Local stability implies global stability for the 2-dimensional Ricker map ´ Ferenc A. Bartha†‡ , Abel Garab‡ and Tibor Krisztin∗‡? † CAPA group, Department of Mathematics, University of Bergen, Bergen, Norway ‡ Bolyai

arXiv:1209.2406v1 [math.DS] 11 Sep 2012

? Analyis

Institute, University of Szeged, Szeged, Aradi v. tere 1, 6720, Hungary

and Stochastics Research Group of the Hungarian Academy of Sciences,, Bolyai Institute, University of Szeged

Abstract Consider the difference equation xk+1 = xk eα−xn−d where α is a positive parameter and d is a nonnegative integer. The case d = 0 was introduced by W.E. Ricker in 1954. For the delayed version d ≥ 1 of the equation S. Levin and R. May conjectured in 1976 that local stability of the nontrivial equilibrium implies its global stability. Based on rigorous, computer aided calculations and analytical tools, we prove the conjecture for d = 1. Keywords: global stability; rigorous numerics; Neimark–Sacker bifurcation; graph representations; interval analysis; discrete-time single species model 2010 Mathematics Subject Classification: 39A30, 39A28, 65Q10, 65G40, 92D25

1

Introduction

In 1954, Ricker [21] introduced the difference equation xk+1 = xk eα−xk

(1.1)

with a positive parameter α to model the population density of a single species with non-overlapping generations. The function R1 : R 3 x 7→ xeα−x ∈ R is called the 1-dimensional Ricker map. R1 has two fixed points: 0 and α. It is not difficult to show that x = α is stable if and only if 0 < α ≤ 2, and, for 0 < α ≤ 2, x = α attracts all points from (0, ∞); or equivalently, the equilibrium x = α of equation (1.1) is globally stable provided it is locally stable. In 1976, Levin and May [14] considered the case when there are explicit time lags in the density dependent regulatory mechanisms. This leads to the difference-delay equation of order d + 1: xk+1 = xk eα−xk−d , ∗ Corresponding

author. Email: [email protected]

1

(1.2)

where d is a positive integer. The map Rd+1 : Rd+1 3 (x0 , . . . , xd )T 7→ (x1 , . . . , xd , xd eα−x0 )T ∈ Rd+1 is called the (d + 1)-dimensional Ricker map; here T denotes the transpose of a vector. Rd+1 has 2 fixed points in Rd+1 : (0, . . . , 0)T and (α, . . . , α)T . Levin and May [14] conjectured in 1976 that local stability of the fixed point (α, . . . , α)T ∈ Rd+1 implies its global stability in the sense that all points d+1 are attracted by (α, . . . , α)T . As far as we know, the conjecture is still open from Rd+1 + := (0, ∞)

for all integers d ≥ 1. The aim of the present paper is to prove the conjecture for d = 1. Levin and May’s conjecture and many other numerical and analytical studies suggested the folk theorem that ‘The local stability of the unique positive equilibrium of a single species model implies its global stability.’ This claim was recently disproven by a counterexample of Jim´enez L´opez [11, 12] on global attractivity for Clark’s equation [4] when the delay is at least 3. Liz, Tkachenko and Trofimchuk [17] proved that if 0 0 such that kx − pk< δ implies k f n (x) − pk< ε for all n ∈ N. We say that the fixed point p attracts the region U ⊆ D f if for all points u ∈ U, k f n (u) − pk→ 0 as n → ∞. The fixed point p is globally attracting if it attracts all of D f and is globally stable if it is locally stable and globally attracting.

2

The dynamics in the first quadrant (α)

In this section we construct compact squares Si (α) Si

attracts all points of

R2+

(α)

(α)

⊂ R2+ , i ∈ N0 around α so that Fα (Si ) ⊆ Si

and

for all i ∈ N0 and α ∈ (0, 1]. Hence an elementary proof of Theorem 1.1

is obtained for 0 < α ≤ 0.5. Recall Fα (R2+ ) ⊆ R2+ . We can illustrate the image (xk+1 , yk+1 ) of (xk , yk ) as first going horizontally from (xk , yk ) to the diagonal, proceeding upwards if 0 < xk < α, otherwise downwards vertically until we reach the value yk+1 = yk eα−xk . This is shown on Figure 1.

4

Figure 1: Dynamics for x > 0, y > 0 2 Let (x0 , y0 ) ∈ R2+ and 0 < α ≤ 1 be given. Define the sequence (xk , yk )∞ k=0 ∈ R+ by

(xk+1 , yk+1 ) = Fα (xk , yk ), k ∈ {0, 1, . . .}. Consider the following cases: I. 0 < x0 ≤ α ≤ y0 Clearly we have α ≤ x1 ≤ y1 and max{x0 , y0 , x1 , y1 } ≤ y1 . II. α ≤ x0 ≤ y0 In this case α ≤ x1 and y1 ≤ x1 with max{x0 , y0 , x1 , y1 } ≤ y0 . We distinguish two cases depending on y1 ≤ α or not. III. α ≤ y0 ≤ x0 We obtain α ≤ y0 = x1 and y1 ≤ y0 = x1 . During the consequent iterations yi+1 ≤ yi = xi+1 is satisfied as long as α ≤ xi stays true. If α ≤ xi for all i, then yi > 0 for all i, and both (yi )∞ i=0 and (xi )∞ i=0 are monotonically decreasing, and converge. The only possibility is (xk , yk ) → (α, α), since the other fixed point is at (0, 0). Otherwise there is a minimal N > 0 such that 0 < yN ≤ xN < α is satisfied. We note that 0 < yN−1 < α ≤ xN−1 is true. We have maxi∈{0,...,N} {xi , yi } < x0 . IV. 0 < y0 < α ≤ x0 Obviously 0 < y1 ≤ x1 < α, and max{x0 , y0 , x1 , y1 } ≤ x0 . V. 0 < y0 ≤ x0 ≤ α Here we have 0 < x1 ≤ α, 0 < x1 ≤ y1 and max{x0 , y0 , x1 , y1 } ≤ y1 . We distinguish two cases depending on α ≤ y1 or not.

5

VI. 0 < x0 ≤ y0 ≤ α Then x1 ≤ α and x1 ≤ y1 . Now xi+1 = yi ≤ yi+1 is satisfied as long as xi ≤ α stays true. If ∞ xi ≤ α for all i, then yi > 0 for all i, and both (yi )∞ i=0 and (xi )i=0 are monotonically increasing,

and converge. The only possibility is (xk , yk ) → (α, α), since the other fixed point is at (0, 0). Otherwise there is a minimal N > 0 such that α < xN ≤ yN is satisfied. We note that 0 < xN−1 ≤ α < yN−1 . We have maxi∈{0,...,N} {xi , yi } < yN . We conclude that for any (x0 , y0 ) ∈ R2+ the sequence ((xk , yk ))∞ k=0 converges to (α, α) or enters the (α)

(α)

triangle H(h0 ) = {(x, y) : h0

(α)

< y ≤ x ≤ α}, with h0

= 0, that is the region denoted by V in

Figure 2. We mark the possible transitions with arrows, if it is solid, it refers to transition in one step, otherwise it is possible to have multiple iterations before entering the next region.

Figure 2: Dynamics for x > 0, y > 0 (α)

Assume now that (x0 , y0 ) ∈ H(h0 ). Either the sequence converges to (α, α) or there exists an N > 1 such that (xN−1 , yN−1 ) ∈ V ∪V I, (xN , yN ) ∈ I and (xN+1 , yN+1 ) ∈ II. So the following stands: (α)

yN+1 = yN eα−xN = yN−1 e2α−xN −xN−1 ≤ αe2α−2h0 . (α)

(α)

(α)

(α)

This implies that (xN+1 , yN+1 ) ∈ G(g0 ), where G(g0 ) = {(x, y) : α ≤ x ≤ y ≤ g0 }, with g0 αe

(α) 2α−2h0

=

. Now there exists M ≥ N + 2 such that (xM−1 , yM−1 ) ∈ II ∪ III, (xM , yM ) ∈ IV and

(xM+1 , yM+1 ) ∈ V . We get the following inequality: (α)

yM+1 = yM−1 e2α−xM −xM−1 ≥ αe2α−2g0 . (α)

(α)

Therefore (xM+1 , yM+1 ) ∈ H(h1 ), with h1 (α)

(α)

set G(g1 ), with g1

(α)

= αe2α−2g0 . Similarly, the sequence will enter the

(α)

(α)

= αe2α−2h1 . Repeating this argument we get two sequences (hi )∞ i=0 and

6

(α)

(α)

(gi )∞ i=0 defined recursively by h0 It is easy to see that (α) gi

&

(α) g∞

(α) (hi )

(α)

= 0, hi

increases and

(α)

(α)

= αe2α−2gi−1 for i ≥ 1, and gi = αe2α−2hi

(α) (gi )

decreases. We have the limits

(α) hi

%

(α) h∞

for i ≥ 0. ≤ α and

≥ α. Define (α)

Si

(α)

= {(x, y) : hi

(α)

(α)

≤ x ≤ gi , hi

(α)

≤ y ≤ gi }.

(2.1)

We sum our observations in the following Lemma: Lemma 2.1. For every α ∈ (0, 1] and for every i ∈ N0 (α)

(α)

1. Fα (Si ) ⊆ Si , (α)

2. Si

attracts all points of R2+ .

(α)

If S∞ = {(α, α)}, then we have already established that the fixed point is globally attracting. (α)

(α)

This is not the case however for α > 0.5. For a given α, we can view the sequences (hi ) and (gi ) as even and odd iterates of the function τα (t) = αe2α−2t , (α)

starting from t = 0. That is, h0

(α)

= τα0 (0), g0

(2.2) (α)

= τα1 (0), h1

(α)

= τα2 (0), g1

= τα3 (0), . . . . The non-

rigorous bifurcation diagram on Figure 3 shows that the unique fixed point α of τα is stable for 0 < (α)

α < 0.5, it is unstable for α > 0.5, and there is an attracting 2-cycle for α > 0.5. Thus S∞ = {(α, α)} (α)

(α)

can be expected only for α ∈ (0, 0.5). For α > 0.5 the two points of the 2-cycle give h∞ and g∞ .

Figure 3: Bifurcation diagram of τα

Proposition 2.2. If 0 < α ≤ 0.5, then for every (x, y) ∈ R2+ we have Fαn (x, y) → (α, α) as n → ∞. Proof. Observe that τα0 (t) = −2τα (t), τα (t) > 0, and if t > α and α ∈ (0, 0.5], then 2τα (t) = 2α ≤ 1 is satisfied. Since

d2 (τα (τα (t)) − t) = 8τα (τα (t))τα (t) (2τα (t) − 1) , dt 2

7

(2.3)

therefore τα (τα (t)) − t is a concave function for t > α. The first derivative at α ∈ (0, 0.5] is d (τα (τα (t)) − t) = 4τα (τα (α))τα (α) − 1 = 4α 2 − 1 ≤ 0. dt t=α These imply that

d dt

(2.4)

(τα (τα (t)) − t) < 0 for all t > α. Thus the only zero of τα (τα (t)) −t in t ∈ [α, ∞) (α)

(α)

(α)

is α. This gives us that g∞ = α which implies that h∞ = α and consequently S∞ = {(α, α)} in this parameter region. We assume α ∈ [0.5, 1] in the remaining part of the paper.

3

Attracting neighbourhood In this section we are going to give ε(α) > 0, such that

Let us consider map Fα .

infα∈[0.5,1] ε(α) > 0 and K(α; ε(α)) belongs to the basin of attraction of α for α ∈ [0.5, 1], that is, Fαn (x0 , y0 ) → α as n → ∞ for all (x0 , y0 )T ∈ K(α; ε(α)). Introducing the new variables u = x − α, v = y − α, map Fα can be written in the form ! ! u u 7→ A(α) + fα (u, v), v v

(3.1)

where the linear part is A(α) =

0

1

!

−α 1

and the remainder is fα (u, v) =

!

0 v(e−u − 1) + α(e−u − 1 + u)

.

√

The eigenvalues of A(α) are µ1,2 (α) = 1±i 24α−1 ∈ C, and the corresponding complex eigenvectors  √ T  √ T 1−4α are q1,2 (α) = 1∓ 2α , 1 = 1∓i 2α4α−1 , 1 ∈ C2 , respectively for α > 41 . Let q(α) = q1 (α) and µ(α) = µ1 (α). Let p(α) ∈ C2 denote the eigenvector of A(α)T corresponding to µ(α) such that hp(α), q(α)i = 1,. This leads to √  T iα 4α − 1 + i √ √ p(α) = − , . 4α − 1 2 4α − 1 We introduce a complex variable z = z(u, v, α) = hp(α), (u, v)T i =

  1 i(v − 2uα) v− √ . 2 −1 + 4α

We also have an explicit formula for (u, v)T in terms of z, which reads as √  T Rez + 4α − 1Imz T , 2Rez . (u, v) = zq(α) + zq(α) = α Our original system (3.1) is now transformed into the complex system D E z 7→ G(z, z, α) = p(α), A(α)(zq(α) + zq(α)) + fα (zq(α) + zq(α)) = µ(α)z + g(z, z, α),

8

(3.2)

(3.3)

(3.4)

where g is a complex valued smooth function of z, z and α, defined by (A.1). It can also be seen that for fixed α, g is an analytic function of z and z and the Taylor expansion of g with respect to z and z has only quadratic and higher order terms. That is, g(z, z, α) =

where gkl (α) =



∂ k+l g(z, z, α) ∂ zk ∂ zl z=0

1 gkl (α)zk zl , k! l! 2≤k+l

∑

with k, l = 0, 1, . . . ,

for k + l ≥ 2, k, l = {0, 1, . . . }.

Proposition 3.1. Let α ∈ [0.5, 1) be fixed and ( r √ ) 1 4α − 1 9(4α − 1)(1 − α) ε(α) = min , . √ √ 20 2 + α 20(1 + 2 α) 2 + α  Then (x, y)T ∈ R2 : |x − α|< ε(α), |y − α|< ε(α) belongs to the basin of attraction of the fixed point α of Fα . Proof. Let us study the map in the form (3.4). First note that (3.3) easily implies inequalities 2 |u|≤ √ |z| and |v|≤ 2|z|. α

(3.5)

Assuming |u|< 1/10 and |v|< 1/10 and applying the inequalities |e−u − 1| ≤ e1/10 |u|≤ |e−u − 1 + u|

≤

2 e1/10 u2

≤

10 9 |u|

and

5 2 9u

we obtain the following estimations: D E  |g(z, z, α)| = p(α), fα zq(α) + zq(α) √ 4α − 1 + i −u v(e − 1) + α(e−u − 1 + u) √ = 2 4α − 1 ! r r 1/10
α α e |v||e−u − 1|+α|e−u − 1 + u| ≤ |uv|e1/10 + α u2 ≤ 4α − 1 4α − 1 2 r √
5 4(1 + 2 α) 2 5 α ≤ u2 + 2|uv|) ≤ · p |z| . 9 4α − 1 9 α(4α − 1) √ Now, since |µ(α)|= α, hence ! √ √ 5 4(1 + 2 α) |G(z, z, α)|≤ α+ ·p |z| |z|< |z| 9 α(4α − 1) provided that |z|6= 0 is so small that |u|< (3.5) imply that |z|
|zn |> . . . ≥ 0 with |zn |→ c > 0 as n → ∞. Since G is continuous we have that max|z|=c |G(z, z, α)|< c and consequently |zk |< c also holds if k is large enough which is a contradiction.

9

From equation (3.2) one obtains that if |u|< δ , |v|< δ , then r α(2 + α) |z|< δ . 4α − 1

(3.7)

From (3.7) we infer that if |u|< ε(α) and |v|< ε(α) then |z(u, v, α)|< εG (α) which completes our proof. Note that ε(α) goes to 0 as α goes to 1. This means that when α is close to 1, then the constructed region K(α; ε(α)) becomes very small. Thus it is impossible to show by interval arithmetic tools that every trajectory enters into it eventually. Nevertheless ε(α) might be used in the case α ∈ [0.5, 0.999]. However, our following method is not only capable to give an attracting neighbourhood for all α ∈ [0.999, 1], whose size is independent of the concrete value of the parameter, but it also serves a better estimation of the attracting region even if we assume only α ∈ [0.875, 1]. The main results of the section are the following two propositions. Here, we only present the proof of Proposition 3.3. The whole argument can be repeated to get a universal attracting neighbourhood when only [0.875, 1] is assumed. The differences only appear in concrete values in the given estimations.  Proposition 3.2. For all fixed α ∈ [0.875, 1], (x, y)T ∈ R2 : |x − α|< 1/37, |y − α|< 1/37 belongs to the basin of attraction of the fixed point α of Fα .  Proposition 3.3. For all fixed α ∈ [0.999, 1], (x, y)T ∈ R2 : |x − α|< 1/22, |y − α|< 1/22 belongs to the basin of attraction of the fixed point α of Fα . Proof. We follow the steps of finding the normal form of the Neimark–Sacker bifurcation, according to Kuznetsov [13]. In our calculations and estimations we use symbolic calculations and built in symbolic interval arithmetic tools of Wolfram Mathematica v. 7 or 8. According to Kuznetsov [13], we are aiming to transform system (3.4) to a system which takes the following form. w 7→ µ(α)w + c1 (α)w2 w + R2 (w, w, α),

(3.8)

where c1 and R2 are smooth, real functions such that for fixed α, R2 (w, w, α) = O(|w|4 ). We are going to show that there exists ε0 > 0 such that for all |w|< ε0 and α ∈ [0.999, 1], µ(α)w + c1 (α)w2 w + R2 (w, w, α) < |w| holds, which implies that Bε0 belongs to the basin of attraction of the fixed point 0 of the discrete dynamical system generated by (3.8). From this, we shall be able to show that the fixed point α of Fα 1 attracts all points of K(α; 22 ).

To be more precise, for a fixed α, we are looking for a function hα : C → C, which is invertible in a neighbourhood of 0 ∈ C and which is such that in the new coordinate w = h−1 α (z), our map (3.4) takes the form 2 w 7→ h−1 α (G(hα (w), hα (w), α)) = µ(α)w + c1 (α)w w + R2 (w, w, α),

10

(3.9)

where R2 (w, w, α) = O(|w|4 ) for fixed α. According to [13], such a function can be found in the form hα (w) = w +

h20 (α) 2 h02 (α) 2 h30 (α) 3 h12 (α) 2 h03 (α) 3 w + h11 (α)ww + w + w + ww + w . (3.10) 2 2 6 2 6

Clearly, hα has an inverse in a small neighbourhood of 0 ∈ C. A formula for h−1 α can be given in the form h−1 α (z) = hinv,α (z) + R3 (z, α), where hinv,α (z) = z +

∑

akl (α)zk zl

2≤k+l≤4

and R3 (z, α) = O(|z|5 ). The coefficients can be obtained by substituting w = h−1 (z) into z = hα (w) and equating the coefficients of the same type of terms up to the fourth order. The result for hinv,α in terms of h20 (α), . . . , h03 (α) is given in (A.14). The coefficients h20 (α), . . . , h03 (α) are determined such that h−1 α (G(hα (w), hα (w), α)) has the form µ(α) + c1 (α)w2 w plus at least fourth order terms in w, that is, the transformation kills all second and third order terms with one exception. This requires the condition   µ(1) k 6= 1 for all k ∈ {1, 2, 3, 4}. |µ(1)| As µ(1) = 12 + i

√ 3 2 ,

this is clearly satisfied. Formulae (A.15)-(A.20) contain the obtained results.

In which region is the transformation valid? We are going to show that hα is injective on B1/3 ⊂ C and that h−1 α is defined on B1/5 . Let us suppose that z ∈ C is fixed and h20 (α), . . . , h03 (α) are given for a fixed α ∈ [0.999, 1]. Let Hα,z : C 3 w 7→ w + z − hα (w) ∈ C. By this notation, Hα,z (w) = w holds if and only if hα (w) = z. Now, we have the following |Hα,z (w1 ) − Hα,z (w2 )| = |w1 − hα (w1 ) − w2 + hα (w2 )|   |h20 (α)| |h02 (α)| ≤|w1 − w2 |· + |h11 (α)|+ (|w1 |+|w2 |) + 2 2   
|h30 (α)| |h12 (α)| |h03 (α)| 2 2 + + |w1 | +|w1 ||w2 |+|w2 | . 6 2 6 If |h20 (α)|/2 + |h11 (α)|+|h02 (α)|/2 < δ1 , |h30 (α)|/6 + |h12 (α)|/2 + |h03 (α)|/6 < δ2 , |w|≤ δ3 and |z|≤ δ4 hold, then we have |Hα,z (w1 ) − Hα,z (w2 )| ≤ |w1 − w2 |(2δ1 δ3 + 3δ1 δ32 ) and |Hα,z (w)| ≤ δ4 + δ1 δ32 + δ2 δ33 . By interval arithmetics we obtain that the first two inequalities are fulfilled if δ1 = 0.76 and δ2 = 0.52. Now, if we choose δ3 =

1 3

and δ4 =

1 5

we obtain that Hα,z : B1/3 → B1/3 is a contraction. Hence

11

for all fixed z ∈ B1/5 there exists exactly one w = w(z) ∈ B1/3 such that Hα,z (w(z)) = w(z), that is hα (w(z)) = z. This means that h−1 α can be defined on B1/5 . The obtained estimations on hα are going to be used in the sequel. These were |h20 (α)|/2 + |h11 (α)|+|h02 (α)|/2 < 0.76, |h30 (α)|/6 + |h12 (α)|/2 + |h03 (α)|/6 < 0.52 and in particular |w|−0.76|w|2 −0.52|w|3 < |hα (w)|< |w|+0.76|w|2 +0.52|w|3

(3.11)

for all α ∈ [0.999, 1]. Let H = {(α, w) ∈ R × C : α ∈ [0.999, 1], w 6= 0 and |w|< 1/20}. In the sequel, we shall always assume that (α, w) ∈ H. From this assumption and inequality (3.11) we readily get that |w|< 1.05|hα (w)| from which we get in particular that |z|= |hα (w)|< 1/19. Our goal now is to give ε0 ∈ (0, 1/20], independent of α such that for every α ∈ [0.999, 1], if    0 < |w|< ε0 , then h−1 G h (w), h (w), α < |w| holds which guarantees that Bε0 belongs to the α α α basin of attraction of the fixed point 0 of the discrete dynamical system generated by (3.9). For, we turn our attention to the estimation of function R2 in (3.9). First, we go back to (3.4). Let us consider g(z, z, α) =

gkl (α) k l z z + R1 (z, z, α). k+l=2,3 k! l!

∑

The explicit formulae for g20 (α), . . . , g03 (α) can be found in equations (A.2)–(A.8). By interval arithmetics, one may obtain that for all α ∈ [0.999, 1], p |g02 (α)| 1 − α + α(2 + α) |g20 (α)| + |g11 (α)|+ = p < 1.01, 2 2 α(−1 + 4α)

|g30 (α)| |g21 (α)| |g12 (α)| |g03 (α)| + + + = 6 2 2 6

s

(6 + α) + 9α(4α − 1)

s

2 + (α − 2)α < 1.09. α 2 (4α − 1)

(3.12)

(3.13)

(α) k l We also have that R1 (z, z, α) = ∑k+l=4 gklk!l! z z + R˜1 (z, z, α), where R˜1 (z, z, α) = O(|z|5 ) for fixed

α. For explicit formulae of the fourth order coefficients see equations (A.9)-(A.13) in the appendix. It is clear from equations (3.1), (3.3) and (3.4) that ! √ ∞ ∞ k k 4α − 1 − i (−u) (−u) R˜1 (z, z, α) = √ v∑ +α ∑ , 2 4α − 1 k=4 k! k=5 k! where u and v are defined by equation (3.3). Using 0 < |z|< 1/19 and (3.5) we have that |u|< 1/8, |v|< 1/8 and obtain ! r r   1/8 1/8 α α 8 e e 16 32 4 5 R˜1 (z, z, α) ≤ |v| |u| +α |u| < 2|z| + |z| |z|4 4α − 1 4! 5! 4α − 1 7 24α 2 α 3/2 120 s r   α 8 4 4 64 1 4 ≤ + |z| = |z|4 . 4α − 1 7 57α 2 285α 2 665 (4α − 1)α 3

12

Now for all (α, w) ∈ H with z = hα (w) we get that p √ gkl (α) k l 6 − 3α + α(12 + α) + 4 3 + α 2 ˜ p |R1 (z, z, α)| ≤ ∑ + R˜1 (z, z, α) z z + R1 (z, z, α) = 12 α 3 (−1 + 4α) k+l=4 k! l! p √ 4758 − 1995α + 665 α(12 + α) + 2660 3 + α 2 p . < 7980 α 3 (−1 + 4α) By interval arithmetics we obtain for all α ∈ [0.999, 1] that |R1 (z, z, α)| < 0.76|z|4 .

(3.14)

We also need a similar estimation on h−1 α (z) . Let us recall that z = hα (w) and w = h−1 α (z) = hinv,α (z) + R3 (z, α). As hinv,α (z) is a polynomial of z and z of degree four (see formulae (A.14)–(A.20)), we denote the coefficient corresponding to zk zl by hkl inv (α). Calculating these coefficients and using interval arithmetics, we obtain that for all α ∈ [0.999, 1], kl hinv < 0.76,

∑ k+l=2

∑

kl hinv < 1.06

and

k+l=3

kl hinv < 1.39.

∑

(3.15)

k+l=4

See formulae (A.21) – (A.27). Let us recall that for all (α, w) ∈ H, |w|< 1.05|hα (w)| holds. Now, for the fifth and higher order terms in R3 , first we give an estimation of type |R3 (hα (w), α)|< K3 |w|4 and then we get that |R3 (z, α)|< K3 1.054 |z|4 , with z = hα (w). From the definition of hinv,α , it follows that R3 (hα (w), α) = w − hinv,α (hα (w)) is a polynomial of w and w and it has only fifth and higher order terms. Let r3kl (α) denote the coefficient of R3 (hα (w), α) corresponding to wk wl . We use a bit rougher estimation for ∑5≤k+l rk+l (α)wk wl . Namely, first we 3

give the estimations |h20 (α)|< 1.01;

|h11 (α)|< 0.001;

|h30 (α)|< 0.89;

|h12 (α)|< 0.45;

|h02 (α)|< 0.51

(3.16)

|h03 (α)|< 0.89

for all α ∈ [0.999, 1]. Now, in R3 (hα (w), α) we replace w and w by |w|, hnm (α) by the estimates given in (3.16) (for 2 ≤ n + m ≤ 3, (n, m) 6= (2, 1)), and then we turn every − sign into + to get a real polynomial Rˆ 3 (|w|), with nonnegative coefficients rˆ3k (independent of α) corresponding to |w|k . If we use 0 < |w|< 1/20, then we get that

∑

|r3kl (α)||w|k+l
0 and the set K([α]; ε) is in the basin of attraction of α for every α ∈ [α]. Observe that, using subintervals with α + − α − ≤ 10−3 and the ε0 obtained from Propositions 3.1, 3.2 and 3.3, α + − α − < ε0 is satisfied. After each step in the main cycle in Algorithm 3, |V | is a rigorous enclosure of all the nonwandering points of Fα in [S] \ {α} for every α ∈ [α]. This is easy to see since vertices are removed for two possible reasons which are both checked in line 11 of Algorithm 3. First if v ∈ / T , then v does not contain non-wandering points for any Fα , α ∈ [α] as we have seen in the proof of the correctness of Algorithm 1. Second if v ⊆ [U] or F[α] (v) ⊆ [U] that is v is inside or mapped into the small attracting neighbourhood of every fixed point. Note that if a vertex is inside the basin of attraction of a fixed

23

point α, then it cannot contain any other non-wandering point of Fα , not even on the boundary. This is a similar criterion to what we have used in line 11 of Algorithm 2. The difference is that now we remove these vertices, consequently we do not have to collect them into a list. If the procedure ends in finite time, then we have established, that there are no other non-wandering points in [S], thus the fixed point is globally attracting for all α ∈ [α]. We state this as Proposition 5.1. If Algorithm 3 ends in finite time with input parameters ([α], 10−1 ), that is, after finite number of steps, |V |= 0/ is satisfied, then α is a globally attracting fixed point of the two dimensional Ricker-map Fα for every α ∈ [α]. We implemented our program in C++, using the CAPD Library [5] for rigorous computations, and the Boost Graph Library [22] for handling the directed graphs. The recursion number in Tarjan’s algorithm was very high, therefore we converted it into a sequential program, using virtual stack structures from the Standard Library in order to avoid overflows. Instead of simulating the Rickermap itself, we used its third iterate, the formula is still compact enough not to cause big overestimation in interval arithmetics and it considerably speeds up the calculations. As an example, we ran our program for the parameter slice [0.9, 0.90001], with δ = 10−1 as the initial diameter for the partition. We show the evolution of the enclosure during the first 8 iterations on Figure 5.

Figure 5: Enclosure after 1, 3, 4, 6, 8 steps The small rectangle is the attracting neighbourhood [U]. After 6 iterations the diameter of the partition is sufficiently small in order to have some boxes removed from the inside even though they are in directed cycles. This happens because they are contained in, or get mapped into the basin of attraction of the fixed point. We used different sizes for the parameter intervals and ran the computations on a cluster of the NIIF HPC centre at the University of Szeged (48 cores, 128 GB memory / cluster) parallelising it

24

with OpenMP. We summarise some technical details in Table 1. parameter

size of slices

# of CPU

max. memory

wall clock time

total time

[0.5, 0.875]

10−3

48

2.30 GB

1.2s

38.5s

[0.875, 0.95]

10−3

48

3.05 GB

2.4s

52.3s

[0.95, 0.99]

10−4

48

3.07 GB

33.7s

22m 5.9s

[0.99, 1]

10−5

20

65.30 GB

204m 42.6s

1800m 37.1s

Table 1: Resources used by the program Remark 5.2. Here wall clock time (real) refers to the actual running time of the process, whilst the total time (user + sys) is the sum of the time spent on individual CPUs. The complexity of the computations for some parameter slices is shown in Table 2. parameter slice

[S]

# iteration

max # of vertices

max # of edges

[0.875, 0.876]

[2.072e − 04, 5.049031]2

13

242

1, 676

[0.999, 0.99901]

[2.928e − 06, 7.369087]2

27

729, 528

4, 193, 329

[0.99999, 1]

[2.822e − 06, 7.389015]2

33

3, 105, 304

118, 751, 916

Table 2: Complexity of the computations The program ran successfully, thus we established that the fixed point is globally attracting for α ∈ [0.5, 1]. Combining this with Proposition 2.2 and Proposition 3.4, the proof of Theorem 1.1 is completed.

Acknowledgements The first author was supported by the Bergen Research Foundation. The second and third authors were sup´ ported by the Hungarian Scientific Research Fund, Grant No. K 75517. Abel Garab was also supported by the European Union and co-funded by the European Social Fund. Project title: “Broadening the knowledge base and supporting the long term professional sustainability of the Research University Centre of Excellence at ´ the University of Szeged by ensuring the rising generation of excellent scientists.” Project number: TAMOP4.2.2/B-10/1-2010-0012. The computations were performed on the HPC center provided by the Hungarian National Information Infrastructure Development Institute [10] at the University of Szeged. The authors thank Warwick Tucker from the CAPA group [3], Daniel Wilczak and Tomasz Kapela, members of the CAPD group [5] for useful suggestions and their help.

25

A

Appendix  √ √    √ z−iz −1+4α+¯z+i −1+4α z¯ 1 2α −i + −1 + 4α g(z, z¯, α) = √ −1 + e− (z + z¯) 2 −1 + 4α (A.1)   √ √ z+i −1+4α z¯ 2i¯z 2z − z−iz −1+4α+¯ 2α √ √ + + α −1 + e + 1 + i −1 + 4α i + −1 + 4α 1 3i g20 (α) = − + √ 2 2 −1 + 4α g11 (α) =

2(−1 + α) √ −1 + 4α + i −1 + 4α


√ iα 3i + −1 + 4α
√ g02 (α) = √ −1 + 4α i − 2iα + −1 + 4α
√ i 1 + α − i −1 + 4α √ g30 (α) = − α −1 + 4α √ −3i + 2iα + −1 + 4α √ g21 (α) = 2α −1 + 4α
√ 2 3i − 2iα + −1 + 4α g12 (α) = √
2 √ −1 + 4α i + −1 + 4α
√ 4iα 5i + −1 + 4α g03 (α) = √
3 √ −1 + 4α i + −1 + 4α
√ 8 −2 + α − 2i −1 + 4α g40 (α) = √
3 √ −1 + 4α −i + −1 + 4α
√ 2 −2i + iα + −1 + 4α
√ g31 (α) = − α −i + 4iα + −1 + 4α g22 (α) =

2(−2 + α)
√ α −1 + 4α + i −1 + 4α


√ 8 2 − α − i −1 + 4α g13 (α) = √
3 √ −1 + 4α i + −1 + 4α
√ 8α 7i + −1 + 4α g04 (α) = √
4 √ −1 + 4α i + −1 + 4α

26

(A.2)

(A.3)

(A.4)

(A.5)

(A.6)

(A.7)

(A.8)

(A.9)

(A.10)

(A.11)

(A.12)

(A.13)

 1  h20 (α)z2 1 3  + z 3h20 (α)2 −h30 (α)+3h11 (α)h02 (α) + z4 −15h20 (α)3 +10h20 (α)h30 (α) 2 6 24  2 − 30h11 (α)h20 (α)h02 (α) − 3h02 (α)h02 (α) + 4h11 (α)h03 (α) − 12h11 (α)h02 (α)h11 (α)  1  − h11 (α)z¯z + z2 3h11 (α)h20 (α) + h02 (α)h02 (α) + 2h11 (α)h11 (α) z¯ 2 1 3 + z −15h11 (α)h20 (α)2 + 4h11 (α)h30 (α) − 12h11 (α)2 h02 (α) + 3h12 (α)h02 (α) 6 − 6h02 (α)h20 (α)h02 (α) + h02 (α)h03 (α) − 12h11 (α)h20 (α)h11 (α) − 6h02 (α)h02 (α)h11 (α)  2 − 6h11 (α)h11 (α) + 3h11 (α)h12 (α) − 3h11 (α)h02 (α)h20 (α) z¯  h02 (α)¯z2 1  − + z 2h11 (α)2 − h12 (α) + h02 (α)h20 (α) + 2h02 (α)h11 (α) + h11 (α)h20 (α) z¯2 2 2 1 2 + z −12h11 (α)2 h20 (α) + 3h12 (α)h20 (α) − 3h02 (α)h20 (α)2 + h02 (α)h30 (α) + h03 (α)h02 (α) 4 − 9h02 (α)h11 (α)h02 (α) − 12h11 (α)2 h11 (α) + 4h12 (α)h11 (α) − 6h02 (α)h20 (α)h11 (α)

hinv,α (z) = z−

2

− 6h02 (α)h11 (α) + 2h02 (α)h12 (α) − 3h11 (α)h20 (α)h20 (α) − 3h02 (α)h02 (α)h20 (α)   1 −h03 (α) + 3h02 (α)h11 (α) + 3h02 (α)h20 (α) z¯3 − 6h11 (α)h11 (α)h20 (α) z¯2 + 6 1  + z −6h11 (α)3 + 6h11 (α)h12 (α) + h03 (α)h20 (α) − 9h02 (α)h11 (α)h20 (α) − 3h02 (α)2 h02 (α) 6 + 3h03 (α)h11 (α) − 18h02 (α)h11 (α)h11 (α) − 6h11 (α)2 h20 (α) + 3h12 (α)h20 (α)  2 − 3h02 (α)h20 (α)h20 (α) − 12h02 (α)h11 (α)h20 (α) − 3h11 (α)h20 (α) + h11 (α)h30 (α) z¯3 1  + 4h03 (α)h11 (α) − 12h02 (α)h11 (α)2 + 6h02 (α)h12 (α) − 3h02 (α)2 h20 (α) − 12h02 (α)2 h11 (α) 24  2 + 6h03 (α)h20 (α) − 18h02 (α)h11 (α)h20 (α) − 15h02 (α)h20 (α) + 4h02 (α)h30 (α) z¯4 (A.14)
√ 2 1 + 1 − 4α − α
√ h20 (α) = α 1 − 4α + i −1 + 4α h11 (α) =

2(−1 + α)
√ α −1 + 4α − i −1 + 4α


√ 2α 3i + −1 + 4α h02 (α) =
2
√ √ i + −1 + 4α −i + 4iα + α −1 + 4α

(A.15)

(A.16)

(A.17)

h30 (α) =      √ √ √ 16 −12 − 12i −1 + 4α + α 45 + 21i −1 + 4α + α −27 + 2α − 5i −1 + 4α  4 √  √   −1 (A.18) √ √ 1 + 1 − 4α −1 + 4α i + −1 + 4α + α −2i + α −5i + −1 + 4α

h12 (α) =       √ √ √ 4i + α −4 5i + 3 −1 + 4α + α 28i + 18 −1 + 4α + α −31i − 9 −1 + 4α +       √ √ √ α 9i + −1 + 4α α 3 −i + −1 + 4α + α 6i + α −13i − 15 −1 + 4α + (A.19)  −1 √ 2α 10i + −1 + 4α h03 (α) =

27

     √ √ √ 2 5i + 7 −1 + 4α + α −29i − 14 −1 + 4α + α 17i − iα + 3 −1 + 4α  √     √ √ i + −1 + 4α + α −2 7i + 6 −1 + 4α + α 70i + 46 −1 + 4α +     −1 √ √ α −6 23i + 10 −1 + 4α + α 73i − 4iα + 13 −1 + 4α

(A.20)

02 20 hinv (α) + h11 inv (α) + hinv (α) = q −2(−1 + α)α + α 3 (2 + α) +

s

α 5 (2 + α) −1 + α(4 + α)

! q −1 5 2 α (−1 + 4α)

30 12 03 hinv (α) + h21 inv (α) + hinv (α) + hinv (α) = 1 6

(A.21)

(A.22)

s 36 + 15α − 12α 2 + 4α 3 6 − 40α + 75α 2 − 21α 4 + 4α 5 + 4α 6 + 5 2 α (−2 + 7α + 4α ) (2 − 15α + 22α 2 + 23α 3 + 4α 4 )4α 6 s 12 − 54α + 66α 2 − 13α 3 + 4α 4 + 4α 5 1 + 6 α 3 (−1 + 13α − 57α 2 + 88α 3 − 18α 4 + 7α 5 + 4α 6 ) s 1 2 − 16α + 13α 2 + 124α 3 − 198α 4 + 15α 5 + 54α 6 + 9α 7 + 2 α 6 (−1 + 4α) (−1 + 4α + α 2 )2

s

|h40 inv (α)|= 1  −1140 + 10874α − 12660α 2 − 143073α 3 + 423177α 4 24 − 211261α 5 + 1356α 6 + 30869α 7 − 13766α 8 − 2238α 9 + 1489α 10 1 
3
− 21 2 + 256α 11 (1 − 4α)2 α 6 −1 + 4α + α 2 −1 + 5α − 2α 2 + α 3

(A.23)

|h31 inv (α)|=  −36 + 912α − 9354α 2 + 49026α 3 − 133548α 4 + 155248α 5 + 24851α 6 − 182342α 7 + 127014α 8 − 47122α 9 − 12543α 10 + 34528α 11 − 329α 12 + 1925α 13 + 2452α 14 1
− 1 
−1 2 + 361α 15 −1 + 5α − 2α 2 + α 3 2 6α 5 2 − 15α + 22α 2 + 23α 3 + 4α 4

(A.24)

|h22 inv (α)|=  −16 + 4α + 2240α 2 − 14868α 3 + 19782α 4 + 91297α 5 − 309731α 6 + 259610α 7 + 81080α 8 − 147591α 9 + 12815α 10 + 19871α 11 − 8236α 12 + 241α 13 + 2132α 14 
−1
− 1  + 361α 15 −1 + 5α − 2α 2 + α 3 2 4α 5 2 − 15α + 22α 2 + 23α 3 + 4α 4

(A.25)

|h13 inv (α)|= 1  36 − 840α + 7728α 2 − 35454α 3 + 84157α 4 − 96139α 5 + 39017α 6 6 + 15361α 7 − 22836α 8 + 10489α 9 + 5142α 10 − 397α 11 + 922α 12 + 529α 13  1  − 1
3  2 2 + 64α 14 α 9 −1 + 4α + α 2 2 − 17α + 35α 2 + 4α 3 − α 4 + 4α 5

28

(A.26)

|h04 inv (α)|= 1  36 − 396α + 1350α 2 − 1422α 3 + 318α 4 + 441α 5 − 145α 6 + 100α 7 24 1 
2
− 21 2 + 25α 8 α 6 −1 + 4α + α 2 1 − 9α + 22α 2 − 9α 3 + 4α 4

(A.27)

|r240 (α)|= 1  −432 + 2808α − 5016α 2 + 4692α 3 − 2584α 4 24 1 2 + 930α 5 − 210α 6 + 7α 7 + 27α 8 − 6α 9 + α 10 
2
− 12 α 4 −1 + 4α + α 2 1 − 9α + 22α 2 − 9α 3 + 4α 4

(A.28)

|r231 (α)|= 1 90 − 810α + 1635α 2 + 1485α 3 − 3462α 4 + 4056α 5 − 659α 6 6 1 2 − 108α 7 − 320α 8 + 24α 9 + 87α 10 − 102α 11 + 36α 12 + α 13    − 1 2 α 7 2 − 17α + 33α 2 + 13α 3 + 4α 5 + α 6 (−1 + 4α)2

(A.29)

|r222 (α)|= 1 64 − 608α + 1348α 2 + 728α 3 − 1692α 4 + 1948α 5 − 688α 6 − 2008α 7 + 1184α 8 4 1 2 (A.30) + 742α 9 − 355α 10 − 136α 11 + 34α 12 + 14α 13 + α 14 
2 − 12 α 9 (−1 + 4α) −2 + 7α + 6α 2 + α 3 |r213 (α)|= 1 −54 + 918α − 5973α 2 + 18921α 3 − 32250α 4 + 32742α 5 − 15643α 6 + 2506α 7 6 1 2 + 3246α 8 − 3551α 9 + 1327α 10 − 156α 11 − 48α 12 + 59α 13 + 16α 14 + α 15

− 1 α 7 −2 + 9α + α 2 + α 4 (6 − 48α + 90α 2 + 4α 3 )2 2

(A.31)

|r240 (α)|= 1  −432 + 2808α − 5016α 2 + 4692α 3 − 2584α 4 24 1 2 + 930α 5 − 210α 6 + 7α 7 + 27α 8 − 6α 9 + α 10 
2
− 12 α 4 −1 + 4α + α 2 1 − 9α + 22α 2 − 9α 3 + 4α 4 c1 (α) =     √ √ 2i − 2 −1 + 4α + α 2 −7i + 5 −1 + 4α     √ √ √ + α 25i − 13 −1 + 4α + α −25i + 7 −1 + 4α − α −7i + −1 + 4α    −1 √ √ √ 2α 3 −1 + 4α −i + −1 + 4α iα + −1 + 4α

29

(A.32)

(A.33)

References [1] http://www.math.u-szeged.hu/~krisztin/ricker. [2] Goetz Alefeld. Introduction to interval analysis. SIAM Rev., 53(2):380–381, 2011. [3] CAPA: Computer-aided proofs in analysis group. http://www2.math.uu.se/~warwick/ CAPA/. University of Uppsala, University of Bergen. [4] Colin W. Clark. A delayed-recruitment model of population dynamics, with an application to baleen whale populations. Journal of Mathematical Biology, 3:381–391, 1976. [5] Computer Assisted Proofs in Dynamics group. CAPD Library. http://capd.ii.uj.edu.pl. a C++ package for rigorous numerics. [6] Michael Dellnitz and Andreas Hohmann. A subdivision algorithm for the computation of unstable manifolds and global attractors. Numer. Math., 75(3):293–317, 1997. [7] Michael Dellnitz, Andreas Hohmann, Oliver Junge, and Martin Rumpf. Exploring invariant sets and invariant measures. Chaos, 7(2):221–228, 1997. [8] Attila D´enes and G´eza Makay. Attractors and basins of dynamical systems. Electron. J. Qual. Theory Differ. Equ., pages No. 20, 11, 2011. [9] Zbigniew Galias. Rigorous investigation of the Ikeda map by means of interval arithmetic. Nonlinearity, 15(6):1759–1779, 2002. [10] NIIF: National Information Infrastructure Development Institute. http://www.niif.hu. [11] V´ıctor Jim´enez L´opez. A counterexample on global attractivity for Clark’s equation. In Proceedings of the Workshop Future Directions in Difference Equations, volume 69 of Colecc. Congr., pages 97–105. Univ. Vigo, Serv. Publ., Vigo, 2011. [12] V´ıctor Jim´enez L´opez and E. Parre˜no. L.A.S. and negative Schwarzian derivative do not imply G.A.S. in Clark’s equation. [13] Yuri A. Kuznetsov. Elements of applied bifurcation theory, volume 112 of Applied Mathematical Sciences. Springer-Verlag, New York, second edition, 1998. [14] Simon A. Levin and Robert M. May. A note on difference-delay equations. Theoret. Population Biology, 9(2):178–187, 1976. [15] Eduardo Liz. Local stability implies global stability in some one-dimensional discrete singlespecies models. Discrete Contin. Dyn. Syst. Ser. B, 7(1):191–199 (electronic), 2007. [16] Eduardo Liz. Stability of non-autonomous difference equations: simple ideas leading to useful results. J. Difference Equ. Appl., 17(2):203–220, 2011.

30

[17] Eduardo Liz, Victor Tkachenko, and Sergei Trofımchuk. Global stability in discrete population models with delayed-density dependence. Math. Biosci., 199(1):26–37, 2006. [18] Stefano Luzzatto and Paweł Pilarczyk. Finite resolution dynamics. Found. Comput. Math., 11(2):211–239, 2011. [19] Ramon E. Moore. Methods and applications of interval analysis, volume 2 of SIAM Studies in Applied Mathematics. Society for Industrial and Applied Mathematics (SIAM), Philadelphia, Pa., 1979. [20] N. S. Nedialkov, K. R. Jackson, and G. F. Corliss. Validated solutions of initial value problems for ordinary differential equations. Appl. Math. Comput., 105(1):21–68, 1999. [21] W. E. Ricker. Stock and recruitment. Journal of the Fisheries Research Board of Canada, 11(5):559–623, 1954. [22] Jeremy G. Siek, Lie-Quan Lee, and Andrew Lumsdaine. The Boost Graph Library User Guide and Reference Manual (With CD-ROM). 2002. [23] Robert Tarjan. Depth-first search and linear graph algorithms. SIAM Journal on Computing, 1(2):146–160, 1972. [24] Victor Tkachenko and Sergei Trofimchuk. A global attractivity criterion for nonlinear nonautonomous difference equations. J. Math. Anal. Appl., 322(2):901–912, 2006. [25] Warwick Tucker. A rigorous ODE solver and Smale’s 14th problem. Found. Comput. Math., 2(1):53–117, 2002. [26] Warwick Tucker. Validated numerics. Princeton University Press, Princeton, NJ, 2011. A short introduction to rigorous computations. [27] Daniel Wilczak. Uniformly hyperbolic attractor of the Smale-Williams type for a Poincar´e map in the Kuznetsov system. SIAM J. Appl. Dyn. Syst., 9(4):1263–1283, 2010. With online multimedia enhancements.

31