The authors of the manuscript would like to point out the following.
The statement of Theorem 3.1 is incorrect because condition $(3.2)$ does not imply in general condition $(3.3)$ which is used in the proof. If instead of $(3.2)$ we assume
$$0<\lambda_i \le \frac{(1-c) \beta_i}{f_i(\beta_1,\beta_2)\gamma_{i}^{*}}, \quad \mbox{for some $(\beta_1,\beta_2)\in (0,B_1)\times (0,B_2)$,}$$
then the result of Theorem 3.1 is true with the same proof. This change in Theorem 3.1 does not affect the example $(3.4)-(3.5)$.
Corrigendum
8 July 2015
The authors of the manuscript would like to point out the following.
The statement of Theorem 3.1 is incorrect because condition $(3.2)$ does not imply in general condition $(3.3)$ which is used in the proof. If instead of $(3.2)$ we assume
$$0<\lambda_i \le \frac{(1-c) \beta_i}{f_i(\beta_1,\beta_2)\gamma_{i}^{*}}, \quad \mbox{for some $(\beta_1,\beta_2)\in (0,B_1)\times (0,B_2)$,}$$
then the result of Theorem 3.1 is true with the same proof. This change in Theorem 3.1 does not affect the example $(3.4)-(3.5)$.
Competing interests
None