locust con-
trol from Yucatan migrated and caused damage to corn and bean
crops and cultivated grasslands and caused defoliation in a nature
reserve and on urban trees from August-September. In 2006, lo-
cust control operations were undertaken in 1,381 ha.

These swarms were able to oviposit in Quintana Roo and com-
plete the second generation. They also returned to Yucatan at the
end of 2006 as swarms. In that generation, 173 ha required con-
trol operations. In subsequent years, the locust population was
present at a low density in the solitarious phase.

Table 5. Matrix (one-step transition) for CAL upsurges over a 24-
year period.

Period Quintana Roo Years Upsurge Upsurge
Matrix (one-step transition) years average
0-0 O-1 1-0 1-41
1995-2006 0.9 0x1 0 0 12 1 1
2007-2018 1 0 0 0 hes 0 0
Total 24 i 1

Classic probabilistic model

Fig. 1 shows a tendency for upsurges to peak every four years
(2006, 2010, 2014, and 2018), except for an additional two peaks
in 2004 and 2009, from 2010 to 2018, this tendency was clearer
than from 2003 to 2009.Therefore and consequently it was as-
signed values from years 1 to 4, starting with the maximum peak,
e.g: year 2006 (year value 1), 2007 (2), 2008 (3), 2009 (4), 2010
(1), 2011 (2) so on (Table 6); from 1952-1955 and 1977-1989,
the tendency of the maximum peaks was less clear, and the treated
area was larger than in recent years.

This information was used to identify the upsurge years
(Table 6). Then, the probability of an upsurge in the past and cur-
rent periods was determined (Table 7).

M.A. POOT-PECH

13,261

8,045

™
Li
lo)
oo
Lo wT
Ly
o
Cael Mm
nm m n m : oo
in i wn a st to
uw = wn = fe- = me
| | : |

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

me io

Mmm 1.164
Mm 1311
Mm 1481

Mm 2.01

o
a

Treated area (ha)

18,000
16,000
14,000
13,246

rh
J
ive]
nm
m4
to =
m ‘.
on om
Ln
i
on
N oo an
n i
st ia)
Lo ‘o
i
° 9 od F
Me =s a e ee
: zi | | -

1952 1953195419551977 197819791980 19811982 1983 19841985 19861987 19881989
Years
Fig. 1. Treated area in Yucatan State from 2003-2018 (above) and
1952-1955/1977-1989 (below).

ME 2000

mmm 2,570

For the period 2003-2018 in Yucatan, the highest probability
of upsurge was in year 4 (P: 1). There was no probability of an up-
surge (P: 0) in year 1 and a minimal probability (P: 0.25) in years
2 and 3. These results were different from the data from the past
(1952-1955 and 1977-1988), where years 1 and 2 were similar
(P: 0.5) and years 3 and 4 were similar (P: 0.75). For Campeche,
the P(1) was 1 in year 1, and years 2-4 had no values. In Quintana
Roo, there was only a remote probability of an upsurge, but the
probability increased if Yucatan had an upsurge so intense (simi-
lar to 2006) that it resulted in locust invasion.

Validation.—According to SENASICA (Table 8), the area treated for
locust control was 2,224 ha in 2019 and 3,714 ha in 2020. This
information was used to validate the model.

Validation of Markov chain probabilistic model

In Yucatan state starting in 2018, which had a value of 1, the
transition 1-0 (one step: 0.66, n,:0.6) had the highest value.
Therefore, the following year, 2019, an upsurge value of 0 would
be expected and starting in 2019 as upsurge value 0, the transi-
tion (0-0: 0.58, 0.61) was the highest value, so it would also be
expected an upsurge value 0 in 2020, both results obtained of the
probabilistic model correspond to the results of the low treated
area (Table 8). In Campeche in 2018, value = 0, the transition 0—0
(0.75; 0.81) had the highest value, a year with value 0 was expect-
ed, but it was not fulfilled, in 2019 according to the treated area
the value was 1. The value 1—0 (1, 0.75) was highest for 2020 and

Table 6. Upsurge value for 2003-2018 in the YP and 1952-1955/1977-1989 in Yucatan state.

States Upsurge value per year

2003. 2004 2005 2006 2007 =+(.2008 2009 2010 2011 £2012 2013 #2014 2015 2016 2017 2018

Yucatan 0 1 0 1 0 0 1 1 0 0) 0 1 0 0) 0 1

Campeche 1 0 0 0 1 0 0 0 1 0 0 0 1 0 0 0

Q. Roo 0 0 0 1 0) 0 0 0 0 0 0 0 0) 0 0 0
Yucatan 1952. 1953 1954 1955 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

1 1 1 1 0) 0 1 1 1 0 0 0 0 1 1 1

Year value 1 2 3 4 J 2 3 4 1 2 3 4 1 2 3 4

JOURNAL OF ORTHOPTERA RESEARCH 2023, 32(1)

M.A. POOT-PECH

it was the upsurge value, through the treated surface (Table 8). In
Quintana Roo, despite good results in the probability of 0—0 for
2019 and 2020, no application was made in transition n, with an
origin of 1 (1—1 or 1-0) because it is an invasion area and not a
locust breeding zone.

Validation of Classic probabilistic model

The comparison of the probability of an upsurge P(1), with
previous information 2003-2018, with values 2019-2020 indicate
that in Yucatan it was according to the model in 2019 and very
close to 2020 (P(1)=0.25, value=0), in Campeche and Quintana
Roo as predicted.

Table 7. Probability of an upsurge, P(1), from 1952-1955/1977-
1988 and 2003-2018.

Year 1952-1955 and 1977-1988 2003-2018

value Yucatan Campeche Q. Roo
Upsurges P(1) Upsurges P(1) Upsurges P(1) Upsurges P(1)

1 1,0,1,0 O.3e) 20;0;050 0 1,1,1,1 1 0,0,0,0 0

2 1,0,0,1 05. 10/00 01255, °.0;0;0;0 0 0,0,0,0 0)

3 1,1,0,1 0.75 0,1,0,0 0.25 0,0,0,0 0 0,0,0,0 0

4 1,1,0,1 Osa, “hla 1 0,0,0,0 0 1,0,0,0 0.25

Table 8. Treated area (ha) in the Yucatan Peninsula in 2019 and 2020.

State Years
2019 2020
Yucatan 813 3,420
Campeche 1411 294
Quintana Roo 0 0
Total 2224 3714
Discussion

Markov Chain Model

Yucatan State.—The CAL upsurge from 1948 to 2018 was a sign
that the frequency of occurrence had changed, as there were no
equal periods. The transition 1—1 had the highest value in the
first period (1948-1959); therefore, it was called a “plague pe-
riod” in which “favorable breeding conditions are present and con-
trol operations fail... two or more regions are affected simultaneously”
(Symmons and Cressman 2001). Upsurge had a higher value in
the past than in recent years because there was no organization
in charge of a permanent locust control program. Control meas-
ures were more recently undertaken over gregarious populations,
mainly swarms (Marquez 1963). From 1941-1942, CAL swarms
invaded Mexico and Central America,originating from some area
of this same region. In 1947, the International Committee for Lo-
cust Control (CICLA) was formed; in 1955, the name was changed
to OIRSA to include locust surveillance (Marquez 1963, Trujillo
1975). In 1948, the Mexican General Direction of Agricultural
Protection was created, an institution that consolidated technical
cooperation with CICLA. The first period (1948-1959) in Mexico
was characterized by the creation of national and international
institutions for locust management (Ortiz and Zuleta 2020).

In the second period (1960-1971), the transition 1—1 de-
creased, perhaps as an effect of more organized locust control and
unfavorable weather conditions, and the number of upsurges per

37

period was reduced from 7 to 5. In period 3, 1972-1983, the tran-
sition 0-0 was the highest (0.71). Consequently, because of the
fewer number of upsurges (4), the locust was declared to be in
its recession period, i.e., a period of several years when the locust
population is low (Symmons and Cressman 2001).

In periods 4 (1984-1995) and 5 (1996-2007), the values of
the transition matrix were very similar, the transition 0-1 and
1—0 were high, that is the upsurges occurred approximately every
2 years. When starting with a solitarious population, CAL needs
three generations to reach the gregarious phase (Barrientos et al.
1992). The CAL presents two generations per year (Astacio 1966,
Trujillo 1975); therefore, a third generation is achieved within two
years, which concurs with the calculations of Marquez (1963),
who indicated that “the locust does not form suddenly; it takes two
or more years to develop.” In periods 4 and 5, transition 1—1 was
reduced because locust management was strengthened with the
creation of state-run locust control programs. For example, in
1988, the Ministry of Agriculture and Water Resources of Mexico
granted greater autonomy to phytosanitary programs in each state,
including the locust control program. The Federal Plant Health
Law (Ley Federal de Sanidad Vegetal 1994) created plant health
committees to develop multiple crop protection programs with
government funding. These programs encompassed the locust
control campaign. In Yucatan, such a committee was created in
1997 and is currently responsible for the permanent monitoring
and control of this pest.

Although Yucatan periods 1 and 5 had similar upsurge years
(7), the severity differed. The overall area treated with insecticides
in period 1 (1952-1955) was 56,000 ha (Marquez 1963), while it
was 26,363 ha in period 5 (2003-2006), which was 47% less than
in the first period.

Only 1 out of 5 correlations between the probability n, and
the matrix was positive and significant. This may be because the
outcome of the Markov chain depends on the outcome of previ-
ous events, meaning that the next state of the system depends on
the present state, and locust outbreaks are erratic events (Sym-
mons 1992). Showler (2002) outlined broad strategic approaches
to locust outbreaks based on intervention timing (prevention or
reaction) and tactics (survey, control, economic effects, inse-
cure areas, and contingency planning). Management, weather,
climate change [World Meteorological Organization-Food and
Agriculture Organization (WMO-FAO) 2016], and political de-
cisions differ in different periods, making it difficult to obtain
reliable probability.

Markov models have limitations, are problematic with short
time intervals, cannot be derived rigorously from deterministic,
dynamic models, and rarely provide the range of time for which
the model is appropriate (DelSole 2000). However, they permit
knowledge of the outbreak population in each period without
the necessity of searching for causation through correlative proce-
dures (Kemp 1987). Weather conditions that could have an influ-
ence on outbreaks include the amount of rain in the year (Stein-
bauer 2011), temperature, relative humidity (Al-Ajlan 2007), wind
for migration (WMO-FAO 2016), ENSO (Contreras and Galindo
2014), and precipitation in previous years (Skinner and Child
2000, Chiconela et al. 2003).

In some species of Orthoptera, there has been a decrease in
outbreaks; for example, outbreaks of rangeland grasshoppers in
Wyoming are highly erratic events, with instances of infestations
persisting for multiple years being quite low, so there is little basis
for prorating the benefits of control beyond the year of treatment
(Zimmerman et al. 2004).

JOURNAL OF ORTHOPTERA RESEARCH 2023, 32(1)

38

Campeche state.—Historically, the Campeche state has been invad-
ed by swarms from Yucatan (Marquez 1963). The entry and migra-
tion routes are found in the north, but these swarms can quickly
move to the south of the state (Chi 2000). This movement is made
by the second generation and at the beginning of the year (Cullen
et al. 2017). In the second period (1995-2006), there was a reduc-
tion in the number of upsurges per year, probably because the lo-
cust campaign of Yucatan Plant Health established control opera-
tions in 1997, reducing the possibility of migration to Campeche.
Before this period, the largest number of upsurges occurred, with
the transition 1—1 indicating several continuous years with an
upsurge or a reduced recession period (0-0). In the third period,
there was an increase in the recession period 0-0 because the
upsurge was dependent on outbreak and migration from Yucatan
every 4 years (2006, 2010, and 2014). The probability of upsurge
in 2019 was expected because of the migration of swarms from
Yucatan in 2018.

With the information obtained on the control and migration
of swarms, it was possible to construct the migration route of CAL
to Campeche. At the beginning of the year, Campeche is an inva-
sion zone (Fig. 2) for swarms originating from places in Yucatan,
where it is difficult to undertake survey and control operations.
Therefore, it is important to carry out locust management tasks in
Yucatan state with new methods to reduce migration to Campe-
che. Despite the large number of swarms formed in Yucatan in the
upsurge year, the damage was minimal for two reasons: a) in the
maximum period of migration that occurs at the end-beginning
of the year, there is less cultivated area, as most of the agriculture
occurs in June-September when there is a greater amount of pre-
cipitation to help rainfed crops, and b) the size of the swarms,
according to the classification of Cressman (2001), corresponds
to very small swarms (<1 km?) due to preventive management of
first stage nymphs.

Campeche

M.A. POOT-PECH

Quintana Roo State.—Vegetation is very important for locust de-
velopment (Sword et al. 2010). Its distribution (Despland et al.
2000), coverage, and status (green or dry) influence locust gregari-
zation (Cisse et al. 2013). Models could be enhanced by using as-
sociations between plant communities and insects to predict risk
areas (Van Der Werf et al. 2005). The CAL is highly associated with
the grass Panicum maximum (Poot-Pech et al. 2018); however, in
Quintana Roo, the vegetation is 90% tropical forest, 3.3% agricul-
ture, 3.2% grasses, and 3.5% other uses (INEGI 2011). According
to survey information from the locust campaign, P. maximum is
not the dominant grass in Quintana Roo.

Classic probabilistic model

In Yucatan in 1952-1955 and 1977-1988, there were 10 up-
surges and 6 recession years, while in 2003-2018, there were 6
upsurges and 10 recession years. This may be because the structure
of the locust program was modernized during the latter period,
with greater autonomy and economic resources for developing the
program and prevention strategies (Rodriguez 2000). The locust
program in Mexico receives technical support from OIRSA (Ortiz
and Zuleta 2020), and in general, biological and ecological knowl-
edge about locusts has increased, which has helped create better
management practices for this pest (Lecoq 2021).

Periods of 4 years (Table 7) were formed, and a linear regres-
sion of areas treated for CAL was obtained (Figs 3, 4), finding a
positive relationship. Generally, the first year had the least amount
of treated area, and the fourth year had the most amount of treated
area. In the fourth year, there was a high migration of swarms to
Campeche (Fig. 2), so the gregarious area remained at low infes-
tation for three generations. Two years later, the population be-
came high once again (Marquez 1963, Barrientos et al. 1992).
However, this pattern occurs when the swarms are controlled

2010-2011

United States of America

Months of Swarm
Migration
November
December
January
February
March

April

> @Heeo

Campeche

Fig. 2. Swarms detected and treated in the Yucatan Peninsula from November-December (2006, 2010, 2014, and 2018) and January-—

April (2007, 2011, 2015, and 2019).

JOURNAL OF ORTHOPTERA RESEARCH 2023, 32(1)

M.A. POOT-PECH

15000
|
i

R?: 0.69 P:0.31

i il
sf, Ed
QoQ
a _|
Qo
Ww
a
2003 2004 2005 2006
o
oOo
S 4
= R2: 0.86 P:0.13 e
ie |
oo
oO
Ps
Qo
Ww
o Je ° uf
I
2011 2012 2013 2014
Years

39

5
| R2: 0.93 P:0.06
s al
i wal
oOo
ao |
oO
ma e
o
= Tae
2007 2008 2009 2010
oO
oOo
re)
= R2: 0.82 P:0.17
oO
oo
Lon]
= 6
uo
®
2
2015 2016 2017 2018
Years

Fig. 3. Linear regression of the treated area in periods of 4 years from 2003-2018 in Yucatan (SENASICA information).

R?: -0.95 P:0.04

20000
|

ha
10000
|
|

0
|

20000
|

ha

0 10000
| |
/
Lo

1983

Years

R*: 0.88 P:0.12

20000

ha

10000
|

0
4@
+»

] ]
1979 1980

R*: 0.13 P:0.82

20000
|

ha

10000

0
i
6

1987 1988

Years

Fig. 4. Linear regression of the treated area in periods of 4 years from 1952-1955 (Marquez 1963) and 1997-1988 (Pereyra 1991) in Yucatan.

along their migration route; if the swarms are not controlled, then
year one would have a high infestation when the swarms oviposit
and return in the next generation, flying in the opposite direction
(Cullen et al. 2017) and repeating the pattern presented in previ-
ous periods (1952-1955 and 1977-1988; Tables 6, 7) with at least
two years of high infestation. This has been seen recently, such as
in the upsurge in 2019 and 2020 of the desert locust in Pakistan
(Sultana et al. 2021) and Africa (Peng et al. 2020, Salih et al. 2020).

There were intermediate years—year 2 in 2004 and year 3 in
2009-which were likely the result of suitable weather conditions
such as precipitation in the breeding zone. A marked increase in
locust numbers on a local scale due to concentration, multiplica-
tion, and gregarization can lead to the formation of hopper bands
and swarms (Roffey and Popov 1968).

In 1952-1955, there was a “plague” of CAL, requiring 4 years
of intense control; gradually, the size of the controlled area was re-
duced. The opposite situation occurred in 1977-1980 and 2007-
2010, with an intermediate rebound in 1979. This situation lasted
until 1981, and the treated surface area was reduced until 1984.

There was another plague from 1985-1989, with 1986 having the
largest controlled area.

From 1952-1955, as shown in Fig. 4, there was a negative
regression, indicating a plague of CAL (Symmons and Cressman
2001). Efforts were made to reduce the population each year, but
control was aimed at the gregarious phase and mostly swarms
(Marquez 1963). International cooperation among countries was
still in development (Yam and Zuleta 2020). In 1977-1980, there
was a positive regression; an increase in the population occurred
in 1979-1980, but in 1981-1984, there was again a negative re-
gression, with a rebound in 1981. Finally, in 1985-1989, the re-
gression value was low because in 1986, there was a higher peak;
however, in the period 1986-1987, there was a plague.

In his book, An Account of the Things of Yucatan, written in
1566, Diego de Landa (2003), bishop of Yucatan, wrote about the
consequences of locusts and noted the 4-year recession period,
saying, “the locust developed for 5 years, which left no green thing, it
was so hungry that people fell dead on the roads ... with 4 good years
after the locust population had improved somewhat.”

JOURNAL OF ORTHOPTERA RESEARCH 2023, 32(1)

40

Table 9. Results of Markov matrix in 2018 compared to results
from 2019 and 2020 upsurge values.

State (value Matrix Values Upsurge
year 2018) value
0-0 O—1 1-0 1-1 2019 2020
Yucatan One-step transition 0.58 0.42 0.66 0.34 0O 0
(2018=1)
n2 0.61 0.39 0.6 £O0.4
Campeche One-step transition 0.75 0.25 il 0 1 0
(2018=0)
n2 0.81 ..0;19° *0.75. 0.25
Q. Roo One-step transition 1 0 0 0 0 0
(2018=0)

Table 10. Probability of an upsurge P (1) and results of upsurge
values 2019 and 2020 in the YP.

Year Years Yucatan Campeche Q. Roo
value P(1) Value P(1) Value P(1) ~~ Value
1 2019 0) 0) 1 1 0) 0
2 2020 O25 0 0 0 0 0
3 0.25 0 0
4 1 0 0.25

Infestation of the desert locust, S. gregaria, in Africa occurred in
four out of five years between 1860 and 1963, and subsequently,
in one year out of six (Symmons 1992). In India, from 1863-
1962, there were 10 locust infestations, with 5-6 consecutive years
of widespread reproduction, swarm production, and damage to
crops, followed by 1-8 years of low activity; from 1963 to 2012,
there were 18 upsurges (Sharma 2014).

The WMO-FAO (2016) listed the effects of the following fac-
tors on the development of a locust outbreak or upsurge: manage-
ment; the failure to implement a preventive control strategy; inex-
perienced field survey teams and campaign organizers; insufficient
or inappropriate resources; lack of training of the field officers;
inaccessibility to important breeding areas; the weather (precipi-
tation, temperature, and wind); and climate change. These factors
are likely to have influenced the high occurrence of CAL in the
intermediate years (years 2 or 3).

Additionally, limited financial capacity and ongoing armed
conflicts could have rendered some of the locust breeding areas
inaccessible, and the coronavirus pandemic lockdown has fur-
ther hampered control efforts (Salih et al. 2020). Political and
socio-economic conditions (Meynard et al. 2020) can also greatly
influence the likelihood of locust outbreak by affecting manage-
ment decisions (Gay et al. 2018). In period 2003-2018 there was
an important change of government in 2007 in Yucatan State
(https://www.yucatan.gob.mx/?p=cronologia) that included a
change in locust-control decision-makers. A clear difference be-
tween the periods 2003- 2010 and 2011-2018 can be seen: Four up-
surges were present in the first period, with only two in the second.
In the second period, upsurges only occurred every four years as a
result of the preventive actions and the incorporation of new tech-
nologies in the locust-control program (Barrientos et al. 2021).

Plagues arise when locusts breed frequently and successfully
over a period of one or more years, with repeated and widespread
rains in successive, often widely separated, seasonal breeding ar-
eas. This allows swarms to form and invade the agricultural zones
surrounding the recession area. Pest control, drought, and migra-
tion to unsuitable areas can have an effect on ending plagues,

M.A. POOT-PECH

although their relative importance is not always clear (Pedgley
1981). The uncertainty of pest outbreaks and associated damage
(risk) can be reduced by preventive practices and by the selective
use of pesticides based on monitoring and forecasting (Daamen
and Rabbinge 1991). It is very important to predict and identify
locust recession periods because these populations begin the inva-
sion when ecological conditions become favorable (Lazar et al.
2016), and early prevention strategies have been proven to prevent
damage to major agricultural zones in the invasion area (Sharma
2014). However, the response to an alert must be quick; otherwise,
swarms will invade several regions, making the forecast pointless
(Ceccato 2007).

Curiously, from 2003-2018, the largest treated area against
CAL (Fig. 3) coincided with the years in which a soccer champi-
onship was held (2006 Germany, 2010 South Africa, 2014 Brazil
and 2018 Russia), so in Yucatan these CAL upsurges are known as
“soccer world cup locusts”.

Both probabilistic models are functional as long as the per-
manent monitoring system is sustained and allocated resources
are maintained or increased. A reduction in budget would risk the
development of the pest, as has occurred in the past. The results of
the models discussed here provide insights into the probabilities
of CAL upsurges in the YP, and this information can be incorpo-
rated into ecological models to improve CAL monitoring and aid
management decisions.

Acknowledgements

I thank the field locust officers of the YP, SENASICA, and OIR-
SA. Special thanks to Mario Marin-Correa and Eudaldo Pereyra-
Cuevas, retired YP locust officers who helped rebuild history. I also
thank CONACyT (Consejo Nacional de Ciencia y Tecnologia) and
ITC (Instituto Tecnolégico de Conkal) for their support of my PhD
studies. Special thanks to the reviewer of this document for his
insightful comments and corrections.

References

Al-Ajlan AM (2007) Relationship between desert locust Schistocerca gregaria
(Forskal), infestation, environmental factors and control measures in
Gazan and Makkah Regions, Saudi Arabia. Pakistan Journal Biological
Sciences 10: 3507-3515. https://doi.org/10.3923/pjbs.2007.3507.3515

Aldama AC, Hipdlito CG, Garcia MFJ, Mata CF (2014) Caracterizacion
de los brotes de langosta Centroamericana (Schistocerca piceifrons
piceifrons (Walker) en el Sureste de México. In: Galindo MMG,
Contreras SC, Ibarra ZE (Eds) La plaga de la langosta, Schistocerca
piceifrons piceifrons (Walker) una vision multidisciplinaria desde la
perspectiva del desastre fitosanitario en México. Colecci6n Sanidad
Vegetal Tomo 1. Vigilancia Epidemiol6gica Fitosanitaria. Universidad
Autonoma de San Luis Potosi, México, 160-172.

Astacio CO (1966) Observaciones Sobre la Biologia, Ecologia y Control
del Chapulin (Schistocerca paranensis Burn) en Nicaragua. OIRSA,
Nicaragua, 26 pp.

Barber A, Tun J, Crespo MB (2001) A new approach on the bioclimatol-
ogy and potential vegetation of the Yucatan Peninsula (Mexico).
Phytocoenologia 31: 1-31.

Barrientos LL, Astacio CO, Alvarez BE Poot MO (1992) Manual Técnico
Sobre la Langosta Voladora (Schistocerca piceifrons piceifrons Walker,
1870) y Otros Acridoideos de Centroamérica y Sureste de México.
FAO-OIRSA, El Salvador, 162 pp.

Barrientos LL, Song H, Rocha SAY, Torres CJA (2021) State of the art man-
agement of the Central American Locust Schistocerca piceifrons piceifrons
(Walker, 1870). Agronomy 11: 1024. https://doi.org/10.3390/agrono-
my11061024

JOURNAL OF ORTHOPTERA RESEARCH 2023, 32(1)

M.A. POOT-PECH

Bautista F, Palacio-Aponte G, Quintana P, Zinck JA (2011) Spatial distribu-
tion and development of soils in tropical karst areas from the Penin-
sula of Yucatan, Mexico. Geomorphology 135: 308-321. https://doi.
org/10.1016/j.geomorph.2011.02.014

Bhat UN (1972) Elements for Applied Stochastic Process. Wiley, New York,
488 pp.

Ceccato P, Cressman K, Giannini A (2007) The desert locust upsurge in
West Africa (2003-2005): Information on the desert locust early
warning system and the prospects for season climate forecasting.
International Journal of Pest Management 53: 7-13. https://doi.
org/10.1080/09670870600968826

Chi QG (2000) Problematica Agricola de la Langosta Schistocerca piceifrons
(Walker) en el Estado de Campeche, Periodo 1995-1998. Memo-
ria de Experiencia Profesional, Universidad Autonoma Chapingo,
México, 76 pp.

Chiconela T, Chongo D, D‘Uamba P, Ngazero A, Santos L (2003) Predict-
ing the occurrence of red locust outbreak in Mozambique. African
Crop Science Conference Proceedings 6: 224-230.

Cisse S, Ghaout S, Mazih A, Ould BEMA, Benahi AS, Piou C (2013) Effect
of vegetation on density thresholds of adult desert locust gregariza-
tion from survey data in Mauritania. Entomologia Experimentalis et
Applicata 149: 159-165. https://doi.org/10.1111/eea.12121

Contreras SC, Galindo MMG (2014) La plaga de la Langosta Centroameri-
cana Schistocerca piceifrons (Walker) en México: Reconstrucci6n ambi-
ental 1592-2000. In: Galindo MMG, Contreras SC, Ibarra ZE (Eds)
La plaga de la langosta, Schistocerca piceifrons piceifrons (Walker) una
visi6n multidisciplinaria desde la perspectiva del desastre fitosani-
tario en México. Coleccidn Sanidad Vegetal Tomo 1. Vigilancia Epi-
demiologica Fitosanitaria. Universidad Autonoma de San Luis Potosi,
México, 88-110.

Cullen DA, Cease AJ, Latchininsky AV, Ayali A, Berry K, Buhl J, De Key-
ser R, Foquet B, Hadrich JC, Matheson T, Ott SR, Poot-Pech MA,
Robinson BE, Smith JM, Song H, Sword GA, Vanden Broeck J, Ver-
donck R, Verlinden H, Rogers SM (2017) From molecules to manage-
ment: Mechanisms and consequences of locust phase polyphenism.
Advances of Insect Physiology. 53: 169-285. https://doi.org/10.1016/
bs.aiip.2017.06.002

Cressman K (1996) Current methods of desert locust forecast-
ing at FAO. Bulletin OEPP/EPPO 26: 577-585. https://doi.
org/10.1111/j.1365-2338.1996.tb01500.x

Cressman K (2001) Desert locust Guidelines. 2 Survey. FAO, Rome, 33 pp.

de Landa D (2003) An account of the things of Yucatan: written by the
bishop of Yucatan, based on the oral traditions of the ancient Mayas.
Reprint. Monclem Ediciones. Mexico, 174 pp.

Daamen RA, Rabbinge R (1991) Risks, forecasting and tactical decision-
making in integrated pest and disease management. Bulletin OEP-
PO/EPPO 21: 461-467. https://doi.org/10.1111/j.1365-2338.1991.
tb01276.x

DelSole T (2000) A fundamental limitations of Markov models. Jour-
nal of the Atmospheric Sciences 57: 2158-2168. https://doi.
org/10.1175/1520-0469(2000)057%3C2158:AFLOMM%3E2.0.CO;2

Despland E, Collett M, Simpson SJ (2000) Small-scale processes in de-
sert locust swarm formation: How vegetation patterns influence
gregarization. OIKOS 88: 652-662. https://doi.org/10.1034/j.1600-
0706.2000.880322.x

Duran GR, Garcia CG (2010) Distribucidn espacial de la vegetacion. In:
Duran R, Méndez M (Eds) Biodiversidad y Desarrollo humano en
Yucatan. CICY, PPD-FMAM, CONABIO, SEDUMA, México, 131-135.

Flores GF (2011) Las plagas de langosta en el area maya: Ambiente e histo-
ria de una calamidad en la época prehispanica. Peninsula 2: 27-46.

Galindo MMG, Ibarra ZE, Hipolito CG, Aldama AC (2014) Sistema de
alerta temprana para el control de la Langosta Centroamericana
(Schistocerca piceifrons piceifrons (Walker) en la Peninsula de Yuca-
tan aplicando modelos multicriterio e imagenes de satélite NOAA_
AVHRR. In: Galindo MMG, Contreras SC, Ibarra ZE (Eds) La plaga de
la langosta, Schistocerca piceifrons piceifrons (Walker) una visi6n mul-
tidisciplinaria desde la perspectiva del desastre fitosanitario en Mé-
xico. Colecci6n Sanidad Vegetal Tomo 1. Vigilancia Epidemioldgica

Al

Fitosanitaria. Universidad Aut6noma de San Luis Potosi, México,
173-215.

Garcia QA (2012) La langosta, los mayas y el colonialismo en Yucatan,
México, 1883. Relaciones 129: 215-249.

Gay P-E, Lecog M, Piou C (2018) Improving preventive locust manage-
ment: insights from a multi-agent model. Pest Management Science
74: 46-58. https://doi.org/10.1002/ps.4648

Harvey AW (1983) Schistocerca piceifrons (Walker) (Orthoptera: Acrididae),
the swarming locust of tropical America: a review. Bulletin of
Entomological Research 73: 171-184. https://doi.org/10.1017/
S0007485300008786

Henschel JR (2015) Locust times - monitoring populations and outbreak
controls in relation to Karoo natural capital. Transactions of the Royal
Society of South Africa 2: 135-143. https://doi.org/10.1080/003591
9X.2015.1046974

Hernandez-Zul MI, Quijano-Carranza JA, Yanhez-Lopez R, Ocampo-
Velazquez RV, Torres-Pacheco I, Guevara-Gonzalez RG, Castro-
Ramirez AE (2013) Dynamic simulation model of the Central
American Locust Schistocerca piceifrons (Orthoptera: Acrididae). Florida
Entomologist 4: 1274-1283. https://doi.org/10.1653/024.096.0405

Hunter DM, Spurgin PA (1999) Origin of outbreaks of Austracris guttulosa
(Walker) in Australia and Nomadacris septemfasciata (Serville) in
Southern Africa and its implications for their management. Inter-
national Journal Tropical Insect Science 19: 307-311. https://doi.
org/10.1017/$1742758400018932

INEGI (2011) Uso de suelo y vegetaci6n. Conjunto de datos vectoriales
de la carta de uso de suelos y vegetacion. Escala 1: 250, 000. Serie V.
Quintana Roo, México, 1-200.

Infante GZ, Zarate LGP (1990) Métodos estadisticos. Un enfoque interdis-
ciplinario. Editorial Trillas, México, 643 pp.

Kemp WP (1987) Probability of outbreak for rangeland grasshop-
pers (Orthoptera: Acrididaes) in Montana: Application of Marko-
vian principles. Journal of Economic Entomology 6: 1100-1105.
https://doi.org/10.1093/jee/80.6.1100

Lockwood JA (2015) Locust: An introduction. In: Shroder J, Sivanpillai R
(Eds) Biological and Environmental Hazards, Risk and Disasters.
Elsevier, 63-66. https://doi.org/10.1016/B978-0-12-394847-2.00004-8

Lazar M, Piou C, Doumandji-Mitiche B, Lecog M (2016) Importance of
solitarious desert locust population dynamics: lessons from historical
survey data in Algeria. Entomologia Experimentalis et Applicata 3:
168-180. https://doi.org/10.1111/eea.12505

Lecog M (2021) Scientific knowledge is no more the weakest link to fight
the locust plague. ACADEMIA Letters 1409. https://doi.org/10.20935/
AL1409

Ley Federal de Sanidad Vegetal (1994) Camara de Diputados del Congreso
de la Union. México, 36 pp. http://www.diputados.gob.mx/LeyesBib-
lio/ref/Ifsv.htm [Accessed on 01.12.2021]

Marquez DA (1963) La Lucha Contra la Langosta en México. Colegio de
Ingenieros Agronomos de México A.C, México, 220 pp.

Meynard C, Lecoq M, Chapuis M-P, Piou C (2020) On relative role of
climate change and management in the current desert locust outbreak
in East Africa. Global Change Biology 26(7): 3753-3755. https://doi.
org/10.1111/gcb.15137

Naumova EN, MacNeill IB (2005) Signature-forecasting and early out-
break detection system. Envirometrics 16: 749-766. https://doi.
org/10.1002/env.734

Ortiz YI, Zuleta MC (2020) Asuntos de vecinos: Langosta, defensa agricola
y la construcci6n de la sanidad vegetal en México y Centroamérica, sig-
lo XX. Historia Mexicana 70(1): 313-373. https://doi.org/10.24201/
hm.v70i1.4081

Pedgley DE (1981) Desert Locust Forecasting Manual. Vol. 1. Centre for
Overseas Pest Research, London, UK, 268 pp.

Peng W, Ling MN, Zhang D, Zhou Q, Yue X, Ching KS, Yang H, Guan R,
Chen H, Zhang X, Wang Y, Wei Z, Suo C, Peng Y, Yang Y, Shiung LS,
Sonne C (2020) A review of historical and recent locust outbreaks:
Links to global warming, food security and mitigation strategies.
Environmental Research 191: 110046. https://doi.org/10.1016/j.
envres.2020.110046

JOURNAL OF ORTHOPTERA RESEARCH 2023, 32(1)

A2

Pereyra CE (1991) Biologia, Ecologia y Control de la Langosta Schistocerca
piceifrons (Walker) (Orthoptera: Acrididae) en el Estado de Yucatan.
Tesis, Universidad Autonoma de Chapingo. México, 75 pp.

Poot-Pech MA, Ruiz-Sanchez E, Gamboa-Angulo M, Ballina-Gomez
H, Reyes-Ramirez A (2018) Population fluctuation of Schistocerca
piceifrons piceifrons (Orthoptera: Acrididae) in the Yucatan Peninsula
and its relation with the environmental conditions. Biologia Tropical
66: 403-414. https://doi.org/10.15517/rbt.v66i1.29502

Prospatin P (2012) Multisensor monitoring system for assessment of locust
hazard risk in the Lake Balkhash Drainage Basin. Environmental
Management 50: 1234-1246. https://doi.org/10.1007/s00267-012-
9950-2

R Core Team (2019) R: A language and environment for statistical comput-
ing. R Foundation for Statistical Computing, Vienna, Austria. https://
www.R-project.org/

Retana BJ (2000) Relacion entre algunos aspectos climaticos y el desarrollo
de la langosta centroamericana Schistocerca piceifrons piceifrons en el
Pacifico Norte de Costa Rica durante la fase calida del fen6dmeno El
Nino-Oscilacién Sur (ENOS). Tépicos Meteoroldgicos y Oceanogra-
ficos 2: 73-87.

Rodriguez VJ (2000) Historia de la Fitosanidad en México siglo XXI.
SAGAR, México, 180 pp.

Roffey J, Popov GR (1968) Environmental and behavioural process-
es in a desert locust outbreak. Nature 219: 446-450. https://doi.
org/10.1038/219446a0

Salih AA, Baraibar M, Mwangi KK, Artan G (2020) Climate change and
locust outbreak in East Africa. Nature Climate Change 10: 584-585.
https://doi.org/10.1038/s41558-020-0835-8

SENASICA (2019) Avanzan acciones de control contra la plaga de lango-
sta en 10 entidades del pais. https://www.gob.mx/senasica/prensa/
avanzan-acciones-de-control-contra-la-plaga-de-langosta-en- 10-enti-
dades-del-pais [Accessed on 19.08.2021]

SENASICA-DGSV (2016) Ficha Técnica Langosta Centroamericana
Schistocerca piceifrons piceifrons (Walker, 1870) (Orthoptera: Acrididae).
Servicio Nacional de Sanidad, Inocuidad y Calidad Agroalimentaria
— Direcci6n General de Sanidad Vegetal - Centro Nacional de Refer-
encia Fitosanitaria - Grupo Especialista Fitosanitario. México, 18 pp.

Sharma A (2014) Locust control management: Moving from traditional to
new technologies — An empirical analysis. Entomology, Ornithology
& Herpetology: Current Research 4: 1. https://doi.org/10.4172/2161-
0983.1000141

Showler AT (2002) A summary of control strategies for desert locust,
Schistocerca gregaria. Agriculture, Ecosystems and Environment 90:
97-103. https://doi.org/10.1016/S0167-8809(01)00167-0

Showler AT (2019) Desert locust control: The effectiveness of proactive in-
terventions and goal of outbreak prevention. American Entomologist
65: 180-191 https://doi.org/10.1093/ae/tmz020

Simpson SJ, Sword GA (2008) Locusts. Current Biology 9: 364-366.
https://doi.org/10.1016/j.cub.2008.02.029

M.A. POOT-PECH

Skinner KM, Child RD (2000) Multivariate analysis of the factors
influencing changes in Colorado grasshopper abundances. Journal
of Orthoptera Research 9: 103-109. https://doi.org/10.2307/3503640

Steinbauer MJ (2011) Relating rainfall and vegetation greenness to biology of
spur-throated and Australian plague locusts. Agricultural and Forest Ento-
mology 13: 205-218. https://doi.org/10.1111/j.1461-9563.2010.00518.x

Sultana R, Kumar S, Samejo AA, Soomro S, Lecog M (2021) The 2019-2020
upsurge of the desert locust and its impact in Pakistan. Journal of Or-
thoptera Research 30: 145-154. https://doi.org/10.3897/jor.30.65971

Sword G, Lecoq M, Simpson S (2010) Phase polyphenism and preven-
tive locust management. Journal of Insect Physiology 56: 949-957.
https://doi.org/10.1016/j.jinsphys.2010.05.005

Symmons PM (1992) Strategies to combat the desert locust. Crop Protection
3: 206-212. https://doi.org/10.1016/0261-2194(92)90038-7

Symmons PM, Cressman K (2001) 1. Biology and Behavior. Desert Locust
Guidelines. FAO, Rome, 25 pp.

Trujillo GP (1975) El problema de la langosta Schistocerca paranensis
Burm. Sociedad de Geografia y Estadistica de Baja California, Tijuana,
México, 77 pp.

Van Der Werf W, Woldewahid G, Van Huis A, Butrous M, Sykora K (2005)
Plant communities can predict the distribution of solitarious desert
locust Schistocerca gregaria. Journal of Applied Ecology 42: 989-997.
https://doi.org/10.1111/j.1365-2664.2005.01073.x

Van Huis A, Cressman K, Magor J (2007) Preventing desert locust plagues:
Optimizing management interventions. Entomologia Experimeta-
lis et Applicata 122: 191-214. https://doi.org/10.1111/j.1570-
7458.2006.00517.x

Walker F (1870) Catalogue of the Specimens of Dermaptera Saltatoria in
the Collection of the British Museum, London 3: 578. https://archive.
org/details/catalogueofspeci03britrich

WMO-FAO (2016) Weather and Desert Locust. WMO-No. 175, Geneva,
Switzerland, 38 pp.

Xiang-Hui Y, Huai L, Jin-Jun W, Zhi-Mo Z, Fangneng H, Deng-Fa C (2010)
Population forecasting model of Nilaparvata lugens and Sogotella
furcifera (Homoptera: Delphacidae) based on Markov Chain Theory.
Environmental Entomology 6: 1737-1743. https://doi.org/10.1603/
EN10018

Yam IO, Zuleta MC (2020) Asuntos de vecinos: langosta, defensa agricola
y la construccion de la sanidad vegetal en México y Centroamérica,
Siglo XX. Historia Mexicana 70: 313-373. https://doi.org/10.24201/
hm.v70i1.4081

Zhang L, Lecog M, Latchininsky A, Hunter D (2019) Locust and grass-
hopper management. Annual Review of Entomology 64: 15-34.
https://doi.org/10.1146/annurev-ento-011118-112500

Zimmerman K, Lockwood JA, Latchininsky AV (2004) A spatial,
Markovian model of rangeland grasshopper (Orthoptera: Acrididae)
population dynamics: Do long-term benefits justify suppression of
infestations. Environmental Entomology 2: 257-266. https://doi.
org/10.1603/0046-225X-33.2.257

JOURNAL OF ORTHOPTERA RESEARCH 2023, 32(1)